Wilms tumor 1 gene mutations are associated with a higher risk of recurrence in young adults with acute myeloid leukemia

A Study From the Acute Leukemia French Association

Abstract

BACKGROUND:

Wilms tumor 1 (WT1) is a transcription factor that is overexpressed in most acute myeloid leukemias (AMLs). Recently, 2 groups reported that WT1 mutations occur in approximately 10% of normal karyotype AMLs and are an independent predictor of poor outcome in this subgroup of patients with AML.

METHODS:

The authors studied a cohort of 268 young adults (ages 15-50 years) with AML who were treated on the Acute Leukemia French Association 9802 trial. WT1 exon 7 and 9 mutations were screened retrospectively by polymerase chain reaction and direct sequencing. The patients also were assessed for the presence of the fms-related tyrosine kinase 3 internal tandem duplication (FLT3-ITD), FLT3-D835/I836, nucleophosmin 1 (NPM1), and CCAAT/enhancer binding protein α (CEBPA) mutations.

RESULTS:

WT1 mutations were identified in 14 patients (5%) and were associated with a younger age (P = .02) and an FLT3-ITD (P = .03). No mutation was detected in patients who had favorable cytogenetics. Patients who had WT1 mutations had a shorter overall survival at 4 years (22% vs 56%; P = .01) and a higher risk of recurrence at 4 years (82% vs 46%; P = .0008) compared with patients who had wild-type WT1. Within the subgroup of patients who had normal karyotype AML (n = 106), WT1 mutation was identified as an independent adverse prognostic factor for the risk of recurrence.

CONCLUSIONS:

The current results indicted that WT1 mutations represent an adverse prognostic factor in young adults with AML. Prospective trials should confirm the clinical relevance of WT1 mutations in relation to other prognostic factors in patients with AML. Cancer 2009. © 2009 American Cancer Society.

During the last decade, the clinical outcome of patients with acute myeloid leukemia (AML) has been refined by the identification of several genetic alterations, including nucleophosmin 1 (NPM1) mutations, fms-like tyrosine kinase-3 internal tandem duplications (FLT3-ITD), CCAAT/enhancer binding protein α (CEBPA) mutations or mixed-lineage leukemia partial tandem duplications. The prognostic relevance of these oncogenic events has been assessed mainly in cytogenetically normal AML (CN-AML). In this last group, FLT3-ITD constitutes an adverse prognostic factor, whereas NPM1 or CEBPA mutations are associated significantly with a favorable outcome.1-8

The Wilms tumor 1 (WT1) gene is located at chromosome band 11p13 and consists of 10 exons. The WT1 gene encodes a transcriptional regulator with an N-terminal domain (exons 1 to 6) and a C-terminal domain that contains 4 zinc finger motifs (exons 7 to 10).9, 10WT1 initially was identified as a tumor suppressor gene in Wilms tumor, a pediatric kidney malignancy in which WT1 mutations are identified in approximately 5% to 10% of cases.11 WT1 can be involved in transcriptional activation or repression, depending on the cell type, the WT1 protein isoform, and the target gene. WT1 has a functional duality, because it may behave either as a tumor suppressor gene or as an oncogene. Although WT1 has been implicated in the proliferation and differentiation of hematopoietic cells, to our knowledge its precise role in hematopoiesis and its contribution to leukemogenesis remain uncertain.10, 12-16 The WT1 gene is highly expressed in 70% to 90% of patients with AML and, thus, is a useful marker for monitoring minimal residual disease. The prognostic impact of the WT1 expression level at the time of AML diagnosis has been evaluated in several studies but remains controversial.17-20

WT1 gene mutations in AML were described first more than 10 years ago in approximately 10% to 15% of patients with AML.21-23 Recently, 2 groups reported that WT1 mutations occur in approximately 10% of CN-AMLs and cluster to exons 7 and 9 and that these mutations independently predict poor a outcome in patients with CN-AML.24-26

In the current study, we evaluated the frequency, the main associated features, and the prognostic significance of WT1 mutations in a retrospective cohort of young adult patients with AML who were treated uniformly on the Acute Leukemia French Association 9802 (ALFA-9802) trial.

