of amplification facilitator. Taken together, these data show that only the addition of 0.04% BSA would allow a quantitative follow-up of the MRD. In the case of addition of 0.2% of BSA, there was a decrease in ABL CN (Figure 2a) and an increase in TEL-AML1 CN (Figure 2b) forbidding any accurate quantification. Actually, this effect was due to a variation of cycle threshold (C t ) value for the cDNA measurements with no modification of the plasmid standard curve (not shown).We studied 12 'poor-quality' samples of patients who have a clinical and/or biological disease and possess TEL-AML1 FG. In 70% of the cases with or without ABL amplification, we recovered TEL-AML1 expression by the addition of 0.04 and/or 0.2% BSA. However, only one TEL-AML1-positive sample had an increase in ABL CN higher than 1000 by the addition of BSA in RQ-PCR, allowing quantification of the TEL-AML1 expression 6 (Table 3).So, first, we have shown that the addition of BSA to TaqManbased RQ-PCR analysis improves amplification of 'poor-quality' samples. 6 Secondly, the addition of BSA is better than cDNA dilution or cDNA purification on column. Third, in certain cases, it allows improvement of relevant levels of FG like TEL-AML1.In a clinical laboratory, each patient sample set for molecular MRD analysis is very valuable, so any improvement of RNA quality even in some samples is suitable to apply in practice.Therefore, based on our data, this suggests that 0.04% BSA could be added systematically in ABL and TEL-AML1 RQ-PCR. This BSA concentration does not modify the RQ-PCR results of amplifiable samples and allows recovery to correct amplification of transcripts in some 'poor-quality' samples. Finally, a large prospective study is warranted to definitively assess the use of addition of 0.04% BSA in CG and FG RQ-PCR in a quantitative follow-up of the MRD.