Effects of Acute Treatment With Pravastatin on Cerebral Vasospasm, Autoregulation, and Delayed Ischemic Deficits After Aneurysmal Subarachnoid Hemorrhage
A Phase II Randomized Placebo-Controlled Trial
Background and Purpose— Statins may improve cerebral vasomotor reactivity through cholesterol-dependent and -independent mechanisms. A phase II randomized controlled trial was conducted to examine the hypothesis that acute pravastatin treatment could improve cerebrovascular autoregulation and reduce vasospasm-related complications after aneurysmal subarachnoid hemorrhage (SAH).
Methods— A total of 80 aneurysmal SAH (aSAH) patients (18 to 84 years of age) within 72 hours from the ictus were randomized equally to receive either oral pravastatin (40 mg) or placebo daily for up to 14 days. Primary end points were the incidence, duration, and severity of cerebral vasospasm, and duration of impaired autoregulation estimated from transcranial Doppler ultrasonography. Secondary end points were the incidence of vasospasm-related delayed ischemic deficits (DIDs) and disability at discharge.
Results— Prerandomization characteristics were balanced between the 2 groups. No treatment-related complication was observed. The incidences of vasospasm and severe vasospasm were reduced by 32% (P=0.006) and 42% (P=0.044), respectively, and the duration of severe vasospasm was shortened by 0.8 days (P=0.068) in the pravastatin group. These measurements were maximal on the ipsilateral side of ruptured aneurysms. The duration of impaired autoregulation was shortened bilaterally (P≤0.01), and the incidence of vasospasm-related DIDs and mortality were decreased by 83% (P<0.001) and 75% (P=0.037), respectively, in the pravastatin group.
Conclusion— Acute treatment with pravastatin after aSAH is safe and ameliorates cerebral vasospasm, improves cerebral autoregulation, and reduces vasospasm-related DID. Unfavorable outcome at discharge was reduced primarily because of a reduction in overall mortality. This is the first demonstration of clinical benefits with immediate statin therapy for an acute cerebrovascular disorder.
Aneurysmal subarachnoid hemorrhage (SAH) affects 10 per 100 000 population in the United Kingdom each year. For survivors of the initial hemorrhage, cerebral vasospasm and related delayed ischemic deficits (DIDs) are the major causes of subsequent morbidity and mortality.1,2 Of the many pharmaceutical attempts to reverse vasospasm-related DIDs, only nimodipine and triple-H therapy (hypertensive hypervolemic hemodilution) have proven utility.3,4
Other potential candidates include statins.5 The neuroprotective properties of statins may include improvement in cerebral vasomotor reactivity and fibrinolytic activity combined with reduced thrombogenic potential, platelet activation, and cytokine responses to cerebral ischemia.6,7 Of interest in the context of the pathophysiology of aneurysmal SAH (aSAH) is the ability of statins associated with upregulation of endothelial NO synthase to improve vasomotor reactivity, increase cerebral blood flow (CBF), and potentially modulate the development of cerebral proliferative vasculopathy.8 Statins given 24 hours after experimental cerebral ischemia can enhance CBF, angiogenesis, neurogenesis, and synaptogenesis and improve neurological recovery.9 Thus, acute statin therapy might be beneficial after aSAH, when there is a potential therapeutic window of several days between the ictus and the development of cerebral vasospasm and related DID.2 In this study, we examined the hypothesis that acute pravastatin therapy could improve cerebral autoregulation and reduce the incidence of vasospasm-related complications after aSAH using a phase II randomized controlled trial.
Approval was granted from the local research ethics committee (No. 03/353) and the Medicines and Healthcare Products Regulatory Agency (No. MF/8000/13164). Adult patients (18 to 84 years of age) with aSAH admitted to the neurosurgical department at Addenbrooke’s National Health Services were recruited. Baseline data included age, gender, medical history, and initial aSAH grade according to the World Federation of Neurological Surgeons (WFNS).10 Radiological information included the Fisher grade on computerized tomography (CT),11 presence of hydrocephalus or intraventricular hemorrhage, and aneurysm location on cerebral angiography. Exclusion criteria were nonaneurysmal SAH, pregnancy, preictal statin therapy, and contraindications to statin use (eg, history of liver or renal dysfunction or alanine aminotransferase [ALT] >50 U/L).
