Minimal-Residual Disease Guided Treatment with Azacitidine in MDS/AML Patients at Imminent Risk of Relapse: Results of the Prospective RELAZA2 Trial

Background: Although many patients (pts) with myelodysplastic syndrome (MDS) or acute myeloid leukemia (AML) can achieve a complete remission (CR) by intensive chemotherapy, the majority will eventually relapse even after subsequent allogeneic hematopoietic stem cell transplantation (HSCT). Monitoring of minimal residual disease (MRD) in morphological CR has been shown to allow for predicting subsequent hematological relapse, thus opening a window for pre-emptive therapeutic interventions. We report on the results of the RELAZA2 trial (ClinicalTrials.gov NCT01462578) by the Study Alliance Leukemia (SAL) group evaluating azacitidine (AZA) as treatment of MRD to prevent relapse in patients with MDS or AML in CR after completion of standard treatment.

Methods: From 2011 to 2015, 205 patients with either advanced MDS (n=27) or AML (n=178) in CR after either conventional chemotherapy only (n=58) or consecutive allogeneic HSCT (n=147) have been prospectively studied in 11 centers in Germany. MRD was centrally monitored in bone marrow or peripheral blood by starting on day 56 after completion of the last therapy followed by monthly intervals for a period of 2 years. Markers used for MRD detection confessed either the quantification of NPM1 mutation level (n=77), leukemia-specific fusion genes (DEK-NUP214 n = 1, t(8;21) n = 9, inv16 n = 10) or by a sensitive donor chimerism analysis of CD34(+) peripheral blood cells (n=108) in patients undergoing allogeneic HSCT. Patients with measurable MRD above a threshold defining imminent relapse but still being in CR preemptively received 6 cycles of AZA (75mg/m², s.c. days 1-7). This was followed by an MRD risk-adapted AZA-based therapy for up to 18 additional months. Patients developing a hematological relapse went off study. The primary endpoint of the trial was the rate of hematological relapse after 6 cycles of AZA. We report the confirmatory analysis of the primary endpoint of those 53 patients who entered the treatment phase of the study.

Results: At a median of 100 days (range 3-601) after start of screening, 53 out of 205 (26%) prospectively screened patients became MRD positive as defined by either a decrease of CD34(+) donor chimerism to <80% (n=19) or a detectable mutation level >1% (NPM1 n=31, t(8,21) n=3) while being still in hematological CR. All of these patients started AZA-based preemptive treatment. Six months after start of MRD-guided therapy, 31 out of 53 patients were still in CR (58%, 95%-CI 44-72%, p<0.001, one-sided binomial test for H₀: p_exp≤0.3) while a total of 22 patients (42%) relapsed after median of 3 AZA cycles. In fact, 21 patients (40%) responded with either a decline of MRD below a predefined threshold (increasing CD34(+) donor chimerism to ≥80% or mutation level <1%), while a stabilization in the absence of relapse was achieved in 10 patients (19%). Overall response rate was different between patients with (71%) or without (48%) antecedent allogeneic HSCT (p=0.007). After 6 months, 24 patients (45%) continued to receive a median of 9 (range 1-15) subsequent AZA cycles. Eventually, hematologic relapse occurred in 8 of those patients (33%), but was delayed until a median of 397 days (range 196-584 days) after initial MRD detection. Treatment was well tolerated with infections (n=4) and pneumonia (n=3) being the main severe side effects during the first 6 cycles. With a median follow-up of 13 months after start of MRD-guided preemptive treatment the actual overall and progression free survival rate is 76% and 42%, respectively.

Conclusion: Pre-emptive MRD-guided therapy with AZA can prevent or substantially delay hematologic relapse in patients with MDS and AML at high-risk of relapse. Success of AZA-based MRD therapy seems to be context-dependent emphasizing the potential immunomodulatory effect of hypomethylating agents.

View largeDownload slide

View largeDownload slide

Close modal