Integrative genomic analysis of drug resistance in MET exon 14 skipping lung cancer using patient-derived xenograft models

Abstract

Background: Non-small cell lung cancer (NSCLC) driven by MET exon 14 skipping (METex14) occurs in 3-4% of NSCLC cases and defines a subset of patients with distinct characteristics. While MET targeted therapy has led to strong clinical results in METex14 patients, acquired drug resistance seemed to be unavoidable during treatment. Limited information is available regarding acquired resistance during MET targeted therapy, nor has there been any report on such patient-derived xenografts (PDXs) model facilitating the research.

Methods: We describe a patient case harboring METex14 who exhibited drug resistance after treatment with crizotinib. Subcutaneous xenografts were generated from pretreatment and post-resistance patient specimens. PDX mice were then treated with MET inhibitors (crizotinib and tepotinib) and EGFR-MET bispecific antibodies (EMB-01 and amivantamab) to evaluate their drug response in vivo. DNA and RNA sequencing analysis was performed on patient tumor specimens and matching xenografts.

Results: PDXs preserved most of the histological and molecular profiles of the parental tumors. Drug resistance to MET targeted therapy was confirmed in PDX models through in vivo drug analysis. Newly acquired MET D1228H mutations and EGFR amplificated were detected in patient-resistant tumor specimens. Although the mutations were not detected in the PDX, EGFR overexpression was observed in RNA sequencing analysis indicating possible off-target resistance through the EGFR bypass signaling pathway. As expected, EGFR-MET bispecific antibodies overcome drug resistant in the PDX model.

Conclusions: We detected a novel MET splice site deletion mutation that could lead to METex14. We also established and characterized a pair of METex14 NSCLC PDXs, including the first crizotinib resistant METex14 PDX. And dual inhibition of MET and EGFR might be a therapeutic strategy for EGFR-driven drug resistance METex14 lung cancer.

Keywords: EGFR-MET bispecific antibody; MET exon 14 skippings; TKI resistance; lung cancer; patient-derived xenograft.

Figure 1

Chest computed tomography (CT) scans and histopathology of patient tumor before and after crizotinib treatment. (A) CT scans of the primary tumor of patient at baseline (left), best response (middle) and progression disease (right) after crizotinib treatment. Red arrows point to malignant loci. (B) H&E staining and IHC staining of the tumor biopsies from patient before crizotinib treatment, which was CD56 negative, CK positive and partially positive for TTF1, P40 and NapsinA; the scale bar represents 40 μm. (C) H&E staining and IHC staining of tumor biopsy specimens from a patient in disease progression, which was negative for CD56, NapsinA and TTF-1, positive for CK and P40; the scale bar represents 40 μm.

Figure 2

Tumor-bearing mice and histopathology staining. (A) Representative images of tumor-bearing PDX mice established from patient pretreatment tumor (left) and drug-resistant tumor (right), respectively. (B) H&E staining and IHC staining of tumor derived from pretreatment PDX (passage 3), which was negative for CD56, NapsinA and TTF-1, and positive for CK and P40; the scale bar represents 40 μm. (C) H&E and IHC staining of tumor derived from drug- resistant PDX (passage 2), which is negative for CD56, NapsinA and TTF-1, and positive for CK and P40; the scale bar is 40 μm.

Figure 3

The effect of TKIs on two PDX models. (A) Tumor change in pretreatment PDX mice treated with vehicle, 50 mg/kg crizotinib and 10 mg/kg tepotinib once a day for 3 weeks. (B) Tumor change in drug resistant PDX mice treated with vehicle, 50 mg/kg crizotinib and 10 mg/kg tepotinib once a day for 5 weeks. Tumor volume was measured and expressed as the mean ± standard error.

Figure 4

Genetic and transcriptomic characterization of tumors from the patient and the corresponding PDX models. (A) Co-mutation plot of genomic alterations across tumor specimens from patient and PDX models. Mutations and frequency of each gene are listed in Supplementary Table S1 . (B) Sashimi plot of MET exon14 skipping event in patient and PDX tumor specimens. The Y axis indicates read density, the number on arc represents junction reads, and portion of schematic MET transcripts with exon number is the plot at the bottom. The figure was adapted from IGV (19). (C) Expression level and copy number of RTK related genes in patient and PDX model tumor specimens. Numbers are presented in TPM.

Figure 5

Representative pictures for EGFR IHC staining and the effect of EGFR/cMET bispecific antibody on drug resistant PDX model. (A) IHC staining of EGFR in patient-resistant sample. (B) IHC staining of EGFR in PDX-resistant sample. (C) Tumor change in drug resistant PDX mice treated with vehicle, 50mg/kg crizotinib once daily, 16mg/kg EMB-01 once a week, 16mg/kg Amivantamab twice a week for 3 weeks. Tumor volume was measured and expressed as the mean ± standard error. ***p< 0.001 by two-way ANOVA Bonferroni’s correction.