Discontinuation of tyrosine kinase inhibitors based on BCR-ABL1 monitoring by digital droplet PCR in pediatric chronic myeloid leukemia.

Abstract

Lifelong treatment of pediatric chronic myeloid leukemia (CML) patients with tyrosine kinase inhibitors (TKIs) can affect their growth and development. For these reasons, clinical trials have explored the feasibility of TKI discontinuation in children with a sufficient TKI response. We evaluated the analytical performance of digital droplet PCR (ddPCR) to quantify BCR-ABL1 and compared the results with reverse transcription quantitative polymerase chain reaction (RT-qPCR). We further investigated whether ddPCR could be used to determine TKI discontinuation in a clinical setting. Performance of ddPCR was evaluated using standard materials for BCR-ABL1, and a total of 197 clinical samples from 45 pediatric CML patients was included for comparison with RT-qPCR. ddPCR showed excellent analytical sensitivity with 0.001% international scale (IS) and linearity with R2 > 0.99 in log scale. BCR-ABL1 % IS results correlated well with those of RT-qPCR (R2 = 0.9435), however, they showed a moderate strength for agreement with a Cohen's kappa of 0.41 due to higher sensitivity of ddPCR. Among 45 pediatric CML patients, 42 were treated with first-line TKIs including imatinib (n = 27, 64%) and dasatinib (n = 12, 29%), and three patients that were started with imatinib were switched to dasatinib. When we evaluated whether follow-up samples fulfilled ABL1 copies ≥ 10,000 required for deep molecular response (DMR), all samples were acceptable by ddPCR, whereas 18% by RT-qPCR did not reached acceptable ABL1 copies. Moreover, 52 and 13% reached ABL1 copies ≥ 32,000 required for MR4.5 by ddPCR and RT-qPCR, respectively. Seven patients discontinued TKI and the median TKI treatment duration was 73 months prior to discontinuation. Prior to discontinuation, the median duration of sustained undetected BCR-ABL1 was 60 months. Two patients experienced loss of major MR (MMR) during follow-up and restarted dasatinib 5 months after discontinuation. They achieved MMR again and maintained better than DMR afterward. Results from those patients demonstrated that RT-qPCR did not match the need for adequate ABL1 copies for MR4.5 while majority of ddPCR could. Therefore, ddPCR was technically more acceptable to decide and monitor pediatric CML patients before and after TKI discontinuation.