Identification of prognostic risk factors and construction of nomograms for elderly patients with stage III–IV endometrioid endometrial carcinoma: a study based on the SEER database

Introduction

Endometrial carcinoma (EC) is a clinically prevalent gynecological malignant tumor, and its global incidence is on a yearly increasing trend. Existing data indicate that EC is among the most frequently diagnosed cancers in women, a trend that is particularly evident in developed countries, where its incidence and disease burden continue to rise (1,2). This increasing trend is partially attributable to population aging and the rising prevalence of metabolic disorders. Endometrioid EC represents dominant histological variant of EC, comprising nearly 75–80% of the total (3).

Age significantly influences the development as well as the prognostic profile of EC. Previous studies have shown that the majority of EC cases are diagnosed after menopause, with a substantial proportion occurring in women aged ≥50 years (4). Elderly patients are frequently accompanied by multiple chronic comorbidities, including hypertension, diabetes, and cardiovascular disease, which may influence treatment selection and adversely affect overall survival (OS). Furthermore, elderly patients with EC tend to present with high-grade tumors and aggressive histological subtypes, which are associated with poorer prognosis (5,6). In this context, endometrioid EC, the most common histological subtype, became the primary concern of the present study.

According to the established tumor staging system, stage III–IV EC is classified as advanced disease, characterized by local invasion and/or distant metastasis, and it is associated with a markedly worse overall clinical outcome. In large-scale oncological databases, tumor stage is commonly recorded and routinely defined in accordance with the American Joint Committee on Cancer (AJCC) staging classification. Past research has demonstrated that subjects with advanced-stage disease have markedly poorer 5-year OS than those diagnosed at early stages, particularly in stage IV disease, in which long-term survival remains extremely poor (7). Despite continuous advances in surgical techniques, radiotherapy, chemotherapy, and novel targeted therapies recently, the overall outcome for patients with advanced EC remains unsatisfactory (8). Therefore, identifying key prognostic risk factors associated with survival in advanced disease is crucial for guiding clinical treatment decisions.

In cancer prognostic assessment, nomograms are widely used as visual predictive tools that integrate multiple clinical variables to quantitatively estimate an individual survival probability. Compared with traditional staging systems, nomograms offer a fuller assessment of the combined effects of multiple prognostic risk factors and have therefore been extensively applied in oncological prognostic studies (9,10). Nomogram-based predictive instruments for different types of tumors have recently emerged, demonstrating favorable predictive performance and clinical utility. However, prognostic models specifically tailored for elderly patients with advanced endometrioid EC remain relatively limited.

With the rapid advancement of medical big data, large-scale cancer databases have become invaluable resources for population-based oncological research. Accounting for approximately 30% of the overall population in the United States, the Surveillance, Epidemiology, and End Results (SEER) program is maintained and regularly updated by the National Cancer Institute (NCI). It is currently one of the most widely used databases in cancer epidemiological research (11). Its large-scale nature enables a more comprehensive evaluation of prognostic factors associated with cancer prognosis, thereby enhancing the reliability and representativeness of study findings.

Based on these considerations, the present study systematically analyzed elderly patients with endometrioid EC (stage III–IV) employing SEER database data, seeking to identify prognostic indicators independently linked to OS and cancer-specific survival (CSS). Furthermore, models constructed with nomograms were developed to facilitate individualized prognostic assessment and support clinical decision-making in this specific patient population. We present this article in accordance with the TRIPOD reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-0824/rc).

Methods

Data source

The full original records incorporated into our research were provided by the SEER database, an authoritative nationwide cancer monitoring platform maintained by the NCI. Its registered population coverage extends to approximately 30% of the total U.S. population, providing strong representativeness for population-based epidemiological analyses. Clinical data of patients with a confirmed diagnosis of endometrioid EC between 2010 and 2015 were retrieved via SEER Stat software (version 9.0.42.2) for the present analysis. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Since the SEER database includes publicly accessible de-identified patient data, this study did not require informed consent or institutional review board approval.

