# Chromosomal Microarray Analysis: Copy Number Variants in Clinical Practice

## Platform Technologies

### Array Comparative Genomic Hybridization (aCGH)

In array comparative genomic hybridization, patient and reference DNA are differentially labeled with fluorescent dyes (such as Cy3 and Cy5) and co-hybridized to an array of genomic probes. The signal ratio at each probe indicates relative copy number: a ratio greater than 1 indicates a gain, less than 1 indicates a loss, and equal to 1 indicates normal copy number. A key limitation of aCGH is that it cannot detect copy-neutral changes such as uniparental disomy or absence of heterozygosity. The resolution of the platform depends on probe density and spacing; clinical arrays typically have probes spaced every 50 to 100 kilobases genome-wide, with denser coverage in known disease regions.

### SNP Arrays

SNP arrays hybridize patient DNA to an array of single nucleotide polymorphism probes, providing both copy number data and genotype information through allele-specific signal intensities. The B-allele frequency (BAF) plot is a powerful visualization tool that reveals regions of homozygosity (long contiguous stretches of homozygosity), uniparental disomy (either whole chromosome or segmental), triploidy and tetraploidy, and low-level mosaicism through shifts in BAF patterns. Critically, SNP arrays can detect copy-neutral absence of heterozygosity (CN-LOH), which aCGH alone cannot.

### Combined Platforms

Many clinical laboratories now use combined aCGH plus SNP platforms, such as Agilent CGH+SNP or Affymetrix CytoScan arrays, which provide the advantages of both technologies: high-resolution CNV detection and genotype data. The ACMG recommends chromosomal microarray as the first-tier cytogenetic test for developmental delay, intellectual disability, autism spectrum disorder, and multiple congenital anomalies.

| Feature | aCGH | SNP Array | Combined aCGH+SNP |
|---|---|---|---|
| Copy number detection | Yes | Yes | Yes |
| Genotype/allele data | No | Yes | Yes |
| Detects UPD (CN-LOH) | No | Yes | Yes |
| Detects consanguinity | No | Yes | Yes |
| Detects low-level mosaicism | Limited | Yes (BAF shifts) | Yes |
| Detects balanced rearrangements | No | No | No |
| Probe density control | Flexible | Fixed SNP positions | Both |

## Copy Number Variant Interpretation

### ACMG/ClinGen CNV Classification Framework

Copy number variants are classified using a five-tier system analogous to sequence variant classification. Pathogenic CNVs are clearly disease-causing based on established evidence. Likely pathogenic CNVs have strong supporting evidence. Variants of uncertain significance (VUS) lack sufficient evidence for definitive classification. Likely benign CNVs have strong evidence supporting a benign nature. Benign CNVs are clearly benign based on frequency data and literature.

### Evidence for CNV Classification

Multiple lines of evidence inform CNV classification. Larger CNVs containing more genes are more likely to be pathogenic, though small CNVs encompassing critical dosage-sensitive genes can also be pathogenic. Gene content is evaluated using haploinsufficiency and triplosensitivity scores curated by ClinGen. Population frequency in databases such as the Database of Genomic Variants (DGV) or gnomAD-SV helps establish whether a CNV is common and likely benign. Inheritance patterns matter: de novo CNVs are more likely pathogenic, though inherited CNVs may still be pathogenic due to variable penetrance. Overlap with known syndromes is assessed through resources like DECIPHER and the ClinGen dosage sensitivity map. Finally, the phenotypic fit -- whether the genes within the CNV are consistent with the patient's clinical presentation -- is considered.

