Acute erythroid leukemia (AEL) can be separated into distinct prognostic subsets based on cytogenetic and molecular genetic characteristics

Article metrics

Acute erythroid leukemia (AEL) (=AML FAB M6) comprises <5% of adult acute myeloid leukemia (AML), but becomes more frequent with higher age.1 AEL patients were described to have more frequently poor risk cytogenetics and worse survival than other AML subtypes.2 The cytogenetic risk group was suggested to be prognostically relevant for AEL patients, but the diagnosis of AEL was no independent prognostic parameter when the cytogenetic risk group or the history of the disease were considered.3 AEL was described to differ from overall AML, for example, by a lower FLT3 mutation rate.2 It remains unclear whether a diagnosis of AEL per se or the association with poor risk cytogenetic features or other poor risk characteristics causes the more adverse prognosis.3 So far, mutations of FLT3 and NRAS2 only have been studied in more detail in AEL. We previously suggested that the different AML and myelodysplastic syndrome (MDS) subtypes with predominant erythropoiesis may be combined into one category.4 Others proposed that the clinical and genetic features of AML with erythroid predominance are closer to high-grade MDS than to other types of AML.5

We investigated 14 candidate genes combined with the cytogenetic background and the prognostic impact in 92 AEL patients (31 female/61 male; median age, 68.8 years; range: 21.3–88.3 years; Supplementary Table S1). Survival data was available in 74 patients. All patients had erythroleukemia (erythroid/myeloid leukemia),6 pure erythroid leukemia was not considered. Patients received induction and consolidation chemotherapy according to AML Cooperative Group (AMLCG)7 or comparable protocols. Bone marrow samples were sent for diagnosis to the MLL Munich Leukemia Laboratory from August 2005 to May 2012. All patients gave their written consent for genetic analyses and research studies. Part of the patients had been included in a previous study.4 All cases were investigated by cytomorphology, chromosome banding analysis and, in case of a normal karyotype, by array comparative genomic hybridization (CGH). Because of limitations in sample material, we combined data from 454 amplicon deep sequencing (Roche, Branford, CT, USA)8 with Sanger sequencing and melting curve analyses (for details see Supplementary Appendix). The various screening techniques for the different genes are given in Supplementary Table S2. Of note, despite the incongruent molecular screening programme, no significant bias was detectable in terms of detection of minor subclones by deep sequencing in comparison to other assays. Thus, no influence on the subsequent results is expected.

Cytomorphologic characteristics of the cohort are described in Supplementary Appendix (see also Supplementary Table S3). The majority of patients had aberrant karyotypes (n=61/92; 66.3%). Forty-seven cases (51.1%) presented an intermediate, 45 (48.9%) cases an adverse cytogenetic MRC (Medical Research Council) category.9 Complex karyotypes (3 clonal abnormalities) were detected in 37 cases (40.2%; Supplementary Table S4). By array CGH, most cytogenetically normal karyotype cases (30/31 cases; 96.8%) investigated retained a normal karyotype, one case showed a duplication of 11q13.3 to 11q25 including the ATM and MLL genes.

Mutations were detected in 85 of the 92 patients (92.4%); 56 of the 85 (65.9%) mutated patients carried one and 29 of the 85 (34.1%) cases harbored two (n=23) or more (n=6) mutations. TP53 was most frequently mutated (40/92, 43.5%) followed by NPM1 (15/92; 16.3%) and DNMT3A (12/92; 13.0%). Less frequent were ASXL1mut (7/88; 8.0%), RUNX1 (8/92; 8.7%), MLL-PTD (7/90; 7.8%), WT1 (7/88; 8.0%), IDH1 (6/80; 7.5%), IDH2 (4/85; 4.7%), NRAS (3/91; 3.3%), KRAS (3/92; 3.3%), FLT3-ITD (3/91; 3.3%), FLT3-TKD (3/85; 3.5%), SF3B1 (2/80; 2.5%) and CEBPA (1/92; 1.1%) (Figure 1).

Figure 1

Illustration of the distribution of molecular markers, cytogenetic MRC risk groups and aberrant karyotypes in the cohort. Each column represents one patient. Red rectangles indicate cases with mutations; light grey rectangles indicate wild-type cases; orange rectangles indicate intermediate; light blue rectangles adverse MRC9 karyotypes; dark grey rectangles indicate aberrant karyotypes; dark blue rectangles normal karyotypes; white rectangles depict lack of mutation analysis for the respective gene in the respective patient.

