Molecular biomarkers in uterine sarcomas: a review of reviews with genomic data integration.

BACKGROUND

Uterine sarcomas (US) account for 3-7% of uterine malignancies but cause disproportionate high cancer-related mortality. Despite expanding molecular profiling, evidence remains fragmented across histological categories, limiting translation into validated prognostic tools and actionable therapeutic targets.

METHODS

We conducted a comprehensive review, searching MEDLINE, Scopus, Web of Science, and ProQuest. Eligible studies were systematic or high-quality narrative reviews evaluating molecular biomarkers with diagnostic, prognostic, or predictive relevance in US. Methodological quality was assessed using AMSTAR-2 and SANRA, and pooled prevalence estimates derived using random-effects models, complemented by publicly available genomic dataset analysis.

RESULTS

Twenty-nine reviews met inclusion criteria, encompassing leiomyosarcoma (LMS), endometrial stromal sarcoma (ESS), adenosarcoma, perivascular epithelioid cell tumor (PEComa), and uterine tumor resembling ovarian sex cord tumor; after de-duplication, they contributed 71 distinct primary studies comprising 1812 unique patients. Thirty-seven molecular biomarkers with potential clinical relevance were identified. In ESS, grade-stratified analysis separated the two entities sharply: hormone receptor estrogen (ER) and progesteron (PR) were expressed in 82% and 92% of low-grade ESS (LG-ESS) versus 38% and 47% of high-grade ESS (HG-ESS) (between-stratum p < 0.001 and p = 0.002) defining a hormone-responsive, therapeutically actionable profile in LG disease. HG-ESS was characterized by Cyclin D1 overexpression (68%) and by YWHAE-NUTM2A/B (42%) and ZC3H7B-BCOR rearrangements (36%) rearrangements. LMS was characterized by TP53 (68%) and RB1 (48%) co-inactivation. PEComa demonstrated mTOR pathway activation (MelanA 79%), consistent with established mTOR inhibitor sensitivity. High clinical actionability (ESCAT tiers I-II) was confined to rare kinase fusions, while most biomarkers remained at lower evidence levels.

CONCLUSIONS

Molecular profiling is essential for histotype classification in US and increasingly informative for prognostic stratification, yet predictive utility remains largely confined to a subset of tumor types. Prospective biomarker-driven trials are needed to bridge the persistent gap between biological discovery and clinical implementation.