What is aCGH? A Look at This Genetic Test

Array comparative genomic hybridization, or aCGH, is a laboratory technique that scans a person’s entire genome for missing or extra stretches of DNA. It works by comparing a patient’s DNA against a reference sample to spot gains and losses of genetic material, sometimes catching changes too small for older chromosome tests to see. The test has become a workhorse in diagnosing unexplained developmental delays, birth defects, and certain cancers, though it comes with some interpretive challenges that are worth understanding.

How aCGH Works in Plain Terms

The basic idea behind aCGH is a competition between two DNA samples. A patient’s DNA is labeled with one fluorescent color (typically green), and a reference sample from a chromosomally normal individual is labeled with another color (typically red). Both samples are then placed together onto a glass slide that has thousands of tiny DNA fragments fixed to its surface in a grid pattern. Each spot on the grid corresponds to a known location in the human genome.

The patient’s DNA and the reference DNA compete to bind to those spots. If the patient has the normal two copies of a particular stretch of DNA, the green and red signals balance out. If the patient has an extra copy, the green signal is stronger at those spots. If the patient is missing a copy, red dominates instead. A computer reads the color ratios across the entire grid and produces a map showing where along every chromosome the patient has too much or too little DNA.¹ These gains and losses are called copy number variants, or CNVs, and they range from tiny segments covering a single gene to large chunks spanning millions of DNA base pairs.

Why aCGH Was a Big Step Forward

Before aCGH, the standard approach for looking at someone’s chromosomes was a technique called karyotyping. A lab technician would grow cells, stain the chromosomes, photograph them under a microscope, and visually inspect the banding patterns for anything out of place. Karyotyping is great for catching large-scale problems like an extra chromosome 21 (which causes Down syndrome), but its resolution has a hard floor. It can only reliably detect changes that are roughly 5 to 10 million base pairs or larger.²

Many clinically important genetic changes are smaller than that. aCGH pushed the resolution down dramatically, detecting deletions and duplications in the range of tens of thousands of base pairs. That improvement opened a window into a whole class of conditions that had been invisible to older methods. One study comparing the two approaches in children with developmental delays found that aCGH achieved a diagnosis rate about two and a half times higher than conventional karyotyping.³

Diagnosing Developmental Delay and Intellectual Disability

One of the most common reasons clinicians order aCGH is for children who have unexplained intellectual disability, developmental delay, or certain birth defects like congenital heart defects or unusual facial features. In many of these cases, standard genetic tests come back normal, yet the child clearly has a condition with a genetic basis. aCGH fills that gap by revealing submicroscopic deletions or duplications that disrupt important genes.

In a large Czech cohort of 542 children with intellectual disability or developmental delay, aCGH identified clearly harmful CNVs in about 18% of patients.⁴ That might sound modest, but for families who have been searching for answers, a definitive genetic diagnosis can transform their situation. It may explain the child’s condition, give clinicians a basis for monitoring associated medical risks, guide therapy choices, and help parents understand the likelihood of the same condition appearing in future children.

The screening of large patient groups by aCGH has also led to the discovery of entirely new microdeletion and microduplication syndromes. These conditions were first recognized because multiple unrelated patients turned up with overlapping small chromosomal changes and similar clinical features, a pattern that only became visible once the resolution of the testing technology improved enough to find the changes in the first place.⁵

Prenatal Testing

aCGH has also become an important tool during pregnancy, particularly when an ultrasound reveals a structural abnormality in the fetus. In those situations, conventional karyotyping will catch many major chromosomal problems, but aCGH picks up additional clinically meaningful findings that karyotyping misses. A study of over 3,000 pregnancies found that aCGH added roughly 8% to the diagnostic yield compared with karyotyping alone for fetuses with abnormal ultrasound results.⁶ The test proved especially useful when ultrasound found a balanced translocation or a small marker chromosome that karyotyping could see but not fully characterize.

A separate prospective study of over 1,000 pregnancies with structural anomalies found that the highest rates of meaningful CNV detection were in fetuses with heart defects or problems involving multiple organ systems. Among those findings, a particular microdeletion on chromosome 22 (22q11.2, associated with DiGeorge syndrome) accounted for a striking 40% of the harmful CNVs detected in cardiac anomaly cases.⁷ That kind of specific diagnostic hit directly changes how the pregnancy and the child’s early care are managed.

