Received: 10 January 2023 Accepted: 18 January 2023 DOI: 10.1002/jha2.651 SHORT REPORT Regions of homozygosity confer a worse prognostic impact in myelodysplastic syndrome with normal karyotype Mar Mallo1,2 Heinz Tuechler3 Leonor Arenillas4 Sophie Raynaud5 Thomas Cluzeau5 Lee-Yung Shih6 Chiang Tung-Liang6 Christina Ganster7 Katayoon Shirneshan7 Detlef Haase7 Martí Mascaró8 Laura Palomo9 José Cervera10 Esperanza Such10 Nicola Trim11 Sally Jeffries11 Emma Ridgway11 Giovanni Marconi12 Giovanni Martinelli12 Francesc Solé1 1 MDS Research Group, Institut de Recerca Contra la Leucèmia Josep Carreras (IJC), ICO-Hospital Germans Trias i Pujol, Universitat Autònoma de Barcelona, Badalona, Spain 2 Microarrays Unit, Institut de Recerca Contra la Leucèmia Josep Carreras (IJC), ICO-Hospital Germans Trias i Pujol, Universitat Autònoma de Barcelona, Badalona, Spain 3 Boltzmann Institute for Leukaemia Research and Hematology, Vienna, Austria 4 Hematological Cytology Laboratory, Pathology Department, Hospital del Mar, GRETNHE, IMIM (Hospital del Mar Research Institute), Barcelona, Spain 5 Hematology Department, Cote d’Azur University, CHU of Nice, Nice, France 6 Division of Hematology, Chang Gung Memorial Hospital-Linkuo, Chang Gung University, Taoyuan City, Taiwan 7 Clinics of Hematology and Medical Oncology, University Medical Center Göttingen, Göttingen, Germany 8 Hematology Service, Hospital Son Llàtzer, Palma de Mallorca, Spain 9 Experimental Hematology, Vall d’Hebron Institute of Oncology (VHIO), Vall d’Hebron Barcelona Hospital Campus, Barcelona, Spain 10 Hematology Service, Hospital Universitario La Fe, Valencia, Spain 11 West Midlands Regional Genetics Laboratory, Birmingham Women’s Hospital, Birmingham, UK 12 IRCCS Istituto Romagnolo per lo Studio dei Tumori (IRST) “Dino Amadori”, Meldola, Italy Correspondence Francesc Sole, MDS Research Group, Institut Abstract de Recerca Contra la Leucèmia Josep Carreras (IJC), ICO-Hospital Germans Trias i Pujol, Half of the myelodysplastic syndromes (MDS) have normal karyotype by conventional Universitat Autònoma de Barcelona, Badalona, banding analysis. The percentage of true normal karyotype cases can be reduced by Spain. Email: fsole@carrerasresearch.org 20–30% with the complementary application of genomic microarrays. We here present a multicenter collaborative study of 163 MDS cases with a normal karyotype (≥10 Funding information metaphases) at diagnosis. All cases were analyzed with the ThermoFisher® microarray GRC, Grant/Award Number: 2017 SGR288; CERCA Programme/Generalitat de Catalunya; (either SNP 6.0 or CytoScan HD) for the identification of both copy number alter- Fundació Internacional Josep Carreras; Italy ation(CNA) and regions of homozygosity (ROH). Our series supports that 25 Mb cut-off Ministry of Health as having the most prognostic impact, even after adjustment by IPSS-R. This study high- lights the importance of microarrays in MDS patients, to detect CNAs and especially to detect acquired ROH which has demonstrated a high prognostic impact. KEYWORDS chromosome, cytogenetics, microarrays, molecular cytogenetics, myelodysplastic syndromes This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2023 The Authors. eJHaem published by British Society for Haematology and John Wiley & Sons Ltd. 446 wileyonlinelibrary.com/journal/jha2 eJHaem. 2023;4:446–449. MALLO ET AL . 447 1 INTRODUCTION Myelodysplastic syndromes (MDS) constitute a group of heteroge- neous disorders, whose current diagnosis requires integration from different disciplines. Cytogenetics is the current gold standard for cytogenetic diagnosis and it is one of the main prognostic factors in the revised international prognostic scoring system (IPSS-R) [1]. The mutational profile has been shown to be an important prognostic factor and has been incorporated into the new molecular prognostic scoring system (IPSS-M) [2]. Here, we report the prognosis role of microarrays in MDS with normal karyotype, especially for regions of homozygosity (ROH). Cytogenetics includes several techniques: chromosome banding analysis (CBA), fluorescence in situ hybridization, genomic microar- rays, and, more recently, optical