RESUMO
BACKGROUND: The current study sought to demonstrate the advantages offered by fluorine 18-fluorodeoxyglucose ((18)F-FDG) positron emission tomography/computed tomography (PET/CT) in patients supported with continuous-flow left ventricular assist devices (CF-LVADs) in detecting infection and the consequent effect on clinical decisions. METHODS: Between April 2009 and September 2013, 40 PET examinations were performed in 31 patients (78.1% men; mean age, 51.0 ± 14.9 years) supported with a CF-LVAD. In group A (19 examinations in 14 patients), PET/CT was performed to detect infectious focus in patients without external signs of driveline involvement but with at least two of the following infection signs: recurrent fever, positive blood culture, or elevated infectious indicators. In group B (21 examinations in 17 patients), PET/CT aimed to assess the internal extension of infection in patients with external signs of driveline infection. RESULTS: In 50% of the cases of the patients in group A, abnormal (18)F-FDG uptake (9 patients) was related to VAD components. Matching the results with the final diagnosis, we reported 9 true-positive, 8 true-negative, no false-negative, and 2 false-positive findings. New information unrelated to VAD was found in 9 patients (50%): pneumonia in 3, colon diverticulitis in 3, sternal dehiscence in 1, paravertebral abscess in 1, and erysipelas in 1. In group B, superficial abnormal (18)F-FDG uptake was found at the piercing site of the driveline in 2 patients, deeper extension of infection along the driveline in 10, initial involvement of the pump housing in 2, and full involvement of the device in 4. These findings contributed to changing the clinical management in 84.2% of group A patients and in 85.7% of group B patients: 16 patients were scheduled for urgent transplantation, 2 underwent surgical revision of the driveline, 7 required prolonged antibiotic therapy, and 3 underwent colonoscopy. CONCLUSIONS: This single-center experience highlights the diagnostic value of PET/CT in detecting the localization and internal extension of infection to internal VAD components. Moreover, this information notably influences the therapeutic management.