After admission, the clinical management of each patient was standard for the department, including nimodipine and moderate intravenous fluid supplement.12 For symptomatic vasospasm salvage, triple-H therapy was commenced. None of these patients underwent endovascular angioplasty (balloon angioplasty or endovascular vasodilator therapy). Factors that might affect the outcome were documented, including external ventricular drainage, ventriculitis, sepsis, mode of aneurysm treatment (endovascular versus clipping), and immediate postoperative deficits.
Randomization and Drug Administration
Identical capsules containing either pravastatin sodium 40 mg (Lipostat; Bristol-Myers Squibb) or placebo (lactose) were provided by the Ipswich Pharmacy Manufacture Unit (PMU). The PMU used simple randomization to number bottles containing either pravastatin or placebo. Each bottle was assigned to consenting patients on enrollment, with the contents known only by the PMU. Trial medications were commenced within 72 hours of ictus for up to 14 days or until discharge. Adverse clinical events were reported to the safety officer and study statistician (H.R.).
Primary End Point Measures
The predetermined primary end points were: (1) incidence, severity, and duration of vasospasm on transcranial Doppler (TCD) indices; and (2) duration of impaired cerebral autoregulation.
Secondary End Point Measures
Secondary end points were: (1) incidence of vasospasm-related DIDs; and (2) disability at discharge.
Daily TCD (DWL Multi-Dop X4) using a 2-MHz probe mounted on a purpose-designed head frame was adopted as a surrogate measure of cerebral vasospasm and performed through the temporal window by a single user (M.-Y.T.). Systolic, diastolic, and mean flow velocities (FVs) in both middle cerebral arteries were captured (50 to 60 mm depth) and recorded via the in-house designed BioSan.13 Vasospasm was defined as >1.2 m/s with Lindegaard ratio (LR) >3. This definition has been shown to have a high negative predictive value (94%) for significant vasospasm on cerebral angiography.14 Severe vasospasm was defined as >2.0 m/s with LR >3, which has a high positive predictive value (87%) for severe vasospasm on cerebral angiography.14
Determining Cerebral Autoregulatory Status
The detailed method of assessing cerebral autoregulation using the transient hyperemic response test (THRT) has been described in previous publications.13,15 Briefly, after a minimum of a 10-s baseline data recording, 2 carotid compressions lasting 5 s with a 2-minute interval were performed. The criteria for an acceptable THRT included a sudden decrease in middle cerebral artery FV at the onset of compression, a stable TCD signal during compression, and a minimum of 30% decrease in FV with no blood pressure instability. The THRT ratio (THRR) is calculated as THRR=FVsystolic (hyperemia)/FVsystolic (baseline). THRR was defined as normal (≥l.10) or impaired (<1.10).13
Definition of DID and Outcome
DID was defined as development of focal neurological deficits or a drop in the Glasgow Coma Scale by ≥2 points.3 DID was also defined as vasospasm-related if it was associated with severe vasospasm on TCD. Other possible conditions causing neurological deterioration (eg, hydrocephalus, intracerebral hemorrhage, surgical complications, metabolic abnormalities, or infection) were excluded by repeated imaging (CT, xenon CT, or cerebral angiography) and metabolic screening. Immediate postoperative deficits were defined as new neurological deficits evident on recovery from general anesthesia. Patient disability at discharge was recorded according to the Modified Rankin Scale (MRS) as favorable (MRS 1 and 2) or unfavorable (MRS 3 to 6).16,17
Detailed serum lipid profiles, including total and low-density lipoprotein (LDL) cholesterol, muscle creatine kinase (CK), and ALT were taken at admission and repeated every 3 days. Only safety and quality control issues are reported in this article.
