Background Tyrosine kinase inhibitors (TKIs) have significantly improved outcomes in chronic myeloid leukemia (CML), leading to prolonged overall survival (OS). Seminal first-line trials demonstrated the efficacy of various TKIs, typically through head-to-head comparisons with imatinib. However, real-world practice often involves TKI switching due to intolerance, suboptimal response, insurance barriers, and comorbidities, factors not fully captured in clinical trials. To date, no observational study has directly compared ≥ 3 TKIs simultaneously within a single cohort. In this comprehensive single-center experience, we evaluate the efficacy of all approved TKIs across multiple lines of therapy, reflecting real-world complexity.

Methods We conducted a retrospective study of all adults with chronic or accelerated phase CML (1/2007–12/2022) at Cleveland Clinic. Data were collected longitudinally from diagnosis to death or last follow-up, last updated in 3/2025. Patients with any TKD mutations were excluded. Baseline variables included demographics, comorbidities, cytogenetics, and hematologic parameters. Responses were assessed using BCR::ABL RT-qPCR. Outcomes included OS, event-free survival (EFS), and therapy duration per ELN 2020 guidelines. EFS was defined from treatment start to failure to meet milestones (NCCN 2024), relapse, progression to advanced CML, discontinuation (intolerance, financial), or death. Multivariable Cox regression (MVR) adjusted for age, gender, comorbidities, smoking, and CML phase.

Results In the 1st-line, 347 patients received imatinib (n=183, 53%), dasatinib (n=105, 30%), nilotinib (n=51, 15%), or bosutinib (n=8, 2%). Median ages: 52–60 years (P=0.3); most were White (88–100%, P=0.9); high-risk cytogenetics [i.e., complex, +8, +19, +Ph, i(17q), inv/t(3)] were present in 0–2% and accelerated disease was present in 88%-96% (P=0.2). Bone marrow blasts averaged ~1% (P=0.15). Median duration of these drugs before switching was 16 months (mo) for imatinib, 11 for dasatinib, 16 for nilotinib, and 8 for bosutinib (P=0.2). Reasons to stop first-line therapy were intolerance in 32% of patients on imatinib, 47% on dasatinib, 43% on nilotinib, and 38% on bosutinib. With a median follow-up of 7.2 years (range 0.5–16.9), the 5-year OS and hazard ratio (HR) were: imatinib 84% (ref), dasatinib 78% (HR 1.2, 95CI 0.7–2), nilotinib 90% (HR 0.6, 95CI 0.3–1.1), bosutinib 75% (HR 0.9, 95CI 0.1–6.8) (Log-rank P=0.5, MVR P=0.2). The 5-year EFS and HR were: imatinib 23% (ref), dasatinib 20% (HR 1, 95CI 0.8–1.3), nilotinib 41% (HR 0.6, 95CI 0.4–0.9), bosutinib 25% (HR 1.2, 95CI 0.5–2.7) (Log-rank P=0.06, MVR P=0.05).

In the 2nd-line, 182 patients received dasatinib (n=73, 40%), nilotinib (n=59, 32%), imatinib (n=18, 10%), bosutinib (n=16, 8%), ponatinib (n=6, 3%). With a median follow-up of 5.4 years (range 0.2–15.3), the 5-year OS was: dasatinib 88%, imatinib 79%, nilotinib 87%, bosutinib 93%, ponatinib 31% (P=0.04). MVR-OS was not feasible due to few events. The 5-year EFS and HR were: dasatinib 22% (ref), imatinib 33% (HR 0.6, 95CI 0.3–1.1), nilotinib 25% (HR 0.72, 95CI 0.4–1.2), bosutinib 17% (HR 1.3, 95CI 0.7–5.8), ponatinib 0% (HR 2.1, 95CI 0.98–1.01) (P=0.06, MVR P=0.1).

In the 3rd-line, 93 patients received dasatinib (n=30, 32%), bosutinib (n=17, 18%), imatinib (n=15, 16%), nilotinib (n=14, 15%), ponatinib (n=13, 14%), and asciminib (n=4, 4%). With a median follow-up of 3.6 years (range 0–12.6), the 5-year OS and EFS were: imatinib 79% and 29%, dasatinib 78% and 16%, nilotinib 92% and 8.2%, bosutinib 81% and 12%, ponatinib 100% and 7.7% and asciminib 67% and 25% (OS P=0.2, EFS P=0.3).

Conclusion In our real-world study, 2nd-generation TKIs used 1st-line were frequently discontinued due to intolerance, limiting treatment duration. The availability of alternative TKIs may contribute to early discontinuation, resulting in lower EFS than clinical trials. Despite this, CML patients showed high 5-year OS, indicating TKI effectiveness. Among all TKIs, only 1st line nilotinib was associated with significantly improved EFS compared to imatinib; other agents showed no statistically significant differences in outcomes. These discrepancies from clinical trial data likely reflect real-world treatment dynamics. Imatinib's favorable tolerability and affordability supported its sustained efficacy across lines in patients without TKD mutations, as reflected by its high EFS estimates even when used in the 2nd or 3rd line setting.

This content is only available as a PDF.
2025
Sign in via your Institution