High-dose intraarterial verapamil in the treatment of cerebral vasospasm after aneurysmal subarachnoid hemorrhage

J Neurosurg 108:458 463, 2008 High-dose intraarterial verapamil in the treatment of cerebral vasospasm after aneurysmal subarachnoid hemorrhage JANINE KEUSKAMP, M.D., 1 RAJ MURALI, M.D., 2 AND KUO H. CHAO, M.D., M.P.H. 1 Departments of 1 Radiology and 2 Neurosurgery, New York Medical College, Valhalla, New York Object. Because oral calcium channel blockers appear to reduce the severity of cerebral vasospasm after aneurysmal subarachnoid hemorrhage (SAH), interest in their applica intraarterially has emerged for cases in which noninvasive means of alleviating vasospasm are unsuccessful. Studies to date have been limited to the administra of low intraarterial doses because of concerns about hemodynamic stability and changes in intracranial pressure. These doses, although effective in cases of milder vasospasm, were inadequate in severe cases. The authors present a series of 10 patients with cerebral vasospasm who underwent 12 procedures in which they received 20 mg of intraarterial verapamil per procedure. Methods. A retrospective review was undertaken of all patients who underwent endovascular treatment for cerebral vasospasm due to aneurysmal SAH by the senior author between February 2005 and October 2006. Ten patients were identified who had undergone a total of 12 procedures during which 20 mg of intraarterial verapamil had been administered. From angiography reports, anesthesia records, and nursing records, we obtained pre- and postverapamil mean arterial blood pressures (MABPs), heart rates, intracranial pressures (ICPs) (when available), and visible changes in the degree of vasospasm. Results. No statistically significant changes in MABP, heart rate, or ICP were observed after administra of 20 mg of intraarterial verapamil, and the degree of improvement in vasospasm was statistically significant based on our grading system. No correla was found between the change in hemodynamic parameters and the total dose of verapamil. Conclusions. This study indicates that high-dose intraarterial verapamil may be used to treat cerebral vasospasm without compromising hemodynamic stability or increasing ICP. (DOI: 10.3171/JNS/2008/108/3/0458) KEY WORDS cerebral vasospasm subarachnoid hemorrhage verapamil N Abbrevias used in this paper: CPP = cerebral perfusion pressure; ICP = intracranial pressure; MABP = mean arterial blood pressure; SAH = subarachnoid hemorrhage. EUROLOGICAL deficits incurred as a result of cerebral vasospasm after SAH can sometimes be more devastating than those resulting from the initial aneurysmal rupture. Although noninvasive management via hypervolemia, hypertension, and hemodilu (triple-h therapy) has become widely accepted as a first-line mode of treatment, endovascular treatment of vasospasm can sometimes be successful in cases in which medical management is not effective. The advent of angioplasty and its applica to the cerebral vasculature have improved the ability to treat vasospasm of the more proximal vessels. However, the existing ops for endovascular treatment of vasospasm in the smaller, more distal arteries incorporate chemical rather than mechanical means of dila. Although several oral and intraarterial agents have shown promise as candidates for intraarterial use in relieving vasospasm, most have either potentially dangerous or unpredictable side effects, or have produced improvement that is too transient in nature to be considered clinically effective. 8 10,13 16 In light of the reported success that oral nimodipine has had in limiting cerebral vasospasm, much interest has emerged regarding the intraarterial selective applica of calcium channel blockers during cerebral angiographic studies. 1 7,12 Preliminary results from studies assessing the efficacy and safety of selective intraarterial injec of calcium channel blockers have been encouraging. Compared with papaverine, for example, calcium channel blockers appear to have a more prolonged effect on arterial dila and have not been associated with increases in ICP. 2,4 6 Furthermore, studies have indicated that the sustained effect on vasodila has been achieved without sacrificing systemic hemodynamic stability. Although these results are promising, they are based on the use of calcium channel blockers at doses that proved to be insufficient in some cases of severe vasospasm. 5 We studied a series of patients in whom the starting dose of verapamil was 5 mg per vessel and the total dose per procedure was 20 mg. Because the most common dose- 458 J. Neurosurg. / Volume 108 / March 2008

