IL1RAP Expression As a Measure of Leukemic Stem Cell Burden at Diagnosis of Chronic Myeloid Leukemia Predicts Therapy Outcome

Letters to the Editor 253 mutations, both spontaneously and particularly after treatment 3 Peterson LF, Zhang DE. The 8;21 translocation in leukemogenesis. Oncogene 2004; with genotoxic agents. We also present evidence that the strength 23: 4255–4262. of the mutator phenotype is related to the expression level of the 4 Mulloy JC, Cammenga J, MacKenzie KL, Berguido FJ, Moore MA, Nimer SD. fusion protein. As with all types of AML, relapse is a major cause of The AML1-ETO fusion protein promotes the expansion of human hematopoietic 99 – mortality in t(8;21) AML. If remission–induction chemotherapy fails stem cells. Blood 2002; :15 23. 5 Martinez Soria N, Tussiwand R, Ziegler P, Manz MG, Heidenreich O. Transient to eliminate all the RUNX1/ETO-positive cells, our data suggest depletion of RUNX1/RUNX1T1 by RNA interference delays tumour formation that surviving cells are predisposed to mutations that could drive in vivo. Leukemia 2009; 23: 188–190. relapse. One consideration for chemotherapy should therefore be 6 Ptasinska A, Assi SA, Martinez-Soria N, Imperato MR, Piper J, Cauchy P et al. to limit mutation of surviving leukemic/preleukemic cells. As such, Identification of a dynamic core transcriptional network in t(8;21) AML we hypothesise that targeting the mutator phenotype associated that regulates differentiation block and self-renewal. Cell Rep 2014; 8: with expression of the RUNX1/ETO fusion gene could impede the 1974–1988. acquisition of co-operating mutations required for disease relapse. 7 Yuan Y, Zhou L, Miyamoto T, Iwasaki H, Harakawa N, Hetherington CJ et al. AML1-ETO expression is directly involved in the development of acute myeloid Our data demonstrate the need to develop novel therapeutic 98 strategies that avoid increasing the mutation burden of AML cells. leukemia in the presence of additional mutations. Proc Natl Acad Sci USA 2001; : 10398–10403. 8 Wang YY, Zhou GB, Yin T, Chen B, Shi JY, Liang WX et al. AML1-ETO and CONFLICT OF INTEREST C-KIT mutation/overexpression in t(8;21) leukemia: implication in stepwise leukemogenesis and response to Gleevec. Proc Natl Acad. Sc USA 2005; 102: The authors declare no conflict of interest. 1104–1109. 9 Araten DJ, Krejci O, Ditata K, Wunderlich M, Sanders KJ, Zamecheck L et al. The rate of spontaneous mutations in human myeloid cells. Mutat Res 2013; 749: ACKNOWLEDGEMENTS 49–57. This work was supported by grants from Leukaemia and Lymphoma Research (#13044 10 Krejci O, Wunderlich M, Geiger H, Chou FS, Schleimer D, Jansen M et al. to JMA and #10033, #12055 to OH) and the JGW Patterson Foundation (to JMA and OH). p53 signaling in response to increased DNA damage sensitizes AML1-ETO cells to VJF designed and performed research, analysed data and wrote the manuscript. MHN, stress-induced death. Blood 2008; 111:2190–2199. NM-S, AKB, AP, SEF, HM and SAA performed research. CB and OH designed research and 11 Alcalay M, Meani N, Gelmetti V, Fantozzi A, Fagioli M, Orleth A et al. Acute myeloid analysed data. JMA conceived the project, designed research, analysed data and wrote leukemia fusion proteins deregulate genes involved in stem cell maintenance and the manuscript. All authors approved the final version of the manuscript. DNA repair. J Clin Investig 2003; 112: 1751–1761. 12 Chen R, Knez JJ, Merrick WC, Medof ME. Comparative efficiencies of C-terminal 1 1 1 1 signals of native glycophosphatidylinositol (GPI)-anchored proproteins in conferring VJ Forster , MH Nahari , N Martinez-Soria , AK Bradburn , GPI-anchoring. J Cell Biochem 2001; 84:68–83. 2 2 1 1 2 A Ptasinska , SA Assi , SE Fordham , H McNeil , C Bonifer , 13 Liddiard K, Hills R, Burnett AK, Darley RL, Tonks A. OGG1 is a novel prognostic 1 1 O Heidenreich and JM Allan indicator in acute myeloid leukaemia. Oncogene 2010; 29: 2005–2012. 1Northern Institute for Cancer Research, Newcastle Cancer Centre, 14 The Cancer Genome Atlas Research Network. Genomic and epigenomic Newcastle University, Newcastle-upon-Tyne, UK and landscapes of adult de novo acute myeloid leukemia. N Engl J Med 2013; 368: 2School of Cancer Sciences, Institute of Biomedical Research, College 2059–2074. of Medical and Dental Sciences, University of Birmingham, 15 Suzuki T, Harashima H, Kamiya H. Effects of base excision repair proteins on Birmingham, UK mutagenesis by 8-oxo-7,8-dihydroguanine (8-hydroxyguanine) paired with cyto- sine and adenine. DNA Repair 2010; 9: 542–550. E-mail: [email protected] or [email protected]

