Objectives: Metabolic tumor volume (MTV) on 18F-fluorodeoxyglucose-positron emission tomography/computed tomography (FDG-PET/CT) is a promising prognostic predictor in pancreatic ductal adenocarcinoma (PDAC). However, the optimal segmentation method and threshold value to determine MTV for PDAC are still unclear. We explored the optimal method and threshold value for the prognostic value of MTV measured on pre-treatment 18F-FDG-PET/CT. Methods: Seventy-three patients with resected PDAC who underwent 18F FDG-PET/CT before surgical resection were enrolled. MTV values of the tumor were measured on FDG-PET/CT by the two fixed-threshold methods using threshold values as 2.0, 2.5, 3.0, and 3.5 for the absolute method and 35%, 40%, 42%, 45%, and 50% for the relative method. Receiver operating characteristic curve analysis for prediction of 1-year survival rates was conducted for determining the optimal threshold values, and we selected the optimal method and threshold value considering area under the curve. The prognostic values of each FDG-PET/CT parameter for disease-specific survival and recurrence-free survival were assessed with Kaplan–Meier method and Cox proportional hazard models. Results: In receiver operating characteristic curve analysis, MTV by the fixed-absolute threshold method based on a threshold value of 3.5 (MTV3.5) performed best in our study with area under the curve 0.724, sensitivity of 65%, and specificity of 75%. In univariate and multivariate analyses, MTV3.5 was significantly associated with disease-specific and recurrence-free survival. Conclusions: MTV3.5 by absolute threshold on pre-treatment FDG-PET/CT was the best independent prognostic predictor in resectable PDAC compared with other absolute threshold values and relative threshold values.
ASJC Scopus subject areas