High-dose intraarterial verapamil for cerebral vasospasm limiting factor seemed to be concern about adverse systemic effects, we focused on changes in MABP, heart rate, and whether any interven was required. Addially, due to the reported changes in ICP seen with other agents, such as intraarterial papaverine, we sought to assess whether verapamil would have similar side effects at higher doses. 9,13 Our final considera was whether higher doses of verapamil resulted in a greater reduc in the degree of vasospasm, as determined by angiographic evalua. Clinical Material and Methods We reviewed the charts of all 10 patients who underwent endovascular treatment performed by the senior author (K.C.) for vasospasm due to aneurysmal SAH between February 2005 and October 2006. Each of these patients received 20 mg of intraarterial verapamil in a total of 12 procedures because lower doses were not effective. Angioplasty, a higher risk procedure, was considered only if verapamil did not achieve optimum arterial dila. In each procedure, verapamil was selectively administered in increments of 5 mg through a 5- or 6-F guiding catheter posied in the proximal cervical internal carotid artery and/ or vertebral artery depending on the loca of vasospasm. A follow-up angiographic study was performed 1 2 minutes after each bolus was given to evaluate for improvement. Addial 5-mg doses were given until it was believed that the maximum benefit had been achieved. For each patient, age, sex, Hunt and Hess grade, Fisher grade, and aneurysm size and loca were recorded. One of the 10 patients had undergone 3 separate procedures for endovascular treatment of vasospasm, and 20 mg of verapamil was administered during each procedure. From angiography reports, anesthesia records, and nursing records, we obtained informa regarding the pre- and postverapamil MABP, heart rate, ICP when available, and visible changes in the degree of vasospasm. We were also able to calculate the verapamil dose per unit time for each patient after noting the total verapamil dose given and dura of the procedure. Evalua of Hemodynamic Effects and Changes in ICP As we evaluated each record, a number of factors needed to be considered in conjunc with the patient s preand postprocedure hemodynamic status. The preangiography MABP and heart rate were defined as the values obtained by staff when the patient arrived in the angiography room. In cases in which addial seda or anesthesia was required to perform the angiography, we noted whether these had any effect on the initial MABP or heart rate and considered the values obtained after seda but before the first verapamil dose as the initial values. When possible, pre- and postprocedure ICP values were noted, so that changes in CPP could be calculated. Addially, as triple-h therapy has become the standard approach to medical management of cerebral vasospasm, some patients arrived having already been placed on a continuous intravenous infusion of neosynephrine, ephedrine, and/or dopamine. We evaluated these cases to determine if changes in the rate of infusion or addial boluses of medica were required to maintain the MABP or heart rate. Finally, we evaluated each chart to see whether any patient J. Neurosurg. / Volume 108 / March 2008 who had not previously been placed on continuous intravenous pressors required the initia of blood pressure support after verapamil administra. Evalua of Changes in Vasospasm Angiographic images and reports were reviewed for informa regarding the loca and degree of vasospasm before treatment. The degree of vasospasm before treatment was based on comparison with the original cerebral angiogram obtained at the time of admission for SAH. The verapamil dose per vessel, total verapamil dose, and change in degree of vasospasm were also recorded. For vessels requiring repeated verapamil doses during a single procedure, the individual doses were added and considered as 1 dose. Each vessel treated underwent a