This work is licensed under a Creative Commons Attribution 4.0 REFERENCES International License. The images or other third party material in this 1 Wang HY, Tirado CA. t(8;21)(q22;q22) Translocation involving AML1 and ETO in article are included in the article’s Creative Commons license, unless indicated B lymphoblastic leukemia. Hum Pathol 2010; 41:286–292. otherwise in the credit line; if the material is not included under the Creative Commons 2 He G, Wu D, Sun A, Xue Y, Jin Z, Qiu H et al. B-Lymphoid and myeloid lineages license, users will need to obtain permission from the license holder to reproduce the biphenotypic acute leukemia with t(8;21)(q22;q22). Int J Hematol 2008; 87: material. To view a copy of this license, visit http://creativecommons.org/licenses/ 132–136. by/4.0/

Supplementary Information accompanies this paper on the Leukemia website (http://www.nature.com/leu).

IL1RAP expression as a measure of leukemic stem cell burden at diagnosis of chronic myeloid leukemia predicts therapy outcome

Leukemia (2016) 30, 253–257; doi:10.1038/leu.2015.135 initiating cells, also termed leukemic stem cells.1,2 The treatment of CML was greatly improved with the introduction of tyrosine kinase inhibitors more than a decade ago, but not all patients Chronic myeloid leukemia (CML) is characterized by the BCR/ABL1 respond optimally to the initial treatment and even after obtaining fusion gene formed by a t(9;22) chromosomal translocation. a deep molecular remission, discontinuation of treatment is often The t(9;22) is thought to arise in a hematopoietic stem cell followed by a relapse.3–5 creating a hierarchically arranged leukemia with a subset of cells The leukemic stem cell (LSC) burden of CML patients at in the CD34+CD38low fraction comprising the primitive leukemia diagnosis could potentially provide an important disease variable

Accepted article preview online 12 June 2015; advance online publication, 30 June 2015

