Home Molecular Testing Methods Chromosomal Microarray Test: Copy Number Variants, Deletions, and Duplications

Molecular Testing Methods

Chromosomal Microarray Test: Copy Number Variants, Deletions, and Duplications

August 5, 2026 Modified date: August 5, 2026

Understand how chromosomal microarray detects deletions and duplications, what pathogenic and uncertain CNV results mean, what CMA misses, and when follow-up testing is needed.

A chromosomal microarray test scans DNA for missing or extra segments called copy number variants. It can find many deletions and duplications that are too small to appear on a routine karyotype, including changes linked to developmental delay, intellectual disability, autism, congenital anomalies, pregnancy ultrasound findings, and some reproductive problems. The test is also called chromosomal microarray analysis, CMA, molecular karyotyping, array comparative genomic hybridization, or SNP array testing. Its strength is broad, genome-wide copy-number detection without requiring the laboratory to suspect one particular syndrome first. Its limits are equally important: a microarray usually cannot detect balanced translocations, most single-letter DNA variants, many repeat expansions, or all forms of low-level mosaicism. Results may be pathogenic, likely pathogenic, uncertain, likely benign, or benign. The size of a finding alone does not determine its meaning; gene content, inheritance, population evidence, and the person’s clinical features all matter.

CMA detects DNA gains and losses across the genome: It is especially useful for unexplained developmental differences, multiple birth defects, and selected prenatal indications.
A deletion means DNA is missing; a duplication means extra DNA is present: Either can be harmless, disease-causing, or uncertain depending on the region and genes involved.
A normal microarray does not rule out a genetic condition: Sequence variants, balanced rearrangements, repeat expansions, and some mosaic changes require other methods.
Parental testing can clarify a child’s result: A new, or de novo, CNV often carries different evidence than the same finding inherited from an unaffected parent.
Prenatal CMA is diagnostic for detected copy-number changes: It is performed on chorionic villus or amniotic fluid cells, not on routine maternal blood screening.
Variants of uncertain significance should not be treated as a diagnosis: Reanalysis, family studies, and clinical follow-up may change classification over time.

How Chromosomal Microarray Works
When Chromosomal Microarray Is Used
Array Types, Resolution, and Laboratory Design
Copy Number Results and Classification
Inheritance, Mosaicism, and Family Testing
What a Microarray Cannot Reliably Detect
Chromosomal Microarray During Pregnancy
Next Steps After a CMA Result

How Chromosomal Microarray Works

Human DNA is packaged into 23 chromosome pairs. Most people have two copies of each autosomal region, one inherited from each parent, plus the expected number of X and Y chromosome regions for their chromosomal sex. A copy number variant, or CNV, is a segment present in fewer or more copies than expected. A one-copy loss is commonly called a deletion. A three-copy gain is commonly called a duplication, although larger gains can contain four or more copies.

Chromosomal microarray analysis compares DNA from the tested person with a reference pattern at hundreds of thousands or millions of locations. The laboratory measures the signal from probes placed across the genome. A lower-than-expected signal across neighboring probes suggests a deletion; a higher signal suggests a duplication. The software defines the boundaries, estimates size, maps the interval to a genome build, and identifies genes and regulatory regions within it.

A CMA report usually describes a finding with chromosome coordinates, such as arr[GRCh38] 16p11.2(approximately 29.6–30.2 Mb)x1. The notation identifies the genome reference version, chromosome band, approximate interval, and observed copy number. Exact coordinates matter because two deletions with the same chromosome-band label may include different genes and have different consequences.

Unlike chromosome analysis by karyotype, CMA does not rely on visually examining stained chromosomes under a microscope. It can therefore detect much smaller imbalances, often in the tens to hundreds of kilobases depending on probe density and laboratory reporting policy. It also does not need growing cells in culture for most postnatal blood samples, which can shorten laboratory processing.

The test is genome-wide, but it is not truly uniform. Probe density is higher in clinically important regions and lower in repetitive or technically difficult areas. Laboratories also apply reporting thresholds so that tiny common variants are not listed unnecessarily. The reportable resolution therefore differs from the smallest theoretical change a platform can measure.

