Acute Leukemias

Comprehensive analysis of genetic alterations and their prognostic impacts in adult acute myeloid leukemia patients

Article metrics

Abstract

To clarify the cooperative roles of recurrently identified mutations and to establish a more precise risk classification system in acute myeloid leukemia (AML), we comprehensively analyzed mutations in 51 genes, as well as cytogenetics and 11 chimeric transcripts, in 197 adult patients with de novo AML who were registered in the Japan Adult Leukemia Study Group AML201 study. We identified a total of 505 mutations in 44 genes, while only five genes, FLT3, NPM1, CEBPA, DNMT3A and KIT, were mutated in more than 10% of the patients. Although several cooperative and exclusive mutation patterns were observed, the accumulated mutation number was higher in cytogenetically normal AML and lower in AML with RUNX1-RUNX1T1 and CBFB-MYH11, indicating a strong potential of these translocations for the initiation of AML. Furthermore, we evaluated the prognostic impacts of each sole mutation and the combinations of mutations and/or cytogenetics, and demonstrated that AML patients could be clearly stratified into five risk groups for overall survival by including the mutation status of DNMT3A, MLL-PTD and TP53 genes in the risk classification system of the European LeukemiaNet. These results indicate that the prognosis of AML could be stratified by the major mutation status in combination with cytogenetics.

Introduction

Acute myeloid leukemia (AML) is a clinically and genetically heterogeneous disease.1, 2 Although about 80% of younger adults with AML achieve complete remission (CR) with induction chemotherapy, more than half of the CR patients relapse, even if they receive intensive consolidation therapies. Allogeneic hematopoietic stem cell transplantation (allo-SCT) is applied to the patients who have risk factors for relapse, and it has been demonstrated by meta-analysis that allo-SCT at the first CR improves the long-term prognosis of the cytogenetically intermediate- and adverse-risk groups.3 Cytogenetic-risk classification for AML is well established and commonly used as criteria for the application of allo-SCT at the first CR, whereas there is clinical heterogeneity in the intermediate-risk group, particularly cytogenetically normal (CN)-AML.4 Recent advances and the accumulation of information on the prognostic relevance of recurrent genetic alterations have made more detailed risk stratification possible in AML patients.5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 The European LeukemiaNet (ELN) has recommended a novel risk classification system on the basis of the cytogenetic and genetic status.2 In this system, CN-AML is stratified into two risk groups according to the mutation status of FLT3, NPM1 and CEBPA: patients with NPM1 mutation but not FLT3-ITD and those with CEBPA mutation are included in the favorable-risk (FR) group, and patients with FLT3-ITD and those with neither NPM1 mutation nor FLT3-ITD are categorized into the intermediate-I-risk (IR-I) group. Long-term prognosis according to the ELN classification system was retrospectively evaluated in well-established cohorts, and it has been demonstrated that the ELN system is useful for further risk stratification of younger adult patients with CN-AML.20, 21 However, it has been reported that another genetic status, such as mutations in epigenetic modifier-encoding genes, could more precisely distinguish the prognosis in each ELN-risk group.5, 22 In addition, the prognostic impacts of recently identified mutations in spliceosome and cohesin complex genes on AML remain unclear.

The Japan Adult Leukemia Study Group (JALSG) conducted six phase III trials for adult de novo AML from 1987 (AML87, AML89, AML92, AML95, AML97 and AML201).23 In the JALSG AML201 study, we prospectively compared a standard dose of idarubicin (IDR) with a higher dose of daunorubicin (HiDNR) in combination with cytarabine (Ara-C) as induction therapy, and three courses of high-dose Ara-C (HiDAC) with four courses of conventional standard-dose multiagents as consolidation therapy in CR patients.24, 25 We demonstrated that HiDNR was equivalent to IDR, as induction therapy and HiDAC was of benefit only to patients with core-binding factor (CBF)-AML as consolidation therapy. Although the CR rate remained at 75–80% during the six JALSG studies, 7-year overall survival (OS) was improved to 48% in the AML201 study from 29% in the AML87 study. Allo-SCT was conducted in only 7.1% of registered patients in the AML87 study, whereas 45.8% of registered patients received allo-SCT not only at the first CR but also after relapse or primary induction failure in the AML201 study, indicating that active application of allo-SCT even after relapse or primary induction failure might contribute to the improvement of OS. These results collectively suggested that a novel risk stratification system for decision making of allo-SCT at the first CR is required.

In this study, we comprehensively analyzed mutations in 51 genes that have been recurrently identified in myeloid neoplasm as well as cytogenetics, and evaluated the association of genetic status with prognostic and clinical features in patients who were registered in the AML201 study.

Patients and methods

Patients and samples

The study population included 197 newly diagnosed de novo AML patients, except for those with acute promyelocytic leukemia, who were registered in the JALSG AML201 study (UMIN Clinical Trials Registry C000000157, http://www.umin.ac.jp/ctrj/). The diagnosis of AML was on the basis of the French–American–British (FAB) classification.26 Median follow-up time was 32.5 months. The age distribution is presented in Table 1. In the AML201 study, patients were randomly assigned to receive either IDR or HiDNR for induction therapy, and those who achieved CR were again randomized to receive either four courses of conventional consolidation therapy or three courses of HiDAC therapy.24, 25 Of the 197 patients, 98 and 99 patients were assigned to IDR and HiDNR arms for induction therapy, respectively. CR was achieved in 161 of 197 (81.7%) patients, and 80 and 77 patients were assigned to HiDAC and conventional consolidation therapies, respectively (Table 1).

