The Diagnostic Accuracy of 18F-Fluorodeoxyglucose Positron Emission Tomography/Computed Tomography for Infections in...

Infection remains a major complication in patients receiving contemporary left ventricular assist devices (LVADs). This retrospective single-center study evaluated the diagnostic accuracy of 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) for detecting different types of infections. Seventy-five scans from 40 HeartMate 3 patients performed before May 2024 were included and blindly assessed by an experienced nuclear medicine specialist. Both visual and semi-quantitative analyses of 18F-FDG uptake were performed. Compared with the final clinical diagnosis established by a multidisciplinary team, the sensitivity of 18F-FDG PET/CT for driveline and central device infections was 96% (95% confidence interval [CI], 81-99) and 81% (95% CI, 48-97), with corresponding specificities of 48% (95% CI, 29-67) and 37% (95% CI, 23-53), respectively. In semi-quantitative analyses, a cut-off maximum standardized uptake value (SUVmax) of 4.15 provided optimal accuracy for driveline infections (sensitivity 78%, specificity 80%), whereas a cut-off SUVratio-liver of 1.29 yielded the best accuracy for central device infections (sensitivity 73%, specificity 69%). 18F-fluorodeoxyglucose uptake was also noted at various device regions in patients without infection or with non-mechanical circulatory support-specific infections, with false-positive rates of 80% and 63%, respectively. 18F-fluorodeoxyglucose PET/CT shows high sensitivity but considerable false positives and should be used as a complementary tool within a multidisciplinary framework for infection management.