Cerebral vasospasm is one of the major causes of morbidity and mortality in patients with subarachnoid hemorrhage (SAH) (1). Angiographic vasospasm, defined as a focal or diffuse narrowing of the major cerebral arteries, appears most often around 7 days after the onset of SAH and continues for 2-3 weeks (2). Symptomatic vasospasm brings about permanent or temporal delayed neurologic deficits. Cooperative study of intracranial aneurysms and subarachnoid hemorrhage reported that symptomatic vasospasm occurred in 40% of patients with SAH and angiographic vasospasm in 62% (3). However, the detailed mechanisms producing cerebral vasospasm after SAH are still obscure.
Numerous humoral and neural factors are responsible for cerebral vasospasm. Recently particular attention has been given to nitric oxide (NO) as one of the relaxation factors. Several reports have demonstrated that endothelial function to produce/release NO is impaired in rabbit experimental SAH models (4,5). In addition, it has been reported that NO and NO donors effectively improve the cerebral vasospasm (6,7). Heros et al. (6) demonstrated that intravenous infusion of sodium nitroprusside was effective in relieving acute and delayed cerebral vasospasm in dogs, but the large doses required to achieve complete relaxation or nearly complete relaxation frequently produced significant hypotension. Therefore the drug's therapeutic uses become limited. Several investigators reported that the intracarotid infusion of NO and NO donors improved the cerebral vasospasm in a primate model of SAH (8,9). However, it is hardly acceptable for humans because it is invasive.
In the previous study, we demonstrated that the sensitivity to NO was increased when the endothelial function producing/releasing endothelium-derived relaxing factor (EDRF)/NO was impaired in the basilar artery isolated from SAH rabbits (5). Therefore our experiments were designed to investigate whether low-dose nitroglycerin as an NO donor effectively improves cerebral vasospasm in the rabbit SAH model. Low-dose nitroglycerin was applied onto ear skin with tape that is clinically available and causes only minimal invasion.
We used nitroglycerin tape (Vasolator tape; Sanwa Kagaku Kenkyusho, Nagoya, Japan) as an NO donor. This tape (20 cm2) contains 27 mg nitroglycerin, and reportedly releases 5.4 ± 1.2 mg nitroglycerin during 24 h (10). The plasma concentrations of nitroglycerin in human samples were 0.64 ± 0.24 ng/ml when 20-cm2 tape was used, but under the assay limit (<0.05 ng/ml) when half of the tape was used. On the basis of the rabbit body weight of ∼3 kg, which is estimated to be 1/20 of the human body weight of 60 kg, 1 cm2 tape was regarded as "usual dose" in the rabbit. In these experiments, we used 0.5-cm2 tape containing 0.675 mg nitroglycerin, which is assumed to release 0.125 mg nitroglycerin during 24 h, and defined it as "low dose."
Experimental SAH model and angiography
Twenty-one Japanese White male rabbits weighing 2.8-3.2 kg were used in the experiments. They were anesthetized with intravenous sodium pentobarbital (25 mg/kg), and supplemental anesthetic agent was administered during the experiment to ensure that the rabbits were as free of stress as possible. The experimental SAH was induced by injecting 3 ml nonheparinized autologous arterial blood into the cisterna magna according to the method described previously (11). Rabbits injected with 3 ml saline instead of the autologous blood served as controls. Animals were placed in the following three groups: (a) injected with autologous blood and applied low-dose nitroglycerin tape (nitroglycerin group, n = 7); (b) injected with autologous blood and applied placebo tape (placebo group, n = 7); and (c) injected with saline (saline group, n = 7). In the saline group, effects of low and usual doses of nitroglycerin tape on the diameter of the basilar artery also were determined. Arterial blood pressure and heart rate were measured every 24 h. The animals were placed in a supine position, and a catheter (Fastracker 325 infusion catheter; Target Therapeutics, Fremont, CA, U.S.A.) was inserted selectively into the left vertebral artery via a femoral artery by the Seldinger method, as previously described (12). Angiograms of the basilar artery were obtained by manual injection of 0.5 ml contrast medium containing 306 mg iopamidol (Iopamiron300; Schering AG, Berlin, Germany) for 2 s. Angiography was performed twice, once before injecting blood into the cistern on day 0 and once more on day 2. The angiograms obtained were transferred to an analytic processing system, and the diameter of the basilar artery was measured at five points (at the midpoint of the basilar artery, at 1 mm central and peripheral from the midpoint, and at 2 mm central and peripheral from the midpoint), as described previously (12). Then the mean diameter at these five points was determined. All angiograms were obtained by one investigator (Y.I.), and the diameters of basilar arteries were measured by another investigator (E.I.) without knowledge of the group.