© 2016 Macmillan Publishers Limited Leukemia (2016) 238 – 260 Letters to the Editor 254 for predicting response to tyrosine kinase inhibitor (TKI) treatment. population from the same patients.6 IL1RAP expression showed We previously estimated the LSC burden at the diagnosis of CML a strong correlation with BCR/ABL1 positivity in the CML CD34+ by quantification of primitive CD34+CD38low BCR/ABL1-positive CD38low fraction (r = 0.8078, Po0.0001), with a slope of regression cells, which was predictive of response to TKI therapy.6 However, line close to 1 (Y = 0.8876 × − 6.6) (Figure 2c). This indicates that this procedure would be challenging to implement clinically, almost all CD34+CD38low BCR/ABL1-positive cells express IL1RAP because it requires both sorting of CD34+CD38low CML cells and and that the BCR/ABL1-negative cells lack IL1RAP. Only a weak fluorescent in situ hybridization (FISH) analysis to enumerate BCR/ correlation was seen between BCR/ABL1-positive cells and CD25 ABL1-positive cells. Here, we aimed at establishing a flow expression (r = 0.4532, P = 0.0391) with a slope of regression of cytometry-based protocol to estimate the LSC burden at diagnosis 0.36 (Y = 0.3584 × +2.6), indicating that CD34+CD38low BCR/ABL1- by exploring the expression of cell surface markers reported to be positive cells often lack CD25 expression (Figure 2d). We then used upregulated on primitive CML CD34+CD38low cells. a previously described Flow-FISH sorting technique to sort CD34+ Several markers have recently been proposed to be upregu- CD38low cells according to CD25 expression and evaluated lated on candidate CML stem cells, including IL1RAP,7 CD25 the fraction BCR/ABL1-positive cells in four CML patients. (IL2RA),8 CD26 (DPP4),8 CD123 (IL3RA),9 CD117(KIT)10 and ST2 We found that of the CD25 expressing cells, a mean of 98% were (IL1RL1 or IL33R).11 Of these markers, only CD26 and IL1RAP, BCR/ABL1-positive cells (ranging between 94 and 100%), a co-receptor of the interleukin 1 receptor (IL1R1) with unknown whereas the CD25-negative population of the same patients still function in normal and malignant hematopoiesis, have so far been contained a mean of 85% BCR/ABL1-positive cells (range between proven to separate BCR/ABL1-positive from negative cells within 46 and 99%). Consistent with our previous findings,7 sorted the CD34+CD38low fraction.7,8 Here, we first used a staining IL1RAP-positive CD34+CD38low cells contained a great majority protocol to simultaneously analyze the expression of these seven (99%) of BCR/ABL1-positive cells (Supplementary Figure 2), markers in bone marrow aspirates from five CML patients and two whereas IL1RAP-negative cells were almost all BCR/ABL1-negative normal bone marrows. The LSC-containing CML CD34+CD38low (5%). Taken together, these data show that IL1RAP is a marker for compartment was defined as shown in Figure 1a. We found that BCR/ABL1-positive cells in the CML CD34+CD38low fraction, IL1RAP, CD25 and CD26 were distinctly upregulated on primitive whereas CD25 is only expressed on a subfraction of BCR/ABL1- CML cells relative to corresponding cells in normal bone marrows, positive cells. whereas the other markers showed a variable expression pattern The response of CML to TKI treatment is currently monitored by or were expressed also on CD34+CD38low normal bone marrows cytogenetic analyses and real-time quantitative reverse PCR cells (Figure 1b and Supplementary Table 1). CD123, CD117 and analysis of BCR/ABL1 transcript levels.4 However, so far no easily IL1R1 were all expressed on the immature CML CD34+CD38low implementable laboratory-based method is available to estimate cells but showed expression also on the normal hematopoietic the LSC burden at diagnosis as a possible predictor of treatment stem cell. In contrast to a recent study reporting low ST2 response. We therefore investigated whether IL1RAP or CD25 expression on CML CD34+CD38low cells,11 this marker could only expression in the CML CD34+CD38low compartment, as a measure be detected in one of five CML patients using the same antibody, of the LSC burden at diagnosis, could provide clinically relevant possibly because of a lower sensitivity of our flow cytometry prognostic information. In our previous study, the presence of protocol. Thus IL1RAP, CD25 and CD26 were specifically expressed ⩾ 80% BCR/ABL1-positive cells by FISH in the CD34+CD38low on CML CD34+CD38low cells compared with corresponding fraction at diagnosis of CML was used to define a group of normal cells. patients with ‘high’ LSC burden, whereas o80% were designated To study the expression of LSC markers in more detail, IL1RAP, as ‘low’ LSC burden.6 Based on IL1RAP expression, we used a CD25 and CD123 were chosen for further analysis using an similar definition, that is, patients were classified as IL1RAPlow optimized panel of antibodies. As CD25 and CD26 displayed (n = 13, defined as o80% of the CD34+CD38low cells expressing similar expression patterns in primitive CML cells (Figure 1c), IL1RAP) or IL1RAPhigh (n =8, ⩾ 80% of CD34+CD38low cells consistent with previous studies,8 we included only CD25 in the expressing IL1RAP). Following this subdivision, we investigated if staining protocol. Instead we selected CD123 since antibodies the two groups differed in the probability of achieving optimal targeting this receptor currently are being evaluated as a therapy response as defined by the European Leukemia Net guidelines, in CML.9 These three markers were analyzed in a cohort of following TKI treatment (Supplementary Table 3).4 All IL1RAPlow 21 CML patients included in the NordCML006 study, in which and 7 of 8 IL1RAPhigh patients achieved BCR/ABL1 transcript levels patients were randomized to receive treatment either with ⩽ 10% within 3 months. However, IL1RAPlow patients had a higher imatinib or dasatinib.12 A summary of patient data is provided chance of reaching transcript levels o1% after 6 months in Supplementary Table 2. Using the optimized protocol, we (P = 0.0475), a definition of optimal response according to obtained a clear separation of both IL1RAP-positive and CD25- European Leukemia Net.4 Since this cohort consisted of mainly positive cells within the CML CD34+CD38low compartment, optimal responders at 3 months, we looked at even deeper levels whereas CD123 expression failed to divide cells into distinct of molecular and cytogenetic response, previously shown to positive and negative cell populations (Figure 2a). In normal bone predict long-term outcome at this time point.13,14 Interestingly, marrows, we did not observe expression of IL1RAP or CD25 in the IL1RAPlow patients (13 out of a total of 21) had a higher chance of CD34+CD38low population, while CD123 was weakly expressed achieving complete cytogenetic response (P = 0.0009) and major (Supplementary Figure 1). While analyzing co-expression of molecular response (BCR/ABL1 transcript levels ⩽ 0.1%) after IL1RAP and CD25 in the CML CD34+CD38low compartment, we 3 months (P = 0.023) (Supplementary Table 3). No significant found one population co-expressing IL1RAP and CD25 and one correlations were seen between IL1RAP expression and Sokal or expressing only IL1RAP. No cells were found to express only CD25 Hasford risk scores. Altogether, this suggests that IL1RAP (Figure 2b). Although the cell fractions expressing IL1RAP and expression as a measure of LSC burden at diagnosis predicts CD25 showed a significant correlation (r = 0.8558, Po0.0001), cytogenetic and molecular response to TKI treatment although IL1RAP was present on a consistently larger fraction of CML CD34+ larger prospective studies are needed before such measurement CD38low cells (mean of difference 25.62%, confidence interval 95% can be implemented clinically. 15.81–35.44, Po0.0001). Previously, light scattering properties together with CD34 To further delineate the difference between cells expressing and CD45 expression have been shown to differentiate between IL1RAP and CD25, we correlated the flow cytometry data on BCR/ABL1-positive and -negative CML CD34+CD38low cells.15 Using IL1RAP and CD25 expression from the current study to previously that protocol, it was shown that the LSC burden at diagnosis could reported BCR/ABL1 FISH data on the CML CD34+CD38low predict the complete cytogenetic response after 1 year of

