Adverse Prognosis in Acute Myeloid Leukemia with Abnormality Abn(3q): Does EVI1 Matter?

Introduction: The ELN classification of cytogenetic aberrations in acute myeloid leukemia (AML) distinguishes favorable risk, intermediate risk I and II and adverserisk. The adverse-risk group contains patients (pts) with inv(3) and t(3;3). These pts have a significantly poorer outcome compared to other cytogenetic aberrations. The MRC classification considers both pts with inv(3) and t(3;3) as well as patients with other abn(3q) as adverse risk, but excludes t(3;5). Pts with inv(3) or t(3;3) have breakpoints located on the long arm of chromosome 3 at q21 and q26. As a result of these chromosomal modifications, an enhancer-protein is deregulated and the stem-cell regulator zinc finger protein EVI1 on 3q26 is over expressed. Other 3q aberrations do not involve EVI1. We conducted a comparative analysis on the impact of abn(3q) with likely EVI1 alteration versus abn(3q) without EVI1 involvement. Analyses were done both in the entire group of abn(3q) pts and in the subgroup of pts treated with allogeneic hematopoietic stem-cell transplantation (HSCT).

Methods: We performed a retrospective analysis on 163 patients with an abnormality on the q arm of chromosome 3 (abn(3q)). These pts were treated between 1996 and 2009 in three multicenter studies by the German SAL study group (AML2003, AML96, AML60+). Pts with t(3;5) were excluded (n=11). The remaining 152 patients were divided into two groups. Group 1 (EVI1) contained 56 patients with a chromosomal aberration likely to alter EVI1, i.e. t(3;3), inv(3) and abn(3)(q26). Group 2 (noEVI1) comprised the remaining 96 patients displaying other abn(3q) aberrations. We compared groups for baseline characteristics, complete remission (CR), relapse-free survival (RFS) and overall survival (OS) in total and stratified for treatment.

Results: Descriptive comparison of the groups (EVI1 vs noEVI1) revealed a significantly higher WBC count (14.3 vs 4.6 Gpt/l), PLT count (62 vs 47 Gpt/l) and -7 incidence (29% vs 16%) in the EVI1 group, whereas in the noEVI1 group, complex aberrations (25% vs 74%) and 17p alterations (0% vs 24%) occurred in a higher proportion of pts. CR rates (52% vs 47%), median RFS (7 vs 6 months) and median OS (6 vs 7 months) did not differ significantly between the two groups. In order to explore the clinical behavior of the different abn(3q) aberrations in relation to allogeneic HSCT, we compared EVI1 pts (n=21) versus noEVI1 pts (n=38) who received an allogeneic HSCT at any time during treatment. Patients with aberrant EVI1 were significantly younger (median age 44 vs 52 years), had a higher incidence of -7 (29% vs 13%), but less frequent karyotype complexity (10% vs 74%) or 17p alterations (0% vs 24%). More patients in the EVI1 group achieved a first CR before HSCT (95% vs 84%). Amongst CR pts, median RFS was slightly higher in the EVI1group (9 vs 6 months). In all abn(3q) pts with allogeneic HSCT, median OS was 30 months in the EVI1 group and only 12.5 months in the noEVI1 group. According to the log-rank test, this difference did not reach statistical significance (p=0.137). The advantage in mean OS for EVI1 patients is most likely due to the higher proportion of patients transplanted in CR while the accumulation of complex karyotypes in the noEVI1 group caused more primary resistant AML cases with a rapid progression even after allogeneic HSCT.

Conclusions: Although AML development may be based on different molecularbiological mechanisms in patients with different abn(3q) aberrations depending on EVI1 alteration, the prognosis of the two groups is very similar. The most likely reason is the equal balance of favorable and adverse prognostic factors between the two groups such as age, karyotype complexity, 17p alteration and -7. Patients of both groups benefit from allogeneic HSCT to a similar extent. Confirmation of these results on larger data sets is desirable and under way.