Table 1 Characteristics of the 197 patients

High molecular weight DNA and total RNA were extracted from bone marrow samples using standard methods.27, 28, 29

Cytogenetic G-banding analysis was performed by standard methods. We also examined 11 chimeric gene transcripts (Major: BCR-ABL1, Minor: BCR-ABL1, PML-RARA, RUNX1-RUNX1T1, CBFB-MYH11, DEK-NUP214, NUP98-HOXA9, MLL-MLLT1, MLL-MLLT2, MLL-MLLT3and MLL-MLLT4) by reverse transcriptase-mediated quantitative PCR as previously reported.30

Morphological diagnosis, the FAB classification and karyotypes were reviewed and confirmed by the central review committees of the JALSG using the BM samples obtained at diagnosis.

We obtained informed consent from all patients to use their samples for banking and molecular analysis, and approval was obtained from the ethics committees of the participating institutes.

Screening for mutations in 51 genes

A custom-made oligonucleotide probe library was designed to capture the exons of 51 genes that have been recurrently identified in myeloid neoplasm (Supplementary Table 1). Captured and enriched exons were subjected to sequencing on an Illumina HiSeq (Illumina, San Diego, CA, USA).31, 32, 33 Sequence variation annotation was performed using known polymorphism databases, followed by mutation characterization. Each predicted variant sequence was confirmed by Sanger sequencing. Internal tandem duplication of the FLT3 gene (FLT3-ITD) and partial tandem duplication of the MLL gene (MLL-PTD) were examined as previously reported.28, 34

Statistical analysis

Differences in continuous variables were analyzed by the unpaired t-test or the Mann–Whitney U-test for distribution between two groups. Analysis of frequencies was performed using Fisher's exact test for 2 × 2 tables or Pearson's χ2 test for larger tables. A multivariate analysis to identify risk factors for achieving CR was performed by the logistic regression model. Survival probabilities were estimated by the Kaplan–Meier method, and differences in the survival distributions were evaluated using the log-rank test. OS was defined as the time from the date of entry into the AML201 study to death due to any cause or last follow-up. Disease-free survival (DFS) was defined as the time from the day of achieving CR to relapse, death due to any cause or last follow-up. Patients undergoing SCT were not censored at the time of transplantation. The prognostic significance of the clinical variables was assessed using the Cox proportional hazards model. These statistical analyses were performed with Prism 5 (GraphPad Software, La Jolla, CA, USA) and JMP Pro10 (SAS Institute Japan, Tokyo, Japan). For all analyses, the P-values were two-tailed, and a P-value of less than 0.05 was considered statistically significant.

Results

Cytogenetic analysis

Cytogenetic analysis revealed a normal karyotype in 72 (36.5%) patients and an abnormal karyotype in 106 (53.8%) patients, including 41 AML with t(8;21) (q22;q22); RUNX1-RUNX1T1 and 14 AML with inv(16) (p13q22); CBFB-MYH11. However, karyotypes could not be determined in 19 (9.6%) patients because we could not obtain sufficient mitotic cells. On the basis of the G-banding karyotype and chimeric transcript analyses, patients were assigned to favorable- (n=55, 27.9%), intermediate- (n=100, 50.8%) and adverse-risk (n=23, 11.7%) groups according to the refined MRC criteria (Table 1).4

Frequencies of mutations

We identified mutations in 44 of 51 genes analyzed in the 197 AML patients. However, there were only five genes (FLT3, NPM1, CEBPA, DNMT3A and KIT) that were mutated in more than 10% of the patients (Figure 1a and Supplementary Table 1). Each position and type of mutation is presented in Supplementary Figure 1. As germ-line controls were available in a limited number of patients, we could not completely confirm that all identified mutations were somatic mutations. Therefore, there is a possibility that a part of identified mutations might be rare SNPs. FLT3 mutation was the most frequently identified (50 patients, 25.4%), followed by NPM1 (38 patients, 19.2%), DNMT3A (32 patients, 16.2%), CEBPA (31 patients, 15.7%) and KIT mutations (28 patients, 14.2%). Of the 50 patients with FLT3 mutations, 36 (18.3%) and 17 (8.6%) patients harbored FLT3-ITD and FLT3-KDM, respectively, and three patients harbored both mutations. Of the 31 patients with CEBPA mutations, 19 (9.6%) and 12 (6.1%) patients harbored double CEBPA (CEBPA-D) mutations and a single CEBPA (CEBPA-S) mutation, respectively. Of the 28 patients with KIT mutations, 4, 2 and 23 patients harbored mutations in exon 8, exons 10–11 and exon 17 of the KIT gene, respectively, and one patient harbored mutations in both exons 10–11 and exon 17.

Figure 1
figure1

Frequencies and associations of mutated genes. (a) Frequencies of analyzed gene mutations. Frequency of each mutated gene is shown. Mutations were identified in 44 genes of 51 genes analyzed in 197 AML patients. Only five genes (FLT3, NPM1, CEBPA, DNMT3A and KIT) were mutated in more than 10% of the patients. (b) Mutated gene numbers according to the cytogenetics. Mean mutation number±s.d. is indicated by horizontal lines. Mean mutation number per one patient in patients with normal karyotype (3.18±0.16) was significantly higher than in that with aberrant karyotype (2.10±0.15) (P<0.0001). Those in AML with RUNX1-RUNX1T1 (1.68±0.17) and CBFB-MYH11 (1.57±0.20) were significantly lower than that in all samples (P=0.0008 and P=0.0123, respectively). That in AML with MLL-translocation tended to be lower but not statistically significant (2.28±0.57, P=0.5854). (c) Frequencies of mutations according to the conceptual classification. Mutations in Class I, Class II and epigenetic modifying genes were frequently identified. (d) Association of mutated genes according to the conceptual classification. Circos plot of mutated genes according to the function is shown.45 Overlap mutations between Class I, Class II and epigenetic modifying genes mutations were frequently observed. These major mutations were widely coexistent with another family genes, such as cohesin complex, BCOR family and spliceosome genes.

