Impact of race and age on intrinsic subtype distribution and treatment decisions in metastatic breast cancer: Preliminary analysis of HARMONY (LCCC1829).

Abstract

1055 Background: Gene expression profiling (GEP) is used to classify breast cancer (BC) into four intrinsic subtypes: Luminal A (LumA), Luminal B (LumB), HER2-Enriched, and Basal-like. Non-LumA subtypes are associated with poorer outcomes. Studies in early BC show that hormone receptor-positive (HR+)/HER2-negative (HER2-) tumors have a higher frequency of non-LumA subtypes in Black and younger (<50) women. Similarly, triple negative BC, which carries the poorest prognosis and is mostly Basal-like, are overrepresented in Black and younger women. These variations in intrinsic subtype contribute to outcomes disparities. The extent to which these racial and age differences in subtypes and mortality occur in metastatic breast cancer (MBC) is unknown. Methods: HARMONY (NCT03769415) is a prospective clinical trial in patients with newly diagnosed MBC. Bulk tumor GEP (PAM50 intrinsic subtyping) is performed on primary and metastatic tumors. A primary objective is to determine the extent and treatment implications of molecular discordance (defined a priori as treatment differences by intrinsic subtype compared to clinical subtype); other endpoints to be reported later include GEP comparisons between primary and metastatic samples. Here we examine differences in a) intrinsic subtype and b) treatment patterns in MBC by race and age. Results: At the time of analysis, the study had accrued 220 participants. Mean age was 58, 24% were Black women. Sixty-five percent had HR+/HER2- MBC, 14% HER2-positive, and 21% triple negative. Unlike in early BC, LumB exceeded LumA (LumA 27%, LumB 35%, HER2-Enriched 12%, Basal-like 25%). Similar to early BC, subtype varied by race and age with more non-LumA subtypes in young and Black women. Discordance between intrinsic and clinical subtype was seen in 24% of Black, 20% of White women, in 29% of younger and 18% of older women. Treatment patterns also differed by race and age, where a higher percentage of younger Black women received first-line chemotherapy compared to younger White women overall (45% vs. 10%, p=0.0025), and in HR+/HER2- disease (33% vs 8.7%, p=>0.05). No difference was seen in older women (24% vs 17%, p=0.44). Median progression free survival was worst in younger Black women compared to all other participants (5.4 vs 19.5 months, p=0.027), which may be driven by clinical subtype. Conclusions: Intrinsic subtype of tumors in a metastatic population has a different distribution by race and age than early BC, with more poor-prognosis non-LumA subtypes, and approximately 1 in 5 have discordance between clinical and molecular subtype, which may be therapeutically relevant. Clinical trial information: NCT03769415 . [Table: see text]