High risk CNIs, race and early stage breast cancer

Abstract

We determined tumor copy number imbalances in 971 stage I/II breast tumors treated at MDACC from 1985–2000 that included 123 Hispanic and 125 African American cases. CNI were obtained using Affymetrix Oncoscan™. Models of relapse containing CNIs significantly outperformed models with clinical characteristics alone. For triple negative tumors, the combination of CNIs + clinical covariates yielded a significantly better prediction of relapse (C‐index; 0.79 ± 0.023) than the best clinical model (C‐Index = 0.64 ± 0.034). Race/ethnicity was not associated with relapse risk after accounting for clinical covariate but high risk CNIs were more frequent in AA and Hispanic cases in luminal and TNBC models. For example, a narrow region of 7q36.1 was more commonly lost in AA than NHW and Hispanic LUM cases (p=0.01). Furthermore, loss at Xq11–12 a risk factor for recurrence in TNBC was more common among AA and Hispanic women than NHWs (p=0.009). Measures of specific CNIs add significant additional predictive information for risk of recurrence and suggest additional opportunities to refine prognostication for early stage luminal and triple‐negative breast cancers. Further, specific CNIs differ in proportion by race/ethnicity, which perhaps reflects differential germline susceptibility or distinct etiologic risk factors that contribute to observed disparity among race/ethnic groups in breast cancer specific outcomes.