### Recurrent Microdeletion/Microduplication Syndromes

Recurrent microdeletion and microduplication syndromes are mediated by non-allelic homologous recombination (NAHR) between flanking segmental duplications, also known as low-copy repeats. These syndromes have well-characterized critical regions. The 22q11.2 deletion (DiGeorge/velocardiofacial syndrome) spans 1.5 to 3 megabases, occurs in about 1 in 4,000 births, and features cardiac defects, palatal anomalies, immune deficiency, hypocalcemia, and learning difficulties. The 7q11.23 deletion (Williams syndrome) spans about 1.5 megabases and causes supravalvular aortic stenosis, hypercalcemia, a gregarious personality, and intellectual disability. The 15q11.2-q13 deletion causes Prader-Willi syndrome when the paternal copy is deleted or Angelman syndrome when the maternal copy is lost. The 17p11.2 deletion causes Smith-Magenis syndrome, characterized by intellectual disability, sleep disturbance, and self-injurious behavior, while the reciprocal duplication at the same locus causes the milder Potocki-Lupski syndrome. The 16p11.2 deletion and duplication are associated with autism spectrum disorder, obesity in deletion cases, and being underweight in duplication cases, with variable expressivity. The 1q21.1 deletion is associated with microcephaly and cardiac defects while the duplication is associated with macrocephaly, both with variable neurodevelopmental phenotypes. The 15q13.3 deletion causes intellectual disability, epilepsy, and behavioral issues with incomplete penetrance.

### Variants of Uncertain Significance (VUS)

VUS constitute a significant proportion of microarray findings, representing 5 to 10% of all results. Parental testing through trio analysis is essential because de novo status substantially upgrades the classification, whereas inherited status may not resolve the significance. VUS should not be used for clinical decision-making unless supporting evidence accumulates over time. Periodic reassessment as new data emerge is recommended. Reporting findings to databases such as ClinVar and DECIPHER contributes to the broader knowledge base, and referral to research studies for functional characterization can be considered.

## Clinical Applications and Guidelines

### First-Tier Testing Indications (ACMG 2010 Guideline)

Chromosomal microarray is recommended as the first-tier test for intellectual disability or developmental delay of unknown etiology, autism spectrum disorder, and multiple congenital anomalies. The diagnostic yield in these populations is 15 to 20%, which is significantly higher than the approximately 3% yield from conventional karyotyping.

### Prenatal Applications

Microarray is recommended when structural anomalies are detected on prenatal ultrasound, providing an additional yield of about 6% over karyotype. In structurally normal pregnancies with advanced maternal age or positive screening, the additional yield is approximately 1.7% over karyotype. ACOG and SMFM recommend offering microarray as an option to all patients undergoing invasive prenatal diagnosis. Challenges in the prenatal setting are significant: VUS findings create a substantial counseling burden when pregnancy decisions may be time-sensitive, susceptibility loci with incomplete penetrance are particularly problematic in the prenatal context, and parental studies should ideally be performed concurrently when possible.

### Limitations of Microarray

Chromosomal microarray cannot detect balanced rearrangements such as reciprocal translocations, inversions, or balanced insertions. It also cannot detect point mutations or small insertions and deletions. Low-level mosaicism below approximately 15 to 20% may be missed, though this threshold varies by platform. Microarray does not assess epigenetic changes such as methylation, which is relevant for imprinting disorders. Noncoding CNVs located in gene deserts may be difficult to interpret, and some platforms have reduced coverage in certain genomic regions.

## Regions of Homozygosity (ROH)

### Clinical Significance

Long contiguous stretches of homozygosity greater than 5 to 10 megabases may indicate consanguinity (when multiple ROH regions are scattered across multiple chromosomes), uniparental disomy (when whole chromosome or segmental ROH appears on a single chromosome), or ancestral homozygosity (common in populations with founder effects or genetic isolation). Clinical follow-up for UPD depends on which chromosome is involved. For chromosomes harboring known imprinted genes (6, 7, 11, 14, 15, and 20), UPD may directly cause disease. For other chromosomes, UPD may unmask autosomal recessive conditions.

### Consanguinity Detection

Total ROH exceeding 1% of the genome is suggestive of parental relatedness. First-cousin parents are expected to share approximately 6.25% of the genome as homozygous regions (inbreeding coefficient F = 1/16). This information must be discussed with sensitivity, as it may reveal non-paternity or unexpected family relationships.

