Creating Models to Identify New Therapeutic Options for Aggressive African Breast Cancers

Background: Population-based incidence rates of breast cancer (BC) that does not express the estrogen receptor (ER), progesterone receptor (PR) or overexpress the human epidermal growth factor 2 HER2/ neu (triple negative breast cancer, TNBC) are higher among Africans compared with white women. However the underlying biologic and genetic differences among different ethnicities are poorly understood and there are currently very few ethnically diverse BC models available for identifying new therapeutic options. Aim: Establish an international collaboration to: i, characterize African breast tumors ii, create models for studying these tumors and iii, identify biomarkers for early detection and treatment personalization. Methods: We have collected tumors from 154 white Americans WA, 76 African Americans, AA, 190 Ethiopians, Eth, and 286 Ghanaian (Gh) BC patients. We then established a unique resource of patient derived xenografts (PDX) from these tumors. The PDXs were then fully characterized using whole exome and RNA sequencing for the primary tumor, matched normal DNA, and corresponding low passage PDXs. Using immunohistochemistry, we evaluated the ER, PR, HER2/ neu, androgen receptor (AR), and ALDH1 (cancer stem cell marker) expression among these tumors. Based on biomarker expression the PDXs were then tested against a panel of IND drugs, either alone or in combinations, in an ex vivo organoid culture system to discover potential new therapeutic options. Results: Mean age at BC diagnosis was 43; 49; 60; and 57 years for the Eth; Gh; AA; and WA patients, respectively. The proportion of TNBC was significantly higher for the AA and Gh patients (41% and 54%, respectively) compared with the WA and Eth patients (23% and 15%, respectively); P < 0.001. Significant differences were observed for distribution of AR positivity, which was 71%; 55%; 42% and 50% for the WA; AA; Gh; and Eth cases, respectively ( P = 0.008). The Gh breast tumors exhibited the highest number of loss of function and missense mutations that are likely to impact therapy with a high frequency of P53, APC, and FGFR mutations. These mutations were maintained in the corresponding PDXs that were developed, and were thus used as biomarkers for drug screening. These tumors exhibited a gene expression signature based on the ethnicity of the patients with 2385 genes differentially expressed between Gh and AA, 1573 between AA and CA and 1317 between GH and CA. Results from our ongoing drug screening and biomarker identification will be available soon. Conclusions: Establishing the molecular and genetic platform of aggressive breast cancers occurring in women with African ancestry will help in identifying biomarkers for early cancer detection and targeted treatment stratification for optimum patient outcome. The availability of tumor models based on tumors from diverse African populations is the important missing pieces that have to be incorporated into current drug discovery efforts.