1. The WT1 Gene in AML
WT1 was initially identified as a tumor-suppressor gene involved in the pathogenesis of childhood renal Wilms’ tumor
[1]. The gene is located on chromosome 11 (band 11p13) and encodes for a zinc finger DNA-binding protein with four major isoforms, each of which plays a significant role in normal gene function
[2][3][4][5]. Physiologically, WT1 acts as a transcriptional factor, regulating the transcription of growth factors (such as PDGF-A chain, CSF-1, and IGF-II), growth factor receptors (IGF-IR), and other genes (such as RARA, c-myc, bcl-2)
[2][6][7][8][9][10][11]. WT1 can either enhance or repress the expression of its target genes or constructs. In normal human bone marrow, WT1 expression is detected at extremely low levels and is restricted to the primitive CD34+ cell population
[12]. In mouse models, it is thought to be involved in the self-renewal of early hematopoietic cells
[2]. In normal human hematopoietic cells, WT1 appears to be a tumor-suppressor gene; indeed, its overexpression induces growth arrest, reduces colony formation, and promotes spontaneous differentiation
[13][14].
Studies analyzing directly leukemic cells have shown that WT1 is highly expressed in the majority of AMLs, and even in blast crisis of chronic myeloid leukemia, whereas its expression is undetectable in normal blood cells
[15][16][17]. Combining the results of several studies, WT1 RNA levels, as assessed by RT-PCR and Northern blot, were elevated in about 80% of AML patients
[2][16][17][18][19][20][21]. In contrast to normal bone marrow (BM), in AML, WT1 overexpression appears to act as an oncogene and its reduction results in cell death
[2]. Furthermore, its effect seems to be dependent on multiple protein partners such as p53, altering the pro-apoptotic behavior of both proteins, or growth factor signaling proteins such as FLT3, as AML with the FLT3-ITD mutation has been shown to be associated with the highest levels of WT1
[2][22][23]. It is currently unknown what drives WT1 overexpression in AML or if it is an early or late event in the disease onset. Moreover, it is unclear how a pro-apoptotic factor becomes an oncogene as it is unmutated, but it is likely due to a complex pattern of interactions
[2]. Finally, the prognostic significance of WT1 overexpression is also matter of debate as several studies have yielded conflicting results
[21][24][25][26].
2. The Crucial Role of MRD in AML
There is growing evidence that MRD detection is critical for assessing prognosis in AML, particularly in patients undergoing an intensive chemotherapy program
[27]. The MRD monitoring is crucial to guide treatment after complete remission (CR), to define the need for consolidation with allogeneic stem cell transplantation (Allo-SCT), to detect impending relapse allowing early intervention, and to provide reliable post-transplant surveillance
[27].
The MRD working group of the European LeukemiaNet (ELN) recently reviewed the main scientific evidence on MRD monitoring in AML and reached a consensus on thresholds and best timing to detect MRD, and also promoted the standardization of the different methods used
[27][28].
To date, MRD can basically be measured using two methods, namely multiparametric flow cytometry (MFC) and real-time quantitative polymerase chain reaction (qPCR), for specific target genes or fusion transcripts: mutated NPM1, RUNX1-RUNX1T1, CBFB-MYH11, and PML-RARA. MRD surveillance based on next-generation sequencing (NGS) is an emerging and appealing technique but still under development
[28][29].
MFC exploits the leukemia-associated immunophenotype (LAIP). It is applicable to more than 80% of patients, has a fast turnaround time, but is still operator-dependent. The approved cut-off is 0.1% but a MRD quantification below this threshold may be consistent with residual leukemia, and several studies have shown prognostic significance at lower cut-off levels of MFC-MRD
[27][30][31]. The quantitative PCR for mutated NPM1, RUNX1-RUNX1T1, and CBFB-MYH11 is highly standardized and has a higher sensitivity than MFC (detection of one abnormal cell in 10
3–10
6 normal cells), but it can be used in only about 40% of AML patients. The most informative time points are after two cycles (in peripheral blood-PB samples) and at the end of treatment (in bone marrow-BM samples).
