The goal of this observational study is to evaluate whether combining postoperative circulating tumor DNA (ctDNA) testing with a novel genomic risk model can better predict recurrence risk and guide adjuvant therapy decisions in patients with endometrial cancer.

The main questions it aims to answer are:

• Can the Clinico-Genomic Risk Score (CGRS) model, which integrates POLE hypermutation, TP53 mutation, FIGO grade, and hormone receptor status, effectively stratify patients into different prognosis groups?
• Is postoperative ctDNA positivity associated with a higher risk of disease recurrence?
• Does combining CGRS and ctDNA provide more accurate risk assessment than either alone?

Researchers will compare the high-risk and low-risk groups defined by the CGRS model, as well as ctDNA-positive and ctDNA-negative groups, to see if these markers can identify patients who need more intensive or less intensive treatment.

Participants will:

• Undergo radical hysterectomy and lymph node dissection as their standard surgical treatment
• Provide blood samples and tumor tissue samples for genetic testing (next-generation sequencing) and immunohistochemistry after surgery
• Receive routine follow-up care with imaging examinations at regular intervals (every 3 months for the first 2 years, every 6 months for years 3-5, and annually thereafter) to monitor for recurrence

Endometrial cancer is one of the most common gynecological malignancies worldwide, with rising incidence and mortality. Traditional prognostic assessment based on clinicopathological factors such as FIGO stage, histologic grade, and hormone receptor status has limitations in accurately stratifying recurrence risk, leading to overtreatment in some patients and undertreatment in others. The TCGA molecular classification has refined prognostic evaluation, but challenges remain due to tumor heterogeneity.

This is a retrospective observational cohort study conducted at the Department of Oncology, Guizhou Provincial People's Hospital, China. A total of 52 patients who underwent radical hysterectomy with bilateral salpingo-oophorectomy and lymph node dissection for endometrial cancer between September 2019 and December 2024 were enrolled. All participants had available formalin-fixed paraffin-embedded (FFPE) tumor tissues and postoperative plasma samples.

Tumor tissue genomic DNA was extracted using the QIAamp DNA FFPE Tissue Kit. Plasma cell-free DNA (cfDNA) was extracted and purified using the QIAamp Circulating Nucleic Acid Kit. Genomic DNA from peripheral blood leukocytes served as the normal control. Comprehensive genomic profiling was performed using a targeted next-generation sequencing (NGS) panel covering 425 cancer-related genes (GeneseeqPrime™). Libraries were sequenced on the DNBSEQ-T7 platform. Immunohistochemical analysis was conducted to assess p53 expression and hormone receptor status.

The Clinico-Genomic Risk Score (CGRS) model was constructed based on TCGA binary decision principles, incorporating four prognostic variables: POLE hypermutation status, TP53 mutation status, FIGO grade, and hormone receptor status. POLE hypermutation serves as the primary stratification indicator. Patients with POLE hypermutation are classified into the low-risk group. For patients without POLE hypermutation, secondary stratification is based on TP53 mutation status. Patients with TP53 mutations are classified into the high-risk group. For patients without POLE hypermutation and with TP53 wild-type, final classification is determined by FIGO grade and HR status. Only patients with FIGO Grade 1 and HR-positive status are classified into the low-risk group; all others are included in the high-risk group.

All patients are followed according to a standardized schedule: imaging re-examination at 4 weeks post-treatment; every 3 months within 1-2 years; every 6 months within 3-5 years; and annually after 5 years. Disease-free survival (DFS), defined as time from surgery to disease recurrence or progression, is the primary endpoint. Overall survival (OS) is the secondary endpoint.

Statistical analyses include Kaplan-Meier survival curves, log-rank tests, and univariable and multivariable Cox proportional hazards regression models to identify independent prognostic factors. External validation of the CGRS model is performed using the TCGA endometrial cancer cohort. All statistical tests are two-sided, with significance set at P < 0.05. Analyses are performed using R software (version 4.4.3).

This study was approved by the Ethics Committee of Guizhou Provincial People's Hospital

Inclusion Criteria:

• 1. Female patients aged ≥18 years with histologically confirmed endometrial cancer.

  2. Underwent radical hysterectomy with bilateral salpingo-oophorectomy and lymph node dissection.

  3. Available formalin-fixed paraffin-embedded (FFPE) tumor tissue specimens. 4. Available postoperative peripheral blood samples for plasma ctDNA analysis. 5. Complete clinicopathological data and follow-up records.

Exclusion Criteria:

• 1. Patients with stage IV endometrial cancer or distant metastasis at diagnosis.

  2. Patients who received neoadjuvant chemotherapy, radiotherapy, or targeted therapy prior to surgery.

  3. Insufficient or poor-quality tissue or plasma samples for NGS analysis. 4. History of other primary malignancies within the past 5 years (except adequately treated non-melanoma skin cancer or carcinoma in situ).

  5. Incomplete clinical data or loss to follow-up.

This study population consists of patients diagnosed with endometrial cancer who underwent radical hysterectomy with bilateral salpingo-oophorectomy and lymph node dissection at Guizhou Provincial People's Hospital, China, between September 2019 and December 2024. All patients had histologically confirmed endometrial cancer and available matched tumor tissue and postoperative plasma samples for NGS and IHC analysis. Patients with stage IV disease or prior neoadjuvant therapy were excluded. The cohort includes FIGO stages I-III, various histologic grades, and all molecular subtypes. A total of 52 patients were enrolled with a median follow-up of 37.5 months.

Individual participant data will not be shared with other researchers due to patient privacy and genomic data confidentiality concerns. De-identified summary results will be made available in the article and supplementary materials. Requests for additional data may be directed to the corresponding author.