DOI: 10.1002/pd.3863
EXPERT OPINION
Integration of microarray technology into prenatal diagnosis: counselling issues generated during the NICHD clinical trial Ronald J. Wapner1*, Deborah A. Driscoll2 and Joe Leigh Simpson3† 1
Department of Obstetrics and Gynecology, Columbia University College of Physicians and Surgeons, New York, NY, USA Hospital of the University of Pennsylvania, School of Medicine, Philadelphia, PA, USA 3 College of Medicine, Florida International University, Miami, FL, USA *Correspondence to: Ronald J. Wapner E-mail:
[email protected] † Present address: March of Dimes Foundation, White Plains, New York, NY, USA 2
ABSTRACT Cytogenetic microarray analysis (CMA) in prenatal testing detects chromosome abnormalities and new genetic syndromes that would be missed by conventional cytogenetics and has the potential to significantly enhance prenatal genetic evaluation. A large Eunice Kennedy Shriver National Institute Of Child Health and Human Development (NICHD)sponsored multicentered trial to assess the role of CMA as a primary prenatal diagnostic tool has been completed, and results will soon be available. Integration of this technology into clinical care will require thoughtful changes in patient counseling. Here, we examine four cases, all ascertained in the NICHD prenatal microarray study, to illustrate the challenges and subtleties of genetic counseling required with prenatal CMA testing. Although the specifics of each case are distinct, the underlying genetic principles of uncertainty, variable expressivity, and lack of precise genotype–phenotype correlation are well known and already part of prenatal counseling. Counselor and practitioner education will need to include both the science of interpreting array findings as well as development of improved approaches to uncertainty. A team approach to interpretation will need to be developed, as will standardized guidelines by professional organizations and laboratories. Of equal import is additional research into patient attitudes and desires, and a better understanding of the full phenotypic spectrum of copy number variants discovered in utero. © 2012 John Wiley & Sons, Ltd. Funding sources: Support for this work was partially provided by the National Institutes of Health under grant no. R01 HD055651. Conflicts of interest: None declared
INTRODUCTION Microarray copy number analysis has proven to be a powerful tool in the evaluation of infants and children with congenital anomalies, dysmorphic features, and neurocognitive disabilities including developmental delay and autism spectrum disorder, with clinically significant findings reported in up to 15% of cases with normal conventional karyotypes.1–3 On the basis of this experience, the use of cytogenetic microarray analysis (CMA) in prenatal testing is being utilized in some centers, especially for the evaluation of fetal structural anomalies identified by ultrasound.4–13 Further, a large National Institute Of Child Health and Human Development-sponsored multicentered trial to assess the role of CMA as a primary prenatal diagnostic tool has been completed, and results will be available shortly. Reports to date suggest that CMA is a more sensitive diagnostic test and adds incremental value to conventional karyotyping with the potential to replace karyotyping in the initial cytogenetic evaluation of the fetus. As the technical aspects of CMA testing continue to evolve and prospective studies are completed to evaluate the clinical validity and utility of CMA in the prenatal setting, integration of this information into clinical care will require thoughtful changes in patient counseling. Although the paradigm for genetic testing in the postnatal setting has shifted to CMA as the first-line test in many Prenatal Diagnosis 2012, 32, 396–400
situations, there are distinct differences in the prenatal use of genetic information. Although infants and children are identified for testing because of structural or developmental anomalies allowing correlation of the CMA results with the clinical features, prenatal results frequently require counseling with incomplete or absent knowledge of the phenotype because of the inherent limitations of ultrasound investigations. This is further complicated by the present literature describing the clinical effects of microdeletions and duplications on the basis of postnatal recognition of an abnormal phenotype and likely biased toward the more severe end of the phenotypic spectrum as those children are more likely to come to clinical attention and be referred for diagnostic testing. One of the unique aspects of prenatal testing is the option to use the information to make difficult decisions about continuation of a pregnancy, whereas postnatal results provide parents with an explanation for the child’s phenotype and are used to plan future management of the child and future reproductive planning. Parents choosing prenatal testing are frequently using the information to assure the birth of an unaffected child. In most cases, determination of the clinical significance of a copy number variant (CNV) discovered during prenatal testing is straightforward. Many of these have been previously described © 2012 John Wiley & Sons, Ltd.
