Everitt, Alex Daniel; (2003) The structural basis of the epilepsies: An MRI and epidemiological study. Doctoral thesis (Ph.D), UCL (University College London).

Text The_structural_basis_of_the_ep (1).pdf Download (12MB)

Abstract

Contemporary knowledge of the epidemiology, aetiology and consequences of epilepsy are essential in order to effectively manage this common condition. Brain magnetic resonance imaging (MRI) helps determine the underlying cause in up to 75% of hospital based epilepsy patients, whilst at the population level the aetiology remains cryptogenic in the majority of cases. In particular, the frequency of hippocampal sclerosis (HS), the commonest structural correlate of epilepsy in tertiary referral centres, is unknown in the community. The cause(s) of HS, and whether it is progressive, are also poorly understood. This thesis consists of a series of community based studies of epilepsy employing high resolution MRI, in addition to a single MRI study of severe intractable epilepsy amongst the inhabitants of a large residential epilepsy centre. The main aim was to characterise the nature, frequency and severity of structural brain abnormalities, especially HS. A population base of 207,553 persons was established in Buckinghamshire, England. During a 2 year period, 165 adults with newly diagnosed and 279 adults with chronic active epilepsy were prospectively identified. The age-corrected annual incidence of all afebrile seizures was 52.8 per 100,000 (95% CI 42.9-62.6). In a sub-population of 159,388 adults, 279 had chronic active epilepsy (epilepsy of at least 4 years duration, with at least 1 seizure during the year preceding prevalence day), yielding a prevalence of 1.8 per 1000 (95% CI 1.55-1.96). High resolution MRI was performed in 110/165 (66.7%) of newly diagnosed adults, in 174/279 (62.3%) with chronic active epilepsy, and in 170 neurologically normal control subjects. 130/170 (76.5%) of the control subjects had normal MRI. Small (WML) were the most frequent visible abnormality (18.8%). No control had HS, MCD, or focal cortical brain damage. Patients with newly diagnosed partial onset seizures (37.9%) were significantly more likely to harbour focal brain lesions (focal scars, HS, MCD, cavernomas and neoplastic lesions) detectable by MRI than those with generalised seizures (7.0%) (p=0.0003). Focal brain abnormalities were also found more frequently in those with chronic active partial onset seizures (54.4%) than in those with generalised seizures (4.5%) (p<0.0001) The most frequent abnormalities overall were multiple small white matter lesions (WML) and focal neocortical damage. HS was an infrequent finding in the newly diagnosed partial seizures group (2.7%), but more common in the chronic active partial cohort (13.6%). However, quantitative hippocampal abnormalities were more frequent and were found in approximately one quarter of patients with newly diagnosed seizures, in one third with chronic active epilepsy, and in only 4.7% of control subjects. It was possible to assign a precise aetiology (including the idiopathic category) in more than three quarters of patients with newly diagnosed seizures, and in nearly two thirds with chronic active epilepsy. Hippocampal damage, as determined by the group comparisons of mean and smallest corrected hippocampal volumes, and mean and highest hippocampal T2 relaxation time, with control values, was statistically significant only in those with newly diagnosed partial onset seizures and chronic active partial epilepsy. Patients with idiopathic seizures did not have hippocampal damage, although those with non-idiopathic generalised seizures tended to have lower hippocampal volumes and higher hippocampal T2 values. Of the 263 residential epilepsy centre inhabitants, MRI showed abnormalities of aetiological relevance in 66%; HS/HA without other focal pathology, 26%; focal neocortical brain damage and HS/HA, 16%; isolated focal neocortical brain damage, 14%; malformations of cortical development, 7% (of whom 55% had HS/HA); other, 3%. Quantitative hippocampal abnormalities were present in 47.5%. High resolution MRI revealed previously undetected, aetiologically relevant lesions in 47% of the 203 residents previously scanned. MRI increased the proportion of residents with a definable aetiology from 50% to 68%. Logistic regression analyses in both the community and residential centre based subjects failed to reveal any significant correlations between the extent of hippocampal damage and the severity of the epilepsy.

Download activity - last month

Export as CSVXMLJSON

Download activity - last 12 months

Export as XMLCSVJSON

Downloads by country - last 12 months

Export as XMLCSVJSON

Archive Staff Only

View Item