Abstract Breast cancers arising in younger women are associated with worse survival outcome when compared with stage-matched breast cancers diagnosed in older women; to date, however, the poorer outcome of these younger onset breast cancer cases has not been fully explained by prognostic biomarkers or biological mechanisms. Collecting age cohorts of node-negative, estrogen receptor (ER)-positive sporadic breast cancers, we recently showed that younger onset cases can be distinguished from older onset ER-positive cases by expression profiling but not by genomic profiling, revealing a predictive age signature and enrichment of a proliferation gene expression signature (Perou et al., 2000) among younger ER-positive breast cancer cases (Yau et al., 2007). To further explore the impact of age on the outcome and biology of early stage breast cancers stratified by ER status, we pooled outcome and expression microarray data on adjuvant-untreated, node-negative cases obtained from three sources (van de Vijver et al., 2002; Wang et al., 2005; Desmedt et al., 2007). Dichotomized age-at-onset cohorts were defined as either younger (Y) ≤ 39 y or older (O) ≥ 40 y cases. Gene expression data from 447 ER-positive (61 Y, 386 O) and 178 ER-negative (40 Y, 138 O) cases, generated from either of two microarray platforms (Affymetrix, Agilent), were mapped by gene symbol and combined using distance weighted discrimination (DWD), to produce pooled datasets for each ER subtype containing 6209 unique genes. Distant metastasis-free survival (DMFS) was plotted by Kaplan-Meier analysis out to 15 years for 563 of the pooled 624 cases (400 ER-positive: 52 Y and 348 O; 163 ER-negative: 36Y and 137 O), and log rank significance testing was performed. DMFS of all 163 ER-negative cases was significantly worse than that of all 400 ER-positive cases (p = 0.05); and DMFS of all 88 Y cases was significantly worse than that of all 475 O cases (p = 0.02). DMFS of ER-positive Y cases was significantly worse than ER-positive O cases (p = 0.02), but DMFS of ER-negative Y cases was not significantly different than that of ER-negative O cases (p = 0.66). Consistent with our earlier results, expression of a 38-gene proliferation signature was significantly enriched in ER-positive Y cases relative to ER-positive O cases. In contrast, this proliferation signature was not differentially enriched in ER-negative Y cases relative to ER-negative O cases. While high proliferation gene signature expression correlated with worse DMFS in ER-positive Y and O cases, it did not correlate with ER-negative DMFS in either Y or O cohorts. Thus, metastatic outcome analyses indicate that for node-negative breast cancers without systemic treatment, age is a prognostic factor only for ER-positive and not for ER-negative disease. Moreover, gene expression analyses indicate that the better prognosis of ER-positive breast cancers arising after age 39 may be attributed to their lower proliferative potential while the worse DMFS of younger onset ER-positive breast cancers may be attributed to their higher proliferative potential. In contrast, the metastatic outcome of ER-negative breast cancer appears independent of both age-at-onset and tumor proliferative potential. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 4051.