The rabbits were killed on day 2, and basilar artery was isolated to measure mechanical responses. The tape without nitroglycerin or containing nitroglycerin was applied onto the inner ear skin immediately after inducing experimental SAH and exchanged every 24 h for 2 days. All surgical procedures were performed in a humane manner. The experimental protocol was evaluated and approved by the Committee on Animal Experimentation of Tokyo Medical and Dental University. The care of the animals was performed in accordance with the Guidelines for Animal Experimentation of Tokyo Medical and Dental University.
The following chemicals were used in these experiments: acetylcholine (Ovisot for injection; Daiichi Pharmaceutical, Tokyo, Japan); indomethacin (Merck-Banyu Pharmaceutical, Tokyo, Japan); calcium ionophore A23187, serotonin, methylene blue, sodium nitroprusside (all from Sigma, St. Louis, MO, U.S.A.); and NG-nitro-L-arginine and endothelin-1 (Protein Research Foundation, Osaka, Japan). All chemicals were dissolved in distilled water immediately before use except for indomethacin and A23187, which were dissolved in dimethyl sulfoxide and kept frozen at −20°C until use (10−2 M stock solution). Dimethyl sulfoxide was present in a final concentration of 0.5% in the experiments using the agents, and this concentration had no effect on any parameters tested.
Measurement of mechanical responses
Basilar arteries were isolated and immersed in cold modified Krebs solution. After the adherent connective tissue of the vessels was trimmed under microscopy, rings of the basilar artery 2 mm long were cut with a razor blade. A pair of stainless hooks, of which one was fixed and the other connected to a force-displacement transducer (TB-611T; Nihon Kohden Kogyo, Tokyo, Japan) was inserted in the lumen of the ring preparations. Chambers were filled with 4.5 ml modified Krebs solution maintained at 37 ± 0.5°C and continuously bubbled with 95% O2/5% CO2. Special care was taken to avoid unintentional rubbing of the intraluminal surface. Isometric tension was measured and recorded on a polygraph, as previously described (11). Before the experiments were started, the preparations were equilibrated in bathing solution for ≥60 min, with their length adjusted until a stable tension of 125 mg was obtained. The bathing solution was replaced with fresh solution every 20 min during the equilibration period. The compositions of the modified Krebs solution used were as follows (in mM): NaCl, 115.0; KCl, 4.7; MgSO4 · 7H2O, 1.2; CaCl2 · 2H2O, 2.5; KH2PO4, 1.2; NaHCO3, 25.0; and glucose, 10.0. After equilibration, each preparation was tested for the presence of functioning endothelial cells by an addition of 10−6 M acetylcholine during contraction induced by 10−5 M serotonin. After the control response induced by 60 mM KCl was obtained, ring preparations of basilar artery were exposed to increasing concentrations of the following contractile agonists (in the order listed): serotonin (10−9 to 3 × 10−5 M) and endothelin-1 (10−11 to 3 × 10−8 M). The results were expressed as a ratio to the control response induced by 60 mM KCl and compared among groups. To determine the relaxations, we used a different set of rings from the same animals. The rings were precontracted with 10−5 M serotonin, and increasing concentrations of vasodilators were added to the organ chamber until maximal relaxation was obtained. The vasodilators used (in the order listed) were acetylcholine (10−9 to 3 × 10−5 M), sodium nitroprusside (10−9 to 3 × 10−6 M), and A23187 (10−9 to 3 × 10−6 M). Acetylcholine was selected because it produces endothelium-dependent relaxation through a receptor-mediated mechanism (13); A23187 was used because it is an endothelium-dependent relaxant that is receptor independent in action (14); sodium nitroprusside directly relaxes the vascular smooth muscle by releasing NO as an active substance from its own structure (15). Relaxation induced by each agent was expressed as a percentage of the 10−5 M serotonin-induced contraction. After maximal response to each agent was obtained, the rings were washed repeatedly with modified Krebs solution and maintained in the bathing solution until stable tension was regained. The involvement of EDRF/NO in acetylcholine (10−6 M)-, sodium nitroprusside (10−6 M)-, and A23187 (10−6 M)-induced relaxations was confirmed by repeating these responses in the presence of NOARG (10−4 M), methylene blue (10−5 M), or without endothelium. Any effect of prostanoid formation during endothelium-dependent relaxation was ruled out by comparing responses with or without addition of indomethacin (10−5 M). Specimens were incubated with these antagonists for 20 min before adding 10−5 M serotonin. The maximal response (Emax) and the concentration producing 50% of Emax (EC50) were obtained from the log concentration-response curves.
