Abstract
Evidence that recurrent epileptic seizures may cause neuronal injury in some patients has been inferred from clinical observation, neuropsychological assessments, and neuroimaging studies. Cross-sectional investigations have yielded conflicting results and it is not possible to draw conclusions regarding causation, rather than merely association, from such designs. However, there is also evidence from in vivo biochemical studies that seizures may cause neuron injury. The heterogeneity of the epilepsies, epileptic seizures, co-morbidities, treatment regimens, and individual patient susceptibility all complicate the picture and inhibit the drawing of conclusions that are uniformly applicable. Longitudinal neuroimaging studies have the potential to objectively identify structural changes in the brain that are markers of neuronal injury. Such studies are a major undertaking, requiring age-matched control groups and consistent image acquisition and analysis techniques. One needs to analyze not only changes in group means but also the number of patients who show significant changes in imaging parameters that exceed the limits of test–retest reliability and changes in age-matched controls. Quantitative analysis of MRI T1-weighted volumetric datasets can reliably identify changes in cerebral and hippocampal volumes of 1–3% in individual subjects. The sensitivity of such quantitative analysis of structural data to identify functionally significant changes is not yet certain. Functional imaging techniques such as MR spectroscopy, PET, and SPECT may be more sensitive for detecting cerebral abnormalities, but their test–retest reliability is inferior. Other MRI tools, such as diffusion tensor imaging, may be useful for evaluating secondary cerebral damage after seizures, both acutely and chronically. Present evidence suggests that, to detect significant treatment effects, longitudinal studies of putative neuroprotective agents, using neuroimaging methods as a surrogate end point, would require at least a 3-year observation period, include large numbers of patients, and provide stratification for important clinical variables.
The inquiry “do seizures cause neuronal injury in humans?” leads to a range of questions. The epilepsies form a heterogeneous group of conditions in which overt seizures are only one manifestation. Apparently similar seizures may cause cerebral damage in the context of one form of epilepsy but not in another. Subclinical seizures and interictal epileptiform activity might also cause cerebral damage. Some medications might increase or decrease the risk for secondary neuron injury. In addition, genetic factors may predispose some individuals to be at greater risk than others.
It is desirable to be able to identify those patients who are at risk for secondary cerebral damage. If therapies can be designed to prevent or reduce these effects, methods to determine whether the treatments are effective will be crucial.
The assessment of neuropsychological functions can measure the consequences of cerebral damage. Serial EEG studies have not been shown to be a sensitive indicator in this regard. There is a need for sensitive in vivo assessment methods that are quantitative, reliable, reproducible, and safe. For repeated use over a period of years, they must be acceptable to patients and to a healthy control group. If large numbers of patients are to be evaluated, it is beneficial if the methods can be reliably applied in multiple sites.
Neuropathologic evidence for seizure-induced neuron injury.
Epilepsy surgery programs provide fresh material for neuropathologic study. This is most commonly from patients with medically refractory temporal lobe epilepsy, and includes hippocampus and temporal neocortex as well as neoplastic, dysplastic, and acquired lesions. The process of selection for surgical treatment generates material from a restricted group of patients, studied at one point in time, and it is therefore not possible to draw definite conclusions about the cause and effect of observed changes.
Severe hippocampal neuron loss is associated with a long history of epilepsy and secondarily generalized seizures.1-3⇓⇓ These studies have implied that damage develops over many years and is in part programmed cell death (apoptosis).4 The density and branching of dendritic spines are reduced.5 Sprouting of mossy fibers can be seen in the dentate gyrus in resected hippocampi. Increased expression of polysialylated neural cell adhesion molecule (PSA-NCAM) indicates axonal growth and plasticity in the dentate gyrus and the CA1 region of hippocampi and the entorhinal cortex.6,7⇓ There is controversy as to whether there is neurogenesis in sclerotic hippocampi.6,8⇓ Astrocytosis is commonly found in resected tissue.9 The presence of activated microglia suggests continuing injury, and increases in the CA1 and CA3 subregions of resected hippocampi imply a continuing process of damage.10 There are much fewer data on the effects and consequences of epileptic seizures and the epilepsies on the neocortex and cerebellum. The well-known association of cerebellar atrophy with chronic epilepsy has often been ascribed to phenytoin use, but this is not the only factor.11
The available evidence suggests that chronic epilepsy and recurrent seizures are associated with neuronal injury and reactive changes in human hippocampi. Anecdotally, however, there are examples of individuals who experienced frequent complex partial and generalized seizures but in whom both in vivo MRI and postmortem neuropathologic examination were remarkable for the normality of both hippocampus and neocortex, at least using qualitative assessment. This underlines the heterogeneity of individual susceptibility to neuronal damage from severe epilepsy, presumably from as yet undefined genetic factors.
