The diagnosis of infective endocarditis (IE) in prosthetic valves and intracardiac devices is challenging because both the modified Duke criteria (DC) and echocardiography have limitations in this population. The added value of 18 F-fluorodeoxyglucose ( 18 F-FDG) positron emission tomography (PET)/computed tomography (CT) and 18 F-FDG PET/CT angiography (PET/CTA) was evaluated in this complex scenario at a referral center with a multidisciplinary IE unit.
Ninety-two patients admitted to our hospital with suspected prosthetic valve or cardiac device IE between November 2012 and November 2014 were prospectively included. All patients underwent echocardiography and PET/CT, and 76 had cardiac CTA. PET/CT and echocardiography findings were evaluated and compared, with concordant results in 54% of cases (κ=0.23). Initial diagnoses with DC at admission, PET/CT, and DC+PET/CT were compared with the final diagnostic consensus reached by the IE Unit. DC+PET/CT enabled reclassification of 90% of cases initially classified as possible IE with DC and provided a conclusive diagnosis (definite/rejected) in 95% of cases. Sensitivity, specificity, and positive and negative predictive values were 52%, 94.7%, 92.9%, and 59.7% for DC; 87%, 92.1%, 93.6%, and 84.3% for PET/CT; and 90.7%, 89.5%, 92%, and 87.9% for DC+PET/CT. Use of PET/CTA yielded even better diagnostic performance values than PET/nonenhanced CT (91%, 90.6%, 92.8%, and 88.3% versus 86.4%, 87.5%, 90.2%, and 82.9%) and substantially reduced the rate of doubtful cases from 20% to 8% ( P <0.001). DC+PET/CTA reclassified an additional 20% of cases classified as possible IE with DC+PET/nonenhanced CT. In addition, PET/CTA enabled detection of a significantly larger number of anatomic lesions associated with active endocarditis than PET/nonenhanced CT ( P =0.006) or echocardiography ( P <0.001).