This policy is based on a Blue Cross Blue Shield Association (BCBSA) Technology Evaluation Center (TEC) Special Report on array comparative genomic hybridization (aCGH). Since that TEC Report was written, the technology has rapidly increased in resolution, and chromosomal microarray (CMA) has become the term of general use to accommodate all variations in the technology. Increased resolution arrays have been quickly translated to clinical services with a resulting increase in diagnostic yield, but also an increase in the potential for results of undetermined significance. Surveys conducted two to three years ago indicated that there is a lack of consensus between laboratories in the interpretation and reporting of copy number variants (CNVs), particularly those that are challenging (Tsuchiya et al., 2009). The International Standards for Cytogenomic Arrays (ISCA) database now offers increased standardization and classification of CNVs that have been previously reported and should improve consensus in reporting.
Diagnosis of developmental delay/intellectual disability or autism spectrum disorder
The diagnosis of developmental delay (DD) is reserved for children younger than age 5 years who have significant delay in two or more of the following developmental domains: gross or fine motor, speech/language, cognitive, social/personal, and activities of daily living (Moeschler, 2008). The diagnosis implies DD that may be significant and may predict life-long disability, although not all children diagnosed with DD will later be diagnosed with intellectual disability.
Intellectual disability (ID) is a life-long disability diagnosed at or after age 5 when intelligence quotient (IQ) testing is considered valid and reliable. The Diagnostic and Statistical Manual of Mental Disorders of the American Psychiatric Association (DSM-IV), defines patients with ID as having an IQ less than 70, onset during childhood, and dysfunction or impairment in more than two of areas of adaptive behavior or systems of support.
According to the DSM-IV, pervasive developmental disorders (PDD) encompass five conditions: autistic disorder, Asperger disorder, pervasive developmental disorder-not otherwise specified (PDD-NOS), childhood disintegrative disorder, and Rett syndrome. While the term autism spectrum disorder (ASD) is not mentioned in the DSM-IV, it is now accepted to include the first three in this list. However, ASD, PDD, and autism are often used interchangeably (Caronna et al., 2008). These conditions are characterized by varying degrees of restrictions in communication and social interaction, and atypical behaviors.
Some children present with features of both DD/ID and of autism. For example, Yeargin-Allsopp et al. reported that nearly 70% of children with a validated diagnosis of ASD, sampled from five metropolitan Atlanta counties, had cognitive impairment. The evaluation pathway depends on the pediatrician, consulting specialists, and their consensus on the primary neurodevelopmental diagnosis.
Post-natal CMA analysis
Several studies have conducted CMA analysis on samples with known chromosomal abnormalities by standard karyotyping (BCBSA TEC Report, 2009). In general, currently available CMA clinical services achieve near 100% sensitivity for known chromosomal abnormalities. False-positive rates (i.e., CNVs of undetermined clinical significance) on known normal samples were inconsistently reported and could not be summarized. One study evaluated the analytic validity of an oligo array and reported 99% sensitivity and 99% specificity, with a resolution of 300–500 kilobases (Kb) for 10 selected cases with different known chromosomal abnormalities (Xiang and Valentin, 2008).
Several studies reported the diagnostic yield of CMA analysis in DD/ID or ASD patients with normal standard karyotype and in several cases normal FMR1 gene analysis and/or subtelomere FISH screening (BCBSA TEC Report, 2009). Overall, diagnostic yield ranged from 5% to 16.7% in DD/ID patients and from 3.4% to 11.6% in patients with ASD; for this compilation, studies differed considerably in array resolution and in patient selection criteria. This compares well with a synthesis of studies recently published by the ISCA Consortium, reporting an average diagnostic yield of 12.2% across 33 studies (Miller et al., 2010). Hochstenback et al. (2009) reported a CMA diagnostic yield of 19% for 36,325 DD/ID cytogenetic referrals in the Netherlands; and Shen et al. (2010) reported a 7% diagnostic yield among 933 ASD referrals. Cooper et al. (2011) studied CMA analyses from over 15,000 individuals with DD/ID, ASD, and/or various congenital abnormalities and compared them to CMA analyses from over 8,000 unaffected controls, finding a significant excess of large CNVs among cases compared to controls. Using a common cutoff for CNV size, about 26% of cases had a CNV larger than 400 kilobases (kb) compared to about 12% of controls, suggesting that CNVs of this size account for approximately 14% of cases. CNVs larger than 400 kb were also significantly more common among cases with multiple congenital abnormalities.
Since the introduction of CMA analysis in about 2005, 18 new genomic disorders have been described, more than doubling the number of disorders described in the previous 20 years (Mefford, 2009). Using CMA in place of conventional cytogenetic testing would have missed 0.6-0.8% of all cases, i.e., those with balanced translocations (Hochstenback et al., 2009; Rauch et al., 2006).
A portion of the increased diagnostic yield from CMA analysis comes from the discovery that some chromosomal rearrangements that appear balanced (and therefore not pathogenic) by G-banded karyotype analysis are found to have small imbalances with greater resolution. It has been estimated that 40% of apparently balanced de novo or inherited translocations with abnormal phenotype are associated with cryptic deletion if analyzed by CMA (Schluth-Bolard, et al., 2009). This contradicts earlier assumptions about inherited, apparently balanced rearrangements and shows that microarray analysis can allow for a less subjective and more accurate interpretation of an abnormal banding pattern (South et al., 2010).
