Bevacizumab with or without erlotinib as maintenance therapy in patients with metastatic colorectal cancer (GERCOR DREAM; OPTIMOX3): a randomised, open-label, phase 3 trial

Background The combination of an anti-VEGF or an anti-EGFR-targeted monoclonal antibody with chemotherapy has shown clinical activity in patients with metastatic colorectal cancer. However, combining both anti-VEGF and anti-EGFR antibodies with chemotherapy in first-line treatment resulted in adverse outcomes. We assessed whether the combination of erlotinib, an EGFR tyrosine kinase inhibitor, with bevacizumab could increase the efficacy of maintenance therapy in patients with unresectable metastatic colorectal cancer. Methods This randomised, open-label, phase 3 study was undertaken in 49 centres in France, Austria, and Canada. Eligible patients were aged 18-80 years with histologically confirmed, unresectable metastatic colorectal cancer, WHO performance status 0-2, had received no previous therapy for metastatic disease, and had adequate organ function. Patients without disease progression after bevacizumab-based induction therapy were randomly assigned (1: 1) by a minimisation technique to bevacizumab (7.5 mg/kg every 3 weeks) or bevacizumab plus erlotinib (150 mg once daily) as maintenance therapy until progression. All patients were stratified by centre, baseline performance status, age, and number of metastatic sites. The primary endpoint was progression-free survival on maintenance therapy analysed by intention to treat. We report the final analysis. This trial is registered with ClinicalTrials.gov, number NCT00265824. Findings Between Jan 1, 2007, and Oct 13, 2011, 700 eligible patients were enrolled; following induction treatment, patients without disease progression were randomly assigned to bevacizumab (n=228) or bevacizumab plus erlotinib (n=224). At the final analysis, median follow-up was 51.0 months (IQR 36.0-60.0) in the bevacizumab group and 48.3 months (31.5-61.0) in the bevacizumab plus erlotinib group. In the primary analysis (after 231 progression-free survival events), median progression-free survival from randomisation was 5.1 months (95% CI 4.1-5.9) in the bevacizumab plus erlotinib group compared with 6.0 months (4.6-7.9) in the bevacizumab group (stratified hazard ratio [HR] 0.79 [95% CI 0.60-1.06]; p=0.11; unstratified HR 0.76 [0.59-0.99]; p=0.043). In the final analysis, median progression-free survival from randomisation was 5.4 months (95% CI 4.3-6.2) in the bevacizumab plus erlotinib group compared with 4.9 months (4.1-5.7) in the bevacizumab group (stratified HR 0.81 [95% CI 0.66-1.01], p=0.059; unstratified HR 0.78 [0.68-0.96], p=0.019). At the final analysis, median overall survival from maintenance was 24.9 months (95% CI 21.4-28.9) in the bevacizumab plus erlotinib group and 22.1 months (19.6-26.7) in the bevacizumab group (stratified HR 0.79 [95% CI 0.63-0.99], p=0.036; unstratified HR 0.79 [0.64-0.98], p=0.035). The most frequent grade 3-4 adverse events were skin rash (47 [21%] of 220 patients in the bevacizumab plus erlotinib group vs none of 224 patients in the bevacizumab alone group), diarrhoea (21 [10%] vs two [<1%]), and asthenia (12 [5%] vs two [<1%]). Interpretation Maintenance bevacizumab plus erlotinib might be a new non-chemotherapy-based maintenance option for the first-line treatment of patients with unresectable metastatic colorectal cancer after bevacizumab-based induction therapy.