Background: Given the variability in response to novel immunotherapeutic agents such as anti-PD1/PD-L1 therapies and the desire to extend their long-term benefit to more patients (pts), there is an increased need for the development of biomarkers that can help predict treatment outcomes and ensure that these treatments, which may have significant toxicities, are offered to pts more likely to benefit. We evaluated Bim (BCL-2-interacting mediator of cell death) in peripheral blood (PB) tumor-reactive CD11a^highPD-1⁺CD8⁺cytotoxic T lymphocytes (CTL) as a marker of the status of PD-1 engagement and T-cell reversibility to identify pts with MM who are more likely to benefit from anti-PD-1. Methods: MM patients treated with Pembrolizumab (pembro) 2 mg/kg every 3 weeks through an Expanded Access Program (EAP) had PB collected at baseline and at radiographic tumor evaluation through a separate biomarker sub-study. Frequencies of Bim⁺ T cells and Bim median fluorescence intensity (MFI) were measured by flow cytometry in gated CD11a^highPD1⁺ CD8⁺T cells. Non-parametric t-test was used to compare baseline Bim and percent change in Bim levels in pts who had a radiographic response (CR/PR/SD) compared to those who had progressive disease (PD) at 12 wks. Results: 29/40 ptsenrolled in the EAP had baseline samples and 14/29 pts had serial samplesavailable. Clinical benefit (CR/PR/SD) was observed in 9/27 evaluable pts at 12 wk; 18/27pts had PD. Two pts discontinued therapy for adrenal insufficiency (attribution to pembro unknown) and were not evaluable. Pts with clinical benefit after 4 cycles had higher frequency of Bim+/PD-1+ CD8 T cells at baseline compared to pts with radiographic PD (mean 60% vs. 49%, P = 0.04). In 9/9 responders who had serial PB samples available, the levels of Bim in PD-1+ CD8 T cells decreased after the first 3 months of treatment, and they increased/did not change in all 5/5 nonresponders with serial samples evaluable (P = 0.003). Conclusions: Measurements of Bim levels in tumor-reactive PD-1+ CD8 T cells may select patients likely to benefit from anti-PD-1 therapy, and provide a new non-invasive way to monitor response to anti-PD-1 blockade in MM. These results are being validated on a larger prospective cohort.