Lower incidence of CNS leukemia using dexamethasone instead of prednisone for induction in acute lymphocytic leukemia

Patients with acute lymphocytic leukemia were randomized to receive vincristine and either prednisone 40 mg/m²/day or dexamethasone 6 mg/m²/day with or without L asparaginase for their initial therapy. Two mth after the onset of remission all patients received intrathecal methotrexate 12 mg/m²/wk for 6 doses alone or combined with cranial radiation 2400 rad. Of the 223 who received dexamethasone there were 20 who have had a CNS relapse as compared to 44/257 treated with prednisone. This represents 1.13 occurrences of CNS leukemia per 100 mth of patient observation for prednisone as compared to 0.56 for dexamethasone. In independent analyses of induction, maintenance and CNS prophylaxis a consistently lower incidence of CNS relapses occurred in the dexamethasone group. The effect cranial radiation within each steroid group was not as marked as the effect of dexamethasone. The marrow remission rate is the same for each of the steroids, namely 85% and the remission durations were identical. Thus, there are no apparent disadvantages to using dexamethasone as the induction steroid and definite advantages in terms of subsequent reduction in incidence of CNS leukemia.