Symptomatic vasospasm, or delayed cerebral ischemia associated with arteriographic evidence of arterial constriction, is currently the most important cause of morbidity after acute subarachnoid hemorrhage. The development of vasospasm is directly correlated with the presence of thick blood clots in the basal subarachnoid cisterns, which can be detected by an early computed tomographic scan. Symptomatic vasospasm usually develops between 4 and 12 days after subarachnoid hemorrhage. The onset is gradual, occurring over hours or days. There is typically a gradual deterioration of the level of consciousness, accompanied by focal neurological deficits that are determined by the arterial territories involved. Hyponatremia frequently occurs and may exacerbate the symptoms. The patients are usually volume depleted, and therefore many authorities now treat them with replenishment and expansion of their intravascular volume with colloid and blood. Volume expansion, together with elevation of the systemic blood pressure and reduction of the intracranial pressure when elevated, constitute the only currently available effective therapy for symptomatic vasospasm. The cause of vasospasm remains obscure. Mechanisms of smooth muscle cell contraction and relaxation and experimental efforts to elucidate the nature of vasospasm are reviewed.

Angiographic spasm of cerebral arteries was produced in dogs by successive injections of cisternal blood 48 hours apart. Angiograms were taken before and after each cisternal injection. There was progressively greater angiographic vasospasm of the basilar artery. Intravenous aminophylline, 10 mg/kg/hr, reversed vessel constriction on the 1st and 3rd day after one injection of cisternal blood. On the 5th day after two blood injections (on Day 1 and Day 3), intravenous aminophylline, nifedipine (1 mg/kg), and intra-arterial bolus injection of 2 mg/kg papaverine failed to reverse the constriction. The intractable constriction produced in this model resembles that found in patients. The calcium antagonist, nifedipine, is as ineffective as the more traditional vasodilators in reversing vasospasm in this model.

Sequential cisternal blood injections in dogs reproduce some of the morphological and physiological features seen in man after subarachnoid hemorrhage-induced vasospasm. This study reports the morphological features observed in cerebral vessels in areas exposed to subarachnoid blood. Subarachnoid hemorrhage was produced in dogs by two cisternal injections of non-heparinized autologous blood 48 hours apart. Dogs were sacrificed 48 hours after the second injection. Angiographic narrowing of the basilar artery was routinely present 48 hours after the second injection, and there was a good correlation between angiographic vasospasm and a narrowed arterial lumen at postmortem examination. All basilar arteries showed structural changes with electron microscopic examination; these included endothelial cell vacuoles, early smooth-muscle cell necrosis, intimal changes, and adventitial erythrocytes, leukocytes, and mast cells. The finding that accompanied vessel constriction most uniformly was packing of the adventitial cerebrospinal fluid spaces with erythrocytes. Angiographically visible spasm was resistant to vasodilators. These data suggest that infiltration of blood elements into the arterial wall is an important concomitant feature of morphological and angiographic vasospasm.

A morphologic and morphometric examination of the major cerebral blood vessels in the dog was carried out to determine whether there were vasa vasorum in these arteries and what features might be associated with them. True vasa vasorum confined to the media were not seen in any of the vessels examined. Microvessels confined to the adventitia, however, were found in the internal carotid and vertebral arteries but not in the basilar, middle cerebral, or anterior spinal arteries. Animal size, vessel size as determined by adventitial and medial area, and the number of smooth muscle cell lamellae were not associated with the presence of these adventitial vessels; they occurred only in arteries with both an intra- and extradural portion. It therefore appears that most canine cerebral arteries do not have vasa vasorum.