The purpose of this study was to, assess the efficacy of glycosylated recombinant human granulocyte colony-stimulating factor (lenograstim) in the prevention of neutropenia and infection in patients receiving dose-intensive chemotherapy for non-Hodgkin's lymphoma (NHL). A second objective was to determine clinical predicators of delay to cytotoxic chemotherapy administration. One hundred-sixty two patients with intermediate- or high-grade NHL and at least one poor prognostic factor received a total of 4 cycles of the LNH-84-regimen every 2 weeks, with an open randomization to treatment with anthracyclines. Patients were randomized to receive subcutaneous lenograstim 5 micrograms/kg/day (n = 82) or placebo (n = 80) from day 6 to day 13 of each cycle. The incidence of severe neutropenia (absolute neutrophil count (ANC) < 0.5 x 10(9)/L) was reduced in the lenograstim group compared with placebo (52% vs 75%). A significant reduction (p < 0.001) in the median duration of ANC < 0.5 x 10(9)/L was also observed in patients treated with lenograstim during each cycle of chemotherapy (0-1 day vs 2-4 days in placebo recipients). Fever occurred in 66 patients in each treatment group. Thirty-four percent of placebo recipients had documented infections during ANC < 1.0 x 10(9)/L compared with 18.5% of lenograstim-treated patients (p < 0.05). Infections of > or = 2 severity were significantly less frequent (p = 0.001) among lenograstim recipients compared with placebo (25 vs 49). The most common adverse events among lenograstim recipients were headache, mild bone pain and injection site reactions. Although lenograstim significantly increased (p = 0.0001) relative dose intensity compared with placebo (93% vs 80%), no difference in CR rate (67% vs 71%) or 3-year survival (63% vs 55%) was observed. The results of this study suggest that patients treated with a chemotherapy regimen that induces severe neutropenia can benefit from treatment with lenograstim. Furthermore, lenograstim permits treatment to be delivered at full dose intensity at 2 week intervals, even in patients with bone marrow involvement, and may permit further dose escalation of the chemotherapeutic regimen used.