<image>A schematic diagram showing SNP array data visualization for three clinical scenarios arranged vertically: (1) Normal diploid: log2 ratio at 0 (two copies) and B-allele frequency showing three bands at 0, 0.5, and 1 representing AA, AB, and BB genotypes; (2) Heterozygous deletion: log2 ratio shifted down to approximately -0.5 in the deleted region and BAF showing loss of the middle heterozygous band (0.5) with only AA and BB bands visible; (3) Copy-neutral loss of heterozygosity (UPD): log2 ratio normal at 0 but BAF showing absence of the heterozygous band across a chromosomal segment, indicating two copies from a single parent. Each scenario has a chromosome ideogram below with the affected region highlighted.</image>

<image>A genomic map showing the 22q11.2 region with common deletion breakpoints labeled LCR-A through LCR-D (low copy repeats shown as colored arrows). The typical 3 Mb deletion (LCR-A to LCR-D), the nested 1.5 Mb deletion (LCR-A to LCR-B), and the distal deletion (LCR-D to LCR-E) are shown as horizontal bars. Key genes within the region are labeled (TBX1, COMT, PRODH, DGCR8). Below, a clinical features panel lists the major phenotypic features associated with the deletion: conotruncal cardiac defects, palatal anomalies, thymic hypoplasia/immune deficiency, hypocalcemia, renal anomalies, and neurodevelopmental differences. Arrows indicate which genes contribute to specific features where known.</image>

<image>A decision algorithm flowchart for interpreting a CNV identified on chromosomal microarray. Starting from "CNV Identified," the flow branches based on: (1) Is it in a well-established syndrome region? If yes, classify as pathogenic/likely pathogenic. (2) If not, assess gene content -- does it contain dosage-sensitive genes (ClinGen haploinsufficiency score 3)? (3) Check population frequency in DGV/gnomAD-SV -- present at greater than 1% suggests benign. (4) Determine inheritance -- de novo supports pathogenicity, inherited requires further assessment of parent phenotype. (5) Size consideration -- larger CNVs with multiple genes carry higher prior probability of pathogenicity. Terminal nodes show the five-tier classification outcomes (pathogenic, likely pathogenic, VUS, likely benign, benign) with recommended clinical actions for each.</image>

## Clinical Pearls

Chromosomal microarray is the recommended first-tier cytogenetic test for unexplained developmental delay, intellectual disability, autism spectrum disorder, and multiple congenital anomalies, with a diagnostic yield of 15 to 20%. SNP arrays provide genotype data that aCGH alone cannot, enabling detection of uniparental disomy, consanguinity, and low-level mosaicism. Parental trio testing is essential for interpreting CNVs of uncertain significance, as a de novo CNV is much more likely to be pathogenic. In the prenatal setting, microarray provides approximately 6% additional diagnostic yield over karyotype when fetal structural anomalies are present, but VUS findings create significant counseling challenges. Microarray cannot detect balanced rearrangements, so when a balanced translocation or inversion is suspected based on family history, conventional karyotype or genome sequencing is required. Recurrent microdeletion and microduplication syndromes are mediated by NAHR between flanking segmental duplications and have relatively predictable, though variable, phenotypes. Regions of homozygosity on SNP array may reveal consanguinity or UPD, both of which have distinct clinical implications requiring sensitive counseling.

## References

- Miller DT et al. Consensus statement: chromosomal microarray is a first-tier clinical diagnostic test for individuals with developmental disabilities or congenital anomalies. Am J Hum Genet. 2010;86(5):749-764.
- Riggs ER et al. Technical standards for the interpretation and reporting of constitutional copy-number variants: a joint consensus recommendation of ACMG and ClinGen. Genet Med. 2020;22(2):245-257.
- Wapner RJ et al. Chromosomal microarray versus karyotyping for prenatal diagnosis. N Engl J Med. 2012;367(23):2175-2184.
- South ST et al. ACMG Standards and Guidelines for constitutional cytogenomic microarray analysis. Genet Med. 2013;15(11):901-909.
- Kaminsky EB et al. An evidence-based approach to establish the functional and clinical significance of copy number variants in intellectual and developmental disabilities. Genet Med. 2011;13(9):777-784.