The aim of this pilot study was to assess the clinical and immunological effects of human allogeneic liposomal melanoma vaccine alone or combined with Interleukin-2 (IL-2) in patients with metastatic melanoma. Four concurrent treatment arms were included: vaccine alone (A); vaccine combined with systemic IL-2 (B); vaccine combined with low-dose liposomal regional IL-2 (C); and low-dose regional IL-2 as in group C but without vaccine (D). Vaccine was prepared from semisynthetic phospholipids (dimyristol phosphatidylcholine and dimyristol phosphatidylglycerol) and membranes of six human melanoma cell lines. The latter were chosen as expressing MHC class I and II antigens and a "mosaic" of melanoma-associated antigens (MAAs) as detected by MoAbs R24, p97, CF21 and TA99. Nine of the 24 patients had objective clinical responses: of the ten patients treated with liposomal vaccine and low dose regional IL-2 (arm C), three had complete responses (CR) and three had partial responses (PR); of the five patients treated with liposomal, low-dose regional liposomal IL-2 only (arm D), three had PRs. No clinical responses were seen in patients treated by vaccine alone (A) nor in patients treated by vaccine and systemic IL-2 (B). Patients' in vivo and in vitro cellular immune responses were closely monitored. Conversion to positive cutaneous delayed type hypersensitivity (DTH) to membrane vaccine (without liposomes) was induced only in the six clinical responders of arm C. Positive DTH correlated with augmented in vitro proliferative lymphocyte responses stimulated by melanoma cell lines and membrane preparation and with the augmented cytolytic activity against melanoma cell lines.