TO THE EDITOR
PCR-based methods for molecular analysis of minimal residual leukemia (MRL) are increasingly used in routine diagnostic laboratories. The recent development of a real-time quantitative PCR (RQ-PCR) method based on fluorescent Taqman assay provides a new powerful tool for the accurate quantitation of a number of DNA and RNA templates. With this methodology, the PCR product is measured during the exponential phase of the amplification process when none of the reaction components is rate-limiting, showing highly reproducible estimates. TEL/AML1 is a fusion mRNA derived from t(12;21)(p13;q22) cryptic translocation, the most frequent genetic lesion in pediatric B cell precursors acute lymphoblastic leukemia (BCP-ALL).1 The prognosis of patients with TEL/AML1 BCP-ALL varies among different series. Some investigators reported a better clinical outcome for this leukemia subtype,2 while others could not confirm this observation.3 We describe here a RQ-RT-PCR assay to quantitate TEL-AML1 fusion transcripts in BCP-ALL. We first optimized the assay for both primers and experimental conditions by using the t(12;21)-expressing human cell line Reh in spiking experiments. We found that the main parameter affecting the amplification efficiency was the primer grade synthesis. As shown in Figure 1, for a given amount of input RNA and the same optimized RT-PCR conditions, we observed a fall of more then 2 Ct units of TEL/AML1 target message when the PCR reaction was conducted with highly purified primers rather than non-purified primers. This led to a gain in sensitivity equivalent to at least a four-fold increase as in the example given.
Reh cells express the two major fusions transcripts variants resulting from the alternative splicing of AML1 exon 2 at levels similar to those observed in virtually all diagnostic samples so far analyzed. Therefore, we chose total RNA from Reh cell line as a standard for TEL/AML1 quantitation. Quantitation of fusion transcripts in unknown samples can then simply be done by measuring the sample's Ct (threshold cycle) and by using a standard curve constructed either (1) by serial dilutions of 1 μg of Reh cell line RNA-equivalent cDNA into 1× RT buffer or, to reproduce more closely clinical sample analysis, (2) with serial dilutions of 1 μg of Reh cell line RNA in a constant amount of mouse RNA, prior to cDNA synthesis. For each standard curve the amounts of the diluted total Reh RNA or cDNA equivalent varied from 100 to 0.01 ng. In both cases we found a strong linear relationship between the Ct and the starting copy number of the TEL/AML1 transcripts (R2 ⩾ 0.99) over a range of at least four orders of magnitude. Also, the PCR efficiency of TEL/AML1 amplification did not differ significantly with the two methods, as indicated by the slope of the curves obtained with the RNA and cDNA point dilutions (−3.76 and −3.66 corresponding to a PCR efficiency value (E) of 84% and 87% respectively). The TEL/AML1 fusion transcripts could be detected after 104-fold dilution of 1 μg of Reh total RNA in mouse total RNA, which is equivalent to 10 pg of target RNA. The average Ct of the TEL/AML1 target at this dilution was 37.8 with a within-run coefficient of variation (CV) of Ct values varying from 0.8 to 2.4 in different PCR runs. Thus the sensitivity of the method was comparable to that of conventional end-point RT-PCR used for diagnostic purposes. At lower RNA concentration (5 × 104 and 105-fold dilutions), positive samples were still detected but the variation among sample replicates became too high for accurate measurement. Indeed, as the Ct values of TEL/AML1 increased to over 40, the CV among sample replicates was greater than 2.5%, indicating that the quantitation of positive results was not reliable. As the CV for Ct values of triplicates samples indicates the reproducibility of the measure, the limit of accurate quantitation can be easily deduced from the CV.
Because the limit of target detection and quantitation was not the same, we introduced a ‘reference sample’ to define the quantitation limit. The reference sample represents the smallest amount of target messenger that can be detected and accurately quantified. This limit was a dilution corresponding to 50 pg of human cell line Reh total RNA in 1 μg of mouse RNA. At this concentration the Ct values of TEL/AML1 varied between 35 and 36, but the within-run variation of the averaged Ct was always below 2%. Moreover, the reference sample provides a quantitative internal control, both for the efficiency of the RT-PCR procedure and for RNA and cDNA recovery (Figure 2).