Plasma concentration of nitroglycerin and its metabolites
Blood samples were taken every 24 h until day 2. The samples for determining the concentration of nitroglycerin and its metabolites were prepared according to the methods of Han et al. (16). In brief, the blood was collected in glass tubes and centrifuged at 3,000 rev/min, 0°C, for 15 min. Plasma was immediately frozen at −80°C until the determination. Plasma concentration of nitroglycerin, 1,2-, and 1,3-glyceryl dinitrate (1,2-GDN and 1,3-GDN, respectively) and 1- and 2-glyceryl mononitrate (1-GMN and 2-GMN, respectively) was measured by the gas chromatography method. The residual nitroglycerin on the used tape was dissolved in methanol and determined by the high-performance liquid chromatography method.
Results are given as mean ± SEM. Statistical significance of each value was assessed by analysis of variance (ANOVA). The values were considered significantly different at p < 0.05.
We investigated whether or not low-dose nitroglycerin tape had influences on the systemic blood pressure and heart rate. There was no significant difference in mean arterial blood pressure and heart rate between the placebo and nitroglycerin groups (Table 1). Because the total plasma concentration of nitroglycerin and its metabolites was under the assay limit (<0.05 ng/ml), we measured the residual nitroglycerin content in the tape and assumed the decreased amount of nitroglycerin as the dose that had been absorbed percutaneously. Apparent daily consumption of nitroglycerin through the skin was estimated to be 14.8% (n = 7) of the total content (0.675 mg; Table 2).
The diameters of the basilar artery were measured angiographically before the blood injection and on day 2. In the placebo group, the diameter of the basilar artery before the blood injection was 0.57 ± 0.013 mm (n = 7), and that on day 2 was 0.40 ± 0.015 mm (n = 7), which was significantly (p < 0.0001) reduced as compared with that before injection. In the nitroglycerin group, the diameters of basilar artery before the blood injection and on day 2 were determined to be 0.54 ± 0.012 mm (n = 7) and 0.48 ± 0.008 mm (n = 7), respectively. These values were also significantly (p < 0.0005) different, but the angiographic vasospasm was significantly (p < 0.0001) less as compared with that of the placebo group. The ratio of the diameter on day 2 to that before injection was calculated as 69.6 ± 4.4% (n = 7) for the placebo group and 89.4 ± 2.5% (n = 7) for the nitroglycerin group. The angiographic vasospasm of the basilar artery was significantly (p < 0.0001) reduced in the nitroglycerin group (Fig. 1). Diameter of the basilar artery remained unaffected in the sham-operated rabbits when 0.5 cm2 or 1.0 cm2 nitroglycerin tape was applied.
Acetylcholine, sodium nitroprusside, and A23187 produced concentration-dependent relaxations in the ring preparations of basilar artery. Relaxations produced by acetylcholine and A23187 were abolished after endothelial removal and greatly inhibited by 10−5 M NOARG or 10−5 M methylene blue. Indomethacin (10−5 M) did not modify these relaxations. None of these treatments except for 10−5 M methylene blue modified the sodium nitroprusside-induced relaxation, which was significantly attenuated by 10−5 M methylene blue. These responses were compared among the following three groups on day 2: saline group, placebo group, and nitroglycerin group. Comparisons were made in terms of Emax and EC50 (Table 3). As shown in Table 3, the Emax value for acetylcholine was significantly (p < 0.0001) decreased in the placebo group as compared with the saline group. In the nitroglycerin group, the impaired acetylcholine-induced relaxation was significantly (p < 0.0001) but not completely restored, because there were significant differences between the placebo group and the nitroglycerin group (p < 0.0001) and between the nitroglycerin group and the saline group (p < 0.01; Fig. 2). The EC50 value for sodium nitroprusside was significantly increased in the placebo and low-dose nitroglycerin groups. In the placebo group, the Emax value for A23187 had a tendency to increase compared with the saline group. In the nitroglycerin group, it also had a tendency to increase compared with the saline group and the placebo group, although it was not significant (Table 3).
