Abstract

Lymphoid enhancer-binding factor-1 (LEF1) is a key transcription factor of Wnt signaling. We recently showed that aberrant LEF1 expression induces acute myeloid leukemia (AML) in mice, and found high LEF1 expression in a subset of cytogenetically normal AML (CN-AML) patients. Whether LEF1 expression associates with clinical and molecular patient characteristics and treatment outcomes remained unknown. We therefore studied LEF1 expression in 210 adults with CN-AML treated on German AML Cooperative Group trials using microarrays. High LEF1 expression (LEF1high) associated with significantly better relapse-free survival (RFS; P < .001), overall survival (OS; P < .001), and event-free survival (EFS; P < .001). In multivariable analyses adjusting for established prognosticators, LEF1high status remained associated with prolonged RFS (P = .007), OS (P = .01), and EFS (P = .003). In an independent validation cohort of 196 CN-AML patients provided by the German-Austrian AML Study Group, LEF1high patients had significantly longer OS (P = .02) and EFS (P = .04). We validated the prognostic relevance of LEF1 expression by quantitative PCR, thereby providing a clinically applicable platform to incorporate this marker into future risk-stratification systems for CN-AML. Gene-expression profiling and immunophenotyping revealed up-regulation of lymphopoiesis-related genes and lymphoid cell-surface antigens in LEF1high patients. In summary, we provide evidence that high LEF1 expression is a novel favorable prognostic marker in CN-AML.

Introduction

The Wnt signaling pathway is a critical regulator of stem cell function in healthy tissues and cancer, including acute myeloid leukemia (AML).1  Studies in humans and mice have demonstrated that the Wnt pathway is essential for maintenance, activation, and proliferation of normal hematopoietic stem cells,1  and that aberrant Wnt signaling can lead to expansion of leukemic stem cells in myeloid and lymphoid neoplasias.2-5  Activation of the transcriptional coactivators β-catenin and γ-catenin, the main downstream effectors of canonical Wnt signaling, is linked to enhanced leukemogenic potential and self-renewal of putative leukemic stem cells.5-7 

Lymphoid enhancer-binding factor-1 (LEF1) is a member of the LEF1/T-cell factor (TCF) family of transcription factors. During canonical Wnt signaling, LEF1/TCF proteins directly interact with β-catenin to induce expression of target genes, including the cell-cycle regulators cyclin D1 and c-myc.1  In addition, LEF1 has roles in normal hematopoiesis and leukemogenesis that are independent of its involvement in Wnt signaling.8,9  Our group previously demonstrated that ordered expression of Lef-1 is necessary for normal hematopoietic stem cell function in mice, and that Lef-1 overexpression induces AML which is propagated by leukemic stem cells with lymphoid characteristics.10  A subset of patients with cytogenetically normal AML (CN-AML) express high levels of LEF1,10  but associations between LEF1 expression and other clinical and molecular patient characteristics and outcomes are unknown so far. Here, we demonstrate for the first time that high LEF1 expression is a favorable prognostic factor in patients with CN-AML in 2 independent, relatively large AML patient cohorts when measured by microarray techniques or quantitative PCR (qPCR), and is associated with distinct molecular and immunophenotypic characteristics.

Methods

Patients

We studied 210 patients with previously untreated CN-AML (median age, 59 years; range, 17-83 years) who were treated on 2 consecutive phase 3 trials of the German AML Cooperative Group (AMLCG-1992 and AMLCG-1999 [clinicaltrials.gov identifier NCT00266136]) between 1999 and 2004.11,12  One hundred nine patients (52%) were aged < 60 years (younger patients) and 101 patients (48%) were ≥ 60 years (older patients). All patients received cytarabine-based intensive induction and consolidation chemotherapy.11-13  The diagnosis of a normal karyotype was based on conventional cytogenetic examination of at least 20 metaphases from bone marrow (BM). Patients were characterized for NPM1, CEBPA, IDH1, and IDH2 mutations, FLT3-internal tandem duplications (FLT3-ITD), tyrosine kinase domain mutations (FLT3-TKD [D835]), and MLL partial tandem duplications (MLL-PTD), as described previously.14-16  An independent validation cohort of 196 younger (< 60 years) CN-AML patients was provided by the German-Austrian AML Study Group (AMLSG; trials AMLSG-HD98A [NCT00146120]17  and AMLSG 07-04 [NCT00151242]). This patient group was characterized for the gene mutations previously mentioned, and for TET2, ASXL1, DNMT3A, RUNX1, and WT1 mutations. All study protocols were in accordance with the Declaration of Helsinki and approved by the institutional review boards of the participating centers, and all patients provided written informed consent.

Microarray analyses

For the primary (AMLCG) cohort, pretreatment BM samples were studied using Affymetrix HG-U133A (n = 154) or HG-U133plus2.0 (n = 56) oligonucleotide microarrays as described previously (for details on microarray data processing, see supplemental Methods, available on the Blood Web site; see the Supplemental Materials link at the top of the online article).18,19  Patients with LEF1 expression values above the median of all patients were classified as having high LEF1 expression (LEF1high), and all other patients were considered to have low LEF1 expression (LEF1low). The choice of the median as the threshold for separating patients into 2 groups was based on analyses of outcomes according to quartiles of LEF1 expression (data not shown). ERG, BAALC, MN1, and EVI1 expression levels were also determined from the microarray data.19  In the independent validation cohort provided by the AMLSG, pretreatment BM (n = 142) or peripheral blood (n = 54) samples were studied by independent investigators using cDNA microarrays (n = 130) or Affymetrix HG-U133plus2.0 oligonucleotide microarrays (n = 66), as previously described.20,21 LEF1 expression levels were dichotomized at the median of all samples. Analysis of differentially expressed genes and gene set enrichment analysis (GSEA) are described in supplemental Methods. Microarray data are available at the Gene Expression Omnibus database (GEO; accession no. GSE12417, GSE8043, and GSE15434).22 

Measurement of LEF1 expression by qPCR

LEF1 expression was measured by real-time qPCR in a subgroup of 122 CN-AML patients from the primary cohort for which sufficient material was available. cDNA had been prepared at the time of initial diagnosis using random hexamer primers, and cryopreserved at −80°C. LEF1 expression was measured using a TaqMan probe-based qPCR assay recognizing all 4 major human LEF1 isoforms (Hs01547250_m1; Applied Biosystems), and normalized to ABL expression (Hs01104724_mH).23  To allow comparison of our expression data and thresholds with future studies, expression values were standardized to a calibrator sample (cDNA from the K562 cell line), using the comparative threshold cycle (CT) method.24  An analysis of Martingale residuals from univariable Cox models identified a ΔΔCT value of −2.50 as a clinically informative threshold to separate LEF1high and LEF1low patients.