MATERIALS AND METHODS

Patients and Treatment

We considered patients ages 15 years to 50 years with previously untreated, de novo AML enrolled on the French ALFA-9802 trial. This study was approved by the Ethics Committee of Lyon B-Hotel Dieu Hospital (no. 99017B). Details of the treatment modalities have been reported previously.27 Overall, 405 patients were included in this trial. To perform this retrospective study, we selected a cohort of 268 patients for whom frozen material collected at time of AML diagnosis was available for molecular analysis. This cohort is representative compared with the overall study population in terms of age, sex, white blood cell (WBC) count, French-American-British (FAB) classification, and cytogenetic risk-group allocation. In total, 216 bone marrow (BM) samples and 52 peripheral blood (PB) samples were used. The median leukemic blast percentage at sampling was 66% in BM and 43% in PB (range, 20%-96% and 20%-92% for BM and PB, respectively). The percentage of blasts after enrichment by Ficoll density gradient centrifugation was >60% for all samples that were included in this study. Informed consent was obtained from all patients in accordance with the Declaration of Helsinki. The main characteristics of the patients studied are listed in Table 1.

Table 1. Patient Characteristics and Clinical Outcome According to the Mutational Status of the Wilms Tumor 1 Gene

	No. of Patients (%)
Characteristic	Total, n=268	WT1wt, n=254	WT1mut, n=14 (5%)	P
WT1 indicates Wilms tumor 1; wt, wild-type; mut, mutant; NS, not significant; WBC, white blood cells; FAB, French-American-British; ND, not determined; FLT3, fms-related tyrosine kinase 3; ITD, internal tandem duplication; FLT3D835/I836, FLT3 mutations; NPM1, nucleophosmin 1; CEBPA, CCAAT/enhancer binding protein α; CR, complete remission; RR, risk of recurrence; CI, confidence interval; OS, overall survival.
Median age (range), y	38 (15-50)	39 (15-50)	31 (19-47)	.02
Sex ratio, men:women	0.93	0.93	1.4	NS
Median WBC count (range), ×10^9/L	17 (0.8-273)	17 (0.8-273)	23 (1.5-131)	NS
FAB subtype				NS
M0	7 (3)	7 (3)	0
M1	55 (21)	53 (21)	2 (14)
M2	79 (29)	77 (30)	2 (14)
M4	48 (18)	43 (17)	5 (37)
M5	52 (19)	50 (20)	2 (14)
M6	9 (3)	8 (3)	1 (7)
ND	18 (7)	16 (6)	2 (14)
Cytogenetics				NS
Favorable	51 (19)	51 (20)	0
Intermediate	144 (54)	133 (52)	11 (79)
Unfavorable	56 (21)	53 (21)	3 (21)
Failure	17 (6)	17 (7)	0
Gene mutations
FLT3-ITD	35/251 (14)	30/237 (13)	5/14 (36)	.03
FLT3D835/I836	15/247 (6)	14/233 (6)	1/14 (7)	NS
NPM1 mutation	63/247 (26)	62/233 (27)	1/14 (7)	NS
CEBPA mutation	21/248 (8)	20/234 (9)	1/14 (7)	NS
Outcome
CR	243/268 (91)	232/254 (91)	11/14 (79)	NS
4-y RR [95% CI, %]	47 [41-54]	46 [39-53]	82 [59-100]	.0008
4-y OS [95% CI, %]	54 [48-61]	56 [48-62]	22 [0-47]	.01

Cytogenetic Analysis

Cytogenetic G-banding analysis was performed according to standard methods. The definition of a cytogenetic clone and descriptions of karyotypes followed the International System for Human Cytogenetic Nomenclature.28 Abnormalities were categorized according to the Medical Research Council's classification.29