3. European LeukemiaNet Standardized Method for Quantitative Evaluation of WT1 Expression
In 2009, ELN researchers validated a quantitative WT1 assay and established reference ranges for WT1 expression in PB and BM analyzing a large number of control samples, to allow transcript levels indicative of residual leukemia to be distinguished from normal background levels
[32]. The selected standardized assay (
Ipsogen WT1 ProfileQuant, QIAGEN) is commercially available and includes exons 1 and 2, which are less prone to mutation than exons 7 and 9 (to reduce false-negative results). The upper normal values were set at 250 WT1 copies/10
4 Abelson (ABL) for BM and at 50 WT1 copies/10
4 ABL for PB, with a sensitivity of 10
−4–10
−5 [32].
4. Role of WT1 Expression Monitoring in Bone Marrow as MRD Marker
All studies reported in this section were performed in patients undergoing an intensive chemotherapy program (such as 3 + 7 or fludarabine-based regimens, followed by cytarabine-based consolidations, with or without allogeneic stem cell transplantation). WT1 overexpression was evaluated using the standardized method of Cilloni et al., but the proposed thresholds of normality in BM samples (less than 250 WT1 copies/10
4 ABL) have not been widely accepted
[32]. In
Table 1, the selected studies together with the cut-off used were summarized in here.
With regard to pre-Allo-SCT setting, Cilloni et al. analyzed 91 patients with WT1 levels >20,000/10
4 ABL at diagnosis and reported that the magnitude of WT1 log reduction after induction chemotherapy provided an independent predictor of relapse in the multivariate analysis, which remained highly significant even when patients were censored at the time of transplantation
[32]. They also showed that detecting WT1 transcripts at levels above the upper limit of the normal value (WT1-MRD positive) at the end of consolidation predicted a significantly increased risk of relapse (67% vs. 42% at 5 years;
p = 0.004).
The second is a study, in which researchers analyzed 122 AML patients overexpressing WT1 at diagnosis and that underwent Allo-SCT in the first CR
[33]. In this study, patients with WT1 levels within the normal range before Allo-SCT (WT1-MRD negative) had improved overall survival—OS (median not reached vs. 9 months,
p < 0.0001) and disease-free survival—DFS (median not reached vs. 8 months,
p < 0.0001) than those WT1-MRD-positive
[33]. In addition, the relapse rate after Allo-SCT was 15% in patients WT1-MRD-negative and 44% in WT1-MRD-positive patients (
p = 0.00073). WT1-MRD negativity was the only independent prognostic factor for improved OS and DFS in this study
[33]. The same prognostic relevance of WT1-MRD negativity before Allo-SCT in a subsequent study performed in FLT3-mutated AML has been confirmed, in which the median OS and DFS in the WT1-MRD-negative group were not reached and were 10.2 and 5.5 months, respectively, in the WT1-MRD-positive group (
p = 0.0005 and
p = 0.0001, respectively)
[34]. It should be noted that, in this study, patients in CR who were WT1-MRD-positive had the same negative outcome as those without morphological CR
[34].
In another study, Frairia et al. analyzed 255 AML patients who were overexpressing WT1 at diagnosis and who were tested for WT1 levels after induction and before Allo-SCT
[35]. The authors reported that the median OS and DFS were significantly shorter in patients with >350 WT1 copies/10
4 ABL after induction than in those with ≤350 WT1 copies (
p = 0.018 and
p = 0.025, respectively). Moreover, patients with WT1 > 150 copies before Allo-SCT had a significantly higher 2-year cumulative incidence of relapse (CIR) compared to those with WT1 ≤ 150 copies (HR 4.61,
p = 0.002)
[35].
