The isocitrate dehydrogenase (IDH1/IDH2) genes are recently identified to be frequently mutated in acute myeloid leukemia (AML) patients. The IDH1 is located in the cytosol and the IDH2 in the mitochondrion and are normally involved in citrate metabolism in the tricarboxylic acid cycle. The mutant IDH1 (R132) and IDH2 (R140 or R172) acquire neomorphic enzymatic activity by converting α-ketoglutarate (α-KG) to an accumulation of 2-hydroxyglutarate (2-HG) and receive epigenetic defects. TET2, a DNA demethylase and α-KG dependent enzyme and essential for demethylation, interrupt the function when the IDH1 or IDH2 genes are mutated and results in hypermethylation and impaired hematopoietic differentiation.


To investigate the frequency and outcome of the acquired IDH1/IDH2 mutations in cytogenetically normal AML (CN-AML) cohort.

Materials and Methods

207 de novo AML patients were genotyped for IDH1 and IDH2 mutations. The analysis was performed by a PCR reaction and the PCR product was purified and used for direct sequencing. Kaplan-Meier survival analysis with log rank test was used to evaluate the impact of different IDH1 and IDH2 genotypes on overall survival (OS) calculated as time from diagnosis until death, date of the latest follow-up, or date of bone marrow transplantation.


IDH1 mutations in exon 4 codon132 were found in 16/207 (7.7%) of the patients and four different amino acid exchanges were detected (R132C, R132H, R132G, R132L). IDH2 mutations in exon 4 in codon 140 and codon 172 were found in 21/207 (10.1%) and 6/207 (2.9%) patients, respectively. Two different amino acid exchanges were identified in codon 140 (R140Q and R140G) and in codon 172 all exchanges were to the same amino acid (R172K). There was a significant impact on OS for patients carrying mutated IDH2 gene in codon 140 compared with patients carrying wildtype IDH2 gene (p=0.009), the mean OS was 3 vs. 21 months. This was seen in the intermediate risk patient group with CN-AML ((IDH2 mutated n=9, IDH2 wildtype n=67). Due to the low frequency with codon 172 mutations (n=6), no independent effect could be detected. Neither in the entire patient group nor subdivided in different risk groups, IDH1 mutations had any significance compared with the wildtype IDH1 patients.


Overall, IDH1 and IDH2 mutations were identified in 20.8% in de novo AML in this cohort. IDH1 and IDH2 mutations were mutually exclusive. A significant impact in overall survival with adverse outcome was seen in patients with IDH2 mutations (R140) in the intermediate risk patient group with CN-AML. Previously, no other studies have reported any impact on overall survival for patients with IDH2 (R140) mutations. Other studies have reported that the accumulation of 2-HG is enough to promote leukemogenesis when IDH1 or IDH2 mutations arise. Further other studies have demonstrated that it is possible to target the 2-HG metabolite and get the AML blasts to differentiate. Thus, AML patients with IDH1 or IDH2 mutations may be a new patient group for risk stratification and new treatment options like upfront allogeneic stem cells transplantation.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.