Clin Lymphoma Myeloma Leuk. 2026 Apr 14:S2152-2650(26)00101-1. doi: 10.1016/j.clml.2026.04.001. Online ahead of print.
ABSTRACT
BACKGROUND: Optimal timing for switching tyrosine kinase inhibitors (TKIs) in chronic myeloid leukemia (CML) remains uncertain. While early molecular milestones inform treatment decisions, real-world evidence evaluating how the timing of TKI switch influences survival and toxicity is limited.
METHODS: Using the TriNetX US Collaborative Network, we identified adults with CML treated with first-line imatinib, dasatinib, or nilotinib between 2000 and 2022. Patients were categorized as early switchers (≤3 months), late switchers (>12 months), or monotherapy continuers. Propensity score matching (1:1) was performed for three comparisons per TKI: early switching versus monotherapy, late switching versus monotherapy, and early versus late switching. The primary outcome was overall survival; secondary outcomes included hematologic toxicity, cardiovascular events, metabolic complications, and healthcare utilization.
RESULTS: Among 15,967 patients, 3309 (20.7%) underwent TKI switching (776 early, 2,533 late). Early switching was associated with numerically higher mortality without statistical significance (imatinib HR: 1.34; dasatinib HR: 1.15), but significantly higher mortality than late switching in imatinib-treated patients (HR: 2.38). Late switching showed no consistent survival difference across any TKI; imatinib late switchers had increased stroke (HR: 2.09) and pleural effusion (HR: 1.72). Agent-specific complications included heart failure and pleural effusion with imatinib early switching, and anemia, diabetes, and hyperlipidemia with dasatinib early switching.
CONCLUSION: Timing of TKI switching was associated with outcome differences in CML without consistent survival effects. Early switching in imatinib-treated patients carried higher mortality than late switching, suggesting timing reflects underlying disease risk rather than a direct treatment effect. These findings support individualized decision-making incorporating response trajectory, comorbidities, and cumulative toxicity over fixed timepoint thresholds.
PMID:42108129 | DOI:10.1016/j.clml.2026.04.001