Flow cytometry

Multiparameter flow cytometric analysis was performed on BM mononuclear cells from 124 AMLCG patients, using a FACSCalibur flow cytometer (BD Biosciences). Patients were characterized for 28 antigens using 17 triple monoclonal antibody combinations, including CD34/CD2/CD33, CD7/CD33/CD34, CD4/CD13/CD14, and cyTdT/cyCD79a/cyCD3 (conjugated with the fluorochromes FITC, PE, and PE–cyanin-5.1, respectively; Immunotech), and 20 000 events were acquired per specimen.

Statistical analyses

Definitions of clinical endpoints (complete remission [CR], relapse-free survival [RFS], overall survival [OS], and event-free survival [EFS]) are listed in supplemental Methods. The association between LEF1 expression as a continuous numerical variable and patient outcomes was studied using univariable Cox proportional hazards and logistic regression models. Baseline clinical and molecular characteristics were compared between LEF1high and LEF1low patients using the Fisher exact test for categorical variables and the Wilcoxon rank-sum test for continuous variables. Time-to-event variables were analyzed according to the Kaplan-Meier method, and P values were calculated by the log-rank test. A multivariable logistic regression model was constructed for factors associated with achievement of CR, and multivariable Cox proportional hazards models were used to study factors associated with survival endpoints. No variable selection technique was used, and all variables remained in the multivariable model. All statistical data analyses were performed using the R 2.14.1 software package (R Foundation for Statistical Computing).

Results

Association of LEF1 expression levels with pretreatment patient characteristics

In the primary (AMLCG) cohort, patients with high LEF1 expression had lower pretreatment white blood cell counts (WBC; P < .001), lower BM blasts percentages (P = .02), and higher platelet counts (P = .04) than LEF1low patients (Table 1). LEF1high patients were less likely to carry a FLT3-ITD (P < .001) than LEF1low patients. We found no association between LEF1 expression and other gene mutations, but LEF1high patients were less likely to have high expression of ERG (P = .03). LEF1 expression levels did not differ significantly between younger and older patients (P = .20).

Table 1

Pretreatment characteristics in the primary cohort of 210 CN-AML patients according to LEF1 expression levels

VariableLEF1high, n = 105LEF1low, n = 105P
Median age, y (range) 57 (17-77) 59 (18-83) .56 
Female sex, no. (%) 59 (56) 63 (60) .67 
Secondary or treatment-related AML, no. (%) 4 (4) 6 (6) .75 
FAB subtype, no.   .21 
    M0  
    M1 22 35  
    M2 32 33  
    M4 27 19  
    M5 12 13  
    M6  
    RAEB-2  
Median WBC, 109/L (range) 21.0 (0.9-216) 44.6 (0.85-486) < .001 
Median BM blasts, % (range) 80 (11-100) 87.5 (17-100) .02 
Median platelet count, 109/L (range) 64 (6-280) 48 (9-471) .04 
FLT3-ITD present, no. (%) 25 (24) 61 (58) < .001 
NPM1 mutated, no. (%) 52 (50) 61 (58) .27 
CEBPA mutated, no. (%) 11 (11) 12 (12) > .99 
    Monaoallelic CEBPA mutation  
    Biallelic CEBPA mutation  
    Missing data  
ELN genetic group,25  no. (%)   .01 
    Favorable 49 (48) 31 (30)  
    Intermediate-I 53 (52) 71 (70)  
    Missing data  
IDH1 mutated, no. (%) 8 (12) 10 (12) > .99 
    Missing data 40 21  
IDH2 mutated, no. (%) 11 (17) 15 (18) > .99 
    Codon R140 mutation 11 15  
    Codon R172 mutation  
    Missing data 40 21  
FLT3-TKD (D835) mutated, no. (%) 9 (9) 8 (8) .81 
    Missing data  
MLL-PTD present, no. (%) 15 (15) 11 (11) .41 
    Missing data  
High ERG expression,* no. (%) 19 (18) 34 (32) .03 
High BAALC expression,* no. (%) 52 (50) 53 (50) 1.0 
High MN1 expression,* no. (%) 53 (50) 52 (50) 1.0 
VariableLEF1high, n = 105LEF1low, n = 105P
Median age, y (range) 57 (17-77) 59 (18-83) .56 
Female sex, no. (%) 59 (56) 63 (60) .67 
Secondary or treatment-related AML, no. (%) 4 (4) 6 (6) .75 
FAB subtype, no.   .21 
    M0  
    M1 22 35  
    M2 32 33  
    M4 27 19  
    M5 12 13  
    M6  
    RAEB-2  
Median WBC, 109/L (range) 21.0 (0.9-216) 44.6 (0.85-486) < .001 
Median BM blasts, % (range) 80 (11-100) 87.5 (17-100) .02 
Median platelet count, 109/L (range) 64 (6-280) 48 (9-471) .04 
FLT3-ITD present, no. (%) 25 (24) 61 (58) < .001 
NPM1 mutated, no. (%) 52 (50) 61 (58) .27 
CEBPA mutated, no. (%) 11 (11) 12 (12) > .99 
    Monaoallelic CEBPA mutation  
    Biallelic CEBPA mutation  
    Missing data  
ELN genetic group,25  no. (%)   .01 
    Favorable 49 (48) 31 (30)  
    Intermediate-I 53 (52) 71 (70)  
    Missing data  
IDH1 mutated, no. (%) 8 (12) 10 (12) > .99 
    Missing data 40 21  
IDH2 mutated, no. (%) 11 (17) 15 (18) > .99 
    Codon R140 mutation 11 15  
    Codon R172 mutation  
    Missing data 40 21  
FLT3-TKD (D835) mutated, no. (%) 9 (9) 8 (8) .81 
    Missing data  
MLL-PTD present, no. (%) 15 (15) 11 (11) .41 
    Missing data  
High ERG expression,* no. (%) 19 (18) 34 (32) .03 
High BAALC expression,* no. (%) 52 (50) 53 (50) 1.0 
High MN1 expression,* no. (%) 53 (50) 52 (50) 1.0 

CN-AML indicates cytogenetically normal acute myeloid leukemia; FAB, French-American-British classification; WBC, white blood cell count; ITD, internal tandem duplication; ELN, European Leukemia Net; TKD, tyrosine kinase domain; and PTD, partial tandem duplication.

