Abstract 2761: Germline polymorphism discovered via a cell-based genome-wide approach predicts platinum response in ovarian and head and neck cancers

Abstract

Abstract Identifying patients prior to treatment that are less likely to benefit from or most likely to experience adverse events from chemotherapeutic agents is essential. Even though humans are the most relevant system, pharmacogenomic discovery in the field of oncology is plagued by difficulties in executing large clinical trials and confounding factors such as co-morbidities, dosage, and concomitant medications. Therefore, we developed a genome-wide cell-based approach evaluating single nucleotide polymorphisms (SNPs) and gene expression to predict chemotherapy-induced response and toxicity and then validated our findings in clinically relevant patient cohorts. Our model utilizes the International HapMap lymphoblastoid cell lines (LCLs). Genome-wide association studies were performed to identify SNPs significantly associated with carboplatin sensitivity through their effects on mRNA expression. The significant findings were evaluated in an independent LCL replication set and in patient samples obtained from the Australian Ovarian Cancer Study (AOCS) and two Phase II head and neck clinical trials (UC12019 and UC13881) conducted at the University of Chicago. Four hundred and nine ovarian cancer patients receiving carboplatin and paclitaxel were analyzed in AOCS; while 60 and 32 head and neck cancer patients were evaluated in UC12019 and UC13881 trials. Carboplatin was used as induction and concomitant chemo-radiation therapy in UC12019 and UC13881 trials, respectively. Using LCLs, our genome-wide model identified 65 SNPs that are associated with at least one carboplatin sensitivity phenotype through the expression of 61 genes. Five of them were replicated in a separate set of LCLs. In AOCS, SNP (rs1649942) was significantly associated with progression-free survival (Plog-rank=0.009) and overall survival (Plog-rank=0.03). In the head and neck cancer trials, 2 other SNP-phenotype associations were identified and replicated in both trials, including the association between SNP (rs4946514) and overall response to carboplatin, and the association between SNP (rs7134205) and post treatment platelet changes. Given the obstacles to performing large, replicable pharmacogenomic studies in patients, the cell-based model is proven to be an effective alternative in novel pharmaco-SNP discovery. We demonstrate germline SNPs identified through the cell-based genome-wide approach are clinically important predictors of chemotherapy response and toxicity. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 2761.