genome mapping [3]. Of these, only chromosome-level events that may be discovered with classical CBA are included as a prognostic parameter in the IPSS-R [1] and it has been retained in the new IPSS-M [2]. With traditional techniques, half of MDS cases are found to have a normal karyotype. In-depth analy- ses, with advanced techniques, have shown that the percentage of true F I G U R E 1 Circos plot displaying genetic aberrations found by normal karyotype cases can be reduced by 20–30% by the comple- microarrays. Blue bars indicate gains; red bars, losses; and green bars, mentary application of genomic microarrays [4, 5]. Thus, the addition regions of homozigosity. of genomic microarrays can increase the diagnostic yield and correctly classify a substantial portion of MDS patients that may otherwise be misclassified with the standard approach of karyotyping alone. ing altered cases, the median size aberration was 404 Kb (range: 24.1 Kb–190 Mb). Focusing on the type of alteration, 56 (34.4%) cases har- bored CNA and 26 (15.95%) ROH. Considering the size of the affected 2 MATERIALS AND METHODS genome in aberrant cases, the median value for CNA was 429 Kb (range: 69 Kb–190 Mb), for ROH, 25 Mb (range: 2–138 Mb); and for We here present a multicenter collaborative study of 163 MDS de CNA plus ROH, 1 Mb (range: 69 Kb–190 Mb). Microarray results are novo cases with a normal karyotype (15 of them had more than 10 given in detail in Figure 1. metaphases completely analyzed) at diagnosis. All cases were ana- Regarding clinical variables, no differences were observed between lyzed in parallel with the ThermoFisher® microarray (either SNP altered and normal cases, in terms of age, sex, blood counts, and IPSS-R. 6.0 or CytoScan HD) for the identification of both CNA and ROH. OS and time to transformation were analyzed according to microarray We collected microarray results information from different centers results for 138 patients (Table 1). which were integrated into a common database for statistical evalua- A positive microarray result indicating ROH showed a borderline tion. CNA and ROH results were submitted to the common database significant negative impact on OS (Figure 2A), while categorization following their own’s center criteria. of ROH in up to 25 Mb versus more than 25 Mb showed a strong First, we performed descriptive analyses of demographic data. For and significant negative impact of ROH on both OS and AML evo- the comparison of subgroups, we applied Chi-square and Kruskal– lution (Figure 2B,C). To compensate for differing risk distributions, Wallis tests. Overall survival (OS) was defined as the time from we performed analyses adjusting for IPSS-R in 110 patients (Table 1). diagnosis to the last available follow-up or death. Time to acute myeloid Microarray results in general (considering CNA and ROH) only show a leukemia (AML) transformation was defined as the time from diagno- tendency for a lower OS in patients with a positive abnormal microar- sis to AML date. Statistical comparisons between different curves were ray result (p = 0.063) (Figure 2D). The presence of ROH greater than based on log-rank tests and were calculated for patients with available 25 Mb, however, maintains its prognostic impact for OS (p = 0.032) follow-up data (n = 138). (Figure 2E). 3 RESULTS 4 DISCUSSION According to SNP-A analysis, 73/163 (44.8%) cases showed alterations In the next-generation sequencing era, microarrays are still useful (CNA and/or ROH). Table S1 summarizes patients’ characteristics. The for routine clinical investigation. Several studies have shown their median number of alterations per patient was 0 (range: 0–8). Consider- successful applicability in MDS, especially in cases with a low number 448 MALLO ET AL . TA B L E 1 Prognostic impact of microarray alterations for both standard analysis and analysis adjusted by IPSS-R Standard analysis (n = 138) Analysis adjusted by IPSS-R (n = 110) AML Overall survival evolution AML evolution (mo) p value (mo) p value Overall survival (mo) p value (mo) p value Normal versus altered 59.73 versus 46.65 0.504 NR 0.778 53.59 versus 44.55 0.063 NR versus 65.22 0.407 Harboring