*

High BAALC and MN1 expression were defined as an expression level above the median of all samples, respectively, and high ERG expression was defined as an expression level above the 75th percentile.19,26 

High LEF1 expression levels associate with favorable treatment outcomes

The median OS of the entire AMLCG patient cohort was 12.7 months, and the median follow-up according to Korn was 47 months.27  When LEF1 expression was analyzed as a continuous variable (supplemental Table 1), higher levels were significantly associated with higher CR rate (P = .03), longer RFS (P = .005), longer OS (P < .001), and longer EFS (P < .001). When patients were dichotomized according to their LEF1 expression levels, LEF1high patients showed a trend toward a higher CR rate (70% vs 57%, P = .09), and had significantly longer RFS (P < .001; Figure 1A), OS (P < .001; Figure 1B), and EFS (P < .001; Figure 1C) than LEF1low patients (Table 2).

Figure 1

Survival of CN-AML patients according to LEF1 expression levels. Kaplan-Meier plots of (A) RFS, (B) OS, and (C) EFS in the primary cohort.

Figure 1

Survival of CN-AML patients according to LEF1 expression levels. Kaplan-Meier plots of (A) RFS, (B) OS, and (C) EFS in the primary cohort.

Table 2

Treatment response and survival according to LEF1 expression in the primary cohort

All patients, n = 210
Patients younger than 60 y, n = 109
Patients 60 y and older, n = 101
LEF1highLEF1lowPLEF1highLEF1lowPLEF1highLEF1lowP
CR rate, % 70 57 .09 70 68 > .99 69 46 .03 
RFS   < .001   .01   .004 
    Median RFS, mo 26.8 6.7  nr 9.8  13.5 5.7  
    Estimated RFS at 3 y, % (95% CI) 46 (36-59) 22 (14-36)  61 (48-79) 32 (20-52)  29 (17-49) 8 (2-31)  
OS   < .001   .006   < .001 
    Median OS, mo 36.2 8.6  75.8 10.7  22.8 7.8  
    Estimated OS at 3 y, % (95% CI) 51 (42-62) 20 (14-30)  60 (48-74) 27 (17-42)  42 (30-58) 14 (7-28)  
EFS   < .001   .04   < .001 
    Median EFS, mo 10.5 5.5  13.7 7.7  9.9 2.4  
    Estimated EFS at 3 y, % (95% CI) 32 (24-42) 13 (8-21)  43 (31-58) 22 (13-37)  20 (11-35) 4 (0-13)  
All patients, n = 210
Patients younger than 60 y, n = 109
Patients 60 y and older, n = 101
LEF1highLEF1lowPLEF1highLEF1lowPLEF1highLEF1lowP
CR rate, % 70 57 .09 70 68 > .99 69 46 .03 
RFS   < .001   .01   .004 
    Median RFS, mo 26.8 6.7  nr 9.8  13.5 5.7  
    Estimated RFS at 3 y, % (95% CI) 46 (36-59) 22 (14-36)  61 (48-79) 32 (20-52)  29 (17-49) 8 (2-31)  
OS   < .001   .006   < .001 
    Median OS, mo 36.2 8.6  75.8 10.7  22.8 7.8  
    Estimated OS at 3 y, % (95% CI) 51 (42-62) 20 (14-30)  60 (48-74) 27 (17-42)  42 (30-58) 14 (7-28)  
EFS   < .001   .04   < .001 
    Median EFS, mo 10.5 5.5  13.7 7.7  9.9 2.4  
    Estimated EFS at 3 y, % (95% CI) 32 (24-42) 13 (8-21)  43 (31-58) 22 (13-37)  20 (11-35) 4 (0-13)  

CR indicates complete remission; RFS, relapse-free survival; OS, overall survival; CI, confidence interval; EFS, event-free survival; and nr, not reached.

Age is one of the most important risk factors in CN-AML. The impact of molecular markers may vary with age,28  and in our primary cohort, younger (< 60 years) and older patients (≥ 60 years) received different chemotherapy dosages during induction therapy. Therefore, we studied the association of LEF1 expression with outcomes in both age groups separately. Among younger patients, LEF1 expression was not associated with attainment of CR (P > .99; Table 2). However, younger LEF1high patients, compared with younger LEF1low patients, had significantly longer RFS (P = .01; Figure 2A), longer OS (P = .006; Figure 2B), and longer EFS (P = .04; Table 2). Among older patients, those with high LEF1 expression had a higher CR rate (69% vs 46%; P = .03), longer RFS (P = .004; Figure 2C), longer OS (P < .001; Figure 2D), and longer EFS (P < .001) than LEF1low patients (Table 2).

Figure 2

Survival of CN-AML patients according to age group and LEF1 expression. (A) RFS and (B) OS of CN-AML patients younger than 60 years of age. (C) RFS and (D) OS of CN-AML patients aged 60 years or older.

Figure 2

Survival of CN-AML patients according to age group and LEF1 expression. (A) RFS and (B) OS of CN-AML patients younger than 60 years of age. (C) RFS and (D) OS of CN-AML patients aged 60 years or older.

Varying proportions of nonmalignant cells in the BM specimens might confound LEF1 expression measurements and thereby influence the outcome analyses. Exploratory analyses restricted to patients with a pre-Ficoll BM blast count of > 80% (n = 136) revealed that high LEF1 expression was still associated with higher CR rate (P = .008), longer RFS (P < .001), and longer OS (P < .001; data not shown).

Prognostic impact of LEF1 expression in molecular subsets of CN-AML

According to the European Leukemia Net (ELN) reporting system for genetic changes in AML, CN-AML patients are categorized into the ELN Favorable or ELN Intermediate-I genetic category, depending on their CEBPA, NPM1, and FLT3-ITD mutation status.25 LEF1 expression differed between the ELN genetic groups. Forty-eight percent of LEF1high, but only 30% of LEF1low patients belonged to the ELN favorable genetic category (P = .01; Table 1). Therefore, we evaluated the impact of LEF1 expression within the 2 ELN genetic groups separately (supplemental Table 2). In the ELN favorable genetic group, LEF1 expression was not associated with CR rate (P = .30), RFS (P = .59; Figure 3A), or EFS (P = .15). However, OS of ELN favorable/LEF1high patients was significantly longer than for ELN favorable/LEF1low patients (P = .04; Figure 3B). Within the ELN intermediate-I genetic group, LEF1 expression was not associated with CR rate (P = .59). ELN intermediate-I patients with high LEF1 had longer RFS (P = .004; Figure 3C), longer OS (P = .01; Figure 3D), and longer EFS (P = .01) than those with low LEF1.

