Background: AMP-activated protein kinase (AMPK) can phosphorylate BRAF and block its activation of MEK. AMPK is activated by biguanides such as metformin and phenformin (phen). Preclinical data indicated that phen can enhance the effects of BRAF inhibitors (BRAFi) in sensitive and BRAFi-resistant melanoma (mel) cell lines whereas metformin does not. Phen also can decrease myeloid-derived suppressor cells (MDSCs) enhancing immune reactivity against mel. For these reasons, we designed a phase I trial combining phen with dabrafenib/trametinib (dab/tram) in patients (pts) with metastatic BRAF V600-mutated mel. Methods: Bulk phen was provided by the Division of Cancer Treatment and Diagnosis of the National Cancer Institute under a materials transfer agreement. Phen was encapsulated at the Pharmaceutical Product Facility at MSKCC under cGMP procedures and used under an IND held by MSKCC. Eligible pts had metastatic BRAF V600-mutated mel. For the dose-escalation phase, prior BRAFi therapy was allowed. A 3+3 dose escalation design was used starting at a phen dose of 50 mg bid and standard dosing of dab/tram (150 mg bid/2 mg qd). Subsequent planned dose levels were 100 mg bid, 200 mg bid, and 300 mg bid with de-escalation to intermediate doses for toxicity. Plasma PK samples were collected and assayed by MRI Global Inc using a validated, GLP-compliant, HPLC-MS/MS assay. PBMC were collected to measure circulating MDSCs. At the recommended phase 2 dose (RP2D), we planned to treat 10 BRAFi-naïve pts. Results: We accrued 18 patients from March 2017 through December 2019. Dose-escalation to 200 mg bid was not tolerable nor was 150 mg bid. Therefore, we determined 100 mg bid to be the RP2D and accrued 2 patients on the dose expansion cohort before further accrual became impossible due to the COVID-19 pandemic. The most common toxicities were gastrointestinal – nausea, vomiting, anorexia, and transaminase and alkaline phosphatase elevation – accounting for 9 DLTs. Two patients were hospitalized due to lactic acidosis. There were also constitutional/inflammatory and skin/mucosal adverse events thought to be related to dab/tram. Weight loss was seen in most patients (median 3.7% at week 8) thought to be related to phen. There were 8 PRs and 2 CRs. Responses were observed in 2/7 pts who had received prior therapy with BRAFi. One CR lasted > 1 year. There were 8/11 (73%) responders among the BRAFi-naïve pts. At the RP2D of 100 mg bid, day 8 median phen plasma level was 73.4 ng/ml (range 32.8-253.2). At the the 50 mg bid dose level, it was 40.8 ng/ml (range 29.7-110.2). MDSC could be analyzed from the 150 mg bid and 200 mg bid dose levels (N = 7 pts). In all but 1 pt, MDSCs decreased while on phen. Conclusions: We observed 10/18 (55.5%) total responses with 2/7 among previously treated pts. Circulating MDSCs decreased on treatment. Phen at 100 mg bid can be combined safely with dab/tram and should be explored in future randomized phase II trials. Clinical trial information: NCT03026517.