Right Ventricular Systolic Function in Dogs with Preclinical and Clinical Myxomatous Mitral Valve Disease

E.H. Chapel1, B.A. Scansen2, K.E. Schober1, J.D. Bonagura1

1The Ohio State University College of Veterinary Medicine, Columbus, OH, USA, 2Colorado State University, Fort Collins, CO, USA

Right ventricular (RV) dysfunction strongly and independently predicts outcomes in people with myxomatous mitral valve disease (MMVD). The purpose of this prospective pilot study was to assess RV systolic function (RVSF) in dogs with different stages of MMVD. We hypothesized that RVSF would differ among stages of disease, and decrease in decompensated MMVD.

Client‐owned dogs with MMVD not previously receiving cardiovascular medications and without moderate‐or‐severe pulmonary hypertension were eligible for enrollment. Dogs presenting in acute congestive heart failure received a single dose of furosemide prior to echocardiography. Five validated indices of RVSF were measured: tricuspid annular plane systolic excursion (TAPSE), fractional area change, lateral tricuspid annular longitudinal peak systolic velocity (S’), and longitudinal global strain (Strain) and strain rate of the RV free wall. Groups were compared using one‐way ANOVA and Sidak's test. Frequencies of cases with cardiac remodeling falling outside previously‐established reference intervals were compared using Fisher's exact test.

Thirty‐five dogs at 3 stages of MMVD were recruited based on statistical power analysis: B1 (n = 12), B2 (n = 12), and C (n = 11). Dogs in stage B2 had higher values for TAPSE, S’, and Strain (P ≤ 0.05) compared to stage B1; and TAPSE compared to stage C (P = 0.019). TAPSE exceeded reference limits in 67% of B2 dogs compared with 9% in stage C (P = 0.009). This study indicates that compensated MMVD with remodeling is associated with hyperdynamic RVSF, but with cardiac decompensation, TAPSE, S’ and Strain often normalize. These results demonstrate that RVSF differs across stages of MMVD.

High‐Pressure Balloon Valvuloplasty for Severe Pulmonic Stenosis: A Prospective Observational Study in 25 Dogs

C. Belanger1, C.T. Gunther‐Harrington2 , S. Nishimura1, M.S. Oldach2, L.C. Visser3, J.A. Stern3

1Veterinary Medical Teaching Hospital, University of California Davis, Davis, CA, USA, 2University of California, Davis Veterinary Medical Teaching Hospital, Davis, CA, USA, 3Department of Medicine and Epidemiology, University of California Davis, Davis, CA, USA

We aimed to evaluate the safety and efficacy of high‐pressure balloon valvuloplasty (HPBVP) for treatment of canine pulmonic stenosis.

Twenty‐five dogs with severe valvular PS (echocardiographically‐derived transpulmonic peak/maximum pressure gradient (EDPG) ≥80 mmHg) were recruited. All patients underwent complete echocardiography before and 16–24 hours after balloon valvuloplasty (BVP) was performed using a high‐pressure balloon (Atlas Gold PTA Dilatation Catheter) with rated burst pressures ranging from 14–18 mmHg. Balloon sizes were 12–24 mm and 4 cm in length. The pulmonic valve annulus diameter was multiplied by a mean sizing factor of 1.4 ± 0.15 to determine balloon size. Procedural success was defined as a post‐HPBVP EDPG reduction of at least 50% or reduction into at least the moderate category of PS (50–79 mmHg). Optimal result was defined as a post‐procedural EDPG ≤ 30 mmHg.

Initial median (IQR) EDPG for all dogs was 96 (88, 127)mmHg with a post‐operative median of 48 (36, 65)mmHg. HPBVP produced a median pressure gradient reduction of 63 (39, 68)% and procedural success rate of 92% (23 dogs). Optimal results were achieved in 56% (14 dogs). There were no significant correlations between pressure gradient reduction and valve morphology (Type I and Type II) or severity of right ventricular hypertrophy. Pulmonic valve annulus diameter was the only tested variable that was significantly correlated to pressure gradient reduction (P = 0.02; R2 = 0.22). No patients experienced suicide right ventricle and all patients survived the procedure.

These findings suggest that HPBVP is an appealing treatment option for routine cases of severe PS. Considering the procedural success rate and high number of optimal results, further studies comparing HPBVP to traditional BVP are warranted.

Exit Block as a Mechanism of Sinus Node Dysfunction Evidenced by Geometric Heart Rate Variability

F. Giacomazzi1, R. Pariaut2, R. Santilli3, N. Otani4, N.S. Moise3

1Cornell University, Ithaca, NY, USA, 2Cornell University College of Veterinary Medicine, Ithaca, NY, USA, 3Clinica Veterinaria Malpensa, Samarate, Lombardia, Italy, 4Rochester Institute of Technology College of Science, Rochester, NY, USA, 5Department of Clinical Sciences, Section of Cardiology, Cornell University College of Veterinary Medicine, Ithaca, NY, USA

The purpose of this study was to determine if beat‐to‐beat patterning as judged by geometric heart rate variability (HRV) was compatible with the hypothesis of impulse exit block as a mechanism for sinus node dysfunction (SND) in dogs.

Twenty‐four dogs with SND and 24 age/size matched control dogs were studied. Geometric HRV was assessed using Poincaré plots and tachograms from Holter recordings. Autonomic tone was assessed using time and frequency domain HRV. A computer model was designed to assess beat‐to‐beat behavior. Inputs included (1) sinusoidal oscillations of vagal tone with a 4‐second period (respiration), slowly varying component of sleep‐wake cycle, and randomly varying noise; (2) oscillator of sinus node with variation dependent on vagal tone and a random component based on previous exit block; (3) probability of exit block with threshold dependent on vagal tone.

SND dogs had higher indices of HRV than control dogs. In contrast to control dogs, recordings from SND dogs showed >2 beat‐to‐beat clusters that approximated multiples. Depending on input, control and SND dog beat‐to‐beat patterns could be mimicked by the model. The computer model verified a pattern representing exit block (Frames A,B,C). SND dogs had Poincaré plots that failed to reveal an inherent sinus node rate (90–100 bpm) as did the control dogs. Instead, the shortest intervals were faster (120–170 bpm) with multiple bifurcations as the rate slowed.

Beat‐to‐beat patterning of SND dogs is consistent with block in the sinus node supporting the hypothesis of impulse block in the exit pathways from the sinus node.

Is the Left Atrium Prothrombotic in Cats with Aortic Thromboembolism?

T. Kurosawa1, W.C. Cheng2, L. Wilkie2, J. Sargent3, D.J. Connolly4, V.L. Fuentes1

1Royal Veterinary College, Hertfordshire, UK, 2Royal Veterinary College, Hatfield, UK, 3Royal Veterinary College/Southern Counties Veterinary Specialists, Bath, UK, 4Royal Veterinary College, Hatfield, UK

Left atrial (LA) endothelial damage or stretch results in a procoagulable environment in association with down‐regulation of thrombomodulin expression. The role of the endothelium in aortic thromboembolism (ATE) in cats is unknown. The aim of our study was to compare LA thrombomodulin gene expression in normal cats, cats with congestive heart failure (CHF) and cats with ATE.

Myocardium was collected from 5 normal cats, 8 cats with CHF and 12 cats with ATE within 15 minutes of euthanasia. LA myocardium was placed in RNA later and frozen in −80°C until batch analysis. The rest of the heart was formalin‐fixed for histopathology. RNA was extracted and GeXP eXpress Profiler software was used to design primers. Multiplex qRT‐PCR was performed to quantify thrombomodulin gene expression. All multiplex results were normalized to ribosomal protein S7. Results were expressed as median (range) and compared between groups using the Kruskal‐Wallis and Mann Whitney tests.

There was no difference in sex (P = 0.22) or age between groups (P = 0.38). Median age was 79 months (4–184 months) in normal cats, 77 months (7–130 months) in CHF cats and 103 months (44–222 months) in ATE cats. Normalized thrombomodulin concentration in the ATE group was 0.10 (0.01–0.19) compared with 0.14 (0.08–0.32) in the CHF group and 0.2 (0.08–0.22) in the normal group (P = 0.09).

Although no significant difference in thrombomodulin expression was identified, further investigations with a larger sample size would be justified.

Characterization of Gene Expression Profiles Linked to Degenerative Mitral Valve Disease in Small Breed Dogs

E.M. Oxford1, N.S. Moise1, E.J. Rice2, A.R. Boyko3, C.G. Danko2

1Department of Clinical Sciences, Section of Cardiology, Cornell University College of Veterinary Medicine, Ithaca, NY, USA, 2Department of Biomedical Sciences, Baker Institute for Animal Health, Ithaca, NY, USA, 3Department of Biomedical Sciences, Cornell University College of Veterinary Medicine, Ithaca, NY, USA

The purpose of this study was to use a genome‐wide association study (GWAS) and precision nuclear run‐on and sequencing (PRO‐seq), to identify novel gene candidates and expression profiles linked to DMVD in small breed dogs.

Blood samples from 249 (95 control, 154 DMVD) client‐owned dogs (<20 kg) were used for the GWAS. Tissue samples from the anterior mitral valve leaflets were available from 2 DMVD and 5 control dogs for Western blot and PRO‐seq analysis. DMVD dogs had a murmur and mitral regurgitation associated with leaflet thickening and/or prolapse on echocardiography. Control dogs were >9 years of age with no murmur, and no abnormalities on echocardiography.

A GWAS was performed with genomic DNA using a custom single nucleotide polymorphism (SNP) genotyping array containing 185,850 markers. Results were analyzed by calculating the odds ratio and P value for each SNP. A gene of interest was identified and overlapping primer pairs were designed to incorporate exon segments between 300–400 base pairs in length. Polymerase chain reactions (PCR) were performed using a touchdown program. Sanger sequencing was performed on a total of 20 samples (10 DMVD, 10 controls). Sequence alignments were performed using Clustal Omega software.

A total of 5 control and 2 DMVD valves were available for Western blot analysis. Samples were probed for antibodies recognizing alpha‐catenin and beta‐catenin (P‐Y142 isoform), and normalized to GAPDH signal.

PRO‐seq was performed on one control and one DMVD valve, in order to directly detect changes in gene transcription (expression) through the genome‐wide mapping of RNA polymerase. Data were analyzed for genome wide up‐or‐down regulation using the PANTHER classification system (pantherdb.org). Genes were grouped based on their biological function.

The GWAS revealed the strongest association signal on Chromosome 17 (P = 7.7 × 10−6), near the CTNNA2 gene, which codes for alpha‐catenin. Fine mapping of CTNNA2 did not detect a mutation within the coding regions of the gene. However, Western blot analysis revealed a 53% decrease in normalized alpha‐catenin levels and an 86% decrease in beta‐catenin (P‐Y142 isoform) compared to control samples. PRO‐seq experiments did not reveal a clear signal for regulatory enhancers upstream of alpha‐catenin near the highest association signal. The PRO‐seq data additionally revealed genome‐wide expression profiles for genes significantly up‐or‐down regulated (P < 0.05 after Bonferroni correction) in the DMVD sample. A > 2‐fold change was noted in 187 genes; with functions such as wound healing, regulation of muscle contraction, and muscle structure development.

This study identified a candidate gene (CTNNA2), which codes for alpha‐catenin, an essential protein of the adherens junctions and modulator of the canonical Wnt pathway. Although a mutation was not detected within the coding regions of the gene or within upstream enhancers of CTNNA2, Western blot revealed decreased levels of alpha‐catenin (and the phosphorylated form of beta‐catenin) in DMVD valves. These preliminary data may support the presence of a mutation located within the non‐coding regions controlling CTNNA2 transcription and further investigation is warranted.

Prognostic Value of MAPSE and TAPSE in Feline Hypertrophic Cardiomyopathy

I. Spalla1, J.R. Payne2, K. Borgeat2, V.L. Fuentes3, D.J. Connolly1

1Royal Veterinary College, Hatfield, UK, 2Royal Veterinary College/ Highcroft Veterinary Referrals, Bristol, UK, 3Royal Veterinary College, Hertfordshire, UK

Hypertrophic cardiomyopathy (HCM) has a variable prognosis depending on many factors, including left atrial size, the presence of clinical signs and left ventricular (LV) systolic function. Mitral annular plane systolic excursion (MAPSE) and tricuspid annular plane systolic excursion (TAPSE) assess longitudinal ventricular systolic function and may have prognostic value in cats with HCM.

We sought to investigate whether MAPSE and TAPSE were associated with survival time in cats with HCM.

Medical records were searched for cats diagnosed with HCM by echocardiography (LV wall thickness ≥6 mm), images of adequate quality to measure MAPSE/ TAPSE and survival information. Of 185 included cats, 77 were showing clinical signs (53 CHF, 12 aortic thromboembolism [3 with concurrent CHF], 9 syncope, 3 exertional dyspnoea). MAPSE and TAPSE were measured by anatomical M‐Mode at the atrioventricular annular level of the LV free wall (MAPSE‐FW), interventricular septum (MAPSE‐IVS) and right ventricular wall (TAPSE).

Cats with MAPSE‐FW < 3.6 mm had a median survival time (MST) of 352 days vs. 2171 days in MAPSE‐FW ≥ 3.6 mm (P < 0.01). Cats with MAPSE‐IVS < 3.2 mm had a MST of 352 days vs. 2171 days for MAPSE‐IVS ≥ 3.2 mm (P < 0.001) and cats with TAPSE < 6.4 mm had a MST of 380 days vs 2171 days for TAPSE ≥ 6.4 mm.

Lower MAPSE and TAPSE were associated with worse survival, suggesting systolic longitudinal dysfunction has prognostic value in cats with HCM

Echocardiographic Assessment of Aortic Root Rotation in Dogs with Congenital Heart Disease

N.R. Wyatt1, B.A. Scansen2

1Colorado State University Veterinary Teaching Hospital, Fort Collins, CO, USA, 2Colorado State University, Fort Collins, CO, USA

Rotation of the aortic root, defined as abnormal position of the aortic valvar sinuses relative to surrounding cardiac structures, has been described in children with congenital heart disease but not dogs. Children with tetralogy of Fallot (TOF) show clockwise aortic root rotation and have an increased prevalence of coronary artery anomalies. Coronary artery anomalies are also common in brachycephalic dogs with pulmonary valve stenosis (PS). This study evaluated aortic root rotation in dogs with normal hearts, PS, subaortic stenosis (SAS), and TOF. Aortic root rotation in normocephalic compared to brachycephalic PS dogs was also investigated.

Echocardiograms were retrospectively evaluated from normal (n = 31), PS (31), SAS (16), and TOF (5) dogs. Theta was defined as the angle between interatrial septum and commissure of left coronary and noncoronary sinuses, while rho was the angle between interatrial septum and commissure of right coronary and left coronary sinuses; each angle was measured in triplicate for all dogs and averaged. Groups were compared by ANOVA or Welch's test with alpha set at 0.05.

Differences in angles theta (P < 0.001) and rho (P = 0.021) were apparent between groups. Dogs with PS showed mild clockwise aortic root rotation, while dogs with TOF showed severe rotation, compared to normal or SAS dogs. In PS dogs, clockwise rotation was greater in brachycephalic compared to normocephalic dogs (P = 0.0004).

We report clockwise aortic root rotation in dogs with congenital heart disease, particularly PS and TOF. Ongoing studies are investigating the relationship between aortic root rotation and coronary artery anomalies in brachycephalic PS dogs.

Echocardiographic Right Atrial Volumes in Cats with Cardiac Disease and Pleural Effusion vs. Pulmonary Edema

M. Wilson, K.F. Scollan

Oregon State University, Corvallis, OR, USA

Cardiomyopathies, including hypertrophic and restrictive, are the most common acquired cardiac diseases in cats. Many affected cats develop pulmonary edema secondary to left‐sided congestive heart failure (L‐CHF). In addition, cats with primarily left‐sided cardiomyopathy can develop pleural effusion without the presence of pulmonary edema. It is well known, but not well understood, why cats with left‐sided CHF can develop both pulmonary edema and pleural effusion. A previous study identified that cardiomyopathic cats with pleural effusion had larger right atrial (RA) diameters than cats with pulmonary edema. The aim of this study was to compare right atrial volumes in cats in L‐CHF with pleural effusion, L‐CHF with pulmonary edema, and a group of healthy cats.

Medical records and echocardiographic studies from 2008 to 2015 were reviewed and measured. Cats were categorized into three groups: 1) L‐CHF cats with pleural effusion 2) L‐CHF cats with pulmonary edema, and 3) healthy cats. Right atrial volumes were measured from both the right long‐axis four‐chamber (RLA) and left apical four‐chamber (LAP) views. Volumes were calculated using both the modified Simpson's method of disks (MOD) and the area‐length formula (A‐L) and indexed to body to yield mL/kg values. The LA:Ao ratio was measured from the right parasternal short‐axis view. The RA volume methods (MOD vs. A‐L) and views (RLA vs. LAP) were compared using Wilcoxon matched‐pairs signed rank tests. Group medians of the MOD RLA volumes and LA:Ao ratios were compared using Kruskal Wallis tests. Receiver operator curves were generated to assess sensitivity and specificity of various cut‐off values of indexed RA volumes for cats to develop pleural effusion.

Sixty‐nine cats were included in the study, 27 cats with L‐CHF and pleural effusion, 21 cats with L‐CHF and pulmonary edema, and 21 healthy cats. The median indexed RA volume of the pleural effusion group (0.41 mL/kg, range 0.12–2.4) was significantly larger than the pulmonary edema (0.16 mL/kg, range 0.09–0.6; P < 0.0001) and normal (0.23 mL/kg, range 0.12–0.39; P = 0.007) groups. RA volumes were not significantly different between the pulmonary edema and normal groups. Using the MOD method, RA volumes were not different from the RLA and LAP views, however values obtained by MOD and A‐L methods were significantly different (P = 0.0013). Median LA:Ao values were not statistically different between the pleural effusion (2.17, range 1.63–3.14) and pulmonary edema groups (2.23, range 1.59–3.2), though both were different than the normal group. Using RA volumes to differentiate pleural effusion from pulmonary edema CHF, the area under the curve was 0.83. Sensitivity (Se %) and specificity (Sp %) for selected cut‐off values of indexed RA volumes were: 0.3 mL/kg, Se 74, Sp 82; 0.35 mL/kg, Se 63, Sp 87; and 0.4 mL/kg, Se 56, Sp 95.

The results of this study indicate that cats that developed pleural effusion had significantly larger indexed RA volumes than those that developed pulmonary edema and healthy cats. Cats with CHF in both groups had LA enlargement, though this study would indicate that concurrent RA enlargement is more likely to cause pleural effusion as a manifestation of CHF than LA enlargement alone. RA volumes greater than 0.3 mL/kg offer adequate sensitivity and specificity for the development of pleural effusion in cardiomyopathic cats. Measurement of RA volumes are useful in estimating the risk of developing pleural effusion in cats with cardiomyopathy and normal values of RA volumes from a larger population should be investigated.

Reader Performance of Radiographic Left Atrial Enlargement in Dogs and Comparison to Echocardiographic Left Atrial Assessment

L. Duler, N. LeBlanc, K.F. Scollan, S. Cooley, S. Nemanic

Oregon State University, Corvallis, OR, USA

Accurate assessment of left atrial (LA) size is integral to the evaluation of left‐sided cardiac disease in dogs and provides risk stratification for the development of left‐sided congestive heart failure (CHF). Thoracic radiographs are routinely used to subjectively assess the presence and severity of left atrial enlargement (LAE). The primary aim of this study was to assess interreader agreement of subjective radiographic LAE between readers with different levels of experience and expertise. A second aim was to compare subjective degrees of radiographic LAE with echocardiographic measurements of LA size.

This retrospective study identified dogs that underwent thoracic radiographs and an echocardiogram on the same day at the Oregon State University Veterinary Teaching Hospital between 2012 and 2014. All thoracic radiographic views were anonymized and reviewed by 2 board‐certified cardiologists, 2 board‐certified radiologists, and 2 small animal rotating interns. Radiographs were evaluated for the presence or absence of LAE, severity of LAE if present (mild, moderate, or severe), and presence or absence of 7 pre‐defined subjective LAE criteria. On the lateral projection these included elevation of the left caudal mainstem bronchus, loss of the normal gentle curvature of the caudal margin of the cardiac silhouette, straightening of the caudodorsal margins of the cardiac silhouette, and increased height of the caudodorsal border of the heart. On the ventrodorsal and dorsoventral projections, these criteria included divergence (or splaying) of the principal bronchi, a double opacity sign on 5.00–7.00 o'clock position on the cardiac silhouette, and evidence of left auricular enlargement at the 2:30–3 o'clock position.

Echocardiographic LA size was evaluated objectively by the left atrial‐to‐aortic ratio (LA:Ao) and LA volume via the monoplane modified Simpson's method of discs (MOD). LA volumes (LAV) were indexed to body weight as mL/kg. Objective measures of LA size were stratified into categories of normal (LA:Ao < 1.5; LAV < 1.3), mild (1.5 ≤ LA:Ao < 1.7; 1.3 ≤ LAV < 2.35), moderate (1.7 ≤ LA:Ao < 2; 2.35 ≤ LAV < 3.4), or severely enlarged (LA:Ao ³ 2; LAV ³ 3.4). Interreader agreement of radiographic LAE evaluation was assessed by linearly weighted kappa (k) and intraclass correlation (ICC). Sensitivity (Se) and specificity (Sp) of identifying varying degrees of echocardiographic LAE on radiographs was assessed using both LA:Ao and LAV.

101 dogs were included with a median weight of 8.6 kg (range 1–65 kg) and a median age of 8 years old (range 0.2–15 years). Echocardiographic diagnoses included normal (n = 13), chronic degenerative valve disease (n = 43), dilated cardiomyopathy (n = 12), pulmonic stenosis (n = 13), left‐to‐right patent ductus arteriosus (n = 11), subaortic stenosis (n = 4), and single cases of other cardiac diseases. Agreement for the presence or absence of radiographic LAE was excellent between all readers (ICC = 0.7828) The interreader agreement for the degree of radiographic LAE ranged from moderate to substantial (k = 0.493–0.701) dependent on the combination of readers. For the 7 distinct subjective LAE criteria, the interreader agreement ranged from fair to good (ICC = 0.264–0.586) with the strongest interreader agreement for increased height of the caudodorsal heart margin, splaying of the mainstem bronchi, and enlargement of the left auricle. Agreement between specialists was higher than agreement between all readers for each individual criterion, with the exception of elevation of the left caudal mainstem bronchus.

Overall there was strong agreement between methods of echocardiographic LAE stratification: The agreement between LAV and subjective LAE was almost perfect (k = 0.939), while LA:Ao had moderate agreement with the LAV (k = 0.587). The agreement between subjective radiographic LAE and objective echocardiographic LAE was moderate and varied slightly with echocardiographic method of classification (LAV, k = 0.573; LA:Ao, k = 0.428). The sensitivity and specificity of thoracic radiographs to identify LAE was slightly better using LAV than LA:Ao (LAV Se = 84% and Sp = 70%; LA:Ao: Se = 78% and Sp = 71%).

The results of this study confirm the hypothesis that interreader agreement for radiographic classification of LAE is strong, even across specialty disciplines and levels of experience. There was variable agreement between readers for application of distinct, subjective criteria of LAE and this study indicated that increased height of the caudodorsal heart margin, splaying of the mainstem bronchi, and enlargement of the left auricle were the most consistent criteria used by cardiologists, radiologists, and interns to identify LAE. Subjective radiographic LAE severity had only moderate agreement with echocardiographic LAE severity regardless of objective LA size measure, however LAV and subjective LAE showed stronger agreement than LA:Ao.

Echocardiographic Effects of Butorphanol and Atenolol in Dogs with Pulmonary Valve Stenosis

S. Nishimura1, L.C. Visser2, C. Belanger1, C.T. Gunther‐Harrington3, M.S. Oldach4, J.A. Stern2

1Veterinary Medical Teaching Hospital, University of California Davis, Davis, CA, USA, 2Department of Medicine and Epidemiology, University of California Davis, Davis, CA, USA, 3University of California, Davis Veterinary Medical Teaching Hospital, Davis, CA, USA, 4University of California, Davis, Davis, CA, USA

We sought to determine the effects of butorphanol and atenolol on maximum transpulmonary pressure gradient (maxPG), pulmonary valve effective orifice area indexed to body size (EOAi), and right ventricular (RV) systolic function in dogs with pulmonic stenosis (PS).

Dogs (n = 18) with PS (maxPG > 50 mmHg) were prospectively recruited. Dogs had 3 echocardiograms: baseline, 5‐minutes post‐butorphanol (0.2–0.25 mg/kg IV), and 2–3 weeks post‐atenolol (1–1.5 mg/kg q12). In a blinded fashion, maxPG, EOAi, pulmonic‐to‐aortic velocity time integral ratio (VTI ratio) and RV function (assessed via tricuspid annular plane systolic excursion, RV fractional area change, and RV systolic velocity [S’]) were measured. MaxPG, VTI ratio, EOAi and RV function indices were compared pre‐ and post‐drug using a paired t‐test. Percent changes post‐drug were compared with an unpaired t‐test.

Post‐butorphanol, there were no significant differences in indices of RV function or severity of PS but heart rate was significantly reduced (mean %change ± SD = −21% ± 16; P < 0.001). Post‐atenolol, heart rate and maxPG was significantly (P ≤ 0.01) reduced with mean %changes ± SD of −17% ± 13 and −15% ± 9, respectively. There were no significant changes in EOAi, VTI ratio and RV function post‐atenolol. Percent change in VTI ratio post‐atenolol (6% ± 12) was significantly (P = 0.01) different than maxPG (−15% ± 9). Three dogs previously diagnosed with severe PS (maxPG > 80 mmHg) were diagnosed with moderate PS (maxPG 50–80 mmHg) post‐atenolol.

Atenolol reduces PS severity assessment via maxPG. EOAi and VTI ratio may be useful flow‐independent methods for PS severity assessment, particularly post‐drug or intervention.

Prospective Clinical Trial Evaluating Spironolactone in Dobermans with Congestive Heart Failure due to Dilated Cardiomyopathy

L. O'Sullivan1, S. Fonfara1, A. Laskary2, H. Chambers1

1University of Guelph, Guelph, ON, Canada, 2Ontario Veterinary College/University of Guelph, Guelph, ON, Canada

The aldosterone antagonist spironolactone was recently shown to improve survival in dogs with congestive heart failure (CHF) due to chronic mitral valvular disease. Whether spironolactone has beneficial effects in dogs with dilated cardiomyopathy (DCM) is not known. The primary objective of the study was therefore to evaluate the effect of spironolactone on survival time in Dobermans with CHF due to DCM.

The trial design was prospective, randomized, single‐blinded, and placebo‐controlled. 67 dogs were randomized to receive 50 mg of spironolactone twice daily (34 dogs) or a placebo (33 dogs), in addition to standard therapies. Clinical follow up was every 1–4 weeks until endpoint. Quality of life questionnaire and physical examination were performed at every recheck, while renal biochemistry, ECG, echocardiography, and thoracic radiography were reassessed as needed. The primary endpoint was time to cardiac death, defined as death from CHF, euthanasia (euth) for CHF, or sudden death (SD). Kaplan‐Meier survival estimates were computed and compared.

The 67 dogs comprised 20 females and 47 males with average (range) age 7.9 years (2.0–14.3). One dog in the spironolactone group was withdrawn by the owner. 59 dogs (89%) reached the primary endpoint (5 CHF death, 26 CHF euth, 28 SD). 7 dogs (11%) were censored (6 euth non‐heart, 1 lost to follow‐up). The median survival time in the spironolactone group (183 days) was not statistically significantly different than that for the placebo group (124 days) (Log‐Rank P = 0.35, Wilcoxon P = 0.19). Results need to be verified in larger trials given the relatively small sample size.

Heart‐Fatty Acid Binding Protein – A Potential Marker for Canine Cardiac Diseases

C. Lam1, D.C. Sorrosal2, G. Monteith3, S. Fonfara4

1Ontario Veterinary College, University of Guelph, Guelph, ON, Canada, 2Dick White Veterinary Referrals, Cambridge, UK, 3Ontario Veterinary College, Guelph, ON, Canada, 4University of Guelph, Guelph, ON, Canada

Heart‐Fatty Acid Binding Protein (hFABP) is produced primarily in cardiomyocytes, mainly responsible for fatty acid homeostasis, and has been proposed as an early and sensitive biomarker in human cardiac diseases. Its potential use as a cardiac biomarker in veterinary medicine has yet to be investigated. This study aims to assess hFABP as a prospective biomarker for canine cardiac diseases compared to cardiac Troponin‐I (cTnI).

Twenty‐one healthy dogs (control), 23 dogs with degenerative mitral valve disease (MVD), and 13 dogs with dilated cardiomyopathy (DCM) of various cardiac disease stages were studied. Serum measurements of HFABP and cTnI, routine bloodwork and echocardiography were performed at initial presentation for all dogs and at 1 and 3 months for the diseased dogs, and followed up until 12 months to verify survival.

Levels recorded during the initial presentation showed significantly elevated hFABP with highest levels found in ACVIM stage C disease regardless of cause, followed by stage B and controls. cTnI differed stage C from both B and controls, but was unable to differentiate between stage B and controls. DCM patients had higher hFABP and cTnI levels over MVD patients, but cTnI was indistinguishable between MVD dogs and controls. Reduction of hFABP and cTnI during the follow up period was also observed from initial presentation among both DCM and stage C dogs. hFABP was the most significant predictor for death when multivariate study was performed.

This study showed the potential value of hFABP as a cardiac marker to differentiate stages and disease type in dogs.

Non‐Gated Computed Tomography Angiography for Evaluation of Coronary Artery Anomalies in Bulldogs with Pulmonic Stenosis

C.T. Gunther‐Harrington1, K. Phillips2, L.C. Visser3, J.A. Stern3

1University of California, Davis Veterinary Medical Teaching Hospital, Davis, CA, USA, 2UC Davis, Davis, CA, USA, 3Department of Medicine and Epidemiology, University of California, Davis, Davis, CA, USA

Prepulmonic coronary anomalies (CAs) have been reported in bulldogs and represent an increased risk when performing balloon valvuloplasty (BVP). Identification of CAs have been reported using high‐slice cardiac‐gated computed tomography angiography (CTA). However, the utility of the more accessible, non‐gated, 16‐slice CTA is unknown. The purpose of this study was to evaluate the feasibility of non‐gated CTA to identify CAs in anesthetized bulldogs with severe pulmonic stenosis (PS) prior to BVP.

In this prospective observational study, bulldogs and bulldog mixes with severe PS were recruited. Full echocardiograms were performed to diagnose PS. Dogs were placed under general anesthesia and a non‐gated 16 slice CTA was performed. CTAs were reviewed by a board‐certified radiologist and cardiologist and coronary anatomy identified. When normal coronary artery anatomy was noted on CTA, a right ventricular outflow tract angiogram was performed and evaluated during levophase to confirm normal coronary anatomy prior to BVP. Dogs with prepulmonic CAs were recovered from anesthesia and no BVP was performed.

Five bulldogs and 2 bulldog crosses were enrolled. All dogs had clear, interpretable images from the non‐gated CTA. Coronary anomalies were identified in 4 dogs based on non‐gated CTA, all of which were consistent with R2A anomaly. Three dogs had normal coronary anatomy based on non‐gated CTA which was confirmed during levophase angiography. Successful BVP was performed without incident in these 3 dogs.

We conclude that non‐gated CTA represents an accessible, noninvasive method for diagnosing CAs in bulldogs with PS.

Ventricular Remodelling Marker Levels Differ in Cats with Hypertrophic Cardiomyopathy with and without Atrial Thrombus

S. Fonfara1, S. Kitz2, A. Kipar2

1University of Guelph, Guelph, ON, Canada, 2University of Zurich, Zurich, Switzerland

Atrial thrombosis (AT) is a known complication in cats with hypertrophic cardiomyopathy (HCM). Slow blood flow, endothelial damage and hypercoagulability, caused by an enlarged left atrium (LA) and cardiac dysfunction are suspected to be involved in thrombus formation. However, not all cats with HCM and LA enlargement develop AT, and how cats with AT differ from cats without AT is not known.

We therefore compared the myocardial transcription of cytokines (interleukin [IL]‐1, ‐2, ‐4, ‐6, ‐8, ‐10, ‐12, ‐18, tumour necrosis factor‐α, interferon‐γ, transforming growth factor [TGF]‐β), matrix metalloproteinases (MMP)‐2, ‐3, ‐9, ‐13, and tissue inhibitors of metalloproteinase (TIMP)‐1, ‐2, and ‐3 from atria and ventricular samples from 10 cats with HCM and 5 cats with HCM and AT using quantitative real time PCR. The cats had been clinical patients, and diagnosis of the disease was obtained by echocardiography.

Cats with HCM showed generally higher atrial than ventricular marker transcription, which applied for cats with HCM and AT only for IL‐1, TGF‐β, MMP‐2, MMP‐9 and TIMP‐1. Cats with HCM and AT exhibited also significantly lower atrial IL‐12 and lower ventricular MMP‐2 and TIMP‐2 transcription than atria and ventricular samples, respectively, from cats with HCM without AT.

These results suggest different remodelling processes primarily in the ventricles of cats with HCM and AT, which might contribute to cardiac dysfunction and AT formation in cats with HCM. Lower IL‐12 and TIMP‐2 have been reported in association with increased left ventricular hypertrophy, fibrosis and dysfunction.

Normal Echocardiographic Parameters in Clinically Healthy Adult Captive Tigers with and without Medetomidine Sedation

K.T. Rodriguez, A. Cushing, E. Ramsay, R. Gompf

University of Tennessee Small Animal Clinical Sciences, Knoxville, TN, USA

This study aimed to describe echocardiographic findings in healthy adult tigers and determine the difference between echocardiographic variables under sedation with medetomidine and after reversal with atipamezole. Eight adult tigers (Panthera tigris) from an exotic cat sanctuary were included.

Each tiger underwent a complete echocardiographic examination by the same operator following sedation with medetomidine, midazolam and ketamine and induction into general anesthesia. Following completion of the first exam, the medetomidine was reversed with atipamezole. A second echocardiographic examination was then performed 30 minutes later by the same operator. Measurements were tested for statistical differences between the two examinations and between gender and weight.

One tiger was excluded from analysis due to the finding of a ventricular septal defect, leaving a total of 7 tigers. Physiologic regurgitation of the tricuspid and pulmonic valves was seen in 86% and 100% of the tigers. There was a statistically significant difference in left ventricular internal dimension at end‐diastole based on weight and interventricular septum thickness at end‐systole between pre and post‐echoes. Fractional shortening and left ventricular internal dimension at end‐systole approached, but did not reach, statistical significance between pre and post‐echoes.

In conclusion, physiologic tricuspid and pulmonic regurgitations are common findings in healthy adult tigers. The echocardiographic measurements obtained during this study can be used as references for future examinations in adult tigers. The effects of medetomidine on these measurements, specifically the left ventricular internal dimension at end‐diastole and interventricular septum thickness at end‐systole, should be taken into account.

Evaluating Clopidogrel Effect in Cats Using Plateletworks (Arachidonic Acid) and Platelet Function Analyzer 200 (P2Y)

A. Abrams‐Ogg1, L. O'Sullivan1, S. Fonfara1, S. Raheb1, S. Saati1, K. Ho2, S. Blois1, R.D. Wood1

1University of Guelph, Guelph, ON, Canada, 2Central Toronto Veterinary Referral Clinic, Toronto, ON, Canada

Platelet function testing may be useful to identify if clopidogrel resistance occurs in cats. Plateletworks® (PW) consistently detects clopidogrel effect in normal cats using ADP as agonist. In canine PW studies, arachidonic acid (AA) is as useful an agonist as ADP to detect clopidogrel effect, but there are no data for cats. Platelet Function Analyzer® (PFA) inconsistently detects clopidogrel effect in normal cats using the collagen‐ADP cartridge. This also occurs in humans, prompting development of the INNOVANCE® P2Y cartridge, which consistently detects clopidogrel effect in humans and dogs. The objective of this study was to investigate PW‐AA and PFA‐P2Y in cats. Nine normal cats and 13 cats with cardiovascular diseases were studied. All had normal hematocrits and platelet counts. Blood samples were collected by jugular venipuncture. The 9 normal cats showed PW‐AA %Aggregation (%Agg) = 83–97% and PFA‐P2Y Closure Time (CT) = 38–111 seconds. For cats with cardiovascular diseases (one or both tests performed): a) 7 cats already receiving clopidogrel: PW‐AA %Agg = 4–19% (n = 3) and PFA‐P2Y CT >300 seconds (n = 6). b) 6 cats prior to starting clopidogrel: PW‐AA %Agg = 21–96% (n = 5) and PFA‐P2Y CT = 50–118 seconds (n = 3). For 3 of these 6 cats one or both tests were repeated while receiving clopidogrel: PW‐AA %Agg = 0–12% (n = 3) and PFA‐P2Y CT >300 seconds (n = 1). For both tests, differences between normal cats and cats being treated with clopidogrel were significant (P < 0.05). PW with AA agonist and PFA with P2Y cartridge may be highly sensitive to detect clopidogrel effect in cats.

Postoperative Changes in Aortic Characteristics Affect the Low Fractional Shortening in Canine Patent Ductus Arteriosus

T. Aoki1, Y. Fujii1, H. Sunahara1, K. Sugimoto1, T. Uehara1, Y. Wakao2

1Azabu University, Sagamihara, Kanagawa, Japan, 2Azabu University, Sagamihara‐shi, Kanagawa, Japan

Our objective is to elucidate the influence of changes in diastolic aortic pressure and aortic compliance on the fractional shortening (FS%) in dogs that undergo patent ductus arteriosus (PDA) occlusion. 62‐client owned dogs that underwent percutaneous PDA occlusion were retrospectively used for our study. The dogs were evaluated using echocardiography and a catheter for the measurement of their direct blood pressure during PDA occlusion. The diameter of descending aorta was measured using digital fluoroscopic cine loops. After PDA occlusion, the systolic blood pressure did not change; however, pulse pressure significantly decreased (P < 0.01) and diastolic blood pressure (DAP) significantly increased (P < 0.01). In a univariate analysis (n = 49), the low postoperative FS% was associated with an older age (P < 0.05), the increased postoperative diastolic blood pressure, and decreased pulse pressure (both with P < 0.01). In a multivariate analysis (n = 49), low postoperative FS% was associated with an older age (P < 0.05) and a decreased postoperative pulse pressure (P < 0.05). The diameter of the descending aorta significantly increased both in systole and diastole (both with P < 0.01) and the aortic FS% was significantly decreased (P < 0.05) just after PDA occlusion. In conclusion, left ventricle might not be able to contract after PDA occlusion because of increased DAP and decreased aortic compliance, i.e., the aortic valve more strongly closed and the aorta was rapidly dilated like a rupturing balloon due to trapping a large amount of blood into aorta.

Congestive Heart Failure Associated with DOCP Therapy in 12 Dogs with Hypoadrenocorticism

C. Bourguignon1, T.C. DeFrancesco1, S. Tou1, R. Hanel2, S. Marks1, B. Keene1

1North Carolina State University, Raleigh, NC, USA, 2BluePearl Veterinary Partners, Raleigh, NC, USA

Steroid‐associated congestive heart failure (CHF) has been documented in both humans and cats. Episodes of possible DOCP‐associated CHF in dogs with hypoadrenocorticism (HA) prompted a search of medical records from 2003 to 2016 to evaluate the potential relationship between DOCP administration and CHF.

Of 146 canine HA cases identified during this time period, 12 (8%) also experienced CHF. Eight of these dogs had chronic mitral valve disease; the other 4 had dilated cardiomyopathy. Five of the 12 dogs experienced CHF after the first injection of DOCP; 3 of these had known pre‐existing heart disease. The median time between the DOCP and CHF was 11 days (range 1‐ 29 days).

The 12 HA/DOCP/CHF cases were matched to historical controls (CHF without HA, matched by age, breed, sex, cardiac diagnosis and year of diagnosis). The number and time interval of documented CHF episodes and survival after CHF was evaluated in both groups.

Median survival after the onset of CHF was 448 days in controls vs. 953 days in the HA/DOCP group, but this difference was not significant (P = 0.43). The number and time interval of CHF episodes was likewise not different between groups.

The small number of cases and the large variability in the parameters measured limited our ability to identify differences between groups. Despite the negative findings in this small, retrospective study, further study of the relationship between DOCP administration and CHF seems warranted.

An Impaired Oculocardiac Reflex Identifies the Autonomic Imbalance in Dogs with Symptomatic Mitral Valve Disease

B.C. Brüler1, T.C.S. Vieira1, M. Wolf2, S.B. Lucina2, M.G. de Sousa3

1Federal University of Paraná – UFPR/ Department of Veterinary Science, Curitiba, Parana, Brazil, 2Federal University of Paraná – UFPR, Curitiba, Parana, Brazil, 3University of Estadual Paulista UNESP/JABOTICABAL, JABOTICABAL, Sao Paulo, Brazil

The oculocardiac reflex is the heart's physiological response to digital compression of the eyeballs, producing negative chronotropic and inotropic responses in healthy subjects. In this study, we investigated the autonomic imbalance in dogs with CHF due to myxomatous mitral valve disease (MMVD) by means of their individual response to that vagal maneuver. Fifty‐two client‐owned MMVD dogs (stages B1/B2/C/D) and twelve healthy dogs admitted for regular cardiac evaluation were recruited for a prospective observational study. Every dog had an ECG recorded for 5 minutes, and during the last minute both globes were compressed into the orbits using digital pressure until the 5 minute‐recording was over. Twenty RR intervals were used to calculate the mean RR interval, the standard deviation of RR intervals, the root mean square of the successive differences in RR intervals and the vasovagal tonus index before eyeball compression and at approximately twenty seconds after compression started. The percent change of every surrogate was calculated. Although a significant difference was only documented for VVTI (P = 0.0419), the other surrogates did behave differently in stages C and D dogs as compared to B1 and B2 animals (Figure 1). An explanation for this might be the parasympathetic withdraw in CHF, in addition to a subtle sympathetic activation due to discomfort, which was likely suppressed in dogs with preserved vagal response. Also, the sustained increased sympathetic activation rather than transitional sympathetic activation might be involved with the distinct response to OCR in symptomatic dogs as compared to the asymptomatic ones.

Instability of the QT Interval in Dogs with Myxomatous Mitral Valve Disease

B.C. Brüler1, F.S. Jojima1, G. Dittrich1, A.T. Giannico1, M.G. de Sousa2

1Federal University of Paraná – UFPR/ Department of Veterinary Science, Curitiba, Parana, Brazil, 2University of Estadual Paulista UNESP/Jaboticabal, Jaboticabal, Sao Paulo, Brazil

In people, QT prolongation and instability are known to relate to ventricular arrhythmias. Since dogs with myxomatous mitral valve disease are prone to developing ventricular arrhythmias, this study investigated whether indices derived from QT interval could be used to predict an increased likelihood of arrhythmias regardless of clinical signs. 167 client‐owned MMVD dogs (stages B1/B2/C) admitted for cardiac evaluation were recruited for a prospective cross‐sectional study. Every dog underwent an echocardiogram and ECG recording. Fifty consecutive QT intervals were measured for each dog. QTs were corrected for heart rate and used to calculate the average QT (QTa), QT variance (QTv), total instability (TI), short term (STI) and long term (LTI) instabilities. A significant difference between asymptomatic and symptomatic animals was documented for every parameter (P < 0.0001),with higher values found for stage C dogs. Significant difference between B1 and B2 also existed for QTa. An AUC >0.5 was found when QTa and STI were used to discriminate dogs with or without ventricular arrhythmias (Figure 1). Dogs with QTa >310 (sensitivity:56%/specificity:71%), and STI >11 (sensitivity:56%/specificity:85%) were shown to be slightly (15%) and moderately (30–45%) more prone to developing ventricular arrhythmias, respectively. Also, an AUC > 0.5 was documented when every QT index was used to distinguish animals with remodeled (B2/C) and non‐remodeled hearts (B1). QTa and QTv best discriminated B1 from B2/C dogs and B1/B2 from C dogs, respectively. Because an increased QT instability was shown in symptomatic patients, QT indices might translate into prognostic surrogates for arrhythmias and CHF in MMVD dogs.

Cardiovascular Changes in Brachycephalic Dogs

R.M. Canola1, J. Braz1, E. Zacché1, A.Z. Restan1, J.C. Silva Filho1, B.C. Rodrigues1, M. Fenerich1, E.R. Carvalho1, M.G. de Sousa2, A.A. Camacho1

1UNESP – Jaboticabal, Jaboticabal, Sao Paulo, Brazil, 2University of Estadual Paulista UNESP/Jaboticabal, Jaboticabal, Sao Paulo, Brazil

Brachycephalic syndrome is characterized by primary and secondary upper respiratory tract abnormalities, which may result in significant upper airway obstruction. Cardiovascular and hemodynamic changes may also be present. However, unlike upper respiratory abnormalities which have been intensely studied, there is a paucity of information regarding cardiovascular and hemodynamic changes related to the syndrome.

To that purpose, one evaluated 19 animals selected in two distinct groups: Control Group (CG) comprised of 6 beagles with mean age of 2 years, and Brachycephalic group (BG) consisted of 13 French bulldog with brachycephalic syndrome, aging 2.5 years. Thoracic radiographs, arterial blood gas analysis and echocardiographic examination were performed. Data were submitted to normality test and statistical differences in between groups were assessed by using t‐test (P ≤ 0.05).

BG presented PaO2 (89.4 ± 9.9 mmHg) and PaCO2 (34.7 ± 2.1 mmHg) below and above CG (104.9 ± 5.2 mmHg; 32.4 ± 1.7 mmHg), respectively (P ≤ 0.05). VHS index in BG (11.9 ± 0.9) was higher when compared to CG (10.9 ± 0.2) (P ≤ 0.05). BG showed better performance on M‐mode regarding LVIDs (15.4 ± 3.5 mm) and FS (44.8 ± 9.5%) when compared to CG (19.1 ± 3.0 mm; 33.2 ± 4.7%, respectively) (P ≤ 0.05). The LA/Ao ratio was also higher in BG (1.43 ± 0.1) than in CG (1.26 ± 0.1) (P ≤ 0.05).

Based on our results, brachycephalic dogs presented signs of cardiorespiratory disorders, demonstrated by the lower PaO2 values. Furthermore, their heart silhouette was considered bigger than CG (higher VHS index), with signs of right‐side cardiac remodeling. However, no significant variations were observed on echocardiographic examination which could indicate signs of cardiac remodeling in both groups. The best performance obtained by the BG based on LVIDs and FS% variables could be explained by catecholamine release since there was no chemical restrain during examination procedure and these animals were more stressed at the examination table. It is believed that, since the animals on BG group were all young adults, with a mean age of 2.5 years, they did not have enough time to manifest all the complications, including cardiovascular and hematological, related to the syndrome. It is our belief that with aging thus, with progression of the condition, right‐side cardiac remodeling, and alterations in blood gas and arterial blood pressure would be more evident. Even so, depending on severity, development of pre‐capillary pulmonary arterial hypertension may also occur. Our results suggest brachycephalic syndrome may have effects over the cardiovascular system. Therefore, we encourage the need of a thorough investigation to better understand the cardiovascular changes related to brachycephalic syndrome in order to distinguish them from specific heart conditions, especially when attaining predisposing dog breeds.

Dobutamine Stress Test in Dogs with Degenerative Mitral Valve Disease

B.C. Rodrigues1, E. Zacché1, J.C. Silva Filho1, R.M. Canola1, J. Braz1, M. Fenerich1, E.R. Carvalho1, A.Z. Restan1, M.G. de Sousa2, A.A. Camacho1

1UNESP – Jaboticabal, Jaboticabal, Sao Paulo, Brazil, 2University of Estadual Paulista UNESP/Jaboticabal, Jaboticabal, Sao Paulo, Brazil

Mitral valve myxomatous degenerative disease (MVMD) is the main heart disease in dogs, whose evolution leads to significant secondary hemodynamic changes with activation of cardiac and neurohormonal compensatory mechanisms. Pharmacological stress with dobutamine was performed aiming detection of possible changes in the myocardial reserve and systolic dysfunction, which are usually not detected during the conventional echocardiographic exam. Therefore might help veterinarians with early diagnosis of these disorders and thus, establishing a prognosis and appropriate treatment for the patient.

For this purpose, 18 beagles, male and female, older than 6 years old were evaluated. These animals were selected in 3 groups: control group (G1) comprised of 6 animals and 2 other groups composed by dogs with MVMD. The second group (G2) with 7 animals, had the left ventricular internal diameter ratio in diastole indexed to aorta (DIVEd / Ao) less than 2.8 and the third group (G3) with 5 animals and DIVEd / Ao above 2.8.

Echocardiography was performed before and after pharmacological stress with dobutamine. The protocol for pharmacological stress was a continuous infusion of 5 μg/kg/minute during 5 minutes, followed by 10 μg/kg/minute for 10 minutes. The results were based on the difference between pre and post pharmacological stress values of shortening fraction (FS%) and ejection fraction (EF%) variables, obtaining means of each groups and comparing them to each other.

Statistical analysis was performed by variance analysis followed by the Tukey′s test. Regarding FS%, there was no statistical difference (P < 0.05) among G1 and G2 (0.44 ± 0.14 and 0.49 ± 0.13, respectively). However, G1 and G3 (0.44 ± 0.14 and 0.19 ± 0.14) and G2 and G3 (0.49 ± 0.13 and 0.19 ± 0.14) showed statistical difference. Along with EF% variable, there was no statistical difference among G1 and G2 (0.25 ± 0.08 and 0.30 ± 0.08) and G1 and G3 (0.25 ± 0.08 and 0.13 ± 0.09), but there was difference between G2 and G3 (0.30 ± 0.08 and 0.13 ± 0.09).

Based on our results, it was possible to infer that pharmacological stress test with dobutamine did not evidence changes in systolic function in G2 animals with mild cardiac remodeling when compared to G1. However, animals with severe remodeling (G3) the systolic function differed when compared to other groups, suggestive of systolic dysfunction.

Speckle Tracking in Dogs with Myxomatous Mitral Valve Disease Submitted to Pharmacological Stress Test

J.C.S. Filho1, M.G. de Sousa2, E. Zacché1, B.C. Rodrigues1, R.M. Canola1, J. Braz1, M. Fenerich1, E.R. Carvalho1, A.Z. Restan1, A.A. Camacho1

1UNESP – Jaboticabal, Jaboticabal, Sao Paulo, Brazil, 2University of Estadual Paulista UNESP/JABOTICABAL, Jaboticabal, Sao Paulo, Brazil

Myxomatous mitral valve disease (MMVD) is a more common heart disease in dogs, leading to valve heart failure and congestive heart failure.

The evaluation of myocardial function through two‐dimensional speckle tracking has been shown to be an effective method for the early identification of ventricular dysfunction, being safe and feasible for evaluation of myocardial function in dogs with suspected of some form of latent myocardial dysfunction not evidenced by conventional echocardiographic examination. Stress echocardiography with dobutamine is commonly indicated to detect cardiac dysfunction in asymptomatic patients, evaluating myocardial contractile reserve and viability. Pharmacological stress test was performed with dobutamine to evaluate the myocardial function via two‐dimensional speckle tracking in healthy dogs and in different stages of the MMVM. The study included 15 Beagle dogs, distributed in 3 groups and selected according to the internal diameters of the left ventricle in diastole and the aortic ratio (DIVEd/Ao). The groups were constituted: Group I (Control) – 5 healthy dogs without MMVD; Group II – 5 dogs with MMVD (DIVEd)/Ao ratio lower than 2.8; Group III – 5 dogs with MMVD and DIVEd/Ao ratio higher than 2.8. The animals were submitted to the echocardiographic examination performed by a single operator and evaluation was performed before and after dobutamine infusion. Dobutamine was diluted in 5% glucose solution, following the progressive scheme: 5 μg/kg/minute for 3 consecutive minutes; 7.5 μg/kg/minute for 3 consecutive minutes; 10 μg/kg/minute for 3 consecutive minutes. The images were analyzed using the Syngo Velocity Vector Imaging (VVI) algorithm optical flow algorithm (SIEMENS®). Data were submitted to the normality test and to repeated measures analysis of variance and Tukey's test.

Cardiovascular Echocardiography Aspects During Pregnancy in Female Dogs

V.T. de Almeida1, R.A.R. Uscategui1, V.J.C. dos Santos1, A.A. Camacho2, M.G. de Sousa1, R.M. Canola2, A.P.R. Simões1, M.L. Avante1, W.R.R. Vicente1

1University of Estadual Paulista UNESP/JABOTICABAL, Jaboticabal, Sao Paulo, Brazil, 2UNESP – Jaboticabal, Jaboticabal, Sao Paulo, Brazil

The aim of this study was to evaluate the adaptations involving the left ventricle, across the gestational process and its diastolic function, using tissue Doppler echocardiography. Animals: 15 pregnant bitches, healthy aged 2–6 years and weighing between 5 and 15 kg. Heart rate and blood pressure clinical evaluation was performed, as well as echocardiography M‐mode, spectral and tissue Doppler, in the estrous (EZ), every 15 days during pregnancy (PD 15, 30, 45 and 60) and 45 days post partum (PP). The variables from the evaluation were submitted to statistical analysis. Heart rate (HR) (P = 0.149) and systolic blood pressure (SAP) (P = 0.564) did not show significant changes during pregnancy. The parameters that have undergone significant changes were EF (P = 0.0016) and FS (P = 0.0019). AVVmax, PVmax, wave E’(s) S’ (s) and S’ (lat) being influenced by the gestational period (P < 0.05). These changes could be explained by cubic regression model. The E:A (P = 0.041) IVRT (0.044) relation and A ‘(lat) (0.023) also suffered influence of pregnancy, but had a nonparametric residual distribution. Pregnancy produces changes in maternal hemodynamics also influencing diastolic function. Myocardial contractility increases as a result of an improvement in systolic performance, while diastolic function changes significantly revealing a more compliant ventricle, which is the physiological response to increased volume and metabolic requirements due to gestation.

TAPSEidx Echocardiographic Evaluation in Dogs with Pulmonary Hypertension Pre‐Capillary and Post‐Capillary

J. Braz1, R.M. Canola1, A.Z. Restan1, J.C. Silva Filho1, E. Zacché1, M.G. de Sousa2, A.A. Camacho1

1UNESP – Jaboticabal, Jaboticabal, Sao Paulo, Brazil, 2University of Estadual Paulista UNESP/JABOTICABAL, JABOTICABAL, Sao Paulo, Brazil

Pulmonary hypertension (PH) is a hemodynamic and pathophysiologic condition defined an increase in systolic pulmonary artery pressure greater than 30 mmHg. Classically, this condition can lead to right ventricular dysfunction, which is associated to poor outcomes, especially in patients with chronic congestive heart failure. Unfortunately, right ventricular function in dogs with pulmonary arterial hypertension has been poorly investigated. Thus, the aim of the present study was to assess the right ventricular systolic function in dogs with tricuspid regurgitation and pulmonary hypertension by echocardiographic evaluation.

To that purpose, one evaluated 35 dogs with echocardiographic diagnosis of tricuspid regurgitation and PH, calculating the transtricuspid gradient regurgitation through Bernoulli modified equation and then adding an assumed right atrial pressure. Animals were selected into 5 groups: Pos‐capillary PH group, composed by dogs with myxomatous mitral valve disease and clinical or echocardiographic congestion signs; Pre‐capillary PH group including dogs with pulmonary disease. Each group was subdivided in Mild PH group, including dogs with pulmonary arterial pressure gradient estimated below 50 mmHg; and moderate/severe PH group, comprised by animals with pulmonary arterial pressure gradient over 50 mmHg; Pre‐capillary mild group was composed by 14 dogs; pre‐capillary PH moderate/severe group 6 dogs; pos‐capillary PH mild group 6 dogs and pos‐capillary PH moderate/severe group 9 dogs. All PH groups were comprised by animals between 5 and 16 years, 2.8 and 27.6 Kg and pulmonary arterial pressure estimated between 36 and 96 mmHg. These animals did not receive antihypertensive treatment until the time of examination. Control group with 6 clinically healthy dogs with no echocardiographic abnormality, 2 years old and weight between 9.1 and 12.7 Kg.

Echocardiographic evaluation of RV was performed in all dogs and followed the recommendation of the Guidelines for the echocardiographic assessment of the right heart endorsed by the European Association of Echocardiography. The parameter assessed was tricuspid annular plane systolic excursion (TAPSE), acquired by placing an M‐mode cursor through the tricuspid annulus and measuring the amount of longitudinal motion of the annulus at peak systole an apical 4‐chamber window. Once TAPSE varies according patient′s weight it was indexed by aortic diameter (TAPSEidx). The data was submitted to normality test, the analysis between groups was performed by one way ANOVA test (P < 0.05) and means were compared by Tukey test.

There was statistical difference between control and pre‐capillary moderate/severe groups (P < 0.05), and from pos‐capillary moderate/severe and control groups (P < 0.05) when evaluating TAPSEidx parameter. Results are summarized in Table 1.

Statistical difference detected by TAPSEidx among control group and pre and post‐capillary moderate/severe groups may suggest that patients with moderate/severe PH have right ventricular systolic dysfunction.

Pre Clinical Results of LASSBIO294: An Inodilator Compound

R. de Oliveira Alves Carvalho1, A.P.A. Costa1, P.R. Nasciutti1, R.V. Silva1, B.D. Matos1, N.G. Melo1, G.M.M. de Andrade1, E. de Jesus Barreiro2, V. de Oliveira1

1Federal University of Goias, Goiania, Goias, Brazil, 2Federal University of Rio de Janeiro, Rio de Janeiro, Rio de Janeiro, Brazil

LASSBio 294 is a compound, developed by the Laboratory of Evaluation and Synthesis of Bioactive Substances (LASSBio®) of the Federal University of Rio de Janeiro (UFRJ) made from the substrate safrole, a compound extracted from sassafras oil, found in Brazilians plants such as the cinnamon‐white (Ocotea pretiosa). The inodilator effect of the compound has been demonstrated in bioassays, using isolated and intact aortic rings, and in whole hearts and fractions of myocardial of rats, suggesting that the mechanism of action for its vasodilatory effects is related to inhibition of phosphodiesterases, and the inotropic effect on the increased Ca 2 + uptake by the sarcoplasmic reticulum. In rats with experimentally induced myocardial infarction, LASSBio 294 decreased collagen deposition, presumably regulating inflammation mediators. In addition, it increased cardiac lusitropism, decreasing left ventricular filling pressure, modulating the action of the SERCA protein, increasing Ca 2 + uptake by the sarcoplasmic reticulum, inducing myocardial relaxation and improving diastolic function. In previous studies performed by these authors on healthy beagle dogs, the oral administration of the compound lowered blood pressure, without causing hypotension or other toxic effects and increased the shortening fraction by 19.1%. Faced with the need to test the compound in a non‐rodent mammal with myocardial disease, it was proposed in this study to verify the action of LASSBio 294 (2 mg/kg, n = 7) on the cardiovascular parameters of New Zealand rabbits with dilated cardiomyopathy induced by doxorubicin, using pimobendan (3 mg/kg, n = 6) as a positive control, both treatments were administered every 12 hours for 30 days. DCM was induced by intravenous administration of 1 mg/kg of doxorubicin, twice a week for three weeks and weekly thereafter until FS = 25% or less was achieved. The animals were evaluated weekly by: electrocardiography, echocardiography, blood pressure measurement, chest X‐ray and biomarkers of renal and hepatic function and myocardial injury. This study was approved by the Ethics Committee of Animal Use of the Federal University of Goias (UFG) under the protocol number 006 / 14. The data were submitted to analysis of variance and compared by means of Tukey test, adopting 5% of significance. Under these conditions, LASSBio 294 increased systolic function, increasing FS by 35% in the second week, and improved diastolic function in a superior manner to pimobendan, maintaining the E /A ratio <1 and E ‘/ A’ <1 and reducing IVRT, without altering the arterial pressure or producing proarrhythmogenic or toxic effects and also, reduced the serum concentration of creatinine. However, it did not prevent the evolution of the congestive condition, since two animals presented pulmonary edema and effusions. Thus, the pre‐clinical in vitro and in vivo testing stages of LASSBio 294 was concluded, and it was tested in rodent and non‐rodent mammals, healthy and with induced DCM, with a positive inodilator and lusitropic effect. Therefore, it can be concluded that LASSBio 294 is a promising compound, which needs to be used in clinical trials with individuals with naturally occurring DCM, so that it can complete the necessary trials for its rise as drug, allowing an increase in the existing therapeutic protocols.

Clinical and Echocardiographic Features, and Survival of French Bulldogs with Pulmonic Stenosis (66 Cases: 2005–2016)

C. Damoiseaux1, L. Desquilbet2, V. Gouni3, M. Lavennes1, C. Poissonnier1, S. Bertrand4, M. Chenebaux4, J.‐L. Pouchelon1, V. Chetboul3

1National Veterinary School of Alfort/Alfort Cardiology Unit, Maisons‐Alfort, Ile‐de‐France, France, 2National Veterinary School of Alfort/ Department of Biostatistics and Clinical Epidemiology, Maisons‐Alfort, Ile‐de‐France, France, 3Alfort National Veterinary School/Alfort Cardiology Unit, Maisons Alfort, Ile‐de‐France, France, 4Alfort National Veterinary School, Maisons Alfort, Ile‐de‐France, France

Pulmonic stenosis (PS) is one of the most common congenital heart defects in dogs, affecting various canine breeds including English and French Bulldogs (FB). PS has been extensively described in English bulldogs. However, to the best of our knowledge, no specific study has been dedicated to PS in FB. The objective of this prospective observational study was therefore to document the epidemiological, clinical, and echocardiographic features, and risk factors for death in FB with PS. The study population consisted of 66 FB. In most cases (53/66, 80%), at least 2 obstructive lesions were observed (most commonly (42/66, 64%) valvular and supravalvular), using color‐flow Doppler mode and 3‐dimensional echocardiography. The median Doppler‐derived peak transvalvular pulmonary gradient (DG) was high: 170 mmHg (lower‐upper quartiles: 139–223 mmHg), with a DG ≥200 mmHg in 33% of FB. Among the 51 FB for which a follow‐up was available, 21/51 (41%) died, 67% (14/21) of all deaths being cardiac‐related (CD) at a median age of 3.5 years. In univariate analyses, age (HR=2.2 per 1 year; P = 0.02), clinical signs at presentation (HR=3.7; P = 0.03), DG (HR=1.2 per 10 mmHg; P = 0.01), right ventricle dilation (HR=5.0; P = 0.04), moderate to severe tricuspid regurgitation (HR=7.6; P < 0.01), and right‐sided congestive heart failure (HR=4.8; P = 0.05) were significantly associated with time to CD. After adjustment for age and DG, tricuspid regurgitation remained significantly associated with time to CD (HR=5.1; P = 0.02). In conclusion, PS in FB are commonly severe and complex, with tricuspid regurgitation being an independent predictor of CD.

Comparative Multidimensional Imaging of Patent Ductus Arteriosus in Dogs

K.R. Doocy, A. Saunders, S. Gordon

College of Veterinary Medicine & Biomedical Sciences, Texas A&M University, College Station, TX, USA

Accurate measurements are integral for device selection and sizing to minimize complications associated with minimally invasive transcatheter closure of patent ductus arteriosus (PDA), in particular regarding minimal ductal diameter (MDD) to device ratio depending on method of measurement. Standard measurements are typically obtained with two dimensional (2D) angiography and echocardiography. The aims of this study were to characterize PDA dimensions including MDD, ampulla diameter 3 mm above the MDD and morphology based on right lateral angiography, transthoracic echocardiography (TTE) from the right and left imaging windows and 2D, biplane, and 3D transesophageal echocardiography (TEE).

The study population consisted of 25 dogs with left‐to‐right shunting PDA; 20 had minimally invasive closure and 5 had surgical ligation (see table).

The 2D TTE and angiographic MDD measurement is acquired in the same orientation as 3D TEE measurement A. A was the smallest of the 2 perpendicular 3D measurements in 23/24 (96%) dogs. The ratio of A/B indicated a circular (0.05 mm difference between A and B) shaped ampulla in 5/24 (21%) and MDD in 7/24 (29%). Real‐time 3D TEE documented a dynamic change in the pulmonary ostium dimensions throughout the cardiac cycle. In 21 dogs with angiography performed prior to catheter‐based closure, morphology was classified as type IIA (16), type IIB (3), type III (1, subsequently ligated), and other (1, similar to Krichenko type D). All 7 dogs with circular MDD on 3D TEE were classified as type IIA on angiography. Median Amplatz® canine duct occluder (ACDO) size used in 20 dogs was 7 (range, 3–10). Median (range) CDO to MDD ratio was 2.04 (1.48–3.07) with angiography and 2.09 (1.25–4.00) with 2D TEE. Median (range) CDO proximal disc to ampulla ratio was 1.68 (1.26–2.78) with angiography and 1.76 (1.07–2.63) with 2D TEE. The measurement deemed most accurate including 3D TEE was used to make device selection in individual cases. No complications occurred in any dog.

In conclusion, 3D TEE provides a clear cross sectional image of the PDA pulmonary ostium and documented an elliptical shape MDD in the majority (71%) with the minor axis (A) similar in location to the angiographic and 2D TEE measurement which could contribute to device under sizing. Based on standard 2D TEE and angiographic imaging, the ACDO was oversized by approximately 2 times the MDD as previously reported, and the proximal disc to ampulla ratio was always at least 1.

Changes In Neurohumoral Factors in 9 Dogs That Received Carperitide and Presented with Pulmonary Edema

M. Enokizono1, N. Kanno2, H. Suzuki3, S. Yamada4, K. Matsuura3

1Japan Small Animal Medical Center, Tokorozawa, Saitama, Japan, 2Nihon university, Fujisawa, Kanagawa, Japan, 3Animal cardiovascular and Thoracic Surgery Center, Sayama, Saitama, Japan, 4Animal Cardiovascular and Thoracic Surgery, Fuchuu, Tokyo, Japan, 5Animal cardiovascular and Thoracic Surgery Center, Sayama‐shi, Saitama, Japan

The purpose of this study was to clarify the treatment effects of carperitide in veterinary cases. Nine dogs presenting with pulmonary edema due to mitral valve insufficiency were treated with a continuous infusion of carperitide (0.05 μg/kg/min). Dogs with hypotension were excluded from the study. Blood sampling was performed twice: once before the administration of carperitide (Pre) and then after the administration of carperitide (Post). Neurohumoral factors, such as atrial natriuretic peptide (ANP), N‐terminal pro b‐type natriuretic peptide (NT‐proBNP), and aldosterone levels, were measured at both time points. In addition, chest radiography, echocardiography, and blood pressure measurements were performed depending on the specific cases. Seven of the 9 dogs showed improvements post‐treatment, although 2 cases did not show any improvement. ANP levels increased post‐treatment in all cases (mean ± standard deviation: pre = 292 ± 163 vs. post = 908 ± 866). NT‐proBNP levels increased in 4 cases and decreased in 5 cases (pre = 6563 ± 2061 vs. post = 5627 ± 3026). Aldosterone level increased only in one case (pre = 232 ± 172 vs. post = 124 ± 130) In addition, the left atrial aortic root ratio (LA/Ao), E‐wave, vertebral heart score (VHS), and blood pressure decreased. Carperitide is a human ANP agent, which is why we expected higher ANP levels post‐treatment. We observed a decrease in NT‐proBNP levels in all cases that were treated for more than 7 hours. This is thought to be due to a time lag until the half‐life period of NT‐proBNP and the stretch stimulation of the ventricular muscle, which are reflected in the blood concentration of NT‐proBNP. In addition, the ventricular muscle stretch was inhibited by the decrease in left atrial pressure following the administration of carperitide, which lasted for more than 7 hours. Lastly, the decrease in aldosterone levels was thought to be due to the inhibitory effects of carperitide on the renin‐angiotensin‐aldosterone system. This study suggested that carperitide administration is effective since pulmonary edema is improved and aldosterone is suppressed.

Myocardial Remodeling in Canine Dilated Cardiomyopathy (DCM)

S. Gasparini1, S. Fonfara2, U. Hetzel3, A. Kipar1

1University of Zurich, Zurich, Zurich, Switzerland, 2University of Guelph, Guelph, ON, Canada, 3Institute of Veterinary Pathology, Vetsuisse Faculty, University of Zurich, Switzerland, Zurich, Zurich, Switzerland

Dilated cardiomyopathy (DCM) is one of the most common cardiac diseases in dogs. However, its pathogenesis is still not well understood. Based on the detailed histological and transcriptional assessment of the myocardium in DCM, the present study aimed to investigate the involvement of remodeling processes in the pathogenesis of DCM.

The myocardium of 16 dogs with DCM was examined by histology and immunohistology. Alongside, quantitative reverse transcriptase PCR (qPCR) for cytokines and remodeling markers was performed in four of these cases, in relation to their transcription in six dogs with non‐cardiac diseases. Laser microdissection with subsequent qPCR served to confirm the cardiomyocytes as the source of these markers.

The histological examination confirmed the occurrence of degenerative myocardial changes (“fatty infiltration‐degenerative type” and “attenuated wavy fiber type”), but also revealed a general increase in interstitial cells. In addition, focal lesions were observed. These comprised accumulations of fibroblast‐like cells, mild fibrosis and edema, with neovascularization and degenerating cardiomyocytes. The majority of interstitial cells appeared to be of macrophage origin and were often found to express intracellular adhesion molecule (ICAM‐1), transforming growth factor (TGF)‐b and vascular endothelial growth factor, which also cardiomyocytes expressed frequently. The transcription of ICAM‐1, matrix metalloproteinase (MMP)‐2, tissue inhibitor of MMP (TIMP)‐1, TIMP‐2, lysyl oxidase and tumor necrosis factor (TNF)‐a was significantly increased, and cardiomyocytes were shown transcribe most relevant markers.

The results confirm remodeling processes as a major component in the pathogenesis of canine DCM and suggest that cardiomyocytes at least contribute to the pathogenic processes.

Retrospective Evaluation of Effect of Heart Rate on Survival in Dogs with AF

B. Pedro1, J. Dukes‐McEwan2, M. Oyama, M.S. Kraus3, A.R. Gelzer3

1Willows Veterinary Centre and Referral Service, Solihull, UK / Small Animal Teaching Hospital, University of Liverpool, UK, Solihull, UK, 2Small Animal Teaching Hospital, University of Liverpool, UK, Neston, UK, 3Department of Clinical Studies‐Philadelphia, School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA, USA

Atrial fibrillation (AF) is a common arrhythmia in dogs, usually associated with a rapid ventricular response rate, and typically managed with pharmacological rate control. This retrospective study aimed at assessing the effect of heart rate (HR) on survival time in dogs with AF. Forty‐six dogs from 2 sites with AF and at least one 24 hours Holter recording were included from records reviewed between 2008 and 2016. The mean heart rate (meanHR, 24 hours average), minimum HR (minHR, 1 minute average), maximum HR (maxHR, 1 minute average) were recorded at the time when adequate rate control was presumed and no further adjustments of rate control therapy were made. Long‐term outcome was ascertained and the primary study endpoint was all cause mortality.

All dogs had structural heart disease, 31/46 were in congestive heart failure, 44/46 were receiving anti‐arrhythmic drugs. MeanHR was 125 bpm (range 62–203 bpm), minHR was 82 bpm (range 37–163 bpm) and maxHR was 217 bpm (range 126–307 bpm). Sex, age, weight, cardiac diagnosis, and study site had no significant impact on survival. For every 10 bpm increase in meanHR, the risk of all‐cause mortality increased by 35% (hazard ratio 1.35; 95% CI 1.17–1.55; P < 0.001). Median survival time of dogs with meanHR < 125 bpm (n = 23) was significantly longer (1037 days; range 524‐open) than dogs with meanHR ≥ 125 bpm (n = 23) (105 days; range 67–267 days) (P = 0.0012). Only meanHR was independently associated with survival time (P < 0.003) by multivariate regression. Our results suggest that aggressive rate control in dogs with AF leads to reduced mortality and prospective studies targeting meanHR of 125 bpm or lower are indicated.

Left Atrial Volume Can Detect Congestive Heart Failure in Dogs with Chronic Mitral Valve Disease

M.M. Mantovani1, J. Ribeiro de Castro1, A.M. Gimenes1, L.C. Petrus1, C.N. Duarte1, M.Y. Ueda1, G.T. Goldfeder2, M.H.M.A. Larsson2, D.S. Schwartz1

1School of Veterinary Medicine and Animal Science – University of São Paulo (USP), São Paulo, Sao Paulo, Brazil, 2University of Sao Paulo, São Paulo, Sao Paulo, Brazil

This study aimed to evaluate left atrial volume (LAV) in dogs in different stages of chronic mitral valve disease (CMVD), and to evaluate the association of these values with congestive heart failure (CHF). Our hypothesis was that LAV increases with severity of CMVD and can be used to detect CHF in dogs with CMVD. Eighty dogs were included in a prospective cross‐sectional clinical study, grouped according to the staging of CMVD proposed by the American College of Veterinary Internal Medicine (A, B1, B2 and C), based on clinical signs and echocardiographic evaluation. Left atrial volume was calculated using the biplane area‐length method from the left apical two and four chamber views at the end of ventricular systole. The LAV was indexed to body weight. LAV/kg showed high values (P25%–75%=3.91–6.81) compared with groups A (0.82 mL/kg, P25%–75%=0.64–0.95), B1 (0.91 mL/kg, P25%–75%=0.47–1.29) and B2 (2.23 mL/kg, P25%–75%=1.93–2.85). Analysis of ROC curve for CHF diagnosis showed area under curve of 0.92 for LAV/kg. A LAV/kg >3.26 mL/kg was the optimal cut‐off identified for detecting CHF in dogs with CMVD with a sensitivity and specificity of 90%. In addition, when LAV is greater than this cut‐off, it increases to 90% the probability to detect CHF (an increase of 40% in diagnostic capacity than pre‐test), which allows for the decision to treat for CHF. In conclusion, left atrial volume increases with severity of CMVD and helps the detection of CHF in these dogs.

Serum Levels of ST2 Soluble Fraction in Clinically Healthy Dogs

A.M. Gimenes1, L. Beccari1, J. Ribeiro de Castro1, M.M. Mantovani1, C.N. Duarte1, L.C. Petrus1, M.H.M.A. Larsson2, D.S. Schwartz1

1School of Veterinary Medicine and Animal Science – University of São Paulo (USP), São Paulo, Sao Paulo, Brazil, 2University of Sao Paulo, São Paulo, Sao Paulo, Brazil

The soluble fraction of ST2 (sST2) is the circulating isoform of ST2, a member of the interleukin‐1 (IL‐1) receptor family. The sST2 is induced when cardiomyocytes and cardiac fibroblasts undergo mechanical strain, making it a promising biomarker of fibrosis and cardiac remodeling. sST2 has been widely studied in human medicine, and characterized as an important prognostic biomarker and an independent mortality predictor. However, there are no studies determining reference values or evaluating the influence of gender and age on sST2 in healthy dogs. Therefore, this study aimed to measure serum levels of sST2 in healthy dogs, in order to determine normal values for the study population and to evaluate the influence of gender and age on this biomarker concentration for future assessment of dogs with heart disease. With this purpose, a prospective cross‐sectional clinical study was conducted, and included a sample of 60 healthy small‐to‐medium size dogs, males and females, adults of different ages. There was no significant difference in sST2 serum levels between males (22) and females (38) (P = 0.980) or between age groups in the study population (P = 0.358). The reference values found for the serum levels of sST2 in healthy dogs ranged from 36.94 to 44.67 pg/mL (CI 95%), with a mean value of 40.80 pg/mL. There was no influence of gender or age group on serum levels of sST2 for the assessed dog population. It was possible to suggest a reference range for the biomarker in healthy dogs.

Soluble ST2 Biomarker in Heart Failure Secondary to Chronic Mitral Valve Disease in Dogs

A.M. Gimenes1, J. Ribeiro de Castro1, M.M. Mantovani1, C.N. Duarte1, L.C. Petrus1, M.Y. Ueda1, G.T. Goldfeder2, M.H.M.A. Larsson2, D.S. Schwartz1

1School of Veterinary Medicine and Animal Science – University of São Paulo (USP), São Paulo, Sao Paulo, Brazil, 2University of Sao Paulo, São Paulo, Sao Paulo, Brazil

The soluble fraction of ST2 (sST2) is considered a cardiac biomarker, which is induced mechanically by the stretch of cardiomyocytes and fibroblasts. This biomarker has prognostic value in human cardiology, and it is considered an important independent mortality predictor. However, studies assessing sST2 in dogs are still lacking. Therefore, this study aimed to determine the utility of sST2, compared to NT‐proBNP and cardiac troponin I (cTnI), for assessment of dogs with chronic mitral valve disease (CMVD). With this purpose, 151 dogs in different stages of CMVD were submitted to clinical exam, laboratory tests, cardiologic exams and measurement of sST2, NT‐proBNP and cTnI. The results demonstrate that sST2 was higher in dogs in CHF compared to non‐CHF dogs. However, when dogs were stratified according to CMVD stages, it was not possible to differentiate the stages based on sST2 concentrations. NT‐proBNP and cTnI were higher in CHF dogs, as well as in more advanced stages of CMVD. There was correlation between sST2 and NT‐proBNP and cTnI. Analysis of ROC curve for CHF diagnosis showed areas under the curve (AUC) of 0.66 for sST2, 0.86 for NT‐proBNP and 0.76 for cTnI. As for prognosis assessment, sST2 concentrations higher than 40 pg/mL indicated higher mortality compared to groups with lower sST2 values. Results suggest that sST2's performance as a standalone diagnostic test is limited and inferior to the other studied biomarkers. However, sST2 seems to be a useful independent prognostic marker, and mortality predictor in dogs with CMVD.

Prevalence of Heart Disease in Dogs at a Teaching Hospital: A Retrospective Populational Study

V. Dayoub, P. Itikawa, G.T. Goldfeder, A.E.M. Gonçalves, B. Silva, R.B. Pessoa, N. Cardoso, S. Enumo, M.H.M.A. Larsson

University of Sao Paulo, São Paulo, Sao Paulo, Brazil

Chronic valvular heart disease (CVHD) is the most common morbidity, of great relevance in the internal medicine of small animals. The scope of the present study was to evaluate the main diagnoses performed in an 11‐year (2005–2015) period at the cardiology unit of a teaching hospital. In the period of the study, 11,921 dogs were attended, with gender balance: 51% males and 49% females. Mongrel dogs and Poodles represented both 25% of the animals. The mean age of the patients was 129 months, of which 45% (5,387) sought the unit for pre‐anesthetic evaluation. Excluding pre‐anesthetic evaluation, CVHD accounted for 64% of the diagnoses, healthy dogs 19%, congenital anomalies 3%, arrhythmias 3%, pericardial effusion 3% (178), being 148 (83%) with cardiac base neoplasia, dilated cardiomyopathy 2%, and 5% left including tracheal collapse, bronchitis, and dirofilariasis. Between 2005 and 2010, 54 animals were diagnosed with cardiac base neoplasia, while between 2010 and 2015, 94 dogs were diagnosed, representing an increment of 74% of cardiac neoplasias. From 141 cases with dilated cardiomyopathy, 97 were attended between 2005 and 2010, and 44 between 2010 and 2015, showing a decrease of 121% in the last period. The data obtained are in agreement with the literature that reveals CVHD and congenital anomalies as the most frequent cardiopathies. The significant decrease in the prevalence of CMD may be caused by despopularization of the breeds predisposed to DCM, while the increase of cardiac base neoplasias may be theorized by the improvement in canine nutritional management increasing survival of companion animals.

Assessment of Left Atrial Function by Speckle‐Tracking Echocardiography in Dogs with Myxomatous Mitral Valvular Disease

M.H. Kang, S.‐W. Han, H.‐M. Park

College of Veterinary Medicine, Konkuk University, Seoul, Seoul‐t'ukpyolsi, Republic of Korea

Left atrial function is well known prognostic factor in dogs with myxomatous mitral valvular disease (MMVD). Left atrial functional analysis with advanced echocardiographic technique such as speckle‐tracking echocardiography (STE) have not been performed in affected dogs. The purpose of this retrospective study was to evaluate left atrial STE variables associated with disease severity and to compare it with conventional left atrial and ventricular echocardiographic parameters.

Retrospective review of medical records for dogs with MMVD was performed. Client‐owned forty‐eight dogs were included in this study. Every dog was presented for evaluation of heart disease: 34 dogs with MMVD (according to mitral regurgitation(MR) severity: mild, n = 12; moderate, n = 11; and severe groups, n = 11) and 14 healthy age and weight‐matched controls. The conventional echocardiographic parameters including fractional shortening (FS), ejection fraction (EF), E velocity, E/A ratio, E/E’ ratio, left atrium to aorta ratio (LA/Ao), end‐systolic volumes indexed to body surface area in B‐mode (EDVIB), end‐diastolic volumes indexed to body surface area in B‐mode (ESVIB) and PW TDI parameters were recorded. The left atrial speckle tracking imaging including global peak atrial longitudinal strain (PALS) and global peak atrial contractile strain (PACS) was acquired from two‐dimensional cine loops of apical four‐chamber view.

Global PALS was increased in the mild MR group, reduced in the moderate MR group and further reduced in the severe MR group. Global PACS was reduced in moderate and severe MR groups. Global PALS and PACS correlated with LA/Ao, EDVIB, ESVIB, FS, EF, E velocity, and E/E’. Dogs with diastolic dysfunction had reduced global PALS when plotted for left atrial size.

The STE is repeatable and reproducible tool and it can have more important role in managing advanced heart disease in dogs than the left atrial size parameters, which may underestimate the severity of disease in end‐stage heart disease.

Acute Echocardiographic Effects of Sotalol on Ventricular Function and Ectopy in Dogs with Ventricular Tachyarrhythmias

J.L. Kaplan1, L.C. Visser2, S. Nishimura3, C.T. Gunther‐Harrington4, C. Belanger3, M.S. Mueller5, J.A. Stern2

1University of California, Davis, Davis, CA, USA, 2Department of Medicine and Epidemiology, University of California, Davis, Davis, CA, USA, 3Veterinary Medical Teaching Hospital, University of California Davis, Davis, CA, USA, 4University of California, Davis Veterinary Medical Teaching Hospital, Davis, CA, USA, 5University of California, Davis/Blue Pearl Specialty and Emergency Pet Hospital, Renton, WA, USA

We sought to determine the acute effects of sotalol (administered orally) on left ventricular (LV) systolic function and ventricular ectopy in dogs with ventricular tachyarrhythmias.

Dogs (n = 27) diagnosed with stable ventricular tachyarrhythmias were prospectively enrolled. Dogs had an echocardiogram and 5‐minute ECG before and 2–3 hours post‐sotalol. Systolic function was assessed during sinus rhythm via fractional shortening (FS), LV shortening area (LVSA), and modified Simpson's ejection fraction (EF). Number of ventricular premature complexes (VPCs)/5‐min were tabulated and ventricular arrhythmias were graded: 0 = no VPCs, 1 = single VPCs, 2 = accelerated idioventricular rhythm, 3 = bigeminy/trigeminy, 4 = couplets/triplets, and 5 = ventricular tachycardia or R‐on‐T phenomenon. Echocardiographic and ECG measurements were performed in a blinded fashion. Number of VPCs/5‐min, complexity grade, indices of LV systolic function and sinus heart rate were compared using a paired t‐test (or Wilcoxon signed‐rank test).

Baseline heart rate, FS, LVSA, and EF were all significantly (P < 0.001) decreased post‐sotalol (2.5 ± 0.2 mg/kg PO) with mean percent changes ± SD of −18 ± 16%, −14 ± 11%, −10 ± 12%, −13 ± 16%, respectively. Median, IQR of VPCs/5‐min (18, 9–32 vs. 4, 0–17) and arrhythmia grade (1, 1–4 vs. 1, 0–1) were also significantly (P < 0.01) decreased post‐sotalol.

The reduction in LV systolic function post‐sotalol is relatively mild in most dogs and percent change post‐sotalol may not exceed previously reported within‐day coefficients of variation for echocardiographic LV systolic function indices. Based on a 5‐minute ECG, sotalol acutely improved ventricular ectopy.

Speckling‐Tracking Echocardiographic Alterations in Dogs with Chronic Mitral Valvular Disease and Concurrent Renal Disorders

Y. Kim1, C. Park2, J. Kim3, M.‐Y. Kim4

1College of Veterinary Medicine, Chonbuk National University, Curi, Kyonggi‐do, Republic of Korea, 2College of Veterinary Medicine, Chonbuk National University, Iksan, Cholla‐bukto, Republic of Korea, 3Taegu Plus Animal Medical Center, Buk‐gu, Taegu‐jikhalsi, Republic of Korea, 4Chonbuk National University, Chonbuk, Taejon‐jikhalsi, Republic of Korea

Chronic mitral valvular disease (CMVD) is the most common acquired canine heart disease, characterized by resultant volume overload firstly leading to left atrial enlargement. Occasionally, dogs with CMVD concomitant with renal insufficiency are found. Pathological interactions exist between cardiovascular diseases and renal disorders, which was recently defined as cardiovascular‐renal disorders (CvRD) in veterinary medicine. Although the mechanisms of CvRD are not yet fully understood, cardiac diseases and renal dysfunction have recognized as both important determinant of prognosis to each other. Of tools for evaluation of cardiac function, speckle‐tracking echocardiography (STE) is newer technique that allows more accurate assessment of left atrial (LA) as well as left ventricular (LV) functions. In the present study, the aims were to evaluate LA and LV function using STE, to assess the correlation between the class of cardiac and renal disease, and to investigate the correlation among the class of renal disease, echocardiographic variables, and laboratory variables in different stage of CMVD dogs and CMVD dogs concomitant with renal disorders.

Forty‐one client‐owned dogs with CMVD were included. All the dogs underwent physical examination, blood works, radiographs, abdominal ultrasound, urinalysis, electrocardiogram, and echocardiography. Dogs were categorized according to American College of Veterinary Internal Medicine (ACVIM) consensus statement and International Renal Interest Society (IRIS) staging system. 2D, M‐mode, and Doppler examinations were performed for conventional echocardiography. Using STE, strains of circumferential, radial, and longitudinal motions, LV twist, LA volume, and LA strain were assessed. Systolic twist increased to moderate stage (ACVIM C class) and decreased in severe stage (ACVIM D class). Dogs in advanced stage had increased circumferential and radial strain, whereas decreased longitudinal strain compared with earlier stages. LA maximal and minimal volume increased, and left atrial active emptying index, LA‐FACact and peak LA strain rate during atrial contraction were decreased with severity. Higher concentrations of serum creatinine and serum urea were found in ACVIM class D. IRIS stage had statistically significant correlation with ACVIM class and several echocardiographic variables including early diastolic untwisting rate and E/A.

In this study, the significant correlation among IRIS stage, serum creatinine concentration, echocardiographic variables, and ACVIM class was shown. This finding suggests connection between chronic mitral valve disease and renal disorder.

Comparative Evaluation of Dogs with Chronic Mitral Valve Regurgitation Treated with Amlodipine or Pimobendan

O. Kamakura, P. Itikawa, G.T. Goldfeder, R.B. Pessoa, M.H.M.A. Larsson

University of Sao Paulo, São Paulo, Sao Paulo, Brazil

This prospective, randomized, double blinded study has the purpose to evaluate the echocardiographic parameters in dogs with chronic mitral valve degeneration (CMVD) stage C treated with pimobendan or amlodipine in addition to conventional therapy. The use of pimobendan in stage C dogs is well established, but has been debated due to the apparent preservation of the myocardial inotropism in these patients. Amlodipine is a vasodilator drug with no inotropic effects, usually used in stage D dogs. Ten dogs were randomized to group A and eleven to group B. Both groups were treated with furosemide and enalapril maleate, but group A dogs also received pimobendan (0.25 mg/kg BID) and the group B ones, amlodipine (0.1 mg/kg BID). All dogs were evaluated before the beginning of the medications (T0), thirty days later (T30) and sixty days after T0 (T60), observing changes in echodopplercardiographic and electrocardiographic examinations, as well as systolic blood pressure. There were no significant differences in electrocardiographic or systolic blood pressure measurements between the groups. The echodopplercardiographic examinations showed a statistically significant difference in speeds in Em waves, between groups at T0 and in the amlodipine group between T0 and T60 and speed of Am wave, between T30 and T60 times. These parameters may indicate a change in early myocardial relaxation. This is the first prospective randomized double blinded study comparing pimobendan and amlodipine in stage C CMVD dogs. Despite the relative small sample size, with the findings of this study, we cannot conclude superiority of one drug over another.

Systolic Blood Pressure Obtained by Doppler Method in Dogs: Comparison Between Anatomical Positions and Evaluators

A.E.M. Gonçalves1, P. Itikawa1, G.T. Goldfeder1, A.M. Gimenes2, M.H.M.A. Larsson1

1University of Sao Paulo, São Paulo, Sao Paulo, Brazil, 2School of Veterinary Medicine and Animal Science – University of São Paulo USP, São Paulo, Sao Paulo, Brazil

Because obtaining accurate systolic blood pressure (SBP) measurement is important in clinical practice, a comparison of SBP measurements obtained by Doppler method in dogs when cuff was placed on the left forelimb and tail, by three examiners with different degrees of experience, was made in order to evaluate the correlation of values between the anatomical positions and reliability of the measurements of the evaluators. Agitation of the animals in each situation was also considered, in order to correlate it to different evaluators and anatomical cuff positions. For this purpose, 36 dogs, males and females, up to 20 kg were used. Each patient had their SBP measured by Doppler method by three evaluators, both on the left forelimb as the tail, in randomly selected order. Each evaluator classified the animal as calm or agitated. The results showed no difference in SBP measured between forelimb (136.8 ± 26.37, P = 0.3549) and tail (133.2 ± 25.64, P = 0.3062), with concordance between the anatomical positions. Measures obtained between evaluators with different degrees of experience did not differ (Forelimb: evaluator A: 136.9 ± 31.52, P = 0.3928; evaluator B: 134.8 ± 24.44, P = 0.1257; evaluator C: 138.8 ± 31.14, P = 0.1322. Tail: evaluator A: 132.9 ± 28.33, P = 0.5157; evaluator B: 127.2 ± 24.50, P = 0.4182; evaluator C: 132.4 ± 21.09, P = 0.6715). Agitation of patients did not change between the anatomical positions and evaluators. SBP values and agitation decreased as the dogs get used to the measurements in forelimb. It was concluded that the measurement of SBP by Doppler method can be performed reliably in both forelimb and tail, and adequate instruction is enough to examiners measured precisely SBP.

Effects of Anti‐Inflammatory Glucocorticoids on Hemodynamics and Cardiac Function in Clinically Healthy Dogs

A. Masters1, D.J. Berger1, W.A. Ware1, J.F. Coetzee2, N. Langenfeld3, J.L. Ward1

1Iowa State University, Ames, IA, USA, 2Kansas State University College of Veterinary Medicine, Manhattan, KS, USA, 3University of Iowa, Centralia, IA, USA

The use of glucocorticoids in veterinary patients with heart disease is presently limited due to concern for possible precipitation of congestive heart failure. The objective of this study was to investigate previously proposed mechanisms by which glucocorticoids could cause progression of heart disease in dogs, including fluid retention from mineralocorticoid effect, transient hyperglycemia causing intravascular fluid shift, increased afterload due to systemic hypertension, and direct cardiac remodeling.

In this prospective clinical trial, clinically healthy dogs with allergic dermatitis (n = 11) were given anti‐inflammatory doses of oral prednisone (1 mg/kg/day) for 14 days, followed by a taper and washout period. Hemodynamic, biochemical, and echocardiographic variables were measured prior to prednisone administration (d0) and at days 7, 14, and post‐washout (d35). A matched control population was evaluated concurrently. We hypothesized that prednisone administration would cause no changes in measured variables at any time point. Linear mixed models were used to compare changes in variables from baseline to d7, d14, and d35 between treatment groups.

Results demonstrated no significant changes in sodium/potassium ratio, blood glucose, or echocardiographic variables. A significant increase in blood pressure from baseline occurred in the treatment group when compared to the control group at d7 (P = 0.014). Expected changes in hematologic and biochemical variables due to prednisone administration (neutrophilia, eosinopenia, isosthenuria, and elevated ALP and ALT) occurred in the treatment group. Overall, this study demonstrated that anti‐inflammatory doses of glucocorticoids in clinically healthy dogs have the potential to impact cardiac function by causing elevated blood pressure and thus increased afterload.

Effect of Ultrafiltration during Cardiopulmonary Bypass in Dogs Undergoing Mitral Valve Plasty

K. Matsuura1, S. Yamada2, M. Enokizono3, H. Suzuki4, D. Nagakubo5, N. Kanno6

1Animal Cardiovascular and Thoracic Surgery Center, Sayama‐shi, Saitama, Japan, 2Animal Cardiovascular and Thoracic Surgery, Fuchuu, Tokyo, Japan, 3Japan Small Animal Medical Center, Tokorozawa, Saitama, Japan, 4Animal cardiovascular and Thoracic Surgery Center, Sayama, Saitama, Japan, 5The University of Tokyo, Sayama, Saitama, Japan, 6Nihon University, Fujisawa, Kanagawa, Japan

Cardiopulmonary bypass(CPB) is necessary to perform mitral valve plasty(MVP) in dogs. Ultrafiltration (UF) has major advantages such as reducing the amount of blood transfusion and removal of various cytokines in human medicine. We performed MVP for dogs with spontaneous mitral valve regurgitation(MR) under CPB and investigated the effect of ultrafiltration. Twenty‐two dogs diagnosed with severe MR that underwent MVP under CPB were included. Depending on whether or not ultrafiltration was used, it was divided into a non‐UF group and a UF group. We performed CUF (Conventional Ultrafiltration) and MUF (Modified Ultrafiltration). The lowest hematocrit (Ht) value and Ht after the end of CPB in both groups were compared. The average of the lowest Ht during the operation was 23.0% in the non‐UF (n = 11) group and 24.0% in the UF group (n = 11), and no significant difference was observed between the two groups. The average Ht after the end of the CPB was 29.0% in the non‐UF group and 35.5% in the UF group. Thus the average Ht was significantly higher in the UF group (P = 0.01). We also analyzed cases with low body weight (5 kg or less) separately. There were 6 such cases in the non‐UF group and 9 low body weight cases in the UF group. The lowest Ht and the Ht after the end of the CPB was significantly higher in the UF group. Notably, the average Ht after the CPB was 10% higher in UF group (36.0%) than in non‐UF group (26.0%).The amount of cardioplegia sufficient to produce cardiac arrest varies depending on the specific case. This often results in excessive hemodilution. Since it helps to drain excess fluids under such circumstances, CUF is thought to contribute to the difference in Ht in this study. In addition, during MUF, residual blood in the extracorporeal circuit can be returned to the patient's body. As a result, the Ht after the end of the CPB were higher in the UF group than in the non‐UF group. We conclude that UF during CPB increases the lowest Ht value during CPB and increases Ht after end of the CPB in dogs undergoing MVP. These difference are particularly noticeable in small dogs.

Right Ventricular Dysfunction and Dyssynchrony Assessed by Echocardiography in Dogs with Pre‐Capillary Pulmonary Hypertension

T. Morita1, K. Nakamura2, T. Osuga1, K. Morishita2, N. Sasaki2, H. Ohta2, M. Takiguchi2

1Hokkaido University, Sapporo, Hokkaido, Japan, 2Department of Veterinary Clinical Sciences, Hokkaido University, Sapporo, Hokkaido, Japan

Right ventricular (RV) dysfunction and dyssynchrony have been associated with decreased cardiac function and clinical worsening in human patients with pre‐capillary pulmonary hypertension (PH). However, few reports are available on RV function and dyssynchrony in dogs with pre‐capillary PH. The objective of this retrospective study was to investigate RV morphological change, function and dyssynchrony assessed by echocardiography in dogs with pre‐capillary PH.

Medical records were retrospectively reviewed in 75 client‐owned dogs; 21 dogs with pre‐capillary PH, and 54 healthy dogs. RV morphological variables, including RV internal diameter in end‐diastole (RVIDd), RV end‐diastolic area (RVEDA) and RV wall thickness in end‐diastole (RVWTd), and echocardiographic indices of RV function, including RV strain by speckle tracking echocardiography, and dyssynchrony index, the standard deviation of the time to peak longitudinal strain of RV (RV‐SD) by speckle tracking echocardiography, were evaluated. In addition, the relationship between echocardiographic variables, QRS duration and RV‐SD was validated in dogs with pre‐capillary PH.

The dogs with pre‐capillary PH had larger RVIDd, RVEDA, RVWTd, lower RV strain, and higher RV‐SD compared with healthy dogs. There was a positive correlation between RV‐SD and RVIDd, RVEDA and TR velocity (RVIDd, r = 0.55, P = .012; RVEDA, r = 0.81, P < .001; TR velocity, r = 0.73, P < .001), but RVWTd and QRS duration were not related with RV‐SD.

In conclusion, the dogs with pre‐capillary PH had RV dysfunction and dyssynchrony, and RV dyssynchrony may be caused by RV dilation and elevated pulmonary arterial pressure in dogs with pre‐capillary PH.

Short‐term Results of Quality‐of‐Life and Right‐Ventricle Function after Palatoplasty and Rhinoplasty in Brachycephalic Dogs

J. Munoz‐Pérez1, J.E. Rojas‐Aguirre1, S.E. Linares‐Villalba1, S. Cardona‐Ramirez2

1Departamento de Salud animal, Universidad de Caldas, Manizales, Colombia, Manizales, Caldas, Colombia, 2Department of Pathobiology, University of Missouri, Columbia, Missouri, Manizales, MO, USA

Brachiocephalic syndrome (BS) is the result of a combination of anatomical abnormalities in predisposed breeds including stenotic nares and elongated soft palate among others. The signs of BS include laryngeal stridor, rales, exercise intolerance, and higher risk of heat stroke, cyanosis, vomiting, pulmonary hypertension, and right ventricle congestive heart failure. The English bulldog has been reported as a natural model for the study of sleep obstructive apnea syndrome in people, because this breed show alterations like snoring, fragile sleep and hypoxemia. In addition, upper respiratory tract obstruction and inflammation are known to have direct effects on the cardiovascular system and health‐related quality of life (HRQoL). The aim of this study was to compare quality of life and right ventricular global function before and after corrective surgery for upper respiratory tract obstruction.

For this longitudinal prospective study, all patients with clinical evidence of Upper Airway Obstruction candidates for corrective Palatoplasty (PP) and Rhinoplasty (RP) were recruited during December 2015 and December 2016. To evaluate the HRQoL, the Spanish version of the “FETCH™ Questionnaire” was used (FETCHSV2‐Q™). The global right ventricular function was evaluated with parameters such as the Tricuspid Annular Plane Systolic Excursion “TAPSE”. Because there is a curvilinear relationship between TAPSE and weight, the following correlation equation was used. TAPSE = 0.17 + 0.52 Weight1/3. The maximal velocity of the S’ wave and E′ early diastolic wave of the tricuspid annulus were evaluated by tissue Doppler. In addition, systolic time intervals and the Index of myocardial performance (RVIMP) were evaluated from the spectral signal of the transtricuspid and pulmonary blood flow. All patients were evaluated between 10 and 2 days before surgery and 1–6 months afterward. Student t‐test and the Wilcoxon rank test were used to compare parametric and non‐parametric variables, respectively. The results were considered statistically significant with a P‐value < .05.

Twenty dogs of different brachiocephalic breeds with signs of upper airway obstruction were recruited. No difference was found before and after surgery in the total score obtained with the FETCHSV2‐Q™ instrument: 30.1 ± 11.2 vs. 29.3 ± 6.7 (P > 0.05). In this sample, TAPSE values were reduced compared to the established reference ranges in dogs (population of this research: 0.9 ± 0.3 cm vs reference range: 1.5–2.0 cm). In addition, there was difference between TAPSE obtained 10 and 2 days before and 2–6 months after corrective surgery: 0.9 ± 0.3 vs. 1.4 ± 0.9 (P < 0.05). The maximum velocity of the S’ wave of the tricuspid annulus before surgery was 8.4 ± 2.3 cm/s and after surgery was 9.25 ± 0.9 cm/s, being a statistically significant difference (P‐value < 0.05). The values of the E′ early wave velocity was similar and before the procedure was 7.75 ± 1.09 cm/s vs. 8.50 ± 0.89 cm/s afterward (P < 0.05). RVIMP was 0.19 ± 0.7 before surgery and similar after intervention 0.25 ± 0.2 (P‐value > 0.05). The Pre‐ejection time (PEP) was 81 ± 7.9 ms before surgery and markedly diminished post‐recovery 64 ± 11 ms (P‐value < 0.05).

In conclusion, corrective surgery for upper airway obstruction may release the right ventricle from an increased afterload and stabilize systolic function values. However, apparently does not improve the quality of life of patients undergoing to these interventions.

Prognostic Factors in Cats with Restrictive Cardiomyopathy: A Retrospective Study of 92 Cases (2001–2015)

P. Passavin1, V. Chetboul2, V. Gouni2, C. Damoiseaux1, J.‐L. Pouchelon1, E. Trehiou‐Sechi1, A. Petit1, C. Poissonnier1, M. Lavennes1, L. Desquilbet3

1National Veterinary School of Alfort/Alfort Cardiology Unit, Maisons‐Alfort, Ile‐de‐France, France, 2Alfort National Veterinary School/Alfort Cardiology Unit, Maisons Alfort, Ile‐de‐France, France, 3National Veterinary School of Alfort/ Department of Biostatistics and Clinical Epidemiology, Maisons‐Alfort, Ile‐de‐France, France

Restrictive cardiomyopathy (RCM) is the second most common feline primary cardiomyopathy. However, few data are available regarding prognostic variables in large feline RCM populations. The aim of this retrospective study was therefore to analyze the clinical course of cats diagnosed with RCM between 2001 and 2015 at the Alfort Cardiology Unit, and to identify factors associated with cardiac death (CD). RCM diagnosis was based on the following echocardiographic criteria: left atrial (LA) or biatrial enlargement associated with normal or mildly altered left ventricle (LV). Cats with mildly altered LV should also have a restrictive LV filling pattern. Median survival time to CD and adjusted hazard ratios (HR) were estimated by Kaplan‐Meier method and multivariable Cox models, respectively. The study population consisted of 92 cats with RCM. Most RCM cats (64/92, 70%) were symptomatic at the time of diagnosis, with dyspnea related to congestive heart failure in 57/64 cats (89%). Fifty‐eight of the 72 RCM cats for which follow‐up was available died (81%). CD represented 64% of all death causes, with a median survival time from diagnosis to CD of 667 days. Independently of age, biatrial enlargement and arrhythmia, the risk of CD was increased by both the LA on aorta (Ao) ratio and the presence of severe LA enlargement (LA:Ao ratio ≥2): HR=4.2 for a 1‐unit increase, 95% confidence interval (CI)=[1.7–10.7] (P = 0.002) and HR=2.5, 95% CI=[1.1–5.6] (P = 0.003), respectively. In conclusion, CD is common in RCM cats, and LA enlargement is significantly associated with decreased survival time.

Interindividual Variability in Clopidogrel Active Metabolite Concentrations in Feline Plasma

M.C.L. Reinelt1, P.M. Lee1, M.H. Court2

1Washington State University, Pullman, WA, USA, 2Washington State University, Program in Individualized Medicine (PrIMe), Pullman, WA, USA

Arterial thromboemboli (ATE) are a devastating consequence of heart disease in cats. Although the antiplatelet medication clopidogrel has shown superior efficacy to aspirin for prevention of cardiogenic ATE in cats, treatment failures with clopidogrel still occur. In humans, treatment failures are often attributed to variation in metabolism of clopidogrel to the clopidogrel active metabolite (CAM). The primary objectives of this project were to quantify variation in plasma CAM concentrations between cats and determine if these differences are correlated with variability in clopidogrel metabolism and/or body weight.

Thirty‐five cats were screened, and 19 healthy cats were enrolled.

A single oral dose of 18.75 mg clopidogrel was administered to each cat. Blood was collected 2 hours post‐administration. Plasma concentrations of clopidogrel, CAM, and clopidogrel acid (major inactive metabolite) were measured using high performance liquid chromatography with tandem mass spectrometry.

A significant correlation was appreciated between the percentage of drug circulating as CAM (CAM%; an indicator of the extent of clopidogrel metabolism to CAM) and CAM concentrations (simple linear regression: R2=0.47, P = 0.001; multiple linear regression including weight as covariate: R2=0.55, P = 0.001). No significant correlation was seen between weight and CAM concentrations (simple linear regression: R2=0.12, P = 0.147; multiple linear regression including CAM% as covariate: R2=0.55, P = 0.121). One cat with high CAM% and high CAM concentration and four cats with low CAM% and low CAM concentrations were identified.

Substantial variation in circulating CAM concentrations exists between cats, and 47% of this variation could be attributed to individual differences in clopidogrel metabolism to CAM.

