Towards DNA-Based Monitoring of Therapy In Chronic Myeloid Leukemia

Abstract 2284 Monitoring of CML is conventionally performed by RT-qPCR but the use of RNA as the starting material leads to a number of disadvantages, which include propensity of RNA to degradation, a low level of endogenous BCR-ABL transcripts, requirement for reverse transcription, somewhat complex laboratory standardisation, and uncertain relationship between transcript level and leukemic cell number. The use of genomic DNA to quantify BCR-ABL sequences would overcome these disadvantages but it presents 2 major problems - the need to determine the breakpoint sequence for each patient and the need to synthesise and test specific primers for each patient. We developed a simple and rapid method for isolating the BCR-ABL sequence which involved a primary highly multiplexed PCR to amplify across the breakpoint and 2 subsequent PCRs using specially designed primers to eliminate the non-specific material produced in the primary PCR. The need to develop primers specific for each patient was bypassed by synthesising a library comprising 11 pairs (to enable nested PCR) of forward BCR primers and 493 pairs of reverse ABL primers which were all tested to ensure efficient amplification. Minimal residual disease (MRD) was quantified by a one round qPCR if the level was anticipated to be > 10-3, by nested qPCR if the level was anticipated to be < 10-4, and by one or both methods if the level was anticipated to be in the range 10-3 – 10-4. Quantification involved use of a standard curve comprised of various dilutions of diagnosis DNA and amplification was measured using a Taqman hydrolysis probe directed to the BCR region of the BCR-ABL sequence. MRD levels quantified by RT-qPCR were also available for 75 of the 79 samples. To date, 217 MRD estimations have been performed by DNA-qPCR on 79 samples obtained from 30 patients during treatment. MRD was quantified down to approximately 10-6. There was a highly significant correlation (p < 0.001) between the results of DNA-qPCR and RT-qPCR for the 54 samples in which MRD was quantified by both methods. MRD was quantified by DNA-PCR but was not detected by RT-qPCR in another 15 samples and was not detected by either technique in a further 6 samples. There were no samples in which MRD was detected by RT-qPCR but not by DNA-qPCR. Overall, the MRD result was significantly lower (p < 0.01) by a mean of 0.38 log units (2.4 fold) when quantified by RT-qPCR as compared to when quantified by DNA-qPCR. Replicate DNA-qPCR estimations on the same sample showed that the SD of a single estimation was 0.25 log units (95% CI = 3.2 × result to 0.31 × result) and there was no significant difference between a one-round qPCR and a nested qPCR for the 14 samples estimated by both methods. Analysis of the difference between the RT-qPCR result and the DNA-qPCR result for each sample drawn from an individual patient showed systematic differences between patients which were almost certainly attributable to differences in BCR-ABL expression. The Figure shows the difference between the RT-qPCR result and the DNA-qPCR result for the 17 patients for whom two or more samples had been quantified by both techniques. Each point refers to an individual sample. Some patients e.g. the 3 on the left of the Figure, appeared to be high expressors and other patients e.g. the 2 on the right, appeared to be low expressors. The level of expression differed from the mean by up to approximately 1 log unit (10 fold). Conclusions 1. DNA-qPCR provides MRD results which are broadly comparable to RT-qPCR results when MRD is > 10-4. 2. DNA-qPCR is more analytically sensitive than RT-qPCR by up to 2 log units. This may become increasingly important as more patients achieve complete molecular response as defined by RT-qPCR. 3. DNA-qPCR provides a more accurate measure of leukemic cell number than RT-qPCR as it is not influenced by the level of BCR-ABL expression. 4. DNA-qPCR is more expensive than RT-qPCR owing to the requirement for determining the breakpoint sequence and obtaining specific primers. However this initial expense can be amortised if multiple MRD estimations are performed during the course of monitoring –if this is the case then the extra cost/assay is minimal. Disclosures: Morley: Monoquant P/L: Equity Ownership, Research Funding. Branford: Novartis Pharmaceuticals: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Bristol Myers Squibb: Membership on an entity's Board of Directors or advisory committees, Research Funding. Hughes: Novartis Pharmaceuticals: Honoraria, Research Funding; Bristol-Myers Squibb: Honoraria, Research Funding.