Abstract 5655: AP26113 possesses pan-inhibitory activity versus crizotinib-resistant ALK mutants and oncogenic ROS1 fusions

AP26113 is a potent, reversible inhibitor of ALK fusions and mutant (but not native) EGFR. To overcome mutation-based resistance, AP26113 was designed to maintain activity against crizotinib-resistant ALK variants such as the gatekeeper ALK mutant L1196M. In an ongoing phase 1 dose-escalation study, AP26113 has achieved steady-state trough concentrations in excess of 1 μM and demonstrated promising clinical activity in both crizotinib-resistant and naïve ALK-positive NSCLC patients and preliminary evidence of activity in patients with mutant EGFR (#439O, ESMO 2012). To further assess the activity of AP26113 against crizotinib-resistant ALK mutants, and the structurally related ROS1 fusions recently identified in NSCLC, we engineered Ba/F3 cells to express the appropriate oncogenic drivers and evaluated their sensitivities to crizotinib and AP26113. Ba/F3 cell lines expressing clinically-identified EML4-ALK mutants (T1151T insertion, S1206Y, D1203N, L1196M, G1202R, F1174C and C1156Y) had substantially reduced sensitivity to crizotinib (viability IC50= 363-1296 nM) compared with native EML4-ALK (IC50= 137 nM). These data are consistent with the identification of these mutants in patients with acquired crizotinib resistance. AP26113 potently inhibited both native (IC50= 21 nM) and crizotinib-resistant ALK mutants (IC50= 26-254 nM) at concentrations substantially below the clinically achievable trough levels of AP26113 (1 μM), suggesting it may possess a pan-ALK inhibitory profile. AP26113 also effectively inhibited the viability of Ba/F3 cells expressing CD74-ROS1 (IC50= 18 nM), FIG-ROS1 (IC50= 31 nM), SDC4-ROS1 (IC50= 16 nM) and EZR-ROS1 (IC50= 41 nM) thus demonstrating that it is an equipotent inhibitor of ALK and ROS1. In a Ba/F3 CD74-ROS1 xenograft model, AP26113 inhibited tumor growth in a dose-dependent manner, with 50 mg/kg AP26113 inducing substantial tumor regression. Since ROS1 and ALK are structurally related and similarly sensitive to crizotinib, we reasoned that acquired drug resistance in ROS1-driven tumors may also occur via mutation. We therefore generated Ba/F3 cells driven by L2026M gatekeeper mutant forms of CD74-ROS1 and FIG-ROS1 and tested their drug sensitivity. Interestingly, the inhibitory capacity of AP26113 was unaffected by L2026M. In contrast, crizotinib potencies were reduced approximately 4-fold. In conclusion, we have demonstrated that AP26113 inhibits clinically relevant crizotinib- resistant ALK mutants and oncogenic ROS1 fus