Biological Variation Of Galectin‐3 In Clinically Healthy Dogs

J. Ribeiro de Castro1, B.B. Real1, A.M. Gimenes1, M.M. Mantovani1, C.N. Duarte1, L.C. Petrus1, G.T. Goldfeder2, M.H.M.A. Larsson2, D.S. Schwartz1

1School of Veterinary Medicine and Animal Science – University of São Paulo USP, São Paulo, Sao Paulo, Brazil, 2University of Sao Paulo, Sao Paulo, Sao Paulo, Brazil

Galectin‐3 (Gal‐3) is novel serum biomarker associated with fibrosis and inflammation of heart failure in human patients. However, there are no studies determining reference values or evaluating the influence of gender and age on Gal‐3 in healthy dogs. This study aimed to establish reference intervals for Gal‐3 in a healthy dog population and to evaluate the influence of gender and age on this biomarker concentration. A prospective cross sectional clinical study was conducted, and included a sample of 60 healthy small‐to‐medium size dogs, males and females, adults of different ages. Considering the gene homology of approximately 85% (http://www.ncbi.nlm.nih.gov/homologene) between canine and human Gal‐3, human ELISA kit (KitRD#DGAL30, Quantikine ELISA, Fenix, USA) and canine (dog Galectina‐3, EIAab®, Wuhan EIAAB Science, China) were tested. Gal‐3 concentration recovery for healthy dogs obtained with canine kit was significantly lower, with low repeatability and reproducibility, compared to the human kit, suggesting lower sensitivity of the canine Gal‐3 kit used. Reference values obtained for human Gal‐3 were 7.548 ng/mL (P25%–75%= 8.933–10.960). There was no significant difference in Gal‐3 serum levels between males and females (P = 0.2297) or between age groups ≤6 years and >6 years in the study population (P = 0.6936). The evaluation of the biological variation of this biomarker showed that concentrations of Gal‐3 were unaffected by age and gender. These data suggest a reference range for the biomarker human Gal‐3 in healthy dogs.

Galectin‐3 as Biomarker in Heart Failure Secondary to Chronic Mitral Valve Degeneration in Dogs

J. Ribeiro de Castro1, A.M. Gimenes1, M.M. Mantovani1, C.N. Duarte1, M.Y. Ueda1, L.C. Petrus1, G.T. Goldfeder2, M.H.M.A. Larsson2, D.S. Schwartz1

1School of Veterinary Medicine and Animal Science – University of São Paulo (USP), São Paulo, Sao Paulo, Brazil, 2University of Sao Paulo, São Paulo, Sao Paulo, Brazil

Chronic mitral valve degeneration (CMVD) is a highly prevalent heart disease seen mainly in older small breed dogs. In order to follow the progression of heart failure (HF), galectin‐3 (Gal‐3) has been applied as a biomarker to identify pre‐clinical cardiac diseases, progression and decompensation in human patients. This study aimed to establish the utility of this new biomarker, isolated or in association with Type B natriuretic pro‐peptide (NT‐proBNP) and cardiac troponin I to estimate short term prognosis in dogs with HF caused by CMVD. It was designed as a prospective cross‐sectional clinical study with a longitudinal arm. One hundred thirty nine dogs were distributed among five groups with rigorous selection criteria, according to ACVIM CMVD staging (Control group: stage A‐ 60 healthy small breed dogs, predisposed to CMVD; 28 dogs in stage B1, 20 dogs in stage B2, 20 dogs in stage C and 11 dogs in stage D), recruited from a Veterinary Teaching Hospital. Groups B1, B2, C and D had a second blood sampling at day 60. Measurements were obtained for human Gal‐3, NT‐proBNP and cTnI. Reference values obtained for group A for human Gal‐3 were 7.548 ng/mL (P25%–75% = 8.933–10.960). We concluded that the magnitude and variation observed in Gal‐3 did not allow for detection of differences between stages of CMVD nor were capable of identifying patients in HF, compared to the other measured biomarkers, NT‐proBNP and cTnI, already established for canine HF evaluation.

Assessment of Longitudinal Systolic Function Using Tissue Motion Annular Displacement in Healthy Dogs

M. Wolf1, B.C. Brüler2, G.L.R. Tuleski1, S.B. Lucina1, A.P.S. Lopes3, F.S. Jojima2, M. Gonçalves de Sousa4

1Federal university of Paraná – UFPR, Curitiba, Parana, Brazil, 2Federal University of Paraná – UFPR/Department of Veterinary Science, Curitiba, Parana, Brazil, 3Pontifical Catholic University of Parana, Curitiba, Parana, Brazil, 4University of Estadual Paulista UNESP/Jaboticabal, Jaboticabal, Sao Paulo, Brazil

Although not assessed in standard echocardiography, longitudinal myocardial fibers play an active role in ventricular contraction. Longitudinal Strain (LSt) and Tissue Motion Annular Displacement (TMAD) are techniques that evaluate these fibers. TMAD, which has not been investigated in dogs, is calculated as the dislodgement of a virtual point automatically determined between mitral's septal and lateral annulus toward the left ventricular apex. One of the reported advantages of TMAD over LSt is the possibility of having it calculated using images not suitable for strain analysis, as well as the reduced processing time. In this prospective cross‐sectional observational study twenty‐seven healthy dogs (1–14 years/2–32 kg) underwent an echocardiogram. Apical 4‐chamber (AP4) and 2‐chamber (AP2) images were obtained, which allowed the calculation of TMAD and LSt obtained from both images. Data underwent the Shapiro‐Wilk test to check for a normal distribution. Results are shown in table 1. Similar TMAD and LSt was calculated from AP4 and AP2 images. No differences existed between males and females for TMAD‐AP4, TMAD‐AP2 and LSt. A positive correlation existed between TMAD‐AP2 and LSt‐AP2, Global LSt (combined AP4 + AP2 LSt) and body weight. Interestingly, TMAD correlated significantly with echocardiographic surrogates of volume and LV filling pressure, but LSt did not. In addition to being technically easier to perform, TMAD required less image enhancement and operator experience as compared to LSt. Further studies are warranted to clarify whether TMAD might be a reliable substitute for global LSt in the assessment of longitudinal systolic function in dogs.

Coupling Interval and Prematurity Index in Dogs with Mitral Valve Disease and Ventricular Arrhythmias

E.R. Carvalho1, R.A.N. Ampuero1, A.A. Camacho1, M. Gonçalves de Sousa2

1UNESP Jaboticabal, Jaboticabal, Sao Paulo, Brazil, 2University of Estadual Paulista UNESP/Jaboticabal, Jaboticabal, Sao Paulo, Brazil

Ventricular arrhythmias (VA) were already demonstrated to be a concern in dogs with myxomatous mitral valve disease (MMVD). The coupling interval (CI) and the prematurity index (PI) (Figure 1) were shown to accurately differentiate benign and malignant VA in human beings, in which VA are known to be associated with an increased risk to either evolve into signs of heart failure or die suddenly. In this study, we investigated how the CI and PI perform in dogs with MMVD.

In a retrospective, cross‐sectional investigation that included dogs with either symptomatic (stages C/D; n = 41) or asymptomatic (stages B1/B2; n = 29) MMVD, we reviewed the electrocardiographic tracings to calculate both the CI and PI. In eight dogs we also compared these indices obtained from both a Holter recording and a standard ECG tracing and no statistical differences existed (CI, P = 0.97; PI, P = 0.17). Even though CI and PI were determined in all animals enrolled in the study, VPCs’ characteristics were only compared between symptomatic and asymptomatic dogs when a Holter recording was available (n = 49).

The PI was different (P = 0.01) between symptomatic (0.65 ± 0.17) and asymptomatic (0.56 ± 0.18) dogs, but CI was considered similar (P = 0.91). Also, the symptomatic dogs had more polymorphic VPCs (P = 0.002) and supraventricular arrhythmias (P = 0.0002) than the asymptomatic animals.

In conclusion, VPCs in dogs with symptomatic MMVD are more premature and more commonly associated with supraventricular arrhythmias and polymorphic ventricular premature complexes than asymptomatic animals.

Computed Tomographic Angiography and 3D Printing of Aberrant Coronary Arteries in Dogs with Pulmonic Stenosis

S.M. Stieger‐Vanegas1, R. Cunningham1, K.F. Scollan2

1College of Veterinary Medicine, Oregon State University, Corvallis, OR, USA, 2Oregon State University, Corvallis, OR, USA

Pulmonic stenosis is one of the most common congenital cardiac defects in dogs and can occur concurrently with type R2A aberrant coronary artery anatomy, most commonly reported in English Bulldogs and Boxers. Computed tomography angiography (CTA) allows to non‐invasively assess coronary artery anatomy. Three dimensional (3D) models improve the understanding and visualization of complex cardiac anatomy.

Our hypothesis is that CTA a can be used to evaluate the presence and anatomy of anomalous coronary arteries in dogs with pulmonic stenosis and that 3D models of anomalous and normal coronary arteries can be created using specialized software and printed.

Retrospective study. CTA images of the coronary arteries from 6 dogs with confirmed R2A coronary artery anomalies and 5 dogs with normal coronary artery anatomy were segmented using two different 3D modeling software programs and printed using a 3D printer.

Normal and aberrant coronary arteries could be visualized using CTA and segmented using both 3D software programs. All R2A cases had a single coronary ostium in the right aortic sinus that gave rise to a normal right main and circumpulmonary left main coronary artery with all other visualized branches following normal anatomic courses. All normal cases had normal coronary artery anatomy. 3D printed models were of excellent quality and allowed direct visualization of normal and abnormal coronary anatomy.

CTA is effective for identification of R2A coronary artery anomaly. Non‐gated CTA is not suitable for creating printable 3D models. 3D models increase understanding of the complex anatomy of normal and anomalous coronary arteries.

Radial Strain and Strain Rate Assessed by Speckle‐Tracking Echocardiography in Turkish Kangal Dogs

O.S. Terzi, H. Albasan

Department of Internal Medicine, Faculty of Veterinary Medicine, University of Ankara, Ankara, Turkey, Diskapi, Ankara, Turkey

Cardiac diseases especially dilated cardiomyopathies are common in dogs. Quantitative analysis of left ventricular deformation by two‐dimensional speckle tracking echocardiography provides better diagnosis. The purpose of this study was to specify radial strain and strain rate values of left ventricle by two‐dimensional speckle‐tracking echocardiography in Turkish Kangal dogs.

Study groups of A and B consisted of 3 dogs with dilated cardiomyopathy and 4 dogs with valve disease, respectively. Control group (C) consisted of 7 healthy dogs. The mean ages of 3.8 ± 2.9, 5.5 ± 2.4, and 2.1 ± 1.7 years in groups of A, B and C, respectively were not significantly (P > 0.05) different.

Right parasternal short‐axis view of the left ventricle at the level of the papillary muscles was used to determine radial strain and strain rates by two‐dimensional speckle‐tracking echocardiography. Results of this study revealed that group A was significantly (P < 0.05) different from group C and group B. However, there was no significant difference between group C and group B.

Two‐dimensional speckle tracking echocardiography revealed the decreased left ventricular myocardial systolic performance in dogs with dilated cardiomyopathy while similar strain and strain rates between healthy dogs and dogs with valve disease. Therefore, this procedure can be applied to detect dilated cardiomyopathy in dogs with cardiac disease.

CT‐Based Anatomical Features of Large Airway and Heart Volume in Dogs with Different Body Size

T. Uehara, Y. Fujii, K. Sugimoto, H. Sunahara, T. Aoki

Azabu University, Sagamihara, Kanagawa, Japan

Chronic valvular heart disease is commonly seen in small‐breed dogs and a cough is a common clinical sign. We hypothesized that anatomical features of the heart and large airway were different, influenced by body size, and would relate to coughing when the heart enlarged. The purpose of this study was to compare relative heart volume and anatomical features of large airway in different body sizes using CT scan. Retrospective analyses performed on medical records of dogs without heart and lung disease that underwent CT scanning. Dogs (n = 209) were divided into 3 groups on the basis of body weight; small (<7 kg), medium (7 kg–20 kg), and large (≥20 kg). The following parameters were calculated using OsiriX: thoracic volume, heart volume, relative heart volume to thoracic volume, heart length, distance from main stem bronchi to vertebra, tracheal diameter, and angle of bronchus. Relative heart volume was significantly large in small dogs (0.29 ± 0.070) and medium dogs (0.28 ± 0.061), compared with large dogs (0.23 ± 0.075). The relative distance from main stem bronchi to vertebra was significantly short in small dogs (0.10 ± 0.037), compared with medium (0.15 ± 0.043) and large dogs (0.19 ± 0.050). There was no significant difference in the angle of bronchus among the groups. Findings revealed that small dogs had a large heart relative to the thorax, and the distance from the heart to vertebra was relatively short, compared with larger dogs. This anatomical feature may be potentially predisposed to occurrence of cough.

Long‐term Outcome of Irish Wolfhound Dogs With Preclinical Cardiomyopathy Treated With Pimobendan, Methyldigoxin, Or Benazepril

A. Vollmar1, P. Fox2

1Private practice, Bonn and Wissen, Germany, 2Animal Medical Center, New York, NY, USA

Dilated cardiomyopathy (DCM) is a common cause of morbidity and mortality in Irish Wolfhounds (IW), but clinical data reporting treatment efficacy are limited. We aimed to evaluate long‐term pimobendan, benazepril, and methyldigoxin therapy in 121 IW with preclinical DCM.

This retrospective, open label, cohort study evaluated IW receiving pimobendan (n = 25; 0.24–0.26 mg/kg PO q12 h); benazepril (n = 24; 0.25–0.5 mg/kg PO q 12 h); methyldigoxin (n = 21; 0.005 mg/kg PO q 12 h); methyldigoxin+benazepril (n = 40); or pimobendan+benazepril (n = 11), for up to 7.8 years. Time to events were estimated by Kaplan Meier method for first onset of CHF, cardiac death, and all‐cause mortality using pair‐wise comparisons between groups. Pooled data from 86 IW with echocardiographic follow‐up were dichotomized based upon improvement in %FS, LV end‐systolic, and LV end‐diastolic dimensions into references ranges (‘responders’), vs. no echo improvement (‘nonresponders’), to compare cardiac survival.

Time to CHF was not significantly different between groups (all P > 0.2). Time to cardiac death was longer with benazepril‐vs‐methyldigoxin (P = 0.034), pimobendan‐vs‐methyldigoxin (P = 0.007), benazepril‐vs‐pimobendan+benazepril (P = 0.005), pimobendan‐vs‐pimobendan+benazepril (P = 0.004), and methyldigoxin+benazepril‐vs‐pimobendan+benazepril (P = 0.027), respectively. Survival from all‐cause mortality was significantly longer for pimobendan‐vs‐methyldigoxin (0.029) and benazepril‐vs‐pimobendan+benazepril, respectively (P = 0.045). Time to cardiac death was longest for ‘responders’ (median‐4.8 years) vs (median‐2.9 years) (P = 002); time to all‐cause death was not different (P = 0.067).

Larger, prospective, blinded studies are needed to clarify clinical benefits of these drugs.

Evaluation of Lung Ultrasound for the Diagnosis of Congestive Heart Failure in Dogs with Cough

J.L. Ward1, G.R. Lisciandro2, W.A. Ware1, T.C. DeFrancesco3

1Iowa State University, Ames, IA, USA, 2Hill Country Veterinary Specialists, Spicewood, TX, USA, 3North Carolina State University, Raleigh, NC, USA

Lung ultrasound (LUS) is an emerging imaging technique that can suggest presence of cardiogenic pulmonary edema (CHF) by identifying certain artifacts (B‐lines) caused by interstitial or alveolar fluid. This study aimed to determine the accuracy of LUS for diagnosing CHF in dogs with cough.

Sixty coughing dogs were enrolled. Patients underwent protocolized LUS quantifying the presence and number of B‐lines at 4 sites on each hemithorax. Medical records, including thoracic imaging, were evaluated for final diagnosis.

Seven dogs were diagnosed with CHF and 53 with noncardiac (NC) causes of cough (most commonly dynamic airway disease). Dogs with CHF had a greater number of B‐lines compared to NC dogs (P < 0.0001) and more sites containing B‐lines (P < 0.0001). Using the cutoff criterion that at least 4 out of 8 sites contain B‐lines, sensitivity and specificity of LUS for diagnosing CHF were 100% and 92%, respectively. Diagnostic accuracy was higher among the subset of dogs with heart murmurs (n = 33; sensitivity 100%, specificity 96%) and dogs with moderate to severe left atrial enlargement (n = 19; sensitivity and specificity 100%). Notably, 23% (12/53) of dogs with NC cough had been treated previously with diuretics due to prior CHF misdiagnosis.

LUS was useful in predicting CHF as the cause of cough in this study population, particularly among dogs with concurrent heart disease. LUS is a potentially valuable diagnostic tool to identify (and avoid misdiagnosis of) CHF in coughing dogs.

Common Cytogenetic Alterations Define Choroid Plexus Tumors in Dogs

D. Ancona1, M. Aguilar2, D.L. Bannasch3, P.J. Dickinson3

1UC Davis Veterinary Medical Teaching Hospital, Davis, CA, USA, 2UC Davis Population Health and Reproduction: Veterinary Medicine, Davis, CA, USA, 3University of California Davis, Davis, CA, USA

Defining commonly occurring cytogenetic abnormalities in canine choroid plexus tumors (CPTs) is essential to an understanding of oncogenesis, and the definition of appropriate therapeutic strategies. We used Illumina 173K HD SNP arrays to determine copy number alterations (CNAs) and allelic imbalances in 12 histologically confirmed spontaneous canine choroid plexus tumor samples (3 papillomas/CPP, 8 carcinomas/CPC, 1 recurrent CPP) and tumor matched normal tissue. Copy number calls and allelic events were determined by matched paired analysis using BioDiscovery Nexus Copy Number™ software.

Hierarchical clustering by aberration profile did not distinguish between CPPs and CPCs. Choroid plexus tumors were characterized by highly recurrent single copy, whole chromosomal losses with specific losses involving chromosomes 2, 5, 8 and 20, seen in 90–100% of tumors. The pattern of predominantly whole chromosomal losses was most similar to findings in human CPCs compared to CPPs, and several recurrent chromosomal losses in dog tumors were syntenic to commonly reported regions within whole chromosomal losses in human CPTs. Highly recurrent losses in dog CPTs encompassed key tumor suppressor genes and genes associated with chromosomal instability including TP53 and VHL.

The strikingly high prevalence of specific CNAs in canine tumors is supportive of their biological relevance in CPT oncogenesis and progression. If the characteristic CNAs of whole chromosomal losses reflects a common oncogenic mechanism in human and canine CPTs, comparative cytogenetics and more detailed analysis of candidate genes within orthologous regions may provide insight into both novel and common pathways in choroid plexus tumor biology.

Pullout Properties of Monocortical and Bicortical Pins and Screws in Canine Lumbar Vertebral Bodies

D. Ancona1, B. Sturges2, A. Kapatkin2, T. Garcia‐Nolen3, S. Stover3

1UC Davis Veterinary Medical Teaching Hospital, Davis, CA, USA, 2Surgical and Radiological Sciences: Veterinary Medicine: Veterinary Medicine, Davis, CA, USA, 3Anatomy, Physiology and Cell Biology: Veterinary Medicine, Davis, CA, USA

Implants used in surgical vertebral stabilization are typically placed bicortically due to a theoretical increase in pullout strength and construct stiffness. Bicortical implants, however, have an increased risk of impingement of regional neurovascular structures. Recent studies involving cervical vertebrae suggested that monocortical pin/screw plate constructs tested in bending are biomechanically equivalent to bicortical constructs. There is no data directly comparing the pullout strength of each implant type/method in canine lumbar vertebrae.

A full factorial design was used to assess pullout strength of 3.2 mm × 4.0 mm positive profile pins, 3.5 mmg cortical screws, and 3.5 mm locking screws, inserted monocortically or bicortically in predrilled holes in 6 lumbar vertebrae (L1‐L6) from 6 canine cadavers of similar body weight and size. Implants were distracted in a mechanical testing system in a direction along their long axis at a constant rate of 0.5 mm/s until screw/pin pullout. Displacement and load data were acquired at 128 Hz. The effects of implant and insertion depth on implant pullout were assessed with a repeated measures ANOVA that accounted for cadaver and vertebra.

Pullout strength was greater with bicortical (4531 N +/‐ 459 N) than monocortical implants (1783 N +/‐ 459 N) regardless of implant type. There were no significant differences in pullout strength between implant type or vertebra, however, this may be due to low sample number. These data provide evidence that bicortical implant placement is a greater factor than the nature of the implant with respect to pullout strength in lumbar vertebral bodies, for those implants tested.

Pharmacokinetic Analysis of Single Dose Extended Release Levetiracetam Per OS in Healthy Cats

L.R. Barnard1, H.L. Barnes Heller1, D.M. Boothe2

1University of Wisconsin‐ Madison, Madison, WI, USA, 2Auburn University, Auburn, AL, USA

The purpose of this study was to determine the pharmacokinetic profile and side effects following a single dose (500 mg) of extended release levetiracetam (XRL) in healthy cats. Seven healthy cats, weighing ≥5 kg were included. Health was determined by normal physical and neurologic examination, PCV/ TP and serum biochemistry evaluation. XRL was administered per os, with food, at time 0. Timed blood collection occurred over 36 hours via jugular catheter. Serum levetiracetam was quantitated by high performance liquid chromatography using a technique previously validated in cats. Data is reported as mean +/‐ SD. C max = 89.7 + /−25.7 μg/ ml, adjusted C max = 1.03 + /−0.27 μg/ml, T max =291.4 + /−94.4 minutes, MRT = 534.0 + /−61.7 minutes, CLss (compensated for F) = 0.001 + /−0.0005, C12 hour = 43.7 + /−18.4 ug/ ml, C24 hour = 4.9 + /−3.4 μg/ml, Vd (adjusted for F) = 0.51 + /−0.11 L/kg, dose adjusted AUC = 676.5 + /−205.0 μg · h/ ml. All cats had levetiracetam concentrations above the minimum human interval at 12 and 18 hours; 1 cat maintained levetiracetam concentrations above the interval at 24 hours. No adverse effects were noted throughout the study period and all cats maintained normal neurologic examinations. In conclusion, side effects appear to be minimal and the mean concentration at 24 hours was below the minimum human therapeutic interval by 0.1 μg/ml suggesting that once daily dosing may be possible however individual peak and trough serum concentrations should be considered when administered once daily.

Surgical Decompression, with or without Adjunctive Therapy, for Treatment of Primary Vertebral Osteosarcoma in Dogs

S. Dixon1, A. Chen‐Allen1, B. Sturges2, K. Vernau3, J. Levine4, A.J. Otamendi4, P.J. Early5, L. Curtis6, A. Partnow7, S.A. Thomovsky8, H. Greatting1, T. Jukier1

1Washington State University, Pullman, WA, USA, 2Surgical and Radiological Sciences: Veterinary Medicine: Veterinary Medicine, Davis, CA, USA, 3UC Davis College of Veterinary Medicine, Davis, CA, USA, 4Texas A&M University College of Veterinary Medicine, College Station, TX, USA, 5North Carolina State University, Raleigh, NC, USA, 6VCA Veterinary Specialty Center of Seattle, Lynnwood, WA, USA, 7VCA Veterinary Specialty Center of Seattle, Kenmore, WA, USA, 8Purdue University College of Veterinary Medicine, West Lafayette, IN, USA

Osteosarcoma is the most common primary vertebral tumor in dogs, however previous studies examining outcome after surgical decompression of these tumors are limited. The goal of this study was to determine survival time after decompressive surgery alone and combined with adjunctive therapies such as definitive radiation or chemotherapy.

Records were collected retrospectively from six academic and private practice institutions, and were reviewed for data regarding surgical procedure, neurologic function, adjunctive therapy and survival. Sixteen cases were included, all of which received decompressive surgery. Of those cases, 1 dog was treated with chemotherapy and 3 dogs were treated with both chemotherapy and definitive radiation therapy. All dogs either improved neurologically following surgery or remained stable neurologically with improved pain control. All were large breed dogs with an average age of 9 years. Median survival time for dogs treated with surgery alone (n = 12) was 43.5 days (range 3–1333). One outlier was still alive 4 years after surgery alone. The one dog treated with chemotherapy survived 56 days. Dogs treated with radiation therapy and chemotherapy (n = 3) had a median survival time of 257 days (rage 223–270). In most cases where follow up was available, dogs were euthanized due to recurrence of the primary tumor or return of original clinical signs.

In conclusion the results of the study indicate that overall survival time with surgical decompression alone is poor. Adjunctive therapy appears to prolong survival in cases of primary vertebral osteosarcoma that have had decompressive surgery.

The Effects of Stabilizing Agents, Fetal Calf Serum and Vetstarch, on Canine Cerebrospinal Fluid Analysis

L.N. Peterson1, S.A. Thomovsky1, J.A. Christian2, G.E. Moore1, R.T. Bentley1

1Purdue University College of Veterinary Medicine, West Lafayette, IN, USA, 2Purdue Department of Comparative Pathobiology, West Lafayette, IN, USA

A common challenge of CSF analysis is time‐dependent rapid degradation of nucleated cells. This degradation leads to an inability to accurately interpret CSF, diagnose disease and institute treatment. CSF additives, fetal calf serum and the hydroxyethyl starch, Hetastarch, have been used to delay cell degradation but some alter microprotein levels. The purpose of this research is to determine if a more readily available and less expensive hydroxyethyl starch, VetStarch™, is effective at preserving CSF nucleated cell morphology without altering microprotein levels.

CSF samples were collected from 9 dogs undergoing neurologic workup to date with a target sample size of 28. Samples were divided into 4 aliquots. One aliquot was analyzed immediately (control). The remaining three aliquots received one stabilizing agent each [saline, fetal calf serum or VetStarch™] before storage at 4°C. Samples were analyzed at 48 hours; total nucleated cell count, microprotein, and cell morphology scores were recorded. A morphology score of 1 indicated no observed cellular degeneration; a score of 4 indicated severe degeneration.

Samples stored in VetStarch™ and fetal calf serum exhibited improved morphology scores, median 2.46 (1.42–3.20) and 2.25 (1.13–3.33), respectively, as compared to those in stored in saline, 3.0 (1.14–4.00). Microprotein levels for samples stored in VetStarch™ were similar to those recorded in control samples, 18.6 ± 4.50 and 18.8 ± 2.81 (mean ± standard deviation), respectively. Fetal calf serum significantly altered microprotein levels, 464.3 ± 109.75.

VetStarch™ is a commercially available, inexpensive, effective stabilizing agent for canine CSF without erroneous protein elevation.

GM2 Gangliosidosis in Shiba Inu Dogs with an In‐Frame HEXB Deletion and Autofluorescent Storage Granules

N.A. Villani1, A. Kolicheski1, G.S. Johnson1, D.P. O'Brien2, T. Mhlanga‐Mutangadura1, D.A. Wenger3, J.F. Taylor4, R.D. Schnabel5, M.L. Katz6

1Department of Veterinary Pathobiology, University of Missouri, Columbia, MO, USA, 2Department of Veterinary Medicine and Surgery, University of Missouri, Columbia, MO, USA, 3Department of Neurology, Jefferson Medical College, Philadelphia, PA, USA, 4Division of Animal Sciences, University of Missouri, Columbia, MO, USA, 5Division of Animal Sciences and Informatics Institute, University of Missouri, Columbia, MO, USA, 6Mason Eye Institute, University of Missouri, Columbia, MO, USA

This study was performed to identify the mutation causing a progressive neurodegenerative disease in Shiba Inu dogs with storage body autofluorescence. Two related Shiba Inu dogs were suspected to have neuronal ceroid lipofuscinosis (NCL) based on clinical signs, antemortem diagnostics, disease progression and autofluorescence of storage material on histopathology. A whole genome sequence (WGS) was generated with DNA from one affected dog using previously described techniques. The data were screened for potentially causal mutations in known NCL genes but no candidate variants were found. However, the WGS contained a novel homozygous deletion of a conserved codon in the HEXB gene (which encodes the beta subunit of beta‐hexosaminidase). The deletion was not found in the WGS from 105 unaffected dogs of various breeds. The other affected Shiba Inu was also homozygous for this variant. Mutations in HEXB have been implicated in GM2 gangliosidosis (Sandhoff disease) in humans and in other dog breeds. Beta‐hexosaminidase activity was measured in brain tissue from one of the affected dogs using previously described synthetic substrates. Activity against two different substrates was markedly reduced compared to a normal dog, consistent with Sandhoff‐like disease (HEXB deficiency). Thus, we conclude that the HEXB mutation in these dogs caused GM2 gangliosidosis and led to their clinical signs. To our knowledge, this is the first report that GM2 gangliosidosis‐associated storage material is autofluorescent. This study also demonstrates the utility of WGS for the efficient diagnosis of potentially heritable diseases in veterinary medicine by identifying a mutation that abolishes HEXB activity in Shiba Inu dogs.

Exosome‐Associated Integrins as Liquid Biopsy Biomarkers for Canine Glioma

C.‐F. Li1, M. Mellema2, R. Carney2, B. Sturges3, D. York2, R. Liu4, K. Lam5, P.J. Dickinson5

1UC Davis, VMTH, Davis, CA, USA, 2University of Davis, California, Davis, CA, USA, 3Surgical and Radiological Sciences: Veterinary Medicine: Veterinary Medicine, Davis, CA, USA, 4University of Davis, California, Sacramento, CA, USA, 5Department of Biochemistry & Molecular Medicine Professor of Medicine, Division of Hematology & Oncology Harold Albin Johnson Endowed Chair in Biomedical Research, UC Davis Cancer Center, Sacramento, CA, USA

Easily accessible, validated biomarkers for gliomas are needed to aid in diagnosis and therapeutic monitoring. Exosomes are extracellular membranous vesicles (30–100 nm) that are released by cells into the microenvironment and may carry signatures of the cell of origin. We hypothesized that integrin targeting cyclic nonapeptides (LXW7/αvβ3, LXY30/α3β1) derived from screening of human and canine glioma cells would identify canine glioma cell lines and cell line derived exosomes in vitro.

Three glioma cell lines (J3TBg, G06A, SDT3G) were used to validate targeting of peptides to live cells using flow cytometry and immunofluorescent labelled peptide. Canine glioma cell derived exosomes were isolated from serum free media using ExoQuick precipitation and centrifugation, and characterized using nanoparticle tracking analysis (NTA) technology (NanoSight® LM10HS) and peptide conjugated quantum dots. Non‐glioma targeting peptide (LLP2A/α4β1) was used as a negative control.

All cell lines were labelled by the LXW7 and LXY30 peptides with no LLP2A labelling. Exosomes were isolated from media collected from all cell lines and confirmed by NTA analysis and CD9 immunocytochemistry. Quantum dot conjugated labelling of exosomes was demonstrated for LXW7 and LXY30 peptides, but not for LLP2A.

Expression of cell surface αvβ3, and α3β1integrin in canine glioma cells was demonstrated using integrin specific peptide targeting, consistent with human cell lines and expression in dog tumour samples. Importantly, these integrin surface markers are also expressed in cell derived exosomes. Characterization of LXW7 and LXY30 binding exosomes in CSF and serum is therefore a rational strategy to define accessible biomarkers for gliomas in vivo.

Investigation of Immune or Infectious Etiologies in Canine Meningitis/Meningoencephalitis/ Meningoencephalomyelitis

G.E. Moore1, M. Podell2, R. Windsor2, Z. Niman2, J. Ochsenhirt2, E. Dubovi3

1Purdue University College of Veterinary Medicine, West Lafayette, IN, USA, 2MedVet Chicago, Chicago, IL, USA, 3Cornell University, Ithaca, NY, USA

Meningitis, meningoencephalitis, and meningoencephalomyelitis (M/ME/MEM) are inflammatory conditions which may be of immune or infectious origins. A perceived increase in these conditions at a specialty referral practice led to two investigative studies. The first study was a retrospective survey of MRI diagnoses made during a 3‐year period (2013–2015) to assess increased prevalence and/or temporal trends. The second study was a 7‐month prospective study of possible immune or infectious causes of M/ME/MEM diagnosed at the practice.

In the first study, monthly incidence of M/ME/MEM was calculated for the study period using computer searches for “immune”, “immuno”, “meningitis”, “meningoencephalitis”, or “meningoencephalomyelitis” within recorded MRI diagnoses. Of the total MRIs, 229 (9.3%) were annotated as possible cases. A possible seasonal trend was seen as 143 of 229 (62.4%) occurred April‐September.

In the second study, 64 dogs were considered M/ME/MEM cases based on MRI and/or CSF results. Median age was 7.8 years (0.2–16.0 years) and median weight 7.7 kg (1.8–56.3 kg); more than 30 breeds were represented. Infectious diseases (one each Bartonella, Anaplasma, Ehrlichia) were identified via routine tests in 3 dogs. Vaccination histories were available on 61/64 dogs; 21/61 (34.4%) of cases had received one or more vaccinations within 35 days of the onset of clinical signs. Multivalent Bordetella vaccines (IN) were most commonly administered in the 35‐day window, followed by Leptospira vaccines. The number of vaccines administered most recently did not differ between dogs vaccinated.

A cause of seasonal increases in M/ME/MEM was not identified. An immune origin may play a role, mediated in some dogs by undefined mechanisms related to immunological booster from multivalent or monovalent vaccinations.

Imaging Findings at 3T and Clinical Outcome in Deep‐Pain Negative Dogs with Intervertebral Disc Herniation

A.J. Otamendi1, B. Boudreau, L. Gilmour, J. Griffin, J. Levine, N. Jeffery

Texas A&M University College of Veterinary Medicine, College Station, TX, USA

Non‐invasive prognostic indicators are minimal for dogs with absent pelvic limb deep‐pain perception resulting from acute thoracolumbar intervertebral disc herniation. We examined dogs presented to one veterinary hospital between 2011 and 2015 with paraplegia and absent pelvic limb deep‐pain perception undergoing 3T MRI. Our purpose is to correlate MRI findings in these dogs with clinical outcome.

Medical records were reviewed to identify dogs meeting the above criteria. Dogs were included if their MRI study comprised T2‐weighted sagittal and transverse images covering the T3‐L3 region, and acute disc herniation within this region was confirmed at surgery or necropsy. Surviving dogs were prospectively followed through follow‐up neurological examinations 4–6 weeks post‐operatively. MRI studies were reviewed by two board‐certified radiologists blinded to outcome.

Fifty‐two dogs met our inclusion criteria, and outcome data were available for 47. Twelve developed myelomalacia, 10 failed to recover deep‐pain perception, and 25 recovered deep‐pain perception and voluntary pelvic limb movement by 4–6 weeks post‐operatively. Estimates of relative intramedullary T2‐weighted hyperintensity length in the sagittal plane were well‐correlated between observers (r = 0.76, P < 0.001). Median length of average T2‐weighted hyperintensity did not differ among outcome groups. 45% of dogs with T2‐weighted hyperintensity >6 × L2 body length had a good recovery. 42% of dogs that developed myelomalacia had T2‐weighted hyperintensity <6 × L2 body length.

We conclude that sagittal length of intramedullary T2‐weighted hyperintensity associated with acute thoracolumbar disc herniation in dogs with absent deep‐pain perception, as apparent on 3T MRI scans, does not correlate with the likelihood of functional recovery or the development of ascending‐descending myelomalacia.

Efficacy of a Therapeutic Diet on Dogs with Signs of Cognitive Dysfunction Syndrome

Y. Pan1, G. Landsberg2, I. Mougeot2, S. Kelly3, H. Xu1, S. Bhatnagar1, N.W. Milgram2

1Nestlé Purina Petcare Global Resources, St Louis, MO, USA, 2CanCog Technologies, Toronto, ON, Canada, 3Vivocore Inc, Fergus, ON, Canada

The objective of this clinical study was to evaluate the effects of dietary supplementation with medium chain triglyceride oil (MCT) and a proprietary brain nutrient blend (BPB containing DHA, EPA, and enhanced levels of B vitamins, antioxidants and arginine) on pet dogs with cognitive dysfunction syndrome (CDS). Previous studies have shown that diets supplemented with MCT or BPB enhance learning, memory and problem‐solving ability in senior dogs.

Participating veterinary clinics screened senior dogs for signs of CDS as determined by a Senior Canine Behavior and Health Questionnaire, which included six categories (disorientation, altered social interaction, sleep‐wake cycle disturbance, loss of house training, anxiety and altered activity). Out of 102 dogs screened, after ruling out potential medical causes, 87 were randomly enrolled into one of three diet groups with 29 dogs per group: Control (0% MCT), 6.5% MCT, 9% MCT. Each dog's signs were re‐evaluated at day 30 and day 90.

All 6 categories of the CDS signs were significantly (P < 0.05, Fisher least significant difference test) improved in the dogs given the 6.5% MCT oil and BPB diet at the end of the 90‐day study. By contrast, the control diet did not significantly improve disorientation and social interactions scores. The 9% MCT diet was observed to produce significant improvements in only those dogs whose owners had positive feedback on the diet, suggesting possible palatability issues.

These findings support the use of the 6.5% MCT + BPB supplemented diet for treatment of clinical signs in senior dogs with signs consistent with CDS.

Cognitive Impairment, Health Status and Overall Adaptation in Aged Dogs: An Epidemiological Study

A.S. Mani1, J. Bowen2, J. Fatjó2, C. Torre1

1Affinity Petcare, Barcelona, Catalonia, Spain, 2Ethometrix, East Sussex, UK

Previous epidemiological studies using validated scales have provided very useful information about changes in behaviour and cognitive competence in senior dogs. However, they usually present 3 limitations: 1) they don't include information on the overall health status of the dog, 2) they target a pre‐selected population of senior dogs, thus missing information from other age groups with whom comparisons could be made, and 3) behavioural measurements are focused on very specific behaviour changes linked to the ageing process, as opposed to the impact that cognitive impairment could have on overall adaptation and functioning.

A study was designed to explore the impact of the ageing process in a population of owned dogs, including information on the overall health status and the overall adaptation to the physical and social environment.

An online questionnaire was made available to dog owners through social media and a selected group of veterinary clinics. To avoid an age‐related selection‐bias, the study was presented as an online tool for all dog owners to estimate the overall level of adaptation of a dog to its environment, with each participant being given a score for their dog (which we called the “global functioning score”). Questions were divided into 4 sections: demographic data (including age and weight, a validated scale to assess cognitive dysfunction (“Canine Cognitive Dysfunction Rating scale”, Salvin et al), physical health conditions and how worried owners were about them, and how well the dog performed in a wide range of common situations (such interacting with people, family members and other dogs).

To estimate the prevalence of mild, moderate and severe cases of cognitive dysfunction, two methods were used. Firstly, scores from the cognitive dysfunction rating scale were calculated according to the authors’ scheme. Secondly a two‐step cluster analysis was performed on the cognitive function scores. Kruskal‐Wallis was used to explore differences in cognitive dysfunction scores between weight groups and to compare Global Functioning Score (GFS) between age groups.

Two thousand one hundred and fourteen valid questionnaires were obtained. More than 28% of the dogs were 8‐years‐old or older (n = 605). The overall prevalence of cognitive dysfunction in that population was 19.7%. No severe cases of cognitive dysfunction were detected below 7 years of age. However, global functioning score increased until the age of 7 and then declined, with the lowest scores being seen in dogs aged 14 and above. The reported prevalence of cognitive dysfunction by the owners was just 2.3%, being ranked last in a list of 10 health conditions reported by the owner, including arthritis, tumours, hair loss, deafness and breathing problems. When owners were asked to rate the seriousness of each of a list of 5 health conditions, cognitive dysfunction ranked third, behind cancer and heart disease but above arthritis. Older owners ranked cognitive dysfunction as more serious than younger owners.

Our findings indicate that although cognitive dysfunction is common, it is often not detected by the owner. This may be due to a combination of a lack of knowledge of the signs of the condition and a false perception that it is less serious than other some other health problems. Both will need to be tackled if we want to improve take up of treatments for this condition. Global functioning score, as an indicator of the dog's everyday performance as a pet, offer added information that could be refined to help in early detection of cognitive dysfunction and other age‐related conditions.

Salvin, H.E., McGreevy, P.D., Sachdev, P.S., Valenzuela, M.J., (2011). The canine cognitive dysfunction rating scale (CCDR): a data‐driven and ecologically relevant assessment tool. Vet. J. 188, 331–336.

Non‐Surgical Treatment of Intracranial Empyema and Abscesses in 5 Dogs

M. Tensley1, B. Young2, S. Schatzberg3

1The Animal Neurology and Imaging Center, Rio Rancho, NM, USA, 2VCA Alameda East Veterinary Hospital, Denver, CO, USA, 3The Animal Neurology and Imaging Center, Algodones, NM, USA

The gold standard treatment recommendations for both veterinary and human patients with intracranial infection includes surgical drainage and antibiotic therapy. The purpose of this retrospective case series was to describe the clinical and MRI findings, as well as the medical treatment protocols, for dogs with confirmed or presumed intracranial bacterial infections that were treated without surgical intervention.

Five dogs that were treated successfully between July 2011 and May 2015 are described. Medical treatment decisions were based upon the diagnoses of intracranial bacterial infection and/or owner reluctance for surgical intervention. MRI images for each patient were reviewed by a board‐certified veterinary neurologist and radiologist. Success was defined by post‐treatment MRI confirming resolution of the primary lesion(s) or a survival time of greater than 6 months past the discontinuation of treatment without relapse of clinical signs.

All dogs received enrofloxacin in combination with a second antimicrobial, most frequently amoxicillin‐clavulanic acid. Prednisone also was prescribed at a variable dosage and duration. Treatment in 3 animals included an initial period of intravenous medications, while the other 2 dogs were treated with oral medication only. Treatment duration ranged from 6 weeks to 4 months. Initial clinical signs varied in severity from mild neurological deficits noted on examination to severe clinical signs such as altered mental status and evidence of increased intracranial pressure. Four dogs were alive at the time of manuscript composition, 10–28 months after diagnosis. The remaining dog died from unrelated cause 36 months after diagnosis.

We concluded that medical treatment without surgery is a feasible option for patients with presumed intracranial infection. Early and aggressive therapy with broad‐spectrum antibiotics is recommended. Clinical severity did not affect outcome in this group of dogs.

Cytarabine Constant Rate Infusion as Treatment for Meningoencephalomyelitis of Unknown Etiology in 22 Dogs

R.L. Arnold1, R.L. Winter2, J. Narak3, A.R. Taylor4

1Tuskegee University/Auburn University/Southeast Veterinary Neurology, Miami, FL, USA, 2Department of Clinical Sciences, Auburn University College of Veterinary Medicine, Auburn, AL, USA, 3Veterinary Referral Surgical Practice, Roswell, GA, USA, 4Auburn University, Auburn, AL, USA

Meningoencephalomyelitis of unknown etiology (MUE) is a noninfectious inflammatory disease of the central nervous system. Cytarabine, an antimetabolite, has been used in combination with glucocorticoids to treat MUE in the dog.

Medical records of 22 dogs (age 6.27 + /‐ 2.83 years) with MUE that received cytarabine were reviewed. MRI and CSF findings were used to confirm the diagnosis of MUE. Data was reported as mean+/‐ SD or median (range) as appropriate. Student's t‐test or Wilcoxon rank sums were used for comparisons, and correlation was assessed by Spearman rank correlation and simple linear regression.

All dogs received at least 1 CRI treatment of cytarabine. Fourteen dogs also received at least 1 SQ treatment. The number of CRI treatments per dog was median of 5.5 (1 – 38). Survival of all dogs was median of 258.5 days (61 – 1312). Survival was not affected by age at diagnosis (P = 0.289), duration of clinical signs (P = 0.825), or sex (P = 0.413). There was a positive and significant correlation with number of CRI treatments and survival (r = 0.87, P = < 0.001). The number of SQ treatments did not affect survival time (r = −0.05, P = 0.819). In dogs with a relapse, there was a positive and significant correlation between number of CRI treatments and the longest period between clinical signs relapse (r = 0.82, P = 0.002).

Dogs with MUE may live longer and have longer periods of time between clinical relapse with increasing number of cytarabine CRI treatments.

Hereditary Myotonia in Pigs Associated with Chloride Voltage‐Gated Channel 1 Gene Partial Deletion

A.S. Borges1, C. Araújo2, J. Barbosa3, M.E. McCue4, C. Oliveira3, L.A. Resende5, J.P. Oliveira‐Filho2

1Sao Paulo State University – FMVZ – Unesp, Botucatu, Sao Paulo, Brazil, 2FMVZ‐Unesp‐Botucatu, Botucatu, Sao Paulo, Brazil, 3UFPA, Castanhal, Para, Brazil, 4University of Minnesota, St Paul, MN, USA, 5FMB‐Unesp, Botucatu, Sao Paulo, Brazil

Mutations in the CLCN1 gene were previously associated with hereditary myotonia in several species but not in pigs. The aim of this study is to characterize hereditary myotonia in pigs caused by a CLCN1 gene mutation.

Five pigs from the same litter, presenting clinical myotonia and two non‐myotonic pigs were submitted to clinical examination, electromyography (EMG), muscle biopsy (hematoxylin and eosin), and molecular investigation (muscle and blood samples were respectively used for RNA and DNA extractions). RT‐PCR, PCR‐sequencing reactions and alignments of obtained sequences from all clinically affected, clinically normal, and the NCBI reference pig sequence (GenBank: XM_003134570.3) were performed.

All affected animals presented muscle hypertrophy and stiffness. Action myotonia episodes, especially after startle and warm‐up phenomenon were observed. The EMG of the affected animals evidenced spontaneous myotonic discharges with a characteristic sound. Dystrophic muscle abnormalities were not observed. All affected animals had a 348 base pairs (bp) deletion (c.[1582_1930del]) corresponding to exons 15, 16, and 17 in the mRNA. Using the DNA, an absence of a 4165 bp fragment (g. [7153448_7157612del]) including the aforementioned exons, introns 15 and 16 entirely, and introns 14 and 17 partially was verified. In silico analysis suggested that this mRNA deletion results in a chloride voltage‐gated channel protein without 116 amino acids.

Pedigree analysis indicated an autosomal recessive mode of inheritance of this form of hereditary myotonia associated with a CLCN1 gene partial deletion in an animal species not yet reported with this disease.

Population Pharmacokinetics of Extended‐Release Levetiracetam in Epileptic Dogs When Administered Alone, with Phenobarbital or Zonisamide

K. Munana1, A.J. Otamendi2, J.A. Nettifee3, M.G. Papich3

1North Carolina State University, College of Veterinary Medicine, Raleigh, NC, USA, 2Texas A&M University College of Veterinary Medicine, College Station, TX, USA, 3North Carolina State University, Raleigh, NC, USA

The extended release formulation of levetiracetam (LEV‐XR) has gained acceptance as an antiepileptic medication in dogs, but there are no studies evaluating its disposition in dogs with epilepsy. The objective of this study was to evaluate the pharmacokinetics of LEV‐XR in epileptic dogs when administered alone, or concurrently with other AEDs. A population pharmacokinetic approach and non‐linear mixed effects modeling (NLME) was used to examine the variation and source of variation in our population. Eighteen client‐owned dogs on maintenance therapy with LEV‐XR only (Group L, n = 6), LEV‐XR and phenobarbital (Group LP, n = 6) or LEV‐XR and zonisamide (Group LZ, n = 6) were enrolled. All drugs were examined at steady state concentrations. Blood samples were collected at 0, 2, 4, 8 and 12 hours after the morning dose of LEV‐XR was administered with food. The LP group had significantly lower CMAX compared to the L and LZ groups (13.48 ± 9.48 vs. 33.93 ± 29.84 and 34.13 ± 9.01 μg/mL, respectively), significantly lower AUC (134.15 ± 153.67 vs. 353.10 ± 284.11 and 452.80 ± 116.89 h∙μg/mL, respectively), and significantly higher CL/F (0.17 ± 0.20 vs. 0.08 ± 0.07 and 0.07 ± 0.02 L/kg/hours, respectively), demonstrating that concurrent administration of phenobarbital significantly lowers the LEV concentration in epileptic dogs. Elimination half‐life was similar among groups (L, 1.46 ± 2.08 hours; LP, 1.54 ± 1.97 hours; LZ, 1.57 ± 1.13 hours), but shorter than previously reported in healthy dogs. These findings warrant consideration when utilizing LEV‐XR as a treatment for epilepsy in dogs.

Steroid‐Responsive Myelitis in Five Dogs: Different Characteristics in Comparison with Steroid‐Responsive Meningitis‐Arteritis

S.‐J. Park1, D.‐W. Chang1, J.‐H. Kang2, M.‐P. Yang1, B. Kang1

1College of Veterinary Medicine, Chungbuk National University, Cheongju, Ch'ungch'ong‐bukto, Republic of Korea, 2Chungbuk National University College of Veterinary Medicine/Veterinary Medical Center at Chungbuk National University, Cheongju, Ch'ungch'ong‐bukto, Republic of Korea

The most frequently occurring non‐infectious inflammatory lesion of the canine spine is steroid‐responsive meningitis‐arteritis (SRMA), while pure myelitis is a rare disease in dogs. Although meningomyelitis of unknown cause and granulomatous meningomyelitis are the most common diagnoses in canine cases, non‐infectious myelitis responsive to glucocorticoid treatment has not been well described. This report describes the clinical and diagnostic features of steroid‐responsive myelitis (SRM), which were different to those of SRMA, in five dogs.

Five dogs with SRM were identified retrospectively. Signalment, history, body temperature, neurological signs, diagnostic workup, treatment, and prognosis were analyzed. Affected breeds were small‐ to medium‐sized: Miniature Pinscher (n = 2), Maltese (n = 1), Toy Poodle (n = 1), and Cocker Spaniel (n = 1). The mean age of the dogs was 4.2 years (range 2–8 years). Paraspinal hyperesthesia, paresis, and paralysis were acutely observed according to the affected region of the spinal cord. However, all dogs were non‐febrile, and only one dog had neutrophilic leukocytosis on hematological analysis. Magnetic resonance imaging (MRI) of each dog revealed diffuse spinal lesions with T2 and FLAIR‐hyperintense and T1‐hypo‐ or isointense signal intensities at different regions: cervical (n = 1), thoracolumbar (n = 1), lumbosacral (n = 1), and multifocal (n = 2) spinal parenchyma. Contrast enhancement was not detected in the meninges. Four dogs had cerebrospinal fluid (CSF) pleocytosis: mononuclear (n = 3) and mixed (n = 1). In both serum and CSF, the levels of C‐reactive protein (CRP) and immunoglobulin A (IgA) were within respective normal limits, except in two dogs with elevated serum CRP. No dog tested positive on CSF for infectious disease. A presumptive diagnosis of non‐infectious inflammatory myelitis was established based on the clinical signs, MR scanning, CSF analysis, and laboratory tests.

Treatment was initiated with prednisolone (4 mg/kg/day orally) for immunosuppression, and then doses were tapered according to follow‐up results. The mean duration of immunosuppressive therapy was 202 days (range 131–264 days), and medication was ceased in four dogs. Improvement in spinal lesions was confirmed in all dogs by neurological evaluation, CSF analysis, and MR follow‐up scans. Because clinical signs did not recur after discontinuation of glucocorticoid, the type of myelitis present in all cases was considered to be steroid‐responsive.

In general, SRMA is characterized by fever, neck pain, neutrophilia, increased levels of IgA and CRP, neutrophilic pleocytosis, and contrast enhancement of the meninges on MR scanning. The present cases did not have these features of SRMA, while mononuclear or mixed pleocytosis, diffuse spinal lesions on MR images, and excellent responsiveness to glucocorticoid were commonly noted in all dogs.

The Pharmacokinetics of Cytarabine Administered Subcutaneously to Dogs with Meningoencephalomyelitis of Unknown Etiology

B. Pastina1, P.J. Early2, R.L. Bergman3, K. Munana4, J.A. Nettifee2, K. Bray4, A. Castel4, R. Waldron5, A. Maller4, M.G. Papich2, K. Messenger4

1North Carolina State University College of Veterinary Medicine, Raleigh, NC, USA, 2North Carolina State University, Raleigh, NC, USA, 3Carolina Veterinary Specialist, Charlotte, NC, USA, 4North Carolina State University, College of Veterinary Medicine, Raleigh, NC, USA, 5Upstate Veterinary Specialist, Greenville, SC, USA

The objective of this study was to evaluate the pharmacokinetics (PK) of cytarabine (CA) after subcutaneous (SC) administration to dogs with meningoencephalomyelitis of unknown etiology (MUE). Twelve client‐owned dogs, from 22 months to 10 years old, and weighing 2.5–39.2 kg, received a single SC dose of CA at 50 mg/m2 for treatment of MUE. A sparse sampling technique was used to collect pharmacokinetic data. Four samples were collected from each dog to cover the time period from 0 to 360 minutes after administration. All dogs were concurrently on an oral prednisone dose of 0.5–2 mg/kg/day. Plasma CA concentrations were measured by HPLC and pharmacokinetic parameters were estimated using nonlinear mixed effects modeling (NLME). Plasma drug concentrations ranged from 0.05 to 2.8 μg/mL. The population estimate (CV%) for elimination half‐life and Tmax of cytarabine in dogs was 0.7 (46.2) h and 0.8 (11.3) h, respectively. The volume of distribution per fraction absorbed was 0.7 (38.4) L/kg. Mean plasma concentration of CA for all dogs was at or above 1.0 μg/mL at the 30, 60, 90, and 120‐minute time points. In this study, the pharmacokinetics of CA in dogs with MUE after a single 50 mg/m2 SC injection in dogs was similar to what has been previously reported; there was moderate variability in the population estimates in this small clinical population of dogs.

Cavernous Sinus Syndrome in Dogs and Cats: 15 Cases (2002–2015)

H. Rylander1, E. Bentley2, A. Jones3

1University of Wisconsin Madison, Madison, WI, USA, 2School of Veterinary Medicine, UW Madison, Madison, WI, USA, 3College of Veterinary Medicine and Biomedical Sciences, Veterinary Teaching hospital, Fort Collins, CO, USA

The purpose of this study was to identify the presenting complaints, neurologic findings, diagnosis, and outcomes in dogs and cats with cavernous sinus syndrome (CSS).

Medical records between 2002 and 2015 were reviewed. Inclusion criteria were neurologic signs consistent with CSS and advanced imaging and/or post mortem examination.

Thirteen dogs and 2 cats were included. Twelve dogs received advanced imaging. Post mortem examination was performed on 2 cats and 3 dogs.

Dogs were 6–13 years (mean= 10.8 years) of different breeds. Both cats were male neutered domestic shorthair, 3 and 14 years respectively.

Presenting complaints included mydriasis (N = 4), behavior changes (N = 3), hyporexia (N = 3), ptosis (N = 2), ataxia (N = 2), pain (N = 2), weakness (N = 2), lethargy (N = 2), and one each of epiphora, ocular swelling, polydipsia, seizures, facial muscle atrophy, dysphagia, and head tilt.

Neurologic signs included ophthalmoparesis/plegia (N = 13), reduced/absent pupillary light response (N = 11), mydriasis (N = 10), reduced/absent corneal sensation (N = 7), ptosis (N = 6), reduced facial sensation (N = 2), and enophthalmos (N = 1).

Thirteen patients had a mass lesion within the cavernous sinus, 6 of which were neoplastic on histopathology.

Median survival time for 4 patients treated with radiation therapy was 1035 days (range 150–2280), for 4 patients receiving medical treatment 360 days (range 7–1260 days), and for 5 non‐treated patients 14 days (range 0–90 days).

In conclusion mydriasis and ophthalmoplegia are common signs of CSS. A mass lesion is the most common cause. Survival time may improve with radiation therapy.

Effects of Chloride in the Diet on Serum Bromide Concentrations in Dogs

G. Togawa1, M. Saito1, S. Iwata2, N. Watanabe3, K. Orito1

1School of Veterinary Medicine, Azabu University, Sagamihara, Kanagawa, Japan, 2School of Veterinary Medicine, Azabu University, Sagamihara, Kagawa, Japan, 3Watanabe Animal Hospital, Shimada, Shizuoka, Japan

Potassium bromide (KBr) is an antiepileptic drug that is widely used for seizure control in dogs. Diets with high contents of Cl‐ have been shown to decrease serum Br‐ concentrations in dogs. However, the numerical relationship between dietary Cl‐ intake and serum Br‐ concentrations currently remains unknown. Therefore, the aim of the present study was to quantify serum Br‐ concentrations in dogs fed diets with different contents of Cl‐ and reveal the underlying relationship in order to contribute to the therapeutic drug monitoring of KBr.

Steady‐state serum Br‐ concentrations were measured using a gold chloride method in 23 dogs treated with KBr. The content of Cl‐ (per 1 g) was measured using Mohr's method or calculated from the ingredients label provided by the manufacturer. A regression analysis was used to evaluate the relationship between daily Cl‐ intake (mg/kg/day) and the serum Br‐ concentration per dose (μg/ml per mg/kg).

Dogs with a higher Cl‐ intake had lower serum Br‐ concentrations. A strong negative correlation was observed between daily Cl‐ intake (mg/kg/day) and the serum Br‐ concentration per dose (μg/ml per mg/kg) (P < 0.01).

These results suggest the importance of considering the dietary content of Cl‐ for KBr‐treated dogs and may also assist with the selection of KBr doses and appropriate diets by clinicians for dogs treated with KBr.

Systematic Optimization of Therapy for Canine Intracranial Neoplasia Using PAC‐1, a Novel and Potent Procaspase‐3 Activator (VCS Award Winner)

L.J. Schlein1, H.C. Podenis1, E.J. Roy1, S. Lezmi2, J.S. Looper3, M. Podell4, P.J. Dickinson5, T.M. Fan6, P.J. Hergenrother7

1University of Illinois, Urbana, IL, USA, 2Ipsen Pharmaceuticals, Urbana, IL, USA, 3Chicago Veterinary Specialty Group, Chicago, IL, USA, 4MedVet Chicago, Chicago, IL, USA, 5University of California, Davis, Davis, CA, USA, 6Department of Veterinary Clinical Medicine, University of Illinois at Urbana‐Champaign, Urbana, IL, USA, 7University of Illinois at Urbana‐Champaign, Urbana, IL, USA

Canine central nervous system tumors can develop deeply within the brain parenchyma, precluding surgical resection and limiting therapeutic options. PAC‐1 is a novel, blood‐brain barrier penetrant, pro‐apoptotic small molecule activator of procaspase‐3 (PC‐3), with orphan drug status for the treatment of human glioblastoma multiforme. PC‐3 is frequently overexpressed in malignantly transformed tissues, providing an opportunity to selectively induce apoptosis in cancer cells with dysregulated upstream apoptotic circuitry. This study evaluates the in vitro activity of PAC‐1 against a panel of brain tumor cells, and the feasibility of combining PAC‐1 with conventional therapies in dogs with spontaneous brain cancer.

Immunohistochemical characterization of PC‐3 was performed in 21 normal canine brains and over 450 canine and human intracranial neoplasms. Murine, canine, and human glioma cell lines were evaluated for PC‐3 expression and in vitro sensitivity to PAC‐1 and radiation monotherapies. Dogs with spontaneously‐arising glioma and meningioma were treated with oral PAC‐1 in combination with conventional therapies to determine tolerability, with serial clinical and MRI outcome assessments.

PC‐3 is overexpressed in canine intracranial neoplasms and in high grade human astrocytomas relative to normal brain tissues. Immortalized canine, human, and murine glioma cell lines show in vitro sensitivity to PAC‐1 and radiation monotherapies at biologically relevant exposures. Six dogs (3 gliomas 3 meningiomas) received PAC‐1 combinatorial therapies and achieved objective responses (1 complete response, 3 partial responses, 2 with stable disease).

PAC‐1 shows therapeutic promise in the treatment of canine intracranial neoplasia, as well as high grade human astrocytoma. Investigation of therapeutic approaches that combine PAC‐1 with radiation therapy and/or temozolomide will further elucidate its therapeutic potential in murine models and canine patients.

Development of a Histopathologic and Flow Cytometric Subclassification System In Canine B‐Cell Lymphoma (VCS Award Winner)

A. Wolf‐Ringwall1, L. Lopez1, B. Charles1, R. Elmslie2, B. Fowler3, J. Lori4, G. Sfiligoi5, A. Skope6, E. Arnold7, D.H. Thamm8, E.J. Ehrhart1, A. Avery1, S. Lana1

1Colorado State University, Fort Collins, CO, USA, 2Veterinary Referral Center of Colorado, Englewood, CO, USA, 3Aspen Meadow Veterinary Specialists, Longmont, CO, USA, 4Animal Emergency and Specialty Clinic, Parker, CO, USA, 5Wheat Ridge Animal Hospital and Veterinary Specialists, Wheat Ridge, CO, USA, 6VCA Alameda East Veterinary Hospital, Denver, CO, USA, 7Animal Hospital Specialty Center, Highlands Ranch, CO, USA, 8Colorado State University Flint Animal Cancer Center, Fort Collins, CO, USA

Canine B‐cell lymphoma is a common hematologic malignancy and a clinically heterogeneous disease. Little progress has been made to improve remission duration and survival. The primary goal of this study was to correlate modified WHO histopathologic subclassification and flow cytometry (FC) with clinical outcome. We hypothesized that histologically distinct subclassifications would have unique cell surface markers as determined by FC, and these subcategories of disease would have variable outcomes when treated with a standardized treatment protocol.

In this prospective study, dogs with naïve multicentric B‐cell lymphoma were treated with a 19‐week CHOP chemotherapy protocol at the Flint Animal Cancer Center and six specialty clinics. All patients had lymph node biopsies and FC collected prior to therapy. Outcome was correlated with patient characteristics, treatment, histopathologic subclassification, and FC features.

A total of 74 dogs were enrolled. Flow cytometry diagnosis of B‐cell lymphoma was consistent with histopathologic diagnosis in all cases. Seventy‐seven percent of dogs were diagnosed with diffuse large B‐cell lymphoma (DLBCL). There were no phenotyping differences between the subcategories of lymphoma detectable by our FC panel. All DLBCLs fell above a minimum forward light scatter value, indicating that cell size, as established by FC, may be useful to rule out this disease. The median overall survival time (MST) for DLBCL (316 days) was statistically significant compared to that of all other histopathologic subgroups combined (114 days, P = 0.0231).

Prognostic factors identified on univariate analysis as significant for median progression free survival time and/or MST included cell size on FC, presence of lymphocytosis, presence of hypercalcemia, need for dose reduction, experiencing complete remission as best response to therapy, and completing CHOP. Further studies are needed to determine factors that may allow for further patient stratification to eventually modify therapy to improve outcome.

Incidence and Prognostic Significance of HER2/neu Expression in Canine Osteosarcoma

M. MaloneyHuss1, C. Gross2, J. Engiles1, N. Mason1

1University of Pennsylvania, Philadelphia, PA, USA, 2Pieper Memorial Veterinary Center, Middletown, CT, USA

Canine osteosarcoma (OSA) is a highly aggressive mesenchymal cancer accounting for 85% of canine bone tumors. OSA commonly arises in the long bones of large breed dogs, and 95% of patients are thought to have micrometastases at diagnosis. Standard of care treatment for canine OSA consists of amputation followed by chemotherapy, with median survival times ranging from 207 to 366 days. Development of metastatic disease limits long‐term survival, with 9.7%–28% of patients alive two years from diagnosis. Treatment of macroscopic metastasis is unrewarding and given this, therapy is focused on prevention of metastatic disease.

HER2/neu is an epidermal growth factor receptor tyrosine kinase that is expressed in human cancers of the prostate, breast, ovary, stomach, and colon. HER2/neu expression is reported in up to 60% of human pediatric osteosarcomas, where increased expression correlates with aggressive biologic behavior and an increased risk of metastasis. Contrary to the membranous expression seen in carcinomas, HER2/neu expression in OSA is principally cytoplasmic. In canine patients, one previous retrospective study reported HER2 expression in 4/10 primary tumor samples, with a suggestion that HER2 expression correlated with decreased survival. We therefore hypothesized that HER2 expression in canine OSA denotes a more aggressive disease phenotype.

To evaluate this hypothesis, we reviewed the medical records of 115 OSA patients with sufficient formalin‐fixed tumor tissue available for immunohistochemical (IHC) analysis. Samples with at least 10% of malignant cells staining positive for HER2/neu were classified as HER2/neu+. Based on these criteria, 95/115 (82.6%) primary tumor samples were HER2/neu+. Additionally, 54 metastatic lesions from a total of 13 dogs were evaluated and 39/54 (72.2%) of these lesions were HER2/neu+. There was no correlation between HER2/neu expression by the primary tumor and HER2/neu expression by metastatic foci.

Samples were then scored based on the percentage of neoplastic cells staining for HER2/neu (distribution score; 50% = 3) and the intensity of HER2/neu staining (weak = 1, moderate = 2, strong = 3). A combined score was obtained by multiplying the distribution score by the intensity score. Based on the combined score, patients were grouped into negative/low (combined score of 1–4), intermediate (> 4, < 7.5), or high (7.5 or greater) categories of HER2/neu expression for Time To Progression (TTP) and Overall Survival (OS) analysis.

To determine whether expression of HER2/neu has prognostic significance, TTP and OS of dogs with negative/low, medium or high combined HER/neu scores were determined. Only dogs treated with standard of care amputation and chemotherapy (n = 55) were included in this analysis. The median TTP and OS of dogs with negative/low HER2/neu expression (n = 22) was 190 and 299 days respectively compared with a TTP and OS of 154 and 253 days respectively for dogs with intermediate HER2/neu expression (n = 23). TTP and OS for dogs with high HER2/neu expression (n = 10) was 127 days and 187 days respectively.

When previously reported (ACVIM 2014), the original data set showed no correlation between HER2/neu expression and TTP or OS. However, in this report, only patients that underwent standard of care amputation and chemotherapy were included in the TTP and OS analysis, revealing a trend toward shorter TTP and OS for dogs with tumors that highly express HER2/neu.

Prospective studies will be required to determine whether a statistically significant correlation exists between HER2/neu status and patient outcome. Nevertheless, taken together, these data show that the majority of canine OSA express HER2/neu, and that dogs with tumors that have high HER2/neu combined scores have a tendency toward earlier metastasis and decreased OS. These data support the evaluation of HER2/neu targeted immunotherapies in both the prevention and possible treatment of metastatic OSA.

MicroRNA‐34a Regulates the Invasive Capacity of Canine Osteosarcoma Cell Lines

C.M. Lopez1, P. Yu2, X. Zhang3, C.A. London4, J.M. Fenger5

1Department of Veterinary Clinical Sciences, The Ohio State University, Columbus, OH, USA, 2The Ohio State University College of Medicine, The Ohio State University Wexner Medical Center, Columbus, OH, USA, 3Center for Biostatistics, The Ohio State University, Columbus, OH, USA, 4The Ohio State University/Tufts University, Columbus, OH, USA, 5The Ohio State University, Columbus, OH, USA

Osteosarcoma (OSA) is the most common bone tumor in children and dogs; however, despite aggressive treatment, little improvement in survival times have been achieved in either dogs or people over the past 15 years. MicroRNAs (miRNAs) are small non‐coding RNAs that regulate gene expression and play a fundamental role in cancer. Our laboratory previously identified a unique miRNA expression signature associated with canine OSA and found miR‐34a expression to be significantly decreased in canine OSA tumors compared to normal canine osteoblasts. These data are concordant with recent studies demonstrating that miR‐34a functions as a tumor‐suppressor gene in human OSA and that loss of miR‐34a is associated with a poor prognosis in human OSA patients. We hypothesize that loss of miR‐34a generates a pattern of gene dysregulation that contributes to the pathogenesis of canine OSA.

MiR‐34a expression was evaluated in primary canine OSA tumors, OSA cell lines and normal canine osteoblasts (Cell Applications, Inc.) using Taqman miRNA assays. To investigate the effects of miR‐34a on OSA cell behavior, canine OSA8 and Abrams cell lines were stably transduced with empty control or pre‐miR‐34a lentiviral constructs (Systems Biosciences) and cell proliferation, cell cycling, cell invasion and migration was evaluated. RNA sequencing of canine OSA8 cells overexpressing miR‐34a was performed and changes in gene expression were validated using real‐time PCR.

Real time PCR demonstrated that miR‐34a expression levels were significantly lower in primary canine OSA tumors and OSA cell lines compared to normal canine osteoblasts. Furthermore, miR‐34a expression was found to be significantly decreased in distant metastases compared to primary OSA tumors, suggesting that loss of miR‐34a is directly involved in the metastatic process. To address the effects of miR‐34a on OSA cell behavior, the canine OSA8 and Abrams cell lines which express low basal levels of miR‐34a were transduced with miR‐34a lentiviral constructs resulting in high levels of miR‐34a. In transformed canine OSA8 and Abrams cell lines, miR‐34a overexpression reduced cellular invasion and migration but had no effect on cell proliferation or cell cycling. Transcriptional profiling of OSA8 cells overexpressing miR‐34a demonstrated dysregulation of several genes, including down‐regulation of putative miR‐34a target genes associated with cellular invasion. Consistent with these findings, enhanced expression of these genes correlates with decreased miR‐34a expression in primary canine OSA tumors.

Taken together, these findings suggest that miR‐34a may play an important role in regulating OSA invasion and loss of miR‐34a in vivo may promote the aggressive behavior of canine OSA.

Outcome of Cats with Intermediate or Large Cell Gastrointestinal Lymphoma Treated with Surgical Mass Resection

K. Tidd1, D. Brown1, A. Durham2, S. Veluvolu3, E.L. Krick3

1Department of Clinical Studies, University of Pennsylvania School of Veterinary Medicine, Philadelphia, PA, USA, 2Department of Pathobiology, University of Pennsylvania School of Veterinary Medicine, Philadelphia, PA, USA, 3University of Pennsylvania School of Veterinary Medicine, Philadelphia, PA, USA

Standard of care treatment for feline gastrointestinal lymphoma consists of adminstration of systemic chemotherapy. The use of surgery except in cases of intestinal obstruction or perforation/peritonitis is largely unknown. The study purpose was to determine the outcome of cats with intermediate and large‐cell gastrointestinal lymphoma following surgical resection.

Records of 34 cats diagnosed with intermediate or large cell gastrointestinal lymphoma based on biopsy of a mass from 2005 to 2015 were retrospectively reviewed. Cases were included if the cat survived surgery and had a known two week post‐operative status. Data collected included anatomic site, surgical margins, surgical complications, chemotherapy, and post‐operative and long‐term outcome.

Within the included cases, affected sites were small intestines (n = 18), large intestines (n = 7), and stomach (n = 5). Complete margins (n = 13) were unrelated to anatomic site of lymphoma. Thirty cats survived to hospital discharge, and 26 cats survived to suture removal approximately two weeks after surgery. Reported causes of post‐operative death included peritonitis, sepsis, and persistent or worsening clinical signs. Overall MST for cats surviving to suture removal was 213 days (95% CI: 48, 465). MST for cats with large intestinal masses was significantly longer than for cats with small intestine or stomach masses for all cats assessed (675 days, 51 days, 96 days; P = 0.03) and for cats that survived to suture removal (675 days, 83 days, 96 days; P = 0.03). Adjunctive systemic chemotherapy (n = 16) and completeness of the surgical margins did not significantly affect survival.

Serious complications were overall uncommon following resection of gastrointestinal lymphoma with the majority of cases found to survive the peri‐operative period. Given the significantly longer survival times of cats with large intestinal lymphoma, it may be reasonable to recommend surgical removal for cats with solitary lymphoma tumors in this location.

Retrospective Evaluation of Toceranib Phosphate (Palladia®) Use in the Treatment of Canine Gastrointestinal Stromal Tumors

E.P. Berger, C.M. Johannes, A.E. Jergens, K. Allenspach, L.E. Fox

Iowa State University, College of Veterinary Medicine, Ames, IA, USA

The purpose of this study was to solicit data regarding current clinical use of toceranib in dogs with gastrointestinal stromal tumors (GIST) to provide initial assessment of biologic activity. The American College of Veterinary Internal Medicine Oncology and Small Animal Internal Medicine listservs were used to solicit data pertaining to cases in which toceranib was used to treat canine GIST. Case data from 28 dogs with histopathogically confirmed GIST were received from 17 participating institutions. Dogs undergoing treatment other than surgical excision prior to toceranib were excluded. Clinical benefit (CB) was observed in 93% (26/28; 24 complete response [CR]/no evidence of disease [NED], 1 partial response [PR], 1 stable disease [SD]). All dogs with microscopic disease experienced CB (21/21; 21 NED), while 71% (5/7; 3 CR, 1 PR, 1 SD) with gross disease experienced CB. Median duration of treatment in dogs experiencing CB was 36 weeks (1–159). At the time of data submission, 8/26 responding dogs were still receiving treatment and an additional 4/26 were experiencing continued NED following completion of prescribed therapy. In dogs with CB, the median toceranib dose was 2.6 mg/kg (0.5–3.5) and in 81% (21/26) was given three days per week. Treatment was well‐tolerated with 57% (16/28) experiencing adverse events, a majority of which were Grade 1 or 2 gastrointestinal toxicities. Biologic activity of toceranib is evident in studied dogs; however, prospective studies are needed to refine the role of toceranib in treatment of GIST in microscopic and gross disease settings.

Association of Tumor‐Infiltrating Regulatory T Cells With Adverse Outcome in Dogs With Solid Tumors

K. Sakai1, S. Maeda, T. Yonezawa, N. Matsuki

Graduate School of Agricultural and Life Sciences, The University of Tokyo, Bunkyo‐ku, Tokyo, Japan

Regulatory T cells (Tregs) play an integral role in suppression of inflammation and the maintenance of immunological tolerance. In addition, Tregs infiltrate into a variety of tumor tissues and thereby suppress antitumor immunity. However, data on the association between tumor‐infiltrating Tregs and prognosis in dogs with various tumors are limited. The purpose of this study was to examine the number of tumor‐infiltrating Tregs and their association with prognosis in dogs with solid tumors.

This study included 113 canine tumors that were surgically excised in the Veterinary Medical Center of the University of Tokyo between May 2009 and December 2015. On the basis of the histopathology, the following diagnoses were determined: squamous cell carcinoma in 14 dogs; hepatocellular carcinoma in 28 dogs; oral malignant melanoma in 29 dogs; lung adenocarcinoma in 16 dogs; and transitional cell carcinoma in 26 dogs. Normal tissue samples were obtained from 6 to 13 healthy dogs. Foxp3‐positive Tregs in the normal and tumor tissues were examined by immunohistochemistry. Cells with clear lymphocyte morphology, distinct nuclear staining for Foxp3, and an unstained cytoplasm were classified as Tregs. Foxp3‐positive Tregs were quantified in different compartments (intratumoral and peritumoral areas). For survival analysis, cases were classified as having a high or low density of Foxp3‐positive Tregs according to the median Treg number in each tumor. Overall survival was compared between cases with a high Treg number and those with a low Treg number using log‐rank test.

Compared to the normal tissues, the number of intratumoral and peritumoral Tregs in the tumor tissues were significantly increased in dogs with squamous cell carcinoma (P < 0.0001), oral malignant melanoma (P < 0.0001), lung adenocarcinoma (P < 0.0001), and transitional cell carcinoma (P < 0.0001). No significant difference was observed in Foxp3‐positive Tregs between normal liver and hepatocellular carcinoma. The overall survival for cases with a high number of intratumoral Tregs was significantly shorter than for cases with a low number of intratumoral Tregs in dogs with lung adenocarcinoma (P = 0.0074), squamous cell carcinoma (P = 0.0145), and transitional cell carcinoma (P = 0.0011). In dogs with hepatocellular carcinoma and oral malignant melanoma, the number of Foxp3‐positive Tregs was not significantly related to the prognosis.

This study demonstrated that Foxp3‐positive Tregs infiltrated into various canine tumor tissues. The higher number of intratumoral Tregs was associated with negative outcomes in dogs with lung adenocarcinoma, squamous cell carcinoma, and transitional cell carcinoma, suggesting that tumor‐infiltrating Tregs contribute to the suppression of antitumor immunity in dogs as well as in humans.

Adoptive T Cell Transfer in Healthy Dogs, and Dogs with Osteosarcoma

B.K. Flesner1, G.W. Wood2, P.A. Gayheart‐Walsten3, F.L. Sonderegger2, C. Henry4, S.M. Bechtel1, K.A. Selting1, D.J. Tate1, J.N. Bryan1, N. Reyes5

1University of Missouri, Columbia, MO, USA, 2TVAX Biomedical, Inc., Olathe, KS, USA, 3Xenometrics LLC, Stillwell, KS, USA, 4Professor of Oncology, Columbia, MO, USA, 5ELIAS Animal Health, Olathe, KS, USA

Osteosarcoma is the most common primary bone malignancy in the dog. Based on favorable findings in infected canine limb‐spare patients, stagnant survival times despite a plethora of adjuvant chemotherapy options, and promising recent preclinical and human immunotherapy outcomes, novel immunotherapy approaches have become attractive treatment options.

We hypothesized that a T cell infusate made up of peripheral blood mononuclear cells (PBMCs) activated and stimulated to proliferate ex vivo by exposure to superantigens could be safely administered to healthy dogs and dogs with osteosarcoma. Additionally, we evaluated the safety of a course of low‐dose human interleukin‐2 (IL‐2) following T cell infusion.

Four dogs were used in the healthy dog study (2 female hounds and 2 male mongrels). Three dogs with spontaneously occurring osteosarcoma have been enrolled in an initial phase I/II clinical trial at the University of Missouri. Two underwent the course of autologous vaccinations and leukapheresis, with one also receiving ex vivo‐activated T cells.

Healthy dogs underwent leukapheresis to isolate PBMCs. T cells were selectively grown ex vivo using IMDM media and superantigens. Following culture, superantigens were removed and the product was then infused into the dogs after premedication; they also received five subcutaneous doses of IL‐2. Dogs’ vitals were analyzed, aerobic/anaerobic cultures were performed, and cytokines were analyzed.

Minimal side effects were noted in the healthy dogs. Vitals showed normal variation during the study; gastrointestinal signs (most often noted as random emesis) were documented in three of four dogs. Pro‐inflammatory cytokines (IL‐6, IFNγ, TNFa, and CCL2) and anti‐inflammatory cytokine IL‐10 were produced, although significant inflammatory‐type symptoms were not noted in the healthy dogs. Both clinical patients were discharged the same day of the leukapheresis procedure. The only clinical patient who received ex vivo‐activated T cells had not been pre‐medicated and experienced a severe inflammatory response and became septic. The dog recovered and is still alive 11 months after diagnosis with no signs of metastasis and with no further cancer therapy.

Based on the encouraging safety data in our healthy canine patients, this adoptive T cell transfer approach has the potential to change treatment recommendations and outcome in dogs diagnosed with osteosarcoma. A phase I/II clinical trial is ongoing to determine if ECI‐OSA‐1 after amputation is indeed safe and results in similar or enhanced efficacy compared to surgery and chemotherapy.

Dysregulation of Healthy Osteoblasts by Canine Osteosarcoma Exosomes

A. Schlueter1, C. Goodall1, R. Troyer1, K. Bathke1, L. Yang2, S. Bracha1

1Oregon State University College of Veterinary Medicine, Corvallis, OR, USA, 2Oregon State University, Corvallis, OR, USA

Exosomes are small lipid vesicles that mediate cell‐to‐cell communication. Cancer cells exploit exosomes to promote metastasis, evade host immune responses, and facilitate angiogenesis. Presently, the effects of osteosarcoma (OSA) exosomes on healthy osteoblasts have not been studied. This is the first study to investigate the proteomic composition of canine OSA exosomes and their impact on osteoblasts. We hypothesized that OSA exosomes carry a unique cargo which manipulates healthy osteoblasts, resulting in alteration of bone turnover.

Exosomes were extracted from OSA (POS, HMPOS, COS) and healthy osteoblast (CnOb) supernatants. The size and concentration of the exosomes were validated via nanoparticle tracking analysis. Exosomal proteomic content was determined by mass spectrometry and validated by Western blot. In vitro assays were performed by co‐culturing OSA exosomes with CnOb. Treated CnOb, untreated CnOb, and exosomes from untreated CnObs were also subjected to proteomic analysis. Fluorescent calcification assays for assessment of hydroxyapatite deposition were performed with OSA cells, osteoblasts alone, and osteoblasts treated with OSA exosomes.

OSA exosomes contained over 200 proteins. In comparison to CnOb derived exosomes, OSA exosomes revealed additional proteins associated with ossification (gelsolin, osteocalcin, vitamin D binding protein). CnOb treated with OSA exosomes showed an upregulation of 129 proteins and a downregulation of 42 proteins compared to untreated CnOb.

OSA cells exhibited osteoblastic characteristics on fluorescent calcification assays. When CnOb were challenged with OSA exosomes, their secretion of hydroxyapatite was increased significantly (P = 0.018).

These findings demonstrate that OSA exosomes contain pro‐osteoblastic cargo which promotes healthy osteoblasts to increase bone deposition. Our study provides a preliminary understanding of the role of OSA‐secreted exosomes and their impact on the bone microenvironment.

Quantification and Prognostic Significance of PD‐L1 Expression in Canine Diffuse Large B‐cell Lymphoma (VCS Award Winner)

L. Ambrosius1, D. Dhawan2, A. Ruple2, J. Ramos‐Vara2, M. Childress2

1Purdue University, Lafayette, IN, USA, 2Purdue University, West Lafayette, IN, USA

Increased expression of programmed cell death ligand 1 (PD‐L1) has been noted in a subset of human lymphomas and recently has been targeted with antibody therapy. The purpose of this study was to measure the expression of PD‐L1 in canine diffuse large B‐cell lymphoma (DLBCL) utilizing quantitative real time polymerase chain reaction (qPCR), and to determine if PD‐L1 expression is correlated with progression free survival (PFS).

A biorepository maintained by the Purdue Comparative Oncology Program was searched for RNA samples of histopathologically‐confirmed cases of canine DLBCL diagnosed between 2010 and 2015. Archived lymph node biopsy samples, processed in TRIzol reagent, from treatment‐naïve dogs were retrieved. Paired RNA samples from lymph node from clinically healthy dogs were retrieved to serve as controls. Multiplex qPCR was performed using TaqMan probes and primers against canine PD‐L1 and GAPDH. The data were analyzed using the 2‐ΔΔCT

method to calculate the fold change in PD‐L1 expression in DLBCL relative to normal lymph node.

Forty‐two dogs met inclusion criteria. Median progression free survival time was 252 days (range 23–466) days. PD‐L1 expression in the DLBCL samples ranged from 0.21 to 7.45 fold different from that measured in normal canine lymph node. 21/42 (50%) dogs had PD‐L1 fold change less than 1.0, and 21/42 (50%) had PD‐L1 fold change more than 1.0 (compared to normal lymph node average). When PD‐L1 expression was dichotomized as fold change greater than or less than 1.0, it was not found to be significantly associated with PFS using the log‐rank test. However, when modeled continuously, PD‐L1 expression was found to be negatively correlated with PFS in both univariate and multivariate Cox proportional hazards models.

These findings demonstrate that PD‐L1 is variably expressed in canine DLBCL, and that higher expression of PD‐L1 correlates with reduced PFS in affected dogs.

In vitro Inhibition of Monocarboxylate Transporters Display Synergistic Activity with Doxorubicin or Metformin in Canine Osteosarcoma (VCS Award Winner)

H.L. Gardner1, J.M. Fenger1, V. Sandanayaka2, C.A. London3

1The Ohio State University, Columbus, OH, USA, 2Nirogyone Therapeutics, Boston, MA, USA, 3The Ohio State University/Tufts University, Columbus, OH, USA

Introduction: Monocarboxylate transporters (MCT) transport lactic acid across plasma membranes, providing a mechanism for tumor cells to meet their bioenergetic needs during glycolysis in a variety of microenvironments. The objective of this work is to explore the therapeutic potential of modulating cellular metabolism through inhibition of MCT1 and MCT4 in canine osteosarcoma, with the ultimate goal of providing proof of principle data to support future clinical studies.

Materials & Methods: MCT1 and MCT4 expression in canine osteosarcoma were assessed with qRT‐PCR, western blotting and IHC. Proliferation, apoptosis and oxygen consumption were evaluated using shRNA and small molecule inhibitors directed against MCT1 and MCT4. Metformin and doxorubicin were evaluated for their ability to synergize with MCT1/4 small molecule inhibitors.

Results: MCT1 is expressed at higher levels relative to MCT4 in canine osteosarcoma cell lines. Small molecule inhibitors directed against MCT1/4 do not significantly decrease cell proliferation. Synergistic activity was observed when cells were treated with doxorubicin or metformin in the presence of MCT1 and MCT4 small molecule inhibitors.

Conclusions: MCT1 and MCT4 are expressed in canine osteosarcoma. Drug combinations with synergistic activity are necessary to realize the full potential of MCT1/4 inhibition.

Expression of PDGFR‐β, VEGFR2, and KIT in Canine Transitional Cell Carcinoma

L.S. Walters1, O. Martin, M.‐J. Sula

University of Tennessee College of Veterinary Medicine, Knoxville, TN, USA

Transitional cell carcinoma (TCC) is the most common neoplasia of the urinary tract in dogs. It tends to be locally invasive and has a moderate metastatic rate. Systemic medical therapy with a combination of a non‐steroidal anti‐inflammatory drug (NSAID) and a chemotherapeutic agent is the current mainstay of treatment. Receptor tyrosine kinases (RTKs) play an important role in promoting cell growth, differentiation and regulation of cell function. RTK inhibitor Palladia has been used anecdotally to treat TCC.

The goal of this study was to evaluate archived normal urinary bladder, TCC, and cystitis bladder samples for expression of Palladia targets: VEGFR2, PDGFR‐β, and stem cell factor receptor (KIT).

Formalin‐fixed paraffin wax‐embedded samples of 30 canine urinary bladders with transitional cell carcinoma, 10 canine urinary bladders with cystitis and 10 normal canine urinary bladders were obtained from the University of Tennessee Department of Pathobiology histologic archives for use in this study. Staining for PDGFR‐β, VEGFR2, and KIT was carried out using standard immunohistochemical procedures developed for use in formalin fixed tissues. Staining intensity was evaluated at 20× in representative sections. Overall staining intensity was scored from 0 to 3 for the entire section. Immunohistochemical staining (% cells stained) was evaluated over ten 40× fields and given a score from 1 to 4. A final scoring product was calculated by multiplying the intensity score by the immunohistochemical staining score

A significant number of TCC samples (30/30) expressed PDGFR‐β compared to cystitis (5/10) and normal bladder samples (3/10). A large number of the TCC samples (11/30) also had mild to moderate staining for KIT compared to cystitis (0/10) and normal bladder (0/10) samples, but the difference was not statistically significant.

Based on this proof of target study, further investigation is warranted to determine clinical response of TCC to Palladia.

Clinical Survey of Veterinary Specialists Evaluating Management of Chemotherapy Induced Vomiting and Inappetence in Dogs

N.P. Punt1, C.M. Johannes2, L.R. Hackbarth3, L.E. Fox2

1Kansas State University, Manhattan, KS, USA, 2Iowa State University, College of Veterinary Medicine, Ames, IA, USA, 3Iowa State University, Ames, IA, USA

Chemotherapy‐induced vomiting and inappetence (CIVI) are common adverse events experienced by cancer patients. However the incidence and recommended treatment following administration of specific chemotherapeutics has not been established in dogs. The aim of this study was to assess perceived emetogenic and inappetence potential of chemotherapeutics and evaluate the clinical use of antiemetics and appetite stimulants in dogs receiving chemotherapy.

A survey was issued on the American College of Veterinary Internal Medicine Oncology and Small Animal Internal Medicine listservs to assess the perceived incidence and clinical management of acute and delayed CIVI associated with 20 chemotherapeutics as defined to VCOG‐CTCAE v1.1.

Eighty‐five completed surveys were evaluated with responses from oncologists (83.5%) in private practice (73%) with >80% of their practice being medical oncology (70.6%). The oral chemotherapeutics evaluated were associated with a minimal/low (≤50%) risk of acute and delayed vomiting independent of VCOG grade. Injectable chemotherapeutics most associated with moderate/high (>50%) risk of causing clinically significant (≥Grade 2) acute or delayed vomiting included cisplatin, dacarbazine, epirubicin, paclitaxel and streptozocin. Of the nine antiemetics evaluated, maropitant and ondansetron were most frequently administered. Inappetence following chemotherapy administration was most commonly treated with maropitant and mirtazapine.

The perceived incidence and management of CIVI in dogs widely differs. Dogs receiving chemotherapeutics with currently perceived lower potential for CIVI may also benefit clinically from treatment with antiemetics and/or appetite stimulants. Further research to understand the true incidence and clinical impact of CIVI will contribute to development of guidelines to appropriately manage these effects in dogs.

Immunohistochemical Expression of P‐Glycoprotein in Dogs with Epitheliotropic and Nonepitheliotropic Cutaneous Lymphoma

S. Lucas1, A.L. Alves1, V.M. Winkel2, L. Rodrigues1, B. Cogliati1

1University of São Paulo, São Paulo, Sao Paulo, Brazil, 2School of Veterinary Medicine and Animal Science – University of São Paulo, São Paulo, Sao Paulo, Brazil

Cutaneous lymphoma represents 1% of skin tumors in dogs and is classified in epitheliotropic and nonepitheliotropic. Unlike the other anatomical forms of lymphoma, cutaneous lymphomas have a low response to the treatment, and dogs with nonepitheliotropic lymphoma have a lower survival time when compared to dogs with epitheliotropic form. The resistance of neoplastic cells to cytotoxic drugs is the major cause of failure of chemotherapy in cancer and one of multiple drug resistance mechanisms (MDR) is the active drug efflux mediated by ATP‐Binding Cassette superfamily (ABC). Among these transporters is P‐glycoprotein (ABCB1).The aim of this study was to investigate ABCB1 immunoexpression in cutaneous epitheliotropic and non‐epitheliotropic lymphoma. Immunohistochemistry was performed in 18 paraffin‐embedded tumor samples of cutaneous lymphoma in dogs (11 nonepitheliotropic and 7 epitheliotropic) using CD20 and CD3 markers for immunophenotyping and ABCB1 antibody (C494 – Enzo Life Sciences®). It was found 5 epitheliotropic T‐cells and 2 B‐cells and 10 nonepitheliotropic T‐cells and 1 B‐cell lymphoma. Two samples of epitheliotropic lymphoma were negative for ABCB1 and the median percentage of labeled cells was 20% (0–40%) whereas in nonepitheliotropic lymphoma, all cases were positive for ABCB1 and the median percentage of labeled cells was 50% (20–90%) and statistical analysis showed a significant difference expression of this marker (P = 0.0162). In conclusion, the majority of cases with cutaneous lymphoma expressed P‐glycoprotein and the subtype nonepitheliotropic showed higher labeled cells than epitheliotropic.

Cell Proliferation Markers Ki‐67 and Mitotic Index in Cats with Intestinal Lymphoma

V.M. Winkel1, A.L. Alves2, B. Cogliati2, A. Reche2, S. Lucas2

1School of Veterinary Medicine and Animal Science – University of São Paulo, São Paulo, Sao Paulo, Brazil, 2University of São Paulo, São Paulo, Sao Paulo, Brazil

Intestinal lymphoma is the main anatomical form of lymphoma diagnosed in cats. Cell proliferation indicators as Ki‐67 expression and mitotic index may be important in assessing the clinical behavior of the tumor and to determine the degree of malignancy. The objectives of this study were to determine the expression of Ki‐67 and mitotic index in intestinal lymphoma of cats and its relation with overall survival time. Forty‐seven samples of cats with intestinal lymphoma were studied and immunohistochemical was performed for: CD3, PAX5 and Ki‐67. The analysis of Ki‐67 was made by counting 1,000 infiltrating lymphocytes in the tumor microenvironment and mitotic index was ranked as mild, moderate or high, by the evaluation of 10 microscopic fields at 40×. The results showed a significant difference in the immunohistochemical expression of Ki‐67 in B and T‐cell Lymphoma (P = 0.001) with a median of 47.2% and 8.5%, respectively. The same was observed when comparing the histological type (lymphoblastic and lymphocytic), with a median of 40.5% and 8.2%, respectively (P = 0.002). Regarding mitotic index, two cats with lymphoblastic B‐cell and one with lymphoblastic T‐cell lymphoma showed high expression, while three animals showed moderate expression and the majority of cases, 41 animals (87.2%) had mild expression. In conclusion, Ki‐67 expression was a better proliferative marker when compared to mitotic index in cats with intestinal lymphoma. Higher expression of Ki‐67 was observed in those animals affected by lymphoblastic or B‐cell and is related to low therapeutic response and shorter survival time.

Molecular Analyses of Pituitary Somatostatin and Dopamine Receptors in Feline Hypersomatotropism (ESVE Award Winner)

C. Scudder1, S. Mirczuk2, K. Richardson2, R. Gostelow1, Y. Forcada1, I. McGonnell2, D. Church1, P. Kenny2, M. Korbonits2, R. Fowkes2, S. Niessen4

1Royal Veterinary College, Hatfield, UK, 2Royal Veterinary College, North Myms, Hatfield, UK, 3The Royal Veterinary College, Hatfield, UK, 4Royal Veterinary College/Newcastle Medical School, Hatfield, UK

Feline hypersomatotropism (HS) (which results in acromegaly) is typically caused by a growth‐ hormone (GH) secreting pituitary adenoma and is thought to be a common cause of feline diabetes mellitus. Medical pituitary inhibition has been challenging in some humans and cats with this condition. This could be related to variable expression of somatostatin (SSTR) and dopamine (DR) receptor expression. There are five different SSTR subtypes (SSTR1‐5) and five different subtypes of DRs, with D2R being the predominant DR subtype in the human pituitary. Pasireotide, a somatostatin analogue with high binding affinity to SSTR1, SSTR2 and SSTR5, is effective at controlling GH secretion in feline acromegalics but somatostatin analogues with high binding affinity to SSTR2 (octreotide and lanreotide) are not.

The aim of this study was to document the receptor expression in feline somatotrophinomas and to compare this to the healthy feline pituitary. We utilised a novel customised multiplex RT‐qPCR assay to investigate the expression profile of the somatostatin and dopamine receptors in GH‐secreting pituitary adenomas. Pituitary adenoma RNA was extracted from 20 GH‐secreting adenomas and eight control cats followed by multiplex RT‐qPCR analyses for multiple genes, including somatostatin receptors (SSTR1‐5), dopamine receptor (D2R), growth hormone receptor, growth hormone secretagogue receptor and the housekeeper RPL18.

The normalised gene expression of SSTR1‐5 in acromegalic cats was SSTR1 = SSTR5 > SSTR2 vs SSTR5 = SSTR1 = SSTR2 in control cats. SSTR3 and SSTR4 were undetectable. Expression of SSTR1, SSTR2 and SSTR5 were upregulated in acromegalics vs controls (SSTR1: P = 0.006, SSTR2: P = 0.005, SSTR5: pSSTR1 increased x9 [95% CI 4 – 12], SSTR2 increased x6.5 [95% CI 2.5 – 8.5] and SSTR5 increased x4 [95% CI 2 – 4]. D2R gene expression was not significantly different between tumorous or normal pituitary tissue; within the hypersomatotropism group D2R exhibited a moderate negative correlation with pituitary volume (r = ‐0.62, P < 0.01, Rsquare 0.39).

Feline GH‐secreting adenomas expressed SSTRs are similar to human GH‐secreting adenomas, having higher SSTR1, SSTR2 and SSTR5 expression than SSTR3 and SSTR4. However, the higher expression of SSTR1 and SSTR5 vs SSTR2 than in most humans may explain the greater biochemical response to pasireotide than octreotide. Decreased D2R expression in larger tumours may be a mechanism allowing unchecked adenoma growth. The receptor expression profiles of feline GH‐secreting adenomas will help the search for improved medical management options of feline hypersomatotropism.

Comparison of Two Doses for ACTH Stimulation Testing in Dogs Suspected of Hypoadrenocorticism

A.S. Botsford1, H.P. Lee2, R. Kemppainen2, E.N. Behrend2

1Auburn University, Auburn, AL, USA, 2Auburn University College of Veterinary Medicine, Auburn, AL, USA

The ACTH stimulation test, using cosyntropin at 5 mcg/kg IV, is the preferred method for diagnosis of hypoadrenocorticism. Previous studies have shown maximal stimulation of the adrenal glands using 1 mcg/kg cosyntropin in normal dogs. No studies have evaluated the efficacy of the lower dose in dogs suspected of hypoadrenocorticism. Our objective was to compare 1 mcg/kg (low‐dose) to 5 mcg/kg (high‐dose) cosyntropin IV to determine if both doses result in a similar adrenocortical response.

Testing was prospectively performed in dogs suspected of hypoadrenocorticism. Dogs suspected of having hypoadrenocorticism received 1 mcg/kg cosyntropin IV followed four hours later by 5 mcg/kg cosyntropin IV. Preliminary studies confirmed the validity of performing two ACTH stimulation tests on the same day. Blood samples were obtained pre‐ and one hour post‐ACTH for each dose (4 measurements total). Cortisol was measured using a previously validated radioimmunoassay. To detect differences in cortisol concentration between cosyntropin doses (1 and 5 mcg/kg) and between time points (baseline and 60‐minutes), data were analyzed using a paired t‐test and equivalence testing by commercial statistical computer programs. Significance was set at the P ≤ 0.05 level.

Twelve dogs were included. Administration of both ACTH doses caused a significant increase in cortisol concentration (P < 0.001 for both). A significant difference was detected between the post‐ACTH cortisol concentrations. The mean post‐ACTH cortisol concentration after the high‐dose was significantly greater than the mean post‐ACTH cortisol concentration obtained after the low‐dose (one‐tailed P value, P = 0.061). Using equivalence testing, the results of the two tests were not equivalent (P > 0.05).

Therefore, the 1 mcg/kg IV dose of cosyntropin does not cause maximal adrenal response as does the standard 5 mcg/kg IV dose in dogs suspected of hypoadrenocorticism. The lower dose is not sufficient for ACTH stimulation testing in those patients suspected of hypoadrenocorticism.

Performance Evaluation of a Veterinary‐Specific Enzyme‐Linked Immunosorbent Assay for Measurement of Canine Serum Cortisol Concentrations

M. Lane1, B. Flatland2, S. Olin2, K. Fecteau2, M. Rick3, L. Giori2

1University of Tennessee, Knoxville, TN, USA, 2College of Veterinary Medicine, University of Tennessee, Knoxville, TN, USA, 3Michigan State University College of Veterinary Medicine, East Lansing, MI, USA

A veterinary‐specific enzyme‐linked immunoassay (ELISA) (SNAP Cortisol Test Kit, IDEXX, USA) is available to veterinarians for in‐house diagnosis and management of canine hypo‐ and hyperadrenocorticism. However, analytical performance of the assay has not been reported in peer‐reviewed literature. The purpose of this study was to investigate the precision of the ELISA, agreement of the ELISA with two chemiluminescence assays (CAs) at two reference laboratories, and the clinical implications of any identified bias.

Stored, frozen canine serum samples were used. ELISA precision was prospectively evaluated by measuring three canine serum pools having low, medium, and high cortisol concentrations, respectively, in triplicate on five consecutive days. A prospective method comparison study was performed by measuring cortisol concentrations in sixty‐three independent canine serum samples using the ELISA and two CA. Methods were compared using Pearson's Correlations, Passing‐Bablok regressions and Bland‐Altman analyses. Clinical implications of identified bias were investigated using an error grid.

Within‐laboratory coefficients of variation for the low, medium, and high concentration pools were 21%, 28.9%, and 13.0%, respectively. Constant and proportional biases between the ELISA and CA methods were present at all concentrations. There was good overall correlation with both CA (r = 0.92 and r = 0.94). Despite identified biases, there was clinical agreement for results in 74.6% of samples.

There is clinical agreement between the ELISA and CA in most cases. Clinical disagreement would occur in 25.4% of cases. Equivocal results reported by the ELISA should be confirmed by a reference laboratory.

Investigation of a Novel Modified Fixed Dose Determination Protocol for Radioiodine Treatment of Feline Hyperthyroidism

W.A. Morre, D. Panciera, G. Daniel, W. Monroe

Virginia Maryland College of Veterinary Medicine, Blacksburg, VA, USA

While radioiodine is the treatment of choice for feline hyperthyroidism because of its safety and efficacy, dose determination of 131I in cats remains controversial. The goal is to achieve a euthyroid state with one treatment, avoiding persistent hyperthyroidism or hypothyroidism. No method of dose determination has been able to consistently achieve this goal. We tested the hypothesis that a radioiodine dose determination based on thyroid scintigraphy would be associated with reduction in hypothyroidism without an increase in persistent hyperthyroidism at 6 months post radioiodine treatment compared with a fixed dose of 131I.

The purpose of this study was to compare a novel method of radioiodine dose determination that utilized thyroid scintigraphy (novel dose group) with a standard fixed dose method (fixed dose group) for the treatment of hyperthyroidism in cats. Specifically, this study focused on determining radioiodine dose by measuring thyroid gland size and percent of 99mTcO4‐ uptake using thyroid scintigraphy. Hyperthyroidism was diagnosed in all cats by finding a serum T4 concentration above the reference interval and characteristics diagnostic of hyperthyroidism on nuclear scintigraphy. Scintigraphy was performed by obtaining images 30 minutes after IV administration of 3.0 mCi 99mTcO4. In cats in the novel dose group, percent dose uptake was calculated by drawing a region of interest around the thyroid gland lobe(s) and dividing the corrected counts (CPM) in the thyroid gland by the 99mTcO4 dose (CPM). Cats with a percent dose uptake < 5%, 5–10%, and > 10% were calculated to receive 3 mCi, 3.5 mCi, or 4.5 mCi of 131I, respectively administered subcutaneously. Thyroid size was assessed by a ratio of the thyroid ROI size (# of pixels) to the salivary ROI size (# of pixels) (TS). Cats with a TS ratio of <5, 5–10, and >10 were calculated to receive 3 mCi, 3.5 mCi, and 4.5 mCi of 131I. The dose of 131I to be given was first determined by measuring percent dose uptake. If the thyroid size fell into a higher dosing category than percent uptake, the dose was increased accordingly. Cats in the fixed dose group received a mean of 4.3 mCi of 131I subcutaneously. Cats were evaluated 6 months after treatment to determine the functional status of the thyroid gland. Hypothyroidism was defined as having a serum T4 concentration below the reference interval and persistent hyperthyroidism was defined as having a serum T4 concentration above the reference interval 6 months post treatment. Fifty‐eight cats were included in the novel dose group and the fixed dose group consisted of 25 cats treated prior to those in the novel dose group. Univariate analysis using Chi‐square and Fisher's exact analysis was used to determine associations between dose determination method and outcome with the level of significance set at P < 0 .05.

Hypothyroidism developed in 32% of cats in the fixed dose group compared to 21% in the novel dose group (P = 0.27). Persistent hyperthyroidism occurred in 8% of cats in the fixed dose group compared to 16% of cats in the novel dose group (P = 0.35).

Results of this study showed no statistical difference in the number of cats remaining persistently hyperthyroid or becoming hypothyroid at 6 months post radioiodine treatment using the novel modified fixed dose protocol compared to the standard fixed dose protocol. More studies investigating this novel radioiodine dose determination are needed that incorporate additional doses of 131 I in addition to the 3 mCi, 3.5 mCi, and 4.5 mCi described in this study.

Comparison Between Glucose Measurements in Canine Whole Blood, Serum and Plasma

E.N. Behrend, N.D. Suchowersky, E.A. Carlson, H.P. Lee

Auburn University College of Veterinary Medicine, Auburn, AL, USA

Blood glucose concentration is frequently measured in dogs as many diseases are associated with hypoglycemia and hyperglycemia such as diabetes mellitus. Use of whole blood, serum or plasma can affect results when using a human glucometer, with serum and plasma yielding more accurate results (Tauk et al; J Am Vet Med Assoc, 2015). The purpose of this study was to investigate the correlation between glucose concentrations in canine whole blood, serum and plasma measured by the AlphaTrak2 (Zoetis, Parsippany, NJ), a veterinary glucometer, and glucose concentration measured by a biochemical analyzer.

In this prospective study, blood samples were collected from 31 dogs. Reason for presentation was not an absolute entrance criteria, but samples from diabetic dogs were sought. Immediately following each blood draw, blood was placed in a heparinized tube for submission to the laboratory and a drop of blood was analyzed by an AlphaTrak2 specifically designated for the study; blood was also used to fill a heparinized capillary tube and a nonheparinized capillary tube. The capillary tubes were centrifuged within 2 minutes after collection to acquire serum (nonheparinized tube) and plasma (heparinized tube) to obtain a glucose concentration measured by the AlphaTrak2. Within 10 minutes following each blood draw, blood was submitted to the Auburn University Clinical Pathology Lab for determination of the glucose concentration by an automated biochemical analyzer. The blood glucose concentration in whole blood, serum and plasma measured by the AlphaTrak2 were compared to the blood glucose concentration measured by the automated analyzer. The Lin concordance correlation coefficient (pc) and Bland‐Altman plots were used to determine correlation and agreement between the results obtained in the different sample types.

