Discovery and characterization of selective, FGFR1 sparing, inhibitors of FGFR2/3 oncogenic mutations for the treatment of cancers

Targeting FGFR genetic alterations using small molecule inhibitors is a validated therapeutic strategy for urothelial carcinoma and cholangiocarcinoma. However, the current FDA-approved pan-FGFR inhibitors, erdafitinib and pemigatinib, are subject to FGFR1-mediated dose-limiting toxicities (e.g., hyperphosphatemia). These treatments necessitate a high rate of dose reductions, interruptions, and discontinuations, thereby potentially limiting efficacy. In addition, drug-resistant mutations (e.g., gatekeeper) in FGFR2 and FGFR3 genes rapidly emerge in patients treated with these drugs. Our research goals are to reveal the full spectrum of oncogenic FGFR2 and FGFR3 mutations that drive tumor growth and to discover an inhibitor that selectively targets these mutations together with FGFR2 and FGFR3 gene fusion and drug-resistance mutations, while minimizing FGFR1 activity and associated toxicities. We hypothesize that this will deliver an FGFR precision medicine with enhanced anti-tumor activity, an improved drug resistance profile, and broader mutational coverage. Applying the Mutation-Allostery-Pharmacology (MAP) platform technology developed by Black Diamond Therapeutics, we define a spectrum of 20 allosteric FGFR2/3 oncogenic mutations, including 10 previously uncharacterized mutations that we now show to be oncogenic. We demonstrated how this mutation family is activated due to disulfide-bond mediated dimerization; hence we term these mutations locked-dimer (Lo-Di) FGFR oncogenes. Herein, we report the discovery of a series of orally available, selective Lo-Di-FGFR2/3 inhibitors that 1) shows antiproliferative potency across all 20 mutations; 2) spares FGFR1-wild-type; 3) is active against gatekeeper mutations and 4) shows favorable selectivity versus a subset of closely related kinases in the human kinome. When dosed orally, one example was well tolerated and exhibited dose-dependent PK/PD and anti-tumor efficacy and regression in the UM-UC-14 xenograft model in mice. Our data support the development of rationally designed selective inhibitors targeting a spectrum of FGFR2/3 mutations while sparing dose limiting FGFR1 activity.