Shanshan Wu
Xu Zhang
Suhan Lai
Xiaohui Yang
Jing Wan- Department of Obstetrics and Gynecology, The Third Affiliated Hospital of Sun Yat-sen University, Guangzhou, China
Background: Endometrial atypical hyperplasia (EAH) and endometrial cancer (EC) are increasingly detected in asymptomatic postmenopausal women. This often leads to delayed treatment. Risk stratification remains challenging, and single-factor models may not accurately identify high-risk individuals. This study aimed to develop and validate a multivariable prediction model for identifying EAH or EC in asymptomatic postmenopausal women.
Methods: This retrospective cohort study included asymptomatic postmenopausal women with endometrial pathology records from the Third Affiliated Hospital of Sun Yat-sen University, China (2021–2024). Candidate risk factors included demographics, clinical characteristics, and hematological parameters. The primary outcome was a composite of histologically-confirmed EAH or EC. Multivariable Poisson regression with robust variance was then employed to identify independent risk factors for this composite outcome. Risk - stratified models were developed by calculating predicted probabilities for key combinations of risk factors.
Results: Among 928 patients [median age: 59 years, IQR (interquartile range): 55–65; median BMI: 23.4 kg/m2], the overall prevalence of EAH and EC was 2.59% (24/928). Key independent risk factors included endometrial thickness (EMT) > 8 mm [vs. ≤ 4 mm: adjusted Relative Risks (aRR) = 11.34, 95% confidence interval (95% CI): 4.35–39.56; p < 0.001], diabetes (aRR = 2.54, 95% CI: 1.12–5.78; p = 0.026), and platelet count per 109/L increase (aRR = 1.01, 95% CI: 1.01–1.02; p < 0.001). EMT > 8 mm was the strongest predictor, with the highest aRR. In the stratified analysis, the combination of EMT > 8 mm and diabetes was associated with the highest observed prevalence (33.33%). The risk-stratified model demonstrated clinical utility: using a ≥ 5% risk threshold, biopsy would be recommended for 17% of patients (sensitivity 70.8%, specificity 84.4%); at a ≥ 10% threshold, biopsy would be recommended for 4.3% of patients (requiring 3.64 biopsies per true positive case).
Discussion: The risk of EAH and EC among asymptomatic postmenopausal women varies significantly based on clinical factors. This risk-stratified modeling approach delivers individualized risk estimates to inform endometrial biopsy decisions, facilitating personalized patient management.
1 Introduction
Endometrial cancer (EC) is one of the most common gynecological cancer, predominantly occurs in postmenopausal women. Driven by population aging and rising obesity rates, the global incidence of EC continues to rise. with 48,931 incident cases documented in China during 2024 (1, 2), while the number of new cases worldwide exceeds 420,000 annually (3). Although 90% of EC patients present with postmenopausal bleeding (4), approximately 5%–15% are diagnosed in asymptomatic women (5). In this subgroup, the absence of alarm symptoms frequently leads to delayed diagnosis, which is strongly associated with adverse prognostic outcomes.
The American College of Obstetricians and Gynecologists (ACOG) (6) and the British Gynaecological Cancer Society (BGCS) guidelines (7) recommend an EMT threshold of >4 mm for evaluating postmenopausal bleeding (PMB). Invasive endometrial sampling procedures should be employed to assess the risk of EC including aspiration (includes pipelle and endometrial biopsy), dilatation and curettage (D&C), and hysteroscopy. These procedures may induce patient anxiety (8), cause moderate-to-severe pain(8, 9), and entail risks of complications including endometritis, pelvic inflammatory disease (PID), and uterine perforation (9–11). However, the threshold of EMT for asymptomatic postmenopausal women remains controversial. Direct applicability of >4 mm EMT cut-off may lead to substantial overuse of invasive biopsy due to their lower baseline risk of malignancy (12). Some have suggested that an EMT of >8 mm measured by ultrasound (13) in asymptomatic populations may be a more optimal threshold (14, 15), but no consensus has been reached.
In addition to EMT, various clinically validated risk factors for EAH or EC include obesity, diabetes, metabolic syndrome, unopposed estrogen exposure (e.g., hormone replacement therapy), and parity. At the same time, blood markers (such as platelet count) (16) also show their prognostic potential.
However, relying solely on a single factor (e.g., EMT threshold) for risk assessment has obvious limitations and may lead to over-biopsy or missed diagnoses. Multifactorial risk prediction models may mitigate these limitations and demonstrate clinical utility in early cancer detection. There have been previous studies to develop EC multivariable prediction models for use in pre- and postmenopausal women. Both showed moderate to good discriminatory ability (C statistic 0.64 and 0.70) (17, 18), However, their development primarily using Western cohorts and incorporation of complex or costly biomarkers (e.g., genetic markers). Furthermore, they are used for primary EC prevention and failed to include EMT as the most important variable.
Therefore, this study aims to: (1) identify independent risk factors and develop a multivariable prediction model for EAH or EC in Chinese postmenopausal women, and then (2) generate individualized risk stratification estimates.
