TEL-AML1 fusion transcripts are found in 25% of children with B-cell precursor ALL (BCP-ALL). From June 1993 to December 1999, 1195 children with BCP-ALL were included in the FRALLE 93 protocol. Out of these, 792 were evaluated for TEL-AML1 transcript expression. There is no difference in terms of initial features, DFS, EFS, survival between evaluated (792) and non evaluated (403) patients. Out of the 792 pts, 191 (24%) expressed TEL-AML1 transcripts at diagnosis. To assess the potential prognostic value of TEL-AML1 transcripts quantification, we have retrospectively analysed follow up marrow samples using Europe Against Cancer procedures for real time quantitative RT-PCR assay, on ABI PRISM 7700 (2 reference labs) and Light Cycler apparatus (1 reference lab). Out of the 191 TEL-AML1+ve pts, 83 were evaluated for MRD at different time points after induction therapy (median = D41 (34–55) (53 evaluable pts), at D111 (62–158) (62 pts), at D216 (159–325) (33 pts) and at D838 (365–1287) (49 pts). According to normalized Ct values, samples were attributed to 4 MRD level ranging from 0 to 3 and defined as follows: 0: Ct>40 ; 1 : 36<Ct≤40 ; 2 : 33<Ct≤36 ; 3 : Ct≤33, corresponding respectively to undetectable MRD ; MRD<10-4 ; 10-4≤MRD<10-3 ; MRD≥ 10-3, with respect to dilution of REH cDNA.

Distribution of pts according to MRD level at different time points after induction treatment are summarized in the following table. Seventeen relapses have occurred at a median time of 41 months (17–73)(bone marrow: 7, BM + other: 5, testis: 3, CNS: 2). A level 2 positivity at the end of induction was associated with an increased risk of relapse of 3.31(95%CI:1.02 – 10.76, p =.047) while level 3 positivity was associated with a relative risk of 9.52 (95%CI: 2.91 – 31.08, p =.0002). Positivity at D111 was associated with an increased risk of relapse of 8.6 (2.0 – 38.5, p = 0.0042), whatever the level. Combination of data obtained at D41 and D111 allows to distinguish 3 subsets of pts with decreasing relapse-free survival: from 97.5% (95%CI: 85–100%) in pts with no positivity at D111 whatever the D41 result, to 75% (95%CI: 58–92%) in pts with MRD +ve at D111 with low level at D41 and 42% (95%CI: 14–69%) in pts with MRD +ve at D111 with level 2 or 3 at D41 (p<.0001). No other prognostic factor was found (age, sex, WBC, D8 steroid response, D21 bone marrow response) which renders the MRD profile unique in this matter.

Conclusion: RQ-PCR-based MRD detection is a powerful prognostic tool in TEL-AML1+ve leukemia. Combination of two time points allows a relevant stratification of pts according to the risk of relapse, compatible with clinical decision making towards intensification or deescalation in the setting of controlled trials

FU time pointNumber of ptsin MRD classes(number ofrelapses)
 Not evaluated 
D41 27 (3) 11 (2) 10 (4) 5 (4) 30 (4) 
D111 40 (2) 14 (6) 7 (2) 1 (1) 21 (6) 
D216 29 (2) 2 (1) 1 (0) 1 (1) 50 (13) 
D838 47 (8) 1 (0) 1 (1) 34 (8) 
FU time pointNumber of ptsin MRD classes(number ofrelapses)
 Not evaluated 
D41 27 (3) 11 (2) 10 (4) 5 (4) 30 (4) 
D111 40 (2) 14 (6) 7 (2) 1 (1) 21 (6) 
D216 29 (2) 2 (1) 1 (0) 1 (1) 50 (13) 
D838 47 (8) 1 (0) 1 (1) 34 (8) 

Author notes

Corresponding author

Sign in via your Institution