A randomized phase II study of capecitabine-based chemoradiation with or without bevacizumab in resectable locally advanced rectal cancer.

3571 Background: The addition of bevacizumab (BEV) to capecitabine (CAP)-based chemoradiation (CRT) has shown encouraging efficacy in locally advanced rectal cancer (LARC), in nonrandomized studies. This randomized phase II study investigated the effect of adding BEV to preoperative CAP-based CRT in patients (pts), with LARC. Methods: The primary end point was pathologic complete response (pCR). A two-stage design was used. Assuming a minimum pCR rate of at least 15% in one of the arms, a difference between the two arms of 10%, and accepting a probability of correct selection of 87%, 41 pts per arm were needed. Patients with LARC (Stages II-III assessed by MRI) and ECOG PS <2 were randomized to concurrent radiotherapy 45Gy/25f/5 weeks + CAP (825mg/m²/bid) + BEV every 2 weeks (5 mg/kg for 3 doses) (arm A) or the same schedule without BEV (arm B). Surgery was scheduled 6-8 weeks after completing CRT. Results: 90 pts were randomized (arm A/B: 44/46). Patient’s characteristics were well balanced between both arms: male 61%, median age 62 years, median distance from anal verge 7 cm, T3 79%, N+ 87%. 40 (91%)/43 (93%) of pts (arm A/B) finalized the planned CRT + surgery treatment. Overall grade 3-4 toxicity rates were 18 % and 13% (arm A/B, p=0.50); no grade 3-4 hematological toxicity was reported. Postoperative complications were 19(43%)/17(37%)(arm A/B). Efficacy data on patients who actually underwent surgery are reported in the table. Conclusions: The addition of BEV to CAP-based preoperative CRT has shown to be feasible and safe in the local control of LARC. No differences in pCR were observed and longer follow-up is needed to assess the impact on survival endpoints. [Table: see text]