FSC-A CD34

CD34+CD38+ = 80%

CD38 CD34+CD38low = 5%

= isotype

CD34 = stain

IL1RAP CD25 CD26 CD123 IL1R1 CD117 ST2

CML cells low CD38 + NBM CD34

CML CD34+CD38low CD25

CD26

Figure 1. Flow cytometry analysis of stem cell surface marker expression in CML. (a) Gating algorithm used to determine progenitor and stem cell compartments. (b) Representative histograms showing expression of IL1RAP, CD25, CD26, CD123, IL1R1, CD117 and ST2 in the CD34+ CD38low compartment of a CML patient and a normal bone marrow. Isotype control (red line) and staining antibody (blue line). (c) Dotplots of co-expression of CD25 and CD26 in the CD34+CD38low compartment in five CML patients, isotype control (red) and staining antibody (blue). treatment as well as major molecular response after 1.5 years. No sensitivities of the two assays to detect IL1RAP expression. All correlations were made to response after 3 months of therapy. patients in our study clearly expressed IL1RAP; even in a patient in Other investigators have used CD26 as a marker for primitive CML which FISH analysis of sorted CD34+CD38low cells revealed o1% cells and showed a correlation with leukocyte counts at diagnosis BCR/ABL1-positive cells, a fraction of cells expressed IL1RAP but not with Sokal or Hasford risk scores, possible association with (patient no. 10, Supplementary Table 2). cytogenetic and molecular response were not reported.8 In In conclusion, we identified IL1RAP, CD25 and CD26 as the most contrast to our results, the authors could not detect over- specific markers for primitive CML cells relative to the correspond- expression of IL1RAP on primitive CML cells in all patients. Most ing normal cells and demonstrate that the percentage of IL1RAP likely, the difference in the two studies is a result of different expressing cells within the CD34+CD38low compartment can be

CML CD34+CD38low

IL1RAP CD25 CD123

CML CD34+CD38low CML CD34+CD38low

100 r = 0.8078 80 p < 0.0001 60 Y = 0.8876X - 6.592 cells (%) + 40

20 CD25 IL1RAP 0 0 50 100 BCR/ABL1+ (%)

IL1RAP CML CD34+CD38low 98% 85% CML CD34+CD38low 100 100 r = 0.4532

p = 0.0391 (%) 80 + n=4 Y = 0.3584X + 2.601 60 50 cells (%)

+ 40 BCR/ABL1 20 CD25

0 0 0 50 100 CD25+ CD25- BCR/ABL1+ (%)

