Molecular detection of Bartonella spp. DNA in dogs with hemangiosarcoma.

Introduction: The potential role of pathogens, particularly vector-transmitted infectious agents, as a cofactor or cause of neoplasia has not been intensively investigated. We previously reported a potential link between Bartonella spp. bacteremia and splenic hemangiosarcoma (HSA) in dogs living in the United States. The purpose of this study was to: 1/ further determine the prevalence of Bartonella spp. DNA in dogs with splenic HSA from throughout the United States; 2/ assess the impact of sample preservation methods on Bartonella spp. DNA amplification using characterized tissue samples from dogs diagnosed with HSA.

Methods: In a blinded manner, we determined the presence of Bartonella spp. DNA in scrolls from biorepository formalin-fixed paraffin-embedded (FFPE) spleens from dogs living in three distant locations geographically transecting the United States. DNA extracted from non-lesional spleens (n = 249), nodular lymphoid hyperplasia spleens (n = 248), and splenic HSA (n = 330) were tested by quantitative polymerase chain reaction (qPCR), and droplet digital PCR (ddPCR). Subsequently, Bartonella PCR results from FFPE tissues and formalin-fixed tissues were compared using previously tested fresh frozen tissues from an additional 48 dogs with HSA.

Results: There was no significant difference in the proportion of Bartonella PCR positive FFPE tissues from dogs diagnosed with an alesional spleen, nodular lymphoid hyperplasia, and splenic HSA. Regardless of the histological diagnosis, the most common Bartonella species identified was B. henselae (32/38). Bartonella spp. DNA was detected in a significantly larger proportion of fresh frozen tissues compared to FFPE tissues, when tested by qPCR (22/48 versus 1/48; p <0.0001) or ddPCR (19/48 versus 1/48; p <0.0001). Using ddPCR, Bartonella DNA was more often amplified from formalin-fixed tissues compared to FFPE tissues (15/39 versus 1/39; p <0.0001). The sensitivity of qPCR on FFPE samples and formalin-fixed samples, when comparing to fresh frozen samples as the reference standard, was 4.5% and 11.8%, respectively.

Conclusion: Due to decreased DNA amplification efficiency, FFPE scrolls should not be used for the detection of Bartonella infection in spleen samples from dogs with HSA. PCR testing of fresh frozen tissues substantially improves the detection of Bartonella spp. infection. If fresh frozen tissues are not available, formalin-fixed tissues should be tested with digital PCR to enhance Bartonella DNA detection.