O-021 Tailored strategy for locally-advanced rectal carcinoma: preliminary results of a phase II multicenter trial (GRECCAR 4)

Abstract

Purpose: locally‐advanced rectal carcinomas (LARC) raise the issues of oncologic control and therapeutic morbidity. Preoperative chemoradiotherapy (CRT) followed by total mesorectal excision has become the standard in LARC treatment. We designed a multicenter randomized trial to evaluate the feasibility of a tailored management of LARC according to the early tumor response to a short and intensive induction trichemotherapy (TCT), respecting a minimal 90% R0 resection rate in all arms. Methods: Tumor response after induction TCT (FOLFIRINOX) was evaluated using magnetic resonance imaging. Good responders had a >75% reduction of tumor volume and were randomized between immediate surgery (arm A) versus standard CRT (Cap 50) plus surgery (arm B). Poor responders had a predictive CRM<1 or a volumetric response <75%, and were randomized in CRT (Cap50) (arm C) versus intensive CRT (Cap 60) (arm D) both followed by surgery. Results: 206 patients with LARC (T3>c, T4, predictive CRM<1) were included between 05.11 and 10.14 in 16 French centers. Compliance to TCT was generally good (6 pts stopped CT). After TCT, 194 patients were evaluated, 30 (15%) were good responders and 164 (85%) showed unfavorable response. Once the required number of patients was reached in unfavorable stratum, randomization was stopped for this stratum but recruitment continued in favorable stratum. The trial was stopped before the required number of patients was reached in favorable stratum. Distribution of the population was: 16, 14, 113 and 51 in arms A, B, C and D. 133 Patients were analyzed according to the treatment received and the planned number in the unfavorable stratum: 11 arm A, 19 arm B, 52 arm C and 51 arm D. Good responders presented lower tumors (p<0.001) while poor responders were bulky (23cm3/45cm3;p<0.001). Among poor responders, 23% had a volumetric reduction <50% and 65% had a reduction between 50‐75%. The surgical procedure was similar between the 4 groups regarding hospitalization length, conversion rate, abdomino‐perinal resection rate, postoperative morbidity and re‐intervention. The R0 resection rates [IC95%] were 90% [61‐99], 100% [85‐100], 83% [71‐90], and 88% [77‐95] in arms A, B, C and D, respectively. Mesorectum resection was complete (grade III) for 78%, 81%, 78%, 76% patients in arms A, B, C and D, respectively. Patients had positive CRM<1 in 11%, 0%, 14% and 7%, and positive distal margins in 0%, 0%, 6%, 2%, respectively. Conclusions: Early response to induction CT (4 cycles of FOLFIRINOX) enables to adapt preoperative radiotherapy for LARC. A tailored strategy seems feasible in terms of safety and pathologic results for bad responders; despite a poor accrual, this strategy seems promising for good responders. Long‐term clinical results are needed to confirm the efficacy of such a tailored strategy. (Figure Presented).