Neither standard cytogenetic nor CMA analysis have been systematically studied for impact on clinical outcomes other than diagnosis (Subramonia-lyer, et al., 2007; Moeschler and Shevell, 2006); Schaefer and Mendelsohn (2008) acknowledge, for example, that a genetic diagnosis “typically will not change interventions for the [autism] patient.” Rather, clinical utility of genetic testing is primarily inferred based on the value of diagnosis to the family, estimation of recurrence risk, and on the importance of early detection and early intervention (Moeschler and Shevell, 2006). Two studies indirectly addressed clinical outcomes other than diagnosis as a result of CMA analysis.
Saam et al. (2008) interviewed 14 physicians (two neurologists, 12 medical geneticists) regarding management changes as a result of positive CMA test results from the University of Utah Cytogenetics Laboratory for 48 patients with DD or ID and normal karyotypes. Only 29% of patients had no management changes reported. For significant proportions of patients, the diagnostic odyssey was ended. However, this study was only a survey and did not attempt to quantitate the diagnostic tests avoided. Saam et al. also reported that 14.6% of patients with genetic diagnoses were referred to medical specialists, and 25% had improved access to insurance and educational services, but the study did not assess the benefits of specialist referrals or screening for comorbidities on patient outcomes, or describe and quantitate the improvement in access to community services.
Coulter et al. (2011) identified and reviewed, over the course of one year, the medical records of all patients at a tertiary children’s hospital who had CMA results showing an abnormal variant or a variant of possible significance. A board-certified medical geneticist reviewed the clinical notes from the ordering provider and abstracted recommendations for clinical actions (a specialist referral, imaging study, diagnostic test, or medication prescription) made specifically as a result of the CMA result. Of 1,792 patients for whom CMA was ordered during the year reviewed, 131 had an abnormal variant and 104 had a variant of possible significance. Of these, 121 and 73 patients were included in the analysis. Overall, patients with an abnormal variant had a significantly higher rate of recommended clinical action (54%) than patients with a variant of possible significance (34%; p=0.01). Among patients with an abnormal variant and a diagnosis of DD/ID or congenital anomalies, about two-thirds of patients were referred for additional clinical action based on the CMA results, whereas referrals were made for 27% of patients with ASD and an abnormal variant. Referral rates were similar for patients with a CMA result of a variant of possible significance, with the exception of patients with congenital anomalies, who were referred for additional clinical action only 17% of the time. Patients younger than 2 years were significantly more likely to have clinical anomalies and were significantly more likely to have abnormal variants. Cases were described in which ancillary CMA results suggested clinical interventions for the present or future regarding possible co-morbid conditions. In no patients, however, were referrals linked to actual patient outcomes; the authors report that this study is ongoing.
Risk estimates for recurrence of disease in future births can be altered considerably by information from the genetic diagnosis. For example, the average sibling recurrence risk in ASD is 5% (Freitag et al., 2010). However, if the cause is a dominant single gene disorder with full penetrance and a parent is a carrier, the sibling risk is 50%. If the disorder is recessive but characteristics are otherwise the same, the sibling risk is 25%. If the cause is Fragile X, the recurrence risk in a brother is 50%, while a sister may be only mildly affected but will have a carrier risk of up to 50%. However, in the case of a de novo CNV (i.e., not carried by either parent), the sibling risk remains low, at the population average.
Knowledge of recurrence risk is expected to lead to improved future reproductive decision making in families with children affected with DD/ID or ASD associated with specific mutations. Turner et al. (2008) studied the reproductive decisions of women from 38 families characterized by male members with and a pattern consistent with chromosome X-linked transmission. Most of the women in these families spent many years knowing that they were at some risk of being carriers and of having a boy with mental retardation/intellectual disability. Prior to the availability of pathogenic mutation analysis, the birth rate for these families was below average for the district (United Kingdom-New South Wales), 1 in 27 versus 1 in 11 per year, respectively. After pathogenic mutation status was determined, both carriers and non-carriers (previously thought to be at risk) of the mutation had children at same rate, with 74% of carriers choosing prenatal genetic evaluation. While the results of this study are suggestive, they do not show that knowledge of recurrence risk directly affected reproductive decisions. Saam et al., (2008) in the survey described previously, reported that recurrence risk evaluation was possible in about one-third of families after positive aCGH results but did not study the impact of recurrence risk evaluation on reproductive planning.