final angiographic run before the catheter was removed so that the maximum degree of change could be assessed. For cases in which angioplasty was also performed, the degree of change in vasospasm was based on vessels in which verapamil was the sole form of therapy or on interval runs performed between verapamil and angioplasty. The degree of vasospasm before and after verapamil was graded on the following scale of 1 to 5 by the senior author (K.C.): 1 indicated mild constric ( 20%); 2, mild moderate constric (21 40%); 3, moderate constric (41 60%); 4, moderate severe constric (61 80%); and 5, severe constric ( 81%). The median score for degree of vasospasm was calculated for both before and after treatment with verapamil. Verapamil Dose Per Minute and Rela to Hemodynamic Stability As it was our goal to evaluate the hemodynamic effects of high-dose intraarterial verapamil, it was necessary to consider the amount of verapamil administered over a given time period and whether a relaship to MABP and heart rate could be established. Because in some cases a given dose of verapamil was administered during the course of a 2-hour procedure and in others that same dose was given during a 1.5-hour procedure, we used the verapamil dose per minute to compare results from each procedure. Anesthesia records were reviewed to obtain the amount of time that lapsed between the initial preverapamil MABP and heart rate and the final MABP and heart rate values obtained after the procedure. Addially, we noted whether any transient changes in hemodynamic parameters occurred during the procedure. For each patient, a verapamil dose per minute was calculated and compared with the overall change in MABP and heart rate. In patients with external ventricular drains in place, we were also able to determine whether the dose of verapamil had any significant relaship with changes in ICP or CPP. Changes in Neurological Status Although the main focus of our investiga was to assess the safety of high-dose intraarterial verapamil administra, we also sought to determine whether it resulted in an improvement in neurological func. A chart review yielded informa regarding the neurological deficits that prompted interven and whether the patient s neurological status improved, remained unchanged, or worsened after treatment. We then compared the degree of angiograph- 459

J. Keuskamp, R. Murali, and K. H. Chao ic improvement in vasospasm with the change in neurological examina to determine whether any correla emerged. Statistical Analysis Statistical analysis of data regarding pre- and postverapamil heart rate, blood pressure, and ICP was performed using the Student t-test with significance set at p 0.05. Analysis of verapamil dose per unit time was also performed using the Student t-test with significance set at p 0.05. Analysis of angiographic vasospasm grade pre- and postverapamil treatment was performed using the Wilcoxon signed-rank test with significance set at p 0.05. Results are expressed as mean standard devia unless otherwise noted. Results Between February 2005 and October 2006, 10 patients (6 men and 4 women) underwent a total of 12 procedures for endovascular treatment of cerebral vasospasm during which 20 mg of verapamil was administered. The mean age was 48 years (range 36 59 years.) At the time of initial presenta, 1, 1, 5, and 3 patients had Hunt and Hess Grade 1, 2, 3, and 4 aneurysms, respectively (Table 1). In terms of aneurysm loca and size, 4 patients had aneurysms at the A1/A2 junc of the anterior communicating artery (size range 4 10 mm), 1 patient had an anterior communicating artery aneurysm (3 mm), 2 had a posterior communicating artery aneurysm (6 and 4 mm), 1 had a supraclinoid internal carotid artery aneurysm (16 mm), 1 had a pericallosal artery aneurysm (4 mm), and 1 patient had a middle cerebral artery bifurca aneurysm (6 mm). The mean verapamil dose administered per procedure was 41 29 mg (range 20 120 mg). Hemodynamic Stability and ICP TABLE 1 Patient demographics, aneurysm size and loca, Hunt & Hess and Fisher grades, posthemorrhage day at vasospasm treatment* Case Age (yrs), Aneurysm Aneurysm H & H Fisher PHD When No. Sex Loca Size (mm) Grade Grade Treated 1 50, F lt A 1 /A 2 junc- 5 3 15 2 36, M lt PCoA 6 3 11 3 59, F rt A 1 /A 2 junc- 5 3 3 11 4 53, M