Molecular Recurrence-Free Survival (MRecFS) Following Imatinib Vs. Nilotinib in Patients with Chronic Myeloid Leukemia in Chronic Phase (CML-CP): Matched Analysis of Patients in EURO-SKI and Enestfreedom

Background: Deep molecular response is achieved in a higher proportion of patients on nilotinib compared to imatinib. Stable MR^4.5 (BCR-ABL1^IS≤ 0.0032%; IS: measurement on the international scale) was the criterion for discontinuing therapy in an attempt to achieve treatment-free remission (TFR) in the ongoing ENESTfreedom study. However, there are no trials comparing the success rate of TFR among patients discontinuing nilotinib vs imatinib. The majority of patients in the EURO-SKI study were treated with imatinib prior to attempting TFR.

Aims: To compare matched patients from EURO-SKI and ENESTfreedom to assess the proportion of patients maintaining TFR, defined as molecular recurrence-free survival (MRecFS): no loss of major molecular response (MMR; BCR-ABL1^IS ≤ 0.1%) at 12 months after stopping treatment.

Patients and methods: Criteria for enrollment in EURO-SKI (NCT01596114) and ENESTfreedom (NCT01784068) have been described in prior publications. To reduce bias when comparing data from 2 different studies, a 2-step patient selection and matching process was undertaken:

[1] For the EURO-SKI group, patients who received imatinib as frontline treatment were eligible for inclusion. To be consistent with the ENESTfreedom eligibility criteria, only patients in MR^4.5 at the time of treatment stop, without prior interferon treatment, and without switch to a second-line therapy qualified for analysis. The result of MRecFS at 12 months had to be available.

[2] Using the EURO-SKI subpopulation identified in step [1], matched pairs of patients from each trial were selected through 1:1 propensity score matching for the following variables: duration of TKI treatment, age, and sex.

"At 12 months" was defined by the time interval 11.25 to 13.5 months after stopping treatment. Any loss of MMR or death prior to 13.5 months was considered as an event. Patients without an event but without confirmation of MMR at 12 months, or with restart of therapy prior to loss of MMR within 13.5 months were not eligible.

In the resulting matched sample, between the two trials, proportions of MRecFS at 12 months were compared by the odds ratio of the logistic regression model stratified for the matching strata.

Results:

Of 868 patients included in the EURO-SKI study, 308 were identified in step [1] as eligible for propensity score matching with 188 patients who attempted treatment stop in ENESTfreedom. With 13 of 321 (EURO-SKI) and 2 of 190 (ENESTfreedom) excluded for the sole reason of lacking an eligible 12-month result, monitoring was remarkably efficient in both studies. Propensity score matching led to identification of 75 matched pairs. No significant differences in the matching variables were detected.

Among the matched patients from ENESTfreedom and EURO-SKI, respectively, 49% were male in both studies, median age was 57 vs 52 years, and median TKI treatment duration was 48 vs 50 months. At 12 months after treatment stop, the odds ratio for MRecFS was 1.50 (95%-confidence interval (CI): 0.76-2.95, nominal p=0.24). The raw (unstratified) probabilities of MRecFS were 51% (95%-CI: 39-62%, ENESTfreedom) and 41% (95%-CI: 30-53%, EURO-SKI).

Conclusions:

Considering patients without prior use of interferon, with MR^4.5 at the time of treatment stop and matching for treatment duration, age, and sex, the rate of TFR at 12 months was numerically higher after frontline nilotinib. The methods used in this analysis reduce, but do not eliminate, the possible bias inherent in any comparison of results from separate trials which was not prospectively powered, and thus the results must be interpreted with caution. Nevertheless, together with ENESTnd, these data suggest that potentially more patients treated with frontline nilotinib will be candidates for stopping treatment and be successful at achieving TFR.