Figure 2. Expression of IL1RAP, CD25 and CD123 in CD34+CD38low cells from 21 CML patients. (a) Histograms of IL1RAP, CD25 and CD123 in the CD34+CD38low compartment of a representative CML patient. Isotype control (red line) and staining antibody (blue line). (b) Co-expression of IL1RAP and CD25 in CD34+CD38low cells. (c) Flow cytometry-assessed IL1RAP expression in CD34+CD38low cells shows a strong correlation to previously reported FISH data on BCR/ABL1 expression from 21 CML patients. (d) Flow cytometry-assessed CD25 expression in CD34+ CD38low cells shows a weaker correlation with the previously reported FISH data on BCR/ABL1.(e) CD34+CD38low cells from four CML patients sorted according to CD25 expression and analyzed with FISH with probes detecting BCR/ABL1. Mean values are plotted, error bars show range.

used to predict the response to TKI treatment. The described flow ACKNOWLEDGEMENTS cytometry-based protocol may become a valuable prognostic tool We thank all investigators of the NordCML006 study for sharing samples and clinical data. in the management of CML patients. This work was supported by the Swedish Cancer Society, the Swedish Children’sCancer Foundation, the Swedish Research Council, the Inga-Britt and Arne Lundberg Foundation, the Gunnar Nilsson Cancer Foundation, the Medical Faculty of Lund University, the strategic CONFLICT OF INTEREST research program BioCARE, the Finnish Cancer Institute and the Academy of Finland. MJ and TF are cofounders and have equity ownership in Cantargia AB (Ideon Medical 1 1 1 1 1 Village, Lund, Sweden) formed with Lund University Bioscience AB. JR has stock N Landberg , N Hansen , M Askmyr , H Ågerstam , C Lassen , 1 2,3 4 1 options in Cantargia AB and has received honoraria from Novartis and Bristol-Myers M Rissler , HH Hansen , S Mustjoki , M Järås , 5 1 Squibb. SM has received research funding from Novartis, Bristol-Myers and Squibb J Richter and T Fioretos 1 and Pfizer and honoraria from Novartis and Bristol-Myers and Squibb. The remaining Department of Clinical Genetics, Lund University, Lund, Sweden; 2 authors declare no conflict of interest. Department of Hematology, St Olavs Hospital, Trondheim, Norway;

Leukemia (2016) 238 – 260 © 2016 Macmillan Publishers Limited Letters to the Editor 257 3 Department of Cancer Research and Molecular Medicine, 8 Herrmann H, Sadovnik I, Cerny-Reiterer S, Rulicke T, Stefanzl G, Willmann M et al. Norwegian University of Science and Technology (NTNU), Trondheim, Dipeptidylpeptidase IV (CD26) defines leukemic stem cells (LSC) in chronic Norway; myeloid leukemia. Blood 2014; 123: 3951–3962. 4Hematology Research Unit Helsinki, Department of Clinical 9 Nievergall E, Ramshaw HS, Yong ASM, Biondo M, Busfield SJ, Chemistry and Hematology, University of Helsinki and Vairo G et al. Monoclonal antibody targeting of IL-3 receptor with CSL362 123 Comprehensive Cancer Center, Helsinki University Hospital, Helsinki, effectively depletes CML progenitor and stem cells. Blood 2014; : 1218–1228. Finland and 5 10 Florian S, Sonneck K, Hauswirth AW, Krauth M-T, Schernthaner G-H, Department of Hematology and Vascular Disorders, Skåne Sperr WR et al. Detection of molecular targets on the surface of University Hospital, Lund, Sweden CD34+/CD38– stem cells in various myeloid malignancies. Leuk Lymphoma E-mail: [email protected] or thoas.fi[email protected] 2006; 47: 207–222. 11 Levescot A, Flamant S, Basbous S, Jacomet F, Feraud O, Anne Bourgeois E et al. BCR-ABL-induced deregulation of the IL-33/ST2 pathway in CD34(+) REFERENCES progenitors from chronic myeloid leukemia patients. Cancer Res 2014; 74: 1 Ren R. Mechanisms of BCR–ABL in the pathogenesis of chronic myelogenous 2669–2676. leukaemia. Nat Rev Cancer 2005; 5: 172–183. 12 Hjorth-Hansen H, Stenke L, Söderlund S, Dreimane A, Ehrencrona H, 2 Sloma I, Jiang X, Eaves AC, Eaves CJ. Insights into the stem cells of chronic Gedde-Dahl T et al. Dasatinib induces fast and deep responses in newly myeloid leukemia. Leukemia 2010; 24: 1823–1833. diagnosed chronic myeloid leukaemia patients in chronic phase: clinical results 3 Deininger M, Buchdunger E, Druker BJ. The development of imatinib as a from a randomised phase-2 study (NordCML006). Eur J Haematol 2014; 94: therapeutic agent for chronic myeloid leukemia. Blood 2005; 105: 2640–2653. 243–250. 4 Baccarani M, Deininger MW, Rosti G, Hochhaus A, Soverini S, Apperley JF et al. 13 Marin D, Ibrahim AR, Lucas C, Gerrard G, Wang L, Szydlo RM et al. Assessment of European LeukemiaNet recommendations for the management of chronic BCR-ABL1 transcript levels at 3 months is the only requirement for predicting myeloid leukemia. 2013Blood 2013; 122:872–884. outcome for patients with chronic myeloid leukemia treated with tyrosine kinase 5 Mahon FX, Etienne G. Deep molecular response in chronic myeloid leukemia: the inhibitors. J Clin Oncol 2012; 30:232–238. new goal of therapy? Clin Cancer Res 2014; 20:310–322. 14 Hanfstein B, Muller MC, Hehlmann R, Erben P, Lauseker M, Fabarius A et al. Early 6 Mustjoki S, Richter J, Barbany G, Ehrencrona H, Fioretos T, Gedde-Dahl T et al. molecular and cytogenetic response is predictive for long-term progression-free Impact of malignant stem cell burden on therapy outcome in newly diagnosed and overall survival in chronic myeloid leukemia (CML). Leukemia 2012; 26: chronic myeloid leukemia patients. Leukemia 2013; 27: 1520–1526. 2096–2102. 7 Jaras M, Johnels P, Hansen N, Agerstam H, Tsapogas P, Rissler M et al. Isolation 15 Janssen JJ, Deenik W, Smolders KG, van Kuijk BJ, Pouwels W, Kelder A et al. and killing of candidate chronic myeloid leukemia stem cells by antibody tar- Residual normal stem cells can be detected in newly diagnosed chronic myeloid geting of IL-1 receptor accessory protein. Proc Natl Acad Sci USA 2010; 107: leukemia patients by a new flow cytometric approach and predict for optimal 16280–16285. response to imatinib. Leukemia 2011; 26:977–984.

