Abstract Background A genomic-driven therapeutic strategy in metastatic breast cancer (MBC) was recently demonstrated as feasible in the clinical practice, but its actual impact on patient outcome remains elusive. SAFIR02 study is an ongoing national multicentric phase II randomized trial evaluating targeted therapies matching specific genomic alterations (GA) administered as maintenance after objective response and/or stable disease obtained with chemotherapy in HER2-negative MBC patients. This analysis reports on feasibility of the procedure and the rate of identified actionable targets. Methods Eligible MBC patients (PS=0/1, first- or second-line of chemotherapy, HER2-negative/hormone receptor (HR)-negative or endocrine resistant HR-positive; measurable per RECIST 1.1; accessible to tumor biopsy; no bone metastases-only disease, no major organ dysfunction) were subjected to tumor biopsy for genomic analysis (CGH arrays, Affymetrix Cytoscan; NGS, Ion Torrent PGM, AmpliSeq, panel of around 50 genes). Actionable GA were identified and corresponding targeted therapies were proposed by a multidisciplinary tumor board (MTB). Patients received cytotoxic-based treatment at physician's choice and those with stable or responding disease after 6 to 8 cycles (or at least 4 if stopped for toxicity reason) and targetable GA, were offered randomization between targeted therapy or chemotherapy maintenance until progression or intolerance (main study). Since January 2016, an amendment was made to propose to patients without targetable alteration a randomization between anti-PD-L1 (MEDI4736) or standard chemotherapy maintenance (substudy). Results Between March 2014 and May 2016, 457 patients have been enrolled at 21 centers. Genomic analyses could not be obtained in 107 cases (23%) due to either biopsy failure (n= 40; 9%) or low cellularity (n=67; 14%). Of the 307 patients reviewed by the MTB, 197 (64%) had an actionable GA, including PIK3CA-PIK3CB-PIK3R1 (n=51), FGF4 or FGFR1/2 (n= 42), BRCA1/2 (n=15), AKT1/2/3 (n=13), BRAF/KRAS/NRAS (n=13), HER2/3 (n=10), NF1-FRS2 (n=10), MTOR-RPTOR-TSC2 (n=8), PTEN (n=7), STK11 (n=7), IGF1R (n=7), EGFR (n=5). Therapeutic proposals by MTB included AZD5363 (n=71), AZD4547 (n=42), AZD2014 (n=23), selumetinib (n=23), olaparib (n=16), AZD8931 (n=15), vandetanib (n=5), bicalutamide (n=2). In an exploratory analysis involving 157 patients, the rate of targeted therapy proposal by MTB markedly differed between triple-negative patients (TNBC; 24 of 48, 50%) and HER2-negative/HR-positive patients (92 of 109, 84%; p=6.14. 10-6, Chi-2 test). At the time of the analysis, 85 patients have been randomized (main study, 68; substudy, 17). Causes of randomization failure (n=108) included disease progression (n=45) or death (n=25), non-eligibility criteria (n=27), patient/physician's decision (n=11). Conclusion A large number of patients had identified targetable GA. Of note, the rate of targeted therapeutic proposal was significantly lower in TNBC than in HER2-negative/HR-positive patients. Rapidly progressing disease may impede ultimate randomization. Citation Format: Gonçalves A, Bachelot T, Lusque A, Arnedos M, Campone M, Bièche I, Lacroix L, Pierron G, Dalenc F, Filleron T, Sablin M-P, Jimenez M, Ferrero J-M, Lefeuvre-Plesse C, Bonnefoi H, Attignon V, Soubeyran I, Jezequel P, Commo F, André F. High-throughput genome analysis and therapeutic decision for patients with HER2-negative metastatic breast cancer: First feasibility and molecular results of the randomized phase II study SAFIR02 BREAST (UCBG-0105/1304) [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr PD1-08.