Genetic Testing for Epilepsy
Genetic testing for infantile- and early childhood-onset epilepsy syndromes in which epilepsy is the core clinical symptom (see Policy Guidelines section) may be considered medically necessary for individuals age 5 and younger if positive test results may:
Genetic testing for epilepsy is considered investigational in all other situations.
This policy addresses testing for the common epilepsies, which are also called idiopathic epilepsies. These are defined as epilepsy syndromes that present in childhood, adolescence, or early adulthood, in which epilepsy is the only clinical manifestation and for which there is not a structural or metabolic defect predisposing to epilepsy.
This policy also addresses the rare epilepsy syndromes that present in infancy or early childhood, in which epilepsy is the core clinical symptom (Dravet syndrome, early infantile epileptic encephalopathy, generalized epilepsy with febrile seizures plus, epilepsy and intellectual disability limited to females, Nocturnal frontal lobe epilepsy, and others). Other clinical manifestations may be present in these syndromes, but are generally secondary to the epilepsy itself.
This policy does not address testing for genetic syndromes that have a wider range of symptomatology, of which seizures may be one, such as the neurocutaneous disorders (e.g., neurofibromatosis, tuberous sclerosis) or genetic syndromes associated with cerebral malformations or abnormal cortical development, or metabolic or mitochondrial disorders. Genetic testing for these syndromes may be specifically addressed in other medical policies (see Related Policies)
Note: This policy does not address the use of genotyping for the HLA-B*1502 allelic variant in patients of Asian ancestry prior to considering drug treatment with carbamazepine due to risks of severe dermatologic reactions. This testing is recommended by the U.S. Food and Drug Administration (FDA) labeling for carbamazepine. (1)
Medically Necessary Statement Definitions and Testing Strategy
The medically necessary statement refers to epilepsy syndromes that present in infancy or early childhood, are severe, and are characterized by epilepsy as the primary manifestation, without associated metabolic or brain structural abnormalities. As defined by the International League Against Epilepsy, these include epileptic encephalopathies, which are electroclinical syndrome associated with a high probability of encephalopathic features that present or worsen after the onset of epilepsy. Other clinical manifestations, including developmental delay and/or intellectual disability may be present secondary to the epilepsy itself. Specific clinical syndromes based on the International League Against Epilepsy classification include:
Mutations in a large number of genes have been associated with early onset epilepsies. Some of these are summarized in Table 1.
Table 1: Single-Gene Mutations Associated With Epileptic Syndromes
Application of Medically Necessary Policy Statement
Although there is not standardization in the definition of epileptic encephalopathies, they are generally characterized by at least some of the following:
(1) Onset in early childhood (often in infancy);
(2) Refractory to therapy;
(3) Associated with developmental delay or regression; and
(4) Severe electroencephalogram (EEG) abnormalities.
There is a challenge in defining the population appropriate for testing given that specific epileptic syndromes may be associated with different EEG abnormalities, which may change over time, and patients may present with severe seizures prior to the onset or recognition of developmental delay or regression. However, for the purposes of this policy, the medically necessary policy statement would apply for patients with:
There is clinical and genetic overlap for many of the electroclinical syndromes previously discussed. If there is suspicion for a specific syndrome based on history, EEG findings, and other test results, testing should begin with targeted mutation testing for the candidate gene most likely to be involved, followed by sequential testing for other candidate genes. In particular, if an SCN1A-associated syndrome is suspected (Dravet syndrome, GEFS+), molecular genetic testing of SCN1A with sequence analysis of the SCN1A coding region, followed by deletion/duplication analysis if a pathogenic variant is not identified, should be obtained.
Given the genetic heterogeneity of early-onset epilepsy syndromes, a testing strategy that uses a multigene panel may be considered reasonable. In these cases, panels should meet the criteria outlined in a policy. (see Related Policies) (General Approach to Evaluating the Utility of Genetic Panels).
If the specific gene being tested has been codified in CPT, the appropriate CPT code would be reported. If the specific gene has not been codified in CPT, the unlisted molecular pathology code 81479 would be reported. If a panel of tests that has not been codified in CPT is performed, code 81479 would be reported once.
Here is a list of some of the tests related to epilepsy that are listed under CPT Tier 2 codes:
NOTE: Commercially available tests include GeneDx®, GeneDx® Childhood Onset epilepsy panel, GeneDx® Infancy Panel. The Courtagen epiSEEK®, Emory Genetics® Epilepsy and Seizure Disorders Sequencing panel. The tests are not cleared for marketing by the U.S. Food and Drug Administration (FDA). Each is available under the auspices of the Clinical Laboratory Improvement Act (CLIA). (See Regulatory Status).
Epilepsy is a disorder characterized by unprovoked seizures. It is a heterogeneous condition that encompasses many different types of seizures and that varies in age of onset and severity. The common epilepsies, also called idiopathic epilepsy, are thought to have a complex, multifactorial genetic basis. There are also numerous rare epileptic syndromes that occur in infancy or early childhood and that may be caused by a single gene mutation. Genetic testing is commercially available for a large number of genetic mutations that may be related to epilepsy.
