Discontinuation of Tyrosine Kinase Inhibitor in Children with Chronic Myeloid Leukemia (JPLSG STKI-14 study)

Background: Tyrosine kinase inhibitor (TKI) has now enabled patients with chronic myeloid leukemia (CML) to live a normal lifespan. However, its long-term side effects such as growth impairment are an issue of concern especially for children. Recent clinical trials in adults have suggested that a subset of CML patients with deep molecular response on TKI therapy may have chance to discontinue TKI without molecular relapse (Saussele et al. Lancet Oncol. 2018). However, the biology of CML in children may differ from adults with more aggressive presentation, and data of TKI discontinuation in CML children are limited (Hijiya et al. Blood. 2019; Bruijn et al. British Journal of Haematol. 2019).

Methods: The Japan Pediatric Leukemia/Lymphoma Study Group (JPLSG) CML committee recruited 22 Japanese patients diagnosed with CML-chronic or accelerated phase at < 20 years of age, treated with TKI for ≥ 3 years, and sustained molecular response (MR4.0) for ≥ 2 preceding years. Patients who relapsed after hematopoietic stem cell transplantation (HSCT) were included if total duration of TKI treatment was ≥ 3 years after rejection and relapse, and also met the criteria mentioned above. We prospectively analyzed treatment-free remission rate (TFR) at 12 months. Molecular relapse was defined as at least one loss of major molecular response (MMR), and TKI treatment was restarted as prescribed before discontinuation.

Results: All the patients were diagnosed with CML-chronic phase. Median age at diagnosis of CML was 9 years (range, 1 to 14 years), and median age at discontinuation of TKI was 16 years (range, 5 to 26 years). Initial TKI was imatinib in 21 patients, and imatinib was switched to second generation (2G)-TKI in 2 patients because of intolerance or poor response. One patient was treated only with 2G-TKI. Median treatment duration of TKI before discontinuation was 100 months (range, 42 to 178 months), and median duration of MR4.0 was 53.5 months (range, 25 to 148 months). TFR at 12 months was 50.0% (90% CI=31.7-65.8%). Eleven patients experienced molecular relapse and restarted TKI at median of 102 days (range, 67 to166 days) after discontinuation of TKI (Figure. 1). No progression was observed during study, and all 11 patients reachieved MR4.0 at median of 64 days (range, 25 to 196 days) after restart of TKI. All the patients who lost MR4.0 within 3 months after stopping TKI failed to maintain MMR thereafter and restarted TKI. On the other hand, a single patient who lost MR4.0 after 32 months of stopping TKI reachieved MR4.0 without TKI. Plasma trough concentration level of imatinib was determined just before stopping imatinib in 18 patients. In univariate analysis, plasma concentration level of imatinib was associated with TFR after stopping TKI (p=0·005). Higher plasma concentration of imatinib was associated with higher risk of relapse. In contrast, treatment duration of TKI, duration of MR4.0, Sokal, Hasford, or EUTOS scores had no impact on TFR. Of 3 patients who underwent HSCT, 2 patients succeeded in maintenance of MR4.0 without TKI. In terms of adverse effects, grade 1 musculoskeletal or joint pain observed in 3 patients, and grade 1 to 3 elevation of creatine phosphokinase observed in 6 patients regressed after discontinuation of TKI. No withdrawal syndrome as reported in adults was observed in this study. In 3 males who discontinued TKI before puberty (5, 9, and 13 years of age with testicular size < 3mL), slight recovery in growth associated with increase in both serum bone type alkaline phosphatase and urine N-Telopeptides of Type I collagen was observed soon after discontinuation of TKI. However, these 3 patients relapsed and restarted TKI within 4 months after discontinuation of TKI.

Conclusion: These data indicate that as in adults, TKI may be safely discontinued also in children younger than 15 years of age at diagnosis, treated with TKI for ≥ 3 years, and sustained MR4.0 for ≥ 2 years, including patients who relapsed after HSCT. We suggest that plasma trough concentration level of imatinib may predict TFR for children with CML and sustained MR4.0 with imatinib. Further study of TKI discontinuation in children with CML is expected.

View largeDownload slide

View largeDownload slide

Close modal