Chromosomal Microarray Analysis in the Era of Optical Genome Mapping: Clinical Implications in Detecting Copy-Neutral Events

1. Introduction

Classical cytogenetics has served as the cornerstone for diagnostic and prognostic classification of cancers for decades but remains limited by relatively low-resolution technologies and high labor intensity[1,2]. Identification of hallmark chromosomal abnormalities is integral for clinical diagnosis, risk stratification, and therapeutic decision-making, particularly in hematologic malignancies[3,4]. Technological advances such as chromosomal microarray (CMA) have broadened the scope of cytogenetic testing, enabling genome-wide detection of copy number (CN) variants and copy-neutral events, such as copy-neutral loss-of-heterozygosity (CN-LOH)[5,6]. Several clinically significant regions of CN-LOH have been identified on 4q, 9p, 13q, and 17p, among others, and their accurate recognition has become essential for routine testing and prognostic stratification [5,7].

Current gold-standard practice in cytogenetics laboratories relies on a combination of conventional karyotyping and fluorescence in-situ hybridization (FISH)[1,3]. Identification of recurrent abnormalities, such as deletions of 5q and 7q in acute myeloid leukemia (AML), provides important prognostic information to guide treatment selection[4]. However, both techniques have limitations. For example, FISH is limited to predefined genomic regions, while karyotyping is technically challenging, time-consuming, and constrained by low-resolution capabilities[7,8,9]. These challenges have driven the need for high-resolution, genome-wide solutions.

CMA offers a genome-wide, high-resolution capability for detecting submicroscopic CN changes and regions of allelic imbalance[6]. Importantly, CMA can also detect small CN-LOH events within clinically significant regions, aiding prognostic stratification[10]. Currently, CMA has become an established complementary method to conventional karyotyping and FISH for both neoplastic and constitutional testing[3,6,10]. However, its inability to detect balanced rearrangements necessitates the use of additional molecular methods.

The limitations of current cytogenetic techniques have created demand for comprehensive, streamlined technology. Optical Genome Mapping (OGM) has emerged as such a platform, offering direct, high-resolution imaging of ultra-high-molecular-weight DNA to visualize structural variants [11,12]. Often described as “next-generation cytogenetics,” OGM is increasingly adopted by clinical laboratories for the characterization of structural variants in cancer[11,12,13,14,15]. Multiple studies have highlighted OGM’s key strengths, including excellent breakpoint resolution and high sensitivity for variants >500 bp in size[16,17,18]. Further, complex karyotypes can be easily resolved, and workflows are streamlined compared to conventional karyotyping. These features position OGM as an attractive alternative to traditional techniques, promising to consolidate multiple workflows into a single assay.

Despite these advantages, OGM’s sensitivity for CN-LOH detection is a key limitation and could be particularly problematic in disorders where biallelic inactivation is significant[15,19]. For example, homozygosity for driver mutations in genes such as JAK2, CBL, TET2, or TP53 can be essential findings for clinical management[20,21,22,23]. While OGM excels at characterization of structural rearrangements, its lack of allelic resolution represents a critical diagnostic and prognostic constraint. Accurate detection of CN-LOH remains vital for comprehensive genomic interpretation and optimal clinical management in the age of precision medicine.

As laboratories transition toward OGM adoption, the question remains whether CMA should still have a place in modern workflows. To address this, we retrospectively analyzed 53 neoplastic CMA cases collected at the University of Kansas Health System (TUKHS) that included at least one CN-LOH event. Our findings demonstrate that, although OGM would enhance structural variant detection, clinically significant focal copy-neutral events may go undetected. These results highlight the continued importance of performing CMA as a complementary modality in the ascendant era of OGM.

2. Materials and Methods

2.1. Patient Samples

327 patients with hematological malignancies at TUKHS from 2025 were included in the study. The cancer diagnosis was performed on blood or bone marrow by morphological and flow cytometry evaluations, as a standard clinical practice. Next-generation sequencing (NGS) using the QIAseq Targeted DNA Human Myeloid Neoplasms 141 gene Panel was also performed. Clinical significance of NGS variants was reported as per published guidelines [24]. The results of these evaluations were retrospectively reviewed from the patient's medical records.