rt ICA bifur- 16 4 6 ca 5 46, F lt A 1 /A 2 junc- 4 1 3 3 6 43, M lt ACoA 3 2 3 6 7 56, F lt MCA 6 4 4 7 8 44, F rt A 1 /A 2 junc- 10 3 4 10 9 52, F rt pericallosal 4 4 3 4 10 45, M lt PCoA 4 3 3 6 * ACoA = anterior communicating artery; H & H = Hunt and Hess; ICA = internal carotid artery; MCA = middle cerebral artery; PCoA = posterior communicating artery; PHD = posthemorrhage day of endovascular treatment with verapamil. Fisher grades for cases 1, 2, and 4 could not be obtained during chart review, and initial head computed tomography scans were unavailable. Procedures 4, 5, and 6 were performed in the patient in Case 4 in subsequent tables. At the onset of each of the 12 procedures, the mean heart rate was 92 23 bpm and the mean MABP was 106 13 mm Hg. All patients displayed a MABP 80 mm Hg, and only 1 patient was bradycardic with a heart rate of 50 bpm. The postprocedure values were a mean heart rate of 92 14 bpm and a mean MABP of 103 14 mm Hg, with no patient displaying an MABP 80 mm Hg. The mean change in heart rate was 0 15 bpm, and the mean change in MABP was 3 15 mm Hg (Table 2). Neither the change in heart rate (p = 0.49) nor the change in MABP (p = 0.24) was statistically significant. The patient who had initially presented with bradycardia had a postprocedure heart rate of 65 bpm, and no new cases of bradycardia were found. Our review of anesthesia records also revealed that, although a few patients experienced a transient decrease in heart rate or blood pressure, these decreases lasted for 5 minutes at most, and no patient became hypotensive or newly bradycardic. The patient who presented with bradycardia before treatment did not display any further decrease in heart rate during treatment with verapamil. In 6 procedures, the ICP measurement was made possible by the presence of an external ventricular drainage catheter (Table 3). In 5 of the 6 cases, the patient s ICP remained at or lower than 15 mm Hg throughout the entire procedure, and 1 patient presented with an ICP of 20 mm Hg and experienced no change after the procedure. The mean change between pre- and postprocedure ICP values was 0.17 2.8 mm Hg, which was not statistically significant (p = 0.37). Based on the pre- and posttreatment ICP and MABP values, the mean CPP values before and after verapamil were found to be 89 17 and 89 9 mm Hg, respectively. The mean change in CPP was 0.67 17 mm Hg, which was not statistically significant (p = 0.46). Before 7 of the 12 procedures, the patients were being treated with continuous infusion of either neosynephrine or dopamine as one component of triple-h therapy. In 3 of 7 cases, no change in the rate of infusion was required after endovascular treatment for vasospasm with high-dose verapamil. One of 7 patients required an increase in the rate of continuous dopamine from 4 g/kg/min to 15 g/kg/ min, so that the blood pressure could be increased from a normotensive to a hypertensive level as part of triple-h therapy. The patient who required 3 separate procedures for treatment of vasospasm initially presented with a Hunt and Hess Grade 4 aneurysm and accounts for 3 of the 7 procedures performed in patients who required pressors before intraarterial verapamil. After the first procedure, the patient required a lower dose of neosynephrine but a higher dose of dopamine. Before the second procedure, the patient was on continuous dopamine only, required a bolus of neosynephrine during the procedure, and required a higher rate of continuous dopamine infusion after the procedure. During the third procedure, the patient received a one-time bolus of neosynephrine but eventually required a lower rate of continuous dopamine infusion following treatment with a total of 120 mg of intraarterial verapamil. Throughout all 3 procedures, the patient never experienced any episodes of hypotension, sustained MABP 80 mm Hg, or bradycardia. Of the 12 procedures reviewed, there were no cases in which a patient required initia of continuous neosyn- 460 J. Neurosurg. / Volume 108 / March 2008

