# Prognostic Differences and Survival Predictive Models for Mucinous Versus Usual‐Type Adenocarcinoma of the Uterine Cervix

## TL;DR

This study compares the survival outcomes of mucinous and usual-type adenocarcinoma of the uterine cervix and develops predictive models for prognosis.

## Contribution

The study introduces survival prediction models specifically for mucinous adenocarcinoma of the cervix.

## Key findings

- Mucinous adenocarcinoma had lower overall and cancer-specific survival rates compared to usual-type adenocarcinoma.
- Predictive nomograms for overall and cancer-specific survival were developed and validated for mucinous adenocarcinoma.
- Seven clinical variables were identified as key predictors for survival outcomes in mucinous adenocarcinoma.

## Abstract

There is significant histological heterogeneity between the endocervical adenocarcinoma (EA) subtypes. Usual‐type carcinoma (adenocarcinoma) and mucinous carcinoma (mucinous adenocarcinoma, MA) are the most common types of EA.

Demographic and clinical variables were collected from the SEER database for selected patients between 2004 and 2021. The effect of confounding variables was reduced by propensity score matching (PSM). Survival data were analyzed using the Kaplan–Meier method and Cox regression models. A risk prediction model nomogram for MA was developed and validated.

The median age for MA patients was 46 years compared to 45 years for adenocarcinoma (p = 0.021). The 1‐, 3‐, and 5‐year overall survival (OS) rates for MA were 88.2%, 74.5%, and 68.4%, respectively, significantly lower than those for adenocarcinoma (89.0%, 79.0%, and 74.9%, p < 0.0001). Cancer‐specific survival (CSS) showed a similar trend (p < 0.0001). Seven variables, including age, primary site, T, N, combined stage, surgery, and chemotherapy, were selected to create the nomograms for predicting OS, while age, primary site, tumor size, T, N, combined stage, and surgery were selected for CSS. The validations of all predictive models were satisfactory.

This study revealed MA's poorer prognosis compared to adenocarcinoma using the SEER database. It developed predictive models for OS and CSS of MA, offering a more accurate prognosis assessment tool for clinical practice.