Diagnostic Accuracy of 18F-FDG PET/CT in Infective Endocarditis and Implantable Cardiac Electronic Device Infection: A Cross-Sectional Study.

Abstract

A prospective study with 80 consecutive patients with suspected IE and ICED infection (65 men and 15 women with a mean age of 68 ± 13 y) between June 2013 and May 2015 was performed in our hospital. The inclusion criteria were clinically suspected IE and ICED infection at the following locations: native valve (NV) (n = 21), PV (n = 29), or ICED (n = 30) (automatic implantable defibrillator [n = 11] or pacemaker [n = 19]). Whole-body 18F-FDG PET/CT with a myocardial uptake suppression protocol with unfractionated heparin was performed in all patients. The final diagnosis of infection was established by the IE Study Group according to the clinical, echocardiographic, and microbiologic findings. A final diagnosis of infection was confirmed in 31 patients: NV (n = 6), PV (n = 12), and ICED (n = 13). Sensitivity, specificity, positive predictive value, and negative predictive value for 18F-FDG PET/CT were 82%, 96%, 94%, and 87%, respectively. 18F-FDG PET/CT was false-negative in all cases with infected NV. 18F-FDG PET/CT was able to reclassify 63 of 70 (90%) patients initially classified as possible IE by modified Duke criteria. In 18 of 70 cases, 18F-FDG PET/CT changed possible to definite IE (26%) and in 45 of 70 cases changed possible to rejected IE (64%). Additionally, 18F-FDG PET/CT identified 8 cases of septic embolism and 3 of colorectal cancer in patients with a final diagnosis of IE. 18F-FDG PET/CT proved to be a useful diagnostic tool in suspected IE and ICED infection and should be included in the diagnostic algorithm for early diagnosis. 18F-FDG PET/CT is not useful in the diagnosis of IE in NV but should be also considered in the initial assessment of this complex scenario to rule out extracardiac complications and possible neoplasms.