Epilepsy is defined as the occurrence of 2 or more unprovoked seizures. It is a common neurologic disorder, with approximate 3% of the population developing the disorder over their entire lifespan. (2) The condition is generally chronic, requiring treatment with one or more medications to adequately control symptoms. Seizures can be controlled by anti-epileptic medications in most cases, but some patients are resistant to medications, and further options such as surgery, vagus nerve stimulation, and/or the ketogenic diet can be used. (3)
Epilepsy is heterogeneous in etiology and clinical expression and can be classified in a variety of ways. Most commonly, classification is done by the clinical phenotype, i.e., the type of seizures that occur. The International League Against Epilepsy (ILAE) developed the classification system shown in Table (2,4) which is widely used for clinical care and research purposes. Classification of seizures can also be done on the basis of age of onset:
Table 2. Classification of Seizure Disorders by Type (condensed from Berg et al.) (4)
More recently, the concept of genetic epilepsies has emerged as a way of classifying epilepsy. Many experts now refer to “genetic generalized epilepsy” as an alternative classification for seizures that were previously called “idiopathic generalized epilepsies.” The ILAE report published in 2010 offers the following alternative classification (4):
For the purposes of this policy review, this classification is most useful. The policy will focus on the category of genetic epilepsies in which seizures are the primary clinical manifestation. This category does not include syndromes that have multiple clinical manifestations, of which seizures may be one. Examples of syndromes that include seizures are Rett syndrome and tuberous sclerosis. Genetic testing for these syndromes will not be assessed in this policy, but may be included in separate policies that specifically address genetic testing for that syndrome.
Genetic epilepsies can be further broken down by type of seizures. For example, genetic generalized epilepsy refers to patients who have convulsive (grand mal) seizures, while genetic absence epilepsy refers to patients with nonconvulsive (absence) seizures. The disorders are also sometimes classified by age of onset.
The category of genetic epilepsies includes a number of rare epilepsy syndromes that present in infancy or early childhood. (2,5) These are syndromes that are characterized by epilepsy as the primary manifestation, without associated metabolic or brain structural abnormalities. They are often severe and sometimes refractory to medication treatment. They may involve other clinical manifestations such as development delay and/or intellectual disability, which in many cases are thought to be caused by frequent uncontrolled seizures. In these cases, the epileptic syndrome may be classified as an epileptic encephalopathy, which is described by ILAE as disorders in which the epileptic activity itself may contribute to severe cognitive and behavioral impairments above and beyond what might be expected from the underlying pathology alone and that these can worsen over time. (4) A partial list of severe early-onset epilepsy syndromes is as follows:
Dravet syndrome (also known as severe myoclonic epilepsy in infancy or polymorphic myoclonic epilepsy in infancy) falls on a spectrum of SCN1A-related seizure disorders, which includes febrile seizures at the mild end to Dravet syndrome and intractable childhood epilepsy with generalized tonic-clonic seizures at the severe end. The spectrum may be associated with multiple seizure phenotypes, with a broad spectrum of severity; more severe seizure disorders may be associated with cognitive impairment or deterioration. (6) Ohtahara syndrome is a severe early-onset epilepsy syndrome characterized by intractable tonic spasms, other seizures, interictal EEG abnormalities, and developmental delay. It may be secondary to structural abnormalities but has been associated with mutations in the STXBP1 gene in rare cases. West syndrome is an early-onset seizure disorder associated with infantile spasms and the characteristic EEG finding of hypsarrhythmia. There are other seizure disorders that present early in childhood and may have a genetic component but which are characterized by a more benign course, including benign familial neonatal seizures and benign familial infantile seizures.
Genetics of Epilepsy
The common genetic epilepsies are primarily believed to involve multifactorial inheritance patterns. This follows the concept of a threshold effect, in which any particular genetic defect may increase the risk of epilepsy, but is not by itself causative. (7) A combination of risk-associated genes, together with environmental factors, determines whether the clinical phenotype of epilepsy occurs. In this model, individual genes that increase the susceptibility to epilepsy have a relatively weak impact. Multiple genetic defects, and/or particular combination of genes, probably increase the risk by a greater amount. However, it is not well understood how many abnormal genes are required to exceed the threshold to cause clinical epilepsy, nor is it understood which combination of genes may increase the risk more than others.
Early-onset epilepsy syndromes may be single-gene disorders. This hypothesis arises from the discovery of pathologic mutations in small numbers of patients with the disorders. Because of the small amount of research available, the evidence base for these rare syndromes is incomplete, and new mutations are currently being discovered frequently. (8)
Some of the most common genes that have been associated with both the common epilepsies and the rare epileptic syndromes are listed in Table 3.