Gene Mutations Detection

Genomic DNA was extracted from mononuclear cells from diagnostic BM or PB using standard procedures. In this study, WT1 mutational analysis focused on WT1 exons 7 and 9, because these regions have been recognized previously as hot spots for mutations in AML.21 Exons 7 and 9 of WT1 were amplified from genomic DNA by polymerase chain reaction (PCR) using the HotStar hi-fidelity polymerase kit (Qiagen, Valencia, Calif) and the following intronic forward (F) and reverse (R) primers: 7F, 5′-CAGTGCTCACTCTCCCTCAA-3′; 7R, 5′-TAGCAGTGTGAGAGCCTGGA-3′; 9F, 5′-TGCA GACATTGCAGGCATGGCAGG-3′; and 9R, 5′-GCACTATTCCTTCTCTCAACTGAG-3′. The purified PCR products were sequenced directly in both directions using a BigDye Terminator Cycle Sequencing Kit (Applied Biosystems, Foster City, Calif) and were analyzed on the Applied Biosystems 3130xl Genetic Analyzer. Positive controls were included in each series. To evaluate the sensitivity of the technique, DNA isolated in cells from a patient with Denys-Drash syndrome who had a constitutional heterozygous mutation in WT1 exon 9 (ie, 50% of mutated allele) was diluted serially by mixing with DNA isolated from PB lymphocytes from normal donors to make 40%, 30%, 20%, and 10% positive DNA and then was analyzed as described above. The sensitivity level for detecting a WT1 exon 9 mutation was such that at least 20% mutant DNA could be detected confidently in a sample. FLT3-ITD, FLT3-D835/I836, NPM1, and CEBPA mutations were detected as described previously.3, 30-32

Definition of Clinical Endpoints

A complete remission (CR) required BM cellularity ≥20% with maturation in all cell lineages, <5% BM blast cells, no Auer rods, recovery of neutrophils to ≥1500/μL and of platelets to >100,000/μL in PB, and no evidence for circulating blasts and/or extramedullary leukemia, all of which had to persist for at least 1 month. Relapse or recurrence was defined as a reoccurrence of >5% leukemic blasts in BM, the reappearance of circulating blasts, or the development of extramedullary leukemia.33 Overall survival (OS) was calculated from the date of diagnosis until the date of death, and patients who remained alive at last follow-up were censored. The risk of disease recurrence was measured from the date of CR until the date of recurrence, and patients who remained alive at last follow-up or who died in CR were censored.

Statistical Analysis

Patient characteristics and CR rate comparisons were performed using the Fisher exact test for binary variables and the Mann-Whitney test for continuous variables. OS and the risk of disease recurrence were estimated by using the Kaplan-Meier method and were compared using the log-rank test. The effects of continuous variables on OS and risk of recurrence were assessed using a Cox proportional hazards regression. In the subgroup of patients with CN-AML, the variables that were tested in univariate analysis were age, WBC count, molecular risk group (NPM1 mutated/FLT3-ITD negative, or CEBPA mutated/FLT3-ITD negative vs other genotypes), and WT1 mutational status. Variables that were associated significantly with OS or the risk of recurrence in univariate analysis (ie, WBC count, molecular risk group and WT1 mutational status) were used as covariates in multivariate Cox models and were tested by using the likelihood ratio test. A P value ≤.05 was considered to indicate statistical significance. Informed consent was obtained from all patients in this study.

RESULTS

Frequency and Types of WT1 Mutations

Fourteen of 268 patients with AML (5%) harbored a WT1 mutation. Among the 106 patients who had CN-AML, WT1 mutations were detected in 9 patients (8.5%). All WT1 mutations that were detected within this cohort of patients were located in exon 7. Nevertheless, we identified a WT1 exon 9 mutation in 2 patients who were not included in this cohort (1 patient was treated according to the ALFA-9802 trial protocol but was not enrolled on the trial, and another patient was treated on the ALFA-9801 trial, a treatment protocol for AML patients aged 50 to 70 years).

The main clinical and biologic features of the patients who had mutated WT1 are listed in Table 2. With only 1 exception, all of these patients were heterozygous for the mutation. All WT1 exon 7 mutations identified were frame-shift mutations that corresponded mainly to small duplications, deletions, or combined insertions/deletions and were predicted to result in the production of a truncated protein that was missing the normal zinc finger domain (Table 2).

