Predicting pathological response to neoadjuvant chemoradiotherapy in locally advanced rectal cancer using ¹⁸FDG-PET/CT

Background. Pathologic complete response (pCR) after neoadjuvant chemoradiation (CRT) has been observed in 15-30% of patients with locally advanced rectal cancer (LARC). The objective of this study was to determine whether PET/CT can predict pCR and disease-free survival in patients receiving CRT with LARC. Methods. This is a retrospective review of patients with EUS-staged T3-T4, N + rectal tumors treated with CRT, who underwent pre/post-treatment PET/CT from 2002-2009. All patients were treated with CRT and surgical resection. Standardized uptake value (SUV) of each tumor was recorded. Logistic regression was used to analyze the association of pre-CRT SUV, post-CRT SUV, %SUV change, and time between CRT and surgery, compared with pCR. Kaplan-Meier estimation evaluated significant predictors of survival. Results. Seventy patients (age 62 years; 42M:28F) with preoperative stage T3 (n = 61) and T4 (n = 9) underwent pre- and post-CRT PET/CT followed by surgery. The pCR rate was 26%. Median pre-CRT SUV was 10.8, whereas the median post-CRT SUV was 4 (P = 0.001). Patients with pCR had a lower median post-CRT SUV compared with those without (2.7 vs. 4.5, P = 0.01). Median SUV decrease was 63% (7.5-95.5%) and predicted pCR (P = 0.002). Patients with a pCR had a greater time interval between CRT and surgery (median, 58 vs. 50 days) than those without (P = 0.02). Patients with post-CRT SUV <4 had a lower recurrence compared with those without (P = 0.03). Patients with SUV decrease ≥63% had improved overall survival at median follow-up of 40 months than those without (P = 0.006). Conclusions. PET/CT can predict response to CRT in patients with LARC. Posttreatment SUV, %SUV decrease, and greater time from CRT to surgery correlate with pCR. Post-CRT, SUV <4, and SUV decrease ≥63% were predictive of recurrence-free and overall survival.