The prognostic significance of single‐nucleotide polymorphism array‐based whole‐genome analysis and uniparental disomy in myelodysplastic syndrome

Abstract Introduction Myelodysplastic syndrome (MDS) is a group of heterogeneous hematological diseases characterized by ineffective hematopoiesis and dysplastic morphology. Single nucleotide polymorphism array (SNP‐A)‐based whole genome analysis has a much higher resolution for chromosomal alterations when compared with conventional cytogenetic tools. In the present study, we evaluated the diagnostic value and prognostic significance of SNP‐A in MDS patients with normal karyotypes. Methods A total of 127 patients with MDS and myeloproliferative neoplasms or acute myeloid leukemia with myelodysplasia‐related changes were included in our study. The advantages and disadvantages of SNP‐A were compared with those of traditional metaphase cytogenetic analysis (MC). The Kaplan‐Meier analysis and COX regression analysis were used to investigate the prognostic value of SNP‐A and uniparental disomy (UPD) in MDS patients with normal karyotype. Furthermore, the chromosomal abnormalities detected by SNP‐A in patients with specific gene mutations were explored. Results SNP‐A was more sensitive toward meaningful chromosomal aberrations (58.2% vs 36.9%; P < .05) than MC. Among the patients with normal karyotype, those who were detected with new chromosomal abnormalities via SNP‐A presented with inferior survival compared with those without the abnormalities (P = .003). Additionally, the presence of UPD was an independent prognostic factor in patients with normal karyotype (P = .01). TP53 and RUNX1 mutations often occurred with abnormalities in chromosomes 17p and 21q, respectively. Conclusions Compared with MC, SNP‐A capable of detecting UPD can offer more diagnostic and prognostic information; TP53 and RUNX1 gene mutations are often accompanied by abnormalities in their chromosomes (17p, 22q).


| INTRODUC TI ON
Myelodysplastic syndrome (MDS) is a group of hematological diseases characterized by ineffective hematopoiesis and a dysplastic morphology; some patients with excessive marrow blasts may present with acute myeloid leukemia (AML). 1 MDS is a heterogeneous type of disease, that is, the course of the disease may vary in the patient based on their clinic-pathological features (including age and history of chemotherapy, among others).
Besides the blast percentage, patients with MDS may be classified into different risk groups using prognosis scoring systems. The International Prognostic Scoring System (IPSS) and the revised IPSS (IPSS-R) have been most commonly used to evaluate the prognosis of patients with MDS, and the cytogenetic index occupies a significant position in these systems. 2,3 However, the clinical manifestations of this disease may differ among the patients in the low-risk and intermediate-1 (int-1) groups when evaluated by IPSS or IPSS-R. Some lowrisk patients may present with a progressive course, whereas several patients in the int-1 group may have a better prognosis. Several other important scoring systems for MDS, including the MD Anderson Cancer Center MDS model, mainly target patients belonging to the low-risk and int-1 groups but do not yield idealistic effects. 4 Nearly half (50%) of patients with MDS are known to present with chromosomal abnormalities. 5,6 Cytogenetic alteration has been determined as an essential prognostic factor in these patients.
Conventional cytogenetic tools include metaphase cytogenetic analysis (MC) and fluorescence in situ hybridization (FISH). A routine MC can be used to examine numeric and structural alterations in the chromosomes at the single-cell level. Despite its low sensitivity and the need for technical proficiency, MC has a remarkable advantage in detecting novel cytogenetic abnormalities. 7 Additionally, FISH can be used to detect numeric and structural alterations in the chromosomes, but it cannot discover novel alterations, despite its superior sensitivity and ease of handling. Thus, MC remains as the main tool used to examine alterations in the chromosomes, whereas FISH is used as a supplement for sensitivity.
Single nucleotide polymorphism array (SNP-A)-based genomewide analysis technology can be used to examine the imbalanced alterations of somatic or clonal cells in hematopoietic diseases. 1 This method has a much higher resolution when compared with conventional cytogenetic tools and has the advantage of discovering unknown potential alterations when compared with FISH. 8,9 Additionally, SNP-A could detect uniparental disomy (UPD), which could not be detected by MC or FISH. 1 UPD refers to a pair of homologous chromosomes in which one (paternal or maternal chromosome) is duplicated and the opposite one is deleted. 10 Primary UPD is associated with errors in meiosis and may lead to some growth-and development-related diseases. 11 Acquired UPD, such as other chromosomal abnormalities, can be used as a marker of cloning in malignant tumors. 12 An increasing number of studies have found that cancer cells may gain clonal advantages through acquired UPD, such as homozygous mutation of a part of the JAK2 V617F gene associated with UPD 9p. 13,14 Some gene mutations often occur with abnormalities of the chromosome in which they are located. 15 However, the prognostic significance of the presence of UPD in hematologic malignant diseases remains to be determined. Additionally, the association between chromosome abnormalities and molecular genetics should be examined.
In the present study, we evaluated the cytogenetic characteristics of MDS patients, compared the detection yields of the chromosome aberrations between the SNP-A and conventional cytogenetic examinations, and determined the prognostic significance of SNP-A and UPD.

| Data collection and follow-up
The patients were selected using the Hospital Information System (HIS), and general clinical data, including gender, age, and time of onset of the disease, were collected. The Laboratory Information System was used to obtain the results of the complete blood counts (including the hemoglobin, platelet count, white blood cell count, neutrophil percentage, and peripheral blood blast cell percentage) and bone marrow smear image analysis (marrow blast cell percentage and presence of myelodysplasia). The MC, FISH, SNP-A, and myeloid neoplasm-related gene mutation examinations were conducted by second-generation sequencing. The clinical data were graded according to IPSS and IPSS-R, and the survival conditions were followed up via telephone calls.

| Statistical analysis
The R language (Bell Laboratories; Lucent Technologies) was used for statistical analysis. The chi-square test and Fisher's exact probability method were used to compare the merits and demerits of SNP-A and the conventional method of cytogenetic examination. The Kaplan-Meier (KM) method and COX regression analysis were used ("survival" package 17 ) to analyze the overall survival (OS) of the patients in the different subgroups, classified according to the cytogenetic results and to draw the survival curves. Chisquare and Fisher's exact probability tests were used to test the independence of the chromosomal abnormalities and hot-spot gene mutations. The "karyoploteR" package was used to map the genomic alterations. 18 Hypothesis testing was statistically significant when P < .05.

| Cytogenetic and genetic examinations
In the present study, one hundred and twenty-two patients completed MC, whereas sufficient cytogenetic results were not obtained from five patients because of the lack of the mitotic phase, and detailed MC results are shown in Figure 1B

| Merits and demerits of SNP-A compared with MC
The results of SNP-A and MC were improved in 122 patients. As shown in Table 1, the positivity of SNP-A for significant chromosomal aberrations was higher than that observed with the MC

| Prognostic significance of SNP-A and UPD
The effect of the newly discovered abnormalities via SNP-A on the

| Compared with the MC, the resolution of SNP-A is superior to detect significant chromosome aberrations
The SNP array is designed using a comparator DNA, RNA, or tissue that is arrayed on a glass slide or glass beads, instead of a normal human control. 24

ACK N OWLED G EM ENTS
The authors thank Miss Long Yichen of the Institute of Biophysics at the China Academy of Sciences for her help with the methodology and statistics.

CO N FLI C T O F I NTE R E S T
The researchers claim no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
Author elects to not share data.