Background: Breast cancer (BCa) is the most common malignancy in women where incidence, distribution, clinical outcomes, and mortality rates vary widely among women of different racial and/or ethnic backgrounds.
 
Methods: To explore the metabolome of normal and malignant breast tissues and try to discriminate BCa disease in a cohort of African American women, biospecimens (n=48) were analyzed by untargeted metabolomics using ¹H nuclear magnetic resonance (NMR) spectroscopy. Breast specimens were characterized as control or malignant. Control samples were obtained from reduction mammoplasty surgeries, including those with benign fibrocystic changes or non-malignant fibroadenoma, whereas malignant BCa samples encompassed grades I-III and four clinical subtypes. Multivariate and statistical analyses were performed to identify important spectral bins for differentiation of phenotypes and corresponding metabolite annotation by library-matching was done in Chenomx. Pathway analysis was performed using MetaboAnalyst. Metabolite association with demographic, clinical, and pathological risk factors were assessed using rank-based tests across all 212 spectral bins, with p-value adjusted for multiple testing using the Benjamini-Hochberg False Discovery Rate (BH-FDR) method. 

Results: Preliminary metabolic differences were observed between reduction mammoplasty biospecimens, based on the presence or absence of fibrocystic changes or fibroadenoma. Control biospecimens were then classified as mammoplasty-normal, mammoplasty-with fibrocystic changes and mammoplasty-with fibroadenoma. Thus, following initial comparisons, subsequent subgroup analyses comparing BCa biospecimens were performed using only mammoplasty-normal tissues as the control reference. Distinct endogenous pathway profiles were also observed across all preliminary comparisons derived by multivariate analysis. Multivariate analysis also distinguished BCa from control tissues and captured exploratory differences across histological BCa groups by stage and grade. However, following Benjamini-Hochberg False Discovery Rate (BH-FDR) correction, only 12 bin regions remained significant. No bins survived FDR analysis in correlation analyses with risk factors, while several subgroup models showed weak predictive validity. 

Conclusions: This exploratory pilot study identified several metabolites that differed among breast tissue types from African American women using NMR. These metabolites represent promising candidates for further investigation through larger cohorts and independent validation studies to determine their contributions to BCa etiology, disease progression and potential metabolic contributions to health disparities. 

Metabolomics, Biomarkers, Breast cancer (BCa), Reduction mammoplasty, Fibrocystic breast disease, Fibroadenoma, African American (AA), Health disparity