Frequency and clinical impact of WT1 mutations in the context of CEBPA-mutated acute myeloid leukemia

ABSTRACT Introduction Several studies have confirmed that mutations in the Wilms tumor 1 (WT1) gene occur in adult acute myeloid leukemia (AML). However, few data are available regarding the incidence of WT1 mutations in CEBPAmut AML and their impact. Methods We retrospectively analyzed the frequency and clinical impact of WT1 mutations in 220 newly diagnosed AML patients with CEBPA mutations(CEBPAmut). Chromosome karyotype analysis was performed by R or G banding method and further confirmed either by fluorescence in situ hybridization (FISH) and/or by multiple reverse transcription polymerase chain reaction (multiple RT–PCR). Mutations were detected with a panel of 112mutational genes using next-generation sequencing (NGS). Results Overall, 30 WT1 mutations were detected in 29 of the 220 CEBPAmut AML patients (13.18%) screened. These mutations clustered overwhelmingly in exon 7 (n=16). WT1 mutations were found to be significantly more frequent in AML patients with double-mutated CEBPA (CEBPAdm) than in AML patients with single-mutated CEBPA (17.36%vs. 8.08%, P = 0.043). Among WT1-mutated patients, the most common co-mutation was FLT3-ITD (n = 7, 24.14%), followed by NRAS (n = 5, 17.24%), CSF3R (n = 4, 13.79%), GATA2 (n = 4, 13.79%), and KIT (n = 4, 13.79%). The most frequent functional pathway was signaling pathways inas many as 62.07% of cases. Notably,the concomitant mutations in epigenetic regulatorswere inversely correlated with WT1 mutations(P = 0.003). CEBPAdm AML patients with WT1 mutations had inferior relapse-free survival, event-free survival and overall survival compared with patients CEBPAdm AML without WT1 mutations (P = 0.002, 0.004, and 0.010, respectively). Conclusion Our data showed that WT1 mutations are frequently identified in CEBPAmut AML, especially in CEBPAdm AML. CEBPAmut AML patients with WT1 mutations show distinct spectrum of comutations. In the context of CEBPAdm AML, WT1 mutations predict a poor prognosis.


Introduction
CCAAT/enhancer binding protein a (CEBPA) plays a crucial role in the repression of self-renewal, cell cycle arrest, and myeloid differentiation of mature neutrophils during normal hematopoiesis [1,2]. Mutations in the CEBPA gene can occur across the whole coding region and have been described in approximately 6-24% of all acute myeloid leukemia (AML) patients [3][4][5]. CEBPA-mutated (CEBPA mut ) patients can be subdivided into two different subgroups: (1) AML carrying one mutation on one allele (singlemutated CEBPA, CEBPA sm ) and (2) AML with two CEBPA mutations (double-mutated CEBPA, CEBPA dm ), typically showing an N-terminal mutation and a basic leucine zipper gene mutation. However, only patients who harbor CEBPA double mutations are associated with favorable outcomes [6,7]. In the 2016 World Health Organization (WHO) classification of leukemia, AML with CEBPA dm was included as a distinctive disease entity due to its unique biological and clinical profiles [8].
With the application of next-generation sequencing (NGS), >85% of CEBPA mut AML patients were observed to have other known concurrent mutations, such as WT1, CSF3R, and GATA2 [9,10]. However, studies have shown conflicting data on the impact of CEBPA mutations on prognosis when identified with other concomitant mutations [9][10][11]. These findings suggest that it is necessary to further explore co-mutations in the context of CEBPA mutations in AML.
Recently, mutations in the Wilms tumor 1 (WT1) gene have been observed in adult and childhood leukemia, commonly in AML [12,13]. In several studies investigating more than 1000 AML samples total, mutations in WT1 were found in 2%−8.3% of the patients [12][13][14][15]. However, in AML patients with CEBPA dm , the frequency of WT1mutations increased to 13.6%−18.52% [10,14]. To the best of our knowledge, few data are available regarding the impact of WT1 mutations on CEBPA mut AML. We report here the frequency and types of WT1 mutations and their association with clinical features and outcomesin a cohort of 220 CEBPA mut AML patients.

