Abstract

3507 Background: In the phase 3 PARADIGM trial (NCT02394795), adding panitumumab (PAN) to FOLFOX6 improved overall survival (OS) compared with bevacizumab (BEV) in patients (pts) with RASWT mCRC. In this exploratory analysis, we investigated the role of acquired gene alterations in pts who experienced progressive disease (PD) and the impact of these alterations on post-progression survival (PPS). Methods: PPS was defined as the interval from PD to death. Pre- and post-treatment cell-free DNA (cfDNA) was analyzed using the custom PlasmaSELECT-R 91 PGDx panel (NCT02394834). PFS, OS, and PPS were compared by patterns of acquired gene alterations. Results: Among the enrolled 802 pts, 390 discontinued treatment due to PD, of which 276 (70.8%) had evaluable pre- and post-treatment cfDNA samples (PAN, 126; BEV, 150). PAN pts with acquired RTK/RAS alterations had shorter PPS than those without alterations (13.2 vs 18.8 mo, HR 1.88 [95% CI: 1.28–2.76]) with the same trend in OS. BEV pts with acquired CIMP pathway alterations had shorter PPS than those without (14.9 vs 18.6 mo; HR 1.58 [1.09–2.29]); OS had similar results. Co-occurrence of gene alterations was more frequent in PAN vs BEV (overall 52.4% vs 43.3%, RTK/RAS gene co-alterations 25% vs 10%). Co-occurrence of gene alterations resulted in shorter PPS only in PAN pts. Conclusions: Distinctive patterns of acquired gene alterations were observed in PAN- and BEV-treated pts with PD. Co-occurrence of gene alterations, particularly in the RTK/ RAS pathway, was more prevalent with PAN than BEV. Clinical trial information: NCT02394834 .