Study population

Individuals with a confirmed diagnosis of endometrioid EC during the period 2010–2015 were retrieved from the SEER database using the SEERStat software. Eligible participants were selected according to predefined inclusion and exclusion criteria. Briefly, patients were required to have a primary endometrial tumor with endometrioid histology, complete staging information based on the 7th edition of the AJCC staging system, and a single primary malignancy. A total of 47,750 patients with endometrioid EC were initially identified from the SEER database. Patients were excluded if they had missing key clinicopathological or survival information (n=20,711), AJCC stage I–II disease (n=23,614), age <60 years (n=1,560), or missing data for all study variables (n=569). After applying these criteria, 1,296 patients were found to be eligible and included in the final analysis. The detailed inclusion and exclusion criteria are presented in Table S1, and the patient selection process is illustrated in Figure 1.

Figure 1 Flowchart of participant selection. AJCC, American Joint Committee on Cancer; SEER, Surveillance, Epidemiology, and End Results.

Outcomes

OS and CSS were designated as the primary endpoints of this study. OS was defined as the interval from diagnosis to death from any cause or the final follow-up date, while CSS was defined as the interval from diagnosis to death specifically attributed to EC. All survival durations were calculated in months. Outcome data were obtained from the SEER database, in which vital status and cause of death are recorded using standardized coding procedures based on death certificates, independent of study variables. Therefore, no additional blinding or subjective outcome assessment was required.

Candidate variables

Candidate predictors were prespecified based on clinical relevance and data availability within the SEER database. All baseline variables were measured at the time of diagnosis, and treatment variables were defined based on the first course of therapy. Predictor data were obtained from standardized SEER coding procedures, independent of study outcomes, with no subjective assessment involved; thus, blinding of predictor assessment was not applicable. Demographic variables included age (continuous, years), race [black, other (including Asian, Pacific Islander, and American Indian), or white], and marital status (divorced/separated, married, single, or widowed). Tumor characteristics comprised tumor grade (grade I–IV), tumor size (≤20 mm, 21–50 mm, >50 mm), and AJCC stage (stage III or IV). Tumor extent was further characterized using the TNM classification (T1–T4, N0–N2, and M0–M1). Treatment variables included surgery, radiotherapy, and chemotherapy (all categorized as yes vs. no/unknown), defined by receipt during the initial treatment course. Detailed information on specific regimens or doses is not available in the SEER database. Metastatic status at diagnosis was assessed using SEER site-specific metastasis variables and included the presence of bone, brain, liver, and lung metastases (yes vs. no).

Statistical analysis

All statistical analyses were performed using R software (version 4.5.1). Baseline patient characteristics were summarized through descriptive statistical analysis. Categorical variables were presented as frequencies and percentages, and differences between the training and validation cohorts were compared using the chi-square test (χ² test). To develop the prognostic models, all patients were randomly assigned to the training and validation cohorts at a ratio of 7:3. In the training cohort, univariate Cox proportional hazards regression analysis was first performed to identify potential factors associated with survival outcomes. Variables with statistical significance in univariate analysis were subsequently included in the multivariate Cox proportional hazards regression model to identify independent prognostic factors. The proportional hazards assumption of the final multivariable Cox regression models was assessed using Schoenfeld residuals. The statistical testing indicated no violation of the assumption (Table S2). Nomograms were then constructed based on the results of the multivariate Cox regression analysis to predict 1-, 3-, and 5-year OS and CSS. Individual linear predictors and corresponding survival probabilities were calculated using regression coefficients from the training model without model re-estimation. No risk stratification or risk group categorization was performed. Model performance was evaluated using the receiver operating characteristic (ROC) curve and area under the curve (AUC)/concordance index (C-index). Calibration curves were plotted to assess the agreement between predicted and observed survival outcomes. In addition, decision curve analysis (DCA) was performed to evaluate the potential clinical utility of the models. The validation cohort was generated by randomly splitting the same SEER dataset used for model development.