Figure 3

Survival of CN-AML patients according to ELN genetic group and LEF1 expression. (A) RFS and (B) OS of CN-AML patients in the ELN favorable genetic group. (C) RFS and (D) OS of CN-AML patients in the ELN intermediate-I genetic group.

Figure 3

Survival of CN-AML patients according to ELN genetic group and LEF1 expression. (A) RFS and (B) OS of CN-AML patients in the ELN favorable genetic group. (C) RFS and (D) OS of CN-AML patients in the ELN intermediate-I genetic group.

LEF1 expression associates with favorable RFS, OS, and EFS in multivariable analyses

We performed multivariable analyses to determine the prognostic significance of LEF1 expression after adjusting for the impact of other known risk factors. In a multivariable model for CR achievement, LEF1high patients showed a trend toward higher odds for achievement of remission (P = .08; odds ratio, 1.82; 95% confidence interval, 0.94-3.55). The only factors significantly associated with a higher chance of reaching CR were younger age (P = .04) and presence of an NPM1 mutation (P = .04). In a model for RFS, LEF1high patients had a 50% lower risk of relapse or death compared with LEF1low patients (P = .007; Table 3). Other factors significantly associated with longer RFS were younger age and the NPM1-mutated/FLT3-ITD–negative genotype. In a multivariable model for OS, high LEF1 expression was associated with a 40% reduction of the risk of death (P = .01; Table 3). Other factors associated with longer survival were younger age, the NPM1-mutated/FLT3-ITD–negative genotype, and biallelic CEBPA mutations. Finally, high LEF1 expression was significantly associated with longer EFS (42% risk reduction; P = .003) together with younger age, the NPM1-mutated/FLT3-ITD–negative genotype, and wild-type IDH1 (Table 3).

Table 3

Multivariable analyses in the primary cohort of 210 CN-AML patients

VariableRFS, n = 133
OS, n = 210
EFS, n = 210
HR (95% CI)PHR (95% CI)PHR (95% CI)P
LEF1 expression, high vs low 0.50 (0.30-0.83) .007 0.60 (0.40-0.90) .01 0.58 (0.40-0.83) .003 
Age, per 10-y increase 1.28 (1.08-1.53) .006 1.26 (1.09-1.46) .002 1.31 (1.14-1.49) < .001 
Sex, male vs female 1.13 (0.69-1.86) .62 1.08 (0.73-1.59) .70 1.20 (0.84-1.71) .31 
WBC, logarithmic, per 10-fold increase 1.39 (0.91-2.13) .13 1.33 (0.96-1.87) .09 1.24 (0.92-1.70) .15 
Secondary or therapy-related AML, vs de novo AML nd*  1.43 (0.68-3.02) .34 0.94 (0.45-1.95) .87 
NPM1 mutated/FLT3-ITD negative, vs other genotypes 0.27 (0.14-0.51) < .001 0.43 (0.25-0.73) .002 0.38 (0.24-0.62) < .001 
CEBPA mutation       
    Monoallelic vs absent 0.38 (0.14-1.05) .06 0.52 (0.24-1.13) .10 0.53 (0.26-1.07) .08 
    Biallelic vs absent 0.52 (0.16-1.69) .27 0.38 (0.14-1.00) .05 0.79 (0.37-1.72) .55 
IDH1 mutation, present vs absent 1.91 (0.81-4.49) .14 1.80 (0.90-3.61) .10 1.87 (1.01-3.44) .04 
IDH2 mutation, present vs absent 0.98 (0.49-2.00) .97 0.78 (0.43-1.41) .41 0.90 (0.54-1.51) .69 
MLL-PTD, present vs absent 1.15 (0.51-2.62) .73 1.20 (0.66-2.16) .56 1.38 (0.80-2.37) .24 
ERG expression, high vs low 1.57 (0.86-2.86) .14 1.42 (0.91-2.22) .13 1.42 (0.94-2.14) .10 
BAALC expression, high vs low 0.93 (0.53-1.62) .79 1.28 (0.80-2.03) .29 1.07 (0.71-1.63) .73 
MN1 expression, high vs low 1.59 (0.92-2.77) .10 0.95 (0.61-1.47) .81 1.28 (0.86-1.92) .23 
EVI1 overexpression, present vs absent nd*  1.15 (0.55-2.41) .71 1.23 (0.63-2.40) .54 
VariableRFS, n = 133
OS, n = 210
EFS, n = 210
HR (95% CI)PHR (95% CI)PHR (95% CI)P
LEF1 expression, high vs low 0.50 (0.30-0.83) .007 0.60 (0.40-0.90) .01 0.58 (0.40-0.83) .003 
Age, per 10-y increase 1.28 (1.08-1.53) .006 1.26 (1.09-1.46) .002 1.31 (1.14-1.49) < .001 
Sex, male vs female 1.13 (0.69-1.86) .62 1.08 (0.73-1.59) .70 1.20 (0.84-1.71) .31 
WBC, logarithmic, per 10-fold increase 1.39 (0.91-2.13) .13 1.33 (0.96-1.87) .09 1.24 (0.92-1.70) .15 
Secondary or therapy-related AML, vs de novo AML nd*  1.43 (0.68-3.02) .34 0.94 (0.45-1.95) .87 
NPM1 mutated/FLT3-ITD negative, vs other genotypes 0.27 (0.14-0.51) < .001 0.43 (0.25-0.73) .002 0.38 (0.24-0.62) < .001 
CEBPA mutation       
    Monoallelic vs absent 0.38 (0.14-1.05) .06 0.52 (0.24-1.13) .10 0.53 (0.26-1.07) .08 
    Biallelic vs absent 0.52 (0.16-1.69) .27 0.38 (0.14-1.00) .05 0.79 (0.37-1.72) .55 
IDH1 mutation, present vs absent 1.91 (0.81-4.49) .14 1.80 (0.90-3.61) .10 1.87 (1.01-3.44) .04 
IDH2 mutation, present vs absent 0.98 (0.49-2.00) .97 0.78 (0.43-1.41) .41 0.90 (0.54-1.51) .69 
MLL-PTD, present vs absent 1.15 (0.51-2.62) .73 1.20 (0.66-2.16) .56 1.38 (0.80-2.37) .24 
ERG expression, high vs low 1.57 (0.86-2.86) .14 1.42 (0.91-2.22) .13 1.42 (0.94-2.14) .10 
BAALC expression, high vs low 0.93 (0.53-1.62) .79 1.28 (0.80-2.03) .29 1.07 (0.71-1.63) .73 
MN1 expression, high vs low 1.59 (0.92-2.77) .10 0.95 (0.61-1.47) .81 1.28 (0.86-1.92) .23 
EVI1 overexpression, present vs absent nd*  1.15 (0.55-2.41) .71 1.23 (0.63-2.40) .54 

Multiple imputations using a predictive mean matching algorithm were used in the case of missing covariables.