ROH alone 64.69 versus 19.19 0.054 NR 0.336 45.54 versus 19.19 0.098 NR versus 65.22 0.265 (no vs. yes) ROH (<25 vs. >25 Mb) 64.69 versus 17.71 0.001* NR 0.033* 46.62 versus 19.19 0.032* NR 0.067 Abbreviations: CNA, copy number alteration; mo, months; NR, not reached; ROH, region of homozygosity. *Statistically significant. F I G U R E 2 (A) OS curve for the presence or absence of ROH alterations. (B) OS curve considering the presence of ROH greater than 25 Mb. (C) AML evolution considering the presence of ROH greater than 25 Mb. (D) OS curve, after IPSS-R adjustment, for microarray result (normal vs. altered). (E) OS curve, after IPSS-R adjustment, considering the presence of ROH greater than 25 Mb. Abbreviations: AML, acute myeloid leukemia; IPSS-R, revised international prognostic scoring system; OS, overall survival; ROH, region of homozygosity. of metaphases and/or unsuccessful cytogenetic studies [6, 7]. The first Tiu et al. [4], with 430 myeloid patients (118 MDS patients with a nor- studies appeared in 2009 [8] and showed the utility of germline control mal karyotype). In 2016, Volkert et al. [5] presented the largest series samples, especially for ROH. However, this approach is not always so far, of 520 MDS with normal karyotype. However, this study was feasible because the costs per analysis are doubled (normal plus tumor performed with Array Comparative Genomic Hybridization (aCGH), DNA). Maciejewski et al. established a cut-off of 25 Mb to distinguish which is unable to detect ROH. Compared to our series, the detection germline from acquired ROH [9]. Our series supports that 25 Mb rate of abnormalities was rather low, with 11% of cases only. The cut-off as having the most prognostic impact, even after adjustment by microarray used was 270K aCGH, whereas our series was performed IPSS-R. The prognostic impact of ROH was also stated by Palomo et al. with SNP arrays that includes more than 1.8 million probes, combining in a series of 128 chronic myelomonocytic leukemia with normal kary- copy number and SNP probes. We postulate that the difference in the otype [10]. One of the largest studies in MDS was published in 2011 by incidence of alterations (11% vs. 45%) could be due to the difference in MALLO ET AL . 449 the resolution of arrays used. The range of alterations was 174 Kb–3.4 plastic syndromes and related myeloid malignancies. Blood. Mb for Volkert et al., and 24 Kb–190 Mb for our series. However, 2011;117(17):4552–60. 5. Volkert S, Haferlach T, Holzwarth J, Zenger M, Kern W, Staller M, assessing the prognostic impact of small alterations is difficult because et al. Array CGH identifies copy number changes in 11% of 520 MDS of the lack of recurrent aberrations. Volkert et al. highlighted the prog- patients with normal karyotype and uncovers prognostically relevant nostic impact of deletions [5], but we were not able to confirm their deletions. Leukemia. 2016;30(1):257–60. prognostic impact in our cohort, in which the main adverse prognostic 6. Simons A, Sikkema-Raddatz B, de Leeuw N, Konrad NC, Hastings RJ, Schoumans J. Genome-wide arrays in routine diagnostics of hemato- factor in our cohort was the presence of ROH greater than 25 Mb. logical malignancies. Hum Mutat. 2012;33(6):941–8. Microarray guidelines are essential for diagnostic applications [11]. 7. Arenillas L, Mallo M, Ramos F, Guinta K, Barragán E, Lumbreras E, In 2018, Kanagal-Shamanna et al. published a review article focused et al. Single nucleotide polymorphism array karyotyping: a diagnostic on the clinical utility of microarrays, pointing out the utility of ROH and prognostic tool in myelodysplastic syndromes with unsuccess- ful conventional cytogenetic testing. Genes Chromosomes Cancer. detection [12]. Although molecular studies are becoming more and 2013;52(12):1167–77. more essential for MDS diagnosis and stratification [2], SNP microar- 8. Heinrichs S, Kulkarni RV, Bueso-Ramos CE, Levine RL, Loh ML, Li rays are valuable complementary techniques. An AML study using SNP C, et al. Accurate detection of uniparental disomy and microdele- microarrays showed that 50–100% of cases harboring deletions or tions by SNP array analysis in myelodysplastic syndromes with normal ROH, in distinct myeloid genes, show point mutations of the same cytogenetics. Leukemia. 