Patients
A total of 220 selected de novo AML patients with CEBPA mutations from five medical institutions of hematology (Affiliated Changzhou Second Hospital of Nanjing Medical University, Affiliated Hospital of Jiangnan University, Wuxi No. 2 People's Hospital, The Third Affiliated Hospital of Soochow University, and The First Affiliated Hospital of Soochow University) between August 2014 and November 2020 were analyzed. Of the CEBPA mut AML patients, 114 were male and 106 were female. The median age was 39 years (range: 18-88years), with 190 patients being <60 years and 30 patients being ≥60 years. According to the French-American-British (FAB) classification, 14, 96, 33, 62, 5, and 10 patients were diagnosed with M1, M2, M4, M5, M6 and undetermined types, respectively. All CEBPA mut AML patients gave their written informed consent for genetic analysis and for the use of the laboratory results for scientific studies. The study was approved by the research ethics board of each participating hospital and adhered to the tenets of the Declaration of Helsinki.

Cytomorphology, cytogenetics and immunophenotyping
Cytomorphologic assessment was performed by May-Grünwald-Giemsa staining, myeloperoxidase (MPO) reaction, and nonspecific esterase (NSE) staining using alpha-naphthyl-acetate following the FAB and WHO classifications. Karyotyping by the G or Rbanding method was performed following standard methods on bone marrow (BM) cells after short culture. When possible, at least 20 metaphases were analyzed for each case. Immunophenotyping was performed as previously described [16].

Mutation screening
BM smear or peripheral blood (PB) samples at the time of initial diagnosis were collected. A sensitive next-generation amplicon deep-sequencing assay was used with an Illumina next-generation sequencer. A high depth of coverage (1000×) was obtained for 112 genes, including the whole coding regions that are known to be frequently mutated in hematologic malignancies. Altered DNA sequences were deemed mutations or variants in IGV software analysis and were identified by the COSMIC database and needed to exclude dbSNPs. Mutations in NPM1, FLT3, and CEBPAwere identified by polymerase chain reaction (PCR) followed by direct Sanger sequencing, as previously described [17][18][19]. The mutations of the 112 genes analyzed in this study are listed in supplemental Table 1 (Table S1).

Fusion gene and WT1 mRNA expression
Forty-one fusion genes were detected by multiple reverse transcription PCR amplification, which was performed as 8 parallel multiplex reactions on a 7500 Realtime PCR System (Applied Biosystems). We evaluated WT1 mRNA expression in BM or PB using a WT1 mRNA assay kit (Otsuka Pharmaceutical, Tokyo, Japan), as previously reported [20].
Complete remission (CR) was observed in patients after one course of induction chemotherapy. Relapsefree survival (RFS) was defined as the time from diagnosis to relapse or the last follow-up in CR. Event-free survival (EFS) was calculated as the time interval from the date of diagnosis to the date of first evidence of an event, which included relapse, progression, recurrence, change, and death. Overall survival (OS) was defined as the time from diagnosis to death or the last follow-up.

Statistical analysis
Student's t-test was used to analyze continuous factors with a normal distribution, and the Mann-Whitney U test was used for comparisons of data that failed the normality test between different groups. Analysis of categorical variables between different cohorts was performed by the chi-square test and Fisher's exact test. Survival curves were calculated for RFS, OS and EFS based on the Kaplan-Meier method and compared using the two-sided log rank test. Patients who had undergone allogeneic bone marrow transplantation were censored at the time of transplantation. SPSS software version 20.0 (SPSS version 20.0; SPSS Institute, Chicago, IL, USA) was used for statistical analysis, and the results were considered significant at P < 0.05.

