Nomogram Predicts Peritoneal Metastasis in Ovarian Cancer Before Surgery

Peritoneal metastasis is the most common form of spread in ovarian cancer, and patients whose disease has reached the peritoneum have a significantly lower five-year survival rate than those without it.

The team analyzed records of ovarian cancer patients who underwent surgery at three hospitals between January 2019 and October 2023 and who received pathological diagnoses of ovarian cancer.

In total, 673 patients with ovarian cancer were included. Of these, 471 formed the training set used to build the model, and 202 formed an internal validation set. Data from patients at a third center were held back as an independent external test set to evaluate how the model performed on patients it had never seen.

The inclusion of D-dimer is notable because the authors conclude it can serve as a predictor of peritoneal metastasis in this setting.

Performance was assessed with standard discrimination and calibration measures. The area under the receiver operating characteristic curve exceeded 0.80 in both the training and internal validation sets, indicating discrimination better than chance by a comfortable margin. In the training set, the maximum Youden index of the ROC curve was 0.416, corresponding to a sensitivity of 0.740 and a specificity of 0.760, meaning the model correctly flagged roughly three quarters of patients with peritoneal metastasis while correctly clearing a similar proportion of those without it.

The most demanding test was the external validation cohort of 132 patients from a center not involved in model construction. There the model achieved an AUC of 0.87, and Hosmer–Lemeshow test results indicated satisfactory goodness-of-fit, suggesting that predicted probabilities aligned reasonably well with observed outcomes in this independent group.

The work is retrospective, relying on patients already operated on at the three participating hospitals during the study window, so its findings reflect the populations and measurement practices of those centers. The article was published as early-view peer-reviewed accepted research in BMC Medical Imaging, citable with a permanent DOI and subject to further edits before the final version of record.

Within those limits, the authors argue the model has practical value: a preoperative estimate of peritoneal metastasis risk could inform clinical treatment decisions in ovarian cancer management, where knowing the likelihood of peritoneal spread before surgery may shape planning.

For clinicians, the study adds D-dimer to the familiar preoperative picture of ovarian cancer, alongside imaging signs such as ascites and peritoneal thickening and established markers such as CA125 and HE4. Whether the model’s accuracy holds in broader populations and prospective use will require further evaluation, but the reported AUCs above 0.80 in training, internal validation, and an independent external cohort of 132 patients mark it as a candidate tool for preoperative risk stratification in ovarian cancer.

Subject of Research: A risk prediction model and nomogram for preoperative prediction of peritoneal metastasis in ovarian cancer

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

Article References: Li, X., Duan, C., Li, Y., Hu, B., Zhang, C., & Zhou, Y. (2026). Develop and validate a risk prediction model and nomogram for peritoneal metastasis in patients with ovarian cancer. BMC Medical Imaging. https://doi.org/10.1186/s12880-026-02941-9

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02941-9

Keywords: ovarian cancer, peritoneal metastasis, nomogram, risk prediction model, D-dimer, CA125, HE4, transvaginal ultrasonography, ascites, external validation, BMC Medical Imaging, Develop