Two novel prognostic models for ovarian cancer respectively based on ferroptosis and necroptosis

Abstract

Abstract Background Platinum-resistant cases account for 25% of ovarian cancer patients. Our aim was to construct two novel prognostic models based on gene expression data respectively from ferroptosis and necroptosis, for predicting the prognosis of advanced ovarian cancer patients with platinum treatment. Methods According to the different overall survivals, we screened differentially expressed genes (DEGs) from 85 ferroptosis-related and 159 necroptosis-related gene expression data in the GSE32062 cohort, to establish two ovarian cancer prognostic models based on calculating risk factors of DEGs, and log-rank test was used for statistical significance test of survival data. Subsequently, we validated the two models in the GSE26712 cohort and the GSE17260 cohort. In addition, we took gene enrichment and microenvironment analyses respectively using limma package and GSVA software to compare the differences between high- and low-risk ovarian cancer patients. Results We constructed two ovarian cancer prognostic models: a ferroptosis-related model based on eight-gene expression signature and a necroptosis-related model based on ten-gene expression signature. The two models performed well in the GSE26712 cohort, but the performance of necroptosis-related model was not well in the GSE17260 cohort. Gene enrichment and microenvironment analyses indicated that the main differences between high- and low- risk ovarian cancer patients occurred in the immune-related indexes, including the specific immune cells abundance and overall immune indexes. Conclusion In this study, ovarian cancer prognostic models based on ferroptosis and necroptosis have been preliminarily validated in predicting prognosis of advanced patients treated with platinum drugs. And the risk score calculated by these two models reflected immune microenvironment. Future work is needed to find out other gene signatures and clinical characteristics to affect the accuracy and applicability of the two ovarian cancer prognostic models.