Results

Baseline clinical and pathological characteristics

The final analytic cohort consisted of 1,296 elderly patients with stage III–IV endometrioid EC. Following randomization according to a 7:3 proportion, 909 patients were included in the training cohort, whereas the remaining 387 patients constituted the validation cohort (Table 1). In the overall study population, 527 patients (40.66%) were alive at the end of follow-up, whereas 769 patients (59.34%) had died. The median follow-up duration was 68 months [interquartile range (IQR), 21.0–107.25 months]. The median age of the study population was 68 years (IQR, 63–75 years). In terms of baseline demographic characteristics, the majority of patients were of White race and were married. Regarding tumor characteristics, most tumors were of grade II or III, with tumor sizes predominantly in the ranges of 21–50 mm and >50 mm. According to the 7th edition of the AJCC classification system, stage III disease was most frequently observed. In terms of tumor extent, T3 was the most common T stage, whereas N0 and N1 were the predominant N stages. The majority of patients presented with M0 disease. With respect to treatment, the vast majority of patients underwent surgery, while a subset also received radiotherapy or chemotherapy. The proportions of distant organ metastases were 2.08% for bone, 0.69% for brain, 1.93% for liver, and 5.56% for lung. Overall, most baseline characteristics did not exhibit significant differences among the training versus validation cohorts (all P>0.05).

Kaplan-Meier survival analysis

To evaluate OS and CSS across treatment groups, Kaplan-Meier survival analysis was conducted. Our findings showed marked differences in both OS and CSS among groups stratified by chemotherapy, surgery, and radiotherapy (all P<0.001, Figure 2). In particular, patients underwent surgical intervention exhibited significantly improved OS and CSS relative to those who did not receive surgery (Figure 2C,2F). Notably, the survival benefit associated with surgery appeared to be more pronounced than that observed for chemotherapy and radiotherapy across the corresponding subgroups (Figure 2A,2B,2D,2E).

Figure 2 Kaplan-Meier survival curves for OS and CSS according to treatment modality. (A) OS for chemotherapy; (B) OS for radiotherapy; (C) OS for surgery; (D) CSS for chemotherapy; (E) CSS for radiotherapy; and (F) CSS for surgery. CSS, cancer-specific survival; OS, overall survival.

Results of univariate Cox proportional hazards regression analysis

Univariate Cox regression analysis (Table 2) revealed that age, AJCC stage, T/M stage, surgery, radiotherapy, chemotherapy, and distant metastasis (bone, brain, liver, lung) were significantly correlated with both OS and CSS (all P<0.05). Patients aged ≥68 years exhibited a significantly higher mortality risk relative to those younger than 68 years [OS: hazard ratio (HR) =1.71; CSS: HR =1.82; P<0.001]. Similarly, patients with AJCC stage IV disease had markedly higher risks of death compared with those with AJCC stage III disease (OS: HR =2.59; CSS: HR =2.78; P<0.001). Surgical intervention was linked to a marked reduction in mortality risk, with significantly lower risks of death observed in patients who underwent surgery relative to those who did not (OS: HR =0.19; CSS: HR =0.15; P<0.001). Meanwhile, both chemotherapy (OS: HR =0.54; CSS: HR =0.52; P<0.001) and radiotherapy (OS: HR =0.56; CSS: HR =0.58; P<0.001) were also associated with lower risks of death. Among distant metastases, brain metastasis showed the strongest association with poor prognosis (OS: HR =21.18; CSS: HR =18.12; P<0.001).

Results of multivariate Cox proportional hazards regression analysis

Multivariate Cox regression analysis (Table 3) demonstrated that age, tumor grade, N stage, surgery, radiotherapy, chemotherapy, and brain metastasis were still significantly correlated with both OS and CSS (P<0.05). Patients aged ≥68 years continued to demonstrate significantly increased risks of death (OS: HR =1.57; CSS: HR =1.84; P<0.001). Higher tumor grade correlated with an unfavorable prognostic profile. Relative to grade I tumors, Grade III (OS: HR =2.02; CSS: HR =2.48; P<0.001) and grade IV tumors (OS: HR =2.78; CSS: HR =3.71; P<0.001) were associated with significantly elevated mortality risks. Compared with patients with N0 disease, those diagnosed with N2 disease exhibited significantly elevated mortality risks (OS: HR =1.4; CSS: HR =1.42; P<0.05). Surgery remained a prominent protective factor, with markedly reduced mortality observed in surgically treated patients relative to their non‑surgical counterparts (OS: HR =0.22; CSS: HR =0.19; P<0.001). In addition, both chemotherapy (OS: HR =0.55; CSS: HR =0.52; P<0.001) and radiotherapy (OS: HR =0.67; CSS: HR =0.72; P<0.001) were associated with lower risks of death. Moreover, brain metastasis remained one of the strongest adverse prognostic risk factors (OS: HR =17.31; CSS: HR =15.40; P<0.001).

