If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The field of epilepsy biomarkers remains is in its infancy.
•
All the potential biomarkers discussed are preliminary work requiring validation.
•
The most promising genetic biomarkers are IL-1β rs1143634 and A1AR rs10920573.
Abstract
Introduction
Posttraumatic epilepsy (PTE) is caused by traumatic brain injury (TBI) and is an important contributor to the overall social and economic burden of epilepsy. Epidemiological studies suggest that there is a genetic contribution to the development of PTE. Identification of clinically useful genetic biomarkers is important for advancements in diagnosis and treatment of PTE.
Methods
A systematic review was performed on the existing literature of genetic biomarkers of posttraumatic epilepsy (PTE). A multi-database search yielded 4 articles deemed suitable for review. Potential genetic biomarkers were identified and critically evaluated.
Results & discussion
Biomarkers identified included single nucleotide polymorphism (SNP) rs1143634 of the interkeukin-1β (IL-1β) gene, SNPs rs3828275, rs3791878, and rs769391 of the glutamic acid decarboxylase 1 (GAD1) gene, SNPs rs3766553 and rs10920573 of the adenosine A1 receptor (A1AR) gene, and the functional variant C677T of the methylenetetrahydrofolate reductase (MTHFR) enzyme. The most promising biomarkers identified were IL-1β rs1143634 and A1AR rs10920573. Both had heterogenous at risk genotypes (CT). Those with IL-1β rs1143634 CT genotype developed PTE in 47.7% of cases (p = 0.008) and those with A1AR rs10920573 CT genotype developed PTE in 19.2% of cases (p = 0.022).
Conclusion
The majority of articles were preliminary with a need for validation of results. There is a need for continued high calibre research in order to validate the currently identified genetic biomarkers as well as to discover new genetic biomarkers in PTE.
Traumatic brain injury (TBI) is a complex and heterogeneous condition that results directly from an external mechanical force to the head. Causes of TBI vary considerably, ranging from motor accidents to gunshot wounds and explosions. TBI is associated with a broad spectrum of lifelong disorders, such as epilepsy, and is cited as a leading cause of death and disability worldwide, placing a considerable economic and social burden on society [
]. PTE is a heterogeneous condition that can vary tremendously in time from injury to onset, from weeks to years. Three classifications of post-traumatic seizures exist : immediate seizures occurring less than 24 h post-injury, early seizures occurring less than 1 week post-injury, and late seizures occurring more than 8 days post-injury [
Acquired epilepsies develop in a relatively predictable pattern that includes three phases. The initial brain insult will usually lead to epileptogenesis during the latency period, defined as the seizure-free period directly following the cerebral insult, which in turn results in recurring and unprovoked epileptic seizures [
]. The time from injury to seizure onset is very important in determining prognosis and treatment. Immediate and early post-traumatic seizures (PTS) appear to be low risk in terms of seizure recurrence because they are usually attributed to cortical inhibition resulting directly from the injury as opposed to epileptic events, whereas late seizures are more highly correlated with the development of PTE [
]. Most individuals who experience a single posttraumatic seizure are started on anti-epileptic medications following the first seizure event, although there is no evidence this prophylactic treatment works to prevent the subsequent development of PTE [
PTE occurs in both children and adults, particularly children under 7 years of age and adults over 65 years of age. In the general population, PTEs account for 5% of all epilepsy cases and 20% of symptomatic epilepsy [
]. This number is understandably higher in the military population where almost 50% of TBI victims, more specifically those who sustain penetrating head injuries, go on to develop PTE [
]. Many risk factors have been identified, including but not limited to severity of initial TBI, length of posttraumatic amnesia, presence of intracranial haemorrhage, penetrating head injuries, depressed skull fractures, retention of metal fragments, location of lesion(s), cerebral contusions, age, and comorbid chronic alcoholism [
Mechanisms of posttraumatic epileptogenesis are incompletely understood. Studies suggest that hippocampal and dentate gyrus involvement and early and selective cell loss in the CA3 (Cornu Ammonis sub-region 3) of the hippocampus and in the hilus of the dentate gyrus are key to the epileptogenic process [
]. In addition, significant hippocampal mossy fibre sprouting as well as neurodegeneration, axonal injury, astrocytosis, and dysfunction of the blood-brain barrier appear to play a role [
Biomarkers that may be clinically useful in the prediction, diagnosis, prognosis, and treatment of posttraumatic epilepsy are currently under-investigated. Methods employed to discover clinically relevant biomarkers include imaging and diagnostic methods as well as transcriptional profiling, proteomic, genomic, and metabolomic approaches [
]. Consequently, all that can be offered is symptomatic management once a pattern of unprovoked seizures is established. Biomarkers offer a potential way of predicting the onset of PTE, a better understanding of the molecular mechanisms associated, and identification of possible preventative measures. The purpose of this review is to investigate and synthesise the literature on the subject of potential genetic biomarkers of PTE with the aim to evaluate the strength of evidence for individual biomarkers in addition to highlighting areas of potential for advancement or areas where evidence is currently lacking.