High-dose intraarterial verapamil for cerebral vasospasm TABLE 2 Hemodynamic parameters before and after high-dose verapamil* Heart Rate (bpm) MABP (mm Hg) Procedure No. Preverapamil Postverapamil Change Preverapamil Postverapamil Change 1 68 90 22 113 93 20 2 80 70 10 96 88 8 3 120 108 12 112 105 7 4 115 90 25 105 95 10 5 92 92 0 85 105 20 6 118 110 8 100 97 3 7 69 95 26 121 125 4 8 50 65 15 118 101 17 9 115 112 3 118 97 21 10 94 93 1 105 91 14 11 85 87 2 83 102 19 12 96 91 5 116 134 18 mean SD 92 23 92 14 0 15 106 13 103 14 3 15 p value 0.49 0.24 * SD = standard devia. ephrine or dopamine infusion during or after intraarterial verapamil therapy. Improvement in Vasospasm In 10 of 12 procedures performed, a decrease in the vasospasm score was observed (Fig. 1, Table 4), with the median change in score equal to 2. No patient experienced an increase in vasospasm, and the patient who underwent 3 separate procedures experienced no change in score after the first and second procedures, with a change in score of 1 following the third procedure. Based on the Wilcoxon signed-rank test, with the number set at 10 and a t value of 0, the change in score was statistically significant. Verapamil Dose Per Minute and Relaship to Hemodynamic Stability Whereas the mean verapamil dose per procedure was calculated at 41 29 mg, the mean verapamil dose over time was 0.24 0.09 mg per minute. When the verapamil dose per minute was compared with changes in postprocedure MABP and heart rate, no discernible pattern emerged. For example, 1 patient who received 0.36 mg/minute of verapamil displayed a postprocedure MABP that was 8 mm Hg lower than the initial MABP, and another patient who received 0.17 mg/minute had a 14 mm Hg decrease in MABP postprocedure. Similarly, 1 patient received 0.21 mg/minute of verapamil and exhibited a decrease in heart rate of 12 bpm postprocedure, while another patient who received 0.21 mg/minute experienced a 15-bpm increase in heart rate. As was the case for MABP and heart rate, no relaship was found between verapamil dose per minute and ICP in those cases where ventricular drains were in place. Change in Neurological Status The neurological deficits that prompted endovascular treatment improved after 8 of 12 procedures and remained unchanged after 4 of 12 procedures (Table 5). Varia in the documenta of postprocedure neurological status precluded us from making a valid assessment regarding the degree to which each patient improved. As a result, a correla between the degree of improvement in vasospasm and the degree of improvement neurologically could not be gauged. None of the patients displayed further neurological decline following our interven. In the patient in Case 4, who underwent 3 procedures, only the third treatment resulted in an angiographic improvement in vasospasm by 1 grade, and none of these treatments resulted in neurological improvement. In Case 7 (Procedure 9), the degree of angiographic vasospasm improved by 3 grades; however, no neurological improvement was observed. TABLE 3 Intracranial pressure before and after high-dose verapamil ICP (mm Hg) CPP (mm Hg) Procedure No. Preverapamil Postverapamil Change Preverapamil Postverapamil Change 3 4 2 2 108 103 5 4 20 20 0 85 75 10 5 15 15 0 70 90 20 6 9 13 5 91 88 3 9 10 9 1 108 88 20 11 13 10 3 70 92 22 mean SD 12 5.5 12 6.1 0.17 2.8 89 17 89 9.0 0.67 17 p value 0.37 0.46 J. Neurosurg. / Volume 108 / March 2008 461

J. Keuskamp, R. Murali, and K. H. Chao FIG. 1. Case 10. Angiograms obtained before (left) and after (right) administra of a high-dose (30 mg) verapamil injec into the left vertebral artery, showing a change in vasospasm. Discussion Previous studies have suggested that intraarterial calcium channel blockers, with or without angioplasty, can be an effective and relatively safe means of treating cerebral vasospasm. 2,4 6 However, concerns regarding the ability to maintain hemodynamic stability have limited prior use to relatively low doses. Individual cases have been described in which doses as high as 8 mg per vessel were used, and some authors have proposed that higher doses may be more effective in treating more distal vessels with severe vasospasm. 