Search

Safety and efficacy of the thrombolytic agent urokinase (URO) in the elimination of subarachnoid clot and prevention of chronic vasospasm was compared with tissue-type plasminogen activator (rt-PA) in a blind, randomized placebo-controlled trial. Twenty monkeys were randomly assigned to one of five groups of four. Each group underwent baseline cerebral angiography followed by bilateral craniectomy and experimental subarachnoid hemorrhage. An Ommaya reservoir was inserted on the right side with its catheter placed into the ipsilateral subarachnoid space. Twenty-four hours later, depending upon group assignment, the animals received 100,000 IU URO, 200,000 IU URO, 1 mg rt-PA, 2 mg rt-PA, or the equivalent volume of normal saline (control group). On Day 7, angiography was repeated and the animals were killed. One animal died as a result of complications during the baseline angiography, presumably due to blood loss and prolonged anesthesia, and a replacement animal was obtained. No animals demonstrated any delayed neurological deficits. The study demonstrated that a single intracisternal bolus injection of rt-PA, 2.0 mg in 2 ml sterile water, or URO, 200,000 IU in 2 ml sterile water, 24 hours after induction of experimental subarachnoid hemorrhage in primates, was equally effective in thrombolysing ipsilateral clot, but neither dosage prevented angiographic vasospasm. Vasospasm occurred bilaterally in all groups. Whereas gross subarachnoid clot was found bilaterally in all animals in the placebo group and both smaller-dose URO and rt-PA groups, right-sided subarachnoid clot was virtually absent and left-sided clot reduced in both higher-dose URO and rt-PA groups.(ABSTRACT TRUNCATED AT 250 WORDS)

This paper reviews our recent studies on the mechanism of cerebral vasospasm following subarachnoid hemorrhage (SAH) in monkeys. Middle cerebral artery (MCA) vasospasm was maximal at 7 days, resolving by 14 days, and absent at 28 days after SAH. Arterial fibrosis was not detected during vasospasm, although there was intimal hyperplasia with fibrosis 28 days after SAH. On scanning electron microscopy, smooth muscle cells from vasospastic arteries had corrugated cell membranes and appeared similar to cells contracted pharmacologically, suggesting that vasospastic smooth muscle is contracted. Morphometric analysis of arteries obtained 7 days after SAH showed no significant increases in arterial wall area of vasospastic arteries compared with normal MCAs. The results suggest vasospasm in monkeys is not due to hypertrophy, hyperplasia, or fibrosis in the arterial wall. Vasospasm may be mainly vascular smooth muscle contraction, which damages the arterial wall, leading to secondary structural changes in the arterial wall which occur after angiographic vasospasm.

Cerebral vasospasm may result from lipid peroxidation induced by oxyhemoglobin in the subarachnoid space after subarachnoid hemorrhage. To test this theory, vasospasm was induced in monkeys by intrathecal injections of oxyhemoglobin or supernatant fluid from autologous blood incubated in vitro. Concentration of malondialdehyde (MDA), a product of lipid peroxidation, was elevated in cerebrospinal fluid (CSF) in association with vasospasm caused by oxyhemoglobin and supernatant fluid. Intrathecal injections of methemoglobin or bilirubin did not cause vasospasm or increased CSF MDA. Activity of glutathione peroxidase in CSF increased significantly after injection of oxyhemoglobin and methemoglobin. There were no significant changes in CSF superoxide dismutase activity although there was a trend towards higher activities in animals treated with oxyhemoglobin, methemoglobin, bilirubin, and supernatant fluid. These results show oxyhemoglobin-induced vasospasm is associated with MDA and lipid peroxidation in the subarachnoid space. Furthermore, detection of peroxidation products after injection of oxyhemoglobin in the absence of erythrocyte membranes indicates that oxyhemoglobin may directly damage cerebral arteries and brain by inducing lipid peroxidation in these structures. Depletion of free-radical scavenging enzymes in CSF did not seem necessary for development of vasospasm. In fact, there was a tendency for vasospasm to elevate enzyme activities, as if production of scavengers was induced by excess free radicals in the subarachnoid space.