Table 2. Biologic and Clinical Features of Patients With Acute Myeloid Leukemia and Wilms Tumor 1 Mutations*

					WT1 Mutation
Patient	Age at AML Dx, y	Sex	Karyotype	FAB Subtype	DNA Change	Protein Change	Other Gene Mutations Detected	Response to Induction Chemotherapy	Recurrence	Status at Last Follow-up
WT1 indicates Wilms tumor 1; AML, acute myeloid leukemia; Ds, diagnosis; FAB, French-American-British; del, deletion; CR, complete remission; ins, insertion; A, adenine; C, cytosine; T, thymine; G, guanine; dup, duplication; NA, not available; NPM1, nucleophosmin 1; ASCT, allogeneic stem cell transplantation; CR1, first complete remission; FLT3, fms-related tyrosine kinase 3; ITD, internal tandem duplication; +, addition; t, translocation; −, deletion; inv, inversion; der, derivative; CEBPA, CCAAT/enhancer binding protein α. * Sequence numbering is according to Genbank accession number NM_024426 and Swiss-Prot protein number P19544. The sequence variations are designated according to the current recommendations of the Human Genome Variation Society (available at: http://www.hgvs.org/mutnomen/ Accessed on September 3, 2008). The protein changes were deduced theoretically. † The stop codons created by these WT1 exon 7 mutations do not occur until the 5′ part of exon 9. ‡ A CR was achieved after a second cycle of salvage chemotherapy.
1	32	Woman	46,XX[20]	M4	1340_1346del	A314RfsX65*†	None	CR	Yes	Death
2	25	Man	46,XY[25]	M1	1331_1332insACCACTCTTG	V311DfsX9	None	CR	Yes	Death
3	22	Man	46,XY[21]	M6	1305_1336dup	S313VfsX5	None	Failure to achieve CR	NA	Death
4	47	Man	46,XY[20]	M1	1300_1305del insCCCTTCCTATGATTTCTCCT	R301PfsX11	None	CR	Yes	Death
5	22	Woman	46,XX[20]	M2	1337_1340dup	A314VfsX4	NPM1	CR	No	Alive in persistent CR (ASCT in CR1)
6	38	Woman	46,XX[20]	M5	1303_1325del insGCCCCTCTCCCCCGT	R302PfsX12	FLT3-ITD	CR	Yes	Alive in persistent CR (ASCT in CR2)
7	19	Woman	46,XX[20]	M5	1337_1340dup	A314VfsX4	FLT3-ITD	CR	Yes	Death
8	37	Man	46,XY[20]	M2	1334_1338dup	A314GfsX69*	FLT3-ITD	CR	Yes	Death
9	38	Man	46,XY[20]	ND	1304dupC	R302PfsX15	FLT3-ITD	Failure to achieve CR	NA	Alive in persistent CR (ASCT in CR1)‡
10	19	Woman	47,XX,+8[20]	M4	1334-1337dup	A314VfsX4	FLT3-ITD	CR	Yes	Death
11	45	Woman	46,XX,t(3;5)(q25;q34)[20]	M2	1334_1338dup	A314GfsX69*	None	CR	No	Alive in persistent CR
12	29	Man	45,XY,−7[3]/46,XY[27]	M4	1326_1336del insAGT	T309KfsX5	FLT3-D835	CR	Yes	Death
13	23	Man	45,XY,t(3;3)(q21;q26),−7,inv(9)[25]	M4	1337_1340dup	A314VfsX4	None	Failure to achieve CR	NA	Death
14	34	Man	46,XY,del(3)(q?13),der(18)t(3;18)(q?21;q?21)[3]/45,idem,−7[17]	M4	1306dupT	V303CfsX14	C-terminal CEBPA	CR	Yes	Death