Although mutations in the 51 analyzed genes were not identified in 14 (7.1%) patients, 183 (92.9%) patients harbored one or more mutations; one mutation in 36, two mutations in 56, three mutations in 40, four mutations in 27, five mutations in 17, six mutations in five and seven mutations in two patients. The mean mutation number per patient was 2.56±0.11 in all patients, whereas it was significantly higher in patients with a normal karyotype (3.18±0.16) than in those with an aberrant karyotype (2.10±0.15) (P<0.0001). Furthermore, mean mutation numbers per patient in AML with RUNX1-RUNX1T1 (1.68±0.17) and CBFB-MYH11 (1.57±0.20) were significantly lower than that in all samples (P=0.0008 and 0.0123, respectively) (Figure 1b).

The mean mutation number per patient aged 60–64 years (3.18±0.41) tended to be higher, although there was no significant difference between the mean mutation number and age (Supplementary Figure 2).

Genetic alterations found in AML have been conceptually grouped into class I mutation, which causes constitutive activation of intracellular signals that contribute to the growth and survival, and class II mutation that blocks differentiation and/or enhance self-renewal by altered transcription factors.35, 36, 37 Recently, it has been suggested that mutations that modify the epigenetic status generate a new class because of their overlap mutations both with class I and class II mutations.13, 38 In this study, Class II mutations (NPM1, CEBPA, RUNX1 and GATA2 mutations, and RUNX1-RUNX1T1 and CBFB-MYH11) were the most frequently identified (138/197; 70.1%), followed by Class I mutations (FLT3, KIT, N/KRAS, PTPN11, JAK1/3 and TP53 mutations) (116/197; 58.9%) and mutations that modify the epigenetic status (ASXL1, ATRX, EZH2, TET2, PBRM1, DNMT3A, IDH1/2, KDM6A, MLL and DOT1L mutations) (91/197; 46.2%). Furthermore, mutations of NOTCH family genes (NOTCH1 and NOTCH2), cohesin complex genes (STAG2, SMC1A, SMC3 and RAD21), BCOR family genes (BCOR and BCORL1), NCOR family genes (NCOR1, NCOR2 and DIS3) and spliceosome genes (SF3B1, U2AF1, SRSF2 and ZRSR2) were identified in 19 (9.6%), 22 (11.2%), 17 (8.6%), 22 (11.2%) and 9 patients (4.6%), respectively (Figure 1c).

Association between gene mutations and cytogenetics

The prevalence of each gene mutation differed among the cytogenetic-risk groups. KIT mutations were preferentially identified in the favorable cytogenetic-risk group. FLT3-ITD, NPM1, CEBPA and DNMT3A mutations were preferentially identified in the intermediate-risk group, particularly in patients with a normal karyotype. BCORL1 and TP53 mutations were preferentially identified in the poor-risk group; in particular, TP53 mutations were frequent in patients with a complex karyotype. In addition, PHF6 mutations were also frequently identified in patients with a complex karyotype (Figure 2 and Supplementary Table 2).

Figure 2
figure2

Mutation status according to the cytogetics-risk groups. Identified mutations in analyzed AML patients are shown according to the cytogenetic-risk groups. Pink boxes indicate single mutations and orange boxes indicate double mutations.

Overlap mutations

Several patterns of overlap mutations were identified in this comprehensive mutation analysis (Supplementary Figures 3 and 4). Significantly overlapped mutations were observed between FLT3 mutations and NPM1, DNMT3A and MLL-PTD mutations; NPM1 mutations and DNMT3A, IDH1 and IDH2 mutations; CEBPA mutations and TET2 mutations; ASXL1 mutations and spliceosome gene mutations; DIS3 mutations and MLL mutations; DNMT3A mutations and PTPN11 mutations; GATA2 mutations and CEBPA-D mutations; K/NRAS mutations and WT1 mutations and BCOR/BCORL1 mutations; RUNX1 mutations and U2AF1, MLL-PTD, BCOR/BCORL1 and PHF6 mutations; SF3B1 mutations and NRAS mutations; and TET2 mutations and STAG2 mutations. In contrast, mutually exclusive mutations were observed between FLT3 mutations and KIT, K/NRAS and CEBPA-D mutations; NPM1 mutations and CEBPA-D and RUNX1 mutations; and CEBPA mutations and IDH1/2 mutations.

According to the conceptual classification of the mutated genes, overlap mutations between Class I, Class II and epigenetic modifying gene mutations were frequently observed. However, these major mutations widely coexisted with other family gene mutations, such as the cohesin complex, BCOR family and spliceosome gene mutations (Figure 1d). Although biological functions of mutated genes have not been fully clarified, we also present frequencies and associations of mutated genes according to the provisional gene function in the Supplementary Figure 5.

Association of gene mutations with clinical characteristics

Several associations between mutations and clinical characteristics were observed. DNMT3A mutations and MLL-PTD were more frequently identified in patients over 50 years old than in those less than 50 years old (P=0.0064 and P=0.0121, respectively), whereas the other mutations were not significantly associated with age (Supplementary Table 3).

Several mutations were associated with the white blood cell count at diagnosis. FLT3-ITD, NPM1, DNMT3A and NOTCH1 mutations were significantly associated with the high white blood cell count. In contrast, ASXL1 and IDH1 mutations were associated with a lower white blood cell count (Supplementary Table 4).