Karyotyping, FISH, and Neoplastic CMA were performed at TUKHS Cytogenetics Laboratory. No patients were excluded. All testing was performed as part of routine clinical care. The retrospective review was approved by an institutional review board (Study ID: STUDY00160638, approved on 20^th June 2024).

2.2. Karyotyping and FISH Analysis

Conventional karyotyping and FISH analyses were performed on cultured peripheral blood or bone marrow specimens using standard cytogenetic protocols. A minimum of 20 metaphases were evaluated for each case. FISH testing included acute myeloid leukemia (AML) FISH probes RUNX1T1::RUNX1, KMT2A, PML::RARA, MYC, and 7q probes from Abbott, and 5q, CBFB::MYH11, MECOM, and PML::RARA from Cytocell and TP53 from MetaSystems. B-cell acute lymphoblastic leukemia (B-ALL) -associated FISH probes included BCR::ABL1, KMT2A, CDKN2A, IGH, and MYC from Abbott, ETV6::RUNX1, ABL2, PDGFRB, JAK2, EPOR, CRLF2, and P2RY8 from Cytocell, and IKZF1 from Empire Genomics. 200~500 interphase cells were scored for FISH. Chromosome and FISH analysis was performed utilizing Cytovision Software (Leica). Results were interpreted and reported using the International System for Human Cytogenomic Nomenclature (ISCN 2024).

2.3. DNA Extraction

Genomic DNA was extracted from peripheral blood or bone marrow using the QIAamp DNA Blood Mini Kit (Qiagen) according to the manufacturer’s instructions. DNA was quantified by Qubit fluorometry. At least 200 ng of total DNA was utilized for CMA.

2.4. Chromosomal Microarray

Microarray-based chromosome analysis was performed using the iScan System with the Global Diversity Array-8 (GDACyto) v1.0 Array BeadChip (Illumina). Criteria for designating reportable aberrations include gains or losses larger than 50 kb involving clinically significant cancer genes, gains >2 Mb, and losses >1 Mb outside known clinically significant oncology regions spanning at least one annotated RefSeq gene. Smaller aberrations are reported only if the regions are likely to be clinically significant. Copy-neutral loss of heterozygosity (CN-LOH) are reported when the region exceeds 3 Mb. CMA analysis and visualization were performed using NxClinical 6.2 (Bionano).

All NGS and CMA findings are annotated using the human genome reference build GRCh37/hg19.

2.5. Statistics

Descriptive statistics, including frequencies, median values, and distribution counts, were generated in Microsoft Excel or R. Figures were generated using R (v4.5.1).

3. Results

3.1. Assessing the Impact of CN-LOH Events in Routine Neoplastic Testing

TUKHS Cytogenetics Laboratory performed neoplastic CMA on 327 patients aged 18 to 92 years in 2025. 184 patients were male, 142 were female, and one identified as other. Among this cohort, 53 patients were selected who demonstrated at least one CN-LOH event, and retrospective analysis was performed to evaluate the prospective clinical value OGM could have provided in lieu of CMA. The average patient age was 64 years, with 28 male and 25 female cases. The full demographic, cytogenetic, and molecular results of our patient cohort are summarized in Table 1. These cases served as the foundation of our comparative analysis of CMA and potential OGM coverage.

A genome-wide map of CN-LOH events from our cohort is shown in Figure 1. A total of 85 CN-LOH calls were detected (median ~2 per case), ranging from focal (<1 Mb) to whole arm (>110 Mb) in size. 42 CN-LOH calls (49% of total) were <25 Mb in length, a common benchmark identified by multiple previous studies, where OGM would likely not detect these calls[14,19,25]. The distribution of CN-LOH calls by chromosome number in our cohort is highlighted in Figure 2A. The most common CN-LOH event detected involved the JAK2 gene on 9p, observed in 37% of all cases evaluated. Frequently, multiple adjacent regions of CN-LOH on 9p are identified, such as in patient 1 (Figure 2B).