Construction and validation of the nomograms

As shown in Figure 3, nomograms were built for the prediction of 1-, 3-, and 5-year OS and CSS in elderly patients with endometrioid EC. In the OS nomogram, age, tumor grade, T stage, and brain metastasis had the greatest impact on patients’ survival probabilities (Figure 3A). In the CSS nomogram, brain metastasis, T stage, and surgery were identified as the most influential predictors (Figure 3B). A specific score was allocated to each variable, and the cumulative score was calculated as the sum of individual points. This cumulative score was subsequently used to estimate the probabilities of OS and CSS at 1, 3, and 5 years. The results revealed that a higher cumulative score predicted a worse prognostic profile.

Figure 3 Nomograms for predicting survival outcomes based on multivariate Cox regression analysis: (A) nomogram for predicting OS; (B) nomogram for predicting CSS. Model covariates comprised age, race, tumor grade, T stage, N stage, surgery, chemotherapy, radiotherapy, and brain metastasis. CSS, cancer-specific survival; N, regional lymph node; OS, overall survival; T, primary tumor.

ROC curve analysis

The discriminative performance of the nomograms was evaluated using the C-index and time-dependent ROC curves in both the training and validation cohorts. For the OS model, the C-index was 0.73 (95% CI: 0.70–0.75) in the training cohort and 0.71 (95% CI: 0.68–0.75) in the validation cohort. The corresponding 1-, 3-, and 5-year AUCs were 0.80 (95% CI: 0.76–0.84), 0.79 (95% CI: 0.75–0.82), and 0.78 (95% CI: 0.76–0.81), respectively, in the training cohort, and 0.88 (95% CI: 0.84–0.93), 0.78 (95% CI: 0.73–0.84), and 0.76 (95% CI: 0.71–0.81), respectively, in the validation cohort (Figure 4A,4B). These results indicate good discriminative ability and stability of the OS nomogram. For the CSS model, the C-index was 0.74 (95% CI: 0.72–0.77) in the training cohort and 0.73 (95% CI: 0.69–0.77) in the validation cohort. The corresponding 1-, 3-, and 5-year AUCs were 0.84 (95% CI: 0.80–0.88), 0.81 (95% CI: 0.78–0.84), and 0.79 (95% CI: 0.76–0.82), respectively, in the training cohort, and 0.83 (95% CI: 0.77–0.90), 0.77 (95% CI: 0.71–0.82), and 0.80 (95% CI: 0.75–0.85), respectively, in the validation cohort (Figure 4C,4D). Overall, the CSS nomogram also demonstrated satisfactory predictive performance and maintained stable discrimination across both cohorts.

Figure 4 ROC curve analysis of the predictive performance of the nomograms. (A) OS (training cohort); (B) OS (validation cohort); (C) CSS (training cohort); (D) CSS (validation cohort). AUC, area under the curve; CI, confidence interval; CSS, cancer-specific survival; OS, overall survival; ROC, receiver operating characteristic.

Calibration curve analysis

Calibration curves for 1‑, 3‑, and 5‑year survival were plotted in both the training and validation cohorts to assess the consistency between nomogram‑predicted survival probabilities and actual observed outcomes. For the OS model (Figure 5A,5B), the calibration curves at all time points (1, 3, and 5 years) closely approximated the ideal reference line in both cohorts, indicating excellent calibration between predicted and observed survival. Although minor deviations were observed in certain probability ranges, the overall fit was satisfactory, suggesting that the model had good calibration ability. Similar results were observed for the CSS model (Figure 5C,5D), with calibration curves at all time points in both cohorts closely overlapped with the ideal line, demonstrating good predictive accuracy. In particular, the deviations between predicted and observed values were minimal across the moderate‑to‑high probability ranges, further confirming the stable calibration performance of the models.

