Background: IL-2 plays a central role in immunity affecting the proliferation and survival of effectors of antitumor response. IL-2 at high bolus IV dosage gives a 16% response rate and durable responses in 4-5% of patients (pts) with metastatic melanoma. VEGF plays a critical role in angiogenesis and host innate and adaptive immunity. VEGF blocks maturation of dendritic cells and inhibits priming of T cell responses. High baseline serum VEGF was reported to be a predictor of non-response to HD IL-2. We developed a strategy to deplete VEGF prior to HD IL-2 to reverse the immunosuppressive impact of VEGF and enhance antitumor T cell response. Z (known as aflibercept outside the US) is a high-affinity soluble decoy VEGF receptor and potent angiogenesis inhibitor. Our phase II study of Z alone for advanced melanoma showed median OS of 16.3 months and PFS 3.7 months. Methods: NCI 8628 is a phase II trial of Z and HD IL-2 (Arm A) versus HD IL-2 alone (Arm B) randomized 2:1 respectively with accrual goal 105 pts. Arm A: consists of 3 courses (maximum) and each course consists of 2 cycles of HD IL-2 at 600,000 IU/kg IV every 8 hours for up to 14 doses (1^st cycle), followed by a rest period of 1-2 weeks’ (wk) rest and readmission for HD IL-2 (2^nd cycle). Z is given concurrently at 3 mg/kg IV every 2 wk, starting 2 wk prior to IL-2 in course 1. In the absence of disease progression, maintenance Z is given at 4 mg/kg every 2 wk after completion of IL-2. Arm B: patients receive HD IL-2 alone for a maximum of 3 courses (6 cycles). Eligible pts: Stage III inoperable or Stage IV (any M) melanoma . ECOG performance status of 0/1 and adequate organ function for HD IL-2. Up to two prior regimens for metastatic melanoma are allowed, and stable treated brain metastases. Response assessment follows RECIST v.1.1. Blood and tumor specimens are being collected prospectively for biomarker and mechanistic studies. Clinical trial information: NCT01258855.