A gel consisting of agarose and oxyhemoglobin (OxyHb) was developed so that, when placed in the subarachnoid space, OxyHb would be slowly released, simulating lysis of erythocytes after subarachnoid hemorrhage. The system was used to investigate the importance of reactions mediated by free radicals in the genesis of OxyHb-induced vasospasm in monkeys. Seventeen monkeys were randomly assigned to have subarachnoid placement, on Day 0, of one of the following: 1) agarose gel alone (n = 2); 2) agarose plus OxyHb (n = 3); 3) agarose plus OxyHb plus intrathecal administration of superoxide dismutase and catalase (n = 6); and 4) agarose plus OxyHb plus intrathecal administration of placebo (n = 6). Vasospasm was assessed by comparison of angiograms performed on Day 0 and 7 days after subarachnoid placement of compounds, and by electron microscopy. OxyHb alone caused significant reduction in the diameter of the middle cerebral artery (40 +/- 8%, P less than 0.005, paired t test), which was associated with ultrastructural damage to smooth muscle. Treatment with superoxide dismutase plus catalase or with placebo attenuated vasospasm of the middle cerebral artery, although significant narrowing persisted in both groups (27 +/- 12% and 26 +/- 13%, respectively, P less than 0.05, paired t test). Analysis of variance showed no difference in the degree of vasospasm between groups exposed to subarachnoid placement of OxyHb. Cerebrospinal fluid aspirated from the cisterna magna on Day 7 contained elevated activity of superoxide dismutase in animals that received treatment. Malondialdehyde was undetectable in cerebrospinal fluid after subarachnoid placement of agarose alone, although it was present in similar amounts in all groups that received subarachnoid placement of OxyHb. Since intrathecal superoxide dismutase and catalase failed to protect against OxyHb-induced vasospasm, mechanisms mediated by free radicals may not be important in its genesis. As only one combination of doses of superoxide dismutase and catalase was administered, however, it may be that other dosage schedules might be efficacious.

It is unclear if vasospasm after subarachnoid hemorrhage (SAH) is predominantly due to smooth-muscle contraction, proliferative vasculopathy, or other changes within the arterial wall such as fibrosis or change in smooth-muscle phenotype. In this study, immunohistochemistry was used to examine changes in extracellular and cytoskeletal proteins in cerebral arteries after SAH that might support one of these mechanisms. Following baseline cerebral angiography, bilateral SAH was created in nine monkeys. Three animals each were killed 7, 14, or 28 days after SAH. Cerebral angiography was repeated on Day 7 in all animals and immediately prior to sacrifice in animals killed on Days 14 and 28. Both middle cerebral arteries and four control basilar arteries were examined using fluorescent antibody techniques with antisera to alpha-actin, myosin, fibronectin, fibrinogen, vimentin, desmin, laminin, and collagens (types I, III, IV, and V). Angiography showed that vasospasm was most severe on Day 7, present but resolving on Day 14, and completely resolved by Day 28. Microscopic study of arterial sections and blinded review of microphotographs of arterial sections by five independent observers did not reveal changes in intensity of density of staining for collagens, desmin, myosin, laminin, or alpha-actin in the tunica media of tunica adventitia. Fibronectin immunoreactivity increased 14 days after SAH. Seven days after SAH, occasional areas of tunica media showed immunoreactivity to fibrinogen. On Day 28, intimal thickening was observed in four of six middle cerebral arteries and this tissue demonstrated immunoreactivity to alpha-actin, myosin, vimentin, desmin, fibronectin, laminin, and each type of collagen. No significant increases in the number of intimal cells showing immunoreactivity to alpha-actin were seen and no significant changes in the hydroxyproline content of cerebral arteries developed at any time after SAH. These results suggest that rigidity and lumen narrowing of vasospasm are not due to increased arterial collagen, although other proteins in the arterial wall or an alteration in cross-linking of existing proteins could produce these changes. There is no indication that smooth-muscle contractile proteins change during vasospasm or that increases in the number of alpha-actin-containing myointimal cells contribute to vasospasm. The occurrence of intimal thickening and increased tunica media fibronectin after vasospasm suggests that vasospasm damages smooth muscle, possibly as a result of intense prolonged smooth-muscle contraction.

Normal cytoarchitecture of smooth muscle cells of monkey cerebral arteries was studied using scanning electron microscopy after removal of adventitial connective tissue by hydrolysis with HCl. Cerebral arteries were also examined after contraction in vitro with prostaglandin F2 alpha (PGF2 alpha). Anterior cerebral arteries were studied after exposure for 6 days in vivo to whole blood, oxyhemoglobin, methemoglobin, bilirubin, mock cerebrospinal fluid, or supernatant fluid from an incubated mixture of autologous blood and mock cerebrospinal fluid. Normal smooth muscle cells were spindle-shaped and oriented circumferentially around the vessel. They were often grouped into bundles of 5 to 10 cells; bundles were recognizable because cells within them were joined by multiple intercellular contacts. Groups of smooth muscle cells oriented longitudinally were present outside the circular layers of cells. The adventitial surface of muscle cells was smooth apart from fine longitudinal striations in some areas. Arteries contracted with PGF2 alpha had markedly convoluted and folded cell membranes. Muscle cells of vasospastic arteries and of arteries exposed to oxyhemoglobin and supernatant fluid appeared identical to cells contracted with PGF2 alpha. The outer surface of cells of arteries exposed to bilirubin, methemoglobin, and mock cerebrospinal fluid were normal. Marked similarity between vasospastic smooth muscle cells and smooth muscle cells from arteries contracted with PGF2 alpha suggest that smooth muscle contraction occurs during "vasospasm" due to whole blood and to intrathecal injection of oxyhemoglobin.

