During the 2023 European Haematology Association (EHA) congress, novel findings of the EURO-SKI trial shed light on the prognostic factors impacting molecular remission maintenance after tyrosine kinase inhibitor (TKI) treatment discontinuation in patients with chronic myeloid leukaemia.
The goal of discontinuing long-term therapy with tyrosine kinase inhibitors (TKIs) has become feasible for an increasing number of patients with chronic phase chronic myeloid leukaemia (CP-CML). The EURO-SKI trial, conducted across eleven countries, represents one of the largest studies addressing TKI treatment discontinuation in CML patients.1
This new analysis of the EURO-SKI trial aimed at the identification of prognostic factors associated with maintaining molecular relapse-free survival three years after stopping TKI treatment. The trial included patients who had received first-line or second-line (due to toxicity to first-line) TKI treatment for at least three years, had the transcripts e13a2a and/or e14a2, and were in MR4 (BCR:ABL≤ 0.01) for at least one year. TKI treatment was discontinued upon confirming the MR4 status, followed by routine RQ-PCR analyses over three years to assess molecular response.
A total of 510 patients from the EURO-SKI trial were included in the prognostic modelling at 3 years, as well as 184 patients from the STIM 2 trial for validation. At 36 months, 45% and 41% of the STIM 2 and EURO-SKI cohorts achieved a major molecular response (MMR) maintenance rate, respectively. Univariate analysis identified several significant factors associated with a higher probability of having a maintained MMR at 3 years. This includes a longer duration of TKI treatment (OR=1.124), a longer duration of a deep molecular response (DMR) under TKI therapy (OR=1.102), and the presence of the e14a and e13a2 transcript types compared to e13a2 alone (OR=2.064). Additionally, a higher blast ratio at the time of diagnosis was associated with a lower likelihood of maintaining MMR at 3 years (OR=0.889). These findings were confirmed by the STIM 2 trial. Multiple modelling revealed three significant models in the EURO-SKI trial, including the duration of TKI treatment combined with blast ratio at diagnosis (model a) or transcript type (model b). Model c included DMR duration under TKI, time to DMR under TKI, and blasts in peripheral blood. Only the first two models were confirmed by the STIM 2 trial.2
In summary, this analysis identified four prognostic factors for MMR maintenance: duration of TKI treatment, DMR duration under TKI treatment, transcript type, and the number peripheral blasts at diagnosis. These factors were further confirmed in univariate modelling by the STIM 2 trial. In addition to the duration of TKI treatment, transcript type and peripheral blasts at diagnosis were prognostic factors for MMR maintenance also in multiple modelling.2