|Home | About | Journals | Submit | Contact Us | Français|
Mutations of fms-like tyrosine kinase 3 (FLT3) and nucleophosmin (NPM1) exon 12 genes are the most common abnormalities in adult acute myeloid leukemia (AML) with normal cytogenetics. To assess the prognostic impact of the two gene mutations in Chinese AML patients, we used multiplex polymerase chain reaction (PCR) and capillary electrophoresis to screen 76 AML patients with normal cytogenetics for mutations in FLT3 internal tandem duplication (FLT3/ITD) and exon 12 of the NPM1 gene. FLT3/ITD mutation was detected in 15 (19.7%) of 76 subjects, and NPM1 mutation in 20 (26.3%) subjects. Seven (9.2%) cases were positive for both FLT3/ITD and NPM1 mutations. Significantly more FLT3/ITD aberration was detected in subjects with French-American-British (FAB) M1 (42.8%). NPM1 mutation was frequently detected in subjects with M5 (47.1%) and infrequently in subjects with M2 (11.1%). FLT3 and NPM1 mutations were significantly associated with a higher white blood cell count in peripheral blood and a lower CD34 antigen expression, but not age, sex, or platelet count. Statistical analysis revealed that the FLT3/ITD-positive group had a lower complete remission (CR) rate (53.3% vs. 83.6%). Survival analysis showed that the FLT3/ITD-positive/NPM1 mutation-negative group had worse overall survival (OS) and relapse-free survival (RFS). The FLT3/ITD-positive/NPM1 mutation-positive group showed a trend towards favorable survival compared with the FLT3/ITD-positive/NPM1 mutation-negative group (P=0.069). Our results indicate that the FLT3/ITD mutation might be a prognostic factor for an unfavorable outcome in Chinese AML subjects with normal cytogenetics, while NPM1 mutation may be a favorable prognostic factor for OS and RFS in the presence of FLT3/ITD.
Acute myeloid leukemia (AML) is a phenotypically and genetically heterogeneous disease. Cytogenetics is regarded as an important prognostic factor for AML patients, which are classified into three risk groups: favorable, intermediate, and unfavorable. Normal cytogenetics is relegated to the intermediate risk group. In recent years, some somatic alterations have been identified in patients with AML (Schichman et al., 1994; Nakao et al., 1996; Pabst et al., 2001; Falini et al., 2005). Most of these abnormalities occur in cytogenetically normal AML (CN AML) and have been confirmed as important prognostic factors. They contribute to dividing CN AML into distinct prognostic subgroups, and represent potential targets for gene therapies. The fms-like tyrosine kinase 3 (FLT3) gene encodes a member of class III receptor tyrosine kinase family, which affects the proliferation, differentiation, and survival of hematopoietic stem cells. Many studies have found that FLT3 internal tandem duplication (FLT3/ITD) mutation is associated with an adverse prognosis in CN AML patients (Yanada et al., 2005; Baldus et al., 2006; Thiede et al., 2006; Colovic et al., 2007). Nucleophosmin (NPM1), as a nucleolar phosphoprotein, usually impacts ribosomal protein assembly and transport, and prevents protein aggregation in the nucleolus. In multivariable analysis, the status of NPM1 mutation without FLT3/ITD is an independent favorable prognostic factor on overall survival (OS) and relapse-free survival (RFS) in patients with CN AML (Döhner et al., 2005).
The objective of this study was to assess the prevalence and prognostic impact of FLT3 and NPM1 gene mutations in adult CN AML patients in China. We also evaluated the association between the two gene mutations and their clinical characteristics, such as age, white blood cell (WBC) count in peripheral blood, and French-American-British (FAB) subtype.
A total of 76 newly diagnosed patients with CN AML (except for FAB M3), who entered the First Affiliated Hospital of Zhejiang University from 2003 to 2005, were investigated in this study. AML was diagnosed according to the FAB classification (Bennett et al., 1985). Cytogenetic G-banding analysis was preformed with standard methods. Mononuclear cells were isolated by Ficoll density gradient centrifugation, and cryopreserved in 10% (v/v) dimethylsulphoxide (DMSO) at −80 °C. The subject characteristics are given in Table Table11.
