Prevention of cerebral vasospasm by a capsaicin derivative, glyceryl nonivamide, in an experimental model of subarachnoid hemorrhage

Vascular

Prevention of Cerebral Vasospasm by a Capsaicin Derivative, Glyceryl Nonivamide, in an Experimental Model of Subarachnoid Hemorrhage Chih-Lung Lin, M.D., Yi-Ching Lo, Ph.D.,* Chih-Zen Chang, M.D., Aij-Lie Kwan, M.D., Ph.D., Ing-Jun Chen, Ph.D.,* and Shen-Long Howng, M.D., Ph.D. Department of Neurosurgery and *Pharmacology, Kaohsiung Medical University, Kaohsiung, Taiwan, Republic of China

Lin C-L, Lo Y-C, Chang C-Z, Kwan A-L, Chen I-J, Howng S-L. Prevention of cerebral vasospasm by a capsaicin derivative, glyceryl nonivamide, in an experimental model of subarachnoid hemorrhage. Surg Neurol 2001;55:297–301. BACKGROUND

Cerebral vasospasm after aneurysmal subarachnoid hemorrhage (SAH) remains a major complication in patients suffering from SAH. In our previous study, we reported that stimulating vascular K⫹ channel activity prevented the development of cerebral vasospasm. Recent evidence indicates that glyceryl nonivamide (GLNVA), a capsaicin derivative, has a vasorelaxant effect on the aortic vascular smooth muscle due to the release of coronary calcitonin gene-related peptide, which in turn stimulates K⫹ channel opening. The purpose of the present study was to examine the preventive effects of GLNVA on vasospasm. METHODS

New Zealand white rabbits were subjected to experimental SAH by injecting autologous blood into the cisterna magna. GLNVA or vehicle was injected intrathecally immediately after the induction of SAH. All animals were killed by perfusion-fixation at 48 hours after SAH. The basilar arteries were removed and sectioned, and their cross-sectional areas were measured. RESULTS

The average cross-sectional areas of basilar arteries were reduced by 69% and 71% in the SAH only and SAH plus vehicle groups, respectively, when compared with the healthy controls. After treatment with 0.35, 1.75, and 3.5 mg/kg GLNVA in rabbits subjected to SAH the average cross-sectional area was decreased by 46%, 12% and 2%, respectively, when compared with the healthy controls. The protective effect of GLNVA achieved statistical significance at all dosages. Morphologically, corrugation of the internal elastic lamina of vessels was often observed Address reprint requests to: Dr. Aij-Lie Kwan, Kaohsiung Medical University, Department of Neurosurgery, No. 100, Shih-Chuan 1st Rd., Kaohsiung 80708 Taiwan, R.O.C. Received October 19, 2000; accepted February 27, 2001. © 2001 by Elsevier Science Inc. 655 Avenue of the Americas, New York, NY 10010

in the vehicle-treated group, but was not prominent in the GLNVA-treated groups or healthy controls. CONCLUSION

The findings showed that GLNVA dose-dependently attenuated cerebral vasospasm after SAH in the rabbit. These results suggest that intrathecal administration of GLNVA could be an effective strategy for preventing cerebral vasospasm after SAH. © 2001 by Elsevier Science Inc. KEY WORDS

Cerebral vasospasm, glyceryl nonivamide, subarachnoid hemorrhage.

