Association with the single-nucleotide polymorphism (Glu785Lys) of the granulocyte colony-stimulating factor receptor with myelodysplastic syndromes and acute myeloid leukemia with multlineage dysplasia

Myelodysplastic syndromes (MDSs) comprise a heterogeneous group of disorders. Although our understanding of the pathophysiology of this disease has been greatly improved within recent years, the underlying factors relevant for the development of MDS have not been identified yet. Wölfler et al.1 have recently shown that a single-nucleotide polymorphism (SNP) of the cytoplasmic domain of the granulocyte colony-stimulating factor (G-CSF) receptor at position 785 (G-CSF-R_Glu785Lys) predisposes for the development of high-risk MDS as defined by more than 5% marrow blasts. As a matter of fact, the SNP was detected in 7/72 (9.7%) of MDS patients, including five patients with secondary acute myeloid leukemia (AML) derived from MDS, compared to a frequency of 0.9% in normal controls. Interestingly, further analyses revealed the occurrence of the SNP in only 2.4% of patients with de novo AML (n=84). Together with in vitro experiments, showing that this SNP results in the inhibition of colony formation of G-CSF-receptor-deficient mice progenitor cells when exposed to G-CSF, the authors concluded that this new constitutional polymorphism renders normal individuals at increased risk of high-risk MDS.

Recently, the WHO defined several new categories of AML2 including AML with multilineage dysplasia (AML-MD), which displays dysplastic features in bone marrow as seen in patients with MDS. We speculated whether this SNP occurs more frequently in this new category, which has been underrepresented in the report by Wölfler et al.1 We therefore analyzed a larger cohort of AML patients (n=445) including 346 patients with AML-MD treated within the AML-SHG 96 study.3 Given the increased risk of high-risk MDS in patients carrying the polymorphism, we further investigated 80 MDS patients including a defined (n=21) population (refractory anemia: n=4, RA with excess blast (RAEB): n=10; RAEB in transformation: n=7) with a documented interval (median 7.1 months, range 0.2–73) from diagnosis to progression to AML from the Düsseldorf MDS registry. The patient characteristics are provided in Table 1.

Table 1 Patient characteristics

It is of note that, in comparison to the study by Wölfler et al.,1 our cohort of MDS patients displayed comparable characteristics in regards to age (median 66 years, range 23–97) and gender, whereas the median age for AML patients was 59 years (range 17–83). Using denaturing high-performance liquid chromatography (DHPLC), we detected only nine patients with abnormal DHPLC profile (Figure 1), which is much lower than in the initial report. Therefore, our data do not support an association with the SNP Glu785Lys of the G-CSF receptor and the development of high-risk MDS. One might speculate that our MDS patients were somewhat younger, which might influence a direct comparison between both studies. In fact, given the comparable number of patients with excess of blasts in our study and the SNP characteristics, we believe that the frequency of this polymorphism is still much lower than recently suggested. However, an analysis of a larger cohort of MDS patients is needed in order to delineate finally its role in this heterogeneous group of patients. One might speculate that the different methodology used (RFLP vs DHPLC) could have an impact on the detection frequency, but DHPLC has been extensively used in clinical genetics and appears to be a highly accurate procedure, once the assay conditions are established. As we were able to detect the predescribed non-synonymous G2591A mutation in seven of the nine cases with aberrant DHPLC profile, this would argue against a relevance of the detection method used. In addition, two novel mutations were observed, a silent mutation at position 2665 (G–A) and a 2-bp deletion of bases 2509–2510 (CA) with consecutive frame shift and a predicted premature stop of translation at codon 763. Interestingly, all of the six G-CSF-R_Glu785Lys-positive AML patients were diagnosed with AML-MD. Although not statistically significant, one might speculate a link between this polymorphism and dysplastic features in AML patients. Additionally, all patients displayed rather ‘mature’ French–American–British subtypes, which again emphasizes the link to the dysplastic pattern. Indeed, 5% of AML M6 patients carried the SNP, but this has to be clearly interpreted with caution due to the small number of patients. Possibly, an analysis of a larger cohort of those patients might further delineate the role in this rare group of AML patients. In this context, it might be interesting to categorize MDS patients according to the WHO classification, which discriminates between dysplasia affecting single or multiple lineages.

Figure 1

DHPLC profile (a) and sequence analysis (b and c) of patients analyzed for CSF3R mutations. (a) DHPLC chromatogram of three patients. After PCR (using published primer1 and standard PCR conditions with Optimase Taq-Polymerase; Transgenomic, Omaha, NE, USA, 7 μl of the 277 bp PCR product were injected into a WAVE 3500 System. Elution was performed using a standard linear gradient over 5 min at 63.9°C oven temperature. The blue curve denotes a wt sample, the green curve corresponds to a G2591A-positive sample and the red curve shows the profile of the G2665A mutation. (b) Sequence analysis of a sample with a G2591A mutation. Direct sequencing was performed using BigDye-chemistry and the reverse primer, and runs were performed on an ABI 377 HT sequencer. (c) Sequence analysis of patient DNA with the G2665A mutation. In this case, the forward primer was used.

Taken together, our data do not support a clinical and ethical relevant increased prevalence of the polymorphism in patients with AML or MDS compared to the general population, but might hint to a potential association of the G-CSF-R G2591A SNP with dyplastic features in AML. Thus, they do not exclude that the polymorphism has a biological function. However, as we did not find an increased frequency in patients with AML compared to the MDS data presented by Wölfel et al.,1 the results strongly argue against a major causative role in the transformation of MDS to AML.


  1. 1

    Wölfler A, Erkeland SJ, Bodner C, Valkhof M, Renner W, Leitner C et al. A functional single-nucleotide polymorphism of the G-CSF receptor gene predisposes individuals to high-risk myelodysplastic syndrome. Blood 2005; 105: 3731–3736.

  2. 2

    Vardiman JW, Harris NL, Brunning RD . The World Health Organization (WHO) classification of the myeloid neoplasms. Blood 2002; 100: 2292–2302.

  3. 3

    Thiede C, Steudel C, Mohr B, Schaich M, Schakel U, Platzbecker U et al. Analysis of FLT3-activating mutations in 979 patients with acute myelogenous leukemia: association with FAB subtypes and identification of subgroups with poor prognosis. Blood 2002; 99: 4326–4335.

Download references


We thank Marika Karger, Ulrike Löwel, Marita Hartwig and Peggy Grassmel for excellent technical assistance.

Author information

Correspondence to U Platzbecker.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Platzbecker, U., Germing, U., Schäkel, U. et al. Association with the single-nucleotide polymorphism (Glu785Lys) of the granulocyte colony-stimulating factor receptor with myelodysplastic syndromes and acute myeloid leukemia with multlineage dysplasia. Leukemia 20, 2188–2189 (2006).

Download citation

Further reading