2 Materials and methods
2.1 Population selection
This cross-sectional retrospective study included naturally postmenopausal women who underwent endometrial sampling at the Gynecology Department of The Third Affiliated Hospital of Sun Yat-sen University between 1 January 2021, and 31 December 2024.
Medical records of all eligible patients were systematically reviewed. Participants meeting all inclusion criteria and none of the exclusion criteria were enrolled. Inclusion criteria were: (1) age ≥ 45 years with natural menopause [defined as ≥ 12 months of amenorrhea, excluding cases of bilateral oophorectomy or medications (e.g., GnRH analogs)]; (2) complete endometrial sampling pathology report available; (3) asymptomatic status (absence of vaginal bleeding, vaginal discharge, and pelvic pain). Exclusion criteria were: (1) duplicate encounters or individuals (retaining only the encounter with the most severe pathology); (2) missing EMT data; (3) history of hematological or immunological diseases; (4) current or recent tamoxifen use.
Among 3,003 patients aged ≥ 45 years who underwent endometrial sampling during the study period, we first identified naturally postmenopausal individuals. After excluding 1,422 women reporting vaginal bleeding, abnormal discharge, or pelvic pain (symptomatic), 1,060 asymptomatic subjects remained. We then excluded 71 duplicate records, 30 women with missing EMT data, nine with hematologic or immune disorders, and 22 who had used tamoxifen, leaving 928 unique asymptomatic postmenopausal women for final analysis (Figure 1).
Figure 1. Flowchart of participant selection. EMT, endometrial thickness. Diagram illustrates the process of identifying the final study cohort of asymptomatic postmenopausal women from the initial patient population.
All the procedures performed in studies involving human participants were in accordance with the ethical standards of Ethics Committee of IPMCH and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
2.2 Definition of outcome
Our primary outcome comprised histologically-confirmed endometrial EAH or EC. We have reviewed the complete pathology reports, and categorized the findings into the specific pathological types listed, which includes Normal endometrium/Atrophic endometrium, Endometrial polyp, Simple/complex hyperplasia without atypia, Endometrial hyperplasia with polyp, EAH and EC. The distribution and classification of pathological results are shown in Table 1.
Table 1. Pathological classification of endometrial sampling results.
2.3 Statistical analyses
Descriptive statistics were used to report frequencies and percentages of baseline characteristics. Pathology Prevalence refers to the proportion of patients within each characteristic category who had EAH or EC.
To address the rare outcome, we used Poisson regression with robust variance to analyze risk factor, which is suitable for modeling rare events and adjusting for overdispersion, and less biased than logistic regression for common outcomes.
The following candidate variables were considered: Demographic characteristics [age (years), ethnicity (han ethnicity/other ethnic minorities), occupation (employed/unemployed/retired), education level (illiterate or primary school/secondary school/college or above)]; Clinical characteristics [BMI, maternal history (gravidity/parity), menopausal parameters (age/years), metabolic disorders (diabetes/hypertension), hormone replacement therapy (HRT), history of breast cancer]; Sonographic measure (EMT); and Hematological indicators (complete blood count parameters, coagulation profile, neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), lymphocyte-to-monocyte ratio (LMR)). Univariate analyses were initially conducted to estimate crude relative risks (RR) with 95% confidence intervals (CI). Variables demonstrating a univariate association with P < 0.05 were subsequently entered into a multivariable Poisson regression model to derive adjusted relative risks (aRR) and 95% CIs. Multicollinearity was assessed using variance inflation factors (VIF), with a VIF < 5 indicating no substantial collinearity (Supplementary Table 4).
Sensitivity analyses included: (1) Testing interaction effects between EMT and diabetes mellitus, age at menopause, and platelet count; (2) Restricting outcomes to EC (excluding EAH) to test robustness against premalignant lesion heterogeneity; and (3) Analyzing EMT’s effects in different models: first adjusting only for demographic covariates (age, ethnicity, occupation, education), then metabolic and hormone exposure factors (BMI, diabetes, hypertension, age at menopause, duration of menopause), and finally all covariates from the multivariable analysis.
In the risk stratification analysis, continuous variables were categorized for stratification: platelet count (cut-off: 300 × 109/L, based on prior studies demonstrating its prognostic relevance in solid tumors (19–21)), and age at menopause (cut-off: 50 years, aligned with the median menopausal age in the study population). For each combination of these stratified factors alongside EMT categories (≤4, >4 to ≤8, >8 mm), the following were calculated: (i) Prevalence of EAH or EC and corresponding 95% CI using the Clopper-Pearson exact method; (ii) Predicted prevalence based on the multivariable Poisson regression model; and (iii) Proportion of patients exceeding ≥5% or ≥10% risk thresholds to evaluate screening strategy efficacy.
Given the retrospective design and use of de-identified data, the study protocol was approved by the Ethics Review Committee of the Third Affiliated Hospital of Sun Yat-sen University, exempting informed consent.
All analyses were performed using SPSS 25.0 and R 4.4.2 (statistical significance: two-tailed P < 0.05).
