BACKGROUNDDespite the molecular heterogeneity of standard-risk acute myeloid leukemia (AML), treatment decisions are based on a limited number of molecular genetic markers and morphology-based assessment of remission. Sensitive detection of a leukemia-specific marker (e.g., a mutation in the gene encoding nucleophosmin [NPM1]) could improve prognostication by identifying submicroscopic disease during remission. METHODSWe used a reverse-transcriptase quantitative polymerase-chain-reaction assay to detect minimal residual disease in 2569 samples obtained from 346 patients with NPM1-mutated AML who had undergone intensive treatment in the National Cancer Research Institute AML17 trial. We used a custom 51-gene panel to perform targeted sequencing of 223 samples obtained at the time of diagnosis and 49 samples obtained at the time of relapse. Mutations associated with preleukemic clones were tracked by means of digital polymerase chain reaction. RESULTSMolecular profiling highlighted the complexity of NPM1-mutated AML, with segregation of patients into more than 150 subgroups, thus precluding reliable outcome prediction. The determination of minimal-residual-disease status was more informative. Persistence of NPM1-mutated transcripts in blood was present in 15% of the patients after the second chemotherapy cycle and was associated with a greater risk of relapse after 3 years of follow-up than was an absence of such transcripts (82% vs. 30%; hazard ratio, 4.80; 95% confidence interval [CI], 2.95 to 7.80; P<0.001) and a lower rate of survival (24% vs. 75%; hazard ratio for death, 4.38; 95% CI, 2.57 to 7.47; P<0.001). The presence of minimal residual disease was the only independent prognostic factor for death in multivariate analysis (hazard ratio, 4.84; 95% CI, 2.57 to 9.15; P<0.001). These results were validated in an independent cohort. On sequential monitoring of minimal residual disease, relapse was reliably predicted by a rising level of NPM1-mutated transcripts. Although mutations associated with preleukemic clones remained detectable during ongoing remission after chemotherapy, NPM1 mutations were detected in 69 of 70 patients at the time of relapse and provided a better marker of disease status. CONCLUSIONSThe presence of minimal residual disease, as determined by quantitation of NPM1-mutated transcripts, provided powerful prognostic information independent of other risk factors.

Rigorous sequential RQ-PCR monitoring provides the strongest predictor of RFS in APL and, when coupled with pre-emptive therapy, provides a valid strategy to reduce rates of clinical relapse. This provides a model for development of a more individualized approach to management of other molecularly defined subtypes of acute leukemia.