NPM1 Mutations Have a High Impact On the Development of Secondary AML.

Abstract 999

Poster Board I-21

NPM1 mutations are frequently reported to be typical for de novo AML and are regarded as prognostically favorable if not associated with FLT3-ITD. These mutations have rarely been reported in secondary AML after myelodysplastic syndrome (MDS) or after myeloproliferative neoplasms (MPN). We have detected NPM1 mutations in 37/283 patients with AML after a previous MDS (s-AML) (13.1%) and in 6/67 after a previous MPN (9%). Here we describe the characteristics of these 43 NPM1 mutated s-AML cases to show the involvement of NPM1 mutations in development of secondary AML. The total cohort of 43 cases was composed of 22 males and 21 females with a median age of 71.3 years (range: 29.3-87.7 years). Cytogenetics was available in 40 of the 43 cases (93%). 27 of these had a normal karyotpye whereas 13 revealed one of these aberrations: +4 (n=3), t(1;14)(p34;q32) (n=1); -7 (n=1), del(9q) (n=2), +13 (n=1); +21 (n=1), -Y (n=1); i(X)(p10) (n=1), [+1,der(1;13)(q10;q10),+i(5)(p10),+8] (n=1) and a t(5;12)(q33;p13) (n=1). All 43 samples were analysed for MLL-PTD, FLT3-ITD, FLT3-TKD, NRAS, CEBPA, RUNX1 mutations as well as for KITD816 and JAK2V617F mutations. The incidence of additional cooperating mutations was similar to de novo AML. FLT3-ITD was detected in 14/37 AML after MDS (37.8%) and only once (1/6) after MPN. FLT3-TKD was observed in 3/37 case after MDS (8.1%) and never after MPN. In addition there was one case with RUNX1 and 4 cases (10.8%) with NRAS mutation after MDS. In none of the cases a CEBPA mutation or MLL-PTD was observed. Thus a total of 18/37 cases (48.8%) after MDS revealed a further molecular mutation in addition to NPM1. Of those without additional molecular mutations (only NPM1) 4 cases revealed cytogenetic aberrations resulting in 22/37 cases (59.5%) with additional cytogenetic or molecular mutations. Also in the 6 cases with NPM1 after MPN we detected a high proportion of additional mutations. Two of these 6 cases defined to be after MPN had a history of KITD816V mutated mastocytosis. Two further cases had preceding JAK2V617F mutated MPN and one additional carried an ETV6-PDGFRB rearrangement. In all these 5 transformed MPN cases the initial typical MPN mutation was retained in AML (blast crisis) whereas the NPM1 mutation was acquired and may have served as a second hit in the development to AML. One of the two JAK2+/NPM1+ cases in addition also acquired an FLT3-ITD. From 11 of the s-AML cases a paired sample from the timepoint of MDS was available. Retrospectively the NPM1 mutations was retraced by mutation specific realtime PCR and also all other markers were analysed. Three different patterns were observed: 1) in two cases the NPM1 mutation was not detectable in MDS (analysed 35 and 11 months before diagnosis of s-AML). In one case an NPM1/ABL1 level of 1.6% was detectable 6 months after diagnosis of MDS and a level of 2129% eleven months after diagnosis of MDS. 2) In six cases the NPM1 mutation was not detectable with standard methods in MDS, but with sensitive Real time PCR a ratio of 1-4 log below the s-AML level was already detectable 6-17 months before onset of s-AML. 3) In three further cases a high NPM1 level comparable to that in s-AML was already detectable in MDS 2-12 months before s-AML evolved. These three cases gained an FLT3-ITD at the time point of transformation from MDS to AML. These pattern show that NPM1 can be an early or a late event in transformation to s-AML and although the acquisition of mutations seems to be important in the transformation to AML the sequence of the single events seem to be secondary. As NPM1 have a favourable prognosis in de novo AML if not associated with FLT3-ITD we did a respective analysis for overall survival (OS) and (EFS) for our cohort of s-AML after MDS. For this analysis 278 s-AML patients were available: NPM1-/FLT3- (n=223); NPM1+/FLT3- (n=20), NPM1-/FLT3+ (n=20) and NPM1+/FLT3+ (n=12). The total cohort revealed a bad outcome (median OS: 56.6 days and median EFS: 43.5 days; range 2-1049 days for both). The median time for MDS until transformation to AML was 316 days (range: 15-6310 days). No difference with respect to outcome was detected between the four different molecular genetic subgroups. In conclusion, these data 1) show that NPM1 mutations play a major role in the evolution of AML following MDS or MPN. 2) NPM1 mutations can be the first as well as the second hit during transformation. 3) Support the theory of a multistep genetic principle in development of secondary AML. 4) s-AML with a NPM1+/FLT3-ITD- status can not be regarded as prognostically favorable.