Outcome in patients with subarachnoid hemorrhage treated with antiepileptic drugs

Transcription

1 See the Editorial and the Response in this issue, pp J Neurosurg 107: , 2007 Outcome in patients with subarachnoid hemorrhage treated with antiepileptic drugs AXEL J. ROSENGART, M.D., PH.D., 1,2 DEZHENG HUO, M.S., M.D., 3 JOCELYN TOLENTINO, M.P.H., 1 ROBERTA L. NOVAKOVIC, M.D., 1 JEFFREY I. FRANK, M.D., 1,2 FERNANDO D. GOLDENBERG, M.D., 1 AND R. LOCH MACDONALD, M.D., PH.D. 2 Department of Neurology, 1 Neurocritical Care and Acute Stroke Program, 3 Health Studies and Surgery, and 2 Section of Neurosurgery, The University of Chicago Pritzker School of Medicine, Chicago, Illinois Object. Prophylactic use of antiepileptic drugs (AEDs) in patients admitted with aneurysmal subarachnoid hemorrhage (SAH) is common practice; however, the impact of this treatment strategy on in-hospital complications and outcome has not been systematically studied. The goal in this study was twofold: first, to describe the prescribing pattern for AEDs in an international study population; and second, to delineate the impact of AEDs on in-hospital complications and outcome in patients with SAH. Methods. The authors examined data collected in 3552 patients with SAH who were entered into four prospective, randomized, double-blind, placebo-controlled trials conducted in 162 neurosurgical centers and 21 countries between 1991 and The prevalence of AED use was assessed by study country and center. The impact of AEDs on in-hospital complications and outcome was evaluated using conditional logistic regressions comparing treated and untreated patients within the same study center. Results. Antiepileptic drugs were used in 65.1% of patients and the prescribing pattern was mainly dependent on the treating physicians: the prevalence of AED use varied dramatically across study country and center (intraclass correlation coefficients 0.22 and 0.66, respectively [p 0.001]). Other predictors included younger age, worse neurological grade, and lower systolic blood pressure on admission. After adjustment, patients treated with AEDs had odds ratios of 1.56 (95% confidence interval [CI] ; p = 0.003) for worse outcome based on the Glasgow Outcome Scale; 1.87 (95% CI ; p 0.001) for cerebral vasospasm; 1.61 (95% CI ; p 0.001) for neurological deterioration; 1.33 (95% CI ; p = 0.04) for cerebral infarction; and 1.36 (95% CI ; p = 0.03) for elevated temperature during hospitalization. Conclusions. Prophylactic AED treatment in patients with aneurysmal SAH is common, follows an arbitrary prescribing pattern, and is associated with increased in-hospital complications and worse outcome. (DOI: /JNS-07/08/0253) KEY WORDS antiepileptic drug complication outcome subarachnoid hemorrhage A NEURYSMAL SAH affects approximately 30,000 people per year in the US. The majority of patients who survive have significant functional and cognitive disabilities. The high disability rate, in combination with the relatively young age of these patients in comparison to those with ischemic stroke, results in an important economic impact on our nation. Investigations performed between the mid-1950s and the mid-1970s supported the suggestion that the risk of seizures was as high as 27% in these patients. 5,35 This concept Abbreviations used in this paper: AED = antiepileptic drug; BP = blood pressure; CI = confidence interval; CT = computed tomography; GCS = Glasgow Coma Scale; GOS = Glasgow Outcome Scale; OR = odds ratio; SAH = subarachnoid hemorrhage; SD = standard deviation; WFNS = World Federation of Neurosurgical Societies. J. Neurosurg. / Volume 107 / August, 2007 led to the current and generally accepted medical practice of prophylactic AED use in patients with SAH. In these studies, approximately 40% of seizures occurred at the time of aneurysm rupture or following rebleeding. 10 In more recent studies of SAH, a seizure risk of only 5 to 8% has been suggested; the difference is possibly a reflection of more modern modalities of treatment, such as early aneurysm clip occlusion, endovascular coil treatment, and designated neurocritical care units. 2,3 All AEDs can impair cognitive function 1 in healthy volunteers and in patients with brain impairment. 23,24,34 Phenobarbital may have the greatest adverse cognitive profile, whereas phenytoin and carbamazepine demonstrate similar adverse effects. 1 In reports on children it has been suggested that the negative effects of AEDs can persist after drug withdrawal. 17 Even short-term exposure to AEDs at a vulnerable time period, such as prenatally, can lead to reduced full-scale IQ scores 29,33 and impaired neurobehavioral de- 253