Prenatal microarray counseling issues
in the literature and/or curated databases. Some CNVs are benign variants that have no significant clinical impact. Others have been confirmed as being pathogenic, but the penetrance and expressivity are variable. If insufficient evidence is available to unequivocally determine if a CNV is of either benign or pathogenic clinical significance, then the CNV is reported by the lab as a CNV of uncertain clinical significance. Although with more experience uncertain findings are becoming less frequent, they still occur in up to 2% of prenatal cases,14 with the frequency varying depending on the content and resolution of the array. Although high-resolution whole genome arrays are likely to discover more clinically significant findings, they are also more apt to uncover findings of uncertain significance.9 Some arrays are designed to emphasize detection of known microdeletion/ microduplication syndromes, such as the 22q11.2 deletion or Smith–Magenis syndrome, as well as other known loci of inherited Mendelian disorders. These targeted arrays may also provide less dense coverage throughout the genome to detect large changes in the ‘nontargeted’ (backbone) regions. This can not only minimize the chance of finding a variant of uncertain significance but also may fail to detect a rare/novel genomic imbalance. However, regardless of which array design is used, the need to counsel patients about uncertainty will be necessary. When a variant of uncertain significance is found, predictive counseling will require determination of the likelihood that a CNV is pathogenic and likely to result in a clinically significant phenotype on the basis of size, gene content, review of the literature, and information in databases. Although clinical information such as ultrasound findings, family history, and inheritance of the CNV can be taken into consideration, none of these will allow definitive prediction of the phenotype. Though sonographic anomalies may be suggestive of an association, it is also possible that they are coincidental. Likewise, although inheritance from a normal parent could be considered reassuring and a de novo event more suspicious, incomplete penetrance and variable expressivity must be considered as a possibility even for inherited CNVs. Early experience with the interpretation of array data in prenatal diagnosis has already demonstrated some of the challenges that practitioners and counselors will face as CMA is incorporated into clinical practice. Here, we outline some of these areas with clinical examples and discuss potential counseling dilemmas and options. We also suggest topics that should be reviewed with patients during a pretest counseling session. Although CMA raises new questions and will require the use of new tools and knowledge, the prenatal genetic counseling approach and issues are similar to those we address for any type of genetic testing. The following four cases, all ascertained in the National Institute Of Child Health and Human Development prenatal microarray study, illustrate the challenges and subtleties of genetic counseling in this new era of genetic testing.
Case 1: Invasive testing performed for ultrasound finding of a unilateral club foot. Array results: A de novo 1.60-Mb gain (microduplication) arr7q11.23 (72,404,248–73,976,981) 1 Review of the literature reveals a well-characterized 7q11.2 microduplication syndrome, the reciprocal duplication of the Williams–Beuren syndrome critical region.15 The probands were Prenatal Diagnosis 2012, 32, 396–400
397
identified in a series of patients with idiopathic intellectual disability. An association with autism is also present. Both familial and de novo cases have been described. Variable speech delay is a constant finding and cognitive abilities range from normal to moderate mental retardation. Hypotonia and congenital anomalies, including club feet, were noted in some probands. Further confirmation of the pathogenic potential of this microduplication comes from review of two large databases. Of the 17 cases of this microduplication reported in the experience of the ISCA consortium, 16 were discovered in patients studied because of developmental delay and only one was seen in a large series of “normal” individuals.16 This results in an odds ratio of 10.3 for a pathogenic phenotype (p =0.0046).17 Cooper et al. have reported similar findings. On the basis of this information, the couple can be counseled that the array finding is pathogenic. Predicting the impact on future development is less certain. Although all probands had significant developmental delay leading to their testing, many of the parents having the same microduplication were employed and functioning well, albeit with a history of learning difficulties, language delay, or both. This raises the question of whether some patients outgrow some of their childhood cognitive problems or whether the phenotype in the specific adult patient group studied represents the mild end of the phenotypic spectrum. If it were not for the more severely affected children, parents would never have been diagnosed. It, thus, remains to be determined how frequently the duplication is present in cohorts of unaffected or mildly affected controls. Long-term follow-up studies of pregnancies with this microduplication but without in utero abnormalities are required to address this question before one can definitively counsel patients.
