The many facets of WT1 in acute myeloid leukemia: clarity remains elusive

The Wilms’ tumor gene (WT1) located on chromosome band 11p13 was among the first tumor suppressor genes to be cloned [1]. Originally named for its role in the pediatric kidney malignancy, Wilms’ tumor, it has since been implicated pathogenetically in many other cancers including hematologic malignancies. In acute myeloid leukemia (AML), WT1 appears to play an oncogenic rather than tumor suppressor role [2]. This apparently contradictory biology typifies the complexity of the still emerging WT1 story.

WT1 encodes a zinc finger transcriptional regulator, the function of which is dependent on the cellular context. It can either enhance or inhibit transcription of its target genes which are involved in cellular growth and differentiation. Multiple WT1 protein isoforms are generated through a variety of transcriptional modifications, complicating the assessment of WT1 function considerably.

Acute myeloid leukemia is a heterogeneous disease characterized cytogenetically by recurrent abnormalities which provide powerful prognostic information [3]. Advances in the molecular profiling, particularly of normal karyotype (NK) AML, have led to new paradigms of risk and treatment stratification [4]. The field of genomic interrogation of AML is burgeoning, and promises advances in pathogenetic understanding, identification of new drug targets and refinement of prognosis assessment. Emerging from the global efforts to document individual genomic variation and its influence on disease [5], single nucleotide polymorphisms (SNPs) have been studied in AML, and several have been reported as prognostically significant.

Many facets of WT1 have been investigated in AML, including the prognostic impact of mutations and expression levels at diagnosis; its utility as a marker of minimal residual disease; as a therapeutic target for anti-sense molecules, antibodies and vaccine strategies; and more recently the prognostic relevance of the synonomous WT1 SNP rs16745. This SNP is located in the mutational hotspot exon 7 and results in adenine (A) or guanine (G) containing alleles, the frequency of which varies between ethnic groups. G is the minor allele in Western – and the major allele in Asian – populations [5]. How “silent” SNPs such as this may affect disease susceptibility or treatment outcome has been the subject of much debate. Potential mechanisms include alterations in RNA expression, stability, splicing and binding and changes in translational kinetics, which in turn may impact on drug susceptibility [6]. The WT1 SNP rs16745 has been investigated in several AML cohorts. It does not appear to be a disease-susceptibility SNP, with the frequency in healthy volunteers similar to that of patients with AML [7]. With regard to prognostic impact, however, results have been conflicting. Damm et al. reported improved relapse-free survival (RFS) and overall survival (OS) in younger European adult patients with NK-AML with at least one copy of the minor allele (WT1^GG/AG), and noted this to be an independent favorable-risk marker discriminatory in NPM1/FLT3 high-risk patients [7]. In a predominantly Caucasian pediatric population, Ho et al. reported improved outcome in the same WT1^GG/AG subset, although it was more discriminatory in low-risk patients [8]. In Chinese children with AML, the homozygous WT1^GG genotype predicted better outcomes [9], while Becker et al. found a favorable outcome only in the small subgroup of Western adults with WT1^GG and high-risk FLT3/NPM1 status [10]. In contrast, four other studies in AML found no prognostic impact of this SNP among European children [11] and French [12], Spanish [13] and Korean adults [14].

WT1 mutations were first described in AML almost two decades ago, and occur in approximately 10% of patients with NK-AML. They are postulated to result in dysfunctional protein with respect to DNA binding and transcriptional regulation [15]. The impact of WT1 mutational status on prognosis remains unclear. Initial reports associated mutations with poor outcomes in both adult and pediatric patients [16,17]. Subsequently, however, large well-characterized cohorts were analyzed and no prognostic impact could be identified [15,18]. In several of these studies mutational status was correlated with WT1 expression levels, and no consistent relationship emerged.

Reports on the impact of WT1 expression levels at diagnosis on outcome have also been contradictory. High levels predicted a worse outcome in several studies [19,20], while in others no prognostic significance was identified [7,11]. Conflicting results for the correlation between WT1 expression levels and SNP rs16754 status have also been reported [7,8,11]. The expression level of WT1 has more consistently been shown to be a reliable marker of MRD with clinical utility in predicting outcome [20,21], although the most informative time-points have yet to be standardized.

In this issue of Leukemia and Lymphoma, Luo et al. have correlated WT1 SNP rs16754 status, mutational status and expression levels at diagnosis with outcome in a cohort of 122 adult patients of Chinese origin with non-M3 AML [22]. Their findings further the debate on the utility of these WT1 assays to inform AML prognosis. The SNP analysis confirmed that G is the major allele in this Asian cohort, occurring at a frequency of 65.4% (compared to 13.7% in a German population [7]), and was seen at a similar frequency in healthy controls. Results demonstrated that the WT1^GG was an independent predictor of better disease-free survival (DFS) and OS compared with that of patients with at least one minor (A) allele (WT1^AA/GA) in a multivariate analysis that included FLT3 status but no other molecular markers. This is in agreement with the findings of Damm and Ho [7,8], recognizing, however, that the G allele was the minor allele in those studies, and therefore WT1^GG/AG were pooled together making direct comparisons difficult. The presence of WT1 mutations found in 8.2% of the cohort was not associated with any differences in remission rates, relapse rates or overall survival, adding to the negative studies previously reported. High WT1 levels at diagnosis predicted worse OS and DFS, however were higher in the better-prognosis WT1^GG subgroup. The authors acknowledge that the patient numbers were relatively small, and this study cannot be regarded as conclusive, but adds further data from Asian patients.

The discrepancies in the results between studies of WT1 in AML are difficult to reconcile. The variable prognostic impact of SNP rs16754 is not explained by ethnicity, age or karyotype. The functional impact of SNP rs16754 is not understood, nor the mechanism by which any impact is derived. One hypothesis is that drug sensitivity may be altered either directly through changes in RNA properties or indirectly through association with other SNPs in genes which affect drug metabolism. Thus, different treatment strategies may potentially influence the resultant impact on outcome conferred by SNP rs16754. In search of support for this hypothesis, gene expression profiles of patients with and without a G allele were reported in two studies, with differences implicating RNA metabolism and DNA binding/transcription identified in one [7] but no distinct differences in the other [10]. Further studies into the basic biology of WT1 SNP rs16754 and studies of larger patient cohorts allowing analysis of the impact of each individual genotype are required.

Discrepant findings may result from differences in patient characteristics between cohorts. The complex interactions between WT1 SNP status, mutation status and expression levels and the multitude of other known molecular prognostic factors may confound results if not accounted for in the multivariate analyses. Assay variability may also affect results, as was clearly demonstrated in a comparison between WT1 expression methods [23].

Indeed a myriad of technical, epidemiologic and therapeutic confounders compound the acknowledged complexities of the WT1 gene, its variations and resultant products and functions. Until greater clarity of the basic biology emerges, the many contradictory results may remain unexplained. That WT1 is important in AML is not in question, so in the mean time much larger international studies of molecularly well-characterized patients with AML of both Asian and Caucasian ethnicity should incorporate standardized studies of WT1 to further clarify the clinical utility of its many facets.

Potential conflict of interest:

A disclosure form provided by the author is available with the full text of this article at www.informahealthcare.com/lal.