Incidence and prognostic value of TET2 alterations in de novo acute myeloid leukemia achieving complete remission

In acute myeloid leukemia (AML), both cytogenetic and molecular abnormalities are strongly associated with prognosis. In particular, in cytogenetically normal AML, FLT3-ITD (internal tandem duplication) carries adverse prognostic factor, whereas NPM1 or CEBPA mutations are associated with favorable outcome.^1-3 

Recently, mutations of the ten eleven translocation 2 gene (TET2) have been reported in 20% to 26% of myelodysplastic syndromes (MDSs) or secondary AML,^4,5  in 14% of myeloproliferative neoplasms (MPNs),^4,6,7  and in 12% of de novo AML.⁸  We evaluated the frequency and type of TET2 alterations in 111 de novo AML patients who had achieved complete remission (CR) with intensive chemotherapy. The incidence of mutations was compared with that observed in 36 AML patients who failed to achieve CR with intensive chemotherapy. In patients who achieved CR, the presence of TET2 mutation was correlated with other molecular abnormalities and outcomes.

A total of 147 patients 15 to 69 years of age with previously untreated de novo AML from 7 French centers (Avicenne Hospital, Cochin Hospital, Dunkerque Hospital, Lille Hospital, Saint Louis Hospital, Roubaix Hospital, and Versailles Hospital) were analyzed. Among them, 111 had reached CR using anthracycline-cytosine arabinoside induction treatment, followed by high-dose cytosine arabinoside consolidation courses (most of them had been enrolled in the French ALFA 9801 and 9802 cooperative group trials: 28 received an allogeneic bone marrow transplantation in first CR). The remaining 36 patients had failed to achieve CR using the same induction chemotherapy.

In all patients who achieved CR, bone marrow DNA was available both at diagnosis and after CR achievement, so that diagnostic and CR samples could both be analyzed, separating acquired mutations from polymorphisms, whereas in patients who failed to achieve CR, only diagnostic samples were studied, and no germline DNA was available. Blast percentage, after Ficoll density gradient centrifugation, was evaluated by May-Grunewald-Giemsa staining: all diagnosis samples contained more than 80% of blast cells complying with single nucleotide polymorphism (SNP) array analysis sensitivity criteria. Informed consent was obtained from all patients or their parents in accordance with the Declaration of Helsinki. This project was approved by the Institutional Review Board of Centre Hospitalier Regional Universitaire de Lille. Main characteristics of the patients studied are listed in supplemental Table 1 (available on the Blood Web site; see the Supplemental Materials link at the top of the online article). The median follow-up in patients who achieved CR was 2.5 years.

Analysis of TET2 sequence variations was performed as described previously⁴  by direct sequencing of polymerase chain reaction products at diagnosis and, in patients who reached CR, after CR achievement. Frameshift and nonsense variations were all scored as mutations, whereas missense variations were considered as mutations only when observed at diagnostic but absent in the paired sample obtained after CR achievement. Previously identified SNPs were not considered. TET2 anomalies were numbered according to GenBank reference FM992369. TET2 variations were also researched by direct sequencing in 36 DNA samples obtained at diagnosis from patients with refractory AML.

Paired diagnosis and CR genomic DNAs were analyzed using Affymetrix Genome-Wide Human SNP Array Version 6.0 (Affymetrix). Data were analyzed using Gene Chip Genotyping Console Version 3.0.2 (GTC, Affymetrix), DChip (www.biosun1.harvard.edu/complab/dchip), and Partek Genomics Suite (www.partek.com). Array data had been uploaded in the European Genome-Phenom archive at the European Bioinformatics Institute (www.EBI.ac.uk/ega). All microarray data have been deposited in the European Molecular Biology Laboratory-European Bioinformatics Institute under accession no. E-MTAB-278.

Comparisons were made by Fisher exact test for binary variables and Mann-Whitney test for continuous variables. In patients who achieved CR, disease-free survival (DFS) and overall survival (OS) were calculated according to the Kaplan-Meier method, respectively, from the date of CR achievement to the date of relapse or last follow-up, censoring patients alive in first CR and from the day of first diagnosis to the day of death or last follow-up censoring patients alive at last follow-up. Patients were not censored at the allogenic transplantation time. Comparisons regarding DFS and OS were performed with the log-rank test. A P value less than or equal to .05 was considered to indicate statistical significance.

In the 111 patients who had achieved CR, comparison between diagnosis and germline samples allowed us to distinguish between SNP and acquired alterations. We found 11 different SNPs, including 8 SNPs already present in National Center for Biotechnology Information Single Nucleotide Polymorphism database (www.ncbi.nlm.nih.gov/projects/SNP) and the 3 other (L34F, Y867H, and Q1084P) previously reported by Abdel-Wahab et al.⁸  Twenty-five acquired mutations of the TET2 coding sequence were observed in 19 of the 111 (17%), suggesting alteration of the 2 TET2 alleles in 6 patients (Table 1). They included 24 different events: 6 frameshift mutations, 7 nonsense mutations, and 11 missense mutations. Six of the missense mutations were located in conserved regions and 5 outside of those regions. All those mutations were heterozygous and detected in the diagnostic sample but absent in the paired remission sample. Except for 1 missense mutation (S282F) detected in 2 patients, no recurrent TET2 mutation was observed (Table 1).

