Larotrectinib, a selective TRK tyrosine kinase inhibitor (TKI), has demonstrated histology-agnostic efficacy in patients with TRK fusion-positive cancers. While responses to TRK inhibition can be dramatic and durable, duration of response may eventually be limited by acquired resistance. LOXO-195 is a novel, selective TRK TKI designed to overcome acquired resistance mediated by recurrent kinase domain (solvent front and xDFG) mutations identified in multiple patients who have developed resistance to TRK TKIs. Activity against these acquired mutations was confirmed in enzyme and cell-based assays and in vivo tumor models. As clinical proof of concept, the first two patients with TRK fusion-positive cancers that developed acquired resistance mutations on larotrectinib were treated with LOXO-195 on a first-in-human basis, utilizing rapid dose titration guided by pharmacokinetic assessments. This approach led to rapid tumor responses and extended the overall duration of disease control achieved with TRK inhibition in both patients.

Introduction: Novel rearranged in transfection (RET)-specific tyrosine kinase inhibitors (TKIs) such as selpercatinib (LOXO-292) have shown unprecedented efficacy in tumors positive for RET fusions or mutations, notably RET fusionpositive NSCLC and RET-mutated medullary thyroid cancer (MTC). However, the mechanisms of resistance to these agents have not yet been described.

No abstract

In May 2020, selpercatinib became the first FDA-approved selective RET inhibitor, indicated for patients (pts) with RET fusion-positive NSCLC and thyroid cancer as well as RET-mutant medullary thyroid cancer. Despite the durable activity of selpercatinib, pts can eventually develop acquired resistance. Previous studies in pts treated with selpercatinib or pralsetinib have reported the shared emergence of recurrent RET G810 mutations at the solvent front of the ATP pocket, which lead to a steric clash and loss of binding potency for both drugs. In some pts, these RET G810 solvent front mutations have been observed to co-occur with RET V804 gatekeeper mutations, demonstrating the importance of identifying agents that can maintain potency against both solvent front and gatekeeper resistance mutations. While multi-kinase inhibitors with potency against RET G810 mutations have been reported, these agents lack activity against V804 mutations and also carry off-target toxicities that further limit efficacy. We have identified a series of potent and selective next-generation RET inhibitors to address the emerging unmet need of pts who relapse on selective RET inhibitors. To assess potency, we evaluated the molecules in HEK293 cell lines engineered to express an M918T RET mutation or a KIF5B-RET fusion, as well as G810S or V804M resistance mutations. One compound, LOX-18228, exhibited nanomolar potency across the M918T RET, KIF5B-RET wild type, KIF5B-RET G810S, KIF5B-RET V804M, and KIF5B-RET V804M/G810S cell lines, with cellular IC50 values of 1.2, 0.9, 5.8, 31 and 51 nM, respectively. LOX-18228 also demonstrated high selectivity for RET when compared to a broad enzyme panel of off-target kinases, followed by relevant cellular assay screening. LOX-18228 also showed robust selectivity against a diverse panel of receptors, transporters and enzymes at a concentration of 10 μM, and exhibited a hERG IC50 of >30 μM. LOX-18228 demonstrated in vitro ADME properties predictive of good in vivo exposure with low predicted intrinsic clearance (range, 2.1-6.6 µL/min/million cells) in human, mouse, rat, and dog hepatocytes, and a measured high bin permeability of Papp = 15 (10-6 cm/s) in an intestinal MDCKII permeability assay. As predicted, LOX-18228 demonstrated high oral exposures in mouse and rat. In a pt-derived xenograft (PDX) model harboring a CCDC6-RET G810S mutation, LOX-18228 demonstrated complete regression at doses ≥30 mg/kg. Similarly, in a PDX model harboring a CCDC6-RET V804M mutation, LOX-18228 exhibited dose-dependent tumor inhibition with 100% tumor growth inhibition observed at 60 mg/kg. These data suggest that LOX-18228, as well as closely related compounds, represent promising next-generation RET inhibitor candidates that could be used to further extend durable disease control for pts with RET-altered cancers following the development of acquired resistance to current agents. An IND is planned for 2021. Citation Format: Gabrielle R. Kolakowski, Erin D. Anderson, Joshua A. Ballard, Barbara J. Brandhuber, Kevin R. Condroski, Eliana B. Gomez, Thomas C. Irvin, Manoj Kumar, Nisha A. Patel, Faith D. Watson, Steven W. Andrews. Pre-clinical characterization of potent and selective next-generation RET inhibitors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 1464.