2. Methods
A multi-database search was performed using the following search terms: epilep* (epilepsy, epileptogenesis, epilepsies), seizure, convulsions, posttraumatic epilep*, traumatic brain injur* (injury, injuries), acquired brain injur*, and genetic biomarkers. Databases utilised were PubMed, Science Direct, Cochrane Library, Web of Science, and Google Scholar. The references of each relevant review were manually screened to ensure no important articles were missed (n = 8). Exclusion criteria were defined as follows: papers addressing general epileptogenesis not specifically related to TBI, papers addressing the pathophysiology of epileptogenesis as opposed to biomarkers, papers addressing treatment of epilepsy, animal studies, and reviews. The database search was not limited by language or time period, but only English studies from the years 1980–2015 were included in this review due to the constraints of the exclusion criteria. This review adhered to PRIMSA guidelines. This review was originally part of a much larger and broader body of work, a Master’s dissertation detailing the CSF/serum, genetic, imaging, neurophysiologic, and clinical biomarkers of PTE, reflected by Fig. 1.
Many factors are at play in individuals who develop PTE after a TBI versus those who do not, including but not limited to predisposing genetic traits. With the completion of the Human Genome Project (HGR) in 2003 as well as recent improvements in genotyping technology, the investigation of genetic factors and their role in the development of disease has become progressively more accessible to researchers, resulting in the increase of genetic association studies and their reliability. In the field of PTE, most genetic associations examined have been single-nucleotide polymorphisms (SNPs), a common variation within a population that can be defined as a variation of a single nucleotide, adenine (A), guanine (G), thymine (T), or cytosine (C), between individuals of the same species. Within each SNP there is a protective and an at-risk genotype for PTE development (Table 1).
Table 1Summary of genetic biomarkers as reported in the current literature.
Biomarker
Reference
Key Findings
N
Main Strengths
Main Limitations
rs1143634
Diamond et al., 2014
•
serum IL-1β significantly different based on rs1143634 genotype
•
CSF/serum IL-1β ratio showed a trend based on rs1143634 genotype
•
heterozygous (CT) genotype associated with PTE, lower IL-1β serum levels, and higher CSF/serum IL-1β ratio
256
-
only late-onset PTS were included, increasing the likelihood that an epileptogenic process was taking place as opposed to investigating non-epileptic seizures
-
small sample size
GAD1 gene rs3828275 rs769391 rs3791878
Darrah et al., 2013
•
rs3828275: homozygous wild-type (CC) is defined as the protected genotype and has significantly higher odds of not developing PTS <1 week
•
rs3791878: homozygous wild-type (GG) is defined as the at-risk genotype and has significant associations with PTS occurring 1 week – 6 months
•
rs769391: homozygous wild-type (AA) is defined as the at-risk genotype but is not significantly associated with PTS occurring 1 week – 6 months
257
-
well-controlled
-
clearly defined variables
-
seizure assessment exclusion criteria, time-specific groupings, and outcome measures are such that incidence of PTE is, if anything, understated
-
small sample size
MTHFR C677T
Scher et al., 2011
•
TT is defined as the at-risk genotype and is significantly associated with PTE compared to the CC group
800 (experi-mental) 800 (control)
-
well-designed with sufficient controls and variable parameters
-
samples from all subjects were genotyped twice
-
sample population was randomly selected, relatively large, and multi-ethnic
-
sample size is technically too small
-
exclusion of medical encounters while subjects were deployed to combat zones (authors cite inability to access records)
-
possible additional misclassification of seizure status due to use of the ICD-9-CM diagnostic codes without further verification from a clinician (however, this is likely to have weakened rather than strengthened the results)