Patient Characteristics in Relation to WT1 Mutational Status

Patients who carried WT1 mutations were significantly younger (median age, 31 years vs 39 years; P = .02). No association was observed between the presence of WT1 mutations and FAB subtype or WBC count at diagnosis (Table 1). The karyotype of WT1-mutated AML predominantly was normal (9 of 14 patients; 64%). None of these patients belonged to the favorable cytogenetic risk group defined by the presence of either a t(8;21)(q22;q22) or an inv(16)/t(16,16)(p13;q22) (Tables 1, 2). Patients who had FLT3-ITD significantly more often had WT1 mutations (5 of 35 patients vs 9 of 216 patients; 14% vs 4% [P = .03]). Considering the subgroup of patients with CN-AML, the frequency of NPM1 mutations in patients who had WT1-mutated AML was significantly lower than in patients without WT1 mutation (1 of 9 patients vs 43 of 90 patients; 11% vs 48% [P = .04]). There was no association between CEBPA and WT1 mutations either in the global population (P = .99) or in the CN-AML subgroup (P = .59) (Table 1).

Prognostic Impact of WT1 Mutations in the Whole Cohort of Patients

Of 268 patients, 243 (91%) achieved CR after induction chemotherapy. The CR rate was lower in WT1-mutated patients compared with patients who had the wild-type WT1, but the difference did not reach significance (79% vs 92%; P = .13). However, patients who had WT1-mutated AML appeared to be more resistant to chemotherapy than patients with unmutated AML, because significantly more patients with WT1 mutation failed to achieve CR after 1 cycle of induction chemotherapy (36% vs 14%; P = .05).

The median follow-up for the patients who remained alive was 3.8 years. The estimated OS rate at 4 years was 54%. WT1-mutated patients had a shorter survival (P = .01; 4-year OS rate, 22% vs 56%), as illustrated in Figure 1A. This difference was explained in part by the higher rate of induction failure and also by a higher risk of recurrence associated with WT1 mutation. Among the 243 patients in first CR, the risk of recurrence at 4 years was 47%. This rate increased dramatically in WT1-mutated patients (P = .0008; 4-year risk of recurrence, 82% vs 46%) (Fig. 1B). Among the 14 patients who had WT1 mutations, 12 patients received conventional chemotherapy as first-line therapy, 2 patients underwent allogeneic stem cell transplantation (ASCT) in first CR, and 1 patient underwent ASCT in second CR.

Figure 1. (A) Overall survival and (B) risk of disease recurrence in young adult patients with acute myeloid leukemia according to the mutational status of the wild–type (wt) and mutant (mut) Wilms tumor 1 gene (WT1) are shown.

Prognostic Impact of WT1 Mutations in the Cytogenetically Normal Acute Myeloid Leukemia Subgroup

In the subgroup of patients with CN-AML (n = 106), the presence of a WT1 mutation also was associated with an increased risk of disease recurrence (P = .0006; 4-year risk of recurrence, 86% vs 40%). Although WT1-mutated patients had an inferior survival compared with patients who had the wild-type WT1, the difference did not reach statistical significance (P = .08; 4-year OS rate, 30% vs 57%). In patients with CN-AML, it has been demonstrated that NPM1, CEBPA, and FLT3-ITD mutations are powerful prognostic factors.1-8 In the current analysis, patients who had the genotypes that were NPM1-mutated/FLT3-ITD–negative or CEBPA-mutated/FLT3-ITD–negative were considered to have a favorable molecular risk. This molecular classification (favorable genotypes vs others) was associated strongly with clinical outcome, because patients who had favorable genotypes had a significantly decreased risk of recurrence (P = .02; 4-year risk of recurrence, 41% vs 66%). Among other variables commonly associated with outcome, WBC count (P = .02) also was associated with a higher risk of recurrence, but age was not.

In multivariate analysis that included WBC count, molecular risk group, and WT1 mutational status as covariates, the presence of a WT1 mutation remained an independent adverse prognostic factor for the risk of recurrence (Table 3). Even after adjustment on other covariates, WT1 mutation was not predictive of OS.