3 Research results
3.1 Characteristics of the study population
This study ultimately enrolled 928 asymptomatic postmenopausal women. Baseline characteristics were as follows: median age 59 years [interquartile range (IQR): 55–65], median age at menopause 50 years (IQR: 49–52), median duration of menopause 8 years (IQR: 4–15), and mean body mass index (BMI) 23.4 kg/m2 (Table 2). Among patients with diabetes, the prevalence of EAH or EC was 6.67%, while among those with hypertension, it was 2.52%. Median EMT is 4.0 mm (IQR: 3.00–6.96 mm). Among the 16.49% (n = 153) of patients with EMT > 8 mm, the prevalence of EAH or EC was 10.46%. Within the overall cohort (n = 928), 24 cases of EAH or EC were identified, yielding an overall prevalence of 2.59%. This comprised 19 cases of EC (2.05%) and 5 cases of EAH (0.54%).
Table 2. Baseline characteristics of the study population.
3.2 Risk factors for EAH or EC
Univariate Poisson regression analysis assessed demographic, clinical, EMT measured via ultrasound, and hematological indicators for their association with EAH or EC risk. Significant predictors identified included age at menopause, diabetes, EMT > 8 mm, and platelet count (all p < 0.05). The results of the univariate analysis for all candidate variables are presented in Table 3. In the subsequent multivariable Poisson model incorporating these significant variables, it was further confirmed that EMT was the strongest predictive factor—patients with EMT > 8 mm had a 11.34-fold increased risk of EAH or EC compared to the EMT ≤ 4 mm group (p < 0.01). Diabetes increased the risk of EAH or EC (aRR = 2.54, 95% CI: 1.12–5.78). Blood markers had limited predictive value, with only platelet count showing a weak association (aRR = 1.01, 95% CI: 1.01–1.02). The complete results of the multivariable analysis are summarized in Table 3. Although age at menopause was significant in the univariate Poisson regression, it did not show significance in the multivariate analysis. Considering its clinically observed impact, we will still include age at menopause in the subsequent model. Other hematological and coagulation parameters analyzed (e.g., PT, APTT, WBC) showed no significant association with EAH or EC risk in univariate analysis (Supplementary Table 1).
Table 3. Univariate and multivariate Poisson regression analysis of endometrial atypical hyperplasia and cancer risk factors.
3.3 Sensitivity analyses
Sensitivity analyses confirmed the robustness of the primary findings. First, poisson regression revealed significant interactions between EMT and diabetes (RR = 1.07, 95% CI: 1.05–1.09), age of menopause (RR = 1.003, 95% CI: 1.001–1.004), and platelet count (PLT) (RR = 1.001, 95% CI: 1.001–1.002) (all p-values < 0.05). For the EMT × diabetes interaction, the RR of 1.07 indicates that for every 1-mm increase in EMT, the risk of EAH or EC is an additional 7% higher in diabetic women than in non-diabetic women, demonstrating a synergistic effect. Second, when the outcome was restricted solely to EC, the key predictive factors remained stable. Finally, cross models with stepwise covariate adjustment, EMT > 8 mm consistently demonstrated the strongest predictive power (all p-values < 0.02). Across stepwise-adjusted models, EMT > 8 mm consistently showed the highest predictive power (all p-values < 0.02). The metabolic-hormonal model yielded an aRR (13.07) closest to the primary model (Supplementary Table 2), supporting synergistic risk amplification by metabolic factors and hormonal exposure duration.
3.4 Observed prevalence of EAH or EC stratified by independent risk factors
Endometrial thickness emerged as the strongest independent predictor for diagnosing EAH or EC in this cohort. Subsequent stratification by diabetes status, elevated platelet count (≥300 × 109/L), and menopausal age (≥50 years) revealed distinct risk patterns. In the group with EMT ≤ 4 mm, the prevalence of EAH or EC was below 1.1% (range 0%–1.08%). When EC increased to 4–8 mm, the prevalence of EAH or EC in patients with diabetes significantly rose to 5.26%, and the prevalence in all subgroups also increased compared to the EMT ≤ 4 mm group (except menopausal age ≥ 50 years). For the EMT > 8 mm group, the baseline prevalence of EMT was 10.46% (95% CI: 6.10%–16.43%); when combined with diabetes, the prevalence sharply increased to 33.33% (95% CI: 13.34%–59.01%); if combined with platelet count ≥ 300 × 109/L, the prevalence reached 24% (95% CI: 9.36%–45.13%) (Table 4).
Table 4. Observed prevalence of endometrial atypical hyperplasia (EAH) or endometrial cancer (EC) stratified by endometrial thickness (EMT) and comorbidities.
3.5 Predicted prevalence and risk threshold benefit analysis
Using a robust multivariate Poisson regression model, we provide a visual representation of the predicted probabilities of EAH or EC for each combination of risk factors, including EMT, diabetes, PLT and age of menopause. These predictions are displayed visually in Figure 2. The highest-risk profile (EMT > 8 mm combined with diabetes, menopausal age ≥ 50 years, and platelet count ≥ 300 × 109/L) yielded a predicted prevalence of 43.58%. Conversely, the lowest-risk profile (EMT ≤ 4 mm without diabetes, menopausal age < 50 years, and platelet count < 300 × 109/L) demonstrated a predicted prevalence of 0.47%. The discriminative ability of the risk-stratified model, as measured by the area under the receiver operating characteristic curve (AUC), was 0.83 (95% CI: 0.73–0.92).
