Randomized Phase 2 Study Comparing Pathological Responses of Resected Colorectal Cancer Metastases after Bevacizumab with mFOLFOX6 or FOLFIRI (BEV-ONCO Trial)

Simple Summary Nowadays, the surgery of liver metastases remains the only hope of a cure for patients with colorectal cancer. Pathological responses evaluated after preoperative treatment strongly influences the risk of relapse and patient survival. Previous studies reported that preoperative bevacizumab combined with an oxaliplatin-based chemotherapy provided a higher pathological response rate compared with an irinotecan-based regimen or chemotherapy alone. This prospective trial, having recruited 65 patients with resectable colorectal liver metastases, ambitioned to report a higher major pathological response rate after mFOLFOX6-bevacizumab compared to FOLFIRI-bevacizumab. Among the 57 patients with 159 resected metastases, no difference in major pathological response rate was observed between treatments. Nevertheless, the trial prospectively confirmed the pathological response of resected colorectal liver metastases as a significant biomarker for tumor recurrence, justifying its implementation in clinical practice. Interestingly, we observed that the homogeneity of the pathological response and histological growth pattern of liver metastases was also strongly associated with patient's survival. Retrospective studies reported that preoperative oxaliplatin-based chemotherapy increased pathological response (PR) in patients resected for colorectal liver metastases (CRLM). This multicenter prospective randomized (1/1) phase II trial evaluated PR on resected CRLM after preoperative mFOLFOX6 (arm A) or FOLFIRI (arm B) + bevacizumab. The primary endpoint was the major pathological response rate (MPRR), defined as the percentage of patients presenting CRLMs with mean tumor regression grade (TRG) < 3. Secondary endpoints included safety, progression-free survival (PFS) and overall survival (OS). Out of 65 patients, 57 patients (28 and 29 in arm A/B) were resected for CRLM (one patient with lung metastases). Clinical and treatment characteristics were similar in both arms. One-month postoperative complications were 39.3%/31.0% in arm A/B (p = 0.585). MPRR and complete PR were 32.1%/20.7% (p = 0.379) and 14.3%/0.0% (p = 0.052) in arm A/B, respectively. PFS and OS were not different. Patients with PR among all CRLMs (max TRG <= 3; 43.8% of patients) had a lower risk of relapse (PFS: HR = 0.41, 95%CI = 0.204-0.840, p = 0.015) and a tendency towards better survival (OS: HR = 0.34, 95%CI = 0.104-1.114, p = 0.075). The homogeneity of PR was associated with improved PFS/OS. This trial fails to demonstrate a significant increase in MPRR in patients treated with mFOLFOX6-bevacizumab but confirms PR as an important prognostic factor.