Abstract 2101A: CNX-2006, a novel irreversible epidermal growth factor receptor (EGFR) inhibitor, selectively inhibits EGFR T790M and fails to induce T790M-mediated resistance in vitro

Introduction: EGFR mutant lung cancers are highly sensitive to first generation EGFR tyrosine kinase inhibitors (TKIs; gefitinib and erlotinib), but resistance eventually develops. In the majority of patients, such acquired resistance (AR) is mediated by a second-site T790M “gatekeeper” mutation. Second generation TKIs (e.g. afatinib, dacomitinib, neratinib) are more potent but have minimal efficacy as single agents in patients with AR. Notably, second generation TKIs still inhibit the wild type receptor which limits dose escalation, and similar to gefitinib/erlotinib, they still select for T790M-mediated resistance in vitro due to selectivity for drug-sensitive mutations (exon 19 deletions and L858R) compared to T790M. In this study, we characterize the inhibitory properties of CNX-2006, a novel irreversible EGFR TKI developed to inhibit specifically T790M. CNX-2006 is the prototype for CO-1686 which is currently in phase I clinical trials for the treatment of EGFR-mutant lung cancer. Methods and Results: CNX-2006 potency was assessed via multiple methods: i) surrogate kinase assays using immunoblotting of lysates from 293 cells transfected with cDNAs encoding various EGFR mutations and treated with drug, ii) growth inhibition assays using various erlotinib-resistant EGFR mutant cell lines or engineered Ba/F3 cells, and iii) H1975 xenografts expressing EGFR(L858R/T790M). CNX-2006 exhibited specificity and potent in vitro and in vivo activity against T790M. The drug also showed activity against uncommon EGFR mutations including G719S, L861Q, an exon 19 insertion mutant (I744-K745insKIPVAI), and T854A, but not an exon 20 insertion (H773-V774HVdup). In an in vitro resistance model, CNX-2006 significantly inhibited the emergence of resistant cells compared with erlotinib (1/24 vs 11/24, p = 0.0013). Chronic exposure to escalating doses of CNX-2006 failed to select for and/or enhance T790M-mediated resistance using PC-9 or HCC827 cells (both harboring exon 19 deletions), or PC-9/ER and HCC827/ER cells with existing T790M and resistance to erlotinib. In PC-9 and HCC827 cells with acquired resistance to CNX-2006, MET amplification or other known mutations in the RAS/MEK/ERK signaling pathway have not been detected. Conclusions: These results demonstrate that the profile of CNX-2006 is different from first and second generation EGFR TKIs. CNX-2006 selectively targets T790M, shows activity against other common and uncommon EGFR mutations, and does not select for r