Change of maximum standardized uptake value slope in dynamic triphasic [18F]-fluorodeoxyglucose positron emission tomography/computed tomography distinguishes malignancy from postradiation inflammation in head-and-neck squamous cell carcinoma: a prospective trial.

Abstract

To evaluate dynamic [(18)F]-fluorodeoxyglucose (FDG) uptake methodology as a post-radiation therapy (RT) response assessment tool, potentially enabling accurate tumor and therapy-related inflammation differentiation, improving the posttherapy value of FDG-positron emission tomography/computed tomography (FDG-PET/CT). We prospectively enrolled head-and-neck squamous cell carcinoma patients who completed RT, with scheduled 3-month post-RT FDG-PET/CT. Patients underwent our standard whole-body PET/CT scan at 90 minutes, with the addition of head-and-neck PET/CT scans at 60 and 120 minutes. Maximum standardized uptake values (SUV(max)) of regions of interest were measured at 60, 90, and 120 minutes. The SUV(max) slope between 60 and 120 minutes and change of SUV(max) slope before and after 90 minutes were calculated. Data were analyzed by primary site and nodal site disease status using the Cox regression model and Wilcoxon rank sum test. Outcomes were based on pathologic and clinical follow-up. A total of 84 patients were enrolled, with 79 primary and 43 nodal evaluable sites. Twenty-eight sites were interpreted as positive or equivocal (18 primary, 8 nodal, 2 distant) on 3-month 90-minute FDG-PET/CT. Median follow-up was 13.3 months. All measured SUV endpoints predicted recurrence. Change of SUV(max) slope after 90 minutes more accurately identified nonrecurrence in positive or equivocal sites than our current standard of SUV(max) ≥2.5 (P=.02). The positive predictive value of post-RT FDG-PET/CT may significantly improve using novel second derivative analysis of dynamic triphasic FDG-PET/CT SUV(max) slope, accurately distinguishing tumor from inflammation on positive and equivocal scans.