Patient-driven discontinuation of tyrosine kinase inhibitors: single institution experience.

Abstract With improved outcome for patients with chronic myeloid leukemia (CML) treated with tyrosine kinase inhibitors (TKIs), treatment discontinuation has become increasingly attractive to patients. We analyzed the outcomes of patients who chose to discontinue TKI therapy regardless of their ongoing response. Thirty-five patients with chronic phase CML discontinued TKI in complete cytogenetic response. Of them, 51% discontinued due to adverse effects, 23% due to long complete molecular response (CMR) (> 5 years), 9% due to pregnancy and 17% due to financial problems. After TKI discontinuation, patients were followed for a median of 16 months. Among 27 patients (77%) who discontinued TKIs in CMR, 11 (41%) had a molecular relapse after a median of 3.5 months. In univariate analysis we observed that patients with ≥ 64 months of CMR before TKI discontinuation had superior cumulative proportions of sustained CMR and major molecular response (MMR) at 12 months after discontinuation: 88.9% vs. 45.5% (p = 0.02) and 100% vs. 75% (p = 0.05), respectively. Patients treated with high dose imatinib or second generation TKIs had a higher cumulative proportion of sustained MMR at 12 months after discontinuation than patients treated with standard dose imatinib: 100% vs. 72.2% (p = 0.03), respectively. Of the five patients who stopped TKI in MR(4.5) (molecular response of 4.5-log reduction) one lost cytogenetic response. All three patients who discontinued TKIs in MMR lost cytogenetic response; one progressed to accelerated phase. Thirteen patients (37%) restarted TKIs after loss of response: 11 improved their response, and for two it is too early to assess. Treatment discontinuation can lead to sustained CMR in some patients, but risk of relapse is higher if patients discontinue TKIs when not in CMR.