Standard tyrosine kinase inhibitor dose versus dose de-escalation prior to treatment-free remission attempt in chronic myeloid leukemia: A single-center 25-year retrospective study

Abstract Background As frontline treatment for CML, tyrosine kinase inhibitors (TKIs) have significantly improved patient prognosis and can even pave the way for treatment-free remission (TFR) for patients achieving deep molecular remissions. Recommended doses for TKIs, however, were derived from phase I clinical trials that did not necessarily consider their chronic indication and potential long-term health and financial impact. At present, dose adjustments are mainly in response to occurrence of overwhelming side effects and deleterious adverse events. In cases where TKIs are tolerated and standard doses are effective, there is currently no reference for systematic approach to dose de-escalation and ultimately TKI discontinuation. This study explores the impact of the use of recommended standard doses versus dose de-escalation of various TKIs on patients who subsequently attempted TFR. Methods This single-center, retrospective study included 131 patients with chronic phase CML who attempted their first TFR from 2000 to 2025. BCR-ABL1 levels were monitored at baseline and at clinically indicated intervals. Specific TKI used prior to first TFR attempt, and corresponding dose history were recorded. The primary goal is to measure time on TFR. Secondary goals include measurement of magnitude of dose de-escalation and time to achieving major molecular response (MMR). Cox proportional hazards regression hazard ratio was used to analyze covariates and predictors. Results This study screened a total of 456 CML patients for eligibility. One hundred thirty-one patients attempted TFR and were included in the final analysis. The median age was 60 y/o (18-33 y/o) with N=68 (51.9%) males and N=63 (48.1%) female, and the median follow-up time was 74 months (95% CI: 61 – 83). Immediately prior to TFR, 68 (51.9%) patients were on imatinib, (22.1%) on nilotinib, 21 (16%) on dasatinib, 8 (6.1%) on ponatinib, 29, 4 (3.1%) bosutinib and 1 (0.8%) on asciminib. The median duration of the last TKI taken prior to TFR was 7.4 years (0.1-22 years), the median time from MMR to TFR attempt was 7.9 years (0.6-21.2 years), and the median age at TFR attempt was 60 years old (18-83 years old). Fifty-one patients (38.9%) remained on recommended standard dose while 80 (61.1%) patients underwent dose de-escalation. The 1-year TFR rate was 82.4% (68.8 - 90.4) and 73.4% (61.9 - 82.0) (HR: 1.34 (0.72– 2.49, p: 0.355) while the 2-year TFR rate was 74.2% (95% CI: 59.7% - 84.1%) and 69.0% (95% CI: 57.0% - 78.2%) (HR: 1.34 (0.72– 2.49, p: 0.354) for the standard dose and dose de-escalation group, respectively. The median magnitude of dose de-escalation was 50% (16.67%-92.83%) of the starting dose for all TKIs and in many cases accompanied by improvement of both side effects and adverse events, such as arthralgia, fatigue, gastrointestinal symptoms, and cardiopulmonary events. As of last follow-up, 88 out of the 131 patients remained on TFR with 52 (59.09%) having undergone dose de-escalation. Of the remaining 43 patients who relapsed and went back to TKI therapy, 20 attempted subsequent TFR: 8 achieved and remained on TFR as of their last follow-up, 6 of which underwent dose de-escalation. Conclusion In CML, BCR-ABL1 TKIs represent the era of targeted therapy while individualized dose de-escalation, a departure from rigid dosing, underscores personalized medicine. The results of this 25-year retrospective study show that TFR rates and duration between patients who received TKI doses recommended based on phase I clinical trials and those who had their doses de-escalated over time did not differ significantly. Moreover, such observation spans different generations of TKIs. Our results showcase the flexibility of dose adjustment and argue for the utilization of minimum effective dose in order to reduce financial burden and proactively prevent occurrence of adverse events affecting patient outcome and quality of life.