Non-A type nucleophosmin 1 gene mutation predicts poor clinical outcome in de novo adult acute myeloid leukemia: differential clinical importance of NPM1 mutation according to subtype
- First Online:
- Cite this article as:
- Koh, Y., Park, J., Bae, E. et al. Int J Hematol (2009) 90: 1. doi:10.1007/s12185-009-0350-1
- 156 Views
Mutations of nucleophosmin gene (NPM1) are known to be related to good prognosis in AML patients lacking FLT3 internal tandem duplication (FLT3-ITD). Recently, NPM1 mutations other than type A were reported, but their clinical significance is not well known. Retrospective medical record review of 106 de novo AML patients lacking FLT3-ITD, who received induction chemotherapy from three centers in Korea between 1997 and 2007, was performed. Direct sequencing of NPM1 and RT-PCR for FLT3-ITD was performed on genomic DNA derived from blood samples of patients before induction chemotherapy for detection of mutations. NPM1 mutation was detected in 18 patients, where 13 were type A mutants and 5 were non-type A mutants. CR, CR1-D and OS was not different according to NPM1 mutational status overall. But, non-type A NPM1 mutation was related to shorter CR1-D when compared with NPM1 wild types and NPM1 type A mutation (p = 0.004). OS was shorter in non-type A mutants when compared with NPM1 wild-type patients and NPM1 type A mutants (p = 0.001). The type of mutation of NPM1 is important for prognosis in de novo AML lacking FLT3-ITD. Non-A type NPM1 mutation is a poor prognostic factor.
Clinical characteristics and treatment results of non-A type NPM1 mutants
Subtype of NPM1 mutationa
Mutation sequence (exon 12)
MRC risk group
Initial BM blast %
2 Materials and methods
2.1 Patients and treatment plan
We included adult (age ≥ 15) consecutive de novo AML patients referred to three centers in Korea between 1997 and 2007, who received induction chemotherapy. One hundred and sixty-six patients agreed to donate blood samples for molecular testing on research base with informed consent and were analyzed. Existence of FLT3-ITD was tested and 106 FLT3-ITD-negative patients were finally included in the study. Patients with acute promyelocytic leukemia were excluded. Data regarding characteristics of patients were obtained by medical record review.
Induction regimens consisted of cytarabine and idarubicin. An average of 1.7 cycles ranging from 0 to 3 cycles of consolidation chemotherapy comprised of combinations of idarubicin, daunorubicin, and cytarabine was given to the patients after complete remission (CR). Allogeneic stem cell transplantation was performed in 17 patients after CR according to donor availability and risk stratification .
Risk grouping based on karyotype was done by medical research council (MRC) criteria [7, 8]. Overall survival (OS) was calculated from diagnosis to death from any cause. Duration of first CR (CR1-D) was defined as a period from first successful induction chemotherapy to documentation of disease relapse. The study protocol was approved by the institutional review board of each hospital.
2.2 Molecular study for NPM1 mutation
For NPM1 mutational analysis, PCR was performed in a mixture of 1.25 pmol of each primer NPM1-ex12F (5′-TCTGAGTATAAATTTTCTTGGAGTCA-3′) and NPM1-ex12R (5′-TGGGAACACAGCACTAAATCC-3′), 50 ng genomic DNA, 250 μM dNTPs, and 0.15 U Taq DNA Polymerase (Applied Biosystems, Foster City, CA, USA) in the buffer provided by the manufacturer. Amplification was carried out in a GeneAmp PCR System 9700 thermal cycler (Applied Biosystems, Foster City, CA, USA) under touchdown conditions . The PCR fragments were purified using Millipore PCR Cleanup Filter Plates (Millipore, Billerica, MA, USA). Sequencing reactions were run on an ABI 3730 sequencer with BigDye, Dye Terminator chemistry (Applied Biosystems, Foster City, CA, USA) and an internal sequencing primer NPM_seqR (5′-TTGGACAACACATTCTTGGCAATA-3′). Sequence variants were verified by chromatograms.
2.3 Statistical analysis
The included variables for analysis in this study were age, gender, cytogenetic analysis result, initial white blood cell (WBC) count, platelet (PLT) count, absolute neutrophil count (ANC), bone marrow cellularity, bone marrow blast percentage, chemotherapy regimen, NPM1 mutation status, outcome of induction chemotherapy, CR1-D and OS.
Statistical analysis of 2 × 2 contingency tables of categorical variables were performed using Fisher’s exact test. Median durations of CR1-D and OS were calculated using the Kaplan–Meier method and comparisons between groups were made using log-rank tests. Multivariate analysis was performed using a logistic regression model for response and Cox regression models for CR1-D and OS. All analyses were performed using SPSS for Windows Version 12.0 (SPSS Inc.).
3.1 Patient characteristics and NPM1 mutation status
This study included 43 females and 63 males with a median age of 47.0 (range 16.0–76.5) years. 99 patients had information about karyotype, where 51 with standard 23 with poor risk disease.
