Improvement of diagnostic accuracy of 18fluorine-fluorodeoxyglucose PET/computed tomography in detection of infective endocarditis using a 72-h low carbs protocol

Background The major challenge in diagnosing infective endocarditis using 18fluorine-fluorodeoxyglucose (18F-FDG) PET/computed tomography (CT) is the high physiological myocardial 18F-FDG uptake, which may interfere with the detection of pathological uptake. Our aim was to retrospectively analyse the impact of a low-carbohydrate, high-protein, high-fat (LCHPHF) diet starting 72 h before the examination with classic overnight fasting. Methods We included 92 patients (51 with LCHPHF diet 72 h before the examination and 41 with overnight fast only). Left ventricular (LV) myocardial 18F-FDG uptake was visually evaluated, to estimate the effect of LCHPHF on myocardial 18F-FDG uptake, we used a three-point visual scale: 0, negligible uptake and evaluable examination; 1, mild uptake and doubtful examination and 2, intense uptake and not evaluable examination. SUVmax and SUVmean were calculated putting a region of interest on LV myocardium at the level of the largest cardiac footprint and on the site of suspected lesion. Thus, a ratio between SUV of the suspect lesion and SUV of LV was calculated as SUV ratio. Results By visual assessment, in LCHPHF diet group 46 patients (96.1%) had score 0, 3 patients (5.9%) had score 1 and nobody had score 2. In control group, 23 patients (56.1%) had a score 2, 12 (29.3%) patients obtained a score 1 with a doubtful examination and 6 patients (14.6%) were classified with a score 0 with reportable examination. Conclusion We demonstrated that an LCHPHF diet can adequately suppress the physiological myocardial uptake leading to a significant improvement in the interpretability and diagnostic accuracy of 18F-FDG PET/CT in infective endocarditis