Table 3. Selected Genes Most Commonly Associated With Genetic Epilepsy (adapted from Williams 2013) (2)
For the severe early epilepsy syndromes, the disorders most frequently reported to be associated with single-gene mutations include GEFS+ syndrome (associated with SCN1A, SCN1B, GABRG2 mutations), Dravet syndrome (associated with SCN1A mutations, possibly modified by SCN9A mutations), and epilepsy and intellectual disability limited to females (associated with PCDH19 mutations). Ohtahara syndrome has been associated with mutations in STXBP1 in cases where patients have no structural or metabolic abnormalities. West syndrome is often associated with chromosomal abnormalities or tuberous sclerosis, or may be secondary to an identifiable infectious or metabolic cause, but when there is not an underlying cause identified it is thought to be due to a multifactorial genetic predisposition. (9)
Pharmacogenomics of Epilepsy
Another area of interest for epilepsy is the pharmacogenomics of anti-epileptic medications. There are a wide variety of these medications, from numerous different classes. The choice of medications, and the combinations of medications for patients who require treatment with more than one agent, is complex. Approximately one-third of patients are considered refractory to medications, defined as inadequate control of symptoms with a single medication. (10) These patients often require escalating doses and/or combinations of different medications. At present, selection of agents is driven by the clinical phenotype of seizures, but has a large trial and error component in many refractory cases. The current focus of epilepsy pharmacogenomics is in identifying genetic markers that identify patients who are likely to be refractory to the most common medications. This may lead to directed treatment that will result in a more efficient process for medication selection, and potentially more effective control of symptoms.
Genetic Testing for Epilepsy
Commercial testing is available from numerous companies. Testing for individual genes is available for most, or all, or the genes listed in Table 3, as well as for a wider range of genes. Because of the large number of potential genes, panel testing is available from a number of genetic companies. These panels typically include large numbers of genes that have been implicated in diverse disorders.
No FDA-cleared genotyping tests were identified. The available commercial genetic tests for epilepsy are offered as laboratory-developed tests. Clinical laboratories may develop and validate tests in-house (“home-brew”) and market them as a laboratory service; such tests must meet the general regulatory standards of the Clinical Laboratory Improvement Act.
Medical policies are systematically developed guidelines that serve as a resource for Company staff when determining coverage for specific medical procedures, drugs or devices. Coverage for medical services is subject to the limits and conditions of the member benefit plan. Members and their providers should consult the member benefit booklet or contact a customer service representative to determine whether there are any benefit limitations applicable to this service or supply. This medical policy does not apply to Medicare Advantage.
The evaluation of a genetic test focuses on 3 main principles:
The genetic epilepsies will be discussed in two categories: The rare epileptic syndromes that may be caused by a single-gene mutation and the common epilepsy syndromes that are thought to have a multifactorial genetic basis.
Early-Onset Epilepsy Syndromes Associated with Single-Gene Mutations
There are numerous rare syndromes that have seizures as their primary symptom which generally present in infancy or early childhood and may be classified as epileptic encephalopathies. Many of them are thought to be caused by single-gene mutations. The published literature on these syndromes generally consists of small cohorts of patients treated in tertiary care centers, with descriptions of genetic mutations that are detected in affected individuals.
Table 4 lists some of these syndromes, with the putative causative genetic mutations.
Table 4. Early-Onset Epilepsy Syndromes Associated with Single-Gene Mutations
Other less commonly-reported single-gene mutations have been evaluated in childhood-onset epilepsies and in early onset epileptic encephalopathies, including ASAH1, FOLR1, GRIN2A, SCN8A, SYNGAP1, and SYNJ1 mutations in families with early-onset epileptic encephalopathies (11) and SLC13A5 mutations in families with pedigrees consistent with autosomal recessive epileptic encephalopathy. (12)
These syndromes can be evaluated by single-gene analysis, which is generally performed by direct sequencing. Direct sequencing is the criterion standard for identifying specific mutations. This testing method has an analytic validity of greater than 99%. They can also be evaluated by genetic panel testing, which is generally done by next-generation sequencing. This method has a lower analytic validity compared with direct sequencing, but is still considered to be very accurate, in the range of 95% to 99%.
The literature on the clinical validity of these rare syndromes is limited, and for most syndromes, the clinical sensitivity and specificity is not defined. Dravet syndrome is probably the most well-studied, and some evidence on the clinical validity of SCN1A mutations is available. The clinical sensitivity has been reported to be in the 70% to 80% range. (13,14) In one series of 64 patients, 51 (79%) were found to have SCN1A mutations. (14)
The false positive rate and the frequency of variants of uncertain significance, is not well characterized.
For the other syndromes, the associations of the genetic mutations with the syndromes have been reported in case reports or very small numbers of patients. Therefore, it is not possible to determine the clinical validity of the putative causative genetic mutations.
One potential area of clinical utility for genetic testing may be in making a definitive diagnosis and avoiding further testing. For most of these syndromes, the diagnosis is made by clinical criteria. However, there may be significant overlap across syndromes in terms of seizure types. It is not known how often genetic testing leads to a definitive diagnosis when the diagnosis cannot be made by clinical criteria.