2 A. J. Rosengart et al. velopment. 30 Furthermore, cognitive dysfunction worsens with increases in AED dosage, 32 higher AED levels, and polytherapy, 1,21 all of which are common in brain-injured patients. 7 In contrast to patients with epilepsy, however, in whom the adverse cognitive effects of AEDs are offset in part by the beneficial effects of seizure reduction, the risk of seizures in patients with SAH may be low enough that the risks of treatment outweigh the benefits. There are, however, few studies addressing the impact of prophylactic AED use on in-hospital complications, and outcome in patients with SAH has not been systematically studied. The raw data collected in approximately 3500 patients with SAH were used to explore several questions related to AED use: 1) what is the pattern of AED use in patients admitted with SAH; 2) what, if any, are the predictors of AED use; and 3) what are the effects of prophylactic AEDs on inhospital complications and outcome at 3 months in patients with SAH? Patient Population Clinical Material and Methods Data collected in 3552 patients with aneurysmal SAH who were recruited into four randomized, double-blind, placebo-controlled trials to evaluate the usefulness of the drug tirilazad were analyzed (Fig. 1). 8,9,12,14,15 These trials were conducted in neurosurgical centers in 21 countries in Europe, Australia, North America, and Africa between 1991 and Adult patients with ruptured saccular aneurysms were included. Detailed inclusion and exclusion criteria have been published. 12,14,15 Definition of Clinical and Radiological Variables Neurological grade was classified using the WFNS scale. The patient s BP and body temperature were recorded on admission and on Day 8 post-sah. Preexisting medical conditions were documented. Prophylactic and therapeutic hemodynamic therapy (hemodilution, induced hypertension, and hypervolemia) for symptomatic vasospasm after Day 2 was recorded. Data on in-hospital complications such as neurological worsening, cerebral infarction and edema, hydrocephalus, and pulmonary edema were collected. Symptomatic vasospasm was defined using clinical criteria and included the following: 1) symptoms between Days 5 and 12 after SAH (worsening headache, stiff neck, and/or low-grade fever; insidious onset of confusion and/ or decline in level of consciousness; and focal deficit); 2) a head CT scan excluding other causes of worsening; and 3) no other identifiable cause of neurological worsening. Neurological worsening was defined as a decrease of two or more points in the modified GCS score or an increase of two or more points in the motor score of the National Institutes of Health Stroke Scale lasting for 8 or more hours. 21 Outcome was graded 3 months post-sah by using the GOS. Unfavorable outcome was defined as death, vegetative state, or severe disability. All AEDs prescribed during the hospital course were noted, although seizure occurrence was not recorded. Radiographic data included grading of the amount of subarachnoid blood on the initial CT scan as none, diffuse or localized thin SAH, or diffuse or localized thick SAH. The presence of hydrocephalus, intraventricular hemorrhage, intracerebral hemorrhage, or vasospasm on admission angiograms was documented. The location and size of the ruptured aneurysm was classified as previously described Statistical Analysis First, the likelihood ratio tests in a random-effect logistic regression were used to examine whether AED use varied by study country and center. The extent of variation in AED use was described using an intraclass correlation coefficient. The intraclass correlation coefficient is equal to 1 when the use of AEDs is fully determined by a study center (all patients in some study centers used