Figure 2. Predicted prevalence of endometrial atypical hyperplasia or cancer based on risk stratification models. EMT, endometrial thickness; PLT, platelet count. Predicted probabilities were derived from the multivariable Poisson regression model.
Considering a ≥ 5% risk threshold, 17% of patients would have met criteria to undergo endometrial sampling. All patients with EMT > 8 mm (risk range 5.32%–43.58%) met this threshold. Among the patients in the EMT 4–8 mm group, those with diabetes who also have either a platelet count ≥ 300 × 109/L or an age at menopause ≥50 years are also included in this risk threshold model. This strategy achieved a sensitivity of 70.8% (17/24 true positive cases detected) and a specificity of 84.4% (763/904 true negative cases correctly spared biopsy). The positive predictive value (PPV) was 10.8% and the negative predictive value (NPV) was 99.1%. Compared to the strategy based solely on EMT > 4 mm, the biopsy rate was significantly reduced. Furthermore, compared to the strategy based solely on EMT > 8 mm, sensitivity increased by approximately 4%. (Supplementary Table 3)
Considering a ≥ 10% risk threshold, 4.3% of patients would have met criteria to undergo endometrial sampling. This strategy specifically targets the highest-risk subgroup, selecting only patients with EMT ≥ 8 mm who also meet at least one of the following criteria: (1) with diabetes (any age or platelet level, minimum risk 13.42%), or (2) platelet count ≥ 300 × 109/L (any age or diabetes status, minimum risk 10.49%). Compared to the strategy based solely on EMT > 8 mm, the biopsy rate was further reduced by 12.2% (absolute reduction), and the number of biopsies required per true positive case detected was optimized to 3.64. However, this improvement in efficiency and reduction in biopsy burden came at the cost of a reduction in sensitivity (13 positive cases missed). For this strategy, the PPV was 27.5% and the NPV was 98.5%. (Supplementary Table 3).
4 Discussion
In this retrospective cohort of 928 asymptomatic postmenopausal women, we performed Univariate Poisson regression analysis to assessed the associations between EAH or EC risk and various factors, including demographic features, clinical variables, ultrasound-EMT, and hematological parameters. Among these, EMT > 8 mm, menopausal age, diabetes mellitus, and platelet count emerged as significant predictive factors.
Endometrial thickness is consistently reported as the strongest independent predictor of EAH or EC, as demonstrated in multiple studies (13, 22). Current ACOG and BJOG guidelines (5, 6) for women with post-menopausal bleeding regard an EMT < 4 mm as carrying a ≥ 99% negative predictive value for EC and therefore recommend biopsy only when EMT exceeds this limit. Whether the same 4 mm rule can be applied to asymptomatic women remains prompting a growing body of research aimed at defining the optimal EMT threshold in the absence of bleeding (13, 23). An Italian prospective cohort suggested that incidentally detected EMT > 8 mm warrants biopsy (11) and a large Chinese cross-sectional study identified 8 mm as the cut-off that maximized the AUC (0.715) (22). Some articles even advocate for a greater endometrial cutoff value for diagnosis (24, 25). To date, however, no consensus has been reached. In our study, EMT > 8 mm emerged as the strongest independent predictor of endometrial atypical hyperplasia or cancer (EAH or EC; aRR = 12.27).
Population-based studies show that women with diabetes have roughly a 70% higher risk of EC (26). Diets that protect against diabetes—low in refined carbohydrates, saturated fat and added sugar—also appear to reduce EC incidence.(27) The common denominator is insulin resistance, which amplifies estrogen signaling and creates a pro-tumor micro-environment.(28) Hyperinsulinaemia acts directly on endometrial epithelium via the insulin and IGF-I receptors to trigger mitogenic pathways (29) and indirectly by raising IGFs, altering sex-steroid availability and perturbing adipokine profiles that govern proliferation and apoptosis (30–33). Our finding of a marked synergistic increase in EC risk is therefore biologically plausible and consistent with the current literature.
A meta-analysis restricted to Asian women showed that every additional year after age 50 at menopause raises endometrial-cancer risk by ≈ 5% (34). The mechanism is thought to be prolonged, unopposed estrogen: persistent proliferative signaling expands the endometrial stem-cell pool, accumulating mutations that heighten malignant risk (35, 36). Anovulatory cycles common in the perimenopause aggravate the imbalance by depriving the endometrium of the countervailing action of progesterone (37, 38). Although age at menopause did not reach statistical significance in the multivariable model, we observed a trend toward increased risk with later menopause, which is consistent with the biological hypothesis that prolonged, unopposed estrogen exposure contributes to endometrial carcinogenesis.