When Chromosomal Microarray Is Used

CMA is widely used when a person has features suggesting a chromosome-level imbalance but no single syndrome is certain. Common postnatal reasons include:

Global developmental delay or intellectual disability.
Autism with additional developmental, neurologic, growth, or physical findings.
Multiple congenital anomalies.
Unexplained short stature, overgrowth, microcephaly, or macrocephaly.
Seizures plus developmental differences or birth defects.
Dysmorphic features that do not fit one recognized condition.
A suspected microdeletion or microduplication syndrome.
An abnormal karyotype requiring more precise definition of gained or lost material.

Diagnostic yield varies with the population and how strongly the phenotype suggests a genomic disorder. In broad neurodevelopmental cohorts, clinically significant CNVs are often found in roughly 10% to 20%, with higher yields in people who have multiple congenital anomalies or a more syndromic presentation. A precise number cannot be promised for one person because referral criteria, platform design, and classification policies differ.

CMA may also be used in stillbirth evaluation, products of conception, and prenatal diagnosis. In prenatal care, it can identify clinically important submicroscopic imbalances in addition to whole-chromosome conditions. The decision to use it depends on ultrasound findings, prior screening, family history, and whether an invasive specimen is being obtained.

For isolated features strongly associated with one gene or one repeat expansion, another test may be more efficient. A child with classic fragile X features, for example, needs a dedicated FMR1 repeat test because a standard microarray does not measure the CGG expansion reliably. A person with a recognizable single-gene syndrome may need sequencing or a multigene panel rather than CMA alone.

Microarray testing can still be useful after another test. Exome sequencing may find a sequence variant but miss or incompletely assess some CNVs. Conversely, a normal CMA may be followed by exome or genome sequencing when the clinical suspicion remains strong. The methods answer overlapping but different questions.

Array Types, Resolution, and Laboratory Design

Two main signal strategies are used, sometimes on the same platform.

Array comparative genomic hybridization

Array comparative genomic hybridization, or aCGH, compares differently labeled patient and reference DNA. Relative signal intensity across probes reveals gains and losses. It is strong for copy-number detection but does not independently show which parental chromosome contributed each allele.

Single-nucleotide polymorphism arrays

SNP arrays measure both total signal and the pattern of alleles at common single-nucleotide polymorphisms. This allows copy-number analysis plus detection of long stretches with no heterozygosity. Those stretches may suggest uniparental disomy, parental relatedness, or a recessive-disease region, depending on their size and distribution.

A SNP array can identify some copy-neutral abnormalities that aCGH alone cannot. For example, if a chromosome region has two copies but both came from one parent, the total copy number looks normal. The allele pattern may still show a long region of homozygosity. Further testing is required because a SNP array generally cannot prove the parent of origin by itself.

Test	Best at detecting	Typical scale	Important blind spot
Karyotype	Whole-chromosome changes and large balanced or unbalanced rearrangements	Usually several megabases	Most small deletions and duplications
Chromosomal microarray	Genome-wide deletions and duplications; SNP arrays also show homozygosity patterns	Often tens to hundreds of kilobases, depending on region	Most balanced rearrangements and sequence variants
FISH	A predefined chromosome region in individual cells	Target dependent	Unexpected changes elsewhere in the genome
Exome sequencing	Sequence variants in protein-coding genes; some CNVs	Single bases to selected exon-level CNVs	Many noncoding regions and some structural changes

Resolution should be interpreted as a range, not a guarantee. A laboratory may place dense probes over known disease genes and sparse probes in regions with no established clinical significance. Centromeres, telomeres, segmental duplications, pseudogenes, and repetitive sequences can be difficult. A deletion that removes one well-covered exon may be detected, while a similarly sized change in a poorly covered region may not be.

Clinical laboratories validate their own thresholds. They may report smaller changes in a recognized dosage-sensitive gene than in a gene-poor region. This is why two laboratories using similar platforms can issue different reports from the same raw biological finding.

Copy Number Results and Classification

CMA findings are classified using evidence about dosage sensitivity, known syndromes, gene function, population frequency, inheritance, published cases, and overlap with curated databases. Constitutional CNVs are commonly placed into five categories.

Pathogenic or likely pathogenic

A pathogenic CNV has strong evidence that the gain or loss causes disease. Examples include a recurrent 22q11.2 deletion associated with 22q11.2 deletion syndrome or a deletion involving a well-established haploinsufficient gene. “Likely pathogenic” means the evidence strongly favors disease relevance but does not reach the laboratory’s highest confidence threshold.

