02/17/2026

Clinicogenomic and Immune Profiles of Male Breast Cancer by Race: Insights from a Large Real-World Cohort

SABCS 2025 PRESENTATION
Authors Y. Abdou; E. Jaeger; S. Fragkogianni; E. Williams; M. Ciampricotti; K. Reeder-Hayes; P. Spanheimer

Background: Male breast cancer (MBC) is a rare disease with limited data on potential racial differences in clinical and molecular characteristics, despite the fact that such differences are well-recognized in females. Understanding differences in the immune microenvironment and clinicogenomic features of MBC by race could inform future treatment strategies and improve health equity. This study leveraged the largest real-world (rw) cohort of MBC patients (pts) to date to characterize these features in Black/African American and White men.

Methods: Male pts with primary breast cancer who underwent xF or xT testing in the Tempus multi-modal database (Tempus AI, Inc., Chicago, IL) were identified using Tempus Lens. Statistical comparisons used Kruskal-Wallis and Chi-squared/Fisher’ exact tests where appropriate. Pts with xR testing (RNA) were classified into PAM50 subtypes and immune infiltration was quantified using quantiseq. Hormone receptor status was extracted from clinical documentation within 60 days of sample collection. Tumor mutational burden (TMB; mutations/megabase) and microsatellite instability were assessed by DNA sequencing. Real World overall survival (rwOS) was defined from the time of sample collection to death, last follow-up, or a study cutoff of 7 years, and was estimated using Kaplan-Meier methodology; comparisons used the log-rank test.

Results: The study cohort included 291 MBC pts: 128 White (44%) and 36 Black (12%). Among pts with available ethnicity, 93% were not Hispanic or Latino. Clinical subtypes were HR+, HER2- (39%), HER2+ (7.6%), triple-negative (7%) and 50% unknown or not otherwise specified. Median age at diagnosis was 63 years, and 87% of staged pts presented with stage IV disease. PAM50 subtyping identified Luminal A (62% of White and 68% of Black) and Luminal B (23% of White and 14% of Black) as the most common subtypes, with no significant racial differences in molecular subtypes (p=0.6). TMB was low (<10 mt/mB) and all tumors were microsatellite stable. AR (41%), PIK3CA (27%), TP53 (26%), ESR1(12%), TERT (11%) and GATA3 (11%) were amongst the most frequent somatic mutations found in the cohort overall, with no racial differences identified. BRCA2 was the most frequent germline mutation identified overall. PD-L1 negativity (<1% CPS) was more frequent in Black vs. White pts (71% vs. 50%, p=0.5). Notably, M2 macrophage infiltration was significantly lower in Black vs. White pts (median 4.58 vs 5.62, p=0.036); CD8+ T cell infiltration was also not statistically significantly lower in Black vs White pts (median 0.03 vs. 0.15, p=0.3). Median rwOS was numerically lower in Black vs White pts, though the difference was not statistically significant (19.1 vs 23.1 months; p=0.5).

Conclusions: This real world analysis of MBC pts revealed largely similar clinicogenomic profiles between Black and White pts, with the notable exceptions of significantly lower M2 macrophage infiltration and numerically lower CD8+ T cell infiltration and higher rates of PD-L1 negativity in Black pts. These immune microenvironment differences may have implications for immunotherapy response, underscoring the need for further research into the biological underpinnings and potential clinical impact of immune differences in racially diverse MBC populations.

VIEW THE PUBLICATION

VIEW THE POSTER