Characterization of 1,387 NSCLCs with MET exon 14 (METex14) skipping alterations (SA) and potential acquired resistance (AR) mechanisms.

Abstract

9511 Background: METex14 SA are oncogenic drivers in NSCLC. Due to the numerous sites around ex14 that bind the spliceosome complex, many variations can result in deleterious alterations (alts). We present a comprehensive overview of these ex14 SA across 1,387 NSCLCs and characterized potential AR mechanisms. Methods: Hybrid-capture based comprehensive genomic profiling (CGP) was performed on samples from 60,495 NSCLC patients (pts). A scoring system was applied leveraging our large database of samples with METex14 SA to optimize accurate reporting of these variants. Paired samples were collected ≥ 60 days apart (median 462). Results: 1,393 METex14 SA were identified in samples (1,278 tissue, 109 circulating tumor DNA (ctDNA)) from 1,387 NSCLC pts (2.3%) spanning multiple functional sites: donor (42%), acceptor (4.7%), poly-pyrimidine tract (15%), acceptor and polypyrimidine tract (13%), D1010 (23%), Y1003 (2.1%), and whole exon deletions (0.3%). 6 samples (5 tissue, 1 ctDNA) harbored 2 METex14 SA, each including a mutation (mut) at the donor or acceptor site. MDM2 and CDK4 amplifications (amps) co-occurred in 32% and 19% of METex14 samples, respectively, but were more common with polypyrimidine tract (37% and 23%) vs donor site (32%, p = 0.07 and 18%, p = 0.07) alts. MET co-amp was present in 12% of cases and frequency did not significantly differ by functional site. 66 (4.8%) cases (57 tissue, 9 ctDNA) had known NSCLC co-drivers, including KRAS (68%) and EGFR (14%) mut, a subset of which may represent AR. Paired samples with a METex14 SA in the 1st sample were available for 36 pts. The METex14 SA was detected in the 2nd sample for 32 pts, excluding 3 with low ctDNA. 22/36 (61%) had reportable acquired alts detected including 9 with ≥1 acquired MET muts [D1228X (4), Y1230X (3), Y1003F (1), D1228A/E/H + L1195V (1)] and 3 with acquired MET amp. Other acquired alts included ERBB2 amp and mut (1 each), EGFR ex19ins (1), KRAS amp (1), PIK3CA mut (1), AKT2 amp (1) and others with unknown functional significance. Potential AR alts were present with primary METex14 SA spanning all functional sites. Conclusions: In a dataset of > 60,000 advanced NSCLCs, METex14 SA were present in 2.3% of cases, and represented 6 major subtypes. Among paired cases, potential AR mechanisms included secondary MET alts (33%), and acquired alts in EGFR, ERBB2, KRAS, and PI3K pathways. Acquired alts were independent of the type of METex14 SA. Characterizing common co-occuring may be critical for predicting responses to MET inhibitors and informing rational combination strategies.