Abstract 1451: Modeling patient-derived lung cancer in mice: Preclinical tool for drug development

Abstract

Abstract Background: Genotype directed treatment is the standard of care for patients with advanced non-small cell lung cancer (NSCLC). However, acquired drug resistance invariably develops. There is a need to develop clinically relevant animal models from patients who have developed acquired resistance in order to study mechanisms of resistance and validate new therapeutic strategies. We have generated a portfolio of clinically relevant, early passage, patient-derived xenograft (PDX) models of lung cancer. Methods: Tumor biopsies (core biopsies (n = 8), pleural effusions (n = 28) and surgical samples (n = 11)) were implanted subcutaneously (n = 39) or into the sub-renal capsule (n = 8) of NSG mice under an IRB approved protocol. After implanting 35 specimens, mostly from pleural effusions, several mice exhibited weight and fur loss. Histopathology revealed infiltration of the liver, skin and lungs by human derived T-cells consistent with graft-versus-host disease (GVHD). Subsequent pleural effusions were then subjected to CD45 depletion using a human CD45 microbead separation kit which eliminated the incidence of GVHD. Tumors from all three sources were allowed to grow and subsequently passaged only as subcutaneous implants in NSG mice. Molecular characterization of PDX models was performed using targeted next generation sequencing and tumor histology confirmed by H&E staining. Some established tumor models have also been propagated as cell lines in vitro. Results: 47 specimens were implanted with an overall take rate of ∼30%. The initial take rate was 22% but increased to 66% following CD45 depletion. We have established 11 NSCLC adenocarcinoma and 3 small-cell lung cancer (SCLC) PDX models from patient samples. The models established thus far range from erlotinib resistant EGFR mutant models with T790M mutation (n = 3), MET amplification (n = 1), SCLC transformation (n = 2) or an unknown mechanism (n = 1); to models including SCLC with an NRAS Q61K mutation (n = 1); EML4-ALK resistant to crizotinib and ceritinib (n = 1); KRAS G12C mutation (n = 1); NUT-mid line carcinoma with BRD4 rearrangement (n = 1); pan wild-type (n = 2) and a PIK3CA mutation (n = 1). We have used these models to study therapeutic interventions. Treatment of the NRAS Q61K mutant SCLC with trametinib resulted in complete tumor regression. Pharmacodynamic analyses revealed inhibition of pERK 1/2 and induction of apoptosis. Data will be presented with further characterization and platform development of the lung PDX models. Notably, >50% of the models do show pulmonary metastases from the subcutaneous implant site. Conclusions: PDXs can be established from lung cancer patients undergoing routine clinical biopsies. Establishment of models from pleural effusions is also feasible following CD45 depletion to prevent GVHD. These models will provide unique information on mechanisms of drug resistance and can be used to evaluate pre-clinical therapeutic strategies. Citation Format: Prafulla C. Gokhale, Sangeetha Palakurthi, Man Xu, Antonio Calles, Shruti D. Shah, Allison O'Connell, Melissa Messineo, Yanan Kuang, Masahiko Yanagita, Mosab Ali, Cam Anh Tran, Jihyun Choi, Atsuko Ogino, Mohit Butaney, Amanda J. Redig, Cloud Paweletz, Paul T. Kirschmeier, Pasi A. Jänne. Modeling patient-derived lung cancer in mice: Preclinical tool for drug development. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 1451. doi:10.1158/1538-7445.AM2015-1451