Discussion
The WT1 gene, located on chromosome 11p13, encodes a zinc-finger protein that exists in multiple  isoforms and functions as a transcription factor involved in cellular growth, proliferation and differentiation and may therefore act as an oncogene [21][22][23]. The WT1 gene was initially identified as a tumor suppressor gene linked with nephroblastoma [24,25]. Over the past years, numerous studies have highlighted that WT1 is aberrantly expressed or mutated in hematopoietic malignancies, including acute leukemia (AL), myelodysplastic syndrome (MDS) and chronic myelogenous leukemia (CML) [26][27][28][29][30]. However, few data are available regarding the frequency of WT1 mutations and their impact on CEBPA mut AML.
Recently, WT1mutations were detected in 13.6% −18.52% of biallelic CEBPA mut AML patients [3,10,14]. We revealed that WT1 mutations occurred in 17.36% of CEBPA dm AML patients and 8.08% of CEBPA dm AML patients. This percentage of CEBPA dm AML cases is slightly higher than that reported by Krauth MT et al. [14]and Fasan Aet al. [3], but similar to report fromSu L et al. [10]. The difference could be explained by the fact that in the present study, we analyzed WT1 mutations in all exons, while in the study by Krauth MT et al., only mutations in exons 7 and 9 were analyzed. Similar to the findings of Fasan A et al. [3]. We also found that WT1 was significantly more frequently mutated in CEBPA dm patients than in CEBPA sm patients, suggesting that the implication of the WT1 mutation may differ in CEBPA dm and CEBPA sm AML.
Becker H et al. found that in WT1-mutated patients, FLT3-ITD more frequently occurred inolder patients with primary normal karyotype AML (NK-AML) [31]. Krauth MT and colleagues found that WT1 mutations were rarely concomitant with DNMT3A (4.4%, P = 0.014), ASXL1 (1.7%, P = 0.001), IDH2R140 (1.7%, P = 0.001) and IDH1R132 (0.9%,P = 0.001) mutations [14]. In this study, concomitant mutations in epigenetic regulatorswere inversely correlated with WT1 mutations in CEBPA mut AML patients(P = 0.003). This observation suggests that WT1 mutations are anticorrelated with epigenetic regulatormutationsunder the background of CEBPA mut AML. Rampal R et al. observed that WT1-mutant AML patients had reduced 5-  There are still some controversies regarding the prognostic significance of WT1 mutations in patients with AML. Studies from Hou HA and Virappane P demonstrated that the WT1 mutation was associated with poor prognosis in NK-AML and nonselective AML patients [34,35]. Tien Feng-Ming and colleagues found that WT1 mut patients with CEBPA dm AML tended to have a lower CR rate, a higher relapse rate, and significantly shorter OS and disease-free survival (DFS) than those with wild-typeWT1 [36]. Krauth MT et al. indicated thatWT1 mut was associated with shorter EFS in NK-AML patients, but it had no impact on prognosis in subgroups with high WT1 mut incidences (CEBPA dm , PML-RARA) [14]. An analysis based on the TCGA database showed that the WT1mutated group had lower rates of CR but higher rates of minimal residual disease (MRD) after one and two courses of induction chemotherapy. In NK-AML pediatric patients, those with WT1 mutations had significantly worse OS and EFS in both univariate and multivariate survival analyses [37]. In our study, CEBPA mut AMLpatients with WT1 mutations had significantly inferior 3-year RFS, EFS and OS rates than WT1 wt patientsin the context of CEBPA double mutations. This result is consistent with that reported by Tien Feng-Ming et al. but different from that reported by Krauth MT et al. This inconsistency may be attributed to differences in the characteristics of the populations investigated, such as sex, karyotype, age, combination with other mutations such as FLT3-ITD, or different ethnicities.
In conclusion, this study confirms that WT1mutations are frequently identified in CEBPA mut AMLpatients, especially in CEBPA dm AMLpatients. CEBPA mut AML patients with WT1mutations show distinct spectrum of comutations. In the context of CEBPA dm AML, WT1mutations predict a poor prognosiscompared to WT1wild-type patients.  Abbreviations: CR, complete remission; n, numbers; RFS, relapse-free survival; EFS, event-free survival; OS, overall survival; mo, months; NA: not available; P, p-value; NR, not reached; SE, standard error. P-value is from Fisher's exact test for CR rate and from the log-rank test for survival endpoints; Statistically significant results are highlighted in bold font. *, including an early death case during induction.