Negative hyperselection and mechanisms of acquired resistance to first-line chemotherapy plus anti-EGFR in patients (pts) with pMMR RAS/BRAF wild-type (wt) metastatic colorectal cancer (mCRC): A translational analysis of the TRIPLETE trial.

3572 Background: Anti-EGFRs plus doublets are the standard first-line treatment for patients with left-sided RAS/BRAFwt mCRC. Anti-EGFR sensitivity seems restricted to negatively hyperselected RAS/BRAFwt cases, according to PRESSING1/2 panels. Analysing circulating tumor DNA in liquid instead of tissue biopsies is a promising strategy to optimize upfront therapy in RAS/BRAF wt mCRC pts and to identify mechanisms of acquired resistance. TRIPLETE is a phase III trial where 435 pts with untreated RAS/BRAF wt – per local assessment - mCRC were randomized to receive first-line FOLFOX/panitumumab (pan) or mFOLFOXIRI/pan. Methods: Patients enrolled in the TRIPLETE trial with available formaline-fixed paraffin embedded (FFPE) tissue and plasma at baseline (BP), and plasma at the time of disease progression (PDP) were included. FFPE tissue samples were profiled using FoundationOne CDx and Oncomine Focus Assay, and plasma samples were analyzed by means of Oncomine Pan-Cancer Cell-Free Assay. Patients with pMMR tumors were grouped as PRESSING-positive or negative based on the detection of alterations included in the PRESSING panel. Results: 134 (31%) pts had available next-generation sequencing (NGS) data from FFPE tissue, BP and PDP. FFPE tissue analysis revealed locally unknown RAS mutations in 6 (4%) cases. After excluding RAS mutated and dMMR/unknown cases (N=26), RAS was found mutated in 2 BP out of 102 pts (2%) with RAS/BRAF wt and pMMR tumors. PRESSING alterations were found in 15 (15%) FFPE cases ( HER2ampl: 9, HER2mut: 3, AKT1mut: 1, PIK3CAex20: 2). Matching results were achieved in 8 ( HER2 ampl: 6; HER2 mut: 2) BP, and 2 additional alterations were found ( AKT1mut and PTENmut). No differences between PRESSING-positive and negative patients were reported in terms of ORR, PFS or OS, regardless of primary tumor location. In PDP, RAS mut were detected in 9 pts (9%), and BRAFV600E mut co-occurred in 2 cases. Patients with RAS mutated ctDNA at PDP showed a significative shorter OS compared to those with RAS wt ctDNA [HR 0.3 95% CI 0.13-0.9, p=.03]. As additional mechanism of acquired resistance, MAP2K1 mut were found at PD in 2 cases, while no other PRESSING alteration was found. Conclusions: Inconsistently with previous data, negative hyperselection both through tissue and plasma analysis failed to demonstrate a prognostic role among pMMR RAS/ BRAF wt mCRC pts treated with first-line chemotherapy plus pan. The confounding effect of the associated chemotherapy backbone is a potential explanation for this finding. FFPE tissue and plasma analyses at baseline failed to provide fully concordant results. The occurrence of RAS mutations in ctDNA at the time of PD was significantly correlated with worse post-progression survival.