Clonality analyses were performed by estimating the mutation load by deep sequencing in cases with >1 mutation. When we analyzed ASXL1mut, which were found in coincidence with DNMT3A, RUNX1 or TP53 mutations, the mutation loads were similar for the different genes. This also holds true for other combinations of mutations (Supplementary Figures S1A–E). Although it has to be seen that different molecular technologies were combined, those results suggest that AEL represents a rather monoclonal disease than a composition of different clones as this was recently supposed for MDS.10 Alternatively, AEL may have a strong hyperproliferative potential after acquisition of all clonal abnormalities thus being able to overgrow clones with fewer mutations.

Ring sideroblasts were more frequent in NPM1wt as compared with NPM1mut cases (P<0.001). The proportion of dysplastic erythropoietic cells (P=0.002) and of ring sideroblasts (P=0.034) were higher in cases with complex karyotypes than non-complex karyotypes (Supplementary Table S3). Therefore, a subset of AEL patients shows vicinity to MDS cases.

Only genes that were found mutated in three or more cases were included in the correlation studies. TP53mut were found in 40 of 77 (43.5%) NPM1wt cases, whereas no TP53mut was found in the 15 NPM1mut cases (P<0.001). TP53mut were found in 38 of 83 (42.2%) MLL-PTD-negative cases, whereas no TP53mut was found in 7 MLL-PTD-positive cases (P=0.020). TP53mut were only found in WT1wt and not in WT1mut cases (39/80; 44.8% vs 0/7; 0.0%; P=0.015; Figure 1). NPM1 (15/47; 31.9% vs 0/45; 0.0%; P<0.001), RUNX1 (8/47; 17.0% vs 0/45; 0.0%; P=0.006) and WT1mut (7/43; 16.3% vs 0/44; 0.0%; P=0.006) were more frequent in MRC intermediate as compared with adverse karyotypes. In contrast, TP53mut were more frequent in adverse than in intermediate karyotypes (37/45; 82.2% vs 3/47; 6.4%; P<0.001) and in complex than in non-complex karyotypes (35/37; 94.6% vs 5/55; 9.1%; P<0.001) (Figure 1; Supplementary Table S5). Correlations of genetic features with demographic parameters are shown in Supplementary Appendix.

Median overall survival (OS) of the cohort was 14.3 months (median follow-up: 23.5 months). Intermediate MRC cases had a better outcome than adverse MRC9 karyotype cases (median OS not reached (n.r.) vs 6.4 months, P=0.002; median event-free survival (EFS) n.r. vs 5.5 months, P<0.001; Supplementary Figure S2A). OS (median n.r. vs 9.3 months; P=0.012) and EFS (25.2 vs 5.6 months; P=0.001) were better for non-complex karyotype cases. Cases with a normal karyotype showed higher OS (P<0.001) and EFS (P<0.001) than aberrant karyotypes (Table 1A).

Table 1A OS and EFS in 74 patients with AEL with available survival data comparing different cytogenetic subgroups

NPM1mut cases showed better outcome than NPM1wt (median OS n.r. vs 9.3 months, P<0.001; median EFS n.r. vs 6.1 months, P=0.002; Supplementary Figure S2B). TP53mut cases had shorter OS (median 6.4 months vs n.r.; P=0.001) and EFS (median, 5.7 vs 25.2 months; P=0.001) than TP53wt (Supplementary Figure S2C). RUNX1mut patients had shorter OS (median, 4.1 vs 19.5 months; P=0.007) and EFS (median 4.1 vs 11.7 months; P=0.030) than RUNX1wt patients; Supplementary Figure S2D). ASXL1mut cases showed a trend of shorter OS (median 5.4 vs 19.5 months; P=0.082), whereas EFS did not differ significantly (median 5.4 vs 12.4 months; P=n.s.; Supplementary Figure S2E; Table 1B).

Table 1B OS and EFS in the different molecular genetic subgroups

When the intermediate MRC risk group was separately analyzed, the favorable prognostic impact of NPM1mut as compared with NPM1wt was confirmed (OS: P=0.013; EFS: P=0.075). RUNX1mut (OS: P=0.001; EFS: P=0.001) and ASXL1mut (OS: P=0.024; EFS: P=0.042) conferred worse outcomes. Survival of TP53mut cases was worse but statistical comparison was hampered by the limited number of TP53mut cases in this subgroup.