Case 2: Chorionic villus sampling done for previous child with developmental delay secondary to dup 5q31.2-31.3 (5.7 Mb). CMA analysis of this case: Normal 5q but a de novo 450-kb duplication arr1q22 (154,348,185–154,799,248) 3 Review of the literature and DECIPHER database18 reveals no similar cases. The UCSC genome browser19 shows 18 genes in the duplicated segment; two of which are potentially diseasecausing. Neuronal acetylcholine receptor subunit beta-2 is a protein that in humans is encoded by the CHRNB2 gene that resides in the duplicated segment. Mutations in this gene are associated with autosomal-dominant nocturnal frontal lobe epilepsy. The RNA-specific adenosine deaminase gene is also located in this segment, and in both humans and animals, mutations of this gene are associated with dyschromatosis symmetrica hereditaria (OMIM DSH; 127400), a benign autosomal-dominant skin disorder. On the basis of these findings, the patient was informed that the finding was of uncertain clinical significance. When a result is uncertain, the clinician should assist the patient in understanding the relative likelihood of the CNV being pathogenic or benign. Any such speculation must be well supported, but the uncertainty of this classification must still be communicated. In this case, the small size of the CNV, the fact that similar cases have not been reported, and the much lower likelihood that a duplication will be pathogenic compared with © 2012 John Wiley & Sons, Ltd.
398
a deletion are all reassuring. However, there is biologic plausibility of a potential phenotypic consequence on the basis of the gene function. All of these factors make counseling this patient difficult. Cases such as this one demonstrate the value of a multidisciplinary team approach composed of laboratory personnel familiar with CMA technology, a clinical geneticist, and genetic counselors. Each team member should be well educated in the unique aspects of CMA interpretation, and if such a resource is not locally available, referral is appropriate. Recall, however, that uncertain results are not unique to CMA technology and arise in interpretation of all forms of genetic testing, including evaluation of a karyotype. As whole genome arrays with higher resolution are designed, more findings of uncertain clinical significance will be identified.9 This has led to the suggestion that these not be reported to the patient.20 Others have found this approach too paternalistic and believe that with counseling, patients are capable of understanding uncertain or inconclusive results. Alternatively, as part of the pretest counseling and consent process, patients choose whether they wish to be made aware of uncertain laboratory results. At present, there is no consensus on the best approach or practice guidelines; hence, it is recommended that laboratories discuss this with referring physicians to develop local policies.
Case 3: Patient has an amniocentesis for positive Down syndrome screening. CMA analysis: A de novo 751-kb deletion in 22q11.2. arr22q11.2 (19,084,421-19,835,417) 1. A fetal echocardiogram is normal Another difficult area of interpretation results when a CNV is discovered in a known pathogenic region, but the size or location is different than the previously described cases. In this case, the deletion was within the well-characterized DiGeorge syndrome region but is smaller than the 3 Mb deletion found in 90% of patients with features of DiGeorge and velocardiofacial syndromes. CMA results indicate that the deletion is located within the distal end of the 3 Mb common deletion, between two of the eight low-copy repeat sequences in this region, which predispose to nonallelic homologous recombination and result in deletions. Review of the literature revealed that about 7% of patients with DiGeorge/velocardiofacial syndrome have a smaller, nested deletion of about 1.5 Mb within the common deletion, and the remaining patients have smaller ‘atypical’ deletions within the region or distal to the common deletion region.21–23 The size and location of deletions within this region do not correlate with the phenotype, which is highly variable.24 In this case, a detailed anatomic survey and fetal echocardiogram are indicated to look for evidence of congenital anomalies associated with the deletion. The absence of a heart defect on fetal echocardiogram does not exclude the potential for this CNV to be pathogenic. Cardiac defects occur in only 75% of patients with 22q11.2 deletions, and subtle cardiac lesions could be evident later in pregnancy or postnatally.25 Although this is an atypical 22q11 deletion, and does not include the major candidate gene (TBX 1), the most prudent approach would be to counsel the parents about the potential for any of the features of the 22q11.2 deletion syndrome Prenatal Diagnosis 2012, 32, 396–400