NPM1 mutation results were not available in one patient due to low sample quality. NPM1 mutation was detected in 18 patients. 17 patients had previously reported NPM1 mutation, whereas 13 had type A mutation and 4 had B mutation . One patient showed novel CCGA insertion in exon 12 of NPM1, which has not been reported yet and we designated it as Q6 mutation. Overall, five patients were non-A type mutant. NPM1 mutation was more frequently detected in standard risk patients, whereas only one patient had poor risk disease. Ratio of A and non-A type mutation was not different according to risk group.
3.2 Mutation status of NPM1 on treatment outcome
Risk group, NPM1 mutation status was not important for CR. No other single parameter predicted CR. The patient with Q6 mutation did not achieve CR. TRM during first induction chemotherapy was observed in seven patients and among them, two patients had non-A type NPM1 mutants. TRM rate was significantly high in non-A type NPM1 mutants (p = 0.021).
For 81 patients who achieved CR, median CR1-D was 16.8 months. Median CR1-D was not different between NPM1 mutants in general and NPM1 wild type (p = 0.644). But, patients with non-A type NPM1 mutation had significantly shorter median CR1-D when compared with NPM1 wild-type patients (7.0 vs. 16.8 months, p = 0.004), and type A NPM1 mutants (7.0 vs. 23.7 months, p = 0.001). There was no significant difference in CR1-D between patients with wild-type NPM1 and type A NPM1 mutation. In multivariate analysis regarding NPM1 mutation status and risk group, non-A type NPM1 mutation and type A mutation had divergent impact on CR1-D and predicted shorter and longer CR1-D, respectively (p = 0.023 and 0.034, respectively).
Median OS was 19.4 months, and NPM1 mutation in overall was not a prognostic factor (p = 0.969). Non-A type predicted shorter survival such that median OS was shorter in non-A type NPM1 mutants when compared with NPM1 wild-type patients (6.8 vs. 14.8 months, p = 0.001) and type A NPM1 mutants (6.8 vs. 33.8 months, p < 0.001). Lower WBC and higher ANC predicted longer OS (Cox, p = 0.003 and 0.047, respectively). In multivariate analysis, no single factor was associated with OS and non-A type NPM1 mutation predicted shorter OS without reaching statistical significance (p = 0.079).
During follow-up of median 51.7 months, 43 patients experienced relapse and 61 patients died.
In this study, we introduced new insight into the nature of NPM1 mutation for the first time. Before large-scale analyses for confirmation of meaning of numerous molecules in AML, there were anecdotal reports with relatively small number of patients suggesting the existence of such molecules. From this point, overall 106 patients and 18 patients with NPM1 mutations are not small numbers when compared to those anecdotal studies.
Nucleophosmin control of cell growth and proliferation is assumed to be a result of several activities which include interactions with p53 and ARF proteins [11–13]. How NPM1 mutation disrupts this cell control and induces leukemia is not well known. Generally, type A to D mutations of exon 12 in NPM1 are believed to share common mechanism in leukemogenesis, and thus to have a similar clinical impact. For other mutations, the significance is rarely known. In this report, we present those minor NPM1 mutations, B and Q6—other than type A—confer poor prognosis and predict shorter CR1-D in unselected AML patients lacking FLT3-ITD. Patient with type Q6 mutation had especially unfavorable prognosis with OS of 0.67 months, without achievement of CR. Besides, type B mutation also conferred poor prognosis in contrast to conventional conjecture. Interestingly, poor prognosis of non-A type NPM1 mutants was mainly due to treatment-related mortality.
The frequency of NPM1 mutation in our study was 17.0% (18/106), and this is relatively low when compared with previous western studies [1, 2]. However, our study included only FLT3-ITD-negative AML patients and the result concur with the findings of Japanese study which showed low-positive rate (13.2%) of NPM1 mutation in FLT3-ITD negative patients . Similarly, low-positive rate of NPM1 mutation in these two studies may be a reflection of different pathophysiology of AML between Asian and western countries which may be attributed to ethnic diversity.
In contrast to previous studies, our results do not support favorable outcome of NPM1 mutations on FLT3-ITD-negative patients. Only tendency for favorable outcome with type A NPM1 mutation was observed without reaching statistical significance. This may be attributed to the small number of patients, but this may be also due to ethnic diversity. It is well known that characteristics of AML in Asian patients are different from that in western patients. Especially, frequency [15, 16] and prognostic impact of molecular markers can be different . Previous Japanese study failed to show favorable impact of NPM1 mutation in AML patients .
In conclusion, subtypes of NPM1 mutation should be considered while analyzing clinical significance of NPM1 mutations in parallel with karyotype and FLT3-ITD status. Further study with larger number of patients is needed to ascertain the exact clinical meaning of this finding.
This work was supported by a grant from Seoul National University Hospital Research Fund (No. 03-2006-003-0). This work was also supported by a grant from Cancer Research Institute, Seoul National University College of Medicine (No.800-20070121).