Background: Interleukin-12 (IL-12), a master regulator of the immune system, results in anti-tumor responses in preclinical models, but safe use requires tightly controlled production. It was conditionally produced in Ph1 “main” study (NCT02026271) in subjects with recurrent glioblastoma (rGBM) using a replication-incompetent adenovirus modified to express IL-12 under transcriptional control of the proprietary RheoSwitch Therapeutic System (Ad-RTS-hIL-12, Ad) regulated by dose of veledimex (V). Monotherapy resulted in sustained intra-tumor influx of activated cytotoxic T cells, consistent with immune-mediated anti-tumor effect, improving overall survival (OS). This correlated with increased circulating CD8⁺/FoxP3⁺ T-cell ratio (“cytoindex”), an emerging biomarker of OS. While widely used with neurosurgery, dexamethasone (dex) blunts response to immunotherapies, nevertheless median mOS of subjects who received 20mg V of 12.7 mo (n=15) at 13.1 mo follow-up. However, subanalysis (n=6) showed low-dose dex (total ≤20 mg) during V dosing improved mOS (17.8 mo). We report a 36 subject substudy in rGBM with limited dex, total rGBM treated (n=70+). Methods: Ongoing Phase 1 substudy (NCT03679754) assesses safety and tolerability of local, inducible IL-12 by single intratumoral injection of Ad (2 x 10¹¹ viral particles) + V (20 mg PO QD x15 doses Days 0-14) in subjects not receiving dex 4 wks prior to Ad. Results: As of 03Jan19, the majority of new subjects received low-dose dex (total ≤20mg Days 0-14). The initial impact of dex on mOS will be reported. As in the main study, Ad+V 20 mg respectively increased (median) serum IL-12 and downstream IFN-g from Days 0-3: 0.8 to 8.8 pg/mL and 0 to 8.6 pg/mL. Between Days 0-14, there was net increase in cytoindex (from 20 to 46). The safety profile was similar to the main study with the main adverse reaction (AR) being mild to moderate cytokine release syndrome (CRS) characterized by flu-like symptoms. No grade 4 CRS was noted; all ARs were manageable and reversable upon holding V. Conclusions: Local, controlled IL-12 production using the Ad + V platform in subjects with rGBM safely activates the immune system and when dex is limited, appears to further improve mOS, which warrants continued investigation. Clinical trial information: NCT03679754