We studied clinical characteristics and coexisting causes of stroke in 305 patients admitted to a population-based primary care center with an initial ischemic stroke and a potential cardiac source of embolism (PCSE). Using systematic standardized cardiac, arterial, and cerebral investigations and the logistics of the prospective Lausanne Stroke Registry, we found that nonprogressive onset, hemianopia without hemiparesis or hemisensory disturbances, Wernicke's aphasia, ideomotor apraxia, involvement of specific territories (posterior division of middle cerebral artery, anterior cerebral artery, cerebellum, multiple territories), and a hemorrhagic component were associated with the presence of a PCSE, as compared with 1,006 initial ischemic stroke patients without PCSE. Although age and sex did not differ, the frequency of hypertension, diabetes, cigarette smoking, elevated blood cholesterol, and deep hemispheric or brainstem infarcts was higher in the patients without a PCSE. Nearly one-fourth of the patients with a PCSE had a coexisting potential arterial cause of stroke (large artery greater than or equal to 50% stenosis or small-vessel disease). In the majority of patients with a PCSE (76.7%), cardioembolism was the most likely cause of stroke, although a direct source of embolism was uncommon (4.3%) and intracranial embolic occlusions were present in less than one-half of the patients who were angiographed.

In patients with acute stroke, electrical epileptic activity occurs more frequently than previously suspected.

All studies concerning the detection of patent foramen ovale (PFO) have compared transthoracic or transesophageal echocardiography (c-TEE) to transcranial Doppler ultrasound after contrast injection (c-TCD), but combining both techniques in the search of PFO has received no consideration. Our study aims to substantiate this claim in 37 patients with cryptogenic stroke. It includes two protocols for the detection of PFO to assess the complementarity of c-TCD and c-TEE performed simultaneously or separately. Firstly, we used a standardized protocol, performing c-TCD alone. Secondly, we used a standardized and a simultaneous protocol which associated c-TCD with c-TEE. When c-TCD and/or c-TEE found right-to-left shunts, they were classified as: minimal, intermediate and massive. c-TCD revealed all PFO detected by c-TEE in 24 patients out of 37 (65%). Furthermore, c-TCD was positive for a PFO in 5 other patients whereas c-TEE was negative. The degree of right-to-left interatrial shunting varied according to the protocol: c-TCD performed alone found 15 massive, 4 intermediate and 5 minimal shunts whereas 10, 9 and 5, respectively, were detected by c-TCD when it was combined with c-TEE. In contrast, c-TEE revealed 8 massive, 8 intermediate and 8 minimal shunts. c-TCD can identify minimal shunts missed by c-TEE and could be more relevant to detect massive shunts, particularly when not performed simultaneously with c-TEE because no sedation is required for c-TCD alone as opposed to c-TEE: thus patients are more cooperative and produce a better Valsalva strain. c-TEE confirms pulmonary shunts suspected by c-TCD and determines the morphologic characteristics of the interatrial septum. While previous studies opposed c-TEE against c-TCD for the detection of a PFO, we think that both techniques are complementary and that it is interesting to associate them, particularly when they are deferred, to increase the ability of detecting PFO and to specify the degree of right-to-left shunting.

SUMMARY Five patients with "locked-in" syndrome and dysconjugate palsy of horizontal gaze were studied. In all cases internuclear ophthalmoplegia due to dysfunction or destruction of the median longitudinal fasciculus was combined with an ipsilateral gaze palsy, producing the "oneand-a-half" syndrome. Clinical and electro-oculographic examination suggested involvement of the paramedian pontine reticular formation when all ipsilateral saccades were abolished, when exotropia of the contralateral eye was present, and when vestibular stimulation showed full conjugate deviation to the damaged side. Involvement of the abducens nucleus was suggested when the palsy of ipsilateral gaze was not dissociated on vestibular stimulation. In three cases these clinical deductions were confirmed by the pathological study, which showed a corresponding destruction -of the median longitudinal fasciculus, paramedian pontine reticular formation and abducens nucleus. In one case the one-and-a-half syndrome evolved into a total horizontal gaze palsy, which corresponded to involvement of the abducens nucleus contralateral to the initially destroyed paramedian pontine reticular formation. Vertical oculocephalic response disappeared, because of destruction of the median longitudinal fasciculus on both sides (bilateral internuclear ophthalmoplegia). Patients with the locked-in syndrome provide a unique situation in which complex pontine oculomotor disturbances may be studied, because consciousness is preserved. In these patients, dissociated and dysconjugate oculomotor palsy may have been underestimated.