Abstract 3075Poster Board III-12The Wilms' tumor 1 (WT1) gene, located at chromosome band 11p13, encodes a transcriptional regulator involved in normal hematopoietic development. WT1 mutations have been identified in approximately 10 % of acute myeloid leukemia (AML), where it has recently been found to predict poor outcome, but also in T-cell acute lymphoblastic leukemias (T-ALL). Our aim was to evaluate the frequency, the main associated features and the prognostic significance of WT1 mutations in a cohort of pediatric patients with T-ALL treated according to EORTC-CLG trials. A total of 146 children, aged 7 months to 17 years, with newly diagnosed T-ALL were included in this study. Patients were treated according to 2 consecutive EORTC trials: 58 881 and 58 951. Immunophenotypic subtypes were classified according to the EGIL. Standard karyotype as well as molecular screening of HOX11/TLX1, HOX11L2/TLX3 and HOXA10 overexpression, SIL-TAL, NUP214-ABL, CALM-AF10 fusions were performed at diagnosis. WT1 transcript level was quantified by real-time PCR (RQ-PCR). Mutations of NOTCH1 exons 26, 27, 34, FBXW7 exons 9, 10, and WT1 exon 7, 9 were screened by direct sequencing. At least one WT1 mutation was found at diagnosis in 15 out of 146 (10%) T-ALL. WT1 mutations were predominantly exon 7 frameshift mutations (14/15 cases), consisting of small duplications, deletions or combined insertions/deletions, and were predicted to result in the production of a truncated protein missing the normal zinc finger domain. The remaining mutated patient harbored a somatically acquired missense mutation in exon 9 (C388Y), previously described in Denys-Drash syndrome. Only 4 out of 15 (27%) patients had 2 WT1 mutations and all WT1 mutations identified showed retention of the wild-type allele. Clonal evolution was investigated by analysis of 12 diagnostic-relapse pairs. Identical WT1 mutation was found at relapse in 3/4 mutated patients whereas 1/4 patients acquired an additional WT1 exon 7 mutation at relapse. One of the 8 patients with WT1 wild-type T-ALL at diagnosis acquired a WT1 exon 7 mutation at relapse. WT1 mutated and wild-type patients did not significantly differ in terms of age, gender, white blood cell count, or mediastinal involvement. Interestingly, WT1 mutated patients had significantly higher WT1 mRNA expression levels (median: 84% [25-837] for WT1 mutants vs 17% [0.007-657] for WT1 wild-type cases, p=0.005). This is in line with the trend for earlier developmental stage arrest observed in our WT1 mutated T-ALL as compared with WT1 wild-type T-ALL. Indeed, WT1 is preferentially expressed in immature hematopoietic progenitors and down-regulated in more differentiated cells. No association was found between the presence of WT1 mutations and NOTCH1 activating lesions. WT1 mutation was associated with HOX genes deregulation. HOX11 or HOX11L2 were overexpressed in 10/15 (67%) WT1 mutated ALL versus 29/123 (24%) WT1 wild-type ALL (p=0.001). In addition, HOXA overexpression and MLL-AF6 were found in one WT1 mutated T-ALL each. Overall, HOX deregulation was demonstrated in 12/15 (80%) WT1 mutated ALL at diagnosis and was also found in the T-ALL that acquired WT1 mutation at relapse. Despite being subclonal lesions strongly associated with HOX11 and HOX11L2 overexpression in T-ALL, WT1 mutations and NUP214-ABL fusion were found independent. A possible impact of WT1 mutation on outcome was investigated. The incidence of very high risk features was similar for patients with WT1 mutated and wild-type T-ALL. No significant differences were found between the WT1 mutated and wild-type group regarding 5-year event free survival (71.6% vs 74.1%; Wald test stratified for protocol: p=0.8) and overall survival (81.8% vs 81.3%; p=0.9). Notably, HOX112 overexpression, which is found in half of WT1 mutated T-ALL, has no pejorative impact either on outcome in EORTC trials. In conclusion, our study confirms that the type and incidence of WT1 mutations are very similar in pediatric T-ALL and AML, although the frequency of bi-allelic alterations may be lower in T-ALL. However, in contrast with AML, no pejorative outcome was associated with WT1 mutation. Moreover, we found that WT1 mutations are highly associated with direct or indirect aberrant HOX genes expression. DisclosuresNo relevant conflicts of interest to declare.