Breast cancer remains the most common cancer and the leading cause of cancer-related death among women worldwide, with incidence rates continuing to rise. However, known risk factors, including genetics and lifestyle, do not fully explain the upward trend. Researchers are increasingly turning to metabolomics-the large-scale study of small molecules in biological systems-as a promising avenue for identifying new risk factors and improving prediction methods for breast cancer.
In a new study at Columbia University Mailman School of Public Health, researchers conducted a metabolome-wide association analysis using data from the New York site of the Breast Cancer Family Registry (BCFR). The findings are published in the journal Breast Cancer Research.
The case-control study included 40 breast cancer cases and 70 age-matched controls, with a median follow-up of 6.3 years. The analysis focused on women unaffected by breast cancer (BC) at enrollment, many of whom had a family history of breast and/or ovarian cancer-a population known to be at two to four times greater risk for BC compared to those without such a history.
Participants were followed to gather data on cancer incidence and vital statistics, with diagnoses confirmed through pathology reviews and medical records. The mean ages were 45.2 years for cases and 46.4 years for controls, with the average age of BC diagnosis at 51.6 years. Most cases and control groups were premenopausal at enrollment.
The study identified eight metabolic features significantly associated with breast cancer risk. Of these, four metabolites were negatively associated with risk, and four metabolites were positively associated with increased risk.
"If replicated, the identification underscores potential pathways and environmental exposures that may contribute to BC risk," said Hui-Chen Wu, DrPH, assistant professor of Environmental Health Sciences at Columbia Mailman School, and first author.
"Importantly, integrating these metabolic features into existing risk prediction models-such as those using age and the Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm risk score-improved predictive accuracy from 66 percent to 83 percent," noted Wu. "This suggests a substantial enhancement in the ability to predict breast cancer when metabolic data is incorporated.
The study provides the first human evidence linking breast cancer to the chemical compound 1,3-dibutyl-1-nitrosourea-used as an anticancer medication and known to cause mammary tumors in animal models- underscoring the value of metabolomics in identifying potential environmental risk factors for the disease. It also highlights metabolic pathways related to diet and lifestyle, echoing the need for further biomarker studies, particularly concerning exposures such as caffeine, whose relationship with breast cancer remains unclear.
"Our findings, if replicated in a larger cohort, point to a promising future for targeted, quantitative metabolomics analyses in refining breast cancer risk prediction and identifying previously unrecognized environmental exposures," said Mary Beth Terry, PhD, Columbia Mailman School professor of Epidemiology in Environmental Health Sciences, and senior author. Dr. Terry is also an associate director at Columbia Herbert Irving Comprehensive Cancer Center and the Executive Director of Silent Spring Institute.
Co-authors are Yunjia Lai, Yuyan Liao, Maya Deyssenroth, Gary W. Miller, and Regina M. Santella of Columbia University's Mailman School of Public Health and Herbert Irving Comprehensive Cancer Center.
The New York site of the Breast Cancer Family Registry is supported by grant 5U01CA164920-13 from the U.S. National Cancer Institute and ES009089. Additional support was provided by the Breast Cancer Research Foundation. The authors also acknowledge the Irving Institute Biomarkers Core Laboratory, supported by the National Center for Translational Sciences (2UL1 TR001873).
