Potential of Digital PCR in CML Calibration

Objectives and background:

This study describes the first attempt to extend the definition of the BCR-ABL1 international scale (IS) with copy number (CN) ratios using digital droplet PCR (ddPCR). Digital droplet PCR allows for the determination of the absolute copy number of BCR-ABL1 as well as reference gene(s) in CML cDNA samples. A conversion factor (CF_ddPCR) between BCR-ABL1_CN/reference gene(s)_CN and %IS may be calculated by combining ddPCR and BCR-ABL1^IS. Such a CF_ddPCR would be static and once agreed upon would enable calibration of the BCR-ABL1 qPCR analysis e.g. using the certified ERM-AD623 BCR-ABL1 calibrator plasmid from the Institute of Reference Material and Measurement, Belgium.

Methods:

Six EUTOS-standardised laboratories supplied cDNA from 14 to 20 BCR-ABL1 positive patient samples (n=100) and their locally obtained BCR-ABL1^IS. Copy numbers of BCR-ABL1, ABL1, BCR and GUSB in the cDNA samples were measured and calculated in Vejle by ddPCR (duplicates) and by qPCR (triplicates). Samples with BCR-ABL1^IS values exceeding 10 %, samples negative for BCR-ABL1 (due to cDNA dilution) or below 0.01 % and samples lacking a ddPCR or qPCR reference gene value due to technical issues were excluded, leaving 70 samples fulfilling the requirements for inclusion in the study. The BCR-ABL1^is range of the 70 samples was 1-10 %IS: 25 samples, 0.1-1 %IS: 35 samples and 0.01-0.1 %IS: 10 samples.

Results:

From the ddPCR analysis (n=70), average reference gene ratios of ABL1:GUSB (1:2.0); ABL1:BCR (1:2.8) and GUSB:BCR (1:1.3) were obtained. Combining the copy numbers obtained by ddPCR of BCR-ABL1 and reference genes with the locally obtained BCR-ABL1^IS values, tentative CF_ddPCR = 70 for BCR-ABL1/ABL1, CF_ddPCR = 140 for BCR-ABL1/GUSB, CF_ddPCR = 200 for BCR-ABL1/BCR and CF_ddPCR = 170 for BCR-ABL1/GUSB^BCR (geometric mean) were calculated. The CF_ddPCR and the copy numbers calculated from the qPCR analysis performed in Vejle were used to calculate a BCR-ABL1^IS for the 4 reference gene combinations of each sample. The concordance between the supplied local BCR-ABL1^IS and BCR-ABL1^IS calculated using the CF_ddPCR was determined as r²= 0.81 for BCR-ABL1/ABL1, r²= 0.61 for BCR-ABL1/GUSB, r²= 0.73 for BCR-ABL1/BCR and r²= 0.71 for BCR-ABL1/GUSB^BCR.

Conclusions:

Digital droplet PCR was used to calculate BCR-ABL1^IS conversion factors based on copy number ratios between BCR-ABL1 and the 3 WHO reference genes ABL1, GUSB, BCR and in addition the geometric mean of BCR and GUSB. Using these CF’s, a high degree of concordance to the BCR-ABL1^IS obtained in 6 European EUTOS-standardised laboratories was obtained, taking into account the variability introduced from differences in reverse transcription efficiencies and the use of different reference gene combinations. Use of a CF_ddPCR could improve BCR-ABL1 analysis precision by enabling local IS calibration when appropriate e.g. when changing batches of critical reagents.