Our purpose was to determine the maximum tolerated dose and toxicity associated with soluble Chinese hamster ovary [s(CHO)] recombinant human interleukin (IL) 1 receptor (IL-1R; Immunex, Seattle, WA) administration in humans and to determine the effective biological dose and/or maximum tolerated dose of the s(CHO) IL-1R in combination with high-dose IL-2 as determined by reduction in IL-2 toxicity and modulation of its biological effects. Twenty-seven patients with metastatic cancer were treated with escalating doses of s(CHO) IL-1R at 1, 1, 5, 10, 20, 40, and 55 mg/m2 i.v. on days -6 (except cohort 2), 1, and 15 and IL-2 at doses of 300,000 IU/kg (cohort 1) and 600,000 IU/kg (cohorts 2-7) i.v. every 8 h on days 1-5 and 15-19. No toxicity directly attributable to s(CHO) IL-1R was observed. The median number of IL-2 doses was 23. Hypotension and neurotoxicity were the major dose-limiting toxicities for the IL-2/s(CHO) IL-1R combination. Of the 24 patients treated with full-dose IL-2, there were six responses, three complete and three partial (response rate, 25%). Three patients developed thyroid dysfunction, and all 3 responding melanoma patients exhibited vitiligo. The t1/2 of s(CHO) IL-1R alone was 24-30 h and was not significantly altered by coadministration with IL-2. Whole-blood functional assays indicated that sufficient s(CHO) IL-1R was present in the circulation at top dose levels to inhibit the in vitro effects of IL-1beta on IL-8 induction; however, no effect on IL-2-induced IL-8 induction, or on the IL-1beta- or IL-2-induced tumor necrosis factor production, was observed. Suppression of IL-2-mediated tumor necrosis factor alpha and IL-6 induction in vivo during the first 24 h after IL-2 administration was observed, and the neutrophil chemotactic defect normally seen with IL-2 was not observed. IL-1R antagonist induction far exceeded that seen previously with IL-2 alone. No inhibition of either serum C-reactive protein induction or enhanced urinary nitrate excretion and no consistent effect on IL-2-related changes in peripheral blood mononuclear cell phenotype or endothelial adhesion molecule expression were seen. The coadministration of s(CHO) IL-1R produced no apparent reduction in IL-2 clinical toxicity manifested by either the ability to administer more IL-2 than anticipated or a reduction in the toxicity associated with a given amount of IL-2. Therefore, no effective biological dose could be identified for the s(CHO) IL-1R.