Platelet counts contribute to tumor progression through the release of pro-angiogenic factors and have been widely utilized as a prognostic indicator in various tumors in recent years (39). Elevated platelet levels (300 × 109/L) are associated with poor prognosis in colorectal cancer (40), thyroid cancer (41), and gynecological cancer (42–44). In EC, previous studies have further demonstrated that elevated platelet counts correlate with advanced disease, lymph node involvement, advanced FIGO stage, and poor disease-specific survival (45, 46). Our findings thus align with this evidence, confirming that platelet count serves as an independent predictive factor within our risk-stratified model.
The results of this study demonstrate no statistically significant association between BMI and EC incidence. This lack of association contrasts with the conclusions of most previous studies. First, while about 40% of EC cases were due to overweight and obesity in developed or industrialized countries (47), the mean baseline BMI in this study cohort was comparatively low (23.4 kg/m2), resulting in a significantly lower obesity prevalence and likely contributing to the null finding. Second, Asian populations exhibit a distinct pattern of central adiposity (48), suggesting that waist circumference or waist-to-hip ratio may represent more sensitive indicators of EAH or EC risk than BMI.
Although numerous studies have identified associations between various risk factors and EAH or EC, current diagnostic protocols for asymptomatic postmenopausal women remain limited to ultrasound-EMT as the sole criterion for biopsy referral (14, 24). Few EC risk prediction models which intergrated multi- epidemiological factors have previously been published. The E2C2 (17) and the Predicting risk of EC in asymptomatic women model (PRECISION) (18), showed good performance in quantifying a woman’s 10-year risk of EC. However, E2C2 model (17), which incorporates transcriptomic markers, was derived from a cohort of White women aged 45–85 years. Its reliance on costly genetic testing hinders widespread clinical adoption. The PRECISION model (18), was also developed in Western cohorts with a high prevalence of obesity, limiting its generalizability. Their use could determine eligibility for primary EC prevention. Neither model incorporates EMT, failing to provide direct guidance on which specific populations should be referred for biopsy. While some studies (49) in Asian populations have investigated postmenopausal biopsy strategies, their focus has largely remained on evaluating the predictive value of EMT as a single indicator. In our study, we developed the first stratified risk model for an Asian population to predict the risk of EAH or EC and guide biopsy decision making in asymptomatic postmenopausal women. Our model overcomes the limitation of single-indicator reliance by integrating EMT with other key clinical factors, demonstrated superior discriminative performance, with an AUC of 0.82.
We propose two distinct risk-stratified clinical strategies based on this model. Had the ≥5% risk threshold been applied (endometrial biopsy is recommended when the EC or EAH estimate risk ≥ 5%), only 158 patients would have required biopsy. It achieves higher specificity than the EMT > 4 mm strategy (reducing unnecessary biopsies) and than the EMT > 8 mm strategy (detecting more cancers). It identifies high-risk individuals in the EMT 4–8 mm group who would be missed by the EMT > 8 mm rule, without substantially expanding the biopsy pool. This strategy is particularly suitable for tertiary hospitals or specialized gynecology centers, where comprehensive risk assessment can be readily implemented, including platelet count measurement, diabetes status confirmation, and ensured access to follow-up care. By applying the ≥10% risk threshold strategy, only 4% of patients in our cohort would have required endometrial sampling. This approach maintains acceptable cancer detection rates while significantly reducing the biopsy burden, making it particularly suitable for resource-limited settings such as community clinics or primary care facilities.
This study has several limitations. First, its retrospective design and restriction to hospitalized patients may have introduced selection bias, potentially distorting estimates of EAH or EC prevalence in the full cohort and in subgroups. Second, the cohort’s mean BMI was 23.4 kg/m2, indicating a predominantly non-obese population; consequently, the results may not apply to severely obese individuals (BMI ≥ 30 kg/m2) and need validation in higher-BMI cohorts. Third, the findings are preliminary, and the predictive model requires internal and external validation before it can be considered for clinical application. Future work should recruit larger, prospective cohorts and include rigorous validation to clarify risk factors and incidence of EAH or EC in asymptomatic postmenopausal women.
5 Conclusion
Using multivariable Poisson regression with robust variance, we derived a preliminary risk model that incorporates EMT, diabetes, age at menopause and platelet count to estimate individual probabilities of EAH or EC in asymptomatic postmenopausal women. This exploratory tool may assist clinicians in more selectively considering endometrial biopsy, potentially reducing unnecessary procedures, but requires prospective validation before routine use.
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
The studies involving humans were approved by the ethics committee of the Third Affiliated Hospital, Sun Yat-sen University. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.
Author contributions
SW: Conceptualization, Methodology, Formal analysis, Investigation, Writing – original draft, Writing – review & editing, Supervision. XZ: Conceptualization, Formal analysis, Writing – original draft, Writing – review & editing, Funding acquisition. SL: Investigation, Resources, Writing – review & editing. XY: Writing – review & editing. JW: Conceptualization, Methodology, Writing – review & editing, Supervision.
Funding
The author(s) declare financial support was received for the research and/or publication of this article. This work was supported by the Reproductive Special Project of Clinical Research Fund of Guangdong Medical Association (Grant No. 2025SZ-A1003).