CN-AML indicates cytogenetically normal acute myeloid leukemia; RFS, relapse-free survival; OS, overall survival; EFS, event-free survival; HR, hazard ratio; CI, confidence interval; WBC, white blood cell count; ITD, internal tandem duplication; nd, not done; and PTD, partial tandem duplication.

*

This variable could not be included in the model for RFS due to the small patient number.

Validation in an independent cohort of CN-AML samples

To validate our findings, we studied an independent cohort of 196 previously untreated younger CN-AML patients provided by the AMLSG. Follow-up for survival was 61.5 months. In the validation cohort, LEF1high patients had a significantly lower WBC, lower incidence of concomitant FLT3-ITD, and trended toward lower BM blast percentage and younger age at diagnosis, compared with LEF1low patients (supplemental Table 3). The CR rate was 80% for LEF1high patients and 69% for LEF1low patients (P = .14; supplemental Table 4). LEF1high patients showed a nonsignificant difference in RFS (P = .18; supplemental Figure 1A) and had a significantly longer OS (P = .02; supplemental Figure 1B) and EFS (P = .04) than LEF1low patients in the validation cohort.

LEF1 expression measured by qPCR discriminates prognostic subgroups in CN-AML

We confirmed the prognostic relevance of LEF1 expression when measured by qPCR in a subgroup of 122 CN-AML patients from the primary cohort with available material (supplemental Table 5). Microarray and qPCR expression measurements showed a highly significant correlation (r = 0.69; P < 1 × 10−15).29  Higher LEF1 expression, measured by qPCR and considered as a continuous variable, associated with higher CR rate (P = .05) and longer RFS (P = .001), OS (P = .005), and EFS (P < .001; supplemental Table 1). To facilitate further validation and potential clinical application of LEF1 expression as a prognostic marker, we identified an informative threshold (a ΔΔCT value of −2.5) for stratifying patients according to LEF1 transcript levels. LEF1high expressers, as defined by qPCR, had a trend toward a higher CR rate (P = .07) and longer RFS (P < .001; Figure 4A), OS (P = .002; Figure 4B), and EFS (P < .001; Figure 4C) than LEF1low expressers (supplemental Table 6). The limited number of patients available for qPCR analysis restricted our ability to study subgroups or evaluate multivariable models. However, in a model for EFS, high LEF1 expression by qPCR significantly associated with favorable outcomes (P = .05) after adjusting for age, sex, WBC, NPM1, and FLT3-ITD mutational status, CEBPA mutations, and ERG, BAALC, and MN1 expression (supplemental Table 7).

Figure 4

Survival of CN-AML patients according to LEF1 expression levels measured by quantitative PCR. (A) RFS, (B) OS, and (C) EFS in 122 patients from the primary cohort with available material for qPCR analysis. A ΔΔCT value of −2.5 was used as the threshold between high and low LEF1 expressers.

Figure 4

Survival of CN-AML patients according to LEF1 expression levels measured by quantitative PCR. (A) RFS, (B) OS, and (C) EFS in 122 patients from the primary cohort with available material for qPCR analysis. A ΔΔCT value of −2.5 was used as the threshold between high and low LEF1 expressers.

High LEF1 expression associates with expression of lymphoid antigens in CN-AML

To gain further insights into biologic differences that are associated with varying LEF1 expression levels in CN-AML patients, we studied global gene expression profiles. We identified a signature of 4958 genes whose expression significantly correlated with LEF1 transcript levels, 1885 genes showing a positive correlation and 3073 genes showing a negative correlation (Figure 5, supplemental Table 8). One of the genes most closely correlated with LEF1 was TCF7, encoding the TCF1 protein, another member of the LEF1/TCF-transcription factor family. Notably, other high-ranking genes showing a strong positive correlation with LEF1 expression encode surface antigens of T lymphocytes, including CD2, CD3, CD5, CD6, CD8, CD247 (the T-cell receptor ζ chain), and IL2RB (the interleukin-2 receptor β chain). These results indicate that high expression of LEF1 is associated with up-regulation of genes linked to lymphoid phenotypic differentiation.

Figure 5

Heatmap image of the 200 genes showing the strongest absolute correlation with LEF1 expression levels. Each column represents 1 of the 210 CN-AML patients, ordered from left to right according to increasing LEF1 expression. LEF1 expression levels are displayed on top of the heatmap. Each row represents 1 gene, ordered from top to bottom according to their Spearman correlation coefficient in descending order.

Figure 5

Heatmap image of the 200 genes showing the strongest absolute correlation with LEF1 expression levels. Each column represents 1 of the 210 CN-AML patients, ordered from left to right according to increasing LEF1 expression. LEF1 expression levels are displayed on top of the heatmap. Each row represents 1 gene, ordered from top to bottom according to their Spearman correlation coefficient in descending order.

Gene set enrichment analysis (GSEA), using a collection of 794 gene sets representing functional biologic pathways, was used to explore biologic mechanisms that might link LEF1 expression to patient outcomes (supplemental Table 9). Eleven gene sets were significantly associated with high LEF1 expression, and most of them are associated with T-lymphocyte differentiation and activation. Expression of 41 gene sets was significantly associated with low LEF1 levels, including gene sets linked to DNA replication, cell-cycle progression, mitosis, and DNA repair. Flow cytometric data from 124 patients confirmed that patients with high LEF1 had a higher percentage of cells staining positive for CD2 (P < .001; Figure 6A) and cytoplasmic CD3 (cyCD3; P < .001; Figure 6B). Expression of CD2 in > 20% of cells was observed in 13% of LEF1high and 6% of LEF1low patients. Expression of cyCD3 in > 10% of cells was observed in 40% of LEF1high patients, compared with 30% of LEF1low patients.

Figure 6

Expression of lymphoid antigens in LEF1high CN-AML. (A) CD2 and (B) cytoplasmic CD3 (cyCD3) expression, assessed by flow cytometry, in patients with low and high LEF1 expression. Each dot represents a patient, and the y-axis shows the percentage of cells within the blast gate staining positive for the respective marker. The red horizontal lines indicate the median for each of group, and the dotted line marks the cutoff for marker expression according to the European Group for Immunophenotypic characterization of Leukemias (EGIL) criteria.

