Epigenetic modifiers mutated in AML

Conventional and molecular cytogenetics are essential components of risk stratification and therapeutic decision making in the clinical management of patients with acute myeloid leukemia (AML). In addition to structural chromosomal alterations, molecular analysis of mutations have been incorporated into the World Health Organization (WHO) and European Leukemia Net (ELN) AML classification; specifically, mutations in FLT3 , NPM1 , and CEBPA are now broadly accepted in routine clinical practice. In particular, mutations in these three genes allow the stratification of patients with cytogenetically normal AML (CN-AML) into prognostic risk categories, and can be used to guide the use of postremission therapy, including allogeneic stem cell transplantation. Currently, patients with CN-AML with either a CEBPA or a NPM1 mutation without a concurrent FLT3-ITD are classified as having favorable-risk AML in the ELN Classification along with patients with core-binding factor leukemias (t(8;21)(q22;q22), inv(16)(p13.1q22), or t(16;16)(p13.1;q22)). Similarly, adult AML with mutations in NPM1 or CEBPA has been classified as a provisional disease entity in the current WHO classification of AML.

In addition to these three clinically used molecular abnormalities, an increasing number of additional mutations are being identified in patients with myeloid malignancies, including AML. Mutations in a large proportion of these newly discovered genetic abnormalities are in genes that impact the epigenetic regulation of gene expression, including mutations in TET2 , IDH1 , IDH2 , and DNMT3a. A precedence for the clinical importance of mutations in epigenetic modifiers in acute leukemia is already well established with the characterization of chromosomal translocations and partial tandem duplications in MLL1 ( MLL -PTD), the main human histone lysine methyltransferase. The MLL- PTD mutation was the first adverse prognostic molecular marker identified in CN-AML. Two initial correlative studies identified that patients with CN-AML with the MLL -PTD mutation experienced significantly shorter remission compared with wild-type counterparts. Subsequent studies have reported variable effects of the MLL- PTD on outcome, possibly related to the more-intensive consolidation therapy administered to patients in recent studies.

This article reviews the progression of clinical understanding of the clinical importance of mutations in TET2 , IDH1 , IDH2 , DNMT3a in AML and discusses the potential use of analyzing these mutations in refining prognostication and affecting choice of induction or postremission therapy in AML.

Mutations and deletions in TET2 : a potentially important prognosticator in CN-AML

TET2 is member of the Tet family of enzymes, Fe(Ii)-, α-ketoglutarate–dependent enzymes with a previously unknown function, specially the ability to convert 5-methylcytosine to 5-hydroxymethylcytosine, an enzymatic activity that is now believed to be critical in mediating DNA demethylation. Somatic mutations in TET2 were first identified in 2009 and occur throughout the coding region of TET2 as deletion mutations, missense, nonsense, and frameshift mutations in patients with myeloproliferative neoplasms (MPN) and myelodysplastic syndromes (MDS). The presence of deletions, nonsense, and frameshift mutations in TET2 provided genetic data suggesting that TET2 functions as a tumor suppressor in myeloid malignancies.

Shortly after the initial description of TET2 mutations, the authors reported that TET2 was mutated in 10.1% of patients with AML and that TET2 mutations were associated with significantly decreased overall survival in a single-institution cohort of 119 patients with AML (including de novo and secondary AML). Since then, a series of studies have been performed to identify the clinical correlates and potential prognostic importance of TET2 mutations in AML in more detail ( Table 1 ). The first study to focus exclusively on the clinical effect of TET2 mutations on prognosis in de novo AML was performed by Nibourel and colleagues. This study found a trend toward a higher frequency of TET2 mutations among patients with AML who did not experience a complete remission compared with those who did (27% vs 17%, respectively; P = not significant). However, the authors found that TET2 mutations had no effect on disease-free or overall survival. Unfortunately, the survival analyses were restricted to only 54 patients, all of whom experienced a complete remission, likely masking any effect of TET2 mutations on outcome.

Four larger subsequent studies evaluating the clinical correlates of TET2 mutations in de novo AML have been reported. The first, presented in abstract form at the 2010 American Society of Hematology Annual Meeting, evaluated the effect of TET2 mutations in 783 patients who were prospectively treated in the AML HD98A Study of the German AML Study Group. Although TET2 mutations were found in 7.6% of patients (60 patients), the investigators did not observe a significant impact of TET2 mutations on clinical outcome in the entire cohort or the subset of patients with CN-AML (this included analysis of overall, event-free, and relapse-free survival). When interpreting the results of this study, one must consider that all patients analyzed were treated using an intense response-adapted double-induction and first consolidation therapy, which is distinct from the commonly used standard induction chemotherapy regimens consisting of an anthracycline plus cytarabine, which are standard of care in the United States, United Kingdom, and elsewhere.

