Reaching measurable residual disease (MRD) negativity in acute myeloid leukemia (AML) was associated with improved survival outcomes and could be used as an endpoint to speed up trials of novel therapies, according to findings from a systematic review and meta-analysis.
In more than 11,000 patients across over 80 studies, those who achieved MRD-negative status following treatment had superior disease-free survival (DFS) compared to those who were MRD-positive (average HR 0.37, 95% bayesian credible interval [CrI] 0.34-0.40), and better overall survival (OS) as well (average HR 0.36, 95% CrI 0.33-0.39), reported Farhad Ravandi, MD, of MD Anderson Cancer Center in Houston, and colleagues.
Writing in JAMA Oncology, the team showed that the MRD-negative group had an estimated 5-year DFS of 64%, as compared to 25% for the MRD-positive group. Estimated 5-year OS was 68% versus 34%, respectively.
The association “was observed across ages, disease subtypes, time of assessment, specimen source, and most MRD detection methods,” the researchers wrote, with the exception being cytogenetics/fluorescence in situ hybridization (though only two studies used this method).
“Assessment of MRD in AML in cytomorphologic remission provides important prognostic information,” Ravandi and co-authors argued. “Given the robustness of the association of MRD with long-term outcomes across studies, use of MRD status as an eligibility criterion and/or an endpoint in clinical trial design could lead to more efficient assessment of the efficacy of new drugs and combination therapies in AML.”
Comparing the MRD-negative versus MRD-positive groups, the 5-year restricted mean survival time difference for OS was 15.37 months (95% CrI 13.58-17.19), and was 19.61 months for DFS (95% CrI 17.33-21.92).
“Given that the standard of care therapy for AML of ‘3+7’ has not changed in over 40 years and yet cures only a minority of adults with AML, more rapid drug development of therapies for AML is needed,” wrote Deepa Jeyakumar, MD, and Susan O’Brien, MD, both of the University of California-Irvine in Orange, in an accompanying editorial. “A lesson learned from the RATIFY trial is that it took over a decade from study conception to U.S. Food and Drug Administration approval of midostaurin based on using overall survival as a primary endpoint.”
Jeyakumar and O’Brien noted that the “suboptimal outcomes” for patients who fail to achieve MRD negativity support novel treatment approaches, with hypomethylating agents having shown potential benefit in clinical trials in this population so far.
MRD status is routinely used to guide treatment in acute lymphoblastic leukemia, Ravandi and co-authors noted, but while numerous publications have suggested its clinical value in AML, the optimal use either as a risk stratification tool or for clinical decision-making has been ill-defined, in part due to the heterogeneity of these earlier reports.
To further evaluate the role of MRD status in AML, the group performed a systematic review and meta-analysis that ultimately examined 81 studies and included a total of 11,151 patients, with MRD evaluated using various methods, including next-generation sequencing, polymerase chain reaction (PCR), and multiparameter flow cytometry (MFC).
The DFS analysis included 64 studies while the OS analysis included 60. Most of the studies were in adults (80-82%), examined MRD status after induction (64-66%), and relied on bone marrow (87-89%) as the specimen source. MFC was the most frequently used method for detecting MRD status (40-44%), followed by PCR gene/fusion (31-35%) and PCR for Wilms tumor 1 (11-12%). Most patients had the non-core binding factor (CBF) AML subtype (81-85%).
Limitations cited by the authors included the fact that they were relying on pooled rather than patient-level data — the non-CBF group is a heterogeneous population, for example, with various subgroups “in which MRD status could have variable effects on long-term outcomes.”
The authors were also unable to assess how MRD information was used by treating physicians, as this could have led to imbalances between groups. Finally the researchers highlighted that most of the assessments involved patients on cytarabine- and anthracycline-based induction chemotherapy, so the findings may be less generalizable to less-intensive therapies.
The study was supported by grants from MD Anderson Cancer Center, the K12 Paul Calabresi Clinical Oncology Scholar Award, the American Society of Hematology Junior Faculty Scholar Award in Clinical Research, and by the Intramural Research Program of the National Heart, Lung, and Blood Institute of the National Institutes of Health.
Ravandi had no disclosures. Co-authors reported financial relationships with Jazz Pharmaceuticals, Sellas, and Merck; and one co-author disclosed being the co-owner of a company that designs bayesian and adaptive clinical trials.
Jeyakumar and O’Brien reported no conflicts of interest.