Adenosine A1 receptor gene: rs3766553 rs10920573
Wagner et al., 2010
•
rs3766553: significantly associated with early, late, and delayed onset PTS – AA genotype associated with early PTS and GG genotype associated with both late and delayed onset PTS
•
rs10920573: CT heterozygous genotype significantly associated with late and delayed onset PTS
•
risk variants for both SNPs were independently associated with late and delayed onset PTS and had a cumulative effect on susceptibility to PTS
206
-
only used subjects who had sustained a severe TBI (GCS of 8 or less), increasing the likelihood of PTE development as severity of TBI is a known clinical predictor
-
all subjects were monitored in a controlled hospital environment, reducing the possibility of confounding factors
-
appropriately controlled DNA collection, genotyping methods, outcome measures, and statistical analysis
-
small sample size
-
penetrating head injury is listed as an exclusion criteria when it is a risk factor for PTE
CSF – cerebrospinal fluid; GCS – Glasgow Coma Scale; ICD-9–International Statistical Classification of Diseases and Related Health Problems Version 9; PTE – post-traumatic epilepsy; PTS – post-traumatic seizures; TBI – Traumatic Brain Injury.
Interkeukin-1β (IL-1β) is a pro-inflammatory cytokine that is released in the central nervous system (CNS) by activated astrocytes and microglia and in the periphery by macrophages and other immune cells as a response to injury or other pathology in the brain [
]. The first of its kind, a study examined the potential link between IL-1β and PTE on the basis of evidence supporting IL-1β involvement in both TBI and other kinds of epilepsies [
]. This study investigated genetic variations in the gene coding for IL-1β, located in the region of 2q12-13, and their possible role in the prediction of PTE onset in moderate-to-severe TBI [
]. Only late PTS (onset one or more weeks post-injury) were included. Of the SNPs investigated, rs1143634 was the only one with a significant association with the eventual development of PTE (Table 2). In addition to being significantly associated with the development of PTE, it seems that rs1143634 also has effects on CSF and serum levels of IL-1β. Serum levels of IL-1β are diminished in the CT genotype (the at-risk genotype), a finding that aligns with decreased levels observed in PTE patients compared to controls [
The results of this study implicate CSF and serum IL-1β levels as well as the SNP rs1143634 in the development of PTE. A significant association between IL-1β CSF and serum levels and rs1143436 is also apparent. As the functionality of the rs1143634 genotype of the IL-1β is unknown, the exact reasons why the CT genotype would be considered the at-risk group while the TT genotype is suggested as protective has not been determined.
3.2 GAD1 gene: rs3828275, rs3791878, & rs769391
Many of the gene association studies in the field of PTE are the first to identify their respective SNPs with the occurrence of post-traumatic seizures. One gene in particular that seems to contribute to PTS in general is the glutamic acid decarboxylase gene (GAD), more specifically GAD1 [
]. One study examined the previously unexplored link between GAD1 and PTE based on multiple studies that confirm its involvement via the GABAergic neurotransmission pathway in TBI pathophysiology and seizure susceptibility individually [
Kinetics of glutamate and gamma-aminobutyric acid in cerebrospinal fluid in a canine model of complex partial status epilepticus induced by kainic acid.
]. GAD1 SNPs were found to be significantly associated with PTS. One of the three SNPs discussed, rs3828275, was correlated with early PTS only, that is seizures occurring within one week of the TBI (Table 2). It was further found that when TT was defined as the at-risk genotype, the CC genotype had a statistically significant chance of not developing PTS when compared to the other two genotypes (p = 0.055) [
]. With regard to this particular SNP, it is important to note that seizures occurring within the first week of injury may be a direct result of the injury as opposed to being epileptic in nature.
