A Next-Generation TRK Kinase Inhibitor Overcomes Acquired Resistance to Prior TRK Kinase Inhibition in Patients with TRK Fusion-Positive Solid Tumors

Abstract

Larotrectinib, a selective TRK tyrosine kinase inhibitor (TKI), has demonstrated histology-agnostic efficacy in patients with TRK fusion-positive cancers. Although responses to TRK inhibition can be dramatic and durable, duration of response may eventually be limited by acquired resistance. LOXO-195 is a 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 2 patients with TRK fusion-positive cancers who 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.Significance: LOXO-195 abrogated resistance in TRK fusion-positive cancers that acquired kinase domain mutations, a shared liability with all existing TRK TKIs. This establishes a role for sequential treatment by demonstrating continued TRK dependence and validates a paradigm for the accelerated development of next-generation inhibitors against validated oncogenic targets. Cancer Discov; 7(9); 963-72. ©2017 AACR.See related commentary by Parikh and Corcoran, p. 934This article is highlighted in the In This Issue feature, p. 920.

Figure 1. TRK inhibitor binding to acquired resistance mutations

A, structural models showing steric interactions (circled red dashes) between larotrectinib (upper) and solvent-front (i.e. G595R, G623R) or xDFG (i.e. G667C, G696A) resistance mutations in TRKA and TRKC. In contrast, LOXO-195 (lower) can accommodate solvent front and xDFG resistance mutations. B, IC₅₀ values for each agent in purified kinase assays at K_M ATP are shown as mean +/− standard deviation (SD) of the indicated replicates. The ratio of IC₅₀ values for each mutant IC₅₀ to the corresponding unmutated kinase is shown in parentheses. The IC₅₀ values for LOXO-195 for TRKC WT and G696A were below the lower limit of the assay (2.5 nM).

Figure 2. Effect LOXO-195 and larotrectinib on TRK cellular activity

A, NIH 3T3 cells expressing the indicated TRK proteins were treated with LOXO-195 or larotrectinib for one hour, followed by quantification of cellular phospho-TRK levels by ELISA (for ΔTRKA, ΔTRKA-G595R and ΔTRKA-G667C) or cellular phospho-ERK levels by flow cytometry (for ETV6-TRKC, ETV6-TRKC-G623R and ETV6-TRKC G696A). IC₅₀ (nM) values are displayed as mean +/− SD of at least 3 replicates. B, KM12 (TPM3-TRKA) xenografts or NIH 3T3 ΔTRKA (−/+ G595R or G667C) allografts following oral treatment with larotrectinib or LOXO-195 at the indicated doses (n = 7 animals per treatment group). Data represent mean tumor volume +/− Standard Error of the Mean (SEM).

Figure 3. Clinical testing strategy and proof of concept of LOXO-195 activity against acquired resistance mutations

A, Design of single patient clinical protocols. B, Measured C_max and C_min (upper and lower black symbols, respectively) for larotrectinib and LOXO-195 for each patient are displayed relative to patient-specific TRK protein target coverage: TRKA/TRKC (green rectangles), TRKA-G595R/TRKC-G623R (red rectangles). The lower edge of each rectangle = IC₅₀ for the indicated inhibitor/target, the upper edge = IC₉₀. xDFG (i.e. TRKA-G667C, TRKC-G696A) mutant target coverages (blue rectangles) are included for comparison. C, Upper CT-PET axial images from adult patient with LMNA1-NTRK1 G595R colorectal cancer demonstrating reduction in hypermetabolic liver and adnexal masses with LOXO-195 treatment. Lower decrease in NTRK1 (TRKA) G595R cell-free DNA allelic fraction preceded the decrease in tumor burden by RECIST 1.1. D, Post-contrast axial T2 MRI images from the pediatric patient with ETV6-NTRK3 G623R MASC before (left) and after 28 (right) and 66 (lower) days of LOXO-195 treatment.

Figure 4. Traditional vs. PK-Guided, Accelerated Phase 1 Design

Traditional Phase 1 “3+3” dose escalations studies enroll between 3–6 patients at each dose level based on observed toxicity with dose increments becoming smaller as dose increases (top). This conservative design, developed in the cytotoxic era, risks under-dosing patients with validated targets. In a first-in-human Phase 1 study of a rationally designed agent against a validated target, the risk of sub-therapeutic dosing may represent a greater harm to patients. For this situation, we propose a PK-Guided, accelerated Phase 1 design (bottom). This design combines multiple efficiencies including: 1) biomarker selection during dose escalation, 2) PK-guided rapid intra-patient dose escalation during a single cycle of therapy, 3) sequentially escalating starting dose and, 4) early dose expansion upon identification of a therapeutic dose.