Purpose: Since 1985, multiple centers have demonstrated that interleukin-2 (IL-2) and lymphokine-activated killer (LAK) cells produce durable anticancer responses in patients with metastatic renal cell carcinoma. High-dose recombinant IL-2 (rIL-2) has been administered by intravenous bolus injection (Rosenberg SA, et al: N Engl J Med 313:1485-1492, 1985) and by continuous intravenous infusion (West WH, et al: N Engl J Med 316:898-905, 1987) combined with lymphokine-activated killer (LAK) cells, with both methods producing responses in patients with advanced renal cell carcinoma. The Extramural IL-2/LAK Working Group has conducted a randomized phase II trial of two intravenous high-dose rIL-2 regimens (bolus three times daily or 24-hour continuous infusion) to determine if either one manifests greater anticancer activity or a more acceptable toxicity profile.
Patients and methods: Ninety-four patients with measurable advanced renal cell carcinoma were enrolled on this study: 46 to the bolus injection arm and 48 to the continuous infusion arm. On both arms, patients underwent a priming phase of rIL-2 administration, four daily lymphocytaphereses to harvest mononuclear cells that were placed in 3- to 4-day culture for generation of LAK cells, and an rIL-2/LAK coadministration phase. Patients were then observed monthly for evidence of response to this therapy and were offered up to two additional courses of treatment every 3 months if evidence of response was detected.
Results: Twenty percent of patients on the bolus injection arm experienced objective responses (three complete responses and six partial responses); 15% of patients on the continuous infusion arm responded (two complete responses and five partial responses). Complete responses were durable, persisting for 310+ to 700+ days. The incidence of severe life-threatening toxicities typical of high-dose rIL-2 therapy was similar in both arms (eg, patients with hypotension requiring pressors: bolus 71%, continuous 63%; oliguria less than or equal to 200 mL/8 hours: bolus 65%, continuous 71%). More episodes of fever, infection, and serum alkaline phosphatase elevation were associated with the continuous infusion arm, while more thrombocytopenia occurred on the bolus injection arm. Four patients (three bolus injection, one continuous infusion) died of respiratory and circulatory failure while under treatment. No clinical or laboratory parameter accompanying treatment on either arm was, by univariate or multivariate analysis, associated with an increased likelihood of response.
Conclusions: Both methods of high-dose rIL-2/LAK cell administration produce nearly equivalent anticancer activity and toxicity in the treatment of renal cell carcinoma. The ability to predict responding patients based on patient or treatment characteristics is not possible.