Glucose concentration in whole blood measured by the AlphaTrak2 was more strongly correlated with the glucose concentration measured by the biochemical analyzer (pc = 0.92) compared to serum and plasma glucose concentrations (pc = 0.59 and 0.61 for canine serum and plasma, respectively). The mean difference between the glucose concentrations in whole blood, serum and plasma measured by the AlphaTrak2 and the glucose concentration determined by the biochemical analyzer were 15, 119 and 113 mg/dL, respectively.

Our findings suggest that when using an AlphaTrak2, the accuracy and reliability of a glucose measurement obtained may be increased when using whole blood compared to serum or plasma. Thus, use of whole blood would allow for a more correct assessment and diagnosis which are necessary for appropriate therapeutic intervention.

Stereotactic Irradiation of Canine ACTH‐Secreting Pituitary Tumors: Clinical and MRI Time‐Related Response Evaluation

M. Dolera1, L. Malfassi2

1La Cittadina Fondazione Studi E Ricerche Veterinarie, Romanengo, Lombardia, Italy, 2La Cittadina Fondazione Studi e Ricerche Veterinarie, Crema, Lombardia, Italy

Optimal irradiation schedule for canine adrenocorticotropic hormone (ACTH) secreting pituitary tumors has not been defined and response evaluation criteria has not been established.

The aim of this study was to evaluate time‐related clinical and magnetic resonance imaging (MRI) response in dogs suffering from pituitary hyperadrenocorticism with a measurable mass at MRI treated with stereotactic radiotherapy (RT).

A single institution prospective single arm observational study was conducted. Dogs naive of any endocrine treatment suffering from pituitary hyperadrenocorticism diagnosed by standard laboratory and imaging tests were irradiated with 6 MV LINAC using Volumetric Modulated Arc Therapy with a dose of 38 Gy delivered in 10 fractions on alternated days basis. Concomitant administration of trilostane were started. Tumor and endocrine response were evaluated by repeated MRI of pituitary and adrenal glands at 0‐2‐4‐6‐8‐12‐18–24 months. Basal and ACTH‐stimulated cortisol were serially measured and the dose of trilostane to achieve eucortisolism was recorded.

14 dogs were enrolled. Median pituitary tumor volume was 1 cmc (range 0.5–8.5 cmc). Mean basal and stimulated cortisol at the diagnosis were 7.8 mcg/dL and 36.1 mcg/dL. Mean initial dose of trilostane was 2.0 mg/kg/12 h. Complete regression of the tumor was observed in 10/14 dogs at 6–18 months (median 8 months), partial regression in 4/4 at the end of the study. No progressions were observed. Normalization of the thickness of adrenal glands was observed in all dogs at 8–18 months (median 12 months). Eucortisolism was obtained in all except 1 dogs. Trilostane was progressively tapered and withdrawn at 14–18 months (median 16.5 months). No radio toxicity were encountered.

The RT schedule used in this study allowed optimal tumor and endocrine control. A gradual progression in the response was observed, starting with a volumetric response of the pituitary tumor, delayed of several months after the treatment, than a response of the adrenal glands and ultimately a reduction of the serum cortisol levels was achieved. This pattern of response must be taken in consideration during follow up evaluation to adjust the dose of trilostane and to establish evaluation criteria.

Pharmacodynamics of Insulin Detemir and Insulin Glargine (300 U/ml) in Healthy Dogs

H. Fink1, C. Gilor2, C. Herbert3

1The Ohio State University, Veterinary Medical Center, Columbus, OH, USA, 2University of California – Davis, Davis, CA, USA, 3The Ohio State University, Columbus, OH, USA

Insulin glargine 300 U/ml (Toujeo®) and insulin detemir (Levemir®) are synthetic long‐acting insulin analogs associated with minimal intra‐ and inter‐patient variability or episodes of hypoglycemia in people. Toujeo® is more predictable and longer acting compared to glargine 100 U/ml (Lantus®) in people. The purpose of this study was to assess the pharmacokinetics or pharmacodynamics of Toujeo® compared to Levemir® in healthy dogs.

Eight healthy purpose‐bred dogs each received 2.4 nmol/kg of Levemir® and Toujeo® (0.1 U/kg and 0.4 U/kg, respectively, SC injection) on two different days, at least one week apart, in random order. Endogenous and exogenous insulin were measured hourly with the Mercodia Isoinsulin and the Mercodia canine insulin ELISA's. Blood glucose (BG) was measured every 5 minutes and glucose was administered intravenously at a variable rate with the goal of maintaining BG within 10% of baseline BG (“Isoglycemic clamp”). Glucose infusion rate and a decline in endogenous insulin were used to define exogenous insulin action.

Median onset of action was significantly delayed with Toujeo® compared to Levemir® (4.0 hours [3.3–5.8 hours] vs. 0.6 hours [0.6–1.2 hours], P = 0.002). There was no significant difference in peak (median [range]=6.3 hours [5.0–21.3 hours] vs. 4.3 hours [2.9–7.4 hours], P = 0.15) or duration (16.3 hours [6.1–20.1 hours] vs. 10.8 hours [8.8–14.8 hours], P = 0.21) of action between Toujeo® or Levemir®, respectively. Toujeo® demonstrated a peakless time‐action profile in 4/8 dogs.

Toujeo® might be better suited than Levemir® as a once‐daily insulin in some dogs based on its long duration of action and peakless time‐action profile. Day‐to‐day variability in insulin action should still be assessed for both formulations.

Impact of an Auditory Stimulus on Baseline Cortisol Concentrations in Clinically Normal Dogs

T. Gin1, M. Puchot2, A.K. Cook3

1Texas A&M College of Veterinary Medicine, Bryan, TX, USA, 2Texas A&M College of Veterinary Medicine, College Station, TX, USA, 3Texas A&M University College of Vet Med, College Station, TX, USA

Baseline cortisol concentrations are routinely used to screen dogs for hypoadrenocorticism (HOC). A result 20%. We hypothesized that exposure to an unpleasant auditory stimulus prior to sample collection would trigger cortisol production and increase concentrations above the 55 nmol/L cut‐off in healthy dogs.

Twenty‐eight healthy client‐owned dogs were evaluated; median age was 4 years (range 2–9 years) and median weight was 20 kg (range 10–27 kg). Dogs were ineligible for inclusion if they had received short‐ or long‐acting glucocorticoids within the previous 30 and 90 days, respectively.

Dogs were randomly assigned to control (n = 7) and test (brief noise: n = 10; and long noise: n = 11) groups. Each dog and owner waited in a secluded area for 10–15 minutes. “Brief noise” dogs were exposed to the sound of a wet‐dry vacuum for 3 minutes at the start of this period. “Long noise” dogs were exposed to random bursts of wet‐dry vacuum noise for 15 minutes. Blood was then collected in an adjacent room. Samples were processed immediately; serum was frozen at ‐80°C prior to submission.

Median cortisol concentrations for the three groups were similar; control: 79 nmol/L (range 33–201 nmol/L); brief noise: 64 nmol/L (range 36–115 nmol/L), and long noise: 65 nmol/L (range 40–177 nmol/L) (Figure 1).

Exposing dogs to an unpleasant noise stimulus for 3–15 minutes prior to sample collection did not increase serum cortisol concentrations and is unlikely to improve the specificity of baseline cortisol determination as a screening test for HOC.

Diabetic Remission in Cats Examined in UK Primary‐Care Veterinary Practices: Occurrence and Risk Factors

R. Gostelow1, D. O'Neill1, D. Brodbelt1, D. Church1, O. Dunham1, S. Niessen2

1Royal Veterinary College, Hatfield, UK, 2Royal Veterinary College/Newcastle Medical School, Hatfield, UK

Identifying predictors of feline diabetic remission aids veterinarians when advising on diabetic cats’ prognosis, and might provide insight into the etiology of feline diabetic remission. Previous investigations have often studied referral patients or utilized relatively small sample sizes, whereas examining larger, diverse populations might identify new predictors. This study aimed to document the remission rate among a large cohort of diabetic cats examined in UK primary‐care veterinary practices, and identify predictors of remission.

An observational, retrospective cohort study was conducted using electronic patient record data collected through the VetCompass Companion Animal Surveillance Programme from 1st September 2009 to 31st August 2014. Cats with newly‐diagnosed diabetes mellitus (DM) were identified and occurrence of remission was recorded. Cumulative probability of remission was assessed using the Kaplan‐Meier estimator, and predictors of remission were identified using multivariable, time‐dependent Cox proportional hazards modeling. Mortality within 30 days of DM diagnosis, and predictors of death, were also assessed using the Kaplan‐Meier estimator and multivariable binary logistic regression, respectively. Significance was set at < 0.05. Average values are given as median (interquartile range).

Five hundred and eighty‐three cats (347 male, 235 female, 1 unknown sex) were recruited from 103 practices. Five hundred and seventeen (88.2%) were crossbred, with Burmese (n = 24; 4.1%) being the most common pure breed. Median age was 13.0 (10.4–15.0) years, median bodyweight 4.5 (3.8–5.3) kg, and median follow‐up time 232 (42–596) days. One‐year cumulative probability of remission was 13.1% (95% confidence interval [CI] 10–17%). Multivariable analysis revealed glucocorticoid therapy in the 90 days before diagnosis (hazard ratio [HR] 5.9; 95% CI 3.3–10.4; P < 0.0001), and Burmese breed compared to crossbred status (HR 3.7; 95% CI 1.6–8.1; P = 0.001) to be independent predictors of remission. Cumulative probability of survival 30 days’ after diagnosis was 83% (95% CI 79–85%). Lack of medical anti‐hyperglycemic therapy (odds ratio (OR) 14.7; 95% CI 7.5–29.0; P < 0.0001) and lower bodyweight (kg) at diagnosis (OR 1.3; 95% CI 1.0–1.7; P = 0.03) were independently associated with increased odds of death within 30 days of DM diagnosis.

Probability of remission was lower than often documented in referral populations. High mortality following DM diagnosis and owner reluctance to pursue medical therapy present substantial barriers to achieving greater diabetic remission rates in primary‐care practice. Increased likelihood of diabetic remission has not been previously identified in Burmese cats. This could result from differences in the pathogenesis of DM in Burmese compared to other breed types and warrants further investigation.

One‐Year Prospective Randomized Trial Comparing Efficacy of Glargine and Protamine Zinc Insulin in Diabetic Cats

R. Gostelow1, C. Scudder1, K. Hazuchova2, Y. Forcada2, D. Church1, S. Niessen3

1Royal Veterinary College, Hatfield, UK, 2The Royal Veterinary College, Hatfield, UK, 3Royal Veterinary College/Newcastle Medical School, Hatfield, UK

Long‐acting insulin treatment is recommended in feline diabetes mellitus (DM) and insulin glargine has frequently been proposed to offer the greatest efficacy and likelihood of diabetic remission. However, a recent systematic review identified that unbiased, adequately powered, long‐term comparisons of insulin types are lacking. A one‐year, prospective, randomized clinical trial was conducted to compare the glycemic control and remission rates associated with glargine (Lantus®, Sanofi) or human‐recombinant PZI (PZIR; ProZinc®, Boehringer Ingelheim) therapy in diabetic cats.

Recently‐diagnosed diabetic cats (duration 6 weeks – 5 months) were recruited through a dedicated feline DM clinic. Cats were transitioned to a low‐carbohydrate diet 10 days before first examination, with overweight cats (body condition score >5/9) receiving caloric intake aimed at weight loss. Cats were excluded if pre‐recruitment screening identified ketoacidosis, clinical pancreatitis, glucocorticoid/progestogen administration, hyperthyroidism, hypersomatotropism, or other conditions causing likely insulin resistance. Diabetic control was assessed using fructosamine, a validated diabetic clinical score (DCS), mean blood glucose (MBG) during 24‐hour blood glucose curves (BGCs) and twice‐daily (BID) insulin dose (U/kg). Pet and owner quality of life was assessed using the validated DIAQoL‐pet score. Cats were randomized to receive BID glargine or PZIR and underwent a one‐year, standardized re‐examination program. Insulin dose was adjusted according to a single, nadir‐led protocol based on BGC results.

Two‐sample t‐tests or Mann‐Whitney U‐tests assessed for differences in MBG, fructosamine, BID insulin dose, DIAQoL‐pet score and DCS between groups at baseline and at 1, 3, 6 and 12 month(s). Proportion of cats entering remission was compared using Kaplan‐Meier product limit estimates and log rank test, and predictors of remission were identified using multivariable Cox proportional hazards modeling. Significance was set at P < 0.05.

Forty‐six cats were recruited (glargine=24, PZIR=22). No significant difference in MBG, fructosamine, BID insulin dose, DIAQoL‐pet score, or DCS was identified between groups at any time point. Eight glargine‐treated cats (33.3%) and five PZIR‐treated cats (23%) achieved remission (P = 0.42). Baseline bodyweight >5.2 kg (hazard ratio (HR) 2.7; 95% confidence interval (CI) 0.9–8.3; P = 0.07), baseline feline body mass index ≥25% (HR 3.3; 95% CI 1.1–10.0; P = 0.04) and ≥2% weight loss at the 1‐month time point (HR 12.9; 95% CI 3.5–47.4; P = 0.0002) were identified as possibly associated with remission during univariate analysis (criterion: P < 0.1). Multivariable analysis revealed a minimum of 2% weight loss at the 1‐month time point to be the only significant predictor of remission (HR 13.5; 95% CI 2.9–62.5; P = 0.0009).

The current study is unprecedented in terms of size and control of bias. It revealed that glargine and PZIR produced comparable glycemic control and no detectable difference in remission rate among recruited cats. Clinicians should consider implementing a controlled weight loss program as a priority in overweight diabetic cats in order to promote remission.

Biochemical Markers of Bone Turnover in Dogs with Naturally Occurring Hyperadrenocorticism

E. Gunn1, R. Shiel1, K. Jankowska2, E. Matthews2, M. Sekiya2, C.T. Mooney2

1University of Glasgow, Glasgow, UK, 2University College Dublin, Dublin, Dublin, Ireland

The effect of hyperadrenocorticism (HAC) on bone in dogs is controversial. The aim of this study was to investigate its effect on bone turnover by assessing intact and whole parathyroid hormone (PTH) concentrations and markers of both osteoblastic (osteocalcin (OC)) and osteoclastic (carboxyterminal cross‐linked telopeptide of type 1 collagen (ICTP)) activity.

Blood samples were collected from dogs with HAC and a hospital control population. Intact and whole PTH by IRMA (Scantibodies), OC by ELISA (Immunodiagnostic Systems) and ICTP by RIA (Orion Diagnostica) were measured using commercial human kits. Data were reported as median (range) with differences between groups assessed by Mann Whitney testing.

Intact and whole PTH concentrations were significantly higher (P < 0.001 in each case) in 41 dogs with HAC (82.27 (12.18–867.73) pg/mL and 41.9 (5.09 – 1138.18) pg/mL, respectively) compared with 24 control dogs (35.00 (18.27 – 183.27) pg/mL and 18.27 (8.00 – 113.55) pg/mL, respectively). The median OC concentration in 35 dogs with HAC was not significantly different from 20 control dogs (3.3 (0.7–118.8) ng/mL and 4.0 (2.0–15.1) ng/mL), respectively). The median ICTP concentration in 34 dogs with HAC was significantly lower (P < 0.001) than in 19 control dogs (3.2 (1.2–12.8) ng/mL and 7.3 (3.7–21.3) ng/mL, respectively).

Whole and intact PTH concentrations are higher in dogs with HAC. This does not appear to be associated with a decrease in bone formation or an increase in bone resorption as assessed by OC and ICTP measurements, respectively.

A Comparison of Intact Parathyroid Hormone Assays for Use in Dogs

E. Gunn1, R. Shiel2, K. Jankowska2, E. Matthews2, M. Sekiya2, C.T. Mooney2

1University of Glasgow, Glasgow, UK, 2University College Dublin, Dublin, Dublin, Ireland

Several assays are available for measurement of intact parathyroid hormone (iPTH) concentrations but there are few reports comparing assays in dogs.

Intact PTH concentrations were measured using an immunoradiometric assay (IRMA) (Scantibodies) and a chemiluminescent assay (CLIA) (Immulite 2000) designed for human use, and a canine specific enzyme‐linked immunosorbent assay (ELISA) (Immutopics). Samples, in duplicate, were assessed using each technique within 4 hours of each other.

45 and 48 samples, respectively, were available for comparison between IRMA and ELISA and IRMA and CLIA. Correlation between ELISA and IRMA was good (r = 0.90 (95% CI 0.83–0.95), P < 0.001). Bland‐Altman difference plot revealed a bias with ELISA results greater than IRMA values by a mean ± sd of 80.74 ± 89.51% (95% CI ‐94.7–256.2)). Using CLIA, 39 (81.3%) samples had iPTH concentrations at or below the reported limit of detection of the assay (3.0 pg/mL). The remaining nine dogs had detectable but low concentrations ranging from 4.82–452.0 pg/mL. By comparison these values were higher by IRMA (range, 125.55 – 867.73 pg/mL).

The CLIA cannot be recommended for measuring canine iPTH as most values were undetectable. This may be related to the antibodies used within the assay. The non‐isotopic ELISA conferred advantages over IRMA with regard to ease of use but yielded higher values that would have impacted clinical decision‐making. Additionally a high prevalence of duplicate errors and gel formation impacted reliability. Whilst more technically demanding, the IRMA proved the most reliable and robust method evaluated.

Predicted Ionized Calcium Can Accurately Confirm Hypo/Hypercalcemia in Dogs, Even with Missing Biochemistry Values

K.L. Boedec1, J. Danner2

1Centre Hospitalier Veterinaire FREGIS, Arcueil, Ile‐de‐France, France, 2University of Illinois College of Veterinary Medicine, URBANA, IL, USA

Abnormal total calcium (tCa) values do not necessarily imply hypo/hypercalcemia based on measured ionized calcium (miCa). Unfortunately, miCa is not readily available in practice.

In dogs with increased or decreased tCa, a multivariate adaptive regression splines model was used to predict miCa from creatinine, albumin, tCa, phosphorus, sodium, potassium, chloride, alkaline phosphatase (ALP), triglycerides, and age. Sensitivity and specificity of predicted ionized calcium (piCa) were compared to those of corrected tCa for true hyper/hypocalcemia (McNemar's test). The impact of missing biochemistry values on predictive model accuracy was assessed by replacing missing biochemistry variables by their mid‐reference interval value. Accuracy was considered altered if piCa became significantly different from miCa (Wilcoxon test).

Out of 53 dogs with elevated tCa, 26 were truly hypercalcemic based on miCa. Sensitivity of piCa (92%) and of corrected tCa (100%) were equivalent to detect hypercalcemia (P = 0.5). Specificity of piCa (81%) was significantly higher than specificity of corrected tCa (29%) to diagnose hypercalcemia (P < 0.001). Out of 19 dogs with decreased tCa, 10 were truly hypocalcemic based on miCa. Sensitivity of piCa (90%) was significantly higher from sensitivity of corrected tCa (20%) to detect hypocalcemia (P = 0.01). Specificity of piCa (78%) and of corrected tCa (100%) were equivalent to diagnose hypocalcemia (P = 0.5). Without missing values, piCa was not significantly different from miCa (P = 0.12). Triglycerides, phosphorus, sodium, and ALP could be missing without altering piCa accuracy.

In dogs with abnormal tCa values, true hyper/hypocalcemia may be assessed by piCa even if triglycerides, phosphorus, sodium, and ALP measurements are not available.

Serum Symmetric Dimethylarginine Concentrations in Hyperthyroid Cats with and without Azotemic Chronic Kidney Disease

T.L. Williams

University of Cambridge, Cambridge, UK

Hyperthyroidism decreases body muscle mass and serum creatinine concentrations which can confound diagnosis of concurrent chronic kidney disease (CKD). Serum symmetric dimethylarginine (SDMA) concentrations are influenced to a lesser extent by muscle mass, and might be a better biomarker of CKD in hyperthyroidism. This study aimed to evaluate serum SDMA concentrations in hyperthyroid cats with and without azotemic CKD.

Hyperthyroid and euthyroid (without azotemic CKD [control]) cats >9 years old were included. Hyperthyroid cats received anti‐thyroid medication and were monitored for azotemia (serum creatinine concentration >1.7 mg/dL) during a three month follow up period. Hyperthyroid cats that were azotemic before treatment or during follow up were included in the hyperthyroid‐CKD group. Kruskal‐Wallis with Dunn's multiple comparisons test was used to compare baseline serum creatinine and SDMA concentrations (SDMA measured by IDEXX Laboratories) between the control, hyperthyroid‐CKD and hyperthyroid non‐azotemic groups. Median [range] values are presented.

Serum creatinine concentrations in control (1.4 [1.1–1.8] mg/dL, n = 8) and hyperthyroid groups were not significantly different, however serum creatinine was significantly higher in the hyperthyroid‐CKD group (1.9 [1.1–2.4] mg/dL, n = 5) than the hyperthyroid non‐azotemic group (1.1 [0.7–1.3] mg/dL, n = 4; P < 0.05). Serum SDMA concentrations were not significantly (P > 0.05) different between the control (11 [9–13] μg/dL), hyperthyroid‐CKD (14 [7–17] μg/dL) and hyperthyroid non‐azotemic groups (13 [11–19] μg/dL).

Serum SDMA concentrations did not distinguish hyperthyroid cats with and without azotemic CKD. The influence of hyperthyroidism on serum SDMA concentrations warrants further investigation.

Increased Adipocyte Size and Leukocyte Infiltration in Adipose Tissue of Obese Dogs

C.Z. Cavalcante1, R. Sgarbossa1, V. Gmyterco1, L. Capriglione1, T. Casagrande2, T. Fuchs2, J. Castro1, J. Engracia1, M. Farias1, A. Nishiyama3

1Pontifical Catholic University of Parana, Curitiba, Parana, Brazil, 2Universidade Positivo, Curitiba, Parana, Brazil, 3Federal University of Parana, Curitiba, Parana, Brazil

Obesity is a frequent nutritional disorder in dogs and it is associated with several metabolic alterations and represents a risk factor for several systemic comorbidities. However, there is controversy about which of the adipose tissues contributes more to these complications in the canine species. The present study aimed to evaluate the subcutaneous and visceral adipose tissue of ten obese female dogs and ten female dogs with eutrophic body condition, regarding adipocyte size and presence of leukocyte infiltration. The study was conducted with routine animals from the Veterinary Hospital, and all of the animals were submitted to a complete clinical examination, classification based on the body condition score and blood exams to exclude concomitant diseases. Fragments of subcutaneous and visceral adipose tissue were obtained during electives ovariosalpingohisterectomy procedures, the samples were stained with Hematoxylin‐Eosin and analyzed with Zeiss Microscope software. It was observed that obese dogs presented remarkably larger adipocyte size when compared to animals with ideal body condition, in a significance level of 5% through Student's T‐test. Likewise, it was noted the presence of significantly higher leukocyte infiltrate in obese animals, in a significance level of 5% in the Kruskal‐Wallis test, compared to animals with normal body condition that did not have leukocyte infiltration detected. However, the size of adipocytes and leukocyte infiltrate in obese animals did not differ significantly between the visceral and subcutaneous adipose tissue samples. It was concluded, therefore, that obesity in dogs is related with the increase of adipocytes and with presence of leukocyte infiltration, but it was not possible to state in this study that different fat deposits have differentiated metabolic and endocrine levels, which could interfere in a specific way in the inherent processes in obese and diabetic patients.

Lipid Peroxidation in Obese Dogs Before and After the Weight Loss Program

C.Z. Cavalcante1, A. Roncoski1, R. Sgarbossa1, V. Gmyterco1, A.P.S. Lopes1, G. Sechi1, A. Nishiyama2

1Pontifical Catholic University of Parana, Curitiba, Parana, Brazil, 2Federal University of Parana, Curitiba, Parana, Brazil

Obesity is a frequent nutritional disorder in dogs and cats and it results in oxidative stress. The evaluation of oxidative stress can be performed through the process of lipid peroxidation, and malonate is one of the most abundant aldehydes. The evaluation method is called “test for substances that react with thiobarbituric acid” (TBARS). Objective of the present study is to evaluate oxidative stress by measuring the TBARS present in samples of obese dogs before and after the weight loss program. The work was carried out with routine animals of the Veterinary Hospital of the Pontifical Catholic University of Parana (located in Parana, Brazil). The animals were submitted to a complete clinical examination, blood count, biochemistry, urinalysis and systolic blood pressure measurement. Subsequently, they were classified as obese on the body scale; body mass index and percentage of body fat. The animals were divided into group 1 (pre‐diet) and group 2 (post‐diet). Ten obese animals were selected and submitted to a hypocaloric diet loss program. TBARS was performed using the Elisa technique. All animals in groups 1 and 2 presented TBARS values above the reference value, and there was a statistically significant reduction of the values from group 1 to group 2 (P = 0.034). Based on the results, it was concluded that obese dogs present higher production of lipid peroxidation than animals with ideal body scale, leading to the development of oxidative stress and elevation of TBARS concentrations.

Evaluation of Corticotropin (Acthelea) Effect in the ACTH Stimulation Test in Dogs with Hyperadrenocorticism

P.V. Furtado1, D.C. Rodini2, R. Soila2, P. Lopes1, R. Fukumori1, C.A. Oliveira1

1FMVZ/USP, São Paulo, Sao Paulo, Brazil, 2PROVET, São Paulo, Sao Paulo, Brazil

One of the most important endocrine tests for diagnosis and control of hyperadrenocorticism (HAC) in dogs is the ACTH stimulation. The objective of this study was to evaluate the efficacy of corticotropin inducing cortisol production in dogs. This possibility can be used as an alternative to tetracosactrin (Synacthen) which is, currently, the most used drug in this adrenal challenge test, mainly in Latin America.

Thirty ACTH stimulation tests were performed on 15 dogs divided into 2 groups: 5 control animals (2 females and 3 males, healthy, mean age 2.5 ± 1.5 years) and 10 treatment animals (6 females and 4 males, previously hyperadrenocorticism diagnosed, mean age 11.7 ± 2.3 years). All animals were controlling them, in other words, first were challenged with Synacthen and then with Acthelea. For dogs on trilostano (1–2 mg/kg/day), ACTH stimulation tests were initiated 4–6 hours postpill administration as per manufacturer instructions. The first stimulation tests were performed using Synacthen as the drug of election to induce cortisol response and the second, which were separated by 15 days period, using Acthelea. Blood samples were obtained just before injection and 60 minute post‐ACTH. The first to assess the cortisol basal concentration of control group (CSynbas and CEleabas) and Treated group (TtoSynbas and TtoEleapos) and the second collection for assessment of cortisol concentration post‐ACTH in the control group (CSynpos and CEleapos) and treated group (TtoSynpos and TtoEleapos).The doses were: 0.04 mL/Kg for Synacthen and 0.08 mL/Kg for Acthelea, IV. Serum cortisol concentrations were measured at the PROVET/BR Hormone Laboratory using previously validated radioimmunoassay (MP Biomedicals).

Results were expressed for their medium‐sized ± SEM. For the control group, the results obtained for the baseline values were 0.89 ± 0.08 μg/dL (CSynbas) and 0.96 ± 0.12 μg/dL (CEleabas), with P = 0.68; and post‐ACTH values were 9.14 ± 0.46 μg/dL (CSynpos) and 9.00 ± 0.71 μg/dL (CEleapos), with P = 0.58 There was no significant difference between the mean baseline and post‐stimulation in the control group, when comparing the results among drugs. For the treated group, the baseline values were 1.71 ± 0.34 μg/dL (TtoSynbas) and 1.82 ± 0.48 μg/dL (TtoEleabas), with P = 0.82; and after ACTH values were 7.85 ± 2.2 μg/dL (TtoSynpos) and 7.90 ± 2.1 μg/dL (TtoEleapos), with P = 0.98. Therefore, no significant difference was observed between the values obtained when comparing the use of these two drugs applied in the tests.

According to the results obtained in this study, we can conclude that the Acthelea was effective in stimulating the production of cortisol in dogs and can be used as an option in the ACTH stimulation test, particularly in places where Synacthen is not available.

Evaluation of Triglycerides, Cholesterol and Glucose in Dogs with Mastocytoma Submitted to Surgery and Chemotherapy

E.V. Januário1, S.R. Melo2, B.C. Miranda2, E.B.M. Zanutto2, J.M. Matera2

1Faculty of Veterinary Medicine and Animal Science, University of São Paulo, São Paulo, Sao Paulo, Brazil, 2University of Sao Paulo, São paulo, Sao Paulo, Brazil

Humans show metabolic changes after chemotherapy, due to the action mechanism of chemotherapy drugs and/or concomitant corticosteroid therapy, with health consequences. The aim of this study is to evaluate similar changes in dogs.

Laboratory tests of 19 dogs diagnosed with cutaneous mastocytoma treated at the Small Animal Surgery Department – Faculdade de Medicina Veterinária e Zootecnia, Universidade de São Paulo, submitted to different chemotherapy protocols after tumor excision were analyzed.

We compared the levels of triglycerides, cholesterol and fasting glucose of these patients submitted to chemotherapy according to the protocols: Intravenous I‐vinblastine (2 mg/m2 – 8 weekly sessions, 4 fortnightly sessions); metronomic dose of oral II‐lomustine (2.84 mg/ m2 SID) for 16 weeks associated with oral prednisone (2 mg/kg SID); Intravenous synthetic III‐phosphoethanolamine (70 mg/m2, 4 weekly sessions) without glucocorticoids. Measurements were performed in a period before (P0); in three periods during (P1, P2, P3: monthly in the Protocols I and II and weekly in Protocol III) and 30 days after completion of chemotherapy (C).

Nineteen dogs were assessed: 11 (57.89%) females and eight (42.11%) males with a mean age of 8.7 years. At P0 the dogs showed mean values of triglycerides (MVTRI) of 100.28 mg/dL (standard deviation = SD = 55.58); cholesterol (MVCHOL) 222.84 mg/dL (SD = 59.97) and glucose (MVGLUC) 74.5 mg/dL (SD = 24.14). P1 = MVTRI 199.728 mg/dL (SD = 187.51), MVCHOL = 230.2 mg/dL (SD = 63.19), MVGLUC = 91.5 (SD = 1206.14). In period P2 = MVTRI =148.47 mg/dL (SD = 138.84), MVCHOL = 217.57 mg/dL (SD = 108.93), MVGLUC = 123.24 (SD = 8.42). In P3, MVTRI = 1293.55 mg/dL (SD = 1471.78), MVCHOL = 279.45 mg/dL (SD = 104.32). In period C, MVTRI = 139.4 (SD = 63.26) and MVCHOL = 242.8 (SD = 88.63).

Hypertriglyceridemia and elevated serum glucose were observed in dogs treated with chemotherapy associated with glucocorticoids. Further studies on these changes are necessary.

Retrospective Incidence of Hyperthyroidism in Domestic Cats at a Veterinary Hospital of São Paulo, Brazil

L. Scalize, M. Jericó, T. Pacini, R. Castilho

Anhembi Morumbi University, São Paulo, Sao Paulo, Brazil

In recent decades, the prevalence of feline hyperthyroidism (FHT) has increased significantly worldwide and is currently a major endocrine disease that causes high morbidity in middle‐aged cats. Due to the lack of studies on this endocrine disease in Brazil, a retrospective study of stored serum samples from all cats older than eight years treated at the Veterinary Hospital of Anhembi Morumbi University (São Paulo, Brazil) was carried out from January 2013 to January 2015, regardless of the main complaint that led them to be treated at the service. Thyroid function was evaluated through total thyroxine (T4T) measurements. A total of 212 cat serum samples were sent to a private laboratory, where T4T was measured by radioimmunoassay (RIA). Of the 212 assessed serum samples, 3.30% confirmed feline hyperthyroidism, 59.91% were within the normal range and 36.79% were below the normal value, suggesting sick euthyroid syndrome. In 2.35% of the samples, T4T values were close to the upper limit value, which raises the possibility that these cats had FHT, but whose comorbidities at the time of blood collection reduced T4T values in blood serum. We conclude that even though the prevalence of FHT in the assessed group was shown to be discreet, this clinical suspicion should be part of the differential diagnosis in cats aged ≥ eight years.

Alterations in the Hormones Regulating Blood Glucose Concentrations in Healthy Dogs Undergoing Hydrocortisone Administration

H. Kim1, J.‐H. Kang2, C.‐J. Sung1, J. Yun1, P. Son1, S. Woo1, B.‐T. Kang1, M.‐P. Yang1

1College of Veterinary Medicine, Chungbuk National University, Cheongju, Ch'ungch'ong‐bukto, Republic of Korea, 2Chungbuk National University College of Veterinary Medicine/Veterinary Medical Center at Chungbuk National University, Cheongju, Ch'ungch'ong‐bukto, Republic of Korea

Endo‐ and exogenous hypercortisolemia is well known to associated with insulin resistance in dogs. However, the types of changes in the levels of hormones involved in maintaining the blood glucose levels of dogs with hypercortisolemia are unknown. The objective of this study was to assess whether insulin sensitivity and the factors regulating blood glucose levels in dogs are altered by chronic administration of hydrocortisone.

Six laboratory dogs received hydrocortisone (8.5 mg/kg) orally every 12 hours for 90 days. All variables were serially examined during the 90 days of hydrocortisone administration. Insulin sensitivity was evaluated by homeostatic model assessment (HOMA)insulin sensitivity. Visceral fat was quantified by computed tomography.

With hydrocortisone administration, HOMAinsulin sensitivity was significantly decreased, although the fasting glucose concentrations were maintained within the reference range; and the visceral fat mass was significantly increased. In addition, HOMAinsulin sensitivity was significantly correlated with visceral fat mass. The changes in serum concentrations of leptin, adiponectin, insulin, glucose‐dependent insulinotrophic polypeptide (GIP), pancreatic polypeptide, glucagon, and ghrelin were significant. Except for KC‐like chemokine and interleukin‐7, differences between the concentrations of inflammatory cytokines were significant. The correlations between visceral fat mass, leptin, adiponectin, ghrelin, GIP, insulin concentrations, and HOMAinsulin sensitivity were significant.

These results suggest that insulin sensitivity decreased by hydrocortisone administration is likely associated with increased visceral fat mass and changes in serum concentrations of the hormones regulating glucose levels in dogs. Further study clarifying our results in naturally occurring hyperadrenocorticism in dogs is needed.

Persistent Hypertonicity in Diabetic Dogs

C. Perez, T. Brandt, T. Schermerhorn

Kansas State University, Manhattan, KS, USA

Tonicity is a function of the serum concentrations of effective osmoles. Sodium (NA), chloride (CL) and glucose (GLU) are the major endogenous osmoles that affect tonicity, but GLU contributes significantly only in hyperglycemic states. The pathophysiological consequences of hypertonicity (HT) are not completely understood but persistent HT may affect disease progression in diabetic humans. Many diabetic dogs experience persistent hyperglycemia but the role of HT in these dogs has not been investigated. The study objective was to evaluate diabetic dogs for evidence of persistent HT.

A retrospective evaluation was performed to compare pertinent solute concentrations and other tonicity indicators between insulin‐treated diabetic dogs (DD; n = 37) dogs and non‐diabetic dogs (NDD; n = 164). For data analysis, comparisons between groups were made using Student's T test, correlations were performed using Spearman's rho test and ROC analysis was used to study the relationship between the mean corpuscular volume difference (dMCV) and fructosamine.

In DD, tonicity (315 vs 311 mOsm/L; P = 0.02) and GLU (295 vs 110 mg/dL; P = 0.0001) were elevated while NA and CL were decreased compared with NDD. HT (tonicity > 315 mOsm/L) occurred more frequently in DD (45.9% vs 31.1% for NDD). The dMCV, a marker for HT, was significantly increased in DD (3.9 vs 1.7 fL for NDD). Increased dMCV predicted an elevated fructosamine (> 450 μmol/L) with 82.6% sensitivity and 42.8% specificity. The results show that stable, insulin‐treated diabetic dogs frequently have HT and that GLU is the major osmole that contributes to HT in DD. An increase in dMCV is a modest predictor for serum fructosamine, suggesting the latter is an indirect marker for hypertonicity in DD. In conclusion, dogs with moderate/poor glycemic control show evidence of persistent HT, although additional studies are needed to investigate the role of persistent HT and its possible consequences for diabetic dogs.

Serum 25‐Hydroxyvitamin D Concentrations in Dogs with Acute Pancreatitis

D.‐I. Kim1, J.‐H. Kang2, P. Son1, H. Kim1, J. Yun1, S. Woo1, C.‐J. Sung1, B.‐T. Kang1, M.‐P. Yang1

1College of Veterinary Medicine, Chungbuk National University, Cheongju, Ch'ungch'ong‐bukto, Republic of Korea, 2Chungbuk National University College of Veterinary Medicine/Veterinary Medical Center at Chungbuk National University, Cheongju, Ch'ungch'ong‐bukto, Republic of Korea

An imbalance in vitamin D status has recently been suggested to be associated with disease progression in humans with pancreatitis. To date, however, no studies have investigated the association between vitamin D status and acute pancreatitis (AP) in dogs. Therefore, the purpose of this study was to examine differences between the concentrations of serum 25‐hydroxyvitamin D (25‐[OH]D) in healthy dogs vs. AP dogs, and, if so, to determine the correlations between 25‐(OH)D, ionized calcium (iCa), and C‐reactive protein (CRP) levels.

This prospective observational cohort study included 20 client‐owned dogs with newly diagnosed AP and 22 otherwise healthy dogs with similar body‐condition scores. The circulating concentrations of serum 25‐(OH)D, blood iCa, and serum CRP were measured.

The serum 25‐(OH)D and blood iCa concentrations were significantly lower, and the serum CRP concentration was significantly higher in dogs with AP than in healthy dogs. There were significant differences between the concentrations of serum 25‐(OH)D in AP survivors and non‐survivors. The differences between the concentrations of blood iCa and serum CRP in AP survivors and non‐survivors were not significant. Concentrations of 25‐(OH)D were positively correlated with the concentrations of iCa and albumin in dogs with AP. Conversely, a negative correlation was observed between serum concentrations of CRP and iCa in the AP group.

This study found that dogs with AP exhibited decreased 25‐(OH)D levels, which might be the result of inflammatory responses rather than a cause of AP. Moreover, alterations in serum 25‐(OH)D concentrations might be associated with disease severity or mortality in dogs with AP.

Clinical Relationship Between Cholestatic Disease and Pituitary‐Dependent Hyperadrenocorticism in Dogs: A Retrospective Analysis

K.‐H. Kim1, W.‐J. Song1, Q. Li2, M.‐O. Ryu1, S.‐C. Park1, H.‐J. Kim1, S.‐Y. Kim1, H.‐M. Yang1, H.‐Y. Youn1

1Seoul National University, Seoul, N/A, Republic of Korea, 2Seoul National University, Seoul, N/A, China (People's Republic)

A high prevalence of cholestatic disease, including gall bladder mucocele (GBM) has been reported in dogs with naturally‐occurring pituitary‐dependent hyperadrenocorticism. Differences exist in the clinical features of dogs with PDH and concurrent cholestatic disease, and also is the management of these dogs with trilostane. In this study, 65 client‐owned dogs with naturally‐occurring PDH were analyzed. Each dog was treated with trilostane for at least 3 months before the study, and had a good clinical response, as determined by owners. Statistical comparisons of clinical signs, results of routine blood tests, basal and post‐ACTH cortisol concentration, and optimal trilostane dosage were made after dogs were separated into the following 3 groups by ultrasonographic imaging: normal on ultrasound (NOU) group, cholestasis group, and GBM group. The GBM group had more severe clinical signs, and significantly different total serum cholesterol concentration and post‐ACTH stimulation cortisol concentration at the time of diagnosis. Dogs that weighed <6 kg had a significantly higher prevalence of cholestatic disease than did the other dogs (P = 0.003). The optimal trilostane dosages for the GBM and cholestasis groups were 2.5 and 1.5 times the dosage of the NOU group, respectively (P < 0.001). Gall bladder disease associated with cholestatic disease is correlated with PDH in dogs, both in its clinical features and drug management. These findings may be associated with hypercholesterolemia, unidentified genetic factors, and the hydrophobic nature of trilostane.

Usefulness of Thyrotropin Stimulation Test in Dogs with Hypothyroidism, Non‐thyroidal Illness, or Receiving Hydrocortisone

C.‐J. Sung1, J.‐H. Kang2, H. Kim1, S. Woo1, P. Son1, J. Yun1, K.‐J. Na1, M.‐P. Yang1

1College of Veterinary Medicine, Chungbuk National University, Cheongju, Ch'ungch'ong‐bukto, Republic of Korea, 2Chungbuk National University College of Veterinary Medicine/Veterinary Medical Center at Chungbuk National University, Cheongju, Ch'ungch'ong‐bukto, Republic of Korea

Hypothyroidism is a common endocrinopathy in dogs; however, the diagnosis of hypothyroidism may be difficult because nonthyroidal illness (NTI) and medication might affect the results of thyroid function tests. The objective of this study was to evaluate the usefulness of a thyrotropin (TSH) stimulation test for diagnosing hypothyroidism.

This case‐controlled prospective study included 41 dogs, 35 client‐owned dogs who were suspected to have altered thyroid function and 6 dogs who had received oral hydrocortisone for 90 days. The TSH stimulation test was performed in each dog. Based on the results of the test, the 35 client‐owned dogs were divided into 3 groups as follows: dogs with hypothyroidism, dogs with NTI, and dogs without hypothyroidism.

The mean basal serum TT4 concentration of dogs with hydrocortisone administration was similar to that of dogs with hypothyroidism (0.32 μg/dL and 0.27 μg/dL, respectively). The mean basal serum TT4 concentrations of dogs with NTI and dogs without hypothyroidism were 0.78 μg/dL and 1.56 μg/dL, respectively. The mean post‐rhTSH TT4 concentrations of dogs administered hydrocortisone, dogs with NTI, and dogs without hypothyroidism were increased to 2.17 μg/dL, 2.19 μg/dL, and 3.49 μg/dL, respectively; while the post‐rhTSH TT4 concentration of dogs with hypothyroidism remained < 1.0 μg/dL. The peak post‐rhTSH TT4 concentration in dogs administered hydrocortisone occurred at 4 hours after stimulation. Those dogs with NTI and dogs without hypothyroidism occurred at 6 hours after injection of rhTSH.

These results indicate that a TSH stimulation test using rhTSH could be a useful method for diagnosing hypothyroidism, even in dogs receiving long‐term, high‐dose hydrocortisone. In clinical practice, the TSH stimulation test is valuable for the diagnosis of hypothyroidism in dogs with a variety of NTIs, hyperadrenocorticism, or dogs administered glucocorticoids.

Effect of Freezing on Biological Activity of Depot Tetracosactide in Healthy Dogs

T.L. Pinto1, A.B. Vieira2, R.S. Barros1, V. Nowosh1, M.F.A. Balaro1, M.C. Salomão1, N.X. de Alencar1

1College of Veterinary Medicine, Universidade Federal Fluminense, Niteroi, Rio de Janeiro, Brazil, 2Ross University School of Veterinary Medicine, Basseterre, Saint George Basseterre, Saint Kitts and Nevis

The adrenocorticotropic hormone (ACTH) stimulation test is an excellent tool for assessing adrenal gland function in dogs. Among the various presentations of ACTH available for testing, the Depot Tetracosactide (Synacthen® Depot) is economically advantageous in some countries, because of the volume and concentration of its ampoule (1.0 mg/1.0 mL). This advantage depends on the reuse of the surplus volume of the ampoule, as the indicated dose for dogs is 0.25 mL per dog. Our aim was to evaluate the effect of freezing on the biological activity of Depot Tetracosactide for the stimulation of adrenal gland in healthy dogs. Thirteen healthy adult dogs were submitted to ACTH stimulation test (0.25 mL/dog/IM) four times at intervals of two months between each test (0, 2, 4 and 6 months). The first stimulation test was done in every dog with the first aliquot of a fresh, refrigerated ampoule, opened at the time of application. The remaining three aliquots were stored at −20°C in plastic syringes of 50 units protected from light. Samples for determination of serum cortisol by radioimmunoassay were obtained before, one and two hours after administration of the ACTH. All samples were thawed once and analyzed in duplicate by ANOVA. Serum cortisol levels did not differ significantly (P < 0.01) before or after freezing for all sampling times. We conclude that Depot Tetracosactide can be stored in plastic syringes at −20°C for up to six months without interfering with the biological stability of the hormone.

Evaluation of the Clinical Accuracy of 6 Portable Blood Glucose Meters in Dogs

M.‐K. Shin1, J.‐H. Kang2, J. Yun1, B. Shin1, H. Kim1, S. Woo1, P. Son1, C.‐J. Sung1, W.‐B. Baek1, B.‐T. Kang1, K.‐J. Na1, M.‐P. Yang1

1College of Veterinary Medicine, Chungbuk National University, Cheongju, Ch'ungch'ong‐bukto, Republic of Korea, 2Chungbuk National University College of Veterinary Medicine/Veterinary Medical Center at Chungbuk National University, Cheongju, Ch'ungch'ong‐bukto, Republic of Korea

Portable blood glucose meters (PBGMs) are widely used because of their usefulness in the control of blood glucose levels. However, the accuracy of PBGMs has been questioned repeatedly, and factors that can interfere with blood glucose measurements have been identified. The objective of this study was to evaluate the factors associated with deviations in the measurements obtained by PBGMs from the blood glucose reference range in dogs.

This prospective study included 67 client‐owned dogs presenting to the veterinary medical center for various diseases, including diabetes mellitus. All of the blood samples (N = 219) were collected from jugular veins. Blood glucose concentrations measured by 6 PBGMs (AlphaTrak2, iPet, OneTouch Ultra, Cerapet, VetMate, Optium Xceed) were compared to those measured by the reference method, and accuracies were evaluated according to ISO 15197:2013. Statistical agreements of PBGM measurements were determined by Spearman correlation analysis, Passing‐Bablok regression analysis, the Wilcoxon signed‐rank test, Bland‐Altman plotting, and error grid analysis. Mann‐Whitney and Kruskal‐Wallis tests were performed to assess the variables expected to affect glucose measurements.

Spearman correlation analysis revealed high correlations between the values obtained by the PBGMs and the reference method. However, Bland‐Altman plotting showed that deviations of the PBGM values from the reference range increased when actual glucose concentrations increased. Decreased packed cell volume (PCV) resulted in overestimation or underestimation of blood glucose values from iPet and OneTouch Ultra, respectively. Hypertriglyceridemia resulted in overestimation of glucose values from AlphaTrak2, iPet, Cerapet, and VetMate. Hyperproteinemia resulted in overestimation of glucose values from OneTouch Ultra and Optium Xceed. Error grid analysis revealed that AlphaTrak2, OneTouch Ultra, Cerapet, and Optium Xceed were adequate, based on minimum accuracy criteria of ISO 15197:2013.

The results suggest that hypertriglyceridemia, hyperproteinemia, and decreased PCV can significantly affect the results obtained by various PBGMs. Clarification of the causes of errors in glucose measurements should lead to more accurate assessment of disease and appropriate glucose control and treatment for diabetic dogs.

Relationship of the Mucosal Microbiota to Inflammation and Disease Activity in Cats with Intestinal Lymphoma

K. Garraway1, C.M. Johannes2, A. Bryan2, J. Peauroi3, G. Rossi4, T. Atherly5, C. Mosher6, K. Allenspach2, A.E. Jergens2

1Small Animal Internal Medicine, Veterinary Clinical Sciences, Iowa State University, College of Veterinary Medicine, Ames, IA, Ames, IA, USA, 2Iowa State University, College of Veterinary Medicine, Ames, IA, USA, 3VDx Veterinary Diagnostics and Preclinical Research Services, Davis, CA, USA, 4School of Biosciences and Veterinary Medicine, University of Camerino, Camerino, N/A, Italy, 5University of Florida, Department of Oral Biology, Gainesville, FL, USA, 6Iowa State University, Ames, IA, USA

Previous studies have shown a relationship between mucosal bacteria and host responses in cats with inflammatory bowel disease (IBD) and lymphoma (LSA). However, these earlier studies have not investigated the specific microbial etiologic factors nor the selective recruitment of mucosal tumor‐infiltrating immune (CD11b+) cells which may promote tumor progression. The objective of the present study was to investigate the relationship between pro‐inflammatory microbiota (including Fusobacterium spp., Helicobacter spp., Enterobacteriaceae) and mucosal CD11b+ cells in cats with small cell LSA vs. IBD. A retrospective analysis of cats diagnosed with intestinal LSA (n = 14) or IBD (n = 14) from 2010–2015 was performed. Following detailed diagnostic evaluation and confirmation of a histopathologic diagnosis, the mucosal microbiota of endoscopic intestinal biopsies of IBD and LSA cats was evaluated by fluorescence in situ hybridization (FISH) targeting the 16S rRNA genes of select pro‐inflammatory microbiota. Host responses to microbial dysbiosis were studied using immunohistochemistry to identify cells of the gastrointestinal mucosa expressing intestinal NF‐kB (p65) and CD11b+. Results indicated that most IBD/LSA cats were middle aged or older having chronic gastrointestinal signs of several months duration. Analysis by FISH showed that mucosal bacteria (EUB) were uncommonly observed in the duodenum of diseased cats. The numbers of mucosal Bacteroides spp. and Clostridium spp. were increased (P < 0.05) in IBD cats as compared to cats with LSA. Conversely, total numbers of Fusobacterium spp. were increased (P < 0.05) in colonic tissues of cats with LSA vs. IBD (LSA: median 30 [range 0–378]; IBD: median 9 [range 0–346]). Moreover, the numbers of Fusobacterium spp. found within adherent mucus of colons in cats with LSA were increased (P ≤ 0.05) relative to other mucosal compartments. Mucosal cells expressing NF‐kB and CD11b+ were increased (P < 0.05) in intestines of cats with LSA vs. IBD suggesting the presence of a pro‐inflammatory microenvironment. In conclusion, alterations exist in the composition of the mucosal microbiota in cats with LSA and IBD. The intestine of cats with LSA is selectively enriched with bacterial species, including Fusobacterium spp. These microbiota may trigger recruitment of tumor‐infiltrating immune (CD11b+) cells contributing to a pro‐inflammatory environment that is conducive for intestinal LSA progression.

Effects of Akkermansia muciniphila Supplementation on Markers of Intestinal Permeability in Dogs Following Antibiotic Treatment

M.C. Jugan1, A. Rudinsky2, O. Paliy3, A. Gordon3, J. Daniels1, P. Boyaka1, C. Gilor4

1The Ohio State University, Columbus, OH, USA, 2The Ohio State University Veterinary Clinical Sciences, Columbus, OH, USA, 3Wright State University, Dayton, OH, USA, 4University of California – Davis, Davis, CA, USA

Diarrhea and other sequelae of gastrointestinal hyperpermeability are common complications of antibiotic therapy. Akkermansia muciniphila is a mucin‐degrading bacterium, positively associated with gastrointestinal epithelial health and decreased permeability. The objectives of this study were to measure effects of Akkermansia administration on markers of gastrointestinal permeability following antibiotic administration.

Eight healthy, purpose‐bred dogs were randomized to receive either Akkermansia (109 CFU/kg; N = 4) or vehicle (N = 4) for 6 days following a 7‐day course of metronidazole. After a 20‐day washout, dogs were crossed‐over to the alternate treatment. After an additional 20‐day washout, the experiment was repeated with amoxicillin‐clavulanate. Fecal Akkermansia qPCR and plasma concentrations (measured by ELISA) of cytokeratin‐18, lipopolysaccharide, and glucagon‐like peptides (GLP‐1, GLP‐2) were measured at baseline (T0), post‐antibiotic (T1), and post‐treatment (vehicle or Akkermansia) (T2). For each antibiotic, absolute or delta concentrations were compared between time‐points using repeated measures ANOVA.

Akkermansia was detected in feces in 7/8 dogs following supplementation (T2) but not at T0 or T1. Delta (T2‐T1) cytokeratin‐18 after metronidazole was significantly lower on vehicle (−0.27 ng/mL) vs. Akkermansia (2.4 ng/ml; P = 0.03). Cytokeratin‐18 concentrations tended to decrease from T0 to T1 on amoxicillin‐clavulanate (P = 0.05). Post‐prandial GLP‐1 concentrations (38.2 pM/ml) were higher than pre‐prandial (15.5 pM/ml) concentrations. No adverse side‐effects or other significant biomarker alterations were noted.

Akkermansia muciniphila PCR detection suggested successful gastrointestinal transit following oral supplementation in dogs, with an effect on gastrointestinal epithelium based on plasma cytokeratin‐18 alterations. Further study is needed to determine impact in dogs with naturally occurring disease.

The Impact of Oral Tylosin on the Fecal Microbiota of Healthy Dogs

A.C. Manchester1, C. Webb2, J.A. Lidbury3, J.M. Steiner4, J.S. Suchodolski3

1Department of Clinical Sciences, Colorado State University, Fort Collins, CO, USA, 2Colorado State University, Fort Collins, CO, USA, 3Gastrointestinal Laboratory, Texas A&M University, College station, TX, USA, 4Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA

Dysbiosis is thought to play a major role in the pathogenesis of intestinal disease. Treatment of intestinal disease may involve the use of antimicrobials, but their effects on the microbiota are not fully understood and several antibiotics have been shown to promote dysbiosis. The aim of this study was to prospectively evaluate the impact of tylosin administration on the fecal microbiota of healthy dogs.

Ten healthy pet dogs were randomly assigned to one of 2 groups in a double‐blinded fashion: 6 dogs treated with oral tylosin at 20 mg/kg and 4 dogs treated with a placebo capsule, each given q12 hr for 7 days. All dogs were maintained on their usual diet. Fecal samples were collected on day 0. A fecal score (range: 1 to 7) was noted daily during drug administration, with additional fecal samples collected on days 7, 10, 21, and 63. Fecal samples were assessed using qPCR for 7 bacterial groups: Faecalibacterium, Turicibacter, E. coli, Streptococcus, Blautia, C. hiranonis, and Fusobacterium. Results were analyzed for each bacterial group and also compiled mathematically into a single numeric value, the Dysbiosis Index (DI). Parameters were compared using a Friedman's test, followed by Dunn's post‐test. A P‐value < 0.05 was considered statistically significant.

None of the dogs in either group developed diarrhea. The DI increased significantly on day 7 (median [range]: (2.9 [+1.7 to +3.8]) compared to day 0 (−6.5 [−8.3 to +1.7]; P = 0.03) with tylosin treatment. The DI remained significantly increased on day 10 (P = 0.03) but not on day 21 (P = 0.06). The abundance of Faecalibacterium was significantly decreased after tylosin treatment. No significant changes in the abundance of any bacterial group or in the DI were noted in dogs receiving placebo.

Tylosin had a profound effect on the fecal microbiota, which in some dogs persisted for at least 2 months. Further studies are warranted to determine if dysbiosis secondary to antimicrobial administration increases patient risk of developing disease, intestinal or otherwise, in the future.

Prognostic Indicators in Dogs with Extrahepatic Bile Duct Obstruction Secondary to Pancreatitis

S. Palermo1, D.C. Brown1, M. Rondeau2

1University of Pennsylvania, Philadelphia, PA, USA, 2University of Pennsylvania School of Veterinary Medicine, Philadelphia, PA, USA

Pancreatitis may cause extrahepatic bile duct obstruction secondary to local inflammation in the vicinity of the common bile duct. To date, there has been limited evidence to support the optimal treatment for this condition. The purpose of this study was to describe the clinical presentation and outcome of dogs that develop extrahepatic bile duct obstruction secondary to pancreatitis, and assess factors associated with survival.

A retrospective review of medical records from the Matthew J. Ryan Veterinary Hospital at the University of Pennsylvania from 1989 to 2016 was performed. Cases were included if clinical signs and ultrasonographic findings were consistent with pancreatitis and extrahepatic bile duct obstruction was suspected based on ultrasonographic common bile duct dilation of greater than 3 mm and presence of hyperbilirubinemia. Cases were excluded if another cause for bile duct obstruction was identified, or if the medical record was incomplete. Medical records were searched for physical examination findings, clinicopathologic data, diagnostic imaging findings, treatment and outcome. Referring veterinarians were contacted when necessary for information regarding survival. Median survival times were determined by the use of the Kaplan‐Meier product limit method and Cox multivariable survival methods were employed to determine which factors were associated with survival time following diagnosis.

Complete medical records were available for 45 dogs. 30 dogs were managed medically and 15 dogs were managed surgically. The overall median survival time was 241 days (95% CI 25, 631). Dogs 9 years of age or older with azotemia had a 9.9 greater hazard for death (95% CI 2.5, 38.1; P = 0.001) compared to dogs younger than 9 years old without azotemia. Dogs with a temperature greater than 102.5°F at admission had a 3.1 greater hazard for death (95% CI 1.3, 7.7; P = 0.013) than dogs with a body temperature of less than 102.5°F at admission. Dogs without gallbladder distension on ultrasound had a 4.4 greater hazard for death (95% CI 1.3, 15.4; P = 0.018) compared to dogs with gallbladder distension.

In dogs with extrahepatic bile duct obstruction secondary to pancreatitis, age of 9 years or older, azotemia at admission, body temperature greater than 102.5°F at admission and lack of gallbladder distension on ultrasound are associated with an increased risk of death. Medical vs. surgical management, total bilirubin at admission, change in total bilirubin during hospitalization, and magnitude of increased liver enzyme activity were not significantly associated with outcome in this group of dogs.

IL‐1beta Protein Expression is Increased in Biopsies of Dogs with Severe Inflammatory Bowel Disease

M. Hawes1, A. Riddle1, J. Kirk1, K. Allenspach2

1Royal Veterinary College, London, North Mymms, UK, 2Iowa State University, College of Veterinary Medicine, Ames, IA, USA

Canine Inflammatory Bowel Disease (IBD) is believed to be a multifactorial disease due to an abnormal immune response to intestinal microbes in genetically susceptible individuals. IL‐1β, a pro‐inflammatory cytokine, has previously been shown to be elevated in biopsies of dogs with food responsive disease in a small number of cases. The purpose of this study was to investigate duodenal levels of IL‐1β protein in a larger number of dogs with different clinical severity of IBD. Duodenal biopsies from ten dogs each with mild IBD (Canine Chronic Enteropathy Clinical Activity Index, CCECAI 0–5), moderate IBD (CCECAI 6–8), severe IBD (CCECAI ≥9) and dogs with protein‐losing enteropathy (PLE), as well as 10 healthy Beagle dogs were included in the study. Biopsy samples were weighed prior to manipulation. Samples were homogenized at 20 Hz for 2 minutes in 500 μL homogenization buffer (PBS containing 0.05% Tween‐20) with protease inhibitor cocktail. Tissue homogenates were centrifuged at 12500 RCF for 2 minutes and the supernatants frozen at −80°C until the day of ELISA. Quantification of IL‐1β protein expression was determined using a commercially available canine‐specific IL‐1β ELISA (Kingfisher Biotech Inc) following the manufacturer's assay instructions. All samples were analysed in duplicate. Median IL‐1β expression (pg/ml/mg tissue) was 121.6 (range 38.3–219.4) in the severe IBD group; 19.7 (range 10.9–24.8) in the moderate severity IBD group; 13.3 (range 8.0–29.8) in the mild severity IBD group, 13.0 (range 8.0–27.9) in the PLE group, and 0.9 (range 0.3–1.4) in the healthy control dogs. There was a statistically significant difference in IL‐1β expression between all IBD dogs and the healthy controls (p2 = 0.47, P < 0.001). The results of this study support previous findings that IL‐1β may be involved in the pathogenesis of canine IBD.

Histologic Findings in Endoscopically Derived Intestinal Biopsies from Clinically Healthy Cats

S. Marsilio1, J.M. Steiner2, J.A. Lidbury3, J.S. Suchodolski3, M. Ackerman4

1Texas A&M University, Gastrointestinal Laboratory, College Station, TX, USA, 2Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA, 3Gastrointestinal Laboratory, Texas A&M University, College station, TX, USA, 4Department of Veterinary Pathology, College of Veterinary Medicine, Iowa State University, Ames, IA, USA

Criteria developed by the World Small Animal Veterinary Association (WSAVA) are commonly used for the assessment of feline intestinal biopsy specimens. However, the histological characteristics of intestinal biopsy specimens collected from healthy cats have not been well characterized using this scoring system. The aim of this study was to describe the histologic findings in endoscopically derived intestinal biopsy specimens from clinically healthy cats.

Twelve clinically healthy, client‐owned cats underwent gastroduodenoscopy at the Veterinary Medical Teaching Hospital at Texas A&M University (Animal Use Protocol 2015‐0276 CA). Tissue specimens were collected from the stomach (10 cats) and the duodenum (12 cats) and were evaluated according to WSAVA criteria by a board‐certified pathologist (MRA) who was blinded to the health status of the cats.

The median age of the cats was 9.5 years (min‐max: 3 to 18 years). Sample quality was reported as very good in all cases. Histologic evaluation of the stomach revealed lymphoplasmacytic infiltration in all specimens evaluated, which was graded as minimal (1 cat), minimal to mild (2 cats), mild (3 cats), mild to moderate (2 cats), or moderate (2 cats). Histologic evaluation of duodenal biopsies revealed evidence of lymphoplasmacytic infiltration in 11 of 12 cats, which was described as minimal to mild (3 cats), mild (1 cat), mild to moderate (4 cats), or moderate (3 cats). On assessment of duodenal specimens 1 cat was diagnosed as a having diffuse low‐grade lymphoma. Architectural changes were present in 8 of 10 gastric and 12 of 12 duodenal sections. These consisted of fibrosis (8 cats) in gastric specimens and lacteal dilation (1 cat), crypt abscesses (1 cat), fibrosis (3 cats), and crypt hyperplasia (11 cats) in duodenal specimens.

In conclusion, intestinal biopsies from clinically healthy cats commonly show mucosal changes that are considered abnormal when evaluated according to WSAVA guidelines.

Effect of the Bile Acid Sequestrant Cholestyramine on Fecal Bile Acid Concentrations in Healthy Dogs

B.C. Guard1, C. Alexander2, J.B. Honneffer1, J.A. Lidbury3, J.M. Steiner1, K.S. Swanson4, J.S. Suchodolski3

1Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA, 2Division of Nutritional Sciences/University of Illinois, Urbana, IL, USA, 3Gastrointestinal Laboratory, Texas A&M University, College station, TX, USA, 4Department of Animal Sciences and Division of Nutritional Sciences/Department of Veterinary Clinical Medicine, Urbana, IL, USA

Cholestyramine is a bile acid sequestrant that acts in the gastrointestinal tract by binding bile acids, thus preventing their reabsorption. In human patients with inflammatory bowel disease (IBD), cholestyramine is often used to treat individuals with symptoms of diarrhea. Preliminary studies in canine patients with IBD suggest that their fecal bile acid concentrations are altered, typified by an increase in primary bile acids and a decrease in secondary bile acids compared to healthy controls. The purpose of this study was to determine the effect of cholestyramine on the fecal bile acids profile in healthy dogs.

Baseline fecal samples were collected from healthy Beagle dogs (n = 12) two weeks before administration of cholestyramine. Additional fecal samples were collected after two weeks of daily cholestyramine administration and again two weeks later after a washout period. All dogs were maintained on the same maintenance diet during the study and were fed 11.4 g/day of cholestyramine powder (8 g active ingredient) suspended in 75 mL of water during the cholestyramine administration period. Fecal concentrations of primary bile acids (i.e., cholic acid and chenodeoxycholic acid) and secondary bile acids (i.e., lithocholic acid, deoxycholic acid, and ursodeoxycholic acid) were evaluated using gas chromatography coupled with mass spectrometry. Data were assessed for normality using the Shapiro‐Wilk test and differences in bile acid concentrations were compared using the Friedman's test. The Dunn's post‐test was used where appropriate. Statistical significance was set at P < 0.05.

Total bile acids concentrations were significantly increased after two weeks of cholestyramine administration (median [min‐max]: 14.2 μg/mg [5.7–25.5 μg/mg]) when compared to baseline values (median [min‐max]: 6.6 μg/mg [5.6–17.7 μg/mg]; P = 0.0062). Total secondary bile acids concentrations were significantly increased after two weeks of cholestyramine administration (median [min‐max]: 13.1 μg/mg [5.5–23.3 μg/mg]) when compared to baseline values (median [min‐max]: 6.2 μg/mg [5.4–17.1 μg/mg]; P = 0.0183). After two weeks of washout there were no parameters that remained significantly different compared to the original baseline values.

In conclusion, the fecal bile acids profile is altered in healthy dogs after cholestyramine administration. Further studies are needed to understand the potential clinical utility of cholestyramine as a therapeutic option in canine patients with gastrointestinal disease.

Fecal Fatty Acid Concentrations in Dogs with Exocrine Pancreatic Insufficiency Receiving Enzyme Supplementation

J.B. Honneffer1, A.B. Blake1, J.C. Parambeth1, O.C. Kennedy2, B.C. Guard1, J.A. Lidbury3, J.M. Steiner1, J.S. Suchodolski3

1Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA, 2Epi4Dogs Foundation, Inc., Farmville, VA, Farmville, VA, USA, 3Gastrointestinal Laboratory, Texas A&M University, College station, TX, USA

Exocrine pancreatic insufficiency (EPI) is a disease characterized by insufficient synthesis and secretion of pancreatic enzymes by the exocrine pancreas, resulting in malassimilation of macro‐nutrients. For example, insufficient pancreatic lipase prevents normal digestion of dietary fat. Consequently, EPI would be expected to be associated with excessive fat (e.g., fatty acids) remaining in the feces. Treatment of EPI includes oral supplementation of pancreatic digestive enzymes and is often effective at decreasing severity of clinical signs, but it is unclear if assimilation normalizes concomitantly. This study evaluated fecal fatty acid (FA) concentrations in dogs with EPI undergoing enzyme supplementation. The hypothesis of this study was that fecal fatty acid concentrations would be increased in dogs with EPI compared to those of healthy dogs, even when being treated with enzyme supplementation.

Fatty acid concentrations were quantified in fecal samples from 34 dogs diagnosed with EPI that were being treated with pancreatic enzyme supplements and from 82 healthy control dogs using an in‐house gas chromatography/mass spectrometry (GC/MS) assay. Target analytes included palmitic acid (16:0), stearic acid (18:0), oleic acid (18:1ω9), linoleic acid (18:2ω6), α‐linolenic acid (18:3ω3), gondoic acid (20:1ω9), and erucic acid (22:1ω9). A Mann‐Whitney U test was used for comparison between groups. P‐values were adjusted for multiple comparisons and statistical significance was set at P < 0.05.

Fecal FAs were significantly increased in the feces of dogs with EPI (all P < 0.001). Concentrations of FAs for dogs with EPI vs. healthy control dogs were (median [min‐max] μg/mg of lyophilized feces): palmitic acid (12.0 [1.6–48.4] vs. 4.2 [1.3–13.4]), stearic acid (6.6 [1.1–43.2] vs. 2.3 [0.9–14.4]), oleic acid (13.8 [1.8–70.2] vs. 4.0 [0.3–16.9]), linoleic acid (10.3 [1.7–34.5] vs. 4.0 [0.4–29.7]), α‐linolenic acid (1.0 [0.2–5.5] vs. 0.4 [0.1–3.5]), gondoic acid (0.69 [0.10–2.36] vs. 0.19 [0.03–0.61]), and erucic acid (0.07 [0.02–0.96] vs. 0.03 [0.01–0.27]). The sum of measured fecal FAs was 46.8 [9.0–174.6] vs. 15.3 [4.0–61.3].

Fecal fatty acid concentrations were increased in dogs with EPI, even while being treated with pancreatic enzyme supplementation. These data are consistent with malassimilation of fat in these patients.

Ki‐67/CD3 Index in Canine Inflammatory Bowel Disease

S. Karlovits1, A. Manz1, K. Allenspach2, I. Walter3, S. Kummer3, U. Reichart3, A. Tichy4, B. Richter5, A. Fuchs‐Baumgartinger5, I.A. Burgener6, N. Luckschander‐Zeller1

1Clinic for Internal Medicine, Department for Companion Animals and Horses University of Veterinary Medicine, Vienna, Austria, Vienna, Wien, Austria2Iowa State University, College of Veterinary Medicine, Ames, IA, USA, 3VetCore Facility for Research, University of Veterinary Medicine, Vienna, Austria, Vienna, Wien, Austria4Department of Biomedical Sciences, University of Veterinary Medicine, Vienna, Austria;, Vienna, Wien, Austria5Department for Pathobiology, Institute of Pathology and Forensic Veterinary Medicine, University of Veterinary Medicine, Vienna, Austria, Vienna, Wien, Austria6Clinic for Small Animals and Infectious Diseases, VetMedUni Vienna, Austria, Vienna, Wien, Austria

Hypersensitivity and proliferation of T‐cells play a key role in the pathogenesis of IBD. In human medicine, Ki‐67 is an indicator for cell growth, but there are only few studies in canine medicine.

The aim of this study was to investigate Ki‐67 in relation to T‐cells as a marker for canine IBD.

According to clinical signs and histopathologic results eleven privately owned dogs were grouped as IBD (IBD). Six healthy beagles served as controls (CO). All dogs were clinically assessed using the Canine Chronic Enteropathy Activity Index (CCECAI) scoring system. Duodenal mucosal biopsy samples were endoscopically retrieved for histopathological examination using the World Small Animal Veterinary Association‐scheme (WSAVA). Double‐marked immunofluorescence microscopy was used to investigate Ki‐67 for identifying proliferating cells in combination with CD3 to detect T‐cells in 4 different areas (epithelium (E) and lamina propria (LP) of villi and crypts, respectively).

In IBD‐patients there was a significant increase in the clinical scoring (M = 5.2 ± 2.7) vs. CO (constant 0) (P < 0.001). Furthermore, double‐positive cells were significantly increased in the LP of the crypt region of IBD‐dogs (M = 1.23 ± 1.33 cells/mm2) vs. CO (M = 0.120 ± 0.143 cells/mm2) (P = 0.022). In this area, a significant correlation was found between CCECAI and the Ki‐67/CD3 index (r = 0.670).

In summary, the Ki‐67/CD3 index is significant upregulated in the LP, specifically in the crypt area of IBD dogs indicating a metabolically active region. In conclusion, Ki‐67/CD3 could be a useful marker for canine IBD activity.

Videofluoroscopic Swallow Study Features of Obstructive Lower Esophageal Sphincter Disorders in Dogs

M. Grobman1, J. Schachtel1, T. Lever2, C. Reinero1

1University of Missouri College of Veterinary Medicine, Columbia, MO, USA, 2University of Missouri School of Medicine Department: Otolaryngology‐Head and Neck Surgery, Columbia, MO, USA

In people, lower esophageal sphincter (LES) disorders causing functional obstruction have been identified as rare causes of megaesophagus (ME). Obstructive LES disorders have been identified in dogs but lack clear diagnostic criteria. Because these conditions may respond to targeted therapy, correct diagnosis is critical. Videofluoroscopic swallow study (VFSS) is the gold‐standard for the diagnosis of dysphagia in dogs. Our objective was to characterize VFSS features of LES achalasia‐like syndrome and dyssynchrony in dogs. We hypothesized that LES achalasia‐like syndrome (LES‐AS) and LES dyssynchrony (LES‐D) could be distinguished from each other and normal dogs using standardized VFSS criteria.

Retrospective study. Sixty dogs presented to the University of Missouri Veterinary Health Center between April 2015‐November 2016 for VFSS using a free‐feeding standing protocol using at least 2 consistencies of iohexol or barium impregnated food (thin liquid, puree, and kibble). Dogs with a functional LES obstruction (n = 11) were further evaluated using subjective and objective metrics: presence of ME; distal esophageal bolus height relative to T12; maximal LES opening; esophageal contractility (in the face of physiologic distention), propulsions (aboral bolus movement), and primary or secondary peristalsis; “bird beak”; and timing of LES opening/closure.

Using VFSS, dogs were determined to be healthy (n = 6), or had LES‐AS (n = 6), LES‐D (n = 5), and “other” (primary esophageal dysmotility/GERD; n = 1). All dogs with LES‐AS and 2/5 dogs with LES‐D had ME, with the ratio of the esophageal bolus column height:height of T12 vertebral body being significantly greater in LES‐AS compared to healthy dogs (P = 0.021). Dogs with LES‐AS had a significantly lower ratio of distal esophageal column height:LES opening height compared to normal (P = 0.010) and LES‐D (P = 0.043) dogs. Subjective contractility of the esophagus was normal (n = 3 LES‐D), hypocontractile (n = 2 LES‐D, n = 1 LES‐AS) or acontractile (n = 5 LES‐AS). Propulsions were completely effective (n = 3 LES‐D) or ineffective (n = 2 LES‐D). Presence of any primary and/or secondary peristaltic waves were observed in all dogs with LES‐D (n = 5) compared to LES‐AS (n = 1). A “bird‐beak” sign was seen in all dogs with LES‐AS. All dogs with LES‐D had abnormal timing of LES opening and closing.

Functional obstruction of the LES due to LES‐AS and LES‐D represent discrete disease states distinguishable by VFSS.

Effect of an Extruded Vegetable Diet on Fecal Microbiota of Dogs with Food‐Responsive Enteropathy

F. Bresciani1, Y. Minamoto2, J.S. Suchodolski2, G. Galiazzo1, C. Vecchiato1, C. Pinna1, G. Biagi1, M. Pietra1

1Department of Veterinary Medical Sciences, University of Bologna, Ozzano dell'Emilia, Emilia‐Romagna, Italy, 2Gastrointestinal laboratory, Texas A&M University, College Station, TX, USA

Intestinal dysbiosis and adverse food reactions are involved in the pathogenesis of food responsive enteropathy (FRE) in dogs. Various options for an elimination diet are available, and a vegetable dry food is one alternative. Dietary interventions are thought to alter gut microbial communities in healthy individuals and the resolution of dysbiosis is expected in diseased animals concurrent with remission of clinical signs. Therefore, the aim of this study was to evaluate changes in fecal microbiota in dogs with FRE before and after an elimination dietary trial with a vegetable diet. The same vegetable diet trial was performed in healthy control dogs (HC) to evaluate changes in fecal microbiota before and after the trial, and to compare them to FRE dogs.

Dogs with FRE (n = 10) and HC (n = 14) were fed the vegetable diet for 60 days. Fecal samples were collected before and after the dietary trial. Fecal genomic DNA was extracted and used for Illumina sequencing of 16S rRNA genes. Sequence data were analyzed using the QIIME pipeline. The dysbiosis index of the sequence data was calculated using a published mathematical model, and a score >0 was considered as dysbiotic. Statistical significance was set at P < 0.05.

Significantly lower alpha diversity was observed in dogs with FRE‐baseline compared to HC‐baseline and FRE‐after trial. Distinct microbial communities were observed in dogs with FRE‐baseline compared to HC‐baseline (ANOSIM P = 0.001) and dogs with FRE‐after trial (ANOSIM P = 0.032). Microbial communities were still different in FRE‐after trial compared to HC‐baseline (ANOSIM P = 0.001). The calculated dysbiosis index was higher in dogs with FRE‐baseline compared to HC‐baseline (P = 0.022), but no significance difference was observed between FRE‐after trial and HC‐baseline. The fecal microbiota in HC did not show any significant differences before vs. after the vegetable dietary trial.

Results of this study suggest that in FRE dogs, treatment with the vegetable elimination diet led to partial recovery of the fecal microbiota by significantly increasing microbiota richness, which was significantly closer to healthy microbiota after treatment. In contrast, no changes were detected in fecal microbiota of healthy control dogs fed with the same vegetable diet.

Evaluation of Hydromorphone and Dexmedetomidine for Emesis Induction in Cats

M. Nystrom, A. Odunayo, C. Okafor

University of Tennessee, Knoxville, TN, USA

Induction of emesis after ingestion of a toxin can be challenging in cats due to the variable efficacy and adverse effects associated with the use of alpha‐2 agonists as emetics. Hydromorphone may be a more effective alternative with fewer adverse effects. The objective of this study was to compare the efficacy of hydromorphone and dexmedetomidine at inducing emesis in cats.

This was a prospective, blinded, randomized crossover study utilizing 12 healthy purpose‐bred cats. Cats were randomly assigned to receive hydromorphone (0.1 mg/kg [0.045 mg/lb.], subcutaneously) or dexmedetomidine (7 mcg/kg [3.2 mg/lb.], intramuscularly). Following administration, the incidence of emesis, number of emetic events, signs of nausea, temperature, heart rate, respiratory rate and sedation score were recorded.

Emesis was successful in 9 of 12 (75%) cats when treated with hydromorphone and in 7 of 12 (58%) cats when treated with dexmedetomidine (P = 0.67). Dexmedetomidine was more likely to cause sedation than hydromorphone (P < 0.001). Heart rate in cats were significantly decreased at 1 and 2 hours post‐hydromorphone (P = 0.003, 0.014 respectively) and at 1, 2, 3, 5, 6 hours post‐dexmedetomidine (P = 0.001, 0.003, 0.038, 0.013, 0.001 respectively). Cats were more likely to develop an increase in body temperature with hydromorphone administration.

Results of this study indicate that hydromorphone is an effective alternative to dexmedetomidine for the induction of emesis in cats. Sedation and decrease in heart rate were observed less with hydromorphone administration. Further investigation into the most adequate dose of hydromorphone for optimizing emesis is warranted.

Identification of Jejunal Lesions in Dogs Using Capsule Endoscopy

J.S. Pomrantz, J.A. Solomon

Infiniti Medical, Menlo Park, CA, USA

Imaging the jejunum in dogs is problematic. Ultrasound frequently misses mucosal lesions and traditional endoscopy does not typically reach the jejunum. ALICAM is a capsule endoscopy system that images the entire bowel. The purpose of this study was to retrospectively assess the frequency of jejunal lesions in dogs given ALICAM for gastrointestinal signs and/or laboratory abnormalities. A board‐certified internist interpreted each study. Eight of 131 (6.1%) dogs had complete studies with lesions identified in the jejunum that were considered significant. Of the eight dogs with positive studies, 6 were given ALICAM for a regenerative anemia, while 2 had signs of overt gastrointestinal bleeding (melena). Five of 8 dogs had a normal gastrointestinal tract on ultrasound prior to ALICAM administration. One dog had a normal gastroduodenoscopy and a second dog had a normal gastroduodenoscopy and laparoscopy prior to ALICAM. ALICAM findings included ulcerated masses in 2/8 dogs, large erosions or ulcers in 4/8 dogs, and focal fluid, blood or mucosal abnormalities in 2/8 dogs. Following ALICAM, two dogs underwent manual push enteroscopy. In one case, the ulcerative lesions seen on ALICAM were not found. In the second case, the suspected lesion was identified and resected. A third dog had an exploratory laparotomy with subsequent identification and resection of the mass seen with ALICAM. ALICAM is useful in imaging the jejunum and should be considered an important diagnostic step in the work up of gastrointestinal disease, especially in cases of unexplained anemia where traditional endoscopy or ultrasound failed to identify a cause.

Platelet Function and Endoscopic Changes after Clopidogrel, Aspirin, Prednisone, or Combination Therapy in Dogs

J. Whittemore1, A. Mooney1, D. Mawby1, J. Thomason2

1University of Tennessee, Knoxville, Knoxville, TN, USA, 2Mississippi State University College of Veterinary Medicine, MSU, MS, USA

The purpose of this study was to determine the effects of sustained therapy with antiplatelet drugs and glucocorticoids on platelet function tests and the upper gastrointestinal tract.

Fifteen healthy dogs were randomized to 1 of 5 groups: aspirin 2 mg/kg/d (A), clopidogrel 2–3 mg/kg/d (C), prednisone 2 mg/kg/d (P) or combined prednisone with aspirin (PA) or prednisone with clopidogrel (PC) therapy PO for 28 days. Platelet count, PFA‐100® analysis (collagen/epinephrine or collagen/ADP as appropriate), and endoscopy were performed at baseline, 14 days (platelet tests only), and 28 days. The presence of ≥15 hemorrhages, ≥5 punctate erosions, ≥1 invasive erosion, or ≥1 ulcer was considered clinically‐significant bleeding. Dogs were categorized as antiplatelet responders if closure times were >300 sec (aspirin) or increased >30% compared to baseline (clopidogrel). Gastrointestinal bleeding and antiplatelet responder status were compared among groups by Fisher's exact test; P < 0.05 was considered significant.

One PA dog was removed after 14 days for severe weight loss; invasive gastric erosions were identified via capsule endoscopy. Endoscopic lesions were present at day 28 in 0/3 C dogs, 1/3 A dogs (>20 hemorrhages, 4 hemorrhagic tracts), 1/3 P dogs (≥12 invasive erosions), 2/2 remaining PA dogs (≥10 invasive erosions, each), and 2/3 PC dogs (1: 8 invasive erosions, diffuse hemorrhage; 1: 1 deep ulcer). All platelet counts were normal. There were 3/3 and 3/3 C dogs, 2/3 and 1/3 A dogs, 3/3 and 1/3 PC dogs, and 0/3 and 0/2 PA dogs classified as responders based on PFA closure times on days 14 and 28, respectively. Frequency of bleeding did not differ significantly among groups. Responder status differed significantly among groups at day 14 (P = 0.03) but not 28 (P = 0.6)

Clinically significant gastrointestinal bleeding did not occur with sustained clopidogrel therapy but occurred in all other treatment groups. Efficacy of antiplatelet therapy decreased over time, particularly in dogs receiving concurrent steroid therapy. Further evaluation in a larger population of dogs is warranted.

Prevalence of Hypoadrenocorticism in Dogs with Chronic Gastrointestinal Signs

S. Unterer1, C. Schiessl1, I.A. Burgener2, B. Kohn3, J. Lechner4, R. Neiger5, T. Rieker6, A. Wehner1, S.S. Schmitz5

1Clinic of Small Animal Medicine, LMU Munich, Germany, Munich, Bayern, Germany, 2Clinic for Small Animals and Infectious Diseases, VetMedUni Vienna, Austria, Vienna, Wien, Austria3Clinic of Small Animals, Faculty of Veterinary Medicine, Freie Universität Berlin, Berlin, Berlin, Germany, 4Head of Tierärztliche Klinik Nürnberg Hafen, Nürnberg, Bayern, Germany, 5Small Animal Hospital Internal Medicine, Justus Liebig University Gießen, Germany, Gießen, Hessen, Germany, 6AniCura Kleintierspezialisten Ravensburg, Germany, Ravensburg, Baden‐Wurttemberg, Germany

Dogs with hypoadrenocorticism (HA) frequently show gastrointestinal signs (GIS). However, the proportion of dogs presented for chronic GIS as their main complaint ultimately suffering from HA is unknown. The aims of this study were to determine the prevalence of HA in dogs with chronic GIS and to identify clinical and laboratory parameters that help to diagnose HA in these dogs.

A standardized work‐up was performed in all dogs with GIS >3 weeks duration presented to one of the specialists of the German gastroenterology working group between November 2014 and December 2015. This included a basal serum cortisol measurement, and an ACTH stimulation test (ACTHST) if basal serum cortisol was < 3 μg/dL.

151 dogs were prospectively included. Basal serum cortisol concentration was < 3 μg/dL in 80/151 (53%) and < 2 μg/dL in 42/151 (28%) dogs. In 6/151 dogs HA was diagnosed based on ACTHST (stimulated serum cortisol concentration < 2 μg/dL), representing a prevalence of 4%. No hyperkalemia/ hyponatremia was observed in any dog with HA. In 5/6 HA dogs, melena or hematochezia was noted. No historical information or laboratory parameter was able to separate HA from other disorders causing chronic GIS.

In conclusion, the prevalence of HA amongst dogs with chronic GIS is considerable. No differentiation between HA and other causes leading to chronic GIS is possible based on history, physical examination or baseline blood work. Determination of basal serum cortisol concentration should be performed as a screening test for HA, followed by an ACTHST in suspicious cases.

Oral Cobalamin Supplementation in Dogs with Exocrine Pancreatic Insufficiency

L. Toresson1, J.M. Steiner2, J.S. Suchodolski3, T. Spillmann4

1Evidensia Specialist Animal Hospital, Helsingborg/, Sweden/University of Helsinki, Finland, Helsingborg, Skane Lan, Sweden, 2Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA, 3Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA, 4Faculty of Veterinary Medicine, Department of Equine and Small Animal Medicine, Helsinki, N/A, Finland

Exocrine pancreatic insufficiency (EPI) is associated with decreased production of intrinsic factor (IF) in dogs and cats. Since IF is required for cobalamin (CBL) absorption in the ileum, CBL deficiency is a prevalent sequel to EPI in dogs and is associated with a negative prognosis.

Current supplementation protocols for CBL in dogs call for parenteral (PE) supplementation. In humans, several studies have reported equal efficacy of PE and oral (PO) supplementation of CBL. Recently, this group has reported successful PO CBL supplementation in dogs with CE and hypocobalaminemia in a retrospective observational study and in a prospective study comparing oral and parenteral CBL supplementation. The purpose of this retrospective study was to evaluate the effect on PO CBL supplementation in dogs with EPI and subnormal or low‐normal serum CBL concentrations. A computerized database search for dogs treated at Evidensia Specialist Animal Hospital, Helsingborg, Sweden during 2010–2016 was performed. Inclusion criteria were dogs diagnosed with EPI based on subnormal serum canine trypsin‐like immunoreactivity concentration (cTLI; reference range 5–35 μg/L), an initial serum cobalamin concentration ≤ 350 ng/L (reference interval: 234–811 ng/L), and daily treatment with CBL tablets (cyanocobalamin 1 mg/tablet; dogs < 20 kg ¼ tablet/10 kg, dogs ≥ 20 kg 1 tablet). Exclusion criteria were PE CBL supplementation in parallel with PO or failure to comply with the treatment protocol.

Ten dogs aged 1.7–9.2 (median 4.0) years matched the inclusion criteria. Serum CBL for follow‐up was analyzed 28–199 (median 44) days after start of supplementation using an automated chemiluminescence immunoassay (Immulite 2000, Siemens for 8/10 dogs and ADVIA Centaur, Siemens for 2/10 dogs). The most common breeds included were German Shepherds (n = 3) and mixed breed dogs (n = 2). Median serum cTLI concentration was 1.6 μg/L (range < 1 to 3 μg/L).

At inclusion, the median (range) serum CBL concentration was 204 ng/L (150–350 ng/L). It increased significantly to 1,113 ng/L (794–2,385 ng/L) after supplementation. This difference was statistically significant (P = 0.002; Wilcoxon matched‐pairs signed rank test). Due to the retrospective nature of the study, metabolic markers of CBL deficiency could not be analyzed. Despite this limitation, our results suggest that oral CBL supplementation appears effective in treating dogs with EPI and subnormal or low‐normal serum cobalamin concentrations. Since, according to the manufacturer, only traces of IF should be present in the PEz, this finding suggests that dogs, as has been demonstrated in humans, may have an IF‐independent pathway for CBL absorption. Whether such an alternative pathway does exist in dogs requires further studies. Additionally, further studies comparing cellular cobalamin status after PO or PE supplementation in dogs with EPI are warranted.

Development of a Canine Intestinal Organoid Model

A. Meneses1, B. Spee1, L.A. Oosterhoff1, M. Sakai1, K. Schneeberger1, L.C. Penning1, I.A. Burgener2

1Faculty of Veterinary Medicine, Utrecht University, Utrecht, Utrecht, Netherlands, 2Clinic for Small Animals and Infectious Diseases, VetMedUni Vienna, Austria, Vienna, Wien, Austria

Three‐dimensional organ‐buds, or organoids, are a novel and organ mimicking in vitro model based on LGR5‐positive stem cells that show realistic micro‐anatomy. Recently, intestinal organoid cultures from various species including mice and humans have been described. However, canine intestinal crypt‐derived organoids are not yet described. Since organoids can be infinitely expanded in culture, remain genetically stable, and can be differentiated to all cell types of the intestine, they represent an ideal in vitro model for intestinal diseases such as inflammatory bowel disease, virus infections, and drug screening. The aim of this study was to establish small intestinal and colon‐based organoid culture systems from dogs, followed by molecular and cellular characterization.

Fresh cadaveric samples were obtained from duodenum, jejunum and colon of seven individual dogs, followed by crypt isolation. Organoids were propagated in 3D‐culture using RPSO1‐based expansion media. Differentiation was achieved by removing proliferation inducing components. We established a crypt isolation protocol and kept organoid lines from three areas of the intestine (duodenum, jejunum and colon) from seven dogs in culture during at least 12 passages. The gene‐expression levels of stem cell markers (LGR5, CD133, ASCL2 and OLFM4) were stable in expansion media. After differentiation, expression of the goblet cell marker MUC2 and Paneth cell marker NEUROG3 were increased, whereas expression levels of stem cell marker LGR5 was decreased.

This novel robust in vitro model of small intestinal and colonic organoids can be applied as a model to study canine intestinal diseases in the future.

Administration of Enterococcus faecium NCIMB 10415 to Normal Dogs Does Not Change Their Fecal Microbiota

L.A. Matthewman1, J.S. Suchodolski2, D. Werling3

1Royal Veterinary College, Hatfield, UK, 2Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA, 3Department of Pathology and Pathogen Biology, Royal Veterinary College, London, UK, Hertfordshire, UK

Claims for probiotics include that they colonise the gastrointestinal tract and support, stabilise or maintain the balance of a healthy microbiota. There is evidence of the effect of administration of Enterococcus faecium SF68 to unhealthy dogs but no information yet on any effect, particularly on the microbiota, of healthy dogs.

Twelve healthy dogs with no history of gastrointestinal disease and normal faecal examinations, received a 2 week course of a preparation containing 2 x 109 CFU of Enterococcus faecium NCIMB 10415 (SF68) per dose, taken once daily. Normally voided faeces were collected before, during, at the end of treatment and 4 days after the end of treatment. The DNA extracted from the faecal samples, was used to detect and quantify the probiotic bacteria by qPCR, and to analyse the faecal microbiota by Illumina sequencing.

The probiotic bacteria were detected in the faeces whilst the dog received the product but were not detectable 4 days after treatment stopped. There was no effect of the probiotic on the microbiota of the faeces of the 12 normal dogs. Principle co‐ordinate analysis showed the microbiota of the faeces to be more similar to the individual dog, than to the treatment and there was no change in species richness or diversity (P > 0.05). There was a small insignificant increase in Lactobacillus spp. (P > 0.05) and there was no effect on the amounts of E. coli or other Enterobacteriaceae. A cladogram showed that the only difference between all the faecal samples of these dogs taken over the trial period was a 100‐fold increase in the amount of E. faecium on day 14 (P < 0.01). Our data clearly show that this probiotic is a transient coloniser in the gastrointestinal tract of healthy dogs and does not appear to affect the faecal microbiota after 14 days of administration.

Comparison of Maropitant to Metoclopramide Prior to Orogastric Administration of Polyethylene Glycol Solution in Dogs

D.C. Grant, M.S. Leib

Virginia Maryland College of Veterinary Medicine, Blacksburg, VA, USA

We hypothesized maropitant is superior to metoclopramide in preventing vomiting and regurgitation when administered prior to orogastric tube administration of polyethylene glycol solution (PEG).

Dogs (n = 46) with naturally occurring diseases for which PEG was administered for bowel cleansing prior to colonoscopy were prospectively enrolled. Dogs were randomly assigned to be administered maropitant or metoclopramide subcutaneously 30 minutes prior to orogastric tube administration of 66 ml/kg of PEG by or under supervision of the blinded authors. The PEG dose was repeated 2 hours later, but the anti‐emetic was not. The incidences of vomiting or regurgitation were 25.8% and 44.9%, respectively. Regurgitation occurred more often in dogs with lower body weight (15.5 kg vs. 24.3 kg, P = 0.021), in dogs with more rapid administration of PEG (Dose 1: 73.7 seconds vs. 112 seconds, P = 0.005; dose 2: 71.6 seconds vs. 111.8 seconds, P = 0.015.), and in female dogs (13/21 vs. 7/25, P = 0.021). When author 2 inserted the orogastric tube, vomiting occurred less often than when Author 1 did so (5/34 vs. 5/9, P = 0.008). There was no difference in incidence of vomiting (Dose 1: 16.7% vs. 23.8%, P = 0.713; Dose 2: 12.5% vs. 21.1%, P = 0.681) or regurgitation (Dose 1: 25.0% vs. 52.4%, P = 0.059; Dose 2: 12.5% vs. 31.6%, P = 0.153) in dogs that received maropitant or metoclopramide, respectively.

Vomiting and regurgitation occur commonly following PEG administration in dogs with naturally occurring large intestinal diseases. Maropitant was not superior in preventing these. PEG administration technique may be important.

The Role of Pore‐Forming Toxins in Dogs with Acute Hemorrhagic Diarrhea Syndrome

K. Busch1, N. Sindern1, K. Hartmann1, C. Leutenegger2, J. Prescott3, L. Proksch1, R. Müller1, S. Unterer1

1Clinic of Small Animal Medicine, LMU Munich, Germany, Munich, Bayern, Germany, 2IDEXX Laboratories, Inc., West Sacramento, CA, USA, 3Department of Pathobiology, University of Guelph, Guelph, ON, Canada

Recently, novel pore‐forming toxins designated as NetE and NetF were identified in a Clostridium (C.) perfringens type A strain isolated from a dog with acute hemorrhagic diarrhea. Pore‐forming toxins, which have cytotoxic activity, could be responsible for the necrotizing mucosal lesions typically present in dogs with acute hemorrhagic diarrhea syndrome (AHDS). Thus, this study aimed to determine the prevalence of C. perfringens encoding netE and netF in feces of dogs with AHDS and to evaluate any association between severity of clinical signs and the presence of these toxin genes.

Fecal samples of 54 dogs with AHDS, 54 dogs with canine parvovirus (CPV) infection and 66 healthy dogs were tested by real‐time PCR for netE and netFgenes. Severity of clinical signs, time to recovery, mortality rate, and selected laboratory parameters were compared between toxin gene‐positive and ‐negative dogs with AHDS.

There was a significant difference between the three groups in the prevalence of the pore‐forming toxins netE and netF (dogs with AHDS: 26/54 (48.1%); dogs with CPV‐infection: 0/54; (0%); healthy dogs: 8/66 (12.1%)) (P < 0.01). In dogs with AHDS, no significant difference was detected in any parameter evaluated between netE‐ and netF‐positive and netE‐ and netF ‐negative dogs.

The results suggest that overgrowth of C. perfringens type A strains encoding for netE and netF is not a general sequel of acute hemorrhagic diarrhea, since these genes were not detected in dogs with CPV infection. A primary pathogenic role of these toxins in AHDS is likely.

Serum 3‐Bromotyrosine Concentrations in Dogs with Chronic Enteropathy

P. Sattasathuchana1, K. Allenspach2, R. Lopes3, J.A. Lidbury4, J.S. Suchodolski4, J.M. Steiner5

1Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, Texas, USA/Department of Companion Animal Clinical Sciences, Faculty of Veterinary Medicine, Kasetsart University, Bangkok, Thailand, Donmaung, Krung Thep, Thailand2Iowa State University, College of Veterinary Medicine, Ames, IA, USA, 3Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College station, TX, USA, 4Gastrointestinal Laboratory, Texas A&M University, College station, TX, USA, 5Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA

Peripheral eosinophilia has limitations as a marker of eosinophil‐mediated disease. The measurement of 3‐bromotyrosine (3‐BrY) in serum has been reported to serve as a monitoring tool for eosinophil‐mediated diseases in humans. In a previous study, we showed serum 3‐BrY concentrations to be increased in dogs with eosinophilic gastroenteritis as well as lymphocytic‐plasmacytic enteritis. However, serum 3‐BrY concentrations have not previously been evaluated more specifically in dogs with food‐responsive diarrhea (FRD) and steroid‐responsive diarrhea (SRD). Therefore, the objectives of this study were to compare peripheral eosinophil counts in dogs with FRD, SRD, and healthy control dogs and to evaluate and compare serum 3‐BrY concentrations in dogs with FRD, SRD, and healthy control dogs.

The association of peripheral eosinophilia with group (FRD [n = 38], SRD [n = 14], and healthy control dogs [n = 46]) was determined using Fisher's exact test. Serum 3‐BrY concentrations were measured using electron impact ionization gas chromatography/mass spectrometry. Serum 3‐BrY concentrations in dogs with FRD, dogs with SRD, and healthy control dogs were compared using the Kruskal‐Wallis test. Post‐test analysis was used to determine the differences of serum 3‐BrY concentrations among groups. Significance was set at P < 0.05.

There was no significant association between peripheral eosinophilia and the group of dogs (P = 0.0692). The median (min‐max) serum 3‐BrY concentrations in dogs with FRD, SRD, or healthy control dogs were 0.99 (0–8.82) μmol/L, 3.27 (0.9–26.23) μmol/L, and 0.62 (0.62–1.79) μmol/L, respectively. Serum 3‐BrY concentrations in dogs with SRD were significantly higher than those in dogs with FRD (P = 0.0070) or in healthy control dogs (P < 0.0001). Serum 3‐BrY concentrations in dogs with FRD were also significantly higher than those in healthy control dogs (P = 0.0247).

In conclusion, serum 3‐BrY concentrations were higher in dogs with SRD or FRD than in healthy control dogs. Also, dogs with SRD had higher serum 3‐BrY concentrations than dogs with FRD. This suggests that eosinophil‐meditated inflammation plays a role in the pathogenesis of these disorders, especially SRD. Further research is needed to determine whether the measurement of 3‐BrY is useful to sub‐classify dogs with chronic enteropathy as having FRD or SRD.

Eosinophil Peroxidase Monoclonal Antibody Labels Intact and Degranulated Eosinophils in the Gastrointestinal Tract

R.J. Washabau1, I. Bastan2, A. Rendahl3, D. Seelig2, M. Day4, E. Hall5, S.P. Rao3, S. Rao3

1University of Minnesota, West Chester, PA, USA, 2University of Minnesota, College of Veterinary Medicine, St. Paul, MN, USA, 3University of Minnesota, St. Paul, MN, USA, 4University of Bristol, Langford, UK, 5University of Bristol, Bristol, UK

We have previously shown that a monoclonal antibody against eosinophil peroxidase (EPX) detects intact and degranulated eosinophils in canine jejunal mucosa (AJVR, January 2017). The aim of the present study was to compare EPX labelling and haematoxylin and eosin (H & E) staining of intact and degranulated eosinophils in endoscopically‐derived samples of canine duodenal mucosa. Our objective was to determine whether EPX could improve the detection of intact and degranulated eosinophils in the duodenal mucosa of dogs with inflammatory bowel disease (IBD).

Tissue samples were obtained from the histopathology archives of the University of Bristol, School of Veterinary Sciences, and dogs were included in the study if they met WSAVA standards for IBD (i.e., clinical signs, upper G.I. endoscopic imaging and biopsy, and histopathology). Samples were divided into three groups as dogs affected by: i) lymphocytic‐plasmacytic enteritis (LPE), ii) eosinophilic enteritis (EE), and iii) a control group based on WSAVA criteria. Serial sections of paraffin wax‐embedded tissues were immunolabelled with EPX mAb or stained by H&E. Intact and degranulated eosinophils in consecutive microscopic fields (x400 magnification) of the upper (villus tips) and lower (between the muscularis mucosae and crypts) regions of duodenal mucosa were counted manually. Inflammation was classified as mild, moderate, or severe by one blinded investigator (MD), and eosinophil counts were carried out by two investigators blinded to the identity of the cases (IB, DS).

EPX mAb readily detected degranulated and intact eosinophils in the duodenal mucosa and there was good agreement in eosinophil quantification between the two observers in both EPXmAb and H&E stained sections. While H&E staining readily detected intact eosinophils in the duodenum, it was ineffective in detecting degranulated eosinophils. Further, across all study groups, EPX‐mAbIHC (immunohistochemistry) detected three times more eosinophils compared to H&E‐stained slides. In terms of disease severity, degranulated and intact eosinophil numbers were up to two‐fold greater in moderate vs. mild disease.

We conclude that H&E staining not only fails to detect degranulated eosinophils effectively but grossly underestimates eosinophil counts in the duodenal mucosa, while EPX IHC detects both intact and degranulated eosinophils. Results of the present study suggest that EPX IHC may serve as a potential diagnostic method for identifying IBD, monitoring disease activity, and evaluating the success of treatment in dogs. It may further constitute a first step toward understanding the role of eosinophil degranulation in the pathogenesis of IBD in dogs.

Pilot Evaluation of the Effect of Omeprazole on the Feline Fecal Microbiome and Metabolome

S.M. Schmid1, J.S. Suchodolski2, E.N. Gould1, J. Price1, M. Katherine Tolbert3

1University of Tennessee, Knoxville, TN, USA, 2Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA, 3University of Tennessee College of Veterinary Medicine, Knoxville, TN, USA

Proton pump inhibitor administration in humans is associated with an increased risk of enteric, particularly Clostridium difficile, infections. Omeprazole also alters the canine microbiome; however, no studies have explored its effect on the feline fecal microbiome and metabolome. We hypothesized that prolonged oral omeprazole administration would alter the fecal microbiome and metabolome in cats.

A within‐subjects, before and after, study was performed whereby all cats received 60 days of consecutive treatment with placebo (250 mg lactose PO q12 h), followed by 5 mg (0.83–1.6 mg/kg PO q12 h) omeprazole. On days 0, 30, and 60 of placebo and omeprazole therapy, free catch fecal samples were collected from each cat. The fecal microbiome was characterized using 16S rRNA gene sequencing. qPCR assays were also used to specifically target Clostridium perfringens, Clostridium hiranonis, and Bifidobacterium spp. The metabolome was characterized by untargeted mass spectrometry‐based methods. Two‐way repeated‐measures ANOVA tests were utilized to determine the effects of phase of treatment, time, and the interaction of phase and time.

No global changes in the microbiome and metabolome were observed based on multivariant analysis (ANOSIM, P > 0.05). However, some bacterial genera were altered following omeprazole administration including an increased relative abundance of Streptococcus, Lactobacillus, Clostridium, and Faecalibacterium spp. and decreased Bifidobacterium spp. in all cats. The results of this pilot study suggest that omeprazole might have little impact on the feline fecal microbiome and metabolome when administered for 60 days or less. Larger studies are needed to confirm these results.

Serum Biomarkers of Oxidative Stress as a Marker of Inflammatory Bowel Disease in Dogs

C. Rubio1, S. Martinez‐Subiela1, J. Hernandez‐Ruiz1, A. Tvarijonaviciute2, J. Ceron1, K. Allenspach3

1University of Murcia, Spain, Espinardo, Murcia, Murcia, Spain, 2University Autonoma of Barcelona, Spain, Barcelona, Catalonia, Spain, 3Iowa State University, College of Veterinary Medicine, Ames, IA, USA

There is evidence that oxidative stress plays an important role in the pathogenesis of IBD in human patients particularly in the initiation and perpetuation of inflammation and in subsequent tissue damage. Oxidative stress occurs when there is a marked imbalance between the production of Reactive Oxygen Species (ROS) and their removal by antioxidants. Recent publications suggest that oxidative stress could also represent a significant factor in the pathogenesis of IBD in dogs. The objective of this study was to compare a panel of various serum biomarkers evaluating both the antioxidant response and the oxidative damage in dogs with inflammatory bowel disease (IBD) and healthy control dogs. Eighteen dogs with IBD and 20 healthy dogs were enrolled in the study. Trolox equivalent antioxidant capacity (TEAC), cupric reducing antioxidant capacity (CUPRAC), ferric reducing ability of the plasma (FRAP), total thiol concentrations, and paraoxonase 1 (PON1) activity were evaluated in serum to determine antioxidant response. To evaluate oxidative status, ferrous oxidation‐xylenol orange (FOX), thiobarbituric acid reactive substances (TBARS) and reactive oxygen species production (ROS) concentrations in serum were determined. Mean concentrations of all antioxidant biomarkers analyzed, with exception of FRAP, were significantly lower (P < 0.0001) in serum of dogs with IBD than healthy dogs. The oxidant markers studied were significantly higher (P < 0.0001) in serum of dogs with IBD than in healthy dogs. These findings support the hypothesis that oxidative stress plays an important role in the pathogenesis of IBD in dogs.

Analytical Validation of an Enzymatic Assay for Measurement of Fecal D‐ and L‐Lactate in Dogs

A.B. Blake1, Y. Minamoto2, J.A. Lidbury2, J.M. Steiner1, J.S. Suchodolski2

1Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA, 2Gastrointestinal laboratory, Texas A&M University, College Station, TX, USA

L‐lactate is generated from pyruvate via lactate dehydrogenase during normal cell metabolism. Lactobacillales (lactic acid bacteria) in the intestinal tract can synthesize both D‐ and L‐lactate. Currently, serum lactate concentration is used as a marker of tissue hypoperfusion and hypoxia. Cases of D‐lactic acidemia have been reported in dogs and humans with intestinal disease. Fecal lactate has also been shown to be increased in humans with short bowel syndrome and calves with diarrhea. Fecal lactate concentrations might serve as a marker of the microbial activity in the intestine. Therefore, the aim of this study was to analytically validate an enzymatic method for the measurement of both D‐ and L‐lactate in canine feces.

A commercial kit was adapted for use in feces (D‐/L‐lactic acid kit, R‐Biopharm). Surplus homogenized canine fecal samples from seven dogs were used for method validation. Validation variables included lower limit of detection, dilutional parallelism, spiking recovery, and inter‐ and intra‐assay variability.

