Mutations in NPM1 (Nucleophosmin-1) have been described in about 35% of adult patients with de novo AML and 45–60 % of AML patients with a diploid karyotype. NPM1 mutations predict for achieving higher complete remission (CR) rates and better outcomes in AML. Few studies have reported on the reliability of mutated NPM1 as a marker for minimal residual disease (MRD) detection in patients with AML.
We conducted a retrospective analysis of patients (n=360) with newly diagnosed AML treated at our institution between 2008 and 2012. The study was approved by the Institutional Review Board. NPM1 mutation status was determined from DNA from unsorted bone marrow (BM) aspirate samples by a PCR-based method at baseline, remission, and relapse. Genomic DNA from bone marrow samples was isolated using the Autopure extractor (QIAGEN/Gentra, Valencia, CA). Mutations in exon 12 of NPM1 were assessed by a DNA-based semi-quantitative polymerase chain reaction capillary electrophoresis (PCR-CE) assay with analytical sensitivity of approximately 2.5%.
Data on remission and relapse samples from 360 newly diagnosed and previously untreated patients with AML, with available NPM1 analysis on their BM at the time of initial diagnosis, was collected (see flow chart below). Median age was 60 years (range 21 – 81 years). 262 patients (72%) had de novo AML, and 98 (27%) secondary AML. Cytogenetics was diploid in 137 (38%) patients, t(8;21) in 17 (5%), inversion 16 in 26 (7%), deletion 5 alone in 27 (7.5%), del 7 and 5 in 26 (7%), deletion 7 alone in 21 (5.8%), trisomy 8 in 17 (5%) and miscellaneous in 89 (24.7%) patients, respectively. Overall, 60 (16.6%) patients including 46 of the 137 (33.5%) diploid patients had NPM1 mutation at baseline. Secondary leukemia was more common in the NPM1 wild type (30%) than in the NPM1 mutated (13%) category.
When analysed by age, in patients < 60 years (n=175), OS (overall survival), EFS (event free survival) and response rates were significantly superior in NPM1 mutated subgroup (p=0.001, 0.007, 0.02 respectively), while among patients ≥ 60 years (n=185) EFS and response rates were significantly higher in the NPM1 mutated subgroup (p=0.008, 0.03 respectively). Among the patients with diploid cytogenetics who were younger than 60 years (n=60) OS, EFS and CR duration was significantly better in the NPM1 mutated subset (p=0.007, 0.007 and 0.02 respectively), while in those ≥60 years (n=77) there was no statistically significant difference in the outcomes for the NPM1 mutated and wild-type subsets.
Among the 60 NPM1 mutated patients 54 (90%) including 41/46 (89%) of those with diploid cytogenetics achieved complete response (CR) or CR without platelet recovery. Thirty nine patients (including 30 with diploid karyotype) had available NPM1 status at the time of CR and all (100%) were negative for NPM1 mutation.
Among the patients with mutated NPM1 at baseline who have achieved a NPM1 negative status at CR, 10/39 overall (25%) and 7/30 diploid (23%) patients relapsed. NPM1 status was available for 6 patients overall including 4 with diploid karyotype at the time of relapse. Among them, 5/6 overall (83%) and 3/4 diploid (75%) patients had mutated NPM1, while 1/6 overall (16%) and 1/4 diploid (25%) patients remained NPM1 wild type. This patient relapsed with extramedullary disease (leukemia cutis) without any BM involvement.
Among the 300 patients (including 91 with diploid karyotype) with wild type NPM1 at diagnosis, none acquired a mutated NPM1 clone, either at CR or at the time of relapse.
These data suggest that mutated NPM1 is a reliable and stable marker for the detection of MRD at the time of CR. Hence, NPM1 mutations can be used to detect MRD and their recurrence may predict pending relapse.
Cortes:Celgene: Research Funding. Ravandi:Johnson and Johnson: Honoraria; Celgene: Research Funding.
Asterisk with author names denotes non-ASH members.