Figure 6

Expression of lymphoid antigens in LEF1high CN-AML. (A) CD2 and (B) cytoplasmic CD3 (cyCD3) expression, assessed by flow cytometry, in patients with low and high LEF1 expression. Each dot represents a patient, and the y-axis shows the percentage of cells within the blast gate staining positive for the respective marker. The red horizontal lines indicate the median for each of group, and the dotted line marks the cutoff for marker expression according to the European Group for Immunophenotypic characterization of Leukemias (EGIL) criteria.

Discussion

Our study is the first report on the prognostic relevance of LEF1 expression in AML, and demonstrates that high LEF1 expression is associated with favorable CR rate, RFS, OS, and EFS in CN-AML. LEF1 is down-regulated in CD34+ cells from patients with myelodysplastic syndromes (MDS) compared with healthy individuals, and lower LEF1 expression associates with increasing BM blast counts and disease progression toward AML.30  In agreement with these results, we found that CN-AML patients with low LEF1 expression had higher WBC and BM blast percentages and more aggressive disease. In our study, high LEF1 expression associated with absence of FLT3-ITD and low ERG expression, which both are favorable molecular characteristics in CN-AML.19,25,26,31 LEF1high CN-AML patients more often belonged to the favorable genetic group as defined by the ELN classification than LEF1low patients, but subgroup analyses showed that high LEF1 expression associated with favorable survival within both genetic subsets. Furthermore, the association of LEF1high status with longer RFS, OS, and EFS was confirmed in multivariable analyses adjusting for the most important clinical and molecular prognosticators in CN-AML. These results indicate that low LEF1 expression is not merely a surrogate marker for other unfavorable genetic lesions such as FLT3-ITD.

Our primary cohort included CN-AML patients across a broad age range, and more than one-half of our patients were aged 60 years or older, the age group where most cases of AML occur.32  Although the prognostic impact of molecular markers can differ between younger and older patients,28 LEF1high status was a favorable prognostic marker in both age groups. High LEF1 expression also associated with longer OS in a second, independent validation cohort of younger CN-AML patients treated on a different protocol and studied by independent investigators. This finding further supports the validity of our results.

Our initial analyses were based on data from gene expression microarrays, which are difficult to use in a clinical setting. However, we were able to reproduce our results when LEF1 expression was measured by real-time qPCR, a routine technique in most diagnostic laboratories. We defined a clinically meaningful threshold for LEF1 expression, which will facilitate further validation of our results by independent investigators and potentially allow future clinical application of this novel genetic marker. Our results suggest that the prognostic impact of LEF1 expression is most pronounced in the ELN intermediate-I genetic group, and thus LEF1 expression may be used to further refine risk stratification for these patients. However, further studies in large and molecularly well-characterized cohorts are needed to develop and validate such an improved risk classification system for CN-AML.

The mechanisms underlying the association between high LEF1 expression and favorable treatment outcomes are unclear. However, our results are of particular interest because LEF1 is an important downstream effector of Wnt signaling, a pathway that is required for self-renewal of normal hematopoietic and leukemic stem cells.2,33,34  Leukemogenic fusion genes and gene mutations can induce Wnt signaling in AML, aberrant activation of the Wnt effector β-catenin has been detected in primary AML samples, and small-molecule Wnt pathway inhibitors are cytotoxic for AML blasts.3,5,35  Our group previously showed that overexpression of Lef-1 in murine BM leads to disturbed hematopoiesis and, ultimately, to the development of myeloid and lymphoid leukemias.10  Interestingly, the myeloid leukemias arising in this model originated from a leukemic stem cell with lymphoid characteristics and showed coexpression of lymphoid markers. In our present study, we analyzed genome-wide gene expression profiles to identify biologic pathways that are associated with LEF1 expression in CN-AML. In line with our observations in mice, LEF1high patients also showed up-regulation of gene sets related to T-lymphoid differentiation. On the other hand, gene sets related to cell proliferation, DNA replication, and DNA repair were down-regulated in LEF1high patients, which might contribute to their favorable outcomes.

Considering reports that linked Wnt pathway activation and high β-catenin expression with inferior patient outcomes,2,36  it might seem counterintuitive that increased levels of the β-catenin interaction partner LEF1 associate with favorable outcomes in CN-AML. There are several possible explanations for our findings: first, the LEF1/TCF-family comprises at least 4 different transcription factors with redundant roles with regard to Wnt signaling.37 LEF1 expression is not closely correlated with Wnt pathway activation as assessed by a LEF1/TCF-reporter assay in primary AML blasts.3  We found that FLT3-ITD mutations associated with low expression of LEF1, while the Wnt pathway generally is activated in such patients.38  Moreover, expression levels of TCF7, another LEF1/TCF protein, showed a much weaker association with patient outcomes than LEF1 expression (data not shown). Therefore, patient outcomes in CN-AML may be specifically associated with LEF1 expression, rather than with deregulation of Wnt pathway activity in general. Second, apart from its involvement in Wnt signaling, LEF1 is also involved in multiple other cellular pathways.8,39 LEF1 is a crucial transcription factor in neutrophilic granulopoiesis. LEF1 expression is low or absent in patients with severe congenital neutropenia, leading to down-regulation of CEBPA and to a block of neutrophilic differentiation.39  Thus, low LEF1 expression may also contribute to the differentiation block in MDS and AML blasts, as reflected by the higher WBC and blast percentages in LEF1low CN-AML and MDS.30  Finally, LEF1 is not only important for granulopoiesis, but is also involved in B- and T-lymphocyte development.40,41  The causes and effects of abnormal LEF1 expression likely depend on the cellular context and differentiation stage. The diverse functions of LEF1 in normal and malignant hematopoiesis are reflected by recent reports that inactivating LEF1 mutations occur in T-cell acute lymphoblastic leukemia,42  while high LEF1 expression is associated with inferior outcomes in B-cell acute lymphoblastic leukemia.43 

In summary, our study for the first time provides evidence that high LEF1 expression associates with favorable outcomes in adult CN-AML patients, even after adjusting for known clinical and molecular risk factors. LEF1 expression can easily be measured by qPCR, and thus may be a valuable new marker for risk stratification of younger and older CN-AML patients. Moreover, our gene expression data from a large cohort of primary CN-AML patients provides insights into the biologic changes associated with varying LEF1 expression levels in AML, and may trigger further mechanistic studies on the role of LEF1 in myeloid leukemias.

Presented in part at the 51st annual meeting of the American Society of Hematology, New Orleans, LA, December 7, 2009.

The online version of this article contains a data supplement.

