Comparison of RNA- and DNA-based methods for measurable residual disease analysis in NPM1-mutated acute myeloid leukemia

Introduction: Reverse transcriptase quantitative PCR (RT-qPCR) is considered the method of choice for measurable residual disease (MRD) assessment in NPM1-mutated acute myeloid leukemia (AML). MRD can also be determined with DNA-based methods offering certain advantages. We here compared the DNA-based methods quantitative PCR (qPCR), droplet digital PCR (ddPCR), and targeted deep sequencing (deep seq) with RT-qPCR. Methods: Of 110 follow-up samples from 30 patients with NPM1-mutated AML were analyzed by qPCR, ddPCR, deep seq, and RT-qPCR. To select DNA MRD cutoffs for bone marrow, we performed receiver operating characteristic analyses for each DNA method using prognostically relevant RT-qPCR cutoffs. Results: The DNA-based methods showed strong intermethod correlation, but were less sensitive than RT-qPCR. A bone marrow cutoff at 0.1% leukemic DNA for qPCR or 0.05% variant allele frequency for ddPCR and deep seq offered optimal sensitivity and specificity with respect to 3 log(10) reduction of NPM1 transcripts and/or 2% mutant NPM1/ABL. With these cutoffs, MRD results agreed in 95% (191/201) of the analyses. Although more sensitive, RT-qPCR failed to detect leukemic signals in 10% of samples with detectable leukemic DNA. Conclusion: DNA-based MRD techniques may complement RT-qPCR for assessment of residual leukemia. DNA-based methods offer high positive and negative predictive values with respect to residual leukemic NPM1 transcripts at levels of importance for response to treatment. However, moving to DNA-based MRD methods will miss a proportion of patients with residual leukemic RNA, but on the other hand some MRD samples with detectable leukemic DNA can be devoid of measurable leukemic RNA.