Background: Breast cancer (BC) patients (pts) expressing low levels of HER2 by (immunohistochemistry (IHC) 1-2+) are not eligible for trastuzumab (Tz). However, in a randomized phase 2b trial, triple negative BC (TNBC) pts demonstrated a significantly better DFS with nelipepimut-S (NPS) + Tz vs Tz alone. Here, we assess the ex vivo and in vivo immune responses (IR) in both arms. Methods: Disease-free pts (n = 275) with HER2 IHC 1-2+, non-amplified BC who were node positive and/or had TNBC were randomized 1:1 to granulocyte-macrophage-colony stimulating factor (GM-CSF) or NPS+GM-CSF. ±NPS was given every 3 weeks x 6 followed by 4 boosters every 6 months (mo). All pts received Tz concurrently for 1 year per label regimen and were followed for recurrence. IR were evaluated ex vivo by clonal expansion of NPS-specific cytotoxic T lymphocytes (CTL) by dextramer-staining/flow cytometry at time points over 3 years. In vivo IR were assessed by delayed type hypersensitivity (DTH) reactions periodically. Results: The trial enrolled 97 TNBC pts; 60 had 4 timepoints available for analysis (37 NPS + Tz pts; 23 Tz pts). The NPS+Tz group exhibited increases in CTL frequencies vs baseline: 208%, 303%, 379% at 18, 24 and 30 mo, respectively. NPS+Tz pts’ mean CTL frequencies increased from 0.029 ±0.001% at baseline to 0.112±0.026% at 30 mo (p = 0.01) compared to Tz pts who were 0.027 ±0.001% at baseline and 0.057 ±0.016% at 30 mo (p = 0.71). Only 4 NPS+Tz pts recurred as compared to 13 in the Tz arm. While limited by low numbers, recurrent NPS + Tz pt did not mount an IR by ex vivo assessment (range: 0.0 - 0.026%) or by DTH (all measurements: 0 mm), while non-recurrent pts mounted both clonal CTL expansion (range: 0.000- 0.33%) and enhanced DTH (range: 0.0- 80.5mm). Conclusions: NPS+Tz combination is more efficacious in generating time-dependent antigen (NPS)-specific CTL by both ex vivo and in vivo measures vs Tz. Based on these preliminary data, it appears that both ex vivo and in vivo IR to NPS were attenuated in pts with TNBC recurrence. Further analysis of read-outs from these assays to validate the relationship of IRs to clinical effect seen with NPS+Tz in TNBC pts is underway. Clinical trial information: NCT02297698