Abstract Background: The heterogeneity of TNBC includes a subtype that is androgen dependent and whose growth is inhibited by antiandrogen therapy. We conducted a multicenter phase II trial which established the activity of bicalutamide for the treatment of patients (pts) with metastatic ER/PR-negative breast cancer TBCRC011 (Gucalp CCR 2013). We now report results of whole genome, next generation sequencing of the 26 evaluable pts to molecularly classify the study population and to identify potential biomarkers of response to bicalutamide. We hypothesized that those pts whose tumors express increased AR output are most likely to benefit from anti-androgen therapy. Methods: Archival formalin-fixed, paraffin-embedded (FFPE) samples were collected from either primary or metastatic site and RNA was isolated using standard techniques. Whole transcriptome sequencing was performed using the Illumina HiSeq 2000 platform. This data set was compared to that of TCGA, the PAM50 and the Lehmann TN subtypes for molecular classification. Bioinformatic analysis defined a model of AR transcriptional output based on androgen stimulated AR+ ER/PR- breast cancer cell lines (HCC202, SUM185, MDA453) that was applied to the TBCRC011 data set as a predictor of response. Gene set analysis was performed to test secondary hypotheses. Results: 21/26 pts provided adequate gene expression estimates for further analysis. Principal component analysis in comparison to whole transcriptomes from breast cancers in the TCGA show that TBCRC011 tumors associate with the basal like (BL) intrinsic subtype, and this was confirmed by PAM50 classification. Relative AR expression of TBCRC011 cases was similar to that of BLBC in the TCGA. There is a weak but statistically significant positive correlation between AR expression by IHC and transcriptome (r=0.395, p=0.034). An expression-based model of AR activity was derived from AR+ ER/PR- cell lines exposed to androgen. TBCRC011 samples segregate with the TCGA samples predicted to be AR responsive by this model. However, the AR output signature did not predict clinical benefit (CR+PR+SD>24 weeks) as defined in TBCRC011. Unexpectedly, by the Lehmann TNBC subtype criteria TBCRC011 tumors were not consistently molecularly classified as luminal AR (BL1-1, BL2-1, Immunomodulatory-5, Luminal AR-4, Mesenchymal-2, Mesenchymal Stem Like-2, Unclassified-6). None of the 4 pts with clinical benefit were classified as BL by the Lehmann model. Gene set analysis revealed that estrogen signal pathways were associated with the number weeks on therapy, and that samples with high scoring AR by IHC exhibited expression patterns of medullary breast cancer. Conclusions: The majority of samples demonstrated expression variation consistent with TCGA basal like breast cancer. Although cell line models predict enrichment of AR activity in both the TCGA basal like and TBCRC011 cohorts, this AR output signature did not predict clinical benefit of bicalutamide. Molecular classification found clinical benefit beyond the luminal AR subtype but exclusive of basal like 1 and 2. Whether IHC or molecular subtype is the optimal biomarker for pt selection for antiandrogen therapy remains uncertain. Citation Format: Tiffany A Traina, Ayca Gucalp, William Polkinghorn, Steven Isakoff, Sara Tolaney, Lisa Carey, James Ingle, Lisle Nabell, Andres Forero, Hope Rugo, Kimberly Blackwell, Minetta Liu, Matthew Soloway, Lisle Mose, Dilip Giri, Agnes Viale, Clifford Hudis, Charles Sawyers, Joel Parker. Whole transcriptome analysis of AR+ ER/PR- metastatic breast cancers treated with bicalutamide on TBCRC011 [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P3-04-01.