Increased Incidence of Osteonecrosis (ON) with a Dexamethasone (DEX) Induction for High Risk Acute Lymphoblastic Leukemia (HR-ALL): A Report from the Children’s Oncology Group (COG).

Treatment with DEX, rather than prednisone (PRED), improves outcome for children with standard risk ALL. However, DEX exposure is strongly associated with the development of therapy-related ON, particularly in adolescents. Previous COG HR-ALL studies have shown a lower ON risk for patients receiving 1 vs. 2 delayed intensification (DI) phases, and for DI using discontinuous DEX (days 1–7 & 15–21) vs. continuous DEX (days 1–21), suggesting a strategy for giving the drug with acceptable toxicity. The HRALL study COG AALL0232 utilizes a modified augmented BFM backbone that compares in a 2x2 randomized design:

1. induction DEX (10 mg/M2/day x14 days) vs PRED (60 mg/M2/day x28 days), and

2. interim maintenance (IM) escalating-dose “Capizzi” methotrexate vs. high-dose (HD) MTX.

Induction rapid early responders (RER) receive single DI while slow responders receive double DI; all patients receive monthly 5-day DEX pulses during maintenance. To limit ON risk in adolescents, in the initial study design children ≥ 13y received discontinuous DEX during single or double DI; those <13y received continuous DEX. In 10/2006 the study was amended due to an unexpectedly high ON incidence in patients 10–12y receiving continuous DEX (28% @ 18 months) compared with historical controls given discontinuous DEX (3.4%). Subsequently all patients ≥ 10y have received discontinuous DEX; it is too early to assess the impact of this change. A comprehensive interim ON analysis was completed 4/2008 (reported as 24-month cumulative incidences). Overall ON incidence is 10.4% (110/1647), and is higher for those age ≥ 10 vs. <10y (15.2 vs 2.6%, p<0.0001, RHR=6.38); 99/110 cases of ON occurred in the older cohort. Among all patients, ON incidence is higher in DEX vs. PRED regimens (11.6 vs 8.7%, p=0.014, RHR 1.64); rates are similar for Capizzi MTX vs. HD-MTX regimens (10.4 vs 9.8%). Among patients ≥ 13y, incidence is higher in DEX vs. PRED regimens (18.9 vs 9.9%, p=0.02, RHR 1.97). There is no difference between regimens for children <10y. Among randomized RER patients ≥ 10y, incidence is higher in DEX vs. PRED regimens (17.2 vs 12.6%, p=0.006, RHR=1.79). For historical comparison, ON incidences by age cohort for RER patients on AALL0232 regimen PC (PRED + Capizzi MTX) vs. CCG-1961 regimen D (double DI with discontinuous DEX) were: <10y 4.1±3.4 vs. 2.0±1.4%, 10–12y 21.9±10.6 vs. 7.1±2.8%, and ≥ 13y 7.1±10.1 vs. 6.6±3.8%. To address these findings, the study was amended 6/2008. Patients ≥ 10y will be non-randomly assigned to induction PRED; the induction steroid randomization will continue for younger patients. Patients of all ages will receive discontinuous DEX during DI and PRED pulses during maintenance. Of note, compared with CCG-1961 the AALL0232 augmented BFM backbone was non-randomly modified in several ways that may affect ON incidence, including the use of pegaspargase during induction, a higher 15 mg intrathecal methotrexate dose for those age ≥ 9y, and monthly DEX instead of PRED maintenance pulses. Heightened awareness among caregivers may also have led to increased recognition and reporting of this toxicity. Using CTCAE v3.0 criteria, clinical ON severity among the 110 patients is: 3% grade 1, 60% grade 2, 35% grade 3, and 2% grade 4. Data regarding surgical intervention are being collected. These findings will directly influence the design of future trials in an effort to lessen the incidence and burden of this toxicity.