Characteristics of patients included in this study
Induction therapy consisted of one or two courses of DA (daunorubicin 45 mg/m2 on Days 1 through 3, cytarabine 100 mg/m2 every 12 h on Days 1 through 7) or HAA (homoharringtonine 2 mg/m2 twice daily for 3 d, cytarabine 75 mg/m2 every 12 h on Days 1 through 7, and aclarubicin 12 mg/m2 on Days 1 through 7). Five patients were treated with HA regimen (homoharringtonine 2 mg/m2 twice daily for 3 d, cytarabine 100 mg/m2 every 12 h on Days 1 through 7) because of inferior general conditions. Consolidation therapy was applied every 1 month and consisted of 100 mg/m2 cytarabine every 12 h on Days 1 to 7 in combination with a second drug. The second drug included 45 mg/m2 daunorubicin by intravenous infusion on Days 1 to 3 (Courses 1, 2, 9, etc.), 10 mg/m2 mitoxantrone by intravenous infusion on Days 1 to 3 (Courses 3, 4, 10, etc.), 75 mg/m2 etoposide by intravenous infusion on Days 1 to 5 (Courses 5, 6, 11, etc.), and 12 mg/m2 aclarubicin by continuous infusion over 2 h on Days 1 to 7 (Courses 7, 8, 12, etc.). Patients who failed to obtain complete remission received second line induction regimes, such as mitoxantrone and cytarabine (MA), and aclarubicin, cytarabine and etoposide (AAE). If a cumulative dose of 540 mg/m2 daunorubicin was achieved, thioguanine would take place of daunorubicin.
Genomic DNA was extracted from approximately 106 mononuclear cells (Gentra Puregene Blood DNA kit, Minneapolis, MN, USA). A multiplex polymerase chain reaction (PCR) procedure was used to detect FLT3/ITD and NPM1 mutations (Huang et al., 2008). The 20 µl PCR reaction solution consisted of 200 ng DNA template, 10× PCR buffer 2 µl, 25 mmol/L MgCl2 1 µl, 10 mmol/L deoxyribonucleoside triphosphate (dNTP) 0.5 µl, 200 nmol/L primer for FLT3 and β-globin, 350 nmol/L primer for NPM1, and 1.5 U Taq DNA polymerase (Promega, USA). Samples were amplified using the following PCR conditions: 95 °C for 2 min, 35 cycles at 95 °C for 30 s, 60 °C for 40 s, and 72 °C for 40 s, the final cycle at 72 °C for 30 min.
PCR products were diluted 1:5 (v/v) in distilled water and analyzed using 3130 genetic analyzer (Applied Biosystems, Foster City, CA, USA) according to the manufacturer’s protocol. A peak equal to or above 50 relative fluorescence units (RFU) in the electropherogram was defined as positive. The results were analyzed with GeneMapper software Version 3.2 (Applied Biosystems). The NB4 cell line was used as a normal control, and the sample from a known positive patient was used as a positive control.
To validate the results from capillary gel electrophoresis, 10 FLT3/ITD- and 10 NPM1-mutated samples were amplified for sequencing analysis. PCR products from FLT3/ITD samples were resolved on a 3.5% (w/v) agarose gel stained by ethidium bromide. Each sample displayed an additional PCR product (>330 bp). The longer PCR products were purified by the standard methods and directly sequenced with the same primers used for amplification. PCR products from NPM1 mutant were cloned into pCR2.1-TOPO vector (Invitrogen, Carlsbad, CA, USA). At least four recombinant colonies were selected, and the plasmid DNA was sequenced by the ABI377 sequencer (Applied Biosystems).
The Fisher extract test was used to compare mutation status with dichotomous variables, and the Mann-Whitney U test was used to compare the mutation status with continuous variables. Survival curves for OS and RFS were calculated according to Kaplan-Meier and compared using two-sided log rank test. OS was defined as the time from diagnosis to death owing to any causes. RFS was defined as the time from achieving complete remission (CR) to the first event of either relapse or death. P<0.05 was considered statistically significant. SPSS Version 16.0 software (Chicago, IL, USA) was used for statistical analysis.
