The therapeutic application of high-dose interleukin (IL) 2 in human malignancy is limited by severe multiorgan toxicities that are mediated, in part, by tumor necrosis factor (TNF) and IL-1. CT1501R (lisofylline; LSF) is one of several methyl xanthine congeners that inhibit the effects of TNF by the interruption of specific signal transduction pathways. This randomized, placebo-controlled trial was designed to assess the activity of LSF in reducing the toxicities of high-dose IL-2 therapy. Fifty-three patients with metastatic renal cancer or malignant melanoma were treated with i.v. bolus IL-2, 600, 000 IU/kg every 8 h for 5 days (14 doses), followed by 9 days of rest and another 5-day course of IL-2. Patients were randomly assigned to LSF, 1.5 mg/kg i.v. bolus, or placebo every 6 h during IL-2 therapy. All patients were to be treated to individual maximum tolerance of IL-2 at the intensive care unit level of support. The end points for statistical analysis were the number of IL-2 doses administered during the first cycle of treatment (maximum, 28) and the toxicities experienced by each group after the first 8 planned IL-2 doses. There was no difference between the LSF and placebo groups in the mean number of IL-2 doses tolerated in the entire first cycle of therapy (19.6 +/- 5.4 versus 19.5 +/- 5.8, P = 0.86) or in the first or second 5-day course of IL-2. The only significant difference in toxicities occurring through the eighth dose of IL-2 was in the maximum elevation of serum creatinine (mean, 1.7 +/- 0.8 for placebo versus 1.5 +/- 0.6 mg/dl for LSF, P = 0.013). A Monte Carlo analysis of major toxicities over the first 14-dose course of therapy showed a statistically significant difference favoring the LSF-treated group (P = 0.025). LSF was well tolerated, associated only with mildly increased nausea (P = 0.006 after eight IL-2 doses, but not significant for the entire first cycle). The antitumor activity was comparable in both groups (objective responses, 2 of 28 with LSF versus 4 of 24 with placebo). The mean peak plasma concentrations of LSF on days 1, 5, and 19 were 6.24, 3.83, and 5.04 micromol/liter, respectively. In conclusion, with this dose and schedule, LSF did not alter the toxicities of high-dose i.v. IL-2 sufficiently to impact the overall dose intensity of IL-2. Successful IL-2 toxicity modulation may require the use of higher doses of LSF, the development of agents with more potent anti-TNF activity, and/or combined modulating agents that function via distinct mechanisms to interrupt cytokine-mediated signaling.