Responsiveness of different MET tumour alterations to type I and type II MET inhibitors

Abstract

Background: Mutations in c-MET receptor tyrosine kinase (MET) can be primary oncogenic drivers of multiple tumour types or can be acquired as mechanisms of resistance to therapy. MET tyrosine kinase inhibitors (TKIs) are classified as type I or type II inhibitors, with the former binding to the DFG-in, active conformation of MET, and the latter to the DFG-out, inactive conformation of MET. Understanding how the different classes of MET TKIs impact tumours with varied MET alterations is critical to optimising treatment for patients with MET altered cancers. Here, we characterise MET mutations identified in patients' tumours and assess responsiveness to type I and II TKIs.

Methods: We used structural modelling, in vitro kinase and in cell-based assays to assess the response of MET mutations to type I and II TKIs. We then translated our pre-clinical findings and treated patients with MET mutant tumours with selected inhibitors.

Results: We detected the emergence of four (three previously uncharacterised and one known) MET resistance mutations (MET^G1090A, MET^D1213H, MET^R1227K and a MET^Y1230S) in samples from patients with multiple solid tumours, including patients who had been previously treated with type I inhibitors. In silico modelling and biochemical assays across a variety of MET alterations, including the uncharacterised MET^G1090A and the MET^Y1230S substitutions, demonstrated impaired binding of type I but not of type II TKIs (i.e., cabozantinib/foretinib). Applying our pre-clinical findings, we then treated two patients (one with a non-small-cell lung cancer and one with a renal cell carcinoma) whose tumours harboured these previously uncharacterised MET alterations with cabozantinib, a type II MET TKI, and observed clinical responses.

Conclusions: Comprehensive characterisation of the sensitivity of mutations to different TKI classes in oncogenic kinases may guide clinical intervention and overcome resistance to targeted therapies in selected cases.

Key points: Kinase mutations in RTKs are primary or secondary drivers in multiple cancer types Some of these mutations confer resistance to type I but not to type II inhibitors in preclinical samples and in patients The biochemical characterization of mutations in oncogenic kinases based on their sensitivity to type I and type II inhibitors is crucial to inform clinical intervention.

Keywords: MET; resistance to targeted therapy; type I and type II TKIs.

FIGURE 1

Identification of previously uncharacterised MET tyrosine kinase inhibitor (TKI)‐resistance mutations in patients. (A) Relevant clinical treatment history of a patient (Pt. 1) with a stage IV, PLEKHA6‐NTRK1 fusion‐positive, MET amplified cholangiocarcinoma. Pt. 1 was treated with the combination of selitrectinib and crizotinib to which she responded. Six months later, she developed resistance. Sequencing of the ctDNA by MSK‐ACCESS at progression detected the presence of 13 new MET mutations, four of which (highlighted in red) revealed novel alterations including a MET^G1090A, a MET^D1213H, a MET^R1227K and a MET^Y1230S substitution. The variant allele frequency (VAF) for each alteration is indicated. (B) Clinical treatment history of a patient (Pt. 2) with stage IV, CD47‐MET fusion‐positive non‐small‐cell lung cancer (NSCLC). The timeline reports just those regimens that are relevant for this study. The emergence of the MET^G1090A found on a pleural effusion sample on crizotinib progression is indicated. Computed tomography (CT) scans pre‐crizotinib, on‐crizotinib and following crizotinib progression are presented. POD, progression of disease; PR, partial response. (C) Clinical history of a patient (Pt. 3) with a MET^Y1230S mutant renal cell carcinoma (RCC). Targeted sequencing results performed by MSK‐IMPACT on the tumour are listed.

FIGURE 2

Sensitivity of MET resistance mutations to type I and type II MET inhibitors. (A) Computational prediction of change in binding free energy (ΔΔG) of MET resistance mutations to representative type I (crizotinib) and type II (cabozantinib) MET inhibitors using FEP/MBAR. The change in free energy (ΔΔG) is highlighted in different colours (Green: ΔΔG = ±.6 kcal/mol; Orange: ΔΔG = .6–2.0 kcal/mol; Red: ΔΔG > 2.0 kcal/mol). (B) Representative models showing type I and type II MET inhibitors in complex with MET WT and MET G1090A. Bound inhibitors (colour sticks) and kinase residue Y1230 (sphere) are displayed. Chemical groups of inhibitors that clash with mutant MET kinase are depicted as spheres for visualisation purposes. (C) In vitro kinase assays showing the activity of the type I MET inhibitors crizotinib and capmatinib and the type II MET inhibitors cabozantinib and foretinib against recombinant WT and mutant (MET^G1090A, MET^Y1230S, MET^Y1230H) MET kinases. (D) Western blot images depicting the activity of crizotinib or cabozantinib against HEK‐293T cells transfected with CD47‐MET, CD47‐MET, MET^G1090A and CD47‐MET, MET^Y1230S‐encoding plasmids. (E) NanoBret analysis following transfection of HEK‐293T cells with CD47‐MET, CD47‐MET, MET^G1090A, CD47‐MET, MET^Y1230S and CD47‐MET, MET^Y1230H‐encoding plasmids and treatment with increasing concentrations of the type I MET tyrosine kinase inhibitor (TKI) crizotinib or the type II MET TKI cabozantinib.

FIGURE 3

(A) Representative models showing type I and type II MET inhibitors in complex with MET WT, MET R1227K and MET D1213H kinases. Bound inhibitors (colour sticks) and relevant kinase residues (sphere) are displayed. (B–E) Western blot images (B, D) and their quantifications (C, E) depicting the activity of crizotinib or cabozantinib (B, C) or capmatinib and foretinib (D, E) against HEK‐293T cells transfected with CD47‐MET, CD47‐MET, MET^R1227K and CD47‐MET, MET^D1213H‐encoding plasmids.

FIGURE 4

(A) List of patients identified whose tumours harbour MET mutations predicted to be sensitive to type II MET inhibitors. (B) PET scans showing the response of Pt. 2 (CD47‐MET, MET^G1090A mutant non‐small‐cell lung cancer [NSCLC]) to cabozantinib. (C) Computed tomography (CT) scans showing the response of Pt. 3 (MET^Y1230S mutant renal cell carcinoma [RCC]) to cabozantinib.

FIGURE 5

Graphic summarising steps taken to conduct this study: from the initial identification of MET mutations in clinical samples, to their pre‐clinical and biochemical characterisation, to the final design of ad hoc treatments for patients based on our laboratory findings.