Outcomes for B-Precursor Patients in Legacy Children’s Cancer Group (CCG) and Pediatric Oncology Group (POG) Studies in Childhood Acute Lymphoblastic Leukemia (ALL): A Children’s Oncology Group (COG) Report.

Background: Improved post induction intensification (PII) has led to improved outcome for children with ALL over past decades. The merger of POG and CCG provided an opportunity to compare strategies, specifically, POG: intermediate dose methotrexate (Mtx, 1–2 g/m²) with leucovorin (lv) rescue; CCG: Berlin Frankfurt Münster (BFM) Protocol Ib and Protocol II (delayed intensification, DI), “augmented” with additional vincristine (Vcr), escalating-dose parenteral Mtx with no lv rescue, and l-asparaginase (Asp).

Methods: Between 1996 and 2005, POG 9904/5/6 trials and CCG 1991/1961 trials accrued 6774 B-precursor patients, age > 1 year. NCI standard risk (SR) patients received induction with Vcr, Asp, dexamethasone; NCI higher risk (HR) patients received induction with Vcr, Asp, daunorubicin, and prednisone. POG allocated PII based on age, gender, presenting WBC and the presence or absence of double trisomy 4+10 or t(12;21). CCG allocated PII based on age, presenting WBC, and the Induction Day 7/14 marrow response. POG SR patients without trisomies 4+10 were randomized +/− DI. CCG SR patients received at least one DI. CCG SR slow early response (SER) patients received daunomycin in induction and augmented intensification. ALL CCG HR patients received at least one DI and many received “augmented” intensification; some POG HR patients received no DI. A subset of POG HR patients, defined by age, gender, and WBC, with very high risk of relapse were designated “Pragmatic Very High Risk” (VHR) and assigned to a modification of CCG Augmented BFM (

Results: The 5-year EFS was 84±2% (n=1831) and 88±2% (n=2539) for POG and CCG SR patients (p<0.003). POG and CCG SR patients with double trisomies 4+10 had 5-year EFS of 91±3% (n=387) with no DI and 95±3% (n=317) with one or two DI’s, respectively. The 5-year EFS was 60±4% (n=992) and 71±2% (n=1412) for POG and CCG HR patients (p< 0.001). Results are similar when adjusted for ethnicity. Both SR and HR patients on CCG trials had a statistically significant 25% reduction in risk of an adverse event.

Conclusions: This is a non-randomized cross-study comparison. Differences in patient population, treatment assignment, and clinical practice may affect results and interpretation. However, our data suggest the overall superiority of the CCG modified BFM strategy and support the COG decision to build current trials, e.g., AALL0331 and AALL0232, on this platform.