Another potential area of clinical utility may be in directing pharmacologic treatment. For Dravet syndrome, the seizures are often refractory to common medications. Some experts have suggested that diagnosis of Dravet syndrome may therefore prompt more aggressive treatment, and/or avoidance of certain medications that are known to be less effective, such as carbamazepine. (14,15) In addition, some experts suggest that patients with Dravet syndrome may be more susceptible to particular antiepileptic agents, including clobazam and stiripentol. (6) However, there are no studies that examine the frequency with which genetic testing leads to changes in medication management, and there are no studies that report on whether the efficacy of treatment directed by genetic testing is superior to efficacy of treatment without genetic testing.
For the early-onset epilepsies that may have a genetic component, interventions to reduce the risk of having an affected offspring may be another potential area for clinical utility. Genetic counseling and consideration of pre-implantation genetic testing combined with in vitro fertilization are available options. For Dravet syndrome, most mutations are sporadic, making the clinical utility of testing for the purposes of counselling parents and intervening in future pregnancies low. However, when there is familial disease with a pathogenic mutation present in one parent, then pre-implantation genetic testing may reduce the likelihood of having an affected offspring. For other syndromes, the risk in subsequent pregnancies for families with one affected child may be higher, but the utility of genetic counseling is not well-established in the literature.
There are numerous rare epileptic syndromes which may be caused by single-gene mutations, but the evidence on genetic testing for these syndromes is insufficient to form conclusions on the clinical validity and clinical utility of genetic testing. The syndrome with the greatest amount of evidence is Dravet syndrome. The clinical sensitivity of testing patients with clinically defined Dravet syndrome is relatively high in small cohorts of patients. There may be clinical utility in avoiding further testing, and directing treatment, and for reproductive planning, but there is only a small amount of evidence to suggest this and no evidence demonstrating that outcomes are improved.
The common epilepsy syndromes, also known as idiopathic epilepsy, generally present in childhood, adolescence or early adulthood. They include generalized or focal in nature and may be convulsant (grand mal) or absence type. They are generally thought to have a multifactorial genetic component.
The common epilepsies are generally evaluated by genetic panel testing. The larger, commercially available panels that include many mutations are generally performed by next-generation sequencing. This method has a lower analytic validity compared with direct sequencing but is still considered to be very accurate, in the range of 95% to 99%. Less commonly, deletion/duplication analysis may be performed; this method is also considered to have an analytic validity of greater than 95%.
The literature on clinical validity includes many studies that report the association of various genetic variants with the common epilepsies. There are a large number of case-control studies that compare the frequency of genetic variants in patients with epilepsy with the frequency in patients without epilepsy. There is a smaller number of genome-wide association studies (GWAS) that evaluate the presence of single nucleotide polymorphisms (SNPs) associated with epilepsy across the entire genome. No studies were identified that reported the clinical sensitivity and specificity of genetic mutations in various clinically defined groups of patients with epilepsy. In addition to these studies on the association of genetic variants with the diagnosis of epilepsy, there are numerous other studies that evaluate the association of genetic variants with pharmacogenomics of anti-epileptic medications.
Diagnosis of Epilepsy
The Epilepsy Genetic Association Database (epiGAD) published an overview of genetic association studies in 2010. (16) This review identified 165 case-control studies published between 1985 and 2008. There were 133 studies that examined the association of 77 different genetic variants with the diagnosis of epilepsy. Approximately half of these studies (65/133) focused on patients with genetic generalized epilepsy. Most of these studies had relatively small sample sizes, with a median of 104 cases (range, 8-1,361) and 126 controls (range, 22-1,390). There were less than 200 case patients in 80% of the studies. Most of the studies did not show a statistically significant association. Using a cutoff of p less than 0.01 as the threshold for significance, there were 35 studies (21.2%) that reported a statistically significant association. According to standard definitions for genetic association, all of the associations were in the weak to moderate range, with no associations reported that were considered strong.
In 2014, the International League Against Epilepsy Consortium on Complex Epilepsies published a meta-analysis of GWAS studies for all epilepsy and two epilepsy clinical subtypes, genetic generalized epilepsy and focal epilepsy. (17) The authors combined GWAS data from 12 cohorts of patients with epilepsy and controls (ethnically matched to cases) from population-based datasets, for a total of 8696 cases and 26,157 controls. Cases with epilepsy were categorized as having genetic generalized epilepsy, focal epilepsy, or unclassified epilepsy. For all cases, loci at 2q24.3 (SCN1A) and 4p15.1 (PCDH7, which encodes a protocadherin molecule), were significantly associated with epilepsy (P=8.71 x 10-10 and 5.44 x 10-9, respectively). For those with genetic generalized epilepsy, a locus at 2p16.1 (VRK2 or FANCL) was significantly associated with epilepsy (P=9.99 x 10-9). No SNPs were significantly associated with focal epilepsy.
