Re-stratification of patients with copy-number low endometrial cancer by clinicopathological characteristics

Abstract

Objective To stratify patients with copy-number low (CNL) endometrial cancer (EC) by clinicopathological characteristics. Methods EC patients who underwent surgery between June 2018 and June 2022 at Peking University People’s Hospital were included and further classified according to TCGA molecular subtyping: POLE ultramutated, microsatellite instability high (MSI-H), CNL, and copy-number high (CNH). Clinicopathological characteristics and prognosis of CNL patients were retrospectively reviewed. The Cox proportional hazards regression model was applied to perform univariate and multivariate analysis, and independent risk factors were identified. Differentially expressed genes (DEGs) according to overall survival (OS) were screened based on the transcriptome of CNL cases from the TCGA program. Finally, a nomogram was established, with an accuracy analysis performed. Results (1) A total of 279 EC patients were included, of whom 168 (60.2%) were in the CNL group. A total of 21 patients had recurrence and 6 patients deceased, and no significant difference in recurrence-free survival (RFS) was exhibited among the four molecular subtypes (P = 0.104), but that in overall survival (OS) was statistically significant (P = 0.036). (2) CNL patients were divided into recurrence and non-recurrence groups, and significant differences (P < 0.05) were found between the two groups in terms of pathological subtype, FIGO stage, ER, PR, glycated hemoglobin (HbA1c), and high-density lipoprotein cholesterol (HDL-C). All the above factors were included in univariate and multivariate Cox regression models, among which pathological subtype, PR, and HDL-C were statistically different (P < 0.05), resulting in three independent risk factors for the prognosis of patients in the CNL group. (3) By comparing the transcriptome of tumor tissues between living and deceased CNL patients from the TCGA database, 903 (4.4%) DEGs were screened, with four lipid metabolism pathways significantly enriched. Finally, a nomogram was established, and internal cross-validation was performed, showing good discrimination accuracy with an AUC of 0.831 and a C-index of 0.748 (95% CI 0.444–1.052). (4) According to the established nomogram and the median total score (85.89), patients were divided into the high score group (n = 85) and low score group (n = 83), and the 8 patients with recurrence were all in the high score group. Survival analysis was performed between the two groups, and the difference in RFS was statistically significant (P = 0.010). Conclusion In the CNL group of EC patients, pathological subtype, PR, and HDL-C were independent prognostic risk factors, the nomogram established based upon which had a good predictive ability for the recurrence risk of patients with CNL EC.