SUMMARY Ninety-six patients with 2550% unilateral vertebral artery (VA) stenosis were followed up for an average of 4.6 years. In 89 patients (93%) at least one VA origin was involved, while the intracranial VA was affected in 3 patients (3%). Seventy-four patients (77%) had 2=50% stenosis of at least one internal carotid artery, of whom 52 underwent carotid endarterectomy. None of the patients had definite vertebrobasilar transient ischemic attacks (VB TIA). Nineteen patients (19.8%) experienced nonlocalizing symptoms possibly compatible with VB TIA, none of whom had a stroke. Twenty-three patients (24%) had strokes. The only two patients (2%) who sustained a brainstem infarction had fatal strokes and both were known to have basilar artery stenosis in addition to their VA stenosis. The observed stroke rate was 8.5 times the expected infarction rate for a normal population. Forty patients died during follow up. The observed 5-year survival rate was 60% compared to 87% in a matched normal population. Eight deaths (20% of all deaths) were caused by stroke and 21 deaths (52.5% of all deaths) were cardiac related.VA stenosis is most frequently located at the VA origin (93%), and is associated with a low incidence of brainstem infarction. Stroke Vol 15, No 2, 1984 THE PRESENT STUDY describes the stroke and survival rates in a group of patients with angiographically proven vertebral artery (VA) stenosis. Little information is available regarding the natural history of patients with VA occlusive disease. Previous studies suggest that stenosis of the distal VA is more dangerous than stenosis at the VA origin, and that hemodynamic factors may be more important than emboli in producing vertebrobasilar (VB) ischemia.1 " 3 However, lack of precise information about the prognosis in these patients makes recommendations for treatment difficult. The problem is further complicated since the clinical symptoms of VB ischemia are often difficult to define, and non-localizing events such as isolated vertigo may be erroneously labeled as "VB ischemia". MethodsThe records of patients who underwent aortic arch and/or selective vertebral artery angiography at the Cleveland Clinic between 1974 and 1978 were reviewed. The angiographic films of all patients with reported stenosis of at least 1 VA were examined. Ninety-eight patients with ^5 0 % unilateral or bilateral VA stenosis were selected for follow up. Patients with unilateral or bilateral VA occlusion only and patients who underwent VA operative procedures were excluded. Fifty-two patients who underwent carotid endarterectomy (CE) were included in follow up since the benefit of CE in VB ischemia is doubtful.4 These cases were analyzed separately. Follow up was accomplished using a standardized questionnaire and telephone interviews. End points were death or stroke. Vertebrobasilar transient ischemic attacks (VB TIA) were noted but were not sufficient reason for stopping follow up. The recommendations of the ad hoc NINCDS Committee were followed in defining VB From the

Forty-four patients with greater than or equal to 50% stenosis of a distal vertebral artery (VA) and/or basilar artery (BA) were followed up for an average of 6.1 years. Angiography was performed for definite vertebrobasilar (VB) transient ischemic attacks (TIA) in 19 (43%), for VB infarcts in 13 (30%) and for non localizing symptoms in 12 (27%). Stenosis in the BA with or without VA involvement was present in 28 patients (64%), while 16 patients (36%) had occlusive disease in one or both distal VA sparing the BA. In follow up, 7 patients (16%) had definite VB TIA and 3 patients had possible VB TIA. Eight patients (18%) sustained a stroke, 5 of which were in the VB territory. The observed stroke rate was 17 times the expected rate for a matched normal population. Eight patients died during follow up, three patients due to stroke (2 brainstem infarctions, one intraventricular hemorrhage). The observed 5 year survival rate was 78% compared to 90% in a matched normal population. In comparing this data with our previous study of 93 patients with proximal VA occlusive disease, distal VB occlusive disease appears to carry a higher risk for brainstem ischemia.

No abstract

OBJECTIVE There is little information on the frequency of symptomatic epidural hematomas after the implantation of paddle spinal cord stimulators (SCSs) in the thoracic spine. The purpose of this paper is to provide this metric and compare it to the frequency of symptomatic epidural hematomas for all other thoracic laminectomies combined. METHODS This study involved retrospectively analyzing the experience of a single surgeon in a consecutive series of patients who underwent the implantation of a thoracic paddle SCS with respect to the occurrence of a symptomatic epidural hematoma. For comparison, the occurrence of a symptomatic epidural hematoma in non-SCS thoracic laminectomies done during the same period of time was determined. RESULTS One hundred fifty-four thoracic paddle SCSs were implanted between May 2002 and February 2015. Despite perfect hemostasis and no preoperative risk factors, 4 of 154 patients (2.60%) developed postoperative lower-extremity weakness caused by an epidural hematoma. There were no other causes of a neurological deficit. In 3 of the 4 patients, the symptoms were delayed. Over the same time period, only 1 of 119 patients (0.84%) developed a postoperative motor deficit from a symptomatic epidural hematoma after a non-SCS laminectomy. CONCLUSIONS The occurrence of epidural hematomas after thoracic paddle SCS implantation may be underreported. Suggestions are given to decrease its incidence. It seems paradoxical that an epidural hematoma occurred 3 times more often after small SCS thoracic laminectomies than after larger non-SCS thoracic laminectomies. If confirmed by future studies, this finding may suggest that the intrusion of instruments into a confined epidural sublaminar space or the presence of a paddle and a hematoma in this restricted space may account for this differential.