Melanoma is the solid tumor that metastasizes most frequently to the central nervous system (CNS). While major progress has been made in treating metastatic melanoma, advances in treating melanoma brain metastases (MBM) lag behind. This is due to near-universal exclusion of patients with untreated brain metastases from clinical trials, and only once a regimen is well established in extra-cerebral disease, is it studied in MBM patients. Moreover, pre-clinical studies of MBM have been hampered by a paucity of pre-clinical models and limited access to CNS specimens for analysis, as these patients are typically treated with radiation rather than surgery. We therefore developed infrastructure and resources which we propose to employ to better understand molecular, vascular and cellular underpinnings of brain metastasis. We started by studying activity of the PD-1 inhibitor pembrolizumab (pembro) in patients with untreated MBM on a phase II trial. Although the response rate was modest (26%), the two-year survival was 48%, and the incidence of radiation necrosis and/or neurologic symptoms related to edema was high. We therefore initiated a sequel trial of pembro plus bevacizumab in MBM; the response rate in the initial cohort was 55%, exceeding our expectations and warranting further evaluation. This response rate is similar to that seen with nivolumab and ipilimumab in MBM, but the toxicity profile is far superior with only minimal added toxicity from the bevacizumab. We propose to further study co-targeting the PD-1 and VEGF/VEGF-R pathways in MBM, and to determine the mechanism by which these pathways crosstalk to promote tumor growth within the brain. We will build on our success with anti-VEGF in a new trial involving pembro and a VEGF-R inhibitor (VEGFRi), lenvatinib. In parallel to our clinical research endeavors we developed novel immune-competent models of MBM that are modestly responsive to anti-PD-1, and developed in vitro models of the blood-brain-barrier to study effects of these drugs on vascular leak. These pre-clinical models will be used to test the hypothesis that the addition of VEGF pathway blockade to anti-PD-1 enhances T cell migration and function and directly decreases tight junction leakiness. We also hypothesize that anti-VEGF has different effects to VEGFRis. We will dissect the mechanism by which this occurs using our novel immune competent murine models and single cell transcriptomic and in situ proteomic strategies (Aim 1) and determine whether VEGF inhibiting drugs should be replaced with VEGF-Ris to treat MBM. We are expanding our ongoing clinical trial of pembro and bevacizumab by enrolling a second cohort of melanoma patients to verify our initial result. We have initiated a new trial of pembro and lenvatinib. We will analyze tumors and blood samples from patients on these trials in an attempt to validate mechanistic results from the murine experiments and identify predictors of response or resistance to this regimen (A...