S

ustained narrowing of large cerebral vessels remains a major complication after aneurysmal subarachnoid hemorrhage (SAH) [8 –10]. The absence of an adequate therapy for cerebral vasospasm continues to motivate preclinical and clinical studies. Capsaicin, a pungent natural product of red pepper, possesses a wide spectrum of biological activities [7]. In blood vessels, acute application results in smooth muscle relaxation and delayed contraction [5,12]. Glyceryl nonivamide (GLNVA) is a nonpungent and antinociceptive derivative of capsaicin [3], but without the capsaicin-derived cell excitability in blood vessel smooth muscles [13]. It is more potent as a vasorelaxant [12,13]. This effect of GLNVA is mediated via calcitonin gene-related peptide (CGRP) release-associated potassium channel opening activity. It has been shown previously that K⫹ channel activators such as cromakalim could be of benefit for preventing cerebral vasospasm after aneurysmal SAH [11]. Taken together, 0090-3019/01/$–see front matter PII S0090-3019(01)00438-4

298 Surg Neurol 2001;55:297–301

these findings indicate that GLNVA might be of value in treating cerebral vasospasm. The goal of the present study was to evaluate the potential therapeutic value of intrathecally administered GLNVA on cerebral vasospasm after SAH.

Materials and Methods MATERIALS GLNVA was synthesized in our laboratory [3]. The ventilator (model 683) used to maintain the rabbits was obtained from Harvard Apparatus (Natick, MA). The Hank’s Balanced Salt Solution (HBSS) was purchased from Sigma (St. Louis, MO). Xylazine and Keta Ved are products of Phoenix Scientific (St. Joseph, MO). The ultramicrotome (model Ultracut E) used to section basilar arteries was obtained from Reichert (Vienna, Austria). Image 1 software, available from Universal Imaging (West Chester, PA), was used in the computer-assisted tissue morphometry. GENERAL DESIGN OF EXPERIMENTS AND TREATMENT GROUPS Fifty-six male New Zealand white rabbits, weighing between 3.2 and 4.0 kg, were used in this study. Experimental SAH was induced as described below. Animals were divided into the following seven groups (8 per group): 1) control (i.e., no SAH); 2) control plus high dose GLNVA (3.5 mg/kg); 3) SAH only; 4) SAH plus vehicle; 5) SAH plus low dose GLNVA (0.35 mg/kg); 6) SAH plus medium dose GLNVA (1.75 mg/kg); and 7) SAH plus high dose GLNVA (3.5 mg/kg). GLNVA was injected intrathecally immediately after the induction of SAH. All injections were performed using a volume of 0.3 mL. All animals were killed by perfusion fixation at 48 hours after SAH. INDUCTION OF EXPERIMENTAL SAH Rabbits were anesthetized by an intramuscular injection of a mixture of 9 mg/kg xylazine and 55 mg/kg KetaVed and intubated endotracheally. A 23gauge butterfly needle was inserted percutaneously into the cisterna magna. After withdrawal of 1 mL of cerebrospinal fluid, 3 mL of nonheparinized blood from the central ear artery was injected into the subarachnoid space. The rabbits were then positioned in ventral recumbency for at least 15 minutes to facilitate the formation of blood clot in the basal cisterns. All animals were monitored closely for respiratory distress and ventilated as necessary. The animals were then allowed to recover from

Lin et al

anesthesia, extubated when awake, and returned to the vivarium. PERFUSION-FIXATION Forty-eight hours after SAH, animals were again anesthetized, intubated, ventilated, and then paralyzed with pancuronium bromide (0.3 mg/kg). The central ear artery was cannulated in order to monitor blood pressure and to determine blood gas levels. Perfusion-fixation was then performed. The thorax was opened, a cannula was placed in the left ventricle, the descending thoracic aorta clamped, and the right atrium opened. Perfusion was begun with 300 mL of HBSS, pH 7.4, followed by 200 mL of a mixture of 2% paraformaldehyde and 2.5% glutaraldehyde in HBSS at 37°C under a perfusion pressure of 120 cm H2O. The brain was then removed and immersed in the same fixative overnight at 4°C. Visual inspection during removal of the brain showed that all animals had subarachnoid clots covering the basilar artery. Two animals were eliminated from the study at this point due to inadequate perfusion-fixation: one was from the control plus high-dose GLNVA group (group 2) and the other from the SAH plus medium-dose GLNVA group (group 6). TISSUE EMBEDDING Arterial segments were removed from the middle third of each basilar artery and washed several times in 0.1 mol/L phosphate buffered saline (PBS; pH 7.4). The specimens were postfixed with osmium tetroxide, rinsed, dehydrated, and embedded in Epon 812. Cross-sections of basilar arteries were cut at a thickness of 0.5 ␮m with an ultramicrotome, mounted on glass slides, and stained with toluidine blue for morphometric analysis. TISSUE MORPHOMETRY AND STATISTICAL ANALYSIS Morphometric measurements were performed by an investigator blinded to the treatment groups. At least five random arterial cross-sections from each animal were evaluated qualitatively for the extent of corrugation of the internal elastic lamina (IEL), and the cross-sectional area of each section was measured using a computer-assisted image analysis system. The areas of the five cross-sections from a given artery were averaged to provide a single value for each animal. Group data are expressed as mean ⫾ SEM. Group comparisons were performed using a one-way analysis of variance (ANOVA) with the Tukey post-hoc test. Differences were considered significant at the p ⬍ 0.05 level.