By univariate analysis, the following parameters were significant for OS: higher age (P=0.001, hazards ratio (HR)=1.54 per 10 years of increase), t-AEL vs de novo AEL (P=0.002, HR=4.79), adverse cytogenetic MRC risk group (P=0.003, HR=2.88), higher white blood cell (WBC) counts (P=0.031, HR=1.11 per 109/l increase), and RUNX1 (P=0.012, HR=4.01) and TP53mut (P=0.001, HR=3.11), whereas NPM1mut (P=0.003, HR=0.11) were favorable (Supplementary Figure 3A). By multivariate analysis, age (P=0.025, HR=1.43 per 10 years of increase), t-AEL (P=0.003, HR=5.02), WBC counts (P=0.001, HR=1.18 per 109/l increase) and RUNX1mut (P=0.008, HR=5.96) remained significant for OS. NPM1 (P=0.051, HR=0.19) and TP53mut (P=0.096, HR=2.54) had borderline significance. Similar results were obtained for parameters influencing the EFS in univariate analysis (Supplementary Figure S3B). By multivariate analysis, t-AEL (P=0.020, HR=3.36), higher WBC counts (P=0.007, HR=1.14 per 109/l increase) and RUNX1mut (P=0.012, HR=5.27) remained prognostically adverse for EFS (Supplementary Table S6).

In conclusion, we could confirm the high frequency of adverse karyotypes9 of 48.9% in our AEL cohort. Santos et al.3 and Hasserjian et al.2 described adverse karyotypes in 61% and 64% of AEL patients, respectively. Unfavorable karyotypes were mostly due to the presence of complex karyotypes, which had a frequency of 40.2% in our study. Patients with intermediate karyotypes had a better outcome than patients with adverse MRC9 karyotypes (OS: P=0.002; EFS: P<0.001). Wells et al.1 reported a median OS of 14 months for patients with standard risk cytogenetics and 2 months for poor risk karyotypes (P=0.005) in AEL. In the study by Hasserjian et al.,2 outcome of AEL patients was significantly influenced by the cytogenetic risk group. In contrast, the blast count had no prognostic relevance when AEL patients were compared with patients with other myeloid subtypes (AML with myelodysplasia-related changes or MDS) with erythroid hyperplasia.2

We found a high frequency of mutations (85/92; 92.4%) in our AEL cohort. The mutation profiles differed significantly from overall AML.11 NPM1mut were found in 16.3% and FLT3-ITD in only 3.3% of our AEL patients, thereby much lower when compared with overall AML patients that presented >30% of NPM112 and 20–25% of FLT3-ITD mutated cases.13, 14 Similarly, Hasserjian et al.2 reported FLT3-ITD to occur in only 6% of the AEL patients. We found high frequencies of TP53mut (43.5%), and further detected DNMT3A (13.0%), ASXL1 (8.0%), MLL-PTD (7.8%), WT1 (7/88; 8.0%) and IDH1 (6/80; 7.5%) mutations in our AEL cohort. We could confirm the association of TP53mut with complex karyotypes11, 15 also in AEL. NPM1mut had a positive prognostic impact, whereas TP53, RUNX1 and ASXL1mut were adverse. The high rate of TP53mut contributes to explain the adverse outcome in AEL. Based on these results, therapeutic decisions in AEL patients should always consider cytogenetics and, in the future, molecular mutation profiles focusing on NPM1, RUNX1 and TP53.


  1. 1

    Wells AW, Bown N, Reid MM, Hamilton PJ, Jackson GH, Taylor PR . Erythroleukaemia in the north of England: a population based study. J Clin Pathol 2001; 54: 608–612.

  2. 2

    Hasserjian RP, Zuo Z, Garcia C, Tang G, Kasyan A, Luthra R et al. Acute erythroid leukemia: a reassessment using criteria refined in the 2008 WHO classification. Blood 2010; 115: 1985–1992.

  3. 3

    Santos FP, Faderl S, Garcia-Manero G, Koller C, Beran M, O'Brien S et al. Adult acute erythroleukemia: an analysis of 91 patients treated at a single institution. Leukemia 2009; 23: 2275–2280.

