Understanding biological driver mutations and mechanism of resistance to EGFR-tyrosine kinase inhibitors in lung cancer patients using a novel translational platform.

Abstract

e20555 Background: Lung cancer causes ~1.69 million deaths globally and 5-year survival rate is < 20%. Surgery is preferred for patients with early stage NSCLC, however, in > 50% cases lung cancers are diagnosed at an advanced stage. EGFR-TKIs, erlotinib and gefitinib have shown benefit in patients harboring sensitizing EGFR mutations, although 20-30% of NSCLC patients do not show good clinical response to EGFR-TKIs despite having the mutations. Several factors have been identified for de novo and acquired resistance to first generation EGFR-TKIs, including secondary mutations such as T790M, aberrant activation of bypass pathways such as amplification of c-MET, or other abnormalities in downstream pathways including KRAS mutations or loss of PTEN among others. There exist significant inter-patient variations in response profiles and mechanisms of resistance, which often confounds treatment selection on a personalized level, especially if first-line therapy fails. Methods: We have developed a patient tumor derived ex vivo platform technology using multi-disciplinary systems biology-based approach that recreates a patient’s tumor microenvironment by preserving the tumor-stromal architecture and immune contexture. Fresh tumor biopsies and 10-ml blood were collected from patients with IRB approval and patient consent. Tumors obtained from 21 lung cancer patients were grown ex vivo on our platform in presence of PBMC and autologous sera. The tumors were treated with standard of care chemotherapies and EGFR inhibitors, and drug efficacy on tumors were measured. Results: Of 18 tumor samples treated with EGFR-TKIs, 4 were predicted responders to gefitinib and 14 were predicted non-responders to EGFR-TKIs (erlotinib or gefitinib). Of the 14 non-responders, 7 were predicted responders to carboplatin/pemetrexed and/or carboplatin/docetaxel. Patients were followed, and clinical data was captured. Whole genome sequencing and mutational analysis was done on the tumors. Hits were validated by immunohistochemical assays. Data from molecular signatures of the tumors with functional, phenotypic assays, were used to delineate mechanism of resistance in individual lung cancer patients to EGFR-TKIs. Conclusions: We have developed an ex vivo, personalized platform, which enables prediction of treatment response on an individual level, and also aids in capturing resistance mechanisms for identifying actionable targets.