Furthermore, we studied whether contractile responses to serotonin and endothelin-1 were influenced by low-dose nitroglycerin. Serotonin and endothelin-1 produced concentration-dependent contractions in the ring preparations of basilar arteries. These contractile responses remained unchanged between the placebo group and the nitroglycerin group (Fig. 3). The EC50 values (−logM) for endothelin-1 in the placebo and nitroglycerin group were determined to be 8.60 ± 0.05 and 8.61 ± 0.14, respectively. In the saline group, the value was 8.93 ± 0.14. There were significant differences between the placebo group and the saline group (p < 0.05) and between the nitroglycerin group and the saline group (p < 0.05), but the Emax values for endothelin-1 remained unchanged among the three groups (Fig. 4)
We determined the effect of nitroglycerin tape on the mechanical responses of basilar artery isolated from the sham-operated rabbits. There were no significant changes in responses to all contraction and relaxation agonists tested, when comparison was made with the corresponding values in the saline group (data not shown).
It has been postulated that control of cerebrovascular tone is achieved by a delicate balance between two opposing forces of relaxation and contraction. Disruption of the balance may be related to the development of delayed cerebral vasospasm after SAH. Despite many decades of research, the pathogenesis of cerebral vasospasm after aneurysmal SAH still remains elusive to neurosurgeons. As stated earlier regarding the variables pertaining to cerebral vasospasm, NO has been given particular attention as a relaxation factor. Isotani et al. (5) suggested that the less-functional muscarinic acetylcholine receptors, which mediate less production/release of EDRF/NO, might be involved in causing the reduced relaxations of basilar artery strips after SAH. In addition, Hatake et al. (17) demonstrated the reduced capability of endothelial cells to release EDRF/NO in human basilar artery after SAH. However, several reports have provided evidence that indicates that cyclic guanosine monophosphate (GMP) production is reduced without change in EDRF/NO release in a canine experimental SAH model (18,19).
In our previous demonstration, the sensitivity of basilar artery to NO was increased in the rabbit mild-SAH model, whereas production/release of EDRF/NO was impaired. We also demonstrated that the relaxations to acetylcholine and A23187 might be dependent on the endothelial cell and mediated by EDRF/NO (5). Based on our speculation about these findings, NO at low doses may effectively improve cerebral vasospasm after SAH without affecting intact systemic arteries. In these experiments, low-dose NO successfully improved the basilar artery vasospasm on day 2 without changing the diameter of the basilar artery in the sham-operated group. Furthermore, mean arterial blood pressure and heart rate were unchanged by the application of low-dose nitroglycerin. Several investigators (6,7) have attempted to reverse cerebral vasospasm in animal models with systemic administration of NO donors such as nitroglycerin and nitroprusside. Their success has been limited, probably partly because of differential features of the metabolic conversion and kinetics of these agents, differences in species and methods of drug delivery, and differences in techniques for assessing vasospasm. Furthermore, it may also concern the dosage of NO administered. The usual clinical dose of NO may cause the steal phenomenon, because it dilates intact arteries. Recently Pluta et al. (9) reported that regional intraarterial infusion of proli-NO with an extremely short half-life limited systemic exposure to NO yet retained its activity in the region of interest, as dilation of the spastic vessel was achieved without systemic hypotension. The pharmacokinetic advantage achieved by regional delivery of a drug is directly related to the total body clearance of the drug. In our experiments, the regional delivery of NO was achieved by the use of low-dose nitroglycerin, which did not affect the normal systemic vessels. In addition, nitroglycerin tape was used as the NO donor; it is both easily administered and less invasive than intraarterial infusion of proli-NO.
We speculated that low-dose nitroglycerin would improve the cerebral vasospasm after SAH by effectively dilating the spastic artery with exogenous low-dose NO. Furthermore, these results do not contradict to the hypothesis that low-dose NO possibly acts on the NO-starved smooth-muscle cells to improve the impaired acetylcholine-induced endothelium-dependent relaxation. Schwarz et al. (20) reported the enhancement of endothelium-dependent vasodilation by low-dose nitroglycerin. They intraarterially administered nitroglycerin at a dose that does not affect resting forearm blood flow. It increased the vasodilatory response to intraarterial administration of acetylcholine in patients with congestive heart failure but not in normal subjects. Although the race and the segment of the artery were different, their results coincided with ours. However, the detailed mechanism is still unknown and remains to be investigated.
Although low-dose NO could improve the cerebral vasospasm, the arteries still remained spastic. Because the mechanisms that control vascular tone are very complex, any single drug seems unable to restore the spastic arteries completely after SAH to normal functional levels. Further findings suggest that in the spastic artery, the Emax value for endothelin-1 is increased after endothelial removal, whereas in the normal artery, it remains unchanged. These findings suggest that reactivity of vascular smooth muscle to endothelin-1 is enhanced while endothelial function producing/releasing EDRF/NO is impaired in the rabbit basilar artery after SAH (E. Isotani, unpublished observation). Therefore more effective improvement will be achieved by suppressing contraction mechanisms in combination with activating dilation mechanisms.