AEDs, but no patients in the other centers used these drugs) whereas it is equal to 0 when the variation by centers is simply due to chance. Second, we explored other predictors of AED use with a conditional logistic regression model in which the study center was the matching set, and in which only patients within the same study center were compared with each other; thus we controlled for confounding due to variation between study centers. Third, we examined the effects of AED use on in-hospital complications and 3-month outcome by using conditional logistic regressions, which, with the same methods described earlier, controlled for confounding data related to the study center. We controlled for study center because it is possible that outcomes may differ between centers due to variations in the patients condition on admission, the population served, and various aspects of medical management. Finally, we examined differential effects on the 3-month outcome of AED use by disease severity at admission in the framework of a conditional logistic regression model. A probability value of less than 0.05 was considered statistically significant. Results Admission Characteristics of the Study Population The mean age of the 3552 patients was years (mean SD) and 83% were female. Their clinical conditions were predominately in good grades (36% of patients were categorized in WFNS Grade I, 29% in Grade II, 12% in Grade III, 10% in Grade IV, and 13% in Grade V). The mean GCS score on admission was (mean SD). Thick SAH was observed on admission CT scans in 66% and intraventricular and intracerebral blood in 45 and 23%, respectively. The ruptured aneurysm was 12 mm or less in diameter in 74% of cases. Overall, 90% of patients underwent surgery to clip the ruptured aneurysm at a median of 29 hours (interquartile range hours) after the ictus. Prescribing Pattern for AEDs Sixty-five percent of patients enrolled in the four clinical trials were given a prescription for at least one AED (Table 1). Phenytoin was the most frequently prescribed drug (52.8%), followed by phenobarbital (18.7%), and only 2.3% of patients received carbamazepine. Concomitant use of two AEDs in an individual patient was found in 8%, and three AEDs were prescribed in only five patients (0.1%). Phenytoin and phenobarbital were less likely to be used simultaneously (OR 0.49, 95% CI ; p ). Nevertheless, there was no significant association between 254 J. Neurosurg. / Volume 107 / August, 2007

3 Outcome in patients with SAH treated with AEDs TABLE 1 Use of AEDs in 3552 patients with SAH Drug Used No. of Patients (%) any AED phenytoin phenobarbital carbamazepine phenytoin only phenobarbital only carbamazepine only phenytoin & phenobarbital phenytoin & carbamazepine phenobarbital & carbamazepine all three drugs 2313 (65.1) 1873 (52.7) 663 (18.7) 83 (2.3) 1587 (44.7) 394 (11.1) 31 (0.9) 249 (7.0) 32 (0.9) 15 (0.4) 5 (0.1) simultaneous use of phenytoin and carbamazepine (OR 0.72, 95% CI ; p = 0.13) or between phenobarbital and carbamazepine (OR 1.39, 95% CI ; p = 0.20). Predictors of AED Use Antiepileptic drug treatment varied dramatically according to study country (Fig. 2). The intraclass correlation coefficient describing the extent of variation in AED use between countries was 0.22 (95% CI , x2 = 1594; p, 0.001). Seven of the 21 study countries enrolled more than 100 patients each, and these groups amounted to 87% of all patients. The rate of AED use varied from 7 to 93% of cases, with a greater than 50% prevalence rate in Australia, Canada, Italy, and the US. In contrast, the intraclass correlation coefficient describing the extent of variation across individual study centers was 0.66 (95% CI , x2 = 2184; p, 0.001), from which we inferred that the use of AEDs was mainly determined by study center and therefore by the treating physician. In other words, even physicians in the same country had different patterns for prescribing AEDs in patients with SAH. Physicians in 15 study centers did not prescribe AEDs at all to the 170 patients enrolled in the trials; in 86 centers, physicians sometimes prescribed these drugs to their 2200 patients; and the remaining physicians in 61 centers always prescribed AEDs to their 1181 patients. Demographic and clinical data for patients enrolled in the four clinical trials are given in Table 2. After controlling for variation across study center, which was the most important determinant of AED use, we explored other demographic and clinical characteristics as predictors of AED use; for those investigations we used conditional logistic regression (Table 3). Patients receiving FIG. 1. Forest plot of end points in studies of prophylactic AED use in patients with SAH. SBP = systolic BP. J. Neurosurg. / Volume 107 / August,

