Treatment of malignant melanoma and renal cell carcinoma with recombinant human interleukin-2: analysis of cytokine levels in sera and culture supernatants.

In this study we evaluated the clinical response of 12 patients with malignant melanoma and renal cell carcinoma (RCC) following administration of recombinant human interleukin-2 (rhIL-2) by continuous infusion. Serum samples taken before, during and following sequential courses of IL-2 were assayed for the presence of tumour necrosis factor alpha (TNF-alpha) IL-1 alpha, IL-6 and interferon gamma (IFN-gamma) and the presence or changes in these cytokines were examined with respect to clinical response data: our results did not show any direct correlation between the parameters measured and clinical outcome. In addition, peripheral blood mononuclear cells (PBMC) derived from 3 RCC patients were cultured in a serum-free environment and the resulting supernatants assayed for the production of these cytokines and compared to the corresponding serum levels. During one or more courses of treatment only 1 patient, who had metastatic bone disease, demonstrated detectable serum TNF-alpha; serum IL-6 levels were elevated in a proportion of all patients studied and a sustained IL-6 response occurred in a patient who had complete disease remission; IL-1 alpha was detected in the serum of 3 RCC patients; IFN-gamma could not be detected in any serum sample tested. Cytokine levels in sera and supernatants derived from 3 RCC patients were compared but no correlation was found: TNF-alpha and IL-6 were shown to be present at much higher concentrations in supernatants when compared to sera whereas the levels of IL-1 alpha were almost undetectable. This lack of correlation is probably due to the presence of "interfering" proteins in sera which either depress or enhance the ability to detect cytokines in sera using enzyme immunoassays.