OUP user menu

Array Comparative Genomic Hybridization of Peripheral Blood Granulocytes of Patients With Myelodysplastic Syndrome Detects Karyotypic Abnormalities

Suzanne M. Vercauteren MD, PhD, Sandy Sung BSc, Daniel T. Starczynowski PhD, Wan L. Lam PhD, Helene Bruyere MD, Douglas E. Horsman MD, Peter Tsang MD, Heather Leitch MD, Aly Karsan MD
DOI: http://dx.doi.org/10.1309/AJCPH27ZIZEJLORF 119-126 First published online: 1 July 2010


The diagnosis of myelodysplastic syndromes (MDSs) relies largely on morphologic and karyotypic abnormalities, present in about 50% of patients with MDS. Array-based genomic platforms have identified copy number alterations in 50% to 70% of bone marrow samples of patients with MDS with a normal karyotype, suggesting a diagnostic role for these platforms. We investigated whether blood granulocytes harbor the same copy number alterations as the marrow of affected patients. Of 11 patients, 4 had cytogenetic abnormalities shown by conventional karyotyping involving chromosomes 5, 8, 11, 20, and X, and these changes were seen in the granulocytes of all 4 patients by using array comparative genomic hybridization (aCGH). Cryptic alterations were identified at a significantly higher level in marrow CD34+ cells compared with granulocytes (P < .0001). These data suggest that aCGH analysis of circulating granulocytes may be useful in detecting gross karyotypic alterations in patients with MDS when marrow examination has failed or not been done.

Key Words:
  • Myelodysplastic syndrome
  • Array comparative genomic hybridization
  • Neutrophils

The myelodysplastic syndromes (MDSs) are a group of clonal stem cell disorders characterized by ineffective hematopoiesis with abnormal differentiation and maturation of myeloid cells, resulting in 1 or more peripheral cytopenias.1,2 MDS is thought to be the most common myeloid malignancy, with an overall incidence of about 3 to 5/100,000 people per year, increasing to approximately 100/100,000 per year in people older than 70 years.3,4

MDS likely occurs due to genetic changes in a primitive CD34+ hematopoietic cell that gives rise to abnormal progeny cells.59 Although genetic abnormalities are thought to be present in all patients with MDS, standard cytogenetics on bone marrow samples only detect an abnormal karyotype in approximately 50% of patients with low-risk MDS.1,2,8 We and others recently identified cryptic copy number alterations in bone marrow cells of a large proportion (18%–68%) of patients with low-risk MDS and a normal karyotype, using high-resolution array platforms, suggesting that structural genomic abnormalities are more prevalent than expected.912 In addition, we showed that the presence of these alterations has prognostic value in low-risk MDS.9 These findings suggest that the use of array-based platforms, such as array comparative genomic hybridization (aCGH) or single-nucleotide polymorphism arrays on bone marrow cells, are useful tools in the diagnosis and prognosis of MDS.

Cytogenetic studies are not always performed or successful in MDS. A recent report showed that only approximately 60% of MDS patients undergo a bone marrow evaluation.4 Therefore, we investigated whether peripheral blood granulocytes can be used as an alternative source of cells for obtaining cytogenetic information using genomic arrays. We used high-resolution bacterial artificial chromosome aCGH analysis to compare the cytogenetic alterations in matched CD34+ bone marrow–derived cells and peripheral blood granulocytes from patients with MDS.

Materials and Methods

Sample Collection

Bone marrow and peripheral blood samples were obtained from 11 patients with MDS (MDS, n = 8; and MDS/myeloproliferative neoplasm [MPN], n = 3) at diagnosis, following informed consent Table 1 (age range, 60–85 years; median age, 66 years; mean age, 70 years). All protocols were approved by the British Columbia Cancer Agency/University of British Columbia Clinical Research Ethics Board (Vancouver).

Cell Isolation

CD34+ and CD3+ cells were positively selected from cryopreserved marrow by immunomagnetic separation according to the manufacturer’s instructions (Stem Cell Technologies, Vancouver, BC) with a final purity of more than 70%. Peripheral blood granulocytes were isolated by using Ficoll-Hypaque density-gradient centrifugation according to the manufacturer’s instructions (Stem Cell Technologies).