Some of the larger GWAS studies are described here. The EPICURE Consortium published one of the larger GWAS of genetic generalized epilepsy in 2012. (18) This study included 3,020 patients with genetic generalized epilepsy (GGE) and 3,954 control patients, all of European ancestry. A 2-stage approach was used, with a discovery phase and a replication phase, to evaluate a total of 4.56 million SNPs. In the discovery phase, 40 candidate SNPs were identified that exceeded the significance for the screening threshold (1 x 10-5), although none of these reached the threshold defined as statistically significant for GWA (1 x 10-8). After stage 2 analysis, there were 4 SNPs identified that had suggestive associations with GGE on genes SCN1A, CHRM3, ZEB2, and NLE2F1.
A second GWAS with a relative large sample size of Chinese patients was also published in 2012. (19) Using a similar 2-stage methodology, this study evaluated 1,087 patients with epilepsy and 3,444 matched controls. Two variants were determined to have the strongest association with epilepsy. One of these was on the CAMSAP1L1 gene and the second was on the GRIK2 gene. There were several other loci on genes that were suggestive of an association on genes that coded for neurotransmitters or other neuron function.
In contrast to the 2 studies previously described, a GWAS published from the UK failed to show any robust associations of SNPs with partial epilepsy. (20) This study included 3,445 patients with partial epilepsies and 6,935 controls of European ancestry. Using a threshold of an odds ratio (OR) greater than 1.3, the authors reported that no SNPs were identified that had a statistically significant association at that level.
In 2012, Heinzen et al. used whole exome sequencing to evaluate the association of genetic variants with genetic generalized epilepsy in 118 individuals with the disorder and 242 control patients of European origin. (21) No variants were found that reached the statistical threshold for a statistical association. From this initial data, the researchers selected 3,897 candidate genetic variants. These variants were tested in a replication sample of 878 individuals with GGE and 1,830 controls. None of the tested variants showed a statistically significant association.
In 2014, Baum et al. conducted a case-control study to evaluate the association of polymorphisms in SCN1A, SCN2A, SCN3A, SCN1B, and SCN2B genes and epilepsy. (22) The analysis included 1529 epilepsy patients and 1935 controls from 4 ethnicities or locations. The SCN1A IVS5N+5G>A polymorphism showed the strongest associated with epilepsy, with an OR of 0.85 for allele G (P<0.009) and 0.73 for genotype GG vs. AA (P=0.003). Other gene polymorphisms significantly associated with epilepsy included SCN1A (rs10188577, OR=1.20 for the C allele, P=0.003) and SCN2A (rs12467383, OR=1.16 for the A allele, P<0.009).
In a case-control study that included 441 epilepsy patients (240 with mesial temporal lobe epilepsy with hippocampal sclerosis, 201 with juvenile myoclonic epilepsy) and 267 non-epileptic controls, Balan et al. reported that the GABAA receptor subunit gene GABARG2 (rs211037) is significantly associated with both types of epilepsy (OR =1.6; 95% confidence interval [CI], 1.22 to 2.08; p<0.000). (23)
In addition to the individual studies, there are a number of meta-analyses that evaluate the association of particular genetic variants with different types of epilepsy. Most of these have not shown a significant association. For example, Cordoba et al. evaluated the association of SLC6A4 gene variants with temporal lobe epilepsy in a total of 991 case patients and 1,202 controls and failed to demonstrate a significant association on combined analysis. (24) Nurmohamed et al. performed a meta-analysis of 9 case-control studies that evaluated the association of the ABC1 gene polymorphisms with epilepsy. (25) There were a total of 2,454 patients with epilepsy and 1,542 control patients. No significant associations were found. One meta-analysis that did report a significant association was published by Kauffman et al. in 2008. (26) This study evaluated the association of variants in the IL1B gene with temporal lobe epilepsy and febrile seizures, using data from 13 studies of 1,866 patients with epilepsy and 1,930 controls. Combined analysis showed a significant relationship between one SNP (511T) and temporal lobe epilepsy, with a strength of association that was considered modest (OR=1.48; 95% CI: 1.1 to 2.0; p=0.01).
Pharmacogenomics of Antiepileptic Medications
Pharmacogenomic of Antiepileptic Drug Response
Numerous case-control studies report on the association of various genetic variants with response to medications in patients with epilepsy. The epiGAD database identified 32 case-control studies of 20 different genes and their association with medication treatment. (16) The most common comparison was between patients who were responders to medication and patients who were non-responders. Some of the larger representative studies are discussed next.
Kwan et al. compared the frequency of SNPs on the SCN1A, SCN2A, and SCN3A genes in 272 drug responsive patients and 199 drug resistant patients. (27) A total of 27 candidate SNPs were evaluated, selected from a large database of previously identified SNPs. There was one SNP identified on the SCN2A gene (rs2304016) that had a significant association with drug resistance (OR=2.1; 95% CI: 1.2 to 3.7; p<0.007).
Jang et al. compared the frequency of variants on the SCN1A, SCN1B, and SCN2B genes in 200 patients with drug resistant epilepsy and 200 patients with drug responsive epilepsy. (28) None of the individual variants tested showed a significant relationship with drug resistance. In further analysis of whether there were gene-gene interactions that were associated with drug resistance, the authors reported that there was a possible interaction of 2 variants, one on the SCN2A gene and the other on the SCN1B gene, that were of borderline statistical significance (p=0.055).