  4. 4

    Bacher U, Haferlach C, Alpermann T, Kern W, Schnittger S, Haferlach T . Comparison of genetic and clinical aspects in patients with acute myeloid leukemia and myelodysplastic syndromes all with more than 50% of bone marrow erythropoietic cells. Haematologica 2011; 96: 1284–1292.

  5. 5

    Zuo Z, Medeiros LJ, Chen Z, Liu D, Bueso-Ramos CE, Luthra R et al. Acute myeloid leukemia (AML) with erythroid predominance exhibits clinical and molecular characteristics that differ from other types of AML. PLoS One 2012; 7: e41485.

  6. 6

    Arber D, Brunning R, Orazi A, Porwit A, Peterson L, Thiele J et al. Acute myeloid leukemia, not otherwise specified. In: Swerdlow S, Campo E, Harris N, Jaffe E, Pileri S, Stein H, et al. (eds) WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues 4 edn. IARC Press: Lyon, 2008; p 130–139.

  7. 7

    Büchner T, Hiddemann W, Berdel WE, Wormann B, Schoch C, Fonatsch C et al. 6-Thioguanine, cytarabine, and daunorubicin (TAD) and high-dose cytarabine and mitoxantrone (HAM) for induction, TAD for consolidation, and either prolonged maintenance by reduced monthly TAD or TAD-HAM-TAD and one course of intensive consolidation by sequential HAM in adult patients at all ages with de novo acute myeloid leukemia (AML): a randomized trial of the German AML Cooperative Group. J Clin Oncol 2003; 21: 4496–4504.

  8. 8

    Kohlmann A, Klein HU, Weissmann S, Bresolin S, Chaplin T, Cuppens H et al. The Interlaboratory RObustness of Next-generation sequencing (IRON) study: a deep sequencing investigation of TET2, CBL and KRAS mutations by an international consortium involving 10 laboratories. Leukemia 2011; 25: 1840–1848.

  9. 9

    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.

  10. 10

    Walter MJ, Shen D, Shao J, Ding L, White B, Kandoth C et al. Clonal diversity of recurrently mutated genes in myelodysplastic syndromes. Leukemia 2013; e-pub ahead of print 26 March 2013; doi:10.1038/leu.2013.58.

  11. 11

    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.

  12. 12

    Falini B, Mecucci C, Tiacci E, Alcalay M, Rosati R, Pasqualucci L et al. Cytoplasmic nucleophosmin in acute myelogenous leukemia with a normal karyotype. N Engl J Med 2005; 352: 254–266.

  13. 13

    Schnittger S, Schoch C, Dugas M, Kern W, Staib P, Wuchter C et al. Analysis of FLT3 length mutations in 1003 patients with acute myeloid leukemia: correlation to cytogenetics, FAB subtype, and prognosis in the AMLCG study and usefulness as a marker for the detection of minimal residual disease. Blood 2002; 100: 59–66.

  14. 14

    Thiede C, Steudel C, Mohr B, Schaich M, Schakel U, Platzbecker U et al. Analysis of FLT3-activating mutations in 979 patients with acute myelogenous leukemia: association with FAB subtypes and identification of subgroups with poor prognosis. Blood 2002; 99: 4326–4335.

  15. 15

    Rücker FG, Schlenk RF, Bullinger L, Kayser S, Teleanu V, Kett H et al. TP53 alterations in acute myeloid leukemia with complex karyotype correlate with specific copy number alterations, monosomal karyotype, and dismal outcome. Blood 2012; 119: 2114–2121.

Download references

Author information

Correspondence to T Haferlach.

Ethics declarations

Competing interests

VG, UB, FP, SW, AR, CE, AF, MZ, MS, and AK are employed by the MLL Munich Leukemia Laboratory GmbH. CH, SS, WK and TH are part owners of the MLL Munich Leukemia Laboratory GmbH.

Additional information

Author contributions

VG, CH, AK and TH performed study design. UB and VG performed data analysis and wrote the first manuscript draft. UB and TH performed cytomorphology. VG, SS, FP, SW, CE, AF, and AK performed molecular analyses. AR was responsible for bioinformatics. CH, MZ, and MS did cytogenetic analyses. WK contributed to statistical analysis. All authors contributed to writing of the manuscript, reviewed the final version, and gave consent to the final version.

Supplementary Information accompanies this paper on the Leukemia website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Further reading