The result may establish or support a diagnosis, but it does not predict every feature with certainty. Many CNV syndromes have variable expression. One person may have major congenital anomalies and another with the same recurrent CNV may have mild learning differences. Penetrance can also be incomplete, meaning some carriers show few or no obvious features.

Variant of uncertain significance

A variant of uncertain significance, or VUS, has insufficient or conflicting evidence. The interval may contain genes not yet linked clearly to disease, overlap incompletely with known cases, or have limited population data. A VUS is not the same as a positive diagnosis. It should not be used alone to make major medical or reproductive decisions.

Parental testing can shift interpretation. A de novo CNV that fits the child’s features may gain pathogenic evidence. Inheritance from an apparently unaffected parent may support reduced penetrance, variable expression, or a benign interpretation, but it does not automatically prove harmlessness. The parent may have subtle features or the CNV may act with other genetic or environmental factors.

Likely benign or benign

Benign CNVs are common in healthy populations or lack genes sensitive to dosage. Laboratories often do not list every benign variant. A “normal” report therefore usually means no reportable pathogenic, likely pathogenic, or uncertain CNV was found, not that every person has exactly the same copy number across the genome.

The classification applies to current evidence. Laboratories may update a VUS after new cases, gene-dosage data, or population studies become available. Families should ask how reclassification is communicated and whether periodic reanalysis is offered. The broader principles in VUS interpretation are relevant, although CNV evidence includes interval-level dosage information in addition to sequence-variant evidence.

Inheritance, Mosaicism, and Family Testing

A CNV can be inherited or arise for the first time in an egg, sperm, or early embryo. A de novo result usually means neither tested parent carries the same change in the sampled tissue. It often increases confidence that the CNV explains a child’s condition, especially when the interval is gene-rich and the phenotype matches known cases.

An inherited CNV may still be clinically important. Autosomal dominant CNVs can show variable expression or reduced penetrance. Recessive conditions can occur when a deletion removes one copy of a gene and a sequence variant affects the other copy. An X-chromosome CNV may affect people differently according to chromosomal sex and X-inactivation. Parent-of-origin can matter for imprinted regions such as 15q11-q13.

Parental studies may use microarray, targeted deletion/duplication testing, quantitative PCR, MLPA, or FISH. The laboratory chooses a method that can reliably detect the exact familial change. Testing both biological parents is often more informative than testing only one.

Mosaic copy-number changes

Mosaicism means two or more cell populations have different genetic findings. A mosaic deletion may be present in 30% of blood cells, for example, while the remaining cells have a normal copy number. CMA can detect some mosaic changes, but sensitivity depends on the size of the CNV, array design, laboratory algorithms, and proportion of abnormal cells. Low-level mosaicism may be missed.

Tissue choice matters. A change confined mainly to skin, buccal cells, placenta, tumor, or another tissue may not be detectable in blood. When the physical findings follow a patchy or segmental pattern, the clinical team may recommend testing an affected tissue. For prenatal samples, confined placental mosaicism can create a discrepancy between chorionic villus and amniotic fluid results.

Long regions of homozygosity on a SNP array can also lead to additional testing. A whole-chromosome pattern may raise concern for uniparental disomy. Multiple regions across several chromosomes may suggest that the parents share ancestry. The amount and pattern can help guide recessive-gene analysis, but a microarray does not identify the causative sequence variant within those regions.

What a Microarray Cannot Reliably Detect

A normal CMA leaves many categories of genetic variation untested or incompletely tested.

Balanced translocations and inversions: DNA is rearranged without a measurable net gain or loss. A karyotype or genome-based structural-variant method may be needed.
Most single-nucleotide variants and small insertions or deletions: These usually require sequencing.
Repeat expansions: Conditions such as fragile X syndrome, Huntington disease, myotonic dystrophy, and many spinocerebellar ataxias need dedicated repeat testing.
Many methylation and imprinting defects: Some SNP patterns may suggest a problem, but methylation-specific testing is often required.
Low-level mosaicism: The abnormal cell fraction may fall below the laboratory’s validated detection limit.
Small exon-level CNVs: Detection varies by probe coverage. A gene-specific deletion/duplication assay may be more sensitive.
Mitochondrial DNA variants: Standard CMA focuses on nuclear chromosomes.
Sequence-level recessive partners: A deletion of one gene copy does not reveal a small variant in the remaining copy.