The other two SNPs investigated, rs3791878 and rs769391 in the GAD gene, were both significantly associated with PTS within the timeframe of one week to 6 months (Table 2). Although this association was not significant on multivariate analysis comparing all three genotypes, there was a significant difference between the AA genotype compared to either the AG genotype or the GG genotype. Through additional statistical analysis, it was shown that for rs3791878, people with the GG genotype had significantly higher odds of developing PTE when compared to GT and TT genotypes (p = 0.0231) [
]. The same was true of the rs769391 SNP, with the AA genotype having higher chances of developing PTE when compared to the other two genotypes, demonstrating an insignificant trend (p = 0.058). Due to the apparent involvement of rs3791878 and rs769391 in PTS occurring within the timeframe of 1 week to 6 months as well as the common gene on which they are located (GAD1), the authors performed a multivariate analysis between the two at-risk genotypes, GG (rs3791878) and AA (rs769391). Results from this analysis proved significant when comparing individuals with zero, one, or both risk variants [
]. Patients with both risk variants (GG and AA) had significantly higher rates of PTS when compared to those with one or none of the risk variants (p = 0.019). Moreover multivariate analysis showed that patients with both risk variants were at increased risk of developing PTS compared to those without any risk variants (p = 0.024) [
]. Interestingly, no notable differences were uncovered in the multivariate analysis of both risk variants compared to one risk variant or of no risk variants compared to one risk variant [
A third gene to be investigated in the search for PTE biomarkers is the adenosine A1 receptor (A1AR) gene and its variants. The A1AR gene is suspected to be involved in the development of PTE due to the location of A1ARs in regions associated with both TBI and seizure pathology, namely the hippocampus and the cortex, and their proximity to N-methyl-D-aspartate (NMDA) receptors [
]. Not only has adenosine been demonstrated to be an important neuroprotective agent in TBI, but compromises in adenosine synthesis due to secondary TBI pathology such as glial scars has been linked to potential posttraumatic seizure development [
]. Based on this evidence, a gene association study was performed to determine if genetic variability within this gene could account for an increased susceptibility to PTE in people with TBI. Of the SNPs analysed, two were identified as having significant associations with the development of PTS, both early and late [
]. Notably, different genotypes of rs3766553 were associated with early (less than one week post injury) PTS as well as late (one or more weeks post-injury) and delayed-onset (6 or more months post-injury) PTS (Table 2).
The second SNP implicated in the development of late and delayed-onset PTS is rs10920573, where the heterozygous (CT) genotype was defined as the at-risk group compared to the two homozygous genotypes (wild-type (CC) and variant (TT)). As both rs3766553 and rs10920573 are either significantly correlated with or are trending towards a significant correlation with late and/or delayed-onset seizures, additional statistical analysis was performed to determine the risk associated with carrying none, one, or both of the risk genotypes (GG (rs3766553) and CT (rs10920573)). Results indicate significant associations with carrying one or both of the risk genotypes and the development of late or delayed-onset PTS [
The fact that two different genotypes of the rs3766553 SNP correlate with two different kinds of PTS is noteworthy. This kind of association within a single SNP has certain implications for neuroprotection as well as susceptibility after TBI. It would seem that the rs3766553 heterozygotes are relatively protected against the development of seizures post-TBI. Conversely, wild-type homozygotes (AA) are susceptible to early PTS but apparently not PTE. Variant homozygotes (GG) seem to be the group at highest risk for developing PTE, as they are statistically more likely than AA or AG genotypes to experience late or delayed-onset seizures post-TBI. When examining the second implicated SNP, rs10920573, it is the heterozygotes that are at greatest risk, indicating that neither allele is protective on its own.
In this study, one of the exclusion criteria was penetrating head injury [
]. Aside from this oversight, the A1AR gene seems to show promising links to PTS, early, late, and delayed-onset, and thus could have implications for individual susceptibility to PTE after TBI.
