The clinical patterns of frontal lobe epilepsy

• Main Author: Manford, Mark Ralph Andrew
• Format: Dissertation
• Language: English
• Published: University of London 1993

The aims of this study were to define patterns of seizure semiology aid see if these can be localised to particular regions, with special emphasis on frontal lobe epilepsies; to establish clinical features differentiating frontal and temporal lobe epilepsies; to assess current classification of frontal lobe epilepsies and to analyse partial epilepsies in a general population. The method; used differ from other studies in this area, especially with respect to breadth of case mix; delineation of clinical syndromes, using statistical techniques and prospective analysis of investigative abnormalities in relation to clinically defined seizure types. Patients were selected from hospital records with evidence of partial seizure onset, on the basis of focal imaging abnormality, interictal or ictal EEG abnormality and clinical seizure pattern. Two hundred and fifty-two patients with 352 seizure types were identified. Clinical seizure manifestations were recorded prospectively and encoded according to sequential occurrence during she seizure. These data were entered into a statistical cluster analysis, which was refined to 14 groupings, each corresponding to a different seizure type. These patterns were displayed in flow charts, allowing assessment of seizure evolution. Interictal EEG spike distribution and ictal EEG onset were related to each seizure type. The 126 lesions identified on neuroimaging were measured by a template technique, aid related to each seizure type, using chi-square analysis. Investigations were also related to each other and the clinical seizure associations of pure frontal and pure temporal lesions were assessed. The database of the National General Practice Study of Epilepsy (NGPSE) was also analysed, using a clinical classification, to determine the relative frequency of different partial seizure types in a general population. Of 14 clinical seizure groupings, 2 had associations with frontal regions and 2 perirolandic associations. Focal clonic and somatosensory seizures were associated with perirolandic lesions and other sensory modalities and experiential phenomena with temporal EEG and imaging abnormalities. Seizures characterised by tonic posturing were strongly associated with lesions of the lateral premotor cortex and frontocentral EEG abnormalities. Seizures characterised by bizarre oi frenetic motor activity "motor agitation" were associated with frontal EEG abnormalities and with lesions of the frontopolar or orbitofrontal cortex. These latter seizures were commonly nocturnal and occurred with higher frequency than those in other groups but no other consistent pattern in diurnal variation or seizure frequency emerged. For all seizure types with statistically significant associations with specific lesion sites, there was a substantial minority of cases associated with lesions at different site. A direct comparison of seizures associated with pure frontal and pure temporal lesions confirmed earl) version, clonic activity and tonic posturing as frontal characteristics but found no consistent differences in characteristics of seizure timing. Seizures with startle sensitivity occurred in 19 cases; MRI suggested lesions in the lateral frontal region or the perisylvian region in this group. The NGPSE data supported a high frequency of seizure types associated with frontal lobe abnormalities in the general population and a good prognosis of all partial seizure categories. These findings lend some support to the localisation of seizure types attributed to orbitofrontal and temporal lobes by the ILAE but suggests that many regions of the frontal lobes do not have specific associated seizure types, and that many seizure types, although being associated with one cerebral region, may relatively frequently be due to lesions in other sites. The predictive value of clinical seizure type is thus less than described from retrospective and highly selected post-surgical series, on which the classification is based. The hospital and population-based studies suggest that seizure frequency, timing and prognosis are more related to the population under study than the anatomical site of the underlying lesion.