Biochemical markers for seizure-induced neuron injury.
Neuron-specific enolase (NSE) is a marker for neuronal injury and can be measured in serum and CSF. There have been several reports of increased levels after spontaneous and precipitated seizures in some but not all patients. Generalized seizures have been associated with greater increases than have partial seizures.12-15⇓⇓⇓
Neuropsychological evidence for seizure-induced neuron injury.
Cognitive function and memory. It is generally accepted that episodes of generalized status epilepticus have a negative effect on cognitive function, particularly memory.16 There is limited literature regarding the progressive effects of TLE on cognition. A recent review of 25 series found inconsistent results, with some finding evidence for and some against progressive cognitive deterioration occurring over the course of TLE.17 There are several confounding variables in addition to seizure frequency and severity, including the duration of epilepsy, the underlying pathology, and AED treatment. A recent review considered the effects of epileptic seizures on cognitive function. Eight cross-sectional studies implied more powerful effects than did longitudinal studies, because the former were affected by co-morbidities such as the severity of the underlying pathology. Nine longitudinal studies showed a mild but definite relationship between the occurrence of generalized tonic–clonic, but not partial, seizures and cognitive impairment. These studies have often been limited by the absence of an appropriate age-matched control group and by too short an interval between assessments. It has been suggested that a 25-year interval would be needed to determine the effect of recurrent seizures on human cognitive function.16
Episodic memory is particularly impaired in mesial TLE and, apart from the fact that mesial functions appear affected in chronic nonmesial TLE, memory decline in TLE is similar to that seen in normal aging but occurs at an accelerated rate. Approximately one-third of such patients have memory decline. The degree of memory impairment in patients with TLE is affected by seizure number and severity, and by physical cerebral damage and initial functional capacity.18
Conduct and behavior.
Children with chronic epilepsy have higher rates of behavioral problems than the general population,19 and those with chronic neurologic conditions more commonly have behavioral problems than children with chronic non-neurologic conditions, such as asthma.20,21⇓ The behavioral consequences of seizures and epilepsy may have a more detrimental effect on a child’s function than the seizures themselves.
It is difficult to differentiate effects of seizures from other causal factors, such as the underlying cerebral pathology, development, and medication exposure, but the results are consistent with the thesis that repeated seizures contribute to disturbed behavior.22 Behavior improved over 3 months after epilepsy surgery in children,23 but the improvement could have been due to removal of abnormal brain tissue, improvement in seizures, or psychosocial factors.
In a 24-month follow-up study of 212 children with newly diagnosed epilepsy, those with recurrent seizures had worsened behavior scores, whereas those with no further seizures had improved scores.21 Again, it was not possible to differentiate between direct effects of epileptic seizures and the effects of neurologic dysfunction, because more frequent seizures occurred in those who had more severe neurologic dysfunction. Furthermore, it is possible that transient cognitive impairment caused by interictal epileptiform activity may have an adverse effect on conduct.24
Neuroimaging evidence for seizure-induced neuron injury.
Most attention has focused on the hippocampus for the following reasons. First, the techniques for making hippocampal measurements from MRI scans have been established for many years. Second, there is a clear association between hippocampal sclerosis and TLE. TLE is associated with cognitive impairment, particularly memory dysfunction. Finally, TLE is the most common form of refractory focal epilepsy, and therefore adequate numbers of patients are available for studies. Volumetric 3D T1-weighted MRI produces images with excellent anatomic definition. The technique of hippocampal volumetry using an interactive process has been established for more than a decade25,26⇓ and has been used to demonstrate the spectrum of severity of hippocampal sclerosis. Hippocampal atrophy is identified with hippocampal volume measures and correlates well with hippocampal neuron loss, particularly in the CA1 subregion.27 T2 relaxometry of the hippocampus is also well established28,29⇓ and correlates with the glia-to-neuron ratio27 and dentate gliosis.30
Cendes et al.31 did not find an inverse correlation between hippocampal volume and the occurrence of generalized seizures, the duration of the epilepsy, and the estimated seizure frequency. A similar study concluded that hippocampal atrophy was stable over the duration of TLE.32 In a contrasting result, the extent of hippocampal damage in patients with hippocampal sclerosis correlated with the estimated total lifetime number of secondarily generalized seizures.33
Other cross-sectional MRI studies have shown an association between severity of hippocampal damage and the estimated total seizure burden, seizure frequency, and duration of habitual epilepsy.34,35⇓ In patients with chronic drug-resistant epilepsy, there were 18% and 14% reductions of the left and right hippocampal volume, respectively, on the side ipsilateral to the seizure focus.34,35⇓ Prolongation of hippocampal T2 relaxation time correlated with the estimated total number of partial and generalized seizures. A linear correlation analysis has suggested that approximately 6,500 seizures were associated with a 50% hippocampal volume reduction.34
In refractory TLE, hippocampal volumes were inversely related to duration of epilepsy, ipsilateral to the epileptic focus but not contralaterally.36 Complex partial seizure frequency was not related, but patients with frequent secondarily generalized seizures had smaller ipsilateral hippocampi. In a population with complex partial and secondarily generalized seizures, duration of epilepsy was associated with smaller hippocampi, implying that, after an initial insult, recurrent seizures may be associated with progressive hippocampal damage.37 Refractory TLE may also be associated with atrophy beyond the hippocampus, including entorhinal cortex,38 amygdala, thalamus, and the caudate and cingulate gyrus.39-41⇓⇓
Other cross-sectional studies.