Low-dose nitroglycerin effectively improved the cerebral vasospasm after SAH without any significant changes in the systemic circulation and normal cerebral arteries. These effects were achieved by applying nitroglycerin tape, which is clinically available and less invasive. Furthermore, these effects seem to be partially due to the effective dilation of the spastic artery and the improvement in the impaired acetylcholine-induced endothelium-dependent relaxation with low-dose nitroglycerin.
Acknowledgment: We thank Sanwa Kagaku Kenkyusho for their excellent technical assistance.
1. Kassell NF, Torner JC, Haley EJ, Jane JA, Adams HP, Kongable GL. The International Cooperative Study on the Timing of Aneurysm Surgery. Part 1: overall management results. J Neurosurg
2. Suzuki R, Masaoka H, Hirata Y, Marumo F, Isotani E, Hirakawa K. The role of endothelin-1 in the origin of cerebral vasospasm in patients with aneurysmal subarachnoid hemorrhage. J Neurosurg
3. Graf CJ, Nibbelink DW. Cooperative study of intracranial aneurysms and subarachnoid hemorrhage: report on a randomized treatment study. Stroke
4. Hongo K, Kassell NF, Nakagomi T, et al. Subarachnoid hemorrhage inhibition of endothelium-derived relaxing factor in rabbit basilar artery. J Neurosurg
5. Isotani E, Azuma H, Suzuki R, Hamasaki H, Sato J, Hirakawa K. Impaired endothelium-dependent relaxation in rabbit pulmonary artery after subarachnoid hemorrhage. J Cardiovasc Pharmacol
6. Heros RC, Zervas NT, Lavyne MH, Pickren KS. Reversal of experimental cerebral vasospasm by intravenous nitroprusside therapy. Surg Neurol
7. Frazee JG, Giannotta SL, Stern WE. Intravenous nitroglycerin for the treatment of chronic cerebral vasoconstriction in the primate. J Neurosurg
8. Afshar JK, Pluta RM, Boock RJ, Thompson BG, Oldfield EH. Effect of intracarotid nitric oxide on primate cerebral vasospasm after subarachnoid hemorrhage. J Neurosurg
9. Pluta RM, Oldfield EH, Boock RJ. Reversal and prevention of cerebral vasospasm by intracarotid infusions of nitric oxide donors in a primate model of subarachnoid hemorrhage. J Neurosurg
10. Fujiwara T. Evaluation of vasolator tape (Japanese). Pharma Medica
11. Azuma H, Obayashi S, Hamasaki H, Koyama T, Aso T. Role of endothelium in the human uterine arteries during normal menstrual cycle. Br J Pharmacol
12. Shishido T, Suzuki R, Qian L, Hirakawa K. The role of superoxide anions in the pathogenesis of cerebral vasospasm. Stroke
13. Furchgott RF, Zawadzki JV. The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature
14. Furchgott RF. The requirement for endothelial cells in the relaxation of arteries by acetylcholine and some vasodilators. Trends Pharmacol Sci
15. Kowaluk EA, Seth P, Fung HL. Metabolic activation of sodium nitroprusside to nitric oxide in vascular smooth muscle. J Pharmacol Exp Ther
16. Han C, Gumbleton M, Lau DT, Benet LZ. Improved gas chromatographic assay for the simultaneous determination of nitroglycerin and its mono- and dinitrate metabolites. J Chromatography
17. Hatake K, Wakabayashi I, Kakishita E, Hishida S. Impairment of endothelium-dependent relaxation in human basilar artery after subarachnoid hemorrhage. Stroke
18. Kasuya H, Weir BK, Nakane M, et al. Nitric oxide synthase and guanylate cyclase levels in canine basilar artery after subarachnoid hemorrhage. J Neurosurg
19. Kim P, Schini VB, Sundt TJ, Vanhoutte PM. Reduced production of cGMP underlies the loss of endothelium-dependent relaxations in the canine basilar artery after subarachnoid hemorrhage. Circ Res
20. Schwarz M, Katz SD, Demopoulos L, et al. Enhancement of endothelium-dependent vasodilation by low-dose nitroglycerin in patients with congestive heart failure. Circulation
This article has been cited
Keywords:© 2000 Lippincott Williams & Wilkins, Inc.
Cerebral vasospasm; Endothelial dysfunction; Low-dose nitroglycerin