All major types of human interferons (IFNs) have been purified and clinically administered as antitumor agents. We summarize here experience to date with toxicity of IFNs in cancer patients. The acute syndrome consists of fever, chills, myalgias, arthralgias, and headache, with some variation according to type of IFN, route of administration, schedule, and dose. Fatigue, perhaps reflecting CNS toxicity, is the most prevalent nonacute symptom. At high doses, IFNs are neurotoxic; the abnormalities seen by EEG resemble those in diffuse encephalitis. Hematologic toxicity consists mainly of leukopenia, but anemia and thrombocytopenia occur in some patients. Nausea, vomiting, and diarrhea are the main gastrointestinal symptoms. Elevation of serum transaminases seems to reflect liver toxicity. Renal function is well preserved, except for rare instances of acute renal failure. Cardiac toxicity remains questionable, although heart failure and arrhythmias have been associated with the administration of IFNs. Most, if not all, of these effects are reversible or can be ameliorated. With IFN alpha, the type most widely used in clinical studies, doses of 1 million to 9 million units (MU) are generally well tolerated, but doses greater than or equal to 18 MU yield moderate to severe toxicity. Doses greater than or equal to 36 MU can induce severe toxicity and significantly alter the performance status of the patient.

We treated seven patients who had progressive hairy-cell leukemia with daily doses of 3 million units of partially pure alpha (leukocyte) interferon by the intramuscular route. Three patients had a complete remission, and four had a partial remission, according to strict criteria for a response. After treatment, bone-marrow aspirates showed an absence of leukemia cells in three patients and 5 per cent or fewer in three others. Normalization of subnormal peripheral-blood values occurred in six of six patients with anemia, in seven of seven with granulocytopenia, and in four of four with thrombocytopenia. Remissions have been maintained for over 6 to over 10 months. Alpha interferon appears to be highly effective in patients with hairy-cell leukemia.

Sixteen patients with advanced cancer were treated with recombinant-DNA-produced pure leukocyte A interferon (IFLrA) intramuscularly in doses ranging from 3 to 198 X 10(6) units. with interval periods of 72 to 96 hours between doses. At the two lowest doses of 3 and 9 million units, there was a cross-over evaluation between IFLrA and partially pure leukocyte interferon (IFN-C) produced from human cells. THe maximum observed serum concentration of IFLrA measured by enzyme immunoassay and bioassay increased with increasing doses. The mean serum concentrations of IFLrA and IFN-C were similar. Clinical effects produced by IFLrA and IFN-C were similar, including fever, chills, myalgias, headache fatigue, and reversible leukopenia and granulocytopenia. Eight patients had transient and mild numbness of the hands or feet, or both. Three patients developed low titers of antibody to IFLrA, Seven of 16 patients showed objective evidence of tumor regression during the study.

Fifty-six patients with metastatic renal cell carcinoma (RCC) were treated with recombinant DNA-derived interferon alpha (rIFN alpha A). The first 30 patients were randomized between doses of 2 X 10(6) U/m2 and 20 X 10(6) U/m2 intramuscularly daily. No complete (CR) or partial (PR) remissions were achieved in 15 patients receiving the low dose. In contrast, 27% of those receiving the high dose achieved CR or PR. Subsequently, 26 additional patients were given the high dose and achieved a 31% response rate. Remissions lasted from 1 to more than 12 months (median, 3 months). Responses occurred predominantly in lung parenchyma or mediastinal node metastases. Toxicity of the high dose required dose reduction in 50% of the patients. Neutralizing antibodies to rIFN alpha A developed in seven of 12 responsive (58%) and nine of 29 (31%) nonresponsive patients (P = greater than .5). The median duration of remission among the antibody-positive and antibody-negative patients were 2 and 10 months, respectively (P = .009). The clinical significance of the antibodies to rIFN alpha A remains unclear, but the coincidence between the detection of antibodies and the early relapse of the disease in some responsive patients suggests that these antibodies may abrogate the biologic activity of rIFN alpha A. This effect, however, was not associated with adverse clinical sequelae.

No abstract