Lack of noncanonical RAS mutations in cytogenetically normal acute myeloid leukemia
- First Online:
- Cite this article as:
- Reuter, C.W.M., Krauter, J., Onono, F.O. et al. Ann Hematol (2014) 93: 977. doi:10.1007/s00277-014-2061-9
- 237 Views
Transforming mutations in RAS genes are commonly found in human malignancies, including myeloid leukemias. To investigate the incidence, spectrum, and distribution of activating K- and N-RAS mutations in cytogenetically normal acute myeloid leukemia (CN-AML) patients, 204 CN-AML patients were screened. Activating K- and N-RAS mutations were detected in 3 of 204 (1.5 %) and 22 of 204 (10.8 %) CN-AML samples, respectively. RAS mutated patients presented with a lower percentage of bone marrow blasts (65 vs 80 %, P = 0.022). RAS mutations tended to occur with nucleophosmin-1 (NPM1) mutations (P = 0.079), and all three samples containing K-RAS mutations had concomitant NPM1 mutations. There was no significant overlap between K-RAS mutations and N-RAS, FLT3, CEBPA, IDH1/2, WT1 or MLL mutations. RAS mutation status did not impact relapse-free or overall survival of CN-AML patients. In contrast to reports of noncanonical RAS mutations in other cancers, including some leukemia subtypes, we only observed K- and N-RAS mutations in codons 12, 13, or 61 in CN-AML samples. Our findings suggest that while K-RAS mutations are infrequent in CN-AML, activating K-RAS mutations may cooperate with mutated NPM1 to induce leukemia.
KeywordsRASMutationNPM1Cytogenetically normal AML
RAS genes encode a family of membrane-associated G proteins which are important for the transduction of receptor signalling into cellular processes such as proliferation, differentiation, and apoptosis . Transforming mutations in the three functional RAS genes—H-RAS, K-RAS, and N-RAS—cause constitutive activation of the RAS proteins, and have been identified in many types of human cancers, including hematologic malignancies such as acute myeloid leukemia (AML) .
AML is a genetically heterogeneous disease thought to evolve from a combination of events in hematopoietic stem and progenitor cells (HSPC), including accumulation of acquired mutations affecting cellular programs controlling proliferation/survival (class I, including activating mutations of c-KIT, FLT3, and RAS genes), hematopoietic differentiation (class II, including nucleophosmin-1 (NPM1), CEBPA, MLL, AML1-ETO, PU.1 mutations), epigenetic programs (e.g., IDH1/2, DNMT3A, TET2, ASXL1), cell-to-cell interactions, DNA repair (e.g., cohesion complex genes), and RNA splicing (spliceosome complex genes) [3, 4]. Preleukemic hematopoietic stem cells (HSC) containing DNMT3A loss of function mutations were shown to survive chemotherapy, to have a repopulation advantage over wild-type HSCs, and to evolve into leukemic cells upon acquisition of additional mutations, such as NPM1 and internal tandem duplications of FLT3 (FLT3-ITD) .
Karyotype analysis at diagnosis allows classification of clinically distinct leukemia subtypes, providing the most important prognostic information in adult AML. Almost half of AML patients are classified as cytogenetically normal acute myeloid leukemia (CN-AML), and prognostically significant mutations identified in this group of leukemias allow further subclassification, which is important for risk-directed therapeutic intervention. For example, NPM1 and CEBPA mutations predict favorable CN-AML patient outcome, while presence of FLT3-ITD is predictive for decreased survival . The prognostic impact of Wilms’ tumor 1 (WT1) gene mutations on CN-AML patient outcome remains unsettled, with some reporting WT1 mutations to be predictive for poor outcome  while others report no prognostic impact . RAS mutations are associated with distinct cytogenetic subgroups but are not independent prognostic indicators for AML patient outcome [6, 9, 10]. However, primary AML patients with RAS mutations had significantly lower relapse occurrence when given high-dose cytarabine (HDAC) as consolidation therapy following chemotherapy .