Acknowledgments
We acknowledge all co-authors for their collaborative efforts in the preparation and review of this manuscript. We also thanked the Third Affiliated Hospital, Sun Yat-sen University for its support and our colleagues for their valuable feedback.
Conflict of interest
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.
Generative AI statement
The authors declare that no Generative AI was used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
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.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmed.2025.1707883/full#supplementary-material
Abbreviations
EAH, endometrial atypical hyperplasia; EC, endometrial cancer; EMT, endometrial thickness; HRT, hormone replacement therapy; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; LMR, lymphocyte-to-monocyte ratio; GnRH- analogues, Gonadotropin-releasing hormone analogues; BMI, body mass index; PLT, platelet count; RR, crude relative risks; 95% CI, 95% confidence intervals; aRR, adjusted relative risks; PMB, postmenopausal bleeding; D&C, dilatation and curettage; PID, pelvic inflammatory disease.
References
1. Wu Y, He S, Cao M, Teng Y, Li Q, Tan N, et al. Comparative analysis of cancer statistics in China and the United States in 2024. Chin Med J. (2024) 137:3093–100. doi: 10.1097/CM9.0000000000003442
2. Arnold M, Jiang L, Stefanick ML, Johnson KC, Lane DS, LeBlanc ES, et al. Duration of adulthood overweight, obesity, and cancer risk in the women’s health initiative: a longitudinal study from the United States. PLoS Med. (2016) 13:e1002081. doi: 10.1371/journal.pmed.1002081
3. Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, et al. Global cancer statistics 2022: globocan estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. (2024) 74:229–63. doi: 10.3322/caac.21834
4. Li JXL, Chan F, Johansson CYM. Can a higher endometrial thickness threshold exclude endometrial cancer and atypical hyperplasia in asymptomatic postmenopausal women? A systematic review. Aust N Z J Obstet Gynaecol. (2022) 62:190–7. doi: 10.1111/ajo.13472
5. Smith-Bindman R, Weiss E, Feldstein V. How thick is too thick? When endometrial thickness should prompt biopsy in postmenopausal women without vaginal bleeding. Ultrasound Obstet Gynecol. (2004) 24:558–65. doi: 10.1002/uog.1704
6. American College of Obstetricians and Gynecologists Committee on Gynecologic Practice. ACOG committee opinion. No. 336: Tamoxifen and uterine cancer. Obstet Gynecol. (2006) 107:1475–8. doi: 10.1097/00006250-200606000-00057
7. Sundar S, Balega J, Crosbie E, Drake A, Edmondson R, Fotopoulou C, et al. BGCS uterine cancer guidelines: recommendations for practice. Eur J Obstet Gynecol Reprod Biol. (2017) 213:71–97. doi: 10.1016/j.ejogrb.2017.04.015
8. Ai F, Wang Y, Wang Y, Wang J, Zhou L, Wang S. Clinicopathological features of endometrial lesions in asymptomatic postmenopausal women with thickened endometrium. Menopause. (2022) 29:952–6. doi: 10.1097/GME.0000000000001993
9. Ghoubara A, Emovon E, Sundar S, Ewies A. Thickened endometrium in asymptomatic postmenopausal women - determining an optimum threshold for prediction of atypical hyperplasia and cancer. J Obstet Gynaecol. (2018) 38:1146–9. doi: 10.1080/01443615.2018.1458081
10. Goldstein SR, Khafaga A. Ability to successfully image endometrium on transvaginal ultrasound in asymptomatic postmenopausal women. Ultrasound Obstet Gynecol. (2021) 58:625–9. doi: 10.1002/uog.23667
11. Giannella L, Mfuta K, Setti T, Boselli F, Bergamini E, Cerami LB. Diagnostic accuracy of endometrial thickness for the detection of intra-uterine pathologies and appropriateness of performed hysteroscopies among asymptomatic postmenopausal women. Eur J Obstet Gynecol Reprod Biol. (2014) 177:29–33. doi: 10.1016/j.ejogrb.2014.03.025
12. Visser NC, Sparidaens EM, van den Brink JW, Breijer MC, Boss EA, Veersema S, et al. Long-term risk of endometrial cancer following postmenopausal bleeding and reassuring endometrial biopsy. Acta Obstet Gynecol Scand. (2016) 95:1418–24. doi: 10.1111/aogs.13022
13. Bertelli G, Valenzano M, Costantini S, Rissone R, Angiolini C, Signorini A, et al. Limited value of sonohysterography for endometrial screening in asymptomatic, postmenopausal patients treated with tamoxifen. Gynecol Oncol. (2000) 78(3 Pt 1):275–7. doi: 10.1006/gyno.2000.5876
