LAUSR

Discontinuation of imatinib treatment after a sustained deep molecular response enables a proportion of people with chronic myeloid leukaemia (CML) to achieve treatment-free remission (TFR). Other patients experience a molecular relapse with rising BCR-ABL1 levels by reverse transcriptase quantitative polymerase chain reaction (Q-PCR), and need to resume treatment to regain a deep molecular response. Since a TFR attempt has become an accepted part of the treatment pathway, physicians need to be able to provide information to their patients in order for them to decide when is the right time to stop. Multiple studies have now explored the question of how much imatinib is enough. Dennis Kim and his co-investigators in the Canadian TRAD study have added additional data and an alternative perspective on how to answer this question. The first part of the TRAD study, involving 131 imatinib-treated patients, was reported in the British Journal of Haematology, showing results that compared favourably with other imatinib discontinuation clinical trials: molecular relapse-free survival was 56 8% at 12 months, compared with 50% at 24 months in both STIM2 and EuroSKI. In the second part of the TRAD study those patients who experienced molecular relapse will resume tyrosine kinase inhibitor (TKI) therapy with dasatinib instead of imatinib with the aim of enabling a successful second TFR attempt in the future. Comparing TFR studies is complicated by varying definitions of eligibility and molecular relapse. Patients entering the TRAD study had a deep molecular response (MR4.5) for two years or more and the trigger to resume TKI was loss of MR4.0 on two occasions or loss of major molecular response (MMR) on a single occasion. These entry criteria are more stringent than in EuroSKI, which required at least 12 months of MR4.0. The cumulative incidence of molecular relapse by the two different definitions used in TRAD was 29 2% and 34 8%, indicating that most people with confirmed loss of MR4.0 will lose MMR. Interestingly, there were no cases of loss of MMR after six months in the TRAD study, unlike in EuroSKI, suggesting that either the deeper response at study entry or the more stringent relapse definition reduces the incidence of late relapses. The median duration of imatinib treatment in the TRAD study was 9 2 years, compared with 7 5 years in EuroSKI. This enabled the authors to calculate the odds of molecular relapse according to duration of imatinib treatment and duration of MR4.0 up to 12 years (Table I). As in EuroSKI, the duration of MR4.0 was a better predictor of TFR than the duration of imatinib treatment. The hazard ratio for molecular relapse for each additional year of MR4.0 was 0 86 in TRAD, and the odds ratio for relapse-free survival was 1 13 in EuroSKI, indicating that the odds of molecular relapse reduce by approximately 13% for each additional year of MR4.0 prior to imatinib discontinuation. The proportion