As noted in the Description, guidelines emphasize the importance of cytogenetic evaluation to look for certain kinds of mutations that may be linked to specific conditions for early diagnosis and intervention. However, the benefits of early intervention for these disorders are uncertain. Few randomized trials have been conducted, and the interventions differ considerably in the available studies, indicating that the field is still early in researching the critical elements of effective early intervention. For well-characterized genetic syndromes, it may be important to incorporate monitoring for comorbidities known to be associated with the condition. For example, 22q11 microdeletion syndrome (includes diGeorge and velo-cardio-facial syndromes) is associated with development of hearing impairment in a significant proportion of patients and subsequent delayed speech (Digilio et al., 1999). Velo-cardio-facial syndrome is also associated with heart defects (Freitag et al., 2010). Klinefelter syndrome may first be detected as developmental delay in early childhood; androgen treatment is an important component of therapy (Freitag et al., 2010). CMA analysis may also predict future conditions for which interventions are possible. In a report of three cases, one patient had a chromosomal deletion that included a gene associated with autosomal dominant Peutz-Jeghers syndrome (PJS); tumor screening protocols for males with PJS generally begin with upper and lower endoscopy with small-bowel follow-through radiographs beginning at age 8 years (Adam et al., 2009). Two other patients had a de novo deletion of chromosome 17p encompassing the TP53 tumor suppressor gene responsible for Li-Fraumeni syndrome (LFS); tumor screening protocols for LFS also begin in childhood. In another report, a child presenting to a neurology service with unusual behaviors was found to have a deletion that included exons of the DMD gene (Duchenne muscular dystrophy) associated with Becker muscular dystrophy (BMD). Additional testing revealed a markedly elevated creatine kinase, and a thorough physical exam was consistent with BMD. This diagnosis explained some of the child’s behavior and prompted a plan for future surveillance for cardiac and other complications of BMD, as well as carrier testing and surveillance of the child’s mother (Coulter et al., 2011).
The American Academy of Neurology and the Practice Committee of the Child Neurology Society updated their guideline regarding the evaluation of unexplained global developmental delay/intellectual disability with information on genetic and metabolic (biochemical) testing in order to accommodate advances in the field (Michelson, et al., 2011). The guidelines conclude that CMA testing has the highest diagnostic yield in children with DD/ID, that the often complex results require confirmation and careful interpretation, often with the assistance of a medical geneticist, and that CMA should be considered the first-line test. The guidelines acknowledge that “Research is sorely lacking on the medical, social, and financial benefits of having an accurate etiologic diagnosis.”
The American College of Medical Genetics (ACMG) published guidelines on array-based technologies and their clinical utilization for detecting chromosomal abnormalities (Manning and Hudgins, 2010). Chromosomal microarray testing for copy number variation is recommended as a first-line test in the initial postnatal evaluation of individuals with the following:
- Multiple anomalies not specific to a well-delineated genetic syndrome
- Apparently non-syndromic developmental delay/ intellectual disability
- Autism spectrum disorders
ACMG also recommends against use of CMA in cases of multiple miscarriages.
Additional ACMG guidelines have been published for the design and performance expectations for clinical microarrays and associated software (Kearney et al., 2011, p.676-9) and for the interpretation and reporting of CNVs, (Kearney et al., 2011, p. 680-5) both intended for the postnatal setting (see Description).
The International Standard Cytogenomic Array Consortium published a Consensus Statement in which they recommend offering CMA as the first-tier genetic test, in place of G-banded karyotype, for patients with unexplained DD/ID, ASD, or multiple congenital anomalies (MCA). “Except in special cases, such as those involving family history of multiple miscarriages, a karyotype is not cost effective in a child with DD/ID, ASD, or MCA and a negative array study. CMA testing is not inexpensive, but the cost is less than the cost of a G-banded karyotype plus a customized FISH test such as subtelomeric FISH, and the yield is greater” (Miller et al., 2010).
CMA analysis offers a higher resolution approach to detecting the presence of chromosomal alterations that have been associated with cases of developmental delay/intellectual disability or autism spectrum disorder compared to karyotyping and ancillary testing. However, the diagnostic yield remains low in unselected populations without accompanying signs and/or symptoms. In individuals with apparent nonsyndromic DD/ID, or suspected ASD and accompanying malformations, the diagnostic yield is much higher and is higher than the yield of karyotype testing.
Evidence on the clinical benefit of CMA testing is largely anecdotal. Cases have been documented in which the information derived from testing ends a long diagnostic odyssey, aids in planning for surveillance or management of associated comorbidities, and assists in future reproductive decision-making. While systematic studies of the impact of CMA analysis on patient outcomes is lacking, the improvement in diagnostic yield has been well-demonstrated, and feedback from physician specialty societies, academic medical centers, and in respected guidelines is consistent in supporting the clinical benefit of CMA testing for defined populations. As a result, chromosomal microarray analysis may be considered medically necessary in individuals with developmental delay or autism spectrum disorders who meet the clinical criteria defined the Coverage section.
Disclaimer for coding information on Medical Policies
Procedure and diagnosis codes on Medical Policy documents are included only as a general reference tool for each policy. They may not be all-inclusive.
The presence or absence of procedure, service, supply, device or diagnosis codes in a Medical Policy document has no relevance for determination of benefit coverage for members or reimbursement for providers. Only the written coverage position in a medical policy should be used for such determinations.
Benefit coverage determinations based on written Medical Policy coverage positions must include review of the member’s benefit contract or Summary Plan Description (SPD) for defined coverage vs. non-coverage, benefit exclusions, and benefit limitations such as dollar or duration caps.