Search

We followed 123 patients with primary intracerebral hemorrhage (ICH), defined as bleeding without known precipitating cause except hypertension, for an average of 4.6 years or until death in order to determine the incidence, prevalence, and type of epileptic seizures. Twenty-five percent had seizures. In one-half of these, the seizures began within 24 hours of the hemorrhage. Survival table analysis predicted a potential cumulative seizure incidence of 50%, had all patients survived 5 years. Seizure incidence was high with bleeding into lobar cortical structures (54%), low with basal ganglionic hemorrhages (19%), and zero with thalamic hemorrhages. Within the basal ganglia, caudate involvement predicted seizures; within the cortex, temporal or parietal involvement predicted seizures. Although seizure incidence was high, prevalence of chronic epilepsy was much lower: 13% in 30-day to 2-year survivors and 6.5% in 2- to 5-year survivors. Seizure incidence is higher than previously reported after ICH because small lobar hemorrhages are the most epileptogenic and are now easily recognized with computed tomography.

Brainstem auditory evoked responses (BAERs) were recorded in 40 patients with clinically definite brainstem infarction, and results were compared to localizations from physical signs and CT scans. The BAER was abnormal in 92% of patients with evidence on physical examination of dysfunction of lateral structures in the pons or midbrain. Normal BAERs were seen with medially-situated or medullary lesions. When both rostrocaudal level and lateralization were considered, the BAER indicated damage in additional areas not evident on physical examination in 25% of patients. However, physical signs indicated damage in areas not reflected by the BAER in 22% of patients. Therefore, the BAER complements the localization obtained from physical findings. BAERs were abnormal in more of these patients than were CT scans, and thus are useful for confirmation of bedside impressions.

We identified a large Georgia kinship in which a hereditary autosomal dominant chorea appeared at an average age of 65 years, much later than usual for Huntington's chorea. Progression was slow. Dementia was not an obvious initial feature. Family members denied that affected persons became demented, and those affected cognitively intact with bedside testing. However, deficits of memory were apparent on formal psychologic testing. In the propositus' generation, five of 12 siblings were affected during their 60s. Of 23 persons in a younger generation, aged 30 to 60 years, none was yet affected. A number of variants of Huntington's chorea have been proposed. In elderly patients without obvious dementia, psychologic testing may be of diagnostic importance, revealing characteristic memory deficits. When the disorder consistently occurs at an advanced age and progresses slowly, the implications for the family may be less grave than with Huntington's chorea of earlier onset.

We examined the records of 147 consecutive patients studied by femoral catheterization to identify factors contributing to angiographic risk in cerebrovascular disease. Cerebral complications occurred in 12.2 percent and were permanent in 5.2 percent. Computer-assisted multivariate analysis of 21 possible risk factors was done. Two of these risk factors correlated strongly with increased risk: number of previous transient ischemic attacks (TIAs) (p less than 0.001), and the presence of arterial stenosis of greater than 90 percent (p less than 0.03). Risk factors of marginal significance were: diabetes, female sex, and number of selective injections. A discriminant function for estimation of risk was derived: D = [8 X number of TIAs] + [6 x number of arteries catheterized] + [14 if diabetic, 0 if not] + [11 if female, 0 if male]. When D was greater than 55, 77 percent of patients had a complication. When D was less than 55, 98 percent of patients had no complication. Unfortunately, patients in whom the study is most indicated tend to be those at greatest risk.

EEGs were recorded with depth electrodes in 8 monkeys undergoing transient middle cerebral artery ligation. Electrodes measured EEG, cerebral blood flow (CBF), and tissue oxygen simultaneously during and after occlusion. An EEG frequency analysis was performed. Electrode sites were examined microscopically, and infarction size, tissue vacuolization index, and neuronal morphology were described quantitatively. Serial neurological examinations were performed. Two patterns of EEG frequency change were delineated, dependent upon degree of ischaemia. Mild ischaemia, as indicated by less severe clinical deficits, higher CBF during occlusion, and minor pathological changes was associated with large increases in slow EEG activity and decreases in fast EEG activity during occlusion, with recovery of slow activities to baseline, but continued suppression of fast activities 24 h later. Severe ischaemia was associated with suppression of both fast and slow frequencies during occlusion, with slow activities returning to baseline and fast activities remaining suppressed 24 h later. The best quantitative EEG indicator of severity of ischaemia was suppression of slow wave activity during occlusion. The best EEG indicator that an ischaemic event had occurred 24 h previously was continued suppression of fast EEG activities. These data may be helpful in the design of EEG frequency analysis studies for monitoring the time course of human cerebral ischaemia and for retrospective diagnosis of transient ischaemic attacks (TIAs).