CMA also cannot determine whether every gene in a duplicated segment is actively expressed or predict an exact developmental outcome. It measures dosage, not gene activity. A report may identify a large region with dozens of genes while only one or a few are responsible for the main phenotype.

Balanced rearrangements deserve special attention in reproductive care. A person carrying a balanced translocation can be healthy but have infertility, recurrent miscarriage, or pregnancies with unbalanced chromosome content. A microarray on that person may be normal because no DNA is gained or lost. A karyotype for infertility is often the more appropriate first method when a balanced rearrangement is suspected.

No test can exclude all genetic causes. The clinical team should select follow-up methods based on the phenotype, pedigree, and exact gaps in the completed assay rather than ordering every available test without a plan.

Chromosomal Microarray During Pregnancy

Prenatal CMA is performed on fetal or placental cells obtained through chorionic villus sampling, usually in the first trimester, or amniocentesis, generally from the second trimester. It is different from cell-free DNA screening, which estimates risk from placental DNA fragments in maternal plasma. CMA on an invasive specimen is a diagnostic analysis for the copy-number changes it can detect.

It is commonly considered when ultrasound shows one or more fetal structural anomalies, when an invasive test is already being performed after an abnormal screening result, or when a prior pregnancy or family history suggests a chromosome imbalance. Some patients choose CMA even with a structurally normal ultrasound because it can detect submicroscopic pathogenic CNVs that a karyotype misses. The chance of an uncertain finding and the scope of possible incidental information should be discussed before testing.

Prenatal interpretation can be difficult because a fetus has limited observable features. A CNV associated with variable neurodevelopmental outcomes may not produce a prenatal ultrasound sign. Reports may describe a range from mild learning difficulties to major disability without being able to predict where one child will fall.

Sample-specific issues include maternal cell contamination, confined placental mosaicism, culture artifacts, and limited DNA. A chorionic villus result may require confirmation in amniotic fluid when mosaicism is suspected. If the CMA detects a large terminal imbalance, a parental karyotype may be needed to determine whether one parent carries a balanced rearrangement that affects recurrence risk.

A normal prenatal CMA does not rule out a single-gene disorder. When major fetal anomalies remain unexplained, prenatal exome sequencing may be discussed after appropriate counseling. More detail about the procedure-specific context is available in prenatal chromosomal microarray testing.

Next Steps After a CMA Result

The next step depends on whether the result is diagnostic, uncertain, or negative.

For a pathogenic or likely pathogenic finding, the clinical team may compare the person’s features with the known syndrome, arrange organ-specific screening, refer to specialists, and offer family testing. A molecular diagnosis can end an unnecessary diagnostic search, but it may also reveal surveillance needs not previously suspected. Recurrence counseling should consider whether the CNV is inherited, de novo, caused by a parental rearrangement, or associated with possible germline mosaicism.

For a VUS, useful actions include detailed phenotyping, testing parents when recommended, reviewing the interval’s genes, and asking about future reanalysis. Avoid labeling every symptom as caused by the VUS. Medical care should continue to address the patient’s actual clinical findings.

For a normal result, ask what was not assessed. Depending on the reason for testing, next options may include sequencing, a repeat-expansion assay, methylation analysis, karyotyping, or testing another tissue. A negative result can also be genuinely reassuring when the suspected condition is one that CMA detects well and the sample quality was adequate.

Questions to bring to a result visit include:

What is the exact chromosome interval, size, genome build, and copy number?
Which genes with established dosage sensitivity are included?
Does the result explain all, some, or none of the observed features?
Was the CNV inherited, de novo, or not yet tested in parents?
Is penetrance complete, and how variable are reported outcomes?
Does the result create health screening needs for the patient or relatives?
Could a balanced rearrangement explain the finding or recurrence risk?
When should the laboratory or clinic revisit an uncertain result?

Keep the original report because coordinate systems, classification, and test versions matter. A summary such as “microarray positive” is not enough for future reproductive or medical decisions. Genetic counseling can translate the laboratory finding into individual risk while acknowledging uncertainty and variation among people with similar CNVs.

References

Disclaimer

Chromosomal microarray findings must be interpreted with the tested person’s features, family history, specimen type, and laboratory methodology. This article is educational and does not replace genetic counseling, prenatal diagnosis, or individualized medical advice. Do not use a VUS alone to make treatment or pregnancy decisions.