The lower limit of detection for D‐ and L‐lactate concentrations were 0.007 and 0.003 mM, respectively. Observed‐to‐expected ratios for dilutional parallelism ranged from 89% to 117% for D‐lactate and 87% to 112% for L‐lactate. Recovery for spiking sample extracts ranged from 95.9% to 103.4% for D‐lactate, 96.4% to 118.5% for L‐lactate, and 97.8% to 112.9% for total lactate. Average intra‐assay coefficients of variation (%CVs) for D‐, L‐, and total lactate were 5.0%, 5.1%, and 4.1%, respectively. Average inter‐assay %CVs for D‐, L‐, and total lactate were 23.5%, 20.1%, and 19.4%, respectively.

This assay was shown to be linear, precise, accurate, and reproducible. Future studies are underway to evaluate the utility of fecal lactate measurement in dogs with gastrointestinal diseases.

Dogs with Exocrine Pancreatic Insufficiency have Dysbiosis and Abnormal Fecal Lactate and Bile Acid Concentrations

A.B. Blake1, B.C. Guard1, J.B. Honneffer1, F.G. Kumro1, O.C. Kennedy2, J.A. Lidbury3, J.M. Steiner1, J.S. Suchodolski3

1Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA, 2Epi4Dogs Foundation, Inc., Farmville, VA, Farmville, VA, USA, 3Gastrointestinal Laboratory, Texas A&M University, College station, TX, USA

It has been reported that dogs with exocrine pancreatic insufficiency (EPI) commonly have intestinal dysbiosis. However, the effects of EPI on microbial metabolism are poorly understood. The aim of this study was to compare fecal dysbiosis as well as fecal lactate and bile acid concentrations between dogs with EPI and healthy control dogs.

Fecal samples were collected from eleven dogs with EPI that had not received antibiotics for at least 3 weeks and had been on enzyme supplementation for 0.5–10 years (median 5 years). Fecal samples from healthy dogs (n = 18), collected for three consecutive days and pooled, served as control samples. DNA was extracted and analyzed by qPCR for selected bacterial groups and data expressed as Dysbiosis Index (as previously reported). Fecal lactate was measured by enzymatic methods (D‐/L‐lactic acid kit, R‐Biopharm) and bile acids were quantified with gas chromatography/mass spectrometry from lyophilized feces. The Mann‐Whitney U test was used to compare the Dysbiosis Index and fecal lactate and bile acid concentrations between dogs with EPI and healthy control dogs. Correlations were assessed using Spearman's correlation coefficient and significance was set at P < 0.05.

Dogs with EPI had a higher Dysbiosis Index (median [min‐max]: +3.08 [−7.29 to +7.62]) than healthy control dogs (−3.81 [−7.57 to +3.32]; P = 0.0232). Total fecal lactate concentrations were increased in dogs with EPI (3.44 mM [0.71–158.30 mM]) compared to healthy control dogs (1.14 mM [0.54–6.64 mM]; P = 0.0037). The proportion of secondary bile acid was lower in dogs with EPI (70% [6–96%]) compared to healthy control dogs (93% [12–97%]; P = 0.0431). There was no correlation between any measurements and duration of enzyme therapy.

In conclusion, this study identified differences in the fecal microbiota as well as fecal lactate and bile acid concentrations between dogs with EPI and healthy control dogs.

Analytical Validation of a Radioimmunoassay for the Measurement of Soluble CD25 Concentrations in Canine Serum

A. Buono1, J.A. Lidbury2, J.S. Suchodolski2, J.M. Steiner3

1Texas A&M University, College Station, TX, USA, 2Gastrointestinal Laboratory, Texas A&M University, College station, TX, USA, 3Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA

CD25, also known as interleukin‐2 receptor α (IL‐2Rα), is a 55 kDa transmembrane cytokine receptor, primarily expressed by activated T cells, regulatory T cells (Tregs), and some types of cancer cells. In humans, increased expression of membrane‐bound CD25 has been reported in patients with autoimmune or inflammatory diseases, during transplant rejection, and some types of neoplasia. A soluble form of CD25 (sCD25) is generated by proteolytic cleavage and has a molecular mass of 45 kDa. Elevated serum concentrations of sCD25 have been reported in human patients with the aforementioned conditions. A positive correlation has been reported between the membrane expression of CD25 and the serum concentration of sCD25. In vitro, activated T cells and Tregs have been shown to shed sCD25, therefore high concentrations of serum sCD25 are thought to be reflective of a sustained immune activation. Moreover, sCD25 has been used as a biomarker to help characterize disease progression, prognosis, and treatment in humans. Thus, the aim of this study was to establish and analytically validate a radioimmunoassay for the measurement of sCD25 in canine serum.

Canine specific CD25 protein and antibodies were commercially available and used to establish this assay. A competitive, liquid‐phase 125I radioimmunoassay (RIA) was set‐up and analytically validated by determination of lower limit of detection (LLOD), dilutional parallelism, spiking recovery, and intra‐ and inter‐assay variability. Surplus canine serum samples were pooled and used for validation. In addition, due to the inability to find canine serum samples with naturally‐occurring high concentrations of sCD25, three samples were spiked with different concentrations of pure canine CD25 and treated as unknown samples to test the assay across the entire working range. Recovery was evaluated by spiking three samples with five different concentrations of CD25 (5, 10, 25, 35, and 40 ng/mL).

LLOD was calculated to be 0.49 ng/mL. Mean (± SD) observed‐to‐expected ratio (O/E) for serial dilutions was 95.8 ± 14.4% for 6 samples, and mean (± SD) O/E for spiking recovery was 93.2 ± 4.2% (mean ± SD) for 3 samples. Coefficients of variation for intra‐assay variability for 8 serum samples were ≤ 12.5% (mean ± SD: 7.5 ± 4.2%), whereas inter‐assay coefficients of variation for 6 samples were ≤ 15.7% (mean ± SD: 11 ± 4.4%), respectively.

In conclusion, the RIA described here is linear, accurate, precise, and reproducible for the measurement of soluble CD25 in canine serum. Further studies are warranted to evaluate the clinical utility of measuring serum sCD25 concentrations for evaluation of dogs with immune‐mediated and inflammatory conditions.

Fecal Microbial Transplantation Decreases the Dysbiosis Index in Dogs Presenting with Chronic Diarrhea

J. Chaitman1, B.C. Guard2, F. Sarwar3, J.A. Lidbury3, J.M. Steiner2, J.S. Suchodolski3

1Veterinary Internal Medicine and Allergy Specialists, New York, NY, USA, 2Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA, 3Gastrointestinal Laboratory/Texas A&M University, College Station, TX, USA

The intestinal microbiome is important for maintaining gastrointestinal health and alterations may contribute to gastrointestinal disease. It is currently unclear whether fecal microbial transplantation (FMT) may present an alternative approach to standard therapy for gastrointestinal disease. The purpose of this study was to evaluate the fecal microbiota before and after FMT in patients that presented for chronic diarrhea.

Fresh or frozen stool from an infection free donor dog on a hypoallergenic diet was administered as an FMT using 5 g/kg of stool diluted with 60 mL saline administered per rectum via a red rubber catheter on an out‐patient basis to unsedated patients. Stool samples were collected from 7 patients with gastrointestinal disease immediately before and one week after administration of FMT. Microbial dysbiosis was determined using a previously described Dysbiosis Index which targets several key bacteria typically associated with maintaining health including: Universal 16S rRNA, Faecalibacterium, Turicibacter, Streptococcus, E. coli, Blautia, Fusobacterium, and Clostridium hiranonis. Data was analyzed by a Wilcoxon signed rank test for matched pairs. Significance was set at P < 0.05.

The Dysbiosis Index significantly decreased in dogs post FMT (median [min‐max]: −5.1 [−7.1 to −0.8]) compared to dogs before FMT (median [min‐max]: −1.3 [−4.8 to +8.1]; P = 0.0156).

This study suggests that FMT alters the abundance of some commonly targeted bacterial groups typically associated with gastrointestinal disease. Further studies are warranted to investigate how FMT alters fecal microbiota in specific enteropathies and its long‐term effect on the microbiome.

Fecal Short Chain Fatty Acid Collection, Storage and Variability in Healthy Dogs

A.‐C. Duchaussoy1, K.L. Boedec2, C. Mansfield3

1UVET Hospital, Faculty of Veterinary and Agricultural Sciences, University of Melbourne, Moorabbin, Victoria, Australia, 2Centre Hospitalier Veterinaire FREGIS, Arcueil, Ile‐de‐France, France, 3UVET Hospital, Faculty of Veterinary and Agricultural Sciences, University of Melbourne, Werribee, Victoria, Australia

Genetic, environmental, defective immune system and dysbiosis have been identified as main factors involved in the pathophysiology of chronic enteropathies. Short Chain Fatty Acids (SCFA) are volatile compounds that are directly correlated to the quantitative and qualitative bacterial population of the gastrointestinal tract. An increasing number of publications assessing gastrointestinal health and inflammation include SCFA measurements in fecal matter. However, there is scant information about the repeatability and reliability of SCFA analysis in dogs and cats. As SCFA are highly volatile and unstable, and our preliminary results in feces from both dogs and cats showed high variability for the same animal and between healthy animals, we investigated repeatability and variability of SCFA measurement.

We started with a prospective study of 4 healthy dogs (normal body condition). Fecal samples were collected upon voiding and divided into aliquots immediately after collection. The aliquots (in duplicate) were stored in ambient temperature or +4C before freezing at −20°C vs. −80°C. We also varied the duration in time between collection and freezing. In all samples, SCFA were measured using a dilution gas chromatography‐mass spectrometry (GC‐MS) assay as previously described. All samples were thawed by batch, and kept on ice until the internal standard solution was added. We focused on the three most prevalent short chain fatty acids identified in dogs, namely propionic acid, butyric acid and valeric acid. Results showed a linear increase in the amount of SCFA over time in between collection and the freezing process when left at room temperature. This was not observed when sampled were stored at +4°C. There was no significant difference in between samples stored at −20°C and −80°C results.

The next component of the study obtained fecal samples from 17 healthy dogs monthly. Samples were frozen immediately after collection at −80°C, and SCFA then measured using GC‐MS assay. Results showed wide variability of results for normal dogs, and also a noticeable degree of individual variability overtime. Potential causes of this variability include variation within a stool sample, as well as the possible presence of undigested debris in the fecal materials.

In conclusion, we showed a high degree of variability within individual dogs as well as with storage methods. This data improves our understanding of SCFA analysis, and the potential interpretation of SCFA in studies of chronic enteropathies in dogs.

Immunohistochemical Distribution of TGR5 Bile Acid Receptors in the Gastrointestinal Tract of Dogs

P.R. Giaretta1, R.R. Rech1, A.K. Blick1, J.M. Steiner2, J.A. Lidbury3, B.C. Guard2, J.S. Suchodolski3

1Texas A&M University, College Station, TX, USA, 2Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA, 3Gastrointestinal Laboratory, Texas A&M University, College station, TX, USA

Beyond their role in digestion, bile acids also act as systemic signaling modulators of several pathways involving glucose and lipid metabolism and play an anti‐inflammatory role through the activation of nuclear and G‐protein coupled receptors (i.e., GPBAR‐1 or TGR5). TGR5 expression has been demonstrated in many gastrointestinal tissues of humans and laboratory animals. The aim of this study was to evaluate the immunohistochemical distribution of TGR5 receptors in the gastrointestinal tract of dogs.

Samples were collected from six adult dogs (3 females and 3 males; median: 4 years, range: 1 to 13 years) that were humanely euthanized due to causes not related to this study. These dogs did not have any clinical signs or histopathological lesions suggesting gastrointestinal disease. Sections of duodenum, jejunum, ileum, colon, cecum, rectum, liver, gallbladder, and pancreas were fixed in 10% formalin, followed by routine histologic processing. For immunohistochemistry, heat‐induced antigen retrieval was performed. Tissues were incubated with a polyclonal TGR5 antibody (directed against amino acids 264 to 330 of human TGR5) that was diluted 1:100 or rabbit IgG as a negative control for 1 hour at room temperature.

In the intestinal mucosa, TGR5 labeling was identified as a granular staining pattern in the membrane of goblet cells, enterocytes, and colonocytes. Enteroendocrine cells in the crypts of the small and large intestine showed granular labeling in the cytoplasm and membrane, as did resident macrophages in the lamina propria and Peyer's patches. Granular and homogeneous labeling was present in the cytoplasm and membrane of ganglion cells in the submucosal and muscularis plexus of the small and large intestine. In the muscularis, both layers of smooth muscle expressed TGR5 in the membrane. In the liver, the membrane of the biliary epithelium and the canalicular surface of the hepatocellular membrane were positive, as were the cytoplasm and membrane of Kupffer cells. The epithelium of the gallbladder showed granular labeling in the membrane and cytoplasm. In the pancreas, membranous labeling was evident in the epithelial cells of the ducts, and a membranous/cytoplasmic pattern was identified in approximately 40% of islet cells. The cytoplasm and membrane of endothelial cells were positive in all the examined tissues.

Our findings show that, based on antigenic expression, TGR5 is ubiquitously distributed in the gastrointestinal tract of dogs, expressed predominantly in the membrane of epithelial cells and both in the membrane and cytoplasm of ganglia, histiocytes, and enteroendocrine cells.

Immunolocalization of TGR5 Bile Acid Receptors in the Colon of Dogs with Chronic Enteropathy

P.R. Giaretta1, R.R. Rech1, A.K. Blick1, J.M. Steiner2, J.A. Lidbury3, B.C. Guard2, Y. Minamoto3, J.S. Suchodolski3

1Texas A&M University, College Station, TX, USA, 2Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College Station, TX, USA, 3Gastrointestinal Laboratory, Texas A&M University, College station, TX, USA

Canine chronic enteropathy (CE) is suspected to have a multifactorial etiology, with an interplay of genetics and environmental factors and an altered gut microbiota and host immune system. Dogs with CE have alterations in the serum and fecal metabolome, including changes in fecal bile acids. Bile acids are important signaling molecules that act through the interaction with their receptors: G protein‐coupled bile acid receptor 1 (TGR5) and farnesoid X receptor (FXR). TGR5 agonists have been shown to suppress TNF‐α production in macrophages from the lamina propria of human patients with Crohn's disease. The aim of this study was to compare the antigenic expression of TGR5 receptors in colonic samples as well as the fecal concentration of bile acids between dogs with CE and control dogs.

The CE group was comprised of seven dogs (3 females and 4 males, median: 7 years, range: 6 months to 10 years) with a history of gastrointestinal signs for more than three weeks, mild to moderate inflammatory infiltration of the colonic mucosa, and no other identified cause for gastrointestinal (GI) inflammation. The control group consisted of six adult dogs (3 females and 3 males; median: 4 years, range: 1 to 13 years) that were humanely euthanized due to causes not related to this study. These dogs did not have any GI signs or histopathologic lesions. None of the dogs in this study received antibiotic or corticosteroid therapy within 3 weeks of colonic and fecal sampling. Formalin‐fixed and paraffin‐embedded sections of colon were used for immunohistochemistry. Heat‐induced antigen retrieval was performed and after blocking, tissues were incubated with a polyclonal TGR5 antibody (directed against amino acids 264–330 of human TGR5) that was diluted 1:100 or rabbit IgG as a negative control for 1 hour at room temperature. Concentrations of fecal bile acids were measured by gas chromatography/mass spectrometry. The Mann‐Whitney U‐test was used to test for statistical significance (P < 0.05).

In both groups, immunolabeling was identified in the membrane of colonocytes, goblet cells, and smooth muscle cells of the muscularis; and in the membrane and cytoplasm of enteroendocrine cells in the crypts, resident macrophages, ganglion cells, and endothelial cells. The infiltrating inflammatory cells in CE dogs were mainly composed of lymphocytes and plasma cells, neither of which expressed TGR5. Immunohistochemical expression of TGR5 in the colon did not differ between control and CE dogs. However, dogs with CE had an increased proportion of primary bile acids in their feces (P = 0.012).

Although dogs with CE showed altered proportions of fecal bile acids, differences in the expression of TGR5 in colonic samples could not be detected by immunohistochemistry. Further investigations using other techniques, including gene arrays, are needed and might contribute to the understanding of bile acid dysmetabolism in dogs with CE.

Enteropathogenic Escherichia coli Is Associated with Diarrheal Mortality and Intestinal Inflammation in Foster‐Age Kittens

V. Watson1, J. Flowers1, S. Strong2, C. DebRoy3, J. Gookin4

1North Carolina State University, Raleigh, NC, USA, 2Wake County Animal Center, Raleigh, NC, USA, 3The Pennsylvania State University E. coli Reference Center, University Park, PA, USA, 4North Carolina State University College of Veterinary Medicine, Raleigh, NC, USA

It is estimated that ˜15% of foster kittens die or are euthanized due to illness before 8‐wks of age. Most reportedly have clinical signs or post‐mortem evidence of gastrointestinal disease. In children, enteropathogenic E.coli (EPEC) is a common cause of diarrhea, is associated with higher hazard of death, and has proven difficult to detect or experimentally model in animals. We previously reported EPEC in a series of kittens that died with diarrhea. The purpose of this study was to determine the prevalence of EPEC infection in kittens and its association with diarrhea, diarrhea‐associated mortality, specific intestinal tract pathology, and possible risk factors.

Live kittens (n = 61) with and without diarrhea and deceased kittens (n = 54) that were euthanized in apparent good health or were euthanized/died due to severe diarrhea were obtained from 2 different shelter facilities. All kittens were ≤ 12‐wks of age and ≤1 kg body weight. Fecal culture for EPEC and extraction of fecal DNA for qPCR amplification of EPEC enterocyte‐attaching‐effacing gene (eae) was performed. Deceased kittens underwent autopsy and sections of the stomach, small intestine, and colon were obtained for histopathology, scoring of lesions, and fluorescence in‐situ hybridization (FISH) to detect enteroadherent E. coli. Medical records were examined for variables of potential association with EPEC infection.

Based on fecal culture the prevalence of EPEC infection was 18% in both live and deceased kittens. Based on qPCR amplification of eae from feces prevalence was 37% in live kittens and 48% in deceased kittens. EPEC was observed in kittens with and without diarrhea. However, kittens that died or were euthanized due to severe diarrhea harbored a significantly greater quantity of live EPEC and eae compared to kittens without diarrhea. Kittens diagnosed with EPEC had significantly greater severity of inflammatory infiltrates in the small intestine and colon and increased epithelial injury in the small intestine. Enteroadherent E. coli was observed by means of FISH in 2 kittens. When live and deceased kitten populations were examined on the basis of EPEC diagnosis, significant associations were observed between EPEC infection and medical history of subcutaneous fluid administration.

These findings identify EPEC as a prevalent infection in kittens and implicate EPEC as a significant primary or contributing cause of intestinal inflammation, dehydration, and diarrhea‐associated mortality in kittens. A high prevalence of EPEC in healthy kittens and children complicates interpretation of EPEC infection in patients with diarrhea. Future efforts to identify mechanisms responsible for susceptibility to EPEC‐induced diarrhea will lead to advanced understanding of EPEC pathogenesis in kittens and children.

Development and Analytical Validation of an Assay for the Quantification of Canine Fecal Bile Acids

B.C. Guard1, M.M. Jonika2, J.B. Honneffer1, J.A. Lidbury2, J.M. Steiner1, J.S. Suchodolski2

1Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA, 2Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA

The gut microbiota is important in maintaining intestinal health. Bile acids are increasingly appreciated to play a role in regulation of gut microbial composition and intestinal health. Bile acids are synthesized from cholesterol, conjugated in the liver, and once secreted into the gastrointestinal tract (GIT), undergo modification by certain members of the intestinal microbiota. Numerous bile acid receptors (e.g., farnesoid X receptor and G protein‐coupled membrane receptor) have been identified along the GIT and are responsible for regulating metabolism and maintaining an anti‐inflammatory environment in the gut. The aim of this study was to develop and analytically validate a gas chromatography/mass spectrometry (GC/MS) assay for the identification and quantification of bile acids in canine feces.

Fecal bile acids (cholic acid [CA], chenodeoxycholic acid [CDCA], lithocholic acid [LCA], deoxycholic acid [DCA], and ursodeoxycholic acid [UDCA]) were measured in their unconjugated form after undergoing butyl esterification for chromatographic separation. A capillary DB‐1 ms Ultra Inert column was used with a 20:1 split sample injection ratio. Validation parameters included the lower and upper limits of quantification (LLOQ and ULOQ, respectively). Additionally, precision of the assay was calculated by assaying 6 aliquots taken from a single fecal sample from 4 dogs on the same run/day followed by calculating intra‐assay coefficients of variation (CV%). Reproducibility of the assay was determined by analyzing 6 aliquots taken from a single fecal sample from 4 dogs on 6 consecutive days followed by calculating inter‐assay variation (CV%).

The LLOQ and ULOQ in μg/mL were as follows for each compound: cholic acid (3.9 and 1000), chenodeoxycholic acid (6.25 and 200), lithocholic acid (1.9 and 500), deoxycholic acid (31.3 and 1000), and ursodeoxycholic acid (0.78 and 50). For intra‐assay variability, the average CV%s were: 6.0, 5.6, 7.1, 7.3, and 8.8% for CA, CDCA, LCA, DCA, and UDCA, respectively. For inter‐assay variability, the average CV%s were: 8.3, 8.0, 4.8, 8.6, and 13.2% for CA, CDCA, LCA, DCA, and UDCA, respectively.

In conclusion, the present assay was found to be both reproducible and precise for the quantification of select bile acids in canine feces.

Longitudinal Characterization of the Fecal Metabolome in Dogs with Inflammatory Bowel Disease

B.C. Guard1, M.M. Jonika2, J.B. Honneffer1, J.A. Lidbury2, J.M. Steiner1, A.E. Jergens3, J.S. Suchodolski2

1Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA, 2Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA, 3Iowa State University, College of Veterinary Medicine, Ames, IA, USA

Canine inflammatory bowel disease (IBD) is a multifactorial disease, the pathogenesis of which includes alterations in gut microbiota and improper activation of the immune system. Recent studies have used untargeted metabolomics of serum and feces to investigate patients with chronic gastrointestinal (GI) disease. However, evidence is lacking about how GI inflammation and ongoing microbial dysbiosis affect metabolites long‐term in patients that undergo immunosuppressive therapy. Therefore, the purpose of this study was to characterize the fecal metabolome in dogs with IBD upon initial diagnosis and after therapy, using an untargeted approach.

Nine dogs that were non‐responsive to dietary or antibiotic therapy and had histologically confirmed intestinal inflammation were enrolled. Fecal samples were collected at baseline, 3 weeks, and 8 weeks. Patients received immunosuppressive therapy after initial diagnosis and clinical signs were scored according to the canine IBD activity index (CIBDAI). Fecal samples were also collected from healthy dogs (n = 13) to serve as controls. Fecal samples were analyzed by an untargeted metabolomics platform using gas chromatography coupled with mass spectrometry. Data were found to be non‐parametric. Therefore, comparisons were made across time points using a Friedman's test for repeated measures. A Dunn's post‐test was used where appropriate. P‐values were adjusted for multiple comparisons using the Benjamini and Hochberg False Discovery Rate and significance was set at P < 0.05.

Multivariate testing consisted of principal component analysis that revealed significant differences between healthy controls and dogs with IBD at baseline, 3 weeks, and 8 weeks after therapy. The median CIBDAI score for patients 8 weeks following initial diagnosis was 0 compared to a median CIBDAI score of 7 at enrollment. Univariate analysis revealed that of the 233 metabolites measured, trans‐4‐hydroxyproline, sinapinic acid, glycine, 2‐methylglyceric acid, stearic acid, heptadecanoic acid, myo‐inositol, and dehydroabietic acid were found to be significantly different among the repeated time points sampled from dogs with IBD (P < 0.05).

In conclusion, distinct changes in metabolic profiles were observed between healthy dogs and dogs with IBD. Despite improvement in clinical activity scores, several metabolites remained altered at 8 weeks follow up.

Longitudinal Assessment of Fecal Sterol and Fatty Acid Concentrations in Dogs with Diarrheal Diseases

J.B. Honneffer1, B.C. Guard1, S. Unterer2, F. Bresciani3, S.A. Wennogle4, J.A. Lidbury5, J.M. Steiner1, A.E. Jergens6, J.S. Suchodolski5

1Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College Station, TX, USA, 2Clinic of Small Animal Medicine, LMU Munich, Germany, Munich, Bayern, Germany, 3Department of Veterinary Medical Sciences, University of Bologna, Ozzano dell'Emilia, Emilia‐Romagna, Italy, 4College of Veterinary Medicine, Colorado State University, Fort Collins, CO, Fort Collins, CO, USA, 5Gastrointestinal Laboratory, Texas A&M University, College station, TX, USA, 6Iowa State University, College of Veterinary Medicine, Ames, IA, USA

Sterols and fatty acids play an essential role as building blocks for structural components, as signaling molecules, and in energy metabolism, yet can also be toxic to cells at increased concentrations. Increased cholesterol and decreased phytosterol concentrations in the feces of dogs with chronic enteropathy have previously been reported. This study aimed to explore concentrations of fecal sterols and fatty acids in dogs exhibiting a wide variety of gastrointestinal disease phenotypes.

Baseline fecal samples were collected from dogs with acute hemorrhagic diarrhea syndrome (AHDS, n = 22), food‐responsive diarrhea (FRD, n = 10), steroid‐responsive diarrhea (SRD, n = 24), and healthy control dogs (n = 82). In a subset of diseased dogs, follow‐up samples were collected 2–3 months after the baseline sample (AHDS, n = 7; FRD, n = 6; SRD, n = 9). Diagnoses of FRD and SRD were based on response to therapy. All dogs with FRD were successfully managed with the same vegetarian diet. Feces were analyzed by gas chromatography/mass spectrometry (GC/MS) using an in‐house assay. At each of the two timepoints, a Kruskal‐Wallis test and the Benjamini‐Hochberg step‐up method to adjust for multiple comparisons were used to identify significantly altered compounds. Dunn's test was used to compare groups, and statistical significance was set at P < 0.05.

At baseline, dogs with AHDS exhibited decreased fecal phytosterol (i.e., β‐sitosterol, campesterol, sitostanol, fucosterol, and stigmasterol, all P < 0.001), increased cholesterol (P < 0.001), and increased fatty acid (i.e., stearic, erucic, gondoic, palmitic, and α‐linolenic, all P < 0.001) concentrations when compared to healthy control dogs. There were no statistically significant differences in any of the compounds between FRD and SRD dogs at baseline. There were several significant differences observed between dogs with SRD and healthy control dogs at baseline: decreased β‐sitosterol, sitostanol, campesterol, and fucosterol (P < 0.001, P < 0.001, P = 0.002, and P = 0.013, respectively), and increased erucic, gondoic, and stearic acids (P = 0.008, P = 0.012, and P = 0.038, respectively). However, at the 2–3 months follow‐up, several sterols were altered between dogs with SRD and FRD: sitostanol, β‐sitosterol, and fucosterol were decreased in SRD (P < 0.001, P = 0.004, and P = 0.033, respectively), and α‐linolenic, erucic, and gondoic acids were also decreased in dogs with SRD (P < 0.001, P = 0.004, and P = 0.004). At the 2–3 months follow‐up campesterol, stigmasterol, cholesterol, stearic acid, and palmitic acid were no longer significantly different between dogs with AHDS and healthy control dogs.

The most profound changes in fecal sterols observed were at baseline between dogs with AHDS and healthy control dogs, but many of these changes normalized rapidly. The observed differences between dogs with SRD and FRD at follow‐up after 2–3 months may be largely due to the effect of diet. Future studies are required to evaluate the utility of these compounds as markers of disease phenotype or response to therapy.

Validation of an Assay for Quantification of Sterol and Fatty Acid Concentrations in Dog Feces

J.B. Honneffer1, B.C. Guard1, J.A. Lidbury2, J.M. Steiner1, J.S. Suchodolski2

1Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA, 2Gastrointestinal Laboratory, Texas A&M University, College station, TX, USA

Cholesterol is the most abundant sterol in mammalian tissues, playing a critical role in cell membrane function and serving as a precursor molecule for hormones, vitamins, and bile acids. Structurally similar to cholesterol, phytosterols have been associated with anti‐inflammatory effects in rodent models of colitis. Previous studies have shown altered fecal sterol concentrations in dogs with chronic enteropathy. Fatty acids are an important source of energy, and along with sterols, they are also critical for intestinal epithelial cell membrane structure and function. The aim of this study was to analytically validate a gas chromatography/mass spectrometry (GC/MS) assay for the measurement of several sterols (i.e., cholesterol, coprostanol, cholestanol, β‐sitosterol, stigmasterol, campesterol, sitostanol, and fucosterol) and fatty acids (i.e., palmitic, oleic, linoleic, α‐linolenic, stearic, gondoic, and erucic acids) in canine fecal samples.

A fecal sample was collected from eight dogs, lyophilized, and used for assessment of intra‐ and inter‐assay variability (%CV). Sample and standard preparation were adapted from published methods, using deuterated analogs of cholesterol, sitostanol, and stearic acid for internal standards. Validation parameters also included lower and upper limits of quantitation (LLOQ and ULOQ, respectively). Quality control procedure for each sample run consisted of blanks, continuing calibration verification standards, and a laboratory control sample of pooled feces.

Intra‐ and inter‐assay variability (median %CV) were 4% and 6% for cholesterol, 5% and 11% for coprostanol, 4% and 9% for cholestanol, 4% and 6% for β‐sitosterol, 4% and 7% for stigmasterol, 4% and 6% for campesterol, 4% and 11% for sitostanol, 6% and 10% for fucosterol, 4% and 7% for palmitic acid, 4% and 9% for oleic acid, 8% and 10% for linoleic acid, 14% and 12% for α‐linolenic acid, 4% and 6% for stearic acid, 5% and 11% for gondoic acid, and 10% and 18% for erucic acid. Limits of quantitation for each compound in μg/mL (LLOQ‐ULOQ) were as follows: palmitic acid (6.25–800), linoleic acid (6.25–800), α‐linolenic acid (12.5–400), oleic acid (12.5–800), stearic acid (6.25–800), gondoic acid (1.56–200), erucic acid (1.56–200), cholesterol (1.56–1600), campesterol (6.25–400), stigmasterol (3.13–100), fucosterol (1.56–200), β‐sitosterol (3.13–800), coprostanol (1.56–200), cholestanol (0.78–200), and sitostanol (3.13–400).

In conclusion, the GC/MS method validated here is precise, reproducible, and sensitive for the quantification of several sterols and fatty acids in canine fecal samples.

Longitudinal Assessment of Fecal Dysbiosis in Dogs with Acute and Chronic Diarrhea

H.L. Klein1, S. Unterer2, A.E. Jergens3, J.A. Lidbury4, J.M. Steiner5, A.L. Ziese6, J.S. Suchodolski4

1Gastrointestinal Laboratory, Texas A&M University, College Station, TX, College Station, TX, USA, 2Clinic of Small Animal Medicine, LMU Munich, Germany, Munich, Bayern, Germany, 3Iowa State University, College of Veterinary Medicine, Ames, IA, USA, 4Gastrointestinal Laboratory, Texas A&M University, College station, TX, USA, 5Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA, 6Ludwig Maximilian University, Munich, Germany, Munich, Bayern, Germany

Previous studies have identified dysbiosis in dogs with idiopathic IBD (IBD) and dogs with acute hemorrhagic diarrhea syndrome (AHDS). The aim of this study was to evaluate whether dysbiosis patterns differ between these disease groups, and over time.

Fecal samples were analyzed from healthy dogs (n = 27), dogs with IBD (n = 19), and dogs with AHDS (n = 23). Follow‐up samples were collected over multiple time‐points up to 8 weeks from 8 dogs in each group. None of the dogs was treated with antibiotics. Microbiota was analyzed by PCR for E. coli, Streptococcus, Turicibacter, Faecalibacterium, Fusobacterium, Blautia, and C. perfringens. Data was expressed individually and numerically combined as Dysbiosis Index (DI). Kruskal‐Wallis and Friedman's tests were used to compare data between groups and over time with significance set at P < 0.05.

At baseline, dogs with IBD (P < 0.001) and AHDS (P < 0.01) had higher DI compared to healthy dogs. While the CIBDAI score decreased in dogs with SRE significantly (P < 0.05), the DI remained significantly increased compared to healthy dogs even at 8 weeks. In dogs with AHDS, the DI was significantly increased at day 7 (P = 0.002), but not at day 14 (P = 0.59). Differences in dysbiosis patterns were also observed between IBD and AHDS, primary an increase in C. perfringens in dogs with AHDS but not in IBD.

In conclusion, although dogs with SRE improved clinically, the DI remained increased. In contrast, dogs with AHDS showed a rapid normalization in microbiota within 14 days, suggesting differences in dysbiosis patterns due to the underlying disease.

Fecal Microbiota and Body Condition in Growing Beagle Dogs Fed Unrestrictedly for 3.5 hours Daily

L. Leclerc1, C. Thorin2, A. Isaiah3, J. Flanagan4, V. Biourge4, J.S. Suchodolski3, P. Nguyen1

1Nutrition & Endocrinology Unit, Oniris – Vet Nantes, Nantes, Pays de la Loire, France, 2Animal Physiopathology and Functional Pharmacology Unit, Oniris – Vet Nantes, Nantes, Pays de la Loire, France, 3Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA, 4Royal Canin – R&D Discover, Aimargues, Languedoc‐Roussillon, France

Several human and animal studies have reported that the gut microbiota (GM) is altered in adult obese compared to non‐obese subjects, but few studies have been conducted in growing ones. The aim of this study was to identify differences in GM in a homogenous population of growing dogs and to identify when these alterations occur.

We evaluated the composition of the fecal microbiota in 24 growing female Beagle dogs 7 times from 5 to 21 months of age by Illumina sequencing and quantitative PCR of the 16S rRNA gene. The population of dogs was homogenous in terms of age, age at neutering (8 months), environmental and feeding conditions. They were fed the same diet ad libitum for 3.5 hours daily. Up to 10.5 months of age, they were fed a puppy growth diet (fat = 20%, total dietary fiber = 8%, crude protein = 29%, on a dry matter basis (DM)) and subsequently an adult maintenance diet (fat = 11%, total dietary fiber = 17%, crude protein = 28% on as DM basis). The dogs were allotted a posteriori into 3 groups according to fat‐free mass, fat mass percentage and pelvic circumference at 24 months: optimal body weight (IW, n = 9), moderately overweight (OW1, n = 6) and overweight (OW2, n = 9). The data were analyzed by a linear mixed‐effect model with the animal as a random term and corrected for multiple comparisons.

The most abundant phyla were Firmicutes, Bacteroidetes, Fusobacteria and Proteobacteria, accounting for 97 to 99% of sequences. During the pre‐neutering period, Proteobacteria were significantly less abundant (P < 0.05) in the OW2 group than in the IW group. However, throughout the study, the relative abundance of Proteobacteria increased significantly more in the OW1 and the OW2 groups than in the IW group (P < 0.05), without any significant difference in abundance at 21 months between groups.

These results suggest that the early low and then increase abundance of Proteobacteria could be related to overweight development. However, Proteobacteria have highly diverse metabolisms and further analyses at genus level and/or metagenomics analyses may explain the underlying link between the microbiota modulation and excessive weight gain.

Evaluation of the Fecal Microbiota in Cats with Chronic Enteropathy Before and After Treatment

S. Marsilio1, B. Chow2, S.L. Hill2, J.A. Lidbury3, J.M. Steiner4, J.S. Suchodolski5

1Texas A&M University, Gastrointestinal Laboratory, College Station, TX, USA, 2Veterinary Specialty Hospital, San Diego, CA, USA, 3Gastrointestinal Laboratory, Texas A&M University, College station, TX, USA, 4Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College Station, TX, USA, 5Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA

The gastrointestinal (GI) microbiota is considered to play a crucial role in the onset, progression, response to treatment, and outcome in many disease processes including chronic enteropathy (CE) in cats. However, only limited information on the fecal microbial composition in cats with CE is available. The aim of this study was to characterize the fecal microbiota in cats with histologically confirmed CE before and after treatment.

Fecal samples were collected from clinically healthy cats (n = 52) and from cats with CE (n = 14) before and 4 weeks after medical treatment. Chronic enteropathy was defined as the display of signs of gastrointestinal disease for ≥ 3 weeks. Cats that received antibiotics or steroids within 2 weeks prior endoscopy and fecal sampling were excluded from the study. All cats with CE underwent standard medical work‐up before collecting intestinal biopsies endoscopically. Fecal DNA was evaluated using quantitative PCR (qPCR) analysis for: total bacteria, Fusobacteria, Streptococcus spp., Blautia spp., E. coli, Faecalibacterium spp., Turicibacter spp., and Clostridium hiranonis. Differences in bacterial abundance among healthy cats and cats with CE (before treatment) were evaluated using a Mann Whitney test. Samples collected before and after treatment from each affected cat were compared using a Wilcoxon signed rank test. The Feline Chronic Enteropathy Activity Index (FCEAI) was compared in cats with CE before and after treatment using a Wilcoxon signed rank test. Statistical significance was set at P < 0.05.

Histologic evaluation of intestinal biopsies was consistent with IBD (n = 10) or small cell lymphoma (n = 4). Treatment consisted of prednisolone, chlorambucil, various antibiotics, cyanocobalamin, or a combination of these. Before treatment, the abundance of Faecalibacterium spp. and Turicibacter spp. was significantly decreased in cats with CE (P = 0.004 and 0.007, respectively) compared to healthy control cats, while Streptococcus spp. and Fusobacteria were significantly increased in cats with CE (P = 0.037, 0.028, respectively). There were no significant differences in the abundance of the other bacterial groups or the total bacterial counts between groups. FCEAI was significantly decreased after treatment in cats with CE (median [range]: 3 [0 – 4]) compared to before treatment (6 [4 – 8]), P = 0.0312). However, no significant differences were found in the abundance of any bacterial group before and after treatment in cats with CE.

In conclusion, fecal dysbiosis is present in cats with CE. Although the FCEAI decreased significantly after 4 weeks of medical treatment, the abundance of selected bacterial groups did not change significantly. Results might suggest that a longer period of treatment or a different approach to medical therapy (i.e., pre‐ or probiotics) may be needed for cats to recover from fecal dysbiosis associated with CE.

Platelet Indices as Markers of Disease Presence and Severity in Dogs with Inflammatory Bowel Disease

S.O. Mehain1, J.M. Haines2, P.M. Lee2

1Washington State University College of Veterinary Medicine, Pullman, WA, USA, 2Washington State University, Pullman, WA, USA

In human medicine, recent studies have shown that individuals with inflammatory bowel disease (IBD) have statistically significant decreases in mean platelet volume (MPV) and platelet distribution width (PDW) and increases in platelet crit (PCT) compared to healthy controls. In remission, PDW and PCT have been found to be significantly greater than in controls. No studies to date have investigated the association between platelet indices in dogs with IBD compared to healthy controls and to determine if platelet indices could be useful predictors of disease presence or severity. Therefore, the purpose of this retrospective study was to determine if any platelet indices are significantly different between dogs diagnosed with IBD and healthy matched control dogs, or if any indices are predictive of severity based on canine chronic enteropathy clinical activity index (CCECAI) scores assigned retrospectively to dogs diagnosed with IBD.

Dogs with a histopathologic diagnosis of inflammatory bowel disease (IBD) that had no other concurrent illness and had not received immunosuppressant therapy were eligible for inclusion in this study. Twenty dogs diagnosed with IBD and 20 healthy matched control dogs were included in the analysis. Variables compared between the two groups included MPV, PDW, and PCT, as well as mean platelet number (PLT), mean platelet component concentration (MPC), mean platelet component concentration distribution width (PCDW), mean platelet mass (MPM), and platelet dry mass distribution width (PMDW). Significant differences between dogs diagnosed with IBD and healthy controls were evaluated using a t‐test (reported as the mean ± standard deviation) and a Receiver Operating Characteristic (ROC) curve. Additionally, the correlation between platelet indices and the CCECAI scores was evaluated using bivariate scatterplots and linear regression models to determine if any variables were predictive of disease severity.

The results indicate that none of the variables compared between the two groups were statistically significant (P‐value < 0.05) in predicting the presence of disease. Although not significant, those variables close to a P‐value of 0.10 all trend towards significance, and could represent possible predictive indices for disease: mean PLT and PCT were noticeably higher in IBD dogs and mean PCDW was noticeably lower in IBD dogs than in healthy controls. The ROC analysis was performed to determine predictive value of each variable using area under the curve (AUC). Based on AUC values, no platelet indices are strongly correlated with disease. However, PLT, PCT, PCDW, MPM, and PMDW all have AUC values between 0.6 and 0.7, indicating they may be somewhat predictive of disease presence. Bivariate scatterplots with moving average demonstrate that only PLT (P‐value 0.02) and PCT (P‐value 0.03) were statistically significant in predicting disease severity. Linear regression showed mild correlation with CCECAI score for these two indices with PLT r2 value of 0.28 and PCT of 0.25.

In conclusion, a strong correlation between platelet indices and presence or severity of disease in dogs with inflammatory bowel disease is not statistically supported. In spite of this, many indices show potential value as predictors of disease in patients with IBD. A larger study population is likely needed to determine true significance.

Comparison of Serum Lipoprotein Profiles Between Cats with Hepatic Lipidosis and Healthy Control Cats

T. Minamoto1, A. Hamilton2, S.L. Hill2, J.S. Suchodolski3, J.M. Steiner4, J.A. Lidbury3

1Gastrointestinal Laboratory Texas A&M University, College Station, TX, USA, 2Veterinary Specialty Hospital, San Diego, CA, USA, 3Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA, 4Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College Station, TX, USA

Lipoproteins are conglomerates of triglycerides, cholesterol, phospholipids, and apolipoproteins that transport lipids in the blood stream and are classified into 5 classes based on their density: chylomicrons, very low density lipoproteins (VLDL), intermediate density lipoproteins (IDL), low density lipoproteins (LDL), and high density lipoproteins (HDL). Based on the human literature, altered proportions of lipoprotein classes have been described in certain disease states, such as cardiovascular disease or metabolic disease. Few published reports are available in veterinary medicine regarding the characteristics of lipoproteins in diseased animals. Feline hepatic lipidosis (FHL) is characterized by excessive accumulation of lipid within hepatocytes, which can lead to intrahepatic cholestasis and severe hepatic dysfunction. We hypothesized that cats with FHL show an altered serum lipoprotein profile due to intensive lipolysis in adipose tissue and alterations in hepatocellular lipid metabolism. The aim of this study was to compare serum lipoprotein profiles as well as serum cholesterol and triglyceride concentrations between cats with FHL and healthy control cats.

Analysis of serum lipoprotein profiles using density gradient ultracentrifugation was performed in 20 cats with FHL (confirmed by cytological and/or histological findings) and 20 healthy control cats. The area under the curve for triglycerides‐rich‐lipoproteins (TRL), LDL, and HDL was calculated. In addition, serum cholesterol and triglyceride concentrations were measured using a clinical chemistry analyzer. Data were analyzed using a Wilcoxon rank sum test. Significance was set at P < 0.05.

LDL content was significantly higher in cats with FHL than in healthy control cats (P = 0.0001), while HDL content was significantly lower in cats with FHL than in healthy control cats (P = 0.0032). TRL content was not significantly different between the two groups (P = 0.0699). Also, serum cholesterol and triglyceride concentrations were not significantly different between the two groups (P = 0.5075 and P = 0.2541, respectively).

In our study, serum lipoprotein profiles were altered in cats with FHL even though serum cholesterol and triglyceride concentrations were not significantly different compared to healthy control cats. The clinical importance of lipoprotein profiling in cats with hepatic lipidosis warrants further study.

Ileal and Colonic Mucosal Microbiota in Dogs with Steroid Responsive Chronic Enteropathy

F. Bresciani1, Y. Minamoto2, J.S. Suchodolski2, G. Galiazzo1, C. Vecchiato1, C. Pinna1, G. Biagi1, M. Pietra1

1Department of Veterinary Medical Sciences, University of Bologna, Ozzano dell'Emilia, Emilia‐Romagna, Italy, 2Gastrointestinal laboratory, Texas A&M University, College Station, TX, USA

Exact aetiology for inflammatory chronic enteropathies in dogs remains unknown. Accumulating evidence suggests a pivotal role for intestinal dysbiosis in disease pathogenesis. Many studies have evaluated the alteration of faecal microbiota in canine chronic gastrointestinal (GI) disease, and less research is focused on mucosal microbiota, especially in the ileum and colon. The objectives of the current study were to evaluate ileal and colonic mucosal microbiota in dogs with steroid responsive enteropathy (SRE) before and after 4 months of treatment, and to compare them to control dogs (CD).

A total of 10 dogs diagnosed with SRE were enrolled. Complete GI endoscopy was performed and samples were collected by a cytology brush at diagnosis (SRE‐Baseline, n = 10) and after 4 month of treatment (SRE‐After, n = 8). Oral laxative and 2–4 water enemas were performed before endoscopy. A total of 6 CD that were euthanized for reasons unrelated to this study, with no GI disease, were included. Samples from CD were obtained during necropsy within 3 hours of death. Mucosal genomic DNA was extracted and used for Illumina sequencing of 16S rRNA genes. Sequence data were analysed using the QIIME pipeline. Statistical significance was set at P < 0.05.

Clinical signs improved significantly after 4 month of treatment in SRE, but no improvement was seen on endoscopic or histological evaluation. Significant differences in microbial communities between SRE‐baseline and CD were observed in the colon (ANOSIM P = 0.002), but not in the ileum (ANOSIM P = 0.180). In dogs with SRE, both ileal and colonic microbial communities remained similar after 4 month of treatment (ANOSIM P = 0.189 and P = 0.637, respectively), and were different from CD (ANOSIM P = 0.001 and P = 0.004, respectively).

Results of this study suggest that the mucosal microbiota in the colon of dogs with SRE is different from that of CD. Although clinical signs improved, colonic mucosal dysbiosis was still present after 4 months of treatment.

Effects of Prednisolone Administration on Pancreatic Tissue and Pancreatic Lipase Immunoreactivity Concentration in Healthy Dogs

H. Ohta, N. Sasaki, K. Morishita, K. Nakamura, M. Takiguchi

Department of Veterinary Clinical Sciences, Hokkaido University, Sapporo, Hokkaido, Japan

The canine pancreatic lipase immunoreactivity (cPLI) concentration is considered the most specific biomarker for the diagnosis of canine pancreatitis. In one report, 4 weeks of oral administration of prednisone (2.2 mg/kg/day) to six healthy dogs had no significant effect on the serum cPLI concentration (Steiner et al., 2009). The aim of this study was to determine the effect of high‐dose prednisolone (4 mg/kg/day) on the pancreatic tissue and serum cPLI concentration in healthy dogs. Six healthy beagles received prednisolone (4 mg/kg/day subcutaneously) for 2 or 3 weeks. Measurement of the cPLI concentration and ultrasonographic examination of the pancreas were performed before and after treatment. Laparoscopic biopsy of the liver and right pancreatic lobe was also performed before and after treatment. The serum cPLI concentration was compared before and after treatment using the Wilcoxon signed‐rank test. The serum cPL concentration was significantly higher after than before treatment. The post‐treatment cPLI concentration was classified as normal (≤200 μg/L) in five dogs and abnormal (≥400 μg/L) in one dog.

Ultrasonographic examination of the pancreas showed no significant change after treatment. There was no histological evidence of pancreatitis in the right pancreatic lobe after treatment. Administration of prednisolone at 4 mg/kg/day for 2 or 3 weeks increased the serum cPLI concentration in six healthy dogs, although a clinically significant increase (≥400 μg/L) was observed in only one dog. Histological evidence of pancreatitis was not observed in any of the six dogs examined.

Corticosteroid Therapy as Initial Treatment for Dogs with Acute Pancreatitis: 65 Cases (2011–2016)

H. Okanishi1, T. Nagata2, S. Nakane2, T. Watari1

1Laboratory of Veterinary Internal Medicine, Department of Veterinary Medicine, College of Bioresource Sciences, Nihon University, Fujisawa, Kanagawa, Japan, 2Nakane Animal Hospital, Kitaku, Tokyo, Japan

In humans with acute pancreatitis (AP), treatment within 48 hours of onset is very important and can profoundly influence the prognosis. In dogs with AP, the efficacy of corticosteroid treatment is unknown. The purpose of this study was to investigate the responsiveness to initial corticosteroid therapy and prognosis in dogs with AP.

Sixty‐five dogs were included in this study. Dogs were divided into the prednisolone (PDL) or non‐prednisolone (NPDL) groups. Prednisolone was administered at a dose of 1 mg/kg/day in the PDL group. Enrofloxacin, maropitant, famotidine, fentanyl, and IV fluids were used as treatments in both groups. The number of days until the CRP level < 2 mg/dl, the clinical sign score (APCS) < 2, duration of hospitalization, mortality, and relapse rate were analyzed. Treatment effect was assessed using relative and absolute risk reduction (RRR, ARR) and number needed to treat (NNT).

CRP value was significantly lower in the PDL group than in the NPDL group after day 3 of hospitalization (4.2 vs. 12 mg/dl, P < .05). Days taken for CRP level < 2 mg/dl and APCS < 2 and the duration of hospitalization were significantly shorter in the PDL group than the NPDL group (4 vs. 8 days, 5 vs. 7 days, and 5 vs. 8 days, P < .01, respectively). Mortality in the PDL group was significantly lower than in the NPDL group (11.3 vs. 46.1%), and had a therapeutic effect (RRR, 0.76; ARR, 34%; NNT, 2.9).

In conclusion, initial corticosteroid therapy may be effective for dogs with AP.

Fecal Composition in Dogs with Naturally Occurring Exocrine Pancreatic Insufficiency

J.C. Parambeth1, A. Isaiah2, J.A. Lidbury2, J.S. Suchodolski2, J.M. Steiner1

1Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College Station, TX, USA, 2Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA

Exocrine pancreatic insufficiency (EPI) is a syndrome characterized by the inadequate synthesis and secretion of pancreatic digestive enzymes, resulting in maldigestion. Dietary trials in dogs with EPI have not demonstrated a benefit for any specific diet. Determination of fecal composition can be used to determine nutrient digestibility. The aim of this study was to evaluate the fecal composition [dry matter (DM), organic matter (OM), ash, crude protein (CP), and total lipid content (acid hydrolyzed fat, AHF)] in dogs with naturally occurring EPI.

Surplus fecal samples from 17 dogs (9 dogs were receiving enzyme supplementation) enrolled in an unrelated clinical trial involving dogs with EPI were utilized. To be included into the study, the dogs had to be at least 1 year of age, have clinical signs of EPI (i.e., polyphagia, weight loss, steatorrhea, and/or loose, voluminous, and/or malodorous stools), have a serum cTLI concentration ≤2.5 μg/L, and be free from any clinically apparent disease other than EPI. Twelve healthy dogs were enrolled as healthy controls. Three naturally voided fecal samples were collected by the owners at home and stored frozen until shipment. Upon receipt, samples were pooled, aliquoted, and stored at −80°C until shipment to the Animal Sciences Laboratory, Department of Animal Sciences and Division of Nutritional Sciences, University of Illinois. Fecal samples were analyzed for DM, OM, ash, CP, and AHF. Fecal composition parameters were compared between the three groups; healthy control dogs, dogs with EPI on enzyme supplementation (EPI‐S), and dogs with EPI without enzyme supplementation (EPI‐NS) using the Kruskal‐Wallis test. Where significant differences were identified, post‐hoc testing was performed using a Dunn's multiple comparison test. Significance was set at P < 0.05.

Median DM content of feces (min–max) were 36% (30–46.1%), 29.8% (26.5–32.4%), and 28.6% (21.2–34.7%) in the healthy, EPI‐S, and EPI‐NS dogs, respectively. Median OM content of feces on DM basis (min–max) were 68.3% (60.9–71.5%), 83.3% (72.9–89.9%), and 89.2% (86–95.9%) in the healthy, EPI‐S, and EPI‐NS dogs, respectively. Median ash content of feces on DM basis (min–max.) were 31.7% (28.6–39.1), 16.7% (10.1–27.1), and 10.8% (4.2–14) in healthy, EPI‐S, and EPI‐NS dogs, respectively. Median CP content of feces on DM basis (min–max.) were 25.1% (21.7–28.1), 34.9% (24.5–40.2), and 32.3% (16.3–51.6) in the healthy, EPI‐S, and EPI‐NS dogs, respectively. Median AHF content of feces on DM basis (min– max) were 7.2% (5.4–9.5), 11.4% (8–20.1), and 20.2% (9–34.1) in the healthy, EPI‐S, and EPI‐NS dogs, respectively. DM, OM, ash, and AHF were significantly different between healthy dogs and dogs with EPI. However, no significant differences were observed in the fecal composition of EPI dogs, regardless of whether they were treated with enzyme supplementation.

This study showed that fecal composition is altered in dogs with naturally occurring EPI even when managed with enzyme supplementation. Larger prospective clinical studies are necessary to elucidate the mechanisms leading to this altered fecal composition despite enzyme replacement therapy.

Genome‐Wide Association Study in German Shepherd Dogs with IBD Reveals Possible Role of TH2 Cytokines

A. Peiravan1, F. Bertolini2, M. Rothschild2, A.E. Jergens3, K. Simpson4, D. Werling5, K. Allenspach3

1Department of Veterinary Clinical Sciences and Services, Royal Veterinary College, London, UK, Hertfordshire, UK, 2Department of Animal Science, College of Agriculture and Life Sciences, Iowa State University, USA, Ames, IA, USA, 3Iowa State University, College of Veterinary Medicine, Ames, IA, USA, 4College of Veterinary Medicine, Cornell University, USA, Ithaca, NY, USA, 5Department of Pathology and Pathogen Biology, Royal Veterinary College, London, UK,, Hertfordshire, UK

German shepherd dogs (GSD) in the UK are at increased risk of developing Inflammatory Bowel Disease (IBD). Similar to IBD in humans, understanding the genes involved in canine IBD is believed to reveal insights into disease pathogenesis in the dog.

The aim of the present study was to identify single nucleotide polymorphisms using a genome wide association study (GWAS) which may confer genetic susceptibility or resistance to IBD. Genomic DNA was extracted from EDTA blood or saliva samples of 96 cases and 98 controls. Genotyping of cases and controls was performed on the Canine Illumina HD SNP array and data generated were analyzed using PLINK. Using Fst and logistic association analyses, we identified 16 candidate genes that have been previously reported to be associated with human IBD or are involved in the pathways that are linked to the disease. Eight of these genes, one (PTPRC; Protein Tyrosine Phosphatase, Receptor Type C) on chromosome 7 and seven (IL4, IL5, IL13, CSF2; Colony Stimulating Factor 2, SLC22A4 and A5; Solute Carrier Family22 member 4 and 5, IRF1; Interferon Regulatory Factor 1) on chromosome 11 have been identified in both approaches. In conclusion we identified 16 genes associated with IBD in GSDs. Our results suggest that Th2 cytokines could be implicated in the pathogenesis of IBD in GSDs. Targeted re‐sequencing of the genes identified in our study will help to identify causative SNPs and consequently functional analysis of the causal SNPs reveal insights into mechanisms involved in pathogenesis on canine IBD.

Capsule Endoscopy Findings in Dogs with Hypoalbuminemia

J.S. Pomrantz, J.A. Solomon

Infiniti Medical, Menlo Park, CA, USA

The purpose of this study was to analyze capsule endoscopy (CE) findings in hypoalbuminemic dogs. The records of 178 dogs given ALICAM for gastrointestinal signs and/or laboratory abnormalities were retrospectively assessed. Seventeen dogs were hypoalbuminemic and had complete studies. The mean ± SD albumin was 1.9 ± 0.4 g/dL, mean ± SD age was 7.5 ± 2.8 years, and mean ± SD weight was 20.5 ± 11.4 kg. Eight of 17 dogs had concurrent anemia. Prior to CE, 15 of 17 dogs had ultrasound examinations. The gastrointestinal tract (stomach, small intestine [SI], and colon) was normal (n = 8) or characterized by nonspecific wall thickening (n = 7). Seventeen of 17 CE studies were abnormal. Findings consisted of irregular/thickened mucosa with or without erosions (n = 6), numerous dilated lacteals (n = 4), severe erosions or ulcers with active bleeding (n = 2), mass (n = 2), hookworms (n = 1), SI nodules (n = 1), and lesions consistent with colonic vascular ectasia (n = 1). Lesions seen on CE (dilated lacteals, colonic vascular ectasia) were confirmed in two dogs that subsequently underwent traditional endoscopy. A third dog had gastroduodenoscopy prior to CE, which showed SI hemorrhage, but missed the SI nodules later identified with CE. A diagnosis of Strongyloides was made based on endoscopic biopsies and fecal Baermann test. A fourth dog with severe melena had a negative gastroduodenoscopy and laparoscopy, yet CE subsequently identified severe gastrointestinal erosions and ulcers thought to be secondary to NSAID administration. CE can be informative in dogs with hypoalbuminemia even when lesions are not identified with traditional endoscopy, surgery, or ultrasound.

Canine Adipose Tissue‐derived Mesenchymal Stem Cells Reduce Severe Acute Pancreatitis by Regulating T cells

W.‐J. Song1, Q. Li2, M.‐O. Ryu1, S.‐C. Park1, H.‐J. Kim1, S.‐Y. Kim1, H.‐M. Yang1, H.‐Y. Youn1

1Seoul National University, Seoul, N/A, Republic of Korea, 2Seoul National University, Seoul, N/A, China (People's Republic)

Severe acute pancreatitis (SAP) is associated with systemic complications and high mortality rate in dogs. Mesenchymal stem cells (MSCs) have been investigated for their therapeutic potential in several inflammation models. This study investigated the effects of canine adipose tissue‐derived (cAT)MSCs in a rat model of SAP induced by retrograde injection of 3% sodium taurocholate solution into the pancreatic duct. cATMSCs labeled with dioctadecyl‐3,3,3′‐tetramethylindo‐carbocyanine perchlorate (1 × 107 cells/kg) were systemically administered to rats and pancreatic tissue was collected 3 days later for histopathological, quantitative real‐time PCR, and immunocytochemical analyses. Greater numbers of infused cATMSCs were detected in the pancreas of SAP as compared to sham‐operated rats. cATMSC infusion reduced pancreatic edema, inflammatory cell infiltration, and acinar cell necrosis, and decreased pancreatic expression of the pro‐inflammatory cytokines tumor necrosis factor‐α, interleukin (IL)‐1β, ‐6, ‐12, ‐17, and ‐23 and interferon‐γ, while stimulating expression of the anti‐inflammatory cytokines IL‐4 and IL‐10 in SAP rats. Moreover, cATMSCs decreased the number of cluster of differentiation 3‐positive T cells and increased that of Forkhead box P3‐positive T cells in the injured pancreas. These results indicate that cATMSCs can be effective as a cell‐based therapeutic strategy for treatment of SAP in dogs.

Likelihood and Outcome of Esophageal Perforation Associated with an Esophageal Foreign Body in Dogs

A. Sterman1, A.K. Cook2, K. Thieman‐Mankin1

1Texas A&M College of Veterinary Medicine, College Station, TX, USA, 2Texas A&M University College of Vet Med, College Station, TX, USA

There is limited information regarding the likelihood and outcome of esophageal perforation secondary to an esophageal foreign body (EFB) in dogs.

The medical records databases at two veterinary hospitals identified 125 dogs with EFBs from Jan 2005‐Jan 2014. Bones (44%) and fishhooks (37%) predominated. Fifteen dogs (12%) were diagnosed with a perforation based on radiography (n = 4), esophagoscopy (n = 10) or surgical exploration (n = 1). Ten of these had a fishhook EFB; five had a bone. Esophageal perforation was not influenced by patient size. Esophageal perforation was more likely in dogs with a fishhook (10/15 v 27/110; odds ratio 6.1; P = 0.002) than those with other types of EFB. For fishhook and bone EFB, the median time from ingestion to presentation was substantially longer for dogs with a perforation (12 hrs and 96 hrs respectively) than those without (1 hr [P = 0.007] and 24 hrs [P = 0.003] respectively). Endoscopic EFB removal was attempted in 13 dogs; eight recovered without surgical intervention and one was euthanized because of esophageal necrosis. Six dogs were treated surgically; five survived and one was euthanized intraoperatively. The two non‐survivors both had long‐standing bone EFBs. Survival rate for dogs with esophageal perforation was 87%.

Esophageal perforation secondary to an EFB is relatively uncommon (12%). However, ingestion of a fishhook markedly increases the likelihood of perforation, as does a delay between EFB ingestion and presentation to the hospital. All ten dogs with perforation secondary to a fishhook were discharged alive, and the majority of the survivors (8/13; 62%) did not require surgical intervention.

The Effect of Prolonged Ambient Temperature Exposure on the Feline Fecal Microbiota

M.T. Gavriel1, D. Gomez‐Nieto2, C. Chau1, J.S. Weese2, A. Verbrugghe1

1Department of Clinical Studies, Ontario Veterinary College, University of Guelph, Guelph, ON, Canada, 2Department of Pathobiology, Ontario Veterinary College, University of Guelph, Guelph, ON, Canada

Research involving fecal collection from client‐owned cats can be very challenging, due to the time difference between defecation and collection, as well as the owners’ compliance to refrigerate/freeze samples immediately following collection. This study examined the effect of prolonged ambient temperature exposure on the feline fecal microbiota composition.

Fecal samples were collected within 15 minutes after defecation from 12 healthy cats, at a boarding facility in Guelph, ON, Canada. Samples were aliquoted and the first aliquot was frozen at −80°C within one hour of defecation (time point (T) 0). Remaining aliquots were maintained at ambient temperature (20–23°C) and frozen at −80°C at 6, 12, 24, 36, 48, 72 and 96 hours after collection. DNA was extracted from all aliquots, and the V4 region of the 16S rRNA gene was amplified using polymerase chain reaction (PCR). The PCR products were sequenced with next‐generation sequencing (Illumina MiSeq). Mothur software was used for bioinformatics analysis.

A total of 9,118,609 sequences passed all filters, with a median of 93,909 sequences per sample (range: 50,315 to 207,127 sequences/sample). Relative abundance of different taxa was calculated, and compared between time points, with no significant differences (P > 0.05 for all taxa between time points, following the Benjamini‐Hochberg adjustment). Alpha‐diversity indexes (Shannon diversity, Simpson diversity and Chao1) were used to evaluate evenness, diversity and richness respectively, and no significant differences were identified (P > 0.05 for all comparisons). Beta‐diversity measurements were applied to assess similarity in community membership and structure. Phylogenetic trees were created, to visually present community membership and structure (Jaccard and Yue & Clayton indexes, respectively) over time between cats, and no significant differences were identified (parsimony and unifrac‐unweighted tests, P > 0.05). Principal coordinate analyses were performed and showed that samples cluster mainly by cat, with no significant differences between time points (AMOVA, P > 0.05; HOMOVA, P > 0.05). Linear discriminant analysis (LDA) effect size (LEfSe) method was performed to identify genomic features, such as enriched taxa, between time points. LEfSe analysis failed to detect any differences between groups. Random forest algorithm analysis identified a 92% error rate, indicating homogeneity across time points.

Contradictory evidence exists in human studies in regards to the effect of storage at different temperatures on the fecal microbiota. A recent study by Weese and Jalali showed limited impact of refrigeration for 14 days on the feline and canine microbiota. This concurs with the current study, which demonstrated that maintaining feline fecal samples at ambient temperature up to 96 hours (4 days) has no demonstrable effect on the bacterial membership and structure. Results from LEfSe and random forest analysis also provides supportive evidence for the lack of difference between time points.

Effects of Donor Sex And Pre‐Conditioning on Cytokine Production of Canine Adipose‐Derived Mesenchymal Stem Cells

T. Webb, S. Smith, R. Timmons, C. Webb

Colorado State University, Fort Collins, CO, USA

Current research suggests that the use of mesenchymal stem cell (MSC) therapy to treat clinical diseases, such as chronic enteropathy, may be optimized through donor selection and pre‐conditioning strategies. We therefore investigated the effects of several pre‐conditioning methods and donor sex on in vitro cytokine production of adipose‐derived canine MSC (cMSC). Passage 6 cMSC were generated from cryopreserved subcutaneous adipose tissue from 6 healthy, adult, intact dogs, 3 female and 3 male. cMSC were subjected to 6 different conditions: control, poly I:C, IFNγ, TGFβ, serum‐free media, and a mixed hypoxic environment. Supernatants from the cMSC were harvested after 1, 6, 12, 18, and 24 hours. Levels of IL‐6, IL‐8, IL‐10, TGFβ, VEGF, and MCP‐1 were determined by ELISA. IL‐10 was below the level of detection in all samples. TGFβ was measured in the wells that were treated with TGFβ but was generally otherwise undetectable. IL‐6 was only detected in relatively low levels in some samples from one male and one female donor. cMSC produced small to moderate amounts of IL‐8 constitutively, which reached peak levels by 18 hours and was altered by pre‐conditioning. cMSC produced large amounts of MCP‐1 and VEGF, which were altered by pre‐conditioning. Generally, the serum‐free and the mixed hypoxic environment decreased production of all measured cytokines. Samples from the male donors showed significant donor variability compared to females suggesting that treatment success might be influenced by careful donor selection as well as pre‐treatment conditioning, depending on the desired outcome.

Effects of Proviable‐DC® on Clinical Signs, Fecal Microbiomes, and Metabolomic Profiles in Cats receiving Clindamycin

J. Whittemore1, J.E. Stokes2, N. Laia3, J. Price4, J.S. Suchodolski5

1University of Tennessee, Knoxville, Knoxville, TN, USA, 2College of Veterinary Medicine, University of Tennessee, Knoxville, TN, USA, 3Nashville Vet Specialists, Nashville, TN, USA, 4University of Tennessee, Knoxville, TN, USA, 5Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA

The purpose of this study was to compare the incidence of antibiotic‐associated gastrointestinal signs (AAGS) and alterations in the fecal microbiome and metabolome of 16 healthy cats randomized to receive 150 mg clindamycin with 1 capsule placebo or synbiotic (Proviable‐DC

®) once daily for 21 days with reevaluation 603 days after antibiotic discontinuation.

Daily food consumption, occurrence of vomiting, and fecal score were recorded. Fecal samples were collected on days 5–7 (baseline), 26–28, and 631–633. Sequencing of 16S rRNA genes was performed, and mass spectrometry was used to determine metabolomic profiles. P < 0.05 was considered significant, with the Benjamini & Hochberg's False Discovery Rate (fdr) used to adjust for multiple comparisons.

All cats developed AAGS. Vomiting was less in cats receiving a synbiotic, but the difference was not significant. Alpha and beta diversity changed significantly during treatment. Decreases in Actinobacteria (Bifidobacterium, Collinsella, Slackia), Bacteriodetes (Bacteroides), Lachnospiraceae (Blautia, Coprococcus, Roseburia), Ruminococcaceae (Faecilobacterium, Ruminococcus), and Erysipelotrichaceae (Bulleidia, [Eubacterium]) and increases in Clostridiaceae (Clostridium) and Proteobacteria (Aeromonadales, Enterobacteriaceae) occurred, with incomplete normalization by days 631–633. Derangements in short‐chain fatty acid, bile acid, indole, sphingolipid, polyamine and cinnaminic acid profiles occurred, some of which persisted through days 631–633 and differed between groups.

In cats, AAGS are accompanied by short‐ and long‐term dysbiosis and alterations in metabolomic profiles. Metabolomic profiles, but not AAGS or bacterial abundances, differed between groups over time. Further investigation is warranted to determine whether antibiotic‐induced dysbiosis is associated with an increased risk of AAGS in cats and whether prior synbiotic administration ameliorates this risk.

Effects of Staggered Proviable‐DC® Administration on Fecal Microbiome and Metabolomic Profiles of Cats receiving Clindamycin

J. Whittemore1, J.E. Stokes2, J. Price1, J.S. Suchodolski3

1University of Tennessee, Knoxville, TN, USA, 2College of Veterinary Medicine, University of Tennessee, Knoxville, TN, USA, 3Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA

The purpose of this study was to compare the fecal microbiome and metabolome of 16 healthy cats randomized to receive 75 mg clindamycin with food, followed 1 hour later by 2 capsules of a placebo or synbiotic (Proviable‐DC

®) once daily for 21 days with reevaluation 6 weeks after treatment.

Fecal samples were collected on days 5–7 (baseline), 26–28, and 68–70. Sequencing of 16S rRNA genes was performed, and mass spectrometry was used to determine metabolomic profiles. P < 0.05 was considered significant, with Benjamini & Hochberg's False Discovery Rate used to adjust for multiple comparisons.

Alpha and beta diversity were altered significantly during treatment, and alterations in the Shannon and dysbiosis indices persisted at days 68–70. Age was significantly associated with the Shannon and dysbiosis indices and relative bacterial abundances. Decreases in Actinobacteria (Adlercreutzia, Bifidobacterium, Collinsella, Slackia), Bacteriodetes (Bacteroides, Prevotella), Ruminococcaceae (Faecilobacterium, Ruminococcus), Veillonellaceae (Megamonas, Megasphaera,Phascolarctobacterium) and Erysipelotrichaceae ([Eubacterium]) and increases in Clostridiaceae (Clostridium), Proteobacteria (Enterobacteriaceae) occurred, with incomplete normalization by days 68–70. Metabolite profiles differed significantly by time, group and time x group for 106, 5 and 10 metabolites, respectively. Microbiome changes were accompanied by derangements in short‐chain fatty acid, bile acid, indole, sphingolipid, and cinnaminic acid profiles, some of which persisted through days 68–70 and differed between groups.

In cats, antibiotic administration causes sustained dysbiosis. Age‐associated changes in the microbiome could affect sensitivity to antibiotic‐induced gastrointestinal effects. Metabolomic profiles differed between treatment groups over time, suggesting a potential mechanism for decreased antibiotic‐induced gastrointestinal effects in cats administered synbiotics.

Serum IL‐17 in Dogs with Immune‐Mediated Hemolytic Anemia

B. Cuq1, S. Blois2, R. Darren Wood2, A. Abrams‐Ogg2, C. Bédard3, G.A. Wood1

1Ontario Veterinary College – University of Guelph, Guelph, ON, Canada, 2University of Guelph, Guelph, ON, Canada, 3Faculté de Médecine Vétérinaire – Université de Montréal, St Hyacinthe, PQ, Canada

Canine immune mediated hemolytic anemia (IMHA) has a high mortality rate, up to 70%, despite aggressive immunosuppressive therapy and supportive care. Development of immune‐mediated disease is likely a multifactorial process leading to a loss of normal immune self‐tolerance. Humans affected by autoimmune hemolytic anemia show several cytokine derangements, including a newly defined subset of T cells, Th17 cells secreting IL‐17. IL‐17 appears to have a critical role in the pathogenesis of AIHA in people, and work in mouse models of the disease indicates IL‐17 as a potential treatment target. It was hypothesized that dogs with IMHA would display significant elevations in IL‐17 concentrations compared to healthy controls.

Serum IL‐17 was measured in thirty dogs with primary IMHA at diagnosis (D0); and 2 (D2) and 4 (D4) days after initiation of immunosuppressive therapy as well as in 25 clinically healthy dogs. IL‐17 was significantly increased at the time of presentation (D0) in dogs with IMHA (P = 0.0158) compared to healthy dogs as well at D2 (P = 0.0404). Over the course of hospitalization, IL‐17 significantly decreased between D0 and D2 (P = 0.0146) and D0 and D4 (P = 0.0010) for the discharged patients (N = 19). No changes in IL‐17 levels were observed between D0, D2 and D4 among the dogs that were euthanized in hospital. There was no correlation between IL‐17 concentration and bilirubin levels, lactate or any of the CBC parameters at the time of presentation or throughout the hospitalization. IL‐17 is elevated in dogs with IMHA and seems to decrease when response to therapy is observed.

Glutathione Peroxidase, Total Antioxidant Capacity, and Urinary F2‐isoprostane Concentrations in Anemic Dogs

A. Kendall1, A. Woolcock2, A. Brooks2, G.E. Moore2

1Purdue University College of Veterinary Medicine, Lafayette, IN, USA, 2Purdue University College of Veterinary Medicine, West Lafayette, IN, USA

Oxidative stress plays a considerable role in the pathophysiology of multiple diseases and has been well documented as a contributor to disease in both the human and veterinary literature. Oxidative stress results from the production of reactive oxygen species that cause cellular injury through direct and indirect mechanisms. One at‐risk cell type is the erythrocyte, however, the role of oxidative stress in canine anemia has not been investigated. Oxidative stress is primarily assessed through the activity of antioxidant enzymes, the concentration of endogenous antioxidants, and byproducts of oxidative damage. The purpose of this study was to determine if oxidative stress is present in anemic dogs. We hypothesized that anemic dogs would have an alteration in the activity of glutathione peroxidase (GPx), a reduction of total antioxidant capacity (TAC), and an increased concentration of urinary F2‐isoprostanes (F2IsoP, a byproduct of lipid peroxidation) when compared to healthy dogs.

In this prospective, cross‐sectional study, we recruited 40 client‐owned dogs with anemia (PCV < 30%) due to any cause and 40 client‐owned age‐matched healthy control dogs. All dogs were evaluated for whole blood GPx activity, plasma TAC and urinary F2‐isoprostane concentrations. Anemic dogs were further stratified into those with a hemolytic cause and those with a non‐hemolytic cause.

Anemic dogs had a significantly lower GPx activity (43.1 x 103 U/L +/‐ 1.6 x 103) than dogs in the control group (76.0 x 103 U/L +/‐ 1.9 x 103) (P < 0.01) indicating inactivation of GPx activity and depletion of antioxidant defense mechanisms. Further, dogs with a hemolytic anemia had lower mean GPx activity when compared to dogs with non‐hemolytic anemia, but this difference did not reach significance (P = 0.51). However, this may suggest that the degree of oxidative stress may be more severe with hemolysis. TAC concentrations (P = 0.10) and urinary F2‐isoprostanes (P = 0.75) did not significantly differ between groups.

This is the first study in veterinary medicine demonstrating a glutathione peroxidase deficiency and subsequent oxidative stress in anemic dogs. Further studies are warranted to determine if antioxidant supplementation would improve survival and overall outcome as part of a therapeutic regimen for anemic states, especially those with hemolysis.

The Effect of Yunnan Baiyao on the Kinetics of Hemostasis in Healthy Dogs

R. MacRae, A. Carr

Western College of Veterinary Medicine, Saskatoon, SA, Canada

Yunnan Baiyao is a popular Chinese herbal medicinal agent that has been used for its anti‐inflammatory properties, hemostatic properties and wound healing properties in people for approximately a century. It has been recommended for use in dogs to control hemorrhage such as epistaxis or with splenic hemangiosarcoma. However there have been no studies evaluating the safety of Yunnan Baiyao for use in dogs and there have been no in vivo studies on the effect of Yunnan baiyao on coagulation parameters in healthy dogs. The objectives of this study were to evaluate the effects of Yunnan Baiyao on the kinetics of hemostasis and to determine if there are any adverse effects associated with the administration of Yunnan Baiyao to healthy dogs. This study used six healthy purpose bred dogs. Baseline evaluation included a complete blood count, serum biochemistry profile, urinalysis, prothrombin time, partial thromboplastin time, and hemostatic kinetics using ROTEM thromboelastrometry (EXTEM, INTEM, FIBTEM). ROTEM parameters assessed included clotting time, clot formation, clot stability and clot lysis. The dogs were administered oral Yunnan baiyao at a dose of 25 mg/kg orally once daily for 6 days. On day 2, day 4 and day 6 ROTEM thromboelastometry was again performed on all dogs. On day 6 a complete blood count, serum biochemistry profile, urinalysis, prothrombin time and partial thromboplastin time was also performed on each dog. No significant changes were noted with treatment in any of the parameters monitored. No significant change in clotting times (P > 0.2), clot formation time (P > 0.2), and clot formation rate (P > 0.2) were noted when pre‐treatment ROTEM values and post treatment ROTEM values were compared. From the results of this study we can conclude that Yunnan Baiyao is safe for administration to dogs and does not result in significant changes to baseline health parameters or coagulation parameters in healthy dogs.

Prestorage Leukoreduction Attenuates Neutrophil Extracellular Trap Formation in Canine Stored Packed Red Blood Cell Units

E.R. McQuinn1, S. Smith2, A.K. Viall1, C. Wang1, D.N. LeVine1

1Iowa State University, Ames, IA, USA, 2Colorado State University, Fort Collins, CO, USA

Neutrophil extracellular traps (NETs), webs of DNA and citrullinated histones extruded from activated neutrophils to trap invading organisms, are a newly recognized cause of transfusion‐related acute lung injury. NETs form during storage of human red blood cells (RBCs); formation is attenuated by prestorage leukoreduction (LR). Inflammatory cytokines that form in stored RBCs may promote NETosis in the neutrophils of the transfusion recipient.

We hypothesized that NETs form during storage of canine RBCs, that prestorage LR reduces NETosis, and that application of supernatant from stored, non‐leukoreduced (NLR) RBC units would induce NETosis in healthy canine neutrophils.

A blood unit was collected from 6 healthy dogs, processed, then divided into two equal subunits of RBCs, one of which was leukoreduced with a commercial filter. Units were stored at 4°C for 42 days and sampled biweekly. NET markers citrullinated histone H3 and cell‐free DNA were quantified in unit supernatants via western blot and fluorescence with DNA dye, respectively. Supernatants were applied ex vivo to neutrophils from healthy dogs and release of extracellular DNA representing NETosis was assessed. Data was analyzed by repeated measures ANOVA with treatment and time as fixed variables.

Both markers of NETs increased during RBC storage and were higher in NLR than LR units (Figure 1). Isolated neutrophils did not form NETS in response to supernatant from stored canine RBC.

NETosis occurs in stored canine NLR RBC units, and is attenuated by prestorage LR. NETs may be mediators of transfusion reactions; the clinical benefits of prestorage LR should be further explored.

Retrospective Analysis of Length of Glucocorticoid Treatment in 77 Cases of Canine Autoimmune Blood Disorders

N. Puza1, S. Hillock2

1Veterinary Medical Center of Long Island, West Sayville, NY, USA, 2Veterinary Medical Center of Long Island, West Islip, NY, USA

The purpose of this study was to describe the length of time glucocorticoids are utilized in the treatment of autoimmune hemolytic anemia (AIHA), immune‐mediated thrombocytopenia (ITP), and Evans syndrome in canines treated at the authors’ institution.

A retrospective analysis was performed of patients that achieved remission and were weaned from glucocorticoid treatment for which full treatment data was available. Seventy‐seven cases were included and analyzed for median dosage, tapering timeline, and length of glucocorticoid therapy. The collected data is presented alongside anecdotal data obtained by survey of Diplomates of the American College of Veterinary Internal Medicine in the specialty of Internal Medicine.

The median starting glucocorticoid dose was 2.3 mg/kg/day (prednisone equivalent). All cases were started on a secondary immunosuppressant. The median time between taper steps was 7 days. The median taper of the original dose at each step was 25%. Median length of glucocorticoid treatment was 49 days (range 0–922). Thirteen cases (16.8%) experienced relapse within the study period. Two occurred within 6 months of discontinuing prednisone while on a secondary immunosuppressant, and two within 6 months of discontinuing secondary immunosuppressive therapy. The remainder occurred 8–26 months after discontinuing secondary immunosuppressive therapy.

Survey data indicated 85.1% of respondents initiate glucocorticoid treatment at 2 mg/kg/d. 80.5% begin to taper glucocorticoids at ≥2 weeks, with each additional step at 2–4 week intervals (97%). 85.4% of respondents decrease by 25% of the original dose at each step, with 79% of responses indicating glucocorticoid treatment for greater than three months. Estimated relapse rates within six months of discontinuing immunosuppressive therapy was < 20% (85.7% of respondents).

The results of this study indicate success in maintaining remission using a rapid glucocorticoid tapering schedule with use of a secondary immunosuppressant. These data are similar to that anecdotally achieved by prolonged treatment with glucocorticoids with a presumed decrease in severity of side effects of prolonged glucocorticoid therapy.

Erythrocyte Damage and In‐Line Pressure Changes Associated with Transfusion of Canine Blood Through Microaggregate Filters

K. Cooley‐Lock1, P. Williams1, M. Williams1, S. Elder2, T. Archer1, A. Mackin1, J. Thomason1

1Mississippi State University College of Veterinary Medicine, MSU, MS, USA, 2Mississippi State University Agricultural and Biological Engineering, MSU, MS, USA

Blood transfusions are common, potentially life‐saving therapies in small animal medicine. Unfortunately, in dogs, previous studies have shown that red blood cells (RBCs) transfused via syringe pump have a significantly shorter post‐transfusion survival time compared to RBCs transfused via gravity flow. One explanation for this reported decrease in RBC survival is that, during a transfusion, there is an increase in shearing stresses that damage RBCs as blood is forced through an in‐line microaggregate filter.

The goal of this study was to determine if blood passing through a microaggregate filter, via a syringe pump, would damage canine RBC. Our hypothesis was that, over time, the filter would become clogged with fibrin and cellular debris, requiring an increase in force to maintain the preferred flow rate and thereby causing an increase in damaged RBCs.

Whole blood (50 mL) was collected from 8 healthy dogs and, using a syringe pump, blood was passed through a standard microaggregate filter at three standard rates used in the clinical setting (12.5, 25, and 50 mL/hr). Pre‐ and post‐filter blood samples were collected at the beginning and end of a simulated transfusion. Samples were analyzed to determine the erythrocyte osmotic fragility (mean corpuscular fragility [MCF]), RBC count, hemoglobin concentration, RBC distribution width (RDW), and RBC morphology. The force required to pass blood through the microaggregate filter was also measured continuously during the simulated transfusion.

Regardless of the transfusion rate, there was no change in the MCF, RBC count, hemoglobin concentration, or RDW between pre‐ and post‐filter samples. Additionally, in 5 dogs, no significant change in pressure within the filter was seen over time. Echinocytes were the most common abnormal erythrocyte morphology noted at the 25 and 12.5 mL/hr rates at the end of the simulated transfusion in both the pre‐ and post‐filter samples. Schistocytes were identified in one sample (post‐filter at 12.5 mL/hr).

Our study suggests that, regardless of the transfusion rate, the microaggregate filter does not alter the fragility of the canine RBC, but may alter the morphology. RBC morphology alterations could contribute to a decrease in RBC survival when administrating blood via a syringe pump.

Analytical Validation of a Flow Cytometric Protocol for Circulating Platelet Microparticle Quantification in Dogs

S.E. Cremer1, A.K.H. Krogh1, M.E.K. Hedström1, L.B. Christiansen2, I. Tarnow3, A.T. Kristensen1

1Department of Veterinary Clinical and Animal Sciences, University of Copenhagen, Frederiksberg C, N/A, Denmark, 2Department of Veterinary Disease Biology, University of Copenhagen, Frederiksberg C, N/A, Denmark, 3Savara Pharmaceuticals A/S, Hørsholm, Hørsholm, N/A, Denmark

Circulating platelet microparticles (PMP) are subcellular vesicles released from platelet and possibly megakaryocyte membranes. In people, altered PMP concentrations are associated with pro‐thrombotic, inflammatory, neoplastic and cardiovascular conditions, but only few canine studies exist. Blood‐processing pre‐analytically influences the PMP concentration. The study purpose was to analytically validate PMP quantification in dogs with a perspective of clinical application.

Using flow cytometry, dually CD61 and Annexin positive PMP of 0.3–1.0 μm were quantified in citrated whole blood (WB) and platelet‐poor plasma (centrifuged at 4500 g for 3 minutes). The coefficient of variance (CV) for duplicate – (intra‐assay) and parallel (inter‐assay) PMP analyses and the detection limit (DL) were calculated, accepting maximally 20% variation.

Data obtained from 13 healthy dogs demonstrated acceptable intra‐assay CV for WB (2.5%, 0.3–10.8% (median, range)) and plasma (5.7%, 0.0–11.1%) and acceptable inter‐assay CV for WB (11.1%, 3.9–21.2%) but not for plasma (24.2%, 4.1–105.0%). Plasma PMP concentrations equaled the noise level of negative controls, and antibody centrifugation (plasma protocol) markedly reduced this noise with a DL of 1560 PMP/μL. Antibody centrifugation was implemented in another seven dogs. The results demonstrated acceptable WB intra‐assay (2.4%, 1.2–6.7%) – and inter‐assay (6.0%, 0.0–12.3%) CV, except in one dog (58.2% inter‐assay CV). Unfortunately, hemolysis and/or lipidemia interfered with plasma PMP quantification in five of the seven dogs.

In conclusion, PMP concentrations were quantified reliably in WB with acceptable analytical precision, suggesting an application in clinical settings. Plasma PMP concentrations were invalid or unobtainable due to DL overlap or hemolysis and/or lipidemia, respectively.

Flow Cytometric Platelet Microparticle Quantification in Cats with Mild Cardiomyopathy

S.E. Cremer1, J. Koch, N. Graversen, A.S. Gravgaard, L. Bochsen, R. Langhorn, J. Willesen, A.T. Kristensen, L.N. Nielsen

Department of Veterinary Clinical and Animal Sciences, University of Copenhagen, Frederiksberg C, N/A, Denmark

Cardiogenic embolism (CE) in cats is a devastating condition primarily associated with cardiomyopathy. Hypercoagulability may pose a risk for thrombus formation; however, no single test can predict development of CE. Platelet microparticles (PMP) are subcellular vesicles released from platelet membranes and are associated with thrombosis in humans. The purpose of this study was to: i) analytically validate flow cytometric PMP quantification in cats and ii) in a pilot study, evaluate the concentration of procoagulant PMP in healthy cats and cats with cardiomyopathy.

The cats underwent physical examination, hematology, serum biochemistry, an extended coagulation screen, and a full cardiac examination including blood pressure measurement, electrocardiography, and echocardiography. Using flow cytometry, dually CD61 and Annexin positive PMP of 0.3–1.0 μm were quantified in citrated whole blood (WB) and platelet‐poor plasma. The coefficient of variance (CV) for duplicate – (intra‐assay) and parallel (inter‐assay) PMP analyses were calculated.

Six healthy cats and five cats with mild hypertrophic cardiomyopathy were examined. PMP concentrations could be quantified with acceptable intra‐assay CV for WB (3.8%, 0.1–12.5% (median, range)) and plasma (7.4%, 0.5–15.3%) and acceptable inter‐assay CV for WB (7.1%, 0.7–22.2%) but not for plasma (27.8%, 8.4–77.1%). The group‐wise comparisons of WB PMP concentrations demonstrated no differences between the healthy cats (362.000 PMP/uL, 224.900–677.100 PMP/uL) and the cats with mild cardiomyopathy (383.600 PMP/uL, 4.700–477.400 PMP/uL, P > 0.99).

In conclusion, WB PMP concentrations could be reliably quantified in cats in a clinical setting. The PMP concentration however, did not differ between healthy cats and cats with mild cardiomyopathy.

Diagnosis and Clinical Outcome Associated with a Leukemoid Response in Dogs: 81 cases (2006–2016)

B.W. Janacek1, L. Cohn1, M. Grobman2

1University of Missouri, Columbia, MO, USA, 2University of Missouri College of Veterinary Medicine, Columbia, MO, USA

The purpose of the study was to review cause and outcome of leukemoid response in dogs. A computerized search for dogs seen over the last 10 years with neutrophil counts ≥60,000 cells/μL at the University of Missouri was completed. Demographic data, hemogram, diagnoses and clinical outcome were extracted from records. Based on recorded diagnosis, leukemoid response was deemed to be due to focal or systemic infection or inflammation, immune‐mediated disease, neoplastic disease, tissue damage/necrosis, or cause undetermined.

A total of 81 dogs were identified, including 43 male (9 intact, 33 castrated, 1 of unknown status) and 38 female (8 intact, 30 spayed) dogs with a mean age of 6.8 years (minimum 0.17, maximum 15). Nine were mixed breed, and German shepherd dogs (7/81; chi square P = 0.035) were over‐represented. The median neutrophil count was 70,940 cells/μL (IQR +/‐ 24,080; maximum 197,680 cells/μL). A definitive diagnosis was available for 56 dogs, and a presumptive diagnosis for 17. Forty‐two (51.9%) had infection/inflammation, 22 (27.2%) neoplastic disease (7 lymphoma or leukemia), 5 (6.2%) immune‐mediated disease, 4 (4.9%) tissue damage, and 8 dogs (9.9%) went undiagnosed. Forty‐four dogs survived to discharge, 27 were euthanized, 10 died in hospital. The median segmented neutrophil counts did not differ between disease group based on Kruskal‐Wallis One Way Analysis of Variance on Ranks. Similarly, survival was not related to neutrophil count.

Magnitude of leukemoid response alone may not predict disease cause or prognosis.

Immediate Post‐Transfusion Survival of Biotinylated Canine Red Blood Cells with Various Transfusion Techniques

P.Y. Martin1, B. Thames2, C. Mulligan2, J. Thomason2, A. Mackin2

1Tuskegee University College of Veterinary Medicine, Tuskegee, AL, USA, 2Mississippi State University College of Veterinary Medicine, Mississippi State, MS, USA

The use of infusion or syringe pumps during canine blood transfusion is controversial, because pumps can potentially damage red blood cells (RBC), rendering the transfusion less effective. A previous study in healthy dogs reported a complete absence of detectable biotinylated RBCs 24 hours post‐transfusion in many animals when blood was given through syringe or infusion pumps. Complete loss of transfused RBCs within one day of transfusion is, however, inconsistent with anecdotal clinical experiences.

The objectives of our current study were to confirm whether a number of different transfusion methods (one syringe and one infusion pump brand, compared to gravity drip) resulted in premature loss of transfused cells, and to explore the timing of early RBC loss if it occurred. Eight healthy hound dogs were phlebotomized (54 ml of whole blood in CPDA‐1), and blood was centrifuged, with supernatant plasma collected and stored. Remaining RBCs were incubated with NHS‐biotin dissolved in 100% DMSO, and then washed in saline. The RBCs were then resuspended in the stored autologous plasma, and 5 mL was retained as a positive biotinylated control for each sample, with the remainder transferred to either a 100 mL dry blood bag (infusion pump and gravity drip) or 60 mL syringe (syringe pump). Assigned transfusion treatments were as follows: gravity drip (2 dogs), infusion pump (Heska Vet/IV 2.2; 3 dogs), and syringe pump with an in‐line filter (Medfusion 2001; Hemo‐Nate Blood Filter; 3 dogs). Autologous whole blood transfusion was performed at a rate of 30 mL per hour. Blood samples were collected in sodium citrate anticoagulant at time 0 (pre‐transfusion, used as negative non‐biotinylated control), and at hours 1, 6, 12, 18, and 24, and days 2, 3, and 4 post‐transfusion. Samples were labeled with phycoerythrin‐streptavidin, and percentage of biotinylated RBCs analyzed by flow cytometry with RBCs gated according to their forward and side light scatter at 10,000 gated events per sample.

Immediate post‐transfusion (1 hour) percent biotinylated RBCs (percent of total RBCs; mean ± SEM) was 1.43 ± 0.23% (gravity), 1.45 ± 0.05% (infusion) and 1.55 ± 0.10% (syringe), consistent with minimal loss of RBCs during transfusion, since estimated post‐transfusion mean target percent biotinylated RBCs was 1.12% (range 1.02–1.24%). Over the first 24 hours, and the subsequent 3 days, there was no significant difference between transfusion techniques, with approximately 1.5% of biotinylated cells present up to 4 days following transfusion (Figure 1).

Our study found acceptable post‐transfusion survival of labeled RBCs regardless of method, a finding that is discrepant with the prior study documenting unmeasurable levels of biotinylated RBCs one day after transfusion using various pumps.

Subsequent exploration of reasons for discrepant results suggests several possibilities. DMSO is the standard solvent used for biotin, since biotin is insoluble in aqueous solutions. Biotin precipitates in DMSO solutions of 50% or less, and high DMSO concentrations are therefore needed for biotin labeling studies. Inadvertent dilution of DMSO in residual plasma during RBC labeling therefore has the potential to render biotinylation ineffective. Additionally, we found that different concentrations of DMSO cause variable lysis of RBCs: 60% DMSO (diluted in PBS) caused marked hemolysis, 50 and 100% caused moderate hemolysis, 30, 40 and 90% caused mild hemolysis, and 10, 20, 70 and 80% caused no hemolysis. Our findings suggest that transfusion studies using biotinylated RBCs may be susceptible to technique artifacts, with subtle variations in DMSO concentrations leading to either failure of biotinylation or inadvertent RBC damage.

Markers of Oxidative Damage during Storage of Feline Packed Red Blood Cells

J.M. Reinhart, J.M. Walker

Department of Medical Sciences, School of Veterinary Medicine, University of Wisconsin‐Madison, Madison, WI, USA

Biochemical and immunologic changes consistent with erythrocyte storage lesion have been described in feline packed red blood cell (pRBC) units. Oxidative damage occurs in stored human erythrocytes as evidenced by decreased glutathione concentrations, but this has yet to be investigated in cats. The purpose of this study was to determine the effect of storage on glutathione, osmotic fragility, and methemoglobin reductase activity in feline pRBCs. Whole blood (60 mL) was collected from five healthy cats into blood bags containing CPDA‐1. Each bag was split in half and pRBC units were prepared. For each cat, additive solution 3 (AS‐3) was added to one of the units and pRBCs were stored at 4°C. Aliquots of each unit were sampled on days 1, 14, 28, 35, and 42 of storage and assayed for glutathione, osmotic fragility, and methemoglobin reductase activity. Friedman's test was used to assess changes over time. Glutathione decreased significantly with time in both AS‐3‐treated and untreated pRBCs. Osmotic fragility increased significantly with time in untreated pRBCs, but did not change from baseline in pRBCs treated with AS‐3. Methemoglobin reductase activity was unchanged in both treated and untreated samples. Storage of feline pRBCs over 42 days leads to glutathione depletion and increased osmotic fragility. These findings are further evidence for storage lesion in feline pRBCs. However, the addition of AS‐3 mitigates the some of these changes and should be considered in the development of future blood product protocols.

Adherence to Human Guidelines for Plasma Transfusions in Dogs and Cats: 154 Cases

T. Taney1, E.A. Rozanski2

1Cummings School of Veterinary Medicine, Tufts University, South Grafton, MA, USA, 2Cummings School of Veterinary Medicine, Tufts University, Grafton, MA, USA

Fresh frozen plasma is used to treat a variety of conditions in critical illness. Plasma transfusion has the potential for causing adverse reactions and is a limited resource. There are currently no evidence‐based plasma transfusion recommendations in dogs and cats. These include documented coagulopathy with bleeding or a planned invasive procedure (DC), warfarin reversal (WR), massive transfusion (MT), and factor deficiency (FD). The aim of this study was to determine if plasma transfusions in animals met guidelines used in human medicine.

The medical record system was retrospectively searched for dogs and cats that had received plasma transfusion. The indication (s) for the transfusion recorded. The percentage in each category was recorded, as well as those cases with no clear indication. Survival was recorded.

Dogs (n = 136) and cats (n = 18) were identified. Seventy‐six percent (n = 117/154) of patients had plasma transfusions that could be classified under an indication in the proposed guidelines for human patients. 15 patients had indications that could be classified into multiple categories. The most common indications were DC at 61% (n = 94/154), MT at 17% (n = 27/154), WR at 5.8% (n = 9/154), and FD at 1.2% (n = 2/154). Twenty‐four percent (n = 37/154) of animals were administered plasma for other reasons, most frequently hypoproteinemia. Survival to discharge was 48% for dogs (n = 65/136) and 44% for cats (n = 8/18).

The majority of plasma transfusions were considered indicated based guidelines for human patients, however, 24% of transfusions were not considered indicated. Similar guidelines should be considered for companion animals.

Antiplatelet Effects of Generic Clopidogrel Compared to Trade Name Clopidogrel (Plavix®) in Cats

M. Malkawi1, A. Woolcock2, D. Hogan3

1Purdue University, West Laffayete, IN, USA, 2Purdue University College of Veterinary Medicine, West Lafayette, IN, USA, 3Purdue University/VTH, West Lafayette, IN, USA

Cats with hypertrophic cardiomyopathy (HCM) are at an increased risk for formation of an aortic thromboembolism (ATE) due to blood stasis and platelet hyper‐reactivity in enlarged cardiac chambers, among other prothrombotic factors. Anti‐platelet medications are commonly used in cats considered at risk of developing thromboembolism. Clopidogrel bisulfate, a thienopyridine drug, causes irreversible inhibition of platelet aggregation in response to ADP and collagen, common initiators of clot formation. Trade name clopidogrel (Plavix®) was previously investigated in healthy cats and was associated with a significant inhibition of platelet activity. In 2012, the FDA approved the generic form of clopidogrel, but its inhibitory effects on platelet aggregation have not been investigated. We hypothesized that generic clopidogrel is associated with equal anti‐platelet effects to those associated with trade name clopidogrel in healthy cats, and that both formulations are associated with minimal oral/gastrointestinal adverse events.

In this randomized, double‐blind, crossover study, five skeletally mature cats were determined to be healthy based on physical examination and diagnostics including CBC, serum chemistry profile, urinalysis, and coagulation times. The cats participated in three treatment phases in which they were randomly assigned to receive either: trade name clopidogrel 18.75 mg PO SID for 7 days, generic clopidogrel 18.75 mg PO SID for 7 days, or a placebo (cellulose) PO SID for 7 days. On day 0, prior to starting the treatment, baseline platelet aggregation was assessed on a multiplate platelet aggregometer. Briefly, blood was collected into a lithium heparin tube, and platelets were activated for the assay using ADP. Platelet aggregation was repeated on day 8 to determine the percentage of platelet inhibition achieved. Information regarding potential adverse effects were recorded daily for each cat during each phase. A minimum four‐week washout period occurred between each phase.

All cats maintained a normal platelet count at each time point of the study. No adverse effects were noted with the administration of trade name or generic clopidogrel in any cat. Two cats reliably showed significant inhibition of platelet aggregation in response to both trade name and generic clopidogrel (full responders). One cat showed slight inhibition of platelet aggregation in response to generic clopidogrel, but not trade name clopidogrel (partial responder). Two cats did not show any inhibition of platelet aggregation in response to both trade name and generic clopidogrel (non‐responders).

While we determined that both the trade name and generic formulations of clopidogrel are safe to administer in cats, there was not reliable inhibition of platelet aggregation in all cats. Cats tended to have a similar response to both formulations of clopidogrel implying that intrinsic factors affect the efficacy of the drug, as opposed to human or assay error. This variable response to treatment may be due to alterations in the cytochrome P450 gene coding for enzymes responsive for clopidogrel metabolism, as several polymorphisms have been identified in humans. This is the first study to investigate the effect of the generic formulation of clopidogrel on platelet aggregation, but due to the low sample size, meaningful conclusions cannot be made about the varying response to the drug. Further studies should aim to increase sample size, and investigate metabolic, genetic, and other intrinsic factors that could affect drug efficacy.

In Vivo Effects of 7.5% Hypertonic Saline Solution on Canine Whole Blood Coagulation Using Thromboelastography

H. Kim1, S. Baek1, A. Nam2, K.W. Seo3

1College of Veterinary Medicine, Chungnam National University, Daejeon, Taejon‐jikhalsi, Republic of Korea, 2Department of Veterinary Internal Medicine, College of Veterinary Medicine, Seoul National University, Seoul, Seoul‐t'ukpyolsi, Republic of Korea, 3Chungnam National University, Daejeon, Taejon‐jikhalsi, Republic of Korea

The objective of this study is to evaluate the effect of 7.5% hypertonic saline solution (HSS) on whole blood coagulation in healthy dogs and to compare the effects of in vivo and in vitro dilution of HSS.

This prospective, controlled, experimental study included twelve healthy beagles. All twelve dogs were included in the treatment group with five of them randomly selected to serve as the control group. The dogs in the treatment group (n = 12) were given a 5 mL/kg 7.5% HSS IV over 1 mL/kg/min. Following a 2‐week washout period, the dogs in the control group (n = 5) received 5 mL/kg 0.9% normal saline solution. Subsequently, 4‐mL blood samples were collected at 0, immediately after infusion, 30, 60 and 90 min after infusion for thromboelastography (TEG) measurement.

The in vitro study was performed with diluted samples, and 2‐mL blood samples were collected at baseline and diffused with 7.5% HSS at a ratio of 1:18, which was equivalent to 5 mL/kg infusion in vivo situation.

Measurements using ANOVA were carried out repeatedly for analysis of the in vivo study, and there were no statistically significant changes (P < 0.01) in any TEG variables between the treatment group and control group. For comparison between in vivo and in vitro studies, paired t‐tests were used.

The comparison between in vitro and in vivo dilution was performed only at a point of time immediately after infusion, which is most similar situation with in vitro dilution because HSS is immediately diffused in vivo.

There were significant changes in maximum amplitude (P = 0.04) and G values (P = 0.04), suggesting that HSS affects whole blood coagulation differently, according to the study design. Unlike in previous in vitro studies, which revealed that 7.5% HSS has anticoagulant effects, a clinically relevant dose (5 mL/kg) of 7.5% HSS had no effect on whole blood coagulation when compared with 0.9% normal saline in vivo.

In Vitro Inhibition and Detection of Canine Complement Activation in Canine Immune‐Mediated Hemolytic Anemia

D.H. Muguiro1, E. Behling‐Kelly, R. Goggs

Cornell University, College of Veterinary Medicine, Ithaca, New York, NY, USA

The purpose of this study is to evaluate the ability of two human complement inhibitors to inhibit canine complement‐mediated hemolyisis in vitro and to validate the use of an ELISA for detection of canine complement activation in dogs.

Human complement inhibitors, C3‐inhibitor (Compstatin) and human recombinant C1‐esterase Inhibitor (C1‐INH) were evaluated using an in vitro hemolytic assay. Antibody‐coated sheep erythrocytes were exposed to canine complement with and without inhibitors. Dose curves were generated by plotting the percentage of hemolysis against inhibitor concentration. Complement activation in samples from 11 healthy and 13 dogs with IMHA was determined using an ELISA for detection of Terminal Complement Complex C5b‐9 (TCC C5b9). The concentration of TCC C5b9 in patient samples was extrapolated from a standard curve generated using the supplied protein standard. Samples treated with snake venom were used as a positive control for complement activation.

C1‐INH showed dose‐dependent inhibition of canine complement‐mediated hemolysis. Compstatin was minimally effective. TCC C5b9 concentrations tended to be higher in control samples compared to dogs with IMHA.

C1‐INH inhibited canine complement activation in vitro. Compstatin showed minimal inhibition of canine complement‐mediated hemolysis at the doses evaluated. The ELISA tested did not consistently detect complement activation.

Novel Thrombogenicity Examination of Whole Blood: Total Thrombus‐Formation Analysis System in Dogs

T. Iwanaga1, S. Yamada2, R. Fukushima3, T. Nagasato1, I. Maruyama1, N. Miura1

1Kagoshima University, Kagoshima, Kagoshima, Japan, 2Animal Cardiovascular and Thoracic Surgery, Fuchuu, Tokyo, Japan, 3Tokyo University of Agriculture and Technology, Fuchuu, Tokyo, Japan

Blood clotting is controlled by platelet activation and the coagulation system. Virchow's three conditions must be met to form a clot in vivo. However, current blood coagulation examinations are fragmentary. The Total Thrombus‐formation Analysis System (T‐TAS) is a new laboratory procedure to evaluate the blood clotting ability in which whole blood flows into a microflow chamber imitating blood vessels.

T‐TAS utilizes two microchips: the PL‐chip and AR‐chip. The PL‐chip predominantly evaluates platelet function, and the AR‐chip evaluates both platelet activity and coagulation simultaneously. The blood clotting ability is evaluated by measuring the internal pressure in the chip, which rises with clot formation. The measurement values are the occlusion start time (OST: time to reach 10 kPa), occlusion time (OT: time to reach 80 kPa), and area under the pressure curve (AUC). In this study, we assessed the reproducibility of each chip in 8 healthy dogs, established the reference range in 54 dogs (PL, n = 19; AR, n = 35), and confirmed the effect of aspirin in 3 dogs.

The reproducibility of each chip was confirmed based on the coefficient of variation (CV) (PL: CVOST=8.69%, CVOT=13.58%, CVAUC=6.45%; AR: CVOST=10.01%, CVOT=7.65%, CVAUC=1.57%). The reference ranges of the OST and OT for the PL‐chip (average ± standard deviation) were 2.26 ± 0.81 and 4.73 ± 1.21 min, respectively (AUC, 406.77 ± 54.36), and those for the AR‐chip were 5.02 ± 1.83 and 6.04 ± 2.14, respectively (AUC, 1955.35 ± 156.67). In all dogs, the graph shifted to the right for the PL‐chip after aspirin administration, but not for the AR‐chip; thus, apparent extension of PL was observed, indicating successful evaluation of platelet suppression by aspirin.

The results of this study indicate that evaluation of blood coagulation in dogs using the T‐TAS is useful and that the blood clotting ability in dogs with various diseases might be evaluated more subtly by T‐TAS in the future.