Association of gene mutations with the CR rate

We analyzed the association of mutations with the CR rate. By Fisher’s exact test, RUNX1-RUNX1T1 or CBFB-MYH11, KIT, NPM1 and CEBPA-D mutations were identified as favorable factors for achieving CR, and TP53 mutation was an unfavorable factor; however, multivariate logistic regression analysis including all analyzed mutations showed that only NPM1 (Hazard ratio (HR): 96.206, 95% Confidence interval (CI): 2.247–411.9, P=0.0172) and TP53 (HR: 22.222, 95% CI: 1.597–333.3, P=0.0172) mutations were identified as favorable and unfavorable factors for achieving CR, respectively (Table 2 and Supplementary Table 5).

Table 2 Gene mutations affecting the CR achievement

Importantly, KIT mutations were closely associated with RUNX1-RUNX1T1 or CBFB-MYH11, whereas the other mutations that confer the achievement of CR were mutually exclusive (Supplementary Figures 3 and 6). In the patients with RUNX1-RUNX1T1 or CBFB-MYH11, NPM1 and CEBPA-D mutations, the CR rate (106/112; 94.6%) was significantly higher than for those with the other genotypes (55/85; 64.7%) (P<0.0001).

Prognostic impacts of mutations

We next analyzed the prognostic impact of each mutation. By univariate analysis, FLT3-ITD (HR: 1.805, 95% CI: 1.130–2.885, P=0.0135), DNMT3A (HR: 1.696, 95% CI: 1.055–2.725, P=0.0291), TP53 (HR: 15.167, 95% CI: 6.555–35.094, P<0.001), MLL-PTD (HR: 3.782, 95% CI: 1.948–7.346, P<0.001) and RUNX1 (HR: 2.301, 95% CI: 1.278–4.146, P=0.0055) mutations and the karyotypes other than RUNX1-RUNX1T1 or CBFB-MYH11 (HR: 2.786, 95% CI: 1.608–4.831, P=0.0003) were identified as unfavorable prognostic factors for OS (Table 3 and Supplementary Figure 7). Multivariate Cox regression analysis with stepwise selection showed that TP53 (HR: 14.803, 95% CI: 6.259–35.009, P<0.001), MLL-PTD (HR: 2.853, 95% CI: 1.401–5.810, P=0.0039) and RUNX1 (HR: 1.965, 95% CI: 1.054–3.663, P=0.0336) mutations and the karyotypes other than RUNX1-RUNX1T1 or CBFB-MYH11 (HR: 2.353, 95% CI: 1.342–4.132, P=0.0028) were independent poor prognostic factors for OS (Table 3).

Table 3 Unfavorable prognostic factors for overall survival (OS)

In this cohort, mutations of NOTCH family, the cohesin complex, BCOR family and spliceosome genes were frequently identified. NOTCH family and BCOR family genes were not associated with the CR rate, OS and DFS. Although mutations of cohesin complex genes were not associated with the CR rate and DFS, the patients harboring those mutations revealed better OS than those without mutations (P=0.0274) (Figure 3). The CR rate and DFS of patients with spliceosome gene mutations tended to be lower than for those without mutations, although both differences were not statistically significant: the CR was achieved in five of the nine (55.6%) and 156 of the 188 (83.0%) patients (P=0.0601), and 3-year DFS were 0% and 38.9% (P=0.1117) in those with and without mutations, respectively (Supplementary Table 3 and Supplementary Figure 8).

Figure 3
figure3

Prognostic impact of mutations in cohesin complex genes. Kaplan–Meier curves for OS and DFS according to the mutations in cohesin complex genes are shown. (a) OS in the total patients, (b) DFS in the total patient, (c) OS in the patients with normal karyotype, (d) DFS in the patients with normal karyotype. Although mutations of cohesin complex genes were not associated with the CR rate and DFS, the patients harboring those mutations revealed better OS than those without mutations (P=0.0274). In the patients with normal karyotype, OS of the mutated patients tended to be better that that of unmutated patients, though statistical significance was not observed (P=0.1731).

When the patients were stratified into the risk groups recommended by ELN, that is, FR, IR-I, IR-II and AR groups included 92, 35, 42 and 28 patients, respectively. The ELN system well stratified the long-term prognosis of adult AML patients, whereas the OSs of IR-I and AR groups were the same in the present cohort (Supplementary Figure 9). Therefore, we analyzed whether another mutations could further stratify the prognosis in each ELN-risk group. MLL-PTD, GATA2 and TP53 mutations were identified as further poor prognostic factors in IR-I, IR-II and AR groups, respectively. Furthermore, we identified that the DNMT3A mutation was a poor prognostic factor in the FR group except for the AML with t(8;21) (q22;q22); RUNX1-RUNX1T1 or inv(16) (p13q22); CBFB-MYH11 (CBF-AML) (Figure 4).

Figure 4
figure4

Mutations that could further stratify the ELN-risk groups into two risk groups. (a) DNMT3A mutation was a poor prognostic factor in the FR group except for the AML with t(8;21) (q22;q22); RUNX1-RUNX1T1 or inv(16) (p13q22); CBFB-MYH11 (CBF-AML). (b) MLL-PTD was a poor prognostic factor for the OS in the ELN IR-I group. (c) GATA2 mutation was a poor prognostic factor for the OS in the ELN IR-II group. (d) TP53 mutation was a poor prognostic factor for the OS in the ELN AR group.

In addition, we also evaluated two recently reported risk stratification systems on the basis of genetic status in our cohort.5, 39 Patel et al.5 reported a risk stratification system on the basis of cytogenetics and genetic status According to their system, our patients were clearly stratified into three risk groups, although the patients in the intermediate cytogenetic risk with a favorable mutational risk profile and those in the favorable cytogenetic-risk profile showed the same probability of OS (Supplementary Figure 10a). Grossmann et al.39 reported a prognostic model solely on the basis of molecular mutations. Although our cohort did not include AML with PML-RARA, our patients were clearly stratified into four risk groups. However, the patients in the very favorable group and those in the favorable group showed the same probability of OS (Supplementary Figure 10b).

