P90.04 RAS Precision Medicine Trans-Atlantic Partnership: Multi-Centre Pooled Analysis of RAS Pathway Mutations in Advanced NSCLC

Abstract

The most common histological subtype of non-small cell lung cancer (NSCLC) is adenocarcinoma (50-55% of patients), of which the most common oncogenic driver is KRAS mutation (∼30% internationally). Genetic data has altered the taxonomy of NSCLC over the past 10 years, highlighting subgroups such as EGFR mutation, which have improved prognosis and are susceptible to targeted therapies. Despite its frequency, targeting RAS has historically been limited by pre-clinical and clinical failures. Emerging preclinical/clinical data suggests there is value looking beyond RAS mutation, analysing subtypes- represented by a limited number of variations in its mutational isoforms, codons and alleles. Through international collaboration our partnership aims to extend existing knowledge of RAS precision medicine in NSCLC, evaluating biological, clinical and treatment effects at subtype level. For this analysis, the first 309/384 patients with stage IIIb/IV RAS- and/or NF1-mutant NSCLC were identified from the Christie NHS Foundation Trust and the Gustave Roussy Cancer Centre between August 2008 and July 2020. DNA was extracted from archival FFPE samples, plasma or both to identify mutations using targeted next generation sequencing. Molecular, clinical, pathological and outcome data were collected on all patients and Kaplan-Meier survival analysis was performed to test for survival differences between subtypes based on treatment arms. For clinical characteristics, Mann-Whitney and Fisher’s exact tests were used for statistical comparisons in continuous and dichotomous datasets, respectively. Of 309 patients analysed, 151 patients were from the Christie NHS Foundation Trust and 158 patients from the Gustave Roussy Cancer Centre. Mean age was 63, range 19-92. 30 patients (10%) had stage IIIb disease with the remainder stage IV, 292 patients (94.5%) demonstrated non-squamous histology vs. 16 (5%) squamous vs 1 (0.5%) adenosquamous. Median PD-L1 status across the cohort was 20%. 209 patients (68%) had a mutation identified by tissue analysis, 38 patients (12%) by plasma and 62 (20%) by both. The most common RAS mutation identified was KRAS in 259 patients (84%), followed by NRAS in 13 patients (4%) and HRAS in 7 patients (2%). Amongst the KRAS mutant population, 111 pts (43%) harboured a mutation in G12C, followed by 52 (20%) G12D and 41 (16%) G12V. Codon 61 was most commonly mutated within this subgroup 16/259 (6%). 40 patients (13%) were identified as carrying a mutation in NF1; 35 non-squamous histology, 4 squamous and 1 adenosquamous. Progression free survival following first line chemotherapy plus immunotherapy or immunotherapy alone was variable across KRAS-, NRAS- and NF1-mutant subgroups (8.3 vs. 14.9 vs. 10.2 months; p=0.86). Median PD-L1 status was found to be 63% vs. 100% vs. 79% respectively. Analysis of RAS codon and allelic subgroups, including their therapeutic vulnerabilities beyond immunotherapy, will be presented following further analysis. The RAS precision medicine Trans-Atlantic partnership is designed to evaluate prognostic/predictive value of RAS and NF1 mutations in advanced NSCLC, focusing specifically on signal seeking for therapeutic vulnerabilities in mutant RAS isoforms, codons, and allelic subtypes. Ongoing analysis is planned to expand the dataset and inform the optimal sequence of therapy for subtypes of RAS- and NF1-mutant patients.