Active Immunotherapy of Human Melanoma Exploiting the Immunopotentiating Effects of Cyclophosphamide

Abstract
Malignant tumors may escape rejection by the immune system because they induce a state of immunological tolerance mediated by tumor antigen-specific suppressor T cells. In animal systems, cyclophosphamide can reverse the tolerance and thereby facilitate immunologically mediated tumor destruction. We have applied these concepts to the immunotherapy of human malignant melanoma. Forty-three patients with metastatic disease were treated with a whole cell vaccine 3 days after intravenous administration of cyclophosphamide, 300 mg/m2. The vaccine consisted of cryopreserved, irradiated autologous melanoma cells, obtained from metastatic masses by dissociation with collagenase and DNAase, mixed with bacillus Calmette-Guerin (BCG) and injected intradermally. The cyclophosphamide (CY) + vaccine combination was repeated every 28 days. Delayed-typed hypersensitivity (DTH) was tested by injecting 1 .times. 106 melanoma cells intradermally and measuring the diameter of induration at 48 h. Most patients had minimal pretreatment DTH responses to melanoma cells (mean .+-. SE, mm = 2.4 .+-. 0.5). After two vaccine treatments, the responses increased significantly (mean increase .+-. SE = 12.1 .+-. 1.6 p < .001) and that level of response was maintained after 4, 6, and 8 treatments. The patients were also skin-tested with a mixture of the enzymes used to dissociate the tumors. No patients exhibited DTH to collagenase + DNAse prior to vaccine injection, but every patient developed DTH to the mixture following two treatments (mean, mm = 26.4 .+-. 3.9). Although extracting viable cells from tumor tissue without the use of enzymes proved difficult, we were able to test DTH to mechanically dissociated tumor cells in 23 patients. After two vaccine treatments, there was a signficant increase in DTH to enzyme-free autologous melanoma cells (mean DTH .+-. SE, mm: 5.4 .+-. 1.0, p < .01). Whereas 5 of 23 patients had positive DTH responses (5 mm induration or greater) before treatment, 11 of 23 were positive after two treatments. Further significant increases in DTH enzyme-free cells were observed after 6 and 8 treatments. Thus, it appears that patients receiving CY + vaccine developed DTH to tumor-associated antigens as well as to residual collagenase and DNAse on the cell surface. Thirty-three patients could be evaluated for antitumor effects of cyclophosphamide + vaccine. There were 3 complete remissions, 1 partial remission, and 2 minor responses. Two complete responders remain alive and free of disease after 57 and 12 months, respectively, and the third died after 39 months. The partial remission consisted of 75% regression of a pulmonary metastasis. Toxicity of the immunotherapy program was minimal and consisted of a local inflammatory response at the site of vaccine injection and mild nausea and vomiting in 25% of patients. These results corroborate the findings in experimental animals and suggest that application of ideas developed through basic investigations in immunoregulation will lead to more effective immunotherapy of human cancer.