Studies investigating RAS mutations often focus on codons 12, 13, and 61, as these sites are the most prevalent loci for activating RAS mutations. Recently, transforming noncanonical N-RAS mutations at codon 60 and K-RAS mutations at codons 14, 74, and 146 were detected in primary leukemia samples, indicating that the prevalence of activating RAS mutations may be underestimated in leukemia . These noncanonical RAS mutations reported by Tyner et al. were identified in samples of atypical AML, atypical chronic myelogenous leukemia (aCML), chronic myelomonocytic leukemia (CMML), and AML with complex or missing cytogenetics . Thus, the aim of our study was to determine the incidence of canonical and noncanonical RAS mutations in a well-characterized cohort of CN-AML patients . Additionally, the distribution and spectrum of RAS mutations in CN-AML patients was evaluated.
Design and methods
This investigation was approved by the Hannover Medical School ethics committee, and patient samples were collected upon informed consent in accordance with the Declaration of Helsinki. Primary leukemia cells were obtained from peripheral blood or bone marrow aspirates from patients (60 years old or younger). Mononuclear cells were purified by Ficoll-Hypaque gradient centrifugation (Pharmacia LKB, Uppsala, Sweden). Samples were evaluated by cytomorphology, cytochemistry, multiparameter flow cytometry, cytogenetics, fluorescence in situ hybridization, and molecular genetics in parallel. The cytomorphologic classification of AML was performed according to the French-American-British (FAB) classification. Cytogenetic fluorescence R-banding analysis was accomplished using standard methods. Only patients with a normal karyotype upon chromosome banding analysis were included in this study.
N- and K-RAS sequencing of codons 8 through 74 was accomplished by specific amplification of cDNA prepared from total RNA as described . Analysis of K-RAS codon 146 was accomplished by direct sequencing of amplicons generated by K-RAS-specific amplification of cDNA (forward primer: 5′-ATGACTGAATATAAACTTGTGG; reverse primer: 5′-TTACATAATTACACACTTTGTC). Analyses of CEBPA , WT1 exons 7 and 9 , NPM1 exon 12 mutations , and MLL-PTD mutations  and detection of FLT3 exons 14 and 15 (ITD mutations) , IDH1 , and IDH2  were performed following published protocols.
The definition of relapse-free survival (RFS) and overall survival (OS) followed recommended criteria . Median follow-up for survival was calculated using the method of Korn . The OS endpoints death (failure) and alive at last follow-up (censored) were measured from date of entry into the prospective study. RFS endpoints relapse (failure), death in complete remission (CR) (failure), and alive in CR at last follow-up (censored) were measured from the date of documented CR. Primary analysis was performed on OS. Descriptive pairwise comparisons were performed by two-sided Mann-Whitney U tests for comparison of two continuous variables and by two-sided chi-squared tests for categorical variables and are provided for exploratory purposes. The Kaplan-Meier method and two-sided log-rank tests were used to estimate the distribution of RFS and OS, and to compare differences between survival curves, respectively. The statistical analyses were performed with the statistical software package SPSS 16.0 (SPSS Science, Chicago, IL).
Results and discussion
RAS mutation frequency in CN-AML
Comparison of pretreatment clinical and molecular characteristics of CN-AML patients (n = 204) with wild-type N/K-RAS versus those with activating N/K-RAS mutations
RAS mutation status
(n = 179)
(n = 25)
Gender, no. (%)
ECOG Performance status, no. (%)
FAB Subtype, no. (%)
Peripheral blood blasts
Missing data, no. (%)
Bone marrow blasts
Missing data, no. (%)
Type of AML, no. (%)
White blood cell count
Missing data, no. (%)
Missing data, no. (%)
Platelet count (×109/l)
Missing data, no. (%)
FLT3-ITD, no. (%)
NPM1, no. (%)
NPM1 mutated/FLT3-ITD negative, no. (%)
CEBPA mutation, no. (%)
IDH1/IDH2 mutation, no. (%)
MLL-PTD, no. (%)
WT1 mutation, no. (%)
Low WT1 expression, no. (%)
WT1 rs16754 minor allele, no. (%)
Types of RAS mutations observed in CN-AML patients
Amino acid change
N-RAS mutations, no. (%)
K-RAS mutations, no. (%)
12.1 G > A
Gly > Ser (G12S)
12.2 G > A
Gly > Asp (G12D)
12.2 G > T
Gly > Val (G12V)
12.2 G > C
Gly > Ala (G12A)
13.1 G > C
Gly > Arg (G13R)
13.2 G > A
Gly > Asp (G13D)
61.3 A > T
Gln > His (Q61H)
Lack of noncanonical RAS mutations in CN-AML
Most investigations of RAS mutations analyze only canonical or “hot-spot” mutations in codons 12, 13, and 61, as these sites are the most often observed activating RAS mutations. However, the recent observation of transforming, noncanonical N-RAS mutations at codon 60 and K-RAS mutations at codons 14, 74, and 146 in primary leukemia samples suggests that it may be important to screen the entire RAS coding sequence in order to more accurately estimate the incidence of activating RAS mutations in leukemia . Our initial sequencing strategy, which allowed analysis of K-and N-RAS codons 8 through 74, identified no noncanonical K- or N-RAS mutations in codons 14, 60, or 74. As others observed K-RAS codon 146 mutations in some leukemia subtypes, we extended our coverage to include this region of K-RAS in CN-AML samples, but we detected no K-RAS codon 146 mutations in any CN-AML samples tested (n = 80). Our findings suggest that while noncanonical K- or N-RAS mutations may be infrequent genetic lesions in complex and abnormal karyotype AML, atypical CML, and CMML [12, 25], they are even less prevalent in CN-AML.