Vaccination Behavior and Adverse Events in Dogs Treated for Primary Immune‐Mediated Hemolytic Anemia

A.K.B. Moon1, J. Veir2

1Oregon State University, Corvallis, OR, USA, 2Colorado State University, Fort Collins, CO, USA

The purpose of this study was to describe the frequency with which dogs treated for primary immune‐mediated hemolytic anemia (IMHA) at veterinary teaching hospitals are vaccinated after diagnosis, and investigate whether an association exists between vaccination and vaccine reactions (including but not limited to relapse of IMHA) in this population. A retrospective descriptive observational study was performed involving 66 dogs at 2 veterinary teaching hospitals: Colorado State University (57) and Oregon State University (9). Medical records databases were searched for cases meeting inclusion and exclusion criteria for primary IMHA. Owners and referring veterinarians of identified dogs were contacted with a survey via postal mail to gather information regarding vaccination of the dog following diagnosis. Referring veterinarians were contacted by phone regarding vaccination history for each dog prior to diagnosis of IMHA. At least one survey was returned in 44 cases. Results indicated 22 dogs did not receive vaccinations after diagnosis, and 21 dogs did receive vaccinations after diagnosis. For the remainder of dogs, no survey was returned or the survey was excluded for inconsistency of information. Of the vaccinated dogs, only 2 possible vaccination reactions were reported (vomiting and urticaria in 1 dog, and worsening anemia in 1 dog). A date of vaccination prior to diagnosis was available for 46 dogs. The mean number of days between vaccination and initial diagnosis of IMHA was 351 days. These results indicate that vaccination occurred among nearly half of dogs diagnosed with IMHA, with rabies being the most common vaccine administered (19 of 21 dogs). This may reflect the legal requirement of rabies vaccination in most locations. Adverse events were uncommon overall, but comparison to the rate of adverse events in the general population was not performed. Finally, this study did not document a temporal relationship between vaccination and initial diagnosis of IMHA, although comparison to a control population is necessary to confirm this finding.

Evaluation of Thromboelastography in Dogs with Heartworm Infection

S. Park1, Y. Kim2, C. Park3, J. Kim4

1College of Veterinary Medicine, Chonbuk National University, Iksan, Cholla‐bukto, Republic of Korea, 2College of Veterinary Medicine, Chonbuk National University, Curi, Kyonggi‐do, Republic of Korea, 3College of Veterinary Medicine, Chonbuk National University, Iksan, Cholla‐bukto, Republic of Korea, 4College of Veterinary Medicine, Konkuk National University, gwangjin‐gu, Seoul‐t'ukpyolsi, Republic of Korea

Canine heartworm disease due to Dirofilaria immitis is an important disease causing hemostatic disorders. Although there are several studies showing the relationship between coagulation pathway and heartworm infection, most of studies have used conventional methods of laboratory assessment of hemostasis. There is a limitation in using conventional coagulation test to evaluate hypercoagulable state in heartworm infected dogs. Thromboelastography (TEG) provides a more truthful reflection of in vivo hemostasis compared to conventional coagulation tests in the diagnosis and monitoring of patients. In this study, we describe TEG findings in dogs with heartworm infection and compare their coagulation status as determined by TEG to clinical presentation, heartworm grade, and conventional coagulation tests.

This study consisted of fifteen healthy dogs and twenty‐five dogs with heartworm infection. Heartworm infection was diagnosed using commercial ELISA test kit. Dog with heartworm infection were graded to four groups according to blood work, radiography, echocardiography. Platelet count, prothrombin time, activated partial thromboplastin time, plasma D‐dimer concentration, plasma fibrinogen concentration, and thromboelastography were performed in the all dogs. In the results, the value of alpha showed a tendency to increase as the grade increased (table 1, P = 0.03), which indicates higher level of fibrinogen activation in more advanced grades. In the rank coefficient correlation between conventional coagulation tests and TEG, platelet had a significantly positive coefficient correlation with MA and G value in heartworm infected grade 1. An interesting result is that a dog diagnosed femoral arterial thromboembolism showed significantly hypercoagulable state in TEG, compared with conventional tests (figure 1). Also, the serial TEG analysis of a dog with disseminated intravascular coagulation represented more hypocoagulable state over time (figure 2). This study suggests that application of TEG is helpful for diagnosis and treatment of heartworm infection in dogs.

Characteristics of Hemostasis During Experimental Ehrlichia Canis Infection in Laboratory Beagles

S. Shropshire, C. Olver, M.R. Lappin

Colorado State University, Fort Collins, CO, USA

Ehrlichia canis infection in dogs can manifest with clinical signs related to bleeding and also commonly causes thrombocytopenia. The exact mechanisms of bleeding and thrombocytopenia are unknown but are thought to be related to processes such as vasculitis and immune and non‐immune processes on platelets. It is also unknown why some dogs show signs of bleeding while other dogs do not despite significant thrombocytopenia. It is possible that platelets become activated during infection and/or blood clots become resistant to fibrinolysis which could prevent a bleeding phenotype. The purpose of this study was to assess the platelet indices of activation, platelet function via whole blood impedance platelet aggregometry, percentage of immunoglobulin associated platelets (percent IgG), and thromboelastography (TEG) measurements including velocity curve variables in dogs experimentally infected with E. canis.

Six healthy laboratory beagles were experimentally infected with E. canis via needle inoculation. Four dogs were inoculated from one naturally infected dog and two dogs were inoculated from a second naturally infected dog. Four dogs became positive for E. canis via serology and PCR but the remaining two dogs never became positive despite repeated testing. As a result, the four dogs were designated as positive and the two dogs were designated as negative. Platelet indices; platelet count, mean platelet volume (MPV), plateletcrit (PCT), platelet volume distribution width (PDW), mean platelet mass (MPM), mean platelet component concentration (MPC) and platelet component distribution width (PCDW) were measured. For platelet aggregometry, the area under the curve (AUC) was recorded using a saline control (AUCsaline) and for each agonist (adenosine diphosphate (ADP) [AUCADP] and arachidonic acid (AA) [AUCAA]). Percent IgG was assessed by direct flow cytometry. Tissue factor (TF)‐activated TEG and TF‐activated TEG + tissue plasminogen (tPA) were performed. TEG variables (reaction time (R), clotting time (K), rate of clot formation (α), and maximum amplitude (MA), LY30 (percent clot lysis 30 minutes after MA), CL30 (clot lysis 30 minutes after MA), LY60 (percent clot lysis 60 minutes after the MA), and CL60 (clot lysis 60 minutes after MA)) were recorded in addition to the velocity curve variables: maximal rate of thrombus generation (MRTG), time to maximum rate of thrombus generation (TMRTG), total thrombus generated (TG), maximum rate of lysis (MRL), time to maximal rate of lysis (TMRL), and total lysis (L). Testing was performed in both the positive and negative dogs prior to inoculation and then on weeks 1, 2, 3, 4, and 5. Due to clinical signs (weight loss, hyporexia, lethargy, fever), the positive dogs were administered doxycycline at 10 mg/kg, PO, daily between week 3 and week 4. For each variable, a repeated measures analysis (2‐way ANOVA) with post‐hoc analysis was performed comparing the positive and negative dogs. For all analysis, P < 0.05 was considered significant.

There were no significant differences between groups (positive vs. negative) for the variables PDW, MPC, PCDW, AUCsaline, R, K, α, CL60, TMRTG and MRL. During infection, the platelet count, PCT, AUCADP, AUCAA, LY30, and LY60 were significantly lower and the MPV, MPM, percent IgG, MA, CL30, MRTG, TG, TMRL, and L were significantly higher in the positive group. No overt signs of bleeding were noted during the study.

The positive dogs had significantly lower platelet counts but had evidence of activated platelets based on an increased MPV and MPM. There was evidence of decreased platelet function in the positive dogs but this could have been related to the degree of thrombocytopenia. The positive dogs also showed evidence of anti‐platelet antibodies during infection. Overall, the positive dogs appeared hypercoagulable based on an increased MA, MRTG, and TG and hypofibrinolytic based on a lower LY30, LY60 and higher CL30 and TMRL. The characteristics of hemostasis during experimental E. canis infection are complicated but platelet indices of activation and evidence of a hypercoagulable, hypofibrinolytic state may explain the lack of a bleeding phenotype in some dogs despite significant thrombocytopenia.

Whole Blood impedance Platelet Aggregometry in Healthy Greyhounds Compared to Healthy Beagles

S. Shropshire, C. Olver

Colorado State University, Fort Collins, CO, USA

Whole blood impedance platelet aggregometry can be affected by hematocrit and platelet count in humans. We sought to evaluate whether this phenomenon exists in dogs by comparing blood from Greyhounds, a breed known to have high hematocrits and low platelet counts, to Beagles, a breed with “normal” hematocrits and platelet counts. We hypothesized that aggregometry in Greyhounds may result in erroneous interpretations of platelet function. The purpose of this study was to compare whole blood impedance platelet aggregometry and other hemostatic assays in healthy Greyhounds to healthy Beagles.

Whole blood impedance platelet aggregometry and tissue factor‐activated thromboelastography (TF‐TEG) were performed in six healthy client‐owned Greyhounds and six healthy laboratory Beagles. For platelet aggregometry, the area under the curve was recorded for each agonist (adenosine diphosphate (ADP) and arachidonic acid (AA)) and a saline control. For TF‐TEG, reaction time (R), clotting time (K), rate of clot formation (α), maximum amplitude (MA), percent clot lysis 30 minutes after MA is reached (LY30), amount of clot lysis 30 minutes after MA is reached (CL30), percent clot lysis 60 minutes after the MA is reached (LY60), and amount of clot lysis 60 minutes after MA is reached (CL60) were recorded. The hematocrit, platelet count, and fibrinogen were also recorded. An unpaired t‐test was performed on ADP, AA, hematocrit, platelet count, fibrinogen, R, α, MA, and a Mann‐Whitney U test was performed on the saline control, K, LY30, CL30, LY60, and CL60. Pearson correlation was also assessed between hematocrit and ADP, AA, and MA, platelet count and ADP, AA, and MA, and fibrinogen and MA. For all tests, P < 0.05 was considered significant.

The ADP, AA, platelet count, fibrinogen, MA, CL30, and CL60 were significantly lower and the hematocrit, LY30, and LY60 were significantly higher in the Greyhounds. There were no statistical differences for the saline control, R, K, and α between groups. In all twelve dogs, there was correlation between hematocrit and MA (r = −0.70), ADP (r = −0.91), and AA (r = −0.81), and platelet count and ADP (r = 0.9), AA (r = 0.72), and MA (r = 0.68) and fibrinogen and MA (r = 0.85).

In agreement with previous studies, healthy Greyhounds had significantly higher hematocrits and lower platelet counts. Based off of the TEG tracings, Greyhounds appeared significantly hypocoagulable (decreased MA) and hyperfibrinolytic (increased LY30, LY60, and decreased CL30, CL60). Future studies should investigate whether hyperfibrinolysis contributes to the propensity to bleed in Greyhounds. The platelet aggregometry results were significantly lower for both agonists in Greyhounds as compared to Beagles which could be erroneously interpreted as decreased platelet function in this breed. Therefore, just as hematocrit and platelet count can affect TEG tracings, the same considerations should be applied to whole blood impedance platelet aggregometry. Both the high hematocrit and low platelet counts in healthy Greyhounds can potentially contribute to lower aggregometry readings due to suspected in vitro effects rather than reflecting true in vivo effects since other studies have demonstrated normal platelet function in Greyhounds. Therefore, it is not recommended to compare whole blood impedance platelet aggregometry measurements of Greyhounds to non‐Greyhounds. Subject‐based or population‐based reference intervals should be established in this breed for accurate interpretation of results.

Thawing Times and Hemostatic Proteins in Canine Fresh Frozen Plasma Using a Commercial Thawing Device

L. Torkildsen1, M. Pashmakova1, M. Bishop2, J. Barr1

1Texas A&M University, College Station, TX, USA, 2WAVE Veterinary Internal Medicine, Naples, FL, USA

The objectives of this study were to compare thawing times between a 37 °C warm water bath (WWB), running water bath (RWB), and Plasmatherm and compare hemostatic protein function in canine fresh frozen plasma (FFP) thawed with WWB or Plasmatherm.

Thawing time was measured for a 240 mL bag of FFP thawed in WWB, RWB, or Plasmatherm 10 times. Fresh canine donor plasma samples were split into 120 mL bags and 3 mL control aliquots prior to freezing. Bags were thawed by WWB or Plasmatherm and aliquots equilibrated to room temperature. Concentrations of hemostatic proteins, albumin, D‐dimers, prothrombin time (PT) and partial thromboplastin time (PTT) were obtained. Repeated measures ANOVA, or Friedman, as appropriate with corresponding post tests were used to compare groups. Alpha was set to 0.05.

There was an overall difference in thawing times between the three groups (P = 0.0274). Post testing revealed significantly shorter thawing time with RWB as compared to Plasmatherm (P = 0.0417). A significant difference in PTT was noted overall (P < 0.001) and post testing revealed longer PTT in Plasmatherm and WWB groups compared to control (P = 0.0010, P = 0.0052, respectively). There was an overall significant difference in factor VII concentration (P = 0.0007) and post testing revealed decreased factor VII in Plasmatherm and WWB groups compared to control(P = 0.0010, P = 0.0041, respectively). A significant difference in factor X concentration was seen overall (P = 0.0317) and post testing revealed decreased factor X in Plasmatherm compared to control (P = 0.0417).

The RWB had the fastest thawing time. Significant differences were detected in hemostatic protein function but these differences are not considered clinically significant.

Clinical Evaluation of the QuickVet/RapidVet Feline Blood Typing Test

C. Weingart, J. Assmann, B. Kohn

Clinic of Small Animals, Faculty of Veterinary Medicine, Freie Universität Berlin, Berlin, Berlin, Germany

Blood transfusions have become an integral part of feline intensive care treatment. For patient safety blood typing is an important pretransfusion test.

Objective of this prospective study was evaluation of an automated feline cartridge blood typing method (QuickVet®/RapidVet®(QVRV) SMB, Farum, Denmark) and comparison of results with slide and tube (reference) methods (not commercially available). Samples with discordant results were additionally tested with Alvedia® (Quick Test A+B, Alvedia, Limonest, France). To determine precision, blood typing of 3 A and B samples were repeated 5 times within 24 hours and over 5 consecutive days.

83 EDTA‐anticoagulated blood samples routinely collected for diagnostic purposes and 16 provided by outside laboratories were available. (One sample with severe hemolysis could not be typed with QVRV.) Slide/tube tests yielded the same results on all 98 samples. 78 samples were blood type A, 14 B and 6 AB. QVRV identified 75/78 of the A and 14/14 of the B samples. Three A samples with low hematocrit (≤ 10%) could only be typed after diluting them. Six AB samples gave discordant results compared to the reference method: 1 AB sample was typed as blood type A with QVRV (and Alvedia®) and 5 as B with QVRV (and AB with Alvedia®). All repeats yielded the same results. Based on all samples run, QVRV agreed in 90.8% with the reference method (exclusively for A and B cats, the agreement was 96.5%). Blood samples with type B should be confirmed by back typing and samples with type AB by tube method.

Characterization of Thrombocytosis in 715 Dogs (2011–2015)

A. Woolcock, A. Keenan, C. Cheung, G.E. Moore

Purdue University College of Veterinary Medicine, West Lafayette, IN, USA

Thrombocytosis is an uncommon hematologic abnormality in canine patients that has been associated with various neoplastic, metabolic, and inflammatory conditions. The aim of this study was to characterize thrombocytosis in dogs by its severity and the diseases to which it is associated.

Medical records of all dogs with thrombocytosis between 2011 and 2015 were reviewed. The thrombocytosis for each patient was characterized as primary or secondary, and then further classified by severity and diagnosis. Diagnostic categories included neoplasia, endocrine disease, inflammatory disease, or miscellaneous. Inflammatory disease was further classified by the body system(s) involved. Serum and plasma potassium concentration, if available, were evaluated for the possibility of pseudohyperkalemia.

A total of 1254 complete blood counts with thrombocytosis from 715 dogs were included in the study. Median platelet count in this population was 582 x 103 / μL (500–1810 x 103 / μL). No correlation between severity of thrombocytosis and diagnosis was identified. Secondary causes included neoplasia (55.7%), endocrine disease (12.0%), and inflammatory disease (46.6%). Of the dogs with neoplasia, carcinoma (52.0%) and lymphoma (27.2%) were the most common diagnoses. Immune‐mediated disease was common (22.2%), associated with frequent glucocorticoid administration, and had a significantly higher median platelet count when compared to the other body systems (P < 0.001).

Thrombocytosis is commonly associated with neoplasia and immune‐mediated disease diagnoses in dogs. Further research is warranted to investigate the role of thrombocytosis in hepatobiliary and renal disease.

Biomarkers of Oxidative Stress and Redox Status of the Liver in Dogs with Liver Disease

C. Barry‐Heffernan1, J.L. Ekena2, S. Dowling3, K. Viviano1

1University of Wisconsin, Madison, WI, USA, 2Department of Medical Sciences, School of Veterinary Medicine, University of Wisconsin‐Madison, Madison, WI, USA, 3Pet Emergency Treatment and Specialties, Inc., Lancaster, PA, USA

Oxidative stress plays a significant role in canine liver disease. The main intracellular antioxidant, reduced glutathione (rGSH), is decreased in dogs with naturally occurring liver disease. It is unknown whether circulating rGSH levels correlate with liver concentrations. The aims of this study were to determine whether circulating rGSH correlates with liver rGSH in dogs with naturally occurring liver disease and to evaluate whether other markers of systemic oxidative stress, plasma vitamin E and urine 8‐isoprostanes/creatinine (8‐IsoP), correlate with liver rGSH.

Dogs with liver disease presenting for liver biopsies as part of their clinical evaluation were included. Urine and blood were collected within 24 hours of liver biopsy. Plasma, erythrocyte, and liver rGSH were measured using high performance liquid chromatography (HPLC); vitamin E was measured by HPLC, and 8‐IsoP was measured by gas chromatography/mass spectrometry.

Thirty‐one dogs, of the targeted enrollment of 34, have completed the study. The median rGSH in the liver, plasma, and erythrocytes were 5.78 mM, 3.90 μM, and 1.88 mM, respectively. No correlations were found between liver rGSH and erythrocyte rGSH, vitamin E, or 8‐IsoP. There was a modest negative correlation between liver and plasma rGSH (ρ – 0.40; P 0.043). Dogs with liver disease had significantly higher urine 8‐IsoP compared with healthy controls (5885 vs. 2980 pg / mg Cr; P < 0.0001). The initial results of this study demonstrate that in dogs with naturally occurring liver disease only plasma rGSH correlates with liver rGSH, albeit negatively. In addition, dogs with naturally occurring liver disease have evidence of increased systemic oxidative stress compared to healthy controls.

Canine Cholangitis/Cholangiohepatitis: A Descriptive Study in 54 Dogs (2004–2014)

J.L. Harrison1, B. Turek2, D.C. Brown3, C.W. Bradley1, J.C. Clark4

1University of Pennsylvania School of Veterinary Medicine, Philadelphia, PA, USA, 2Antech Diagnostics, Irvine, CA, USA, 3University of Pennsylvania, Philadelphia, PA, USA, 4New England Animal Medical Center, West Bridgewater, MA, USA

Canine cholangitis appears to be more common than previously thought; yet our understanding of the disease remains incomplete. The study aim was therefore to describe a population of dogs with biopsy confirmed cholangitis.

Medical records and biopsies of fifty‐four dogs with cholangitis or cholangiohepatitis confirmed via histopathology between January 2004 and December 2014 were retrospectively reviewed.

Presenting clinical signs included vomiting (72.2%), lethargy (70.4%) and inappetence (64.8%). The majority of dogs (49/50) had liver enzyme elevations, hyperbilirubinemia (32/50) and hypercholesterolemia (24/43). Ultrasonographic abnormalities of the liver and/or biliary tree were seen in 84% of cases. Histopathology revealed that 53/54 cases had neutrophilic cholangitis/cholangiohepatitis while 1 case had lymphocytic cholangitis. The majority (42/54) were chronic. Histopathology revealed concurrent gallbladder disease in 46.2% of cases and biliary tract obstruction in 42.6% of cases. Seventeen of 36 biliary and 11/25 liver cultures were positive for bacterial growth; Escherichia coli and Enterococcus spp. were most common. The median patient survival was 671 days (CI 95% 114, 1426). Controlling for age, dogs that did not have a cholecystectomy had a 2.1 greater hazard ratio for death (P = 0.001, CI 95% 1.0, 4.3) compared to dogs that had a cholecystectomy. Controlling for cholecystectomy, dogs greater than 13 years of age had an increased risk of death (P = 0.037, CI 95% 1.9, 13.2) compared to those less than 13 years.

In this population of dogs, chronic neutrophilic cholangitis/cholangiohepatitis was most common. Gallbladder disease, specifically cholecystitis and biliary tract obstruction, occurred often in conjunction with cholangitis. In this study dogs with cholangitis and biliary tract disease that underwent cholecystectomy had a reduced risk of death, thus cholecystectomy may improve patient outcome.

Hepatic Fibrosis is Associated with Shortened Survival Times in Dogs with Chronic Hepatitis

C.R. Dolan1, A.R. Hoffmann2, R.M. Gold2, J.S. Suchodolski1, J.M. Steiner3, J.A. Lidbury1

1Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA, 2Department of Veterinary Pathobiology Texas A&M University, College Stattion, TX, USA, 3Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA

Chronic hepatitis is a common condition in dogs and fibrosis is an important component of this disease process. In human patients with chronic hepatitis prognosis is related to the stage of hepatic fibrosis. Therefore, the aim of this study was to determine if there is a relationship between the stage of hepatic fibrosis and survival times in dogs with chronic hepatitis.

Dogs histologically diagnosed with chronic hepatitis between January 2003 and January 2013 at the Veterinary Medical Teaching Hospital, Texas A&M University were retrospectively enrolled. Only dogs for which sufficiently detailed follow‐up medical records were available were included in the study. For each dog the time of liver biopsy, survival time, and cause of death or euthanasia were recorded. Hepatic fibrosis was staged by a board‐certified veterinary pathologist (ARH) using a 5‐point scoring system adapted from the Ishak scoring system (absent, mild, moderate, marked, very marked). The survival of dogs with absent to moderate and marked to very marked hepatic fibrosis was compared using Kaplan‐Meyer survival curve analysis and the Mantel‐Cox test. Statistical significance was set as P < 0.05.

Medical records from 112 dogs with chronic hepatitis were reviewed. Sufficiently detailed follow up records were available for 47 dogs. Median survival times were 2,438 days for dogs with absent to mild fibrosis (n = 19) and 675 days for dogs with marked to very marked hepatic fibrosis (n = 28; P = 0.0126).

The median survival time of dogs with absent to mild fibrosis were significantly longer than that of patients with marked to very marked fibrosis.

Proteomic Analysis of Liver Tissue from Dogs with Chronic Hepatitis

Y.A. Lawrence1, L.J. Dangott2, J.M. Steiner3, J.S. Suchodolski1, J.A. Lidbury1

1Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA, 2Department of Biochemistry and Biophysics, Texas A&M University, College Station, TX, USA, 3Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA

Chronic hepatitis is a common cause of morbidity and mortality in dogs. The identification of proteins that are differentially expressed in dogs with chronic hepatitis could contribute to the development of diagnostic markers of this disease and provide insights into its pathogenesis. The objective of this study was to identify proteins that are differentially expressed in the liver of dogs with chronic hepatitis.

Hepatic tissue was collected from 8 healthy dogs during ovariohysterectomy and also from 8 dogs with histologically confirmed chronic hepatitis. The proteome of the liver tissue was extracted by mechanical disruption and detergent based cell lysis and was analyzed by 2‐dimensional fluorescence difference gel electrophoresis. Proteins were separated in the first dimension by isoelectric focusing on a 3 – 10 nonlinear immobilized pH gradient and by 12% sodium dodecyl sulfate‐polyacrylamide gel electrophoresis in the second dimension. DeCyder 2‐D Differential Analysis Software® was utilized for data analysis. Proteins with an absolute fold change value > 2.0 were selected for further analysis. Protein identification was achieved by nanoflow liquid chromatography tandem mass spectrometry (NANO LC‐MS/MS). Post‐hoc testing was performed using Bonferroni's multiple comparisons test and significance was set at P < 0.05.

Twenty‐eight protein spots showed an increased abundance and 12 showed a decreased abundance in liver tissue from dogs with chronic hepatitis as compared with liver tissue from healthy control dogs. Sixteen proteins were selected for further analysis. Proteins significantly increased in the liver of dogs with chronic hepatitis included protein disulfide‐isomerase A4, carbamoyl‐phosphate synthase, senescence marker protein 30, formimidoyltransferase‐cyclodeaminase, and complement C3. These proteins are involved in a variety of processes including inflammation, redox homeostasis, as well as detoxification.

Further studies are needed to determine whether some of the proteins identified may prove useful as a biomarker of chronic hepatitis in dogs.

Characterization of the Urine Metabolome of Dogs with Chronic Hepatitis

Y.A. Lawrence1, B.C. Guard2, J.M. Steiner2, J.S. Suchodolski1, J.A. Lidbury1

1Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA, 2Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College Station, TX, USA

Canine chronic hepatitis can present a diagnostic challenge with various etiologies leading to similar clinical signs and laboratory findings. The histopathological assessment of a liver biopsy specimen is often required for a definitive diagnosis to be made and few prognostic markers exist. The emerging science of metabolomics is used to detect changes in endogenous low molecular weight metabolites in biological samples and offers the possibility of identifying non‐ or minimally‐invasive surrogate markers of disease. The objective of this study was to characterize metabolite changes in the urine of dogs with chronic hepatitis using an untargeted methodology.

Urine samples were collected by cystocentesis from 10 adult dogs with chronic hepatitis and 10 healthy adult control dogs. Urine samples were analyzed by gas chromatography time‐of‐flight mass spectrometry (GC‐TOF). Data were found to be non‐parametric. Principle component analysis was used to make multivariate comparisons between dogs with chronic hepatitis and healthy control dogs. A Wilcoxon rank‐sum test was used to make univariate comparisons between groups. P – values were adjusted for multiple comparisons using the Benjamini and Hochberg False Discovery Rate and significance was set at P < 0.05.

Principle component analysis revealed significant differences between dogs with chronic hepatitis and healthy control dogs. The univariate analysis demonstrated statistically significant differences between dogs with chronic hepatitis and healthy control dogs for 25 of 220 named metabolites. The most clinically relevant of these included glutamine, glycylproline, putrescine, and ornithine, which are important stimulators of M2 macrophages that play a role in tissue repair and proliferation.

Significant differences between the urine metabolites of dogs with chronic hepatitis and healthy control dogs were identified. Further targeted assessment of the altered metabolites is needed to validate these observations.

Hepatic Leptospiral Infections without Prominent Renal Involvement

K.E. McCallum1, F. Constantino‐Casas1, A.J. Kortum1, N. Lingley2, A.J. Sterritt3, H. Swales4, J. Warland5, T. Cogan6, P.J. Watson1

1University of Cambridge, Cambridge, UK, 2Suffolk, UK, 3Highcroft Veterinary Group, Bristol, UK, 4University of Liverpool, Neston, UK, 5Animal Health Trust, Suffolk, UK, 6University of Bristol, Bristol, UK

Leptospiral infection is an infectious disease seen in dogs that has been shown to cause dysfunction in multiple organ systems but typically demonstrates a predilection to the liver and kidneys. Reports of chronic hepatitis caused by Leptospira spp. with absence of renal involvement are confined to small case series in individual kennels. To date, the role of this organism in idiopathic chronic hepatitis remains unknown. However, with the recent availability of Fluorescence In‐Situ Hybridisation (FISH), accurate identification of spirochaetes in liver samples is possible. Consequently, this may help elucidate the role of the organism in cases of chronic canine hepatitis. The aim of this study was to describe the clinico‐histopathological presentation of cases of hepatic leptospirosis diagnosed by FISH, with an absence of renal involvement.

Cases included in this study consisted of cases presented to the Queen's Veterinary School Hospital, University of Cambridge from 2013 to 2016 or cases consulted for telephone advice. Dogs were included in the study if they had a histological diagnosis of hepatitis and the presence of leptospires identified by FISH, with clusters suggesting living, reproducing organisms. Liver histology in all cases was reviewed by a single pathologist. Cases without histopathological diagnosis were excluded. The presence of clinical evidence of renal involvement: – azotaemia (serum creatinine > 140 mmol/L – based on IRIS guidelines) and glycosuria were further criteria of exclusion.

Ten dogs were included in the study. Age at presentation ranged from 2.8 years to 10 years (median 8.5 years). Six out of ten were spaniels: breeds represented included Cavalier King Charles Spaniel (n = 1), Labrador Retriever (n = 2), Cocker Spaniel (n = 2), English Springer Spaniel (n = 3), West Highland White Terrier (n = 1) and Jack Russell Terrier (n = 1). Eight of the cases were female and two were male. Clinical signs varied but most commonly included anorexia (n = 6), abdominal pain (n = 2), vomiting (n = 2), lethargy (n = 4), gingivitis (n = 2), pyrexia (n = 3), and polydipsia/polyuria (n = 2). Two dogs were siblings that lived together and developed clinical signs at the same time. Three dogs were asymptomatic but had persistent elevations in liver enzymes. Mild thrombocytopenia was reported in four cases and all cases had an elevation in liver enzymes. Histopathology of the liver documented chronic hepatitis in all cases with granulomatous hepatitis in seven cases, one of which showed sheets of macrophages, and three cases demonstrated lymphoplasmacytic/neutrophilic hepatitis. Three dogs underwent repeat liver biopsy 3, 15 and 16 months following initial diagnosis, respectively. Fibrosis and vacuolar hepatopathy was diagnosed in the first case and cirrhosis in the second case (both of these cases remained positive for leptospira by FISH). The case re‐biopsied at 16 months was diagnosed with granulomatous hepatitis and had become negative by FISH. One patient with a histopathological diagnosis of neutrophilic, lymphoplasmacytic gingivitis and vasculitis demonstrated the presence of leptospira by FISH. Leptospirosis speciation in one case of granulomatous hepatitis documented the presence of Leptospira interrogans.

All dogs were treated with a minimum of 10 mg/kg doxycycline PO SID for 14 days. Clinical outcome was known for all cases: – six dogs were alive at the time of writing and four dogs were euthanised as a result of progressive liver disease. This study suggests that the presence of leptospiral organisms within the liver may be associated with chronic hepatitis (often granulomatous) without clinical evidence of renal involvement. We intend to undertake further studies that perform sequencing of Leptospira spp. organisms as well as performing FISH on control liver biopsies with an absence of hepatitis.

Demographics and Histological Changes in a Cohort Group of Dogs with Abnormal Hepatic Copper Concentrations

T. Ullal, D. Twedt

Colorado State University, Fort Collins, CO, USA

Liver disease associated with abnormal hepatocyte copper (Cu) concentrations occurs commonly in dogs. Excessive hepatic Cu results in hepatocyte damage from oxidative stress, resulting in cell death and subsequent necroinflammatory changes. Genetic mutations in Cu metabolism, impaired biliary excretion of Cu, and increased dietary Cu are thought to contribute to excessive hepatic Cu.

The aim of this study was to document trends in age, sex, and breed for dogs having abnormally elevated hepatic Cu and to then correlate the abnormal copper content to the degree of histopathological inflammatory change. We hypothesized that abnormal hepatic Cu concentrations would be more common in middle aged to older female dogs with certain breeds being over represented. We also hypothesized that the higher the Cu content, the more severe the inflammatory liver changes would be.

We examined our diagnostic laboratory records for cases having both liver histopathology and hepatic Cu quantitation in the year 2014. Fresh liver tissue was analyzed for Cu via flame atomic absorption spectroscopy and expressed as μg/g dry weight liver ([dwl], normal 120–400). We elected to evaluate a subset of dogs having hepatic Cu > 1000 μg/g because those cases would more likely reflect a primary copper storage disorder. Histopathologic diagnoses were grouped as inflammatory or non‐inflammatory. Inflammatory samples were sub‐classified as mild, moderate or severe. Non‐inflammatory samples were characterized as hepatocellular swelling, hyperplasia, lipidosis, nonspecific reactive hepatopathies, fibrosis, or neoplasia.

628 cases met the criteria of having both liver histology and Cu quantitation. 125 cases (20%) had hepatic Cu > 1000 (median 1670 μg/g dwl). Ages ranged from 1 to 14 years (median 7.0 years). 56% were female. 47 different breeds were represented. The Labrador retriever (21.6%), Doberman pinscher (8%), West Highland white terrier (5.6%), and standard poodle (4%) were the four most common pure breeds. Median age for Labrador retrievers was 8 compared to 5 years for Doberman pinschers. 52% of Labrador retrievers, 70% of Doberman pinschers, 80% of standard poodles, and none of the West Highland white terriers were female. Of the 326 cases having normal copper levels, the Labrador retriever (10%), Chihuahua (5.2%), Maltese (4.9%), standard poodle (4.6%), shih tzu (4.3%), and Yorkshire terriers (4%) comprised the most cases.

108 cases (86%) having hepatic Cu > 1000 also had inflammatory changes on histopathology. The inflammation was classified as mild (16.7%), moderate (58.3%), or severe (24%). Dogs with inflammatory changes were far more likely to have elevated hepatic Cu (P < 0.001). Specifically, Cu dogs with moderate to severe inflammation had significantly higher hepatic Cu concentrations (P < 0.0001). The Labrador retriever was the breed most represented in the moderate and severe groups.

Our findings indicate that the Labrador retriever is a common breed with copper associated liver disease. However, this may be an overrepresentation due to the popularity of the breed. More female cases had elevated hepatic Cu in all breeds with the exception of the West Highland white terrier, in which only male cases were identified. Affected dogs in this study were on average 7–8 years old at the time of diagnosis except for Doberman pinschers, which were on average younger. Higher levels of hepatic Cu were more likely associated with more severe inflammatory changes on histopathology.

Coagulation Abnormalities and their Correlation to Use of Gelfoam® to Aid in Hemostasis During Laparoscopic Liver Biopsy in Dogs

S.A. Wennogle1, C. Olver2, D. Twedt2

1College of Veterinary Medicine, Colorado State University, Fort Collins, CO, Fort Collins, CO, USA, 2Colorado State University, Fort Collins, CO, USA

Coagulation abnormalities are common in human and veterinary patients with hepatobiliary disease. Due to the liver's complex and varied influence on hemostasis these derangements can result in a multitude of outcomes in the patient, including a propensity toward bleeding and/or thrombosis. Hepatic biopsy is commonly indicated in the diagnostic work‐up of hepatobiliary disease in dogs, and increasingly performed via laparoscopy. Knowledge of a patient's hemostatic status prior to biopsy is considered essential to assess bleeding risk. However, conventional tests are not always accurate predictors of clinical bleeding. Additionally, newer techniques for evaluating the coagulation system have recently become available.

The objective of our study was to evaluate both conventional and modern tests of the coagulation system in a group of dogs undergoing laparoscopic liver biopsy for diagnostic work‐up of hepatobiliary disease. Results of these tests were then correlated with the use of intervention to control clinical bleeding during laparoscopic liver biopsy.

27 dogs undergoing laparoscopic liver biopsy at Colorado State University were prospectively enrolled. Blood was collected prior to the procedure for chemistry profile, platelet count, platelet aggregometry (Multiplate® analyzer), thromboelastography, prothrombin time, activated partial thromboplastin time, antithrombin activity, fibrinogen, d‐dimers, and fibrin degradation products. Bleeding was scored based on whether Gelfoam® (absorbable gelatin sponge, USP) was utilized by the laparoscopist to aid in control of bleeding. Gelfoam® was applied when active bleeding persisted for >30 seconds at one or more biopsy sites.

Histology confirmed hepatic disease in all dogs (Idiopathic chronic hepatitis; n = 12; Copper‐associated hepatopathy: n = 5; Hydropic/Vacuolar: n = 5; Non‐specific reactive: n = 3; Vascular: n = 1; Lymphoma: n = 1). 18/27 (67%) dogs had at least one coagulation abnormality. 5/27 (19%) dogs had Gelfoam® placed to aid hemostasis in one or more biopsy sites. Hypoalbuminemia (P = 0.009), decreased antithrombin activity (P = 0.003), and older age (P = 0.038) were associated with the use of Gelfoam®. 14/27 (52%) dogs had abnormal platelet aggregometry, including all 5 dogs who received Gelfoam®. Abnormal platelet aggregometry was associated with the use of Gelfoam® to aid in hemostasis (P = 0.041). Prolonged PT or APTT were not associated with use of Gelfoam®. Hypofibrinogenemia was also not associated with Gelfoam® use, but was correlated with hypoalbuminemia (P = 0.042). Bleeding was controlled with Gelfoam® intervention alone in all dogs. One dog sustained a splenic laceration during entry into the abdomen, but hemostasis was achieved quickly. Otherwise, no significant complications occurred.

Thrombocytopathia may have contributed to clinical bleeding during laparoscopic liver biopsy in this group of dogs with hepatobiliary disease. Additionally, all dogs with low antithrombin had Gelfoam® placed at biopsy sites to aid hemostasis. Low antithrombin may reflect decreased liver function and/or be due to consumption as a consequence of low‐grade disseminated intravascular coagulation. As hypofibrinogenemia was correlated with hypoalbuminemia, it is possible fibrinogen could be assessed as an additional marker of hepatic function. In these 27 dogs, laparoscopic liver biopsy was safe and adequate samples to achieve a histologic diagnosis were obtained in all cases.

Bleeding Risk Assessment for Percutaneous Ultrasound Guided Hepatic Biopsy in Dogs and Cats

M.K. Pavlick1, C.R. Leveille‐Webster2, D.G. Penninck3

1Cummings School of Veterinary Medicine, Tufts University, Worcester, MA, USA, 2Cummings School of Veterinary Medicine at Tufts University, Grafton, MA, USA, 3Cummings School of Veterinary Medicine, Tufts University, North Grafton, MA, USA

Hepatic biopsy is often necessary for a definitive diagnosis of liver disease. Because the liver plays a central role in hemostasis, bleeding after biopsy procedure is a concern. The aim of this study was to determine risk factors for bleeding following percutaneous ultrasound‐guided liver biopsy. The records of 104 dogs and 30 cats that had a percutaneous ultrasound‐guided liver biopsy were retrospectively reviewed. Based on human guidelines for bleeding, blood loss was classified as minor, moderate and severe by an absolute decrease in the hematocrit of ≤ 6%, 7–14% and ≥ 15%, respectively. Complications were defined as physiologic compromises that necessitated an intervention (transfusion or resuscitative fluids) or death. The relationship between change in hematocrit/complications and initial hematocrit, PT, aPTT, platelet count, serum parameters (liver enzymes, bilirubin) as well as the number of biopsies, biopsy needle gauge, final histological diagnosis and ultrasonographic findings (focal vs diffuse, nodularity, ascites) were assessed using Chi square tests, One Way Analysis of Variance or Pearson's correlation coefficient with a P value < 0.05 considered significant.

Mild, moderate and severe bleeding occurred in 61(58.7%), 33(31.7%) and 10(9.6%) dogs, respectively and complications in 2 (1.9%) dogs. Dogs with a higher initial hematocrit had a greater decrease in hematocrit. Dogs with neoplasia were more likely to have complications. Mild, moderate and severe bleeding occurred in 13(43.3%), 16(53.3%) and 1(3.3%) cat(s), respectively and complications in 5(16.7%) cats. Moderate/severe bleeding was more likely in cats with a nodular liver. Cats with a higher initial platelet count had a greater decrease in hematocrit. Complications were associated with a lower pre‐biopsy hematocrit in cats. In both dogs and cats, there was no correlation between the change in hematocrit/complications and PT, aPTT, bilirubin, serum liver enzyme activity, the number of biopsies, biopsy needle gauge, histological diagnosis, vitamin K administration or the presence of ascites, focal or diffuse disease.

Ultrasound‐guided liver biopsy is a relatively safe procedure. The risk of complications is higher in cats and is associated with a lower pre‐biopsy hematocrit. The reason for the paradoxical findings that higher platelet number and hematocrit in the cat and dog respectively are associated with greater decrease in hematocrit will require further investigation. Typical indicators of hypocoagulability (prolonged PT, aPTT and low platelet count) did not predict the occurrence of complications or the magnitude of change in hematocrit post‐biopsy. This likely reflects the complex state of coagulation that exists in liver disease.

High‐Throughput RNA Sequencing and Differential Gene Expression Analysis in Dogs with Chronic Hepatitis

V.M. Eulenberg1, Y.A. Lawrence1, J.S. Suchodolski1, J.M. Steiner2, J.A. Lidbury1

1Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA, 2Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College station, Texas, USA, College Station, TX, USA

Hepatic fibrosis is the result of chronic hepatic injury and can lead to loss of functional hepatic mass and portal hypertension and thus is an important event in the progression of chronic hepatitis. The pathogenesis of hepatic fibrosis in dogs has not yet been fully characterized. The aim of this study was to identify differentially expressed genes in dogs with chronic hepatitis to further elucidate the molecular mechanisms involved in the development and regulation of hepatic fibrosis.

RNA extraction and high‐throughput RNA sequencing were performed on liver biopsy specimens collected from 16 dogs with chronic hepatitis and 12 healthy control dogs undergoing ovariohysterectomy. The sequencing data were analyzed using CLC Genomics software (CLC bio, QIAGEN) and the included empirical analysis of differential gene expression. Genes with an absolute fold change‐value between groups > 2.5 and a multiple comparisons adjusted P‐value < 0.05 were considered to be differentially expressed. Differentially expressed genes and transcripts potentially related to the pathogenesis of hepatic fibrosis were selected and subcategorized using the genecards.org databases (GeneCards, PathCards, and GeneAnalytics).

Six‐hundred‐and‐fifty‐one genes (1,136 transcripts) were differentially expressed in dogs with chronic hepatitis. A connection with hepatic fibrosis or related pathways was identified for 190 of these genes (269 transcripts). One‐hundred‐and‐sixty‐five of these genes (222 transcripts) were upregulated and 25 (47 transcripts) were downregulated. They were grouped into the following subcategories: extracellular matrix degradation, extracellular matrix organization, hepatic stellate cell activation, apoptosis, detoxification of reactive oxygen species, and components of fibrotic mechanisms mentioned in the human literature.

A large number of genes were found to be differentially expressed in dogs with chronic hepatitis compared to healthy control dogs and some of these genes may potentially be related to the pathogenesis of hepatic fibrosis. Further study of this dataset and confirmatory testing of a subset of the identified genes with quantitative RT‐PCR might provide further insights into the pathogenesis of canine hepatic fibrosis.

In Vitro Effects of Vitamin D on Phagocytosis, TLR4, and Cytokine Production in Healthy Dogs

J. Jaffey, J. Amorim, A. DeClue

University of Missouri Veterinary Health Center, Columbia, MO, USA

Hypovitaminosis D has been identified in dogs with sepsis. In people, calcitriol reverses the immunopathology during sepsis, yet little is known about the effects of calcitriol on innate immune function in the dog. We hypothesized that calcitriol will reduce canine leukocyte TNF production, increase IL‐10 production, augment PMN phagocytosis and reduce TLR4 expression.

Blood was collected from eight healthy, adult dogs and incubated with calcitriol (10−7 M) or control for 24 h. Following incubation, LPS‐stimulated leukocyte production of TNF and IL‐10 were measured using a canine‐specific multiplex assay and PMN phagocytosis of opsonized E. coli and constitutive TLR4 expression were evaluated using flow cytometry. In three dogs, a concentration response curve was performed to evaluate if the cytokine modulation was concentration dependent. Comparisons were made using paired T‐test or Wilcoxon Signed Rank test, with a P < 0.05 considered significant.

Calcitriol significantly reduced LPS‐stimulated leukocyte TNF production (mean±SD, calcitriol, 823 ± 862; control, 1694 ± 1023 pg/mL; P = 0.022) and TNF:IL‐10 ratio (median, Q1, Q3, calcitriol, 0.9, 0.2, 1.1; control, 3.5, 1.4, 6.9; P = 0.039) but not IL‐10 production (calcitriol, 791 ± 386, control, 551 ± 323 pg/mL; P = 0.114). Blunting of TNF production was concentration dependent (concentration range: 10−9 to 10−7 M). Phagocytosis and TLR4 expression were not significantly altered by calcitriol.

These data indicate that calcitriol induces an anti‐inflammatory shift in leukocytes exposed to LPS without altering phagocytosis or TLR4 expression, in vitro, which may be due to alterations in post‐receptor signaling or protein translation. Thus, calcitriol could be a novel candidate immunomodulatory therapy for sepsis in dogs.

Urban Environment and Odds of Canine Immune‐Mediated Disease

U. Jeffery

Texas A & M University, College Station, TX, USA

Urban environment and exposure to air pollution are associated with human auto‐immune diseases including systemic lupus erythematosus and multiple sclerosis. Specific environmental triggers for canine immune‐mediated disease remain largely undiscovered, but their identification would improve the management of high‐risk individuals.

The aim of this study was to determine if dogs living in urban areas have increased odds of immune‐mediated disease compared with those in rural areas.

Case records from a referral hospital in the American Mid‐West were retrospectively reviewed to identify dogs meeting pre‐specified criteria for immune‐mediated hemolytic anemia (IMHA) and/or thrombocytopenia (ITP); immune‐mediated polyarthritis (IMPA) or meningoencephalitis of unknown origin (MUO). Breed‐matched controls without immune‐mediated disease were randomly selected. The census tract in which the dog lived and its population density were retrieved from a government database. Urban environment was classified as >1000 people per square mile, as defined by the US census bureau.

137 cases and 137 controls meet the inclusion criteria. The odds ratio for any immune mediated disease was 0.94 (95% confidence interval: 0.58–1.52, P = 0.81) for urban vs. rural dogs. For specific immune‐mediated diseases, odds ratios were 0.85 (0.45–1.61, P = 0.63, n = 79 cases) for IMHA/ITP; 0.83 (0.25–2.77, P = 0.76, n = 21 cases) for IMPA and 1.26 (0.49–3.23, P = 0.63, n = 37 cases) for MUO.

This study did not reveal significant associations between urban environment and canine immune‐mediated disease. Larger studies, or studies in other geographical areas might still reveal a relationship, but the current study suggests exposures common to both rural and urban dogs merit further investigation as triggers of immune‐dysregulation.

Immune Cells in the Nose and Oropharynx of Cats and Response to Mucosal Immune Stimulation

S. Dow, W. Wheat, E.T. Contreras, M.R. Lappin

Colorado State University, Fort Collins, CO, USA

Little is known regarding the normal immune cell populations resident in the nose and oropharynx of healthy cats. Nor have previous studies investigated the cellular responses of these mucosal tissues to non‐specific immune stimulation. Such information could provide important insights into the nature and timing of mucosal immune responses in the upper respiratory and GI tracts of cats.

Nasal lavage and oropharyngeal swabs were used to obtain resident cells from 9 healthy adult cats. The cell populations were phenotyped and enumerated using flow cytometry. After baseline samples were collected, cats were treated by intranasal and oral administration of a liposome‐TLR mucosal immune stimulant (MucosImmune), and changes in immune cell populations were assessed by nasal lavage and oral swabs obtained at 6 h, 24 h and 72 h after treatment.

We found that the normal cat nasal passages contained a mixed population of CD5 + T cells and CD14 + monocytes, with relatively abundant neutrophils. In contrast, the oropharynx contained numerous CD5 + T cells and few monocytes or neutrophils. Following administration of MucosImmune, there was a sustained influx of neutrophils into the nose over 72 h, while neutrophils only increased in the oropharynx at 6 h after treatment. There was a significant increase in CD14 + monocytes in both sites by 6 h, with gradual declines over 72 h. In both sites, there was a transient disappearance of CD5 + lymphocytes at 6 h, which rebounded to normal numbers by 24 to 72 h.

These data indicate that nasal passages and the oropharynx in cats both contain a relatively abundant population of immune cells capable of responding rapidly to TLR activation, with the oropharynx containing a dense T cell population and the nose containing more neutrophils and likely B cells. Both sites responded rapidly to mucosal delivery of TLR agonists, with early appearance of innate immune cells and disappearance and then reappearance of T lymphocytes. These results indicate that mucosal administration of a liposome‐TLR agonist rapidly activates local immune responses in the feline upper respiratory tract, which may be associated with non‐specific protection from viral or bacterial pathogens.

The Use of a Combination‐Drug Protocol Versus Glucocorticoids Alone for Treatment of Idiopathic Immune‐Mediated Polyarthritis

S. Ravicini1, A.C.C. Kent1, M.D. Dunning2, F. Allerton1

1Willows Referral Centre, Solihull, UK, 2University of Nottingham, Sutton Bonington, UK

Idiopathic immune‐mediated polyarthritis (IMPA) is an inflammatory non‐infectious joint disease caused by immune‐complex deposition in the synovium due to a type‐III hypersensitivity reaction. Oral administration of immunosuppressive drugs is the standard treatment for this condition. The most commonly used drug is prednisolone, often in combination with other immunosuppressive agents. The benefit of combination immunosuppressive therapy introduced at the time of diagnosis has not been previously evaluated. The aim of this retrospective study was to compare outcomes for dogs with IMPA treated with prednisolone alone or in combination with other immunosuppressive drugs introduced at the time of diagnosis. We hypothesised that the use of combination immunosuppressants would not improve outcome in canine IMPA.

The medical records from a single tertiary referral centre were reviewed between January 2014 and May 2016 for dogs diagnosed with idiopathic IMPA on the basis of consistent joint fluid analysis. Cases were excluded if an underlying disease was identified. Thirty‐six dogs met the inclusion criteria: 17 females and 19 males of various breeds, with a median age of 8 years (range 2–12) and median bodyweight of 19 kg (6.3–46 kg). 16/36 cases were presented for lameness or reduced mobility with 20 dogs displaying systemic signs (e.g. lethargy, decreased appetite, pyrexia) of which 7 manifested no mobility issues prior to diagnosis. Median time to diagnosis was 100 days (range 5 – 560) with no seasonal bias. Group 1 (23 dogs) received prednisolone, whilst group 2 (13 dogs) received prednisolone with azathioprine (7) or ciclosporin (6). Outcome data was compared using non‐parametric tests.

13/23 dogs in group 1 relapsed during tapering of medication compared to 4/13 in group 2 (P = 0.07). Three dogs in group 1 were euthanased due to suspected complications arising from IMPA or therapy (one dog died after developing pyothorax and was excluded from calculations). All 13 dogs in group 2 are still alive (P = 0.17). In group 1 the median treatment duration was 397 days (104–1250) compared to a median of 158 days (81–260) in group 2 (P = 012).

There was no significant difference between groups 1 and 2 for duration of therapy, relapse rate or mortality. Thus the hypothesis was accepted. The observations that dogs started on a multi‐drug protocol from the outset required a shorter treatment duration, had lower mortality and a lower relapse rate may reflect a type II error due to low group numbers. A prospective study with greater numbers is ongoing to investigate this further.

Activation of Innate Immune Responses in Cat Leukocytes by a Liposomal TLR Ligand Immune Stimulant

S. Dow, W. Wheat, J. Coy, L. Chow, M.R. Lappin

Colorado State University, Fort Collins, CO, USA

Previous studies have demonstrated the ability of liposome‐TLR ligand complexes (LTLC) to potently activate innate immune responses in several species, but they have not been fully evaluated for activity in cats. There is currently an unmet need for a broad spectrum immune stimulant for prevention or early treatment of viral and bacterial respiratory tract pathogens in cats. The purpose of the present study was to investigate in vitro responses of cat lymphocytes and monocytes to stimulation by LTLC.

Cat peripheral blood mononuclear cells (PBMC) obtained from healthy cats were used to assess immune activation by LTLC. PBMC were incubated with LTLC for 24 to 72 hours, and immune activation was assessed by cytokine release assays, RT‐PCR, and flow cytometry.

We found that LTLC triggered IFN‐g production by PBMC within 24 h of incubation. In addition, upregulated expression of IL‐12 and TNF‐a mRNA was demonstrated using RT‐PCR. With flow cytometry, LTLC induced activation of B cells and T and monocytes, as reflected by upregulated expression of MHCII expression on B cells and monocytes and upregulated expression of OX40 on T cells and B cells. In addition, LTLC induced T cell proliferation.

In a companion abstract, we demonstrate the mucosal immune stimulatory properties of an LTLC‐based immune stimulant administered to healthy cats. Therefore, taken together with the in vitro findings presented here, these results suggest that LTLC may be used as an effective non‐specific immune stimulant for induction of anti‐viral and anti‐bacterial immunity in cats.

Validation of an Electrochemiluminescent Immunoassay for Measurement of Tumor Necrosis Factor‐Alpha in Cat Serum

C.C. Thomas1, S. Marsilio2, J.S. Suchodolski1, J.M. Steiner3, J.A. Lidbury4

1Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA, 2Texas A&M University, Gastrointestinal Laboratory, College Station, TX, USA, 3Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College Station, TX, USA, 4Gastrointestinal Laboratory, Texas A&M University, College station, TX, USA

Tumor necrosis factor‐alpha (TNF‐α) is a cytokine released mainly by macrophages in response to contact with infectious agents, tissue injury, tumor growth, and autoimmune conditions. Thus, measurement of TNF‐α warrants evaluation as a diagnostic and prognostic marker in cats with inflammatory diseases. However, currently no analytically validated assay is commercially available for the measurement of this cytokine in cats. The aim of this study was to develop and analytically validate an electrochemiluminescent immunoassay for the measurement of TNF‐α in cat serum.

Using a commercially available platform (UPLEX, Meso Scale Diagnostics) a feline TNF‐α electrochemiluminescent assay was developed from recombinant feline TNF‐α and polyclonal rabbit anti‐feline TNF‐α antibodies. Validation variables included the lower limit of detection, dilutional parallelism, spiking recovery, intraassay variability, and interassay variability.

The lower limit of detection was 1.3 ng/L. Observed‐to‐expected ratios for dilutional parallelism (n = 6) ranged from 43.3% – 133.3% with a mean (±SD) of 102.9% ± 25.4%. Recoveries for spiked TNF‐α in serum samples (n = 4) ranged from 72.2% – 112.1% with a mean ± s.d. of 93.1% ± 12.5%. Intraassay coefficients of variation (%CVs; n = 4) were ≤ 4.3% with a mean ± s.d. of 3.4% ± 0.7%. Interassay coefficients of variation (n = 4) were ≤ 17.6% with a mean ± s.d. of 12.5% ± 4.9%.

This assay was shown to be sensitive, precise, and accurate for the measurement of TNF‐α in cat serum. However, assay linearity was suboptimal. Further studies are needed to determine whether linearity can be improved with the current assay system. Also, the clinical implications of serum TNF‐α concentrations need to be evaluated in this species.

A Flow Cytometric Study of CD4 + CD25 + FoxP3 +  Regulatory T Cells Changes in Canine Atopic Dermatitis

D.‐j. Lee1, C. Park1, H.‐j. Kim2, Y. Kim3, S. Yoon1

1College of Veterinary Medicine, Chonbuk National University, Iksan, Cholla‐bukto, Korea, 2College of Veterinary Medicine, Chonnam National University, Kwang‐ju, Cholla‐namdo, Korea, 3College of Veterinary Medicine, Chonbuk National University, Curi, Kyonggi‐do, Korea

Canine atopic dermatitis (CAD) is a common, chronic, genetically‐predisposed, and relapsing inflammatory skin disease. CAD is a multifaceted disease resulting from a complex interaction between environmental and genetic factors. Both of these factors can shape skin barrier function and the immunological response of predisposed patients. CD4 + CD25 + FoxP3 + Regulatory T cells (Tregs) become known to prevent excessive immune responses through their immunoregulatory properties and therefore play a leading role in the maintenance of immunological homeostasis. There was only few researches that was performed to analyze the relationship between CAD and proportion of Tregs in periphery blood. In the present study, the aims were to assess the proportion of circulating Tregs (CD4 + CD25 + FoxP3 + , CD4 + FoxP3 + ) in the healthy and the atopic dogs to analyze implication of Tregs to CAD and to investigate the correlation of Tregs with CAD severity and age.

The animals for analysis consisted of eight healthy dogs and nine atopy dogs that had never been treated. The diagnosis of CAD was made with fulfillment of clinical criteria (Farvot's criteria) and exclusion of differential diagnosis including ectoparasitic and infectious dermatoses. Expression of Tregs in peripheral blood was evaluated by flow cytometry. Tregs were identified as CD4 + CD25 + PoxP3 + T cells and CD4 + PoxP3 + T cells. There were statistically significant increase in proportion of both CD4 + CD25 + FoxP3 + T cell and CD4 + FoxP3 + T cell in periphery blood of atopic dogs compared to healthy dogs.(mean 0.91% verse 4.2%, P = 0.012; mean 1.76% verse 10.06%, P = 0.004). The circulating regulatory T cells (Tregs) of atopic dogs had negative correlation with their disease severity. (r=−0.758 P = 0.05) and also negative with age (r=−0.666 P = 0.05).

Both CD4 + CD25 + FoxP3 + Tregs and CD4 + FoxP3 + Tregs may have a relevance to the pathogenesis of CAD in course of atopic dermatitis. This study demonstrated negative association between Treg frequency and disease severity. The numbers of circulating tregs of atopic dogs decreased as they got older.

12/15 Lipoxygenase Deficiency Enhances the Development of Pneumonic Tularemia

C.L. Pratt1, J. Skyberg2, C.R. Brown2

1University of Missouri, Saint Louis, MO, USA, 2University of Missouri, Columbia, MO, USA

Francisella tularensis, a pathogenic Gram negative intracellular bacteria that causes pneumonia, skin lesions, and systemic illness known as Tularemia that can affect cats in certain regions of the United States. 12/15 lipoxygenase (LO) is an enzyme that utilizes arachidonic acid to produce anti‐inflammatory mediators. Deletion would be expected to produce an increase in the severity of disease. Since 12/15‐LO has a role in regulating inflammation, our hypothesis is pneumonic tularemia in mice would produce increases in morbidity and mortality. Francisella tularensis Live Vaccine Strain (Ft LVS) is pathogenic to mice but not to people. C3H wild type (WT) and 12/15 LO knock out (KO) mice were infected intratracheally with a sublethal dose of Ft LVS and their weights were monitored for 10 days. On day 10 post‐infection the mice were euthanized and the lungs, spleens, and liver were collected for analysis. The cellular phenotype of infiltrating inflammatory cells, leukotriene B4 (LTB4) and KC levels were determined from lung tissue. Histopathology was assessed from the lung and liver, and splenic weights were determined. 12/15 LO KO mice had increased morbidity compared to WT mice indicated by their significant weight loss. KC, a chemoattractant for neutrophils, was significantly (P < 0.05) higher in the lung tissue of KO mice. Inflammatory cells (neutrophils, T cells, and macrophages) and pulmonary histopathology scores were significantly higher (P < 0.05) in KO mice. In contrast LTB4, which assists in host defense was significantly decreased (P < 0.05) in 12/15 LO KO mice. Splenic weights were significantly (P < 0.05) higher than WT spleens correlating with the severity of disease. Given this information, 12/15 LO plays a role in regulating inflammation in pneumonic Tularemia. Eicosanoids are highly conserved among species and It is possible that 12/15 LO agonists may be used as natural therapeutics to aide in the resolution of pneumonia in domestic animals, but further studies are needed.

Feline Adipose Tissue‐Derived Mesenchymal Stem Cells Reduce Inflammation by Secreting Soluble Factors

H.‐K. Chae, B.‐Y. Lee, W.‐J. Song, Q. Li, M.‐O. Ryu, S.‐C. Park, H.‐J. Kim, S.‐Y. Kim, H.‐M. Yang, H.‐Y. Youn

Seoul National University, Seoul, Korea

Mesenchymal stem cells (MSCs) have immunomodulatory functions and differentiation capacity, conferring them with various clinical uses. Although MSCs have been applied in the treatment in various inflammatory diseases, mechanistic research on feline MSCs is lacking. Accordingly, in this study, we aimed to analyse the immunomodulatory mechanisms of MSCs isolated from feline adipose tissue (fAT‐MSCs). Cytokine expression in RAW264.7 murine macrophages and allogeneic feline peripheral blood mononuclear cells (fPBMCs) was measured using quantitative real‐time polymerase chain reaction (qRT‐PCR) and compared according to the presence of fAT‐MSCs. We also extracted RNA and collected supernatants from fAT‐MSCs for measurement of soluble factor expression levels by qRT‐PCR and enzyme‐linked immunosorbent assay. The levels of pro‐inflammatory cytokines, e.g., tumour necrosis factor‐α (TNF‐α), inducible nitric oxide synthase and interleukin (IL)‐1β, were significantly decreased in coculture of mitogen‐stimulated RAW264.7 cells with fAT‐MSCs compared with that in RAW264.7 cells. Additionally, the expression of immunomodulatory factors from fAT‐MSCs, including cyclooxygenase‐2 (COX‐2), transforming growth factor (TGF)‐β, indoleamine‐2,3‐dioxygenase (IDO) and hepatocyte growth factor, increased in the presence of RAW264.7 cells. Expression of pro‐inflammatory TNF‐α, interferon‐γ and IL‐6 decreased, and expression of anti‐inflammatory IL‐10 increased during coculture of mitogen‐stimulated allogeneic fPBMCs with fAT‐MSCs. TGF‐β, COX‐2 and IDO expression and prostaglandin E‐2 production from fAT‐MSCs increased in the presence of allogeneic fPBMCs. Our data suggested that soluble factors secreted from fAT‐MSCs play an important role in the immunomodulatory effects of these cells. These findings may be helpful in the application of fAT‐MSCs to feline patients with immune‐related diseases.

Presence of Naturally Occurring Circulating Anti‐ AAV8 and AAV9 Antibodies in Pet Dogs and Cats

B. Lucchetti1, A. Specht2, K. Coleman3, T. Conlon3

1BluePearl Veterinary Partners Arizona, Scottsdale, AZ, USA, 2College of Veterinary Medicine, University of Florida, Gainesville, FL, USA, 3Department of Pediatrics/Powell Gene Therapy Center, University of Florida, Gainesville, FL, USA

Gene therapy is an emerging treatment modality in human medicine and animal models including dogs and cats are often used in preclinical studies to demonstrate efficacy and safety. Research into gene therapies for some conditions in animals has started as well. Adeno‐associated virus (AAV) based vectors are among the most common delivery mechanisms. It is well established that antibodies to AAV vectors can impact the success of gene transfer. The primary objective of this study was to determine if pet populations have circulating antibodies to two AAV vectors used for gene therapy.

An ELISA technique was utilized to assess the titer level of circulating antibodies to serotypes AAV8 and AAV9 in 115 canine and 85 feline patients presenting to a university hospital. Results were variable, but mean values for canines were 2.3x103 (anti‐AAV8) and 1.4x104 (anti‐AAV9). Feline mean values were 1.2x105 (anti‐AAV8) and 4.1x105 (anti‐AAV9). There was very weak, positive correlation between serotypes within species based on a two‐sided Pearson's product‐moment correlation test (Canine r2 = 0.18, Feline r2= 0.37).

Based on these results, it appears dogs and cats in the pet population have some circulating antibodies capable of binding to AAV8 and AAV9 particles. However, without concurrent neutralizing antibody levels, the clinical significance of this finding is unknown. If a correlation between circulating antibody titers and neutralizing antibody titers exists, it would suggest that potential candidates for gene therapy drawn from a pet population might require screening for antibodies or immunosuppressive therapy prior to receiving AAV vector based treatment.

Serum from Some Pet Dogs and Cats Neutralizes AAV8 and AAV9 Mediated Gene Transduction

T. Conlon1, S. Boye2, W. Li2, K. Coleman1, B. Lucchetti3, A. Specht4

1Department of Pediatrics/Powell Gene Therapy Center, University of Florida, Gainesville, FL, USA, 2Department of Ophthalmology, College of Medicine, University of Florida, Gainesville, FL, USA, 3BluePearl Veterinary Partners Arizona, Scottsdale, AZ, USA, 4College of Veterinary Medicine, University of Florida, Gainesville, FL, USA

With the increasing interest in gene therapy as a potential treatment modality for a variety of human diseases, large animal models including dogs and cats are often used in preclinical studies. There has also been interest in using these techniques to treat diseases that affect pet populations. Adeno‐associated virus (AAV) based vectors are among the most common utilized delivery mechanisms and circulating antibodies capable of binding to AAV8 and AAV9 particles have been identified in pet dogs and cats. The primary objective of this study was to identify whether some of these same pets might have neutralizing antibody titers against AAV8 and AAV9 transduction.

Serum samples from seven pet dogs and seven pet cats with varying levels of circulating anti‐AAV8 and anti‐AAV9 antibodies were utilized. Neutralizing antibody assays were performed using a Self‐complementary c‐smCBA‐mCherry reporter packaged in either AAV8 or AAV9 and permissive cells (661W and Lec2, respectively) at a multiplicity of infection of 10,000 vg/cell. Positive controls were serum samples from dogs in a research colony that had been previously treated with AAV8 and AAV9 vectors and negative controls had no serum. Vectors were pre‐incubated with series of serum dilutions (1:10, 1:40, 1:160, 1:640 and 1:2560) before infecting cells. Three days post infection, transduction was assessed by measuring mCherry‐mediated fluorescence. Neutralization of transduction greater than 50% relative to no serum control in the lowest dilution (1:10) was identified in 2/7 cats and 1/7 dogs for AAV8 and 1/7 cats and 5/7 dogs for AAV9. There was no obvious positive correlation between circulating antibody titers and neutralization titers. There was a trend for the presence of feline serum to have a positive impact on transduction efficiency of both vectors.

The clinical significance of these results remains unknown, but they suggest the need for further evaluation to determine if candidates for gene therapy from a pet population might require screening tests or immunosuppressive therapy prior to receiving AAV vector based treatment.

Bartonella Seroepidemiology in Dogs from North America, 2008–2014

E. Lashnits1, M. Correa1, B. Hegarty1, A. Birkenheuer1, E. Breitschwerdt2

1North Carolina State University, Raleigh, NC, USA, 2NCSU‐CVM, Raleigh, NC, USA

Improved understanding of Bartonella spp. serological patterns in dogs may aid clinical decision making and enhance current understanding of naturally‐occurring arthropod vector transmission. We analyzed a large diagnostic laboratory database to better understand Bartonella seroepidemiology of dogs in North America, identifying demographic groups in which Bartonella exposure may be more likely, describing variations in chronologic and geographic patterns of Bartonella seroreactivity, and examining co‐exposure to other vector‐borne pathogens. We hypothesize that there will be no significant demographic risk factors for Bartonella exposure, and that Bartonella exposure will be more common in regions and seasons with high vector prevalence and among dogs with other canine vector‐borne disease (CVBD) exposure.

CVBD serology results (15,451 diagnostic submissions) generated between January 1, 2008 and December 31, 2014 at the North Carolina State University, College of Veterinary Medicine Vector Borne Disease Diagnostic Laboratory were retrospectively reviewed. Bartonella henselae, B. vinsonii subsp. berkhoffii, and/or B. koehlerae seroreactivity was determined by indirect immunofluorescent antibody (IFA) testing. For a subset of accessions a more comprehensive diagnostic CVBD serology panel (Ehrlichia spp., Anaplasma spp., Rickettsia spp., Borrelia burgdorferi¸ Babesia spp. and Dirofilaria immitis) was performed using SNAP® 4Dx® (IDEXX Laboratories) or IFA.

Overall, 3.26% of accessions were Bartonella spp. seroreactive, a percentage that is comparable to seroreactivity for other CVBDs among US canine population‐wide serosurveys. Seroreactivity to B. henselae (2.13%) and B. koehlerae (2.39%) antigens was detected more frequently than seroreactivity to B. vinsonii subsp. berkhoffii (1.42%, P < 0.0001) antigen. Intact male dogs were more likely to be seroreactive (5.04%) than neutered males (2.87%, P < 0.0001; OR 1.80, 95% CI 1.37–2.35, P < 0.0001) or intact or spayed females (3.22%, P = 0.0003; OR 1.59, 95% CI 1.23–2.05, P = 0.0004). Mixed or non‐AKC breed dogs were more likely to be seroreactive (4.45%) than purebred dogs (3.02%, P = 0.0002; OR 1.49, 95% CI 1.21–1.85, P = 0.0002). There was no seasonal seroreactivity trend for any of the three Bartonella spp. Geographic patterns of seroreactivity did not correspond with other regional CVBD patterns. Rather, seroreactivity was distributed broadly across North America with some areas of apparent higher exposure risk (B. henselae in WA and CT, B. koehlerae in MO). Co‐exposure to other vector‐borne pathogens was common, with associations between B. henselae and B. vinsonii subsp. berkhoffii and multiple other CVBD pathogens. B. koehlerae seroreactivity was associated with Ehrlichia spp. (OR 3.31; 95% CI 2.43–4.51, P < 0.0001) and Rickettsia spp. (OR 2.72; 95% CI 1.57–4.71, P = 0.0002) exposures, but was not associated with the pathogens transmitted predominantly by Ixodesscapularis (B. burgdorferi and Anaplasma spp.).

We conclude that Bartonella spp. exposures may occur throughout North America and may be seen at any time of year. Dogs exposed to other CVBD, male intact dogs, and mixed breed dogs are at higher risk for Bartonella exposure. Fleas and multiple tick species are proposed vectors for canine bartonellosis; our seroepidemiological analyses suggest that there may be multiple vectors for Bartonella infection in dogs, or that the major vector may depend on local geographic, environmental, or host factors.

Impact of Immune Suppression on Dogs with Treated Select Vector Borne Disease Agents

M. Sato, J. Veir, S. Shropshire, M.R. Lappin

Colorado State University, Fort Collins, CO, USA

A 4‐week course of doxycycline is the recommended treatment for Ehrlichia canis infection in dogs. Although some studies report clearance of the infection after antibiotic therapy, others have documented persistent infection even after treatment. Therefore, concerns for recrudescence of disease arise when an immunosuppressive drug for treatment of an immune mediated disease in a dog previously infected with E. canis. The purpose of this study was to determine whether administration of prednisolone and cyclosporine at a standard dose would induce clinical or laboratory evidence of reactivated E. canis in dogs previously treated for the infection.

Seven beagles that were experimentally infected with E. canis and administered doxycycline at 10 mg/kg, PO daily for 4 weeks were included in this study. Three of the dogs were also infected with Anaplasma platys and Babesia vogeli and were administered two doses of imidocarb at 6.6 mg/kg, IM 2 weeks apart prior to enrollment. Each dog was administered prednisolone (Dose range 0.94 – 1.25 mg/kg, PO, q12 hours) in conjunction with cyclosporine whose dosages were adjusted based on the trough levels (Final dose range 5.3 – 8.3 mg/kg, PO, q12 hours) for 6 weeks. Clinical signs (fever, lethargy, and anorexia), complete blood count (CBC), polymerase chain reaction (PCR) for DNA of E. canis, A. platys, and B. vogeli in blood, anti‐E.canis indirect fluorescent antibodies (IFA) titers, and flow cytometry for antiplatelet antibodies were monitored.

All dogs were negative on PCR but positive on IFA prior to immunosuppression and completed the immunosuppressive protocol. No recrudescence of E. canis infection was detected based on clinical signs or results of CBC, PCR, IFA and flow cytometry for anti‐platelet antibodies. Anti‐E.canis IFA titers were negative in 5/7 dogs at the end of immunosuppressive protocol and were negative 6 months after the protocol in 5/5 dogs. All dogs were negative via PCR for E. canis, A. platys and B. vogeli throughout immunosuppression.

A 4‐week course of doxycycline with or without supplemental imidocarb cleared E. canis infection for at least 6 months after immunosuppression for 6 weeks. A risk of reactivation of E. canis infection with immune suppression in previously treated dogs was considered minimum.

Immune Complex Dissociation Methods for Detecting Heartworm Antigen and Their Utility in Assessing Patient Samples

R. Alleman1, A. Arguello‐Marin2, J. Drexel2, J. Liu2, M. Beall2, R. Chandrashekar2

1Lighthouse Veterinary Consultants, Alachua, FL, USA, 2IDEXX Laboratories, Inc., Westbrook, ME, USA

Heartworm antigen testing is routinely performed to diagnose heartworm infections in dogs. The original diagnostic tests employed immune complex dissociation (ICD) methods to release antigen that was bound by endogenous canine antibodies. However, ICD methods were discontinued in the majority of commercially available assays due to improvements in assay technology (O'Connor T, et al. “Immunoassay applications in veterinary diagnostics.” In: Wild D, ed. The Immunoassay Handbook. 4th ed. Great Britain: Elsevier, 2013: 623–645). Interference from immune complexes is not unique to heartworm diagnostics and various methods are available for ICD. The purpose of this study was to evaluate different methods of ICD for the detection of free heartworm antigen by microtiter plate ELISA and evaluate the performance of ICD in pet dogs. The original, USDA licensed PetChek® Heartworm Test (IDEXX Laboratories, Inc.) utilized pepsin at an acidic pH for ICD prior to antigen testing. Using a set of experimental samples, the pepsin method was compared to heat treatment, and acid treatment. All methods were found to release complexed antigen and produce positive results when the post‐treatment samples were tested for heartworm antigen by microtiter plate ELISA. While heat treatment required a minimum of 600 μL of either serum or plasma, the pepsin and acid methods needed only 50 μL of the patient sample. To extend this to a field population, samples from 1115 patients were submitted to IDEXX Laboratories between 2014 and 2016 for investigation of discrepant heartworm results. Samples were evaluated with and without pepsin ICD using PetChek® Heartworm. A total of 112 patients became antigen positive with the ICD protocol only, 122 were positive with and without ICD but demonstrated at least a 50% increase in plate absorbance (OD), and 109 were positive both with and without ICD but had minimal change in plate OD following ICD. The remaining 772 patients did not have detectable antigen with and without ICD. In a prospective study, dogs receiving pre‐adulticide heartworm treatment consisting of a macrocyclic lactone in combination with doxycycline were evaluated in the same way. Serial samples from 12 dogs revealed that 3 had a reduction in free circulating antigen within 4 weeks of initiating treatment. In all cases, free antigen levels could be recovered with ICD. Heartworm antigen testing with ICD can be a valuable adjunct diagnostic for those patients who have discrepant results, have intermittently used a macrocyclic lactone, or are receiving treatment.

Fecal Shedding of Parvovirus Following Modified Live Feline Panleukopenia Virus Vaccination in Adult Healthy Cats

M. Bergmann1, S. Schwertler2, U. Truyen3, S. Speck3, S. Reese4, K. Hartmann1

1Clinic of Small Animal Medicine, LMU Munich, Munich, Bayern, Germany, 2Clinic of Small Animal Medicine, Munich, Bayern, Germany, 3Veterinary Faculty (Virology) Leipzig, Leipzig, Bayern, Germany, 4Institute for Anatomy, Histology and Embryology, LMU Munich, Munich, Bayern, Germany

Fecal samples positive for canine parvovirus (CPV) have been reported in apparently healthy dogs after modified‐live vaccination. There is no information about fecal shedding of feline panleukopenia virus (FPV) in cats after modified‐live vaccination. Aim of this study was to provide information about parvovirus DNA‐shedding in recently vaccinated, adult, healthy cats and to assess related factors using Chi‐squared and Fisher′s exact test. Forty cats were included in the study and were vaccinated with a commercial FPV modified‐live vaccine. Feces of cats was tested for the presence of parvovirus DNA on day 7, 14, 21, and 28 by quantitative real‐time PCR. Pre‐ and post‐vaccination serum antibody titers were measured by hemagglutination inhibition on day 0, 7, and 28. Almost one‐third of the cats (12/40; 29.3%; 95% CI: 18.0–45.6) shed parvovirus DNA during the post‐vaccination period. Parvovirus DNA was detectable up to 28 days after immunization. Parvovirus DNA‐shedding was significantly associated with a lack of protective pre‐vaccination antibody titers (P = 0.016; odds‐ratio (OR): 6.44; 95% CI: 1.44–28.89) and with post‐vaccination titer increases (P = 0.029; OR: 5.00; 95% CI: 1.17–21.39). Vaccinated cats can shed of parvovirus DNA at least up to 4 weeks after vaccination and shedding occurs more commonly in cats without protective titers. Post‐vaccinal shedding thus can interfere with diagnosis in recently vaccinated cats that are presented with clinical signs suspicious for of panleukopenia.

Effects of a Liposome‐TLR Mucosal Immune Stimulant on Kittens Infected with Feline Herpesvirus 1

E.T. Contreras, S. Dow, M.R. Lappin

Colorado State University, Ft Collins, CO, USA

Feline herpesvirus 1 (FHV‐1) is the most common feline infectious disease in cats, with over 90% of cats housed in group settings being infected. Liposome‐TLR ligand complexes (LTLC) have been shown in mice to enhance innate immunity and induce non‐specific protection against both viral and bacterial diseases. A new formulation of LTLC to be used as a mucosal immune stimulant (MucosImmune) has been shown to be safe and capable of stimulating local immunity in cats. The objectives of this study were to determine whether MucosImmune administered prior to FHV‐1 experimental inoculation or at the first signs of clinical illness post‐inoculation in kittens could lessen clinical signs and shedding of FHV‐1 when compared to untreated controls.

A total of 12 female and 13 male, 12‐week‐old purpose‐bred kittens that were negative for FHV‐1 antibodies in serum and for FHV‐1 DNA in cells collected from the oropharynx by swab were divided into three groups housed in three separate rooms. Thirteen kittens were randomized into Group A (n = 7; prevention) or Group B (n = 6; treatment). The 12 kittens assigned to Group C served as untreated controls. Group A was administered MucosImmune (0.2 mL each nostril, 0.6 mL into the oropharynx) 24 hours prior to FHV‐1 inoculation. On Day 0, all kittens were sedated and administered 105.3 TCID50 of FHV‐1 divided between the nares and oropharynx. Group B kittens were administered one dose of MucosImmune the first day that clinical signs were noted with a second dose administered 24 hours later. Two trained, masked observers applied a standardized clinical scoring system detailing ocular and respiratory clinical signs to each room for 30 minutes each day for a total of 28 days. Mucosal cells were collected from the caudal pharynx of each kitten from Groups A and C on Days 7, 14, 21, and 28, and from each kitten from Group B after clinical signs were recognized but prior to the first MucosImmune dose, 7, 14, and 21 days after the first treatment, and then on Day 28 after FHV‐1 inoculation. Total DNA was extracted from the oropharyngeal swabs, and quantitative PCR assays were performed for FHV‐1 DNA and GAPDH with results presented and compared amongst groups as a ratio. The Wilcoxon rank‐sum test was used for comparisons amongst groups within the time periods Days 1 to 14, Days 15 to 28, and Days 1 to 28. Time to resolution of illness (total score ≤ 1) was also compared amongst groups.

No significant differences were found in total combined scores in Group A as compared to Group C kittens. When compared to Group C, total ocular scores in Group A kittens were significantly lower in Days 15 to 28 (P = 0.04) and lower conjunctivitis scores were detected in all periods (P = 0.01; P < 0.001; P < 0.001). When compared to Group A, total respiratory scores in Group C kittens were significantly lower in Days 15 to 28 (P < 0.001) and Days 1 to 28 (P < 0.001). Clinical illness was of shorter duration in Group A kittens (median 17.5 days, range 10 to over 28 days) as compared to Group C kittens (median 22, range 20 to 27 days) but the result was not significantly different (P = 0.13). On Days 21 and 28, FHV‐1/GAPDH ratios for Group A were significantly lower (P = 0.01) than those for Group C. When Group B results were compared to those of Group C, no significant differences were found in total combined scores, total ocular scores, total respiratory scores, or time to resolution. However, Group B kittens had significantly less pyrexia on Days 1 to 28 (P = 0.03) as compared to Group C.

Administration of MucosImmune lessened some clinical signs of FHV‐1 in both experiments in this model and decreased FHV‐1 shedding when administered 24 hours prior to inoculation with FHV‐1.

Cell Block Cytology Applied to Sporotrichosis Diagnosis

F.F. Gonsales1, N. Coelho Couto de Azevedo Fernandes2, J.M. Guerra2, A. Moreto1, J. Takahashi2, S.M.P. da Silva3, N. Benites1

1School of Veterinary Medicine, University of São Paulo, São Paulo, Sao Paulo, Brazil, 2Adolfo Lutz Institute, São Paulo, Sao Paulo, Brazil, 3Adolfo Lutz Institute, São Paulo, Espirito Santo, Brazil

Sporotrichosis is the most frequent subcutaneous mycosis detected in urban areas, it is caused by zoonotic pathogenic species of Sporothrix schenckii complex that affects especially cats. In Brazil, due to a large outbreak of sporotrichosis in the city of Guarulhos (State of São Paulo), the disease has become notifiable. Clinically the disease is characterized as a subcutaneous mycosis, with papular lesions with nodular feature or ulcero‐gummy. The definitive diagnosis is obtained by the isolation of the fungus in culture. However, the result can take weeks, resulting in a delay to start the appropriate treatment and disease could spread through other community members. The aim of this study was standardization of the diagnosis of feline sporotrichosis by Cell Block Cytology (CB). CB consists of immediate wet cell fixation in preservative liquid based cytology solution (Surepath, BD), and histological processing. We compared two collecting methods: fine needle aspiration (FNA) and exfoliative brushing (EB) with an endocervical brush. Ten samples of skin lesions of nine suspects cats were collected, (five samples with each method). EB allowed a cell thick sediment, and production of several histological sections, with a slide stained with periodic acid Schiff method. All samples collected by this method showed positive results for fungal structures. FNA cases revealed low cellularity and impossible histological processing. This standardization shows that CB processing is possible with EB of lesions and fixation in cytology preservative solution, allowing storage for up to a month after harvest and further testing, as immunohistochemistry. CB is an alternative for sporotrichosis diagnosis.

A Rapid In‐Clinic Test Detects IgM Antibodies to Four Different Leptospira Serovars

J. Lizer1, M. Loenser2, A. Weber1, P. Meeus1, M. Krecic3

1Zoetis, Kalamazoo, MI, USA, 2Zoetis, Parsippany, NJ, USA, 3Zoetis, Morristown, NJ, USA

Canine leptospirosis, a disease of worldwide distribution and significant zoonotic potential, is an acute bacterial infection of dogs caused by pathogenic spirochetes belonging to the genus Leptospira. Timely diagnosis of canine leptospirosis is important because early initiation of antibiotic therapy provides the greatest opportunity for a successful patient outcome. However, diagnosis is hindered by vague clinical signs and by diagnostic methods that can be difficult to interpret and require a lag time for results to be reported. A lateral flow test (WITNESS® Lepto, Zoetis) has been developed for the detection of IgM antibodies to Leptospira, offering the advantage of a result in 10 minutes at the point‐of‐care without the need for specialized equipment or expertise. Our hypothesis is WITNESS Lepto can detect IgM to four significant Leptospira serovars and will yield positive results earlier or concurrent to that of the microscopic agglutination test (MAT), the serological reference method for diagnosis of leptospirosis.

Thirty‐two beagle dogs approximately six months of age were experimentally infected on days zero, one, and two with serovar Canicola, Grippotyphosa, Icterohaemorrhagiae, or Pomona (n = 8 dogs/serovar). Sera were collected and evaluated for leptospiral antibodies with WITNESS Lepto and MAT at three to four day intervals. A minimum four‐fold increase of antibody titer (≥ 1:400) for any serovar of a 7‐serovar MAT panel indicated seroconversion. This study was approved by the Zoetis Institutional Animal Care and Use Committee.

Seroconversion was detected in all dogs by days 10 and 14 for WITNESS Lepto and MAT, respectively, and as early as day 4 for both methods. At day 7, WITNESS Lepto yielded positive results for 28/32 dogs (87.5%) whereas MAT yielded positive results for only 22/32 dogs (69%). In conclusion, WITNESS Lepto detects IgM to the four serovars included in this study and detects seroconversion earlier or concurrent to that of MAT. At the point‐of‐care, WITNESS Lepto may be useful as a screening test for acutely ill dogs suspected to have leptospirosis.

Relationship Between Histoplasma Capsulatum Urine EIA Results and Outcome in Cats Treated for Histoplasmosis

J. McGill1, G. Norsworthy2, A. Hanzlicek3, K. KuKanich4, A.K. Cook5

1College of Vet Medicine, Louisiana State University, Baton Rouge, LA, USA, 2Alamo Feline Health Center, San Antonio, TX, USA, 3College of Veterinary Medicine, Oklahoma State University, Stillwater, OK, USA, 4College of Vet Medicine, Kansas State University, Manhattan, KS, USA, 5College of Vet Med, Texas A&M University, College Station, TX, USA

Infection with Histoplasma capsulatum is routinely diagnosed in cats using a quantitative enzyme immunoassay (EIA) performed on urine +/or serum. Any level of antigenuria/antigenemia is considered indicative of active infection, and urine EIA has a > 90% sensitivity. The objective of this study was to determine if urine EIA results at the time of diagnosis (prior to the administration of antifungal agents) were predictive of outcome (i.e., survival to 6 months) for cats with histoplasmosis. Pre‐treatment urine EIA results were available for 50 cats examined at four institutions between April 2012 and December 2015; 35 cats were alive at Day 180, 12 had died (median survival time: 24 days; range 2–124 days) and three were lost to follow up. For statistical purposes, positive results below the level of quantification (19 ng/mL) were listed as 19.1 ng/mL. The median EIA at the time of diagnosis for cats alive at 6 months was 5 ng/mL (range: 0–19.1); this was similar to findings for the non‐survivors (median: 7.29 ng/mL; range of 0.78–19.1; P = 0.54) [Figure 1]. Surviving cats were significantly younger (mean age: 6.9 yrs) than non‐survivors (mean age 9.9 yrs; P = 0.03) but median body weights (3.8 kg v 3.6 kg) and rates of pulmonary involvement (22/35 v 9/12) were similar. These findings indicate that urine EIA results at the time of diagnosis are not predictive of outcome in cats with histoplasmosis and should not be used as a prognostic indicator.

Prevalence and Phylogeography of Methicillin Resistant Staphylococcus spp. Carriage in Companion Animals in Brisbane, Australia

E. Meler, R.S. Magalhaes, R. Rochelle, T. Farry, E. Bennett, T. Maguire, J. Gibson

The University of Queensland, Gatton, Qld, Australia

Methicillin resistant Staphylococcus spp. (MRS) carriage rates in companion animals have been well described in North America and Europe but not investigated in Australia. The aims of this study were to quantify the overall carriage rate of MRS in companion animals presenting to two major veterinary hospitals in the Brisbane area during 2016 and to identify geographical differences in carriage rates and risk factors for MRS carriage between a rural (R) and urban (U) veterinary hospital.

Nasal and rectal swabs were taken from animals within 24 h of hospital admission. Following growth on selective media (Brilliance Staph Agar ™), PCR for mec‐A and nuc genes were performed to confirm methicillin resistance and differentiate S. aureus (MRSA) from S. pseudintermedius (MRSP). Medical data collected included residential address, signalment, previous drug use, surgical procedures, and veterinary visits/hospitalisations over the prior year. Risk factors were evaluated using Stata Corp ™. Geographical variation in carriage analysis used semivariograms and OpenBUGS™.

Overall prevalence was 7.2% (19/265) in dogs, 1.6% (1/61) in cats. No (0/45) horses carried MRS. MRSP carriage was found in 19 dogs and none of the cats. MRSA carriage was identified in one dog (also MRSP carrier) and in one cat. We found a statistically significant difference in MRSP carriage rate between site R (3.2%) and site U (24.4%). Significant spatial variation was found in MRSP carrier dogs explained by clinical history and contextual factors of dog's residence.

In conclusion, major difference in canine MRSP carriage rates may exist between urban and rural areas.

Prevalence of Vector‐borne Diseases in Healthy Dogs in Oklahoma

E. Rupert1, L.A. Nafe1, P. DeMars2, L. Holland2, L. Sypniewski2, S. Little2

1Oklahoma State University, Stillwater, OK, USA, 2Center for Veterinary Healthy Sciences, Oklahoma State University, Stillwater, OK, USA

Vector‐borne diseases are common in dogs and result in a multitude of clinical signs with variable severity between individual patients. The purpose of this study was to determine the PCR prevalence of Ehrlichia spp., Anaplasma spp., Rickettsia spp., Babesia spp., Hepatozoon spp., Mycoplasma spp., Babesia spp., Leishmania spp., Neorickettsia spp., and Bartonella spp. in presumptively healthy dogs presenting to the Oklahoma State University VTH for annual wellness evaluation. A secondary objective was to correlate PCR results with historical information, physical examination, and clinicopathologic results. All dogs enrolled in the study had a client survey completed, SNAP® 4Dx Plus® (IDEXX), CBC, chemistry panel, and vector‐borne disease PCR. One hundred and twelve dogs (61 females, 51 males) with a median age of 5 years (range: 0.75–14 years) were enrolled in the study between July 2016 and November 2016. SNAP® 4Dx Plus® results showed that 13 dogs (11.6%) were positive for Ehrlichia spp., while only 1 of 13 was Ehrlichia PCR positive. Eight dogs (7.1%) were PCR positive for a vector‐borne pathogen, with 6/8 infected with Candidatus Mycoplasma haematoparvum, 1 with Mycoplasma haemocanis, and 1 with Ehrlichia ewingii. Four of 8 infected dogs had normal CBC and chemistry results, while 2/8 dogs had clinicopathologic abnormalities potentially attributed to vector‐borne disease (thrombocytopenia 62,000/uL and hyperglobulinemia 4.0 g/dL). Anemia was not documented in any dog infected with Mycoplasma spp. The prevalence of vector‐borne disease in healthy dogs presenting to the Oklahoma State University VTH is fairly low, with Mycoplasma spp. infection being most common.

Prevalence of Anaplasma phagocytophilum Infection in Feral Cats in Massachusetts

E. Galemore, M. Labato, E. O'Neil

Tufts University, North Grafton, MA, USA

Anaplasma phagocytophilum infection has been commonly associated with clinical disease in humans and dogs. However, A. phagocytophilum infection has not routinely been associated with clinical disease in cats. Recent reports have shown that Anaplasma infection can cause significant illness in cats. The objective of this study was to determine the prevalence of A. phagocytophilum infection in adult feral cats in Massachusetts, an endemic area for A. phagocytophilum and its tick vector Ixodes scapularis.

Blood samples were collected between June and December 2015 from 175 apparently healthy adult feral cats that were presented to trap and release spay/neuter centers in Massachusetts. Complete blood count, blood smear evaluation, SNAP® 4Dx® (IDEXX), and A. phagocytophilum PCR were performed on all samples to document acute infection (PCR positive and/or inclusions observed on blood smear) and exposure to A. phagocytophilum (SNAP® 4Dx® positive for A. phagocytophilum antibodies).

The prevalence of exposure to A. phagocytophilum in feral cats in Massachusetts was 9.71%, whereas the prevalence of acute infection was 6.94%. All blood smears were negative for Anaplasma inclusions; therefore, acute infection was defined as testing positive on PCR analysis.

The prevalence of A. phagocytophilum exposure in feral cats approaches 10% and is significantly higher than the previously reported prevalence of 4.3% in the United States (P < 0.001%). A. phagocytophilum infection may be an emerging infectious disease in felines. Further research is needed to determine the prevalence of clinical illness associated with A. phagocytophilum infection in cats living in endemic areas.

Opportunistic Fungal Infections Associated with Cyclosporine Use in Dogs with Immune‐Mediated Diseases

B.B. McAtee1, K. Cummings2, A.K. Cook3, J.A. Lidbury4, J. Heseltine1, M. Willard1

1Texas A&M Veterinary Teaching Hospital, College Station, TX, USA, 2Department of Veterinary Integrative Biosciences, Texas A&M University, College Station, TX, USA, 3Texas A&M University College of Vet Med, College Station, TX, USA, 4Gastrointestinal Laboratory, Texas A&M University, College station, TX, USA

Opportunistic fungal infections (OFIs) have previously been reported in dogs undergoing immunosuppressive therapy. However, the incidence of OFIs and risk factors for their development in dogs undergoing immunosuppressive treatment are unknown. The aim of this study was to determine the proportion of dogs diagnosed with certain immune‐mediated diseases treated with immunosuppressive drugs that develop OFIs during follow‐up and to determine if particular drug(s) were a risk factor for OFIs.

Medical records of dogs presenting to the Texas A&M University, Veterinary Teaching Hospital between January 2008 and December 2015 diagnosed with IMHA, IMPA, ITP, and/or Evans’ syndrome were retrospectively reviewed. Age, sex, and breed of patients were evaluated. Dogs were excluded if they were euthanized, died, or lost to follow up within 30 days of initiation of immunosuppressive treatment. For each dog the immunosuppressive agent(s) that they were treated with and whether or not they developed an OFI during follow up was recorded. Multivariable logistic regression was used to determine the association between particular immunosuppressive agents and the development of OFI. Statistical significance was set at P < 0.05.

Two hundred and forty‐six dogs diagnosed with selected immune‐mediated diseases were retrospectively included in the study: 105 with immune mediated hemolytic anemia, 62 with immune mediated thrombocytopenia, 61 with immune mediated polyarthritis, and 18 with Evans’ syndrome. Sixteen patients (6.5%) were diagnosed with an OFI. Multivariable logistic regression analysis showed that the odds of developing an OFI were significantly higher among dogs that received cyclosporine (OR, 4.4; 95% CI, 1.2–16.2; P = 0.02), relative to dogs that did not receive cyclosporine and that OFI were significantly more likely in male dogs than in female dogs (OR, 4.1; 95% CI, 1.4–12.4; P = 0.01).

OFIs were an important complication in these dogs with select immune‐mediated diseases. Patients were significantly more likely to develop an OFI if they had been treated with cyclosporine and if they were male.

Staging Feline Leukemia Virus Infections Using Quantitative p27 ELISA and Real‐Time PCR Results

M. Beall1, J. Buch1, R. Cahill1, G. Clark1, J. Hanscom1, C. Leutenegger2, R. Chandrashekar1

1IDEXX Laboratories, Inc., Westbrook, ME, USA, 2IDEXX Laboratories, Inc., West Sacramento, CA, USA

Feline leukemia virus (FeLV) is an oncogenic retrovirus of cats. Studies of FeLV infected cats have demonstrated a correlation between outcome of infection and viral RNA and proviral DNA loads. It has been proposed that a similar correlation exists for p27 antigen concentration. The purpose of this study was to utilize a next‐generation microtiter plate ELISA to quantify p27 antigen concentrations and compare that with real‐time PCR results for a population of samples obtained from the field. A total of 353 feline samples submitted between 2010 and 2015 to IDEXX Laboratories for RealPCR™ testing that included FeLV were utilized for the study. Matched plasma that had been stored frozen was tested using the PetChek® FeLV 15 ELISA (IDEXX Laboratories, Inc.) for p27 antigen. Recombinant p27 protein was used to generate a standard curve for quantifying the results. Of the 89 FeLV PCR positive samples, 27 had crossing points (Cp) less than or equal to 21 and all had p27 concentrations greater than 50 ng/mL (average 286 ng/mL). Twenty‐eight PCR positive samples had Cp greater than 30 and all had p27 concentrations less than 10 ng/mL, with 19 having no detectable antigen. Of the 264 PCR negative samples, 24 samples had p27 antigen concentrations from 1–30 ng/mL. This study demonstrates a potential correlation between proviral DNA loads and the concentration of circulating p27 antigen. Results from the combined diagnostic methodologies may provide more objective information for staging (progressive or regressive‐latent or regressive‐focal) and monitoring FeLV infections.

Worldwide Clinic‐Based Serologic Survey of FIV Antibody and FeLV Antigen in Cats

J. Buch, M. Beall, T. O'Connor, R. Chandrashekar

IDEXX Laboratories, Inc., Westbrook, ME, USA

Feline Immunodeficiency Virus (FIV) and Feline Leukemia Virus (FeLV) are globally distributed retroviruses capable of inducing fatal disease in cats. FeLV is spread primarily via saliva among cats in close casual contact whereas FIV is most commonly transmitted between combative cats through deep bite wounds. Rapid test methods for detection of FIV antibody and FeLV p27 antigen in blood are well established as standard practice among veterinarians to screen cats for these diseases. The purpose of this observational study was to evaluate macrogeographic trends in FIV and FeLV test results in pet cats over a 9‐year period between 2008 and 2016. The data were mined from a comprehensive international database of SNAP® Feline Triple®, SNAP FIV/FeLV Combo, and SNAP FIV/FeLV Combo Plus field results and used to assess the frequency of singular and co‐infections of these retroviruses in cats. Results were obtained from more than 2.9 million samples distributed across more than 26,000 unique postal codes from 68 countries. Countries were grouped geographically into 7 global regions: North America (3 countries, n = 2,538,792); Caribbean (8 countries, n = 6,882); Latin America (9 countries; n = 9,984); Northern Europe (20 countries, n = 95,800); Southern Europe (12 countries, n = 206,157); Middle East‐Africa (6 countries, n = 4,787 samples); and Asia‐Pacific (10 countries, n = 81,201).

The weighted average and range for FIV positive rates by region were: North America (Avg = 5.0%, Range = 3.5–5.3%); Caribbean (12.6%, 7.7–27.7%); Latin America (7.4%, 2.5–19.7%); Northern Europe (7.5%, 1.6–19.4%); Southern Europe (12.4%, 5.6–20.2%); Middle East‐Africa (14.2%, 9.2–24.1%); and Asia‐Pacific (13.0%, 3.4–18.8%). The weighted average and range for frequency of FeLV positives by region were: North America (Avg = 3.9%, Range = 1.6–4.1%); Caribbean (9.2%, 0.5–18.7%); Latin America (12.6%, 2.5–19.7%); Northern Europe (7.5%, 1.6–19.4%); Southern Europe (12.4%, 5.6–20.2%); Middle East‐Africa (14.2%, 7.1–42.9%); and Asia‐Pacific (6.1%, 2.7–20.7%). The weighted average and range for incidence of FIV‐FeLV co‐infections by region were: North America (Avg = 0.5%, Range = 0.2–0.5%); Caribbean (1.9%, 0.0–8.5%); Latin America (1.6%, 0.5–9.2%); Northern Europe (0.6%, 0.0–1.9%); Southern Europe (1.7%, 0.0–5.4%); Middle East‐Africa (2.0%, 0.2–6.3%); and Asia‐Pacific (1.8%, 0.6–4.6%).

This study represents the broadest worldwide feline retroviral serologic survey reported to date utilizing data from a single test platform. These results along with future monitoring of this data may be useful in developing a global view of the efficacy of various control and eradication strategies being employed for FeLV and FIV across different regions of the world.

Evaluation of a Rapid Immunoassay for Point‐of‐Care Detection of Bacteria in Cat Urine

J. Daniels1, N. Martin2, J. Byron2

1The Ohio State University, Columbus, OH, USA, 2College of Veterinary Medicine, The Ohio State University, Columbus, OH, USA

Lower urinary tract signs in cats (LUTS) are more frequently due to sterile inflammation rather than infection. Practitioners may forgo microbiological culture due to cost, and prescribe antimicrobial agents empirically, incurring unnecessary selection pressure and side effects on most patients. Therefore, a rapid screening test for detection of bacteria in cat urine could direct appropriate therapy.

A “dip‐stick” style immunoassay to detect bacteria was performed on feline urine (n = 100) submitted for routine bacterial culture in a Veterinary Teaching Hospital. The immunoassay was performed according to manufacturer instructions and culture was performed quantitatively on Columbia agar with 5% sheep's blood, with a maximum 72 hour incubation time at 35°C in ambient air. Bacteria were identified using matrix‐assisted laser‐desorption time‐of‐flight mass spectrometry.

Eleven samples yielded ≥ 100,000 cfu/mL in culture, ten of which (90.9%) were positive on the immunoassay, and consisted of E. coli (n = 5), Enterococcus spp. (n = 5), and Staphylococcus felis (n = 1). Three samples yielded ≤ 200 cfu/mL, one of which (33.3%) was positive on the immunoassay, and consisted of E. coli (n = 2), and Staphylcoccus felis + E. coli (n = 1). The immunoassay was positive for two (2.3%) culture negative samples. The immunoassay correctly categorized the bacteria as gram‐negative or gram‐positive for eleven (100%) of the concordant samples.

When samples with ≤ 1000 cfu/mL were considered negative, sensitivity and specificity of the immunoassay were 90.9% and 96.6% respectively, with positive and negative predictive values of 76.9% and 98.9%. Given the low prevalence of bacterial cystitis in cats, these data suggest that the immunoassay is a useful screening tool for cats presenting with LUTS.

Enterococcus Faecium Inhibits Feline T. Foetus Growth and Adhesion to Intestinal Epithelial Cells

R.E. Dickson, D.A. Bemis, M. Katherine Tolbert

College of Veterinary Medicine, University of Tennessee, Knoxville, TN, USA

Tritrichomonas foetus (Tf) causes chronic diarrhea in cats. Limited treatment options and increasing resistance support investigation of alternative treatments. Probiotics reduce proliferation and pathogenicity of other intestinal pathogens. However, little work has been done to investigate their efficacy against Tf infection. Demonstrating that probiotics reduce Tf infection could have important clinical implications in treating feline trichomonosis.

Entercoccus faecium (Efm) was cultured from a commercially available probiotic. Its effect on Tf proliferation was evaluated throughout log phase growth. The potential of Efm to reduce Tf adhesion to the intestinal epithelium was analyzed with a co‐culture model using porcine intestinal epithelial cells (IPEC‐J2). Tf and IPEC‐J2 were pre‐treated with Efm prior to co‐culture to evaluate both prophylactic and post‐infection treatment with Efm. Tf adhesion and cytotoxicity were evaluated using fluorescent microscopy and crystal violet spectrophotometric analysis. Data were analyzed using SigmaStat.

Efm‐induced inhibition of Tf growth was observed at concentrations as low as 10:1 Tf:Ef (P < 0.001) and was determined to be largely pH dependent (P < 0.01). Tf‐induced cytopathogenicity, when co‐cultured with Efm prior to infection of IPEC, was unaltered. However, pretreating IPEC monolayers with Efm reduced Tf adhesion (P = 0.002), suggesting a benefit of pre‐treatment with Efm for cats at risk of Tf infection.

These results demonstrate that Efm inhibits Tf growth. Pre‐treatment of intestinal epithelial cells with Efm, but not treatment at the time of Tf infection, inhibits adhesion of Tf to intestinal epithelial cells in vitro. Studies investigating other Enterococcus spp, as well as multi‐strain probiotics, in combination with ronidazole are warranted.

A Molecular Epidemiological Survey of Gammaherpesviruses in Domestic Cats in Japan

M. Tateno, M. Takahashi, Y. Endo

Laboratory of Small Animal Internal Medicine, Joint Faculty of Veterinary Medicine, Kagoshima University, Kagoshima, Kagoshima, Japan

Gammaherpesviruses (GHVs) are members of an emerging subfamily of the Herpesviridae. The two major human GHVs, Epstein‐Barr virus (EBV) and Kaposi's sarcoma‐associated herpesvirus (KSHV), are clinically important pathogens because those viruses possess tumorigenesis potential especially in immune‐compromised patients. In feline practice, we often experience tumor cases, especially lymphomas. Feline leukemia virus (FeLV) was a major cause for the development of lymphoma in the past days, however, the majority of lymphoma cases are shifting to FeLV‐free cats. But the mechanism of tumorigenesis in these spontaneously developed lymphomas has not been well understood. Under this situation, GHV became a candidate which explain the lymphomagenesis in domestic cats. Recent study identified a novel GHV in domestic cats (Felis catus GHV, FcaGHV), and epidemiological surveys found that FcaGHVs are distributing in several countries including the United States, Australia, Singapore, Germany and Austria. In the present study, a nation‐wide molecular epidemiological study was conducted to investigate the prevalence of GHVs in Japan.

GHV‐derived DNA in blood samples was detected by degenerate nested pan‐GHV PCR targeting gB gene as previously reported. Nucleotide sequences of obtained DNA fragments were determined and BLAST analysis was carried out. Phylogenic analysis was also conducted. Risk factors relating to GHV infection on clinical parameters, including age, sex, status of retrovirus infections, bite wound history and clinical signs, were statistically analyzed using univariable (Fisher's exact test) and multivariable (Logistic regression model) analyses.

GHV‐derived DNA was detected in 1.4% (23/1,738) of Japanese domestic cats. BLAST analysis revealed that all of them showed high similarity with FcaGHVs which were isolated from domestic cats in the United States. Phylogenic analysis also revealed that all Japanese isolates formed one cluster with FcaGHV. In the univariable analysis, older age (OR, 4.93; 95% CI, 1.43–16.99; P < 0.01), male (OR, 2.51; 95% CI, 0.98–6.40; P = 0.05), FIV infection (OR, 5.35; 95% CI, 2.30–12.46; P < 0.01), FeLV infection (OR, 2.02; 95% CI, 0.74–5.50; P = 0.18) and sick cat (OR, 2.37; 95% CI, 0.80–7.02; P = 0.12) were statistically selected as risk factors for GHV infection. In the following multivariable analysis, older age (OR, 3.84; 95% CI, 1.06–13.86; P = 0.04) and FIV infection (OR, 4.11; 95% CI, 1.52–11.11; P < 0.01) were found to be significant and independent risk factors for GHV infection.

