Diagnostic accuracy of visual analysis versus dual time-point imaging with 18F-FDG PET/CT for the characterization of indeterminate pulmonary nodules with low uptake.

Abstract

OBJECTIVE\nTo determine the accuracy of visual analysis and the retention index (RI) with dual-time point 18F-FDG PET/CT for the characterization of indeterminate pulmonary nodules (IPN) with low FDG uptake.\n\n\nMATERIALS AND METHODS\nA retrospective analysis was performed on 43 patients (28 men, 64 ± 11 years old, range 36-83 years) referred for IPN characterization with 18F-FDG-PET/CT and maximum standard uptake value ≤ 2.5 at 60 minutes post-injection (SUVmax1). Nodules were analyzed by size, visual score for FDG uptake on standard (OSEM 2,8) and high definition (HD) reconstructions, SUVmax1, SUVmax at 180 minutes post-injection (SUVmax2), and RI was calculated. The definitive diagnosis was based on histopathological confirmation (n = 28) or ≥ 2 years of follow-up.\n\n\nRESULTS\nTwenty-four (56%) nodules were malignant. RI ≥ 10% on standard reconstruction detected 18 nodules that would have been considered negative using the standard SUVmax ≥ 2.5 criterion for malignancy. RI ≥ 10% had a sensitivity, specificity, PPV, NPV and accuracy of 75, 73.7, 78.3, 70, and 74.4%, respectively, while for FDG uptake > liver on HD these were 79.1, 63.2, 73.1, 70.6, and 72.1%, respectively. SUVmax1 ≥ 2, SUVmax2 > 2.5 and FDG uptake > liver on standard reconstruction had a PPV of 100%. FDG uptake > mediastinum on HD had a NPV of 100%.\n\n\nCONCLUSIONS\nRI ≥ 10% was the most accurate criterion for malignancy, followed by FDG uptake > liver on HD reconstruction. On standard reconstruction, SUVmax1 ≥2 was highly predictive of malignancy, as well as SUVmax2 > 2.5 and FDG uptake > liver. FDG uptake < mediastinum on HD was highly predictive of benign nodules.