Introduction: Population-based incidence rates of triple negative breast cancer (TNBC) are higher for African American (AA) compared to White American (WA) women, but it is unclear whether TNBC risk is genetically associated with African ancestry because AA women represent an ancestrally admixed population. Higher frequencies of TNBC have also been observed in sub-Saharan African breast cancer (BC) patients, but comparative analyses of biomarker expression among datasets that include AA, WA, and African women are sparse. We report findings from an international registry that features specimens from a diverse patient population in Detroit, Michigan as well as a hospital in Kumasi, Ghana.

Methods: The study dataset included formalin-fixed, paraffin-embedded invasive BC tumors diagnosed between 1998 and 2014 at the Komfo Anokye Teaching Hospital in Ghana and the prospectively-maintained/annotated Henry Ford Health System cohort in Michigan. All Ghanaian tumors underwent pathology confirmation and immunohistochemistry for estrogen receptor (ER), progesterone receptor (PR) and HER2/neu expression at the University of Michigan. Women were classified into five BC phenotypes and dichotomized into two age groups, <50 and ≥50 years. Polychotomous multivariate GLM models were developed to estimate the risk for each BC phenotype. Statistical analyses were performed in SAS v. 9.0 (Carey, NC). This research was approved by the Institutional Review Boards of the participating institutions.

Results: A total of 234 Ghanaian cases with mean age 49 years (range 24-92); 271 AA with mean age 60 (range 27-87); and 321 WA with mean age 62 (range 31-91) (P=0.001) contributed to this study. Prevalence of histologic grade 3 was lowest in WA (n=107, 33.7%) which was statistically significant from the observed prevalence in AA (n=135, 50.4%) and Ghanaians (n=84, 53.8%) (P<0.0001). ER-negative and TNBC were more common among Ghanaian and AA compared to WA cases (frequency ER-negativity 67.5%, 37.1%, and 19.8%, respectively, p<0.0001; frequency TNBC 53.2%, 29.8%, and 15.5%, respectively, p<0.0001). In the age group <50 years, 82 women (42.5%) were diagnosed with ER+/PR+/HER2-, 65 (33.7%) with TNBC, 27 (14.0%) with ER+/PR+/HER2+, 14 (7.2%) with ER-/PR-/HER2+ and 5(2.6%) with ER-/PR+/HER2- phenotypes. In this young age group, prevalence of TNBC remained highest among Ghanaian women (50.8%), followed by AA (34.3%) and WA (15.9%); (P=.0006). In contrast, highest prevalence of ER+/PR+/HER2+ and ER+/PR+/HER2- phenotypes was observed in WA, followed by AA and Ghanaians. On multivariate analysis histologic grade 3 and racial heritage remained statistically significantly associated with the TNBC phenotype (OR for AA vs. WA with TNBC 1.87, 95% CI 1.15-3.04; OR for Ghanaian vs. WA with TNBC 10.63, 95% CI 5.32-21.25; OR for Grade 3 vs Grade 1 histology with TNBC 33.3, 95% CI 13.45-82.4).

Conclusions: This study confirms an association between the TNBC phenotype and African ancestry; furthermore, extent of African ancestry appears to be associated with an increased likelihood of having a TNBC tumor, since frequency of TNBC among AA patients was intermediate between WA and Ghanaian patients.

Citation Format: Newman LA, Jiagge E, Bensenhaver JM, Chitale D, Kleer C, Merajver S, Kyei I, Aitpillah F, Oppong J, Amankwaa-Frempong E, Adjei E, Wicha M, Awuah B, Stark A. Comparative analysis of breast cancer phenotypes in African American, White American, and African patients- Correlation between African ancestry and triple negative breast cancer. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P6-12-14.