Meta-analysis of 18F-FDG PET/CT in the diagnosis of infective endocarditis

Background: The current diagnosis of infective endocarditis (IE) is based on the modified Duke criteria, which has approximately 80% sensitivity for the diagnosis of native valve endocarditis (NVE), with lower sensitivity for the diagnosis of prosthetic valve endocarditis (PVE) and culture-negative endocarditis. There is preliminary evidence that ¹⁸F-FDG PET/CT is an adjunctive diagnostic test with high accuracy reported in small studies to date. We therefore performed a meta-analysis of studies evaluating the use of PET/CT in the diagnosis of IE to establish a more precise estimate of accuracy. Methods: PubMed, Embase, Cochrane library, CINAHL, Web of Knowledge, and www.clinicaltrials.gov were searched from January 1990 to April 2017 for studies evaluating the accuracy of PET/CT for the evaluation of possible IE. Results: We identified 13 studies involving 537 patients that were included in the meta-analysis. The pooled sensitivity of PET/CT for diagnosis of IE was 76.8% (95% CI 71.8–81.4%; Q = 39.9, P < 0.01; I² = 69.9%) and the pooled specificity was 77.9% (95% CI 71.9–83.2%; Q = 44.42, P < 0.01; I² = 73.0%). Diagnostic accuracy was improved for PVE with sensitivity of 80.5% (95% CI 74.1–86.0%; Q = 25.5, P < 0.01; I² = 72.5%) and specificity of 73.1% (95% CI 63.8–81.2%; Q = 32.1, P < 0.01; I² = 78.2%). Additional extracardiac foci of infection were found on 17% of patients on whole body PET/CT. Conclusion: PET/CT is a useful adjunctive diagnostic tool in the evaluation of diagnostically challenging cases of IE, particularly in prosthetic valve endocarditis. It also has the potential to detect clinically relevant extracardiac foci of infection, malignancy, and other sources of inflammation leading to more appropriate treatment regimens and surgical intervention.