Prediction of sustained remission after tyrosine kinase inhibitor discontinuation with BCR::ABL1 digital PCR in chronic myeloid leukemia patients.

ABSTRACT

Precise and reliable predictive parameters to accurately identify chronic myeloid leukemia (CML) patients who can successfully discontinue their tyrosine kinase inhibitor (TKI) treatment are lacking. One promising parameter is depth of molecular response measured by BCRABL1 digital PCR (dPCR). The aim of this study was to validate a previously described prediction cutoff of 0.0023%IS and to assess the value of dPCR for treatment-free remission (TFR) prediction in relation to other clinical parameters. A droplet-based dPCR assay assessed BCRABL1 %IS prior to TKI discontinuation. The primary endpoint was molecular recurrence (MolR) by 36 months. A total of 186 patients from Canada, Germany, and the Netherlands were included. In patients with a first TKI discontinuation attempt (n = 163), a BCRABL1 dPCR < and ≥0.0023%IS had a MolR probability of 33% and 70%, respectively. Patients treated less than 6 years with a BCRABL1 dPCR <0.0023%IS had a MolR probability of 31%. After correction for treatment duration, both high dPCR value and the use of imatinib (vs. second-generation TKI) were significantly associated with a higher risk of MolR (HR of 3.66, 95%CI 2.06-6.51, p < .001; and 2.85, 95%CI 1.25-6.46, p = .013, respectively). BCRABL1 dPCR was not associated with TFR outcome after second TKI discontinuation, however, with the limitation of a small number of patients analyzed (n = 23). In conclusion, BCRABL1 digital PCR based on the cutoff of 0.0023%IS is a valuable predictive tool to identify CML patients with a high probability of TFR success after first TKI discontinuation, including patients treated for less than 6 years.