Molecular subtyping of endometrial cancer via a simplified one-step NGS classifier, ARID1A and ZFHX4 mutations help further subclassify CNL/MSI-H patients
Abstract
Background
Molecular subtyping has changed the prognostic stratification and therapeutic guidance for patients with endometrial cancer (EC). However, simultaneous application of sanger sequencing and immunohistochemistry under ProMisE criteria may be time- and tissue-consuming. This study attempted to measure subtype-specific biomarkers by one-step next-generation sequencing (NGS) resulting in a shorter turnaround time and less requirement of tissue samples.
Methods
FFPE samples from 233 EC patients were retrospectively collected. Overall survival (OS) information was available for 131 patients with a median follow-up of 66 months. Genomic DNA was extracted and subjected to a one-step NGS panel including TP53, POLE and MSI measurement. Further comprehensive genomic analyses were performed on DNA from MSI-H and copy number low (CNL) subtypes.
Results
The molecular typing ratio of the 233 patients was 8.15% for POLE subtype, 18.88% for MSI-H subtype, 11.59% for copy number high (CNH) subtype and 61.37% for CNL subtype. The 10-year OS and disease-specific survival (DSS) rate was 100% in POLE subtype, while only 33.51% and 39.69% in CNH subtype. In patients with CNL and CNL/MSI-H subtypes, ARID1A and ZFHX4 mutations were significantly associated with worse prognosis respectively.
Conclusion
This simplified one-step NGS panel can effectively subgroup EC patients into four prognostically different subtypes. New biomarkers are able to potentially refine the classification of patients with CNL/MSI-H subtypes into groups with distinct clinical outcomes.