Refining Fertility-sparing Treatment in Endometrial Carcinoma Based on Molecular Classification

1.Research Background Endometrial Carcinoma (EC) ranks among the most prevalent malignancies of the female reproductive system, with a globally increasing incidence and a notable trend toward younger age at onset, leading to a growing disease burden. Data from the U.S. National Cancer Institute indicates that the annual incidence of EC has increased by 1.5% over the past decade, with a particularly prominent rise among women under 50 years of age. The situation is equally severe in China: between 2004 and 2017, the crude incidence of EC rose from 6.20 to 10.06 per 100,000 population. In economically developed regions such as Beijing, Shanghai, and Tianjin, EC has become the most common gynecological malignancy. For Chinese women under 40, the crude incidence increased by 68.20% (from 0.63 to 1.07 per 100,000 in rural areas). This trend is closely associated with delayed childbearing, reduced parity, obesity, and the high prevalence of metabolic diseases like diabetes, predicting a continuous increase in EC-related burden over the next decade and highlighting the urgent need for systematic prevention and control strategies.

For young nulliparous EC patients, radical surgery-though the standard treatment-results in irreversible loss of fertility. Consequently, Fertility-Sparing Treatment (FST) has emerged as the core approach to balance tumor control and reproductive needs. However, FST for young EC patients currently faces critical bottlenecks:

1. Narrow indications and limited treatment options: Guidelines including NCCN， ESGO, and BGCS restrict FST eligibility to patients with well-differentiated (G1), FIGO 2009 Stage IA (confined to the endometrium) disease. Clinically, high-dose potent progestogens (medroxyprogesterone acetate [MPA], megestrol acetate [MA], and levonorgestrel-releasing intrauterine system [LNG-IUS]) remain the mainstay, yielding an overall complete response (CR) rate of only 70%-80%⁶.
2. Significant efficacy variability and prominent adverse effects: Even with our research team's modified regimen of "hysteroscopic lesion resection combined with high-dose progestogens"-which elevated the pathological CR rate to 97% (mean treatment duration: 6.42±0.45 months)-approximately 1/3 of patients require over 9 months of treatment or fail to achieve CR. Prolonged treatment not only increases the risk of endometrial damage due to repeated hysteroscopic evaluations and impairs ovarian function (affecting subsequent pregnancy) but also raises the risk of progestogen-related adverse events such as weight gain, liver dysfunction, thrombosis, and breast cancer. Concurrent symptoms including dizziness, headache, and gastrointestinal disturbances further reduce patients' quality of life and treatment adherence.
3. Lack of molecular subtype-guided precision protocols: The EC molecular classification system proposed by The Cancer Genome Atlas (TCGA) in 2013 (currently categorized into POLEmut, MMRd/MSI-H, NSMP, and p53abn subtypes) has revealed significant differences in biological characteristics and prognosis across subtypes (p53abn subtype has the poorest prognosis and is not suitable for FST; POLEmut subtype has the best prognosis; MMRd/MSI-H and NSMP subtypes have intermediate prognosis). While molecular subtyping has been incorporated into international guidelines such as ESGO/ESTRO/ESP¹¹ and ESMO to guide non-fertility-sparing EC management, research on subtype-specific personalized regimens, prediction of treatment non-responders, and alternative therapies in the FST setting remains severely insufficient, failing to meet clinical demands for precision medicine.

Advancements in EC molecular subtyping research and accumulating clinical evidence for immune checkpoint inhibitors (ICIs) have provided new opportunities to address these bottlenecks:

• POLEmut EC exhibits high tumor mutational burden (TMB-H) and robust immune responses, indicating significant potential for ICI treatment;
• Although MMRd/MSI-H EC shows poor response to progestogens, ICIs have demonstrated favorable efficacy in advanced and locally advanced cases;
• As the most common subtype, NSMP EC may benefit from progestogen therapy optimization through biomarker-based stratification.

Against this backdrop, conducting prospective clinical trials of molecular subtype-guided FST for EC has become an urgent priority to optimize FST strategies, improve therapeutic efficacy, and enhance reproductive and oncological outcomes.