Cross-sectional MRI studies have also reported cerebellar11 and cerebral42 volume loss in association with refractory epilepsy. Cross-sectional studies, however, cannot determine whether structural damage was caused by an initial insult that gave rise to the epilepsy, by progression of an underlying pathology, or as the result of cerebral damage from recurrent seizures of varying type and severity. Further, co-morbidity and medication may confound the interpretation of data.
Longitudinal studies.
The question of whether chronic epilepsy results in smaller hippocampi, or whether a reduction in hippocampal volume determines intractability, can be addressed only by longitudinal studies.
There have been case reports of progressive hippocampal sclerosis in patients suffering from recurrent partial and secondarily generalized seizures43 and status epilepticus.44-46⇓⇓ Van Paesschen et al.47 reported significant reductions of hippocampal volume in 8% of 36 patients with partial seizures scanned 1 year apart, without significant group effects. The changes were considered to be either the result of frequent seizures or the resolution of edema after initial seizures. Recently, a decrease in hippocampal volume was noted over 3.5 years in 24 patients with TLE and correlated with the number of generalized seizures.48
In a prospective, community-based, longitudinal follow-up study, 68 patients with newly diagnosed seizures, 122 with chronic active epilepsy, and 90 age-matched control subjects had baseline MRI and follow-up scans 3.5 years later. Serial quantitative MRI measures used region-based hippocampal volumetry T2 relaxometry, automated measures of cerebral hemisphere gray and white matter, CSF and intracranial volumes, and semi-automated cerebellar volumes. These methods reliably detected neocortical, hippocampal, and cerebellar volume changes greater than 1.6%, 3.1%, and 3%, respectively, in individual subjects within the three groups.49,50⇓ Preliminary analysis suggested that, overall, rates of atrophy were not significantly different between the newly diagnosed chronic epilepsy and control groups. A history of prior neurologic insult was associated with an increased rate of cerebral and cerebellar atrophy. Significant atrophy of hippocampus, neocortex, or cerebellum occurred over the 3.5 years in 7% of newly diagnosed and 16% of chronic epilepsy patients compared with 3% of controls. The numbers of overt seizures and other clinical data had no evident effect on brain volumes.51
In vivo imaging studies have the potential to identify and quantify secondary cerebral damage caused by epilepsy before there is any clinical accompaniment, and to act as a surrogate end point for intervention and preventative strategies. Structural MRI, however, may be insensitive to relatively subtle neuron loss and injury.
Other imaging methods to evaluate neuron injury.
MR spectroscopy. MR spectroscopy (MRS) allows the evaluation of both the integrity and function of neurons by measuring N-acetyl aspartate (NAA), a normal byproduct of neuronal cellular metabolism. NAA is a marker of neuron cell dysfunction, not merely of volume loss. Abnormalities of metabolite profiles may be found in temporal lobes with normal MRI,52,53⇓ and bilateral abnormalities have been noted in up to 50% of patients with apparently unilateral structural abnormality,54 indicating that MRS may be more sensitive for detecting pathology.
The ability of MRS to detect abnormalities (83%) is similar to the ability of structural MRI to detect hippocampal volume loss, and when the two methods are combined the ability to detect abnormalities increased to 93%.55 The sensitivity of MRS to abnormality may be greater than that of MRI, but changes are nonspecific and the test–retest coefficient of reliability for the measurement of NAA using MRS is approximately 15–20%. The reliability is poorer for other metabolites.56 This will limit the utility of MRS in identifying significant changes in individual subjects.
MR spectroscopy has shown elevation of lactate during and for a few hours after complex partial seizures57 and NAA decreases after episodes of status epilepticus.58,59⇓ This method is potentially useful to assess the acute effects of seizures on neuron function and to evaluate the efficacy of neuroprotective strategies.
Radio-isotope imaging.