R. J. Wapner et al.
including neurocognitive difficulties, speech delay, and psychiatric and behavioral problems, as well as recommend a postnatal evaluation by a geneticist or individual with expertise with the deletion syndrome.
Case 4: Chorionic villus sampling performed for advanced maternal age. CMA analysis: A 203-kb maternally inherited deletion arr3p26.2. (4,395,412–4,598,515) 1mat The majority of cytogenetic disorders discovered by prenatal karyotype result in significant fetal and neonatal morbidities that commence at birth and result in severe lifetime morbidity and neurocognitive impairment. This is not always the case with CMA, and some microdeletions and duplications may result in adult onset diseases. Review of the literature in Case 4 revealed that this deletion is associated with spinocerebellar ataxia 15, an adult onset disorder with significant variable expressivity.26 Onset is between ages 7 and 66 years, usually with gait ataxia but sometimes with tremor. Affected individuals remain ambulatory for 10–54 years after onset. Appropriate counseling for the fetus and mother in this case is uncertain. Some guidance comes from The American College of Medical Genetics27 that recommends that when CNVs associated with presymptomatic conditions are discovered postnatally, the CNV should be reported to facilitate early access to medical care. This may not always be appropriate or desirable with prenatal testing and may result in heightened parental anxiety. One may query parents during the pretest consent process as to whether they wish to be informed and include their decision in the consent form which will provide referring physicians and genetic counselors some guidance when such findings are discovered. However, fully informing parents of the implications of such findings during a preprocedure counseling session is impractical, and counselors will still have to use their discretion of how to manage each specific situation. For example, when CNVs associated with a treatable or modifiable condition such as neoplasia are discovered and information on penetrance, lifetime risk, disease spectrum, and clinical management are known, these findings should be reported to the parents to maximize future care and prognosis. Whether to inform the parent of their presymptomatic status for a nonmodifiable adult onset diseases is less clear. In the present case, maternal testing was not performed for the purpose of diagnosing asymptomatic maternal SCA15, and such findings were not anticipated by the mother. Experience with the prenatal diagnosis of Huntington disease, in which some parents go to great strides, including assisted reproductive technology, to avoid knowing their disease status, suggests that not all individuals are prepared for or desire this information. In many labs, discovery of a fetal CNV results in reflex CMA of the parents to determine whether the finding is de novo or inherited. For expedience, this is frequently performed without recontacting the family. We suggest, however, that when the CNV, if inherited, can potentially impact parental health, ascertainment of the parents’ preference prior to testing represents the best approach. © 2012 John Wiley & Sons, Ltd.
