No Influence on the Incidence of Osteonecroses When Dexamethasone Replaces Prednisone during Induction Treatment for Childhood ALL: Results of Trial ALL-BFM 2000.

Osteonecroses (ON) are common complications associated with the treatment of acute lymphoblastic leukemia (ALL). For symptomatic ON, incidences of 1.6% to 9.3% have been published in the literature. The range of incidence rates may be related to treatment, data acquisition methods and patients’ age distribution. Between 09/99 and 06/06, 3,048 protocol patients with ALL (age 1–17 years) were enrolled into the multicenter trial ALL-BFM 2000 (Germany, Austria, Switzerland). Data on symptomatic ON were prospectively collected and were reported in 111 patients (3.6%, 95% confidence interval (CI) 3.0–4.4%). The median age of the patients with ON was 12.8 years (range 1.3–17.9 years), the median time from ALL diagnosis until diagnosis of ON was 15 months (range 2–73 months; data available from 92 patients). The incidence in adolescents, in particular adolescent girls, was significantly higher than in younger children (age 2 /day prednisone vs. 10 mg/m 2 /day dexamethasone for three weeks plus tapering) did not influence the ON rate (prednisone 3.2%, dexamethasone 3.0%, p=0.77, 95% CI for difference: −1.2 – 1.6%). Various treatment attempts of ON were reported; 10 patients have been documented to be free of symptoms. At least 38 patients underwent surgical interventions, mainly core decompressions. Total endoprostheses of one or both hips were implanted in at least 10 patients. Final evaluation of all patients in particular with regard to long-term sequelae is not yet possible. The existing data demonstrate the far-reaching consequences of ON for the quality of life in a considerable proportion of the patients. Comparing these results with the ON data of study ALL-BFM 95 (Burger et al., Pediatr Blood Cancer, 2005) which were collected in a predominantly retrospective way, the data suggest an increase in ON incidence in the current trial ALL-BFM 2000 (ALL-BFM 95: 1.6%). As the control treatment arms of the two trials were quite similar and the randomized treatment modifications in ALL-BFM 2000 had no influence on the ON rate, the higher incidence may be attributed to the prospective data collection in the current study and possibly to a higher attention to these complications by the attending doctors.