Prognostic values of tumor size and location in early stage endometrial cancer patients who received radiotherapy
- Author:
Shuning JIAO
1
;
Lichun WEI
;
Lijuan ZOU
;
Tiejun WANG
;
Ke HU
;
Fuquan ZHANG
;
Xiaorong HOU
Author Information
- Publication Type:Original Article
- From:Journal of Gynecologic Oncology 2024;35(6):e84-
- CountryRepublic of Korea
- Language:English
-
Abstract:
Objective:To investigate the correlation between tumor size, tumor location, and prognosis in patients with early-stage endometrial cancer (EC) receiving adjuvant radiotherapy.
Methods:Data of patients who had been treated for stage I–II EC from March 1999 to September 2017 in 13 tertiary hospitals in China was screened. Cox regression analysis was performed to investigate associations between tumor size, tumor location, and other clinical or pathological factors with cancer-specific survival (CSS) and distant metastasis failurefree survival (DMFS). The relationship between tumor size as a continuous variable and prognosis was demonstrated by restricted cubic splines. Prognostic models were constructed as nomograms and evaluated by Harrell’s C-index, calibration curves and receiver operating characteristic (ROC) curves.
Results:The study cohort comprised 805 patients with a median follow-up of 61 months and a median tumor size of 3.0 cm (range 0.2–15.0 cm). Lower uterine segment involvement (LUSI) was found in 243 patients (30.2%). Tumor size and LUSI were identified to be independent prognostic factors for CSS. Further, tumor size was an independent predictor of DMFS. A broadly positive relationship between poor survival and tumor size as a continuous variable was visualized in terms of hazard ratios. Nomograms constructed and evaluated for CSS and DMFS had satisfactory calibration curves and C-indexes of 0.847 and 0.716, respectively. The area under the ROC curves for 3- and 5-year ROC ranged from 0.718 to 0.890.
Conclusion:Tumor size and LUSI are independent prognostic factors in early-stage EC patients who have received radiotherapy. Integrating these variables into prognostic models would improve predictive ability.
