Dueling mutations in normal karyotype AML

Comment on Döhner et al, page 3740; Schnittger et al, page 3733; and Verhaak et al, page 3747

Nucleophosmin 1 (NPM1) mutations occur in about 50% of normal karyotype AML patients, associated with increased probability of complete remission and, paradoxically, of FLT3 (Fms-like tyrosine kinase 3) mutations. Three European leukemia groups now report that NPM1 mutations are also associated with increased survival but that this association is largely nullified by FLT3 internal tandom duplication mutations.

By cytogenetic analysis, currently the primary method for prognostic subclassification of acute myeloid leukemia (AML), 45% of cases have a normal karyotype.¹  Although normal karyotype AMLs (nk-AMLs) are classified as having intermediate prognosis, there is marked heterogeneity in outcome. Recently, 2 subgroups of nk-AML were identified by gene-expression analysis based on treatment outcome, but no defining genetic abnormality was discovered.²  Thus, the recent finding by Falini et al³  that more than 50% of nk-AML cases harbor mutations in the NPM1 gene, the highest incidence of any mutation in AML, was a landmark discovery.

NPM1 mutation in nk-AML was an independent prognostic marker for achieving complete remission (CR) in the study by Falini et al, but it was also associated with 2 adverse risk factors, increased white blood cell (WBC) count and internal tandem duplication (ITD) mutations of FLT3.³  Remarkably, only a few months later, 3 European leukemia groups have provided an extensive analysis of NPM1 and FLT3 mutations, as well as other mutations with prognostic significance in AML,¹  in cohorts of several hundred patients. These expeditious reports are a tribute to the extraordinary leukemia group resources and effort required to consummate these integrated clinical-laboratory studies. There are significant differences among the reports. Döhner and colleagues (report D) and Schnittger and colleagues (report S) report exclusively nk-AML cases (n = 300 and 401, respectively), while Verhaak and colleagues (report V) report 116 nk-AML cases among 285 total AML cases with further embedding of nk-AML cases among other cases with intermediate risk on the basis of cytogenetic characteristics. Thus, restricting direct comparison of essential clinical trial features to reports D and S, significant differences included median time of follow-up (46 vs 16 months), median age (∼48 vs 60 years), male-female ratio (0.78:1 vs 1.02:1), treatment regimens, and possibly outcome. However, the essential conclusions related to NPM1 mutations were concordant in all 3 reports. Of most importance, NPM1 mutations were found to predict for improved overall survival (see figure), relapse-free survival (reports D & S; P = .001 and P < .001), and event-free survival (reports S and V; P = .001 and P = .05) in either nk-AML or intermediate-risk AML by Kaplan-Meier analysis only if FLT3/ITD mutations were not coincidentally present (ie, in the NPM1⁺FLT3^– subgroup). There were no survival differences between NPM1⁺FLT3⁺, NPM1^–FLT3⁺, and NPM1^–FLT3^– subgroups. Other concordant conclusions in the directly comparable D and S reports were that NPM1 mutations are not related to age, occur more often in females, are associated with increased WBC count, are more frequently associated with the myelomonocytic phenotype and low CD34 expression, and occur less frequently with coincident partial tandem duplication of MLL. Apparent discrepancies between reports included the findings of a decreased coincidence of NPM1 and CEBPA mutations and an independent survival advantage for NPM1 mutation alone by multivariate analysis in reports S and V (in overall AMLs) but not in report D, as well as a decreased incidence of NPM1 mutations below age 35 in report V.FIG1 

These founding NPM1 mutation reports seem certain to have a major impact on clinical and molecular investigative approaches in nk-AML. Döhner et al found that allogeneic stem cell transplantation did not improve the long-term 60% relapse-free survival rate in their NPM1⁺FLT3^– cases but did so in the other subgroups, suggesting that this high-risk procedure may not be indicated in first CR for this newly defined, favorable-prognosis subgroup. More speculatively, treatment with recently developed Fms-like tyrosine kinase 3 (FLT3) inhibitors⁴  might be efficacious in NPM1⁺FLT3⁺ patients by producing pharmacologic conversion to a more chemotherapy-sensitive NPM1⁺FLT3^– status. The interaction between NPM1 and FLT3 mutations could also provide important clues to key therapeutic targets in nk-AML. Verhaak et al, in agreement with a recently published gene expression study,⁵  found that NPM1⁺ AML has a defined profile, not restricted to nk-AML cases, that is distinguished by up-regulation of several HOX genes and their TALE partner genes and by down-regulation of the CD34 gene, suggesting that NPM1 mutation–associated leukemogenesis involves transformation of a primitive CD34-negative hematopoietic stem cell. However, in neither report did the therapeutically influential FLT3 mutation status have a defining role in the gene expression profile. Perhaps, then, the interactive effect of FLT3/ITD mutations is related to the primary posttranslational target level of aberrantly increased FLT3 tyrosine kinase activity, where it may affect cellular pathways regulating cell growth and survival.⁶  Schnittger et al discuss the possibility that the cytoplasmic displacement of nucleophosmin 1 (NPM1) protein by a multiplicity of frameshift mutations that introduce a common nuclear export signal may increase susceptibility to chemotherapy-induced apoptosis. This suggests that dynamic molecular analyses at both transcriptional and posttranscriptional levels early after exposure to chemotherapeutic agents with or without modulators, such as FLT3 inhibitors, mightshowdifferencesbetweenNPM1⁺FLT3^– and NPM1⁺FLT3⁺ nk-AML cells that could implicate important target molecules. It is hopeful that these example ideas are harbingers of advances that will lead to improvement in long-static treatment outcome in AML. ▪