Abstract
1838
Objectives To determine the ability of FDG PET in predicting pathologic response and survival in patients with esophageal cancer after pre-operative chemo-radiation therapy (CRT).
Methods Forty patients (37 men, 3 women; median age 56.5 years) with stage II-IVA esophageal squamous cell carcinoma who received CRT followed by esophagectomy between 2002 and 2009 were retrospectively reviewed. FDG PET was performed before and after CRT in all patients. PET variables including pre-CRT maximum SUV, post-CRT maximum SUV, percentage change in maximum SUV, and pre-CRT lymph node status were correlated with pathologic response and overall survival.
Results FDG PET showed complete response in 16 patients, partial response in 19 patients, and equivocal result in 5 patients. Pathologic evaluation revealed complete response (defined as viable tumor of less than 10%) in 17 patients, partial response in 21 patients, and equivocal result in 2 patients. The sensitivity, specificity, accuracy, positive predictive value and negative predictive value of PET in predicting pathologic complete response were 69%, 70%, 70%, 60% and 78% respectively. At a median follow-up of 33.7 months (range 4.2-96.1 months), there was no significant difference in overall survival between patients with PET complete response and PET partial response. PET variables were not predictive of pathologic response. The receiver operating characteristic (ROC) analysis could not identify a threshold value of PET variables for predicting overall survival.
Conclusions Our study does not support previous reports that FDG PET predicts overall survival in patients with locally advanced esophageal cancer. Further evaluation in larger trial is indicated