5 In a recent study, Mazumdar et al. 11 found that total doses of intraarterial verapamil ranging from 2.5 to 10 mg did not significantly increase the diameter of vasospastic vessels. In our series of 12 procedures, we found that intraarterial verapamil can be administered at an average dose of 41 29 mg per patient during a given procedure without producing any clinically significant change in MABP TABLE 4 Improvement in vasospasm after verapamil treatment Vasospasm Grade* Procedure Verapamil Dose/ No. Preverapamil Postverapamil Change Time (mg/min) 1 5 1 4 0.36 2 5 1 4 0.36 3 5 3 2 0.21 4 4 4 0 0.25 5 4 4 0 0.19 6 4 3 1 0.30 7 5 2 3 0.20 8 4 1 3 0.21 9 4 1 3 0.25 10 4 2 2 0.17 11 3 1 2 0.08 12 5 1 4 0.36 * Grading of angiographic vasospasm: 1, mild; 2, mild moderate; 3, moderate; 4, moderate severe; 5, severe. Median change in grade: 2. Mean dose/time: 0.24 0.09 mg/minute. or heart rate. Some patients received as much as 20 mg of verapamil in 1 vessel without experiencing hemodynamic instability. In some cases, intraarterial verapamil was used in conjunc with angioplasty in areas with severe spasm that were accessible via balloon catheter. In these cases, we based our assessment of the degree of improvement on vessels that had been treated only with verapamil. We achieved good results in proximal vessels that did not necessarily require angioplasty or in cases in which tortuosity of the vasculature precluded the use of a balloon catheter. Of the patients who displayed angiographic and neurological improvement after verapamil administra, not 1 required repeated endovascular therapy. As mened, 1 of the major disadvantages observed with papaverine has been that its effects are transient, thus requiring repeated angiography and infusion. 9 Although some degree of vasospasm may have recurred in those cases successfully treated with high-dose verapamil, none of the patients displayed recurrence or worsening of clinical symptoms. Therefore, any cases of vasospasm that might have recurred after initial improvement failed to be clinically significant. Al- TABLE 5 Neurological status before and after verapamil treatment Procedure Presenting Postverapamil No. Neurological Deficit Neurological Status 1 rt facial droop & pronator drift improved 2 aphasia improved 3 lt facial droop & pronator drift improved 4 lt hemiparesis unchanged 5 lt hemiparesis unchanged 6 lt hemiparesis unchanged 7 disorienta, lethargy improved 8 disorienta improved 9 rt hemiparesis unchanged 10 lt lower-extremity weakness improved 11 bilat lower-extremity weakness improved 12 rt hemiparesis improved 462 J. Neurosurg. / Volume 108 / March 2008

High-dose intraarterial verapamil for cerebral vasospasm though daily transcranial Doppler monitoring may be useful in detecting more subtle changes in vessel caliber, the relevance of these changes becomes quesable in the absence of clinical manifestas. It is also unclear whether verapamil resulted in sustained relief of clinically detectable vasospasm due to its mode of ac or the high doses administered in our case series. Further research based on a much larger patient pool, matching patients who receive either low- or high-dose verapamil, may help to elucidate whether a dose response relaship exists between verapamil and the dura of its effects on vasospasm. In patients with SAH, triple-h therapy has become the standard mode of medical treatment when patients begin to display signs of vasospasm. Therefore, concerns regarding the use of verapamil stem from the need to maintain not only hemodynamic stability but also a relatively hypertensive state. Addially, the presence of cerebral edema in some patients further emphasizes the importance of maintaining adequate cerebral blood flow. In those patients for whom CPP could be calculated, the lowest postverapamil value obtained was 75 mm Hg, which was well within the range of safety. Of those patients who were not receiving intravenous vasopressors before verapamil administra, not one experienced a decrease in blood pressure that required their use after the procedure. In those cases in which addial vasopressors were used, blood pressure never reached levels of true hypotension and was manipulated back to desired levels with relative ease. Some patients who displayed improvement in clinical symptoms following treatment with intraarterial verapamil experienced a decrease in MABP 10 mm Hg, suggesting that the improvement in cerebral blood flow after verapamil therapy was able to compensate for the observed decrease in MABP. Although we were able to assess the relative safety of high-dose verapamil use, the retrospective nature of our study limited our ability to better characterize