Cytogenetic and Molecular Failure at 12 Months will be the Optimal Time Point for BCR-ABL1 Tyrosine Kinase Domain Mutation Analysis in Patients with Chronic Myeloid Leukemia: The Analysis Based on 2013 European LeukemiaNet Recommendation

Abstract

Context BCR-ABL1 kinase domain mutations are closely related to tyrosine kinase inhibitor (TKI) resistance in chronic myeloid leukemia (CML). Although mutation analysis has been recommended in CML patients with treatment failure, there is no standard guideline according to landmark responses at specific timepoints of European LeukemiaNet (ELN) recommendation. Objective To determine appropriate timepoints for mutation analysis, the frequency and type of BCR-ABL1 kinase domain mutation were analyzed using Sanger sequencing and were assessed by achieving landmark responses at specific timepoints. Study Patients Nine hundred sixty one peripheral blood samples from 605 newly diagnosed chronic phase CML patients were analyzed for BCR-ABL1 KD mutation and all of patients were treated with frontline imatinib for at least 3 months. The patients with atypical transcripts were excluded. Hematologic, cytogenetic, and molecular responses on 3, 6, and 12 months were also assessed by ELN criteria. Results Of the 605 patients, BCR-ABL1 KD mutations were detected in 28 patients (4.6%) and a total of 33 mutations were detected. Of the 33 mutations, 23 mutations (69.7%) were highly resistant mutations of T315I and P-loop. In cytogenetic response criteria, the mutation frequencies of optimal, warning and failure group were 0.7% (5/671 samples), 1.8% (2/110 samples) and 16.0% (17/106 samples) respectively and were 0.7% (3/425 samples), 3.6% (13/359 samples) and 7.6% (13/172 samples) respectively in molecular response criteria. Under 18 different response criteria of ELN recommendation, there was a higher incidence of mutations among patients with 12-month cytogenetic treatment failure (21.8%; 12/55 patients), 3-month cytogenetic treatment failure (20.0%; 3/15 patients) and 12-month molecular treatment failure (9.3%; 10/107 patients). Interestingly, mutations were also detected in 5 optimal cytogenetic responders (2/280 patients at 6-month criteria and 3/241 patients at 12-month criteria: 1 G250E, 2 Y253H, 1 V280M and 1 M351T). Conclusions We conclude that some patients with treatment failure, especially with cytogenetic criteria, should be warranted for mutation analysis. In addition, the majority of patients with warning criteria may be enough only with a close monitoring without routine mutation analysis. However, as a few patients with optimal response had mutation, mutation analysis should not be totally excluded.