We further investigated the relationship of CN-LOH, karyotype, and FISH results by primary diagnosis, as shown in Table 2. 35 of 53 cases (66%) with CN-LOH detected had a normal karyotype, with the diagnosis of MPN having the highest percentage of cases with normal karyotype occurring in 76%, followed by AML (61%), and MDS (60%). By contrast, six cases showed complex structural abnormalities (≥3 abnormalities) by karyotype. In these cases, OGM would have offered significant advantages for resolving these complex rearrangements while still identifying the large CN-LOH regions. However, CN-LOH events identified by CMA in four of these individuals were <25 Mb in length. For example, in patient 8 (MDS) and patient 53 (AML), a 17p CN-LOH was identified with 20 and 11 Mb in size, respectively, which included the tumor suppressor gene TP53 (17p13.1) (Figure 3A). TP53 p.Y234D and p.Y236C NGS variants were also present at 56% and 86% variant allele fraction (VAF), respectively. These findings are consistent with biallelic TP53 gene inactivation.

Further analysis revealed that 41 of the 53 patients (77.4%) harbored clinically actionable NGS-identified variants within the CN-LOH regions. Genes in these regions included FLT3 (13q12.2) (seen in 2/53 (3.7%) total; 2/14 (14.2%) AML), JAK2 (9p24.1) (20/53 (37.7%) total; 2/14 (14.2%) AML; 1/15 (6.6%) MDS; 16/21 (76.1%) MPN, one case of CMML), TET2 (4q24) (4/53 (7.5%) total; 2/15 (13.3%) MDS; 1/21 (4.7%) MPN; one case of CMML), RUNX1 (21q22.12) (5/53 (9.4%) total; 4/14 (28.6%) AML; one case of CML), MPL (1p34.2) (2/53 (3.7%) total; 2/21 (9.5%) MPN), TP53 (17p13.1) (2/53 (3.7%) total; one case of MDS, one case of AML), EZH2 (7q36.1) (seen in 2/53 (3.7%) total; 2/15 (13.3%) MDS), NOTCH1 (9q34.3) (patient 15 with T-ALL), CDKN2A (9p21.3) (patient 31 with B-ALL), BCOR (Xp11.4) (patient 34 with MDS), SETBP1 (18q12.3) (patient 12 with MDS/MPN), U2AF1 (21q22.3) (patient 40 with MDS). Although TET2 single- and double-hit mutations were detected in 4/14 (28.6%) AML cases by NGS, CN-LOH was not present. CN-LOH of 21q involving RUNX1 mutation, the sole abnormality detected in AML cases, was associated with a normal or intermediate-risk karyotype; an example is shown for patient 18 (Figure 3B). These findings highlight the complementary value of integrating CMA with NGS, as CN-LOH can augment the impact of existing variants.

3.2. Newly Diagnosed B-ALL cases

To better understand OGM’s advantages with respect to our own testing cohort, we identified 14 patients with newly diagnosed B-ALL who had concurrent karyotyping, FISH, and CMA performed among our total CMA testing cohort of 327 patients. All newly diagnosed B-ALL patients at our institution undergo CMA. The full clinical results of our B-ALL cohort are summarized in Table 1 and Table 3. The average patient age was 53 years, with 11 male and 3 female cases.

Most of our B-ALL cohort demonstrated both abnormal karyotype and FISH results. Our laboratory routinely performs BCR::ABL1 “STAT” FISH testing on peripheral blood, followed by a cascade of additional probes, including KMT2A, IKZF1, CDKN2A, ETV6::RUNX1, and IGH. Cases reported positive for BCR::ABL1 reflex to a Ph-like panel, including ABL2, PDGFRB, JAK2, EPOR, and MYC FISH probes. Further, CRLF2-positive immunophenotypes identified by flow cytometry trigger CRLF2 and P2RY8 FISH confirmatory testing. Collectively, this workflow often results in numerous FISH probes per case, even when many assays yield normal results. For example, patient 13 required 11 separate FISH assays, 10 of which were normal, as shown in Table 3.

CN-LOH was only identified in one patient (patient 31, Table 1) and was >25 Mb in length. However, focal copy-number changes <25 Mb were identified in 7 of the 13 successful CMA studies. Of note, one patient had cancelled CMA because a low-hypodiploid karyotype was detected, with high prognostic risk.

