Background: Histone deacetylase inhibitors (HDACi) enhance tumor immunogenicity through several mechanisms and may augment response to immune checkpoint inhibitors (ICI). We report updated results from a phase I/Ib trial testing the combination of oral HDACi vorinostat (V) with PD-1 inhibitor pembrolizumab (P) in mNSCLC. Methods: In phase I, pts with ICI-naïve or ICI-pretreated mNSCLC were treated with P (200mg IV q3 wk) + V (200 or 400 mg PO daily). In phase Ib expansion, pts were required to have progressed on prior ICI treatment. Primary endpoints were safety/tolerability; secondary endpoints included RR, PFS, DOR, and OS. Tissue and blood specimens from pre- and on-treatment were collected for correlative analyses to determine tumor gene expression changes, T cell density and levels of myeloid-derived suppressor cells. Results: Between 3/2016 - 12/2018, Phase I: 13 pts were treated (4 at 200mg, and 9 at 400mg V dose); and Phase Ib: 20 pts were treated. Median age: 68 (range 38-82); Females: 11 (33%); ECOG 1: 31 (94%); and never/former/current smokers: 3/22/8 (9%/67%/24%). PD-L1 expression was < 1% in 8/33 (18%), ≥1-49% in 7/33 (21%), ≥ 50% in 9/33 (27%) and unknown in 11/30 (33%). No DLTs or treatment related deaths were observed. The RP2D was P 200mg and V 400mg. Most common any grade AEs was fatigue (11%) and nausea/vomiting (8%). 2 (6%) patients had treatment discontinued due to toxicity. 30 pts are evaluable for response, 6 ICI-naïve and 24 ICI-pretreated. 4 (13%) had PR (2 confirmed), 16 (53%) had SD, and 10 (33%) had PD for a disease control rate of 67%. In the ICI-pretreated Ib cohort, 3 pts (1 confirmed; 2 unconfirmed) had a PR and 10 had SD (8 confirmed). For ICI-pretreated pts, mPFS was 3.2 and mOS was 7.3 months, and 1-year PFS was 17% (4 pts). For ICI-naïve, mPFS was 7.6 months and mOS was 16 months. CD8 T cell presence in tumor stromal regions was associated with benefit to P + V treatment. Conclusions: P + V were well tolerated. The combination demonstrates preliminary anti-tumor activity despite progression on prior ICI treatment and stromal CD8 T cells may be associated with benefit from P + V treatment. A randomized phase II portion of this study, examining P combined with V vs. placebo in immunotherapy naïve pts, is ongoing. Clinical trial information: NCT02638090