Very recently, Chou and colleagues studied the effect of TET2 mutations on a cohort of 486 patients with de novo AML treated with standard induction chemotherapy. In this study, TET2 mutations were not associated with differences in overall survival, complete remission, or disease-free survival in the entire cohort. However, TET2 mutations were clearly associated with shorter overall survival in 171 patients with intermediate-risk cytogenetics (median not reached in patients with the TET2– wild-type genotype vs 22.1 months in those with the TET2- mutant genotype; P = .0076). This study also found that patients with CN-AML with FLT3-ITD ⁺ /TET2 -mutant and NPM1 –wild-type/ TET2 -mutant genotypes also had worsened outcome compared with their TET2 –wild-type counterparts.

Metzeler and colleagues specifically investigated the prognostic impact of TET2 mutations in CN-AML. In an analysis of 427 patients with CN-AML who received cytarabine/daunorubicin-based first-line therapy, the authors found that TET2 mutations adversely affected survival in patients within the ELN favorable-risk category of CN-AML (patients who have mutated NPM1 and/or CEBPA without FLT3-ITD mutations). TET2 -mutated patients in this subset had significantly shorter event-free and disease-free survival because of a lower complete remission rate only among patients with favorable-risk CN-AML.

The authors recently completed an analysis of the effect TET2 mutations on 398 patients enrolled on the ECOG E1900 trial. This phase III prospective trial randomized adults younger than 60 years with AML to standard-dose (45 mg/m ² ) or dose-intensified (90 mg/m ² ) daunorubicin as part of induction therapy. Although 10% of the patients analyzed had a TET2 mutation, the adverse effect of TET2 mutations was not seen in the overall cohort. However, a clear association was seen between TET2 mutations and adverse overall survival in the subset of patients with intermediate-risk AML. Moreover, TET2 mutations were found to be associated with worsened overall survival in patients with intermediate-risk AML, regardless of the presence of the FLT3-ITD mutation, suggesting that mutations in TET2 may be an important prognosticator in patients with intermediate-risk AML.

In addition to these studies, which focused mostly on patients with de novo AML (many of whom were genotyped pretreatment), a set of smaller studies focused on the effect of TET2 mutations in patients with secondary AML. The authors noted that TET2 mutations are a frequent event at leukemic transformation of MPNs, occurring in 43% patients for whom paired MPN/secondary AML (sAML) samples were available, suggesting that TET2 mutations are a common event in transformation from MPN to AML. Recently, Kosmider and colleagues performed a retrospective study of the effect of TET2 mutation on clinical outcome in 247 patients with AML from an antecedent MDS (n = 201) or therapy-related AML (n = 46). They found no clear effect of TET2 mutations (19.8% of cohort) on survival in this cohort; however, the different therapies administered were very heterogeneous, ranging from supportive care to induction chemotherapy or allogeneic stem cell transplantation. Future studies of large, clinical trial cohorts of homogenously treated secondary AML are needed to assess the prognostic significance of TET2 mutations in this context.

Despite all of these studies, no clearly consistent clinical or molecular features specific to AML patients with TET2 mutations have been shown, except that TET2 mutations are exclusive of IDH1/2 mutations in AML. This robust finding clearly indicates epistasis between TET2 and IDH1/2 mutations, and led to functional studies showing that 2-hydroxyglutarate, the metabolite produced by neomorphic IDH1/2 mutations, inhibits the function of the TET family of enzymes. At least two studies have also noted a positive correlation between TET2 and NPM1 mutations, but this has not been a consistent finding among all studies. Moreover, Chou and colleagues reported a correlation between TET2 and ASXL1 mutations in AML. However, this finding may be confounded by the inclusion in their analysis of a frequent nucleotide variant in ASXL1 , which does not seem to represent a bona fide somatic mutation.