The fluorescently labeled multiplexing PCR products from 76 samples were used to screen for the prevalence of FLT3 and NPM1 mutations by capillary gel electrophoresis. Wild type FLT3 and NPM1 genes were 329 and 167 bp (Fig. (Fig.1a).1a). FLT3/ITD-positive samples showed an additional peak at a range from 350 to 410 bp (Fig. (Fig.1b).1b). NPM1 mutation showed a double peak at positions of 167 and 171 bp (Fig. (Fig.1c).1c). The control gene (β-globin), which was used to verify DNA quality and PCR reaction, had a peak around 469 bp. FLT3/ITD mutation was detected in 15 (19.7%) of 76 subjects, and NPM1 mutation in 20 (26.3%) subjects. Seven (9.2%) cases were positive for both FLT3/ITD and NPM1 mutations. In the FLT3/ITD-positive cases, the NPM1 mutation was significantly more frequent than in FLT3/ITD-negative cases (46.7% vs. 21.3%). Ten randomly chosen NPM1 mutants were confirmed by sequencing analysis. NPM1 mutation variant A, a “TCTG” insertion at position nucleotide (wild type) 959 was detected in 8 (80%) cases. The mutation variants B and O M were detected, respectively, in one case.
Capillary gel electrophoresis for FLT3 and NPM1 genes in three cases
Both FLT3 and NPM1 mutations in CN AML subjects were associated with FAB subgroups. Frequencies of NPM1 mutation were significantly higher in FAB M5 (47.1%) than non-M5 (20.0%) (P<0.001), and lower in M2 (11.1%) than non-M2 (34.7%) (P=0.03) (Fig. (Fig.2b).2b). Frequency of FLT3/ITD was significantly higher in FAB M1 (42.8%) than in other FAB subgroups (14.5%) (P<0.001) (Fig. (Fig.2a).2a). FLT3/ITD-positive subjects had statistically higher WBC counts in peripheral blood and blast counts in bone marrow compared with the FLT3/ITD-negative group (P=0.006 and P=0.05, respectively). In the NPM1 mutation group, higher WBC counts were found (mean 81.3×109 L−1, median 54.9×109 L−1) compared with mutation-negative group (mean 64.5×109 L−1, median 34.6×109 L−1, P=0.022). Multivariate analysis showed FLT3/ITD and NPM1 mutations were significantly associated with lower CD34 antigen expression (P=0.029 and P=0.002, respectively). None of the 20 cases of NPM1 mutations showed CD7 aberrant expression, whereas 15 of 56 cases of wide type NPM1 expressed CD7 antigen (P=0.004). The frequencies of myelomonocytic markers, such as CD11, CD13, CD14, and CD15, did not significantly differ between the four groups.
Distribution of FLT3 (a) and NPM1 (b) mutations in morphologic subtypes
The total CR rate after induction chemotherapy was 77.6% in all 76 AML subjects. The FLT3/ITD-positive group, with or without NPM1 mutation, showed a statistically lower CR rate (53.3%) compared to the FLT3 wide type group (83.6%) (χ 2=6.356, P=0.019). The subjects with NPM1 mutation-positive showed a similar CR rate to NPM1 mutation-negative subjects (70.0% and 80.3%, respectively). When the two gene mutations were simultaneously analyzed, the highest CR rate was achieved in subjects with FLT3/ITD-negative/NPM1 mutation-positive (84.6%), followed by the FLT3/ITD-negative/NPM1 mutation-negative group (83.3%). The CR rate in subjects with FLT3/ITD-positive/NPM1 mutated-positive (42.9%) was significantly lower than those in the other three groups (χ 2=7.193, P=0.015) (Table (Table2).2). Subjects harboring FLT3/ITD-negative/NPM1 mutation-positive had a higher CR rate (76.9%), achieved by the first induction therapy, than the other three groups, but these differences had no statistical significance (χ 2=3.425, P>0.05).