14. Wolfman W, Bougie O, Chen I, Tang Y, Goldstein S, Bouteaud J. Guideline No. 451: asymptomatic endometrial thickening in postmenopausal women. J Obstet Gynaecol Can. (2024) 46:102591. doi: 10.1016/j.jogc.2024.102591
15. Vitale SG, Buzzaccarini G, Riemma G, Pacheco LA, Di Spiezio Sardo A, Carugno J, et al. Endometrial biopsy: indications, techniques and recommendations. An evidence-based guideline for clinical practice. J Gynecol Obstet Hum Reprod. (2023) 52:102588. doi: 10.1016/j.jogoh.2023.102588
16. Ye Q, Wu Z, Xia T, Liu D, Yang Y, Tang H. Pre-treatment thrombocytosis predicts prognosis of endometrial cancer: a meta-analysis of 11 studies. Exp Ther Med. (2020) 19:359–66. doi: 10.3892/etm.2019.8205
17. Shi J, Kraft P, Rosner BA, Benavente Y, Black A, Brinton LA, et al. Risk prediction models for endometrial cancer: development and validation in an international consortium. J Natl Cancer Inst. (2023) 115:552–9. doi: 10.1093/jnci/djad014
18. Kitson SJ, Crosbie EJ, Evans DG, Lophatananon A, Muir KR, Ashcroft D, et al. Predicting risk of endometrial cancer in asymptomatic women (PRECISION): model development and external validation. BJOG. (2024) 131:996–1005. doi: 10.1111/1471-0528.17729
19. Chatzipanagiotou OP, Tsilimigras DI, Catalano G, Ruzzenente A, Aldrighetti L, Weiss M, et al. Preoperative platelet count as an independent predictor of long-term outcomes among patients undergoing resection for intrahepatic cholangiocarcinoma. J Surg Oncol. (2024) 130:1042–50. doi: 10.1002/jso.27806
20. Wang X, Pu X, Wu X, Wei Y, Wei F, Wu F, et al. Prognostic implication of dynamic platelet count in lung cancer patients with thrombocytosis: a retrospective analysis. PeerJ. (2025) 13:e19551. doi: 10.7717/peerj.19551
21. Wang X, Wang P, Tang B, Xu J, Wang B, Gu L, et al. Platelet count as a double-edged sword: the impact of thrombocytosis and thrombocytopenia on long-term outcomes after hepatic resection for hepatocellular carcinoma. Biosci Trends. (2025) 19:351–60. doi: 10.5582/bst.2025.01067
22. Zhang L, Guo Y, Qian G, Su T, Xu H. Value of endometrial thickness for the detection of endometrial cancer and atypical hyperplasia in asymptomatic postmenopausal women. BMC Womens Health. (2022) 22:517. doi: 10.1186/s12905-022-02089-y
23. Xu J, Rao X, Lu W, Xie X, Wang X, Li X. Noninvasive predictor for premalignant and cancerous lesions in endometrial polyps diagnosed by ultrasound. Front Oncol. (2021) 11:812033. doi: 10.3389/fonc.2021.812033
24. Alcázar JL, Bonilla L, Marucco J, Padilla AI, Chacón E, Manzour N, et al. Risk of endometrial cancer and endometrial hyperplasia with atypia in asymptomatic postmenopausal women with endometrial thickness ≥11 mm: a systematic review and meta-analysis. J Clin Ultrasound. (2018) 46:565–70. doi: 10.1002/jcu.22631
25. Ozelci R, Dilbaz B, Akpınar F, Kınay T, Baser E, Aldemir O, et al. The significance of sonographically thickened endometrium in asymptomatic postmenopausal women. Obstet Gynecol Sci. (2019) 62:273–9. doi: 10.5468/ogs.2019.62.4.273
26. Orgel E, Mittelman SD. The links between insulin resistance, diabetes, and cancer. Curr Diab Rep. (2013) 13:213–22. doi: 10.1007/s11892-012-0356-6
27. Esposito G, Bravi F, Serraino D, Parazzini F, Crispo A, Augustin LSA, et al. Diabetes risk reduction diet and endometrial cancer risk. Nutrients. (2021) 13:2630. doi: 10.3390/nu13082630
28. Pollak MN, Schernhammer ES, Hankinson SE. Insulin-like growth factors and neoplasia. Nat Rev Cancer. (2004) 4:505–18. doi: 10.1038/nrc1387
29. Belfiore A, Frasca F, Pandini G, Sciacca L, Vigneri R. Insulin receptor isoforms and insulin receptor/insulin-like growth factor receptor hybrids in physiology and disease. Endocr Rev. (2009) 30:586–623. doi: 10.1210/er.2008-0047
30. Wolf I, Sadetzki S, Catane R, Karasik A, Kaufman B. Diabetes mellitus and breast cancer. Lancet Oncol. (2005) 6:103–11. doi: 10.1016/S1470-2045(05)01736-5
31. Belfiore F, Iannello S. Reduced insulin sensitivity in Mexican-Americans from San Antonio with elevated incidence of type 2 diabetes compared with Mexicans from Mexico City. Diabetes Care. (2002) 25:937–8. doi: 10.2337/diacare.25.5.937-a
32. Poh QH, Rai A, Carmichael II, Salamonsen LA, Greening DW. Proteome reprogramming of endometrial epithelial cells by human trophectodermal small extracellular vesicles reveals key insights into embryo implantation. Proteomics. (2021) 21:e2000210. doi: 10.1002/pmic.202000210