Cancer Genomics

Beyond inherited conditions, aCGH has been applied in cancer research and diagnostics to map the chromosomal gains and losses that drive tumor growth. Cancer cells are notorious for having wildly rearranged genomes, and knowing which genes have been amplified or deleted can help classify tumor types, predict how aggressive a cancer might be, and even inform treatment decisions.

In colorectal cancer, for example, aCGH was used to distinguish between two different types of genomic instability. The technique identified specific gene amplifications on chromosomes 20, 13, and 7, and deletions on chromosome 17, that were preferentially seen in one subtype over the other.⁸ That level of detail helps researchers understand what is driving individual tumors and may eventually help tailor therapy.

What aCGH Cannot See

For all its power, aCGH has real blind spots. The test is designed to find imbalances in DNA quantity (gains and losses), but it cannot detect changes that do not alter the overall amount of genetic material. That means:

Balanced rearrangements: If two chromosomes swap segments without gaining or losing any DNA, the green-to-red ratio stays even and the swap goes undetected.
Point mutations: Single-letter changes in the DNA code, which cause many well-known genetic conditions like cystic fibrosis or sickle cell disease, are invisible to aCGH.
Low-level mosaicism: If only a small fraction of a person’s cells carry a chromosomal change, the signal may be too faint for aCGH to reliably pick up.

These limitations mean aCGH is not a universal genetic test. It answers one specific question very well: does this person have too much or too little DNA at any point in their genome? If the answer to a patient’s condition lies in a different type of mutation, aCGH will come back normal even though a real genetic problem exists.³

The Problem of Uncertain Results

Perhaps the most challenging aspect of aCGH is not the test itself but interpreting what it finds. Because the test scans the entire genome at high resolution, it inevitably turns up variants that have never been seen before or whose clinical significance is genuinely unclear. These get classified as variants of uncertain significance, or VUS, and they can be a source of real anxiety for patients and families.

The classification system used in most labs follows guidelines from the American College of Medical Genetics and Genomics (ACMG) and the Clinical Genome Resource (ClinGen). Each CNV is scored based on multiple lines of evidence: its size, whether it overlaps with known disease genes, whether it has been reported in other patients with similar symptoms, and how common it is in the general population. The scoring system assigns point values ranging from clearly pathogenic (harmful) through likely pathogenic, uncertain significance, likely benign, down to benign.⁹

In practice, clinical databases like ClinVar, DECIPHER, and OMIM serve as repositories where labs can check whether a particular CNV has been seen before and what it was associated with.¹⁰ As more patients are tested and more data accumulates, variants that were once uncertain can get reclassified. One study tracking VUS reclassification over eight years found that periodic reinterpretation did change classifications for some variants as new knowledge emerged.¹¹ This is encouraging but also means that a “no clear answer” result today might have a different interpretation in a few years.

Parental Testing and Incomplete Penetrance

When a CNV is found in a child, one of the first steps is often to test the parents. If neither parent carries the variant, it arose brand new (de novo), and de novo changes are generally treated with more suspicion. If a parent carries the same variant but is perfectly healthy, the picture becomes murkier. The variant might be benign, or it might show what geneticists call incomplete penetrance, meaning it causes problems in some carriers but not others.

Testing the parents, sometimes called trio analysis, also helps detect a condition called uniparental disomy, where a child has inherited both copies of a chromosome or chromosome segment from one parent rather than one from each. Standard aCGH alone can miss this, because the total amount of DNA is normal. SNP-based arrays (discussed below) are better at catching it.¹² In one large study of children with intellectual disability, trio analysis allowed confident separation of benign from pathogenic variants in all but about 2% of cases.¹³

SNP Arrays and How They Differ

aCGH is sometimes discussed alongside a related technology called SNP arrays. Both are types of chromosomal microarray analysis (CMA), and both detect copy number changes, but they work on slightly different principles and have distinct strengths.

Where aCGH relies on directly comparing a patient’s DNA to a reference sample on the same slide, SNP arrays take advantage of naturally occurring single-letter variations (single nucleotide polymorphisms) scattered throughout the genome. By reading which version of each SNP a patient carries, the array can infer both copy number changes and something aCGH cannot: loss of heterozygosity. Loss of heterozygosity occurs when both copies of a gene region are identical (both from the same parent, for instance), and it can be clinically relevant in cancer and in certain inherited conditions like those involving uniparental disomy.¹⁴

In current clinical practice, many labs use platforms that combine elements of both approaches or have shifted to SNP-based arrays entirely. The choice between the two often depends on what clinical question is being asked. For prenatal testing focused purely on copy number, aCGH and SNP arrays perform similarly. When there is a concern about uniparental disomy or consanguinity, SNP arrays have a clear edge.

