ABSTRACT
BACKGROUND: 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG-PET/CT) has emerged as a useful diagnostic tool for suspected infective endocarditis (IE) in patients with prosthetic valves or implantable devices. However, there is limited evidence regarding use of 18F-FDG-PET/CT for the diagnosis of native valve endocarditis (NVE). METHODS: Between 2014 and 2017, 303 episodes of left-sided suspected IE (188 prosthetic valves/ascending aortic prosthesis and 115 native valves) were studied. 18F-FDG-PET/CT accuracy was determined in the subgroups of patients with NVE and prosthetic valve endocarditis (PVE)/ascending aortic prosthesis infection (AAPI). Associations between inflammatory infiltrate patterns and 18F-FDG-PET/CT uptake were investigated in an exploratory ad hoc histological analysis. RESULTS: Among 188 patients with PVE/AAPI, the sensitivity, specificity, and positive and negative predictive values of 18F-FDG-PET/CT focal uptake were 93%, 90%, 89%, and 94%, respectively, while among 115 patients with NVE, the corresponding values were 22%, 100%, 100%, and 66%. The inclusion of abnormal 18F-FDG cardiac uptake as a major criterion at admission enabled a recategorization of 76% (47/62) of PVE/AAPI cases initially classified as "possible" to "definite" IE. In the histopathological analysis, a predominance of polymorphonuclear cell inflammatory infiltrate and a reduced extent of fibrosis were observed in the PVE group only. CONCLUSIONS: Use of 18F-FDG-PET/CT at the initial presentation of patients with suspected PVE increases the diagnostic capability of the modified Duke criteria. In patients who present with suspected NVE, the use of 18F-FDG-PET/CT is less accurate and could only be considered a complementary diagnostic tool for a specific population of patients with NVE.