SPECT is widely available, with tracers that are sensitive to cerebral blood flow. However, the technique is only semi-quantifiable, with the use of an internal reference region, and radiation exposure is a concern (as it is with PET), and the test–retest coefficient of reliability is 15% at best.60 PET with [18F]-fluorodeoxyglucose can give parametric images of regional cerebral glucose utilization that is sensitive but nonspecific for a range of cerebral pathologies. Hypometabolism is a sensitive but nonspecific marker of cerebral dysfunction and is not only caused by neuron loss; there is an additional metabolic disturbance. Regional hypometabolism occurs in about 90% of patients with medial TLE.61 Focal or diffuse regional hypometabolism occurs in about 70% of patients with neocortical epilepsy.62 Limitations of PET for longitudinal studies include concerns about radiation exposure, lack of availability, high cost, the semi-invasive nature of the test, and a test–retest reliability of only 10% at best.63
Diffusion tensor imaging can be used to derive parametric images of the diffusion of water and of the fractional anisotropy of water diffusion in the brain, and can reveal abnormalities that are not evident on conventional MRI.64,65⇓ It is possible that this technique and other emerging methods, such as magnetization transfer ratio imaging, may be useful for identifying neuronal damage in longitudinal studies. Ictal diffusion-weighted imaging has shown reduced diffusivity at an epileptic focus, most likely due to cell swelling.66 Diffusion and perfusion imaging may also be useful to evaluate the acute effects of seizures,67 and therefore could potentially be useful to evaluate acute neuroprotective strategies.
Outstanding issues and future directions: how methods to assess seizure-induced neuron injury might be used in clinical trials.
It has been suggested that longitudinal neuropsychological studies of patients with epilepsy would need to be of 25 years’ duration to identify the adverse cognitive effects of the condition and treatment effects. Such studies would not be feasible for evaluation of a putative neuroprotective strategy. Longitudinal imaging studies have the potential to provide end points with a much shorter time span.
The objectives of longitudinal imaging studies are, first, to detect evidence of brain damage when it occurs, in the context of epilepsy. Patient groups will be heterogeneous in this regard and analysis must be not only of changes in group means but also of the number of patients who show significant changes in imaging parameters that exceed the limits of test–retest reliability and normal age-related changes. Second, in an interventional study, the objective will be to show that treatment is associated with absence of damage that occurs in a nontreated parallel group. Study populations will need to be stratified according to age, epilepsy and seizure types and, for partial seizures, the localization of seizure onset. Longitudinal studies must have adequate power to detect changes of clinical significance and must also be of sufficient duration to identify differences between healthy controls and patients and between active treatment groups and placebo-treated groups of patients.
The optimal duration of longitudinal studies of potentially neuroprotective AEDs requires careful consideration. A longer study increases the chance for cerebral changes and differences becoming evident, but is more costly and is likely to be more difficult, with subject dropouts and with variation of the imaging acquisition and analysis equipment and protocols.
MRI has a number of advantages that make it attractive as a method to evaluate the presence and development of cerebral damage in patients with epilepsy. MRI is more readily available and is comparatively less expensive than either PET or SPECT. In addition, its noninvasive nature and absence of ionizing radiation make it more acceptable to patients and controls participating in long-term studies.
Such serial quantitative MRI studies are being undertaken in Alzheimer’s disease, in which there is a reasonable correlation between the severity of atrophy and the degree of cognitive impairment. For example, the mean rate of brain atrophy for patients with Alzheimer’s disease was 2.37% per year, while in the control group it was 0.41% per year.68 Therefore, to have 90% power to detect a drug effect equivalent to a 20% reduction in the rate of atrophy, 207 patients would be needed in each treatment arm in a 1-year placebo-controlled trial with a 10% patient dropout rate, assuming that 10% of scan pairs were unusable.
Noninvasive methods to assess secondary neuronal damage and that can be used as a measure of treatment efficacy could provide a surrogate marker for the acute evaluation of neuroprotective agents. A potential criticism of serial volumetric studies of cerebral structure is that, although the results are reliable and reproducible, they may be insensitive to the initial development of secondary cerebral damage.
It is possible that acute postictal imaging of cerebral diffusion and perfusion may identify abnormalities that are associated with the development of secondary cerebral damage.
Acknowledgments
Acknowledgments
I am very grateful to my colleagues in the Department of Clinical and Experimental Epilepsy and for the support of the National Society for Epilepsy, Action Research, the Wellcome Trust, and the Medical Research Council.
Footnotes
-
Publication of this supplement was supported by an educational grant from GlaxoSmithKline.
References
Letters: Rapid online correspondence
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
- Article
- Abstract
- Neuropathologic evidence for seizure-induced neuron injury.
- Biochemical markers for seizure-induced neuron injury.
- Neuropsychological evidence for seizure-induced neuron injury.
- Neuroimaging evidence for seizure-induced neuron injury.
- Other imaging methods to evaluate neuron injury.
- Outstanding issues and future directions: how methods to assess seizure-induced neuron injury might be used in clinical trials.
- Acknowledgments
- Footnotes
- References
- Info & Disclosures