4 A. J. Rosengart et al. AEDs were more likely to be younger (p = 0.001) and to have a lower systolic BP on admission (p = 0.007). There was a significant linear trend (p 0.001) for increasing use of AEDs with worsening WFNS grade. Sixty-four percent of patients in Grade I received AEDs compared with 72% of patients in Grade IV, although only 59% of patients in Grade V received AEDs. In the multivariate conditional logistic regression that simultaneously included WFNS grade, age, and systolic BP, all three factors remained statistically significantly associated with AED use. Primary Outcome and AED Use Patients treated with AEDs (33%) had a higher rate of unfavorable GOS scores than patients who did not receive AED treatment (26%, Table 2). The odds of having an unfavorable outcome was significantly higher in patients treated with AEDs relative to those who were not, after adjustment for study center, WFNS grade, age, and systolic BP on admission (OR 1.56, 95% CI ; p = 0.003). A sensitivity analysis was conducted by excluding 191 patients who died within the 1st week after SAH. This was done to exclude the possibility of selection bias that might occur if it is assumed that early death indicates severity of disease. The adjusted OR of unfavorable outcome was 1.83 (95% CI ) for AED treatment. The detrimental effect of AEDs on 3-month outcome existed among patients in WFNS Grades I to III and in Grades IV and V, demonstrating a lack of disease severity specific interaction (Table 4). No significant interaction was observed between age and AED treatment. There were marginally significant interactions between AED use and thickness of SAH (p = 0.057) and between AED use and intraparenchymal blood on CT scans (p = 0.05). The detrimental effect of AED on outcome was stronger among patients with diffuse thick SAH and among patients without intraparenchymal bleeding. In-Hospital Complications and AED Use The percentages of all in-hospital complications examined were higher in patients treated with AEDs than in those without AED use (Table 4). After adjustment for study center, WFNS grade, age, and systolic BP on admission, AED use was significantly associated with increased odds of clinical vasospasm (odds increase 87%; p 0.001) and neurological worsening (odds increase 61%; p 0.001). The odds of having cerebral infarction (p = 0.04) and elevated temperature at Day 8 of hospitalization (p = 0.03) were moderately but significantly increased in patients treated with AEDs. There were no significant differences in hydrocephalus, lung edema, brain edema, and use of cerebral angioplasty between the two groups. Discussion The key new findings of this study were that two of three patients admitted with acute aneurysmal SAH to neurosurgical centers throughout the world between 1991 and 1997 received AED treatment. Dilantin was the most common single AED used, followed by phenobarbital monotherapy. Antiepileptic drug treatment was determined predominately by the individual treating physician and varied greatly among centers and countries worldwide. These prescribing FIG. 2. Bar graph showing the prevalence rate of AED use in patients with SAH in study countries with more than 100 patients enrolled. n = number of patients. patterns are of substantial interest in the context of the main finding of this analysis, which was that the use of AEDs was significantly associated with neurological worsening, symptomatic vasospasm, cerebral infarction, fever 8 days after SAH, and poor outcome at 3 months posthemorrhage. The prophylactic use of AEDs in patients with SAH was introduced after early investigators reported seizure frequencies of 20 to 30%, mostly associated with the initial hemorrhage or following aneurysm rebleeding. 