A DNA Real-Time Quantitative PCR Method Suitable for Routine Monitoring of Low Levels of Minimal Residual Disease in Chronic Myeloid Leukemia

Abstract

The BCR-ABL1 sequence has advantages over the BCR-ABL1 transcript as a molecular marker in chronic myeloid leukemia and has been used in research studies. We developed a DNA real-time quantitative PCR (qPCR) method for quantification of BCR-ABL1 sequences, which is also potentially suitable for routine use. The BCR-ABL1 breakpoint was sequenced after isolation by nested short-range PCR of DNA from blood, marrow, and cells on slides, obtained either at diagnosis or during treatment, or from artificial mixtures. PCR primers were chosen from a library of presynthesized and pretested BCR (n = 19) and ABL1 (n = 568) primers. BCR-ABL1 sequences were quantified relative to BCR sequences in 521 assays on 266 samples from 92 patients. For minimal residual disease detectable by DNA qPCR and RT-qPCR, DNA qPCR gave similar minimal residual disease results as RT-qPCR but had better precision at low minimal residual disease levels. The limit of detection of DNA qPCR depended on the amount of DNA assayed, being 10(-5.8) when 5 μg was assayed and 10(-7.0) when 80 μg was assayed. DNA qPCR may be useful and practical for monitoring the increasing number of patients with minimal residual disease around or below the limit of detection of RT-qPCR as the assay itself is simple and the up-front costs will be amortized if sequential assays are performed.

BCR-ABL1expression, RT-qPCR and treatment decisions in chronic myeloid leukaemia

“…Measurement variation includes both inherent method variation and stochastic Poisson variation, the latter occurring at low MRD levels when the mean number of targets in the assay is very low. For inherent method variation, the SD of RT-qPCR is approximately 0.1 log3 4 and that for DNA-qPCR is approximately 0.2 log 2. The SD of inter-individual variation in gene expression, which is only relevant to RT-qPCR, was 0.36 log in the present study, the 95% confidence limits being 0.27 and 0.47 log.…”

“…When MRD is very low, within approximately 1 log of the limit of detection (LOD), stochastic Poisson variation becomes an additional source of measurement variation. The LOD of RT-qPCR2 12 is approximately 10 −4.6 , and both expression variation and Poisson variation, therefore, influence decisions on discontinuation of therapy. The LOD for DNA-qPCR2 is lower, 10 −6 –10 −7 , depending on the amount of DNA assayed.…”

“…The LOD of RT-qPCR2 12 is approximately 10 −4.6 , and both expression variation and Poisson variation, therefore, influence decisions on discontinuation of therapy. The LOD for DNA-qPCR2 is lower, 10 −6 –10 −7 , depending on the amount of DNA assayed. Consequently, MRD can still be measured by DNA-qPCR in many patients who are negative by RT-qPCR2 15 16 and Poisson variation does not become important until MRD is <10 −5.5 .…”

“…The LOD for DNA-qPCR2 is lower, 10 −6 –10 −7 , depending on the amount of DNA assayed. Consequently, MRD can still be measured by DNA-qPCR in many patients who are negative by RT-qPCR2 15 16 and Poisson variation does not become important until MRD is <10 −5.5 .…”

Comment on: Technical Issues Behind Molecular Monitoring in Chronic Myeloid Leukemia

“…dPCR therefore holds promise for future monitoring of CML patients as it can improve the level of sensitivity, offering a potential new threshold for stratification of patients achieving deep molecular responses. While DNA-based qPCR BCR-ABL1 monitoring may provide superior sensitivity to RNA-based qPCR [9], one of the barriers for implementation is the requirement for characterisation of each patient's intronic fusion junction and design of patient-specific primers. This requires optimisation of each patient-specific assay which must be reconciled against the alternative of a single RNA-based qPCR assay suitable for all patients.…”