View this table:
Table 1

Whole Genome Tiling-Path aCGH Analysis

Details of whole genome aCGH, including DNA extraction, labeling, and hybridization, as well as image analysis, have been described previously.9,13,14 This platform contains 32,433 duplicate bacterial artificial chromosome–derived DNA segments providing tiling coverage of the entire human genome and has a theoretical resolution of approximately 50 kb.14 Sample (MDS) and reference (normal diploid) genomic DNA (50–200 ng each) were separately labeled using cyanine 3 and cyanine 5 deoxycytidine triphosphate fluorescence markers, respectively. The images were captured with a charge-coupled device camera and analyzed using an ArrayWorx scanner and SoftWorx Tracker Spot Analysis software (Applied Precision, Issaquah, WA).13 SeeGH custom software was used to visualize all data as log ratio plots.14,15 Clones with SD values between duplicate spots of more than 0.1 were filtered from the raw data. A region was considered altered when a minimum of 2 overlapping consecutive clones showed the change. A hidden Markov model algorithm was used to verify breakpoints of genomic alterations identified by visual inspection as described.16

Statistical Analysis

Differences in genomic alterations between CD34+ cells and granulocytes of patients with MDS were analyzed by using the Mann-Whitney test. Analysis was carried out using GraphPad Prism4 (GraphPad Software, San Diego, CA).

Results and Discussion

High-resolution aCGH was performed on 11 DNA samples of matched CD34+ cells and granulocytes from patients with MDS or MDS/MPN Table 2. Of 11 patients, 4 had known cytogenetic abnormalities as identified by conventional karyotyping, with good concordance between conventional cytogenetics and aCGH. Case 2 had a deletion 11 from q14.1 to q24.3 detected in 15 of 15 metaphases, which was identified in the CD34+ cells and the granulocytes by aCGH Figure 1B. Case 5 had a deletion 5 from q23.1 to q31.2 detected in 11 of 20 metaphases and +8 in 4 of 20 metaphases by conventional cytogenetics. By aCGH, we identified deletion 5 from q23.1 to q31.2 in the CD34+ cells and granulocytes but could not detect the +8 in the CD34+ cells or granulocytes, consistent with an aCGH detection threshold of 25% to 30% abnormal cells.13 In addition, in case 5, we detected a deletion 20 from q11.21 to 13.33, which was not clearly resolved by conventional karyotyping. Case 9 had +8 in 14 of 14 metaphases, and this abnormality was detected in CD34+ cells and granulocytes by aCGH. Case 10 had an isodicentric chromosome involving the short arm of chromosome X with a breakpoint at q13 in 18 of 25 metaphases, which was seen in the CD34+ cells and granulocytes with aCGH.

Figure 1

A, Array comparative genomic hybridization (aCGH) analysis was used to identify total genomic alterations in matched marrow CD34+ cells and peripheral blood granulocytes of 11 patients with myelodysplastic syndrome (MDS) and MDS/myeloproliferative neoplasm. No significant difference was seen between the 2 cell populations (P = .19). B (Case 2), aCGH analysis of matched CD34+ cells and blood granulocytes of a patient with MDS with a known del(11q) shows the presence of this abnormality in the CD34+ cells and blood granulocytes. Mb, megabase.

Three patients (cases 4, 7, and 8) had a normal karyotype by conventional cytogenetics, and aCGH did not reveal gross (>3 megabases [Mb]) chromosomal alterations in the CD34+ cells or granulocytes. In 4 patients (cases 1, 3, 6, and 11), conventional karyotyping failed or was not performed. One of these patients (case 11) had a partial trisomy 9 from q33.3 to q34.3 and trisomies 19 and 22 shown by aCGH in CD34+ cells and granulocytes, whereas for the other 3 patients, no large chromosomal abnormalities were shown. These findings are in agreement with previous studies showing that circulating granulocytes contain a detectable population that arises from the malignant clone in the majority of patients with MDS.17,18

Because the resolution of aCGH is significantly greater than that of conventional cytogenetics,19 we investigated whether cryptic alterations in CD34+ cells of patients with MDS are also present in their peripheral blood granulocytes. We found no significant difference (P = .19) in total genomic alterations seen in CD34+ cells (mean ± SD total genomic alterations, 57.5 ± 70.2 Mb) and granulocytes (mean ± SD total genomic alterations, 39.3 ± 67.6 Mb) of 11 patients with MDS by using aCGH Figure 1A. However, when genomic alterations that could potentially be detected by conventional karyotyping (>10 Mb) were excluded, a significantly higher level of cryptic alterations was seen in CD34+ cells (mean ± total cryptic alterations, 13.3 ± 10.3 Mb; range, 3.2–30.2 Mb) of patients with MDS compared with granulocytes (mean ± SD total cryptic alterations, 0.3 ± 0.5 Mb; range, 0.0–1.6 Mb) (P < .0003) Figure 2A.

