preheader BJH 1

Discontinuation of tyrosine kinase inhibitor therapy in CML: Final results of EURO-SKI

The EURO-SKI trial investigated the duration of treatment-free molecular response following tyrosine kinase inhibitor (TKI) discontinuation in patients with chronic myeloid leukaemia. Final results of the study demonstrate that 46% of patients continued to experience molecular recurrence-free survival without restarting treatment. TKI treatment duration and duration of deep molecular response were prognostic for major molecular response loss at six months.

With the success of tyrosine kinase inhibitors (TKIs) to treat chronic myeloid leukaemia (CML), the life expectancy of CML patients is now close to that of the general population. In addition, treatment cessation has become a realistic goal for some of these patients. To date, approximately 40-60% of patients with stable deep molecular response (DMR, corresponding to <0.01% BCR-ABL IS) can stop the TKI successfully, e.g. in accordance with recommendations from the European LeukemiaNet (ELN). The main objectives of the EURO-SKI (European Stop TKI) trial were the evaluation of molecular recurrence-free survival (MRecFS) after stopping TKIs in a large Pan-European cohort of CML patients and the definition of prognostic markers to increase the rate of patients in durable DMR after stopping TKI.

In the previously reported interim analysis of the first 200 patients of the EURO-SKI trial, 62% of patients were in major molecular response (MMR: <0.1% BCR-ABL1 IS) at six months. Furthermore, it was found that DMR duration before TKI stop was most predictive for maintenance of MMR. At ASH 2021, Prof. Mahon presented the results of the final analysis of the EURO-SKI trial after 3 years of follow-up.

EURO-SKI study design

EURO-SKI is a prospective, non-randomised trial that enrolled patients with CML at 61 European centres in 11 countries. Eligible patients had chronic-phase CML, had received any TKI for at least 3 years (without treatment failure according to ELN recommendations), and had a confirmed DMR for at least one year (confirmed by three consecutive PCR tests). The primary endpoint was MRecFS, defined by loss of MMR and assessed in all patients with at least one molecular result, after stopping TKI. DMR confirmation was performed in standardised laboratories. According to protocol, a 36-month follow-up was planned. The null hypotheses were that MMR maintenance at 6 and 36 months was less or equal than 40% and less or equal than 35%, respectively.


Of the 728 eligible patients, 46.8% were female. Median age at diagnosis was 52 years (range, 11–85 years). Median duration of TKI treatment was 7.5 years (range, 3.0–14.1 years) and median duration of MR4 before TKI cessation was 4.7 years (range, 1.0–13.3 years). Most patients (93.7%) received imatinib in first-line, while the others received nilotinib (4.5%) and dasatinib (1.8%). Nine patients died without MMR loss (none CML related), 15 patients restarted TKI without prior MMR loss.

At 6 months, data from 713 patients were available. Since 434 patients (61%) remained without relapse during the first 6 months, the null hypothesis of 40% or lower was rejected (p< 0.0001). At 36 months, 678 patients could be analysed; 17 patients had a premature restart of TKI therapy and 33 patients had missing data. Three hundred nine patients were in MMR or better, corresponding to 46%, and the null hypothesis of 35% or less was rejected (p< 0.0001). MRecFS at 36 months resulted in 48% (CI: 44-52%) and molecular recurrence- and treatment-free survival (MRecTFS) in 46% (CI: 43-50%). The cumulative incidence of MMR loss increased over time, with 38%, 44% and 50% of patients at 6, 12 and 36 months, respectively.

Regarding prognostic factors, the investigators could confirm in this final analysis that TKI treatment duration and DMR duration were still the most important factors to predict MMR loss at 6 months. Late MMR loss, i.e., between 6 and 36 months, was observed in 57 patients. In this, TKI treatment duration before therapy stop was the only relevant variable in a preliminary univariate logistic analysis.


The EURO-SKI trial met its first and second primary endpoint. The observed molecular recurrent-free survival probabilities at 6 months and 36 months were 62% and 46%, respectively. For early MMR loss, prognostic analyses support the importance of duration of TKI treatment and duration of DMR prior to therapy stop.  However, for later loss MMR, only TKI treatment duration was of prognostic importance.


Mahon F-X, et al. FINAL Analysis of a PAN European STOP Tyrosine Kinase Inhibitor Trial in Chronic Myeloid Leukemia : The EURO-SKI Study. Presented at ASH 2021; Abstract 633.

Speaker Francois-Xavier Mahon

Francois-Xavier Mahon

Francois-Xavier Mahon, MD, PhD, Institut Bergonié, Bordeaux, France


See: Keyslides


Back to Top