The publication costs of this article were defrayed in part by page charge payment. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

Acknowledgments

This work was supported by a grant from Deutsche Forschungsgemeinschaft (DFG; SFB 1074, project A4, C.B.) and the Bundesministerium für Bildung und Forschung (NGFN2 grant 01GS0448, C.B.).

Authorship

Contribution: K.H.M., B.H., L.B., and C.B. designed the study; K.H.M., M.C.S., and R.F.S. performed statistical analyses; K.H.M., B.H., R.F.S., L.B., and C.B. wrote the manuscript; and all authors participated in collecting clinical and/or molecular data, helped analyze and interpret data, and reviewed and approved the final version of the manuscript.

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

The current affiliation for B.H. and J.B. is Department of Oncology and Hematology, Hospital Barmherzige Brüder, Regensburg, Germany.

Correspondence: Christian Buske, MD, Institute of Experimental Cancer Research, CCC Ulm, University of Ulm, Albert-Einstein-Allee 11, 89091 Ulm, Germany; e-mail: christian.buske@uni-ulm.de.

References

1
Reya
 
T
Clevers
 
H
Wnt signalling in stem cells and cancer.
Nature
2005
, vol. 
434
 
7035
(pg. 
843
-
850
)
2
Ysebaert
 
L
Chicanne
 
G
Demur
 
C
, et al. 
Expression of beta-catenin by acute myeloid leukemia cells predicts enhanced clonogenic capacities and poor prognosis.
Leukemia
2006
, vol. 
20
 
7
(pg. 
1211
-
1216
)
3
Simon
 
M
Grandage
 
VL
Linch
 
DC
Khwaja
 
A
Constitutive activation of the Wnt/beta-catenin signalling pathway in acute myeloid leukaemia.
Oncogene
2005
, vol. 
24
 
14
(pg. 
2410
-
2420
)
4
Qiang
 
YW
Endo
 
Y
Rubin
 
JS
Rudikoff
 
S
Wnt signaling in B-cell neoplasia.
Oncogene
2003
, vol. 
22
 
10
(pg. 
1536
-
1545
)
5
Müller-Tidow
 
C
Steffen
 
B
Cauvet
 
T
, et al. 
Translocation products in acute myeloid leukemia activate the Wnt signaling pathway in hematopoietic cells.
Mol Cell Biol
2004
, vol. 
24
 
7
(pg. 
2890
-
2904
)
6
Zheng
 
X
Beissert
 
T
Kukoc-Zivojnov
 
N
, et al. 
Gamma-catenin contributes to leukemogenesis induced by AML-associated translocation products by increasing the self-renewal of very primitive progenitor cells.
Blood
2004
, vol. 
103
 
9
(pg. 
3535
-
3543
)
7
Jamieson
 
CH
Ailles
 
LE
Dylla
 
SJ
, et al. 
Granulocyte-macrophage progenitors as candidate leukemic stem cells in blast-crisis CML.
N Engl J Med
2004
, vol. 
351
 
7
(pg. 
657
-
667
)
8
Spaulding
 
C
Reschly
 
EJ
Zagort
 
DE
, et al. 
Notch1 co-opts lymphoid enhancer factor 1 for survival of murine T-cell lymphomas.
Blood
2007
, vol. 
110
 
7
(pg. 
2650
-
2658
)
9
Skokowa
 
J
Cario
 
G
Uenalan
 
M
, et al. 
LEF-1 is crucial for neutrophil granulocytopoiesis and its expression is severely reduced in congenital neutropenia.
Nat Med
2006
, vol. 
12
 
10
(pg. 
1191
-
1197
)
10
Petropoulos
 
K
Arseni
 
N
Schessl
 
C
, et al. 
A novel role for Lef-1, a central transcription mediator of Wnt signaling, in leukemogenesis.
J Exp Med
2008
, vol. 
205
 
3
(pg. 
515
-
522
)
11
Büchner
 
T
Hiddemann
 
W
Berdel
 
WE
, 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
, vol. 
21
 
24
(pg. 
4496
-
4504
)
12
Büchner
 
T
Berdel
 
WE
Schoch
 
C
, et al. 
Double induction containing either two courses or one course of high-dose cytarabine plus mitoxantrone and postremission therapy by either autologous stem-cell transplantation or by prolonged maintenance for acute myeloid leukemia.
J Clin Oncol
2006
, vol. 
24
 
16
(pg. 
2480
-
2489
)
13
Metzeler
 
KH
Hummel
 
M
Bloomfield
 
CD
, et al. 
An 86-probe-set gene-expression signature predicts survival in cytogenetically normal acute myeloid leukemia.
Blood
2008
, vol. 
112
 
10
(pg. 
4193
-
4201
)
14
Schneider
 
F
Hoster
 
E
Unterhalt
 
M
, et al. 
NPM1 but not FLT3-ITD mutations predict early blast cell clearance and CR rate in patients with normal karyotype AML (NK-AML) or high-risk myelodysplastic syndrome (MDS).
Blood
2009
, vol. 
113
 
21
(pg. 
5250
-
5253
)
15
Benthaus
 
T
Schneider
 
F
Mellert
 
G
, et al. 
Rapid and sensitive screening for CEBPA mutations in acute myeloid leukaemia.
Br J Haematol
2008
, vol. 
143
 
2
(pg. 
230
-
239
)
16
Noordermeer
 
SM
Tonnissen
 
E
Vissers
 
I
, et al. 
Rapid identification of IDH1 and IDH2 mutations in acute myeloid leukaemia using high resolution melting curve analysis.
Br J Haematol
2011
, vol. 
152
 
4
(pg. 
493
-
496
)
17
Schlenk
 
RF
Dohner
 
K
Mack
 
S
, et al. 
Prospective evaluation of allogeneic hematopoietic stem-cell transplantation from matched related and matched unrelated donors in younger adults with high-risk acute myeloid leukemia: German-Austrian trial AMLHD98A.
J Clin Oncol
2010
, vol. 
28
 
30
(pg. 
4642
-
4648
)
18
Haferlach
 
T
Kohlmann
 
A
Schnittger
 
S
, et al. 
Global approach to the diagnosis of leukemia using gene expression profiling.
Blood
2005
, vol. 
106
 
4
(pg. 
1189
-
1198
)
19
Metzeler
 
KH
Dufour
 
A
Benthaus
 
T
, et al. 
ERG expression is an independent prognostic factor and allows refined risk stratification in cytogenetically normal acute myeloid leukemia: a comprehensive analysis of ERG, MN1, and BAALC transcript levels using oligonucleotide microarrays.
J Clin Oncol
2009
, vol. 
27
 
