Early response to therapy, including assessment of MRD is a powerful prognostic factor in childhood ALL. However, there are many other known prognostic factors, and the precise relationship among these and MRD has been incompletely explored. In COG protocol P9900, patients were assigned to one of 3 risk-directed therapies based on NCI risk group, as well as TEL-AML1 translocations (TEL) and simultaneous trisomies of chromosomes 4 and 10 (DT). Infants and Ph+ patients were excluded. Good risk patients (9904) comprised NCI SR patients with either TEL or DT without CNS disease. Poor risk patients (9906) included a subset of NCI HR patients with particularly unfavorable risk factors based largely on age and white count, while 9905 had both NCI SR and HR patients not eligible for 9904 or 9906, including NCI HR patients with TEL or DT. MRD was satisfactorily measured with at least .01% sensitivity by 4-color flow cytometry in the marrow at end of induction (D29) or in blood at Day 8 (D8). D29 MRD was associated with poor outcome in all studies, and in both HR and SR patients on 9905 as shown:

Study4 y EFS ± SE (n)p value
MRD <.01%MRD >.01%
9904 Overall 94±1%(668) 75±6%(105) <.0001 
 TEL-AML1 94±2%(367) 68±12%(38) <.0001 
 DT 95±1%(316) 80±7%(69) <.0001 
9905 NCI-Standard Risk 89±2%(462) 63±6%(119) <.0001 
9905 NCI-High Risk 83±4%(292) 49±9%(73) <.0001 
9906  74±6%(156) 39±8%(84) <.0001 
Study4 y EFS ± SE (n)p value
MRD <.01%MRD >.01%
9904 Overall 94±1%(668) 75±6%(105) <.0001 
 TEL-AML1 94±2%(367) 68±12%(38) <.0001 
 DT 95±1%(316) 80±7%(69) <.0001 
9905 NCI-Standard Risk 89±2%(462) 63±6%(119) <.0001 
9905 NCI-High Risk 83±4%(292) 49±9%(73) <.0001 
9906  74±6%(156) 39±8%(84) <.0001 

The proportion of D29 MRD positive DT(17.9%) patients was higher than that of TEL patients (9.4%) but MRD was also prognostic when each group was considered separately; similar findings were seen in HR TEL and DT patients. TEL or DT was associated with a significantly better prognosis in both HR and SR patients who were MRD negative, but were not favorable factors in MRD positive patients. D8 blood MRD was also prognostic in all studies, with higher MRD levels generally associated with poorer outcome. Patients who were both D29 and D8 MRD negative had a good outcome overall (93% 4 y EFS); combining all favorable risk factors, including TEL, DT, NCI risk group, D8 and D29 MRD identified a group of patients (about 13% of B-precursor ALL patients) with 98± 1% 4 y EFS. In multivariate analysis, D29 MRD was the most important prognostic factor. NCI Risk group, DT and D8 blood were also significant, but TEL and day 8 morphologic assessment of marrow were not. We conclude that MRD is a major prognostic factor in childhood ALL even accounting for other risk factors. The prognostic significance of TEL and DT depends on MRD status. The combination of good risk clinical and genetic factors, and absence of MRD in both D8 blood and D29 marrow identifies a group of patients with outstanding outcome on limited therapy.

Disclosures: Research support from Becton Dickinson Biosciences.

Author notes

*

Corresponding author

Sign in via your Institution