Given the finding that patients with AML with TET2 mutations are characterized by a DNA hypermethylation, nascent efforts have been made to investigate whether patients with a TET2 -mutated genotype may have differential rates of response to DNA methyltransferase inhibitors (DNMTIs) compared with their wild-type counterparts. A French study of 86 patients with MDS and secondary AML reported that patients with TET2 mutations had a high response rate to 5-azacitidine compared with the TET2 –wild-type group (82% vs 45%, respectively). However, no effect on survival was seen based on mutational status, and the study group was heterogeneous with few additional genetic parameters studied. In contrast, other groups have found that TET2 alterations in a similar cohort of patients may actually predict for decreased responsiveness to demethylating therapies. The small number of patients included in these studies and the limited genetic characterization of the patients in these cohorts must be considered. Larger studies with more comprehensive genetic evaluation will be critical in determining whether mutations in genetic factors suspected to be important in regulating DNA methylation ( TET2 , IDH1/2 , and DNMT3a mutations, among others) affect response to DNMTIs or other epigenetic therapies. Moreover, the questions remain whether TET2 mutations actually result in global hypermethylation of DNA in all diseases in which TET2 mutations exist and, if so, what is the proper methodology to detect this altered methylation. Ko and colleagues found that patients with TET2 mutations with different chronic myeloid malignancies actually had a relative hypomethylation of DNA compared with their wild-type counterparts using Infinium 27K methylation arrays (Illumina, San Diego, CA). These findings will need to be clarified in future studies with more-comprehensive genotyping and multiple methodologies to assay DNA methylation and hydroxymethylation.

Mutations and deletions in TET2 : a potentially important prognosticator in CN-AML

TET2 is member of the Tet family of enzymes, Fe(Ii)-, α-ketoglutarate–dependent enzymes with a previously unknown function, specially the ability to convert 5-methylcytosine to 5-hydroxymethylcytosine, an enzymatic activity that is now believed to be critical in mediating DNA demethylation. Somatic mutations in TET2 were first identified in 2009 and occur throughout the coding region of TET2 as deletion mutations, missense, nonsense, and frameshift mutations in patients with myeloproliferative neoplasms (MPN) and myelodysplastic syndromes (MDS). The presence of deletions, nonsense, and frameshift mutations in TET2 provided genetic data suggesting that TET2 functions as a tumor suppressor in myeloid malignancies.

Shortly after the initial description of TET2 mutations, the authors reported that TET2 was mutated in 10.1% of patients with AML and that TET2 mutations were associated with significantly decreased overall survival in a single-institution cohort of 119 patients with AML (including de novo and secondary AML). Since then, a series of studies have been performed to identify the clinical correlates and potential prognostic importance of TET2 mutations in AML in more detail ( Table 1 ). The first study to focus exclusively on the clinical effect of TET2 mutations on prognosis in de novo AML was performed by Nibourel and colleagues. This study found a trend toward a higher frequency of TET2 mutations among patients with AML who did not experience a complete remission compared with those who did (27% vs 17%, respectively; P = not significant). However, the authors found that TET2 mutations had no effect on disease-free or overall survival. Unfortunately, the survival analyses were restricted to only 54 patients, all of whom experienced a complete remission, likely masking any effect of TET2 mutations on outcome.

Four larger subsequent studies evaluating the clinical correlates of TET2 mutations in de novo AML have been reported. The first, presented in abstract form at the 2010 American Society of Hematology Annual Meeting, evaluated the effect of TET2 mutations in 783 patients who were prospectively treated in the AML HD98A Study of the German AML Study Group. Although TET2 mutations were found in 7.6% of patients (60 patients), the investigators did not observe a significant impact of TET2 mutations on clinical outcome in the entire cohort or the subset of patients with CN-AML (this included analysis of overall, event-free, and relapse-free survival). When interpreting the results of this study, one must consider that all patients analyzed were treated using an intense response-adapted double-induction and first consolidation therapy, which is distinct from the commonly used standard induction chemotherapy regimens consisting of an anthracycline plus cytarabine, which are standard of care in the United States, United Kingdom, and elsewhere.

Very recently, Chou and colleagues studied the effect of TET2 mutations on a cohort of 486 patients with de novo AML treated with standard induction chemotherapy. In this study, TET2 mutations were not associated with differences in overall survival, complete remission, or disease-free survival in the entire cohort. However, TET2 mutations were clearly associated with shorter overall survival in 171 patients with intermediate-risk cytogenetics (median not reached in patients with the TET2– wild-type genotype vs 22.1 months in those with the TET2- mutant genotype; P = .0076). This study also found that patients with CN-AML with FLT3-ITD ⁺ /TET2 -mutant and NPM1 –wild-type/ TET2 -mutant genotypes also had worsened outcome compared with their TET2 –wild-type counterparts.