4. Discussion

Our retrospective study highlights CN-LOH as a frequent and clinically significant finding in neoplastic testing. In our cohort, nearly half of the CN-LOH events observed were <25 Mb in size and below the conventional OGM reporting threshold. Despite their smaller sizes, they encompass mutations in critical driver genes with direct implications for risk stratification and treatment selection, particularly when accompanied by NGS variants within the CN-LOH region. CMA offers distinct advantages derived from its single-nucleotide polymorphism (SNP)- based array design. The platform provides genome-wide coverage, enabling the detection of submicroscopic CN changes and copy-neutral events such as CN-LOH, uniparental disomy (UPD) and regions of homozygosity[6]. For example, CN-LOH involving the TP53 gene, along with a clinically actionable TP53 variant identified by NGS, has significant prognostic value in MDS and AML patients. The resulting biallelic inactivation of TP53 in MDS (patient 8) meets the criteria for the MDS-defining genetic subtype, MDS with biallelic TP53 inactivation (MDS-biTP53), per the World Health Organization (WHO) and National Comprehensive Cancer Network (NCCN.org), and is associated with increased bone marrow blasts and a high risk of progression to leukemia and death, independent of treatment[26,27]. In AML with biallelic TP53 mutations (patient 53), individuals fall into a high-risk subset within the poor prognostic category of recurrent genetic abnormalities. Accurate determination of TP53 biallelic status is increasingly critical for guiding precision medicine decisions, including recommendations for hematopoietic cell transplantation or clinical trial enrollment[28]. Detecting focal CN-LOH events involving the TP53 locus provides essential insight into biallelic involvement, information that would be missed by OGM alone.

In our study, several 9p CN-LOH events co-occurred with JAK2 mutations identified by NGS, a pattern associated with increased mutation burden, clonal expansion, and a higher risk of progression to myelofibrosis or accelerated phase, underscoring the importance of detecting these copy-neutral events[29,30]. Two AML patients harbored FLT3 internal tandem duplications with CN-LOH spanning the FLT3 locus; prior work has linked this combination to higher relapse rates and poorer overall survival[31]. Additionally, MPL W515 mutations accompanied by CN-LOH involving the MPL locus were observed in patients with myelofibrosis. Acquired 1p CN-LOH affecting MPL has been implicated in fibrotic transformation in MPNs carrying MPL mutations[32].

CN-LOH involving 4q, encompassing TET2, was also frequently observed. TET2 mutations are common in MDS, AML, and clonal hematopoiesis, and are associated with worse prognosis in MDS[33]. TET2 inactivation through CN-LOH has been linked to increased proliferation, inflammation, and self-renewal[34]. Although not a distinct disease subtype, patients with high-VAF TET2 mutations and CN-LOH carry a high risk of progressing from early leukemic states, such as clonal hematopoiesis, to overt malignancy. Early identification of these high-risk TET2 events may enable earlier intervention and risk-reduction strategies. CN-LOH of 21q is another recurrent abnormality in AML[31,35]. In our cohort, 21q CN-LOH co-occurring with a pathogenic RUNX1 variant was detected in 28% of AML cases. While one prior study reported favorable outcomes in AML with 21q CN-LOH, RUNX1 mutation status was not evaluated[36]. In contrast, biallelic RUNX1 alteration resulting from CN-LOH combined with a RUNX1 sequence variant has been associated with a worse prognosis compared with monoallelic mutation[31,37]. These findings underscore the importance of integrating CMA and NGS for accurate clinical interpretation and highlight CMA’s high-resolution CN-LOH detection in guiding clinical decision-making.

OGM provides superior structural variant detection, high-resolution breakpoint mapping, and the ability to consolidate multiple cytogenetic workflows into a single assay[12,14,18]. These advantages are especially valuable in cases with complex rearrangements that are difficult to resolve by conventional karyotyping. However, practical considerations, including institutional resources and technical requirements, must be addressed when developing an OGM-based workflow. Implementation requires a significant upfront investment, including ultra-high molecular weight DNA extraction, specialized instrumentation, and substantial computational infrastructure, which may be challenging for smaller laboratories with limited budgets or support. Although these barriers are expected to diminish as the technology matures, they currently remain a constraint for many centers.