5,10,28 However, the risk of early seizures in patients with SAH is currently estimated to be 5 to 8%. 2,3,16 The reason for the decline is unknown, but may be related to different diagnostic criteria or to changes in treatment modalities. Administration of AEDs would seem warranted in at least some patients, for example in those with the following conditions: extensive cortical injury, unsecured ruptured aneurysms, increased intracranial pressure, impaired cerebral hemodynamics, and those in whom a seizure might worsen longterm outcome by producing cerebral metabolic stress, changes in BP, and brain swelling. Butzkueven et al. 4 reported that seizures at onset of SAH were independently associated with poor outcome. Nevertheless, they did not consider all prognostic factors for outcome in their multivariate analysis. 11 In another study it was reported that seizures at onset of SAH predict a significant neurological deficit. 16 On the other hand, in other reports it has been suggested that early seizures do not have a significant impact on outcome but rather are a marker of brain injury severity. 10,28 Several authors reported that prophylactic postoperative AED treatment had no effect on the interval between injury and first seizure or the frequency of early seizures. 26,37 In addition, the occurrence of perioperative seizures had no influence on the development of chronic epilepsy. 31 Most early seizures occurred at the onset of aneurysmal bleeding in the prehospital setting, 16 when treatment was not available and abnormal movements such as tonic posturing may have been misinterpreted as seizure activity. Therefore, these data in conjunction with our results call into question the current practice of prescribing prophylactic AEDs in two thirds of all patients with SAH. Antiepileptic drug use in this database was associated with worse outcome at 3 months. This association was independent of admission characteristics such as neurologi- 256 J. Neurosurg. / Volume 107 / August, 2007

5 Outcome in patients with SAH treated with AEDs TABLE 2 Demographic and clinical characteristics in 3552 patients with SAH who were treated prophylactically with AEDs* Characteristic Trial 1 (823 patients) Trial 2 (897 patients) Trial 3 (817 patients) Trial 4 (1015 patients) demographic mean age in yrs female sex race Caucasian African-American other clinical op to treat aneurysm tirilizad use admission factor WFNS grade I II III IV V mean systolic BP mean diastolic BP temp 38 C medical history hypertension angina/mi diabetes previous SAH radiological status & findings CT scan finding thick SAH intraventricular hemorrhage intraparenchymal hemorrhage ruptured aneurysm location internal carotid artery anterior cerebral artery middle cerebral artery pst circulation other ruptured aneurysm size (mm) vasospasm on admission angiography outcome unfavorable GOS at 3 mos symptomatic vasospasm neuro worsening cerebral infarction use of cerebral angioplasty hydrocephalus lung edema brain edema temp 38 C on Day * Values are expressed as percentages, unless otherwise noted. Abbreviations: MI = myocardial infarction; neuro = neurological; pst = posterior; temp = temperature. These values given as means SDs. Unfavorable outcome was defined as death, vegetative state, or severe disability on the GOS. Neurological worsening was defined as a greater than two-point decrease in the modified GCS score or a greater than two-point increase in the motor score of the National Institutes of Health Stroke Scale lasting 8 or more hours. cal grade and variations in treatment modalities among centers. We did not identify a patient subgroup experiencing disproportionate impairment or actually benefiting from prophylactic AED treatment. There was a more favorable overall outcome in patients with SAH who were not treated with AEDs, despite the likelihood that seizures may have occurred with relatively higher frequency in these patients. There are no randomized studies in which the J. Neurosurg. / Volume 107 / August, 2007 risks and benefits of early prophylactic AED treatment in patients with SAH have been investigated, but in trials conducted in patients with head injury it was reported that AEDs did not reduce early or late seizure incidence or affect mortality rates. 37 The reasons for the association between AEDs and poor outcome are a matter of speculation. Antiepileptic drugs reduce neuronal irritability, which may result in altered 257