Figure 5 Calibration curve analysis of the nomograms. (A) OS (training); (B) OS (validation); (C) CSS (training); (D) CSS (validation). CSS, cancer-specific survival; OS, overall survival.

DCA

DCA was conducted to further validate the clinical applicability of the nomograms. Net benefits at various threshold probabilities were quantified, with comparisons made against the “treat all” and “treat none” reference strategies. For the OS model (Figure 6A-6F), the predictive models for 1‑, 3‑, and 5‑year survival maintained favorable net benefits across a broad range of threshold probabilities in both training and validation cohorts, exhibiting clear superiority over the “treat all” and “treat none” reference strategies. These findings confirm favorable clinical utility of the OS model across different time points. Similarly, for the CSS model (Figure 6G-6L), the net benefit curves of the 1-, 3-, and 5-year prediction models remained consistently above those of the reference strategies in both cohorts. The performance was particularly more prominent within the intermediate threshold probability range, suggesting that the CSS models may provide valuable guidance for clinical decision-making.

Figure 6 1-, 3-, and 5-year DCA of the nomograms. (A-C) OS (training); (D-F) OS (validation); (G-I) CSS (training); (J-L) CSS (validation). CSS, cancer-specific survival; DCA, decision curve analysis; OS, overall survival; prob, probability.

Discussion

On the basis of SEER data, this study systematically analyzed prognostic determinants among elderly patients with stage III–IV endometrioid EC and constructed nomograms to predict OS and CSS. Through multivariate Cox regression analysis, determinants that independently affect prognosis were identified. Subsequently, the performance of the models was comprehensively assessed via ROC curves, calibration plots, and DCA. Our results showed that the proposed models exhibited favorable performance regarding discriminative power, calibration accuracy, and clinical applicability. These observations underscore the ability of the nomograms to facilitate individualized risk assessment for elderly patients with advanced EC.

From an epidemiological perspective, the global incidence of EC has continued to rise, particularly among elderly women (12,13). With the ongoing acceleration of population aging, the proportion of elderly patients with EC is steadily rising. This population is frequently accompanied by multiple comorbidities, more aggressive tumor biological behavior, and generally poorer prognosis (14,15). Endometrioid EC represents the predominant histological subtype. Therefore, prognostic evaluation specifically targeting elderly patients with advanced disease is clinically significant. This present study oriented toward patients with stage III–IV endometrioid EC, a population typically presenting with locally advanced tumors and/or distant metastases. Survival outcomes in this population group are influenced by multiple factors, making it challenging to accurately predict prognosis using a single indicator (16).

Several previous studies have identified important prognostic factors for advanced EC, including age, myometrial invasion, histological subtype, tumor grade, and metastasis (17-20). For example, Zouzoulas et al. (17) reported that cervical involvement and histological subtype were significantly associated with recurrence risk in EC patients, whereas other studies indicated that advanced age, cervical involvement, high-grade tumors (G3), and metastasis were independently linked to poor prognostic endpoints (18-20). Together, these observations emphasize the heterogeneity of prognostic factors in advanced EC and support the development of comprehensive prognostic models. In the present study, consistent findings were observed in elderly patients with stage III–IV endometrioid EC, in whom advanced age, higher tumor grade, more advanced T stage, lymph node metastasis, and brain metastasis were identified as independent adverse prognostic factors. Similar to previous reports (18), age remained a significant predictor of mortality. This association is potentially driven by age-driven deterioration in immune function, reduced physiological resilience, and a higher prevalence of chronic comorbidities, all of which may limit tolerance to aggressive treatments (21). Furthermore, high-grade tumors were associated with significantly worse survival outcomes, consistent with evidence that poorly differentiated tumors exhibit more aggressive biological behavior and a greater propensity for metastatic spread (22). Likewise, the adverse prognostic impact of nodal metastasis observed herein aligns with previous findings, which identify nodal involvement as a key indicator of disease progression and systemic dissemination (23). Notably, brain metastasis was identified in the present study as a leading factor predicting unfavorable prognosis. Although brain metastases are relatively rare in EC, previous studies have similarly reported extremely poor survival once intracranial dissemination occurs (24). In addition, treatment-related factors were also markedly associated with survival outcomes. In accordance with current clinical guidelines and previous evidence, surgery remained linked to better survival even among patients with far-advanced disease (25). Moreover, chemotherapy and radiotherapy were considered protective factors, supporting the importance of multimodal treatment strategies in the management of advanced EC (26).

