Abstract: Objective This study aimed to use platinum resistance-related (PRR) genes to classify ovarian cancer (OV) patients and establish a prognostic risk model,in order to provide reference for individualized clinical treatment and related mechanism research of OV patients. Methods The univariate Cox regression was used to screen PRR genes with prognostic value,unsupervised consensus clustering was used for subtyping,and the prognostic differences were compared between subtypes.LASSO regression analysis was used to further screen prognosis related PRR genes and construct platinum resistance related gene scores (PRR-GS),combined with clinical information,constructed a prognostic risk model for OV patients,and validated it. Results Univariate Cox regression analysis identified 68 platinum resistance related genes that were associated with patient prognosis.Unsupervised consensus clustering identified 2 subtypes of OV,with C1 group patients having a better prognosis.Twelve genes were used to construct prognostic markers for PRR genes score,and the area under the ROC curves at 1,3,and 5 years were all above 0.700.Age,stage,and PRR-GS were independent factors affecting prognosis in multivariate Cox regression analysis.The model′s C-index was 0.719,and the areas under the ROC curves for 1-year,3-year,and 5-year overall survival rates were 0.774,0.758,and 0.768,respectively.The calibration curve and decision curve analysis showed that the model had good predictive performance. Conclusion Two subtypes of OV patients were identified,each with differences in immune microenvironment,prognosis,and other aspects.The platinum resistance gene prognostic score is an independent risk factor,and when combined with clinical variables,it can effectively predict patient survival outcomes.

Key words: Ovarian cancer, Platinum resistance, Classification, Immune microenvironment, Prognosis