Detection of minimal residual disease in childhood acute lymphoblastic leukemia by digital droplet PCR

Abstract

During the last years, the measurement of minimal residual disease (MRD) has become more and more accepted in the clinical and prognostic assessment of various malignant diseases. In childhood acute leukemic leukemia (ALL) as well, the determination of MRD burden during and after therapy is a crucial factor for the prognosis of an affected patient and further therapeutic steps. Although various methods can be used to measure MRD, the molecular approach, mainly using quantitative real-time polymerase chain reaction (qRT-PCT), is the current gold standard. Different target sequences, such as T-cell receptor and immunoglobulin rearrangements and specific fusion genes are used to monitor a specific blast subpopulation. Based on preferably two or more patient-specific target sequences, disease progression can be estimated. For qRT-PCR, each follow-up bone marrow aspiration is compared to a standard curve of blasts obtained at diagnosis onset, allowing to calculate the current blast burden in the patients' bone marrow. The sensitivity of qRT-PCR ranges up to the detection of one leukemia-type DNA copy in up to 10,000 (10^-4) and in some cases even 1 in 1,000,000 (10^-6) healthy copies. In this project, nine pediatric ALL patients and 17 different MRD targets have been investigated. By measuring dilution series of leukemic blast DNA in healthy DNA with dilution factors from 10^-1 to 10^-6 with qRT-PCR, reliable and reproducible MRD determination was possible. Eight targets reached a sensitivity of 10^-5, also eight a sensitivity of 10^-4 and two targets even 10^-6. Subsequently, MRD measurements were established with digital droplet PCR (ddPCR) as well. Here, the sample to be measured is split into up to 20,000 tiny partitions, thus enabling precise detection and absolute quantification of a PCR product. Eleven of our targets reached a sensitivity of 10^-5, however, among them are two samples with grey-zone results at the dilution of 10^-5. Five targets reached a sensitivity of 10^-4 and one target 10^-6. Unspecific background amplification occurred in both methods and hampered accurate sample determination in one but not the same target per method. Within this project it was possible to visualize all MRD sequences of our nine exemplary patients well and to achieve the same sensitivities with ddPCR as with qRT-PCR. ddPCR mainly offers the advantage to absolutely quantify blasts present at any later follow-up time point without the need for a dilution curve. Further experience in our laboratory and international guidelines for MRD measurement by ddPCR are next necessary steps towards the clinical application of this method.