Previous observations in our unit indicated that vasospasm defined by TCD occurred in 64% of SAH patients.18 To demonstrate a 50% reduction in incidence of vasospasm with a power of 80% and a significance level of 5%, examinations of 80 patients were required. The study was not powered to show improvement in clinical outcome, although trends were sought to identify potential effects.
All analyses were performed on an intention-to-treat basis, and P values were 2-sided. Analysis was performed using statistical software STATA Intercooled 8.0 for Windows (Texas 77845). Prerandomization and postrandomization characteristics were compared using the χ2 test (the Fisher exact test was used when any number of the cell was <5), except the age, when the t test was used. For the primary end points, analysis was performed separately to the side ipsilateral and contralateral to the ruptured aneurysm. The incidence of vasospasm and the secondary end points were analyzed as time-to-event and the log-rank test was used. The t test was used to compare the durations of vasospasm and impaired autoregulation. Logistic regression was used to identify factors that may have affected the outcome. Data were presented as mean values and 95% CIs. It was considered significant when the P fell <0.05.
Eighty-six patients were assessed for eligibility during 2004. Six were excluded because of failed inclusion criteria in 5 and refusal to participate in 1. Eighty patients were randomized equally into 2 groups. Trial medication was commenced within 1.8 days (95% CI, 0 to 4.3 days) of the bleed. Thirty-eight patients (47.5%) completed 14 days of trial. Of the remainder, 30 patients (37.5%) were well enough to be discharged before 14 days, 10 died (12.5%), and premature withdrawal occurred in 2 patients (2.5%); 1 (placebo) because of epigastric pain on day 2, and the other (pravastatin) withdrew from TCD examinations on day 6. All 80 patients were included for the final analysis.
The prerandomization and postrandomization characteristics (Table 1) were well balanced in the 2 groups. There tended to be more patients in the pravastatin group, which experienced immediate postoperative deficits (22.5% versus 7.5%; Fisher’s exact test P=0.115; Table 2).
Primary End Points
Analysis of the primary end points (Table 3) showed a relative reduction in incidences of vasospasm and severe vasospasm on TCD in the pravastatin group by 32% (log-rank test P=0.006; Figure 1A) and 42% (log-rank test P=0.044; Figure 1B), respectively. The duration of severe vasospasm on TCD was shortened in the pravastatin group by 0.8 days (t test P=0.068). The average time of onset of vasospasm from ictus was 5.2 days (95% CI, 4.4 to 6.1 days), and no significant difference was seen between the 2 groups, although a delay was apparent in the pravastatin group (Figure 1). For ipsilateral measurements, a relative reduction in the incidences of vasospasm by 42% (log-rank test P=0.003) and severe vasospasm by 58% (log-rank test P=0.012), and an absolute shortening in the duration of vasospasm and severe vasospasm by 1.2 days (95% CI, −0.2 to 2.7 days; t test P=0.082) and 0.8 days (95% CI, 0.2 to 1.5 days; t test P=0.015), respectively, were observed in the pravastatin group. These effects were not seen on the contralateral side. The period of impaired cerebral autoregulation was shorter with pravastatin: ipsilateral side by 2.4 days (95% CI, 0.6 to 4.1 days; t test P=0.011) and contralateral side by 2.1 days (95% CI, 0.6 to 3.6 days; t test P=0.008).
Secondary End Points
Vasospasm-related DIDs occurred in 14 patients (12 placebo and 2 pravastatin) and were associated with new cerebral infarcts on CT scans (Table 4). The incidence of vasospasm-related DIDs was significantly reduced in the pravastatin group by 83% (log-rank test P<0.001; Figure 2A). The average time of onset of vasospasm-related DIDs from ictus was 5.6 days (95% CI, 4.5 to 6.8 days), and no significant difference was seen between the 2 groups; although, again, a delay was apparent in the pravastatin group (Figure 2).