Supplementary Information accompanies this paper on the Leukemia website (http://www.nature.com/leu).

Array CGH identifies copy number changes in 11% of 520 MDS patients with normal karyotype and uncovers prognostically relevant deletions Leukemia (2016) 30, 257–260; doi:10.1038/leu.2015.257 In this study we addressed the question whether MDS with normal karyotype in chromosome banding analysis harbor cytogenetically cryptic gains or losses detectable by array CGH. In myelodysplastic syndromes (MDS), cytogenetic aberrations A total of 520 patients (208 women and 312 men) with a have an important role for classification and prognostication. median age of 72 years (range 8–90 years) and with different MDS The original International Prognostic Scoring System and subtypes were investigated. This cohort is a subset of the 944 MDS the revised International Prognostic Scoring System classifiers patients published by Haferlach et al.4 Bone marrow and have clearly demonstrated the prognostic impact of distinct peripheral blood samples were referred to the MLL Munich cytogenetic abnormalities.1,2 The vast majority of chromosome Leukemia Laboratory between September 2005 and August 2011 aberrations in MDS are gains and losses of chromosomal for diagnosis. All patients gave written informed consent for the material, whereas balanced rearrangements are rare.3 However, use of data for scientific evaluations. The study was approved by more than 50% of MDS and even a higher proportion the Internal Review Board and adhered to the tenets of the of low-risk MDS harbor a normal karyotype.3 Owing to its Declaration of Helsinki. limited resolution, chromosome banding analysis can only For all patients, cytomorphology and chromosome banding detect gains and losses of more than 10 Mb size. Furthermore, analysis had been performed in our laboratory.5 Survival data were chromosome banding analysis is depending on proliferation available for 487 patients. Five hundred and twenty MDS patients of the MDS clone in vitro to obtain metaphases. Array with normal karyotype were analyzed by array CGH according to comparative genomic hybridization (CGH) has a considerably the manufacturer’s recommendations (Human CGH 12 × 270K higher resolution and therefore is capable of identifying Whole-Genome Tiling Array, Roche NimbleGen, Madison, WI, USA). small unbalanced chromosomal aberrations undetectable by Subsequently, recurrently deleted regions detected by array CGH chromosome banding analysis. In addition, array CGH does not were validated using fluorescence in-situ hybridization (FISH) on rely on proliferating cells. interphase nuclei.

Accepted article preview online 22 September 2015; advance online publication, 13 October 2015