2009;23(9):1605–13. 9. Maciejewski JP, Tiu RV, O’Keefe C. Application of array-based whole genes [13]. Detection of CNA and ROH is important, especially to genome scanning technologies as a cytogenetic tool in haematological determine the TP53 status in MDS. Pathogenic TP53 alterations (point malignancies. Br J Haematol. 2009;146(5):479–88. or indel mutations and 17p deletions and/or ROH) are detected in 7– 10. Palomo L, Xicoy B, Garcia O, Mallo M, Ademà V, Cabezón M, et al. 11% of MDS. About two-thirds have multiple mutation hits (multi-hit), Impact of SNP array karyotyping on the diagnosis and the outcome of chronic myelomonocytic leukemia with low risk cytogenetic features consistent with biallelic TP53 alterations. Of these, more than 90% of or no metaphases. Am J Hematol. 2016;91(2):185–92. TP53 mutated patients have complex or very complex karyotypes and 11. Schoumans J, Suela J, Hastings R, Muehlematter D, Rack K, van they can be addressed as AML-equivalent for therapy considerations den Berg E, et al. Guidelines for genomic array analysis in acquired [2, 14, 15]. haematological neoplastic disorders. Genes Chromosomes Cancer. 2016;55(5):480–91. This study highlights the importance of microarrays in MDS 12. Kanagal-Shamanna R, Hodge JC, Tucker T, Shetty S, Yenamandra patients, to detect CNAs and especially to detect acquired ROH which A, Dixon-McIver A, et al. Assessing copy number aberrations and has demonstrated a high prognostic impact. copy neutral loss of heterozygosity across the genome as best prac- tice: an evidence based review of clinical utility from the Cancer Genomics Consortium (CGC) Working Group for myelodysplastic ACKNOWLEDGMENTS syndrome, myelodysplastic/myeloproliferative and myeloproliferative Financial support: 2017 SGR288 (GRC) and CERCA Pro- neoplasms. Cancer Genet. 2018;228–229:197–217. 13. Ibáñez M, Such E, Onecha E, Gómez-Seguí I, Liquori A, Sellés J, gramme/Generalitat de Catalunya. Fundació Internacional Josep et al. Analysis of SNP array abnormalities in patients with de novo Carreras; Italy Ministry of Health. acute myeloid leukemia with normal karyotype. Sci Rep. 2020;10(1): 5904. 14. Haase D, Stevenson KE, Neuberg D, Maciejewski JP, Nazha A, CONFLICT OF INTEREST STATEMENT Sekeres MA, et al. TP53 mutation status divides myelodysplastic syn- The authors declare they have no conflicts of interest. dromes with complex karyotypes into distinct prognostic subgroups. Leukemia. 2019;33(7):1747–58. 15. Bersanelli M, Travaglino E, Meggendorfer M, Matteuzzi T, Sala C, ORCID Mosca E, et al. Classification and personalized prognostic assess- Mar Mallo https://orcid.org/0000-0001-7741-498X ment on the basis of clinical and genomic features in myelodysplastic syndromes. J Clin Oncol. 2021;39(11):1223–33. Christina Ganster https://orcid.org/0000-0002-7566-5985 Esperanza Such https://orcid.org/0000-0001-6892-403X Giovanni Marconi https://orcid.org/0000-0001-6309-0515 SUPPORTING INFORMATION Additional supporting information can be found online in the Support- REFERENCES ing Information section at the end of this article. 1. Greenberg PL, Tuechler H, Schanz J, Sanz G, Garcia-Manero G, Solé F, et al. Revised international prognostic scoring system for myelodys- plastic syndromes. Blood. 2012;120(12):2454–65. 2. Bernard E, Tuechler H, Greenberg PL, Hasserjian RP, Ossa JEA, Nannya Y, et al. Molecular international prognostic scoring system for How to cite this article: Mallo M, Tuechler H, Arenillas L, myelodysplastic syndromes. NEJM Evid. 2022;1(7):EVIDoa2200008. Raynaud S, Cluzeau T, Shih L-Y, et al. Regions of homozygosity 3. Akkari YMN, Baughn LB, Dubuc AM, Smith AC, Mallo M, Dal Cin P, confer a worse prognostic impact in myelodysplastic syndrome et al. Guiding the global evolution of cytogenetic testing for hemato- with normal karyotype. eJHaem. 2023;4:446–449. logic malignancies. Blood. 2022;139(15):2273–84. 4. Tiu RV, Gondek LP, O’Keefe CL, Elson P, Huh J, Mohamedali A, https://doi.org/10.1002/jha2.651 et al. Prognostic impact of SNP array karyotyping in myelodys-