Relapse o f TEL-AML1 -Positive A cute L ymphoblastic Leukemia i n C hildhood: A M atched-Pair A nalysis

Purpose: The aim of this study was to investigate whether, in relapsed childhood acute lymphoblastic leu- kemia (ALL), the frequent genetic feature of TEL-AML1 fusion resulting from the cryptic chromosomal transloca- tion t(12;21)(p13;q22) is an independent risk factor. Patients and Methods: A matched-pair analysis was performed within a homogeneous group of children with first relapse of BCR-ABL-negative B-cell precursor (BPC) ALL treated according to relapse trials ALL-Rezidiv (REZ) of the Berlin-Frankfurt-Munster Study Group. A total of 249 patients were eligible for this study: 53 (21%) were positive for TEL-AML1, and 196 (79%) were negative. Positive patients were matched for estab- lished most-significant prognostic determinants at re- lapse, time point, and site of relapse, as well as age and peripheral blast cell count at relapse. Results: Fifty pairs matching the aforementioned cri- teria could be determined. The probabilities with SE of event-free survival and survival at 5 years for matched TEL-AML1 positives and negatives are 0.63 6 0.10 versus 0.38 6 0.10 (P 5 .09) and 0.82 6 0.09 versus 0.42 6 0.19 (P 5 .10), respectively. These results were confirmed by multivariate analysis, revealing an in- dependent prognostic significance of time point and site of relapse (both P < .001) but not of TEL-AML1 expression (P 5 .09). Conclusion: TEL-AML1 expression does not consti- tute an independent risk factor in relapsed childhood BCP-ALL after matching for relevant prognostic param- eters. It undoubtedly characterizes genetically an ALL entity associated with established favorable prognostic parameters. High-risk therapeutic procedures such as allogeneic SCT should be considered restrictively. J Clin Oncol 19:3188-3193. © 2001 by American Society of Clinical Oncology.