6 A. J. Rosengart et al. TABLE 3 Potential predictors of AED use* Characteristic OR (95% CI) p Value age (yrs) (ref value) ( ) ( ) ( ) male sex 1.00 ( ) 1.0 race Caucasian 1.00 (ref value) 0.11 African-American 1.66 ( ) other 0.67 ( ) admission WFNS grade I 1.00 (ref value) II 1.09 ( ) III 1.64 ( ) IV 2.36 ( ) V 1.47 ( ) admission BP systolic, per 10 mm Hg increase 0.93 ( ) diastolic, per 10 mm Hg increase 0.95 ( ) 0.19 admission temp 38 C 1.20 ( ) 0.40 medical history hypertension 1.03 ( ) 0.82 angina/mi 0.76 ( ) 0.34 diabetes 1.74 ( ) 0.08 SAH 0.86 ( ) 0.47 CT scan finding thick SAH 1.17 ( ) 0.21 intraventricular blood 1.02 ( ) 0.87 intraparenchymal blood 1.26 ( ) 0.09 ruptured aneurysm location internal carotid artery 1.00 (ref value) 0.20 anterior cerebral artery 1.25 ( ) middle cerebral artery 1.24 ( ) pst circulation 1.06 ( ) other 0.29 ( ) ruptured aneurysm size (mm) (ref value) ( ) ( ) vasospasm on admission angiography 1.23 ( ) 0.28 prophylactic hemodynamic therapy 0.77 ( ) 0.09 op to clip aneurysm 1.17 ( ) 0.42 randomly assigned tirilizad 1.07 ( ) 0.56 * ref = reference. Adjusted for study center by using a conditional logistic regression model. Trend probability values were calculated for age, admission WFNS grade, and ruptured aneurysm size. synaptic growth and connectivity, which in turn may be important in neurological recovery. 22 Most AEDs impair cognitive function, 6 both in healthy volunteers and in patients with brain impairment. 23,24,34 Dilantin and phenobarbital were found to impair recovery from brain injury in animal studies. 7 Many AEDs enhance -aminobutyric acidergic inhibition or block glutamate-mediated excitation and are known to cause widespread and dose-dependent apoptotic neurodegeneration in the developing rat brain, 18 even when given only for several hours. 27 Studies performed in animal models identified potential neuroprotective effects of AEDs when given at the time of the ischemic event, 25 but this effect was only present in vitro when the ischemic insult was of short duration. 36 On the other hand, Goldstein 6 reported that phenytoin worsens cerebral ischemia. Compared with controls, children exposed in utero to AEDs had a reduced full-scale IQ score 29,33 and impaired neurobehavioral development. 30 Cerebral vasospasm after SAH is a combination of prolonged contraction and vessel remodeling leading to cerebral ischemia, clinical deterioration, and infarction. 19 Investigators have described apoptosis of cerebral endothelial cells 38 in addition to smooth-muscle cell necrosis 37 in animal models, providing one possible explanation for how AEDs could aggravate cerebral vasospasm. Use of AEDs was also associated with an increased risk of fever 8 days after admission. Although adverse reactions to AEDs were serious enough to discontinue therapy in 10 to 18% 20,26 of patients, fever itself is strongly associated with poor functional and cognitive outcome. 13,14 Conclusions The limitations of this analysis include lack of information on seizure occurrence and frequency among patients who were treated or not treated with AEDs. However, if we use current estimates of 5 to 8% of early seizures in patients with SAH, the adjusted OR would still be 1.44 and 1.37, respectively, for an unfavorable GOS score in patients treated with AEDs. Even in this worst-case scenario, there would be no beneficial effect of prophylactic AED treatment. Within the study, AEDs were used in two thirds of all patients with SAH, and based on current practices, these drugs would most commonly have been administered prophylactically beginning at the time of admission. We also do not have data on the duration or intensity of AED expo- TABLE 4 End points by AED use and ORs from conditional logistic regressions* Adjusted for Study Center Adjusted for Study Center, Neuro Grade, Age, & Systolic BP End Point OR (95% CI) p Value OR (95% CI) p Value unfavorable GOS score at 3 mos 1.52 ( ) ( ) symptomatic vasospasm 1.94 ( ) ( ) neuro worsening 1.57 ( ) ( ) cerebral infarction 1.36 ( ) ( ) 0.04 use of cerebral angioplasty 1.70 ( ) ( ) 0.22 hydrocephalus 0.83 ( ) ( ) 0.10 lung edema 1.48 ( ) ( ) 0.09 brain edema 1.47 ( ) ( ) 0.09 temp 38 C on Day ( ) ( ) 0.03 * See Table 2 for definitions of unfavorable outcome and neurological worsening. 258 J. Neurosurg. / Volume 107 / August, 2007