2 Research Basis and Protocol Design for Each Molecular Subtype

1. POLEmut EC POLEmut EC accounts for approximately 10% of all EC cases, characterized by one of 11 pathogenic mutations in the POLE exonuclease domain (patients with non-pathogenic mutations have poorer prognosis), TMB-H, and dense tumor-infiltrating lymphocytes (TILs) with high PD-1 expression. Clinically, it exhibits excellent prognosis: the PORTEC-3 study showed that high-risk POLEmut EC patients achieved 5-year recurrence-free survival (RFS) rates of 100% and 97% with and without chemotherapy, respectively (p=0.64), and a 5-year overall survival (OS) rate of approximately 98%. The FIGO 2023 staging system classifies all uterine-confined POLEmut EC as Stage IA (Stage II patients are downgraded to Stage I for surveillance), and the RAINBO POLEmut-BLUE trial is currently exploring the feasibility of treatment de-escalation for this subtype. In the FST setting, hysteroscopic lesion resection remains the primary intervention, while systematic research on ICI application is lacking-only isolated retrospective studies have reported progestogen resistance or disease progression/recurrence in some patients. Preliminary data from our team showed that all 14 POLEmut EC patients receiving FST (including FIGO 2023 Stage IA2 G2, Stage IC G3, and progestogen-insensitive cases) achieved CR; an additional 6 POLEmut EC patients treated with ICIs also attained CR. Based on this evidence, this study proposes PD-1 monoclonal antibody monotherapy for POLEmut EC patients with strong fertility preservation demands, while expanding FST eligibility to FIGO 2023 Stage IA. The primary objectives are to evaluate FST efficacy, oncological outcomes, and reproductive outcomes of ICIs.
2. MMRd/MSI-H EC MMRd/MSI-H EC constitutes 25%-33% of all EC cases, featuring TMB-H (≥10 mutations/Mb), dense intratumoral CD8+ T cells, tertiary lymphoid structures, and high PD-1/PD-L1 expression. However, its response rate to progestogens is significantly lower than that of NSMP EC, with higher recurrence rates (potentially associated with activation of progesterone receptor-independent pathways). Currently, ICIs are approved for the treatment of pretreated advanced MMRd/MSI-H EC (objective response rate: 39%-49.1%, with durable efficacy and manageable toxicity). Early-phase clinical trials have further confirmed their potential: in a Phase II trial involving resectable Stage I-III MMRd solid tumors (including EC), 49 patients with MMRd locally advanced rectal cancer achieved clinical CR after 6 months of PD-1 inhibitor (dostarlimab) neoadjuvant therapy and avoided surgery; among 54 patients with non-rectal MMRd solid tumors, 35 achieved clinical CR (33 avoided surgery). Of 103 patients completing treatment across both cohorts, 84 achieved clinical CR (82 avoided surgery), with a 2-year RFS rate of 92% among 117 patients and 95% of adverse events being reversible Grade 1-2. In the EC-specific setting, 12 out of 15 patients with resectable Stage I-III MMRd/MSI-H EC achieved clinical CR after 6 cycles of nivolumab (480mg Q4W); 7 of these patients avoided surgery with no recurrence during follow-up (first enrollment: February 2019). Additionally, literature has reported successful pregnancy in a Lynch syndrome (MMRd-associated) patient after achieving CR with PD-1 monoclonal antibody; our team's preliminary data showed CR in all 4 Lynch syndrome patients (including those refractory to conventional progestogen therapy) treated with PD-1 monoclonal antibodies, suggesting ICIs may outperform progestogens. Notably, MMRd exhibits internal heterogeneity: based on genetic mechanisms, it can be classified into MLH1 promoter methylation subtype (poorer response to immunotherapy) and MMR gene mutation subtype (nearly 100% response rate to immunotherapy in patients with germline mutations, including Lynch and Lynch-like EC). Therefore, this study proposes an "ICI ± progestogen" regimen for MMRd EC patients undergoing FST: patients with confirmed germline or somatic MMR gene mutations will receive ICI monotherapy, while those with MLH1 methylation will receive ICI combined with high-dose potent progestogens. FST eligibility will be expanded to FIGO 2023 Stage II, with objectives to evaluate regimen efficacy, oncological/reproductive outcomes, and analyze efficacy differences and potential mechanisms across MMRd patients with distinct genetic backgrounds.
3. NSMP EC As the most common EC subtype (32%-49% of all cases), NSMP EC can be stratified into high-risk and low-risk groups based on ER expression. The low-risk subtype, predominantly ER-positive endometrioid carcinoma, represents the main population for FST, with high-dose progestogens as the standard regimen. Retrospective studies showed that NSMP EC patients undergoing FST achieved a 6-month CR rate and overall CR rate (72%, 18/25) significantly higher than ER/PR-positive MMRd patients (37.5%, 6/16) in the same period, with a lower recurrence rate (17.4% vs. 83.3%, p<0.05). However, clinically, 70%-80% of patients with early EC/atypical endometrial hyperplasia (AEH) require 9-10 months to achieve CR, and approximately 1/3 of patients either require over 9 months of treatment or remain non-responsive. Our team's preliminary research has identified PTEN mutation²⁷, SIGLEC10 mutation, insulin resistance, and obesity as key predictive factors for progestogen insensitivity, and established a progestogen sensitivity prediction model.

