- Top of page
- MATERIALS AND METHODS
Summary: Purpose: Hippocampal sclerosis (HS) is the most common lesion underlying drug-resistant temporal lobe epilepsy. Whether HS is a developmental or acquired pathology remains unclear. Whereas HS has been causally linked to prolonged febrile convulsions in childhood, evidence also exists that it may coexist with extrahippocampal abnormalities, the concept of “dual pathology.” The aims of this study were to address whether hippocampal abnormality consistent with HS (a) occurs in children with lesional extrahippocampal epilepsy, (b) is more commonly seen in association with developmental rather than acquired extrahippocampal pathologies, and (c) whether any effect of age at seizure onset is found on the occurrence of HS in lesional extrahippocampal epilepsy.
Methods: Clinical and histopathologic data of patients having resective surgery for extrahippocampal epilepsy that included the hippocampus were investigated.
Results: Twenty-nine children were retrospectively included in this study, and 21 (72%) of 29 were found to have a hippocampal abnormality consistent with HS. No relation was noted between developmental or acquired extrahippocampal pathologies and the presence of hippocampal abnormality. Children with normal hippocampi on visual histologic assessment had a significantly younger age at seizure onset (p < 0.001). Duration of epilepsy was not correlated with the presence of hippocampal abnormality.
Conclusions: Hippocampal abnormalities are seen in similar proportions with both acquired and developmental extra-hippocampal pathologies, suggesting that these abnormalities are the result of seizures from the focus that is remote from the hippocampus. In addition, children who have their initial seizure at an early age are less likely to develop seizure-induced hippocampal injury.
The role of hippocampal sclerosis (HS) in epilepsy has been recognized for more than a hundred years. It is associated with intractable temporal lobe epilepsy and is the most common pathology identified in those patients who require surgical intervention. Sommer (1) first described the histologic features of HS in 1880 in a 25-year-old patient. He recognized a well-defined pattern of cell loss in the pyramidal layer of the hippocampus, with greater neuronal loss in CA1 and the prosubiculum (subsequently called Sommer's sector) than in the end folium (CA3 and CA4) or in the dentate gyrus. Falconer and Taylor (2) later used the term mesial temporal sclerosis, recognising that changes were frequently seen in the amygdala, uncus, and temporal lobe in association with HS. Since then, debate has occurred over whether HS is a developmental or an acquired lesion.
It is clear that HS can be acquired in animal models of limbic status epilepticus. Some data show evidence of hippocampal injury that matures into structural abnormalities similar to human mesial temporal sclerosis (3–5). Many of these animals will develop spontaneous recurrent seizures similar to human temporal lobe epilepsy. In humans, the most recognized association with HS is a prolonged febrile convulsion in childhood, and evidence for HS being acquired is accumulating in humans (2,6). In addition, magnetic resonance imaging studies have shown evidence for acute hippocampal injury associated with prolonged febrile convulsions (7–9), suggesting that a prolonged febrile convulsion could cause acute hippocampal injury leading to the subsequent development of HS.
It is likely, however, that more than one type of HS exists (10–12). HS, as described by Sommer, may represent the histologic appearance of the hippocampus at one point in a spectrum of damage that may range from severe neuronal loss in all hippocampal subzones to a minimal lesion in which loss of neurons is restricted to the end folium (13). In addition to HS recognized in children with a history of prolonged febrile convulsion, abnormality consistent with HS may be seen in patients with other lesions, including developmental anomalies in the hippocampus, the temporal lobe, or more extensively within the brain, described as dual pathology (12,14–16). The exact relations between these developmental anomalies and HS also are uncertain. It is possible that both lesions independently predate the onset of epilepsy. It is also possible that subtle areas of dysgenesis within the hippocampus itself may provide the substrate for a prolonged seizure and the subsequent development of HS (14). In pediatric surgical series, few data clarify the relations between developmental anomalies and HS. Subtle magnetic resonance imaging abnormalities, consistent with seizure-induced injury, have been identified in hippocampi that are not the primary seizure focus (9), and histologic data that show that extrahippocampal childhood seizures are associated with neuronal loss and mossy fibre sprouting within the hippocampus (17). The finding of hippocampal abnormality with postnatally acquired ischemic and encephalitic brain insults, as well as with cortical dysplasia, may be evidence against HS preexisting as a developmental lesion.
