Background. Infective endocarditis (IE) poses significant diagnostic and therapeutic challenges, especially in cases of blood culture-negative infective endocarditis (BCNIE). Among patients undergoing surgery for IE, valve tissue may be evaluated for additional microbiological information by performing a broad-range 16S rDNA polymerase chain reaction (PCR) test, tissue culture and tissue histopathology. In patients with BCNIE, these diagnostic tests may identify the causative agents, guide further clinical management and improve local IE-related epidemiological data. The clinical utility of additional analysis of explanted tissue in a local South African (SA) setting has yet to be described. Objectives. To assess the clinical utility of performing PCR, culture and histopathology on the tissue of surgically explanted valves in BCNIE patients in an SA public sector hospital. We assess their diagnostic yield and treatment impact in a cohort of BCNIE patients treated with empirical antibiotic regimens. Methods. We analysed data from the Groote Schuur Hospital (GSH) Infective Endocarditis Registry, a prospective observational study of adult patients with infective endocarditis. Participants for this analysis were selected based on clinical and pathological criteria for IE and negative blood cultures. All participants were treated between January 2017 and March 2021. Results. During the study period, we identified 165 IE cases, 57 (34.5%) of which were blood-culture negative. BCNIE patients had a mean (standard deviation) age of 40.2 (13.4) years, and 41 (71.9%) were male. Twenty-seven of the 57 BCNIE patients underwent cardiac surgery and had tissue analysis performed. Tissue PCR identified an aetiological agent in 17/27 (63%) cases, with Bartonella spp. (12/27, 44%) being the most common organism. Tissue culture was positive in 3/27 (11%) cases, but the organisms identified were thought to reflect sample contamination. Tissue histopathology was performed in 22/27 (81.5%) cases and provided macroscopic confirmation of IE, but did not identify any specific organisms in any of the specimens. Only a small subset of the overall BCNIE cohort (11/57 (19.3%)) had serum serology for Bartonella spp. and Coxiella spp. performed, and 5/11 (45.5%) were positive for Bartonella spp. There was a 100% concordance rate between positive serum serology and tissue PCR. Tissue PCR impacted the antimicrobial regimen in 20/27 (74%) cases. Tissue culture and tissue histopathology did not influence antibiotic regimens in any patients. Conclusion. In this single-centre study, perioperative serum serological testing was underutilised in BCNIE. Tissue PCR was valuable in determining the aetiology of BCNIE, and influenced management. Tissue culture and histopathology had a poor yield and added little value in identifying the microbiological cause of BCNIE. Finally, the most common BCNIE causative organism identified by additional non-culture testing in our setting is Bartonella spp.
Read full abstract