As observed in previous studies in myeloid malignancies, acquired mutations were spread over all exons (Figure 1A). However, missense mutations targeting amino acids located outside the 2 conserved domains detected in our study were mainly found in the N-terminal part of the protein. Similar results were reported in AML by Abdel-Wahab et al.⁸  Interestingly, the missense mutations seem to be less frequent in MPNs and preferentially located in the C-terminal part of the protein, which could point to different pathogenetic mechanisms. In contrast to previous reports in MDS and MPN,^9,10  no case of uniparental disomy (UPD) was observed in accordance with Gupta et al who reported the presence of UPD in 17% of AMLs¹¹  but very rare UPD on chromosome 4. Only 1 patient (patient 20) presented a small deletion of 60 kb in the TET2 gene locus without TET2 mutation

No significant difference was observed between patients with or without TET2 alterations for most pretreatment characteristics, including gender, age, hemoglobin level, white blood cell count, platelet count, French-American-British subtype distribution, and cytogenetics according to Medical Research Council classification (supplemental Table 1). No significant association was observed between TET2 mutations and FLT3 or CEBPA alterations. On the other hand, TET2 alterations were significantly associated with NPM1 mutations (P = .032; supplemental Table 1). This association between TET2 mutations and NPM1 mutations had not been found by Abdel-Wahab et al⁸  in 91 AML patients. This difference was possibly because of the AML population analyzed. Our study included only de novo AML patients who had all achieved CR to clearly distinguish TET2 mutations from potential polymorphisms, whereas 50% of the patients in the cohort reported by Abdel-Wahab et al⁸  had a secondary AML.

A negative impact of TET2 alterations was reported in MPN but was not in MDS. Contrary to Abdel-Wahab et al⁸  who found an overall negative prognostic impact of TET2 alterations in de novo AML, there was no difference in DFS or OS between patients with and without TET2 alteration in our entire cohort of patients with de novo AML who achieved CR.

We also analyzed the influence of TET2 mutations in patients with normal karyotype (NK)–AML. No significant difference was observed in OS or DFS in NK-AML between patients with and without TET2 mutations (3-year DFS, 0%, 95% confidence interval [CI], 1%-66% vs 57%, 95% CI, 41%-73%, respectively, P = .17; 3-year OS, 51%, 95% CI, 19%-83% vs 54%, 95% CI, 37%-72%, respectively P = .63; Figure 1B-C).

Of the 36 AML patients who failed to achieve CR with intensive chemotherapy, we found 16 mutations in 10 patients (27%): 5 nonsense, 1 frameshift, and 10 missense mutations. Five of these missense mutations were located in the conserved domain and 5 outside. Because of the absence of germline DNA in these last cases, we could not verify the exact frequency of TET2 mutations in refractory de novo AML. However, this maximum incidence of 27% mutations in patients who failed to achieve CR was not different from that observed in patients who achieved CR (P = .2).

In conclusion, in de novo AML, point mutations of TET2 appear to be as frequent as in MDSs or MPNs, whereas TET2 deletion or UPD is very rare. In our study, the incidence of TET2 mutations appeared similar in patients who achieved or did not achieve CR with intensive chemotherapy. In patients who achieved CR, where TET2 mutations could be clearly distinguished from polymorphisms, TET2 mutations were associated with NPM1 mutations. In our study, the presence of TET2 mutations did not affect OS and DFS in the overall population and in NK-AML patients. However, the patient number remained relatively low, and those data will have to be confirmed prospectively on a larger number of patients.

The authors thank all patients, families, and clinicians for their participation.

This work was supported by the Association Laurette Fugain, the Fondation de France (Leukemia Committee), and North-West Canceropole (Onco-Hematology axis), France.

Contribution: O.N. analyzed data and wrote the paper; O.K. and M.F. provided samples and data and contributed to writing the paper; M.C. and C.R. analyzed data and wrote the paper; N.B. performed statistical analysis and wrote the paper; A.R., C.R.-L., and J.-M.C. performed the research and contributed to writing the paper; N.P. and S.G. performed the research; S.Q. and J.S. performed the research, analyzed data, and contributed to writing the paper; H.D., F.D., B.Q., P.F., W.V., and O.A.B. contributed to writing the paper; and C.P. conceptualized the idea, designed research, analyzed data, and wrote the paper.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Claude Preudhomme, Centre Hospitalier Regional Universitaire de Lille, Laboratoire d'hématologie, Bd du Professeur Leclercq, 59037 Lille Cedex, France; e-mail: cpreudhomme@chru-lille.fr.