Recently, Lambert et al. analyzed 341 AML patients treated in the ALFA-0702 trial, who were overexpressing WT1 at baseline and with available WT1 quantification after induction
[36]. Both BM and PB were tested for WT1, and WT1-MRD positivity was defined when at least one of the two measurements was above the cut-off value (250 WT1 copies/10
4 ABL for BM and at 50 WT1 copies/10
4 ABL for PB). Post-induction WT1-MRD positivity after induction was predictive of subsequent relapse (4-year CIR 29% in WT1-MRD-negative vs. 61% in WT1-MRD-positive,
p < 0.0001), and was an independent factor for CIR in the multivariate analysis
[36]. In addition, at 4 years, relapse free survival—RFS was 60% in WT1-MRD-negative vs. 26% in WT1-MRD positive patients (
p < 0.0001), and OS was 71% vs. 44% (
p = 0.0005). In this study, WT1-MRD positivity remained independently associated with poorer RFS and OS in multivariate analysis, and the unfavorable prognostic significance of WT1-MRD positivity after induction was independent of Allo-SCT
[36].
Finally, Nomdedéu et al. analyzed 584 patients in the CETLAM protocol (365 with available post-induction WT1 measurement and 287 with available post-consolidation WT1 measurement) and divided them into three groups according to post-induction WT1 levels (<17.5, 17.6 to 170.5, and >170.6/10
4 ABL) and post-consolidation WT1 levels (<10, 10.1 to 100, and >100/10
4 ABL)
[37]. The median OS and DFS of the three post-induction groups were 59 and 59 months, 48 and 41 months, and 23 and 19 months, respectively. The median OS and DFS of the three post-consolidation groups were 72 and 65 months, 59 and 46 months, and 30 and 27 months, respectively. All differences between groups were statistically significant. Both post-induction and post-consolidation WT1 levels were significant for DFS and CIR in multivariate analysis
[37].
WT1 has also been studied as an MRD marker in the post-Allo-SCT setting, mainly to identify and potentially treat early relapse. In this study, researchers analyzed 38 AML patients undergoing Allo-SCT with available quantitative WT1 evaluations before and after transplantation
[38]. researchers observed a rapid decline in WT1 expression levels in all patients who achieved or maintained a CR after SCT. All patients who relapsed (13%) had increased WT1 expression at/or before relapse. researchers also found a complete concordance between WT1 expression levels and other MRD markers, when available
[38]. In a subsequent study including 25 AML patients transplanted with reduced-intensity conditioning (RIC-Allo-SCT), researchers reported that cytological relapse was always anticipated by an increase in WT1 levels, and this increase anticipated the loss of molecular chimerism in 50% of the cases
[39].
In another study, Pozzi et al. analyzed 122 AML patients with available WT1 evaluations before and after Allo-SCT, finding a higher relapse rate (54%) in patients with WT1 overexpression (exceeding 100 copies/10
4 ABL) at any time post-SCT, as compared to patients with post-Allo-SCT WT1 expression <100 copies (16%,
p < 0.0001)
[40]. Similarly, the 5-year OS was 40% vs. 63%, respectively (
p = 0.03). In multivariate analysis, WT1 overexpression post-Allo-SCT was the strongest predictor of relapse (HR 4.5,
p = 0.0001)
[40].
Using the same threshold of 100 copies of WT1/10
4 ABL, Nomdedéu et al. reported that patients with <100 WT1 copies at the first evaluation after Allo-SCT had better outcomes in terms of OS, PFS, and CIR
[41]. Additionally, in this study, patients with sustained WT1 levels under 100 copies showed a clear benefit in terms of OS, PFS, and CIR, even compared to patients with just a single measurement over 100 copies
[41].
Finally, Duléry et al. used the standardized thresholds in BM and PB (250 copies/10
4 ABL and 50 copies/10
4 ABL, respectively) to evaluate 139 patients 3 months after Allo-SCT, and they found that WT1-MRD-positive patients at this time point had a poorer CIR (90% vs. 14.7%), EFS (at 3 years 10% vs. 72.3%), and OS (at 3 years 21.4% vs. 75.4%) than WT1-MRD-negative patients
[42].
Table 1. Summary of studies exploring the role of WT1-MRD monitoring in BM.