Prevalence of hyposegmentation of granulocytes/Pelger-Huët anomaly in different canine breeds: a Bayesian approach

Pelger-Huët anomaly (PHA) is a benign congenital hematological disorder first observed in humans and occasionally reported in dogs. It has been mainly described in Australian Shepherd Dog (ASD) with a prevalence of 9.8-13% where, based on a genetic study, it was recently renamed hyposegmentation of granulocytes (HG). Prevalences in other canine breeds have not been documented. This study aims to: (1) estimate the prevalence of HG/PHA across various breeds, (2) quantify the uncertainty of the estimated values using a Bayesian approach, and (3) identify affected breeds not previously documented. This cross-sectional study was based on the CBC database of the San Marco Veterinary Clinic and Laboratory (Padua, Italy) from 2001 to 2024. Data were collected from dogs diagnosed with HG/PHA and breeds previously reported as affected. To handle limited data and provide reliable estimates, Bayesian analysis was performed to estimate the prevalence of the anomaly and its uncertainty, using posterior probabilities from an informative prior model. The analysis adhered to the Bayesian Analysis Reporting Guidelines (BARG). The study included 5,716 dogs: German Shepherd (GS, 40%), Dachshund (DA, 22.6%), Cocker Spaniel (CS, 17.3%), Border Collie (BC, 9.5%), ASD (5.9%), Samoyed (SA, 2.2%), Boston Terrier (BT, 1.6%), Australian Cattle Dog (ACD, 0.7%) and Basenji (BA, 0.2%). Overall, HG/PHA was found in 0.45% dogs, specifically in ASD (7.1%), SA (0.8%) and DA (0.08%) and not in the other breeds. The prevalence estimates were 6.47% in ASD with 95% Credible Interval (95% CrI) from 4.22 to 9.18%, 0.32% (95% CrI: 0.04, 1.11%) in SA, 0.2% (95% CrI: 0.02, 0.86%) in BA, 0.18% (95% CrI: 0.02, 0.77%) in ACD, 0.16% (95% CrI: 0.02, 0.64%) in BT, 0.11% (95% CrI: 0.02, 0.30%) in DA, 0.10% (95% CrI: 0.01, 0.34%) in BC, 0.08% (95% CrI: 0.01, 0.25%) in CS, 0.05% (95% CrI: 0.01, 0.15%) in GS. HG/PHA was newly identified in DA. This study, using laboratory data collected over two decades and analyzed with Bayesian methods, could be considered representative of the prevalence of HG/PHA in multiple canine breeds. It is the first study to estimate the prevalence of HG/PHA beyond ASD, highlighting breed-specific differences in a real-world setting.