Measurable residual disease as a biomarker in acute myeloid leukemia: theoretical and practical considerations.

Several methodologies that rely on the detection of immunophenotypic or molecular abnormalities of the neoplastic cells are now available to quantify measurable ("minimal") residual disease (MRD) in acute myeloid leukemia (AML). Although the perfect MRD test does not (yet) exist, the strong association between MRD and adverse patient outcomes has provided the impetus to use measures of MRD as biomarker in the routine care of AML patients and during clinical trials. MRD test results may inform the selection of postremission therapy in some patients but evidence supporting the use of MRD as predictive biomarker is still limited. Several retrospective studies have shown that conversion from undetectable to detectable MRD or increasing MRD over time is associated with overt disease recurrence, and MRD testing may therefore be valuable as a monitoring biomarker for early detection of relapse. Interpreting serial MRD data is complex, with open questions regarding the optimal timing and frequency of testing, as well as the identification of test-specific thresholds to define relapse. Importantly, it is unknown whether intervening at the time of MRD detection, rather than at overt disease recurrence, improves outcomes. Finally, using MRD as a surrogate efficacy-response biomarker to accelerate drug development/approval has already been accepted by regulatory authorities in other diseases and is of great interest as a potential strategy in AML. While the prognostic value of MRD in AML is well established, data from prospective clinical trials confirming that treatment effects on MRD directly relate to clinical outcomes are needed to further establish the role of MRD as a surrogate endpoint in AML.