Clinical Utility of Chromosomal Microarray Testing
History of Cytogenetic Testing For Congenital Disorders
Identification of the underlying genetic etiology for congenital abnormalities and developmental disabilities provides useful information for both clinicians and families. Not only can the specific genetic etiology identify potential medical interventions that have important health implications for the patient, it also provides information that allows accurate recurrence-risk counseling and helps families plan for the expected natural history of the disease. Conventional G-banded chromosome analysis, the visual examination of recognized banding patterns on Giemsa-stained metaphase chromosomes (Figure 1), has been the workhorse of clinical cytogenetics for nearly 40 years. The major advantage of this technique is that all 23 pairs of chromosomes are examined, allowing for the analysis of the entire human genome in a single assay. Traditional chromosome analysis detects a pathogenic chromosome rearrangement in approximately 4% of patients and has led to the characterization of many classic syndromes associated with chromosomal rearrangements, such as Down, Cri-du-chat, and Smith-Magenis syndromes. While chromosome studies can easily detect gains or losses of whole chromosomes (aneuploidy) and large structural rearrangements (deletions, duplications, inversions, translocations), this technique is somewhat subjective and gains or losses of less than approximately 5 to 10 megabases (5 million to 10 million base pairs) cannot be consistently detected. To address this, additional techniques to detect smaller chromosome abnormalities are needed.
Figure 1. Conventional G-banded chromosome analysis
Fluorescence in situ hybridization (FISH) technology entered clinical cytogenetics in the 1990s and has complemented chromosome studies due to its relatively high resolution. FISH testing uses fluorescently labeled probes that hybridize to a particular position on a chromosome and can be used to detect chromosomal rearrangements as small as 100 kilobases (100,000 base pairs) in routine clinical practice (Figure 2). Due to its comparatively high resolution, this technology has been used as the definitive diagnostic test for microdeletion/duplication disorders that are not visible by routine chromosome analysis, such as Williams syndrome and 22q11.2 deletion (DiGeorge/velocardiofacial) syndrome. In addition, submicroscopic chromosomal rearrangements involving the subtelomeric regions (the terminal ends of each chromosome) were discovered to be a significant cause of congenital abnormalities and developmental disabilities. For patients with congenital abnormalities, developmental disabilities, and normal chromosome studies, FISH analysis of the subtelomeric regions of each chromosome detected pathogenic changes in approximately 2.5% of patients.1 Despite advances that FISH testing brings to the field of cytogenetics, its utility is limited by the need to know which area of the genome to interrogate prior to testing. In the traditional “phenotype-first” setting, observed clinical features dictate the FISH testing to be performed. For patients with clinical features that obviously fit with a particular syndrome, FISH testing often confirms the suspected diagnosis. However, many patients do not clearly fit into one of the classically defined syndromes. Thus, for patients with nonspecific phenotypic features that lead to clinical suspicion of a chromosome rearrangement, a “genotype-first” approach, one that combines analysis of the entire human genome (like chromosome studies) at a very high resolution (like FISH studies), is desirable.
Figure 2. In this fluorescence in situ hybridization image, 2 green control probes are present, indicating chromosome 22. The missing red signal demonstrates the 22q11.2 microdeletion on 1 homologue. Green = control probe; Red = 22q11.2 probe
Chromosomal Microarray Testing
Recent technological advances that allow high-density printing of oligonucleotide probes on glass slides have revolutionized cytogenetic testing. For chromosomal microarray testing, highly specific oligonucleotide probes are designed and distributed throughout the genome. This allows for a whole genome survey in a single assay with very high resolution analysis that is only limited by the probe density that can be achieved on the array. Chromosomal microarray technology has quickly moved from research to the clinical setting and has emerged as the American College of Medical Genetics-recommended first-tier test for certain patients (Table).2
|ACMG Clinical Recommendations|
Microarray testing is recommended as a first-tier, postnatal test for individuals with:|
• Multiple anomalies not specific to a well-delineated genetic syndrome
|Further studies are needed to determine the utility of microarray testing in individuals with growth retardation, speech delay, and other indications|
|Appropriate follow-up when a chromosomal imbalance is detected includes:
• Confirmation of results by a second method
Table. American College of Medical Genetics (ACMG) clinical recommendations for microarray testing
The clinical implementation of chromosomal microarray testing has allowed the description of many new clinical syndromes caused by the deletion and duplication of chromosomal segments too small to observe by traditional cytogenetic techniques. This testing has been shown to detect a pathogenic rearrangement in 15% to 20% of patients, increasing the diagnostic yield approximately 4- to 5-fold over chromosome studies alone.3 These important advantages are allowing chromosomal microarray testing to virtually replace conventional chromosome analysis in the cytogenetic workup of this patient population.