On the other hand, KIT mutations were frequently identified in CBF-AML, while they were not a poor prognostic factor for either OS or DFS (Supplementary Figure 11a). According to the types of KIT mutations, CBF-AML patients harboring mutations in exon 17 of the KIT gene showed worse prognosis than those harboring the other types of KIT mutation, although this was not statistically significant (Supplementary Figure 11b). Notably, in the CBF-AML patients harboring KIT mutations, OS and DFS of those treated with three courses of HiDAC consolidation therapy tended to be better than those treated with four courses of conventional standard-dose multiagent therapy (Supplementary Figure 11c). In CBF-AML, the accumulated mutation number was lower than in the other types of AML, although many kinds of mutation were identified (Supplementary Figure 12). However, we could not identify a gene that affects the prognosis of CBF-AML.

Taking these results together, we tried to modify the genetic criteria for the ELN stratification system. When the CN-AML patients with DNMT3A mutations of the FR group and the patients with MLL-PTD of the IR-I group were included in the IR-I and the AR group, respectively, we could more clearly stratified the patients into four risk groups for OS than the original ELN system (Figures 5a and b). Furthermore, as the prognosis of the patients with TP53 mutations were very unfavorable as previously reported,39 we could more clearly stratify the patients into five risk groups for OS by classifying the TP53-mutated patients as the very adverse-risk group (Figures 5c and d).

Figure 5
figure5

Risk stratification by modifying the ELN stratification system. When the CN-AML patients with DNMT3A mutations of the FR group and the patients with MLL-PTD of the IR-I group were included in the IR-I and the AR group, respectively (a), we could more clearly stratify the patients into four risk groups for OS than the original ELN system (b). When the patients with TP53 mutations were classified as the very adverse-risk group (c), we could more clearly stratify the patients into five risk groups for OS (d).

Discussion

In this study, we comprehensively analyzed mutations of 51 genes by the targeting sequence, and identified a total of 505 mutations in 44 genes in 197 adult de novo AML patients except for APL. The whole-genome and -exome analysis demonstrated recurrent mutations in a total of 260 genes in 200 AML patients, suggesting that another mutations might be accumulated in the presently analyzed AML cells.40 However, frequencies of most mutated genes were reportedly less than 10%. In consistence, only five genes (FLT3, NPM1, CEBPA, DNMT3A and KIT) were mutated in more than 10% patients in our study, and each mutation frequency was almost the same as previous reports.5, 17, 19, 39 The frequency of KIT mutation was relatively higher in our study than previous reports,5, 40 while this is caused by the higher frequency of CBF-AML (28.0%) in the Japanese patients, in which KIT mutations are frequently identified. Our study, therefore, essentially includes major genetic regions, which may affect the pathogenesis and prognosis of AML. However, mutation analyses were not thoroughly performed in all subtypes of AML, such as acute erythroid leukemia and acute megakaryoblastic leukemia because of their lower frequencies. Further analyses are required to fully clarify the genetic alterations in AML.

The whole-genome and -exome analysis demonstrated that an average of mutated genes in coding regions per sample was 5.24.40 Of note was that there were significant differences in mutated gene numbers among the types of cytogenetics and mutations: the mean mutation numbers were higher in AML with RUNX1-RUX1T1, and are lower in that with PML-RARA and MLL translocations than that of all samples. As analyzed gene numbers were limited, mean mutated gene number per sample (2.56±0.11) was low in our study; however, there were different features from the previous report. In our study, higher mutation number was observed in CN-AML (3.18±0.16), and lower was in CBF-AML. Furthermore, the mutation number in AML with MLL-translocation tended to be lower (2.28±0.57, P=0.5854). These results collectively suggested that recurrent cytogenetic abnormalities, such as RUNX1-RUNX1T1, CBFB-MYH11 and MLL-translocation, have a strong potential for the initiation of AML, and that most of accumulated mutations in AML with these cytogenetics might be passenger mutations.

It has been reported that common mutations in AML, such as DNMT3A, NPM1, CEBPA, IDH1/2 and RUNX1, were mutually exclusive of the transcription-factor fusions, indicating the high potential for leukemia initiation.10, 11, 13, 40, 41 Consistently, DNMT3A, NPM1 and CEBPA mutations were not identified in CBF-AML, but frequent in CN-AML. In addition, we identified that MLL-PTD mutation was also exclusive of CBF-AML. In CBF-AML, KIT mutations were preferentially identified, whereas several types of mutations were also accumulated, suggesting that many mutations could act as a driver mutation for the clonal expansion of the initiating clone with RUNX1-RUNX1T1 and CBFB-MYH11. Further study is required to clarify which combination is necessary for the clonal expansion, and whether different combinations cause clinical and phenotypical varieties.