Logistic regression analysis identified the WFNS grade (odds ratio [OR], 2.37 per grade; 95% CI, 1.46 to 3.83 per grade; P<0.001), immediate postoperative deficits (OR, 7.06; 95% CI, 1.06 to 47.16; P=0.044), and sepsis (OR, 38.69; 95% CI, 3.66 to 408.52; P=0.002) as variables that increased disability at discharge, whereas pravastatin therapy was associated with an 83% decrease (OR, 0.27; 95% CI, 0.08 to 0.95; P=0.041). The mortality of all causes was reduced in the pravastatin group by 75% (log-rank test P=0.037; Figure 2B).
Within 3 days of commencing pravastatin, there was no significant change in the serum muscle CK; ALT levels showed a similar increase in both groups. The total and LDL cholesterol levels were reduced more in the pravastatin group than in the placebo group by 10.1% (95%CI, 0 to 20.1%; t test P=0.048) and 15.5% (95% CI, 4.1 to 26.9%; t test P=0.008), respectively.
Although longer-term benefits have been reported from immediate statin therapy after acute coronary events,19 we believe that this study provides the first evidence that immediate statin therapy reduces potentially adverse physiological and clinical events after an acute cerebrovascular illness. No safety issues arose, including adverse biochemical effects. Similar cerebrovascular effects have been demonstrated in animal studies with higher doses of other statins.5,13,20 With the current dose (40 mg daily of pravastatin), drug tolerance was very high, with only 2 patients withdrawing early; both were because of nonmedication-related issues. The cholesterol-related biochemical changes seen in the pravastatin group were as expected and consistent with good compliance and drug absorption.
Thus, pravastatin therapy after aSAH significantly reduced the incidence, duration, and severity of cerebral vasospasm, improved cerebral autoregulation, and decreased vasospasm-related DIDs. The study design was not powered to detect a clinical improvement, and 12-month outcome data are still awaited. However, multivariate analysis highlighted pravastatin therapy as the only independent predictor of improved outcome at discharge, despite more patients in the pravastatin group experiencing sepsis or procedure-related neurological deficits, 2 major adverse factors contributing to an unfavorable outcome.
What is particularly encouraging is the combination of improved indices, physiological and clinical, in the pravastatin group. The current data are also encouraging in the light of a recent publication, which indirectly addressed the influence of statin therapy in SAH patients.21 Although TCD-measured indices for vasospasm presents difficulties in measurement and interpretation,14,22 the combination of TCD-registered vasospasm and impaired autoregulation is a powerful predictor of vasospasm-related DIDs and poor outcome.18 Both measures were significantly improved in the pravastatin group, particularly on the side of the culprit bleed, where the inflammatory reaction to subarachnoid blood is likely to be maximal. From our previous understanding of TCD surveillance of aSAH, this reduction is protective. In those in whom vasospasm still occurred in the pravastatin group, the time of onset between ictus and vasospasm and vasospasm-related DIDs seemed to be delayed (Figures 1 and 2⇑). Indeed, the vasospasm-related DIDs in the pravastatin group only appeared to occur after the 14-day trial had finished. This observation seemed compatible with the finding that the immediate vascular effects disappear within 48 hours after withdrawal of statins.23 A future trial will need to address this observation and call for a more prolonged treatment period.
We believe the mechanisms of the action of pravastatin are most likely multifactorial, generic, and unlikely to be specific to pravastatin. Statins lower the serum LDL cholesterol levels by inhibiting the hepatic 3-hydroxy-3-methylglutaryl coenzyme A reductase,24 which decreases the substrates for generating reactive oxygen species and oxidized LDL in vascular endothelium.20 By blocking the same enzyme in vascular endothelium, statins decrease intracellular isoprenoid levels,25 which, in turn, immediately leads to a direct increase in basal NO release and endothelial NO synthase8 and a reduction in the generation of reactive oxygen species.26 These mechanisms may be responsible for improved cerebral vasomotor reactivity, CBF, and neurological function after cerebral ischemia.6,8,9,20 Other potential beneficial properties include antiplatelet aggregation and adhesion, anti-inflammation, and decreased activities of peroxidation.8,20 The anti-inflammatory properties may be of particular relevance in attenuating endothelial damage caused by subarachnoid blood clots, maximal in the vicinity of a ruptured aneurysm.26
This study demonstrates significant physiological and clinical benefits from acute pravastatin therapy after aSAH without safety concerns. The benefits seen have justified the notion for a phase III outcome study in aSAH. Optimal dosage, choice of statins, generic criteria of vasospasm, and short-term and long-term clinical outcome measures for such a study are issues needing further address.