the relaship between angiographic results and clinical improvement. Because no patient experienced a decline in neurological func, and hemodynamic stability was not compromised, our study suggests that high-dose verapamil is unlikely to be detrimental and may result in neurological improvement. Conclusions We used high-dose intraarterial verapamil to treat cerebral vasospasm caused by aneurysmal SAH in 10 patients during a total of 12 procedures, without the occurrence of any clinically significant changes in hemodynamic stability, ICP, or CPP. Neurological improvement was documented after 8 of 12 procedures. Acknowledgment We thank Ms. Arlene Stolper Simon for excellent editorial assistance. References 1. Allen GS, Ahn HS, Preziosi TJ, Battye R, Boone SC, Chou SN, et al: Cerebral arterial spasm: a controlled trial of nimodipine in patients with subarachnoid hemorrhage. N Engl J Med 308: 619 624, 1983 2. Badjatia N, Topcuoglu MA, Pryor JC, Rabinov JD, Ogilvy CS, Carter BS, et al: Preliminary experience with intra-arterial nicardipine as a treatment for cerebral vasospasm. AJNR Am J Neuroradiol 25:819 826, 2004 3. Barker FG II, Ogilvy CS: Efficacy of prophylactic nimodipine for delayed ischemic deficit after subarachnoid hemorrhage: a metaanalysis. J Neurosurg 84:405 414, 1996 4. Biondi A, Ricciardi GK, Puybasset L, Abdennour L, Longo M, Chiras J, et al: Intra-arterial nimodipine for the treatment of symptomatic cerebral vasospasm after aneurysmal subarachnoid hemorrhage: preliminary results. AJNR Am J Neuroradiol 25:1067 1076, 2004 5. Feng L, Fitzsimmons BF, Young WL, Berman MF, Lin E, Aagaard BD, et al: Intraarterially administered verapamil as adjunct therapy for cerebral vasospasm: safety and 2-year experience. AJNR Am J Neuroradiol 23:1284 1290, 2002 6. Hui C, Lau KP: Efficacy of intra-arterial nimodipine in the treatment of cerebral vasospasm complicating subarachnoid haemorrhage. Clin Radiol 60:1030 1036, 2005 7. Karinen P, Koivukangas P, Ohinmaa A, Koivukangas J, Ohman J: Cost-effectiveness analysis of nimodipine treatment after aneurysmal subarachnoid hemorrhage and surgery. Neurosurgery 45: 780 785, 1999 8. Kwan AL, Lin CL, Yen CP, Winardi W, Su YF, Winardi D, et al: Preven and reversal of vasospasm and ultrastructural changes in basilar artery by continuous infusion of CGS 35066 following subarachnoid hemorrhage. Exp Biol Med 231:1069 1074, 2006 9. Liu JK, Couldwell WT: Intra-arterial papaverine infusions for the treatment of cerebral vasospasm induced by aneurysmal subarachnoid hemorrhage. Neurocrit Care 2:124 132, 2005 10. Lynch JR, Wang H, McGirt MJ, Floyd J, Friedman AH, Coon AL, et al: Simvastatin reduces vasospasm after aneurysmal subarachnoid hemorrhage: results of a pilot randomized clinical trial. Stroke 36:2024 2026, 2005 11. Mazumdar A, Rivet DJ, Derdeyn CP, Cross DT III, Moran CJ: Effect of intraarterial verapamil on the diameter of vasospastic intracranial arteries in patients with cerebral vasospasm. Neurosurg Focus 21(3):E15, 2006 12. Öhman J, Heiskanen O: Effect of nimodipine on the outcome of patients after aneurysmal subarachnoid hemorrhage and surgery. J Neurosurg 69:683 686, 1988 13. Smith WS, Dowd CF, Johnston SC, Ko NU, DeArmond SJ, Dillon WP, et al: Neurotoxicity of intra-arterial papaverine preserved with chlorobutanol used for the treatment of cerebral vasospasm after aneurysmal subarachnoid hemorrhage. Stroke 35: 2518 2522, 2004 14. Tanaka K, Minami H, Kota M, Kuwamura K, Kohmura E: Treatment of cerebral vasospasm with intra-arterial fasudil hydrochloride. Neurosurgery 56:214 223, 2005 15. Tseng MY, Czosnyka M, Richards H, Pickard JD, Kirkpatrick PJ: Effects of acute treatment with pravastatin on cerebral vasospasm, autoregula, and delayed ischemic deficits after aneurysmal subarachnoid hemorrhage: a phase II randomized placebo-controlled trial. Stroke 36:1627 1632, 2005 16. Wong GK, Chan MT, Boet R, Poon WS, Gin T: Intravenous magnesium sulfate after aneurysmal subarachnoid hemorrhage: a prospective randomized pilot study. J Neurosurg Anesthesiol 18: 142 148, 2006 Manuscript submitted April 3, 2007. Accepted July 11, 2007. Address correspondence to: Kuo Chao, M.D., Department of Radiology, Westchester Medical Center, Valhalla, New York 10009. email: chaok@wcmc.com. J. Neurosurg. / Volume 108 / March 2008 463