A phase I trial involving continuous infusion of both beta- and gamma-interferon (IFN-beta and IFN-gamma) was conducted in 20 patients in order to determine whether combinations of high doses of IFN-beta and IFN-gamma were tolerable when administered under conditions which mimic conditions of in vitro antiproliferative studies. Patients received a 5-day continuous infusion of IFN-beta/IFN-gamma, followed by a 9-day rest period. Two cycles were administered. Doses of IFN-beta/IFN-gamma were escalated between 4 dose levels, with 5 patients per dose level. Dose-dependent side effects, consisting primarily of constitutional symptoms typical of those experienced with IFN, were observed. The maximally tolerated dose of continuous IFN-beta/IFN-gamma infusion was 3 x 10(6) units of IFN-beta and 200 micrograms of IFN-gamma. Dose-limiting side effects consisted of severe headache, fatigue, fever, and hepatic toxicity. No clinical responses were observed. Serum IFN was measurable only at the highest 3 dose levels. Only 5 patients (4 at the highest dose level) had total serum levels which exceeded 50 laboratory units/ml (55, 63, 800, 800, and 550 laboratory units/ml, respectively). In order to confirm the biological effectiveness of this schedule, we measured IFN-inducible proteins prior to therapy, 24 h after the initiation of the infusion, and at the completion of the 5-day infusion. 2'-5'-Oligoadenylate synthetase, serum beta 2-microglobulin, neopterin, and p78 levels all increased significantly, and serum tryptophan decreased significantly within 24 h after the initiation of treatment (P less than 0.0001). A dose-response effect was observed for serum beta 2-microglobulin, neopterin, and p78 (P less than 0.02). We retrospectively compared the results of this trial with those of another IFN-beta/IFN-gamma trial in which IFN-beta and IFN-gamma were administered by i.v. bolus. Within the limitations of a retrospective comparison, continuous infusion was less well tolerated than our previous schedule of bolus administration 3 times/week. However, the continuous infusion schedule appeared to be more effective in enhancing 2'-5'-oligoadenylate synthetase levels in mononuclear cells. We conclude that tolerable doses of IFN-beta and IFN-gamma do not result in serum IFN levels which produce significant synergistic antiproliferative responses in vitro. This study and other findings suggest that, unless higher doses can be achieved, combinations of IFN-beta and IFN-gamma are unlikely to have significant therapeutic activity.