Endometrial carcinoma (EC) is the sixth most diagnosed cancer among women (1). An estimated 66,570 new cases of uterine corpus cancer and 12,940 deaths were reported in the United States in 2021 (2). EC is usually diagnosed during stage I, and patients have a good prognosis as it induces symptoms from an early stage (3). Postmenopausal vaginal bleeding is the most common symptom of EC (3, 4). Additionally, tumor invasion of the cervix can lead to blood or pus in the uterine cavity, causing abdominal swelling and cramping pain. Patients having advanced disease may suffer pelvic and lumbosacral pain due to the invasion of the tumor within the surrounding tissues or nerves (3, 4). EC is classified as type I and type II based on Bokhman’s dualist model⁵. Type I EC is estrogen-dependent and accounts for nearly 80% of all EC. Its pathological type is primarily endometrioid carcinoma (3–5). Endometrial clear cell carcinoma (ECCC) is a type II EC accounting for approximately 2–4% of the total EC, and is more common in older women (6). ECCC is an estrogen-independent tumor whose onset has no apparent relationship with estrogen (4, 5, 7). ECCC is more aggressive and prone to early metastasis than endometrioid carcinoma (4, 5, 7). Many studies report a 5-year survival rate of less than 50%, irrespective of the ECCC clinical stage. However, all these studies included small samples having limited persuasion (4). Only a few small retrospective cohort studies and some case reports have explored the prognostic factors in ECCC due to its low incidence. Notably, there are no systematic analyses of ECCC from a large population sample. Therefore, the current study aimed to perform a comprehensive retrospective analysis depending on the Surveillance, Epidemiology, and End Results (SEER) database to evaluate the survival and prognostic risk factors for ECCC. Moreover, it establishes definitive individualized prognostic prediction models to predict the 3-, 5- and 10-year overall survival (OS) in ECCC patients. The findings contribute to developing appropriate treatment and follow-up strategies for ECCC.

Unlike EC, most ECCC patients have a tumor negative for the estrogen and progesterone receptors. However, positive for hepatocyte nuclear factor 1β and Napsin A. Notably, TP53 is the most commonly mutated gene in ECCC (4, 9–11). The abnormal p53 expression is considered a poor prognostic factor for EC (11). Previous studies observed that the mutation rate of the TP53 gene in POLE wild-type ECCC is 46%, while that of non-POLE endometrioid carcinoma is only 11% (11). ECCC patients are accompanied by high-risk factors for poor prognoses, including advanced clinical stage, deep muscular infiltration, lymphovascular space involvement, and distant metastasis, with a high recurrence rate, high mortality, and poor prognosis than in type I EC (3, 4, 7). Currently, The Cancer Genome Atlas (TCGA) classification is the most authoritative classification system of EC. However, it does not include ECCC patients. Therefore, it is essential to analyze the demographic and clinicopathological characteristics of ECCC patients independently. Moreover, we must comprehensively evaluate their prognosis to develop an adequate treatment guide for ECCC patients.

Our study identified that age was an essential prognostic factor among ECCC patients, positively correlating with the risk of death. The findings of this study concerning the relationship between age and prognosis in EC patients were consistent with previous studies. A retrospective study found that patients aged ≤ 40 include more favorable prognostic factors, such as a higher proportion of non-invasive carcinoma, a lower proportion in the uterine segment involvement, and less invasion of the lymphatic vascular space than in EC patients aged 40-60 years (12). Furthermore, EC patients aged ≤ 40 years had a lower probability of mismatch repair defects due to MLH1 methylation, a mutation associated with poor prognosis, than patients aged 41-60 years (12). Another study also found that ECCC patients aged ≥ 70 had worse progression-free survival time and OS independent of the treatment modality they were subjected to (13). An investigation on the influence of marital status on the diagnosis and prognosis of EC revealed that marriage was a protective prognostic factor for OS and cancer-specific survival among EC patients. Unmarried, divorced/separated, and widowed patients showed a higher risk of death than married patients (14). This phenomenon was because married patients were more likely to be diagnosed early, possibly due to the stability of the endogenous hormone levels in women associated with emotional benefits (14). In this study, separated and widowed patients having early ECCC had a higher risk of death than married patients at the same stage. However, marital status had no significant effect on the prognosis of patients with advanced ECCC. Previous studies have evaluated the relationship between tumor diameter, myometrium invasion depth, and prognosis of EC patients. Nilufer et al. observed that more than half of the ECCC patients with a tumor diameter > 2 cm were prone to myometrial invasion (15). Kohei et al. identified that large tumor size is associated with deeper myometrial infiltration and lymph node metastasis among EC patients (16). In this study, ECCC patients with large tumor sizes and late T stages significantly enhanced the risk of death. These findings were consistent with a retrospective study that inferred that large tumor size and deep muscle invasion could be associated with poor prognosis among ECCC patients (17). Lymphatic metastasis is the main route of EC metastasis. The survival time of patients is significantly shortened once they develop lymph node metastasis, indicating disease progression (18). This study also depicted that OS is significantly decreased in ECCC patients with lymph node involvement.