Figure 3. Black vertical lines represent the oligonucleotide probes: 1 probe is placed approximately every 25 kb for the backbone coverage. Other clinically relevant regions, including telomeres, centromeres, microdeletions and microduplications, and selected clinically relevant genes contain enhanced coverage. The blue bars represent the unique telomere(pter and qter) regions and the red bars represent the unique pericentromeric(cen) regions that are targeted on the array.
Chromosomal Microarray Methodology
Chromosomal microarray technology has evolved quickly with platforms that contain ever-increasing numbers of oligonucleotide probes. This technical evolution, combined with improvements in the strategic placement of probes, has led to chromosomal microarray designs optimized for clinical testing.4 The Mayo Clinic Cytogenetics Laboratory has recently updated chromosomal microarray testing to utilize a platform containing 180,000 (180K) unique oligonucleotide probes. The testing is offered by Mayo Medical Laboratories as #88898 Array Comparative Genomic Hybridization (aCGH), Whole Genome, Constitutional. Probes are spaced evenly throughout the genome at approximately 25 kilobase intervals to provide backbone, genome-wide coverage. In addition, approximately 500 regions throughout the genome contain higher-density, targeted oligonucleotide probe coverage (Figure 3). These regions include the pericentromeric and subtelomeric regions of each chromosome, chromosomal regions known to be associated with specific clinical phenotypes or syndromes, and specific genes implicated in developmental abnormalities. Analysis of the relative ratio of patient DNA to control DNA allows for the detection of gains and losses (copy number changes) at each probe locus (Figure 4). The results of each probe are plotted with respect to their chromosomal location, and deviations of a minimum of 5 contiguous probes are flagged as high-confidence copy number changes. This results in a minimum functional resolution of approximately 100 kilobases throughout the genome and a minimum functional resolution of approximately 5 kilobases for targeted regions.
Figure 4. Oligonucleotide probes that are complementary to specific regions of the genome are synthesized at unique positions on the array slide. Genomic DNA from the patient and a control are labeled independently with different fluorescent dyes, mixed together, and cohybridized to the microarray. Following competitive hybridization, the slides are scanned and the intensities of the 2 fluorophores are measured. The intensity data for each probe on the array are normalized and the relative ratio of the patient to the control DNA is plotted on a log2 scale for each chromosome. The log2 ratio plots are analyzed to determine the relative gain (red plots) or loss (green plots) of patient DNA relative to the control DNA at the genomic location of each probe.
Indications for Chromosomal Microarray Testing
The main benefits of chromosomal microarray testing are the capacity to detect submicroscopic regions of chromosomal imbalance throughout the genome at very high resolution and the delineation of precise breakpoints that enable the identification of specific gene content within the regions of imbalance.
- Chromosomal microarray testing is the preferred first-tier test for individuals with multiple anomalies not specific to a well-delineated genetic syndrome, unexplained developmental delay/intellectual disability, and autism spectrum disorders.2
- Chromosomal microarray testing is appropriate follow-up for individuals with unexplained developmental delay/intellectual disability, autism spectrum disorders, or multiple congenital anomalies with a previously normal conventional chromosome study due to the superior resolution of the chromosomal microarray testing. Chromosomal microarray testing in this situation would be expected to identify a clinically significant abnormality in approximately 10% to 15% of cases.
- Patients with an abnormality previously identified by conventional chromosome studies may also benefit from chromosomal microarray testing. In this situation, chromosomal microarray testing can determine the precise breakpoints of copy number changes, unveil unappreciated complexity that was not observed at the resolution of the chromosome study, and catalogue and investigate the gene content in the region of imbalance. This additional information can lead to a better understanding of the underlying mechanism that led to the rearrangement which can, in turn, influence recurrence-risk counseling. In addition, knowledge of the specific gene content may provide clues to a patient’s phenotype and lead to the diagnosis of an autosomal dominant single-gene disorder, which has the potential to greatly influence clinical care. Based on our experience with approximately 10,000 chromosomal microarray tests performed to date, many examples of deletions including tumor suppressor genes, such as APC or PTEN, or genes implicated in cardiac disorders, such as MYH7 (associated with hypertrophic cardiomyopathy) or KCNH2 (associated with long QT syndrome) have been detected. Clearly, direct knowledge that these genes are deleted in patients is important for clinical care and may indicate a need to follow cancer screening protocols at the appropriate age or to receive appropriate treatment to manage a cardiac disorder. While these findings may be incidental to the primary indication for testing, they can prove to be of vital clinical importance.