After the completion of genetic alterations in AML, the most important issue is to clarify the prognostic impact of each mutation and/or co-occurring mutations.7, 42 The recently recommended ELN classification system is the first system that includes both cytogenetics and mutation status. Several groups reported that the ELN system clearly stratified the long-term prognosis of AML patients. However, the prognosis of FR groups except for the CBF-AML is still controversial. Paschka et al.43 reported that the IDH1/2 mutation was a poor prognostic factor in CN-AML with mutated NPM1 without FLT3-ITD. In contrast, Patel et al.5 reported that the IDH1/2 mutation was a favorable prognostic factor in AML with mutated NPM1 without FLT3-ITD. Furthermore, it has been reported that the TET2 mutation was an adverse prognostic factor in AML with mutated NPM1 or CEBPA without FLT3-ITD.22 In the present cohort, we could not observe the statistically significant effects of IDH1/2 and TET2 mutations on the prognosis of CN-AML with mutated NPM1 or CEBPA without FLT3-ITD, while we identified that DNMT3A mutation is an adverse prognostic factor in CN-AML with mutated NPM1 or CEBPA without FLT3-ITD. In addition, we could not find the better prognostic impact of the CEBPA double mutations on the FR group recommended by the ELN in contrast to previous reports.39, 44

Although different mutations might further stratify the prognosis of AML with mutated NPM1 or CEBPA without FLT3-ITD, it was noteworthy that all mutations belonged to the class modifying methylation status.13, 38 These results collectively suggested that the epigenetic deregulation might contribute the pathogenesis of AML with mutated NPM1 or CEBPA without FLT3-ITD. Prospective and large-scale study is necessary to clarify what genetic alterations influence the prognosis of AML with these genotypes.

In this study, we demonstrated that the prognosis of adult AML patients could be more clearly stratified by including the DNMT3A and MLL-PTD mutation status than the original ELN system, and that TP53 mutations have a very adverse effect on the prognosis of AML patients. However, as most recurrently identified mutations were observed less than 5% of AML, it is highly expected to refine the genetic-based risk stratification system by much larger-scale studies. In addition, it is also important to evaluate the prognostic effects according to the functions of mutated genes rather than each sole mutation.

In the JALSG AML201 study, patients were randomized to the standard dose of IDR+Ara-C or HiDNR+Ara-C induction therapy, and the CR patients were again randomized to three courses of HiDAC or four courses of conventional standard-dose multiagent consolidation therapy. Therefore, we analyzed whether therapeutic regimens affect the CR rate and long-term survivals according to the mutation status and risk groups on the basis of the genetic status, while we could not observe any significant differences between therapeutic regimens and genetic status. Furthermore, we could not demonstrate that allo-SCT could improve the prognosis of the patients falling in the intermediate- and adverse-risk groups because of the small number of patients who underwent allo-SCT in the first CR in this cohort. It is, therefore, required to evaluate whether therapeutic regimens and allo-SCT affect the prognosis according to the genetic status.

In conclusion, we comprehensively analyzed 51 genes mutations in 197 de novo adult AML patients who were registered to a single prospective clinical study, and demonstrated that cooperative and exclusive mutation patterns and their prognostic impacts. Furthermore, we demonstrated that the prognosis of adult AML patients could be more clearly stratified by including the DNMT3A, MLL-PTD and TP53 mutation status than the original ELN system. However, prognostic impacts of some mutation status are different from the previous reports. We must refine the risk stratification system by considering all known-risk factors in a large-scale and well-established cohort, although molecular genetic status has a strong impact on the prognosis of AML patients. We are now conducting a prospective large-scale study to confirm the present results.

References

  1. 1

    Estey E, Dohner H . Acute myeloid leukaemia. Lancet 2006; 368: 1894–1907.

  2. 2

    Dohner H, Estey EH, Amadori S, Appelbaum FR, Buchner T, Burnett AK et al. Diagnosis and management of acute myeloid leukemia in adults: recommendations from an international expert panel, on behalf of the European LeukemiaNet. Blood 2010; 115: 453–474.

  3. 3

    Koreth J, Schlenk R, Kopecky KJ, Honda S, Sierra J, Djulbegovic BJ et al. Allogeneic stem cell transplantation for acute myeloid leukemia in first complete remission: systematic review and meta-analysis of prospective clinical trials. JAMA 2009; 301: 2349–2361.

  4. 4

    Grimwade D, Hills RK, Moorman AV, Walker H, Chatters S, Goldstone AH et al. Refinement of cytogenetic classification in acute myeloid leukemia: determination of prognostic significance of rare recurring chromosomal abnormalities among 5876 younger adult patients treated in the United Kingdom Medical Research Council trials. Blood 2010; 116: 354–365.

  5. 5

    Patel JP, Gonen M, Figueroa ME, Fernandez H, Sun Z, Racevskis J et al. Prognostic relevance of integrated genetic profiling in acute myeloid leukemia. N Engl J Med 2012; 366: 1079–1089.

  6. 6

    Shen Y, Zhu YM, Fan X, Shi JY, Wang QR, Yan XJ et al. Gene mutation patterns and their prognostic impact in a cohort of 1185 patients with acute myeloid leukemia. Blood 2011; 118: 5593–5603.

  7. 7

    Ofran Y, Rowe JM . Genetic profiling in acute myeloid leukaemia—where are we and what is its role in patient management. Br J Haematol 2013; 160: 303–320.

  8. 8

    Delhommeau F, Dupont S, Della Valle V, James C, Trannoy S, Masse A et al. Mutation in TET2 in myeloid cancers. N Engl J Med 2009; 360: 2289–2301.

  9. 9

    Langemeijer SM, Kuiper RP, Berends M, Knops R, Aslanyan MG, Massop M et al. Acquired mutations in TET2 are common in myelodysplastic syndromes. Nat Genet 2009; 41: 838–842.

  10. 10

    Mardis ER, Ding L, Dooling DJ, Larson DE, McLellan MD, Chen K et al. Recurring mutations found by sequencing an acute myeloid leukemia genome. N Engl J Med 2009; 361: 1058–1066.

  11. 11

    Ley TJ, Ding L, Walter MJ, McLellan MD, Lamprecht T, Larson DE et al. DNMT3A mutations in acute myeloid leukemia. N Engl J Med 2010; 363: 2424–2433.

  12. 12

    Nikoloski G, Langemeijer SM, Kuiper RP, Knops R, Massop M, Tonnissen ER et al. Somatic mutations of the histone methyltransferase gene EZH2 in myelodysplastic syndromes. Nat Genet 2010; 42: 665–667.