Using the independent prognostic determinants identified in this study, nomograms were subsequently constructed. As a user-friendly prognostic instrument, nomograms can integrate multiple variables to quantitatively assess the risk of individual patients. Compared with traditional staging systems, nomograms generally provide superior predictive accuracy and enhanced capability for individualized prediction (27). ROC analysis showed that, across both the training and validation cohorts, the AUC of the OS model and that of the CSS model were consistently above 0.75, with several time points even exceeding 0.80, suggesting good discriminative performance. The calibration curves further demonstrated the evidence of strong alignment linking predicted outcomes to actual outcomes, suggesting good model reliability. Additionally, The DCA curves indicated that the nomograms provided greater net benefit than either the treat-all or treat-none strategies across a broad range of threshold probabilities, suggesting good clinical utility (28). Recently, with the increasing adoption of molecular classification in EC, risk stratification based on molecular characteristics has gradually become a major research focus. Molecular subtypes, including POLE-mutated, microsatellite instability-high, and p53-abnormal molecular subtypes, have demonstrated important value in prognostic assessment (29,30). Future studies integrating clinical features with molecular biomarkers could allow for the creation of more accurate predictive tools, thereby further improving predictive accuracy.

Although this study has notable strengths—namely, the use of a large population-based cohort from the SEER database and the development of nomograms with satisfactory predictive performance, multiple limitations need to be admitted. First, considering its retrospective observational design, this analysis inevitably carries a risk of selection bias, information bias, and uncontrolled confounders. Although multivariable Cox regression was employed to account for available covariates, the influence of unknown or unmeasured confounders cannot be completely excluded. In particular, treatment allocation within the SEER database is not randomized. Patients who received surgery, chemotherapy, or radiotherapy may have differed systematically from those who did not with respect to performance status, comorbidities, disease burden, treatment eligibility, and physician decision-making. Therefore, the associations observed between treatment variables and survival outcomes may be affected by residual confounding and confounding by indication. As such, these findings should be interpreted as prognostic associations rather than causal treatment effects. Second, several important clinicopathological, treatment-related, and molecular variables were unavailable in the SEER database and therefore were unable to be added to the analyses. These include lympho vascular space invasion, depth of myometrial invasion, detailed chemotherapy information (e.g., specific regimens and neoadjuvant chemotherapy), radiotherapy characteristics (e.g., radiation dose and target volume), and key molecular biomarkers such as POLE mutations, microsatellite instability, and p53 status. Given the well-established importance of these factors for endometrial cancer prognosis, their absence may have introduced residual confounding and limited the predictive accuracy of the constructed models. Third, stage III–IV endometrial EC represents a highly heterogeneous disease spectrum. Patients classified within the same stage may differ substantially with respect to tumor burden, metastatic patterns, radiological findings, and routes of disease dissemination. However, since the SEER database lacks detailed information regarding these characteristics, clinically distinct subgroups could not be analyzed separately, which may have contributed to residual heterogeneity in survival estimates. Fourth, several clinicopathological variables, particularly bone, brain, and liver metastases, were present in only a small proportion of patients (<3%). Consequently, the corresponding HR estimates may have been subject to limited statistical power, wider confidence intervals, and reduced precision. These findings should therefore be treated cautiously and will need to be confirmed by future research with greater sample sizes. Finally, although the models were developed and internally validated using randomly divided training and validation cohorts, external validation in independent populations was not performed. In addition, the study population was restricted to patients with endometrioid EC. Therefore, the degree to which the developed nomograms can be extended across diverse demographic groups, distinct geographical settings, and non-endometrioid histologic variants (e.g., serous or clear cell carcinoma) remains uncertain and warrants further investigation.