Complete remission survival according to FLT3/ITD and NPM1 mutation statuses
The median follow-up time was 20 months. The median duration of remission showed a statistically significant difference between FLT3/ITD-negative and FLT3/ITD-positive subjects (18 months vs. 7 months, P=0.017). FLT3/ITD-positive subjects showed a trend of worse median OS (10.5 months vs. 20 months in FLT3/ITD-negative subjects), but this difference was not significant (P=0.073). No statistical difference on median OS or RFS was found between NPM1-unmutated and NPM1-mutated subjects (Fig. (Fig.3b).3b). The four-year rates of OS and RFS for subjects with FLT3/ITD without NPM1 mutation were zero, and for those without FLT3/ITD were 33.3% and 29.2%, respectively. Kaplan-Meier survival curves of the effects of FLT3/ITD status on OS and RFS showed significantly shorter OS and RFS in subjects harboring FLT3/ITD (log rank=5.729, P=0.017 for OS and log rank=5.488, P=0.019 for RFS) (Fig. (Fig.3a).3a). Survival analysis of the four genotypes revealed a worse OS for the FLT3/ITD-positive/NPM1 mutation-negative group than for both the FLT3/ITD-negative/NPM1 mutation-negative group and the FLT3/ITD-negative/NPM1 mutation-positive group (P=0.002 and P=0.015, respectively), whereas no difference on OS emerged between the FLT3/ITD-positive/NPM1 mutation-negative group and the FLT3/ITD-positive/NPM1 mutated-positive group (P=0.069) (Fig. (Fig.4a).4a). The same pattern was found for RFS (Fig. (Fig.4b).4b). Another factor impacting OS and RFS was the number of induction therapy course required to reach CR. Subjects who achieved CR after more than one course of induction chemotherapy were revealed to have significantly worse OS and RFS (log ranks were 36.284 and 36.556, respectively, P<0.001).
Kaplan-Meier analysis results according to the gene mutation status in CN AML patients
Survival analysis estimates for probabilities of OS (a) and RFS (b) according to the four mutation statuses
FLT3/ITD and NPM1 mutations have been shown to be the most prevalent somatic alterations in AML, especially in CN AML. In our study, the incidences of FLT3/ITD and NPM1 mutations were 19.7% and 26.3%, respectively, in 76 subjects with CN AML, which were obviously lower than those reported in Germans (31%, 53%) (Döhner et al., 2005; Schnittger et al., 2005; Thiede et al., 2006) and in Japanese (28.0%, 47.4%) (Suzuki et al., 2005; Schlenk et al., 2008), but approximated to those reported by Colovic et al. (2007). The lower detection rate may be due to a higher background of wild type allele, or a lower percentage of FLT3/ITD or NPM1 mutation-positive cells in some cases. In addition, FLT3-TKD mutation was difficult to detect by capillary gel electrophoresis.
Certain associations between the two gene mutations and clinical characteristics have been reported in these past years. We also found a significantly increased leukocyte count in subjects with FLT3/ITD (P=0.006) or with NPM1 mutation (P=0.05). There was a significant difference in blast count observed only between FLT3/ITD-positive subjects and FLT3/ITD-negative subjects (P=0.022). These findings were consistent with some previous reports (Fröhling et al., 2002; Colovic et al., 2007; Huang et al., 2008; Schlenk et al., 2008). Although the effect of FLT3/ITD on inducing leukemogenesis was not directly proved, the ligand-independent constitutive activation of FLT3 induced by ITD mutation could activate some downstream signal molecules including mitogen-activated protein (MAP) kinase, signal transducer and activator of transcription 5 (STAT5), and serine-threonine kinases Akt, which contribute to cell proliferation and survival advantages (Hayakawa et al., 2000; Kiyoi et al., 2002; Tse et al., 2002; Grundler et al., 2005; Rocnik et al., 2006). Kelly et al. (2002) induced a myeloproliferative disease in a murine bone marrow transplant model by transforming FLT3/ITD mutants, but did not observe hematologic disorders. These findings might partially explain the close relationship between FLT3/ITD and higher WBC count. CD34 has been regarded as a distinct surface marker on immature hematopoietic precursor cells. As shown previously (Falini et al., 2005; Schnittger et al., 2005), significantly lower CD34 antigen expression was observed in FLT3/ITD- and NPM1-mutated subjects in our present study, but unlike the findings in other studies (Schnittger et al., 2002; Thiede et al., 2002; Colovic et al., 2007), more frequent FLT3/ITD (42.8%) was found in FAB M1 subtype. Zheng et al. (2002; 2004) reported that FLT3/ITD blocked granulocytic differentiation through suppression of CCAAT/enhance binding protein alpha (C/EBPα) expression in 32D cells transfected with FLT3/ITD. These effects could be inhibited by an FLT3 inhibitor, lestaurtinib (CEP-701). The high frequency of FLT3/ITD in M1 subtype might be explained by the theory that FLT3 aberrant activation blocked the differentiation of myeloblastic cells, most probably granulocytic differentiation. Higher blast count, absent CD34 expression in FLT3/ITD, and higher frequency in M1 project the implication that FLT3/ITD may be associated with the differentiation stasis of granulocytic lineage and the proliferation of leukemic cells. Meanwhile, a statistically higher frequency of NPM1 mutation was identified in subjects with AML M5 and lower frequency in AML M2, which was consistent with findings from previous studies (Fröhling et al., 2002; Schnittger et al., 2005; Suzuki et al., 2005; Verhaak et al., 2005; Thiede et al., 2006; Huang et al., 2008). Mori et al. (2007) provided the evidence that the expression of CD11b and CD14 antigens was significantly associated with NPM1 mutation, suggesting a close association between NPM1 mutation and monocytic features of AML. More frequent NPM1 mutation in AML M5 and a higher frequency of monocytic marker expression in NPM1-mutated subjects indicated a participation of NPM1 mutations in inducing leukemic development towards monocytic features. The relation of NPM1 mutations and higher WBC counts may be due to an accompanying occurrence of NPM1 mutations and FLT3/ITD.