Phosphoinositide 3-kinase alpha (PI3Kα) H1047R mutations are activating oncogenic events that occur in ~15% of advanced breast cancers. While there is one PI3Kα inhibitor FDA-approved for patients with PI3Kα-mutated breast cancer, and many others in clinical development, all of these agents inhibit wild-type PI3Kα and its mutated form with approximate equal potency. As a result, their efficacy is limited by toxicities associated with on target wild-type PI3Kα inhibition, notably hyperglycemia as well as cutaneous and GI toxicity. LOX-22783 is a highly potent, mutant-selective and brain-penetrant allosteric PI3Kα H1047R inhibitor. Here, we describe the preclinical profile of LOX-22783. H1047R selectivity was measured using biochemical kinase activity and cell-titer Glo and signal transduction assays. Tumor growth inhibition, pharmacokinetic and pharmacodynamic effects were assessed in in vivo studies using xenograft and patient-derived xenograft (PDX)-models. LOX-22783 inhibited growth and signaling responses in multiple H1047R-driven breast cancer cell lines and demonstrated high selectivity for H1047R mutated PI3Kα (EC50 values <5 nM) relative to wild-type PI3Kα (EC50 >250 nM) as well as the other wild-type PI3K isoforms (beta, gamma, and delta, all EC50 >250nM). In enzyme and cell-based assays, LOX-22783 dissociated from PI3Kα H1047R at a slower rate (3-6 hrs) compared to alpelisib (≤10 mins), potentially allowing for extended inhibition of PI3Kα H1047R by LOX-22783. LOX-22783 also normalized the EGF-stimulated membrane-localization of PI3Kα H1047R while alpelisib did not. LOX-22783 was highly kinome-selective when assayed at 3 µM, with no inhibitory activity on 17 lipid kinases or 374 protein kinases. In preclinical species, LOX-22783 demonstrated high oral bioavailability, including exposure in the CNS, a common site of metastases for patients with breast cancer. In vivo, LOX-22783 demonstrated dose-dependent tumor regression in H1047R breast cancer models without inducing hyperglycemia or other toxicities. Tumor pharmacodynamic analyses confirmed successful pathway inhibition. At doses resulting in 90% pathway inhibition, tumor regressions of ≥60% were observed. This wide therapeutic index is predicted to allow for maximizing dose intensity and efficacy in patients, without wild-type PI3Kα inhibition limiting target coverage for the H1047R mutant form. These data demonstrate that LOX-22783 potently and selectively inhibits mutant H1047R, but not wild-type PI3Kα, or other PI3K isoforms. LOX-22783 binds to an allosteric pocket distinct from the ATP binding site used by the approved and investigational PI3Kα inhibitors. We hypothesize that this profile will lead to differentiated efficacy and tolerability for patients with PI3Kα H1047R-mutated cancers, with the additional potential to address brain metastases. An IND submission is planned for 2022. Citation Format: Anke Klippel, Rui Wang, Loredana Puca, Andrew Lee Faber, Weihua Shen, Shripad V. Bhagwat, Kannan Karukurichi, Feiyu Fred Zhang, Carmen Perez, Ramon Rama, Ana Ramos, Yi Zheng, Zahid Bonday, James Thomas, Harold B. Brooks, Lisa J. Kindler, Sarah M. Bogner, Parisa Zolfaghari, Mark Hicks II, Sophie Callies, Brian Mattioni, Laurie LeBrun, Jim Durbin, Erin Anderson, Chris Mayne, Edward Kesicki, Gabrielle Kolakowski, Steven W. Andrews, Barbara J. Brandhuber. Preclinical characterization of LOX-22783, a highly potent, mutant-selective and brain-penetrant allosteric PI3Kα H1047R inhibitor [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2021 Oct 7-10. Philadelphia (PA): AACR; Mol Cancer Ther 2021;20(12 Suppl):Abstract nr P142.