May the correlation between Kit-D816 mutation and AML1-ETO level change the use of prognostic factors in t(8;21) AML?