Abstract PO-064: Predicting tumor sensitivity to chemotherapeutic drugs using molecular, cellular and computational methods

Abstract

Abstract Oral squamous cell carcinoma (OSCC) patients respond poorly to chemotherapy. Predicting anticancer drug response before the treatment of OSCC is a significant challenge. Recently, gene expression profiling and cell-based assays reveals drug sensitivity that could be implemented for drug selection for patients. We analyzed the expression of 11 drug response-related genes in 31 OSCC biopsies collected prior to any treatment, using custom designed PCR array. Further, we investigated drug response pattern of selected anticancer drugs using BH3 apoptotic profiling in the primary cells isolated from OSCC tissues. Then, we correlated the results of drug-response gene expression pattern with apoptotic priming to predict tumor response to chemotherapy. The best performing drug (BPD) and response differences (RD) between the drugs were identified using statistical methods to select the best choice of drug in a personalized manner. Based on the Pierson correlation, we classified OSCC tumors as sensitive, moderately responsive or resistant to chemotherapy. We observed that 2 patients were resistant, 17 patients were moderately responsive and 12 patients were sensitive to chemotherapeutic drugs. We found that up-regulation of genes linked to drug resistance facilitates survival of tumor samples, which was revealed by the percentage of apoptotic priming. Moreover, we found that paclitaxel induced 40-45% apoptotic priming compared to other drugs. Average response difference (RD) analysis showed that 80% of tumors responded well to paclitaxel as compared to other drugs studied that were less effective. The computational prediction of drug responses based on the analysis of multiple clinical features of the tumor will be a novel strategy for accomplishing the long-term goal of precision medicine in oncology. The cancer patients will be benefitted if we computationally account all the tumor characteristics (data) for the selection of most effective and precise therapeutic drug. We developed drug efficacy prediction models using multiple tumor features by employing the statistical methods like multi linear regression (MLR), modified MLR-weighted least square (MLR-WLS) and enhanced MLR-WLS. All the three developed drug efficacy prediction models were then validated using the data of actual OSCC samples (train-test ratio 31: 31) and actual Vs hypothetical samples (train-test ratio 31: 30). The selected best statistical model i.e. enhanced MLR-WLS has then been cross-validated (CV) using 341 theoretical tumor data. Finally, the performances of the models were assessed by the level of learning confidence, significance, accuracy and error terms. We observed that the enhanced MLR-WLS model was the best fit to predict anticancer drug efficacy which may have translational applications. Therefore, drug-response gene expression pattern with the status of apoptotic priming in the tumor cells may be translated to the clinic and which will make impact in precision medicine. Citation Format: N. Rajendra Prasad, Kanimozhi Govindaswamy, G.R. Brindha. Predicting tumor sensitivity to chemotherapeutic drugs using molecular, cellular and computational methods [abstract]. In: Proceedings of the AACR Virtual Special Conference on Artificial Intelligence, Diagnosis, and Imaging; 2021 Jan 13-14. Philadelphia (PA): AACR; Clin Cancer Res 2021;27(5_Suppl):Abstract nr PO-064.