Abstract and Introduction
Abstract
Background: Blacks tend to have a stronger inflammatory immune response than Whites. We hypothesized that racial differences in host immunity also manifest in the tumor microenvironment, constituting part of a distinct aggressive tumor biology underlying higher mortality in Black women.
Methods: Pathological and gene expression profiling approaches were used for characterizing infiltrating immune cells in breast tumor microenvironment from 1315 patients from the Women's Circle of Health Study. Racial differences in tumor immune phenotypes were compared, with results validated in a publicly accessible dataset. Prognostic associations of immune phenotypes were assessed in 3 independent cohorts.
Results: We found marked and consistent differences in tumor immune responses between Black and White patients. Not only did tumors from Blacks display a stronger overall immune presence but also the composition and quality of immune infiltrates differed, regardless of tumor subtypes. Black patients had a stronger CD4+ and B-cell response, and further, a more exhausted CD8+ T-cell profile. A signature indicating a higher ratio of exhausted CD8+ T cells to total CD8+ T cells (ExCD8-r) was consistently associated with poorer survival, particularly among hormone receptor–positive patients. Among hormone receptor–negative patients, combinations of the absolute fraction of CD8+ T cells and ExCD8-r signature identified the CD8lowExCD8-rhigh subgroup, the most prevalent among Blacks, with the worst survival.
Conclusions: Our findings of a distinct exhausted CD8+ T-cell signature in Black breast cancer patients indicate an immunobiological basis for their more aggressive disease and a rationale for the use of immune checkpoint inhibitors targeting the exhaustion phenotype.
Introduction
In the United States, breast cancer mortality rates are the highest in African American or Black women.[1] Findings from the last decade suggest that tumor biology may play a role in these racial disparities.[2–5] Whereas previous research has focused on the tumor itself, less attention was paid to immune infiltrates in the tumor microenvironment (TME), increasingly recognized for their roles in carcinogenesis and as a target of cancer immunotherapy.[6]
Multiple lines of evidence indicate possible racial differences in immune responses to tumors. Systemic immune responses, measured by circulating cytokine levels[7] and transcriptomic profiles of cultured immune cells, differ across ancestral populations.[8–10] Further genetic analysis attributed these differences, in part, to immune adaptation to ancestral environments of early human populations.[7,9,10] Two earlier studies also showed differential gene expression in immune-related pathways in prostate and breast cancers from Black and European American orWhite patients.[11,12] Because the sample sizes in these studies were limited and methods predated the advent of advanced profiling of immune infiltrates, it remains to be established whether immune responses in breast cancer differ by race.
Two recent studies based on The Cancer Genome Atlas (TCGA) data showed only modest racial differences in tumor immune gene expression between Black and White patients after accounting for breast cancer subtypes.[13,14] Despite being an invaluable public resource used for many important studies of immune infiltration in TME, TCGA tumor tissues were collected and processed primarily for sequencing of tumor cells but not necessarily immune cells, which might preside outside of tumor foci.[15] Moreover, the design of TCGA, with no matching of patients by race, was not optimal for racial comparisons. To overcome these limitations and to systematically investigate breast TME between Black and White patients, we profiled immune infiltrates using pathological and molecular methods in tumors from the Women's Circle of Health Study (WCHS), which was established specifically to investigate breast cancer racial disparities, with results validated in publicly accessible data.
J Natl Cancer Inst. 2021;113(8):1036-1043. © 2021 Oxford University Press