7 Outcome in patients with SAH treated with AEDs sure; the database was not designed primarily to evaluate the influence of AED treatment on outcome in patients with SAH. Patients treated with AEDs may be sicker, and thus the observed detrimental effects could simply have been due to selection bias. After adjustment for study center, neurological grade, age, and systolic BP on admission, however, we found that AED users still had a higher risk of unfavorable outcome. We cannot rule out unmeasured differences within the same study center between the two groups; such selection bias can only be accounted for in randomized controlled studies. Despite these shortcomings, the strong associations of multiple adverse events with use of AEDs in patients with aneurysmal SAH are of great interest, particularly in view of the continuing devastating overall effects of SAH. Furthermore, if no beneficial effect of prophylactic AEDs was demonstrated after adjustment for seizure occurrence (based on current estimated risk), then the use thereof becomes of questionable benefit. These results support the need for prospective randomized studies. Acknowledgments We gratefully acknowledge Jonathan R. Kaplan and Sandra G. Guy for their assistance in preparing this manuscript. References 1. Anonymous: Behavioral and cognitive effects of anticonvulsant therapy. American Academy of Pediatrics Committee on Drugs. Pediatrics 96: , Baker CJ, Pestigiacomo CJ, Solomon RA: Short-term perioperative anticonvulsant prophylaxis for the surgical treatment of low-risk patients with intracranial aneurysms. Neurosurgery 37: , Bidziński J, Marchel A, Sherif A: Risk of epilepsy after aneurysm operations. Acta Neurochir (Wien) 119:49 52, Butzkueven H, Evans AH, Pitman A, Leopold C, Jolley DJ, Kaye AH, et al: Onset seizures independently predict poor outcome after subarachnoid hemorrhage. Neurology 55: , Cabral RJ, King TT, Scott DF: Epilepsy after two different neurosurgical approaches to the treatment of ruptured intracranial aneurysm. J Neurol Neurosurg Psychiatry 39: , Goldstein LB: Common drugs may influence motor recovery after stroke. The Sygen In Acute Stroke Study Investigators. Neurology 45: , Goldstein LB: Prescribing potentially harmful drugs to patients admitted to hospital after head injury. J Neurol Neurosurg Psychiatry 58: , Haley EC Jr, Kassell NF, Apperson-Hansen C, Maile MH, Alves WM: A randomized, double-blind, vehicle-controlled trial of tirilazad mesylate in patients with aneurysmal subarachnoid hemorrhage: a cooperative study in North America. J Neurosurg 86: , Haley EC Jr, Kassell NF, Torner JC: A randomized controlled trial of high-dose intravenous nicardipine in aneurysmal subarachnoid hemorrhage. A report of the Cooperative Aneurysm Study. J Neurosurg 78: , Hart RG, Byer JA, Slaughter JR, Hewett JE, Easton JD: Occurrence and implications of seizures in subarachnoid hemorrhage due to ruptured intracranial aneurysms. Neurosurgery 8: , Huff JS, Perron AD: Onset seizures independently predict poor outcome after subarachnoid hemorrhage. Neurology 56: , 2001 J. Neurosurg. / Volume 107 / August, Kassell NF, Haley EC Jr, Apperson-Hansen C, Alves WM: Randomized, double-blind, vehicle-controlled trial of tirilazad mesylate in patients with aneurysmal subarachnoid hemorrhage: a cooperative study in Europe, Australia, and New Zealand. J Neurosurg 84: , Kreiter KT, Commichau C, Kowalski RG, Ostapkovich N, Huddleston D, Fitzsimmons BF, et al: Fever burden and cognitive outcome after subarachnoid hemorrhage. Neurology 60:A230, 2003 (Abstract) 14. Lanzino G, Kassell NF: Double-blind, randomized, vehicle-controlled study of high-dose tirilazad mesylate in women with aneurysmal subarachnoid hemorrhage. Part II. A cooperative study in North