Conclusions

The current investigation, through comprehensive analyses of data from the SEER database, systematically identified key prognostic determinants in elderly patients with stage III–IV endometrioid EC and successfully developed nomograms for the prediction of OS and CSS. Age, tumor grade, T stage, N stage, surgery, radiotherapy, chemotherapy, and brain metastasis were identified as important factors influencing patient prognosis. The constructed nomograms demonstrated good discriminative ability and calibration in both the training and validation cohorts, and showed a high net clinical benefit in DCA, suggesting favorable clinical applicability. These models provide a simple and practical tool for personalized survival forecasting and risk grouping, thereby enabling better-informed and more customized treatment choices.

Acknowledgments

We acknowledge the SEER public database for providing the data used in this study. The same dataset is available to other researchers upon completion of the official application and approval procedures.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-0824/rc

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-0824/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-0824/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Siegel RL, Kratzer TB, Wagle NS, et al. Cancer statistics, 2026. CA Cancer J Clin 2026;76:e70043. [Crossref] [PubMed]
2. Zhang L, Zhang J. Cancer statistics, 2026: divergent trends and the implementation gap. Nat Rev Clin Oncol 2026;23:237-8. [Crossref] [PubMed]
3. Rosario-Santos A, Torres-Cintrón CR, López-Rexach AG, et al. A comparative analysis of endometrial cancer disparities in incidence, mortality, and survival between women living in Puerto Rico, Non-Hispanic Blacks, Non-Hispanic Whites, and US Hispanics between 2000-2018. Gynecol Oncol Rep 2023;49:101275. [Crossref] [PubMed]
4. Mager KL, McLean K. Endometrial Cancer: A Review. JAMA 2026;335:1616-27. [Crossref] [PubMed]
5. Cengiz M, Zaim OC, Temiz BE, et al. Perioperative Risk Stratification in Surgically Treated Endometrial Cancer: The Impact of Nutritional Status and Comorbidity Burden on Morbidity and Survival. Ann Surg Oncol 2026; Epub ahead of print. [Crossref]
6. Li T, Jiang D, Fu M, et al. Reported prevalence and risk factors of malignant tumors in the elderly population in China: a nationwide cross-sectional study. Aging Clin Exp Res 2025;37:194. [Crossref] [PubMed]
7. Stone P, Buckle P, Dolan R, et al. Prognostic evaluation in patients with advanced cancer in the last months of life: ESMO Clinical Practice Guideline. ESMO Open 2023;8:101195. [Crossref] [PubMed]
8. Zafar A, Khatoon S, Khan MJ, et al. Advancements and limitations in traditional anti-cancer therapies: a comprehensive review of surgery, chemotherapy, radiation therapy, and hormonal therapy. Discov Oncol 2025;16:607. [Crossref] [PubMed]
9. Iasonos A, Schrag D, Raj GV, et al. How to build and interpret a nomogram for cancer prognosis. J Clin Oncol 2008;26:1364-70. [Crossref] [PubMed]
10. Balachandran VP, Gonen M, Smith JJ, et al. Nomograms in oncology: more than meets the eye. Lancet Oncol 2015;16:e173-80. [Crossref] [PubMed]
11. Hayat MJ, Howlader N, Reichman ME, et al. Cancer statistics, trends, and multiple primary cancer analyses from the Surveillance, Epidemiology, and End Results (SEER) Program. Oncologist 2007;12:20-37. [Crossref] [PubMed]
12. Siegel RL, Kratzer TB, Giaquinto AN, et al. Cancer statistics, 2025. CA Cancer J Clin 2025;75:10-45. [Crossref] [PubMed]