Across myeloid malignancies, including MPN, MDS, and AML, 60–76% of cases showed normal karyotypes with CN-LOH identified only by CMA (Table 2). Recent data indicate limited clinical utility of OGM in MPN: aside from detecting KMT2A-PTD in a subset of karyotypically normal cases, OGM did not reveal additional tier 1 or tier 2 copy-number or structural variants in MDS or MDS/MPN, nor did it identify any new tier 1 abnormalities in myeloid malignancies with complex karyotypes. In AML, tier 2 OGM-detected abnormalities mainly involved MECOM, KMT2A, and NUP98 rearrangements,[38] most of which would be captured by CMA and the AML FISH panel used in this study for adult and elderly patients; NUP98 FISH may be useful in pediatric settings. Based on these findings, we propose that laboratories without OGM perform karyotyping, a heme NGS panel, and CMA for MPN, MDS, and MDS/MPN, with reflex FISH testing based on karyotype findings for future minimal residual disease detection. For adult AML, we suggest karyotyping, targeted FISH for actionable rearrangements requiring rapid turnaround, CMA, and a heme NGS panel. For laboratories implementing OGM, testing should include limited FISH for urgent actionable rearrangements, CMA until segmental or focal CN-LOH detection is available in OGM, a heme NGS panel, and OGM.

Newly diagnosed B-ALL requires extensive cytogenetic evaluation, often using multiple targeted FISH probes that are costly, labor-intensive, and restricted to predefined loci. OGM offers a comprehensive, streamlined alternative. Although it cannot detect focal CN-LOH, these events are less common in B-ALL than in myeloid neoplasms[39]. In contrast, the structural variants, aneuploidies, fusions, and rearrangements that characterize B-ALL are strengths of OGM detection. In our cohort, OGM would have identified all clinically relevant abnormalities targeted by FISH in a single assay, reducing cost, labor, and turnaround time. Many focal copy-number alterations detected by CMA, such as CDKN2A, RB1, and IKZF1 losses, also fall within OGM’s resolution for copy-number gains and losses (~500 bp), and OGM would concurrently detect structural rearrangements missed by CMA. We propose that diagnostic workflows for B- and T-ALL include karyotype, FISH, a heme NGS panel, and CMA for laboratories without OGM. For laboratories with OGM, we propose limiting FISH to rapidly actionable rearrangements while incorporating OGM and a heme NGS panel.

Although whole-genome sequencing (WGS) provides the most comprehensive assay for detecting CN changes, SVs, and CN-LOH, its clinical adoption is limited by the substantial sequencing depth, computational resources, and bioinformatics expertise required. Similar challenges apply to combined NGS-plus-SNP approaches and whole-exome sequencing (WES)[40]. WES can infer CN-LOH using BAF and read-depth modeling, but sensitivity is far lower than with WGS or SNP arrays due to limited exonic coverage. Accordingly, WGS, WES, and hybrid NGS-plus-SNP workflows were not included in the proposed testing recommendations for routine clinical use.

The limitations of this study include its retrospective design and modest cohort size, which constrain generalizability. Because OGM was not directly performed, concordance could not be empirically evaluated; incorporating OGM in future analyses would strengthen the findings. However, prior reports indicating limited OGM sensitivity for focal CN-LOH suggest that many such events in our cohort would likely remain undetected[41]. Larger, multi-institutional studies across diverse hematologic malignancies, integrating karyotyping, FISH, CMA, OGM, and NGS, are needed to better define the complementary roles of these technologies and guide test selection based on available resources.

5. Conclusions

While OGM represents a significant advance in cytogenetic technology, our analysis demonstrates that CMA remains an essential diagnostic resource for detecting clinically relevant CN LOH events. Rather than replacement, we suggest that the future of cytogenetic testing lies in strategic integration, where the strengths of multiple technologies are utilized to deliver a complete and actionable genetic profile. This combined approach ensures that no significant event is overlooked in the pursuit of precision oncology.