- Patients with abnormal phenotypes and apparently balanced rearrangements (inversions or translocations) identified by previous conventional chromosome studies may benefit from chromosomal microarray testing. A large proportion of such rearrangements appear balanced at the resolution of a chromosome study, but may actually be unbalanced when analyzed by higher-resolution chromosomal microarray testing, and such gains or losses may help explain the patient’s clinical phenotype.5,6 As reviewed above, the gene content within these regions of copy number change should be evaluated and may have significant implications for the diagnosis or treatment of the patient.
Limitations of Chromosomal Microarray Testing
Like all genetic tests, chromosomal microarray testing has certain limitations that are important to understand. Chromosomal microarray testing is designed to detect copy number changes involving unique chromosomal material. Therefore, truly balanced chromosomal rearrangements, such as inversions or translocations, are not detected by this assay. In addition, chromosomal microarray testing alone does not provide information about the structural nature of an imbalance. Such structural information is critical for the identification of the underlying rearrangement mechanism and subsequent recurrence-risk counseling. For example, if a patient has a duplicated region of chromosome 18, one cannot tell from the chromosomal microarray data whether this is a tandem duplication residing on chromosome 18 or if the additional material is present on another chromosome due to an insertional translocation (Figure 5). This distinction is critical since a parent carrying a balanced form of this insertional translocation could have a 50% chance of having another child with an unbalanced chromosome complement. For this reason, current American College of Medical Genetics guidelines for chromosomal microarray testing recommend that chromosomal microarray test results be confirmed by a second method.7 For large deletions and duplications that are cytogenetically visible, conventional chromosome analysis is sufficient for confirmation. FISH testing is often employed for smaller abnormalities and can then be used as the preferred method for testing other family members if necessary. For very small abnormalities, particularly small duplications, FISH testing may not be feasible. In these situations, chromosomal microarray testing is required for any additional family studies.
Figure 5. Duplicated region of chromosome 18, which is present on chromosome 2p due to an insertional translocation Green = 18 centromere, Red = 18q23
Interpretation of Chromosomal Microarray Results
The ability to interrogate the human genome at high resolution has led to the discovery of widespread copy number variation, including both pathogenic copy number changes found in patients with abnormal phenotypes and benign variations found in all individuals. The true extent of the diversity of copy number variation within the general population is yet to be fully appreciated. Therefore, the differentiation between pathogenic and benign copy number variation can be challenging and requires an up-to-date knowledge of the genetic literature. The continual discovery of novel copy number variation and new reports classifying specific copy number variations as pathogenic or benign enter the literature frequently. Therefore, the interpretation of any given copy number variation may evolve rapidly as more information becomes available. While most copy number variations observed through chromosomal microarray testing can readily be characterized as pathogenic or benign, a subset have limited data available to support classification into either of these categories. In these situations, a number of considerations must be taken into account to help interpret the data with regard to the phenotypic presentation in the individual patient.
Many factors influence the clinical interpretation of copy number variations including the nature of the imbalance (ie, deletion versus duplication), the size and gene content of the interval, and comparison to both public and internal databases that catalogue copy number variation. In general, deletions tend to be less well tolerated than duplications and are more likely to be pathogenic. This phenomenon is exemplified by several reciprocal deletion/duplication syndromes such as the deletion associated with Smith-Magenis syndrome and the reciprocal duplication associated with Potocki-Lupski syndrome, both involving the same region of chromosome 17, where the phenotype associated with deletion is more severe than the phenotype associated with duplication. Evaluation of the specific gene content of a given copy number variation requires the laboratory and the clinician to be cognizant of what information is available for any given gene, as well as the types of mutations that lead to abnormal phenotypes. The molecular mechanisms leading to disease differ greatly among disorders and may include loss-of-function mutations, gain-of-function mutations, dominant negative mutations (a mutation leading to an abnormal protein that disrupts the function of the normal protein product as well), and dosage-sensitivity of the gene. For example, a deletion including a gene that causes an autosomal dominant disease due to haploinsufficiency (when half of the normal gene product or activity is not sufficient to produce a normal phenotype) or loss-of-function mutations would be expected to result in typical phenotypic presentation of the disease. However, duplication of that same gene would not be predicted to have the same effect. Furthermore, the mode of inheritance must be considered. A deletion (which would be classified as a loss-of-function) would not lead to disease in an autosomal recessive condition unless there was a second mutation on the other allele. Similarly, neither deletion nor duplication of FGFR3 would lead to achondroplasia, since this condition is caused by missense mutations that result in a gain-of-function.