Other representative studies that report associations between genetic polymorphisms and antiepileptic drug response are summarized in table 5.
Table 5: Genetic Polymorphisms and Antiepileptic Drug Response
CI: confidence interval; OR: odds ratio.
Several meta-analyses evaluating pharmacogenomics were identified. Haerian et al. (33) examined the association between SNPs on the ABCB1 gene and drug resistance in 3,231 drug resistant patients and 3,524 controls from 22 studies. The authors reported no significant relationship between variants of this gene and drug resistance (combined OR=1.06; 95% CI: 0.98 to 1.14; p=0.12). There was also no significant association between on subgroup analysis by ethnicity.
In a separate meta-analysis, Sun et al. evaluated 8 studies evaluating the association between polymorphisms in the multidrug resistance 1 (MDR1) gene and childhood medication-refractory epilepsy, including 634 drug-resistant patients, 615 drug-responsive patients, and 1,052 healthy controls. (34) In pooled analysis, the MDR1 C3435T polymorphism was not significantly associated with risk of drug resistance.
The evidence related to the diagnostic yield of genetic testing in patients with epilepsy is limited. Ream et al. reported a retrospective review of a single center’s use of clinically available genetic tests in the management of pediatric drug-resistant epilepsy. (35) The study included 25 newly-evaluated patients with pediatric drug-resistant epilepsy who underwent genetic testing with one or more of the following: karyotype, chromosomal microarray, gene sequencing of specific single genes, gene sequencing using a panel, and/or whole exome sequencing (WES). Genetic testing was obtained based on the clinical judgment of treating providers due to the lack of an alternative nongenetic etiology and clinical suspicion for a genetic cause. Fourteen (56%) of tested patients had epileptic encephalopathies; 17 (68%) had generalized epilepsy syndromes. Of the 25 patients in the newly-evaluated group, 15 had positive findings on genetic testing (defined as a “potentially significant” result), while 10 of those were considered to be diagnostic (consisting of mutations previously described to be disease causing for epilepsy syndromes or variants predicted to be disease causing.) The yield of a diagnostic result for the various testing modalities were as follows: 1 of 7 tested (14.3%) for karyotype; 2 of 12 tested (16.7%) for microarray; 2 of 13 tested (15.4%) for targeted single-gene sequencing; 6 of 13 (46.2%) for epilepsy gene panel testing; 1 of 6 tested (16.7%) for WES (confirmatory of a variant detected on an epilepsy panel). The yield of genetic testing was higher in patients with epileptic encephalopathies (P=0.005) and generalized epilepsy (P=0.028). Patients with a clinical phenotype suggestive of an epilepsy syndrome were more likely to have positive results on testing: 2 of 2 patients with Dravet syndrome phenotype had pathologic mutations in SCN1A; 3 of 9 patients with Lennox-Gastaut syndrome had identified mutations (one with a CDKL5 mutation, one with an SCL9A6 mutation, and one with both SCN1A and EFHC1 mutations).
Pharmacogenomics of Antiepileptic Drug Adverse Effects
Many antiepileptic drugs have a relatively narrow therapeutic index, with the potential for dose-dependent or idiosyncratic adverse effects (AEs). Several studies have evaluated genetic predictors of AEs from antiepileptic drugs, particularly severe skin reactions including Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN).
Chung et al. evaluated genetic variants associated with phenytoin-induced severe cutaneous AEs (SJS/TEN, drug reactions with eosinophilia and systemic symptoms [DRESS]) and maculopapular exanthema. (36) The study entailed a GWAS study including 60 cases with phenytoin-related severe cutaneous AEs and 412 population controls, followed by a case-control study including 105 cases with phenytoin-related severe cutaneous AEs (61 with SJS/TEN, 44 with DRESS) 78 cases with maculopapular exanthema, 130 phenytoin-tolerant control participants, and 3,655 population controls from Taiwan, Japan, and Malaysia. In the GWAS analysis, a missense variant of CYP2C9*3 (rs1057910) was significantly associated with phenytoin-related severe cutaneous AEs (OR =12; 95% CI: 6.6 to 20; P=1.1 x 10-17). In a case-control comparison between the subgroups of 168 patients with phenytoin-related cutaneous AEs and 130 phenytoin-tolerant controls, CYP2C9*3 polymorphisms were significantly associated with SJS/TEN (OR =30; 95% CI: 8.4 to 109; P=1.2x10-19), DRESS (OR =19; 95% CI: 5.1 to 71; P=7.0x10-7), and maculopapular exanthema (OR =5.5; 95% CI: 1.5 to 21; P=0.01).