A primate model was used to determine whether oxyhemoglobin (OxyHb), methemoglobin (MetHb), or bilirubin is likely to be responsible for cerebral vasospasm following subarachnoid hemorrhage (SAH). Forty cynomolgus monkeys were randomly assigned to one of five groups. On Day 0, each animal underwent angiography followed by right craniectomy and placement of an Ommaya reservoir with its catheter adjacent to the right middle cerebral artery (MCA). The animals received intrathecal injections twice a day for 6 days of one of the following solutions: mock cerebrospinal fluid (CSF); OxyHb; MetHb; bilirubin; or supernatant fluid from an incubated mixture of autologous blood and mock CSF. On Day 7, angiography was repeated and the animals were killed. Comparison of angiograms obtained on Day 0 and Day 7 of the experiment showed significant vasospasm of the right MCA and the right anterior cerebral and internal carotid arteries in the animal groups that had received OxyHb or supernatant fluid. There was a smaller reduction in diameter of the same vessels in the bilirubin group (not statistically significant), while no effects were observed in the groups receiving MetHb or mock CSF. Electron microscopy of the right MCA's gave results consistent with the angiographic findings. One monkey in the OxyHb group developed a delayed-onset right MCA infarction. These data suggest that OxyHb is the cause of cerebral vasospasm following SAH.

It is generally believed that surgery in the face of angiographic vasospasm is dangerous due to an increased incidence of postoperative cerebral ischemia. One theory is that arterial narrowing is exacerbated by surgical manipulation of vasospastic vessels during aneurysm dissection and clipping. This theory was tested in a primate model of cerebral vasospasm and the results reported. Six monkeys underwent baseline cerebral angiography, followed by induction of subarachnoid hemorrhage (SAH) on both sides of the circle of Willis. An equal amount of fresh autologous blood clot was placed around each internal carotid, anterior cerebral, and middle cerebral artery. Six days later, angiography was repeated and the right craniectomy was reopened for clot evacuation and surgical manipulation of the right cerebral arteries, including placement of a temporary aneurysm clip on the right middle cerebral artery. The left cerebral arteries were not exposed or manipulated, and served as controls. Twenty-four hours later angiography was repeated, then the animals were killed. Equal and significant vasospasm (greater than 40% reduction in vessel caliber compared to baseline, p less than 0.05) was seen in the middle cerebral arteries on both sides of the circle of Willis in all animals 6 and 7 days after SAH. There was no significant change in the severity of vasospasm on Day 7 compared with Day 6 in the right cerebral arteries. Increased risk of postoperative cerebral ischemia for surgery in the peak vasospasm period may be due to mechanisms other than increased arterial narrowing precipitated by surgical manipulation.

Fifteen patients undergoing surgery within 48 hours of aneurysm rupture were administered recombinant tissue plasminogen activator (rt-PA) directly into the basal subarachnoid cisterns after minimal surgical clot removal and aneurysm clipping. Preoperatively, 13 patients had diffuse or localized thick subarachnoid blood clots on computerized tomography (CT), and two had diffuse thin clots. The rt-PA was given as a single intraoperative injection of 7.5 mg (one patient), 10 mg (nine patients), or 15 mg (five patients). Postoperative cisternal drainage was employed in three patients. All patients except one demonstrated partial to complete cisternal clot clearance on CT scans within 24 hours after surgery. The patient who showed no clot reduction was the only patient in this series to develop symptomatic vasospasm and was the only fatality, dying 8 days after rupture. No vasospasm was seen on follow-up cerebral angiography in six of the 14 responding patients, and mild-to-moderate arterial narrowing was seen in at least one major cerebral artery in the remaining eight patients. Severe angiographic vasospasm was not seen, although the patient who died did not undergo repeat angiography. There was one major complication early in the series which seemed clearly related to treatment, and that was a large extradural hematoma occurring within several hours of craniotomy. Intrathecal fibrinolytic treatment appears effective in clearing subarachnoid clot and reducing vasospasm, and may be associated with acceptable risks if given to patients with large-volume subarachnoid hemorrhages at high risk for severe vasospasm.

The safety, prevention, and treatment of chronic vasospasm by repeated administration of intrathecally applied nimodipine was evaluated in a primate model of chronic cerebral vasospasm. Twenty-four female cynomolgous monkeys were randomized into three groups of 8: sham, clot, and clot + intrathecal nimodipine. All animals underwent a subarachnoid hemorrhage (SAH) induction operative procedure after base line angiography. Nimodipine was administered postoperatively by subcutaneous injection of 1 ml/0.2 mg tid for 6 days through an Ommaya reservoir with the catheter placed in the subarachnoid basal cisterns. Angiography was performed on Day 7 post-SAH induction. Intrathecally applied nimodipine was not effective in the prevention of angiographic vasospasm. It also did not seem to decrease the degree of pathological change when compared to controls. One animal in the nimodipine group died 2 hours after an intrathecal injection secondary to a respiratory arrest. Transient sedation and hypoventilation were common. No adverse pathological effects were noted. Intrathecal nimodipine did not produce a significant dilation of vessels in moderate or severe spasm when assessed by angiography 2 hours after intrathecal injection. Only 1 animal in 8 showed diffuse dilation of vasospastic cerebral vessels after an intrathecal nimodipine injection. Three other animals in this group developed dilation only of the basilar artery, which was in mild spasm.