Discovering plasma proteins associated with breast cancer incidence in post-menopausal women in ARIC

Background Plasma proteomic data can be used to discover breast cancer risk biomarkers beyond established risk factors. Discovery using a large-scale platform in a diverse population is needed. Methods We investigated 4,712 plasma proteins and breast cancer risk among post-menopausal women in a prospective cohort analysis in the Atherosclerosis Risk in Communities study. Proteins were measured by SomaScan® 5K Assay. Incident cases were ascertained primarily from state cancer registries. We estimated hazard ratios (HRs) and 95% confidence intervals (CIs) using Cox regression adjusting for risk factors, and applied the Benjamini-Hochberg method to control false discovery. We determined whether the statistically significant proteins were confirmed in a case-cohort study within the European Prospective Investigation into Cancer and Nutrition (EPIC). Results After median follow-up of 23.3 years, 340 of 4,403 women had an incident breast cancer. Two proteins were statistically significantly associated after p-value adjustment per doubling, the HR of breast cancer was 1.45 (95% CI: 1.23-1.70, p = 7.47*10−6, padjusted=0.0370) for protein LEG1 homolog and 2.52 (95% CI: 1.66-3.83, p = 1.50*10−5, padjusted=0.0371) for ADP-dependent glucokinase. Results were consistent in a lagged analysis and among both Black (28.1%) and White women. In EPIC, both associations were confirmed (protein LEG1 homolog, HR = 1.24, 95% CI 1.14-1.35, p = 9.79*10−7, ADP-dependent glucokinase, HR = 1.13, 95% CI 1.03-1.23, p = .01). Conclusion We identified two plasma proteins associated with increased breast cancer risk over the longer-term in post-menopausal women; both were confirmed in an independent cohort. If further validated, these plasma protein biomarkers could be considered for utility in current risk stratification tools.