Microarray‐based comparative genomic hybridisation reveals additional recurrent aberrations in adult patients evaluated for myelodysplastic syndrome with normal karyotype

Abstract

The myelodysplastic syndromes (MDS) are a heterogeneous group of haematopoietic stem cell disorders characterized by ineffective haematopoiesis, dysplasia, cytopenia and a propensity to evolve towards acute myeloid leukaemia (AML). MDS affect predominantly the elderly with median age at diagnosis above 70 years. According to the European LeukemiaNet, for diagnosis of adult MDS, examinations of peripheral blood, bone marrow aspirate and biopsy and chromosome banding analysis (CBA) are mandatory and fluorescence in-situ hybridisation (FISH) and flow cytometry immunophenotyping are recommended, whereas microarraybased comparative genomic hybridisation (aCGH) and molecular genetics are suggested for special circumstances (Malcovati et al, 2013). CBA detects aberrations in approximately 50% of de novo MDS (Platzbecker et al, 2012). Recent studies have shown that high resolution techniques increase the abnormality rate. To the best of our knowledge, the majority of aCGH studies in MDS involve series of patients with known diagnoses. Therefore, we sought to evaluate the additional information obtained by aCGH compared to CBA in a large cohort of patients undergoing primary diagnostic evaluation for suspected MDS. Bone marrow samples (n = 237) referred for diagnostic evaluation of MDS were analysed. Informed consent was obtained from all patients. Subsequently, MDS was clinically confirmed in 135 (57%) cases (median age, 71 years; range: 14–94 years) but could not be established in 69 (29%) patients with cytopenia (median age, 67 years; range: 5–87 years), (non-confirmed MDS patients). Other haematological malignancies were diagnosed in 33 (14%) patients, who were excluded (Table SI). Chromosome banding analysis was performed according to the guidelines of the Association for Clinical Cytogenetics, and was successful in 97% confirmed MDS and 94 2% nonconfirmed MDS. 33 1% confirmed MDS patients had an abnormal karyotype. This percentage of abnormal cases is lower than the expected rate of approximately 50%. However, the reported rates vary from 30-60% depending on the MDS risk types included. In our cohort 21 5% MDS patients had refractory anaemia with excess blasts type1/2. Thus, our finding is concordant to with similar recently published cohorts (Ciabatti et al, 2017). Nine (13%) non-confirmed MDS patients had an abnormal karyotype. Microarray-based comparative genomic hybridisation was performed using the AFFYMETRIX CYTOSCAN HD ARRAY KIT (Affymetrix, Santa Clara, CA, USA). Single samples were analysed with the CHROMOSOME ANALYSIS SUITE software (Affymetrix) using cut-offs of 500 Kb for copy number alterations (CNA), excluding the known copy number variants (CNVs) listed in the Database of Genomic Variants (http://dgv.tca g.ca/dgv/app/home), and 10 Mb for copy neutral loss of heterozygosity (cnLOH) unless genes or regions known to be relevant for haematological malignancies were involved or mosaicism was observed (O’Keefe et al, 2010; Schoumans et al, 2016). The cut-off for CNA (5 Mb) proposed by Schoumans et al (2016) was not applied, because it is just a proposition to facilitate and standardize the interpretation of somatic results. aCGH was successful in all cases, and was abnormal in 52 6% confirmed MDS cases and 29% patients with non-confirmed MDS (Tables SII and SIII). Microarray-based comparative genomic hybridisation increased the detection rates of cytogenetic aberrations by 19 5% for confirmed MDS, which is within the expected range, and 16% for non-confirmed MDS. However, CBA analysis detected low-level mosaicism or balanced aberrations in 6 confirmed MDS and 2 non-confirmed MDS cases (Table SIV), which highlights the utility of CBA and indicates that it cannot be replaced by aCGH. Microarray-based comparative genomic hybridisation identified abnormalities that were not detected by CBA in 40 (29 6%) MDS cases and 12 (17 4%) non-confirmed MDS cases (Tables I and II). MDS patients with normal karyotype based on CBA seem to have reduced survival if cryptic abnormalities are present. However, in contrast to gains, only the presence of submicroscopic deletions seems to adversely affect outcome in MDS (Volkert et al, 2016). In our study, cnLOH were observed in 18 5% of MDS, involving almost all chromosomes (Fig S1). Some of the deletions and cnLOH identified in our study are known to contain genes that are possibly relevant in myeloid neoplasms or known tumour suppressor genes. It has been suggested that MDS patients with cnLOH in 7q, 17p, or 11q could be classified into higher cytogenetic risk groups (Makishima et al, 2010). cnLOH 7q was detected in 3 MDS cases and one non-confirmed MDS. cnLOH 11q and 17p were seen in 2 and 3 MDS cases, respectively and in no non-confirmed MDS Correspondence