M.-Y.T. was supported by Raymond and Beverly Sackler Studentship, University of Cambridge.
- Received April 27, 2005.
- Revision received May 27, 2005.
- Accepted June 10, 2005.
Pickard JD, Murray GD, Illingworth R, Shaw MD, Teasdale GM, Foy PM, Humphrey PRD, Lang DA, Nelson R, Richards P, Sinar J, Bailey S, Skene A. Effect of oral nimodipine on cerebral infarction and outcome after subarachnoid hemorrhage: British aneurysm nimodipine trial. BMJ. 1989; 298: 636–642.
McGirt MJ, Lynch JR, Parra A, Sheng H, Pearlstein RD, Laskowitz DT, Pelligrino DA, Warner DS. Simvastatin increases endothelial nitric oxide synthase and ameliorates cerebral vasospasm resulting from subarachnoid hemorrhage. Stroke. 2002; 33: 2950–2956.
Sterzer P, Meintzschel F, Rösler A, Lanfermann H, Steinmetz H, Sitzer M. Pravastatin improves cerebral vasomotor reactivity in patients with subcortical small-vessel disease. Stroke. 2001; 32: 2817–2820.
Delanty N, Vaughan CJ. Vascular effects of statin in stroke. Stroke. 1997; 28: 2315–2320.
Endres M, Laufs U, Huang Z, Nakamura T, Huang P, Moskowitz MA, Liao JK. Stroke protection by 3-hydroxy-3-methylglutaryl (HMG)-CoA reductase inhibitors mediated by endothelial nitric oxide synthase. Proc Natl Acad Sci U S A. 1998; 95: 8880–8885.
Ross N, Hutchinson PJ, Seeley H, Kirkpatrick PJ. Timing of surgery for supratentorial aneurysmal subarachnoid hemorrhage: report of a prospective study. J Neurol Neurosurg Psychiatry. 2002; 72: 480–484.
Giller CA. A bedside test for cerebral autoregulation using transcranial Doppler ultrasound. Acta Neurochir. 1991; 108: 7–14.
Vaughan CJ, Delanty N. Neuroprotective properties of statin in cerebral ischemia and stroke. Stroke. 1999; 30: 1969–1973.
Parra A, Kreiter KT, Williams S, Sciacca R, Mack WJ, Naidech AM, Commichau CS, Fitzsimmons BF, Janjua N, Mayer SA, Connolly ES Jr. Effect of prior statin use on functional outcome and delayed vasospasm after acute aneurysmal subarachnoid hemorrhage: a matched controlled cohort study. Neurosurgery. 2005; 56: 476–484.
Minhas PS, Menon DK, Smielewski P, Czosnyka M, Kirkpatrick PJ, Clark JC, Pickard JD. Positron emission tomographic cerebral perfusion disturbances and transcranial Doppler findings among patients with neurological deterioration after subarachnoid hemorrhage. Neurosurgery. 2003; 52: 1017–1022.
Gertz K, Laufs U, Lindauer U, Nickenig G, Bohm M, Dirnagl U, Endres M. Withdrawal of statin treatment abrogates stroke protection in mice. Stroke. 2003; 34: 551–557.
Takemoto M, Liao JK. Pleiotropic effects of 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors. Arterioscler Thromb Vasc Biol. 2001; 21: 1712–1719.