IDH Mutations Can Be Detected In 28.7% of All Normal Karyotype AML and Have Unfavourable Impact on the NPM1+/FLT3-ITD- Genotype

Mutations in IDH1 and IHD2 have recently been shown to play an important role in AML. As they code for enzymes from the citric acid cycle mutations within these genes from the mechanistical point of view are a totally new kind of mutation associated with AML. In IDH1 one mutational hot spot (amino acid R132) and in IDH2 two hotspots (R140 and R172) have been reported. We aimed at further delineating the impact of IDH1 and IDH2 mutations in AML and analyzed the interaction with other mutations in normal karyotype (NK) AML.

526 AML patients were selected according to normal karyotype and availability of mutational status for FLT3-ITD, NPM1 and MLL-PTD. Further mutation analyses were available in subgroups of the cohort (FLT3-TKD: n=318, CEBPA: n=369, RUNX1: n=174, NRAS: n=220). Female/male ratio was 283/243 and age ranged from 20.0–90.1 years (median, 66.9 years). 435 had de novo AML (82.6%), 71 AML following MDS (s-AML,13.5%) and 20 AML after previous treatment of other malignancies (t-AML, 3.8%). The respective base exchanges in R132, R140, and R172 were analysed by a melting curve assay with subsequent sequencing of the positive samples.

Overall, in 151 pts (28.7%) IDH mutations (IDHmut) were detected. In detail, 68 mutations (12.9% of all cases) were detected in IDH1 (R131C: n=35, R131L: n=17, R131H: n=7, R131G: n=6, R131S: n=3) and 83 mutations (15.8%) in IDH2 (R140Q: n=72, R140L: n=2, R140W: n=1, N141G: n=1, R174K: n=7). IDH1mut and IDH2mut were mutually exclusive in this cohort. IDH1mut were more frequent in females (18.2% vs 8.6 % in males, p=0.001), whereas there was no sex difference for IDH2. According to history IDH1 was equally distributed in de novo AML, s-AML and t-AML whereas IDH2 was more frequent in de novo compared to s- and t-AML (19.6% vs. 7.6 vs 11.8%, p=0.048). According to FAB the most prevalent subtype was FAB M1 with IDHmut in 23.2% compared to 9.8% in all other FAB (in detail: IDH1: 44.8% vs. 23.9%, IDH2: 27.0% vs. 15.1%; p<0.001, for both). IDH1 was underrepresented in M4 (4.9% vs. 15.0 % in all other subtypes, p=0.004), whereas the distribution of IHD2 was not different in M4 vs. all others. The immunophenotype (n= 297) of IDHmut cases tended to be more immature and featured a lower expression of monocytic markers. The analyzed 78 IDHmut cases, as compared to 219 IDHwt cases, showed a significantly higher expression of MPO and CD117 while CD116, CD11b, CD14, CD15, CD36. CD56, CD64, CD65 and CD7 were lower expressed. Age, WBC count, and platelet count were not different between IDH1, IDH2 and IDHwt cases. IDH mutations are not mutually exclusive of other mutations. However, the frequency of CEBPAmut in IDHmut compared to IDHwt was decreased (7.7% vs. 13.7, p=0.001) (IDH1: 0% vs 11.7%, p=0.022 and IDH2: 7.7% vs 13.4%, p=0.053). MLL-PTD was more frequent in IDHmut vs. IDHwt (44.7 vs. 5.8%, p=0.039), however, this is restricted to IDH1mut vs. IDH1wt (26.3 vs. 6.3%, p=0.018). RUNX1mut are distributed equally in IDH2mut and IDH2wt (20.0% vs 27.3%) but are underrepresented in IDH1mut compared to IDH1wt (2.2% vs. 28.7%, p=0.068). FLT3-ITDs are equally distributed between IDHmut and IDHwt, however, those IDH1mut with FLT3-ITD have lower FLT3-ITD/FLT3wt ratios compared to FLT3-ITD+ IDH1wt cases (mean: 0.16 vs. 0.72; p=0.005). All other mutations were distributed equally in IDHmut compared to IDHwt. For survival analysis only cases with de novo AML <65 years were included (n=164, IDHmut: n=37, n=, IDHwt: 127). In the total analysis there was no effect on overall survival or event free survival (EFS). However there was a trend for shorter EFS of the IDHmut vs. IDHwt (median: 439 days vs. not reached, p=0.080) in cases with NPM1+/FLT3-ITD- genotype. For IDH2 there was a significant adverse effect in the NPM1+/FLT3-ITD- group (median EFS: 397 vs. 679 days, p=0.045).

IDH mutations belong to the most frequent mutations in NK AML and can occur together with all other known mutations. There is a high preponderance for the FAB M1 subtype and a more immature immunophenotype for both IDH mutations and a strong female preponderance for IDH1. In addition, an adverse prognostic impact of IDH mutations was shown for the NPM1+/FLT3-ITD- genotype. Further analyses should focus on the definition of the role and place of IDH mutations for therapeutic decisions in patients with AML.

Schnittger:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Alpermann:MLL Munich Leukemia Laboratory: Employment. Kern:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership.