Ngs-Based Multi-Factor Prediction Model Of Platinum Sensitivity For Epithelial Ovarian Cancer

Objectives: Platinum sensitivity is the most important predictor of prognosis and response to further treatment in ovarian cancer, while it is still roughly prejudged in line with the interval of progression free survival in clinical practice. Realizing individualized precision medicine require refining predictive model that could enable selection of patients who would benefit from platinum-based or non-platinum containing therapy. The study aims to identify and integrate novel biomarkers for prediction of platinum sensitivity in ovarian cancer. Methods: 60 patients with epithelial ovarian cancer receiving standard initial treatment were recruited from 2016 to 2019. According to interval of progress free survival (PFS), 44 patients (PFS>6 months) and 16 patients (PFS≤6 months) were defined as platinum-sensitive and platinum-resistant recurrence, respectively. We performed next generation sequencing (NGS) analysis of the entire genomic features of ovarian tumor tissue samples obtained from debulking surgery. Multiple genomic variation were compared between platinum sensitive and resistant patients, including single-nucleotide variant (SNV), single base or indel signature, focal copy number variation (CNV), mutational clonality, whole-genome duplication (WGD), and co-variation in DNA damage repair (DDR) pathways. Collectively, a multi-factor-based prediction model was trained and validate with Lasso regression. Download : Download high-res image (244KB) Download : Download full-size image Results: Analysis of single base signature revealed that the contribution of SBS10a in platinum-sensitive patients was significantly higher than that of platinum-resistant patients (P Download : Download high-res image (198KB) Download : Download full-size image Conclusions: Our study successfully established a prediction model containing three novel biomarkers FAM175A LOH, SBS10a mutation and WGD through in-depth mining of NGS, which has an extremely potential of predictive value for platinum sensitivity in epithelial ovarian cancer. However, it requires larger cohort for further validation.