Assessment of BCR-ABL1 Transcript Levels By Digital PCR (dPCR) in CML Patients who Achieved a Deep Molecular Response (DMR: MR^4.0, MR^4.5 And MR^5.0) with Tkis May Improve the Detection of Minimal Residual Disease (MRD) and the Selection of Patients for Treatment Free Remission (TFR)

Monitoring of BCR-ABL1 levels by quantitative PCR (qPCR) is essential for the management of CML patients treated with TKIs. Because of intrinsic limits, qPCR does not appear to be an optimal assay to select the best candidates to TKIs discontinuation. Up to 40% of CML patients treated with TKIs can achieve a Deep Molecular Response (DMR), but only 50% maintain a stable Treatment Free Remission (TFR) after therapy discontinuation. Digital PCR (dPCR), giving an absolute quantification of BCR-ABL1, is expected to be more sensitive and accurate than qPCR in the assessment of molecular MRD. dPCR performed on a QuantStudio 3D Digital PCR System (Life Technologies) by using a TaqMan-MGB probes targeting the BCR-ABL1 transcript was used to comparatively analyse 102 CML patients with Major (MR^3.0 = 31 cases) or Deep (MR^4.0= 33 cases; MR^4.5 = 24 cases and MR^5.0 = 14 cases) molecular response. Preliminary results showed that: a) ≥77% of deep responders (MR^4.0, MR^4.5 and MR^5.0) fell under the value of 0.468 BCR-ABL1 copies/▢l indicated by the ROC analysis as the value below which the patients with lower levels of MRD might be dissected (spec.=71%, sens.=77%; AUC=0,79); b) BCR-ABL1 transcript levels were detectable by dPCR also in cases resulted undetectable by qPCR.

In this study, we analysed the BCR-ABL1 transcript levels by dPCR in 207 samples related to 102 CML patients. Fourteen (14%) out of 102 CML patients discontinued the TKIs therapy. Among them, 3 patients (21%) lost DMR and all of them showed dPCR values > 0.468 BCR-ABL1 copies/▢l (previously described as cut-off for a deep MRD), while 11 (79%) maintained a stable DMR and 9 of them (82%) fell under the value of 0.468 BCR-ABL1 copies/▢l. These latter patients stratified in different DMR classes by qPCR and all had undetectable level of BCR-ABL1 by qPCR. In 149 out of 207, qPCR revealed DMR. They were: MR^4.0= 59 samples; MR^4.5= 61 cases; MR^5.0 = 29 cases. One hundred twenty-five (84%) fell under the value of 0.468 BCR-ABL1 copies/▢l. A linear regression analysis in these samples did not show any correlation between BCR-ABL1 copies/▢l detected by qPCR with the ones detected by dPCR (R² < 0.01).

On the basis of our preliminary results, TFR seems to be correlated to the maintenance of dPCR values < 0.468 BCR-ABL1 copies/▢l. The concordance between qPCR and dPCR quantification in patients with DMR and with BCR-ABL1 copies/▢l value < 0.468 was poor.

Patients with dPCR values < 0.468 BCR-ABL1 copies/▢l may have 75% of probability to maintain TFR status. These results suggest that dPCR may be more sensitive to detect the MRD and could be more useful for DMR monitoring and for dissecting the best candidate to discontinuation of therapy with TKIs.