Previously we reported that cryptic changes of up to 2.6 Mb in total can be found in CD34+ cells of healthy elderly people.9 Therefore, we considered more than 3 Mb of cryptic changes to be disease related. None of the 11 MDS cases analyzed in this study had more than 3 Mb changes in the granulocytes, excluding large karyotypic alterations (range, 0.0–1.6 Mb). For example, Figure 2B shows 2 cryptic alterations seen in the CD34+ cells of MDS case 6, and amplification 1 is clearly absent in the granulocytes. Although we cannot exclude that amplification 2 is present in a proportion of granulocytes, it would not fulfill our criteria for a cryptic alteration (see “Materials and Methods” section). Therefore, in the absence of data from the CD34+ cells, we would not have considered amplification 2 to be present in the granulocytes.

To determine whether the cryptic changes seen in granulocytes and CD34+ cells were somatic or acquired, we compared the aCGH results with aCGH performed on DNA of CD3+ T cells of 4 patients (cases 1, 3, 4, and 6). The majority of cryptic changes seen in granulocytes were also detected in T cells, whereas the majority of cryptic changes present in CD34+ cells but not in granulocytes were also absent in T cells (Table 2). These results suggest that a large proportion of the circulating granulocytes of patients with MDS likely do not arise from the malignant clone, but rather represent residual normal hematopoiesis, in keeping with previous observations by fluorescence in situ hybridization (FISH) showing that even dysplastic circulating neutrophils may represent the normal population.19,20 However, this does not explain why the larger chromosomal aberrations were detectable in granulocytes but the majority of cryptic changes were absent. The most likely reason is that larger changes can be identified with greater confidence using aCGH. For example, the cryptic amplification 2 in Figure 2B cannot be detected with confidence in granulocytes, although it may be present. Another possible explanation is that the larger abnormalities are present at an earlier stage of leukemogenesis and that additional cryptic changes lead to apoptosis of granulocytes within the bone marrow environment.

Figure 2

Cryptic genomic alterations in marrow CD34+ cells by array comparative genomic hybridization (aCGH) are not detected in matched peripheral blood granulocytes. A, aCGH analysis revealed a significantly higher number of cryptic alterations (P < .001) in CD34+ cells compared with peripheral blood granulocytes of 11 patients with myelodysplastic syndrome (MDS) or MDS/myeloproliferative neoplasm. B (Case 6), Two cryptic amplifications are clearly identified in marrow CD34+ cells, while the presence of both amplifications could not be confirmed in the matched blood granulocytes. Although the presence of amplification 2 cannot be ruled out in granulocytes, it does not fulfill the criteria for a genomic alteration.

We did not detect an increased number or larger size of cryptic changes in granulocytes of patients with higher grade MDS (refractory cytopenia with multilineage dysplasia or refractory anemia with excess blasts) compared with lower grade MDS (refractory anemia or refractory anemia with ringed sideroblasts), but this study was done on a relatively small number of cases and is, therefore, not adequately powered to address this issue.

Our findings suggest that in patients with failed standard cytogenetics, aCGH studies on the peripheral blood may be useful because large genomic alterations are usually detected in peripheral blood granulocytes. In particular, detection of deletion 5q in circulating granulocytes may be a useful alternative to a marrow examination in patients being considered for lenalidomide therapy. Although currently FISH on granulocytes is a cheaper alternative to aCGH for detecting deletion 5q, FISH will not detect other chromosome abnormalities with prognostic significance and will not exclude the presence of additional cytogenetic abnormalities in patients with deletion 5q. Nevertheless, cryptic alterations present in CD34+ bone marrow cells of patients with MDS cannot be readily detected in peripheral blood granulocytes and are, therefore, not an optimal diagnostic substitute in patients with MDS with a normal karyotype, even when high-resolution genomic platforms are used.


Upon completion of this activity you will be able to:

  • list 2 advantages of array comparative genomic hybridization (aCGH) over conventional cytogenetics.

  • describe the role of aCGH on circulating granulocytes of patients with myelodysplastic syndrome.

The ASCP is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The ASCP designates this educational activity for a maximum of 1 AMA PRA Category 1 Credit ™ per article. This activity qualifies as an American Board of Pathology Maintenance of Certification Part II Self-Assessment Module.

The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose.

Questions appear on p 166. Exam is located at www.ascp.org/ajcpcme.


  • Dr Karsan is a senior scholar of the MSFHR.

  • Supported by grant MOP 89976 from the Canadian Institutes of Health Research (CIHR), Ottawa, and from the Leukemia and Lymphoma Society (Dr Karsan), White Plains, NY; from Genome Canada, Ottawa, and CIHR (Dr Lam). Dr Vercauteren is supported by a Leukemia and Lymphoma Society clinical research fellowship and Dr Starczynowski by CIHR and Michael Smith Foundation for Health Research (MSFHR), Vancouver, Canada, fellowships.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
View Abstract