Table 3. Results From Univariate and Multivariate Analyses of the Risk of Recurrence in Patients With Cytogenetically Normal Acute Myeloid Leukemia

	Univariate Analysis	Multivariate Analysis
Variable	P	HR	95% CI	P
HR indicates hazards ratio; 95% CI, 95% confidence interval; NPM1, nucleophosmin 1; mut, mutation; CEBPA, CCAAT/enhancer binding protein α; FLT3, fms-related tyrosine kinase 3; ITD, internal tandem duplication; WT1, Wilms tumor 1; WBC, white blood cell. * Continuous variable. † Relative HR ratio for 10 × 10^9/L increment.
(NPM1mut or CEBPAmut) without FLT3-ITD	.02	0.40	0.19-0.81	.01
WT1mut	.0006	3.98	1.59-10.00	.003
WBC count*†	.02	1.12	1.04-1.20	.002
Age*	.18	—	—	—

DISCUSSION

In the current study, we demonstrated the relatively low frequency of WT1 mutations in patients AML and confirmed that these mutations predict a poor outcome in young adults with AML, mainly because of a higher risk of disease recurrence. In approximately 66% of patients, WT1 mutations occurred in CN-AML. In the CN-AML subgroup, WT1 mutations were detected in 8.5% of patients, a frequency comparable to the results from previous studies that were restricted to CN-AML in which WT1 exon 7 and 9 mutations were identified in approximately 10% of patients.25, 26 In our cohort of homogeneously treated patients, we identified only WT1 exon 7 mutations. In previous studies performed in larger cohorts of adult patients aged <60 years with CN-AML, WT1 exon 9 mutations were identified in only 1% (6 of 470) and 3% (6 of 196) of patients, respectively.25, 26 The finding that WT1 exon 9 mutations are less frequent than WT1 exon 7 mutations may explain the absence of WT1 exon 9 mutations in this study. In addition, the use of direct sequencing for WT1 mutations screening may cause a lower detection rate of mutations compared with more sensitive techniques and potentially may give rise to false-negative results. With regard to WT1 exon 9 mutations, the frequency of identified mutations was lower in the study by Virappane et al, who used direct sequencing,26 than that identified by Paschka et al, who used denaturing high-performance liquid chromatography.25

We observed a positive association between WT1 and FLT3-ITD mutations that was pointed out previously but did not reach statistical significance in 2 earlier studies.25, 26 Within the subgroup of patients with CN-AML, we observed an inverse correlation between the presence of WT1 and NPM1 mutations, as reported previously.26

The current study also confirmed the negative impact of WT1 mutations on clinical outcome. We observed a higher rate of primary resistance in patients who had WT1-mutated AML, because significantly more WT1-mutated patients failed to achieve CR after 1 cycle of induction chemotherapy. In the study by Virappane et al,26 the response to induction chemotherapy was significantly inferior in patients who had WT1 mutations compared with patients who had wild-type WT1, although that finding was not confirmed by Paschka et al.25 In agreement with those 2 previous studies, which were conducted in cohorts of patients with CN-AML,25, 26 we suggest that the presence of a WT1 mutation is an independent prognostic factor associated with an increased risk of recurrence. However, most likely because of the low number of patients with WT1 mutations, we were not able to correlate the presence of these mutations with OS in the subgroup of patients with CN-AML.

Considering the low incidence of WT1 mutations and the frequent association with other adverse prognostic factors (high-risk cytogenetics, FLT3-ITD) in approximately 50% of patients, systematic screening for WT1 mutations may be of limited prognostic interest in all patients with AML. However, testing for WT1 mutations in patients who have CN-AML seems important to improve risk stratification and to adjust treatment in this subgroup of patients, as suggested by previous studies that were restricted to CN-AML.25, 26 Prospective trials should confirm the clinical relevance of WT1 mutations in relation to other prognostic factors in AML. It remains unclear how WT1 mutations may cooperate with other gene mutations such as FLT3-ITD to induce leukemia. The relation between WT1 mutation and leukemic cell chemoresistance also remains to be determined. Finally, in addition to their prognostic interest, WT1 mutations represent a candidate for the development of targeted therapeutic approaches.

Conflict of Interest Disclosures

Supported by the Fondation de France (Leukemia Committee) and the North-West Canceropole (Onco-Hematology Axis).