He et al. conducted a case-control study to evaluate the association between carbamazepine-induced SJS/TEN and 10 SNPs in the genes ABCB1, CYP3A4, EPHX1, FAS, SNC1A, MICA, and BAG6. (37) The study included 28 cases with carbamazepine-induced SJS/TEN and 200 carbamazepine-tolerant controls. The authors reported statistically significant differences in the allelic and genotypic frequencies of EPHX1 c.337T>C polymorphisms between patients with carbamazepine-induced SJS/TEN and carbamazepine-tolerant controls (p=0.011 and p=0.007, respectively). There were no significant differences between SJS/TEN cases and carbamazepine-tolerant controls for the remaining SNPs evaluated.
Wang et al. evaluated the association between HLA genes and cross-reactivity of cutaneous adverse drug reactions to aromatic antiepileptic drugs (carbamazepine, lamotrigine, oxcarbazepine, phenytoin, phenobarbital). (38) The study included 60 patients with a history of aromatic antiepileptic drug-induced cutaneous adverse drug reactions, including SJS/TEN and maculopapular eruption, who were re-exposed to an aromatic antiepileptic drug, 10 of whom had recurrence of the cutaneous adverse drug reaction on re-exposure (cross-reactive group). Subjects who were tolerant to re-exposure were more likely to carry the HLA-A*2402 allele than cross-reactive subjects (OR=0.13, 95% CI: 0.015 to 1.108; P=0.040). Frequency distributions for other HLA alleles testing were not significantly different between groups.
Prediction of Sudden Unexplained Death in Epilepsy
Sudden unexplained death in epilepsy (SUDEP) is defined as a sudden, unexpected, nontraumatic, and nondrowning death in patients with epilepsy, excluding documented status epilepticus, with no cause of death identified following comprehensive postmortem evaluation. It is the most common cause of epilepsy-related premature death, accounting for 15% to 20% of deaths in patients with epilepsy. (39) Given uncertainty related to the underlying causes of SUDEP, there has been interested in identifying genetic associations with SUDEP.
Bagnall et al. evaluated the prevalence of sequence variations in the PHOX2B gene in 68 patients with SUDEP. (39) Large polyalanine repeat expansions in the PHOX2B gene are associated congenital central hypoventilation syndrome, a potentially lethal autonomic dysfunction syndrome, but smaller PHOX2B expansions may be associated with nocturnal hypoventilation. In a cohort of patients with SUDEP, one patient was found to have a 15-nucleotide deletion in the PHOX2B gene, but no PHOX2B polyalanine repeat expansions were found.
There is a lack of evidence on the clinical utility of genetic testing for the common genetic epilepsies. Association studies are not sufficient evidence to determine whether genetic testing can improve the clinical diagnosis of GGE. There are no studies that report the accuracy in terms of sensitivity, specificity, or predictive value; therefore it is not possible to determine the impact of genetic testing on diagnostic decision making.
The evidence on pharmacogenomics suggests that genetic factors may play a role in the pharmacokinetics of anti-epileptic medications. However, this evidence does not provide guidance on how genetic information might be used to tailor medication management in ways that will improve efficacy, reduce AEs, or increase the efficiency of medication trials.
The evidence on genetic testing for the common epilepsies is characterized by a large number of studies that evaluate associations of many different genetic variants with the various categories of epilepsy. The evidence on clinical validity is not consistent in showing an association of any specific genetic mutation with any specific type of epilepsy. Where associations have been reported, they are not of strong magnitude, and in most cases, have not been replicated independently or through the available meta-analyses. Because of the lack of established clinical validity, the clinical utility of genetic testing for the common epilepsies is also lacking.
Ongoing and Unpublished Clinical Trials
A search of ClinicalTrials.gov in December 2014 identified a number of studies designed to identify novel genetic variants associated with early infantile epilepsies (NCT01357707), infantile spasms (NCT01723787), responders to a specific marijuana strain among patients with Dravet syndrome (NCT02229032), severe early onset epilepsies (NCT01858285), but No studies that evaluate outcomes for patients managed with epilepsy managed with genetic testing.
Clinical Input Received From Physician Specialty Societies and Academic Medical Centers
While the various physician specialty societies and academic medical centers may collaborate with and make recommendations during this process, through the provision of appropriate reviewers, input received does not represent an endorsement or position statement by the physician specialty societies or academic medical centers, unless otherwise noted.
In response to requests, input was received from 2 academic medical centers and 4 specialty societies, for a total of 8 reviewers while this policy was under review for 2015. The review was limited to input related to the use of genetic testing for infantile and early-childhood onset epileptic encephalopathies. There was consensus that genetic testing for early onset epileptic encephalopathies is medically necessary. Particular areas of clinical utility noted by reviewers included making specific treatment decisions in SCN1A-related epilepsies and avoiding other diagnostic tests and for reproductive planning for multiple types of early-onset epilepsies.
