Approximately 20% of patients with metastatic melanoma respond to IL-2-based regimens. These responses are generally of short duration, and manipulation of the dose and schedule of IL-2 has failed to improve clinical outcome. In preclinical models, tumor-specific monoclonal antibodies (MoAb) can enhance the activity of IL-2 administered alone or in combination with lymphokine-activated killer (LAK) cells. Recently, the murine MoAb R24, which recognizes the GD3 ganglioside on melanoma tumor cells, was shown to be tolerable in phase I trials and produced some clinical responses. Therefore, we initiated a phase I trial of R24 administered by 24 hour continuous infusion on day 0 followed by IL-2 on days 1-5 (4.5 mu/m sq/d) and 11-18 (3 mu/m sq/d) and LAK cells on days 11, 12, and 14. Eight (7 evaluable) patients were treated (6 received 3 mg/m sq and 2 received 10 ig/m sq of R24). Administration of LAK cells was associated with unexpectedly severe pulmonary toxicity in 4 of these patients, suggesting that administration of the R24 prior to priming and pheresis altered the characteristics and toxicity of the LAK cells. Subsequently, we amended the protocol to delete the LAK cells and delayed the administration of IL-2 to day 14 (4.5 mu/m sq/d by IVCI for 108 hours). Seventeen patients have been treated (3 at 10 mg/m sq, 6 at 30 mg/m sq, at 60 mg/m sq, 3 at 100 mg/m sq and 2 at 200 mg/m sq of R24). Dose-limiting toxicity was reached at 200 mg/m sq of R24. No objective responses were seen at any dose level.
