Develop and validate a risk prediction model and nomogram for peritoneal metastasis in patients with ovarian cancer

The most common type of Ovarian Cancer (OC) metastasis is peritoneal metastasis (PM). Patients with OC with PM have a significantly lower 5-year survival rate than those without PM. Preoperative evaluation of PM is critical. We conducted this study to better evaluate OC with PM before surgery. We conducted a multicenter study. The transvaginal ultrasonography, clinical data and laboratory parameters of patients who underwent surgery at three hospitals between January 2019- October 2023 and who received pathological diagnoses of OC were retrospectively analyzed. A risk prediction model and a nomogram for OC with PM were constructed with multimodal data from two of the multicenter OC patients, and another unit of data was used as an external test set to evaluate model. A total of 673 OC patients were included,471 in the training set and 202 in the validation set. Ascites, peritoneal thickening, abnormal CA125, HE4 and D-dimer levels were found to be independent risk factors for OC with PM. We used these data to construct a predictive model, constructed a nomogram to evaluate the model. The area under the curve (AUC) for both the training and validation sets exceeded 0.80. The maximum Youden index of the ROC curve for the training set was 0.416, with a sensitivity of 0.740,specificity of 0.760. The model achieved an AUC of 0.87 in an independent external validation cohort of 132 patients, and the Hosmer‒Lemeshow test results indicated that the external validation set model exhibited satisfactory goodness-of-fit. D-dimer can be used as a predictor of OC with PM. The predictive model developed by our team effectively forecast the likelihood of OC with PM. This model has significant implications for informing clinical treatment decisions in OC management. Not applicable.