Glyceryl Nonivamide Prevents Vasospasm

Surg Neurol 299 2001;55:297–301

Results GENERAL OBSERVATIONS Before perfusion-fixation, there were no significant differences among the treatment groups in the physiological parameters recorded, including body weight, pH, PaCO2, PaO2, and mean arterial blood pressure. A thick subarachnoid clot was observed over the basal surface of the brain stem in each animal subjected to SAH. Morphologically, the basilar arteries in the SAH only and SAH plus vehicle groups exhibited substantial corrugation of the IEL (Figure 1B). Corrugation of the IEL was less prominent in animals treated with GLNVA. The vessels from the healthy control and high dose groups had similar IELs (Figure 1A,C). The cross-sectional area of basilar arteries was significantly reduced in animals subjected to SAH (Figure 2). When compared with animals in the control group, the areas in the SAH only and SAH plus vehicle groups were reduced by 69% and 71%, respectively. The SAH-induced reduction in arterial area was attenuated in a dose-dependent manner in animals treated with GLNVA (Figure 2). When compared with the control group, the average areas were reduced by 46%, 12%, and 2% in the low-, medium-, and high-dose groups, respectively. The cross-sectional areas obtained in all three GLNVA treatment groups differed significantly from that of the SAH plus vehicle group. There was no significant difference between areas in the highest dosage group and the healthy animals.

Discussion Delayed cerebral vasospasm remains an unpredictable and inadequately treated complication of aneurysmal SAH. Despite intensive research, its pathogenesis is still a matter of debate, and adequate pharmacotherapy has been elusive. Consequently, it is of considerable importance to identify and evaluate potential mechanisms and treatments for cerebral vasospasm. Capsaicin (8-methyl-N-vanillyl-6-nonenamide) is a pungent compound produced by chili peppers and related plants of the Capsicum family. It produces a wide spectrum of pharmacological effects [1,2,4,6, 14,15,17], including decrease in the heart rate caused by stimulation of sensory C-fiber to activate the parasympathetic efferent nerve [2] and increase in cardiac contractile force resulting from the release of CGRP in sensory nerves [5,6]. Capsaicin also causes vascular smooth muscle relaxation and delayed smooth muscle contraction. The contrac-

Micrographs of representative cross-sections of basilar arteries from healthy control (A), SAH animal treated with vehicle (B), and SAH rabbit treated with GLNVA (3.5 mg/kg) (C). Calibration bar ⫽ 250 ␮m.

1

tile response is most likely due to direct effects of capsaicin on vascular smooth muscle, whereas the relaxation response is attributed to the release of CGRP [5,12]. The clinical applications of capsaicin are undoubtedly limited by the vagus reflex as well as bradycardia and apnea accompanying its intravenous administration [4]. These untoward effects are attributed to autonomic and sensory stimulation [14]. In an attempt to reduce the strong pungent and

300 Surg Neurol 2001;55:297–301

Effect of intrathecal administration of GLNVA on cerebral vasospasm. GLNVA elicited a dosedependent reduction in SAH-induced vasospasm. Attenuation of SAH-induced reduction of the cross-sectional areas of basilar arteries by GLNVA was statistically significant in all three treatment groups when compared with that obtained in SAH animals treated with vehicle.

