Combined EGFR/MET or EGFR/HSP90 inhibition is effective in the treatment of lung cancers codriven by mutant EGFR containing T790M and MET

Abstract

Tyrosine kinase inhibitors (TKI) that target the EGF receptor (EGFR) are effective in most non-small cell lung carcinoma (NSCLC) patients whose tumors harbor activating EGFR kinase domain mutations. Unfortunately, acquired resistance eventually emerges in these chronically treated cancers. Two of the most common mechanisms of acquired resistance to TKIs seen clinically are the acquisition of a secondary "gatekeeper" T790M EGFR mutation that increases the affinity of mutant EGFR for ATP and activation of MET to offset the loss of EGFR signaling. Although up to one-third of patient tumors resistant to reversible EGFR TKIs harbor concurrent T790M mutation and MET amplification, potential therapies for these tumors have not been modeled in vivo. In this study, we developed a preclinical platform to evaluate potential therapies by generating transgenic mouse lung cancer models expressing EGFR-mutant Del19-T790M or L858R-T790M, each with concurrent MET overexpression. We found that monotherapy targeting EGFR or MET alone did not produce significant tumor regression. In contrast, combination therapies targeting EGFR and MET simultaneously were highly efficacious against EGFR TKI-resistant tumors codriven by Del19-T790M or L858R-T790M and MET. Our findings therefore provide an in vivo model of intrinsic resistance to reversible TKIs and offer preclinical proof-of-principle that combination targeting of EGFR and MET may benefit patients with NSCLC.

Figure 1. Induction of hMET tyrosine kinase at both RNA and protein levels in bitransgenic mice from different founders

(A) Bitransgenic mice were fed a diet without (−) or with (+) doxycycline. After 12 weeks of doxycycline treatment, mRNA from the lungs from two founders (#31 and #16) was subjected to RT-PCR analysis to determine the transcriptional level of hMET. (B) Quantitative RT-PCR analysis of total MET transcript induction in bitransgenic mice fed a diet without (−) or with (+) doxycycline for 12 weeks. Each sample was amplified in triplicate for quantification of both total MET and β-actin transcripts. The endogenous mouse met level from normal mice was arbitrarily designated as 1. Data were analyzed by relative quantitation using the ΔΔCt method with normalization to β-actin. Error bars, S.D. (C) Bitransgenic mouse founders were treated without (-) or with (+) doxycycline for 12 weeks, and lung lysates were subjected to Western blotting for total MET and β-actin. (D) H&E images of normal lung from a CCSP-rtTA/MET bitransgenic mouse fed with a doxycycline-containing diet for 52 weeks. The right micrograph is a magnified view of the area indicated in the low magnification micrograph shown on the left. Scale bars indicate 1000 μm (left) or 100 μm (right), respectively. There was no evidence of malignant tumor observed in six CCSP-rtTA/MET bitransgenic mice that were maintained on a doxycycline-containing diet for more than 40 weeks. (E,F) Quantitative RT-PCR analysis of total hEGFR (E) and hMET (F) transcript in bi- or tri-transgenic mice on a doxycycline-containing diet. Each sample was amplified in duplicate for quantification of EGFR, MET and β-actin transcripts. Mean expression level of human MET or human EGFR from TD/MET mouse was arbitrarily designated as 1. Data were analyzed by relative quantitation using the ΔΔCt method and normalization to β-actin. Each bar represents the relative quantity of mRNA in one mouse.

Figure 2. Histological comparison of the murine mutant EGFR lung tumors with or without co-expression of hMET

Representative hematoxylin and eosin (H&E) staining of cross-sectional views of different transgenic mouse lungs showing no significant difference in histopathologic assessments of lung carcinoma. Picture on the right for each genotype is a magnified view of the area indicated in the picture with low magnification on the left. Scale bars 100μm.

Figure 3. Combination therapy of WZ4002/crizotinib or WZ4002/17-DMAG shows tumor regression in the mice with concurrent hEGFR del19/T790M mutation and MET overexpression

Tumor-bearing mice were subjected to MRI imaging prior to and after two weeks of treatment with the indicated drugs. Representative MRI images show tumor regression in TD mice but not in TD/MET mice after 2 weeks of WZ4002 treatment. Tumors from either TD or TD/MET mice also do not respond to treatment with single agent crizotinib or 17-DMAG. However, the WZ4002/crizotinib or WZ4002/17-DMAG combinations lead to tumor regression.

Figure 4. Waterfall plots of percentage change in tumor area for individual mice

Tumor-bearing mice were treated daily with WZ4002 (50mg/kg) (A), crizotinib (20mg/kg or 50mg/kg, as indicated) (B), WZ4002 (50mg/kg) plus crizotinib (20mg/kg or 50mg/kg, as indicated) (C), 17-DMAG (20mg/kg) (D), or WZ4002 (50mg/kg) plus 17-DMAG (20mg/kg) (E). Tumor area was documented by MRI imaging prior to and after two weeks of treatment. Each color represents a different mouse colony; each bar represents the % tumor area change in one mouse.

Figure 5. H&E staining of lung tumors demonstrates that combination treatments are more efficacious that individual treatments against tumors harboring compound EGFR mutations and MET overexpression

Representative images of H&E staining of lung tumors from TD mice (A), TD/MET mice (B), TL mice (C), and TL/MET mice (D). All mice were sacrificed for pathological analysis after 2 weeks of treatment with vehicle, WZ4002, crizotinib, 17-DMAG, WZ4002/crizotinib, or WZ4002/17-DMAG. Scale bars, 100μm.

Figure 6. IHC staining of lung tumors shows combination treatments are more efficacious than individual agents for suppression of critical signaling pathways in mice with lung cancers harboring EGFR T790M mutation and MET overexpression

IHC for phospho-EGFR (left), phospho-Akt (S473, middle), and phospho-Erk (T202/Y204, right) of lung tumors from (A) TD or (B) TD/MET mice. Mice were treated with the indicated drugs (single agent or combination) for 2 doses over 24 hours and sacrificed 2 hours following the second dose, and lung sections were stained with indicated antibodies. Photos shown are representative fields in each group in low and high magnification. Scale bars in the main images and in the insets indicate 100μm and 50μm, respectively.