# Prognostic Stratification of Multiple-Classifier Endometrial Cancers: Cohort Study and Meta-Analysis

## TL;DR

This study explores endometrial cancers with multiple molecular features, showing they are biologically diverse and challenging to classify, suggesting the need for better standardized testing.

## Contribution

The study identifies and characterizes biologically heterogeneous multiple-classifier endometrial cancers, emphasizing the limitations of current classification systems.

## Key findings

- POLEmut-MMRd/MSI tumors retain POLE-mutated features with favorable clinicopathological characteristics.
- MMRd/MSI-p53abn/TP53mut tumors show adverse features and lower tumor mutational burden.
- Multiple-classifier ECs have a prevalence of ~5.4%, with variability due to testing strategies.

## Abstract

The classification of endometrial cancer has evolved using molecular features, allowing for improved risk assessment and more personalized treatment strategies. However, a small proportion of tumors show more than one relevant molecular alteration, making their classification and clinical management more challenging. This study aimed to characterize molecular and clinicopathological profiles of multiple-classifier endometrial cancers, enhancing our understanding of their biological heterogeneity. In our patient cohort, POLEmut tumors with concurrent MMRd/MSI generally retained the POLE-associated ultramutated profile, while tumors with both MMRd/MSI and p53abn/TP53mut alterations were more often associated with more adverse clinicopathological characteristics compared to MMRd/MSI-only tumors. By integrating our data in a systematic literature review with meta-analysis, we observed that variability in reported incidence is largely driven by differences in testing strategies. These results highlight the limitations of current classification systems and emphasize the importance of more standardized molecular approaches to improve risk stratification and management in endometrial cancer.

Background/Objectives: The integration of molecular classification has significantly refined prognostic stratification in endometrial carcinoma (EC). However, a tumor subset harboring more than one molecular classifier challenges the current hierarchical classification systems, and their biological and clinical significance remains incompletely defined. Methods: We analyzed an expanded cohort of 150 EC patients using next-generation sequencing and immunohistochemistry, integrating molecular and clinicopathological data. Multiple-classifier ECs were identified and compared with those of single-classifier tumors sharing at least one molecular feature. A systematic review and meta-analysis including our cohort were also performed to estimate the incidence and distribution of multiple-classifier ECs. Results: In our cohort, 6% of ECs harbored molecular multiple-classifiers. POLEmut-MMRd/MSI tumors generally retained POLE-mutated features, including ultrahigh tumor mutational burden (TMB), early-stage disease, and favorable clinicopathological characteristics, consistent with a highly immunogenic phenotype. In contrast, MMRd/MSI-p53abn/TP53mut tumors were more frequently associated with adverse clinicopathological features and showed lower and heterogeneous TMB values, suggesting that the coexistence of TP53 alterations may modify the typical intermediate-risk of MMRd/MSI-only tumors. Triple-classifier tumors were exceedingly rare, precluding definitive conclusions regarding prognosis. The meta-analysis demonstrated an overall prevalence of multiple-classifier ECs of approximately 5.4%, with substantial inter-study heterogeneity largely attributable to differences in molecular testing strategies, analytical sensitivity, and variant interpretation criteria. Conclusions: Multiple-classifier ECs represent a small but clinically relevant subset encompassing biologically heterogeneous entities. Our findings highlight the limitations of current molecular classification hierarchies and underscore the need for harmonized molecular testing and standardized reporting to improve risk stratification and the management of multiple-classifier ECs.