# Nomogram Prediction Model and Prognostic Comparison of Cervical Clear Cell Carcinoma and Cervical Endometrioid Adenocarcinoma: A SEER Database Study

## TL;DR

This study creates a predictive model to help doctors assess survival risks for two rare cervical cancers using data from the SEER database.

## Contribution

A novel nomogram model is developed and validated for predicting survival in cervical clear cell and endometrioid adenocarcinomas.

## Key findings

- The model showed strong performance with AUCs of 0.894 and 0.857 for 12- and 24-month survival predictions in the development cohort.
- The model effectively stratified patients into risk groups with significantly different survival outcomes (p < 0.0001).
- Decision Curve Analysis and Clinical Impact Curve confirmed the model's clinical usefulness in identifying high-risk patients.

## Abstract

Cervical clear cell adenocarcinoma (CCAC) and cervical endometrioid adenocarcinoma (CEAC) are rare and aggressive non‐HPV‐associated malignancies. Despite their histological similarities, these subtypes demonstrate distinct biological behaviors, presenting challenges in treatment and prognosis.

To develop and validate a multivariable prognostic model for CCAC and CEAC, utilizing the SEER database to enhance clinical decision‐making.

A total of 775 CEAC and 421 CCAC cases were analyzed using a multivariable nomogram. Patients were randomly allocated to model‐development (n = 838) and validation (n = 358) cohorts in a 7:3 ratio. The model's performance was evaluated through AUC, Brier score, and Calibration. Decision Curve Analysis (DCA) and Clinical Impact Curve (CIC) were assessed in both development and internal validation cohorts.

The model exhibited excellent calibration and discrimination in predicting overall survival (OS). In the development cohort, the 12‐ and 24‐month prediction models had AUCs of 0.894 (95% CI: 0.860–0.928) and 0.857 (95% CI: 0.821–0.892), respectively. In the internal validation cohort, the 12‐ and 24‐month models achieved AUCs of 0.814 (95% CI: 0.788–0.840) and 0.798 (95% CI: 0.775–0.822), respectively. The model effectively stratified patients into low‐, intermediate‐, and high‐risk groups, with significantly different median survival times (p < 0.0001). DCA and CIC further validated the model's clinical utility.

We developed a robust nomogram for quantifying OS risk in CCAC and CEAC patients. This model provides clinicians with a tool for identifying high‐risk patients and implementing timely interventions.