Standardization of Two Dpcr Platforms for Detection of BCR/ABL1 - Minimal Residual Disease (MRD) in Ph+ Chronic Myeloid Leukemia (CML)

Background: despite the international efforts on standardization, qPCR has some intrinsic limitations of accuracy and sensitivity that make it suboptimal to assess lower levels of BCR-ABL1. In recent years, digital PCR (dPCR) has emerged to provide a more sensitive and accurate detection of minimal residual disease (MRD), that accounts for the increasing interest for its use in the clinic. The Qx100/Qx200 Droplet Digital PCR System (ddPCR, Biorad) and Quant Studio 3D dPCR System (Thermofisher) are the most widespread platforms commonly used. However nowadays internationally recognized operational protocols are not available.

Aim: Aim of our study was: to assess the consistency of the results between two different laboratories that relied on two different platforms; to verify and validate the possible existence of a conversion factor (CF) between the results obtained in the two different laboratories.

Material and methods: 16 RNA pools were prepared at different levels of disease (4 pools for 10% level, 4 pools for 1% level, 4 pools for 0.1% level, 4 pools for 0.01% level). dPCR analysis were performed and repeated in 4 separate sessions (hereafter, Exp1, Exp2, Exp3 and Exp4) from the 2 involved laboratories. ddPCR analysis were performed in triplicate for BCR-ABL1 using 200 ng of cDNA for each replicate and in duplicate for ABL1 using 100 ng of cDNA for each replicate; the QuantStudio 3D dPCR analysis were performed in duplicate for both genes using 50 ng of cDNA for each replicate of BCR-ABL1 and 25 ng of cDNA for each replicate of ABL1.

Results: In order to assess the consistency of the results between the two different labs we first analyzed multiple replicates of 8/16 samples (2 samples of each disease level were selected) - Exp 1: 10 replicates by ddPCR and 6 replicates by Quant Studio 3D dPCR were performed. Results of each single replicate were expressed as BCR-ABL1/ABL1%. We assessed the compatibility between the values obtained in the two laboratories by linear regression and we obtained R2 = 0.9869. In order to confirm the compatibility we compared the measures through the Bland-Altman bias-plot and we carried out the Mann-Whitney nonparametric test. The results of these 2 tests confirmed the compatibility of the different measures thus we computed the conversion factor (CF) value as the antilog of the average of the differences and obtained a CF value of 1.41.[IC 95% 1.3597- 1.4661]. The mean of the differences before the conversion was 0.1498 (sd 0.1723) and after conversion became zero. The obtained CF value was then validated on different series of experiments, i.e. Exp2, Exp3, Exp4 and it proved to be very satisfactory. For each experiment we used the fold difference to compare the results before and after conversion. We also computed the 95% confidence interval (CI) for the average difference before and after conversion. We saw that the average differences after conversion were closer to 1 with a narrowing of the width of the data distribution. Finally we computed the percentage of measures before and after conversion which were included in a 2-fold range, in a 3-fold range and in a 5-fold range. We observed that the conversion led to a general improvement of the percentage of measures included in the 2-fold-range (73% in Exp2; 88% in Exp 3; 95% in Exp 4).

Conclusion: In our study we showed that it is possible to obtainequivalent results from different dPCR platforms by the introduction of a conversion factor. This consistency of the data is mandatory in order to introduce this new technology beside qPCR in CML MRD monitoring.