Adverse Prognostic Impact of FLT3 Internal Tandem Duplication (ITD) Is Age-Associated in Older [≥60 Years (Y)] De Novo cytogenetically Normal Acute Myeloid Leukemia (CN-AML) Patients (Pts): a Cancer and Leukemia Group B (CALGB) Study.

Abstract

Abstract 1579▪▪This icon denotes an abstract that is clinically relevant.Poster Board I-605While the prognostic role of FLT3 ITD has been validated in younger CN-AML adults (<60 y), its association with outcome has not been fully investigated in older pts. We studied the frequency and the clinical impact of FLT3 ITD in a cohort of older pts ≥60 y ([n=243; 111 (46%) aged ≥70 y] with de novo CN-AML. Pts were treated on CALGB protocols [10201, 9720, 9420, 8923 and 8525] with cytarabine/anthracycline-based induction and cytarabine-based consolidation. Diagnostic samples were assessed by quantitative fluorescence-based PCR capillary electrophoresis for FLT3 ITD and tyrosine kinase domain (TKD) mutations, and by DHPLC/sequencing analysis for NPM1 and WT1 mutations. Of the 243 pts, 78 (32%) presented with FLT3 ITD, 24 (10%) with FLT3 TKD and 6 (2%) had both. Outcome analyses were restricted to comparison of FLT3 wild-type (WT) pts (n=147) with those with FLT3 ITD alone (n=72). Of these pts, 121 (55%) had NPM1 mutations and 15 (7%) had WT1 mutations. Unfortunately, only a subset of these pts also had available material for BAALC and ERG measurements, thereby preventing analysis of the prognostic significance of FLT3 ITD in the context of a panel of variables that included these markers. Compared with FLT3 WT pts, FLT3 ITD pts had higher WBC counts and % bone marrow and blood blasts (P<.001, all), and higher frequencies of NPM1 (74% v 46%; P<.001) and WT1 mutations (15% v 3%; P=.001). Complete remission (CR) rates were similar, but disease-free (DFS) and overall survival (OS) were shorter in FLT3 ITD v FLT3 WT pts (Table 1). In multivariable analyses (MVA), FLT3 ITD associated with shorter DFS and OS (Table 2). The FLT3 ITD prognostic impact was associated with age. FLT3 ITD pts aged 60-69 y had shorter DFS and OS than FLT3 WT pts, whereas clinical outcomes were not different for ≥70 y FLT3 ITD v FLT3 WT pts (Table 1). In MVA for the 60-69 y subgroup, pts with FLT3 ITD had shorter DFS and OS (Table 2). The reasons for this age-associated effect remain to be explained. In previous studies of younger CN-AML, a higher (≥ median) FLT3 ITD:WT allelic ratio (AR) was associated with worse clinical outcome. In the current study, FLT3 ITD had an adverse prognostic impact on the 60-69 y pts and no significant impact on the ≥70 y pts regardless of the AR levels. However, a 27-microRNA (miR) signature differentiating between FLT3 ITD and FLT3 WT pts and characterized by >2-fold higher miR-155 expression in FLT3 ITD pts, was associated with shorter DFS and OS in the 60-69 y pt subgroup (P=.001, each) but not in the ≥70 y subgroup (P=.26 and P=.89, respectively), suggesting an age-associated prognostic role of the miRs. In summary, our data show FLT3 ITD is an independent marker for poor outcome in CN-AML pts aged 60-69 y but not in those aged ≥70 y. Although the ≥70 y pts with FLT3 ITD had a seemingly better prognosis than the corresponding 60-69 y pts, the outcome for both groups is poor and novel treatment approaches are needed in older pts.Table 1Outcomes in older CN-AML pts with and without FLT3ITDOverall60-69 y Pts≥70 y PtsFLT3 ITD (n=72)FLT3 WT (n=147)PFLT3 ITD (n=41)FLT3 WT (n=78)PFLT3 ITD (n=31)FLT3 WT (n=69)P% achieving CR67%70%.6471%75%.6761%65%.82DFS % disease-free at 3 y10%18%.0077%19%<.00116%18%.94OS % alive at 3 y14%23%<.00110%26%<.00119%20%.71Table 2Variables in Final MVA Models for DFS and OSOverall60-69 y PtsDFSOSDFSOSHR*PHR*PHR*PHR*PFLT3 ITD, positive v negative2.10<.001†1.97<.0012.94<.001†2.79<.001NPM1, mutated v wild-type0.59.0050.54<.001––0.62.021WBC, continuous, 50 unit increase1.44.028†––––––Hemoglobin, continuous1.27.045†––1.50.018†––*HRs <1 (>1) indicate lower (higher) risk for an event for the first category listed for the dichotomous variables and for the higher values of the continuous variables. Variables considered in the models were those significant at á=0.20 in univariable analyses.†Variable did not meet the proportional hazards assumption, a covariate was used to account for time dependence. DisclosuresStone:Cephalon: ad hoc consultancy; Novartis: Research Funding, ad hoc consultancy.