  13. 13

    Shih AH, Abdel-Wahab O, Patel JP, Levine RL . The role of mutations in epigenetic regulators in myeloid malignancies. Nat Rev Cancer 2012; 12: 599–612.

  14. 14

    Chou WC, Huang HH, Hou HA, Chen CY, Tang JL, Yao M et al. Distinct clinical and biological features of de novo acute myeloid leukemia with additional sex comb-like 1 (ASXL1) mutations. Blood 2010; 116: 4086–4094.

  15. 15

    Grossmann V, Tiacci E, Holmes AB, Kohlmann A, Martelli MP, Kern W et al. Whole-exome sequencing identifies somatic mutations of BCOR in acute myeloid leukemia with normal karyotype. Blood 2011; 118: 6153–6163.

  16. 16

    Li M, Collins R, Jiao Y, Ouillette P, Bixby D, Erba H et al. Somatic mutations in the transcriptional corepressor gene BCORL1 in adult acute myelogenous leukemia. Blood 2011; 118: 5914–5917.

  17. 17

    Welch JS, Ley TJ, Link DC, Miller CA, Larson DE, Koboldt DC et al. The origin and evolution of mutations in acute myeloid leukemia. Cell 2012; 150: 264–278.

  18. 18

    Yoshida K, Sanada M, Shiraishi Y, Nowak D, Nagata Y, Yamamoto R et al. Frequent pathway mutations of splicing machinery in myelodysplasia. Nature 2011; 478: 64–69.

  19. 19

    Naoe T, Kiyoi H . Gene mutations of acute myeloid leukemia in the genome era. Int J Hematol 2013; 97: 165–174.

  20. 20

    Rollig C, Bornhauser M, Thiede C, Taube F, Kramer M, Mohr B et al. Long-term prognosis of acute myeloid leukemia according to the new genetic risk classification of the European LeukemiaNet recommendations: evaluation of the proposed reporting system. J Clin Oncol 2011; 29: 2758–2765.

  21. 21

    Mrozek K, Marcucci G, Nicolet D, Maharry KS, Becker H, Whitman SP et al. Prognostic significance of the European LeukemiaNet standardized system for reporting cytogenetic and molecular alterations in adults with acute myeloid leukemia. J Clin Oncol 2012; 30: 4515–4523.

  22. 22

    Metzeler KH, Maharry K, Radmacher MD, Mrozek K, Margeson D, Becker H et al. TET2 mutations improve the new European LeukemiaNet risk classification of acute myeloid leukemia: a Cancer and Leukemia Group B study. J Clin Oncol 2011; 29: 1373–1381.

  23. 23

    Miyawaki S . Clinical studies of acute myeloid leukemia in the Japan Adult Leukemia Study Group. Int J Hematol 2012; 96: 171–177.

  24. 24

    Ohtake S, Miyawaki S, Fujita H, Kiyoi H, Shinagawa K, Usui N et al. Randomized study of induction therapy comparing standard-dose idarubicin with high-dose daunorubicin in adult patients with previously untreated acute myeloid leukemia: the JALSG AML201 Study. Blood 2011; 117: 2358–2365.

  25. 25

    Miyawaki S, Ohtake S, Fujisawa S, Kiyoi H, Shinagawa K, Usui N et al. A randomized comparison of 4 courses of standard-dose multiagent chemotherapy versus 3 courses of high-dose cytarabine alone in postremission therapy for acute myeloid leukemia in adults: the JALSG AML201 Study. Blood 2011; 117: 2366–2372.

  26. 26

    Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton DA, Gralnick HR et al. Proposed revised criteria for the classification of acute myeloid leukemia. A report of the French-American-British Cooperative Group. Ann Intern Med 1985; 103: 620–625.

  27. 27

    Kiyoi H, Naoe T, Nakano Y, Yokota S, Minami S, Miyawaki S et al. Prognostic implication of FLT3 and N-RAS gene mutations in acute myeloid leukemia. Blood 1999; 93: 3074–3080.

  28. 28

    Ozeki K, Kiyoi H, Hirose Y, Iwai M, Ninomiya M, Kodera Y et al. Biologic and clinical significance of the FLT3 transcript level in acute myeloid leukemia. Blood 2004; 103: 1901–1908.

  29. 29

    Yamamoto Y, Kiyoi H, Nakano Y, Suzuki R, Kodera Y, Miyawaki S et al. Activating mutation of D835 within the activation loop of FLT3 in human hematologic malignancies. Blood 2001; 97: 2434–2439.

  30. 30

    Osumi K, Fukui T, Kiyoi H, Kasai M, Kodera Y, Kudo K et al. Rapid screening of leukemia fusion transcripts in acute leukemia by real-time PCR. Leuk Lymphoma 2002; 43: 2291–2299.

  31. 31

    Makishima H, Yoshida K, Nguyen N, Przychodzen B, Sanada M, Okuno Y et al. Somatic SETBP1 mutations in myeloid malignancies. Nat Genet 2013; 45: 942–946.

  32. 32

    Kon A, Shih LY, Minamino M, Sanada M, Shiraishi Y, Nagata Y et al. Recurrent mutations in multiple components of the cohesin complex in myeloid neoplasms. Nat Genet 2013; 45: 1232–1237.

  33. 33

    Sakaguchi H, Okuno Y, Muramatsu H, Yoshida K, Shiraishi Y, Takahashi M et al. Exome sequencing identifies secondary mutations of SETBP1 and JAK3 in juvenile myelomonocytic leukemia. Nat Genet 2013; 45: 937–941.

