Real-world molecular classification, treatment patterns, and outcomes by race in advanced endometrial cancer: a next-generation sequencing linked database study
IGCS 2026
Ebony Hoskins, Nehemiah Kebede, Zulikhat Segunmaru, Jasmine Sze, Stephen Park, Karime Kalil Machado, Sonia Iyer, John Nakayama
Introduction
Targeted and immunotherapies have improved endometrial cancer (EC) outcomes. In clinical practice, management of advanced EC (aEC) may reflect racial differences in histology, molecular biomarkers and outcomes. This retrospective real-world study evaluated associations among molecular biomarker status, race, clinical characteristics and outcomes in aEC, focusing on White and Black patients, the two largest molecularly characterized subgroups.
Methods
Data were obtained from the Tempus AI multi-modal real-world oncology database and included US adult patients with aEC (stage III/IV). Patients had received non-investigational first-line treatment between January 1, 2018–May 31, 2024 and had documented MMR status.
Results
Among 1121 patients, 54.7% were White and 11.8% were Black. Compared with White patients, Black patients more frequently had p53-abnormal tumors (57.6% vs 27.7%) and less frequently had dMMR (11.4% vs 26.3%) and NSMP tumors (14.4% vs 23.3%) by ProMisE classification. Black patients also had lower ER (62.9% vs 76.1%) and PR (46.5% vs 67.4%) expression (Table 1). Chemotherapy alone was the most common first-line treatment in both subgroups. Black patients had shorter median rwPFS: 8.5 (95% CI, 7.4–10.7) vs 13.3 months (95% CI, 11.7–14.7) (Table 2).
Conclusion/Implications
While the difference in histology prevalence among Black and White patients is well known, we describe variation in biomarker profiles between these patients. Black patients exhibited more aggressive disease profiles, fewer actionable biomarkers and poorer survival. These findings highlight the importance of integrating molecular profiling into clinical guidelines to inform personalized treatment strategies for high-risk patients. Further work to understand molecular profile disparities is warranted.
Related publications