Prenatal microarray counseling issues
DISCUSSION Cytogenetic microarray analysis detects chromosome abnormalities and new genetic syndromes that would be missed by conventional cytogenetics and has the potential to significantly enhance prenatal genetic evaluation. Although some may suggest that its use be limited to pregnancies with fetal structural anomalies, it appears that its improved diagnostic ability to detect clinically relevant abnormalities is likely to lead to its adaptation as a primary prenatal diagnostic tool. At first, this will lead to frequent counseling dilemmas, as does the translation of any new technology into clinical care, but with time, education, and the development of appropriate care guidelines, these should become infrequent. Counselor and practitioner education will be important in the implementation of CMA testing and will need to include both an introduction to the science of interpreting array findings as well as development of improved approaches to uncertainty. Although most CNVs can be interpreted on the basis of gene content, size, inheritance, clinical findings, and review of the literature and databases, many practitioners and laboratories will initially require training and experience before becoming comfortable in their interpretation. This transition period will require close collaboration between laboratories performing arrays, clinical geneticists with expert knowledge and expertise in the area, and counselors skilled in relating results to patients. In many centers, a team approach to interpretation will need to be developed, as will local guidelines as to the reporting and management of specific array findings. Professional organizations and laboratories will need to develop standardized guidelines for testing and reporting of results to minimize the uncertainty and ensure that patients and practitioners are provided with accurate information and appropriate counseling tools. Presently, the tools available to assist in interpreting prenatal CMA findings remain incomplete. Most critical
399
is the development of large databases of CNVs discovered during prenatal testing, thus obviating the known ascertainment bias unavoidable in postnatal cases studied only because of existing abnormalities. Only through such data can the full phenotype be sufficiently well defined to provide accurate predictive counseling. Of equal necessity is research into patient attitudes and desires, and an understanding of in which populations the use of CMA technology will be most beneficial. Although each of the aforementioned examples may result in heightened patient anxiety and confusion, fortunately, they will be infrequent if the prenatal diagnosis program insists on thorough pretest counseling and implements a formal consent process. Although the specifics of each case are distinct, the underlying genetic principles of uncertainty, variable expressivity, and lack of precise genotype–phenotype correlation are well known and should be a standard part of routine prenatal counseling. To minimize the impact of these findings, pretest counseling for CMA should cover each of these possibilities. A discussion of the couples’ tolerance for uncertainty and whether they want to be alerted to these findings may be appropriate, but additional research is required to understand the implications of this approach.
WHAT’S ALREADY KNOWN ABOUT THIS TOPIC?
It is known from postnatal evaluation of infants and children with dysmorphic features, neurocognitive difficulties, or structural anomalies that microarray analysis will provide important clinically relevant information and can also result in findings of uncertain clinical significance.
WHAT DOES THIS STUDY ADD?
This study gives an early insight into the counseling quandaries that prenatal array use can generate.
REFERENCES 1. Shevell MI, Bejjani BA, Srour M, et al. Array comparative genomic hybridization in global developmental delay. Am J Med Genet B Neuropsychiatr Genet 2008 Oct 5;147B(7):1101–8. 2. Rosenfeld JA, Ballif BC, Torchia BS, et al. Copy number variations associated with autism spectrum disorders contribute to a spectrum of neurodevelopmental disorders. Genet Med 2010 Nov;12(11):694–702. 3. Miller DT, Adam MP, Aradhya S, 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 May 14;86(5):749–64. 4. Coppinger J, Alliman S, Lamb AN, et al. Whole-genome microarray analysis in prenatal specimens identifies clinically significant chromosome alterations without increase in results of unclear significance compared to targeted microarray. Prenat Diagn 2009 Dec;29(12):1156–66. 5. Kleeman L, Bianchi DW, Shaffer LG, et al. Use of array comparative genomic hybridization for prenatal diagnosis of fetuses with sonographic anomalies and normal metaphase karyotype. Prenat Diagn 2009 Dec;29(13):1213–17. 6. Maya I, Davidov B, Gershovitz L, et al. Diagnostic utility of array-based comparative genomic hybridization (aCGH) in a prenatal setting. Prenat Diagn 2010 Dec;30(12–13):1131–37.
