A randomized, double-blind, phase II study of first-line FOLFOX plus bevacizumab with onartuzumab versus placebo in patients with metastatic colorectal cancer (mCRC)

Background: In mCRC, MET overexpression has been associated with poor prognosis and resistance to anti‐VEGF therapy. We initiated a phase II study to evaluate the combination of onartuzumab (O), a ligand‐blocking monoclonal antibody directed against the MET receptor, plus bevacizumab and FOLFOX, in first‐line mCRC (GO27827; NCT01418222 ). Methods: This double‐blind, randomized, multicenter phase II study randomized patients 1:1 to receive O (10 mg/kg iv) or placebo (P), plus mFOLFOX6 and bevacizumab (5 mg/kg iv). Stratification was by prior adjuvant therapy. All treatments were given on day 1‐3 of a 2‐week cycle. Oxaliplatin was given for up to 8‐12 cycles; all other agents were continued until progression, unacceptable toxicity or death. Primary endpoint was progression‐free survival (PFS) in ITT and MET+ subgroup by immunohistochemistry (IHC). MET status was determined by central laboratory IHC evaluation, with scores of 2+ or 3+ considered MET+. Results: From September 2011 to November 2012, 194 patients were enrolled. A recommendation was made to stop O after an interim efficacy and safety analysis in September 2013, due to lack of efficacy. The final analysis (cut‐off Feb 2014) found that O did not improve PFS vs. P in the ITT (HR 0.75 [0.52‐1.08]; p=0.12) or MET IHC+ populations (n=79; HR 1.03 [0.56‐1.89]; p=0.93), although improvement was noted in the MET IHC‐ population (n=108; HR 0.60 [0.37‐0.97]; p=0.03). Neither overall survival (OS) nor response rate (RR) was improved with O vs. P in any of the groups (OS HR 0.96 [0.61‐1.50], p=0.85 for ITT; OS HR 1.24 [0.63‐2.43], p=0.54 for MET IHC+; OS HR 0.83 [0.44‐1.56], p=0.56 for MET IHC‐; RR 57.3% vs. 57.7% for ITT, 43.2% vs. 57.1% for MET IHC+, 66.1% vs. 60.8% for MET IHC‐). More edema (65.7% vs. 12.9%) and venous thromboembolic events (30.3% vs. 16.1%) were seen with O vs. P, respectively. Grade >3 events were similar (86.9% vs. 84.9%) and events leading to discontinuation were increased (48.5% vs. 37.6%) with O vs. P. Conclusions: Adding onartuzumab to FOLFOX/bevacizumab did not prolong PFS in first‐line unselected or MET IHC+ mCRC. A trend towards PFS benefit was seen in those with MET IHC‐ mCRC, contrary to prior reports in other tumor types.