Prognostic Impact of FLT3 Mutation Load in NPM1 Mutated AML.

Abstract 826

According to the new WHO classification NPM1 mutated AML is defined as a provisional entity. Those AML without FLT3-ITD (also referred to as FLT3-LM) are regarded as prognostically favorable whereas those with detectable FLT3-ITD in addition are defined as unfavorable. However, some previous studies not taking NPM1 status into account have shown that the prognostic impact of the FLT3 mutation is highly dependent on the load of the mutation. To evaluate a possible impact of FLT3 load in NPM1 mutated AML we have quantified the FLT3 load in 641 NPM1 mutated AML.The cohort was composed of 341 females and 300 males with a median age of 65.2 years (range 17.8-88 years). The NPM1 mutation was detected by melting curve analysis whereas the FLT3-ITD was quantitatively analysed by fragment analysis (Gene Scan, 3130 sequence detection system, ABI). The FLT3-ITD load was quantified as the ratio of the mutation compared to the wildtype allele. All ratios >1 are indicative for allelic loss of the wildtype allele (wt). Overall in 242/641 (37.8%) of the NPM1 mutated AML an FLT3- ITD was detected. The length of the mutation (LM) varied between 9 and 579 bp (median: 51 bp). The ratio of FLT3- ITD was in the range between 0.016 and 44.85 (median: 0.565). Using Cox regression analysis we demonstrated that increasing FLT3-ITD/wt ratio has a significantly unfavorable influence on EFS (p=0.028) . At next, for Kaplan Meier analysis 3 groups were defined according to FLT3-ITD/wt ratio: 1) only FLT3wt (n=398); 2) FLT3-ITD/wt ratio ≤1 (n=189) and 3) FLT3-ITD/wt ratio ≥1 (n=53). It was shown that the FLT3wt group had the best median EFS of 230 days compared to 203 day in group 2 (p=0.032) and only 86 days in group 3 (p<0.001). In a next step the FLT3 ratios were further subdivided into smaller groups 1) only WT (n=398), 2) ratio <0.25 (n=63) 3) ratio > 0.25 but <0.5 (n=44); 4) ratio ≥0.5 but <1 (n=82) 5) ratio ≥1 (n=53). Although the median EFS was continuously decreasing from group 1 to 5 (230 vs 200 vs 198 vs 96 vs 86 days, respectively) a significant difference compared to the FLT3wt group was only detected for group 4 (p=0.014) and group 5 (p<0.001). This analysis demonstrated that only FLT3-ITD/wt ratios ≥0.5 led to inferior outcome in NPM1 mutated AML. The impact of the FLT3/wt ratio on overall survival (OS) could also be detected only for group 5 with a median OS of 142 days compared to 293 days in the FLT3-WT group (p<0.002) and for group 4 (median 147 days) (p=0.033), respectively. In addition other known prognostic factors were analysed for EFS: karyotype (p=0.788; n.s.), gender (p=0.042; better for female) age (p<0.001), WBC (p<0.001), FLT3-status irrespective of ratio (p=0.001) and FLT3-ITD length (p=0.002). In a multivariate analysis only age (p<0.001), WBC (p=0.001) and FLT3-ITD/wt ratio (p=0.042) came out to be independent prognostic factors in NPM1 mutated AML. For OS ratios ≥0.5 came out to be more important than FLT3 status per se (0=0.044 vs. p=0.359). Subsequently ratios '0.5 were multivariately tested against age and WBC. All three parameters came out to be of independent significance for OS (p<0.001, p=0.001, and p=0.008, respectively). In conclusion, this study clearly demonstrates that not FLT3-ITD status per se is predictive for survival in NPM1 mutated AML but ratios of FLT3-ITD load has to be taken into account. Only ratios of ≥0.5 demonstrated to have a significant impact on prognosis in NPM1 mutated AML. This data has enormous implication on clinical decision making in AML including the option of allogeneic transplantation in first CR.