# Establishment and validation of a nomogram model for predicting early death in patients with endometrial cancer bone metastases

## TL;DR

This study creates a tool to predict early death in patients with rare endometrial cancer bone metastases, helping doctors make better treatment decisions.

## Contribution

Development of a validated nomogram model for predicting early death in endometrial cancer bone metastases patients.

## Key findings

- Surgery and chemotherapy significantly reduce early death risk in ECBM patients.
- Brain metastases and sarcomatous histological subtype increase early death risk.
- The nomogram model showed high predictive accuracy and strong clinical utility.

## Abstract

Patients with endometrial cancer bone metastases (ECBM) are clinically rare and have a poor prognosis, including a higher incidence of early death (survival ≤ 3 months). Currently, no practical tools exist to predict early mortality in these patients. Thus, there is an urgent need to develop clinically applicable predictive models, such as nomograms, for individualized assessment of early death risk in ECBM.

Relevant clinical and pathological data for ECBM patients from the SEER database (2010-2021). Univariate and multivariate logistic regression analyses were performed to identify risk factors associated with early death in ECBM patients and to construct prognostic nomograms. ROC analysis, calibration curves, and decision curve analysis (DCA) were used to assess the predictive accuracy and clinical utility of the nomogram model.

A total of 1,201 ECBM patients were found in the SEER database. After applying strict exclusion criteria, 769 patients were finally included in this study. Patients were randomly divided into training and validation cohorts in a 7:3 ratio. The results of univariate and multivariate logistic regression analyses revealed several independent predictive factors for early death. For both overall early death (OED) and cancer-specific early death (CSED), protective factors included surgery (OED: OR = 0.22, 95%CI: 0.12-0.41, p<0.001; CSED: OR = 0.33, 95%CI: 0.18-0.61, p<0.001) and chemotherapy (OED: OR = 0.11, 95%CI: 0.06-0.18, p<0.001; CSED: OR = 0.14, 95%CI: 0.09-0.24, p<0.001). Brain metastases increased risk (OED: OR = 2.98, 95%CI: 1.29-6.87, p=0.01; CSED: OR = 2.20, 95%CI: 1.04-4.79, p=0.047). Compared to 0–9 days, longer time from diagnosis to treatment showed protective associations: 10–27 days (OED: OR = 0.51, 95%CI: 0.27-0.98, p=0.042) and ≥28 days (OED: OR = 0.23, 95%CI: 0.12-0.44, p<0.001; CSED: OR = 0.30, 95%CI: 0.16-0.56, p<0.001). Regarding histological type, compared to endometrioid subtype, sarcomatous subtype significantly increased OED risk (OR = 3.04, 95%CI: 1.40-6.57, p=0.005), while radiotherapy reduced CSED risk (OR = 0.55, 95%CI: 0.33-0.92, p=0.022). Based on these variables, nomograms were developed to predict the risk of early death. The ROC curve confirmed the model’s high predictive accuracy, while the calibration curve showed strong alignment between predicted and actual survival. DCA further demonstrated its clinical utility.

In this study, we developed robust nomogram models to predict the probability of early death in ECBM patients.