Correlation of RAS mutations with biological and clinical parameters
In our study, activating N/K-RAS mutations tended to coincide with NPM1 mutations (P = 0.079, Table 1), but not with other abnormalities such as WT1, CEBPA, FLT3-ITD, IDH1, IDH2, or MLL-PTD mutations. RAS-mutated and RAS wild-type patients were similar with respect to gender, ECOG performance status, AML type (de novo vs secondary), WBC count, hemoglobin levels, and platelet counts (Table 1). However, compared to patients with wild-type RAS, patients with RAS mutations tended to have lower median blast levels in peripheral blood (28 vs 57 %, P = 0.068) and bone marrow (65 vs 80 %, P = 0.022) (Table 1).
Using a murine bone marrow transplantation model, Zhang et al. provided evidence that activating K-ras mutations occur in hematopoietic stem cells and that secondary mutations in lineage-committed progenitors cause the final transformation into leukemic stem cells . Interestingly, activating RAS mutations sensitize cancer cells to cytarabine . Furthermore, AML patients with higher levels of activated RAS or RAS mutations achieve higher response rates and have longer overall survival when treated with HDAC [11, 29]. Thus, the efficacy of HDAC may arise from its ability to eradicate both preleukemic cells harboring only activating RAS mutations (“class I” mutations) as well as leukemic cells containing coincident RAS and “class II” mutations. In our study, all three CN-AML patient samples containing K-RAS mutations had concomitant NPM1 mutations but lacked mutations in other genes commonly observed in AML, such as WT1, CEBPA, N-RAS, FLT3, or MLL. While it may be premature to suggest that K-RAS mutations may represent a second genetic event which cooperates with mutated NPM1 to induce leukemia, it is interesting to speculate in this direction based on observations by Zhang et al. cited above . Additional murine transplantation models were also recently used to demonstrate that mutant K-RAS cooperates with other genes such as AML1/ETO to induce acute monoblastic leukemia  or NOTCH1 mutations to cause T cell lymphoblastic leukemia .
The low incidence of K-RAS mutations may be considered a limitation to the usefulness of molecular screening for this target using current routine diagnostics employed for risk stratification in AML. However, technological advancements such as whole-genome sequencing allow simultaneous analysis of all lesions which may contribute to more successful, individualized treatment strategies . Thus, even infrequent mutational events will be routinely detected in the future and easily integrated into risk stratification schemes.
While we agree with the principle that more complete sequencing coverage of the coding regions of genes implicated in leukemogenesis, such as RAS, is desirable, the additional effort may not yield significant insight into every leukemia/cancer subtype as the incidence of noncanonical mutations may be low or nonexistent. However, N-RAS mutations may be useful for detection of minimal residual disease in the approximately 10 % of AML patients presenting with these mutations. In contrast to the findings reported by Tyner et al.  in other leukemia subtypes, our sequencing data do not support a role of noncanonical RAS mutations in CN-AML.
The authors thank E. Lux and K. Görlich for excellent technical support. This work was supported by grants to C.R. from the German José Carreras Leukemia Stiftung (DJCLS R 05/21 and DJCLS R 07/32f).
Conflict of interest
The authors declare that they have no conflict of interest.