Simultaneously, our team discovered that progestogen-insensitive/refractory EC patients can still achieve CR with the "gonadotropin-releasing hormone agonist (GnRHa) combined with aromatase inhibitor (AI, mainly letrozole)" regimen. This regimen inhibits EC proliferation by suppressing estrogen production in the ovaries and peripheral tissues: GnRHa inhibits ovarian function and reduces estrogen levels within 14 days of administration, while letrozole blocks aromatase activity to inhibit the conversion of androstenedione to estrone/estradiol-representing a mechanism distinct from progestogens. Clinical studies have verified its feasibility: Zhang et al. reported a 100% CR rate (CR duration: 3-6 months) in 6 young obese EC patients treated with this regimen; Zhou et al. achieved CR rates of 88.2% (15/17) in early endometrioid EC (EEC) and 100% (12/12) in AEH patients using "GnRHa combined with LNG-IUS or letrozole" (median treatment duration: 18.7 months);A small-sample retro spective study showed that EEC/AEH patients achieving CR with this regimen had significantly longer disease-free survival than those treated with high-dose progestogens (hazard ratio: 2.158; 95% CI: 0.948-4.913). Additionally, this regimen avoids significant weight gain, causes minimal liver/kidney dysfunction, and prevents adverse events such as vaginal bleeding and breast cancer, thereby improving patient adherence. The ovarian protective effect of GnRHa may also enhance pregnancy rates after CR. However, attention should be paid to GnRHa-induced "medical ovarian ablation" complications (hot flashes, night sweats, osteoporosis) and the long-term hypoestrogenic effects of AIs on bone metabolism, which require close follow-up and timely intervention in the study.

Based on the above evidence, this study proposes progestogen sensitivity prediction model-guided stratified precision treatment for low-risk NSMP EC patients: patients predicted to be progestogen-sensitive will receive classical high-dose potent progestogens (MA/MPA), while those predicted to be progestogen-insensitive will receive "GnRHa combined with letrozole".

3 Research Objectives

To evaluate the FST efficacy and reproductive outcomes of different fertility-sparing regimens based on EC molecular subtypes. As the world's first prospective multicenter umbrella clinical trial optimizing EC FST protocols based on molecular characteristics, this study aims to:

1. Develop personalized FST regimens for POLEmut, MMRd/MSI-H, and NSMP EC subtypes-namely "ICI monotherapy", "ICI ± progestogen", and "progestogen sensitivity-stratified therapy"-to optimize existing FST strategies;
2. Expand FST eligibility to FIGO 2023 Stage IA for POLEmut EC and FIGO 2023 Stage II for MMRd/MSI-H EC, breaking the limitations of current indications;
3. Improve FST CR rate, shorten CR duration, reduce adverse events, and enhance patients' reproductive and oncological outcomes;
4. Explore and validate FST efficacy predictors and prognosis-related biomarkers for each subtype, improve the accuracy of disease risk control and prognostic assessment, and provide high-quality clinical evidence for optimizing EC FST protocols.

4 Research Methods Study Design: Prospective, open-label, multicenter, umbrella clinical trial Study Population: Patients with early-stage endometrial carcinoma requiring fertility-sparing treatment

This is a non-randomized controlled, prospective, multicenter, umbrella clinical trial implementing stratified management based on molecular subtypes:

1. POLEmut Stage II Single-Arm Interventional Study: Enroll patients with FIGO 2023 Stage IA POLE-mutant endometrial carcinoma to evaluate the CR rate, CR duration, and reproductive outcomes of ICI monotherapy in FST with expanded indications. Enrolled patients will receive ICI monotherapy.
2. MMRd Stage II Single-Arm Interventional Study: Stratify patients based on MLH1 methylation status/MMR coding gene mutation status. Patients with MMR coding gene mutations (somatic/germline) will receive ICI monotherapy; MMRd patients with MLH1 promoter methylation will receive ICI combined with high-dose potent progestogens (PD-1 + MPA/MA). The study will evaluate the CR rate, CR duration, and reproductive outcomes of FST with expanded indications.
3. NSMP Stage III Two-Arm Interventional Study: Based on the prediction model, progestogen-sensitive patients will receive high-dose potent progestogens (MA/MPA); progestogen-insensitive patients will receive "GnRHa combined with letrozole". The study will evaluate FST efficacy and reproductive outcomes while further validating the predictive value of the model.
4. exclusion from FST: NSMP EC with ER negativity and/or L1CAM positivity (≥10%) or p53abn endometrial carcinoma.