Metzeler and colleagues specifically investigated the prognostic impact of TET2 mutations in CN-AML. In an analysis of 427 patients with CN-AML who received cytarabine/daunorubicin-based first-line therapy, the authors found that TET2 mutations adversely affected survival in patients within the ELN favorable-risk category of CN-AML (patients who have mutated NPM1 and/or CEBPA without FLT3-ITD mutations). TET2 -mutated patients in this subset had significantly shorter event-free and disease-free survival because of a lower complete remission rate only among patients with favorable-risk CN-AML.

The authors recently completed an analysis of the effect TET2 mutations on 398 patients enrolled on the ECOG E1900 trial. This phase III prospective trial randomized adults younger than 60 years with AML to standard-dose (45 mg/m ² ) or dose-intensified (90 mg/m ² ) daunorubicin as part of induction therapy. Although 10% of the patients analyzed had a TET2 mutation, the adverse effect of TET2 mutations was not seen in the overall cohort. However, a clear association was seen between TET2 mutations and adverse overall survival in the subset of patients with intermediate-risk AML. Moreover, TET2 mutations were found to be associated with worsened overall survival in patients with intermediate-risk AML, regardless of the presence of the FLT3-ITD mutation, suggesting that mutations in TET2 may be an important prognosticator in patients with intermediate-risk AML.

In addition to these studies, which focused mostly on patients with de novo AML (many of whom were genotyped pretreatment), a set of smaller studies focused on the effect of TET2 mutations in patients with secondary AML. The authors noted that TET2 mutations are a frequent event at leukemic transformation of MPNs, occurring in 43% patients for whom paired MPN/secondary AML (sAML) samples were available, suggesting that TET2 mutations are a common event in transformation from MPN to AML. Recently, Kosmider and colleagues performed a retrospective study of the effect of TET2 mutation on clinical outcome in 247 patients with AML from an antecedent MDS (n = 201) or therapy-related AML (n = 46). They found no clear effect of TET2 mutations (19.8% of cohort) on survival in this cohort; however, the different therapies administered were very heterogeneous, ranging from supportive care to induction chemotherapy or allogeneic stem cell transplantation. Future studies of large, clinical trial cohorts of homogenously treated secondary AML are needed to assess the prognostic significance of TET2 mutations in this context.

Despite all of these studies, no clearly consistent clinical or molecular features specific to AML patients with TET2 mutations have been shown, except that TET2 mutations are exclusive of IDH1/2 mutations in AML. This robust finding clearly indicates epistasis between TET2 and IDH1/2 mutations, and led to functional studies showing that 2-hydroxyglutarate, the metabolite produced by neomorphic IDH1/2 mutations, inhibits the function of the TET family of enzymes. At least two studies have also noted a positive correlation between TET2 and NPM1 mutations, but this has not been a consistent finding among all studies. Moreover, Chou and colleagues reported a correlation between TET2 and ASXL1 mutations in AML. However, this finding may be confounded by the inclusion in their analysis of a frequent nucleotide variant in ASXL1 , which does not seem to represent a bona fide somatic mutation.

Given the finding that patients with AML with TET2 mutations are characterized by a DNA hypermethylation, nascent efforts have been made to investigate whether patients with a TET2 -mutated genotype may have differential rates of response to DNA methyltransferase inhibitors (DNMTIs) compared with their wild-type counterparts. A French study of 86 patients with MDS and secondary AML reported that patients with TET2 mutations had a high response rate to 5-azacitidine compared with the TET2 –wild-type group (82% vs 45%, respectively). However, no effect on survival was seen based on mutational status, and the study group was heterogeneous with few additional genetic parameters studied. In contrast, other groups have found that TET2 alterations in a similar cohort of patients may actually predict for decreased responsiveness to demethylating therapies. The small number of patients included in these studies and the limited genetic characterization of the patients in these cohorts must be considered. Larger studies with more comprehensive genetic evaluation will be critical in determining whether mutations in genetic factors suspected to be important in regulating DNA methylation ( TET2 , IDH1/2 , and DNMT3a mutations, among others) affect response to DNMTIs or other epigenetic therapies. Moreover, the questions remain whether TET2 mutations actually result in global hypermethylation of DNA in all diseases in which TET2 mutations exist and, if so, what is the proper methodology to detect this altered methylation. Ko and colleagues found that patients with TET2 mutations with different chronic myeloid malignancies actually had a relative hypomethylation of DNA compared with their wild-type counterparts using Infinium 27K methylation arrays (Illumina, San Diego, CA). These findings will need to be clarified in future studies with more-comprehensive genotyping and multiple methodologies to assay DNA methylation and hydroxymethylation.