DNA sequencing-based measurable residual disease (MRD) detection has shown to be clinically relevant in AML. However, the same methodology cannot be applied to fusion gene-driven subtypes of AML such as core-binding factor AML (CBF-AML). A team led by researchers at the University of Toronto evaluated the effectiveness of using DNA and RNA sequencing in MRD detection and in tracking clonal dynamics in CBF-AML. Using RNA-seq, they were able to quantify expression levels of RUNX1-RUNX1T1 and CBFB-MYH11 at diagnosis and their levels of reduction during remission (P < 6.3e−05 and P < 2.2e−13). The level of reduction of RUNX1-RUNX1T1 as measured by RNA-seq and qPCR were highly correlated (R2 = 0.74, P < 5.4e−05). A decision tree analysis, based on 3-log reduction of RUNX1-RUNX1T1 and cKIT-D816mut at diagnosis, stratified RUNX1-RUNX1T1 AML patients into three subgroups. These three subgroups had 2-year overall survival rates at 87%, 74%, and 33% (P < 0.08) and 2-year relapse incidence rates at 13%, 42%, and 67% (P < 0.05). On the other hand, although low residual allelic burden was common, it was not associated with long-term outcome, indicating that mutation clearance alone cannot be interpreted as MRD-negative. Overall, this study demonstrates that the clinical utility of RNA sequencing as a potential tool for MRD monitoring in fusion gene-driven AML such as RUNX1-RUNX1T1 AML.
A prognostic model based on genetic features measured by RNA-sequencing
(a) A prognostic model using clinical and genetic information obtained from RNA-seq. Two variables were selected; 1. 3-log or deeper reduction of RUNX1-RUNX1T1 transcript level and 2. presence of cKIT-D816 mutation at diagnosis. Based these two factors, the algorithm identified three subgroups. (b) Overall survival and (c) cumulative incidence of relapse based on the defined risk groups.