Background: PARADIGM is the first prospective trial to test the superiority of PAN vs. BEV in combination with standard doublet first-line chemotherapy for patients (pts) with RAS WT mCRC and left-sided primary tumors. Methods: This open-label, multicenter trial in Japan (NCT02394795) randomly selected pts with chemotherapy-naive RAS WT mCRC to PAN + mFOLFOX6 or BEV + mFOLFOX6. Overall survival (OS) as primary endpoint was hierarchically tested in patients with left-sided tumors, followed by those in the full-analysis set (FAS) population. Key secondary endpoints included progression-free survival (PFS), response rate (RR), and curative resection (R0) rate. Results: From May 2015 to June 2017, 823 pts were randomized; 12 did not receive protocol treatment and 9 were excluded due to major deviation of inclusion criteria. A total of 400 pts received PAN and 402 pts received BEV as FAS; 312 and 292 pts had left-sided primary tumors, respectively. OS was analyzed after 448 OS events in left-sided pts with a median follow-up of 61 months. PAN significantly improved OS vs. BEV in both populations: left-sided (HR, 0.82; 95.798% CI, 0.68-0.99, p = .031, which crossed the boundary of significance [0.042]), and FAS (HR, 0.84; 95% CI, 0.72-0.98; p = .030, with < 0.05 as the boundary). Although PFS was comparable between treatment groups, RR and R0 resection rates were higher with PAN compared with BEV (Table). HR for OS in the right-sided population was 1.09. No new safety signal was observed. Conclusions: PAN significantly improved OS vs. BEV in combination with mFOLFOX6 in pts with RAS WT and left-sided mCRC, establishing a standard first-line combination regimen for this population. Clinical trial information: NCT02394795.

^aConfidence intervals: 95.798% CI for OS in left-sided pts; 95% CI for other brackets.