Generally, copy number variations observed through chromosomal microarray testing are classified into 3 categories: pathogenic, uncertain, and benign. Pathogenic results include copy number variations that are known to, or are very likely to, explain the patient’s phenotype based on current knowledge. The pathogenic category includes copy number variations that are consistent with established contiguous gene syndromes or which include an individual gene or genes known to cause a particular phenotype due to either haploinsufficiency or additional copies of the gene. In addition, large copy number variations (generally those that are cytogenetically visible) and those containing significant gene content such that they would be expected to cause an abnormal phenotype are typically considered pathogenic. The benign category includes copy number variations that occur at sufficient frequency in the general population to be considered benign polymorphisms, and those polymorphisms described in published studies that support their benign nature. Since chromosomal microarray testing detects benign copy number variation in nearly all individuals, benign copy number variation is generally not included on the clinical report. Thus, a normal chromosomal microarray test report indicates that no clinically significant copy number changes were detected.
Uncertain chromosomal microarray testing results are occasionally encountered and are a by-product of our evolving understanding of the normal copy number variation in humans. Copy number variations of uncertain clinical significance can be thought of as similar to the variants of undetermined significance that are encountered during the DNA sequencing of disease genes in clinical molecular genetics laboratories. In general, uncertain results include copy number changes that are not easily interpretable due to lack of sufficient scientific data to confidently support a categorization into either the pathogenic or benign categories. Based on our experience with many copy number variations of uncertain clinical significance, it is useful to break this category down into 3 subcategories: likely pathogenic, likely benign, and those that are truly uncertain. The “likely pathogenic” category may include copy number variations that have been described in association with a clinical phenotype in a single case report or those that include several genes and have not been observed previously in internal laboratory databases or databases that catalogue normal human copy number variation (eg, Database of Genomic Variants or dbVar). The “likely benign” category may include copy number variations that have not been previously described, that contain no known genes, or for whichthere is insufficient evidence in databases of normal variation to classify it as a common polymorphism. The “truly uncertain” category consists of copy number variations that may have a limited number of genes, or for which there is conflicting data regarding pathogenicity in publications or databases.
For copy number variations of uncertain clinical significance, testing parental samples can be a useful tool to help assess the clinical significance of chromosomal microarray results and is generally performed at no additional charge to the patient or family. De novo (a new mutation not inherited from either parent) rearrangements are generally interpreted as strong evidence for association with a clinical phenotype. Therefore, if a copy number variation of uncertain clinical significance is found to be de novo, the likelihood that the copy number variation is contributing to the patient’s phenotype is increased. However, it is difficult to unequivocally assign clinical significance based on the inheritance pattern in a single family for a variety of reasons, including the unknown background germline mutation rate for copy number variations and potential nonpaternity.
Alternatively, parental testing may reveal that the copy number variation was inherited from a parent. Inherited copy number variations are generally interpreted as strong evidence that the copy number variation is a benign familial variant and not responsible for the patient’s phenotype. However, this conclusion must only be drawn after careful clinical evaluation of the parent carrying the copy number variation since clinical phenomena such as variable expressivity and incomplete penetrance must also be considered when interpreting inherited copy number variations. Variable expressivity describes family members who have the same copy number change but exhibit differences in the severity or type of presenting phenotypic features. A classic example of variable expressivity is the 22q11.2 deletion syndrome where families often exhibit phenotypic variation despite the common underlying genetic cause (Figure 6). Incomplete penetrance describes families where a genetic abnormality leads to a clinical phenotype in some family members but others who carry the same abnormality may be completely normal. The copy number change, in essence, is a genetic predisposition to a phenotype which only a percentage of individuals will exhibit. Overall, parental testing has significant utility in defining the clinical significance of uncertain copy number variations. However, the inheritance data must be interpreted with caution and in the context of an appropriate medical evaluation.