Prenatal Diagnosis 2012, 32, 396–400
7. Sahoo T, Cheung SW, Ward P, et al. Prenatal diagnosis of chromosomal abnormalities using array-based comparative genomic hybridization. Genet Med 2006 Nov;8(11):719–27. 8. Shaffer LG, Coppinger J, Alliman S, et al. Comparison of microarraybased detection rates for cytogenetic abnormalities in prenatal and neonatal specimens. Prenat Diagn 2008 Sep;28(9):789–95. 9. Tyreman M, Abbott KM, Willatt LR, et al. High resolution array analysis: diagnosing pregnancies with abnormal ultrasound findings. J Med Genet 2009 Aug;46(8):531–41. 10. Van den Veyver IB, Patel A, Shaw CA, et al. Clinical use of array comparative genomic hybridization (aCGH) for prenatal diagnosis in 300 cases. Prenat Diagn 2009 Jan;29(1):29–39. 11. Faas BH, van der Burgt I, Kooper AJ, et al. Identification of clinically significant, submicroscopic chromosome alterations and UPD in fetuses with ultrasound anomalies using genome-wide 250 k SNP array analysis. J Med Genet 2010 Sep;47(9):586–94. 12. Park SJ, Jung EH, Ryu RS, et al. Clinical implementation of wholegenome array CGH as a first-tier test in 5080 pre and postnatal cases. Mol Cytogenet. 2011;4:12. 13. Friedman JM. High-resolution array genomic hybridization in prenatal diagnosis. Prenat Diagn 2009 Jan;29(1):20–8.
© 2012 John Wiley & Sons, Ltd.
400
14. Wapner RJ for the Prenatal Microarray NICHD Study Group. A multicenter, prospective, masked comparison of chromosomal microarray with standard karyotyping for routine and high risk prenatal diagnosis. Am J Obstet Gynecol 2012 February 9;206:S2. 15. Van der Aa N, Rooms L, Vandeweyer G, et al. Fourteen new cases contribute to the characterization of the 7q11.23 microduplication syndrome. Eur J Med Genet 2009 Mar-Jun;52(2–3):94–100. 16. Kaminsky EB, Kaul V, Paschall J, 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 Sep;13(9):777–84. 17. Cooper GM, Coe BP, Girirajan S, et al. A copy number variation morbidity map of developmental delay. Nat Genet 2011 Sep;43(9):838–46. 18. Firth HV, Richards SM, Bevan AP, et al. DECIPHER: Database of chromosomal imbalance and phenotype in humans using ensembl resources. Am J Hum Genet 2009 Apr;84(4):524–33. 19. Kent WJ, Sugnet CW, Furey TS, et al. The human genome browser at UCSC. Genome Res 2002 Jun;12(6):996–1006. 20. Berg JS, Khoury MJ, Evans JP. Deploying whole genome sequencing in clinical practice and public health: meeting the challenge one bin at a time. Genet Med 2011 Jun;13(6):499–504.
Prenatal Diagnosis 2012, 32, 396–400
R. J. Wapner et al.
21. Ben-Shachar S, Ou Z, Shaw CA, et al. 22q11.2 distal deletion: a recurrent genomic disorder distinct from DiGeorge syndrome and velocardiofacial syndrome. Am J Hum Genet 2008 Jan;82(1):214–21. 22. Rauch A, Zink S, Zweier C, et al. Systematic assessment of atypical deletions reveals genotype-phenotype correlation in 22q11.2. J Med Genet 2005 Nov;42(11):871–76. 23. Fernandez L, Nevado J, Santos F, et al. A deletion and a duplication in distal 22q11.2 deletion syndrome region. Clinical implications and review. BMC Med Genet 2009;10:48. 24. Emanuel BS, McDonald-McGinn D, Saitta SC, Zackai EH. The 22q11.2 deletion syndrome. Adv Pediatr. 2001;48:39–73. 25. McElhinney DB, McDonald-McGinn D, Zackai EH, Goldmuntz E. Cardiovascular anomalies in patients diagnosed with a chromosome 22q11 deletion beyond 6 months of age. Pediatrics 2001 Dec;108(6):E104. 26. van de Leemput J, Chandran J, Knight MA, et al. Deletion at ITPR1 underlies ataxia in mice and spinocerebellar ataxia 15 in humans. PLoS Genet 2007 Jun;3(6):e108. 27. Kearney HM, Thorland EC, Brown KK, et al. American College of Medical Genetics standards and guidelines for interpretation and reporting of postnatal constitutional copy number variants. Genet Med 2011 Jul;13(7):680–85.
© 2012 John Wiley & Sons, Ltd.