CD7, as a T/NK cell associated antigen, is aberrantly expressed on blast cells of AML. Its function in AML blast remains unclear. Some previous studies have suggested CD7 may participate in the early development of myelopoiesis due to its association with some known immature antigens, such as CD34 and human leukocyte antigen-DR (HLA-DR) (Rabinowich et al., 1994; Baarcenai et al., 1995). Rausei-Mills et al. (2008) reported a higher frequency (73%) of aberrant CD7 co-expression in FLT3/ITD-positive AML subjects with normal karyotype. In comparison, only 4 of 15 FLT3/ITD-positive AML patients displayed CD7 co-expression in the present study. Interestingly, all 20 NPM1-mutated AML subjects showed no aberrant CD7 co-expression. The absent CD7 expression in NPM1-mutated cases coincided with the lower CD34 expression and highlighted the potential effect of mutated NPM1 on blocking the differentiation of hematopoietic precursor cell at monoblastic stage.
We also analyzed the clinical outcomes of 76 Chinese adults with AML and normal cytogenetics, except for FAB M3 subtype. In FLT3/ITD-positive subjects, CR rate was statistically lower than that in FLT3 wide type subjects. However, no significantly higher CR rate was observed in subjects with NPM1 mutation. When the two gene mutations were analyzed together, the significantly lower CR rate was observed in FLT3/ITD-positive/NPM1 mutation-positive subjects. These results were similar to the findings of Döhner et al. (2005). Many studies have shown that FLT3/ITD has an unfavorable prognostic impact in adult patients with AML. FLT3/ITD contributes to a short CR duration (CRD), lower CR rate, and worse disease-free survival (DFS) and OS (Yanada et al., 2005; Baldus et al., 2006; Thiede et al., 2006; Colovic et al., 2007). In our study, FLT3/ITD-positive subjects showed a poorer median duration of remission, a trend towards worse median OS and lower four-year rates of OS and RFS. Survival analyses of the four genotypes revealed that subjects with FLT3/ITD had worse OS and RFS due to the unfavorable outcome of FLT3/ITD-positive/NPM1 mutation-negative subjects. However, subjects with NPM1 mutation showed no better impact on OS or RFS in absence of FLT3/ITD, which was different from other reports (Fröhling et al., 2002; Schnittger et al., 2005; Verhaak et al., 2005). Although there was no statistical difference on OS or RFS between the FLT3/ITD-positive/NPM1 mutation-negative group and the FLT3/ITD-positive/NPM1 mutation-positive group in present study, NPM1 mutation might be a favorable prognostic factor for OS and RFS in the presence of FLT3/ITD. Our results indicate that Chinese AML patients with FLT3/ITD seem to have a worse prognosis. Unlike the findings in Caucasians, NPM1 mutation did not suggest a favorable prognosis in our study. This might be due to a relatively low response rate to chemotherapy in the NPM1 mutation-positive group. Of course, a different genetic background and chemotherapy regimen might also affect the results. Thus, further studies with larger sample sizes are warranted.
We thank Prof. Mao-de LAI, Dr. Jian CHEN, and Dr. Kun SONG (Department of Pathology and Pathophysiology, School of Medicine, Zhejiang University, China) for their excellent technical assistance.