Background and aim of the study: Infective endocarditis (IE) is diagnosed by blood and/or resected valve cultivation and echocardiographic findings, as defined by the Duke criteria. Unfortunately, cultures may be negative due to prior antibiotic therapy or fastidious or slow-growing microorganisms. The study aim was to investigate the value of the broad-range polymerase chain reaction (PCR) in addition to blood and valve culture for the detection of causative microorganisms.
Methods: Between February 2012 and March 2015, valve samples from 36 patients undergoing cardiac surgery were analyzed; of these patients, 26 had a preoperative diagnosis of IE and 10 served as controls. Multiple blood cultures were obtained from 34 patients before antibiotic therapy was commenced. Valve samples were inoculated on bacteriological media and underwent analysis using broad-range PCR (16S rDNA).
Results: IE was confirmed microbiologically in 21 of the 26 patients (80.7%); in 20 cases (76.9%) this was by positive blood cultures and in 16 (61.5%) by positive valves. Valves were positive in 15 blood culturepositive patients, and in one blood-culture negative patient. Broad-range PCR detected a microorganism in valves significantly more frequently (n = 14; 53.8%) compared to valve culture (n = 8; 30.7%) (chisquare 11.5, p <0.001). The predominant microorganisms were Staphylococcus aureus, Streptococcus of the viridans group, coagulasenegative staphylococci and Enterococcus faecalis. Blood, valve cultures and broad-range PCR were negative in five patients (19.3%) with IE, and in all 10 subjects of the control group.
Conclusions: Broad-range PCR on valves was more sensitive than valve culture. However, blood culture, if taken before the start of antibiotic therapy, was the best method for detecting IE.