Figure 6. This figure represents the pedigree of a family with a 22q11.2 deletion exhibiting variable expressivity
The rapid proliferation of array-based technology into clinical laboratories has led to a lack of uniform guidelines for the clinical interpretation of observed copy number variations, at times causing confusing results or conflicting reports between laboratories.8 To address the lack of uniformity, an independent group of experts in this field from laboratories around the world has formed the International Standards for Cytogenomic Array (ISCA) Consortium. The primary goals of the consortium are to develop evidence-based standards for chromosomal microarray design, to build a public database of clinical array data as a resource for the clinical and research communities, and to utilize the database to develop standards and guidelines for the interpretation of copy number changes in the clinical setting. The consortium currently consists of more than 160 international laboratories, including the Mayo Clinic Cytogenetics Laboratory, committed to improving the quality of patient care through the free sharing of clinical chromosomal microarray data.
One of the greatest aids to interpreting chromosomal microarray findings is the availability of comprehensive clinical and family history information about the patient. This may mean the difference in how a copy number change is classified and could lead to more comprehensive and useful clinical recommendations. A detailed clinical information form (available at MayoMedicalLaboratories.com) has been designed to help classify phenotypic information as it is deposited into public databases. Since detailed phenotypic information is extremely useful to the laboratory for the interpretation and classification of copy number variations, we strongly encourage all ordering physicians to use it to provide adequate clinical information for all patients.
Chromosomal microarray testing is now recommended as the first-tier test to detect chromosomal imbalances in individuals with developmental delay/intellectual disability, autism spectrum disorders, or multiple congenital anomalies. Chromosomal microarray testing detects a genetic cause for these clinical features in 15% to 20% of cases, a substantial increase in the diagnostic yield in this patient population. For individuals with multiple miscarriages, infertility, or who are suspected to have sex chromosome abnormalities, such as Turner or Klinefelter syndromes, a conventional chromosome study remains the most appropriate test. The interpretation of chromosomal microarray test results is a complex and evolving process that is aided by collaboration between the clinician and clinical laboratory, particularly regarding the submission of detailed clinical information at the time of test referral. A genetic consultation is often of benefit to ensure the appropriate clinical interpretation of chromosomal microarray test results. The Mayo Clinic Cytogenetics Laboratory is committed to providing cutting-edge chromosomal microarray technology as evidenced by the new 180K chromosomal microarray platform, which provides high-resolution whole-genome analysis for cytogenetic copy number changes (#88898 Array Comparative Genomic Hybridization (aCGH), Whole Genome, Constitutional). In addition, we are an active member of the ISCA Consortium, which is working toward continuous test improvement and clinical testing guidelines that are crucial for optimal patient health care.
Authored by: Karen E. Wain, CGC and Erik C. Thorland, PhD
- Ravnan JB, Tepperberg JH, Papenhausen P, et al: Subtelomere FISH analysis of 11,688 cases: an evaluation of the frequency and pattern of subtelomere rearrangements in individuals with developmental disabilities. J Med Genet 2006 Jun;43(6):478-489 Epub 2005 Sep 30
- Manning M, Hudgins L; Professional Practice and Guidelines Committee: Array-based technology and recommendations for utilization in medical genetics practice for detection of chromosomal abnormalities. Genet Med 2010 Nov;12(11):742-745
- Miller DT, Adam MP, Aradhya S: Consensus statement: chromosomal microarray is a first-tier clinical diagnostic test for individuals with developmental disabilities or congenital anomalies. Am J Hum Genet 2010 May 14;86(5):749-764
- Baldwin EL, Lee JY, Blake DM, et al: Enhanced detection of clinically relevant genomic imbalances using a targeted plus whole genome oligonucleotide microarray. Genet Med 2008 May;10:415-429
- Astbury C, Christ LA, Aughton DJ: Detection of deletions in de novo “balanced” chromosome rearrangements: further evidence for their role in phenotypic abnormalities. Genet Med 2004 Mar-Apr;6(2):81-89
- Gribble SM, Prigmore E, Burford DC, et al: The complex nature of constitutional de novo apparently balanced translocations in patients presenting with abnormal phenotypes. J Med Genet 2005;42:8-16
- Shaffer LG, Beaudet AL, Brothman AR, et al: Microarray analysis for constitutional cytogenetic abnormalities. Genet Med 2007 Sep;9(9):654-662
- Tsuchiya KD, Shaffer LG, Aradhya S, et al: Variability in interpreting and reporting copy number changes detected by array-based technology in clinical laboratories. Genet Med 2009 Dec;11(12):866-873