In animal models of epilepsy, susceptibility to seizure-induced hippocampal damage has been shown to be less in the immature animal than in the more mature animal (18–22). Severe seizures in the immature animal may occur without histologic evidence of hippocampal damage. Given these data, if HS is acquired after the onset of extrahippocampal seizures in humans, one might expect that it would be less common in children with early onset of seizures, particularly as the presence of HS has not consistently been shown to be related to the duration of epilepsy (17,23). Children with extrahippocampal pathologies and epilepsy, who undergo multilobar resection or hemispherectomy in which tissue from the hippocampus is obtained, provide a group in which some of the important questions regarding the relations between HS, coexisting developmental malformations, and age at onset of epilepsy may be addressed.
The aims of this study were therefore
To determine how frequently histologic evidence of HS is seen in children with intractable epilepsy in the context of major extrahippocampal pathologies requiring surgery.
To identify whether the frequency with which HS occurs is different in children with developmental extrahippocampal pathologies when compared with those with acquired extrahippocampal pathologies.
To investigate whether age at seizure onset influences the occurrence of HS in children with lesional extrahippocampal epilepsy.
MATERIALS AND METHODS
- Top of page
- MATERIALS AND METHODS
Clinical and pathologic data of patients having resective brain surgery, including the hippocampus, for lesional extrahippocampal epilepsies were reviewed. All children had been evaluated in the Epilepsy Surgery Programme at Great Ormond Street Hospital for Children NHS Trust, London, U.K. Routine preoperative evaluation included clinical review, documentation of seizures with video-EEG telemetry recording, ictal and interictal single-photon emission computed tomography (SPECT), magnetic resonance imaging, as well as neuropsychiatric and neuropsychological assessments. The decision to proceed to surgery was made after multidisciplinary discussion. Patient notes were reviewed for age at seizure onset, seizure semiology, duration of seizures, age at surgery, and history of febrile convulsion or status epilepticus. Seizures were classified as temporal or extratemporal in origin, based on clinical and EEG data. Clinical localization to the temporal lobe was accepted if any one of the following was present: behavioral arrest, oroalimentary automatisms, stereotyped complex automatisms with postictal confusion, psychic aura such as fear or déjà vu, an aura of a formed auditory or visual hallucination, a distinct epigastric aura, speech impairment before or immediately after the seizure, or nonspecific aura or automatisms followed by postictal confusion (24–26). Localization to the frontal lobe was considered if the seizures were brief, rapid onset or offset to the seizure occurred, nocturnal bias, bizarre behaviour or vocalization, and predilection to clustering (27–29). An occipital origin was suggested if elemental visual hallucinations, contralateral eye deviation, or ictal blindness was found (27). Laterality was based on lateral motor phenomena (e.g., limb jerking or dystonia) or lack of speech disturbance during or speech disturbance immediately after the seizure (30,31). Corroborative EEG data were available in all patients.
At surgery, the temporal lobe structures including the hippocampus were resected en bloc, providing good-quality specimens for histologic analysis. Resected tissue from the extrahippocampal region and the hippocampus was rapidly fixed in 10% formalin and processed for paraffin wax histology. Sections were stained routinely with hematoxylin-eosin, hematoxylin van Gieson, luxol fast blue-cresyl violet, and for immunocytochemistry with a variety of antibodies including glial fibrillary acidic protein (GFAP), neurofilaments, and synaptophysin. The histopathology of the extrahippocampal lesion and hippocampus in each patient was visually assessed by a single observer who was unaware of the clinical data (B.H.), and cases were divided into two groups: those with abnormal hippocampal histology (ranging in severity from histologic evidence of neuronal loss in all hippocampal subzones at the more severe end of the spectrum to the finding of end-plate gliosis at the milder end of the spectrum of HS) and those with hippocampi that were histologically normal on visual assessment (13).
SPSS for Windows (version 11, Chicago, IL, U.S.A.) was used for the analysis. χ2 analysis was used to investigate the relation between type of pathology (i.e., developmental or acquired) and HS. To address whether the age at seizure onset was lower in patients with normal hippocampi compared with patients with HS, the data were transformed and analysed by using an independent samples t test (not assuming equal variances). The relation between duration of epilepsy and age at operation and the presence of HS, and the relation between age at seizure onset and type of pathology (developmental or acquired) were investigated by using a Mann–Whitney U test. χ2 analysis was carried out to investigate the relation between a history of febrile convulsion or status epilepticus and HS.