Discovery and Development of Predictive Biomarkers for the Personalization of Pemetrexed Therapy in Non-Small Cell Lung Cancer

Lung cancer is a major health problem, and is the leading cause of cancer-related deaths. This research proposal was designed to personalize and improve overall response rates to the FDA-approved chemotherapeutic pemetrexed. To do so, clinically-applicable response phenotypes of a large panel of lung cancer lines needed to be defined to identify gene expression profiles driving response to pemetrexed treatment. To mimic clinical exposure and duration in vitro, published pharmacokinetics of patients treated with pemetrexed were extrapolated to develop the dose schedule used in this study. Although two drug response assays were simultaneously performed, liquid colony formation assay most closely mimicked clinical indications for the treatment of patients with pemetrexed, and was non-toxic to normal human bronchial epithelial cells. There were three distinct pemetrexed response phenotypes across this cell line panel: sensitive, intermediate and resistant. Interestingly, large cell carcinomas and a subset of adenocarcinomas are sensitive to pemetrexed treatment. Upon oncogenotype analysis, I found that mutations in EGFR correlated significantly with pemetrexed resistance, and this was confirmed by mRNA expression profiling and reverse phase protein arrays. Furthermore, I determined that mutations in KRAS were significantly more frequent in sensitive lines, and EML4/ALK adenocarcinomas conferred sensitivity to pemetrexed, consistent with recent clinical findings. In order to test whether the in vitro pemetrexed response phenotypes could be recapitulated in vivo, doses from 26.6 mg/kg up to 1000 mg/kg pemetrexed qwx3 were administered to and well tolerated by NOD/SCID mice. Treatment of established lung cancer xenografts with doses of pemetrexed 500 mg/kg or higher qwx3 were consistent with in vitro findings. To elucidate mechanisms of response, intratumoral mRNA levels of TYMS and pemetrexed-related genes did not have any significant correlation with response. However, microarray expression profiling, RNA-Seq and western blot analysis independently highlighted FTCD overexpression can be found only in pemetrexed-sensitive lung cancer lines. Stable knockdown of FTCD in a FTCD-overexpressed lung cancer line resulted in specific and abrupt cell death. I further describe the ongoing analysis of retrospective clinical datasets. Our plans for prospective FTCD biomarker enrollment in future pemetrexed clinical trials highlight the translational potential and progression of this dissertation proposal.