A study revealed that black patients with EC were more likely to develop invasive and non-endometrioid cancer than white EC patients from America (19). However, this study did not observe a correlation between race and prognosis in ECCC patients. The degree of differentiation was not an independent prognostic factor for ECCC. Therefore, we hypothesized that the prognosis was poor irrespective of the degree of differentiation due to the high invasiveness of ECCC and thus had no significant effect on the prognosis of ECCC patients.

The preferred treatment method to cure ECCC is extensive staging surgery, including total uterine and bilateral adnexectomy, pelvic and para-aortic lymph node dissection, more significant omentum biopsy, and examination of the peritoneal washing fluid (20–22). The advantage of surgery is that the tumor stage is more accurately identified, facilitating the subsequent selection of the appropriate adjuvant treatment. Our results revealed that total hysterectomy was a favorable factor for a good prognosis. The risk of death after total hysterectomy was lower than after partial hysterectomy in both early and advanced stages. Therefore, active surgical treatment was recommended for ECCC patients. Patients who cannot undergo radical surgery should also be treated with tumor-reducing surgery, depending on their physical condition. Adjuvant radiotherapy and chemotherapy are fundamental approaches in treating ECCC. Numerous studies underline that the choice of adjuvant radiotherapy and chemotherapy is associated with the stage of ECCC (23). The adjuvant therapy in patients with early ECCC should be chosen based on prognosis-related factors, such as age and the depth of myometrial invasion. Although our results depicted that radiotherapy and chemotherapy had no role in improving the prognosis of patients with early ECCC, this factor did not hinder patients with early ECCC from benefiting from radiotherapy and chemotherapy. Our results were attributed to the SEER database limitations, which did not allow us to know the adjuvant treatment regimen and course, thus preventing the specific stratification study of the enrolled patients.

The FIGO stage of EC represents the pathological surgical stage, which includes factors related to patient prognoses, such as depth of muscular invasion, nodal metastasis, and distant metastasis. It is the primary tool clinicians use to evaluate the prognosis of EC patients. However, the FIGO stage does not include other factors associated with the survival of patients, such as age, marital status, and treatment methods. At the same time, the nomograms contain the demographic, clinical characteristics, and therapeutic approaches of the ECCC patients. Additionally, the DCA curves applied to ECCC patients established that nomograms had better clinical benefits than traditional FIGO stages in stages I/II and III/IV. Therefore, the nomograms had a significant practical value due to their good accuracy, excellent discrimination ability, and clinical benefits.

Compared to the existing prognostic classification, our predictive model demonstrated several strengths. All the clinical variables included in the survival prediction model were easily accessible. This study enrolled ECCC patients; thus, our nomogram was more applicable to ECCC patients than other classification systems. Moreover, the nomogram intuitively and clearly showed 3-, 5- and 8-year survival rates, which is convenient for clinicians. However, this study had several limitations. Firstly, all the variables selected in our study were clinical characteristics. Several genetic and epigenetic features, including Non-Coding RNAs, identified as predictors of EC patients in previous studies (24, 25), were not included in this study due to the limitations of the SEER database. The absence of these new molecular characteristics deteriorated the practicability of the nomogram model. Secondly, this was a retrospective study; thus, the bias significantly affected the results because the information about the patients was partially missing. For instance, the tumor size of 43.2% of the patients was unknown, which significantly reduced the accuracy of the prediction model. Finally, it was unclear whether the patients received neoadjuvant therapy, and the specific information based on surgery, radiotherapy, chemotherapy, and a potential targeted therapy was unknown.

Nomograms for predicting 3-, 5-, and 10-year OS in ECCC patients were successfully constructed. Moreover, new risk stratification systems were further built to stratify patients into different risk groups. These predictive models may be valuable tools to aid ECCC management and treatment in clinical practice.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ZL, XC, and PC conceived and designed the study. YZ and HZ helped in data collection. PC and YZ performed the analysis and wrote the manuscript. All the authors have reviewed the final draft of the manuscript and approved it for publication.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.