Summary of Evidence
Genetic testing for epilepsy covers a wide range of clinical syndromes and possible genetic defects. For infantile- or early-childhood-onset epilepsy syndromes, which may be caused by single-gene mutations, there is only a small body of research, which is insufficient to determine the clinical validity and clinical utility of genetic testing. There may be a potential role in differentiating these syndromes from the common epilepsies and from each other, and in improving the efficiency of the diagnostic work-up. There also may be a potential role for genetic testing in identifying syndromes that are resistant to particular medications, thereby directing treatment, and in reproductive decision-making for family members of affected individuals. At present, the evidence related to specific ways that genetic testing leads to improved outcomes is limited. Clinical input indicated strong support for the use of genetic testing in the evaluation of infantile- and early-childhood-onset epilepsy syndromes associated with encephalopathy. The clinical utility of genetic testing occurs primarily when there is a positive test for a known pathogenic mutation. The presence of a pathogenic mutation may lead to targeted medication management, avoidance of other diagnostic tests, and/or informed reproductive planning.
For the common epilepsies, which are thought to have a complex, multifactorial basis, the role of specific genetic mutations remains uncertain. Despite a large body of literature of associations between genetic variants and common epilepsies, the clinical validity of genetic testing is poorly understood. Published literature is characterized by weak and inconsistent associations, which have not been replicated independently or by meta-analyses. This literature does not permit conclusions on the clinical validity of genetic testing. Because of the lack of conclusions on clinical validity, conclusions on clinical utility are also lacking.
For epilepsy pharmacogenomics, there are numerous studies that evaluate the associations of genetic variants with medication response. This body of evidence also does not show consistent or strong relationships between genetic variants and response to medications. Therefore, the clinical utility of pharmacogenomics in epilepsy has not been demonstrated.
As a result of these limitations in the literature, genetic testing for epilepsy, other than for evaluation of infantile- or early-childhood-onset epileptic encephalopathies, is considered investigational.
Practice Guidelines and Position Statements
In 2010, the European Federation of Neurological Societies issued guidelines on the molecular diagnosis of channelopathies, epilepsies, migraine, stroke, and dementias.40 The guidelines made the following recommendations pertaining to epilepsy:
“There is good evidence to suggest that a thorough clinical and electrophysiological investigation may lead to the choice of the gene to be tested in patients with periodic paralysis (Level B). In myotonic disorders, it is recommended to first search for myotonic dystrophy and use clinical and electrophysiological phenotype characterization to guide for molecular genetic testing (Level B).
Molecular investigations are possible and may help in some cases to diagnose the condition but cannot be considered as a routine procedure with regard to the large number of different mutations in different genes. Furthermore, diagnosis can be made more easily by clinical and physiological investigation (Good Practice Point). One exception of note is the diagnosis of severe myoclonic epilepsy of infancy (SMEI), in which mutations are found in SCN1A in 80% of the patients (Level B).”
U.S. Preventive Services Task Force Recommendations
Medicare National Coverage
There is no national coverage determination (NCD) for genetic testing for epilepsy. In the absence of an NCD, coverage decisions are left to the discretion of local Medicare carriers.
MT-TK (mitochondrially encoded tRNA lysine) (e.g., myoclonic epilepsy with ragged-red fibers [MERRF]), common variants (e.g., m.8344A>G, m.8356T>C)
NHLRC1 (NHL repeat containing 1) (e.g., progressive myoclonus epilepsy), full gene sequence
ARX (aristaless related homeobox) (e.g., X-linked lissencephaly with ambiguous genitalia, X-linked mental retardation), full gene sequence
CHRNA4 (cholinergic receptor, nicotinic, alpha 4) (e.g., nocturnal frontal lobe epilepsy), full gene sequence CHRNB2 (cholinergic receptor, nicotinic, beta 2 [neuronal]) (e.g., nocturnal frontal lobe epilepsy), full gene sequence
ALDH7A1 (aldehyde dehydrogenase 7 family, member A1) (e.g., pyridoxine-dependent epilepsy), full gene sequence
SCN1A (sodium channel, voltage-gated, type 1, alpha subunit) (e.g., generalized epilepsy with epilepsy with febrile seizures), full gene sequence
Unlisted molecular pathology procedure
New Policy. Policy created with literature review through September 30, 2013. Genetic testing for epilepsy is considered investigational.
Update Related Policies. Remove 12.04.91.
Annual Review. Added Related Policy 12.04.92. Added the names of the commercially available tests to the Policy Guidelines. Policy updated with literature review through September 24, 2014. References 3, 5, 8, 10-11, 16, 21-22, 28-38 added. Policy statement unchanged.
Annual Review. Policy renumbered to 12.04.514; previously 12.04.109. Policy updated with results of clinical input. Policy statement added that genetic testing for early-onset epileptic encephalopathy syndromes may be considered medically necessary with conditions. Clarified policy statement and policy statement information. References 1 and 40 added. Removed 12.04.92 from Related Policies section.
Disclaimer: This medical policy is a guide in evaluating the medical necessity of a particular service or treatment. The Company adopts policies after careful review of published peer-reviewed scientific literature, national guidelines and local standards of practice. Since medical technology is constantly changing, the Company reserves the right to review and update policies as appropriate. Member contracts differ in their benefits. Always consult the member benefit booklet or contact a member service representative to determine coverage for a specific medical service or supply. CPT codes, descriptions and materials are copyrighted by the American Medical Association (AMA).