Randomised, Phase 2 Study (OLIVIA) of Bevacizumab Plus mFOLFOX6 or Folfoxiri in Patients with Initially Unresectable Colorectal Cancer Liver Metastases

Abstract

Background: Most patients (75‐90%) with metastatic colorectal cancer are not candidates for upfront surgical resection. However, a subset of patients with unresectable colorectal cancer liver‐only metastases (CLMs) may become resectable after downsizing by chemotherapy and biologic therapy. Although biologics are thought to improve overall response rate (ORR), the optimal combination of a biologic and chemotherapy for resectability remains uncertain. Methods: OLIVIA was a multinational open‐label phase II study in which patients with unresectable CLMs were randomised to bevacizumab (BEV) plus mFOLFOX6 or FOLFOXIRI q2w. Resectability was assessed by interdisciplinary review. Unresectability was defined as ≥1 of the following: no possibility of upfront R0/R1 resection of all hepatic lesions; <30% estimated residual liver after resection; or disease in contact with major vessels of the remnant liver. The primary endpoint was the overall resection rate (R0/R1/R2). Results: From October 2008 to December 2011, 80 patients were randomised to mFOLFOX6‐BEV (n = 39) or FOLFOXIRI‐BEV (n = 41). Baseline patient/disease characteristics in the mFOLFOX6‐BEV vs FOLFOXIRI‐BEV arms, respectively, were as follows: male (46% vs 71%), age ≥60 years (36% vs 63%), ECOG PS of 1 (23% vs 37%), and ≥5 target CLMs (49% vs 49%). Median duration of therapy was 191 vs 231 days in the two arms, respectively. Overall resection rate, ORR, and progression‐free survival (PFS) data are shown in the Table. Grade ≥3 adverse events occurred in 84% and 95% of patients receiving mFOLFOX6‐BEV and FOLFOXIRI‐BEV, respectively, and included neutropenia (35% vs 48%; febrile, 8% vs 13%) and diarrhoea (14% vs 28%). Conclusion: The results suggest that FOLFOXIRI‐BEV improves resection rates, ORR, and PFS vs mFOLFOX6‐BEV in patients with initially unresectable CLMs. Chemotherapy‐ and BEV‐related AEs occurred with the expected incidence and were manageable. FOLFOXIRI‐BEV should be evaluated further as an effective regimen to downsize CLMs. (Table Presented).