Shanping Shi, Zhiping Wu, Beiying Wang, Xiaocui Li
AS and UCEC showed partially overlapping descriptive features of stromal remodeling. SLC24A3 remains a nominal hypothesis-generating candidate; the association did not remain significant after multiple-testing correction, and colocalization evidence was low/inconclusive. These findings do not establish a disease-level causal relationship or a shared causal mechanism.
BACKGROUND: Asherman's syndrome (AS) and uterine corpus endometrial carcinoma (UCEC) involve fibroblast-driven stromal remodeling but are clinically and biologically distinct. We integrated single-cell RNA sequencing (scRNA-seq) and Mendelian randomization (MR) to compare stromal features and assess genetically proxied expression-UCEC associations.
METHODS: scRNA-seq of endometrium from AS, UCEC, and controls was used to map cell states and myofibroblastic stromal features. The 583-gene candidate set informed an exploratory LD-aware MR analysis of 351 testable genes using whole-blood cis-eQTL and UCEC GWAS data.
RESULTS: Eight major lineages and two recurrent fibroblast states were identified, with disease-specific compositional and transcriptional shifts. CellChat outputs were interpreted as expression-inferred and descriptive. In the primary analysis, SLC24A3 showed a nominal inverse association with UCEC risk (2 instruments; OR 0.8771, 95% CI 0.7860-0.9787; P = 0.0190), which did not remain significant after BH correction across 351 testable genes (FDR = 0.5142) or Bonferroni correction across 583 candidates (P = 1.000). Colocalization provided low/inconclusive evidence of a shared causal variant (PPH4 = 0.1492; PPH3 = 0.1220).
CONCLUSIONS: AS and UCEC showed partially overlapping descriptive features of stromal remodeling. SLC24A3 remains a nominal hypothesis-generating candidate; the association did not remain significant after multiple-testing correction, and colocalization evidence was low/inconclusive. These findings do not establish a disease-level causal relationship or a shared causal mechanism.