33. Pollak MN. Insulin-like growth factors and neoplasia. Novartis Found Symp. (2004) 262:84–98; discussion 98-107.
34. Wu Y, Sun W, Liu H, Zhang D. Age at menopause and risk of developing endometrial cancer: a meta-analysis. Biomed Res Int. (2019) 2019:8584130. doi: 10.1155/2019/8584130
35. Pettersson B, Adami HO, Bergström R, Johansson ED. Menstruation span–a time-limited risk factor for endometrial carcinoma. Acta Obstet Gynecol Scand. (1986) 65:247–55. doi: 10.3109/00016348609155179
36. Key TJ, Pike MC. The dose-effect relationship between ‘unopposed’ oestrogens and endometrial mitotic rate: its central role in explaining and predicting endometrial cancer risk. Br J Cancer. (1988) 57:205–12. doi: 10.1038/bjc.1988.44
37. Pike MC, Pearce CL, Wu AH. Prevention of cancers of the breast, endometrium and ovary. Oncogene. (2004) 23:6379–91. doi: 10.1038/sj.onc.1207899
38. Akhmedkhanov A, Zeleniuch-Jacquotte A, Toniolo P. Role of exogenous and endogenous hormones in endometrial cancer: review of the evidence and research perspectives. Ann N Y Acad Sci. (2001) 943:296–315. doi: 10.1111/j.1749-6632.2001.tb03811.x
39. Schlesinger M. Role of platelets and platelet receptors in cancer metastasis. J Hematol Oncol. (2018) 11:125. doi: 10.1186/s13045-018-0669-2
40. Miao Y, Xu Z, Feng W, Zheng M, Xu Z, Gao H, et al. Platelet infiltration predicts survival in postsurgical colorectal cancer patients. Int J Cancer. (2022) 150:509–20. doi: 10.1002/ijc.33816
41. Martin S, Mustata T, Enache O, Ion O, Chifulescu A, Sirbu A, et al. Platelet activation and inflammation in patients with papillary thyroid cancer. Diagnostics. (2021) 11:1959. doi: 10.3390/diagnostics11111959
42. Yang L, Chen H. Establishing the prognostic value of platelet-to-lymphocyte ratio in cervical cancer: a systematic review and meta-analysis. Int J Gynecol Cancer. (2019) 29:683–90. doi: 10.1136/ijgc-2018-000090
43. Njølstad TS, Engerud H, Werner HM, Salvesen HB, Trovik J. Preoperative anemia, leukocytosis and thrombocytosis identify aggressive endometrial carcinomas. Gynecol Oncol. (2013) 131:410–5. doi: 10.1016/j.ygyno.2013.08.032
44. Chen H, Wu Q, Zhang Y, Li Q, Ma J, Kong F, et al. Nomograms based on the novel platelet index score predict postoperative prognosis in endometrial cancer. Gynecol Oncol. (2020) 158:689–97. doi: 10.1016/j.ygyno.2020.05.040
45. Echenique Elizondo M. Preoperative platelet count in pancreas cancer: a prognostic factor? Rev Esp Enferm Dig. (2000) 92:748–57.
46. Kanmaz AG, İnan AH, Beyan E, Budak A. Importance of transvaginal ultrasonography before endometrial sampling in asymptomatic postmenopausal patients. J Gynecol Obstet Hum Reprod. (2019) 48:25–8. doi: 10.1016/j.jogoh.2018.10.018
47. Bray F, Dos Santos Silva I, Moller H, Weiderpass E. Endometrial cancer incidence trends in Europe: underlying determinants and prospects for prevention. Cancer Epidemiol Biomarkers Prev. (2005) 14:1132–42. doi: 10.1158/1055-9965.EPI-04-0871
48. Wong MCS, Huang J, Wang J, Chan PSF, Lok V, Chen X, et al. Global, regional and time-trend prevalence of central obesity: a systematic review and meta-analysis of 13.2 million subjects. Eur J Epidemiol. (2020) 35:673–83. doi: 10.1007/s10654-020-00650-3
Keywords: endometrial cancer, endometrial atypical hyperplasia, postmenopausal women, risk prediction model, endometrial thickness
Citation: Wu S, Zhang X, Lai S, Yang X and Wan J (2025) A risk-stratified model for predicting endometrial atypical hyperplasia and cancer to guide biopsy decisions in asymptomatic postmenopausal women. Front. Med. 12:1707883. doi: 10.3389/fmed.2025.1707883
Received: 18 September 2025; Revised: 22 November 2025; Accepted: 24 November 2025;
Published: 05 December 2025.
Edited by:
Xiaomo Xiong, University of Cincinnati, United StatesReviewed by:
Jieni Li, University of Houston, United StatesYinan Huang, University of Mississippi, United States
Copyright © 2025 Wu, Zhang, Lai, Yang and Wan. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Jing Wan, d2FuamluZ0BtYWlsLnN5c3UuZWR1LmNu
†These authors have contributed equally to this work and share first authorship