The EVI1 gene (3q26) codes for a transcription factor with important roles both in normal development and in leukemogenesis. EVI1 overexpression (OE) through either 3q26 rearrangements or other unknown mechanisms is associated with a particularly unfavorable prognosis in AML. High expression of the splice form EVI1-1D has been reported to be an independent negative prognostic indicator of survival in AML, irrespective of the presence of 3q26 rearrangements. Recently, two studies have provided new information on the prognostic significance of EVI1 splice variants in AML with and without 3q26 lesions, demonstrating that the OE of the EVI1 5′-end variants correlated with significantly reduced event free survival and disease free survival (DFS). However, these results require further validation in other cohorts. Our aim was to validate in a new cohort of 490 patients with AML the prevalence and prognostic significance of EVI1 OE, and to further define the AML subgroup of patients in which it could be useful to distinguish the EVI1+ cases for diagnostic purposes. We first analyzed all EVI1 variants reported in 16 myeloid cell lines and in a series of patients, and confirmed that the EVI1+ cases overexpressed several EVI1 transcripts. In our series, EVI1+ cases always expressed the EVI1-1C transcript. Therefore, we decide to perform the analysis of EVI1- 1C and EVI1-1D in our cohort of 490 well genetically characterized AML patients. All patients were treated with schedules based on anthracycline and cytarabine, as induction therapy. High dose cytarabine, and autologous or allogenic stem cell transplantation when indicated, were used as consolidation therapy. EVI1 was highly expressed in 19% of patients. High EVI1-1D predicted a distinctly worse DFS in the total patient cohort (p=0.0153) and in patients <60 years-old (p=0.007). Multivariate analysis established high EVI1 expression and FLT3-ITD as independent prognostic markers. In our series, 72% (35/48) patients with 3q aberrations had EVI1 OE, including all 27 cases with 3q26 confirmed by FISH. Other cytogenetic aberrations frequently associated with EVI1+ were 11q23 balanced translocations: 73% (11/15), and monosomy 7 as the sole abnormality: 45% (10/22). We did not find EVI1 OE in the 20 cases with trisomy 8. The incidence of EVI1 OE in patients with normal cytogenetics (NK) was 6.5% (11/167), and none of them had ITD-FLT3, suggesting that this could help to define a subgroup of patients with NK and no FLT3 mutations. We found no RUNX1 mutations in the EVI1+ cases analyzed. In conclusion, high expression of EVI1 was significantly associated with shortened DFS in the total patient cohort as well as in the subgroups with intermediate risk or normal cytogenetics. The present study confirms that high levels of EVI1 mRNA 5′-end variants represent an adverse prognostic factor in AML. Our results suggest that the genetic diagnosis of patients with AML and 3q aberrations, normal karyotype, monosomy 7, and 11q23 rearrangements, should include the quantification of EVI1 expression, at least the two splice forms EVI1-1D and EVI1-1C, which would make it possible to detect most EVI1+ cases. These findings suggest that either MLL fusions or monosomy 7 could have a role in the in the regulation of EVI1 expression. Further studies in a larger cohort of patients are needed to confirm the prognostic significance of EVI1+ in patients with these aberrations.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author