Ethical Dimensions in Prenatal aCGH

The increased resolution of aCGH creates a genuine ethical tension in prenatal medicine. Finding a clearly pathogenic CNV during pregnancy gives families crucial information to prepare for a child’s medical needs or, in some jurisdictions, to make decisions about continuing the pregnancy. But finding a VUS can be deeply distressing. The parents are told their baby has a genetic change whose meaning nobody fully understands, and there is no way to resolve the uncertainty before birth.

Genetic counselors working in this space have grappled with several thorny questions: whether it is ever justifiable to withhold certain ambiguous results from a pregnant patient, what truly informed consent looks like when the test can generate results that not even experts can fully interpret, and whether clinicians bear responsibility if an uncertain finding leads to a pregnancy termination that, in hindsight, was medically unnecessary.¹⁵ There are no universally agreed-upon answers, but the consensus in the field has moved toward providing thorough pre-test counseling so families understand in advance that uncertain results are a real possibility, not a hypothetical one.

Cost Considerations

An early concern about aCGH was that it would be substantially more expensive than karyotyping and therefore hard to justify as a routine first-line test, especially in publicly funded healthcare systems. Economic evaluations, however, have generally found that using aCGH as a first-line test for conditions like learning disability is actually cost-saving compared with using it only as a second-line test after karyotyping fails. This is because the two-step approach means paying for two tests in every patient who ends up needing aCGH anyway, while going straight to aCGH eliminates the redundant first step. A UK-based cost-effectiveness study found that the first-line aCGH strategy was both less costly and as effective as the two-step approach.¹⁶

Quality Control Matters

Like any laboratory test, aCGH results are only as good as the sample and the analysis behind them. DNA quality is a major factor: degraded or contaminated samples can produce noisy data with false signals that look like real CNVs. Custom array designs, which target specific genomic regions of interest rather than scanning the whole genome, can be especially sensitive to quality issues because there are fewer probes to cross-check against.

Expert review of the data remains essential. Automated software flags candidate CNVs, but experienced lab scientists still evaluate each one, checking for artifacts and confirming that the signal is real before including it in a clinical report.¹⁷ Variants flagged as potentially pathogenic are typically confirmed by a second, independent technique before being reported to the clinician.

Newer Technologies on the Horizon

While aCGH and SNP arrays remain widely used, newer approaches are beginning to challenge their dominance. Low-pass genome sequencing, which reads a person’s entire genome at a shallow depth, can detect everything chromosomal microarray can detect and then some. In a head-to-head comparison of over 1,000 prenatal cases, low-pass sequencing identified all the pathogenic CNVs that the microarray found, plus 17 additional clinically relevant CNVs that the microarray missed. It also required less DNA input and had a lower rate of technical failures.¹⁸

A validation study of low-pass sequencing across over 400 clinical cases estimated its clinical sensitivity at above 21%, compared with roughly 15 to 18% for microarray, with a resolution limit below 25 kilobases.¹⁹ The technology is not yet standard everywhere, partly because the analysis pipelines are newer and clinical labs need time to validate and accredit them. But the trend is clearly moving in the sequencing direction, and aCGH may eventually be remembered as the transitional technology that bridged the gap between microscope-based chromosome analysis and whole-genome sequencing.

Applications Beyond Human Medicine

aCGH is not limited to human diagnostics. The same technology has been adapted for use in veterinary genetics, agriculture, and evolutionary biology. In livestock breeding, for instance, researchers have used aCGH to map copy number variation across pig breeds, identifying over 250 variable regions spread across the genome. Many of these overlapped with regions linked to traits of agricultural interest such as growth rate, disease resistance, and meat quality.²⁰ Similar studies have been conducted in cattle, dogs, and crop plants. The underlying logic is the same: compare DNA from one sample against a reference, read the color ratios, and map what has been gained or lost. The questions being asked are just different, focusing on breed differences and trait selection rather than disease diagnosis.