Prognostic Implication of Insertion of FLT3 Internal Tandem Duplication in the BETA-1-Sheet of the Tyrosine Kinase Domain-1

Background: Activating FLT3 internal tandem duplication mutations (FLT3-ITDs) occur in approximately 30% of acute myeloid leukemia (AML) patients. Expression of the FLT3-ITD receptor results in autophosphorylation of FLT3 and subsequent activation of downstream signaling. Clinically, FLT3-ITDs are associated with a dismal clinical outcome; previous explorative analyses suggest that not only FLT3-ITD per se but also the mutant/wild-type allelic ratio and/or the length of the FLT3-ITD provide prognostic information.

Aims: To determine ITD insertion sites and length in FLT3-ITD mutated AML and to correlate these findings with clinical outcome.

Methods: In 241 patients, DNA-based amplification of the FLT3-ITD mutation was followed by DHPLC-based separation of FLT3 mutant and wild-type fragments. Single mutated fragments were collected by a fragment collector, reamplified and sequenced. Patients [16 to 60 years of age] were entered on 3 consecutive AMLSG treatment trials [AML HD93, AML HD98A, AMLSG 07-04] all including intensive therapy.

Results: Thirty-four (14.1) of the 241 patients had more than one ITD (two ITDs n=29, three ITDs n=3, four ITDs n=2). In total, 282 ITDs were analyzed. The median length was 52 nucleotides (range 15–180). For further correlations we grouped ITD integration sites according to the functional regions of FLT3: JM-domain (JMD) [amino acid (AA) 572–603, patients n =141, ITDs n=148], hinge region of JMD [AA 604–609, patients n=45, ITDs n=48], beta-1-sheet of the tyrosine kinase domain-1 (TKD1) [AA 610–615, patients n=69, ITDs n=73], and the remaining region of TKD1 [AA >615, patients n=13, ITDs n=13]. Interestingly, ITD length was strongly correlated to functional regions with shortest ITDs being present in the JMD and longest in the TKD1 (p<0.001). Clinical data were available in 239 patients showing no differences in patient characteristics (age, WBC etc.); frequencies of cooperating class II mutations (NPM1-mut n=137, CEBPA-mut n=12, MLL-PTD n=21) were equally distributed among the functionally categorized groups. The logistic regression model on induction success (IS) revealed ITD integration sites in the beta-1-sheet (odds ratio (OR) 0.25, p=0.01) and in the remaining region of TKD1 (OR 0.14, p=0.007) as well as logarithm of WBC count (OR 0.36, p=0.002) and NPM1 mutations (OR 2.04, p=0.04) as significant variables. ITD insertion in the beta-1-sheet was also significant in the Cox regression analysis on overall survival (OS) (Hazard Ratio (HR) 2.7, p=0.01). In univariable analyses, event free survival (EFS) and OS were significantly inferior in patients with ITD in the beta-1-sheet (p=0.005 and p=0.001). Of note, the proportions of patients receiving an allogeneic transplantation were comparable in both groups (61% and 67%, respectively). In multivariable analyses, neither length of ITD nor mutant/wild-type allelic ratio had an impact on OS.

Conclusions: FLT3-ITD insertion sites seems to be an important prognostic marker for induction failure, EFS and OS. Therefore, not only FLT3-ITD mutation status but also ITD integration site should be prospectively analyzed in future clinical trials, in particular in the context of treatment with FLT3-tyrosine kinase inhibitors.