Despite the excellent overall survival (OS) of patients with chronic myeloid leukemia (CML), a significant proportion will not achieve optimal response to imatinib or second-generation tyrosine kinase inhibitors (2GTKI). For patients with inadequate response to 2GTKIs, alternative 2GTKIs or ponatinib are widely available treatment options in daily clinical practice. Treatment decisions should be guided by correct identification of the cause of treatment failure and accurate distinction between resistant from intolerant or nonadherence patients. This review aims to provide practical advice on how to select the best treatment option in each clinical scenario.

Background: It is unknown whether Torque Teno virus (TTV) DNA load monitoring could anticipate the development of infectious events in hematological patients undergoing treatment with small molecular targeting agents. We characterized the kinetics of plasma TTV DNA in patients treated with ibrutinib or ruxolitinib and assessed whether TTV DNA load monitoring could predict the occurrence of Cytomegalovirus (CMV) DNAemia or the magnitude of CMV-specific T-cell responses. Methods: Multicenter, retrospective, observational study, recruiting 20 patients treated with ibrutinib and 21 with ruxolitinib. Plasma TTV and CMV DNA loads were quantified by real-time PCR at baseline and days +15, +30, +45, +60, +75, +90, +120, +150, and +180 after treatment inception. Enumeration of CMV-specific IFN-γ-producing CD8 and CD4 T cells in whole blood was performed by flow cytometry. Results: Median TTV DNA load in ibrutinib-treated patients increased significantly ( P=0.025) from baseline (median, 5.76 log copies/ml) to day +120 (median, 7.83 log copies/ml). A moderate inverse correlation (Rho=-0.46; P<0.001) was found between TTV DNA load and absolute lymphocyte count (ALC). In ruxolitinib-treated patients, TTV DNA load quantified at baseline was not significantly different from that measured after treatment inception ( P ≥0.12). TTV DNA loads were not predictive of the subsequent occurrence of CMV DNAemia in either patient group. No correlation was observed between TTV DNA loads and CMV-specific IFN-γ-producing CD8 and CD4 T cell counts in either patient group. Conclusion: The data did not support the hypothesis that TTV DNA load monitoring in hematological patients treated with ibrutinib or ruxolitinib could be useful to predict either the occurrence of CMV DNAemia or the level of CMV-specific reconstitution.