30
(pg. 
5031
-
5038
)
20
Kohlmann
 
A
Bullinger
 
L
Thiede
 
C
, et al. 
Gene expression profiling in AML with normal karyotype can predict mutations for molecular markers and allows novel insights into perturbed biological pathways.
Leukemia
2010
, vol. 
24
 
6
(pg. 
1216
-
1220
)
21
Bullinger
 
L
Dohner
 
K
Kranz
 
R
, et al. 
An FLT3 gene-expression signature predicts clinical outcome in normal karyotype AML.
Blood
2008
, vol. 
111
 
9
(pg. 
4490
-
4495
)
22
Edgar
 
R
Domrachev
 
M
Lash
 
AE
Gene expression omnibus: NCVI gene expression and hybridization array data repository.
Nucleic Acids Res
2002
, vol. 
30
 
1
(pg. 
207
-
210
)
23
Beillard
 
E
Pallisgaard
 
N
van der Velden
 
VHJ
, et al. 
Evaluation of candidate control genes for diagnosis and residual disease detection in leukemic patients using ‘real-time’ quantitative reverse-transcriptase polymerase chain reaction (RQ-PCR)–a Europe against cancer program.
Leukemia
2003
, vol. 
17
 
12
(pg. 
2474
-
2486
)
24
Schmittgen
 
TD
Livak
 
KJ
Analyzing real-time PCR data by the comparative CT method.
Nature Protocols
2008
, vol. 
3
 
6
(pg. 
1101
-
1108
)
25
Döhner
 
H
Estey
 
EH
Amadori
 
S
, 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
, vol. 
115
 
3
(pg. 
453
-
474
)
26
Marcucci
 
G
Baldus
 
CD
Ruppert
 
AS
, et al. 
Overexpression of the ETS-related gene, ERG, predicts a worse outcome in acute myeloid leukemia with normal karyotype: a Cancer and Leukemia Group B study.
J Clin Oncol
2005
, vol. 
23
 
36
(pg. 
9234
-
9242
)
27
Korn
 
EL
Censoring distributions as a measure of follow-up in survival analysis.
Stat Med
1986
, vol. 
5
 
3
(pg. 
255
-
260
)
28
Marcucci
 
G
Metzeler
 
KH
Schwind
 
S
, et al. 
Age-related prognostic impact of different types of DNMT3A mutations in adults with primary cytogenetically normal acute myeloid leukemia.
J Clin Oncol
2012
, vol. 
30
 
7
(pg. 
742
-
750
)
29
Dallas
 
PB
Gottardo
 
NG
Firth
 
MJ
, et al. 
Gene expression levels assessed by oligonucleotide microarray analysis and quantitative real-time RT-PCR–how well do they correlate?
BMC Genomics
2005
, vol. 
6
 
1
pg. 
59
 
30
Pellagatti
 
A
Marafioti
 
T
Paterson
 
JC
, et al. 
Marked downregulation of the granulopoiesis regulator LEF1 is associated with disease progression in the myelodysplastic syndromes.
Br J Haematol
2009
, vol. 
146
 
1
(pg. 
86
-
90
)
31
Schlenk
 
RF
Dohner
 
K
Krauter
 
J
, et al. 
Mutations and treatment outcome in cytogenetically normal acute myeloid leukemia.
N Engl J Med
2008
, vol. 
358
 
18
(pg. 
1909
-
1918
)
32
Estey
 
E
Acute myeloid leukemia and myelodysplastic syndromes in older patients.
J Clin Oncol
2007
, vol. 
25
 
14
(pg. 
1908
-
1915
)
33
Wang
 
Y
Krivtsov
 
AV
Sinha
 
AU
, et al. 
The Wnt/beta-catenin pathway is required for the development of leukemia stem cells in AML.
Science
2010
, vol. 
327
 
5973
(pg. 
1650
-
1653
)
34
Reya
 
T
Duncan
 
AW
Ailles
 
L
, et al. 
A role for Wnt signalling in self-renewal of haematopoietic stem cells.
Nature
2003
, vol. 
423
 
6938
(pg. 
409
-
414
)
35
Minke
 
KS
Staib
 
P
Puetter
 
A
, et al. 
Small molecule inhibitors of WNT signaling effectively induce apoptosis in acute myeloid leukemia cells.
Eur J Haematol
2009
, vol. 
82
 
3
(pg. 
165
-
175
)
36
Xu
 
J
Suzuki
 
M
Niwa
 
Y
, et al. 
Clinical significance of nuclear non-phosphorylated beta-catenin in acute myeloid leukaemia and myelodysplastic syndrome.
Br J Haematol
2008
, vol. 
140
 
4
(pg. 
394
-
401
)
37
Clevers
 
H
Wnt/beta-catenin signaling in development and disease.
Cell
2006
, vol. 
127
 
3
(pg. 
469
-
480
)
38
Tickenbrock
 
L
Schwable
 
J
Wiedehage
 
M
, et al. 
Flt3 tandem duplication mutations cooperate with Wnt signaling in leukemic signal transduction.
Blood
2005
, vol. 
105
 
9
(pg. 
3699
-
3706
)
39
Skokowa
 
J
Welte
 
K
LEF-1 is a decisive transcription factor in neutrophil granulopoiesis.
Ann N Y Acad Sci
2007
, vol. 
1106
 (pg. 
143
-
151
)
40
Reya
 
T
O'Riordan
 
M
Okamura
 
R
, et al. 
Wnt signaling regulates B lymphocyte proliferation through a LEF-1 dependent mechanism.
Immunity
2000
, vol. 
13
 
1
(pg. 
15
-
24
)
41
Okamura
 
RM
Sigvardsson
 
M
Galceran
 
J
Verbeek
 
S
Clevers
 
H
Grosschedl
 
R
Redundant regulation of T cell differentiation and TCRalpha gene expression by the transcription factors LEF-1 and TCF-1.
Immunity
1998
, vol. 
8
 
1
(pg. 
11
-
20
)
42
Gutierrez
 
A
Sanda
 
T
Ma
 
W
, et al. 
Inactivation of LEF1 in T-cell acute lymphoblastic leukemia.
Blood
2010
, vol. 
115
 
14
(pg. 
2845
-
2851
)
43
Kühnl
 
A
Gökbuget
 
N
Kaiser
 
M
, et al. 
Overexpression of LEF1 predicts unfavorable outcome in adult patients with B-precursor acute lymphoblastic leukemia.
Blood
2011
, vol. 
118
 
24
(pg. 
6362
-
6367
)

Author notes

*

K.H.M. and B.H. contributed equally to this work.