  34. 34

    Ishikawa Y, Kiyoi H, Tsujimura A, Miyawaki S, Miyazaki Y, Kuriyama K et al. Comprehensive analysis of cooperative gene mutations between class I and class II in de novo acute myeloid leukemia. Eur J Haematol 2009; 83: 90–98.

  35. 35

    Speck NA, Gilliland DG . Core-binding factors in haematopoiesis and leukaemia. Nat Rev Cancer 2002; 2: 502–513.

  36. 36

    Swerdlow S, Campo E, Harris N, Jaffe E, Pileri S, Stein H et al WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues 4th edn. WHO Press: Lyon, 2008.

  37. 37

    Vardiman JW, Thiele J, Arber DA, Brunning RD, Borowitz MJ, Porwit A et al. The 2008 revision of the World Health Organization (WHO) classification of myeloid neoplasms and acute leukemia: rationale and important changes. Blood 2009; 114: 937–951.

  38. 38

    Abdel-Wahab O, Levine RL . Mutations in epigenetic modifiers in the pathogenesis and therapy of acute myeloid leukemia. Blood 2013; 121: 3563–3572.

  39. 39

    Grossmann V, Schnittger S, Kohlmann A, Eder C, Roller A, Dicker F et al. A novel hierarchical prognostic model of AML solely based on molecular mutations. Blood 2012; 120: 2963–2972.

  40. 40

    Cancer Genome Atlas Research N. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. N Engl J Med 2013; 368: 2059–2074.

  41. 41

    Tang JL, Hou HA, Chen CY, Liu CY, Chou WC, Tseng MH et al. AML1/RUNX1 mutations in 470 adult patients with de novo acute myeloid leukemia: prognostic implication and interaction with other gene alterations. Blood 2009; 114: 5352–5361.

  42. 42

    Stone RM . Acute myeloid leukemia in first remission: to choose transplantation or not? J Clin Oncol 2013; 31: 1262–1266.

  43. 43

    Paschka P, Schlenk RF, Gaidzik VI, Habdank M, Kronke J, Bullinger L et al. IDH1 and IDH2 mutations are frequent genetic alterations in acute myeloid leukemia and confer adverse prognosis in cytogenetically normal acute myeloid leukemia with NPM1 mutation without FLT3 internal tandem duplication. J Clin Oncol 2010; 28: 3636–3643.

  44. 44

    Dufour A, Schneider F, Metzeler KH, Hoster E, Schneider S, Zellmeier E et al. Acute myeloid leukemia with biallelic CEBPA gene mutations and normal karyotype represents a distinct genetic entity associated with a favorable clinical outcome. J Clin Oncol 2010; 28: 570–577.

  45. 45

    Krzywinski M, Schein J, Birol I, Connors J, Gascoyne R, Horsman D et al. Circos: an information aesthetic for comparative genomics. Genome Res 2009; 19: 1639–1645.

Download references

Acknowledgements

This study was performed as a research program of the Project for Development of Innovative Research on Cancer Therapeutics (P-Direct), Ministry of Education, Culture, Sports, Science and Technology of Japan, and was also supported by Grants-in-Aid from the Scientific Research Program from the Japanese Ministry of Education, Culture, Sports, Science, and Technology, the Ministry of Health, Labor and Welfare for Cancer Research (Clinical Cancer Research H23-004 and H25-006), the National Cancer Center Research and Development Fund (23-A-23) and from KAKENHI (22134006 and 23249052).

Author contributions

H Kiyoi, S Ogawa and TN designed the study, interpreted the data and wrote the manuscript; RK, YN, T Kato, EY, KS and FC performed molecular analysis and interpreted the data; YN, YS, KC, HT, SM and S Ogawa performed bioinformatics; NA, S Ohtake, SM, YM, TS, YO, N Usui, H Kanamori, T Kiguchi, KI, N Uike, FK, KK, CN, MO, AT, FI, HS, YK and HM collected samples and clinical data, contributed to the interpretation of the data and critically reviewed the draft; and all authors approved the final version submitted for the publication.

Author information

Correspondence to H Kiyoi or S Ogawa.

Ethics declarations

Competing interests

H Kiyoi: Research funding from Bristol-Myers Squibb, Novartis Pharma, Chugai Pharmaceutical Co., Ltd. and Kyowa Hakko Kirin Co., Ltd. YM: Honoraria from Bristol-Myers Squibb, Novartis Pharma, Chugai Pharmaceutical Co., Ltd., Kyowa Hakko Kirin Co., Ltd. and Celgene Japan; Research funding from Bristol-Myers Squibb, Novartis Pharma, Chugai Pharmaceutical Co., Ltd. and Kyowa Hakko Kirin Co., Ltd. NU: Consultant for Kyowa Hakko Kirin Co., Ltd.; Honoraria from Bristol-Myers Squibb, Novartis Pharma and Chugai Pharmaceutical Co., Ltd.; Research funding from Bristol-Myers Squibb, Novartis Pharma, Chugai Pharmaceutical Co., Ltd. and Kyowa Hakko Kirin Co., Ltd. TN: Research funding from Bristol-Myers Squibb, Novartis Pharma, Chugai Pharmaceutical Co., Ltd., Kyowa Hakko Kirin Co., Ltd., Dainippon Sumitomo Pharma and Zenyaku Kogyo. The other authors declare no conflict of interest.

Additional information

Supplementary Information accompanies this paper on the Leukemia website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Kihara, R., Nagata, Y., Kiyoi, H. et al. Comprehensive analysis of genetic alterations and their prognostic impacts in adult acute myeloid leukemia patients. Leukemia 28, 1586–1595 (2014) doi:10.1038/leu.2014.55

Download citation

Keywords

  • acute myeloid leukemia
  • gene mutations
  • prognosis
  • risk factor

Further reading