The Evaluation of Residual Disease By Digital PCR, and TKI Duration Are Critical Predictive Factors for Molecular Recurrence after for Stopping Imatinib First-Line in Chronic Phase CML Patients: Results of the STIM2 Study.

Background

TKI discontinuation is now a critical goal of CML management and several studies have demonstrated the feasibility of stopping safely imatinib (IM). A sustained deep molecular response (DMR) on long-term TKI therapy seems critical prior to attempting treatment-free remission (TFR). Results from different studies have been reported recently, but they failed to identify robust and reproducible predictive factors allowing a better selection of candidate patients for successful TFR.

Aims

We prospectively asked whether the use of IM first-line in de novo CP-CML patients and a systematic screen of patients with digital droplet quantitative PCR (ddPCR) for BCR-ABL1 would impact on the TFR rates.

Methods

In a national prospective phase II trial (Eudract # NCT01343173), we enrolled newly diagnosed CP-CML patients, in sustained MR4.5 (≥2 yrs on at least 5 consecutive points) according to the ELN criteria, checked centrally. Further molecular follow-up was performed in local laboratories, all ELN standardized and aligned on the international scale (IS). Molecular recurrence was defined as the positivity of BCR-ABL1/ABL1 transcripts with at least 1-log increase between 2 consecutive assessments or as loss of MMR on a single assessment. Duplex ddPCR was performed on screening samples after RT, using EAC primers and probes. Pre-mixes were transferred in aDG8 cartridge, and generation of droplets was made on the QX200 droplet generator. Each emulsion was transferred into a 96 wells ddPCR plate and amplified on the Biorad-C1000 Touch Thermocycler. Each patient was analyzed in duplicate. ddPCR was considered positive when the BCR-ABL1/ABL1 % was ≥3 SD. BCR-ABL1/ABL1 ratios were aligned on the IS, a conversion factor was calculated for the RT-ddPCR as previously described (S. Branford et al., Blood 2008). Impact of variables on the risk of relapse were assessed through univariate and multivariate analyses using SAS program.

Results

The median follow-up after IM cessation was 23.5 (1-64) Mo. One patient died from unrelated cause, and 107 experienced a loss of MMR. The molecular recurrence-free survival (RFS) was 52% (95%CI: 45-59%) at 6 months, and 50% (95%CI: 43-57%) at 24 months. The Sokal and the EUTOS long-term survival (ELTS) scores failed to discriminate patients but a trend was observed for ELTS score (p=0.0547). Other variables analysed such as gender (p=0.6246), age at diagnosis (p=0.966), age at IM cessation (p=0.6197), interval diagnosis-IM (p=0.9621), interval IM initiation-DMR (p=0.9978) did not impact on the rate of relapse. In contrast, IM and DMR duration before IM cessation significantly impacted on the rate of relapse, as this interval is redundant with IM duration (p=0.008). ddPCR was applied to 174 out of 218 patients at IM cessation. A cut-off of 0.0023%^IS ratio (= median of positive ratios) was also associated with a higher risk of relapse. The RFS estimate according to this cut-off is shown in Figure 1. In a multivariate analysis, the duration of IM (≥74.8 months) and ddPCR (≥0.0023%^IS) were the two identified significant predictive factors with respectively p=0.0366 [HR=0.635, 95%CI: 0.415-0.972] and p=0.0081 [HR=0.635, 95%CI: 0.415-0.972)].

No association was observed between a ddPCR ≥0.0023%^IS and gender, Sokal or ELTS score, age at diagnosis, age at IM discontinuation, duration of IM treatment, DMR duration or the time-lapse for obtaining DMR.

Conclusion

We conclude that the duration of IM and the residual leukemic cell load as determined by a sensitive technique such as ddPCR are key factors for predicting TFR for de novo CP-CML patients who have been treated with IM front-line.

View largeDownload slide

View largeDownload slide

Close modal