In this study, the prevalence of GHVs in Japanese domestic cats was surveyed and risk factor for GHV infection was determined. FcaGHV prevalence in Japan (1.4%) was lower than previously reported prevalence in other countries and areas (US, 16%; Europe, 16.2%). It was unable to determine the cause of this difference. However, Japanese isolates were identified to have close genetic relationship to domestic cat derived FcaGHV. This finding suggested that the virus is highly conserved in each animal species and GHV is harboring in host specific manner as well as other herpesviruses. The risk factor for GHV infection in Japanese cats was FIV infection and it was same as previously reported. But we have to clarify how FIV enhances the GHV infection and/or viremia. The clinical importance of GHVs infection in feline practice is still unknown. Further investigation will be necessary to identify the pathogenicity of GHVs in domestic cats.

Minimal Inhibitory Concentrations of Antimicrobials Against Escherichia coli Isolated from Companion Animal Urinary Tract Infections

L. Johansen1, C. Lindeman2, D. Bade3, G. Rodgers4

1Zoetis, Kalamazoo, MI, USA, 2Formerly with Zoetis, Kalamazoo, MI, Apple Valley, MN, USA, 3Microbial Research Incorporated, Fort Collins, CO, USA, 4Zoetis, Parsippany, NJ, USA

Zoetis has been conducting an ongoing surveillance program to evaluate the susceptibility trends of antimicrobial agents against canine and feline Escherichia coli (E. coli) pathogens isolated from Urinary Tract Infections (UTIs) since 2011. The data were analyzed annually and the first five years of susceptibility data for amoxicillin/clavulanic acid, ampicillin, cefovecin, cefpodoxime, ceftiofur, enrofloxacin, marbofloxacin, orbifloxacin, trimethoprim/sulfamethoxazole and cephalexin are presented here. A total of 3,300 E. coli strains, isolated as the etiological agent in cats and dogs presenting with naturally occurring UTIs and being seen at primary/general care practices were submitted to this surveillance program by 16 veterinary diagnostic laboratories throughout the United States (US) and Canada. The Minimal Inhibitory Concentrations (MICs) for all isolates were tested using a broth microdilution system (Thermo Fisher Scientific, Oakwood Village, OH) that conforms to the Clinical and Laboratory Standards Institute (CLSI) standards.

Data collected during the first 5 years of this program showed that the modes and MIC50 values remained stable (within one doubling dilution) for each antimicrobial tested against E. coli from both dogs and cats for each year of testing; MIC90 values were stable over the years tested with some exceptions. For amoxicillin/clavulanic acid and ampicillin, mode and MIC50 values remained the same (4 μg/mL) and MIC90 values ranged from8‐ > 16 μg/mL. During this time, the in vitro testing for cefovecin, cefpodoxime, and ceftiofur were steady, the mode, MIC50, and MIC90 values ranged from 0.5 to 2 μg/mL, with a few single drug, single year MIC90 values testing outside that range. The mode and MIC50 values for the fluoroquinolones (enrofloxacin, marbofloxacin and orbifloxacin) all tested at ≤0.12 μg/mL, while MIC90 values ranged from 0.06 to 4 μg/mL. For trimethoprim/sulfamethoxazole, mode and MIC50 values were stable (≤0.12 μg/mL) and MIC90 values ranged from ≤0.25 to 1 μg/mL. In vitro testing of cephalexin was included in 2015. The mode and MIC50 values for this single year of testing E. coli from both dogs and cats were both 8 μg/mL, while the MIC90 value was 16 μg/mL in cats and >64 μg/mL in dogs. Screening and phenotypic confirmation only of Extended‐Spectrum β‐Lactamase (ESBL) production in E. coli were done following CLSI standards. Over the 5 years of testing, 1.2% (23 of 1933) of the E. coli isolated from dogs and 0.8% (11 of 1367) of the E. coli isolated from cats were confirmed as ESBL‐producing strains.

Overall, antimicrobial susceptibility data were stable from 2011 to 2015 and did not show progression of resistance during that time frame.

Effect of Doxycycline Administration on Antibodies Against Different Borrelia Burgdorferi Peptides

M.R. Lappin1, S. Moroff2

1Colorado State University, Fort Collins, CO, USA, 2Antech Laboratories, Lake Success, NY, USA

Borrelia burgdorferi is transmitted by Ixodes spp. and is associated with inflammatory arthritis or nephritis in some dogs (Lyme Disease). A number of diagnostic modalities have been evaluated to aid in the diagnosis of Lyme disease in dogs, with serological testing proving to be the most beneficial. Antibodies against outer surface protein A (OspA), OspC, OspF and antibodies against the C6 peptide are currently used most frequently in commercially available serological assays. Antibodies against OspC are often the first detected in serum after a new exposure to B. burgdorferi. In humans with Ixodes spp. infestation, doxycycline treatment has been shown to block progression of B. burgdorferi infection but whether this occurs in dogs is unknown. The purpose of this study is to report the effect of doxycycline administration on serological responses to OspA, OspC, OspF, and C6 peptide in dogs experimentally infested with wild‐caught I. scapularis.

A total of 23 young adult dogs were infested with 25 female and 15 male I. scapularis wild caught in Rhode Island (48% B. burgdorferi carriage rate). All dogs had multiple fed ticks present when removed from the chambers at Day 7. Serum collected prior to infestation and weekly after infestation were assayed in a commercially available in clinic assay (SNAP4DXPlus; IDEXX Laboratories; C6 peptide) and at a commercial laboratory (Accuplex; Antech Diagnostics; OspA, OspC, and OspF). The first 3 dogs to be positive for OspC antibodies but negative for antibodies against the other 3 peptides were administered doxycycline at approximately 10 mg/kg, daily for 28 days.

By Week 10 after I. scapularis infestation, 21 of 23 dogs had developed serological evidence of B. burgdorferi infection. Of the 3 dogs with OspC antibodies alone that were administered doxycycline for 28 days, 2 dogs never developed antibodies against the other peptides and were negative for OspC antibodies by Week 2 or Week 6 after starting doxycycline administration. The other treated dog became positive to the C6 peptide alone during week 1 of doxycycline administration, was positive for 2 of the next 3 weeks, and then was negative for all B. burgdorferi antibodies for 4 weeks after cessation of treatment. In contrast, only 1 of the 18 infected, untreated dogs transiently (1 week) developed OspC antibodies but then stayed negative for all antibodies for 6 weeks, suggesting a self‐limited infection or a false positive OspC antibody titer.

This model induced serological evidence of B. burgdorferi infection of 21 dogs with 17 of 18 untreated dogs developing serological evidence of chronic infection characterized by antibodies against OspF and C6 peptide. While only 1 dog may have spontaneously limited antibody production against all 4 peptides, all 3 dogs administered doxycycline during the acute phase limited antibody production. These results should be confirmed in larger numbers of dogs, but may suggest that treatment when OspC antibodies are first detected may block development antibody responses associated with chronic B. burgdorferi infection.

Performance of a Leishmanin Skin Test for the Diagnosis of Feline Leishmaniosis

L.S.V. Sobrinho1, J.P. Vides1, K.Y. Hirata1, M. Aschar2, V. Matta2, M.D. Laurenti2, M. Marcondes3

1São Paulo State University, Aracatuba, Sao Paulo, Brazil, 2São Paulo University, São Paulo, Sao Paulo, Brazil, 3School of Veterinary Medicine, São Paulo State University, Aracatuba, Sao Paulo, Brazil

A natural resistance of cats to leishmaniosis, primarily due to a cellular immune response, is widely suggested. Leishmanin skin test (LST) has been successfully used in humans and dogs and a positive LST result indicates exposure to Leishmania and a state of cell mediated immunity. A total of 96 cats, seronegative for FeLV and FIV, presenting to a Veterinary Teaching Hospital in an endemic area in Brazil, were evaluated by LST. Leishmania infection was confirmed in 56 cats by bone marrow qPCR, and 40 non‐infected cats composed the control group. Among infected cats, 23/56 presented clinical signs and 33/56 were asymptomatic. Each cat was intradermally injected with a suspension of 4 × 107 L. chagasi promastigotes (inactivated in thimerosal solution), in the inner surface of the right thigh, and thimerosal solution was injected in the opposite limb. All control and infected cats tested negative to LST at any time reading (24, 48 and 72 h after inoculation). We expected that at least part of the 26 asymptomatic (and seronegative) cats presented a positive LST. In canine leishmaniosis (CanL) LST can detect a high proportion of asymptomatic dogs, although some of them have negative LST even in the presence of a cellular immune response. The method employed was previously used for CanL, and its potential immunogenicity was confirmed by promoting positive LST in five naturally infected and asymptomatic dogs from the same endemic area. However, the LST tested do not seem to be an adequate tool for the diagnosis of feline leishmaniosis.

Evaluation of Dried Blood on Filter Paper for Amplification of Ehrlichia Canis DNA from Dogs

Y.H. Ozekin, C. Minor, J. Hawley, M. Brewer, M.R. Lappin

Colorado State University, Fort Collins, CO, USA

Transportation of blood from remote areas can be problematic. While DNA is generally stable and a number of studies have evaluated DNA extracted from dried blood of mammals, there is minimal information determining the sensitivity of different techniques using dog blood. The objective of this study was to determine the optimal technique for extracting DNA from filter paper dried blood spots using Ehrlichia canis DNA as the target.

Blood samples from dogs were collected with owner permission for use in a vector borne agent prevalence study in Mexico. After collection, the aliquot of blood was placed in EDTA, and then 300 μL was blotted onto filter paper (VWR® Grade 413), allowed to dry, and stored in individual sterile bags. The filter paper samples were stored at room temperature, and the EDTA blood was stored at –80°C until assayed at the central laboratory. A glyceraldehyde 3‐phosphate dehydrogenase (GAPDH) quantitative PCR was used to estimate DNA yields in the different experiments. Different buffers and incubation periods were assessed in experiments using the entire filter paper in different washing procedures, and it was determined that an overnight PBS wash gave the best results (Method 1). In Method 2, a commercially available kit (QIAamp® DNA Blood Mini Kit Dried Blood Spot (DBS) protocol) was performed as per the manufacturer's instructions. Once Method 1 was optimized, 12 samples previously shown to be positive for E. canis DNA in blood by conventional PCR assay and 3 known negative samples were thawed and used to make 2 filter paper blots which were allowed to dry overnight before being subjected to DNA extraction Methods 1 and 2. The whole blood and filter paper derived DNA extracts were then assayed in the GAPDH qPCR and the Ehrlichia spp. PCR assay. A total of 11 whole blood samples were PCR positive with bands characterized as strong (n = 4), medium (n = 3), or weak (n = 4). One sample using Method 1 was excluded due to the washing of an incorrect blot.

When the GAPDH qPCR results from whole blood were compared to those from the 2 filter paper extract Methods, it was shown that both Method 1 (74 fold) and Method 2 (15 fold) had lower DNA yields. Ehrlichia canis DNA was amplified from 5 of 11 whole blood positive samples (45.5%) using Method 2 and 2 of 10 samples (20.0%) using Method 1. The filter paper samples that had been dried for several months gave the same results as those filter paper samples that had been dried for 24 hours.

While specific DNA can be extracted from blood dried on filter paper, the sensitivity will be decreased compared to whole blood. If this technique is used, the commercial kit is likely to be more sensitive than an overnight PBS wash.

Clinical and Laboratory Findings in 30 Dogs Naturally Infected with Ehrlichia Ewingii

B. Qurollo1, R. Chandrashekar2, J. Buch2, M. Beall2, C.A. Yancey1, A. Caudill1, A. Comyn1, E. Breitschwerdt1

1NCSU‐CVM, Raleigh, NC, USA, 2IDEXX Laboratories, Inc., Westbrook, ME, USA

E. ewingii is the most prevalent tick‐transmitted Ehrlichia spp. identified in dogs from the Southern US. Fever, lameness, neutrophilic polyarthritis and neurologic abnormalities are considered common clinical findings in dogs naturally‐infected with E. ewingii; however, nonclinical dogs can be PCR+. To further assess the clinicopathologic features in E. ewingii PCR+ cases, we analyzed the medical records of 30 naturally‐infected dogs. To be included in this study, dogs had clinical data, archived whole blood available from North Carolina State University‐College of Veterinary Medicine, and were E. ewingii PCR (+) and PCR (‐) for other vector‐borne pathogens including Ehrlichia spp. (E. canis, E. chaffeensis, and Panola Mountain Ehrlichia), Anaplasma spp., Babesia spp., Bartonella spp., Hemotropic Mycoplasma spp., and Rickettsia spp. Based upon diagnostic and/or retrospective research testing, all 30 dogs satisfied the inclusion criteria. Frequently reported physical exam findings included joint pain and effusions 10/30 (33.3%); heart murmur 8/30 (26.75); organomegaly 7/30 (23.3%); vomiting 6/30 (20%) and fever 5/30 (16.7%). Of the 30 E. ewingii PCR+ dogs, 28, 26 and 20 had corresponding CBCs, biochemistry panels and urinalyses, respectively. Hematological and biochemical abnormalities included increased abnormal lymphocytes 18/28 (64.3%); neutrophilia 15/28 (53.6%), of which 9/15 (60%) were characterized by a left shift; anemia 14/28 (50%), of which 8/14 (57%) were characterized as regenerative; monocytosis 13/28 (46.4%); lymphopenia 12/28 (42.9%); thrombocytopenia 11/28 (39.3%); elevated ALP 13/26 (50%) and ALT 10/26 (38.5%), of which only two dogs were receiving corticosteroids; hypoalbuminemia 9/26 (34.6%); and elevated creatinine 6/26 (23.1%). Urinalyses abnormalities included albuminuria 11/20 (55%); proteinuria with inactive sediment 6/20 (30%); and bilirubinuria 6/20 (30%). Of the 7 dogs with urine/protein/creatinine ratios, 6 (85.7%) were greater than 1.5 (range: 1.74 – 11.7). Canine vector‐borne disease IFA serology results were available for 28/30 (93.3%) dogs (no sera submitted for 2 dogs), of which 13/28 (46.4%) were E. canis IFA seroreactive (titers ≤ 1:256) and 5/28 (17.8%) were R. rickettsii (spotted fever group rickettsia) IFA seroreactive (titers ≤ 1:256). All dogs were Babesia spp., and Bartonella spp. seronegative. Using the SNAP®4Dx® Plus test kit, 23/29 (79.3%) serum or blood samples were Ehrlichia (+), 1/29 (3.4%) was Lyme (+) and all dogs were Anaplasma antibody (‐) and Dirofilaria immitis antigen (‐). The most frequent medical record diagnoses included renal disease 6/30 (20%); IMHA 5/30 (16.7%); neurologic disease 4/30 (13.3%); polyarthritis 3/30 (10%); hepatic disease 3/30 (10%); cancer 3/30 (10%); and other or undiagnosed 3/30 (30%) illnesses. Ehrlichiosis was the sole diagnosis for only 1/30 (0.033%) dogs and was listed concurrently with other diagnoses for 8 (26.6%) dogs. We hypothesize that infection with E. ewingii may contribute to co‐morbidities or function as a precipitating factor in a spectrum of clinical syndromes in dogs. This hypothesis will be tested using inferential statistics and a large diagnostic data set from IDEXX Laboratories, Inc.

Fluorescent in Situ Hybridization in Tissues from Dogs with Proliferative Urethritis

M. Borys1, S. Hulsebosch1, J. Sykes1, F. Charles Mohr1, K. Watson1, K. Simpson2, J.L. Westropp1

1UC Davis School of Veterinary Medicine, Davis, CA, USA, 2College of Veterinary Medicine, Cornell University, Ithaca, NY, USA

The cause of proliferative urethritis (PU) has not been elucidated. Chronic bacterial infections and immune‐mediated processes have been proposed. Aerobic bacterial urine cultures (ABUC) are often positive at diagnosis, but whether this is a cause or consequence of PU is unknown.

The objectives of this retrospective study were to 1) describe the clinicopathologic data of cases with PU, 2) utilize fluorescent in situ hybridization (FISH) to determine whether mucosally associated bacteria are present, and 3) describe whether an association exists between the type of inflammation and the presence or absence of bacteria on FISH evaluation.

Twenty‐two cases of biopsy‐confirmed PU were identified (mean age 8 years; range 1–15 years). There were 14 females, 8 males, and 15 breeds represented. Four dogs had positive ABUC perioperatively, 4 had positive mucosal bacterial cultures, and 11 had negative ABUC. Urine cultures were unavailable or not performed in 7 dogs. FISH analysis was performed in 13 cases; adherent/invasive bacteria were noted in 5 dogs, and tissue‐associated bacteria in 2 dogs. Of these 7 dogs, 3 had positive ABUC at the time of biopsy. The 6 FISH analyses that were negative for bacteria also had negative ABUC. One of the dogs with adherent/invasive bacteria visualized had lymphoplasmacytic and histiocytic urethritis; the remaining 12 had neutrophilic, lymphoplasmacytic, or mixed inflammation reported on histopathology. These results suggest that primary or secondary bacterial invasion may be associated with PU, despite negative ABUC; this knowledge may alter treatment of this disease.

Vitamin D Metabolism in Dogs with Calcium Oxalate Urolithiasis

E. Groth1, J. Lulich1, D. Chew2, V. Parker3, E. Furrow1

1University of Minnesota, St. Paul, MN, USA, 2The Ohio State University, Columbus, OH, USA, 3The Ohio State University Veterinary Clinical Sciences, Columbus, OH, USA

The aim of this study was to measure vitamin D metabolites in dogs with and without a history of CaOx urolithiasis.

Twenty Miniature Schnauzers, 10 Bichons Frise, and 10 Shih Tzus were included in the study. Cases consisted of 20 dogs with a history of CaOx urolithiasis and fasting urinary calcium to creatinine ratios (UCa/Cr) > 0.05. Controls consisted of 20 age‐, sex‐, and breed‐matched dogs without urolithiasis (as determined by screening radiography) and with UCa/Cr 3, 24,25(OH)2 vitamin D3, and total 1,25 (OH)2 vitamin D measured through a commercial laboratory (Heartland Assays, Ames, Iowa).

There were no statistically significant differences between cases and controls when comparing 25(OH) D3, 24,25(OH)2 D3, total 1,25(OH)2, or the ratio of total 1,25(OH)2 to 25(OH) D3. However, cases had a significantly higher ratio of 25(OH) D3 to 24,25(OH)2 D3 compared to controls (median = 1.43 and 1.17, respectively; P = 0.012). Nine of the 10 highest ratios were found in cases, but there was substantial overlap with more than half the cases falling within the range observed in controls. There was a moderate positive correlation between the ratio of 25(OH) D3 to 24,25(OH)2 D3 and UCa/Cr (r = 0.37, P = 0.017).

This data suggests that decreased conversion of 25(OH) vitamin D3 to the inactive 24,25(OH)2 vitamin D3 occurs in a subset of dogs with CaOx urolithiasis. Abnormalities in vitamin D metabolism may contribute to stone risk in some dogs.

Risk Factors for Urinary Tract Infections in Cats After Ureteral Stent and Ureteral Bypass Placement

L. Kopecny1, C.A. Palm1, K.J. Drobatz2, I.M. Balsa1, W.T.N. Culp1

1University of California, Davis, Davis, CA, USA, 2University of Pennsylvania, Philadelphia, PA, USA

Ureteral stenting and bypass are commonly used for managing benign ureteral obstructions. Objectives of this case‐control study were: 1. to determine post‐operative rate of urinary tract infection (UTI) in cats undergoing these procedures and 2. to identify risk factors associated with UTI development.

Forty‐two client‐owned cats undergoing 46 renal decompressive surgeries were included. Cats were included if they had urine cultures performed prior to and after surgery; variables were compared statistically to identify risk factors.

Median duration of follow‐up was 203 days (range 9–2184 days). Of 137 post‐operative urine cultures, 20 (15%) were positive in 14/42 (33%); 3/42 cats had more than one UTI. Nine different bacteria were identified; Enterococcus faecium (n = 5), and E. coli (n = 5) predominated. Cats undergoing bilateral surgery were more likely (OR: 4, 95% CI: 1.15; P = 0.03) to develop a UTI postoperatively compared to cats with unilateral surgery. Anuric cats were also more likely to develop a UTI compared to non‐anuric cats (OR: 5, 95% CI 1.27; P = 0.03). Pre‐existing chronic kidney disease and post‐operative urinary catheter or abdominal drain placement were not significantly associated with infection, either prior to discharge or at long‐term follow‐up.

Post‐operative UTIs occurred relatively commonly in this cohort of cats. Post‐operative urethral catheterization has been speculated to increase UTI risk, but this was not evident in this cohort of cats. Greater case numbers will be required before definitive conclusions can be made. Future work should focus on gathering greater case numbers and on the clinical significance of these infections.

Concentrated Alkaline Urine is Hostile to Uropathogenic E. Coli Growth

L. Thornton, R. Burchell, S. Burton, N. Lopez‐Villalobos, D. Pereira, I. MacEwan, A. Fong, A. Hatmodjo, M. Nelson, A. Grinberg, N. Velathanthiri, A. Gal

Institute of Veterinary, Animal, and Biomedical Sciences, Massey University, Palmerston North, Manawatu‐Wanganui, New Zealand

Uropathogenic E.coli (UPEC) are common bacteria involved in lower urinary tract infections in dogs. This study tested the hypothesis that acidic diluted urine promotes UPEC growth.

The authors obtained early morning urine samples from nine healthy bitches. Following urine filtration through a 0.2 μm biofilter, we obtained a matrix of three levels of pH (5.5, 7.0, 8.5) by three levels of urine specific gravity (USG; 1.010, 1.020, 1.030) by titrating the urine with HCl and NaOH. We inoculated the urine with known quantities of three UPEC strains, incubated at 37°C for four hours, made serial dilutions from 10−4 to 10−7, plated on nutrient agar for 24 hours and subsequently counted the numbers of colonies forming units (CFU). We only counted plates from dilutions that grew 30–300 CFU. All experiments had negative and positive controls.

The number of UPEC CFU was analyzed with a mixed linear model that includes the fixed effect of strain, pH, concentration and the interaction between the three, the random effect of dog, and inoculated dose of UPEC as a covariate. Least Square Means were obtained and used for comparisons. For all comparisons, the alpha probability error was P = 0.05.

There were significant differences in the mean (±SE) log(CFU) between the different pH 7.0 > 5.5 > 8.5 and USG 1.010 > 1.020 > 1.030. The least bacteria grew in pH 8.5 USG 1.030 and the most in acidic and neutral diluted urine. The strain of UPEC had a significant effect on growth. Differences between strains were substantial in concentrated alkaline urine.

The impact that acidifying pet food diets may have on UPEC infections should be determined.

Correlation Between Canine Urine Cortisol‐Creatinine and Protein‐Creatinine Ratios Collected at Home and in Hospital

L.E. Citron1, N.M. Weinstein1, M.P. Littman1, J.D. Foster2

1School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA, USA, 2Friendship Hospital for Animals, Washington, DC, USA

Recently, it has been demonstrated that urine‐protein‐creatinine (UPC) ratios were higher in canine urine samples collected in hospitals (IH) settings than those collected at home (AH). Stress associated in‐hospital procedures and travel to the hospital could play a role in the higher UPC observed. The purpose of this study was to evaluate urine cortisol‐creatinine ratios (UCCR), as a metric of stress, and to correlate to UPC in urine samples obtained at home and in hospital.

This was a prospective, non‐masked study. Clients obtained a free‐catch urine sample at home or off‐campus (< 12 hours prior to their appointment). Time of urine collection and total hospital visit time were recorded. A second urine sample was obtained at the conclusion of the dog's appointment in hospital. Urine creatinine, protein, and cortisol levels, and full urinalysis were measured on both urine samples. Urine protein creatinine ratios and UCCR were compared between AH and IH settings via Wilcoxon signed rank test. Correlation between UCCR and UPC, in any setting, and effect of travel/hospitalization time on change in UCCR and UPC were evaluated with Spearman's analysis.

Thirty‐six dogs were enrolled. The majority (78%) of dogs were nonproteinuric (UPC < 0.2) in both urine samples. The UCCR was significantly lower (P < 0.0001) in urine samples obtained AH (median 17.5, range 1–41) compared to IH (median 20.5, range 10–189). The UPC was not statistically different (P = 0.144) in urine samples obtained AH (median 0.02, range 0.02–2.87) compared to IH (median 0.03, range 0.01–2.64). UCCR was not significantly associated with UPC in all samples and in any collection setting. Travel time, time in hospital, and total time prior to IH urine collection was not associated with change in UCCR or UPC. In the 8 patients that were borderline proteinuric (UPC 0.2–0.5) or proteinuric (UPC >0.5), there was no statistical difference in UPC between settings (P = 0.25), however the UCCR was significantly higher (P = 0.0312) in samples collected IH (median 17.5, range 11–159) compared to AH (median 13, range 1–41).

Contrary to an earlier study, this data demonstrated that UPC was not higher when measured in urine samples collected IH compared to AH. Dogs had higher UCCR IH compared to AH, however the UCCR was not associated with UPC. Stress, as estimated through UCCR, did not affect proteinuria in these patients.

Application of Peritoneal Dialysis in Companion Animals in Turkey

H. Albasan1,2, O.S. Terzi1,2, S.D. Arun1,2, E. Kara1,2, N. Keskin1,2, I.G. Sancak2,3

1Department of Internal Medicine, Faculty of Veterinary Medicine, University of Ankara, Ankara, Turkey, 2Diskapi, Ankara, Turkey, 3Department of Surgery, Faculty of Veterinary Medicine, University of Ankara, Ankara, Turkey

Peritoneal dialysis (PD) is a renal replacement therapy that has been commonly used in human and veterinary medicine for the treatment of kidney insufficiency. The objective of this study was to assess the renal function in animals with renal failure after PD.

Renal cases presented to veterinary teaching hospital at Ankara University were treated with PD between June 30, 2014 and June 29, 2015. A total of 5 cases consisted of 3 dogs with acute renal failure of ethylene glycol toxicity (n = 1) and ingestion of unknown plants (n = 2) and of the remaining one dog and one cat were with chronic renal failure. PD was applied once daily ranged between 4 to 10 days.

Concentration of serum urea in all patients significantly decreased after the second attempt of PD but, not serum creatinine. When PD procedures completed, the concentrations of serum urea and creatinine were significantly lowered in all cases. However, those values were still above the reference range except in one dog with acute renal failure.

Complications identified during PD were peritonitis (n = 4), hypoalbuminemia (n = 4), hypokalemia (n = 3), dialysate retention or leakage from the catheter site (n = 3), edema (n = 2) and pleural effusion/chylothorax (n = 2).

This is the first report of PD procedure successfully applied as life‐saving, easily applicable, and cost effective clinical procedure to companion animals with renal failure that could be adapted by veterinary practitioners in Turkey.

Predicted Ionized Calcium Better Reflects Calcium Homeostasis Than Total Calcium in Dogs with Renal Azotemia

K.L. Boedec1, J. Danner2

1Centre Hospitalier Veterinaire FREGIS, Arcueil, Ile‐de‐France, France, 2College of Veterinary Medicine, University of Illinois, Urbana, IL, USA

Total calcium (tCa) and corrected tCa are considered unreliable predictors of calcium homeostasis, especially in dogs with renal azotemia. Unfortunately, measured ionized calcium (miCa), which truly reflects calcium homeostasis, is not readily available in practice.

Dogs with renal azotemia (i.e., creatininemia > 2.0 mg/dL and urine specific gravity < 1.030) were retrospectively identified from 751 records for inclusion in the study. In included dogs, tCa was corrected for total protein or albumin concentrations by use of published equations. Predicted ionized calcium (piCa) was calculated from routine biochemical values and age via the Canine Ionized Calcium Calculator website (available at: http://vetmed.illinois.edu/study/mars‐model/VetMed.php). Sensitivity and specificity of piCa were compared to those of tCa and corrected tCa for true hypercalcemia and hypocalcemia, as determined by miCa (McNemar's test).

Out of 40 azotemic dogs included in the study, 7 were hypercalcemic and 14 were hypocalcemic based on miCa. Sensitivities of piCa (86%), tCa (86%) and corrected tCa (100%) were equivalent to detect hypercalcemia (P = 1.0). Specificities of piCa (94%) and tCa (88%) were not statistically different (P = 0.5), but specificity of piCa was significantly higher than specificity of corrected tCa (70–76%) to diagnose hypercalcemia (P = 0.008–0.03). Furthermore, sensitivity of piCa (86%) was significantly higher from sensitivities of tCa (29%) and corrected tCa (7%) to detect hypocalcemia (P < 0.01). Specificities of piCa (96%), tCa (100%) and corrected tCa (100%) were equivalent to diagnose hypocalcemia (P = 1.0).

In dogs with renal azotemia, piCa seems more accurate to diagnose true hypercalcemia and more sensitive to detect true hypocalcemia than corrected tCa.

Urine Protein to Creatinine Ratios of Proteinuric Canine Urine Samples Stored in Multiple Container Types

P. Moyle, A. Specht, R. Hill

College of Veterinary Medicine, Univeristy of Florida, Gainesville, FL, USA

Adsorption of protein to the surface of polymer plastic or glass could affect urine protein:creatinine ratios. Homopolymer polypropylene specimen containers mitigate protein adsorption but urine samples collected by owners are sometimes stored in propylene copolymer plastic food storage or glass containers either at room temperature, refrigerated, or frozen. The primary aim of this study was to determine whether protein:creatinine ratios (UPC) are affected by these storage conditions.

Urine was collected over a 24 hour period from one mildly proteinuric and one severely proteinuric dog each with negative urine cultures and a unremarkable urine sediment and stored at 4°C in a homopolymer propylene collection jug designed for 24‐hour urine protein analysis in humans. Five aliquots of each urine sample were subjected to each of nine storage conditions: in three types of container (glass, propylene copolymer, and homopolymer propylene) at either 24°C for 4 hours, 4°C for 12 hours, or −20°C for 72 hours. An additional five aliquots were stored at −80°C in homopolymer as a control. Following experimental storage, samples were then stored at −80°C in homopolymer until analysis. To determine the minimum size of each of the study groups, a power analysis was performed a priori. With a power of 80%, an alpha of 5%, and a standard deviation of 0.3 for severely proteinuric animals and 0.1 for mildly proteinuric animals (based on pilot study), 4 samples would be needed to detect an approximate 10% decrease in UPC in a severely proteinuric sample (from 10 to 9) and an approximate 33% decrease in the UPC in a mildly proteinuric animal (from 0.6 to 0.4). The latter was chosen as it would alter recommendations based on the 2005 ACVIM guidelines. Shapiro‐Wilk test indicated that values were normally distributed. For each urine sample, ratios were compared using a one‐way ANOVA with storage condition as the only factor in the model, and ratios were compared among treatments post‐hoc with a Bonferroni correction (experimentwise type 1 error < 0.05).

Ratios differed slightly but significantly (P < 0.03) among aliquots stored under different conditions for both urine samples. For mildly proteinuric urine, UPC values were slightly higher (mean 0.68 vs 0.66) in aliquots stored in copolymer containers at 24°C than at −20°C, and were slightly lower (0.66 vs 0.68) in aliquots stored in glass containers at 24°C than at −4°C, but were not different from controls (0.67 + /‐ 0.01). For severely proteinuric urine, UPCs were higher in aliquots stored in copolymer at 24°C (9.6 + /‐ 0.9) and glass at −20°C (9.5 + /‐ 0.5) than in controls (8.4 + /‐ 0.2). Nevertheless, most storage conditions did not affect ratios and the detected differences would not have led to different clinical recommendations based on the ACVIM consensus statement or IRIS guidelines.

The results suggest that collecting and storing urine in clean commonly used household containers such as propylene copolymer food storage containers and glass jars is unlikely to lead to changes of sufficient magnitude to be of clinical importance. However, further study would be needed to determine if there might be specific circumstances or variables (i.e. patients with dysproteinemia, different amounts of surface area contact) in which container type might still cause clinically significant changes in measured UPC.

Evaluation of Gastric pH and Serum Gastrin Concentrations in Cats with Chronic Kidney Disease

M. Katherine Tolbert1, S. Olin1, S. MacLane2, J.M. Steiner3, S. Vaden4, E.N. Gould5, J. Price5

1College of Veterinary Medicine, University of Tennessee, Knoxville, TN, USA, 2Appalachian Animal Hospital, Piney Flats, TN, USA, 3Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College Station, TX, USA, 4North Carolina State University, Raleigh, NC, USA, 5University of Tennessee, Knoxville, TN, USA

Chronic kidney disease (CKD) is a prevalent condition in cats. Advanced CKD is associated with hyporexia and vomiting. These signs are typically attributed to uremic toxins and gastric hyperacidity and underlie IRIS recommendations for acid suppressant use with advanced CKD in cats. However, gastric pH has not been measured in cats with CKD. Accordingly, the central objective of this study was to determine if cats with moderate to severe CKD have decreased gastric pH compared to age‐matched, healthy cats. Based on previously published work demonstrating an association of hypergastrinemia and CKD, we hypothesized that cats with CKD would have a reduced gastric pH compared to healthy, aged‐matched cats.

All cats with concurrent disease were excluded on the basis of history, physical examination, CBC, serum chemistry profile, urinalysis and urine culture, serum total thyroxine, and SDMA (controls only) obtained within 24 hours of pH monitoring and assessment of serum gastrin concentrations. Ten cats with IRIS stage II‐IV CKD (mean±SD age: 14.1 ± 5.7 yrs) and nine healthy mature adult cats (mean age: 12.3 ± 1.7 yrs) were prospectively enrolled. Continuous, 12 hr gastric pH monitoring was performed by oral administration of a pH capsule. Serum gastrin concentrations, mean gastric pH, and mean percentage of time gastric pH was strongly acidic (pH < 2) were compared between groups using an independent 2‐sample t‐test.

pH parameters including mean gastric pH (CKD: 1.8 ± 0.5; healthy: 1.6 ± 0.3; P‐value=0.23) were not significantly different between cats with CKD and healthy cats. Mean serum gastrin concentrations (CKD: 110 ± 210; healthy: 57 ± 33 ng/dL) were higher and highly variable in cats with CKD, as previously described, but were not significantly different between groups (P‐value=0.72). pH data suggest that cats with CKD do not have gastric hyperacidity compared to aged‐matched healthy cats. Thus, further study investigating the benefit of acid suppression in cats with CKD is warranted.

Effect of High Sodium Diet on Urine Characteristics in Healthy Adult Dogs

H. Xu1, X. Si1, S. Bhatnagar1, D. Laflamme2

1Nestlé Purina Petcare Global Resources, St Louis, MO, USA, 2Purina Institute, St. Louis, MO, USA

Increasing urine volume and decreasing urine specific gravity (USG) are considered important in managing dogs with crystal‐associated lower urinary tract diseases. The objective of this study was to evaluate the impact of increased dietary salt (sodium chloride) on water intake, urine volume, USG, and mineral excretion in dogs.

Two diets, identical except for salt content, were fed to 12 healthy adult dogs using a 2 × 2 cross‐over design. Control Diet (CON) contained 0.11 g sodium/100 kcal ME, and High Sodium Diet (HNA) contained 0.37 g sodium/100 kcal ME. All dogs were fed CON for 2 weeks as the baseline, then randomly allocated to receive CON or HNA for 2 weeks. Following a 2‐week washout when all dogs were fed CON, a final 2 week feeding period was used with dogs fed the alternate diet from phase 1. Water consumption was measured daily, total voided urine was collected on days 10, 12, and 14 of each phase for determination of volume, USG and mineral excretion. Generalized linear mixed models detected differences among diets with P < 0.05 considered statistically significant.

The increase in water consumption did not reach statistical significance. However, urine volume (mL/kg/day) increased and USG decreased (both P < 0.05) in dogs fed HNA (28.60 ± 13.28 and 1.027 ± 0.008, respectively) vs. CON‐fed dogs (15.66 ± 14.81 and 1.035 ± 0.005, respectively). Total daily excretion of calculogenic components (calcium, magnesium, oxalate, and phosphate) were increased in dogs fed HNA (P < 0.05) but urine concentration of these minerals were not different from baseline or CON (all P > 0.10). Total excretion of citrate was increased in HNA dogs (P < 0.05 vs baseline) but urine concentration differences did not reach significance (P = 0.079) vs baseline, with no difference vs CON. Among urine concentrations, only Na and Cl were increased and K decreased in HNA vs baseline and CON, all P < 0.05.

In conclusion, increased sodium intake resulted in greater urine volume and lower USG without changing urine concentrations of calculogenic components. Moderately increased dietary sodium could be a viable way to increase urination in dogs with lower urinary tract conditions.

SDMA Is a More Reliable Marker for Kidney Function in Cats and Dogs with Cancer

M.V. Yerramilli1, M. Yerramilli1, G. Farace1, J. Robertson1, R. Relford1, D. Jewell2, J. Hall3

1IDEXX, Westbrook, ME, USA, 2Hill's Pet Nutrition, Topeka, KS, USA, 3College of Veterinary Medicine, Oregon State University, Corvallis, OR, USA

While the prevalence of functional impairment and injury in the kidney is high in cancer patients, it is significantly under diagnosed using creatinine. In a study of lymphoma patients, 4 out of 10 had comprised kidney function but, only 1 out of 4 of this cohort was identified using creatinine. The remainder needed kidney biopsies for diagnosis. The most common causes of compromised kidneys include infiltration, nephrotoxic chemotherapeutics, volume depletion, tumor lysis syndrome, obstruction and sepsis. The ineffectiveness of creatinine is likely due to reduced production as a result of cachexia, reduced protein intake and therapeutics to mention few.

In a French epidemiological study with 5000 human cancer patients with different types of solid tumors it was reported that 52% had a reduced GFR with 12% being stages 3–5. While such detailed information is nonexistent for veterinary patients at this time, there is no reason to think that the situation is any different. Aside from the increasing rate of cancer due to longevity; many of the chemotherapeutics are nephrotoxic so there is a significant need for effective and sensitive biomarkers to diagnose kidney disease and effectively manage interventions.

To understand the effectiveness of SDMA over creatinine to diagnose kidney impairment in cancer patients, we performed a retrospective study in cats and dogs using banked samples and correlated results to histopathology. The study included 19 patients (10 dogs and 9 cats) with a variety of malignancies. SDMA ranged from 8–26 μg/dL with 7 patients being above the upper reference limit. Necropsies and histopathology were performed on all patients.

All the patients with increased SDMA had histopathological findings indicating kidney disease or neoplastic infiltration. No findings were observed in patients with normal SDMA. In all patients, creatinine was below the upper reference limit and ranged from 0.4 to 1.3 mg/dL.

In an additional retrospective evaluation using 50 cats and dogs with neoplasia, 15% had increased SDMA while only 2% had elevated creatinine.

In conclusion, the increase in SDMA correlated with histopathologic findings in patients with neoplasia. Thus, SDMA is a more reliable estimate of kidney function in cats and dogs with cancer.

Increased Water Intake and Hydration in the Domestic Cat Ingesting a Nutrient‐Enriched Water

B. Zanghi, C. Gardner, A. Reynolds

Nestle Purina PetCare, St. Louis, MO, USA

his study investigated water intake and hydration in healthy cats and evaluated a nutrient‐enriched water on hydration. Domestic shorthair cats (N = 18; mean age 9.8 yrs±SD 2.5; BCS 5–7 on 9‐point scale) received either tap‐water (W; N = 9) or test‐water containing whey protein and glycerol (TW; N = 9) for 56 days. Prior to treatment, a 7‐d baseline established daily W and food consumption using a 2‐bowl monitoring system with ad libitum water and dry food. Blood and urine samples were collected on days −1, 8, 15, 30, and 56 for urine specific gravity (USG), creatinine, and osmolality measurements. Endogenous creatinine‐based glomerular filtration rate (GFR) was calculated from sample collection on days 28–30 or 31–33. Baseline W and calorie intake were similar between groups. Although no difference was observed between groups during the treatment phase, TW group had a daily liquid increase of 52%, with a median intake across all days that ranged from 5% to 254% increase. Variance was controlled post hoc with cats characterized as responders (R:consumption >25%) and non‐responders (NR) (W: 2R/7NR versus TW: 7R/2NR; P = 0.02). Daily urine volume (mL/kg BW) was higher (P = 0.02) in TW (15.2 ± SE 1.8) versus the W treated cats (10.3 ± SE 0.7), but no difference (P = 0.60) between groups for GFR (1.8 ± SE 0.1 versus 1.9 ± SE 0.2, respectively). Baseline USG was similar between groups (1.052 g/mL), but declined (ANOVA P < 0.01) with TW (1.037 g/mL ±SE 0.005) versus W (1.053 g/mL ±SE 0.002). Serum osmolality was similar for both groups over the entire trial. This study suggests that specific nutrients added to a cat's water can improve liquid intake to significantly improve overall hydration, as determined by greater urine output and dilution, which may offer a health benefit to some cats in need of greater water consumption.

Feline Serum Cytokines and Renal Alpha‐Enolase Immunohistochemical Staining After Repeated Administration of a FVRCP Vaccine

S. Summers1, S.M. McLeland2, J.J. Quimby3, S. Dow3, J. Hawley3, J. Coy3, M.R. Lappin3

1College of Veterinary Medicine, Colorado State University, Fort Collins, CO, USA, 2College of Veterinary Medicine, Iowa State University, Ames, IA, USA, 3Colorado State University, Fort Collins, CO, USA

Progressive interstitial nephritis (IN) is the primary cause of feline chronic kidney disease (CKD) which affects as many as 50% of elderly cats and prevalence increases with age. Vaccination has been shown to be a risk factor for development of feline CKD. The Crandell Rees feline kidney (CRFK) cell line is commonly used to grow viruses for use in feline vaccine production. Repeated administration of a vaccine containing CRFK remnants as a model for interstitial nephritis was shown to induce strong antibody responses against CRFK lysates and against alpha‐enolase, an immunodominant glycolytic pathway enzyme found in all mammalian cells that has been studied as a marker of inflammatory diseases in people. The primary objectives of this study were to further evaluate this potential model for IN by assessing samples from the cats for cell‐mediated immune responses to alpha‐enolase, by evaluating serum cytokine concentrations using a commercially available kit, and by describing changes in renal alpha‐enolase staining patterns.

In the IN model, serum, urine, peripheral blood mononuclear cells, and a renal biopsy were collected from six, one year old, healthy purpose bred, unvaccinated kittens (three male; three female). After sample collection on Week 0 and again on Weeks 2, 4, 6, 8, 10, 12, and 14, all cats were administered a commercially available parenteral FVRCP vaccine shown previously to induce anti‐CRFK and anti‐enolase antibodies. A second renal biopsy was collected on Week 16, avoiding the previous biopsy sites. Blood and urine samples were collected on all sample dates. A single lymphocyte proliferation assay was performed on 4 of 6 hyperinoculated cats and 2 unvaccinated healthy control cats using 4 antigens (kidney, spleen, alpha‐enolase, and CRFK). Sera were assayed using a commercially available kit that detects sFas, Flt‐3 ligand, GM‐CSF, IFN‐gamma, IL‐1beta, IL‐2, IL‐4, SDF‐1, KC, IL‐6, MCP‐1, IL‐13, IL‐8, IL‐12 (p40), PDGF‐BB, RANTES, SCF, IL‐18, and TNF‐alpha (Milliplex MAP Cytokine/Chemokine Feline Panel, Millipore, Darmstadt, Germany). Alpha‐enolase immunohistochemical staining was performed on kidney biopsy samples (week 0 and week 16) and then evaluated and scored (0–4) by a board‐certified pathologist that was masked to the timing of the biopsies.

Lymphocytes from hyperinoculated cats proliferated significantly more in the presence of alpha‐enolase compared to the other three antigens confirming the induction of a cell‐mediated immune response towards alpha‐enolase contained in the vaccine. Glomerular and tubular alpha‐enolase expression was significantly (P < 0.05) greater at week 16 (mean score 3.37 and 1.95, respectively) compared to week 0 (0.37 and 1.01, respectively). Increased cytokine concentrations over the hyperinoculation period were noted for GM‐CSF (3 cats), IFN‐gamma (5 cats), IL‐1beta (3 cats), IL‐4 (4 cats), SDF‐1 (5 cats), IL‐6 (4 cats), MCP‐1 (3 cats), IL‐13 (4 cats), IL‐8 (5 cats), PDGF‐BB (3 cats), RANTES (6 cats), IL‐18 (3 cats), and TNF‐alpha (3 cats).

Hyperinoculation with a market leading FVRCP parenteral vaccine induced both antibody and cell‐mediated immune response towards alpha‐enolase, increased alpha‐enolase expression in both renal tubules and glomeruli, and led to elevations in select inflammatory cytokines and chemokines within this 16‐week model. Although there has been no direct link made between parenteral FVRCP vaccines and feline interstitial nephritis to date, results of urinary cytokines and further assessment of alpha‐enolase expression within feline kidneys will be used to continue evaluation of this potential IN model.

Tamm‐Horsfall Protein in Dogs with Natural Occurring Chronic Kidney Disease: A Marker of Progression?

F. Chacar1, M. Kogika1, T. Sanches2, D. Caragelasco1, C. Martorelli1, C. Rodrigues2, J. Capcha2, L. Andrade2

1University of Sao Paulo/School of Veterinary Medicine and Animal Science/Internal Medicine, Sao Paulo, Brazil, 2University of Sao Paulo/School of Medicine, Sao Paulo, Sao Paulo, Brazil

Tamm‐Horsfall Protein (THP) is exclusively synthesized by the distal nephron under physiological conditions. Therefore, it is expected larger amounts of this protein in the urine of healthy individuals. In chronic kidney disease (CKD) dogs, a few investigations have reported about the quantities of THP in urine. Lower urinary excretion of THP in CKD dogs might be associated with distal tubular injuries and/or to the loss of renal mass, and then, THP may act as a marker of disease progression. The aim of this study was to assess the role of THP as a marker of CKD progression in dogs. Nine clinically healthy dogs were recruited as control group. Forty CKD dogs were enrolled (IRIS staging, Stage 1, n = 10; Stage 2, n = 10; Stage 3, n = 10; Stage 4, n = 10). Western blotting of THP was performed, and the results of densitometry were expressed in mean ± SEM. Urinary immunodetection of THP was 100.0 ± 7.7% in control dogs, and 117.2 ± 10.8%, 86.0 ± 9.0%, 65.4 ± 8.7% and 46.7 ± 9.0% in CKD dogs in stages 1, 2, 3 and 4, respectively. Differences among all groups were observed, particularly between Stages 1 and 4 (P ≤ 0.0001). It is likely that progressive reduction in urinary excretion of THP may contribute to detect gradual loss of nephrons, and it may act as a marker of disease progression in CKD dogs.

Light and Electron Microscopic Analysis of Kidneys from Dogs Infected with Angiostrongylus Vasorum

B. Gerber1, L. Aresu2, M. Schnyder3

1Clinic for Small Animal Internal Medicine, Vetsuisse Faculty, University of Zurich, Zurich, Switzerland, 2Department of Comparative Biomedicine and Food Science, Università di Padova, Legnaro (PD), Veneto, Italy, 3Institute of Parasitology, Vetsuisse Faculty, University of Zurich, Zurich, Switzerland

Infection with Angiostrongylus vasorum (Av) is an emerging disease in dogs with a wide range of clinical signs. To date, it is not clear whether Av is able to cause glomerular immune complex deposition or other renal damage.

The aim of the study was to evaluate the kidneys of dogs experimentally infected with Av.

Sixteen beagle dogs 1 to 1.5 years of age were orally infected with about 266 third stage larvae each. Eight dogs were treated with an anthelmintic against Av after 30 days and the other eight were left untreated. All dogs were euthanized 56 to 59 days after infection. Renal samples were obtained and fixed in neutral‐buffered formalin and 2.5% glutaraldehyde for light‐ (LM) and electron microscopy (EM) analysis.

Of the untreated dogs, 3 showed no significant changes at LM, 2 had occasional fetal glomeruli (< 5%). In 3 dogs multifocal granulomas were observed in the cortex.

Of the treated dogs 4 showed no significant changes at LM, 3 had occasional fetal glomeruli (< 5%), and 1 had focal and segmental mesangial sclerosis and mild increase of mesangial cells associated with occasional mesangial interposition.

Overall, EM confirmed the histological diagnosis and no glomerular immune deposits were found.

We demonstrated that kidneys of dogs infected with Av for 56 to 59 days have only mild LM changes and no immune mediated disease.

Effect of Weight Gain in Dogs on Renal Function and Perfusion Assessed by Contrasted‐Enhanced Ultrasound

D. Liu1, M. Hesta2, E. Stock1, S. Croubels3, M. Devreese3, B. Broeckx2, E. Bogaerts1, S. Daminet4, K. Vanderperren1

1Department of Veterinary Medical Imaging and Small Animal Orthopedics, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Oost‐Vlaanderen, Belgium2Department of Nutrition, Genetics and Ethology, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Oost‐Vlaanderen, Belgium3Department of Pharmacology, Toxicology and Biochemistry, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Oost‐Vlaanderen, Belgium4Department of Medicine and Clinical Biology of Small Animals, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Oost‐Vlaanderen, Belgium

Studies in obese humans and animal models of obesity revealed that complex interactions between metabolic, inflammatory and hemodynamic factors result in altered renal perfusion, vascular dysfunction and histological changes in the kidneys. Early detection of renal dysfunction is crucial in reducing morbidity and mortality. Contrast‐enhanced ultrasound (CEUS) has been suggested to be a sensitive method to study renal microvascular perfusion. Therefore, the objective of this study was to assess the effect of weight gain in dogs on kidney function and perfusion using CEUS.

Sixteen healthy lean adult beagles (2–8 years) with ideal body weight (BW) (average 11.39 ± 1.76 kg) and ideal body condition scores (BCS) (range 4–5/9, median: 4/9) were equally divided into a control group (group C) and a treatment group (group T), based on age and sex. After four weeks adaptation to a commercial adult maintenance diet, group T received 1.3 × maintenance energy requirements to induce weight gain for 47 weeks, whereas group C was fed to maintain an ideal BW and BCS for 47 weeks. All dogs were weighed and BCS was assessed on a weekly basis. At the end of the adaptation (week 0) and at week 47, systolic blood pressure; serum urea, creatinine, and albumin; urinary specific gravity, urinary protein:creatinine ratio (UPC) were measured. Glomerular filtration rate (GFR) was measured by plasma clearance of exo‐iohexol after an intravenous bolus injection of iohexol (Omnipaque 300®, 64.7 mg/kg BW). While sedated (butorphanol, 0.04 ml/kg), CEUS of the left kidney was performed using a bolus injection with Sonovue®. Time intensity curves reflecting tissue perfusion were generated based on manually drawn region of interests on the renal cortex and medulla. Perfusion parameters (rise/fall time, area under curve (AUC), time to peak, wash‐in/out rate (WiR/WoR), wash‐in/out AUC, wash‐in perfusion index (WiPI), mean transit time, and peak enhancement (PE)) were calculated and evaluated. Parametric and non‐parametric tests were used to compare groups and paired data with significance set at P < 0.05.

At week 0, no significant differences between group C and group T were found. At week 47, group T gained on average 37% more weight; and consequently had a higher BCS (P < 0.001) than group C. Despite the weight gain, no significant effect on blood pressure and GFR was observed in group T at week 47. However, a significantly higher median UPC (P = 0.031) was observed at week 47 compared to week 0 in group T. CEUS perfusion parameters of the cortex were not significantly different between the two groups and between week 0 and 47. In the medulla, however, PE, a measure related to blood volume, was significantly higher in group T than in group C at week 47 (P = 0.021). Furthermore, WiR, WiPI and WoR of the medulla, parameters related to blood velocity, were also significantly higher or close to significant when compared to group C at week 47 (P = 0.038, 0.021 and 0.065, respectively) or to week 0 (WiR: P = 0.039; WoR: P = 0.039).

A weight gain of 37% in dogs does not appear to affect most routine markers of kidney function or GFR. Increased UPC does suggest that weight gain results in higher urinary excretion of protein as seen in humans. Additionally, CEUS was able to detect differences in blood volume and velocity of the renal medulla, which may indicate that structural and functional changes may have occurred as a result of weight gain. Further studies utilizing non‐invasive markers or techniques to localize abnormalities in kidney function or injury in obese dog are warranted. Additional studies are also required to evaluate whether these changes are reversible after weight loss.

Evaluation of FGF‐23 and Urinary Fractional Excretion of Phosphorus Ratio in Chronic Kidney Disease Dogs

C. Martorelli1, M. Kogika1, F. Chacar1, D. Caragelasco1, D. Chew2, R. Toribio3, K. Dembek4, C. Rodrigues5, T. Sanches5, L. Andrade5

1University of Sao Paulo/School of Veterinary Medicine and Animal Science/Internal Medicine, Sao Sao Paulo, Sao Paulo, Brazil, 2The Ohio State University, Columbus, OH, USA, 3Veterinary Clinical Sciences, The Ohio State University, Columbus, OH, USA, 4The Ohio State University, Ames, IA, USA, 5University of Sao Paulo/School of Medicine, Sao Paulo, Sao Paulo, Brazil

Fibroblast growth factor 23 (FGF‐23) is a bone‐derived hormone that regulates phosphorus (P) and vitamin D metabolism. It was recently shown that dogs with chronic kidney disease (CKD) have increased FGF‐23 concentrations. FGF‐23 increases urinary fractional excretion of phosphorus (uFEP) in order to limit increases in serum phosphorus. The phosphaturic activity of FGF‐23 is based on suppression of renal P reabsorption in the proximal tubules. Recently human studies showed that uFEP/FGF‐23 ratio may reflect functional nephrons and the phosphaturic actions of FGF‐23, and can be used as prognostic marker of CKD. A low uFEP/FGF‐23 ratio may indicate a reduction in the number of functional nephrons and/or impaired FGF‐23 signaling from Klotho (co‐receptor for FGF‐23) expression. The goal of this study was to evaluate whether uFEP/FGF‐23 ratio reflect severity of CKD and could predict prognosis of CKD dogs. Sixteen CKD dogs were followed‐up for 13 months or until death (IRIS stage 2, n = 5; stage 3, n = 11). Urinary and serum phosphorus as well as serum FGF‐23 were measured at the first visit. The uFEP/FGF‐23 ratio was determined in fifteen healthy dogs (mean ± SEM; 0.083 ± 0.024). Normophosphatemia (3.54 ± 0.07 mg/dL; range 2.3 to 5.4 mg/dL) was detected in all CKD dogs in stage 2 most of the times during the follow‐up, and hyperphosphatemia detected in 8 out of 11 dogs in stage 3 (5.68 ± 0.23 mg/dL; range 2.5 to 16.0 mg/dL). The uFEP/FGF‐23 ratio was 0.027 ± 0.014 in stage 2 CKD, and 0.006 ± 0.003 in stage 3 CKD (P = 0.0122). Serum FGF‐23 concentrations were higher in stage 3 and even though FGF‐23 was increased, there was no increment in phosphaturia. CKD dogs in stage 3 with lower uFEP/FGF‐23 ratios had a poorer prognosis. Decreased survival was predicted when uFEP/FGF‐23 ratios were less than 0.006 (survival up to 300 days). The worst prognosis occurred when uFEP/FGF‐23 ratio was < 0.003 (less than 130 days survival). In conclusion, a low uFEP/FGF‐23 ratio associated with normal to high serum FGF‐23 concentrations supports a reduction in functional nephrons and/or impaired FGF‐23 signaling from renal injury since increased FGF‐23 should promote sufficient phosphaturia. Evaluation of Klotho (serum, urine, renal tissue) will be valuable to this question. Studies with larger population of dogs in various IRIS CKD stages will confirm the prognostic value of uFEP/FGF‐23 in dogs with spontaneous chronic kidney disease.

Alpha‐Enolase Staining Patterns in the Renal Tissues of Cats With and Without Chronic Kidney Disease

S.M. McLeland1, J.J. Quimby2, M.R. Lappin2

1College of Veterinary Medicine, Iowa State University, Ames, IA, USA, 2Colorado State University, Fort Collins, CO, USA

Alpha‐enolase, a ubiquitous, glycolytic enzyme known for catalyzing the conversion of 3‐phosphoglycerate to phosphoenolpyruvate, has been implicated as a self‐antigen in various autoimmune diseases. Autoantibodies in serum are capable of eliciting renal injury and ultimately can contribute to patient morbidity and mortality. Furthermore, endogenous renal alpha‐enolase has been shown to be overexpressed in human patients with systemic lupus nephritis and mixed cryoglobulinemia. Chronic kidney disease (CKD) is common in companion animals and affects a significant proportion of geriatric cats. While antibodies against alpha‐enolase have been shown to be present in sera of cats, patterns of alpha‐enolase protein expression in the kidney has not determined in healthy or diseased cats. The purpose of this study was to identify and characterize alpha‐enolase expression in renal tissues of cats with and without CKD.

None of the cats from which tissues were collected were euthanized because of the study. Renal tissues were collected from 29 cats with IRIS Stage II – IV CKD (mean age 14.7 years; range 8 – 20.5 years of age) and 12 cats with a creatinine concentration < 1.6 mg/dL, and a urine specific gravity of > 1.035. The cats with normal creatinine concentrations were divided into those < 1.5 years of age (n = 8 cats) and > 10 years of age (n = 4 cats). Interstitial nephritis was the predominant histopathological finding in the CKD group and was minimal to absent in the 12 cats with normal creatinine concentrations. An immunohistochemistry technique for semi‐quantitative measurement of alpha‐enolase protein expression in feline tissue was optimized utilizing a commercially available monoclonal alpha‐enolase antibody. Each section of kidney was evaluated for positive immunoreactivity and assigned a score based on signal intensity—immunoreactivity–ranging from 0 to 3. Scoring was defined as follows: 0 = no staining, 1 = light brown/tan, 2 = moderate brown staining, 3 = dark brown which may obscure the nucleus. Mean scores between young, geriatric and CKD were compared using Kruskal‐Wallis with Dunn's post hoc analysis.

Non‐azotemic cats < 1.5 years of age had moderate alpha‐enolase expression in tubular epithelium (median score = 2; range 2 – 3) but staining was absent to mild in glomeruli (median score = 1; range 0 – 2). In contrast, alpha‐enolase expression was moderate to high in tubular epithelium (median score = 2; range 1.95 to 2.05) and glomeruli (median score = 2; range 1.8 to 2.6) in non‐azotemic cats > 10 years of age. Lastly, evaluation of kidneys of cats with CKD showed that alpha‐enolase expression was decreased in tubules that were degenerative or atrophic (median score = 1; range 0 – 3), similar to the other groups of cats in normal tubules (median score = 2; range 1–3), and moderate to highly expressed in glomeruli (median score = 2; range 0–3). While no statistical difference was found between tubular immunoreactivity scores, there was a significant difference between glomerular immunoreactivity scores between non‐azotemic cats > 10 years of age and non‐azotemic cats < 1.5 years of age (P = 0.04) and CKD cats and non‐azotemic cats < 1.5 years of age (P = 0.009).

Alpha‐enolase expression decreases in damaged tubules and increases in the glomeruli of older cats prior to the development of detectable CKD when compared to young, non‐azotemic cats. Further studies will be required to determine whether these findings relate to the pathogenesis of CKD or could be used as an early diagnostic test for CKD.

Metabolomic Evaluation of Feline Urinary Diseases

N. Priymenko1, B. Autuoro2, C. Torre3, I. Jouanin1, M. Tremblay‐Franco1, I. Jeusette4

1Toxalim, Universite de Toulouse, INRA, ENVT, Toulouse, Midi‐Pyrenees, France, 2Ecole Nationale Veterinaire de Toulouse, Toulouse, Midi‐Pyrenees, France, 3Affinity Petcare, Barcelona, Catalonia, Spain, 4Affinity Pet Care SA, L'Hospitalet de Llobregat, Catalonia, Spain

Metabolomics may be used for providing description of endogenous metabolites present in biological samples in response to pathophysiological or external stimuli. The purpose of the present study was to use metabolomics to compare cats with feline idiopathic cystitis (FIC) or struvite plug (SP) with healthy cats.

Male cats presenting either SP (n = 6) or FIC (n = 19) and healthy (n = 10) were recruited through veterinary practices. Urine samples were collected by cystocentesis. NMR spectra were acquired on Bruker DRX‐600 Avance spectrometer operating at 600.13 MHz 1H resonance frequency and equipped with a cryoprobe. After data reduction and normalisation, 751 buckets RMN were analysed with multivariate statistical analyses.

Inspection of the spectra revealed variation in urine composition between the three groups and enabled identification of metabolites responsible for some peaks. After OSC filtration, a valid and robust PLS‐DA model explaining the variability between the three groups was built. Urine from FIC cats contained higher concentrations of indoxylsulfate, hippurate, creatinine, felinine, betain and TMAO compared to SP or healthy cats. Urine from SP cats contained more guanidinoacetate and dimethylamine compared to FIC or healthy cats and more creatine and TMAO than healthy cats. Urine of healthy cats contained higher quantities of carnitine, but less of TMAO and creatinine.

In this study, the use of metabolomics was able to describe biochemical variations between urines of healthy cats and cats presenting FIC or SP. This novel application of metabolomics opens new hypothesis on pathophysiology and treatment in such cats.

Comparison of Single, Averaged, and Pooled Urine Protein:Creatinine Ratios in Proteinuric Dogs Undergoing Medical Therapy

S. Shropshire1, R. Cerda2, J.J. Quimby1

1Colorado State University, Fort Collins, CO, USA, 2Arapahoe Animal Hospital, Boulder, CO, USA

Monitoring the urine protein:creatinine (UPC) ratio in dogs with protein‐losing nephropathy (PLN) to determine response to therapy can be challenging due to day‐to‐day variation in their UPC values. The purpose of this study was to determine if a single sample, averaged sample, or pooled sample are comparable for determination of the UPC ratio, regardless of the degree of proteinuria, in PLN dogs receiving medical therapy.

Twenty‐five client‐owned dogs diagnosed with PLN currently receiving ACE inhibitors and/or ARBs for treatment of proteinuria were enrolled into the study; 16 dogs had UPC ratios less than 4 and 9 dogs had UPC ratios ≥ 4. UPC ratios were measured in each dog utilizing three separate methods; a single sample (day 3), an average sample (day 1–3), and a pooled sample (equal pooling of urine from day 1–3). Bland‐Altman analysis was performed to evaluate a single UPC versus a pooled UPC, a single UPC versus an average UPC, and a pooled UPC versus an average UPC. The correlation coefficient for each of the three analyses was also calculated followed by simple linear regression. Bland‐Altman analysis was also performed on subgroups (UPC ≤ 4 or UPC > 4) to compare the methods in differing degrees of proteinuria.

When evaluating all of the dogs, there was a strong correlation between a single UPC compared to a pooled UPC (r = 0.96, P < 0.0001), between a single UPC and an average UPC (r = 0.98, P < 0.0001), and between the pooled UPC and average UPC (r = 0.99, P < 0.0001). For each Bland Altman comparison 95% limits of agreement (95% of the time the difference between two methodologies will fall within this range) are described below:

There was good agreement between the three methods particularly for dogs with UPC ≤ 4. Poorer agreement was seen between single UPC and pooled or average UPC, with the single Day 3 UPC typically being higher than the pooled or average UPC.

The results of this study show that the UPC ratio from a single sample, averaged sample, and pooled sample are comparable in PLN dogs receiving medical therapy when UPC ≤ 4. In PLN dogs with UPC > 4, more variability between methods exists but a previous study in X‐linked nephropathy dogs not receiving medical therapy showed this degree of variability may not be clinically relevant. However, future studies are needed to verify if this is true in all dogs with PLN while receiving medical therapy.

Retrospective Study of the Efficacy of Oral Potassium Supplementation in Cats with Chronic Kidney Disease

L. Sieberg1, J. Quimby2

1Colorado State University, Fort Collins, CO, USA, 2The Ohio State University, Columbus, OH, USA

The purpose of this retrospective study was to gain insight into the effect of three oral potassium supplements: potassium gluconate tablets (PGT), potassium gluconate powder (PGP) and potassium citrate granules (PCG) on hypokalemia and serum bicarbonate.

Colorado State University Veterinary Teaching Hospital medical records (2006–2016) were retrospectively searched for cats that had been prescribed an oral potassium supplement for management of their CKD‐associated hypokalemia. Data collected included signalment, bodyweight, age, supplement type, dose, directions, concurrent medications, diet, indication owner was unable to give the supplement, and serum chemistry and urinalysis values. For inclusion, laboratory work had to be available at the time of hypokalemia diagnosis, and at recheck within one to six weeks. Treatment response was defined in three ways: any increase in potassium, an increase in potassium to within the normal reference range, and an increase to greater than 4 mEq/L. Not all data was normally distributed thus Wilcoxon Sign Rank test was used to compare pre and post‐supplementation values. Spearman Rank was used to assess correlation between dose and change in serum potassium and bicarbonate.

A total of 205 cats were identified in the search; 167 were ineligible and 38 met inclusion criteria. Of the 38 included cats, 16 were given PGT, 11 were given PGP, and 11 were given PCG. 5/38 had to change the dose of supplement, and 5/38 changed the supplement type. Dosing ranged from 0.21 – 1.6 mEq/kg (49 – 374.4 mg/kg) for PGT, 0.25 mEq/kg (59 – 347 mg/kg) for PGP, and 0.04 – 1.41 mEq/kg (4 – 144 mg/kg) for PCG. 36/38 cats had an increase in potassium after supplementation (14/16 for PGT, 11/11 for PGP, and 11/11 for PCG). 34/38 cats had an increase to within the potassium reference range (13/16 for PGT, 10/11 for PGP, and 11/11 for PCG). 24/38 cats had an increase in potassium to > 4 mEq/L (8/16 for PGT, 8/11 for PGP, and 8/11 for PCG). Pre‐supplementation potassium mean +/SD for PGT was 3.3 + /‐ 0.5 and 4.1 + /‐ 0.6 post‐supplementation. Pre‐supplementation potassium mean +/SD for PGP was 3.3 + /‐ 0.4 and 4.2 + /‐ 0.7 post‐supplementation. Pre‐supplementation potassium mean +/SD for PCG was 3.3 + /‐ 0.2 and 4.1 + /‐ 0.4 post‐supplementation. There was a statistically significant difference in serum potassium post supplementation for all three medications: PGT (P = 0.0001), PGP (P = 0.001), PCG (P = 0.001). There was a positive correlation between PGT dose and change in potassium levels (P = 0.04), but no significant correlation for PGP or PCG. There was no significant difference between pre and post supplementation serum bicarbonate nor a correlation between potassium supplement dose and serum bicarbonate level. There was not enough data to assess effect on urine pH.

All three potassium supplements were effective in treating hypokalemia secondary to CKD in the majority of cats despite quite variable dosing, but treatment response depending on the definition of response used. No indication of alkalinizing effect was seen, and prospective studies are needed for more thorough assessment.

Serum Concentrations of Symmetric Dimethylarginine in Dogs with Dehydration

W. Choi1, J.‐H. Kang2, S. Woo1, P. Son1, J. Yun1, C.‐J. Sung1, H. Kim1, B.‐T. Kang1, M.‐P. Yang1

1College of Veterinary Medicine, Chungbuk National University, Cheongju, Ch'ungch'ong‐bukto, Korea, 2Chungbuk National University College of Veterinary Medicine/Veterinary Medical Center at Chungbuk National University, Cheongju, Ch'ungch'ong‐bukto, Korea

Symmetric dimethylarginine (SDMA) and neutrophil gelatinase‐associated lipocalin (NGAL) have been suggested as useful biomarkers for evaluating renal disease. The objective of this study was to assess the circulating SDMA and NGAL concentrations in dogs before and after correction of dehydration.

Sixty‐nine client‐owned dogs with dehydration and 21 healthy dogs were included in this case‐controlled observational study. The dogs with dehydration were divided into 2 groups, based on the presence of azotemia. In addition, the dogs with dehydration were divided into 3 groups, according to the severity of dehydration. All dehydrated dogs were rehydrated within 24 hours by standardized fluid therapy, and serum concentrations of SDMA and NGAL were measured before and after correction of dehydration.

There were significant differences between the concentrations of SDMA and NGAL measured before and measured after the correction of dehydration in dogs with azotemia, but not in dogs without azotemia. In addition, SDMA and NGAL concentrations showed significant decreases after the correction of dehydration in dogs with moderate or severe dehydration, but not in dogs with mild dehydration. Notably, in 16 of 69 dogs, serum SDMA concentrations before correction of dehydration were higher than the reference interval and were normalized to the reference interval after rehydration. The correlation between serum SDMA and creatinine concentrations was significant, regardless of the presence of azotemia; whereas the serum NGAL concentration was significantly correlated with creatinine concentration only in dogs with azotemia.

This study found that the circulating SDMA concentration can be affected by hydration status, indicating that measurement of SDMA concentration in dehydrated dogs might need to be repeated after correction of dehydration. However, further studies are needed to clarify our results.

Effect of Antioxidant Supplementation on Markers of Oxidative Stress and Clinical Outcome in Hospitalized Dogs

D. Hagen1, J.L. Ekena2, B. Geesaman3, K. Viviano1

1University of Wisconsin, Madison, WI, USA, 2Department of Medical Sciences, School of Veterinary Medicine, University of Wisconsin‐Madison, Madison, WI, USA, 3Carolina Veterinary Specialists, Winston Salem, NC, USA

Oxidative stress is central to the pathogenesis of many systemic diseases. In humans, antioxidant depletion and lipid peroxidation correlate with disease severity and outcome. We have previously shown ill dogs experience oxidative stress, including decreased glutathione and vitamin E, and increased urinary 8‐isoprostane/creatinine (8‐IsoP/Cr). Short‐term N‐acetylcysteine supplementation did not impact clinical score or outcome but prevented further glutathione depletion. The purpose of this study was to evaluate longer‐term, combination antioxidant supplementation on markers of oxidative stress, clinical score, and survival in ill dogs.

Hospitalized client‐owned dogs were eligible for inclusion. Dogs were randomized to no supplementation (NS) or oral Denamarin and vitamin E supplementation (AS) for 30 days. Antioxidants (glutathione, cysteine, vitamin E, and selenium), 8‐IsoP/Cr, and clinical SPI2 scores were determined at enrollment and 30 days. Glutathione, cysteine, and vitamin E were quantified by high‐performance liquid chromatography. Selenium and 8‐IsoP/Cr were measured using atomic absorption spectroscopy and gas chromatography/mass spectrometry, respectively.

Thirty‐two ill dogs (target enrollment 40) have completed the study to date (AS, n = 17; NS, n = 15). Ill dogs had significantly elevated 8‐IsoP/Cr versus healthy controls (3660 vs. 2980 pg/mgCr; P = 0.021). Antioxidant supplementation did not result in significant differences in median endogenous glutathione, cysteine, vitamin E, or selenium concentrations, 8‐IsoP/Cr, SPI2 score, or 30 day survival. Interestingly, the NS group had significantly decreased plasma cysteine concentrations at 30 days when compared to baseline (9.62 to 6.95 μM; P = 0.039). These preliminary findings suggest, in this heterogeneous population of ill dogs, combination antioxidant supplementation did not alter redox state or clinical outcome.

Voluntary Food Ingestion in Dogs Following Intestinal Surgery

C. Kennedy, J. Gladden, J. Babyak

Tufts University, North Grafton, MA, USA

Eating after intestinal surgery in dogs is variable. In humans, early enteral feeding has been shown to decrease morbidity and reduce the duration of post‐operative ileus (POI); POI increases the length of hospitalization. This study evaluated the duration of anorexia and time to discharge in dogs following simple enterotomy (SA), multiple enterotomy (linear foreign body) (LFB) and resection and anastomosis (R&A). Additionally, prokinetic use was evaluated.

The medical records were searched: A convenience sample of fifty cases was identified. Data recorded included: signalment, the first time voluntary consumption of food occurred, use of a prokinetic and the time of discharge. The time until eating (TUE) and time until discharge (TUD) following surgery were calculated from the time of extubation. Incomplete records, dogs with a history of previous intestinal surgery, or cases in which dehiscence occurred, were excluded. A multivariable regression was performed to model the effect of variables after adjusting for population characteristics.

Nineteen dogs had SA, seventeen had LFB and fourteen had R&A. The mean age was 5.26 (+/− 4.37), 5.09 (+/− 4.53) and 6.79 (+/− 2.92) years respectively. All dogs received opioids: 96% received full mu‐agonists (n = 48). Forty‐two per cent (n = 21) received metoclopramide (six SA [32%], nine LFB [53%], six R&A [43%]): these dogs remained in hospital thirty hours longer on average compared to dogs that did not receive metoclopramide (P = 0.007). Reasons for metoclopramide use were not clearly documented. All dogs survived to discharge. The mean TUE was 11.26 (+/− 6.31), 13.88 (+/− 7.65), 11.64 (+/− 7.1) hours respectively. TUE significantly extended TUD independent of surgery type: each additional hour of not‐eating delayed discharge by 0.5 hours (P = 0.01). The mean TUD was 47.16 (+/− 22.06), 50.0 (+/− 16.14), 85.14 (+/− 53.02) hours respectively. R&A dogs were discharged on average thirty hours later than SA (P = 0.005). R&A dogs that received metoclopramide were discharged sixty hours later compared to SA dogs that did not receive metoclopramide.

Eating decreased the time of hospitalization following intestinal surgery in dogs. Early enteral feeding should be encouraged. Metoclopramide was associated with prolonged hospitalization. Its use suggested real or anticipated POI in this study. The use of prokinetics in POI is contested in human medicine: they do not consistently decrease hospitalization, though they may improve feeding tolerance. Resection and anastomosis is expected to lengthen hospitalization.

Measurement of Body Composition and Body Water Using Quantitative Magnetic Resonance in Preweaning Puppies

J.K. Spears, B. Zanghi

Nestle Purina PetCare, St. Louis, MO, USA

The period from birth through weaning represents a critical period in puppy development. Body composition during this preweaning period has not been well‐studied because sedation was previously required for analysis. Quantitative Magnetic Resonance (QMR) is a noninvasive technology that does not require anesthesia and has been validated as a method to assess body fat, lean, and total body water (TBW) in companion animals. The objective of this study was to determine if QMR can be used to measure body composition in puppies during the preweaning period. Healthy newborn puppies (Labrador retrievers and beagles; n = 16/breed) were selected for this study. Body weights were collected and QMR scans were performed on awake and unrestrained puppies within 24 hours of birth, and then at 2, 4, 6, and 8 weeks of age. Differences in weight, percent lean, body fat, TBW, and lean mass hydration were assessed. Body weight was higher for Labrador retrievers at all timepoints (P < 0.05). Body fat percentage was significantly higher in beagle puppies at birth, 2, 4, and 6 weeks of age (P < 0.05). Although lean mass percentage was similar between breeds at birth, Labrador retrievers had significantly higher proportion at Weeks 2 and 4 (P < 0.05). No breed differences were observed in TBW percentage at any time period, but beagles had significantly higher lean mass hydration at birth (P < 0.05). These results indicate QMR can be used to detect differences in body composition during the critical preweaning period. Further research evaluating differences in body composition among dog breeds during growth is required.

Hypocholesterolemia in Cats: A Retrospective Study of 48 Cases

C.A. Bowman, A.K. Viall, J.‐S. Palerme

Iowa State University, Ames, IA, USA

A retrospective study was performed to describe the underlying diseases, clinicopathologic abnormalities and prognosis associated with hypocholesterolemia in a population of cats presenting to a tertiary referral hospital between 2004 and 2016. Case inclusion criteria were: hypocholesterolemia, physical exam, CBC, serum biochemistry and urinalysis. When available, serum total T4, retroviral tests, imaging reports and necropsy reports were reviewed. For analysis, cats were categorized according to their disease by body system or etiology: neoplastic, infectious, sterile inflammatory, hepatobiliary, cardiac, gastrointestinal, malnutrition, toxin, neurologic and unknown.

Forty‐eight cats met inclusion criteria. Neutered males comprised 54% of the cats, intact males 4%, spayed females 38% and intact females 4%. Median age was 5 years (range 0.5–14 years). Domestic mixed breed cats comprised 75% of the study population, 6% were of unknown breed. Diagnosis categories were distributed as follows: 23% infectious, 23% gastrointestinal, 17% neoplastic, 11% hepatobiliary, 6% malnutrition, 6% inflammatory, 6% unknown, 4% cardiac, 2% neurologic and 2% toxin. Anemia was the most common CBC abnormality, present in 37.5% of cats. The most common biochemical abnormalities were hyperglycemia (45.8%), increased BUN (22.9%), hyperproteinemia (22.9%) and increased ALT (22.9%). The median cholesterol concentration was 67.5 mg/dL (range 50–74 mg/dL). Comorbidities were present in 37% of cats with renal and cardiac disease being the most prevalent at 27% and 16%, respectively.

No significant difference was found in the serum cholesterol concentrations between diagnosis categories. No significant difference was found between the serum cholesterol concentrations of survivors and nonsurvivors. Hypocholesterolemia did not appear to be associated with survival in this study.

Effect of a Commercially Available Probiotic on Immune Responses in Healthy Dogs

M.R. Lappin, J. Coy, J. Hawley, S. Dow

Colorado State University, Fort Collins, CO, USA

In a previous study, Enterococcus faecium strain SF‐68 (FortiFlora®, Purina) was fed to puppies from weaning to 1 year of age and differences in some immune responses were detected when compared to a control group. In that study, the first data collection point was after 10 weeks of supplementation and so it is unknown if any potentially beneficial effects may occur earlier. In addition, whether the probiotic induces changes in immune responses of adult dogs is unknown and since the first study was performed, there have been advances in techniques for evaluating the immune system of dogs. The objective of the current study was evaluate select immune responses of adult dogs before and after supplementation with E. faecium strain SF‐68.

A total of 7 age‐matched clinically healthy laboratory reared young adult beagles housed in a research facility were chosen for study. Blood in EDTA and serum were collected from each dog prior to supplementation and then monthly for 12 weeks. Flow cytometry was used to measure percentage of B cells expressing surface bound IgG or MHCII, the percentage of CD4 + T cells expressing MHCII, and the percentage of CD8 T cells expressing MHCII or CD11a. For each of these molecules, the amount of surface expression was also determined by calculating the geometric mean fluorescence index (GMFI). T lymphocyte proliferative responses of all dogs to a non‐specific mitogen (Concanavalin A) were assessed by flow cytometry and the ratio between proliferation of unstimulated cells and stimulated cells determined. Sera were assessed for a panel of cytokines using a commercially available kit (Milliplex MAP Canine Cytokine/Chemokine Magnetic Bead Panel; EMD Millipore). Differences between groups was assessed with significance defined as P < 0.05.

T lymphocyte proliferative responses to mitogen activation in the dogs on Week 12 (Range = 2.12 – 3.9; median = 2.25) was statistically greater (P = 0.03) than Week 0 (1.47 – 2.47; median = 1.92). Compared to Week 0, statistically significant increases in B cells expressing surface bound IgG or MHCII and increased B cell MHCII GMFI were first noted on Week 4 and were maintained through Week 12. Compared to Week 0, numerically increased percentages of CD4 + T cells expressing MHCII, percentages of CD8 T cells expressing MHCII, increased T cell CD4 + MHC Class2 GMFI, and increased T cell CD8 + MHCII GMFI were detected Weeks 4, 8 and 12, but the differences were not statistically significant. Compared to Week 0, T cell CD8 + CD11a+ and T cell CD8 + CD11a+ GMFI were significantly greater on Weeks 4 and 8 but the effect was not apparent on Week 12. Of the cytokines/chemokines, at least 50% of the sera assessed had concentrations within the range of the assay for KC‐like, IP‐10, IL‐7, IL‐8, IL‐18, and MCP‐1; statistically significant differences after supplementation with the probiotic were not noted.

Supplementation of adult beagles with Enterococcus faecium strain SF‐68 can induce immunomodulation as early as 4 weeks in dogs. Additional studies are ongoing to detail clinical benefits.

Behavior and Health in Aged Dogs Fed a Food with Antioxidants, Phytonutrients and Fatty Acids

H. Shiefelbein, J.M. MacLeay, I. Paetau‐Robinson, D. Jewell

Hill's Pet Nutrition, Inc., Topeka, KS, USA

Two out of 3 pet parents do not feed their older pet an age‐appropriate food. In 2005 AAHA published guidelines for care of older dogs which recommend focus areas for client education that includes wellness care and nutrition. Oxidative stress from free radicals and inflammatory processes are considered leading causes of degenerative diseases and cellular aging. Therefore, the purpose of this study was to evaluate the efficacy of a food designed to counter these changes in senior dogs compared to a non‐nutritionally controlled cohort.

This 6‐month study was reviewed and approved by an animal care and use committee and veterinarians and owners were masked as to the study sponsor. To be enrolled small and medium breed dogs must have been ≥ 9 years of age and large breed dogs must have been ≥ 7 years of age with no overt evidence of major organ disease. Dogs from 17 North American veterinary hospitals were screened (physical examination, CBC, serum chemistry, urinalysis) by attending veterinarians. Dogs were randomized to the test or control groups. The control group could consume any dry non‐therapeutic food, excluding raw foods, of the owner's choosing and were able to switch foods at their discretion. The test food contained controlled minerals, added antioxidants including vitamin E, ascorbic acid, beta carotene and lipoic acid, sources of n6 and n3 fatty acids including chicken fat, fish oil and flax seed, carnitine and phytonutrients supplied by the inclusion of carrots, citrus pulp, spinach, tomato pomace and plant extracts. History, physical examination forms, medication records, dietary information and owner questionnaires were filled out electronically at 0, 1, 2, 3 and 6 months. Blood and urine samples were obtained at screening (T0), month 3 (T3) and month 6 (T6). Blood work results but not questionnaire results have been previously reported (Hall et al. Plos One 2016). The test group had significantly decreased blood urea nitrogen (BUN) from T0 vs. T3 and T6 and these were significantly lower than the control group at T3 and T6. BUN was significantly lower in the test compared to the control group at T3 and T6.

Three hundred dogs were screened and 255 were enrolled in the study: 129 Test (59 m, 70f, mean age 9.54y, range 7–15y), 126 Control (58 m, 68f, mean age 9.86y, range 7–15y); 103 Test and 111 Control dogs completed the study. Statistical analysis was a repeated measures regression model with value of interest as the dependent variable and diet, time point and diet interacted with time point as independent variables. Differences in least square means were used to test for significant differences. P value of 0.05 or less was used to determine significance.

Previous foods consumed by enrolled dogs included 44 dry and 8 canned brands of food. The majority of foods were not labeled as being indicated specifically for senior pets. Pet owners reported that the dogs in the test group exhibited significantly more positive responses and greater improvements in vitality, seeking the owner's attention, coat shine, coat softness, decreased stiffness and improved ability to rise from rest compared to the control group and compared to baseline at Month 6.

This study design did not use a standardized control food and therefore may have captured bias toward and/or away from the test food due to loyalty to owner's chosen food(s) or the novelty of the test food. Improvements in BUN are consistent with controlled studies where pets were transitioned onto a masked senior or control food. This study does provide a real world scenario reflecting the pet parent's overall positive experience in transitioning from their current food to the test food. Dogs benefitted from the test food as shown by the decrease in BUN, positive changes in appearance and aging‐associated behaviors such as rising from rest, increased activity and interacting more with family members.

Behavior and Health in Aged Cats Fed a Food with Antioxidants, Phytonutrients and Fatty Acids

H. Shiefelbein, J.M. MacLeay, I. Paetau‐Robinson, D. Jewell

Hill's Pet Nutrition, Inc., Topeka, KS, USA

Two out of 3 pet parents do not feed their older pet an age‐appropriate food. In 2005 AAHA published guidelines for care of older cats which recommends focus areas for client education that includes wellness care and nutrition. Oxidative stress from free radicals and inflammatory processes are considered leading causes of degenerative diseases and cellular aging. Therefore, the purpose of this study was to evaluate the efficacy of a food designed to counter these changes in senior cats compared to a non‐nutritionally controlled cohort.

This 6‐month study was reviewed and approved by an animal care and use committee and veterinarians and owners were masked as to the study sponsor. To be enrolled cats must have been ≥ 9 years of age with no overt evidence of major organ disease. Cats from 11 North American veterinary hospitals were screened (physical examination, CBC, serum chemistry, urinalysis) by attending veterinarians. Cats were randomized by age to test or control groups. The control food could be any dry non‐therapeutic food, excluding raw foods, of the owner's choosing and they were able to switch foods at their discretion. The test food contained controlled minerals, added antioxidants including vitamin E, ascorbic acid, beta carotene and sources of n6 and n3 fatty acids including chicken fat and fish oil, carnitine and phytonutrients supplied by the inclusion of broccoli, tomato pomace and plant extracts. History, physical examination forms, medication records, dietary information and owner questionnaires were filled out electronically at 0, 1, 2, 3 and 6 months. Blood and urine samples were obtained at screening (T0), month 3 (T3) and month 6 (T6). Blood work results but not questionnaire results have been previously reported (Hall et al. PlosOne 2016). The test group had significantly decreased creatinine and blood urea nitrogen (BUN) from T0 vs. T3 and T6 and these were significantly lower than the control group at T3 and T6.

One hundred twenty‐nine cats were screened, 105 were enrolled in the study: 51 Test (19 m, 32f, mean age 11.61y, range 8–18y), 54 Control (20 m, 34f, mean age 11.56y, range 9–17y). 33 Test and 49 Control cats completed the study. Statistical analysis was a repeated measures regression model with value of interest as the dependent variable and diet, time point and diet interacted with time point as independent variables. Differences in least square means were used to test for significant differences. P value of 0.05 or less was used to determine significance.

Previous foods consumed included 20 dry and 12 canned brands of food. The majority were not labeled as being indicated specifically for senior pets. At 6 months the Test group pet owners reported significantly improved scores in vitality, coat texture, less matting, increased interactions with family members and improved ability to run and play than the control group and compared to baseline.

This study design did not use a standardized control food and therefore may have captured bias toward and/or away from the test food due to loyalty of the owner to their chosen food(s) or the novelty of the test food. Improvements in BUN and creatinine are consistent with controlled studies where pets were transitioned onto a masked senior or control food. This study does provide a real world scenario reflecting the pet parent's experience in transitioning from their current food to the test food. Cats benefitted from the senior food as shown by the improved BUN and creatinine concentrations, positive changes in appearance and aging associated behaviors such as increased ability to run and play, increased activity and greater interaction with the family.

Factors in Successful IV Catheterization of Patients by Veterinary Students in the Emergency Room

S. Babyak1, J. Babyak2, E.A. Rozanski3

1Tufts University/Becker College, North Grafton, MA, USA, 2Tufts University, North Grafton, MA, USA, 3Cummings School of Veterinary Medicine, Tufts University, Grafton, MA, USA

Successful placement of IV catheters is essential for treatment of ill and injured pets. Veterinary students often struggle to gain skills in IV insertion. The purpose of this prospective study was to evaluate factors in student success for surrounding placement of IV catheters.

Students were asked to complete a survey following each catheter placement attempt in the ER. The responses were anonymous. The IRB reviewed the study for student participation. Students were asked to self‐assess their level of experience (high, medium and low), and then characteristics of the patient, restrainer, catheter location and type were recorded as well as success or failure in catheter placement. Degree of “coaching” was also recorded. The chi‐square test was used to evaluate differences between insertion in dogs and cats. A univariate logistic regression analysis was used to assess the relationship between success and all other variables.

One hundred and thirty‐five catheters were attempted, 117 in dogs and 18 in cats. 93% of students reported low or moderate prior experience. 65% percent were successful. Veterinary technicians restrained for the majority of the catheter placements (54%), and provided some coaching in 60% of the attempts. There was no relationship between successful catheter placement and prior experience, species, indication for catheterization, brand of catheter, vessel, and restrainer. No student was successful when a faculty member was restraining the patient, although this was not significant due to the small sample size.

Catheterization is vital clinical skill; on‐going efforts in how to improve teaching during clinical instruction are still necessary.

Vasopressor Use in Critically Ill Cats 41 Cases (2007‐ 2015)

N. Licht1, E.A. Rozanski2, J.E. Rush1

1Tufts University, North Grafton, MA, USA, 2Cummings School of Veterinary Medicine, Tufts University, Grafton, MA, USA

Prolonged hypotension may have negative impact on survival in critically ill cats. Hypotension refractory to fluid therapy is an indication for continuous rate infusion (CRI) of vasoactive medications (VM). The purpose of this study was to review the indications and outcome of cats receiving VM.

The electronic medical record database was searched using the term “cat” and any of “dopamine”, “epinephrine”,”norepinephrine” and “vasopressin”. Records were reviewed; cats were excluded if they received VM during CPR or immediately following ROSC or if they received VM only during anesthesia or immediately post‐operatively. The underlying disease, initial BP, the VM(s) used, the rate, duration of VM use and outcome was recorded. Descriptive statistics were used and Fisher exact test to compare groups.

Forty‐one cats met the inclusion criteria for study enrollment. Initial BP was not detectable or < 50 mmHg (Doppler) in all cats. Sepsis (49%), neoplasia (29%) or DKA (10%) were common underlying conditions. The VM used most commonly was dopamine, with cats typically receiving between 6 and 20 mcg/kg/min. Norepinephrine was the second most common VM and cats typically received between 1.1 and 2 mcg/kg/min. Sixteen cats survived. 27 cats received one VM and 10 cats survived, while 14 of cats received 2 vasopressors, and 6 of these cats survived. (P = 0.22) Thirty cats received VM for < 24 hours with 13 surviving, while 11 received VM > 24 hours, with 3 surviving. (P = 0.48). No adverse affects were attributed to VM. Vasopressors may be used in cats with critical illness.

Evaluation of Middle Ear in Cats with Eustachian Tube Obstruction

N. Keskin1,2, P.A. Schachern2, I.G. Sancak2,3, H. Albasan1,2, M.M. Paparella2,4, S. Cureoglu2

1Department of Internal Medicine, Faculty of Veterinary Medicine, Diskapi, University of Ankara, Ankara, Turkey, 2Department of Otolaryngology‐Head and Neck Surgery, University of Minnesota, Minneapolis, MN, USA, 3Department of Surgery, Faculty of Veterinary Medicine, Diskapi, University of Ankara, Ankara, Turkey, 4Paparella Ear Head and Neck Institute, Minneapolis, MN, USA

Eustachian tube obstruction (ETO) is one of the leading causes of inflammation in the middle ear (ME). Undiagnosed and untreated otitis media can progress and result in serious pathological changes. The purpose of this study was to assess the middle ear in cats with ETO. Study materials consisted of a total of 15 temporal bones of cats with ETO at progressive time intervals of 2 days, 1‐week, 2‐week and 4‐week.

The results revealed moderate hyperplasia equally developed throughout the epithelial layer surrounding the ME and neutrophil‐rich mild inflammatory cell infiltration; more pronounced at one week obstruction compared to 2 days. Moreover, the majority of the ME cleft was filled with serous effusion. At 2‐weeks post‐obstruction, hyperplasia, edema and hyper vascularization of the ME epithelial layer were severe and there was an increased number of inflammatory cells within the serous effusion. After 4 weeks, there was a transition of ME effusion from serous to mucoid with an increased infiltration of plasma and other mononuclear cells, as well as very severe epithelial hyperplasia. In addition, granulation tissue appeared to originate from the extracellular matrix in the ME cavity and round window niche.

ME epithelial hyperplasia and characteristics of ME effusion were directly proportional to the duration of ETO. Therefore, patients presented with long‐lasting complaints of ocular and nasal discharge, epistaxis, sneezing, otorrhea, ear itching, head tilt, facial nerve paralysis, and noisy and difficult breathing should be examined for dysfunction or blockage of the Eustachian tube by diagnostic techniques such as tympanometry.

Radial and Coccygeal Artery Indirect Doppler Systolic Blood Pressure Measurements in Privately‐owned, Conscious Dogs

A. Mooney, D. Mawby, J. Price, J. Whittemore

University of Tennessee, Knoxville, TN, USA

The purpose of this study was to assess the effects of age, body condition score (BCS) and muscle condition score (MCS) on indirect radial and coccygeal Doppler systolic arterial blood pressure (SAP) measurements in dogs.

Sixty‐two privately‐owned dogs were enrolled between June and July 2016. BCS and MCS were determined by 2 investigators. Blood pressure was measured per published guidelines using headphones, and the order of measurement site was randomized. Dogs were positioned in right lateral recumbency for radial measurements and sternal recumbency or standing for coccygeal measurements. Associations between SAP and other variables were assessed by correlation coefficients and analysis of covariance.

Radial and coccygeal SAP measurements were moderately correlated (r = 0.45, P < 0.01). Radial SAP was significantly higher than coccygeal SAP (mean difference 9 mmHg, P < 0.01), but discordance occurred in both directions. No difference was observed between the first measurement taken, the average of measurements 2–6, or the average of all 6 measurements for either the radial (128, 129, and 129 mmHg; P = 0.36) or coccygeal (121, 122, and 122 mmHg; P = 0.82) site. Associations were not found between SAP measurements for either site and age, weight, BCS, MCS, anxiety score, or cuff size. Heart rate decreased significantly from the start of acclimation to the end of the first data collection series regardless of site (P < 0.01).

Measurement site can be based on patient and operator preference given lack of associations with patient variables, but the same site should be used for serial SAP measurements given discordant results between sites.

Hemodynamic and Cardio‐Respiratory Effects of an Infusion of Fentanyl, Ketamine and Lidocaine in Canines

L.A. Vergara Saldarriaga1, S.O. Padilla2, S.A. Toro3, L.F. Restrepo Betancur1

1Universidad de Antioquia, Medellín, Antioquia, Colombia, 2Director Veterinary Hospital of the University of Antioquia, Medellín, Antioquia, Colombia, 3Ocularvet, Medellín, Antioquia, Colombia

To evaluate the cardio‐respiratory and hemodynamic effects produced by the continuous infusion of Fentanyl, Lidocaine and Ketamine (FLK), and the use of Isoflurane as a general anesthetizer in ASA I patients undergoing surgery. Study design. Prospective experimental study. Animals. Nine female dogs.

The variables evaluated in the study were heart rate (HR), respiratory rate (RR), non‐invasive blood pressure (NIBP), cardiac output (CO), cardiac index (CI), arterial oxygen saturation (SaO2), end tidal CO2 (EtCO2), temperature (T °). Arterial pressure of carbon dioxide (PaCO2), arterial pressure of oxygen (PO2), bicarbonate, pH, total carbon dioxide (TCO2) and presence of lactate in the blood were also evaluated. In order to assess the influence of the medications in the cardiorespiratory function, hemodynamic function and acid‐base status, measurements were taken at two different time points: time point one (T1) corresponded to anesthesia with Isoflurane and time point two (T2), corresponds to anesthesia with Isoflurane and continuous infusion ofFLK (0.08, 30 y 15 μg/kg/min).

There is a statistically significant difference between the effects of medications on temperature and acid‐base status at T1 and T2 (P < 0.05). In contrast, there is no statistically significant difference between the effects of medications on the cardiorespiratory and hemodynamic variables at T1 and T2 (P > 0.05).

The use of FLK as a continuous infusion at dosage of 0.08, 30 and 15 μg/kg/min in patients anesthetized with Isoflurane generated statistically significant changes in the blood pH and body temperature (P < 0.05). Regarding the cardiovascular and hemodynamic variables, there were no statistically significant difference changes (P > 0.05) at the different time points although it was possible to observe a tendency to improve the tissue perfusion and hemodynamic stability. Keywords: Anesthesia, blood gases, cardiac output, I ´STAT, multimodal analgesia.

Effect of Corticosteroids on Plasma Protein Binding of Mycophenolic Acid in Dogs and Cats

A. Morassi1, S.M. Rivera1, J.E. Slovak2, M.H. Court3, N. Villarino1

1Washington State Univeristy, Pullman, WA, USA, 2College of Veterinary Medicine, Washington State University, Pullman, WA, USA, 3Program in Individualized Medicine (PrIMe), Washington State University, Pullman, WA, USA

Mycophenolic acid, the active form of mycophenolate mofetil (MMF), is effective in controlling dysregulated lymphocyte responses associated with immune‐mediated diseases. This drug has been shown to be promising for the treatment of multiple autoimmune diseases in dogs and cats. In humans, dose optimization and plasma drug concentration monitoring (TDM) are recommended in order to maximize the MPA effect, particularly in patients that require sustained immunosuppression. In humans, MPA is a highly protein bound drug (˜97%). It is recommended to monitor free drug plasma concentration, as the free (unbound) MPA correlates with its immunosuppressive effect. It is likely that dogs and cats would also need TDM and optimization of dosage regimens. However, the extent of MPA plasma protein binding in dogs and cats are unknown. Importantly, the extent of MPA protein binding cannot be extrapolated from other species due to inherent differences in albumin structure and drug binding capacity.

We hypothesized that the extent of plasma protein binding in dogs and cats was >90%. Furthermore, considering that MMF is likely to be co‐administered with corticosteroids, we also evaluated the effect of prednisolone and dexamethasone on the plasma protein binding of MPA.

The extent of plasma protein binding of MPA (15 μg/mL) and the potential interaction with dexamethasone and prednisolone in plasma were evaluated using a high throughput equilibrium dialysis device and a validated chromatographic method. The concentration of drugs used in this study corresponds to plasma concentrations reported for clinical patients treated with therapeutic doses. For determination of plasma protein binding, we used individual plasma samples from clinically healthy dogs (n = 14) and cats (n = 13). For evaluating the impact of corticosteroids on MPA plasma protein binding, we pooled plasma from 6 healthy dogs and 6 cats. All the experiments were run in triplicate, and were repeated on 3 different days. The data from the groups were analyzed by Student t‐test, and a P ≤ 0.05 was considered significant.

The extent of plasma protein binding (% mean and range) was 96 (95–97) and 92 (90–93) for dogs and cats, respectively. The unbound plasma concentration (μg/mL) ranged between 0.51 and 0.94, and 1.15 and 1.61 in dogs and cats, respectively. The mean unbound plasma concentration of MPA was statistically different (P < 0.001) between the two species. Results of the effect of corticosteroids on plasma protein binding of MPA are presented in tables 1 and 2. The mean unbound plasma concentration of MPA was higher when combined with prednisolone (P < 0.01) but not with dexamethasone (P > 0.05).

In conclusion, results of this study suggest that MPA was highly bound to plasma proteins. Furthermore, prednisolone had an apparent in vitro effect on the extent of binding of MPA in plasma from dogs and cats. These results will not be necessarily reflected on higher unbound plasma concentration of MPA in vivo due to the effect of drug distribution and elimination on plasma concentration.

Assessment of Safety, Toxicity, and Pharmacokinetics of Cannabidiol in Healthy Dogs

L. Bartner1, L. Hyatt2, L. Wittenburg3, S. Rao4, S. McGrath5

1Colorado State University, Fort Collins, CO, USA, 2College of Veterinary Medicine, University of Tennessee, Knoxville, TN, USA, 3University of California Davis, Davis, CA, USA, 4Adjunct Faculty of Epidemiology, Colorado School of Public Health, University of Colorado, Denver, Fort Collins, CO, USA, 5Department of Clinical Sciences, College of Veterinary Medicine and Biomedical Sciences, Fort Collins, CO, USA

Our purpose was to determine the pharmacokinetics (PK) and tolerability of cannabidiol (CBD) in healthy dogs. We hypothesized oral and transdermal delivery would be well tolerated and provide measurable systemic levels of CBD.

Thirty, healthy Beagle dogs were randomly assigned to receive one of three CBD formulations (capsule, oil, or transdermal cream), twice daily for a total dose of 150 or 300 mg/day (approximately 10 and 20 mg/kg/day), over 6 weeks. Blood was collected for 12 hours following the first dose for CBD PK. At 2‐, 4‐, and 6‐weeks, blood was collected for CBC, chemistry panel, bile acids, and CBD plasma levels. Daily observational notes and weekly physical examinations were also performed.

Pharmacokinetic analysis demonstrated that the oil formulation resulted in higher plasma concentrations and systemic exposure, with less variability. Cannabidiol exposure was dose proportional. Elimination half‐life of the oil formulation following a dose of 75 and 150 mg was 199.7 ± 55.9 and 127.5 ± 32.2 minutes, respectively. Twice‐daily administration maintained plasma CBD levels until study completion.

There were no significant abnormalities on CBC, urinalysis, or pre‐ and postprandial bile acids. Mild elevations in serum alkaline phosphatase (ALP) occurred at 4 and 6 weeks, more frequently with capsule and oil formulations at 300 mg/day. All dogs experienced diarrhea; presence, onset, and severity did not correlate with formulation or dose.

Cannabidiol appears to be well tolerated in dogs. Exposure is proportional to dose and oil formulation provides more favorable PK. These results provide a framework for future efficacy studies of CBD in dogs.

Concentration‐Dependent Effects of Mycophenolic Acid on Canine Lymphocytes

C. Mulligan1, K.S. Seo2, K. Lunsford2, J. Thomason1, T. Archer1, B. Kaplan2, N. Park2, L. Narayanan3, B. Lee4, A. Mackin1

1College of Veterinary Medicine, Mississippi State University, Mississippi State, MS, USA, 2College of Veterinary Medicine, Mississippi State University, Starkville, MS, USA, 3Mississippi State University, Starkville, MS, USA, 4Colorado State University, Fort Collins, CO, USA

Mycophenolate mofetil (MMF) is the prodrug of mycophenolic acid (MPA), a potent immunosuppressive agent. MPA inhibits inosine monophosphate dehydrogenase (IMPDH), an enzyme in the de novo purine nucleotide synthesis pathway. Lymphocytes lack a purine salvage pathway as an alternative source of nucleotides, and are therefore especially susceptible to the effects of MPA. MMF is reported to inhibit both B and T cell proliferation. MMF was originally developed for use in human transplant patients but, since inexpensive generic versions have become available, the drug has become popular in veterinary medicine for the treatment of immune‐mediated disorders in dogs. While it has been assumed that MMF will have effects in dogs that are similar to established effects in humans, to date very limited data are available regarding the effects of MPA on canine lymphocytes.

The objective of our current project was to evaluate the concentration‐dependent effects of MPA on canine lymphocyte function utilizing a CD4 cell activation assay, and B and T cell lymphocyte proliferation assays. Peripheral blood mononuclear cells (PBMCs) were isolated from six healthy hound dogs using Histopaque® density gradient centrifugation. Once isolated, the PBMCs were plated, incubated in escalating concentrations of MPA (1.25, 2.5, 5, 10, 20, 40, 80, 160, and 320 ng/mL), vehicle (DMSO) control, and media only control, and then stimulated with either pokeweed mitogen (PWM) or concanavalin A (ConA). The CD4 cell activation assay was modeled on a commercially available human assay of activated CD4 positive T cell ATP synthesis that is marketed as a non‐specific test of the efficacy of immunosuppressive agents. Briefly, CD4 positive cells were magnetically sorted from PBMCs following activation with ConA, cells were lysed to release intracellular ATP, a luminescence reagent (luciferin/luciferase) activated by ATP was added, and resultant bioluminescence was measured using a luminometer. B and T cell proliferation were evaluated by measuring incorporation of tritiated thymidine via a scintillation beta‐counter following activation with PWM and ConA, respectively.

Our study revealed that MPA had no concentration‐dependent effect on the CD4 cell activation assay, with no significant difference between DMSO vehicle control (376 ± 25, mean relative fluorescence units ± SEM) and even the highest MPA concentration evaluated of 320 ng/mL (352 ± 24). Lymphocyte proliferation assays, in contrast, revealed that MPA reliably inhibited both B and T cell proliferation in a concentration‐dependent manner, with an IC50 (half maximal inhibitory concentration) of between 5 and 10 ng/mL for both assays (Figure 1).

This study confirmed that MPA reliably inhibits canine B and T cell proliferation in a concentration‐dependent fashion, and that IC50 is achieved at MPA concentrations that are readily achievable at standard MMF dose rates. Based on the results of this in vitro study, oral MMF would be anticipated to have the potential to inhibit both B and T cell proliferation in dogs. Based on the results of our study, the CD4 cell activation assay was not significantly affected by exposure to MPA, suggesting that the assay, although marketed in human medicine as a non‐specific test of drug‐induced suppression of CD4 positive T cells, may not be of clinical value in dogs receiving MMF.

Pharmacokinetics of Mycophenolic Acid after Multi‐Oral Dose Administration of Mycophenolate Mofetil in Healthy Cats

J.E. Slovak1, S.M. Rivera2, J.K. Hwang2, M.H. Court3, N. Villarino2

1College of Veterinary Medicine, Washington State University, Pullman, WA, USA, 2Washington State University, Pullman, WA, USA, 3Program in Individualized Medicine (PrIMe), Washington State University, Pullman, WA, USA

Mycophenolate mofetil (MMF) is an attractive, alternative, commercially available immunosuppressant with a rapid onset of action and minimal side‐effects. Our goal in this study was to evaluate the pharmacokinetics (PK), and various clinical parameters after 1 week of oral dosing of MMF in healthy cats. The PK of mycophenolic acid (MPA), the active form of MMF, was evaluated following oral administration of 10 mg/kg PO BID (n = 3), 15 mg/kg PO BID (n = 3), and 15 mg/kg PO TID (n = 4) of MMF (CellCept® oral suspension, Roche Lab Inc.) for 1 week to a healthy population of cats. Plasma samples were taken prior to the first administration of MMF and at 0.25, 0.5 1, 1.5, 3, 6, 8, 10, 12 hs after the first MMF administration. In addition, plasma samples were taken 24, 72, 121, 145, 157 hs after the first administration (before the next dose).