Variations in starting amount of total RNA or RT efficiency among samples were taken into account by parallel amplification of an endogenous mRNA control. We chose the TBP gene (TATA box-binding protein, a component of the DNA-binding protein complex TFIID) as endogenous control because it is an ubiquitous transcription factor whose expression is constant in normal tissues and is similar to TEL/AML1 expression in the cell line Reh (TBP-Ct and TEL-AML-Ct <22, for 100 ng of total RNA, with a within-run CV 2%). We rejected PBGD gene because of its low expression level compared to TEL/AML1. The endogenous control should be expressed at an adequate level (ideally, at a level equivalent to the standard) in order to avoid underestimating positive samples and obtaining false-negative results. For each experimental sample, the amount of target and endogenous reference was determined from the standard curve. Then the target amount was divided by the endogenous reference amount to obtain a normalized target value.
We validated the RQ-RT-PCR assay on bone marrow (BM) samples from four patients (Table 1). Our study was not aimed to assess the clinical value of TEL/AML1 detection during CR but rather to evaluate the potential use of RQ-RT-PCR for quantitation of MRL to have prognostic significance. Although this small number of patients does not allow us to draw firm clinically relevant conclusions the comparison of quantitative TEL/AML1 results and clinical outcome yield interesting preliminary findings. First, normalized TEL/AML1 values at diagnosis and relapse were always above 290. Even if the normalized values did not reflect the percentage of BM blasts (all the samples tested contained more then 90% of blasts after Ficoll enrichment), it should be possible to establish a threshold value corresponding to the level of morphologically detectable disease. Second, at the time of CR (day 35–45), all the patients had a reduction in TEL/AML1 values of at least two logs. Interestingly, the highest values were found in patients UPN3 and UPN4 who subsequently relapsed. Furthermore these patients developed detectable and quantifiable disease 3 to 5 months after entering the first CR, whereas patients UPN1 and UPN2 had TEL/AML1 below the quantitation limit. Patients UPN3 and UPN4 were consistently MRL positive.
We have shown that our RQ-RT-PCR assay can be used to calculate a slope of disease response to therapy for each patient. Thus, it might be possible to compare directly the relative efficacy of treatment regimens given to patients in different institutions, and therapy could conceivably be adapted to the individual response. The high efficiency and throughput of real-time PCR, as well as the absence of post-PCR manipulations, make Taqman methodology eminently suitable for routine large-scale application and an inter-laboratories standardization.
Romana SP, Poirel H, Leconiat M-Y, Flexor M, Mauchauffé M, Jonveaux P, Macintyre EA, Berger R, Bernard OA . High frequency of t(12;21) in childhood B-lineage acute lymphoblastic leukemia Blood 1995 86: 4263–4269
McLean TW, Ringold S, Neuberg D, Stegmeier K, Tantravahi R, Ritz J, Koeffler HP, Takeuchi S, Janssen JWG, Seriu T, Bartram CR, Sallan SE, Gilliland GD, Golub TR . Tel/Aml1 dimerizes and is associated with a favorable outcome in childhood acute lymphoblastic leukemia Blood 1996 88: 4252–4258
Harbott J, Viehmann S, Borkhardt A, Henze G, Lampert F . Incidence of Tel/Aml1 fusion gene analyzed consecutively in children with acute lymphoblastic leukemia in relapse Blood 1997 90: 4933–4937
Bièche I, Onody P, Laurendeau I, Olivi M, Vidaud D, Lidereau R, Vidaud M . Real-time reverse transcription-PCR assay for future management of ERBB2-based clinical applications Clin Chem 1999 45: 1148–1156
This work was supported by a grant from the Direction de la Recherche Clinique de l'Assistance Publique-Hôpitaux de Paris (AP-HP), Projet TBI 97-297, France (PB).
About this article
Molecular genetic analysis of haematological malignancies: I. Acute leukaemias and myeloproliferative disorders
Clinical and Laboratory Haematology (2005)
Real-time quantification of TEL–AML1 fusion transcripts for MRD detection in relapsed childhood acute lymphoblastic leukaemia
Leukemia Research (2004)
Detection of minimal residual disease in hematologic malignancies by real-time quantitative PCR: principles, approaches, and laboratory aspects
Standardization and quality control studies of ‘real-time’ quantitative reverse transcriptase polymerase chain reaction of fusion gene transcripts for residual disease detection in leukemia – A Europe Against Cancer Program
Evaluation of candidate control genes for diagnosis and residual disease detection in leukemic patients using ‘real-time’ quantitative reverse-transcriptase polymerase chain reaction (RQ-PCR) – a Europe against cancer program