13. Feng T, Li W, Wang Q, et al. Global, regional, and national burden of uterine cancer among women aged 50 years and older from 1990 to 2021: a systematic analysis for the global burden of disease study 2021. J Health Popul Nutr 2025;44:208. [Crossref] [PubMed]
14. Passarello K, Kurian S, Villanueva V. Endometrial Cancer: An Overview of Pathophysiology, Management, and Care. Semin Oncol Nurs 2019;35:157-65. [Crossref] [PubMed]
15. Felix AS, Yang HP, Bell DW, et al. Epidemiology of Endometrial Carcinoma: Etiologic Importance of Hormonal and Metabolic Influences. Adv Exp Med Biol 2017;943:3-46. [Crossref] [PubMed]
16. Keskinkılıç M, Polat G, Bayramoğlu Z, et al. Association of Molecular Classification with FIGO Stage and Survival Outcomes in Endometrial Cancer. Medicina (Kaunas) 2026;62:846. [Crossref] [PubMed]
17. Zouzoulas D, Sofianou I, Markopoulou E, et al. Prognostic Factors in the Treatment of Advanced Endometrial Cancer Patients: 12-Year Experience of an ESGO Certified Center. Cancers (Basel) 2026;18:343. [Crossref] [PubMed]
18. Hag-Yahia N, Gemer O, Eitan R, et al. Age is an independent predictor of outcome in endometrial cancer patients: An Israeli Gynecology Oncology Group cohort study. Acta Obstet Gynecol Scand 2021;100:444-52. [Crossref] [PubMed]
19. Conforti J, Nistor SI, Sadeghi N, et al. Insights from the Pre-Molecular Era in Advanced Endometrial Cancer: Benchmarking Prognostic Indicators in High-Risk Tumours. J Clin Med 2025;14:8726. [Crossref] [PubMed]
20. Yang J, Tian Q, Li G, et al. Identifying risk factors for cancer-specific early death in patients with advanced endometrial cancer: A preliminary predictive model based on SEER data. PLoS One 2025;20:e0318632. [Crossref] [PubMed]
21. Nicholas Z, Hu N, Ying J, et al. Impact of comorbid conditions on survival in endometrial cancer. Am J Clin Oncol 2014;37:131-4. [Crossref] [PubMed]
22. Casanova J, Duarte GS, da Costa AG, et al. Prognosis of polymerase epsilon (POLE) mutation in high-grade endometrioid endometrial cancer: Systematic review and meta-analysis. Gynecol Oncol 2024;182:99-107. [Crossref] [PubMed]
23. Brierley J, Gospodarowicz M, O'Sullivan B. The principles of cancer staging. Ecancermedicalscience 2016;10:ed61. [Crossref] [PubMed]
24. Sambataro D, Gebbia V, Bonasera A, et al. Brain Metastasis in Endometrial Cancer: A Systematic Review. Cancers (Basel) 2025;17:402. [Crossref] [PubMed]
25. Dong Y, Li H, Dai Y, et al. Laparoscopic surgery is effective and safe in high-risk endometrial cancer. Transl Cancer Res 2025;14:8271-84. [Crossref] [PubMed]
26. Ramírez SP, Fidalgo AP, Ginesta MPB, et al. SEOM-GEICO clinical guidelines on endometrial cancer (2025). Clin Transl Oncol 2025;27:4368-80. [Crossref] [PubMed]
27. Lu Y, Sun J, Huang J, et al. Establishment and validation of a prognostic nomogram for overall survival in type II endometrial carcinoma patients. Sci Rep 2025;15:7801. [Crossref] [PubMed]
28. Zhao L, Leng Y, Hu Y, et al. Understanding decision curve analysis in clinical prediction model research. Postgrad Med J 2024;100:512-5. [Crossref] [PubMed]
29. Li ES, Ho R, Tao R, et al. Molecular subtyping of endometrial carcinoma cell lines uncovers subtype-specific targetable vulnerabilities. NPJ Precis Oncol 2025;9:254. [Crossref] [PubMed]
30. Yan G, Zhang P, Cao Y, et al. Universal screening for Lynch syndrome and molecular classification of patients with endometrial cancer. Int J Surg 2026;112:3154-64. [Crossref] [PubMed]