Concentration of MPA and its derivatives in plasma were determined using a validated chromatographic method. Non‐compartmental analysis was done to estimate PK parameters of MPA. All cats bio‐transformed MMF into MPA. Results suggest that cats bio‐transformed MMF to MPA relatively fast, as observed in an intravenous PK experiment done in our laboratory. There was significant variation in the absorption phase of the concentration time profile of the drug as reflected in the dispersion of the Tmax values. In addition, there was significant inter‐individual variability in MPA exposure. The shape of concentration vs. time profiles of MPA after the first dose of MMF suggests that MPA undergoes enterohepatic recycling. Following multiple administrations, the trough MPA levels do not reflect a significant accumulation of drug in plasma for the BID dosage regimens evaluated. There was a small accumulation over time of MPA after the TID administration of MMF. Three metabolites of MPA were detected in cats; phenol glucuronide, phenol glucoside and acyl glucuronide after the administration of 15 mg/kg. No acyl glucuronide metabolites were identified in the 10 mg/kg PO BID group.

No cats vomited during the study, 2 of 10 cats in the 15 mg/kg TID group were hyporexic, but they recovered at discontinuation of MMF. A mild reduction in serum protein levels, PCV, and platelet counts occurred between day 1 and day 7 of the study. No other values were altered by MMF use in the cats. This study describes the clinical parameters and PK of oral MMF administration in healthy cats. Future studies will evaluate the pharmacodynamics of MPA in patients receiving these doses of MMF.

Limited Sampling Strategy to Determine Mirtazapine Pharmacokinetics in Cats with Liver Disease and Age‐Matched Controls

R. Fitzpatrick1, K. Benson2, L. Wittenburg3, R. Hansen2, P. Lunghofer2, J. Quimby4

1VCA Advanced Veterinary Care Center, Fishers, IN, USA, 2Colorado State University, Fort Collins, CO, USA, 3University of California Davis, Davis, CA, USA, 4The Ohio State University, Columbus, OH, USA

The purpose of this study is to use a limited sampling strategy to determine if pharmacokinetics of oral mirtazapine is altered in cats with liver disease compared to healthy age matched control cats (AMC).

Twenty cats were enrolled; 10 with liver disease and 10 apparently healthy age‐matched controls. Cats with liver disease did not have evidence of chronic kidney disease or hyperthyroidism based on complete blood count, serum biochemistry, urinalysis and total T4 as well as no recent exposure to mirtazapine. Apparently healthy AMC cats were screened with a physical examination, complete blood count, serum biochemistry, urinalysis and total T4. Blood was drawn from all cats 1 hour and 4 hours following one 1.87 mg oral dose of mirtazapine. In a subset of 6 cats from each group, blood was also obtained 24 hours post‐administration. Mirtazapine concentrations were measured by liquid chromatography coupled to tandem mass spectrometry. Drug exposure (AUC) and half‐life was predicted using a limited sampling approach based on multiple linear regression analysis from previous full sampling studies, and estimated using non‐compartmental methods. The data was not normally distributed and a Mann Whitney test was used to compare parameters between the two groups and Spearman Rank was used to analyze correlation between liver values and pharmacokinetic parameters.

There was a statistically significant difference in Tmax between liver disease cats and AMC cats (P = 0.03). Tmax was 4 hours in 7 liver cats versus 2 AMC cats and 1 hour in 3 liver cats versus 8 AMC cats. AUC and was not significantly different between the two groups; however, the calculated half‐life of liver disease cats was significantly increased (*P < 0.004) compared to healthy AMC cats. There was a correlation between serum ALP (P = 0.0003; r = 0.89), ALT (P = 0.003; r = 0.80) and total bilirubin (P = 0.001; r = 0.84) when compared to the serum half‐life of mirtazapine.

Liver disease in cats prolongs the half‐life of mirtazapine. Dosing with a decreased frequency should be considered in cats with liver disease. As this effect was quite variable, additional work will need to be done to determine which cats are most effected.

Antiacidic Effect of Intravenous Esomeprazole Twice a Day with Two Different Doses in Beagle Dogs

D.‐H. Seo1, Y. Cho1, K.W. Seo2

1College of Veterinary Medicine, Chungnam National University, Daejeon, Taejon‐jikhalsi, Korea, 2Chungnam National University, Daejeon, Taejon‐jikhalsi, Korea

Esomeprazole is an S‐enantiomer of omeprazole that has been reported to have more favourable pharmacokinetics and efficacious antacid properties in humans. However, there is little information about the dosage of esomeprazole and pharmacokinetic property in conscious dogs. The aim was to evaluate the efficacy of intravenous esomeprazole q12 h and to compare the efficacy between 1 mg/kg and 0.5 mg/kg in healthy dogs.

Six healthy adult colony beagle dogs were enrolled. Randomized, 2‐way crossover design. In each group, the dogs received 1 mg/kg or 0.5 mg/kg, respectively intravenously q12 h for 48 hours. The continuous intragastric pH was measured by Bravo® pH monitoring system and recorded as mean percentage time (MPT) that pH was ≥3 and ≥4 for 24 hours in each group. The comparison between groups was analysed by repeated measured analysis of variance with a posthoc Bonferroni's method.

The MPT values that pH was ≥3 and ≥4 for first 24 hours were 89.78 ± 9.08, 84.93 ± 11.30 for 1 mg/kg group and 88.18 ± 7.47 and 81.16 ± 9.43 for 0.5 mg/kg group, respectively. There are no statistical differences in MPT between groups. Both groups showed no critical adverse effects.

It is the first study to evaluate intravenous esomeprazole twice daily in non‐anesthetized healthy dogs. The results suggest that 0.5 mg/kg IV q12 h administration of esomeprazole had the same efficacy of 1 mg/kg IV q12 h in healthy dogs.

Pharmacokinetics and Antacid Efficacy of Esomeprazole After Intravenous, Oral and Subcutaneous Administration to Beagle Dogs

J.‐h. Hwang1, J.G. Cho1, S. Lim1, T. Koo2, K.W. Seo3

1Veterinary Internal Medicine, Chungnam National University, Daejeon, Taejon‐jikhalsi, Korea, 2Graduate School of New Drug Discovery and Development, Chungnam National University, Daejeon, Taejon‐jikhalsi, Korea, 3Chungnam National University, Daejeon, Taejon‐jikhalsi, Korea

Esomeprazole is an S‐enantiomer of omeprazole that has been reported to have more favourable pharmacokinetics and efficacious antacid properties in humans. However, the pharmacokinetics and effects on intragastric pH of esomeprazole in dogs have not been reported. To determine the pharmacokinetics of esomeprazole administered via various routes (PK study) and to investigate the effect of esomeprazole on intragastric pH using a Bravo pH monitoring system (PD study).

Twelve healthy adult male Beagle dogs were included in this study. Both studies used an open, randomized, crossover design. In the PK study, 7 dogs received intravenous (IV), subcutaneous (SC) and oral doses (PO) of esomeprazole (1 mg/kg). Each treatment period was separated by a washout period of at least 10 days. Esomeprazole plasma concentrations were measured by HPLC/MS/MS. In the efficacy study, intragastric pH was recorded without medication (baseline pH) and following IV, SC and PO esomeprazole dosing regimens (1 mg/kg) in 5 dogs.

The bioavailability of esomeprazole administered as PO enteric‐coated granules and as SC injections was 71.43% and 106.61%, respectively. The half‐life was approximately 1 hour. The mean percent time with intragastric pH was ≥ 3 or ≥ 4 was significantly increased; these results indicate that esomeprazole administration increased intragastric pH regardless of the dosing route (P < 0.05). The PK parameters of for PO and SC esomeprazole administration were favourable and esomeprazole significantly increased intragastric pH following IV, PO and SC administration.

IV and SC administration of esomeprazole may be useful when PO administration is not possible. No significant adverse effects were observed.

Prediction of Platelet Responsiveness to Aspirin and Fluoride Effects on Esterase Activity in Healthy Dogs

J.M. Haines1, R. Hegedus1, P.M. Lee1, M.H. Court2

1Washington State University, Pullman, WA, USA, 2Program in Individualized Medicine (PrIMe), Washington State University, Pullman, WA, USA

Acetylsalicylic acid (ASA), commonly known as aspirin, is a cyclooxygenase inhibitor frequently used to inhibit platelet function and formation of thromboemboli in at risk patients. A potential phenomenon of resistance to these antiplatelet effects has been documented in people and potentially in dogs. Aspirin resistance may be a contributing factor to the development of life threatening thrombi thus increasing the rate of morbidity and mortality in several canine diseases. Understanding the phenomenon of aspirin resistance and developing methods for predicting it could lead to improvements in survival in these patients. The cause of aspirin resistance remains unclear. Individual variability in esterase activity and therefore hydrolysis of aspirin has been suggested as a potential cause of poor bioavailability of aspirin in blood in people.

The objectives of our study were to determine whether the ASA concentration needed to cause 50% inhibition of platelet aggregation (IC50) in vitro in healthy dogs was predictive of the IC50 value measured in the same dogs after administration of oral aspirin (in vivo IC50). Furthermore, we wanted to evaluate whether the addition of fluoride, known to cause non‐specific esterase inhibition, altered the concentrations of ASA or its metabolite, salicylic acid (SA), in blood samples collected after oral aspirin administration.

Twenty‐one healthy client‐owned dogs were enrolled. First, in vitro IC50 testing was performed twice (trial 1 and 2) at least 1 week apart to determine individual repeatability. Whole blood samples were drawn and incubated with 6 different aspirin concentrations. Aggregation was measured in AUC (area under the curve) units using a multiple electrode impedance aggregometer (Multiplate). For in vivo IC50 testing, dogs were given a single oral dose of 4 mg/kg aspirin. Aggregation was assessed prior to aspirin administration then 20, 40, and 180 minutes after administration. At each time point, samples were also collected into EDTA and EDTA‐fluoride blood tubes for measurement of ASA and SA plasma concentrations using liquid chromatography with mass spectrometry.

Based on calculation of in vitro IC50, dogs were separated into 3 different response groups. Eighteen of the 21 dogs (86%) were consistent as aspirin responders with in vitro IC50 values below 250 umol/L (median 66.7 umol/L, range 34.8–151.4). One dog showed a consistently poor response with in vitro IC50 values between 250–500 umol/L for both trials. Two dogs had marked variation in their in vitro IC50 values between the trials and were classified as variable responders. No dogs were found to be aspirin non‐responders consistently for both trials. Between trial variation in in vitro IC50 for all dogs and the 18 dogs with IC50 values below 250 umol/L was mean (range) of 51.9% (3.4–194.3) and 38.6% (3.4–78.8), respectively. After administration of 4 mg/kg aspirin orally, 8 of 21 dogs did not show > 50% platelet inhibition by 3 hours. In vivo IC50 values for the remaining 13 dogs for samples collected in EDTA and EDTA fluoride were median (range) of 31.7 umol/L (0.1–107.6) and 30.9 umol/L (0.1–92.4), respectively. All three dogs identified as poor or variable responders in vitro had > 50% inhibition of platelet aggregation in vivo. The addition of fluoride to blood samples resulted in a mean percent reduction in SA concentrations of 17.3% and mean percent increase in ASA concentrations of 41.4%.

This study suggests individual variation in aspirin response does exist in dogs as assessed by multiple electrode aggregometry. A single low dose (4 mg/kg) of aspirin was not sufficient to cause measurable inhibition of platelet aggregation in all dogs and resulted in ASA concentrations that were substantially less than those used in vitro. Aspirin responsiveness in vitro did not correlate with responsiveness in vivo in this scenario. The use of EDTA fluoride anticoagulant resulted in inhibition of conversion of ASA to SA suggesting successful inhibition of esterase activity. When storing dog plasma samples containing aspirin, the use of EDTA containing fluoride may prevent esterase hydrolysis of the ASA. Further investigation into the effects of esterase activity as a cause of aspirin resistance in dogs is warranted.

Pharmacokinetics of an Orally‐Disintegrating Sildenafil Film After Oral and Rectal Routes Administration to Beagle Dogs

H. Yang1, H. Hong1, K.W. Seo2

1College of Veterinary Medicine, Chungnam National University, Daejeon, Taejon‐jikhalsi, Korea, 2Chungnam National University, Daejeon, Taejon‐jikhalsi, Korea

Sildenafil citrate, a highly selective phosphodiesterase type V inhibitor, is used to treat pulmonary hypertension (PH) in veterinary medicine. Sometimes, patients with PH experience respiratory distress, and the oral administration of any type of drug to these patients can be fatal. For such patients, the rectal route should be carefully considered as an alternative means of administering drugs.

The objective of this study was to compare the pharmacokinetic characteristics of an orally disintegrating film (ODF) formulation of sildenafil with those of a conventional film‐coated tablet (FCT) formulation in healthy dogs. Additionally, the pharmacokinetic profile of the ODF formulation of sildenafil was analyzed after its rectal administration.

Twelve healthy beagle dogs were divided into four groups and given sildenafil per the following specifications: FCT formulation 2 mg/kg PO (per oral), ODF formulation 2 mg/kg PO, ODF formulation 2 mg/kg diluted in distilled water and administered rectally, and ODF formulation 10 mg/kg dissolved in water and administered rectally. Blood samples were collected over the course of 24 h and the concentrations of sildenafil were assayed.

There were no significant differences in T1/2 and Tmax across the four arms. The Cmax of the rectally administered 10 mg/kg dose did not differ significantly from those of the orally administered 2 mg/kg doses. In contrast, the Cmax of the rectally administered 2 mg/kg dose was significantly lower than those of the orally administered 2 mg/kg doses and rectally administered 10 mg/kg dose. The AUC after rectally administering the 2 mg/kg dose was significantly smaller than those obtained after sildenafil was administered orally. The AUC after rectally administering the 10 mg/kg dose was significantly higher than those for the oral administration arms. There were no notable systemic adverse effects in any of the groups.

During emergencies, it may be helpful for sildenafil to be administered rectally to respiratory distress patients. With the exception of an increased AUC, the pharmacokinetic characteristics of the rectally administered 10 mg/kg dose have been shown to be comparable to those of an orally administered 2 mg/kg dose.

Safety of Mirtazapine 2% Transdermal Ointment Administered Topically to Cats at 5 mg/kg for 28 days

A. Avenatti1, W. Buhles2, J.J. Quimby3, D. Labelle4, M.P. O'Banion1

1Kindred Biosciences, Inc, Burlingame, CA, USA, 2Kindred Biosciences, Inc, Fairfax, CA, USA, 3Colorado State University, Fort Collins, CO, USA, 4Kindred Biosciences, Inc, Atherton, CA, USA

Mirtazapine 2% transdermal ointment (2% MZP) achieves clinically relevant serum concentrations in cats and results in increased body weights at a dose of 0.5 mg/kg. The purpose of this study was to evaluate the safety profile at an exaggerated dose.

In this unmasked study, 2% MZP was applied topically to the inner pinna of five healthy adult intact female cats and two intact female cats with chronic kidney disease (CKD) (serum creatinine 1.7 mg/dL and 1.5 mg/dL; urine specific gravity 1.023 and 1.037 respectively) daily for 28 consecutive days at 5 mg/kg. Cats ranged in age from 1.4 to 3.2 years. 2% MZP volume was split equally between ears at each administration. Behavioral observations were conducted twice daily throughout the study and scored on interaction, activity, vocalization, and tremors. Physical examination, ECG, and doppler blood pressure were conducted on study days −5, 13 and 27. Scanning echocardiogram and abdominal ultrasound were conducted on study days −6 and 26. Hematology, and serum chemistry were conducted on study days −6, 13, and 27. Body weights were measured on study days −5, −1, 6, 13, 20 and 27. Body weight, blood pressure, and serum ALT were compared between time points using a Wilcoxon Signed Rank Test for paired samples. Sum of pre‐dose mean behavior scores were compared to mean behavior scores at Day 27 using a Wilcoxon Signed Rank Test for paired samples. Statistical significance was set at P < 0.05 for all tests.

Erythema of the inner pinna was a common finding in multiple cats at various time points with one cat showing an excoriation of one ear on day 5 which subsequently healed uneventfully. One cat with CKD exhibited vomiting between study days 16 and 21 and recovered without treatment. The same cat showed intermittent bouts (2–3 sec.) of spasms and ataxia with tail curled over the back on day 20 only. All cats gained a significant amount of weight between Day 0 and Day 28 (P = 0.016). Blood pressure was analyzed in the healthy cats at a single time point and there was no significant difference compared to baseline at any time point. Hematology and serum chemistry showed no clinically significant changes in either apparently healthy cats or the two cats with CKD. There were no clinically significant changes in cardiac findings. There was no significant difference in behavior scores between the beginning and end of the study despite a 5× dose of mirtazapine.

Mirtazapine 2% transdermal ointment was well tolerated in cats when administered at 5 mg/kg for 28 consecutive days.

Double‐Blind, Placebo‐Controlled, Randomized Study of Transdermal Mirtazapine Ointment for the Management of Feline Weight Loss

K. Longpre1, W. Buhles2, M. Tin1, T. Hu1, J.J. Quimby3, D. Labelle4, M.P. O'Banion1, J. Lee1, S. Markey1

1Kindred Biosciences, Inc, Burlingame, CA, USA, 2Kindred Biosciences, Inc, Fairfax, CA, USA, 3Colorado State University, Fort Collins, CO, USA, 4Kindred Biosciences, Inc, Atherton, CA, USA

Mirtazapine 2% Transdermal ointment (2% MZP) has been demonstrated to achieve clinically relevant serum concentrations in cats. The purpose of this study was to demonstrate the effectiveness and safety of 2% MZP for the management of weight loss in cats.

This study was a multi‐center, double‐blind, placebo‐controlled, randomized study. Cats at least 1 year of age, weighing ≥ 2 kg, with a documented medical history of ≥ 5% body weight loss that was deemed clinically significant were enrolled. Cats could not have a serum creatinine > 5.0 mg/dL or diagnosed neoplasia. Randomization was 1:1 (mirtazapine:placebo). 2% MZP was dosed at 2 mg per cat (equivalent to 0.1 mL volume of ointment) regardless of body weight. The placebo ointment was identical to treatment ointment but without active ingredient mirtazapine, and was dosed at the same volume. The ointments were administered to the inner pinna daily for 14 ± 3 days. Effectiveness was demonstrated by comparison of mean percent change in body weight between the Day 1 and Week 2 visits using a two‐sample t‐test with significance set at P < 0.05 where the percent change in body weight in the 2% MZP group was greater than or equal to 0. Clinical pathology data from all cases receiving the 2% MZP or placebo were included in the safety analysis and each variable was analyzed using Analysis of Covariance, with treatment as a fixed effect and the baseline value as a covariate. Adverse events for safety analysis included any observations, whether or not considered to be product related, that were unfavorable and unintended, and that occurred after the use of study drug on Day 1 through study termination.

Two‐hundred thirty‐one (231) client‐owned cats were enrolled in the study with 230 being administered at least one dose of 2% MZP or placebo and included in the safety analysis. One‐hundred eighty‐seven (187) cats were determined to be evaluable for the effectiveness analysis. Of those, 90 cats were in the 2% MZP group and 97 cats were in the placebo group.

The results are summarized in Table 1. The difference in mean percent change in body weight (95% C.I.) between the 2% MZP group and the placebo group was 3.78% (2.49%, 5.07%), which was statistically significant (P < 0.0001). No clinically relevant statistically significant changes in clinical pathology parameters were seen. Adverse events reported in the 230 cat safety population were mild with the most common being application site reactions observed in 31.3% of placebo treated cats and 14.8% of 2% MZP treated cats. Nine cats had serious adverse events while on study, related to their underlying disease, of which, 3 cats received 2% MZP and 6 cats received placebo.

Mirtazapine 2% transdermal ointment is effective in managing weight loss in cats. Use of 2% MZP resulted in a mean weight gain of 4.07% compared to 0.29% for placebo after two weeks of transdermal treatment. The product was generally well tolerated.

Single Dose Pharmacokinetics Following Transdermal and Oral Administration of Mirtazapine 2% Transdermal Ointment in Cats

B. Mason1, W. Buhles2, J.J. Quimby3, D. Labelle4, D. Jung5, M.P. O'Banion1

1Kindred Biosciences, Inc, Burlingame, CA, USA, 2Kindred Biosciences, Inc, Fairfax, CA, USA, 3Colorado State University, Fort Collins, CO, USA, 4Kindred Biosciences, Inc, Atherton, CA, USA, 5Pharmaceutical Research Services, Cupertino, CA, USA

Mirtazapine has been shown to increase body weight in cats. The purpose of this study was to evaluate single dose pharmacokinetics (PK) of mirtazapine in a novel formulation of mirtazapine 2% transdermal ointment (2% MZP) in cats.

This study was a randomized, masked, cross‐over PK study. Eight apparently healthy adult cats 6–12 years of age were studied. All cats were healthy as determined by daily clinical observations through a 7 day acclimation period, physical examination, and CBC/serum chemistry. All cats wore Elizabethan collars throughout the course of the study. On Day 0, four cats were treated once with 0.5 mg/kg of 2% MZP applied to the inner pinna of the right ear, and four cats were treated once with 0.5 mg/kg of 2% MZP given orally by syringe to the back of the mouth. On Day 5 the cats crossed over to the opposite treatment. Plasma was collected pre‐dose and at 1, 2, 4, 6, 8, 12, 24, 48, 72 and 96 hours after administration and frozen at −20°C prior to shipping and analysis. Plasma mirtazapine concentrations were measured using a validated LC‐MS/MS method. PK parameters were calculated using standard methods WinNonlin Professional 5.3. PK parameters were statistically compared between treatments using Wilcoxon Signed Rank test with significance set at P < 0.05.

The PK parameters are summarized in Table 1. PK parameters were significantly different between the orally and transdermally administered 2% MZP (Tmax P = 0.02; Cmax P = 0.04; Half‐life P = 0.03) with the exception of AUC0‐∞ (P = 0.09).

Transdermal application of 2% MZP to the inner pinna achieves measureable plasma concentrations in cats. Absorption of 2% MZP was faster and more consistent following oral administration as compared to topical application. Additionally, the mean extent of absorption following oral administration was about 2‐fold higher than following topical application of 2% MZP. The longer apparent terminal half‐life observed following topical as compared to oral administration (26.8 vs. 10.1 hr) suggests the presence of flip‐flop pharmacokinetics, where the rate of absorption is significantly slower than the rate of elimination.

Multiple Dose Pharmacokinetics of Mirtazapine 2% Transdermal Ointment in Cats

M.P. O'Banion1, W. Buhles2, J.J. Quimby3, D. Labelle4, D. Jung5

1Kindred Biosciences, Inc, Burlingame, CA, USA, 2Kindred Biosciences, Inc, Fairfax, CA, USA, 3Colorado State University, Fort Collins, CO, USA, 4Kindred Biosciences, Inc, Atherton, CA, USA, 5Pharmaceutical Research Services, Cupertino, CA, USA

Mirtazapine has been shown to increase body weight in cats and is transdermally absorbed. The purpose of this study was to evaluate multiple dose pharmacokinetics (PK) of mirtazapine 2% transdermal ointment (2% MZP) in cats.

This study was a masked, randomized, three‐arm parallel study to determine the plasma PK of two dosages (0.5 and 2.0 mg/kg) of 2% MZP applied topically once daily for 14 days to healthy adult cats (6–13 years old) as determined by CBC/serum chemistry and physical exam. Twenty cats were randomized: eight to 0.5 mg/kg, eight to 2.0 mg/kg, and four cats to no treatment. Cats underwent acclimation for 7 days, a dosing phase of 14 days, and a post‐dosing phase of 3 days for PK sampling. Cats wore Elizabethan collars for the duration of the study. Cats were assessed daily for general health and application site abnormalities. Physical examinations and body weights were recorded at the beginning and end of the study. Plasma samples for mirtazapine concentrations were collected within one hour prior to treatment on the last dosing day, and at 1, 2, 4, 6, 8, 12, 24, 48, 72 and 96 h post final dosing and analyzed using a validated LC‐MS/MS method. Noncompartmental pharmacokinetic analysis was performed with WinNonlin Professional ver 5.3.

The PK profile is summarized in Table 1. Mean ± SD body weight for cats that received 0.5 mg/kg 2% MZP was 5.4 kg ± 1.1 kg prior to treatment and 5.7 ± 1.2 kg after 14 days of treatment. Mean ± SD body weight for cats that received 2.0 mg/kg 2% MZP was 5.3 kg ± 1.1 kg prior to treatment and 5.7 ± 1.2 kg after 14 days of treatment. Mean ± SD body weight for control cats was 5.8 kg ± 1.2 kg at baseline and 6.1 ± 1.2 kg after 14 days. Mild redness of the pinna (application site) was noted following administration but there was no pinnal excoriation or ulceration observed in any cat.

After once daily dosing for 14 days, mirtazapine 2% transdermal ointment achieves clinically relevant serum concentrations in cats.

Oral Mycophenolate Mofetil Dose Escalation Trial Assessing Adverse Effects and Pharmacodynamic Responses in Normal Dogs

C. Mulligan1, K.S. Seo2, B. Kaplan2, K. Lunsford2, J. Thomason1, T. Archer1, N. Park2, L. Narayanan3, K. Alexander2, A. Mackin1

1College of Veterinary Medicine, Mississippi State University, Mississippi State, MS, USA, 2College of Veterinary Medicine, Mississippi State University, Starkville, MS, USA, 3Mississippi State University, Starkville, MS, USA

Mycophenolate mofetil (MMF) is a purine synthesis inhibitor and immunosuppressive agent that was originally developed for use in human organ transplantation, but has recently been increasingly used in veterinary medicine for treatment of canine immune‐mediated disorders. In dogs, however, there are very limited experimental data available to enable establishment of an effective dose rate, and dose rates vary widely in published case series and anecdotal clinical reports. Adverse effects, particularly diarrhea, are frequently reported in both human and canine patients, and often lead to either dose reduction or drug discontinuation.

The objectives of our study were to, utilizing a dose escalation study in healthy dogs, determine the maximum tolerated dose (MTD) of MMF in each animal, and determine the effects of this dose on pharmacodynamic assays (B and T cell proliferation) focused on the established drug mechanism of action. Eight adult hound dogs were given oral MMF at three different escalating dose rates (10 mg/kg bid, 15 mg/kg bid, and 20 mg/kg bid) for one week at each dose, with a 2 week washout period between doses, and monitored for significant diarrhea, vomiting or inappetence. Once MTD was determined, dogs were administered this dose rate for 2 weeks, and blood was collected on Days 0, 7 and 14 for B and T cell proliferation assays measuring incorporation of tritiated thymidine following activation with pokeweed mitogen and concanavalin A, respectively. A complete blood count (CBC), serum biochemistry, and urinalysis were also performed at the beginning and end of the study.

All 8 dogs tolerated oral MMF dose rates of 10 and 15 mg/kg bid for one week without significant side effects. Three dogs developed significant diarrhea (one of which also developed vomiting) at 20 mg/kg bid for one week, leading to a final starting MTD of 15 mg/kg bid (3 dogs) and 20 mg/kg bid (5 dogs) for the subsequent 2 week trial. However, 2 of the dogs on 15 mg/kg bid then developed significant diarrhea at the start of the second week of drug administration, necessitating a further dose reduction to 10 mg/kg bid. Therefore, by the end of the second week, 5 dogs were on 20 mg/kg, 1 dog was on 15 mg/kg, and 2 dogs were on 10 mg/kg bid. There were no significant changes in CBC, serum biochemistry or urinalysis results over the study period.

B and T cell proliferation results after one week of oral MMF at MTD were not significantly different than Day 0 results (pretreatment baseline) (Figure 1). However, after 2 weeks of oral MMF, both B and T cell proliferation were significantly reduced: B cell proliferation (counts/min) was on average reduced to 49% of baseline, and T cell proliferation to 42% of baseline. Interestingly, in all 3 dogs on reduced MMF doses, both B and T cell proliferation were still decreased by more than 50%.

Our study confirmed that oral MMF at maximally tolerated dose rates significantly inhibits both B and T cell proliferation in dogs. This effect, however, appears to be delayed for several weeks, suggesting that MMF may not be the ideal treatment choice for dogs with severe or emergent immune‐mediated diseases that require rapid immunosuppression. The mechanism for this delay in the immunosuppressive effects of MMF has not been elucidated, although another purine synthesis inhibitor, azathioprine, is also reported to have a delayed effect, the cause of which has also not yet been determined. Diarrhea was, as has been previously reported, the most common dose‐limiting side effect in our study. The exact mechanism of MMF‐associated diarrhea has not been determined in either humans or dogs. Interestingly, in some dogs, drug‐induced diarrhea, as with drug effects on lymphocyte proliferation, appears to be delayed until after the first week of therapy. Since a lower oral dose rate (10 mg/kg bid) of MMF still causes significant suppression of lymphocyte proliferation in some dogs, without associated diarrhea, this reduced dose rate warrants exploration in future pharmacodynamic studies.

Functional Evaluation of the Canine CYB5R3 A729G Single Nucleotide Polymorphism Associated with Sulfonamide Hypersensitivity

J.M. Reinhart, J.L. Ekena, L.A. Trepanier

Department of Medical Sciences, School of Veterinary Medicine, University of Wisconsin‐Madison, Madison, WI, USA

The enzyme cytochrome b5 reductase (encoded by CYB5R3) detoxifies reactive metabolites of sulfonamide antibiotics. The single nucleotide polymorphism (SNP) CYB5R3 729A>G was recently associated with sulfonamide hypersensitivity in dogs. This synonymous coding SNP may partially explain risk for this idiosyncratic drug reaction, however, the functional significance of this variant is unknown. Based on in silico predictions, we hypothesized that the 729A>G SNP would decrease protein synthesis and total cellular cytochrome b5 reductase activity without decreasing mRNA expression in an in vitro model. HepG2 cells were transfected with an expression vector containing either wildtype (A) or variant (G) CYB5R3 cDNA and incubated for 48 hours. Cells were harvested and CYB5R3 mRNA, protein, and function were measured with qRT‐PCR, immunoblots, and cytochrome c reduction assay, respectively. There was no significant difference in protein expression between the A and G alleles. mRNA quantification and cytochrome b5 reductase activity assays are underway. Preliminary results do not support CYB5R3 729A>G as a functional variant leading to decreased cytochrome b5 reductase expression. However, immunoreactivity is only a semi‐quantitative measure of protein expression; results of mRNA and enzyme activity assays may better inform the mechanistic significance of 729A>G SNP in sulfonamide hypersensitivity in dogs.

Recovery of T‐Cell Function in Healthy Dogs Following Cessation of Cyclosporine Therapy

J. Thomason1, C. Mulligan1, L. Narayanan2, L. Durso1, B. Thames1, A. Mackin1, T. Archer1

1College of Veterinary Medicine, Mississippi State University, Mississippi State, MS, USA, 2Mississippi State University, Starkville, MS, USA

Cyclosporine is a potent immunosuppressive agent commonly used to treat immune‐mediated disorders in dogs. Cyclosporine targets T lymphocytes by decreasing cytokine expression (IL‐2, IL‐4 and IFN‐γ) of activated T cells. Unfortunately, significant complications, such as life‐threatening secondary infections, can sometimes arise as a result of cyclosporine administration, necessitating cessation of therapy. Once cyclosporine is stopped, the expectation is that the patient's immune system will completely recover, and commence functioning appropriately. Currently, it is unknown how quickly the immune system and T‐cell function recover following the cessation of cyclosporine therapy. The objective of this study was to utilize a pharmacodynamic bioassay to quantify the average time it takes for T‐cell function to return to pre‐treatment levels in healthy dogs after cessation of an immunosuppressive dose of cyclosporine.

Six healthy adult dogs were given cyclosporine (10 mg/kg BID, PO) for 7 days. Whole blood gene expression of IL‐2 and IFN‐γ was evaluated before treatment and again on Day 7 of drug administration, two hours after giving the medication, using quantitative reverse transcription polymerase chain reaction (qRT‐PCR). Samples were activated for 5 hours with phorbol 12‐myristate 13‐acetate and ionomycin, RNA was extracted, and relative gene quantification was performed comparing gene expression after drug administration to pre‐treatment levels. Additionally, for one week after cessation of cyclosporine, daily blood samples were collected for evaluation of IL‐2 and IFN‐ γ gene expression.

After 7 days of oral cyclosporine, both IL‐2 and IFN‐ γ were markedly suppressed, reaching 1% and 2% of pre‐treatment values, respectively. Mean IL‐2 and IFN‐γ expression remained decreased below pre‐treatment levels on Days 1 and 2 of washout (WO), but by WO Day 3 were 115% (SEM, 25%) and 103% (23%) of pre‐treatment values, respectively. By WO Day 4, mean IL‐2 and IFN‐γ expression were 156% (25%) and 125% (25%) of pre‐treatment values, respectively.

Our results suggest that, following cessation of immunosuppressive doses of cyclosporine, T‐cell expression of IL‐2 and IFN‐γ will begin to return to pre‐treatment values within the initial 24 hours. In some dogs, IL‐2 and IFN‐γ expression return to pre‐treatment values within 3 days of stopping cyclosporine, but other dogs may require an additional recovery day. Interestingly, upon recovery, both IL‐2 and IFN‐γ expression exceed pre‐treatment values, suggesting a potential rebound effect following the cessation of cyclosporine therapy. Additional studies are required to determine the timing of post‐cyclosporine recovery of T‐cell function in dogs with naturally occurring immune‐mediated disorders.

Effect of Oral Yunnan Baiyao on Periprocedural Hemorrhage and Coagulation in Dogs Undergoing Nasal Biopsy

L. Adelman, S. Olin, C.M. Egger, J.E. Stokes

College of Veterinary Medicine, University of Tennessee, Knoxville, TN, USA

The hemostatic efficacy and safety of Yunnan baiyao (YB) has been demonstrated across multiple species. Despite the widespread anecdotal use of YB in veterinary medicine for dogs with bleeding tendencies, few studies have evaluated its efficacy in dogs. The primary objective of this study was to evaluate the effect of YB on clinical hemorrhage following nasal biopsy in dogs as measured by blood loss and time to stop bleeding. A secondary objective was to evaluate the effect of YB on measures of coagulation including buccal mucosal bleeding time (BMBT) and thromboelastography (TEG).

Nineteen client‐owned dogs who were presented to the UT‐VMC internal medicine service for nasal biopsy between March 2015 and August 2016 were enrolled prospectively. Dogs were randomized to treatment (YB) or placebo control group. Owners, statisticians and primary investigators were blinded to which treatment the dog received. Each dog received 2 doses of drug, one dose at 6 pm the evening before and one dose at 6 am the morning of the procedure according to body weight. All dogs had baseline blood pressure, platelet count, prothrombin time and activated partial thromboplastin time performed at presentation and a BMBT performed before and after capsule administration. Dogs were excluded from the study if systolic blood pressure was >160 mmHg, platelet count 13 sec, PTT > 90.6 sec or BMBT > 4 minutes. In addition to BMBT, 9/19 had both TEG and blood loss measurements performed, 6/19 had only TEG performed, and 4/19 had only blood loss measurements performed. Measurements of blood loss included total blood loss, blood loss as a percentage of body weight and time to stop bleeding. A standardized anesthetic protocol was followed. A final diagnosis was determined by histopathologic evaluation of biopsy samples, CT findings and/or cytological exam on an individual case basis. Statistical analysis of blood loss measures was performed using a two‐way ANCOVA. A repeated measures mixed effects ANCOVA was used to analyze BMBT and TEG variables (alpha angle, K, R, MA). Backwards variable elimination was performed to evaluate the effect of covariates. A P value < 0.05 was considered statistically significant.

Time to stop bleeding was significantly shorter in the YB group (300 ± 12 sec) compared to the control group (367 ± 9 sec) and was not influenced by underlying disease. There was no significant effect of body weight, weight of biopsies or weight of biopsies as a percentage of body weight on time to stop bleeding. Age and presence of epistaxis at presentation were significantly associated with a longer time to stop bleeding while blood pressure and biopsy number were significantly associated with a shorter time to stop bleeding. Treatment effect remained significant even when the effect of these covariates were considered. Total blood loss and blood loss as a percentage of body weight were not significantly different between groups. There was no effect of YB on measures of coagulation including BMBT or TEG. No adverse effects were noted during or after treatment in either group.

In this study pre‐procedural oral administration of YB to dogs undergoing nasal biopsy resulted in significantly shorter time to stop bleeding following biopsy. The results of this study support the use of YB prior to nasal biopsy in dogs. Additional studies are needed to determine the clinical use of YB for other procedures or medical conditions associated with hemorrhage in dogs. Furthermore, pharmacokinetic studies are needed to determine optimal oral dosing of YB in dogs.

Bacterial Infection in Canine Tracheal Collapse Syndrome Patients Prior to, and Following, Tracheal Stent Placement

S. Lesnikowski1, C. Weisse2, A. Berent2, A.L. Roux2

1ASPCA, New York, NY, USA, 2Animal Medical Center, New York, NY, USA

The purpose of this study was to determine the incidence of positive pathogenic pre‐stent and post‐stent cultures in dogs with tracheal collapse (TC).

This was a retrospective study with review of records of 26 traditional‐type TC and 27 malformation‐type TC dogs receiving tracheal stents at the Animal Medical Center from 2009 to 2014. All dogs with tracheal stent placement and concurrent endotracheal culture were included as pre‐stent cultures. Any subsequent cultures obtained were considered post‐stent cultures. Positive pathogenic cultures were defined as positive bacteriologic cultures with compatible cytologic and/or radiographic evidence of infection.

Positive pathogenic cultures occurred in 25/38 (66%) pre‐stent and 20/31 (65%) post‐stent dogs. Traditional‐type TC dogs had positive pathogenic pre‐stent and post‐stent cultures in 16/26 (62%) and 7/26 (27%), respectively. Malformation‐type TC dogs had positive pathogenic pre‐stent and post‐stent cultures in 9/27 (33%) and 13/27 (48%), respectively. The most common isolates included Escherichia spp (16%), Pseudomonas spp (16%), and Staphylococcus spp (16%) in pre‐stent patients and Escherichia spp (29%) and Pseudomonas spp (12%) in post‐stent patients.

Positive pathogenic cultures were present in a similar majority of dogs with tracheal collapse prior to, and following, tracheal stent placement. Malformation‐type TC dogs had a higher incidence of pathogenic post‐stent cultures compared to traditional‐type TC dogs. The risk of a positive post‐stent culture increased in malformation‐type TC but decreased in traditional‐type TC patients. Tracheal stents may not increase the risk of airway infections in all tracheal collapse patients. Bacterial cultures in TC dogs may reflect poor clearance of pharyngeal and enteric organisms.

Oral Probiotic Bacterial Species Can Colonize the Healthy Feline Airway Microbiota

A.I. Vientos Plotts1, A. Ericsson2, H. Rindt3, M. Grobman1, C. Reinero1

1College of Veterinary Medicine, University of Missouri, Columbia, MO, USA, 2College of Veterinary Medicine, University of Missouri/University of Missouri Metagenomics Center, Columbia, MO, USA, 3Comparative Internal Medicine Laboratory/College of Veterinary Medicine, University of Missouri, Columbia, MO, USA

While probiotics may beneficially alter the gastrointestinal microbiota, no studies in cats have determined if oral probiotics affect the composition of the airway microbiota. Using 16S rRNA amplicon sequencing of bacterial DNA extracted from oropharyngeal (OP) swabs, bronchoalveolar lavage (BAL), feces, and blood, our objective was to characterize how oral probiotics changed the composition of the upper and lower airway, fecal, and blood microbiota. We hypothesized that probiotics would modulate the respiratory microbiota in healthy cats, demonstrated by the detection of the probiotic bacterial species present in fecal samples and altered “core” airway microbiome communities in the respiratory tract.

Six healthy research cats < 1‐year‐old had OP, BAL, fecal, and blood samples collected at baseline and 4 weeks after receiving oral probiotics. Extracted DNA underwent PCR of the 16S rRNA gene. Once sequenced, coverage (number of reads/sample), richness, and relative abundance of representative operational taxonomic units (OTUs) were determined. Principal component analysis (PCA) demonstrated relatedness of samples.

After receiving probiotics, coverage in BAL (mean ± SD, 269 ± 23 to 11,122 ± 23,546 reads/sample; P = 0.002), and blood (269 ± 43 to 1,513 ± 631 reads/sample; P = 0.002) significantly increased. There was no significant difference in coverage in OP and fecal swabs (P = 0.56, and P = 0.59, respectively). Richness in the upper (from 105 ± 8 to 124 ± 13; P = 0.030) and lower airways (from 27 ± 4 to 68 ± 31; P = 0.002) and blood (from 15 ± 6 to 28 ± 11; P = 0.041) significantly increased. In feces richness did not significantly differ after probiotics (P = 0.485). The probiotic capsule resolved to the OTU level consisted of the following taxa: genus Bifidobacterium (36%), Bifidobacterium animalis (23%), genus Streptococcus (26%), Lactobacillus zeae (12%), and genus Lactobacillus (3%). B. animalis and L. zeae were not detected in any sample site at baseline. L. zeae was detected in all sites post‐treatment, and B. animalis was detected in OP, BAL, and feces, but not blood. The overall composition of the upper and lower airway microbiota underwent the most changes in regards to relative abundance of predominant OTUs with 4.8% and 6.9% of OTUs changing significantly in OP and BAL samples, respectively. In contrast only 0.5% of OTUs in feces changed significantly from baseline. PCA revealed distinct microbiomes in each site; samples clustered more tightly at baseline and had more variation after probiotic administration.

This is the first study describing the effect of oral probiotics on the respiratory microbiota of healthy cats. Finding bacterial species present in the oral probiotics in the upper and lower airways provides pilot data suggesting that oral probiotics could be investigated as a tool to target dysbiosis occurring in inflammatory airway diseases such as feline asthma.

Thoracic Radiograph Findings in Feline Histoplasmosis

L. Kuzimski1, L.A. Nafe1, K. Sippel2, A. Hanzlicek3

1Oklahoma State University, Stillwater, OK, USA, 2Gulf Coast Veterinary Specialists, Stillwater, OK, USA, 3College of Veterinary Medicine, Oklahoma State University, Stillwater, OK, USA

Histoplasmosis is a common fungal infection that has the potential to result in severe clinical disease, with the majority of cats having pulmonary involvement. The purpose of this study was to evaluate thoracic radiograph findings in cats diagnosed with histoplasmosis. Medical records of cats presenting to the Oklahoma State University VTH between 2005 and 2015 were reviewed. Cats included in the study had orthogonal thoracic radiographs performed, a complete medical record, and a definitive diagnosis of histoplasmosis based on identification of H. capsulatum, positive urine antigen, and/or positive fungal culture. A radiologist blinded to the clinical severity of each patient evaluated all radiographs to determine the predominant pulmonary pattern and severity score (0‐normal, 1‐mild, 2‐moderate, 3‐severe) for each lung lobe. Forty‐seven cats (27 female, 20 male) were included in the study with a median age of 7 years (range: 0.7 – 19 years) of which 32 (68%) were presumed to be alive at the time of this abstract. The predominant pulmonary pattern was bronchial in 38 cats (81%) and 6 cats (13%) had a mixed pulmonary pattern (combination of bronchial, interstitial, and/or alveolar). Three cats (6%) had bilateral pleural effusion and 3 cats (6%) had normal radiographs. All lung lobes were abnormal in the majority of cats and the median severity score of all lung lobes was 2 out of 3 (moderate). Feline histoplasmosis results in pulmonary abnormalities in the majority of cats infected, and the predominant pulmonary pattern is a moderate bronchial pattern with or without an interstitial component.

Frequency of Bronchoalveolar Lavage Hemosiderosis in 171 Dogs and 34 Cats with Respiratory Disease (2007–2016)

S. Jelovcic1, A. Defarges1, K. Hooi1, S. Nykamp1, J.S. Weese2, D. Bienzle3

1University of Guelph, Guelph, ON, Canada, 2Department of Pathobiology, Ontario Veterinary College, University of Guelph, Guelph, ON, Canada, 3Department of Pathobiology, University of Guelph, Guelph, ON, Canada

To retrospectively evaluate the presence of hemosiderin in canine and feline bronchoalveolar lavage (BAL) samples and possible association with signalment, pre‐BAL treatment, diagnosis, blood parameters, transthoracic fine‐needle aspiration, BAL procedure time and prior immune‐mediated disease or neoplasia.

Medical records of dogs and cats with respiratory disease that underwent BAL between 2007 and 2016 at the Ontario Veterinary College were reviewed.

Records from 171 dogs and 34 cats met inclusion criteria. Hemosiderin was reported in BAL cytology from 13/171 (7.6%) dogs and 18/34 (52.9%) cats (P < 0.001). Cats had a higher risk of BAL hemosiderosis than dogs (OR=7.5). Dogs that had a history of cancer were at higher risk for BAL hemosiderosis (OR=3.64). None of the other factors were associated with BAL hemosiderosis, except for prolonged BAL procedure in dogs (P = 0.044). An increased proportion of BAL eosinophils in cats was associated with absence of hemosiderin (P = 0.0463). Hemosiderin was identified in the BAL of 4/72 (5.5%) dogs with pneumonia, 6/56 (10.7%) dogs with bronchitis and 2/9 (22.2%) dogs with neoplasia. One of 16 non‐diagnostic BAL samples in dogs contained hemosiderin. Hemosiderin was identified in the BAL of 4/6 (66.7%) cats with asthma, 7/12 (58.3%) cats with pneumonia, 5/12 (41.6%) cats with bronchitis and 2/4 cats (50%) with neoplasia. Hemosiderin was present in 2/6 (33.3%) dogs and 1/2 cats (50%) that underwent transthoracic fine needle aspiration prior to BAL.

Hemosiderosis is more common in BAL samples from cats than from dogs, and does not appear to be associated with specific respiratory conditions.

Lung Lobe Torsion in 39 Dogs and Cats (2004–2016)

K. Benavides1, E.A. Rozanski2, T. Oura1

1Tufts University, North Grafton, MA, USA, 2Cummings School of Veterinary Medicine, Tufts University, Grafton, MA, USA

Lung lobe torsion (LLT) is an uncommon pulmonary disease where the lung twists around its pedicle, resulting in necrosis of the lobe. The primary objective of this study was to evaluate the characteristics of animals identified with LLT, with a secondary objective to evaluate any prognostic indicators.

The medical record system was searched using for “Lung lobe torsion,” and individual cases were reviewed for confirmation of LLT either at the time of surgical intervention or on necropsy. Data recorded included signalment, large or small breed, imaging modality(ies), presence of pre‐operative pleural effusion, lobe affected and outcome. Fisher's exact test was used to compare groups.

Thirty‐nine cases were identified including 35 dogs and 4 cats. Pugs (n = 10) were most common, with the remaining dogs being small breed (n = 7), or large breed (n = 18). Three cats were DSH and one was a minskin. One small breed dog and one cat were euthanized at the time of diagnosis. Thoracic radiographs were performed in all cases, CT in 19, and bronchoscopy in 5. Lobes affected included right middle (n = 18), left cranial (n = 18), right cranial (n = 2), and accessory (n = 1) Of those treated, 86.5% survived to discharge. All 16 treated pugs, other small breed dogs and cats survived, while only 12/18 large breeds survived. (P = 0.02). Thirty‐one animals had pre‐operative pleural effusion, 26 survived; 6 animals had no pre‐operative effusion, all survived (P = 0.57).

Small breeds and cats with LLT have an excellent prognosis for survival to discharge; the prognosis in large dogs is more guarded.

Fluoroscopic Estimation of Thoracic Dimensional Changes in Healthy Dogs

J. Chan, L. Johnson, C. Brown, R. Pollard

School of Veterinary Medicine, University of California Davis, Davis, CA, USA

There are limited non‐invasive methods available to evaluate lung mechanics. We hypothesized that inspiratory to expiratory change in lung area as assessed by fluoroscopy would result in reproducible measures that could be used for future evaluation of lung function and mechanics in dogs with respiratory disease. Forty‐four conscious, unsedated dogs with no evidence of respiratory disease were minimally restrained in sternal recumbency. Tidal respiration was recorded fluoroscopically for 10 respiratory cycles, and maximal inspiratory and expiratory images from 3 respiratory cycles were evaluated. The number of pixels in the entire thoracic cavity was measured for both inspiration and expiration, and the average percent change in thoracic dimension was determined for each dog. This process was repeated using a hemithorax measurement that excluded the mediastinum and cardiac silhouette. Reference ranges with 95% confidence were computed using a non‐parametric percentile distribution. Median percent change in thoracic dimension for the total thorax measurement was 12.5% (confidence interval of 8.91–23.97%). Median percent change for the hemithorax measurement (20.8%) was significantly different (P < 0.001) with a larger confidence interval of 14.32–37.61%, rendering it more responsive to lung area changes. Neither measurement technique was correlated with age, sex, thoracic conformation, BCS, or breed but varied with body weight above and below 10 kg. In conclusion, fluoroscopy provides a noninvasive and reliable measure of lung area changes during respiration in small and large dogs. Further investigations of lung area changes are required in dogs with respiratory disease to assess its utility in detecting dysfunction.

Exposure to Second‐Hand Smoke in Pets from Different Socioeconomic Backgrounds

N. Smith1, E.A. Rozanski2, D. Welsh1, G. Wolfus3

1Tufts University, North Grafton, MA, USA, 2Cummings School of Veterinary Medicine, Tufts University, Grafton, MA, USA, 3Tufts University, Worcester, MA, USA

Environmental tobacco smoke (ETS) exposure in more common in children from lower socioeconomic classes. ETS exposure has been linked with a variety of health risks. Urine cotinine may be used to estimate ETS exposure. The purpose of this study is to evaluate the rate of ETS exposure in dogs and cats from different socioeconomic classes in Massachusetts.

Urine samples submitted left‐over from routine testing were collected from two university laboratories. Samples from animals living in Worcester (two zip codes), Webster/West Brookfield, Westborough/Walpole, or Wellesley/Weston were eligible for inclusion, representing low income, lower middle income, upper middle income, and high income areas. No further owner information was recorded. Informed consent was waived by the Clinical Sciences Review Committee. Urine was tested for cotinine and classified as positive or negative. Results were compared using Chi‐square analysis.

One hundred and seventy‐eight dogs and cats were enrolled. Seventeen were positive for urine cotinine (9.5%). Worcester (low‐income) had a significantly higher rate of positive animals. (P = 0.00093). Worcester animals had a 21% rate of ETS exposure, Webster/West Brookfield 11%, Westborough/Walpole 1% and Wellesley/Weston 0%.

Dogs and cats, similar to children, have a higher rate of exposure to ETS in lower socioeconomic settings, and may be at risk of more smoke‐associated diseases.

Quantification of Left Atrial Wall Motion in Healthy Horses Using Two Dimensional Speckle Tracking

C. Eberhardt1, K.J. Mitchell2, C.C. Schwarzwald3

1Vetsuisse Faculty, University of Zurich, Zurich, Switzerland, 2University of Zurich, Zurich, Switzerland, 3Clinic for Equine Internal Medicine, Vetsuisse Faculty, University of Zurich, Zurich, Switzerland

The objective of this study was to validate the use of two‐dimensional speckle tracking (2DST) echocardiography for assessment of left atrial (LA) mechanical function in healthy horses. We aimed to establish test‐retest reliability and calculate normal reference intervals.

26 healthy Warmblood horses were included. 2DST analyses of LA wall motion were performed on digitally stored cineloop recordings of a standardized right‐parasternal 4‐chamber view focusing on the left atrium. A P‐wave trigger was used to define the start of a cardiac cycle. Global longitudinal strain (SL), global longitudinal strain rate (SRL) and time to peak left atrial contraction (TPLC) were measured to characterize LA contractile, reservoir and conduit function. Measurements of 10 randomly selected horses were repeated in a blinded fashion by the same observer and by a second observer on different days.

The mean, SD and lower and upper limit of the reference interval were calculated using Reference Value Advisor software (www.biostat.envt.fr). Intra‐observer and inter‐observer measurement variability was quantified using the coefficient of variation (CV).

Reference intervals (mean ± SD, lower and upper limit) were SL (%):−11.82 ± 3.52, −19.20 to −4.44 and SRL (s−1): −1.31 ± 0.43, −2.23 to −0.40 for LA contractile function; SL (%): 15.33 ± 2.94, 9.17 to 21.50 and SRL (s−1): 0.78 ± 0.12, 0.52 to 1.03 for LA reservoir function; and SL (%):−5.29 ± 1.74, −8.95 to −1.62 and SRL (s−1): −1.53 ± 0.23, −2.01 to −1.06 for LA conduit function. Normal TPLC (ms) was 264 ± 23; 215 to 313. Day‐to‐day intra‐observer and inter‐observer variability (CV%), respectively, for LA contractile function were SL: 9.2 and 10.7 and SRL: 7.6 and 14.2; for LA reservoir function were SL: 7.5 and 13.0 and SRL: 10.4 and 6.4; for conduit function were SL: 11.1 and 15.4; SRL: 6.1 and 12.3; and for TPLC were 2.0 and 5.1.

This study shows that 2DST is a feasible and reliable method to quantify left atrial wall motion, with very low to low day‐to‐day intra‐observer and inter‐observer variability. Reference intervals for healthy Warmblood horses are reported. Future studies will focus on left atrial wall motion in horses with cardiac disease to proof the concept of using 2DST for LA wall motion analysis.

Effect of Flecainide on Atrial Fibrillatory Rate in Horses with Induced Atrial Fibrillation

E.Z. Hesselkilde1, H. Carstensen1, M.M. Haugaard1, J. Carlson2, T. Jespersen3, P.G. Platonov2, R. Buhl1

1Department of Large Animal Sciences, Faculty of Health and Medical Sciences, University of Copenhagen, Taastrup, Hovedstaden, Denmark, 2Department of Cardiology, Clinical Sciences, Lund University, Lund, Skane Lan, Sweden, 3Department of Biomedical Sciences, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Hovedstaden, Denmark

Atrial fibrillation (AF) is the most important arrhythmia in equine medicine and traditionally, characteristics of AF have been assessed by intra‐cardiac measurements of atrial fibrillation cycle length. Atrial fibrillatory rate (AFR) retrieved from surface ECGs can however, be used as a biomarker for atrial electrical remodeling. The aim of this study was to study characteristics of induced AF and its modification by flecainide in healthy horses.

Fourteen horses were electrically stimulated into AF by a multipolar catheter placed in the right atrium and were treated with either isotonic saline (n = 7) or flecainide 2 mg/kg (n = 7) IV. A single horse, with spontaneous persistent AF lasting more than three weeks, was also treated with flecainide. ECGs from all 15 horses were analyzed using spatiotemporal cancellation of QRST complexes and calculation of AFR from the residual atrial signal.

At drug administration AFR was 322 ± 30 fibrillations per minute (fpm) in the flecainide‐treated horses and 299 ± 12 fpm in the control group (ns). Flecainide infusion was followed by significant (P < 0.05) AFR decrease that preceded restoration of sinus rhythm in all induced horses (fig. 1A). Time to cardioversion after flecainide infusion start was 4.3 ± 2.5 min. In the control animals, mean AF duration was 205 min (range 152–273 min) and no alteration in AFR was observed following saline administration. The last 30 min before spontaneous cardioversion, a gradual decrease in AFR was observed (P < 0.05) (fig. 1B). In the horse with spontaneous persistent AF, AFR decreased from 450 fpm to 300 fpm following flecainide administration without restoring sinus rhythm.

Flecainide caused a rapid decrease in AFR that resulted in restoration of sinus rhythm while AFR gradually decreased in the horses that spontaneously cardioverted. AFR in the horse with naturally occurring AF was markedly higher which suggest AFR to be a non‐invasive method to study atrial remodeling and the effect of antiarrhythmic compounds.

Reduced Exercise Capacity in Horses Due to Atrial Fibrillation

H. Carstensen1, E.Z. Hesselkilde1, M.F. Fenner1, B.Z. Klein1, S. Pehrson2, T. Jespersen3, R. Buhl1

1Department of Large Animal Sciences, Faculty of Health and Medical Sciences, University of Copenhagen, Taastrup, Hovedstaden, Denmark, 2Department of Cardiology, The Heart Centre, Copenhagen University Hospital, Copenhagen, Hovedstaden, Denmark, 3Department of Biomedical Sciences, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Hovedstaden, Denmark

Atrial fibrillation (AF) is the most common pathological arrhythmia resulting in poor performance in horses and the purpose of this study was to quantify the impact of AF on exercise capacity.

Nine unfit Standardbred mares were included in the study; six AF horses and three time‐matched controls. All horses performed a maximal exercise treadmill test (Test 1) before (day 0) and after (day 57/58) (Test 2) induction of AF by pacemaker stimulation in the right atrium. During the tests all horses were equipped with surface ECG and a venous catheter for blood sampling.

Maximum speed in the AF group was 1.2 ± 0.4 m/s slower in Test 2 compared to Test 1 whereas the control horses ran 0.7 ± 0.6 m/s faster in their second test (P < 0.05). In the AF group speed at HR 200 bpm was 8.6 ± 0.7 m/s in Test 1 opposed to 6.5 ± 0.8 m/s in Test 2 (P < 0.05) and time for the HR to reach 100 bmp after cessation of the test was 20 ± 22 and 65 ± 44 minutes in Test 1 and 2 respectively (P < 0.05). For horses in AF maximum HR reached 311 ± 27 bpm (range 276–346 bpm) during exercise compared to 226 ± 11 bpm (range 207–242 bpm) for the same horses in sinus rhythm (P < 0.05). No significant differences in blood parameters were found between Test 1 and 2.

Exercise capacity was significantly reduced in horses after developing AF characterized by decreased velocity, increased HR at lower speed, increased maximum HR and prolonged cardiac recovery after exercise.

Antiarrhythmic Effects of Flecainide on Experimentally Induced Atrial Fibrillation of Different Duration in Horses

H. Carstensen1, E.Z. Hesselkilde1, M.F. Fenner1, A.V. Loft‐Andersen1, J.K. Kanters2, S. Pehrson3, T. Jespersen2, R. Buhl1

1Department of Large Animal Sciences, Faculty of Health and Medical Sciences, University of Copenhagen, Taastrup, Hovedstaden, Denmark, 2Department of Biomedical Sciences, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen N, Hovedstaden, Denmark, 3Department of Cardiology, The Heart Centre, Copenhagen University Hospital, Copenhagen, Hovedstaden, Denmark

Atrial fibrillation (AF) is the most common clinical arrhythmia in horses. Only a limited number of drugs are available for pharmacological treatment and their use can be a challenge due to adverse reactions. Flecainide has been tested in equine AF with variable efficacy. The purpose of this study was to test whether the efficacy of flecainide is dependent on AF duration.

Six AF horses and three controls were included in the study. The study was performed during a period of 55 days. On day 0 all horses had flecainide (2 mg/kg IV) administered in sinus rhythm for baseline measurements. AF was then induced in the AF group by means of pacemaker stimulation in the right atrium. On days 3, 9, 27 and 55 flecainide was administered again to all horses.

All horses in AF cardioverted to sinus rhythm on day 3 and 9. Five out of six horses cardioverted on day 27 whereas only 2/6 horses cardioverted on day 55. Flecainide infusion prolonged QRS duration from 0.10 ± 0.01 s to 0.13 ± 0.02 s (P < 0.001) and the corrected QT interval from 0.44 ± 0.03 s to 0.49 ± 0.04 s (P < 0.001). Four out of six horses in the AF group experienced ventricular arrhythmias following flecainide infusion on days 9, 27 and/or 55 whereas none of the controls experienced side effects.

Flecainide efficacy decreases with the length of AF duration but could be used in horses with AF of recent onset. The drug can cause ventricular arrhythmias and should be used with caution.

Beat‐to‐Beat Heart Rate Variability During Ridden Exercise in Healthy Horses

P. Egli1, O. Lorello2, C.N. de Solis2, C.C. Schwarzwald3, K.J. Mitchell4

1Vetsuisse Faculty, University of Zurich, Zurich, Switzerland, 2Department of Veterinary Clinical Sciences, University of Bern, Bern, Switzerland, 3Clinic for Equine Internal Medicine, Vetsuisse Faculty, University of Zurich, Zurich, Switzerland, 4University of Zurich, Zurich, Switzerland

The aim was to quantify beat‐to‐beat heart rate variability (HRV) in healthy horses during a standardised ridden exercise test (SRET) and determine if level of fitness influences the HRV.

SRETs were performed in 13 horses used for pleasure riding (Group P) and 23 used for medium level eventing competitions (Group E). Velocity, HR, blood lactate concentration and ECGs were recorded. The ECGs were corrected for artefacts and arrhythmias noted. HRV analysis was performed on the gallop segments and per cent (%) beat‐to‐beat variability was calculated. Performance indices (velocity at blood lactate 4 mmol/L [V4] and velocity at HR 200 min−1 [V200]) were calculated.

Group P had lower V200 and V4 compared with Group E (mean V200 ± SD: Group P: 10 ± 1 m/s, E: 11 ± 0.78 m/s, P = 0.0372; mean V4 ± SD: Group P: 8.3 ± 0.77 m/s, E: 9.7 ± 0.73 m/s, P < 0.0001). Four Group P and 4 Group E horses had arrhythmias during the gallop phase of the exercise test. The % beat to beat variation during the gallop phase was very low (median [1% and 99% percentiles] Group P: 0% [−3.5 to 3.3%], E: 0% [−3.6 to 3.2%]) and there was no difference between groups (P > 0.9999).

Horses used for pleasure riding had lower performance indicators (V200, V4) than horses used for higher level eventing competition. Fitness level did not influence the very low beat‐to‐beat variability seen in this population of horses. Arrhythmias detected in healthy horses during ridden exercise were infrequent and did not influence beat‐to‐beat HRV. This group of ridden horses had similar‐beat to‐beat HRV to that seen in an additional study (using a different data set) during exercise on a high‐speed treadmill.

Corticotropin Releasing Hormone in Foals with Adrenocortical Insufficiency

K. Dembek1, S. Vijan2, B. David3, B. Barr4, T. Burns2, R. Toribio5

1The Ohio State University, Ames, IA, USA, 2The Ohio State University, Columbus, OH, USA, 3Hagyard Equine Medical Institute, Lexington, KY, USA, 4Rood and Riddle Equine Hospital, Lexington, KY, USA, 5Veterinary Clinical Sciences, The Ohio State University, Columbus, OH, USA

Dysfunction of the hypothalamic‐pituitary‐adrenal axis (HPAA), manifested as relative adrenal insufficiency (RAI) or critical illness related corticosteroids insufficiency (CIRCI) has been associated with sepsis and mortality in newborn foals. HPAA dysfunction at the adrenal gland (cortisol) level is well documented in critically ill foals, but information at the hypothalamic (corticotropin releasing hormone [CRH]) and pituitary (adrenocorticotropin [ACTH]) levels is lacking. We hypothesized that in most septic foals the HPAA will be activated at all three levels and the degree of activation will be linked to disease severity and mortality. We proposed that CRH and ACTH deficiency will occur in few critically ill foals and will be associated with disease severity and mortality.

Blood samples were collected on admission from 37 septic (sepsis score > 12), 38 sick non‐septic (SNS), and 18 healthy foals of < 4 days of age. Blood concentrations of CRH, ACTH, and cortisol were determined by radioimmunoassays. CRH and ACTH deficiency was defined as values below the upper level of the 95% confidence interval from values of healthy foals. Hormone values in the normal range in critically ill foals were considered inappropriate.

CRH, ACTH and cortisol concentrations were higher in septic and SNS than healthy foals (P < 0.01). Nonsurviving foals had higher CRH and ACTH concentrations than survivors (P < 0.05). Twenty one percent, 11% and 13% of septic foals had CRH, ACTH and cortisol deficiency, respectively. Septic foals with hypocortisolemia also had decreased CRH and ACTH concentrations but increased CRH/ACTH ratios compared to septic foals with adequate cortisol release (P < 0.01).

All components of the HPAA were activated in most critically ill foals. Some septic foals demonstrated HPAA dysfunction with hypocortisolemia as well as ACTH, and CRH insufficiency. High CRH/ACTH ratio in septic foals with hypocortisolemia suggested the pituitary dysfunction in addition to adrenocortical insufficiency.

Immunohistochemical Expression of Insulin, Glucagon and Somatostatin in Pancreatic Islets from Insulin Resistant Horses

K. Newkirk1, G. Ehrensing1, A. Odoi1, N. Frank2

1College of Veterinary Medicine, University of Tennessee, Knoxville, TN, USA, 2Cummings School of Veterinary Medicine at Tufts University, North Grafton, MA, USA

Hyperinsulinemia and insulin resistance are associated conditions in horses and insulin secretion increases as tissue insulin sensitivity decreases. We hypothesized that insulin‐resistant (IR) horses would have increased insulin staining of pancreatic islets when compared to insulin‐sensitive (IS) horses.

Twenty‐three horses were assigned to IS (n = 13) and IR (n = 10) groups on the basis of frequently‐sampled intravenous glucose tolerance tests and minimal model analysis, or fasted insulin concentrations. Humane euthanasia was performed for reasons other than this study, and pancreas samples were obtained for immunohistochemical analysis. Pancreas tissues were stained for insulin, glucagon and somatostatin and digital images were analyzed to determine expression of each hormone relative to total islet area.

No significant differences in insulin staining were detected between groups and our hypothesis was not supported, but IR horses had a significantly less glucagon within pancreatic islets than IS horses. Median (interquartile range) percentage of total islet area staining positive for glucagon was 12% (5%, 13%) of total islet area in IR horses, compared to 18% (16%, 22%) in IS horses (P = 0.001). Immunoreactive glucagon, insulin, and somatostatin were present in approximately 15%, 63%, and 8% of total islet area, respectively.

It was concluded that insulin expression did not differ between IR and IS groups and there was no evidence that insulin resistance increases insulin secretion within pancreatic islets. Lower glucagon expression in IR horses may be a result of compensatory down‐regulation of hormone secretion in response to hyperinsulinemia or hyperglycemia.

Equine Breed Specific Insulin/Glucose Dynamics During Frequently Sampled Intravenous Glucose Tolerance and Oral Sugar Tests

J.M. Manfredi1, R.J. Geor2, P.S.D. Weber1, L. Jill McCutcheon2, M.E. McCue3

1Michigan State University, East Lansing, MI, USA, 2Massey University, Palmerston North, Auckland, New Zealand, 3University of Minnesota, St Paul, MN, USA

Insulin dysregulation (ID), specifically hyperinsulinemia, has been identified as a cause of laminitis. Certain breeds are seemingly more susceptible. Understanding breed differences in insulin responses and measures of lipid metabolism/adipokines may be critical for identifying truly “at risk” individuals. Insulin‐modified frequently sampled intravenous glucose tolerance tests (FSIGTT) (N = 90) and oral sugar tests (OST) (N = 82) were performed along with assessment of markers for lipid metabolism/adipokines (N = 90) in 5 breeds (Quarter Horses (QH), Arabians, Morgans, Welsh Ponies (WP), and Thoroughbreds). Minimal model analyses of the FSIGTT for insulin sensitivity (SI), the acute insulin response to glucose (AIRg), disposition index (DI), glucose mediated glucose disposal (Sg), the lowest glucose value (Gmin), and the deflection of glucose below baseline (dGB) were assessed. OST insulin thresholds, trajectories, and area under the curve (AUC) were assessed. Statistics included: multilevel regression analysis, regression modeling of time trajectories, one‐way ANOVA, Kruskal‐Wallis test, ROC curve analysis (significant at P < 0.05). QH had higher SI than all other breeds, lower AIRg than WP and Arabians, and higher DI than Morgans. Arabians had higher AIRg than Morgans, and lower Sg than WP. Morgans had lower AIRg than WPs. Arabians, Morgans, and WP had lower Gmins than Thoroughbreds. Morgans had a greater dGB than QHs. Different OST insulin thresholds for ID existed between breeds. OST glucose and insulin trajectories and AUC were lowest in QH. Significant breed differences existed in nonesterified fatty acids, high molecular weight adiponectin, leptin, and triglycerides. Breed specific differences are important to consider when evaluating metabolic health.

Equine Breed Histologic Differences in Adipocytes and Muscle Related to Insulin Dynamics and Body Fat

J.M. Manfredi1, R.J. Geor2, P.S.D. Weber1, M.E. McCue3, L. Jill McCutcheon2

1Michigan State University, East Lansing, MI, USA, 2Massey University, Palmerston North, Auckland, New Zealand, 3University of Minnesota, St Paul, MN, USA

Insulin dysregulation, specifically hyperinsulinemia, has been identified as a cause of laminitis. Muscle and adipose tissue have large roles in insulin regulation and in the pathology of human metabolic syndrome, but these roles have not been well interrogated in horses. Tailhead adipose tissue (N = 76) and middle gluteal muscle (N = 28) biopsies were performed (4 equine breeds: Quarter Horses (QH); Arabians, Morgans, Welsh Ponies (WP)) to relate adipocyte area (AA) and muscle fiber type percent area and proportions to body condition score (BCS), total body fat mass (TBFM) (using the deuterium dilution technique) and minimal model parameters from a frequently sampled intravenous glucose tolerance test (FSIGTT). Statistics included: Kolmogorov‐Smirnov analysis, MANOVA, and Spearman correlation. Significance was set at P < 0.05. Overall BCS was weakly to moderately correlated to insulin sensitivity (SI), the acute insulin response to glucose (AIRg), and AA. TBFM was not correlated to SI and weakly correlated to AIRg and AA overall, but moderately correlated to AA in QH. QH had a significantly smaller AA than Arabians and WP but not Morgans. AA was weakly related to SI and moderately to AIRg. Baseline insulin concentrations were moderately correlated to Type 1 muscle fiber percent area and proportion. No breed differences existed between muscle fiber type area or proportion. There were breed differences in adipocyte, but not muscle histology. The weak correlations between BCS, TBFM and SI and AIRg suggest that adiposity may not be a key factor in determining metabolic fitness in horses.

Shared Genetic Loci Underlying Equine Metabolic Syndrome in Welsh Ponies and Morgan Horses

E. Norton1, N. Schulz1, R.J. Geor2, J. Mickelson3, M.E. McCue1

1University of Minnesota, St Paul, MN, USA, 2Massey University, Palmerston North, Auckland, New Zealand, 3College of Veterinary Medicine, University of Minnesota, St Paul, MN, USA

Equine metabolic syndrome (EMS), a clustering of metabolic disturbances resulting in insulin dysregulation and derangements in fat metabolism, is the most common cause of laminitis. Previous work from our lab has confirmed that the pathophysiology of EMS is complex, with both genetic and environmental factors contributing to the phenotypic variability in metabolic traits. The objective of this project was to identify genomic regions contributing to EMS by performing a genome‐wide association study (GWAS) in a cohort of 232 Welsh ponies and 286 Morgan horses phenotyped for 11 metabolic traits. Genotyping was performed on one of three SNP arrays (54,000, 670,000 to 1,800,000 SNPs); genotype imputation allowed for generation of a uniform set of makers across platforms (˜1.8 million SNPs). GWAS was performed using a mixed linear regression model with sex and age included as covariates. Within breeds, a total of 67 (Welsh ponies) and 116 (Morgans) significant loci were identified for the metabolic traits, of which 4 were shared between breeds. Loci were defined as shared if SNPs with a P‐value < 1.00e‐05 were within a 500 kb window of each other. Specifically, shared loci were identified on ECA10, ECA18, ECA20, and ECA3 for insulin levels post oral sugar test, and baseline non‐esterified fatty acids, triglycerides, and ACTH, respectively, with a total of 75 candidate genes identified. These data confirm that EMS is a polygenic trait with risk loci unique to individual breeds as well as those shared between breeds. Future directions include interrogation of these regions through whole genome sequencing.

Effect of Age and Diet on Insulin and Glucose Dynamics in Horses

S. Jacob1, R.J. Geor2, P. Harris3, M.E. McCue4

1Michigan State University, East Lansing, MI, USA, 2Massey University, Palmerston North, Auckland, New Zealand, 3Waltham Centre for Pet Nutrition, Leicestershire, UK, 4University of Minnesota, St Paul, MN, USA

The relationship between insulin dysregulation, dietary adaptation and aging is poorly understood yet critical to making dietary recommendations for horses at risk of insulin dysregulation and associated laminitis. In this study, the effect of age and diet on insulin and glucose dynamics was examined in 16 healthy Thoroughbred and Standardbred horses, divided into two groups: Adult (8.8 ± 2.9 years; mean ± SD) and Aged (20.6 ± 2.1 years). Four concentrate isocaloric diets: Base (low starch, low sugar), Starch (base plus kibbled corn), Fiber (base plus beet pulp and soybean hulls), and Sugar (base plus dextrose powder) were each fed for seven weeks (balanced four‐way crossover). Horses were group fed low non‐structural carbohydrate hay once daily, and individually fed concentrate meals twice daily. In the seventh week, insulin‐modified frequently sampled intravenous glucose tolerance tests and oral sugar tests (OST) (0.25 mL/kg corn syrup) were performed. Data were analyzed with multivariable linear mixed regression (significance set at P ≤ 0.05). Acute insulin response to glucose (AIRg), OST_peak insulin, insulin area under the curve, and OST_insulin75 minutes were higher in Aged than Adult independent of diet. Aged had higher AIRg and disposition index on Fiber and Sugar, and glucose effectiveness on Sugar than Adults adapted to the same diet. Adults had lower OST_insulinbasal on Fiber compared to Sugar. Aged horses had higher OST_peak insulin than Adults on Fiber. Age and diet had an influence on insulin and glucose dynamics, and should be considered when evaluating a horse's metabolic status.

Fecal Bile Acids, Sterols and Fatty Acid Concentrations in Healthy and Obese Horses

R.G. Madrigal1, M. Coleman2, N. Cohen2, J.S. Suchodolski3, B.C. Guard4, J.B. Honneffer4

1College of Veterinary Medicine, Texas A&M University, Bryan, TX, USA, 2College of Veterinary Medicine, Texas A&M University, College Station, TX, USA, 3Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA, 4Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College Station, TX, USA

Diagnosing and managing horses with metabolic disorders including obesity is challenging due to a poorly understood pathogenesis and lack of well‐characterized risk factors and therapeutic options. In other species, differences in fecal bile acids (BA), fatty acids (FA), and sterols have been recognized in metabolic disease; however, these have not been reported for horses. Thus, the objective of this study was to compare fecal BA, sterols, and FA between obese and non‐obese horses.

Fecal samples were collected from obese horses with a body condition score (BCS) of ≥ 7 and healthy horses with a BCS of ≥ 3 and ≤ 5. Fecal BA, FA, and sterol concentrations were obtained using gas chromatography coupled with mass spectrometry. Variables were assessed for normality and analyzed using t‐test or Wilcoxon rank sum tests to assess differences between groups.

A total of 40 samples were analyzed, including 21 obese horses and 19 healthy horses. Signalment was similar between groups. Seven FA, 9 sterols and 5 BA were analyzed. No statistical difference in bile acids was noted between groups. Results showed statistically significantly lower concentrations of linoleic acid (P = 0.024), oleic acid (P = 0.006), stearic acid (P = 0.009) and gondoic acid (P = 0.021) and increased cholesterol concentration (P = 0.020) in obese horses.

Understanding the role of fecal FA and sterols in horses may improve our understanding of the pathogenesis of metabolic disease and lead to the development diagnostic tests or novel therapeutic agents.

Eosinophilic Peritonitis in 8 Horses (1995–2015)

F.L. Smith1, K.G. Magdesian2, M.B. Whitcomb3

1William R Pritchard Veterinary Medical Teaching Hospital, University of California, Davis, Winters, CA, USA, 2Department of Medicine and Epidemiology, School of Veterinary Medicine, University of California, Davis, Davis, CA, USA, 3Department of Surgical and Radiological Science, School Veterinary Medicine, University of California, Davis, Davis, CA, USA

Eosinophilic peritonitis has previously not been described in horses. The objective of this study was to identify clinicopathologic features, treatment and outcome of horses with eosinophilic peritonitis.

A retrospective study of medical records (1995–2015) at the Veterinary Medical Teaching Hospital, University of California, Davis identified 8 horses with ≥10000/μL total nucleated cell count and ≥1000 eosinophils/μL in abdominal fluid samples.

Arabians were overrepresented (P = 0.01, OR 8.6 [CI: 2.1–34.4]). The most common presenting complaint was colic (6/8). The median peritoneal total nucleated cell and eosinophil counts were 21,040/μL (range 10,000–75,000/μL) and 3786 cells/μL (range 1329–35,343/μL), respectively. Peritoneal fluid total protein concentration was increased (median 4.5, range 2.3–6.7 g/dL). All horses had peripheral eosinophilia (964, range 540–3144 cells/μL). Abdominal ultrasonography revealed gastrointestinal thickening (5/5) and lymphadenopathy (4/5). Microscopic evidence of eosinophilic infiltration of abdominal organs was present (5/6). Quantitative fecal egg counts were negative in 5/6 horses.

Treatment included anthelmintics (8/8), antimicrobials (7/8), corticosteroids (4/8), flunixin meglumine (3/8) and cetirizine (1/8).

All horses were discharged. Two horses were euthanized for colic 1 and 9 months after discharge. Five horses were alive at follow up (2–24 months). Horses treated with corticosteroids with serial abdominocenteses and CBC had improvement of the peritonitis and eosinophilia (3/3). Only 1/3 horses not treated with corticosteroids with serial abdominocenteses had improvement in abdominal counts, and 2/2 did not have improvement of eosinophilia.

Eosinophilic peritonitis is a newly described hypereosinophilic syndrome of horses, which may involve abdominal organ infiltration. Treatment with corticosteroids may result in remission.

The Effects of Metronidazole on the Equine Cecal and Fecal Microbiome and Metabolome

C.E. Arnold1, A. Isaiah2, J. Coverdale3, J.M. Steiner4, J.S. Suchodolski2

1Department of Large Animal Clinical Sciences/Gastrointestinal Laboratory Texas A&M, Bryan, TX, USA, 2Gastrointestinal Laboratory, Texas A&M University, College Station, TX, USA, 3Department of Animal Science, Texas A&M, College Station, TX, USA, 4Gastrointestinal Laboratory, College of Veterinary Medicine, Texas A&M University, College Station, TX, USA

Antimicrobial administration is considered a risk factor for colitis in horses. The subsequent alteration of the resident population of gastrointestinal bacteria may result in dysmetabolism. The purpose of this study was to investigate the effects of metronidazole, a commonly prescribed antibiotic for anaerobic infections, on the cecal and fecal microbiome and metabolome of horses.

Metronidazole (15 mg/kg po q12 h) was administered to adult horses (n = 5) with indwelling cecal cannulas (AUP 2013‐0123). Fecal and cecal samples were taken on days −52, −28, −14, 0, 3 (fecal only), 7, and 14. Illumina sequencing of 16S rRNA genes was performed on extracted DNA. Data were analyzed using Quantitative Insights Into Molecular Ecology (QIIME) and analysis of similarities (ANOSIM) was performed. Univariate analysis of bacterial groups was performed using a Friedman's test and Dunn's post‐test. Linear discriminant analysis effects size (LEfSe) was used to find bacterial taxa associated with study day. Short chain fatty acids (SCFA) and untargeted metabolomics were analyzed using mass spectrometry platforms.

Alpha diversity was significantly decreased between days −28 and 14 for cecal samples (Chao1, P = 0.043; Observed OTUs, P = 0.001; Shannon, P = 0.002) and between days 0 and 3 for fecal samples (Shannon, P = 0.021). Beta diversity, as measured by weighted Unifrac distances, showed significant differences between days −52, −28, −14, 3 and 7 in cecal and fecal samples (ANOSIM, P = 0.001). The LEfSe analysis revealed that several bacterial taxa (Christensenellaceae, Campylobacteriaceae, and Spirochaetaceae) were significantly altered due to antibiotic administration. SCFA were not significantly affected by metronidazole. Of the 553 metabolites analyzed, 223 were named and 330 were unnamed. Analysis of the named fecal metabolites found 100 that were significantly different among study days 0, 3, 7, and 14. Those metabolites that were most statistically significant (P < 0.05, q = 0.041) were active in the biosynthesis of fatty acids, steroids, proteins, carbohydrates, sucrose, and starches. Of the 223 named cecal metabolites, 93 were statistically different on study days 0, 7, and 14 (P < 0.05) but failed to remain significant after adjustment for multiple comparisons (q = 0.091). In conclusion, metronidazole administration changed the cecal and fecal microbiome and metabolome in horses.

Effect of Age and Cortisol on Equine Dendritic Cells after Exposure to Bacteria Ex Vivo

B. Gorham1, D. Hurley1, S. Giguere2, N. Norton1, S. Giancola1, K. Hart1

1University of Georgia, Athens, GA, USA, 2College of Veterinary Medicine, University of Georgia, Athens, GA, USA

Dendritic cells (DCs) are important in the immune response to bacteria, but foal DC function is not well characterized. Monocyte‐derived DCs (MoDCs) were generated from 8 foals at 1, 7 and 30 days‐of‐age and 9 adult horses. MoDCs were incubated for 48 hours with or without killed Escherichia coli or Staphlococcous aureus antigen in the presence or absence of cortisol (15 μg/dL). MoDC expression of MHCII, CD86, and CD14 was measured with flow cytometry, and supernatant cytokine concentrations (IL‐4, IL‐17, IFN‐gamma, IFN‐alpha, and IL‐10) were quantified with a fluorescent bead‐based immunoassay. Effects of age and cortisol on MoDC variables were determined with mixed‐effects linear models (significance P < 0.05). The percentage of cells expressing MHCII and CD86 was lower and CD14 was higher in 1‐day‐old‐foal MoDCs compared to horse MoDCs (P ≤ 0.006). Bacterial exposure alone did not alter surface marker expression in foal or horse cells, but exposure to bacteria with cortisol further reduced MHCII expression in foal cells (P < 0.001). Following bacterial exposure, 1‐day‐old‐foal MoDCs produced significantly less IL‐4, IL‐17, IFN‐gamma, and IFN‐alpha and more IL‐10 than horse MoDCs (P < 0.001). Following bacterial exposure, 1‐day‐old‐foal MoDCs produced significantly less IL‐4, IL‐17, IFN‐gamma, and IFN‐alpha and more IL‐10 than horse MoDCs (P < 0.001). Foal MoDCs exhibit phenotypic and functional immaturity that persists during the first month of life and is compounded by exposure to cortisol. Thus, foal DCs may be less effective at activating lymphocyte responses, which could contribute to foals’ increased susceptibility to bacterial sepsis.

Blood Neutrophil Activation Markers in Equine Asthma Syndrome and Impact of Dexamethasone

N. Herteman1, C. Grimes2, J.‐P. Lavoie3

1University of Zürich, Zürich, Switzerland, 2Faculté de Médecine Vétérinaire – Université de Montréal/Département de Pathologie et Microbiologie, Saint‐Hyacinthe, PQ, Canada, 3Université de Montréal, Equine Internal Medicine Professor, Saint‐Hyacinthe, PQ, Canada

Pulmonary neutrophilia is a characteristic finding in severe equine asthma (SEA or heaves) and is also common in moderate equine asthma (MEA, or inflammatory airway disease). MPXI, NeutX and NeutY are blood neutrophil activation parameters (activity of myeloperoxidase and neutrophils’ size) automatically provided by the ADVIA 120 Hematology Analyzer. The aim of this study was to evaluate these indices in healthy and asthmatic horses, and how corticosteroids can affect them.

We retrospectively studied MPXI, NeutX and NeutY in the blood of 26 with MEA, 14 horses with SEA, and 12 controls. These parameters were also studied in horses with SEA and 5 controls after the administration of dexamethasone (0.06 mg/kg, PO, SID). Blood was collected by jugular venipuncture in EDTA vacutainers and analyzed with ADVIA 120 Hematology Analyzer. Data were compared using Student t‐tests and one way ANOVA.

All three parameters were significantly increased in the blood of horses with SEA when compared to controls and to MEA. Dexamethasone significantly decreased NeutX values in SEA and in controls, but MPXI and NeutY were only significantly reduced in SEA.

This study describes for the first time the use of MPXI, NeutX and NeutY in equine asthma. Our results indicate that neutrophils are more activated in SEA than in MEA and controls. However, the diagnostic value of these parameters for equine asthma is limited due to data's overlapping between the groups. Furthermore, the effects of dexamethasone on MPXI, NeutX and NeutY should be taken into account when monitoring the parameters in horses.

Use of Serum Amyloid A to Differentiate Between Infectious and Non‐Infectious Diseases in Horses

F.M. Wensley1, K. James2, J. Burges3, D. Carrade‐Holt3, N. Pusterla1

1Department of Medicine and Epidemiology, School of Veterinary Medicine, University of California, Davis, Davis, CA, USA, 2UC Davis Graduate Group in Epidemiology, Boston, MA, USA, 3Clinical Diagnostic Laboratories, School of Veterinary Medicine, University of California, Davis, Davis, CA, USA

Stall‐side serum amyloid A (SAA) assays have recently increased in availability resulting in a more widespread use of SAA to identify disease processes requiring further investigation and diagnostic modalities. Treatment decisions are often made on the basis of a high or change in SAA value. The objectives of this study were to improve correlations of serum amyloid A with plasma fibrinogen and neutrophil count, and to assess the use of SAA as a screening tool to identify infectious disease.

A randomized study on 314 equine patients presented to the UC Davis Veterinary Medical Teaching Hospital with the inclusion criteria of a complete blood count (CBC) submission. SAA samples were run on whole blood using the StableLab Lateral Flow Immunoassay. The horses were classified into 5 groups using clinical and diagnostic modalities from the VMTH Medical records, with no reference to the CBC or SAA results to aid in classification. The 5 groups included healthy (n = 36); healthy recently vaccinated (n = 10); inflammatory, non‐infectious (n = 106); inflammatory, infectious (n = 121); and non‐healthy, non‐inflammatory (n = 41). The median SAA value for inflammatory non‐infectious and inflammatory infectious were significantly different at 9.5 mg/mL and 181 mg/mL, respectively. To refine the model, the SAA values were categorized into normal, low, medium and high ranges. SAA was normal for 86% of the healthy group and for 80% of the non‐healthy, non‐inflammatory group. High elevations between 1,000 and 3,000 mg/mL were mainly distributed in the inflammatory infectious group (78%), with none in the healthy group. SAA values in the medium or high range correlated well with abnormalities to the white blood cell count and fibrinogen.

The data set was further statistically analyzed among only the three unhealthy groups. The inflammatory infectious (n = 121) compared with the two groups of non‐infectious (n = 147). When comparing the infectious to the non‐infectious groups, the SAA value was 2.6 times more likely to be within the medium range and 8 times more likely to be within the high range of 1,000–3,000 mg/mL. The sensitivity of SAA alone was 60% with an overall accuracy of 46%. The combination of an elevated SAA, abnormal neutrophil count and hyperfibrinogenemia, further improves the sensitivity (70%) and overall accuracy (59%) of correctly identifying an underlying infectious inflammatory process.

In conclusion, Serum Amyloid A performance held a greater overall accuracy when compared to neutrophil count and fibrinogen, both individually and combined, for defining infectious versus non‐infectious diseases. When used in conjunction with neutrophil count and fibrinogen, SAA further enhances the accuracy of differentiating disease status. The stall side assay was found to be an excellent screening tool however, a normal SAA does not rule out inflammation, nor is a high SAA definitive of infection, as with all hematological parameters, there is high variability among individuals.

Fecal Shedding Prevalence of Equine Coronavirus In Hospitalized Horses

S.Y. Kwon1, J.R. Gold2, J. Evermann3, F. Bain4, M. Sanz5

1Washington State University, Pullman, WA, USA, 2Department of Veterinary Clinical Sciences, Washington State University, Palouse, WA, USA, 3Washington Animal Disease and Diagnostics Laboratory, Washington State University, Pullman, WA, USA, 4Merck Animal Health, Lexington, KY, USA, 5Department of Veterinary Clinical Sciences, Washington State University, Pullman, WA, USA

Coronavirus infection has been recognized in foals and adult horses with pyrogenic and enteric disease but its significance as an etiologic agent remains unclear. Recently Equine Coronavirus (ECoV) has been shown in feces of horses without clinical illness suggesting that inapparent carriers may shed the virus.

The objective of this current, ongoing study is to investigate the presence of ECoV in feces of horses hospitalized for enteric disease (n = 35) or orthopedic conditions (n = 35). Fecal samples are collected at admission and 48–72 h thereafter and the presence of ECoV is evaluated using electron microscopy (EM) and reverse transcription polymerase (RT‐PCR). Categorical analyses will be performed using Kruskal‐Wallis test to determine the association between observations (age, breed, sex, clinical signs, clinicopathological data, EM and RT‐PCR results).

Our preliminary data shows 17% (3/18) of horses with enteric disease were positive for ECoV during hospitalization. These horses had the following clinicopathologic abnormalities; fever (2/3), elevated liver values (1/3), hepatic lipidosis (1/3), and sand enteropathy (1/3). So far, ECoV has not been detected in the orthopedic group (n = 16). Interestingly, a discrepancy between EM and RT‐PCR results has been noted in our ongoing study and is currently being investigated further.

The results of the current study show that ECoV is detected in a small percentage of horses with enteric disease and is not a common finding in horses with orthopedic conditions. Also, inconsistent agreement between EM and RT‐PCR results emphasizes diagnosis of ECoV is confounded by limited sensitivity and specificity data of currently available diagnostic tests.

Investigation of an Experimental Infection Model of Equine Coronavirus in Adult Horses

E.A. Schaefer1, S. Barnum2, C. Harms3, M. Viner3, N. Pusterla2

1University of California at Davis, Davis, CA, USA, 2Department of Medicine and Epidemiology, School of Veterinary Medicine, University of California, Davis, Davis, CA, USA, 3William R. Pritchard Veterinary Medical Teaching Hospital, School of Veterinary Medicine, University of California at Davis, Davis, CA, USA

To investigate the clinical, hematological, molecular and serological parameters of adult horses experimentally infected with ECoV.

Four horses were experimentally infected with frozen feces containing approximately 109 genome equivalents of ECoV/g of feces collected from the intestinal lumen of a horse diagnosed with ECoV at necropsy. Four additional horses were exposed daily to the feces from the experimentally challenged horses. Monitoring included twice daily physical examinations, daily nasal swab, blood, and fecal collection for quantitative molecular detection of ECoV. Blood was collected every other day for hematologic analysis and prior to experimental infection and weekly thereafter for serological analysis.

All eight horses showed evidence of fecal shedding of ECoV by qPCR. Six of the eight horses (75%) exhibited mild, clinical disease of soft‐formed manure; only one horse exhibited transient pyrexia. All horses maintained a total white cell count within normal limits, though 3 horses developed a mild, transient lymphopenia. There were no statistically significant differences (P = 0.2) in quantity of fecal shedding between experimentally infected horses and horses exposed to ECoV‐containing feces.

The experimental infection of adult horses with ECoV was associated with mild and self‐limiting clinical signs, transient lymphopenia, and fecal shedding of ECoV. The experimental infection mimics natural infection, where morbidity rates vary between 20–80%. Although the number of horses in each group was relatively small, there were no differences between experimentally infected horses and horses being exposed to ECoV‐containing feces. The study supports a feco‐oral route of transmission.

Investigation of Shedding Frequency of Respiratory Pathogens in Horses Recently Imported to the United States

F.L. Smith1, N. Pusterla2, J. Watson2, I. Kilcoyne3, C. Sonder4

1William R Pritchard Veterinary Medical Teaching Hospital, University of California, Davis, Winters, CA, USA, 2Department of Medicine and Epidemiology, School of Veterinary Medicine, University of California, Davis, Davis, CA, USA, 3Department of Surgical and Radiological Science, School Veterinary Medicine, University of California, Davis, Davis, CA, USA, 4Center for Equine Health, School of Veterinary Medicine, University of California, Davis, Davis, CA, USA

Imported horses may represent a significant risk for spreading infectious respiratory pathogens into equine populations. They are significantly stressed, and have been comingled with other horses increasing the likelihood of viral recrudescence and novel exposure. The study objective was to investigate the frequency of shedding of respiratory pathogens in recently imported horses.

168 horses undergoing CEM quarantine between October 2014 and June 2016 had a veterinary exam and nasal secretions collected at the time of entry and if they developed acute respiratory signs during quarantine. Samples were assayed for equine influenza virus (EIV), equine herpes virus 1, 2, 4, and 5 (EHV‐1, ‐2, ‐4, ‐5), equine rhinitis virus A and B (ERVs) and Streptococcus equi subspecies equi using qPCR.

EHV‐2 and EHV‐5 were found in 28.6% and 42.2% of nasal swabs, respectively. EHV‐4 was identified in 4 horses and EHV‐1 (N752) was detected in 2 horses. 34 horses were qPCR positive for multiple EHV types. All EHV‐1 qPCR positive horses were symptomatic at sampling. Horses that were qPCR‐positive for EHV‐4, ‐2 and ‐5 were asymptomatic 57.1% 52.8% and 59.0% of the time, respectively. Clinical respiratory abnormalities were seen 49.1% of qPCR negative horses. EIV and ERVs were not detected in any samples; however, 25% of samples failed quality control. S. equi was not recovered from any horse.

Clinical signs were poorly correlated with being qPCR negative or qPCR positive for EHV‐2, ‐4, ‐5. All EHV‐1 positive horses were clinical affected. While EHV‐2 and EHV‐5 were commonly shed, their significance in acute respiratory disease is questionable. EHV‐1 or EHV‐4 were shed by 3.6% of horses, which could have serious consequences if such animals were introduced into a large mixed population of susceptible horses upon entry into the USA.

Does Vaccinating Mares Prior to Parturition Increase Colostral Antibody Transfer to Foals?

J. Haffner1, S. Grubbs2, J. Carlton3, A. Adams4, S. Reedy4

1Middle Tennessee State University, Murfreesboro, TN, USA, 2Boehringer Ingelheim, Stewartsville, MO, USA, 3MTSU Horse Science Center, Murfreesboro, TN, USA, 4Gluck Equine Research Center, Lexington, KY, USA

Foals are immunologically naïve at birth, meaning they have had no exposure to foreign antigens and have therefore not yet mounted any type of protective immune response or accumulated significant levels of immunoglobulins (Ig). Ingestion and absorption of immunoglobulin‐rich colostrum are the sole means of passive transfer in foals providing temporary protection from infection for the first few months of life. Current recommendations for optimizing colostral antibody transfer from mares to foals includes vaccination of the mare 4 to 6 weeks prior to parturition. This practice has been recommended to increase the serum antibody level in the dam; therefore the mare would concentrate a higher level of Ig in colostrum during the final two weeks of gestation. Although this management practice is widely accepted, it has not been adequately substantiated using evidenced‐based medicine from adequately controlled studies. The purpose of this study was to determine if vaccinating pregnant mares approximately 4 to 6 weeks prior to foaling increases the IgG antibody transferred to foals specific for West Nile Virus (WNV) and Equine Influenza Virus (EIV). Twenty‐seven pregnant adult Tennessee Walking Horse (TWH) mares ranging in age from 5 to 23 years were enrolled in the study. All mares were confirmed in foal via rectal palpation and confirmed via ultrasound. Mares were randomized into three treatment groups: G0 naïve control group (no previous vaccinations; no vaccinations administered during the study); G1 vaccinated group (initial 2 vaccination series and revaccinated 4 to 6 weeks prior to foaling); and G2 vaccinated group (initial 2 vaccination series; no pre‐foaling revaccination). Mares in G1 and G2 were administered Vetera® 4XP + WNV (eastern and western encephalitis, equine influenza clades 1 and 2 + WNV) labeled safe for use in pregnant mares. Following parturition, blood was collected and processed from each foal to determine IgG levels (passive immunity) and serum aliquots frozen at −80°C. Serum was shipped to Gluck Equine Research Center, Lexington, KY for measurement of serum hemagglutination‐inhibiting (HI) equine influenza virus antibody levels against Kentucky/14 (Florida clade 1) and Ayrshire/13 (Florida clade 2) reference strains and samples were shipped to Cornell Animal Health Diagnostic Center, Ithaca, NY for analysis of WNV IgG antibody levels measured by ELISA. The mean antibody level was reported accompanied by standard deviation and 95% confidence intervals. Serological titer values below the quantifiable limit were reported as ‘< 10’ and the value of 5 was substituted for these analyses, which provided a conservative estimate. The null hypothesis there was no vaccination effect was tested with a mixed model repeated measures methodology with subject as a random effect. All pairwise comparisons were reported between vaccination groups. Models were run separately for each reference strain. P values < 0.05 were considered statistically significant.

Foals from G1 mares (initial 2 vaccination series and revaccinated 4 to 6 weeks prior to foaling) had significantly higher WNV antibody levels (P = 0.0012) compared to foals from non‐vaccinated G0 mares. G1 foals also had significantly (P = 0.0321) higher ELISA WNV antibody levels compared to foals from G2 mares only administered the initial two vaccination series. There was no significant difference (P = 0.4429) in ELISA WNV antibody levels in foals from G2 mares and foals from unvaccinated G0 mares. Foals from G1 mares had significantly (P = 0.0012) higher KY/14 (EIV Florida clade 1) antibody levels compared to foals from non‐vaccinated G0 mares. Although the arithmetic means of EIV (Florida clade 1) antibody levels were 24% greater in foals from G1 mares compared to foals from G2 mares, there was no significant difference (P = 0.3891). There was a significant (P = 0.0389) difference in KY/14 (EIV Florida clade 1) antibody levels in foals from G2 mares compared to foals from unvaccinated G0 mares. Foals from G1 mares had significantly (P = 0.0012) higher Ayrshire/13 (EIV Florida clade 2) antibody levels compared to foals from non‐vaccinated G0 mares. Although the arithmetic means of Ayrshire/13 (EIV Florida clade 2) antibody levels in G1 mares were 13% greater than foals from G2 mares, no significant difference (P = 0.6896) was identified. There was a significant (P = 0.0430) difference in Ayrshire/13 (EIV Florida clade 2) antibody levels in foals from G2 mares compared to foals from unvaccinated G0 mares.

The foals of mares vaccinated with the initial series and revaccinated 4 to 6 weeks prior to parturition had significantly higher levels of ELISA WNV and EIV (clades 1 and 2) passively transferred antibodies 24 to 72 hours following parturition. The results of this study supports the importance and recommendation to vaccinate broodmares 4 to 6 weeks before foaling not only for their own protection, but most importantly to maximize concentrations of immunoglobulins in their colostrum to be passively transferred to their foals.

Utility of Acute Phase Proteins in Equine Protozoal Myeloencephalitis and Other Equine Nervous System Diseases

N.S. Mittelman1, A.L. Johnson2

1New Bolton Center, University of Pennsylvania, Kennett Square, PA, USA, 2New Bolton Center, School of Veterinary Medicine, University of Pennsylvania, Kennett Square, PA, USA

The purpose of this pilot study was to determine if C‐reactive protein (CRP) or serum amyloid A (SAA) increase with equine protozoal myeloencephalitis (EPM) compared to other neurologic diseases. Accurate antemortem EPM diagnosis requires evidence of intrathecal antibody production (e.g. serum:CSF SnSAG2, 4/3 titer ratio < 100). Some advocate the use of acute phase proteins in addition to serology, which alone results in substantial false positives. Serum and CSF CRP and SAA were measured in 25 cases of equine neurologic disease: EPM (10), cervical vertebral stenotic myelopathy (CVSM) (10), neuroborreliosis (2), equine motor neuron disease (1), degenerative myelopathy (1), and leukoencephalomalacia (1). Nine of 10 EPM cases had serum:CSF SnSAG2, 4/3 titer ratios ≤ 25. The untested EPM case was confirmed postmortem, as were 4 other EPM cases. Nine of the 10 CVSM cases were confirmed postmortem, and the other case had a positive myelogram and SNSAG2, 4/3 titer ratio of 200. The remaining cases were diagnosed on postmortem exam. Serum CRP was not elevated in any case. Serum SAA was only elevated in the two cases of neuroborreliosis (132 mg/L and 460 mg/L; reference 0–20 mg/L) suggesting SAA may aid its diagnosis, but further testing is warranted. Neither serum CRP nor SAA increased in cases of EPM or CVSM. Cerebrospinal fluid CRP and SAA also failed to differentiate cases of EPM (CRP median 3.35 mg/L, range 0.19–13.43 mg/L; SAA median 0.1 mg/L, range < 0.1–2.4 mg/L) from CVSM (CRP median 4.015 mg/L, range 0.16–9.62 mg/L; SAA median 0.62 mg/L, range < 0.1–2.91 mg/L).

Expression and Sequence of Calcium Regulatory Micropeptides in Equine Skeletal Muscle: Implications for Exertional Rhabdomyolysis

K.M. Soave1, M. Schott1, K. Gardner1, C. Finno2, S. Perumbakkam1, S. Valberg1

1Michigan State University, East Lansing, MI, USA, 2University of California, Davis, Davis, CA, USA

Recurrent exertional rhabdomyolysis (RER) affects 5–7% of racing Thoroughbreds, Standardbreds and potentially Quarter Horses, A genetic basis has been proposed but not substantiated. Dysregulation of sarcoplasmic reticulum calcium cycling has been suggested to cause RER. We hypothesized that calcium dysregulation in RER horses could be impacted by genes encoding newly discovered micropeptides such as sarcolipin (SLN), myoregulin (MRLN), DWORF and phospholamban (PLN) that modify the activity of sarcoplasmic reticulum calcium ATPase (SERCA). The purpose of this study was to determine; 1) gene expression of SLN, MRLN, DWORF and PLN in equine muscle, 2) the coding sequence of these micropeptides and 3) potential differences in their coding sequence between controls and RER Thoroughbreds, RER Standardbreds and RER Quarter Horses. Gene expression was determined in gluteal muscle from 11 horses using Next generation sequencing. Sanger sequencing of SLN, MRLN and DWORF was performed on DNA isolated from blood or muscle tissue of a minimum of 9 RER horses per gene and 9 control horses selected from the Neuromuscular Diagnostic Laboratory.

SLN was the predominant regulatory micropeptide in gluteal muscle with several hundred fold higher gene expression than PLB or MRLN. Unlike all other studied mammalian species, equine SLN encoded a truncated protein (29 vs 31 amino acids) missing key regulatory residues (Ser4, Thr5, Cys9, Tyr31) that are proposed to mediate controlled, reversible SERCA inhibition. In contrast, equine MLN had one conservative amino acid substitution, DWORF 2 non‐conservative amino acid substitutions and PLN was 100% orthologous to other species. Non‐synonymous mutations in SLN, MLN or DWORF were not identified in RER horses compared to control horses.

We conclude that SLN is a key regulator of SERCA in equine muscle. Conservation of SLN's unique regulatory peptides in the horse suggests an evolutionary advantage – slightly elevated myoplasmic calcium – which can enhance muscle power output. RER, however, does not appear to arise from distinct mutations in SLN, MRLN or DWORF.

Identification of Candidate Genes for Recurrent Exertional Rhabdomyolysis in Thoroughbreds and Standardbreds

S.K. Beeson1, J. Mickelson1, M. Binns2, S. Blott3, P. Caputo4, C. Isgren5, A. Moore6, R. Piercy7, J. Swinburne8, M. Vaudin9, M.E. McCue10

1College of Veterinary Medicine, University of Minnesota, Saint Paul, MN, USA, 2Equine Analysis Systems, LLC, Midway, KY, USA, 3School of Veterinary Medicine and Science, University of Nottingham, Leicestershire, UK, 4Paul Caputo DVM, Pompano Beach, FL, USA, 5University of Liverpool, Wirral, UK, 6Moore Equine Services, Cambridge, AB, Canada, 7The Royal Veterinary College, Hertfordshire, UK, 8Animal DNA Diagnostics LTD, Cambridge, UK, 9Animal Health Trust, Newmarket, UK, 10University of Minnesota, St Paul, MN, USA

5–10% of Thoroughbred (TB) and 6% of Standardbred (STDB) racehorses suffer from painful cramping and muscle damage following exercise (recurrent exertional rhabdomyolysis [RER]). Gender, temperament, diet, and activity level influence expression of the clinical phenotype, and heritability ranges from 0.34–0.42% in TBs and 0.39–0.45% in STDBs. The objective of this project is to identify variants associated with RER. In both breeds, total phenotypic variance is explained by < 1% of the SNPs tested. These SNPs were further partitioned based on estimated effect sizes. In the TB, SNPs of small, moderate, and large effect account for 45%, 35%, and 20% of the phenotypic variability, respectively. Similarly, in STDBs, small, moderate, and large effect SNPs explain 47%, 30%, and 23% variability, respectively. Genome‐wide association studies (GWAS) identified seven genomic regions of interest (ROIs) significantly associated with RER in TB (n = 491) and STDB (n = 476) horses: two in TBs, four in STDBs, and one shared locus. Of the 274 genes within these ROIs, 34 were computationally predicted to be candidate genes for rhabdomyolysis in humans. Variants in each ROI were identified using whole genome sequences from 10 RER and 10 controls in each breed. 189,610 total variants were discovered in the ROIs, with an average of 7.5 variants per kb. Of these variants, 47.5% were within genes, and 1.1% had putative functional effects. Variants will be prioritized based on segregation between RER and control horses and predicted functional effect. High‐priority variants will be used to create an assay for genotyping in the entire GWAS population.

