Withdrawal syndrome after tyrosine kinase inhibitors discontinuation in patients with chronic myeloid leukemia

Abstract

Withdrawal syndrome (WS) a musculoskeletal pain after discontinuation of tyrosine kinase inhibitors (TKI) in patients with chronic myeloid leukemia (CML) has been described in the treatment-free remission (TFR) studies. The pathophysiological mechanisms and predisposing factors of WS have not been well established. Our aim was to evaluate clinical features and factors associated with WS in the Russian cohort of CML patients who discontinued TKI therapy. WS was evaluated in total of 183 CML patients with chronic phase and sustained deep molecular response (DMR). WS was defined as a musculoskeletal pain newly observed after TKI cessation or as a worsening of previously observed symptoms. DMR loss free survival at 36 months was 49% and 43% in prospective and retrospective groups respectively (p=0.96) with mеdian (Me) time of observation 33 months (range 1136). WS was observed in 49 (27%) patients: grade 12 was in 45 (92%) patients, grade 3 in 4 (8%) patients. Me time to WS occurrence was 2 months (range 17), Ме duration of WS was 5 months (range 135). WS was resolved in 14 of 15 patients with molecular relapse after 13 months of TKI re-initiation and was decreased in 1 patient. WS was completely resolved in 31 of 34 patients who continued remained in TFR and decreased in 3 patients. WS was resolved spontaneously or with nonsteroidal anti-inflammatory drugs in 14 (45%) and 17 (55%) patients accordingly. Older age (p0.0001), longer duration of TKI therapy (p0.0001) and presence of locomotion system diseases (p=0.022) were observed in patients with WS. No WS was observed in pregnant patients (р0.001). Survival without DMR loss at 12 months after TKI stop was 66 and 42% in patients with and without WS accordingly (р=0.095). The rate of WS was 27% that is in a good concordance with the data of the other TFR studies. A longer period of TKI exposure, older age and the history of locomotion system diseases were associated with the development of the WS. We found for the first time that WS was not observed in patients with pregnancy. There was no association of WS development and the rate of molecular relapses.