Dose escalation of cytotoxic drugs using haematopoietic growth factors: a randomized trial to determine the magnitude of increase provided by GM-CSF.

BACKGROUND The magnitude of chemotherapy dose escalation made possible by the use of recombinant haematopoietic growth factors has not been quantified in a randomized trial. PATIENTS AND METHODS Patients with refractory or relapsing Hodgkin's disease were randomized to receive the Dexa-BEAM regimen with escalating etoposide doses supported by placebo or granulocyte-macrophage colony-stimulating factor (GM-CSF). Using an adaptive sampling method independently in both arms, the etoposide dose was escalated until the maximal tolerated dose for the first cycle was reached. RESULTS Thirty patients were randomized to GM-CSF and thirty to placebo. The etoposide dose could be escalated considerably in both treatment arms. Maximal etoposide dose for the first cycle was 1920 mg/m2 for patients receiving GM-CSF and 1160 mg/m2 for patients receiving placebo (P = 0.045 one-sided), corresponding to a 65% higher etoposide dose and a 13% higher dose intensity with GM-CSF. Dose-limiting events were similar in both arms, consisting mainly of prolonged neutropenia and consecutive infections. Treatment efficacy was not different in the two treatment groups. CONCLUSIONS While GM-CSF permits a somewhat higher dose escalation than placebo, the increase in dose intensity provided by GM-CSF is small. The use of CSF for interval reduction rather than dose escalation is the more effective strategy for dose intensification.