Foal Survival Score Can Predict Disease Severity in Neonatal Foals

K. Dembek1, J. Minuto2, N. Slovis3, S. Reed4, T. Burns2, R. Toribio5

1The Ohio State University, Ames, IA, USA, 2The Ohio State University, Columbus, OH, USA, 3Hagyard Equine Medical Institute, Lexington, KY, USA, 4Rood and Riddle Equine Hospital, Lexington, KY, USA, 5Veterinary Clinical Sciences, The Ohio State University, Columbus, OH, USA

The sepsis score (SS) has been used to diagnose or assume the presence of sepsis in newborn foals over the last thirty years. However, clinical experience and recent validation studies suggest that the SS is not as accurate to predict sepsis as in the original study. Therefore, a novel scoring method – the foal survival score (FSS) was recently developed to predict likelihood of survival in sick newborn foals. Based on sensitivity and specificity, the FSS has good ability to predict survival; however, the association of the FSS with severity of illness remains to be determined. The lower the FSS the more likely is the foal to die. High SS indicates severe disease. We hypothesized that the FSS will be a good predictor of disease severity (compared to SS) and outcome in hospitalized foals.

Foals (< 4 days of age) were classified into 4 type of disease groups based on clinical and laboratory findings: systemic inflammatory response syndrome [SIRS] (n = 45), sepsis (n = 14), multiple organ dysfunction syndrome [MODS] (n = 37), and healthy foals (n = 65). Multinomial regression models and odds ratios (OR) were calculated to predict each type of disease.

The median FSS was inversely associated with the SS; and foals with MODS, sepsis and SIRS had low FSS compared to healthy foals (P < 0.01). The probability of SIRS, sepsis and MODS decreased by 77%, 84% and 86% with increases in FSS compared to healthy foals, respectively (P < 0.01). Increases in SS were associated with higher likelihood of SIRS (OR = 3.5), sepsis (OR = 4.7) and MODS (OR = 5.7) compared to healthy foals (P < 0.01). Foals that survived had higher FSS and lower SS than non‐survivors (P < 0.01). Likelihood of survival increased with higher FSS (OR = 1.9) and decreased with higher SS (OR = 0.74) in hospitalized foals (P < 0.01).

Our study demonstrated that FSS can be used as a scoring method to determined disease severity in addition to predicting survival in critically ill foals. Future research is needed to establish specific cutoff values for FSS to diagnose each type of neonatal disease.

Pharmacokinetic and Biochemical Profiles After Single Administration of Oral Torsemide in Healthy Horses

G.F. Agne1, S. Jung1, A.A. Wooldridge1, A.K. Johnson1, W. Ravis2, S.H. Duran1

1Department of Clinical Sciences, College of Veterinary Medicine, Auburn University, Auburn, AL, USA, 2Department of Pharmacal Sciences, Auburn University, Auburn, AL, USA

Torsemide is a loop diuretic agent more potent than furosemide. Its oral administration has demonstrated high bioavailability across species. The aim of the study was to determine the pharmacokinetic profiles of oral torsemide in the horse.

Torsemide was administered intragastrically at a single dose of 6 mg/kg to six healthy adult mares. Blood samples were collected at predetermined time points over a period of 48 hours. Plasma torsemide concentrations were measured by the use of high performance liquid chromatography, and serum biochemical profiles were examined. Student's t‐test with Bonferroni correction was used to identify differences in pharmacokinetic and biochemical parameters between baseline (pre) and 24 hour post drug administration (post). Statistical significance was set at P < 0.05.

Pharmacokinetic analysis revealed peak concentration (Cmax) of 11.05 ± 5.26 μg/mL, time of maximum concentration (Tmax) of 2.83 ± 1.33 h, area under the curve (AUC) of 92.2 ± 48.9 μg*h/mL, and an elimination half‐life (T1/2) of 9.11 ± 1.5 h. Significantly increased concentrations of creatinine (pre: median of 1.65 mg/dL and range of 1.5–1.8 mg/dL, post: median of 2 mg/dL and range of 1.9–2.3 mg/dL; P = 0.02) and bicarbonate (pre: median of 22.8 mmol/L and range of 17.6–25.3 mmol/L, post: median of 30.7 mmol/L and range of 27.6–34 mmol/L; P = 0.0007) were noted. Chloride concentrations were markedly decreased (pre: median of 99 mmol/L and range of 96–103 mmol/L, post: median of 87.5 mmol/L and range of 85–88 mmol/L; P = 0.0007).

Oral torsemide successfully reached therapeutic concentrations in blood, and resulted in mild pre‐renal azotemia and electrolyte changes. These results suggest that oral torsemide may be employed as a clinically useful diuretic agent for management of congestive heart failure and fluid retention in the horse.

Pharmacodynamic Properties of Oral Torsemide in Healthy Horses

G.F. Agne1, S. Jung1, A.A. Wooldridge1, W. Ravis1, S.H. Duran1, A.K. Johnson1

1Department of Clinical Sciences, College of Veterinary Medicine, Auburn University, Auburn, AL, USA, 2Department of Pharmacal Sciences, Auburn University, Auburn, AL, USA

Diuretic therapy has been limited to injectable formulations in the horse due to poor absorption of oral furosemide. The purpose of this study was to determine whether oral torsemide produced clinically significant diuresis in healthy horses.

Torsemide was administered orally at 4 mg/kg/day to 6 healthy adult mares for 6 days. A 28 Fr Foley urinary catheter with a urine collection bag was used to measure urine output and urine specific gravity (USG) over 12 hours on days 0 and 6. Blood samples were collected daily for serum biochemical analysis. Non‐invasive mean arterial pressure (MAP) was measured daily with a tail cuff. Student's t‐test with the Bonferroni correction was used to analyze for differences in laboratory parameters between Day 0 (pre) and Day 6 (post).

Urine output (median and range) significantly increased with torsemide [pre: 2.2 L, (1.7–4.4 L); post: 11.5 L (8.7–15.4 L); P = 0.0054]. USG markedly decreased [pre: 1.046 (1.032–1.049); post: 1.010 (1.008–1.011); P = 0.002]. Significant alterations in biochemical parameters included hyponatremia [pre: 137 mmol/L (136–141 mmol/L); post: 129 mmol/L (128–134 mmol/L); P = 0.001], hypokalemia [pre: 3.75 mmol/L (3.5–4.1 mmol/L); post: 1.95 mmol/L (1.6–2.4 mmol/L); P = 0.001], and elevated creatinine [pre:1.4 mg/dL (1.3–1.6 mg/dL); post: 1.95 mg/dL (1.8–2.2 mg/dL); P = 0.006]. MAP significantly decreased after torsemide [pre: 76.5 mmHg (70–87 mmHg); post: 56 mmHg (48–74 mmHg); P = 0.02].

Torsemide administered orally induced clinically significant diuresis with mild pre‐renal azotemia and electrolytes changes. These results suggest that oral torsemide therapy could help overcome current therapeutic limitations for horses with excessive fluid retention.

Diarrhea in Horses Treated for Lower Respiratory Tract Disease with Macrolides Versus Other Antimicrobials (2000–2015)

E. John1, G. Hirsch2, H. Staempfli2

1Atlantic Veterinary College, Charlottetown, PI, Canada, 2Ontario Veterinary College, Guelph, ON, Canada

Macrolide antimicrobials have historically been associated with severe adverse gastrointestinal effects when administered to adult horses. However, they are desirable antimicrobials for use in lower respiratory tract disease, and recent anecdotal results suggest efficacy in adult horses without a high incidence of adverse effects.

The objective of this study was to investigate the association between the development of diarrhea in both adult horses and foals administered macrolide antimicrobials versus other classes of antimicrobials for treatment of lower respiratory tract disease.

One hundred and fifty‐eight horses (96 adults, 62 foals) admitted to a referral centre between 2000 and 2015, diagnosed with pneumonia, pleuropneumonia, or pleuritis, and administered systemic antimicrobials were included in this retrospective study.

Information obtained from the medical records included age at admission, breed, sex, antimicrobials administered while in hospital, clinical diagnosis, clinical outcome, and presence or absence of diarrhea. Macrolide treatment (erythromycin, azithromycin, clarithromycin) both with and without rifampin was compared to treatment with any other combination of antimicrobial(s). The relationship between age, antimicrobial(s) used, and development of diarrhea was investigated. Pearson's Chi‐square and Fisher's exact test were applied (P < 0.05), including odds ratios with 95% confidence intervals.

Macrolide administration to adult horses (P = 0.104; OR 3.4 [0.78, 14.38]) and foals (P = 0.729; OR 1.48 [0.38, 5.74]) showed no statistically significant difference in adverse gastrointestinal effects compared to administration of other classes of antimicrobials. Macrolides are potentially safe to administer to adult horses for the treatment of lower respiratory tract disease.

Pharmacokinetics of Intravenous Lithium Chloride and Agreement Between Two Methods of Lithium Measurement in Horses

L.M. Martin1, A. Bukoski1, D. Whelchel2, T. Evans1, C. Wiedmeyer1, P. Johnson2

1University of Missouri, Columbia, MO, USA, 2College of Veterinary Medicine, University of Missouri, Columbia, MO, USA

Pharmacokinetic evaluation of lithium chloride (LiCl), administered to healthy adult horses as an intravenous (IV) bolus, has not been reported. Nor does there exist a validated rapid procedure for determining plasma lithium (Li+) concentrations in horses in order to monitor therapeutic levels and avoid toxicity. The goals of this study were to determine the IV bolus pharmacokinetic parameters of Li+ for healthy adult horses and compare agreement between two methods of measurement of plasma Li+ concentration: spectrophotometric enzymatic assay (SEA) and inductively coupled plasma mass spectrometry (ICP‐MS; reference method).

Lithium chloride was administered as a 0.15 mmol/kg bwt IV bolus. Lithium concentrations were measured at times −30, −15, 0 (pre LiCl administration), 2.5, 5, 10, 20, 40 min and 1, 2, 4, 8, 12, 24, and 48 h (post LiCl administration) in healthy adult horses (n = 8). Performance outcomes of the SEA were assessed using the following criteria: linearity‐of‐dilution, inter‐day coefficient of variation (CV), and spike‐and‐recovery. The two methods (SEA and ICP‐MS) of plasma Li+ concentration measurement were compared for agreement. Pharmacokinetic parameters were determined based on the reference ICP‐MS data.

The IV LiCl bolus was well tolerated and side effects were not observed. The SEA showed acceptable linearity, R2 = 0.9752; inter‐day CV, 2.5%; and recovery, 96.3%. Both non‐compartmental and compartmental analyses (traditional two‐stage and nonlinear mixed‐effects [NLME] modeling) were performed. Geometric mean values of non‐compartmental parameters were: plasma Li+ concentration at time zero, 2.19 mM; terminal elimination half‐life, 1541 min; area under the plasma concentration‐time curve from time zero to infinity, 550 mM min; clearance, 0.273 mL/min/kg; mean residence time, 1875 min; and volume of distribution at steady state, 511 mL/kg. Results of the traditional two‐stage analysis showed good agreement with the NLME modeling approach. Bland‐Altman analyses demonstrated poor agreement between the SEA and ICP‐MS methods over the therapeutic Li+ concentration range (95% limits of agreement = 0.14 ± 0.13 mM).

Lithium chloride dosed at 0.15 mmol/kg bwt IV bolus was well tolerated by healthy adult horses and displayed pharmacokinetic parameters similar to that reported in other species. The SEA displayed acceptable internal validity but did not agree well with ICP‐MS. The results of this study support further evaluation of LiCl for clinical use.

Ocular Tissue Concentrations and Safety after Subconjunctival Injection of a Voriconazole Thermogel in Horses

M.M. Pereira1, E.M. Abarca2, A.A. Wooldridge1, S.H. Duran1, W. Ravis3

1Department of Clinical Sciences, College of Veterinary Medicine, Auburn University, Auburn, AL, USA, 2Department of Clinical Sciences, College of Veterinary Medicine, Auburn University, Bern, Switzerland, 3Department of Pharmacal Sciences, Auburn University, Auburn, AL, USA

Previous in‐vitro studies using a voriconazole‐poly(lactic‐co‐glycolic acid)‐b‐poly(ethylene glycol)‐b‐poly(lactic‐co‐glycolic acid) thermogel (VORI‐GEL) demonstrated sustained‐release of voriconazole above therapeutic concentrations for most equine ocular fungal pathogens. The objective was to evaluate safety and voriconazole concentrations in ocular tissues and aqueous humor (AH) following subconjunctival injection of the VORI‐GEL in horses.

Six normal horses received topical 1% voriconazole (VORI‐TOP) every 4 hours for 2 days, and then subconjunctival injection of 1.5% VORI‐GEL after washout in alternate eyes. Tears and AH were collected day 2 in VORI‐TOP and days 2, 7, 14, and 23 in VORI‐GEL. To evaluate ocular tissue concentrations, 2 horses received a subconjunctival injection of 1.5% VORI‐GEL 2 days and 2 hrs before euthanasia in alternate eyes. Voriconazole concentrations were measured by high performance liquid chromatography, and evaluated using Student's t‐tests and descriptive statistics. Ocular inflammation was assessed using a modified Hackett‐McDonald scoring system.

Neither VORI‐GEL nor VORI‐TOP caused changes in ocular inflammatory scores. Voriconazole concentrations in tears on day 2 were not different between treatments. Voriconazole concentrations in AH for VORI‐GEL were detectable, but not quantifiable on days 2 and 7. Voriconazole concentrations above 0.5 ug/g were detected in the anterior segment tissues 2 hrs and 2 days after VORI‐GEL injection.

Subconjunctival injection of VORI‐GEL is safe. In normal horses, low AH and tear concentrations are likely due to intact ocular barriers and high drug lipophilicity. There is sustained release of therapeutic concentrations of voriconazole from the VORI‐GEL to the anterior segment tissues for at least 2 days following injection.

Comparison of Respiratory Inductive Plethysmography and Forced Oscillatory Mechanics to Measure Airway Hyperreactivity in Horses

C.E. Dixon1, D. Bedenice, M. Mazan

Tufts Cummings School of Veterinary Medicine, North Grafton, MA, USA

A diagnosis of inflammatory airway disease (IAD) is commonly based on abnormal bronchoalveolar lavage fluid (BALF) cytology or airway hyperreactivity (AHR) in horses with poor performance or chronic cough. Our study purpose was to determine whether two modalities of non‐invasive lung function testing (FOM‐forced oscillatory mechanics vs. RIP‐Respiratory Inductive Plethysmography) establish the same clinical diagnosis of AHR, using histamine bronchoprovocation. AHR was defined by the histamine dose needed to double FOM baseline resistance, or to achieve a 35% increase in RIP delta flow.

Nineteen horses (3–25 years, 335–650 kg) with clinical signs suggestive of IAD were randomly assigned to undergo FOM and RIP testing on consecutive days. BALF was subsequently collected and stained with modified Wright's and toluidine blue stains. Categorical data were compared between groups using Chi‐square analysis.

Abnormal BALF cytology confirmed IAD in 14/19 (73.7%) horses. Both FOM and RIP revealed AHR in 7/14 (50%) of these IAD horses. An additional 4/19 (21.1%) horses showed AHR based solely on RIP testing, including 2 horses with normal BALF cytology. A diagnosis of AHR was more often associated with RIP than FOM testing (P = 0.013), although the prevalence of AHR was significantly higher in IAD vs non‐IAD horses, regardless of testing methodology. The phase angle between thoracic and abdominal components of breathing did not differ between patient groups.

In conclusion, RIP diagnosed AHR more frequently than did FOM, including horses with no other diagnostic evidence of IAD. Without further evaluation, these two testing modalities of AHR cannot be used interchangeably.

Effect of Valacyclovir on EHV‐5 Viral Kinetics in Horses Diagnosed with Equine Multinodular Pulmonary Fibrosis

C.A. Easton‐Jones1, J.E. Madigan2, S. Barnum2, N. Pusterla2

1School of Veterinary Medicine, University of California, Davis, Davis, CA, USA, 2Department of Medicine and Epidemiology, School of Veterinary Medicine, University of California, Davis, Davis, CA, USA

Previous studies have revealed that equine herpesvirus‐ 5 (EHV‐5) is commonly isolated from the lungs of horses diagnosed with equine multinodular pulmonary fibrosis (EMPF), suggesting an etiological link. There are currently no studies assessing the impact of valacyclovir treatment on viral kinetics of EHV‐5 in EMPF cases.

The objective of the study was to determine the effect of valacyclovir treatment on viral load of EHV‐5 measured by quantitiative PCR in blood, nasal swabs and broncho‐alveolar lavage fluid. Four horses with EMPF were treated for 10 days with oral valacyclovir. The diagnosis of EMPF was based on histopathological findings and EHV‐5 detection by qPCR on lung tissue. Sequential blood, nasal swabs and broncho‐alveolar lavage fluid samples were collected to analyze the viral kinetics of EHV‐5 during the 10 days of treatment.

There was no statistical difference in median EHV‐5 viral load between the two time points for all of the three sample types tested. EHV‐5 viral load for blood on day 0 ranged from 5,830 to 18,635 (median 7,738) gB gene copies/million cells, and on day 10 ranged from 2,625 to 18,635 (median 12,138) gB gene copies/million cells. EHV‐5 viral load for nasal secretions on day 0 ranged from 23,639 to 3.394 x109 (median 1.522 x106), and on day 10 ranged from 251,186 to 1.697 x107 (median 452,957). EHV‐5 viral load for broncho‐alveolar lavage fluid on day 0 ranged from 614 to 315,655 (median 33,080) gB gene copies/million cells, and on day 10 ranged from 3,562 to 542,417 (median 115060) gB gene copies/million cells.

Valacyclovir is a relatively expensive drug that is routinely used to treat horses with EMPF. This study revealed that 10 days of valacyclovir treatment did not significantly alter viral kinetics of EHV‐5 as measured by qPCR in nasal swabs, blood and broncho‐alveolar lavage fluid samples.

Relationship Between Tracheobronchoscopic EIPH Score and Bronchoalveolar Lavage Erythrocyte Count in Horses

C. Lopez1, W. Bayly1, J.R. Gold2, M. Sanz1, D. Sellon1

1Department of Veterinary Clinical Sciences, Washington State University, Pullman, WA, USA, 2Department of Veterinary Clinical Sciences, Washington State University, Palouse, WA, USA

Exercise‐induced pulmonary hemorrhage (EIPH) is diagnosed and severity assessed by post‐exercise tracheobronchoscopic examination (TBE), and/or enumeration of erythrocytes (RBCs) in bronchoalveolar lavage fluid (BALF). TBE EIPH scores range from 0 to 4, 0 indicating no EIPH. BALF RBC < 1000/μL indicates no EIPH. There is scant information regarding the relationship of TBE EIPH score to BALF RBC number. We hypothesized that BALF RBC was weakly correlated (r2 < 0.20) with TBE score.

TBE followed by BAL was performed 223 times on 151 different horses after treadmill (8 horses, 45 runs), racetrack (6 horses, 24 runs), or barrel racing (143 horses, 155 runs) exercise. TBE EIPH score ranged from 0 (n = 83) to 4 (n = 5). RBC number ranged from 0 to 994,000/μL. Eleven TBE scores were 0 despite RBC count > 1000/μL. There were 61 exams with TBE scores ≥ 1 but BALF RBC counts < 1000/μL. The relationship between TBE score and BALF RBC number was significant but weak (P < 0.001; r2 = 0.19) after Spearman's evaluation. TBE EIPH score was weakly correlated with BALF RBC number and there were large ranges of BALF RBCs associated with each TBE score. An EIPH score of 0 does not rule out the occurrence of EIPH, but neither does a BALF RBC count < 1000/μL. The latter finding was unexpected and may be explained by fast clearance of the respiratory tract or lavage of an airway segment other than the one bleeding. The possibility of technical errors counting RBCs in some samples cannot be discounted.

Prevalence of Exercise‐Induced Pulmonary Hemorrhage in Barrel Racing Horses in the Pacific Northwest

J.R. Gold1, D.P. Knowles2, T. Coffey3

1Department of Veterinary Clinical Sciences, Washington State University, Palouse, WA, USA, 2Animal Disease Research Unit, USDA‐Pacific West Area, Washington State University, Pullman, WA, USA, 3Center for Interdisciplinary Statistical Education and Research, Washington State University, Pullman, WA, USA

Exercise‐induced pulmonary hemorrhage (EIPH) refers to presence of blood in the airways caused by strenuous exercise. It has been documented in racing, polo and eventing horses, but little is known about EIPH in barrel racing horses. The purpose of this study was to test for the presence of EIPH in barrel racing horses, estimate its prevalence in the Pacific Northwest and determine whether a relationship between EIPH and performance exists.

This study enrolled 158 barrel racing horses competing at events in WA., ID., and MT., after obtained owner consent. Data collected included signalment, illness/respiratory history, race division, and race day medications. Tracheobronchoscopy was scored based on quantity of blood in the trachea (0 = no blood to 4 = large amounts throughout the trachea). Erythrocyte counts were obtained from bronchoalveolar lavage fluid. Statistical analysis included logistical regression, Chi squared, Fisher's Exact T‐test, linear regression and calculation of correlation coefficient. Significance was set at P < 0.05.

The prevalence of EIPH in our study was 54%. Horses that ran faster had a significantly greater likelihood of bleeding P < 0.014. Bleeding did not significantly affect performance, although horses that bled finished lower than horses that didn't bleed. Significant (P < 0.001) positive linear relationship between the tracheal score and BAL erythrocyte count was shown. Correlation coefficient between these 2 tests was poor (r2 = 0.14).

This study shows EIPH is present in barrel racing horses in the Pacific Northwest and may impact performance. The long term effects of EIPH and other variables requires further study.

Prevalence of Cardiac Arrhythmias in Competition Draft Horses

L. Gallant1, S. Jacob1, E.M. Tadros2, J. Woodrow3, M. Hines3, S. Ewart1, H.C. Schott, II1

1Michigan State University, East Lansing, MI, USA, 2College of Veterinary Medicine, Diagnostic Center for Population and Animal Health, Michigan State University, Lansing, MI, USA, 3University of Tennessee, Knoxville, TN, USA

Cardiac arrhythmias, most notably atrial fibrillation, have been linked to poor performance. In a recent study Draft and Warmblood breeds were more likely to be presented with atrial fibrillation in comparison to most light breeds, with the exception of Standardbreds. To our knowledge there is no data describing a population of competition ready Draft horses. Our objective was to determine the prevalence of silent cardiac arrhythmias in a population of Draft horses. A convenience sample of 245 Draft horses was studied at the 2015 Michigan Great Lakes International Draft Horse Show. Information gathered on each horse included signalment, history of poor performance, upper airway noise, myopathy, and cardiac arrhythmias or murmurs. All horses were auscultated on each side of the thorax for 30 seconds. A 30‐second electrocardiographic (ECG) recording was then taken using a handheld recording device that transmitted to a smart phone (AliveCor®). ECG recordings were evaluated for rhythm and heart rate, P‐R interval, and R‐R interval were measured. The study population included 82 Percherons, 69 Belgians, 69 Clydesdales, 11 Belgian mules, 8 Shires, and 5 Percheron mules. Arrhythmias were found in four horses: atrial fibrillation (1), ventricular premature depolarizations (1), second degree atrioventricular block (2). None of these horses had a history of poor performance or known cardiac arrhythmias. We concluded that the prevalence of silent arrhythmias in well‐conditioned, competitive Draft horses is low (2/245; 0.8%).

Beat‐to‐Beat Heart Rate Variability During Treadmill Exercise in Healthy Horses and Horses with Cardiac Disease

L. Frick1, C.C. Schwarzwald2, K.J. Mitchell3

1Vetsuisse Faculty, University of Zurich, Zurich, Switzerland, 2Clinic for Equine Internal Medicine, Vetsuisse Faculty, University of Zurich, Zurich, Switzerland, 3University of Zurich, Zurich, Switzerland

The aim was to quantify beat‐to‐beat heart rate variability (HRV) in healthy horses during a high‐speed standardised treadmill exercise test (HSET) and compare with HRV in horses diagnosed with cardiac disease or other causes of poor performance.

HSET were performed in 30 healthy horses (Group H), 9 horses with cardiac disease (Group C) and 11 horses with non‐cardiac causes of poor performance (Group O). Velocity, heart rate, blood lactate and ECGs recorded. The ECGs were corrected for artefacts and arrhythmias noted. HRV analysis was performed on trot and canter segments and % beat‐to‐beat variability was calculated. Performance indices (velocity at blood lactate 4 mmol/L [V4] and velocity at HR 200 min−1 [V200]) were calculated.

Group C had lower V200 and V4 compared with Group H and Group O (meanV200 ± SD: Group C: 7.6 ± 0.7 m/s, H: 8.8 ± 1.2 m/s, O: 8.7 ± 0.9 m/s, P = 0.0053; meanV4 ± SD: Group C: 6.5 ± 0.8 m/s, H: 7.3 ± 1.0 m/s, O: 7.5 ± 1.0 m/s, P = 0.0383). Seven Group C horses had arrhythmias during exercise, while 4 Group H horses had arrhythmias only during the canter phase. No Group O horses had arrhythmias during exercise. The % beat to beat variation during both the trot and canter phases was low (median [1% and 99% percentiles] Trot Group H: 0% [−4.0 to 3.7%], C: 0% [−5.1 to 3.7%], O: 0% [−3.4 to 3.3%]; Canter Group H: 0% [−5.5 to 5.3%], C: 0% [−4.9 to 4.6%], O: 0% [−5.1 to 4.8%]). However, the minimum and maximum % inter‐beat variation was greater in Group C during the trot phase (minimum % variation Group H: −14.9%, C: −28.1%, O: −9.2%, P = 0.0085; maximum % variation Group H: 19.8%, C: 78.1%, O: 14.9%, P = 0.0014).

Horses with cardiac disease have lower performance indicators (V200, V4) than healthy horses or horses with other causes of poor performance. During HSET, horses have very little beat to beat variation. Infrequent arrhythmias are detected in horses during HSET and do not modify the median beat‐to‐beat HRV but can influence the minimum and maximum beat‐to‐beat HRV indices.

Simpson's Method of Discs in Argentina‐Chair Horse: Echocardiographic Reference‐Intervals for End‐Diastolic and End‐Systolic Left‐Ventricle Volumes

J. Munoz‐Pérez1, G. Wess2, S.E. Linares‐Villalba1, C.E. Giraldo3

1Departamento de Salud Animal, Universidad de Caldas, Manizales, Caldas, Colombia, 2Ludwig‐Maximilians‐Universität München, Munich, Berlin, Germany, 3Universidad de Caldas, Manizales, Caldas, Colombia

Frequently the measurement of ventricular volumes by echocardiography in horses has been performed using parameters obtained through the M‐mode. The Teichholz formula has being the most applied for the calculation of the ventricular volumes and the ejection fraction. The American Society of Echocardiography recently does not recommend its use in clinical settings. Instead, the Simpson's method of discs (SMOD) was recommended in the updated of cardiac chamber quantification guidelines with high levels of evidence. In veterinary medicine, the diagnostic superiority of SMOD versus M‐mode measurements has been similar in different breeds of dogs, some as Doberman pinscher, Boxer, Giant Breeds, and Sighthounds. The aim of this study was established the breed‐specific reference intervals for the End‐diastolic volume (EDV) and End‐systolic volume (ESV) by Biplanar Simpson's method of discs (SMOD) in healthy Argentina Chair horses, hypothesizing that in horses the apparent diagnostic success of SMOD is similar, especially in high performance breeds with sports aptitudes.

In this study, 17 Argentina Chair horses with an appropriate health state confirmed by clinical examination, ECG, echocardiography and laboratory test were recruited. Horses with evidence of systemic disease or cardiovascular alterations were excluded. Horses were evaluated without sedation and 2D and M‐mode measurements were performed in the right para‐sternal and left para‐sternal views. The ventricular volumes obtained by SMOD and M‐mode were indexed to body surface area (BSA). Reference intervals for SMOD with and without weight normalization were established using the right para‐sternal and left para‐sternal views. The volumes obtained in each view were compared, as well as between genders and ages using a paired t‐test when normal distribution was apparent. The volumes obtained by SMOD and M‐mode were compared using a T‐test. Reference intervals were reported with 95% confidentiality. A P‐value < 0.05 was considered statistically significant.

All indexed volumes were of normal distribution according to Kolmogorov‐Smirnov test (P = 0.50). The volumes obtained with the SMOD in both views demonstrated modest agreement in the student T‐test (P < 0.05). The breed‐specific references intervals were 98.93 mL/m2 and 47.67 mL/m2 for indexed end‐diastolic volume (EDVI) and indexed end‐systolic volume (ESVI), respectively. There were no difference between the indexed volumes obtained in mares and males (P > 0.05). As well as between the age groups (P = 0.07). However, there was difference between the volumes obtained with SMOD method and One‐dimensional M‐mode (P = 0.03). The volumes derived from M‐mode had significant difference between left and right para‐sternal views (P = 0.001).

In conclusion, this study provided echocardiographic reference intervals for SMOD‐derived left ventricular volumes for these breed with aptitude for equestrian sports. Because some of these sports are under stressful activity, the cardiovascular effort of the athlete horse is greater. The indexed‐volumes provided with this study may possibly have the ability to recognize early, horses with a diminished athletic performance associated to cardiovascular comorbidities.

Equine Pituitary Pars Intermedia Dysfunction (PPID) in the Southern Hemisphere, What is Different?

F.‐R. Bertin

The University of Queensland, Gatton, Qld, Australia

PPID is a well characterized endocrine disorder in the temperate climates of the Northern Hemisphere; however, its description in tropical and subtropical climates of the Southern Hemisphere is scarce. This analysis reviews the cases of PPID reported in the Southern Hemisphere and describes cases from The University of Queensland that reflect a different presentation. Clinical features of the disease and variables associated with laminitis, insulin dysregulation and survival are reported.

Ninety‐nine cases including 34 ponies and 65 horses were diagnosed with PPID with either a basal ACTH (91%) or a dexamethasone suppression test (9%).

The most common clinical signs were tachypnoea (67%), anhidrosis (55%), hypertrichosis/hirsutism (40%), weight loss (38%), polyuria/polydipsia (35%) and lameness (34%). The most common clinicopathologic abnormality was hypophosphataemia (36%). When investigated, laminitis was diagnosed in 82% of cases, insulin dysregulation in 35% and gastrointestinal parasitism in 100%.

Pergolide was administered in 83% of cases at a dose of 0.001 [0.001–0.005] mg/kg increased to 0.002 [0.001–0.015] mg/kg over a period of 1 [0–48] months. Eighty percent showed a clinical improvement and 65% survived.

Factors associated with laminitis were younger age (P = 0.007) and absence of weight loss (P = 0.030). Factors associated with insulin dysregulation were younger age (P = 0.026), duration of clinical signs (P = 0.044) and need for higher doses of pergolide (P = 0.017). The only factor associated with survival was clinical improvement (P = 0.007).

This study shows that commonly described features of PPID are found in the Southern Hemisphere but that more unusual findings also warrant investigation of endocrine disorders.

Hair Cortisol Concentration as a Stress Biomarker in Horses

M.C.D. Graeff1, D. Janz2, C. Waldner1, J. Campbell1, F. Marques1

1WCVM, University of Saskatchewan, Saskatoon, SK, Canada, 2WCVM, Toxicology Centre, University of Saskatchewan, Saskatoon, SK, Canada

Cortisol is commonly used as an indicator of physiological stress. Traditional samples (e.g., blood, urine, saliva) to measure cortisol secretion have proven to adequately assess short term/acute response to stress, while hair cortisol concentration (HCC) is increasingly being used to assess long‐term stress in mammals. The purpose of the study was to analyze HCC in whole hair and hair segments to investigate potential effects of body location and surgical procedure (castration) on HCC in horses.

Mane and tail hair from healthy broodmares (n = 10) was collected, segmented and analyzed. Additionally, yearling male (n = 14) and female horses (n = 14) were used. Separate mane and tail hair samples were collected at the beginning of the study (before castration) and 3 months after from male and female yearlings. Cortisol was extracted into methanol from ground hair and quantified using ELISA. HCC in mane samples from broodmares was greater than in tail samples. In all studied groups, segments further away from the hair root contained lower HCC. The decrease in hair cortisol concentrations between 1 and 2‐months post‐castration was significantly lower in castrated males than intact females. Cortisol accumulation in the month after castration could explain the significantly smaller decrease in HCC concentrations in males than females during the same time period. Results suggest that HCC analysis in horses can be used to address basal and elevated HPA activity in horses. Nevertheless, other factors such as segment location and collection site may also have an effect on measured HCC.

The Association Between Endocrine Disrupting Chemicals and Equine Metabolic Syndrome

S. Durward‐Akhurst1, N. Schulz2, E. Norton2, R.J. Geor3, J. Mickelson4, M.E. McCue2

1University of Minnesota, Center City, MN, USA, 2University of Minnesota, St. Paul, MN, USA, 3Massey University, Palmerston North, Auckland, New Zealand, 4College of Veterinary Medicine, University of Minnesota, St Paul, MN, USA

Our group has demonstrated that up to 49% of Equine Metabolic Syndrome (EMS) phenotypic variability is explained by shared environment; however only 4–18% of this environmental variability is explained by diet, exercise and season, suggesting other environmental factors play a role in EMS. Endocrine Disrupting Chemicals (EDCs) are associated with metabolic syndrome and other endocrine abnormalities in humans, and our preliminary data demonstrated that horses from farms < 31 miles of EDC disposal sites (SF) were more likely to have had laminitis and had higher post oral‐sugar‐challenge insulin concentrations (INS‐OST). These data suggested that EDC exposure is an EMS risk factor.

We have previously reported the measurement of the concentration of EDCs capable of activating the aryl hydrocarbon receptor via a bioassay (AHR‐EDC); and the correlations between these EDCs and metabolic measurements in 301 horses from 32 farms. Here we extend these findings by presenting correlations between EDCs capable of activating the estrogen receptor (ER‐EDC). Significance was P < 0.05. Mean (range) ER‐EDC concentrations were 279.66 (4.35–15,000) pg/mL plasma. EDCs were below the detection limit in approximately half of the horses. AICc statistics were used to determine the best linear multivariable regression model for EDC association with 9 EMS phenotypic variables. ER‐EDC was associated with glucose after an oral sugar test (OST), glucose, insulin and insulin‐OST. Overall, these results suggest that some of the unexplained environmental variance in individuals with EMS is due to EDC exposure, which is mediated in part through the estrogen and arylhydrocarbon receptors.

Clinical Signs Associated with PPID in the Equine Athlete

M. Baus1, J. Kremburg1, K. Kirchherr2, S. Grubbs3, D. Neal4, T. Keefe5

1Grand Prix Equine, Hawleyville, CT, USA, 2Boehringer Ingelheim, Belchertown, MA, USA, 3Boehringer Ingelheim, Stewartsville, MO, USA, 4Boehringer Ingelheim, Platte City, MO, USA, 5Enviro Stats, Windsor, CO, USA

Pituitary Pars Intermedia Dysfunction (PPID) has been described as one of the most common diseases of horses and ponies 15 years of age and older. Recently, the clinical signs of PPID have been divided into early and advanced clinical signs. Establishing a diagnosis of PPID in horses with early clinical signs is currently a difficult challenge facing equine veterinarians. Particularly difficult may be the diagnosis of horses with PPID in the equine athlete. Many of the same clinical signs identified in early or advanced PPID may be recognized in the sport horse along with tendon or suspensory ligament degeneration. Suspensory ligament injuries have been considered a common cause of lameness, in the equine athlete involved in competitive events. A recent histopathological study concluded that an association exists between PPID and suspensory ligament (SL) degeneration. The objective of this study was to identify the most common clinical signs associated with PPID in the sport horse.

Sport horses >10 years of age, any breed, and sex were eligible for study enrollment as long as they were documented to be exhibiting one or more of the early or advanced clinical signs of PPID including suspensory ligament desmitis. Forty‐nine horses were evaluated and included in the final data analysis with at least one clinical sign of early or advanced PPID. Demographic data, signalment, and a physical examination was conducted and each horse was tested for PPID using the TRH stimulation test measuring ACTH at 0 (T0ACTH) and 10 (T10ACTH) min. Insulin and glucose levels were also determined. Normal horses were excluded from the study. Blood samples were shipped overnight to the Animal Health Diagnostic Center, Cornell University, Ithaca, NY for analysis. The association between PPID status, based on ACTH and insulin results, and each of the demographic variables (age, sex, and breed), clinical signs, the two test result variables insulin and glucose were statistically evaluated individually using the Pearson chi‐square test. Odds ratios for significant predictors of PPID status were computed using corresponding 95% confidence intervals when applying multiple logistic regression analysis.

Of the 49 horses, 19 (39%) were PPID+ and 8 (16%) were IR+. Of the 19 PPID+ horses, only 2 (11%) were PPID+ at the T0ACTH time point, whereas 19/19 (100%) were PPID+ at T10ACTH. Six (32%) of 19 PPID+ horses were hyperinsulinemic. Horses’ ages ranged from 11 to 25 years of age (arithmetic mean 17 years) in the PPID+ group. The most common clinical signs observed in the PPID+ horses were delayed regional shedding, loss of epaxial muscle mass, regional adiposity, skeletal muscle atrophy and suspensory desmitis. PPID+ was significantly (P = 0.023) associated with lameness (suspensory desmitis, tendon laxity, superficial digital flexor tendonitis). Of the horses that were lame, 70% were PPID+.

Based on the results of this study, the TRH stimulation procedure was required in 89% of enrolled horses for laboratory confirmation of PPID. In the sport horse, suspensory desmitis was significantly associated with PPID+ status. Veterinarians should include PPID in the list of differential diagnoses when examining sport horses with suspensory desmitis along with early and advanced clinical signs of PPID.

Assessment of Prolactin Concentration as a Screening Test for PPID

H.C. Schott, II1, B. Norby1, D. Thompson2

1Michigan State University, East Lansing, MI, USA, 2Louisiana State University, Baton Rouge, LA, USA

To assess measurement of prolactin (PRL) concentration as a screening test for PPID, serum PRL concentration was compared between 31 horses that were enrolled in and 26 horses that failed inclusion criteria for the clinical efficacy study for pergolide. Enrollment required horses to have a hypertrichosis score ≥1 and either an elevated plasma ACTH concentration (≥ 50 pg/mL) or abnormal overnight dexamethasone suppression test results. PRL was measured again after 3 and 6 months of treatment with pergolide in the horses enrolled in the study. PRL concentrations were compared by rank sum analysis and an ANOVA on ranks for repeated measures. A ROC curve was constructed to assess sensitivity and specificity of PRL at various cut‐off values as a screening test for PPID. PRL was greater (P = 0.02) in enrolled (7.0 ± 8.2 ng/mL, mean ± SD) as compared to excluded (4.3 ± 2.6 ng/mL) horses. Area under the ROC curve was 0.68 (95% CI: 0.55–0.80). Using cut‐off values of ≥3.5 and ≥4.0 ng/mL, sensitivity and specificity were 77 and 54% and 65 and 54%, respectively. Curiously, PRL tended to decrease after 3 months of pergolide treatment (3.8 ± 1.8 ng/mL) but the only significant difference was a higher (P < 0.01) PRL after 180 days of pergolide treatment (9.7 ± 6.4 ng/mL) as compared to the 90 day treatment value. In this cohort of horses, measurement of PRL did not appear to be a useful screening tool for PPID. Further, treatment of PPID‐affected horses with pergolide did not decrease PRL.

Adrenocorticotropic Hormone Levels in Equine Plasma After Centrifugation at Multiple Time‐Points From Collection

K. Shepard1, J. Haffner2, S. Grubbs3, D. Neal4, G. Pierce5

1Boehringer Ingelheim, St Joseph, MO, USA, 2Middle Tennessee State University, Murfreesboro, TN, USA, 3Boehringer Ingelheim, Stewartsville, MO, USA, 4Boehringer Ingelheim, Platte City, MO, USA, 5Innovative Data Resources, Asheville, NC, USA

Pituitary pars intermedia dysfunction (PPID) has been considered the most common endocrinologic disorder of aged horses. The diagnosis of PPID is typical based on history, clinical signs and demonstration of endocrinologic dysfunction via diagnostic assay results. The most common diagnostic test utilized by veterinary practitioners for the diagnosis of PPID is currently measurement of basal plasma adrenocorticotropic hormone (ACTH) concentration. Determination of resting plasma ACTH concentration has been the most practical diagnostic test for PPID likely due to the ease of collection and the availability of seasonal reference ranges that allow for testing any time of the year. Multiple recommendations exist concerning the in vitro stability of ACTH. Most all recommendations state samples should be chilled promptly after collection. Multiple recommendations exist for the timing of centrifugation of chilled samples from sample collection. In the US, a discordance exists between laboratory recommendations concerning centrifugation time from sample collection; anywhere from 2 hours up to 4 hours from sample collection. Realistically, many equine ambulatory practitioners do not have the ability to centrifuge the sample within 2 to 4 hours. Based on this dilemma, the purpose of this study was to determine the stability of ACTH levels in blood samples with increasing duration from collection prior to centrifugation.

Ten horses, 5 PPID+ and 5 PPID− (non‐PPID) horses were enrolled in the study during the autumn time‐period. On Day 0, 5 mL of whole blood was collected from each horse into each of six EDTA tubes and immediately chilled. One tube was centrifuged within 15 minutes of collection (time 0) followed by centrifugation of one tube from each horse at 4, 8, 12, 24 and 36 hours following collection. At each time‐point, plasma was centrifuged and separated into 1.5 mL polypropylene tubes and stored at −80°C. Plasma was shipped frozen with cold packs overnight to the Animal Health Diagnostic Center, Cornell University, Ithaca, NY for analysis. The percent change from baseline (PCFB) was reported to standardize the data due to differing baseline values. Mean PCFB, standard deviation and 95% confidence intervals (CI) were reported. The null hypothesis that there was no time effect and was tested with a mixed model repeated measures methodology with subject as a random effect. P‐values < 0.05 were considered statistically significant.

The absolute changes over time revealed no pattern of change over time. The mean PFCB was 2.8% ± 7.96% (CI −2.88%, 6.99%). There was not significant evidence (P‐value of 0.5056) of a time effect using the repeated measures model. There was no evidence of any significant time effect on the level of ACTH in PPID+ or PPID− (non‐PPID) horses. There did not appear to be any systematic degradation of ACTH level due to delay in measurement through 36 hours post collection. The results of this study supports the recommendation that immediately chilled EDTA blood samples (properly stored) may be centrifuged and plasma separated within 36 hours of collection for shipment to the respective laboratory for analysis without degradation of the ACTH level. Further research is needed to determine ACTH stability in a larger number of horses (PPID+ and PPID−) from separate sites and different geographic regions.

Effects of Dexamethasone on Circulating ACTH and Cortisol Concentrations in Healthy and PPID‐Affected Horses

D. Whelchel1, P. Johnson1, A. Trimble2, L. Schultz1

1College of Veterinary Medicine, University of Missouri, Columbia, MO, USA, 2College of Veterinary Medicine, Kansas State University, Manhattan, KS, USA

The purpose of this study was to investigate effects of systemic corticosteroid administration on circulating ACTH and cortisol concentrations in healthy horses and horses with Pituitary Pars Intermedia Dysfunction (PPID). The hypothesis of this study was that administration of synthetic corticosteroids would decrease circulating ACTH and cortisol concentrations in healthy horses while significant changes would not be detected in horses with PPID.

Seventeen horses meeting the study's inclusion criteria were tested for PPID, using endogenous ACTH plasma concentrations and the thyrotropin releasing hormone (TRH) stimulation test. The study included six healthy horses (group 1: plasma ACTH concentrations < 110 pg/mL post TRH administration) and 6 horses with PPID (group 2: plasma ACTH concentrations above seasonal reference ranges or >110 pg/mL post TRH administration). All 12 subjects were stabled in box stalls with free choice access to grass hay and water for the duration of the study. The study was conducted in accordance of the University of Missouri's Animal Care and Use Committee and between February to July. After a 48‐hour acclimation period, blood samples were collected to measure baseline ACTH and cortisol concentrations in all horses. Dexamethasone (0.05 mg/kg) was subsequently administered IV once. For the next 48 hours, ACTH and cortisol were measured every 6 hours. Repeated measures ANOVA analysis followed by post hoc testing determined significant effects of dexamethasone treatment on ACTH and cortisol concentrations over time in both groups of horses and compared differences between groups.

Significant changes in circulating ACTH (P < 0.0017) and cortisol concentrations (P < 0.0001) were detected at numerous time points after dexamethasone treatment in both groups of horses; however, significant differences between groups were not evident. In conclusion, dexamethasone administration appeared to suppress plasma ACTH concentrations for up to 36 hours and serum cortisol concentrations for at least 48 hours in healthy horses and horses affected by PPID. Studies with larger sample sizes may better distinguish differences in the response of the endocrine axis to dexamethasone treatment in healthy horses compared to horses with PPID.

Evaluation of Dexamethasone to Prevent Post‐Operative Ileus in 61 Horses with and without Intestinal Resection

K.F. McGovern1, B.M. Bladon2

1Donnington Grove Veterinary Group, Hermitage, UK, 2Donnington Grove Veterinary Group, Newbury, UK

Inflammation is key in the development of post‐operative ileus (POI) in rodents, with a similar pathogenesis likely occurring in other species. It is well established that dexamethasone reduces inflammation and therefore may reduce POI in horses. A retrospective study was performed to determine if dexamethasone reduces POI in horses with small intestinal (SI) disease, and to assess effects on incisional health and short‐term survival. Data from all horses that underwent SI colic surgery and received 0.1 mg/kg dexamethasone intravenously during surgery (DEX) was extracted from clinical records (61 horses). Thirty‐two horses underwent (SI) resection and 29 did not. Data from 61 horses that did not receive dexamethasone (NoDEX) was collected sequentially from clinic records for comparison. Horses were matched for the type of resection performed (if any). Fisher's Exact and Student's t tests were used for data analysis. Fewer horses produced gastric reflux when dexamethasone was administered (DEX 13/61; NoDEX 21/61, P = 0.1570). Nasogastric reflux produced (litres) was lower when horses received dexamethasone, (DEX 12.06 + /−35.73, NoDEX 16.85 + /−38.90, P = 0.4841). Differences were not statistically significant. There was no difference in survival to discharge (53/61 DEX, 52/61 NoDEX, P = 1.0000). Fewer surgical incisions became infected in the DEX group with no statistical significant difference (6/61 DEX, 8/61 NoDEX, P = 0.7774). Dexamethasone may have a beneficial effect on the incidence of POI but no statistically significant differences were detected. A single dose of dexamethasone does not have a detrimental effect on short‐term survival or increase the risk of incisional infections. A randomized controlled trial is required.

Acid‐base Disorders in Diarrheic Calves and Horses: Comparison Between the Traditional and Physicochemical Approaches

C. Faso1, J.S. Weese2, W. Sears3, H. Staempfli4, D. Gomez‐Nieto2

1Department of Clinical Studies, University of Guelph, Guelph, ON, Canada, 2Department of Pathobiology, Ontario Veterinary College, University of Guelph, Guelph, ON, Canada, 3Department of Population Medicine, University of Guelph, Guelph, ON, Canada, 4Ontario Veterinary College, Guelph, ON, Canada

The objectives of this study were to evaluate the capacity of 3 different approaches (Henderson‐Hasselbalch (HH), base excess (BE) and strong ion difference (sSID)) to identify acid‐base (AB) disorders in diarrheic calves and horses with normal venous pH and pCO2, and to describe the sSID variables in a subgroup of diarrheic calves and horses without evidence of any AB disorder according to the traditional approach (normal venous pH, pCO2 and BE).

Medical records were reviewed for all calves and horses presented to a teaching hospital between 2010 and 2016 (calves) and 2010 to 2012 (horses). Data of pCO2 (mmHg), pH, and plasma [Na+] (mM/L), [K+] (mM/L), [Cl−] (mM/L), [L‐lactate−] (mM/L), and total protein [TP] (g/L) were recorded. All venous blood gases, electrolytes, and pH measurements were performed using a Radiometer 800 Flex machine. TP was measured as total solids by refractometry. By the HH, BE and sSID approaches, AB disorders in calves were defined when any of the respective variables were outside of the reference ranges: [HCO3] (20 to 30 mM/L), anion gap (AG) (< 20 mM/L), BE (−2 to 6 mM/L), Strong ion difference (SID4) (38 to 48 mM/L), Total plasma concentration of non‐volatile weak acids (Atot) (15 to 22 mM/L) and unmeasured strong ions (USI) (−2 to 3 mM/L). sSID variables for calves were calculated as SID4= (Na++K+)‐(Cl−+L‐Lactate−); Atot= 0.343 × TP(g/L) and USI=Atot/(1 + 10(7.08−pH)) – SID4 – HCO3.AB disorders in horses were defined when any of the following variables were outside of the following reference ranges: BE (1 to 7 mM/L), [HCO3] (24 to 34 mM/L), AG (< 16 mM/L); SID4 (38 to 47), Atot (11 to 16 mM/L) and USI (−2 to 3 mM/L). sSID variables for horses were calculated as SID4= (Na++K+)‐(Cl−+L‐Lactate−); Atot=0.22 × TP(g/L) and USI= Atot/1 + 10(6.65‐pH) – SID4 – HCO3−. McNemar's test was used to evaluate whether pairs of approaches are equally likely in finding an AB disorder. The level of agreement between pairs of approaches in detecting AB disorders was assessed using Kappa coefficient test.

For the first part of the study, 37 calves and 50 horses hospitalized for diarrhea with normal venous pH (mean ± SD; 7.39 ± 0.03 and 7.41 ± 0.03, respectively) and pvCO2 (41 ± 4.5 mmHg and 42 ± 4.5 mmHg, respectively) were included. In these groups, HH approach detected AB disorders in 11 (30%) calves and 15 (30%) horses, BE in 9 (25%) calves and 20 (40%) horses, and sSID in 25 (68%) calves and 33 (66%) horses. McNemar's test demonstrated that HH/BE are equally likely to detect AB disorders in horses and calves, but BE/sSID were not. Kappa agreement test revealed a moderate agreement between HH/BE, and a fair agreement between HH/sSID and BE/sSID, for detection of AB disorders in horses (Table 1.)

For the second part of the study, a subgroup of 27 calves and 30 horses with normal venous pH, pvCO2, and BE was identified. 47% (17/27) of those calves had sSID AB disorders: USI acidosis (6/27), Atot acidosis (4/27), Atot alkalosis (4/27), SID4 acidosis (1/27), SID4 and USI acidosis (1/27), USI acidosis and Atot acidosis (1/27). 47% (14/30) of those horses had sSID AB disorders: SID acidosis (6/14), SID acidosis and Atot alkalosis (5/14), USI acidosis (1/14), USI acidosis and Atot alkalosis (1/14), and USI acidosis and Atot alkalosis (1/14).

Diarrheic calves with apparently normal AB state had underlying mixed metabolic disorders characterized by the acidifying effects of low SID4, USI and total plasma proteins, whereas underlying mixed metabolic AB disorders in horses were characterized by the acidifying effect of low SID4, with concurrent alkalinizing effect of hypoproteinemia. This results suggest that, in addition to venous pH, PvCO2, HCO3 and BE, clinicians should investigate the quantitative effect of the SID4, Atot, and USI on the AB status of diarrheic calves and horses.

Radiographic Evaluation of Compounded and Illegal Over‐the‐Counter Omeprazole Products

M. Wallace

Carolina Equine Hospital, Browns Summit, NC, USA

Compounded and illegal omeprazole paste products often have production quality control problems as documented by incomplete filling of syringes, air pockets and variations in homogeneity. There is one FDA‐approved omeprazole product (GastroGard®) for the treatment of gastric ulcers in horses. Lack of FDA‐oversight can lead to undetermined and unregulated production quality. The objective of this study was to compare the production quality of compounded and illegally manufactured omeprazole products compared to GastroGard® paste.

Syringes of compounded omeprazole paste from 5 pharmacies, 8 illegally manufactured over‐the‐counter omeprazole products and 2 syringes of GastroGard® were acquired. Syringes were radiographed using digital radiographic equipment, labelled to ensure specimen identity in the film. Syringe fill, air pocket presence and product consistency were evaluated. Results were reported using qualitative descriptors for levels of syringe fill and product consistency.

All compounded and over the counter products had incomplete fill and air pockets clearly visible. Eight showed variability in homogeneity. GastroGard® paste showed no visible air pockets and had consistent homogeneity.

The use of compounded omeprazole and illegally produced over the counter omeprazole products in the equine industry is widespread. The lack of an FDA finding on production quality leaves the potential for significant production quality problems. Compounded and illegal omeprazole paste products often have production quality control problems as documented by incomplete filling of syringes, air pockets and variations in homogeneity.

Effects of TLR Agonist and Clarithromycin on Growth of Rhodococcus equi in Adult Equine Macrophages

T. Likavec1, A. Bordin2, S. Giguere3, N. Cohen2

1Texas A&M University, Bryan, TX, USA, 2College of Veterinary Medicine, Texas A&M University, College Station, TX, USA, 3College of Veterinary Medicine, University of Georgia, Athens, GA, USA

The increasing prevalence of macrolide‐resistant Rhodococcus equi and the lack of alternative antimicrobials necessitates identifying novel therapies. Stimulating host innate immune responses can enhance clearance of infectious agents. The objectives of this study were: 1) to determine whether an agonist for Toll‐like receptors (TLR) 2/6 and 9 (viz., PUL‐042) could increase the killing capacity of adult equine monocyte‐derived macrophages (MDMs) against macrolide‐resistant and macrolide‐susceptible R. equi; and, 2) to determine whether PUL‐042 had a synergistic effect with clarithromycin on the killing capacity of MDMs against R. equi. The null hypotheses were that PUL‐042 would not increase the killing capacity of MDMs against either strain of R. equi and that PUL‐042 would not have a synergistic effect on the killing capacity of MDMs when combined with clarithromycin.

To achieve these objectives, MDMs were harvested from 8 apparently healthy, University‐owned adult horses. The MDMs of each horse were then separately infected with a macrolide‐resistant and a macrolide‐susceptible strain of R. equi. Phagocytosis of the bacteria was determined by counting colony forming units (CFUs) from lysed MDMs at baseline (T0 = 45 minutes post‐infection). The CFUs were also determined at 48 hours post‐infection (T48) with 1 of 4 treatments: no treatment (negative control), PUL‐042 alone (PUL), clarithromycin alone (CLR), or a combination of PUL‐042 and clarithromycin (PULC). The percentage of CFU reduction was determined by calculating the ratio of CFUs between T48 and T0 for each treatment and multiplying by 100%. Results showed that phagocytosis varied significantly (P < 0.05) among horses. Additionally, there was no significant difference in phagocytosis (i.e., T0 CFU) between strains (P = 0.27). For both macrolide‐susceptible and macrolide‐resistant R. equi, treatment with CLR or PULC had marked and significant (P < 0.05) reduction of CFU relative to either negative control or treatment with PUL; for both strains, there were no differences between control and PUL or between CLR and PULC. Although the macrolide‐resistant strain tended to be less effectively reduced, this difference was not significant (P = 0.14).

In conclusion, PUL‐042 alone did not effectively kill either a macrolide‐resistant or a macrolide‐susceptible strain of R. equi in equine MDMs. Additionally, there was not a detectable synergistic effect of PUL‐042 in combination with clarithromycin. A similar experiment using adult alveolar macrophages is in progress to determine if more specialized macrophages are affected by PUL‐042 and results will be reported. It is possible that PUL‐042 might exert indirect effects on macrophages through stimulation of respiratory epithelial cells.

Outcome of a Novel Vertebral Interbody Fusion Technique in Horses with Cervical Vertebral Compressive Myelopathy

Y. Nout‐Lomas, J. Easley, E. Aldrich, R. Bayless, H. Seim

Colorado State University, Fort Collins, CO, USA

Current surgical techniques for cervical stabilization are technically challenging and can result in significant complications. Also, they rely solely on compression and do not provide stabilization in tension, which can result in micromotion, potential catastrophic failure, and implant migration. Here we evaluated a novel technique for cervical vertebral interbody fusion in 6 horses with cervical vertebral compressive myelopathy (CVCM). Six horses diagnosed with CVCM with pre‐operative ataxia grades ranging from 2–4/5 on the modified Mayhew scale underwent this novel approach. Two horses were compressed at 1 site and 4 were compressed at 2 sites (Table). The surgical procedure consisted of creation of an intervertebral disc‐space defect and placement of an intervertebral fusion device (IFD), 4 titanium polyaxial pedicle screws and 2 connecting rods secured with set‐screws (Figure). All horses were re‐evaluated at 7 days, 6 months, and 1‐year post‐operative. The procedures were approved by the Colorado State University Institutional Animal Care and Use Committee. Surgery was performed safely and post‐operative radiographs showed good placement of all IFDs, screws and rods in all horses. No implant failure or IFD displacement occurred up to 1‐year post‐operative. All horses improved within 3 months and by 1‐year post‐operative ataxia had reduced by 1–2 grades. The use of this novel fusion system requires less disc space removal, provides stabilization in compression and tension, and was technically straight‐forward to perform. Surgery was performed successfully with improvement of clinical signs in all horses. This technique is a promising alternative to current surgical procedures for treatment of CVCM.

Accuracy of Two Patient Registration Techniques for Neuronavigation‐Guided Brain Biopsy in the Horse

L. Santistevan1, R.A. Packer1, J. Easley2, E. Randall3, A. Ruple4, F. Wininger5

1College of Veterinary Medicine and Biomedical Sciences, Colorado State University, Fort Collins, CO, USA, 2Colorado State University, Fort Collins, CO, USA, 3College of Veterinary Medicine and Biomedical Sciences, Fort Collins, CO, USA, 4Purdue University, West Lafayette, IN, USA, 5Veterinary Specialty Services, Manchester, MO, USA

Neuronavigation‐guided intracranial surgery is increasingly common in veterinary medicine, although its use has not yet been described in equine species. The objective of this study was to determine the accuracy of neuronavigation‐guided intracranial procedures in the horse using a standard fiducial array compared to anatomic landmarks for patient registration. Six equine cadaver heads were used for the study. CT images of each specimen were acquired, with the fiducial array rigidly secured to the frontal bone. Six targets were selected in each specimen. Patient registration was performed separately for 3 targets using the fiducial array, and 3 targets using anatomic landmarks. In lieu of biopsy, 22‐gauge wire seeds were placed at each targeted site. Postoperative images were coregistered with the planning scan to calculate offset from the tip of the seed to the target. No statistical difference in accuracy between registration techniques was identified. The impact of surgeon experience was examined separately for each technique using a two‐sample t‐test and the experienced surgeon was significantly more accurate (1.35 mm, SD 0.69) than novice surgeons (4.04 mm, SD 1.78) using the fiducial array (P < 0.01). Although not statistically significant (P = 0.41), for the experienced surgeon the mean distance to target was smaller when registering with the fiducial array (1.35 mm) than natural landmarks (2.24 mm). In conclusion, surgeon experience impacts degree of accuracy. Data also suggest that registration using anatomic landmarks is feasible, but perhaps should be reserved for cases where a separate planning scan with fiducials cannot be performed.

Causes of Fatalities During FEI Three‐Day Eventing

I. Comyn1, A. Bathe1, A. Foote2, C. Marr3

1Rossdales LLP, Exning, UK, 2Rossdales LLP, Newmarket, UK, 3Rossdales LLP/Equine Veterinary Journal/University of Glasgow/Horserace Betting Levy Board, Exning, UK

Data on equine deaths in Three‐Day Eventing (3DE) is lacking. We aimed to report the prevalence of sudden death and fatal musculoskeletal injury in 3DE and describe causes of sudden death. Numbers of starts, fatalities (death/euthanasia) and fatal musculoskeletal injuries between 2008 – 2014 were obtained from the International Federation for Equestrian Sports (FEI). Post‐mortem reports were reviewed and sudden deaths were classified using categories defined for Thoroughbred racing.

There were 112,958 starts and 66 fatalities thus a prevalence of 5.8 deaths/10,000 starts. There were 41 (62%, 3.6/10,000 starts) musculoskeletal injuries; 16 (24%, 1.4/10,000 starts) sudden deaths, 9 cases (14%) where further details were not provided. There was considerable variation in available information on post‐mortem examination. A definitive cause of death was identified in 6 of the 16 sudden death cases: 3 central nervous system trauma and 3 blood vessel rupture. Presumptive diagnoses were proposed in 5/16 cases, all cardiopulmonary failure. Cause of sudden death was unidentified in 5/16. This distribution of musculoskeletal injury versus sudden death is similar to fatalities in racing. The distribution amongst categories was also similar to Thoroughbred racing fatalities but definitive diagnoses were reached less frequently. A more standardised system of reporting fatalities would be valuable.

Lyle, C. H et al. (2011), Sudden death in racing Thoroughbred horses: An international multicentre study of post mortem findings. Equine Veterinary Journal, 43: 324–331.

Effects of Endothelial Colony Forming Cell/PEG‐Fibrinogen Scaffolds on Wound Healing in Horses

R.L. Winter1, Y. Tian2, W.J. Seeto2, F.J. Caldwell1, D.D. Pascoe3, J.W. Koehler4, E.A. Lipke2, A.A. Wooldridge1

1Department of Clinical Sciences, College of Veterinary Medicine, Auburn University, Auburn, AL, USA, 2Department of Chemical Engineering, Auburn University, Auburn, AL, USA, 3School of Kinesiology, Auburn University, Auburn, AL, USA, 4Department of Pathobiology, Auburn University, Auburn, AL, USA

Endothelial colony forming cells (ECFCs) are progenitor cells which function in neovascularization and may be useful therapeutically in conditions with poor blood supply, such as distal limb wounds in horses. An injectable ECFC/hydrogel scaffold may ensure cell survival and localization to improve neovascularization and healing.

Autologous ECFCs were isolated from 6 adult horses, labeled with quantum nanodots (QD), and a subset of cells encapsulated in poly(ethylene) glycol fibrinogen microspheres (MS). Full‐thickness dermal wounds were created in duplicate on each distal limb and randomly assigned to injections with empty MS, serum, ECFCs, and ECFCs encapsulated into MS (ECFC‐MS). Wounds were either biopsied weekly or at week 4 only. Analysis included wound surface area (WSA), granulation tissue scoring (GS), thermography, and immunohistochemical staining for von Willebrand factor (VWF).

Treatments were well tolerated in all horses. QD labeled cells were identified in wound biopsies up to 3 weeks post‐injection. There were no significant effects of treatment on GS, but the WSA was influenced by treatment (P = 0.0002), with ECFCs alone having the smallest WSA at 4 weeks. GS were greater for wounds biopsied (P = 0.0009) and for hindlimbs (P = < 0.0001). Thermographic analysis revealed that the ECFC‐MS group had the greatest decrease in limb temperature over time (P = 0.011). The vWF immunostaining density was lowest in the ECFC‐MS group.

Treatment effects on wound temperature, vessel density, and wound size may indicate that ECFCs affect wound healing in equine distal limbs. Despite variability within the model, ECFC treated wounds had a significantly positive effect on wound healing.

Measured and Calculated Parameters of Global Oxygenation in Healthy Foals

D.M. Wong1, K.L. Hepworth‐Warren2, J. Howard3, B. Sponseller3

1Iowa State University, Ames, IA, USA, 2None, Farmingdale, NJ, USA, 3None, Ames, IA, USA

Septicemia remains a common clinical entity in neonatal foals with varying degrees of illness associated with the process, including cardiovascular shock. In some instances, clinical parameters such as heart rate, blood pressure and urine output may be unreliable or late indicators of inadequate tissue perfusion and poor global oxygenation status, thus providing limited guidance for timely therapy. Conversely, more invasive monitoring such as measurement of cardiac output (CO) and mixed venous oxygen saturation of hemoglobin have been advocated in some studies to better direct therapy (i.e. early goal‐directed therapy) and improve outcome in people that are critically ill or in septic shock.

Very few investigations have evaluated global oxygenation in foals, despite the frequency of equine neonatal sepsis. Consequently, equine clinicians primarily rely on physical exam parameters and blood lactate to guide therapy. Because placement of a pulmonary artery catheter in neonatal foals to measure mixed venous blood (SvO2) is clinically impractical, the objective of this study was to evaluate parameters associated with global oxygen from central venous blood in healthy neonatal foals. Specifically, in this study, the following parameters were measured from 12 healthy neonatal foals from central venous (cv) and arterial (a) blood: saturation of hemoglobin with oxygen (ScvO2, SaO2), partial pressure of oxygen (PcvO2, PaO2), partial pressure of CO2 (PcvCO2, PaCO2), pH (pHcv, pHa) and L‐lactate (Laccv, Laca). Additionally, the following parameters reflecting global oxygenation were calculated: venous‐to‐arterial CO2 gap (dCO2), oxygen extraction ratio (O2ER), ratio between PaO2 and the fraction of inspired oxygen (PaO2:FIO2 ratio) and central venous and arterial bicarbonate (HCO3cv, HCO3a), base excess (BEcv, BEa), hemoglobin (Hbcv, Hba) and oxygen content (O2Ctcv, O2Cta).

No complications were observed with catheter placement or data collection in this study and in summary, this study reports central venous and arterial blood parameters associated with global oxygenation in healthy foals to establish baseline values at different time points within the neonatal period. These parameters, many utilized in human critical care, might allow better hemodynamic monitoring in critically ill foals. The forthcoming objective is to compare similar global oxygen parameters from critically ill foals, with the hope of improving patient monitoring, therapy and prognostication in equine neonatal sepsis and septic shock.

A Randomized, Blinded, Controlled Crossover Study of Dipyrone to Control Fever in Horses

R. Avenatti1, M. Yin1, C. Reinemeyer2, M.P. O'Banion1, E. Sundman1, T. Hu1

1Kindred Biosciences, Inc, Burlingame, CA, USA, 2East Tennessee Clinical Research, Inc, Rockwood, TN, USA

Dipyrone has well‐described anti‐pyretic effects. This study evaluated the effectiveness of dipyrone administered intravenously to control fever in horses compared to placebo.

Thirty‐one horses with fever (rectal temperature ≥ 102.0°F) from naturally occurring disease were randomized to two groups. Fifteen horses were treated with a single dose of 30 mg/kg dipyrone intravenously (IV), and sixteen horses were treated with a single dose of saline (placebo) IV. Medications or procedures that may have affected the rectal temperature were prohibited. Rectal temperature was assessed six hours following treatment. Twenty‐four hours following dose administration, febrile horses were crossed over to the other treatment group under the same procedures. Treatment response was defined as a temperature decrease of either ≥ 2.0°F from baseline or a return to normothermia (≤ 101.0°F) six hours after administration of dipyrone or saline. Statistical comparison between treatments was performed using the Fisher's Exact test, two‐sided.

Thirty‐one (31) horses completed the first phase and 27 horses completed the second phase. Results are presented in Table 1.

The majority of adverse events were due to infection from underlying disease. Three horses in the dipyrone group had episodes of soft manure, of which two were pre‐existing prior to treatment. One horse in the dipyrone group had several serious adverse events related to severe pleuropneumonia, and was removed from the study following the first phase to receive treatment.

A single administration of dipyrone IV effectively controlled fever 6 hours post administration and was superior to placebo.

A Controlled Study of the Clinical Safety of an Oral Formulation of Dipyrone in Horses

E. Sundman1, C. Reinemeyer2, M.P. O'Banion1

1Kindred Biosciences, Inc, Burlingame, CA, USA, 2East Tennessee Clinical Research, Inc, Rockwood, TN, USA

Dipyrone is available as an injectable formulation in Europe; however, no oral formulation has been developed for use in horses. The commonly accepted dose of dipyrone is 30 mg/kg administered intravenously. This study evaluated the clinical safety of a high dose oral formulation of dipyrone in horses.

Seven horses were randomized in two treatment groups; 5 horses to dipyrone treatment and 2 horses to saline treatment (placebo). The dipyrone group was treated with 80 mg/kg dipyrone per os (PO) twice daily for 15 days. Placebo horses received saline on the same schedule. Gastroscopy was performed and blood samples for hematology and serum chemistry were collected at baseline, Day 8 and Day 15.

All horses tolerated high dose oral dipyrone administration well. Two dipyrone‐treated horses and one placebo horse showed oral mucosal trauma that was related to syringe dosing. One placebo horse had a mild episode of colic on study day 9. Gastroscopy results are summarized in Table 1. below. One dipyrone‐treated horse had a clinically significant increase in creatine kinase (CK) and aspartate aminotransferase (AST) beginning on day 8 and remained unchanged throughout the study. One dipyrone‐treated horse had a clinically significant increase in CK at study termination.

Initial assessment of high dose dipyrone was well tolerated in horses when administered twice daily for 15 days.

In Vitro Effects of Tamoxifen on Equine Neutrophil Respiratory Burst and Phosphatidylserine Expression

C. Olave, N. Morales, C. Henriquez, B. Uberti, J. Sarmiento, H. Folch, G. Moran

Universidad Austral de Chile, Valdivia, Los Rios, Chile

Neutrophils participate in innate immunity as the first line of host defense against microorganisms. However, exacerbated neutrophil activity can be harmful to surrounding tissues; this is important in a range of diseases, including allergic asthma and chronic obstructive pulmonary disease in humans, and equine asthma. Tamoxifen (TX) is a non‐steroidal estrogen receptor modulator with effects on cell growth and survival. TX has been shown to induce early apoptosis of peripheral blood and bronchoalveolar lavage (BAL) neutrophils from horses with acute lung inflammation (Perez et al., 2016). Our study group previously showed that neutrophil respiratory burst and IL‐8‐induced chemotaxis also decrease after treatment with TX, confirming a direct action of the drug on the cell type (Borlone et al., 2017).

We studied the pharmacological effect of TX on equine neutrophils obtained by jugular venipuncture from 6 healthy horses, and purified by Ficoll gradient centrifugation. Respiratory burst was evaluated by chemoluminescence induced by opsonized zymosan in neutrophils. Luminol‐emitted luminescence was measured at 37°C for 3 hours using a luminometer. The Different concentrations of TX and its combination with selective estrogen receptor modulators (SERM) were evaluated (TX 0.5; 1; 2.5; 5; 7.5; 10; 15 and 20 μM; ICI 182,780 [high affinity estrogen receptor antagonist] 1, 10 and 100 nM; G15 [selective G‐protein estrogen receptor antagonist] 1, 10 and 100 nM; β‐estradiol 1, 10 and 100 nM; G1 [selective G‐protein estrogen receptor agonist] 1, 10 and 100 nM; ICI 100 nM + TX 11.5 μM; G15 100 nM + TX 11.5 μM; β‐estradiol 100 nM + TX 11.5 μM; and G1 100 nM + TX 11.5 μM). Phosphatidylserine expression was quantified by flow cytometry using a commercial Annexin V‐FITC kit (BD Biosciences). Cells were incubated at 37°C in a humidity chamber (CO2 5%) for 20 minutes with: DMSO 0.1% (vehicle), TX 11.5 μM,; β‐estradiol 100 nM; ICI 100 nM; G15 100 nM and G1 100 nM.

Results showed that TX inhibits neutrophil respiratory burst induced by opsonized zymosan in a dose dependent manner. Nuclear (17‐β‐Estradiol) and GPR30 cell membrane (G1) estrogen receptor agonists and their antagonists (ICI 182,780 and G15 respectively) do not block nor reproduce TX′s effect. Thus, TX does not inhibit respiratory burst through estrogen receptors. TX 11.5 μM also produced an increase in phosphatidylserine translocation, which was not observed with any of the SERM used. This shows a direct pro‐apoptotic effect on neutrophils. In conclusion, TX′s inhibitory and pro‐apoptotic effect on neutrophils seems to be independent of estrogen receptors.

Effects of Magnesium Sulfate Infusion on Lung Function and Clinical Scores in Severe Asthmatic Horses

L. Tanquerel1, G. Fillon‐Berttrand2, J.‐P. Lavoie3, M. Leclère3

1Université de Montréal/Resident in Equine Internal Medicine, Saint‐Hyacinthe, PQ, Canada, 2Université de Montréal, Saint‐Hyacinthe, PQ, Canada, 3Université de Montréal, Equine Internal Medicine Professor, Saint‐Hyacinthe, PQ, Canada

We hypothesized that magnesium sulfate (MgS04) has bronchodilator effects in severe asthmatic horses and potentiates the effects of inhaled salbutamol.

Six severe asthmatic horses were studied using a crossover experiment with a wash‐out period between treatments. Clinical scores and lung function measured by impulse oscillometry (IOS) were recorded before and after administration of salbutamol (800 μg) and mock procedure (experiment Ia, to document reversibility). The same parameters were also recorded before and after MgSO4 infusion (2.2 mg/kg/min IV over 20 min) and an equivalent volume of saline, adjusted to the same osmolarity (experiment Ib), and before and after MgSO4 infusion and salbutamol, compared to salbutamol alone (experiment II).

Salbutamol significantly improved the lung function in studies Ia and II. MgSO4 significantly improved clinical scores when administered alone or with salbutamol (P < 0.05), with a trend for lung function improvement only observed when administered in combination with salbutamol (P = 0.06). MgSO4 also increased the tidal volume (experiment II). Saline administration and the mock procedure had no effect on lung function or clinical scores.

MgSO4 alone or in combination with salbutamol had a mild positive effect on clinical scores of severe asthmatic horses. It had more effect on clinical scores than IOS parameters, which could be due to an undetected bronchodilatory effect or a change in the breathing strategy.

Comparison of Oral, Intravenous and Subcutaneous Fluid Therapy for Resuscitation of Calves with Diarrhea

V. Dore, D.M. Foster, G.W. Smith

North Carolina State, University College of Veterinary Medicine, Raleigh, NC, USA

Neonatal diarrhea is a major source of economic loss to the cattle industry and the leading cause of calf mortality in most countries. The major goals of treatment are to correct dehydration, increase blood pH, restore electrolyte concentrations and provide nutritional support. It is quite possible that the use of oral electrolytes in calves with moderate dehydration and acidosis is of equal or perhaps even greater benefit than giving large volumes of fluids either intravenously or subcutaneously.

The objective of this study was to compare the ability of a commercially available oral electrolyte solution (OES) alone or in combination with hypertonic saline in comparison to the administration of intravenous or subcutaneous fluid therapy to resuscitate calves with osmotic diarrhea.

Thirty‐two Holstein bull calves 5–27 days of age were used in this study. Calves were acclimated to their housing and normal diet for two days. Osmotic diarrhea and dehydration were induced by adding sucrose to the milk replacer. In addition, spironolactone and hydrochlorothiazide were given orally and furosemide intramuscularly three times per day. Clinical hydration scores, depression status, fecal consistency and body weight were recorded at each feeding period. Treatment was started when the calf had a hydration score of 2, depression score of 1 and fecal score of 3. Calves were randomly divided in four treatment groups of 8 calves each: 1) OES alone (mixed according to label directions); OES with hypertonic saline (4 mL/kg, IV); 3) intravenous fluids (LRS, 2 liters) or 4) subcutaneous fluid (LRS, 2 liters). Treatment was administered at 0 and 12 hours. Changes in plasma volume, acid‐base status, electrolytes concentrations and physical examination parameters were recorded before fluid therapy and again at 1, 2, 4, 8, 12 hours after each treatment.

Decreases in hematocrit, total protein, albumin concentration as well as increase in blood pH between time 0 and 2 were observed in all group. The oral electrolytes group showed greater increase of blood pH at time 1 and 2. All calves showed improvement of their hydration scores, plasma volume, attitude score and fecal score over the period of study. Fecal score did not appear to be affected by treatment.

In conclusion, oral electrolytes with or without hypertonic saline appeared to be an excellent alternative for resuscitation of calves with moderate diarrhea and acidosis and often performed better than small volumes of IV or SC fluids.

Variability of Bovine Leukemia Virus Bulk Tank Milk Antibody Levels over Different Sample Collection Intervals

E. John1, O. Nekouei2, J.T. McClure1, M. Cameron1, G. Keefe1, H. Stryhn1

1Atlantic Veterinary College, Charlottetown, PI, Canada, 2Faculty of Veterinary Medicine, University of Calgary, Calgary, AB, Canada

Bulk tank milk (BTM) samples are used to determine the infection status and estimate dairy herd prevalence for Bovine Leukemia Virus (BLV) using an antibody ELISA assay. BLV ELISA variability between samples from the same herd or from different herds has not been investigated. The objective was to determine the within‐herd and between‐herd variability of a BTM BLV ELISA assay over 1‐month, 3‐month, and 3‐year sampling intervals.

Canadian Maritime region dairy herds (n = 523) contributing to bulk tank milk samples in 2013 and 2016 participated. BLV antibody levels were measured in three BTM samples collected at 1‐month intervals in early 2013 as well as two BTM samples collected over a 3‐month interval in early 2016. Random‐effects models were used to estimate the variability between BTM samples from the same herd and between herds for 3 sampling intervals.

The majority of variability of BTM samples was seen between herds (81.14–92.66%). Unexplained variance between samples from the same herd, on square‐root scale, was greatest for the 3‐year (0.983 ± 0.061), followed by the 1‐month (0.615 ± 0.027) then the 3‐month (0.560 ± 0.035) intervals.

Variability of BTM sample antibody levels within the same herd is present but is much smaller than the variability between herds, and is greatest for the 3‐year sampling interval. The 3‐month sampling interval resulted in the least variability and is appropriate to use to determine the baseline level of within‐herd prevalence for BLV control programs and to determine the effectiveness of control programs when repeated at regular intervals.

Comparison of Six On‐Farm Tests to Estimate Somatic Cell Count at Dry‐Off in Dairy Cattle

S.A. Kandeel, A.A. Megahed, P.D. Constable

University of Illinois at Urbana‐Champaign, Urbana, IL, USA

Somatic cell count (SCC) is the most commonly used method to monitor udder health in dairy cattle. Our objective was to evaluate the clinical performance of different cow‐side tests for SCC (California Mastitis Test {CMT}, Somaticell® test and three PortaSCC® tests) for diagnosing subclinical intramammary infection (IMI) at dry‐off.

Quarter foremilk samples were collected from 117 dairy cows at dry‐off. Quarter SCC was measured using Delaval® cell counter (reference method; IMI=SCC>200,000 cells/mL), CMT, Somaticell®, and three PortaSCC® tests. Spearman correlation coefficient, area under the receiver operating curve (AUC), and kappa coefficient (κ) were calculated. P < 0.05 was considered significant.

Compared to the reference method, CMT had the highest correlation (rs=0.89), and at optimal cut‐point (score ≥ trace), AUC=0.88 and κ=0.77. PortaSCC® color test was the second best performing test (rs=0.82), but required 45 minutes to produce a result, and at the optimal cut‐point (50,098 SCC/mL), AUC=0.91 and κ=0.74. PortaSCC® quick test was the third best performing test (rs=0.80), and at the optimal cut‐point (54,482 SCC/mL), AUC=0.88 and κ=0.62. PortaSCC® reader (rs=0.68) was inferior to PortaSCC® color test, and at the optimal cut‐point (73,419 SCC/mL), AUC=0.82 and κ=0.50. The Somaticell® test didn't perform well (rs=0.44; AUC=0.68, κ=0.24) with an optimal cut‐point of 123,864 SCC/mL.

We conclude that the CMT provides the most accurate, practical, and least cost on‐farm screening test to predict IMI at dry‐off. However, because the CMT is a semi‐quantitative test, we recommend using the PortaSCC color test® if a more quantitative on‐farm screening test is required.

Ability of Milk pH to Predict Increased Somatic Cell Count at Dry‐Off in Dairy Cattle

S.A. Kandeel, A.A. Megahed, P.D. Constable

University of Illinois at Urbana‐Champaign, Urbana, IL, USA

Milk pH is increased in lactating dairy cattle with subclinical mastitis (SCM). Milk pH testing may therefore provide an economic, rapid, and practical method for diagnosing SCM in the field. Our objective was to evaluate the clinical utility of measuring milk pH using a PICCOLO plus® pH meter, Multistix® 10 SG Reagent Strips for Urinalysis (Multistix® strips), and pH Hydrion® paper as on‐farm screening methods for diagnosing SCM in dairy cattle at dry‐off.

Quarter foremilk samples were collected from 117 dairy cows at dry‐off. Quarter somatic cell count (SCC) was measured using a Delaval® cell counter with SCC≥200,000 cells/mL as the reference method for SCM. Milk pH was measured using the pH meter, Multistix® strips, and pH Hydrion® paper. Spearman correlation coefficient, area under the receiver operating curve (AUC), and kappa coefficient (κ) were calculated. P < 0.05 was considered significant.

Compared to the reference method, Multistix® strips had the highest association (rs=0.61) with Delavel® SCC, and at the optimal cut‐point (pH≥7.0), AUC=0.77 and κ=0.42. The pH meter (rs=0.52) was inferior to Multistix® strips, and at the optimal cut‐point (pH≥6.69), AUC=0.74 and κ=0.34. pH Hydrion® paper performed poorly at an optimal cut‐point of ≥6.2 (rs=0.36; AUC=0.59, κ=0.12).

We conclude that milk pH does not provide a clinically useful primary method for diagnosing SCM at dry off in dairy cattle. Combining milk pH with other diagnostic tests may improve the clinical utility of milk pH to detect quarters with increased SCM.

Clinical Utility of Four Point‐of‐Care Enzymatic Tests in Predicting Somatic Cell Count in Dairy Cattle

S.A. Kandeel, A.A. Megahed, P.D. Constable

University of Illinois at Urbana‐Champaign, Urbana, IL, USA

Inflammation of the mammary gland increases the activity of more than 20 enzymes in the glandular secretion. Our primary objective was to characterize the ability of three enzymes in quarter milk samples, catalase, lactate dehydrogenase (LDH), and esterase, to predict the presence of a high somatic cell count (SCC) and intramammary infection (IMI) in lactating dairy cattle.

Quarter foremilk samples were collected from 117 dairy cows at dry‐off and 109 fresh cows within 4–7 days post calving. A Delaval® cell counter was used as the reference method to measure quarter SCC, with IMI being defined as SCC≥200,000 cells/mL. Milk catalase activity was measured using the Accutest Uriscreen® test. Milk LDH activity was measured using Udder‐Check® strips. Milk esterase activity was measured using Multistix® and Serim PeriScreen® strips. Spearman correlation coefficient, area under the receiver operating curve (AUC), and kappa coefficient (κ) were calculated. P < 0.05 was considered significant.

Compared to the reference method, Udder‐Check® strips had the strongest association (rs=0.79) with IMI, and at the optimal cut‐point score≥1, AUC=0.88 and κ=0.73. The Accutest Uriscreen® test had the second highest association (rs=0.72) with IMI, and at the optimal cut‐point foam height≥4 mm, AUC=0.85 and κ=0.54. Serim PeriScreen® strips had the third highest association (rs=0.69), and at optimal cut‐point score≥1, AUC=0.72 and κ=0.42. Multistix® strips had the lowest association (rs=0.45; AUC=0.70, κ=0.19) at an optimal cut‐point score≥1.

We conclude that Udder‐Check® strips and the Accutest Uriscreen® test may provide clinically useful on‐farm enzymatic screening tests for predicting the presence of IMI.

Persistence of Coagulase Negative Staphylococcal Intramammary Infection in Dairy Goats

V.B. Gosselin, J.R. Middleton, P.R.F. Adkins.

University of Missouri, Columbia, MO, USA

The objective of this study was to describe the persistence of species‐specific coagulase negative staphylococcal (CNS) intramammary infection (IMI) in goats from a large dairy goat herd in Missouri.

Udder‐half level milk samples were aseptically collected for bacterial culture, somatic cell count (SCC), and milk components (% fat and % protein) determination from all lactating goats (n = 909) on a commercial goat dairy. Aerobic bacterial culture was performed on Columbia blood agar for 24 h at 37°C. Goats with at least one udder‐half yielding a single colony morphology of CNS were enrolled for two additional once monthly udder‐half level milk samplings. Staphylococcal isolates were identified to the species‐level using matrix assisted laser desorption/ionization time of flight (MALDI‐ToF) mass spectrometry or PCR amplification and sequencing of the tuf gene when MALDI‐ToF yielded an indeterminate result. Intramammary infection status was defined based on the presence of the same species of CNS in 1, 2, or 3 of the samples from each udder half. Differences in prevalence of chronic IMI (CNS‐CNS‐0, CNS‐0‐CNS or CNS‐CNS‐CNS) versus only an IMI on the first sample (CNS‐0‐0) were assessed within species using the Chi‐square test or Fisher's Exact test as applicable (Table 1). Significance was declared at P < 0.05. For species with ≤ 3 IMI within species, data were combined into a single category (Other CNS) for analysis.

Bacterial culture of milk samples from the initial sampling yielded 250 udder‐halves with a CNS IMI on 219 of the 909 goats. Of these 250 udder‐half IMI, 18 were lost to follow‐up after the 1st sample. An additional 15 halves were excluded due to unsuccessful identification of at least one of the CNS isolates from the 3 samplings. The remaining 217 IMI were classified based on the number of samples positive for the same CNS species (Table 1).

The most commonly identified species were S. simulans, S. xylosus, S. caprae, S. chromogenes, and S. epidermidis, and the majority of these IMI persisted over the 3‐month sampling period. On the other hand, most of S. arlettae and S. lentus IMI did not persist after the first sample. Unlike previous studies, Staphylococcus simulans was found to be associated with chronic IMI in the goats in the present study herd. These results suggest that similar to cows, some CNS species might be better host‐adapted whereas other CNS species might have an environmental niche.

Use of a Minimally Invasive Tube Cystostomy for the Management of Obstructive Urolithiasis in Goats

L.F. Wagner1, R. Streeter1, M. Boileau1, R. Oman2, J.L. Halleran1..

1Oklahoma State University, Stillwater, OK, USA, 2University of Missouri, Columbia, MO, USA

Obstructive urolithiasis is the most economically significant urologic disorder in male goats. Depending on the urolith type present, medical or surgical treatment options may be utilized. Medical therapy alone has a high rate of recurrence, while surgical options can be costly and typically require general anesthesia in an already compromised patient. A few low cost urine diversion options have been described, however many of them are associated with significant complications. An initial description of a minimally invasive tube cystostomy technique showed promise in the management of urolithiasis in small ruminants. The following report details our implementation and adaptations of this technique to patients in the United States.

A mixed population of Boer, Boer‐cross and Pygmy bucks and wethers (n = 10) presented to Oklahoma State University for obstructive urolithiasis between July 2015 and October 2016 and underwent a minimally invasive tube cystostomy (MITC) procedure. They ranged in age from 3 months to 8 years (mean = 56 months) and in weight from 7 kg to 90 kg (mean = 53 kg).

Briefly, the MITC procedure was performed with local anesthesia and sedation. A small keyhole incision was performed in the upper left paralumbar fossa, the dorsal aspect of the bladder was punctured with a trocar and a 10F silicone Jackson‐Pratt drain (n = 8) or a 16F pig‐tail latex gastrostomy tube (n = 2) was advanced into the bladder. The body wall was closed in a standard fashion and the tube was secured with a Chinese finger trap ligature where it exited the body wall. A one‐way valve was placed on the end of the tubing. Medical therapy, including urine acidification, antibiotics and non‐steroidal anti‐inflammatory drugs, was instituted after cystostomy tube placement. Depending on the overall health status of the patient, budget, and owner availability for continued care and monitoring, patients were either hospitalized or sent home following tube placement.

Following MITC, resolution of urolithiasis was achieved in 70% of the cases reviewed. The cystostomy tube was in place for an average of 8 days (range of 6–13 days). Complications were associated with the presence of calcium carbonate uroliths (n = 2), displacement of the tube from the bladder and accompanying uroabdomen (n = 3), or hemoabdomen (n = 1). Identifying the urolith type and utilizing a suitable tube and trocar may mitigate potential complications. Based on these cases, MITC shows promise to be a low cost yet effective method of managing obstructive urolithiasis due to phosphatic uroliths in male goats.

Clinical Utility of Measuring Plasma Fructosamine Concentration During Early Lactation in Dairy Cattle

A.A. Megahed1, M. Hiew2, P.D. Constable1

1University of Illinois at Urbana‐Champaign, Urbana, IL, USA, 2Faculty of Veterinary Medicine, Universiti Putra, Putrajaya, Malaysia

Fructosamine is widely used as a long term hyperglycemic biomarker in humans and dogs, but the clinical usefulness of fructosamine as a hypoglycemic biomarker is uncertain. Our primary objective was to evaluate the clinical utility of fructosamine in quantifying the magnitude of hypoglycemia and negative energy balance (NEB) during early lactation in dairy cattle.

Plasma samples were collected weekly from 106 Holstein‐Friesian cattle (34 primiparous, 72 multiparous) on days 4–34 postpartum. Plasma concentrations of glucose ([gluc]), β‐hydroxybutyrate ([BHB]), total protein ([PP]), and other clinicopathologic indices of energy status were determined. Backfat thickness and longissimus dorsi muscle thickness ([LDT]) were measured ultrasonographically. Plasma fructosamine concentration ([FRA]) was measured at approximately 28 days postpartum. Associations between [FRA] and study variables were evaluated using Spearman's rho and stepwise linear regression. P < 0.05 was considered significant.

A positive association was detected between plasma [FRA] and mean plasma [gluc] for days 4–28 postpartum (rs=+0.34, P = 0.0016), and between plasma [FRA] and LDT (rs=+0.31, P = 0.0039). Plasma [FRA] was negatively associated with mean plasma [BHB] for days 4–28 postpartum (rs=−0.31, P = 0.019). Stepwise linear regression identified a positive association between plasma [FRA] and mean [PP] for days 4–28 postpartum (rs=+0.24, P = 0.015). After correcting plasma [FRA] for [PP] at approximately day 28 of lactation, there was only a marginal improvement in the association between plasma [FRA] and mean postpartum plasma [gluc] (rs=+0.37, P = 0.0009).

We conclude that plasma [FRA] is not sufficiently sensitive or specific to provide a clinically useful method for quantifying the magnitude of hypoglycemia or NEB in early lactation.

Evaluation of a Portable Ion‐Selective Electrode Meter for Measuring Sodium Concentration in Cattle Body Fluids

A.A. Megahed1, M. Hiew2, P.D. Constable1

1University of Illinois at Urbana‐Champaign, Urbana, IL, USA, 2Faculty of Veterinary Medicine, Universiti Putra, Putrajaya, Malaysia

A low cost ion‐selective electrode (ISE) hand‐held meter (LAQUAtwin B‐722; Horiba, US) has recently become available for measuring the sodium concentration ([Na+]) in biological fluids. The objective of this study was therefore to characterize the analytical performance of the ISE meter in measuring [Na+] in bovine plasma, urine, milk, and abomasal fluid.

Four method comparison studies were completed using 152 plasma, 80 milk, and 206 urine samples from 16 lactating Holstein‐Friesian cows with experimentally‐induced free water, electrolyte, and acid‐base imbalances, and 94 abomasal fluid samples from 6 healthy male Holstein‐Friesian calves fed fresh milk with or without an oral electrolyte solution. Deming regression and Bland‐Altman plots were used to characterize the meter performance against reference methods (indirect ISE, Hitachi 911 and 917).

The LAQUAtwin ISE meter applied directly in plasma measured [Na+] 11.0 mEq/L (7.9%) lower than the indirect ISE reference method, consistent with the recommended adjustment of +7.5% when indirect ISE methods are used to analyze plasma. The LAQUAtwin ISE meter run in indirect mode accurately measured urine [Na+]. The LAQUAtwin ISE meter run in direct mode measured milk [Na+] 3.1 mEq/L (15.0%) lower than the indirect ISE reference method. The LAQUAtwin ISE meter run in direct mode measured abomasal fluid [Na+] 4.5 mEq/L (8.6%) lower than the indirect ISE reference method.

We conclude that the LAQUAtwin ISE meter provides a practical, low cost, rapid, and accurate point‐of‐care instrument for the on farm measurement of [Na+] in plasma, milk, urine, and abomasal fluid samples from cattle.

Ultrasonographic Determination of Longissimus Dorsi Muscle Thickness and Intramuscular Fat Percentage in Periparturient Dairy Cattle

A.A. Megahed1, M. Hiew2, P.D. Constable1

1University of Illinois at Urbana‐Champaign, Urbana, IL, USA, 2Faculty of Veterinary Medicine, Universiti Putra, Putrajaya, Malaysia

Several studies have examined the rate of backfat mobilization in periparturient dairy cattle, but the rate of intramuscular fat (IMF) mobilization relative to backfat mobilization and muscle tissue mobilization does not appear to have been studied. Our objective was therefore to characterize the changes in IMF relative to changes in muscle thickness and backfat thickness (BFT) in the physiologic response to negative energy balance in periparturient dairy cattle.

One hundred and six periparturient Holstein‐Friesian cattle (34 primiparous, 72 multiparous) at −3 and +28 days relative to parturition were examined. A 5 MHz linear ultrasound probe was used to measure BFT and the maximum thickness and IMF percentage (determined using a thresholding‐segmentation method) of the longissimus dorsi muscle in the thoracic region (LD‐thoracic). Spearman's correlation coefficient (rs) and mixed models analysis were used for statistical analysis and P < 0.05 was declared significant.

The mean decrease in BFT over the 31 day period was similar for primiparous (−46%) and multiparous cows (−45%). The mean decrease in IMF of the LD‐thoracic muscle was similar for primiparous (−22%) and multiparous cows (−30%). Likewise, the mean decrease in LD‐thoracic muscle thickness was similar for primiparous (18%) and multiparous cows (21%). The decrease in IMF was not associated with BFT decrease (rs=0.03), but was associated with the decrease in LD‐thoracic muscle thickness (rs=0.59, P = 0.0011).

We conclude that ultrasonographic measurement of the LD‐thoracic muscle thickness and IMF complements ultrasonographic measurement of BFT when quantifying negative energy balance in periparturient dairy cattle.

Plasma Potassium‐Lowering Effect of Different Hypertonic Infusion Solutions in Hyperkalemic Neonatal Diarrheic Calves

F.M. Trefz1, P.D. Constable2, I. Lorenz3

1Clinic for Ruminants with Ambulatory and Herd Health Services, LMU Munich, Oberschleißheim, Bayern, Germany, 2University of Illinois at Urbana‐Champaign, Urbana, IL, USA, 3Bavarian Animal Health Service (Tiergesundheitsdienst Bayern e.V.), Poing, Bayern, Germany

Hyperkalemia is a frequently observed electrolyte imbalance in dehydrated and acidemic calves with neonatal diarrhea that can result in skeletal muscle weakness and life‐threatening cardiac conduction abnormalities. This study aimed to compare the plasma potassium‐lowering effect of hypertonic sodium bicarbonate, sodium chloride, and glucose infusion solutions in the initial treatment of hyperkalemic calves. A total of 22 diarrheic calves (age < 21 days) with a plasma potassium concentration (cK) > 5.8 mmol/L were included in the study. Calves randomly received one of the following intravenous infusions over a period of 5 minutes: 8.4% NaHCO3 (6.4 mL/kg BW; n = 7), 7.5% NaCl (5 mL/kg BW; n = 8), or 46.2% glucose (5 mL/kg BW; n = 7), and were subsequently allowed to suckle 2 liters of an electrolyte solution. Infusions with NaHCO3 and NaCl provided an identical sodium load of 6.4 mmol/kg BW.

Infusions with hypertonic glucose resulted in marked hyperglycemia and hyperinsulinemia but cK remained unchanged during the first 20 minutes after initiation of treatment. Between 30 and 120 min after start of treatment the most marked decrement of cK was observed in hypertonic NaHCO3, with values that differed significantly (P < 0.05) from hypertonic NaCl, but not from hypertonic Glucose. After 120 min the observed decrements were equivalent to −26 ± 10%, −9 ± 8%, and −22 ± 6% of baseline cK in groups NaHCO3, NaCl, and Glucose, respectively. Results indicate a treatment advantage of hypertonic NaHCO3‐infusions over the use of hypertonic NaCl or glucose in hyperkalemic diarrheic calves, with alkalinization providing the most effective potassium‐lowering mechanism.

Effect of Age on Distribution of Danofloxacin and Tulathromycin in Dairy Calves

D.A. Mzyk1, J.L. Davis2, R.E. Baynes3, G.W. Smith3

1College of Veterinary Medicine – Food Animal Residue Avoidance Databank, North Carolina State University, Cary, NC, USA, 2VA‐MD College of Veterinary Medicine, Blacksburg, VA, USA, 3College of Veterinary Medicine, North Carolina State University, Raleigh, NC, USA

Calfhood diseases have a major economic impact on beef and dairy operations, due to the costs associated with mortality, treatment and the long term effects on growth and performance. With very few approved drugs for use in veal calves and young heifers, questions remain about the distribution and effectiveness of these drugs in young animals. The objective of this study was to compare the distribution of active drug concentrations in the plasma vs. different effector compartments including interstitial fluid (ISF) and pulmonary epithelial lining fluid (PELF) of healthy 3‐week‐old and 6‐month‐old calves.

8 calves in each age group were given a single subcutaneous (SC) dose of either danofloxacin (8 mg/kg) or tulathromycin (2.5 mg/kg). Plasma, ISF, and bronchoalveolar lavage (BAL) fluid were collected and analyzed for drug concentration. Plasma protein binding was measured using a microcentrifugation system in individual animals at 1, 7, and 21 days, 8 weeks and 6 months of age.

6‐month old calves maintained higher ISF/plasma concentration ratios throughout the study period compared to that observed in 3‐week old calves for both danofloxacin and tulathromycin. Non‐linear mixed effect modeling revealed age as a significant covariate in plasma concentration differences seen for calves administered tulathromycin but not for danofloxacin treated calves. Plasma protein binding was assessed in‐vitro in individual dairy calves (n = 5) for 5 drugs, including danofloxacin and tulathromycin. No significant differences were seen in calves from the ages of 1, 7 and 21 days of age but trends toward changes in 8‐week and 6‐month old calves indicate the need to assess protein binding changes for future drug approvals in young calves.

In conclusion, administration of danofloxacin and tulathromycin SC at label doses in 3‐week old and 6‐month old calves demonstrated differences in distribution that are affected by age. More research is needed to determine the impacts of physiological differences and disease on the distribution and efficacy of commonly used drugs in neonatal calves.

Cross‐Sectional Study of Teat and Inguinal Skin Colonization of Dairy Heifers with Various Staphylococcal Species

P.R.F. Adkins1, J.R. Middleton1, S. Dufour2

1University of Missouri, Columbia, MO, USA, 2University of Montreal, Saint‐Hyacinthe, PQ, Canada

The objective of this project was to describe the ecology of host‐adapted Staphylococcus spp. on the teat and inguinal skin of dairy heifers. A cross‐sectional study including Holstein heifers (n = 106) at the University of Missouri Foremost Dairy was conducted. Heifers were randomly selected based on housing type and age group and enrolled in the following groups; preweaned heifers (n = 22), weaned heifers age 2–6 months (n = 15), young pasture raised heifers age 6–13 months (n = 26), pasture raised heifers age 13–18 months (n = 22), and breeding age heifers ˃18 months (n = 21). Body site swabbing samples, which included a composite sample of all four teats and a second composite sample from the inguinal regions of each heifer, were collected. Swabbing samples were mixed with 10 mL of sterile saline, agitated, and plated on mannitol salt agar and incubated at 37°C for 24 h. Up to 10 staphylococcal colonies, including at least one of each morphologically distinct colony type, were saved for analysis. Staphylococcal isolates were speciated using matrix‐assisted laser desorption ionization‐time of flight (MALDI‐ToF) mass spectrometry or PCR amplification and sequencing of the rpoB or tuf gene. The Chi square or Fisher's exact test (as applicable) was used to investigate whether the prevalence of each staphylococcal species differed between the inguinal and teat region. Logistic regression models were used to investigate the relationship between heifer's age (continuous variable) and the probability of having a given body site colonized with a certain staphylococcal species.

A total of 2,042 staphylococcal isolates from 212 body sites of 106 heifers were evaluated. Of the 2,042 isolates, 1,359 were eliminated from the data analysis due to replication of a single species on the same body site, leaving 683 isolates. A total of 17 different staphylococcal species were identified. Staphylococcus aureus was more prevalent on teat skin than in the inguinal region (P < 0.01). All other staphylococcal species were equally distributed between the teat and inguinal region. For most species, colonization of the inguinal and teat region was associated with age of the heifer. Staphylococcus equorum, Staphylococcus lentus, Staphylococcus sciuri, and Staphylococcus vitulinus were more likely to be found on younger heifers, while Staphylococcus chromogenes, Staphylococcus devriesei, and Staphylococcus haemolyticus were more likely to be found on the older heifers. A diverse population of staphylococcal species was found on the teat and inguinal region of heifers, and staphylococcal populations tended to change with age. Some important host‐adapted species frequently associated with intramammary infection, e.g. S. chromogenes, become more prevalent with age.

Quantifying Stress in Primiparous and Multiparous Dairy Cattle Around Parturition

A.A. Megahed1, M. Hiew2, P.D. Constable1

1University of Illinois at Urbana‐Champaign, Urbana, IL, USA, 2Faculty of Veterinary Medicine, Universiti Putra, Putrajaya, Malaysia

Stress can negatively impact the health and productivity of lactating dairy cattle. The primary objective of this study was to determine the degree of stress experienced by dairy cattle around calving, and whether primiparous cattle were more stressed than multiparous cattle in the periparturient period.

One hundred and six Holstein‐Friesian cattle (34 primiparous, 72 multiparous) were studied. Animals were moved indoors into separate calving pens approximately 3 days before the anticipated calving date. Heart rate, respiratory rate, and mean arterial pressure (MAP) were determined daily from 3 days prepartum to 3 days postpartum. The MAP was measured using indirect oscillometry by application of a cuff to the coccygeal artery, and pressures were corrected for the height difference above the scapulohumeral joint. Plasma samples were collected and concentrations of glucose ([gluc]) and cortisol ([cortisol]) determined. Mixed models ANOVA was used for statistical analysis and P < 0.05 was considered significant.

Least squares mean daily heart rates were consistently 10 bpm higher in primiparous (97 bpm) than multiparous cattle (87 bpm, P < 0.0001). Least squares MAP was consistently higher in primiparous (97 mmHg) than multiparous cattle (91 mmHg, P = 0.0004). Least squares mean daily respiratory rates were consistently 8 breaths/min higher in primiparous (52 breaths/min) than multiparous cattle (44 breaths/min, P < 0.0001) Primiparous cattle had higher plasma [gluc] than multiparous cattle on days −2, −1, and 0.5 to 3.0 relative to parturition. Plasma [cortisol] was higher (P = 0.028) in primiparous (10.3 ng/mL) than multiparous (7.7 ng/mL) cattle the day before parturition.

We conclude that primiparous cattle are more stressed around parturition than multiparous cows.

Is Plasma L‐lactate a Reliable Predictor of Non‐Survival in Critically Ill Neonatal Calves?

F.M. Trefz1, I. Lorenz2, A. Lorch1

1Clinic for Ruminants with Ambulatory and Herd Health Services, LMU Munich, Oberschleißheim, Bayern, Germany, 2Bavarian Animal Health Service (Tiergesundheitsdienst Bayern e.V.), Poing, Bayern, Germany

Both in human and veterinary medicine, L‐lactate is a well‐established biomarker of tissue hypoxia, sepsis, disease severity and mortality and has also attracted increasing attention in bovine medicine due to the availability and validation of cheap and portable L‐lactate analyzers. The aim of the present retrospective analysis was therefore to assess the prognostic relevance of increased plasma L‐Lactate concentrations (cL‐lactate) in a large study population of hospitalized neonatal calves. For this purpose, the data of 10,060 neonatal calves admitted to the Clinic for Ruminants, LMU Munich, were retrospectively analyzed. L‐Lactate was determined as part of a routinely performed clinical biochemistry panel on admission.

The overall survival rate in this study population was 65.5% and increased cL‐lactate (> 2.2 mmol/L) were evident in 42.2% of calves. The survival‐rate of calves in the 9th (4.14–6.70 mmol/L) and 10th decile (> 6.70 mmol/L) of cL‐lactate was 52.3% respective 41.5%, opposed to 74.5% in calves of deciles 1–5 which had cL‐lactate in the reference range. A binary logistic regression analysis indicated that the odds for non‐survival increased by a factor of 1.16 (95% CI: 1.15–1.18) for every mmol/L increase of cL‐lactate (P < 0.001). The area under the ROC curve for cL‐lactate was 0.64 (95% CI: 0.63–0.66).

In conclusion, increased cL‐Lactate are associated with an increased risk of non‐survival in hospitalized neonatal calves, but they are not sufficiently reliable for the prediction of a negative outcome. Further work is however necessary to assess whether the prognostic relevance is dependent from the underlying disorder.