Molecular standardization and epidemiological mapping of canine haemoprotozoa in West Bengal, India

Canine vector-borne diseases (CVBDs) impose considerable morbidity in subtropical India, yet the epidemiological burden of haemoprotozoan polyparasitism, diagnostic performance of microscopy versus real-time PCR (qPCR), and clinical impact across five co-circulating pathogens have not been simultaneously quantified in West Bengal.

Peripheral blood from 584 client-owned dogs presenting at the State Animal Health Centre, Salt Lake, Kolkata (July 2024-December 2025), was analyzed in parallel by bright-field microscopy and a five-target qPCR panel targeting 18S rRNA (Babesia gibsoni, B. canis, Hepatozoon sp.) and 16S rRNA (Anaplasma spp., Ehrlichia canis). Representative amplicons were Sanger-sequenced and deposited in GenBank (PV567133, PV578474, PV803122, PV604994 and PZ413127). Statistical analysis included McNemar's test, Mann-Whitney U, multivariable logistic regression, Spearman's ρ, ROC analysis (Youden's J), Kruskal-Wallis H, and PCA.

The molecular prevalence among dogs was 76.54% for B. gibsoni (447/584), 55.75% for Anaplasma spp. (325/583), 28.94% for Hepatozoon sp. (169/584), 27.10% for B. canis (158/583), and 16.61% for E. canis (97/584); polyparasitism was common, with 72.9% of dogs harboring two or more concurrent pathogens and 34.9% infected by three or more species. Microscopy missed 130/158 B. canis positives (McNemar p < 0.0001); discordant samples had significantly higher median CT (32.45 vs. 29.63, Mann-Whitney p = 0.003), confirming failure at low parasitaemia. Microscopy missed 146/169 Hepatozoon sp. positives (McNemar p < 0.0001). Ehrlichia canis was the sole driver of acute multi-systemic morbidity, producing significantly lower platelet counts (105 vs. 186 × 10⁴/mm³, p < 0.0001), haemoglobin (125 vs. 132 g/L, p = 0.016), erythrocyte counts (5.33 vs. 6.10 × 10⁶/mm³, p = 0.008), and elevated ALT (56.70 vs. 41.80 IU/L, p = 0.0004) compared to PCR-negative dogs. Platelet count <109 × 10⁴/mm³ predicted E. canis infection (AUC = 0.675; sensitivity 54.6%, specificity 80.0%). B. gibsoni parasitaemia was significantly amplified in dogs co-infected with E. canis (CT 32.50 vs. 32.88; p = 0.023), indicating immunosuppressive synergy.

The high polyparasitism rate (72.9%) and the poor sensitivity of microscopy, particularly for B. canis (82.3% false-negative) and Hepatozoon sp. (86.4%), demonstrate that real-time PCR should be adopted as the standard diagnostic approach for canine haemoprotozoa in West Bengal. A platelet count below 109 × 10⁴/mm³ can serve as a practical screening threshold to prompt E. canis molecular testing.