Background: ARTEMIS is a double blind, randomized trial to determine if P-NAT impacts rates of excellent pathologic response [Residual Cancer Burden (RCB) 0-I]. P-NAT used a CLIA-certified chemosensitivity mRNA gene signature (GES) and subtyping of triple negative breast cancer (TNBC) by immunohistochemistry (IHC) to select targeted therapy (TT) trials for chemo insensitive tumors. Methods: ARTEMIS planned to randomize 360 TNBC patients (pts) 2:1 to ‘know’ vs. ‘not know’ P-NAT results using a group sequential design with one-sided O’Brien-Fleming boundaries and two interim tests for futility and superiority; overall α set at .05 with 80% power to detect improved RCB 0-1 rate from 50% to 64%. After biopsy, pts began a planned 4 cycles of Adriamycin-based chemo (AC). Volumetric change by ultrasound (US) upon completion of AC (or at progression) combined with GES results (if known) determined sensitivity using a protocol defined algorithm. Pts with sensitive disease received subsequent taxane-based (T) therapy. Pts with insensitive disease were offered phase II trials (TT; table) using IHC results, if known. To gauge impact of TT independently from GES, pts having < 50% volumetric reduction by US to AC were evaluated (US-resistant cohort). Results: The first interim analysis (n = 133 pts with RCB status) revealed a RCB 0-1 rate of 56% (know P-NAT) v 62% (not know P-NAT); p = 1.0; thus, randomization was discontinued for futility. In total 232 pts were enrolled; 168 evaluable for RCB. In the US-resistant cohort (n = 43), RCB 0-I rates were higher in pts treated with TT (n = 30) v AC-T (n = 13); (30% v 8%; odds ratio = 5.1 with 95% CI = (0.6-45.7); Fisher’s exact test 1-sided p-value = 0.11). Conclusions: GES failed to improve rates of RCB 0-I in TNBC; however, in pts with resistant disease identified by US after AC, RCB 0-I rates were higher in pts treated with targeted therapy compared to chemo alone. Clinical trial information: NCT02276443