Thirty-four percent of 182 ischemic stroke patients registered during 1 year in a prospective hospital stroke data base complained of headache within a 72-hour interval of stroke onset. Headache was more common in patients under 70 years of age, in nonsmokers, in those with a past history of migraine, and in subjects presenting transient loss of consciousness, nausea/vomiting, or visual field defects. Headache was more frequent in vertebrobasilar (57%) than in carotid (20%) territory strokes, more so in posterior cerebral artery (90%) and cerebellar infarcts (80%), and was infrequent in subcortical infarcts (7%) and lacunes due to single perforator disease (9%). In multiple regression analysis, vertebrobasilar stroke (odds ratio 6.9), lacuanr stroke (odds ratio 0.06), and past history of migraine (odds ratio 6.7) were significant independent predictors of headache, suggesting that ischemic stroke location is the major determinant of stroke-associated headache, most probably related to activation of the trigeminovascular system, whose threshold may be modified by individual susceptibility.

Sixty (29%) of 205 consecutive patients with transient ischemic attacks registered in a hospital stroke data base had headache within 72 hours of onset. Headache was significantly more common in nonsmokers (odds ratio = 2.8; 95% confidence interval = 6.7 to 1.2). Headache was infrequent in patients with amaurosis fugax, and was not significantly associated with any other particular clinical presentation of transient ischemic attack. Headache was more common in vertebrobasilar (33%) than in carotid distribution (24%) episodes, and was not rare in transient ischemic attacks presenting as lacunar syndromes (29%). Headache was less frequent in patients whose computerized tomograms showed an infarct appropriate to the symptoms (odds ratio = 0.2; 95% confidence interval = 0.02 to 1.4). A diffuse headache was more common in patients with lacunar events than in patients with cortical attacks (odds ratio = 3.0; 95% confidence interval = 13 to 0.07). No other association was found between headache location and the presumed involved vascular territory. Headache in patients with transient ischemic attacks is poorly related/explained by the clinical characteristics of the ischemic event.

In paralytic shellfish poisoning a mollusc contaminated with a toxin (saxitoxin) causes a potentially lethal disease, clinically characterised by gastrointestinal and neurological symptoms, of which possible respiratory depression is the most serious. The toxin acts by blocking the sodium channels. We report 9 Portuguese patients with this disease. The mollusc was identified as Mytilus edulis, contaminated with the dinoflagellate Gymnodinium catenatum, and the toxin saxitoxin. Our patients had a benign clinical course with cerebellar ataxia as the most severe neurological impairment. Eight out of 9 patients had neurophysiological investigations, the largest number so far reported. Motor and sensory conduction velocities and amplitudes were normal. The proximal conduction times, as assessed by F waves, showed delayed conduction and decreased frequency, which returned to normal in few weeks. The somatosensory evoked potentials confirmed normal peripheral and central sensory conduction. The rich vascular supply at root level of the sodium channels of the proximal motor nerves may explain the greater vulnerability to toxin damage. The typically transient and quickly reversible nerve dysfunction caused by ion channel blockade is reported.

Serial cerebral angiograms, computed tomography, and magnetic resonance imaging are among the proposed methods for monitoring disease activity and response to therapy in isolated angiitis of the central nervous system. Cerebrospinal fluid has not proved to be useful in monitoring clinical course. We describe a 45-year-old man with histological diagnosis of isolated angiitis of the central nervous system that was treated with prednisone plus azathioprine and monitored for 2 years. Samples of the cerebrospinal fluid were obtained for cytological and routine chemical examination, as well as albumin and immunoglobulin content. Before treatment, cerebrospinal fluid showed marked plasmatic transudation of albumin and intrathecal synthesis of immunoglobulins. During the first year of immunosuppression no events were noticed, and the previously abnormal aspects of the cerebrospinal fluid showed improvement. During the weaning of azathioprine, a new stroke occurred in conjunction with a marked deterioration of cerebrospinal fluid parameters. Immunosuppression was resumed at previous levels, and during the following year no further events occurred. Once again, abnormal cerebrospinal fluid values improved significantly. We report a case of isolated angiitis of the central nervous system in which the serial cerebrospinal fluid examinations (albumin and immunoglobulin content) showed a close correlation with clinical course. This method may be useful in monitoring response to therapy.

The clinical predictors and potential therapeutic implications of microembolic signals (MES) were investigated in 104 patients with ischemic stroke or transient ischemic attack associated with high-risk (n = 60) or low-risk (n = 44) cardiac emboligenic conditions. Patients with artificial valves and carotid stenosis were excluded. MES counts were based on 30-min bilateral middle cerebral artery transcranial Doppler (TCD) monitoring recordings. MES were detected in 15 subjects (14%). The mean number of MES per hour was 1.9. MES prevalence and counts were influenced neither by age, gender or type of cerebral event, nor by cardiac disease. MES were more frequently detected and greater in number in patients with multiple cerebral ischemic events and in subjects monitored within 1 month of the last event. By multiple logistic regression, TCD recording within 1 month of the last event [Odds ratio (OD) = 13.5; 95% confidence intervals (CI) = 3.3–54.3] and multiple cerebral events (OD = 4.7; 95% CI = 1.3–17.3) were the best MES predictors. MES were detected in 40% (4/10) of patients on heparin, 5% (2/4) of untreated subjects and 13% (5/39) of those on antiplatelet drugs. For patients on warfarin, MES counts and prevalence were similar in subjects with international normalized ratios (INR) below and within/above the therapeutic range. MES were detected in only 2 out of 16 subjects with INR <2 and in 1 with INR >2. MES detection can be potentially relevant to the selection of antithrombotic treatment in acute stroke associated with cardioembolic disease, but further studies are necessary to assess its effectiveness as an additional guide to monitor oral anticoagulant intensity.