Leveraging various extracellular matrix levels to assess prognosis and sensitivity to immunotherapy in patients with ovarian cancer.

Background: Ovarian cancer (OC) is the fifth leading cause of cancer-related deaths among women. Late diagnosis and heterogeneous treatment result in a poor prognosis for patients with OC. Therefore, we aimed to develop new biomarkers to predict accurate prognoses and provide references for individualized treatment strategies. Methods: We constructed a co-expression network applying the "WGCNA" package and identified the extracellular matrix-associated gene modules. We figured out the best model and generated the extracellular matrix score (ECMS). The ECMS' ability to predict accurate OC patients' prognoses and responses to immunotherapy was evaluated. Results: The ECMS was an independent prognostic factor in the training [hazard ratio (HR) = 3.132 (2.068-4.744), p< 0.001] and testing sets [HR = 5.514 (2.084-14.586), p< 0.001]. The receiver operating characteristic curve (ROC) analysis showed that the AUC values for 1, 3, and 5 years were 0.528, 0.594, and 0.67 for the training set, respectively, and 0.571, 0.635, and 0.684 for the testing set, respectively. It was found that the high ECMS group had shorter overall survival than the low ECMS group [HR = 2 (1.53-2.61), p< 0.001 in the training set; HR = 1.62 (1.06-2.47), p = 0.021 in the testing set; HR = 1.39 (1.05-1.86), p = 0.022 in the training set]. The ROC values of the ECMS model for predicting immune response were 0.566 (training set) and 0.572 (testing set). The response rate to immunotherapy was higher in patients with low ECMS. Conclusion: We created an ECMS model to predict the prognosis and immunotherapeutic benefits in OC patients and provided references for individualized treatment of OC patients.