2

irritating activities of capsaicin as well as to eliminate its vasoconstrictive and other adverse effects observed in vivo, structural modification was carried out and GLNVA was synthesized [3,14]. GLNVA is a relatively nonpungent analogue of capsaicin with a masked phenolic hydroxyl group, and induces different cardiovascular changes from those of the parent compound. For example, GLNVA only causes a monophasic depressor response, without the undesirable vagus reflex and vasoconstrictive activity associated with capsaicin [14]. In addition, the mechanism of action of GLNVA on vascular smooth muscle is significantly different from that of capsaicin. GLNVA plays an important role in increasing K⫹ permeability, thus resulting in vasorelaxation [13]. Lo et al reported that GLNVA caused an endothelium-independent vasorelaxant effect [12]. The relaxation of vascular smooth muscle induced by GLNVA was blocked by ouabain, a Na⫹-K⫹ ATPase inhibitor, and also by glibenclamide, an ATP-sensitive K⫹ channel blocker [12]. These results indicate that the ATP-sensitive K⫹ channel activity is involved in the relaxation response of GLNVA, in contrast to the capsaicin-associated cell excitability [12]. Activation of ATP-sensitive K⫹ channels by GLNVA may also be due to CGRP released from perivascular nerves [16]. Thus, GLNVA contributes to relaxation of vascular smooth muscle via both direct and indirect pathways. It has been suggested that K⫹ channel openers might be of value in the management of cerebral

Lin et al

vasospasm [18]. Using electrophysiological recording techniques, Zuccarello and associates have shown that the ATP-sensitive K⫹ channel opener cromakalim could reverse the depolarization of rabbit smooth muscle cells obtained from spastic basilar arteries [19]. Furthermore, topical application of cromakalim relaxes vessels undergoing spasm after SAH [19]. These findings support the concept that activation of the K⫹ channel may be of therapeutic value in vasospasm. Recently, we have demonstrated that systemically administered cromakalim is effective in blocking the development of basilar artery vasospasm in a rabbit model of experimental SAH [11]. Thus, potential clinical value of GLNVA for the treatment of cerebral vasospasm seems promising due to its K⫹ channel opening activity. It remains to be determined whether or not GLNVA can be utilized for systemic treatment in the prevention and/or reversal of SAH-induced vasospasm. This work was supported in part by research grants NSC 89-2314-B-037-062 and NSC 88-2314-B-037-012 from the National Science Council of the Executive Yuan, Taiwan, R.O.C. We appreciate the assistance of Ling-Chin Tsai and Shiou-Jen Wu for their help in preparation of the manuscript.

REFERENCES 1. Buck SH, Burks TF. The neuropharmacology of capsaicin: review of some recent observations. Pharmacol Rev 1986;38:179 –226. 2. Chahl LA, Lynch AM. The acute effects of capsaicin on the cardiovascular system. Acta Physiol Hung 1987; 69:413–9. 3. Chen IJ, Yang JM, Yeh JL, Wu BN, Lo YC, Chen SJ. Hypotensive and antinoceptive effects of ether-linked and relatively non-pungent analogues of N-nonanoyl vanillylamide. Eur J Med Chem 1992;27:187–92. 4. Donnerer J, Lembeck F. Analysis of the effect of intravenously injected capsaicin in the rat. NaunynSchmiedeberg Arch Pharmacol 1982;320:54 –7. 5. Franco-Cereceda A, Laudberg JM, Saria A, Schreibmayer W, Tritthart HA. Calcitonin gene-related peptide: release by capsaicin and prolongation of the action potential in the guinea-pig heart. Acta Physiol Scand 1988;132:181–90. 6. Franco-Cereceda A, Saria A, Lundberg JM. Differential release of calcitonin gene-related peptide and neuropeptide Y from the isolated heart by capsaicin, ischemia, nicotine, bradykinin and ouabin. Acta Physiol Scand 1989;135:173– 87. 7. Holzer P. Capsaicin: cellular targets, mechanisms of action, and selectivity for their sensory neurons. Pharmacol Rev 1991;43:143–201. 8. Kassell NF, Peerless SJ, Drake CG, Boarini DJ, Adams HP. Treatment of ischemic deficits from cerebral vasospasm with high dose barbiturate therapy. Neurosurgery 1980;7:593–7. 9. Kassell NF, Drake CG. Review of the management of saccular aneurysm. Neurol Clin 1983;1:73– 86.