3.4 MTHFR C677T
Methylenetetrahydrofolate reductase (MTHFR) is an enzyme key to the metabolic processes of methionine, an essential amino acid, and is encoded by the MTHFR gene, located on chromosome 1p36.3 [
]. Recent studies suggest that the C677T genotype, one of two common functional variants, is associated with the onset of PTE. Homocysteine is a sulfur-containing amino acid and can result from compromised methionine metabolism [
]. Additionally, there is a suggestion that the TT genotype of the MTHFR C677T gene variant may be disproportionately represented in individuals with epilepsy, although no study has investigated this phenomenon in relation to PTE [
This study investigated the presence of this variant in an age-matched case-control military study (n = 1600). In this study, the authors were investigating epilepsy prevalence and medical encounters suggestive of TBI via anonymous military medical databases. Results of this study suggest that, regardless of genotype, epilepsy was more prevalent in those who had medical encounters that were indicative of TBI (17%) compared to controls (4%) [
]. When the authors genotyped subjects for MTHFR C677T, it was shown that the homozygous wild-type (CC) was relatively protective while the homozygous variant (TT) was defined as the at-risk group [
]. The chances of being diagnosed with any kind of epilepsy, including PTE, was significantly higher in the TT genotype than in the CC genotype, a result that was enhanced when the analysed cases were limited to those cases who had two or more medical encounters for epilepsy [
This study was well-designed and included sufficient controls and variable parameters, such as a randomly selected, relatively large, and multi-ethnic sample population. Due to the lack of clinical verification of seizure assessment inherent in a study using an anonymous database, researchers opted to restrict epilepsy cases to people with two or more medical encounters for epilepsy. Samples from all subjects involved in the study were genotyped twice, with inconsistent results excluded. The exclusion of medical encounters while subjects were deployed to a combat zone is an obvious limitation, due to the fact that subjects would have been at higher risk for TBI (especially penetrating TBI) whilst in a combat zone. The authors cite inability to access these records as the reason for having excluded this data [
]. This limitation is likely to have resulted in an underestimation of PTE cases, or the incorrect classification of PTE cases as non-traumatic epilepsy cases and therefore a potential source of selection bias.
4. Conclusions
From the results of these genetic association studies, it is reasonable to conclude that there is a genetic element to the development of PTE, however understudied. Most promising seems to be the IL-1β SNP rs1143634, as the individuals with TT genotype seemed to be completely protected from developing PTE (Table 2). However, the functionality of the T allele is not fully understood, as it is also part of the at-risk genotype (CT). Similarly, the A1AR SNP rs10920573 seems to be relatively promising and is comparable to the IL-1β SNP rs1143634, with the at-risk genotype being the heterogeneous one (CT). The other genes and their respective SNPs discussed in this review were more powerfully associated with PTS, defined as early seizures that cannot be definitively categorised as PTE.
The subject of genetic biomarkers for PTE is vastly understudied. The majority of potential biomarkers identified have been examined in, on average, one to three primary studies.
The studies that do exist are heterogeneous in their discussion of PTE versus PTS. As previously mentioned, PTS can occur as a direct result of the TBI and are not necessarily epileptogenic while PTE is an epileptogenic process. A single occurrence of a seizure after a traumatic event does not constitute a diagnosis of epilepsy, as there are many seizure-inducing factors that can cause provoked seizures in individuals who do not have epilepsy, or who will not go on to develop epilepsy (PTE). Epilepsy is operationally defined as the occurrence of two or more unprovoked seizures following a cascade of neurochemical changes in the brain, wherein seizures are a symptom of the disorder. The distinction between these two categories of seizures is significant, as it has direct implications for the diagnosis, prognosis and management of the seizures.
Additionally, genetic association studies require very large sample sizes, in the order of 10,000 patients or more to have real significance, which requires collaboration between multiple groups [
]. This general lack of research could be a result of the relatively unknown nature of PTE pathophysiology. However, PTE is a well-recognised clinical consequence of TBI, and as such should be treated with priority due to the projected economic and social burden that may be produced from TBI and PTE.
Conflict of interest
None.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
References
Gupta P.K.
Sayed N.
Ding K.
Agostini M.A.
Van Ness P.C.
Yablon S.
et al.
Subtypes of post-traumatic epilepsy: clinical, electrophysiological, and imaging features.
Kinetics of glutamate and gamma-aminobutyric acid in cerebrospinal fluid in a canine model of complex partial status epilepticus induced by kainic acid.
30097-3&id=gr1.jpg){kind=link}