Background: NPM1 mutation is seen in 25 - 30 % of AML, and its incidence does not decrease with older age. While NPM1 mutation is associated with better prognosis in AML patients treated with intensive chemotherapy (IC), it is unknown if it also confers a good prognosis to patients unfit for IC who are treated with HMA.

Methods: 71 AML patients with mutated NPM1 treated by HMA from 2007 to 2016 in 8 European and 1 US centers were retrospectively analyzed. 48 (68%) received azacitidine (AZA), 16 (22%) received decitabine (DAC), and 7 (10%) guadecitabine (SGI 110).All patients were tested for NPM1 and FLT3 mutations.

Results: M/F was 25/46; median age was 75 years (36-85), 76% being 65 or older. 34 patients, considered unfit for IC, received HMA as first line treatment, and 37 after primary or secondary IC failure. Median WBC count at HMA onset was 5.3 x109/L (1.5 - 83). Sixty patients (84%) had normal cytogenetics, 6 had intermediate prognosis abnormalities, one had a complex karyotype, and two had missing data. In addition to NPM1 mutation, 26 patients (37%) had FLT3-ITD, one FLT3-TKD, and 10 (14%) had IDH1 or IDH2 mutation.

The median number of HMA cycles received was 4. The overall response rate (ORR) was 35.1% (n=25), including 16 (22.5%) CR, 5 (7%) CRi, and 4 (5.6%) PR. The number of previous treatments before HMA, the type of HMA used, and baseline WBC had no impact on ORR. The ORR was 45.5% and 27% respectively in patients receiving HMA as first line or subsequent treatment (p=0.137); 45% and 27% respectively in mutated and non-mutated FLT3 (p= 0.302); 20% and 38.3% respectively in mutated and non-mutated IDH1/2 patients (p= 0.314). Among the 26 FLT3-ITD patients, 11 received concomitant therapy with sorafenib, without any impact on ORR or survival (OS), compared to FLT3-ITD patients who did not receive sorafenib (p= 0.73).

Median OS was 280 days, 1 year OS was 17% and only one patient was alive after 2 years, but he died at 24.5 months. We failed to identify any prognostic markers for OS including whether HMA was administered as first line or subsequent treatment (median 289 vs 269 days, p= 0.90), and FLT3-ITD status (median 217 days vs 312 in FLT3-ITD negative, p=0.44). OS was significantly longer in responders (median 553 days vs 123 days, p<10-4). Four of the 6 patients in whom marrow NPM1 MRD was performed achieved a >3log reduction after six cycles of treatment, but 3 of them relapsed within less than six months, while one was alive in CR 12 months after the onset of HMA.

We compared patients of the present series with 92 NPM1 unmutated AML patients previously treated upfront with AZA by our group (Desoutter, Leukemia, 2016). In this cohort, median age was 74 years, 69% had unfavorable cytogenetics, and 63% TP53 mutation. Despite a different molecular profile, there was no OS difference in patients with mutated or non-mutated NPM1 (median 280 vs 291 days, p= 0.53, respectively). This finding was also true when the analysis was restricted to NPM1 mutated patients treated with HMA as first line therapy (Figure 1, p=0.48)

Conclusion: This is to our knowledge the first study reporting the efficacy of HMA in NPM1 mutated AML. Contrary to what is observed with conventional AML chemotherapy, HMA do not seem to improve the outcome of NPM1 mutated compared to unmutated AML cases, and are unable to provide long term remissions. Thus, in elderly AML patients, presence of a NPM1 mutation could contribute to decide the administration of intensive chemotherapy rather than a HMA, when possible.


Recher: Novartis, Celgene, Jazz, Sunesis, Amgen: Consultancy; Celgene, Sunesis, Amgen, Novartis: Research Funding. Fenaux: Astex: Honoraria, Research Funding; Celgene: Honoraria, Research Funding; Janssen: Honoraria, Research Funding; Novartis: Honoraria, Research Funding; Janssen: Honoraria, Research Funding; Amgen: Honoraria, Research Funding; Amgen: Honoraria, Research Funding; Astex: Honoraria, Research Funding; Celgene: Honoraria, Research Funding; Novartis: Honoraria, Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.