Glyceryl Nonivamide Prevents Vasospasm

10. Kassell NF, Sasaki T, Colohan ART, Nazar G. Cerebral vasospasm following aneurysmal subarachnoid hemorrhage. Stroke 1985;16:562–72. 11. Kwan AL, Lin CL, Yanamoto H, Howng SL, Kassell NF, Lee KS. Systemic administration of the potassium channel activator cromakalim atteuates cerebral vasospasm after experimental subarachnoid hemorrhage. Neurosurgery 1998;42:347–51. 12. Lo YC, Wu SN, Su JR, Chen IJ. Effect of capsaicin on membrane currents in cultured vascular smooth muscle cells of rat aorta. Eur J Pharmacol 1995;292: 321– 8. 13. Lo YC, Wu JR, Wu SN, Chen IJ. Glyceryl nonivamide: a capsaicin derivative with cardiac calcitonin generelated peptide releasing K⫹ channel opening and vasorelaxant properties. J Pharmacol Exp Ther 1997; 281:253– 60. 14. Lo YC, Yeh JL, Wu JR, Yang JM, Chen SJ, Chen IJ. Automonic and sensory cardiovascular activities of nonivamide: intrathecal administration of clonidine. Brain Res Bull 1994;35:15–22. 15. Monsereenusorn Y, Kongsamut S, Pezalla PD. Capsaicin: a literature survey. CRC Crit Rev Toxicol 1982; 10:321–39. 16. Nelson MT, Huang Y, Brayden JE, Hescheler J, Standen NB. Arterial dilations in response to calcitonin gene-related peptide involve activation of K⫹ channels. Nature 1990;344:770 –3. 17. Virus RM, Gebhart GF. Pharmacologic actions of capsaicin: apparent involvement of substance P and serotonin. Life Sci 1979;25:1273– 84. 18. Zhang H, Cook D. Cerebral vascular smooth muscle potassium channels and their possible role in the

Surg Neurol 301 2001;55:297–301

management of vasospasm. Pharmacol Toxicol 1994; 75:327–36. 19. Zuccarello M, Bonasso CL, Lewis AI, Sperelakis N, Rapoport RM. Relaxation of subarachnoid hemorrhageinduced spasm of rabbit artery by the K⫹ channel activator cromakalim. Stroke 1996;27:311– 6. COMMENTARY

This well-written article describes a dose-dependent improvement in basilar artery spasm in a rabbit SAH model with intrathecal administration, immediately after induction of hemorrhage, of the capsaicin derivative GLNVA. With the highest dose there was essentially no difference between the crosssectional area of the artery compared with normal controls. This is potentially important work, in view of the relative lack of adverse effects of GLNVA compared with capsaicin itself. It would be worthwhile, as mentioned in the discussion, to see if this drug has a similar effect with systemic administration. Also, it would be nice to know if its effect continued beyond the 48 hours in this study—say, for 7 days or more. Nicholas Dorsch, FRCS, FRACS Department of Surgery Westmead Hospital Sydney, Australia

very road has two directions.

E

—Russian proverb