America. J Neurosurg 90: , Lanzino G, Kassell NF, Dorsch NW, Pasqualin A, Brandt L, Schmiedek P, et al: Double-blind, randomized, vehicle-controlled study of high-dose tirilazad mesylate in women with aneurysmal subarachnoid hemorrhage. Part I. A cooperative study in Europe, Australia, New Zealand, and South Africa. J Neurosurg 90: , Lin CL, Dumont AS, Lieu AS, Yen CP, Hwang SL, Kwan AL, et al: Characterization of perioperative seizures and epilepsy following aneurysmal subarachnoid hemorrhage. J Neurosurg 99: , Loring DW, Meador KJ: Cognitive side effects of antiepiletic drugs in children. Neurology 62: , Macdonald RL, Kelly KM: Antiepiletic drug mechanisms of action. Epilepsia 36 (Suppl 2):S2 S12, Macdonald RL, Weir BK, Runzer TD, Grace MG, Findlay JM, Saito K, et al: Etiology of cerebral vasospasm in primates. J Neurosurg 75: , Mattson RH, Cramer JA, Collins JF, Smith DB, Delgado-Escueta AV, Browne TR, et al: Comparison of carbamazepine, phenobarbital, phenytoin, and primidone in partial and secondarily generalized tonic-clonic seizures. N Engl J Med 313: , Meador KJ: Cognitive side effects of medications. Neurol Clin 16: , Meador KJ: Neurodevelopmental effects of antiepileptic drugs. Current Neurol Neurosci Rep 4: , Meador KJ, Loring DW, Abney OL, Allen ME, Moore EE, Zamrini EY, et al: Effects of carbamazepine and phenytoin on EEG and memory in healthy adults. Epilepsia 34: , Meador KJ, Loring DW, Allen ME, Zamrini EY, Moore EE, Abney OL, et al: Comparative cognitive effects of carbamazepine and phenytoin in healthy adults. Neurology 41: , Murakami A, Furui T: Effects of conventional anticonvulsants, phenytoin, carbamazepine, and valproic acid, on sodium-potassium-adenosine triphosphatase in acute ischemic brain. Neurosurgery 34: , North JB, Penhall RK, Hanieh A, Frewin DB, Taylor WB: Phenytoin and postoperative epilepsy. A double-blind study. J Neurosurg 58: , Olney JW, Wozniak DF, Jevtovic-Todorovic V, Farber NB, Bittigau P, Ikonomidou C: Drug-induced apoptotic neurodegeneration in the developing brain. Brain Pathol 12: , Pinto AN, Canhao P, Ferro JM: Seizures at the onset of subarachnoid haemorrhage. J Neurol 243: , Reinisch JM, Sanders SA, Mortensen EL, Rubin DB: In utero exposure to phenobarbital and intelligence deficits in adult men. JAMA 274: , Scolnik D, Nulman I, Rovet J, Gladstone D, Czuchta D, Gardner HA, et al: Neurodevelopment of children exposed in utero to phenytoin and carbamazepine monotherapy. JAMA 271: , Shaw MD: Post-operative epilepsy and the efficacy of anticonvulsant therapy. Acta Neurochir Suppl (Wien) 50:55 57, Smith DB: Cognitive effects of antiepileptic drugs. Adv Neurol 55: , Vanoverloop D, Schnell RR, Harvey EA, Holmes LB: The effects of prenatal exposure to phenytoin and other anticonvulsants on in- 259

8 A. J. Rosengart et al. tellectual function at 4 to 8 years of age. Neurotoxicol Teratol 14: , Vining EPG, Mellitis ED, Dorsen MM, Cataldo MF, Quaskey SA, Spielberg SP, et al: Psychologic and behavioral effects of antiepileptic drugs in children: a double-blind comparison between phenobarbital and valproic acid. Pediatrics 80: , Walton JN: The electroencephalographic sequelae of spontaneous subarachnoid haemorrhage. Electroencephalogr Clin Neurophysiol Suppl 5:41 52, Watson GB, Lanthorn TH: Phenytoin delays ischemic depolarization, but cannot block its long-term consequences, in the rat hippocampal slice. Neuropharmacology 34: , Young B, Rapp RP, Norton JA, Haack D, Tibbs PA, Bean JR: Failure of prophylactically administered phenytoin to prevent early posttraumatic seizures. J Neurosurg 58: , Zubkov AY, Tibbs RE, Aoki K, Zhang JH: Morphological changes of cerebral penetrating arteries in a canine double hemorrhage model. Surg Neurol 54: , 2000 Manuscript submitted February 17, Accepted January 26, This work was supported by gifts from The Cancer Research Foundation and The Brain Research Foundation, both of The University of Chicago Pritzker School of Medicine. Address reprint requests to: Axel J. Rosengart, M.D., Ph.D., Neurocritical Care and Acute Stroke Program, The University of Chicago, 5841 South Maryland Avenue, MC 2030, Chicago, Illinois arosenga@neurology.bsd.uchicago.edu. 260 J. Neurosurg. / Volume 107 / August, 2007