Abstract 3783


Some recent studies have reported on the outcome of CML pts who discontinued thyrosin kinase inhibitors (TKI) after achieving sustained undetectable bcr-abl transcript level. Most patients who stop TKI have experienced molecular relapse. Most patients respond after resuming TKIs regaining undetectable bcr-abl transcript levels. These series have prospectively planned treatment discontinuation and included only pts that have sustained complete molecular response (CMR) for at least 2 yrs. However, in many instances pts may want to discontinue TKIs not in CMR. Various reasons may lead patients to discontinue TKI treatment unexpectedly, among them severe adverse effects, pregnancy or economic constraints. This single institution experience reflects the heterogeneous nature of pt-driven TKI discontinuation.


To characterize the outcome and profile of CML pts who chose to discontinue TKI therapy in a single center regardless of their initial response to TKI therapy.


We retrospectively analyzed MDACC data on all patients with CML that were treated with TKIs in our institution and discontinued therapy.


A total of 26 patients with CML-CP managed at MDACC discontinued TKI between 2003 and 2012. The total median follow up time since diagnosis was more than 120 months (mos) (range, 45 mos to 304 mos). The median age at diagnosis was 48 yrs (range, 28–73); 15 pts were female. All pts had been diagnosed and treated in chronic phase. Interferon was initial therapy in 11 pts (42%) and 15 pts received TKI as initial therapy (4 received imatinib 400mg/day, 10 imatinib 600–800mg/day, and 1 bosutinib.) Of the 11 pts initially treated with IFN, 7 then received imatinib 400mg and 4 imatinib 800mg upon IFN failure. Pts treated frontline with TKI started therapy within a median of 0.8 mos from diagnosis (range 0 to 4) and those with previous interferon (n=11) after a median of 60 mos from diagnosis (31 to 164 mos). Before TKI discontinuation 21pts (81%) were receiving their first TKI and 5 (19%) were receiving 2nd (4) or 3rd line (1) TKI. Complete cytogenetic response (CCyR) had been achieved in all 26 pts at a median of 3.5 mos (3–93); Major molecular response (MMR) in all at a median of 9 mos (3–73) and CMR in 17 (65%) at a median of 22 mos (9–120). All patients discontinued therapy in CML-CP, all in CCyR, of them, 17 (65%) had undetectable transcripts, 2 (8%) had MR4.5, and 7 (27%) MMR. The median duration of CMR before TKI cessation was 62 mos, (0– 118). The median duration of total TKI therapy was 101 mos (3– 135).

Fourteen pts (54%) discontinued TKI due to adverse events, 2 pts discontinued to become pregnant, 5 decided to stop after long CMR, and 5 pts discontinued for financial reasons. After TKI discontinuation patients were followed for a median of 11 mos (5–131). Among pts with CMR at discontinuation, molecular relapse occurred in 8 (47%) pts at a median of 4 mos (1–11) from discontinuation with median transcript level at relapse of 0.07 (IS) (range, 0.004–2.17). Six pts with initial INF therapy had CMR at time of TKI discontinuation, 50% of them relapsed. Among 7 pts who discontinued therapy in MMR, after a median follow-up from discontinuation of 21.6 months (range, 4.6–106), 4 remained at MMR, one has minor CyR and one CCyR without retreatment at last follow up after 78 and 105 months from TKI discontinuation, and one transformed to accelerated phase (AP). The 2 pts with MR4.5at discontinuation relapsed to MMR. Three pts had a transient molecular recurrence with spontaneous re-gain of CMR. Seven pts with relapse were treated again with TKI, 3 with nilotinib, 2 with dasatinib, and one each with imatinib and bosutinib (the later in AP). After a median of 13 months on therapy (range 4–52) all patients improved their response, 5 with CMR and 2 MMR (including the pt that had transformed to AP). There were no deaths or transformations to blastic phase of CML. At last follow up 14 (54%) pts were in CMR, 5 (19%) in MMR,5 (19%) in CCyR and 1 each in minor CyR and PCyR.


Pt-driven TKI discontinuation in CML-CP leads to molecular relapse in nearly half of the pts who discontinue therapy in CMR. Some pts who discontinue in MMR may have sustained MMR. Treatment discontinuation should be considered experimental and cannot be recommended to pts as a standard approach.


Ravandi:BMS: Honoraria, Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.