Electrocardiographic Changes Predict Angiographic Vasospasm After Aneurysmal Subarachnoid Hemorrhage

Transcription

1 Electrocardiographic Changes Predict Angiographic Vasospasm After Aneurysmal Subarachnoid Hemorrhage George M. Ibrahim, MD; R. Loch Macdonald, MD, PhD Background and Purpose Early identification of patients at risk of angiographic vasospasm after aneurysmal subarachnoid hemorrhage (SAH) may mitigate its sequelae. One mechanism that may contribute to angiographic vasospasm is increased central sympathetic activity, which is also thought to cause electrocardiographic (ECG) changes after SAH. Here, we perform the first study to determine the association between ECG changes and angiographic vasospasm after SAH. Methods Exploratory analysis was performed on 413 patients from CONSCIOUS-1, a prospective randomized trial of clazosentan for the prevention of angiographic vasospasm. ECGs were obtained within 24 hours of aneurysm rupture and during the vasospasm risk period. Angiographic vasospasm was assessed using catheter angiography at baseline and 7 to 11 days after SAH. Multivariate logistic regression was used to identify significant associations. Results The most prevalent finding on ECG both immediately following SAH and during the vasospasm risk period was QT prolongation (42% and 25%, respectively). A prolonged QT interval and tachycardia on the baseline ECG were associated with angiographic vasospasm (OR, 1.86; 95% CI, ; and OR, 10.83; 95% CI, , respectively). QT prolongation on ECG during the vasospasm risk period was also associated with angiographic vasospasm (OR, 3.53; 95% CI, ). No ECG findings were associated with delayed ischemic neurological deficit, but tachycardia and ST changes were associated with worse clinical outcome. Conclusions QT prolongation and tachycardia on ECG were independently associated with angiographic vasospasm after aneurysmal SAH on multivariate analysis. Clinical Trial Registration URL: Unique Identifier: NCT (Stroke. 2012;43: ) Key Words: subarachnoid hemorrhage electrocardiography vasospasm angiography neurogenic stunned myocardium Angiographic vasospasm is an important cause of morbidity and mortality after aneurysmal subarachnoid hemorrhage (SAH). 1 The strongest known predictor of angiographic vasospasm is the subarachnoid clot burden present on baseline computed tomography scan. 2,3 Although the pathophysiology of angiographic vasospasm following SAH is incompletely understood, pathological hyperactivity of the sympathetic nervous system is 1 mechanism that may be important for this phenomenon. 4,5 An increase in central sympathetic activity is also thought to contribute to electrocardiographic (ECG) abnormalities that may be discernable following SAH. 6 Although several studies have evaluated the association between ECG abnormalities and clinical outcome following aneurysmal SAH, it remains unknown whether an association exists between ECG changes and angiographic vasospasm. The identification of such an association is important for several reasons. First, early identification of patients at risk of angiographic vasospasm may permit closer observation, rapid intervention, and potentially improved outcomes. Second, few predictive models exist for the identification of these patients, and an association between ECG changes and angiographic vasospasm would provide additional insight into the underlying processes mediating both cerebrovascular and cardiovascular changes after SAH. Finally, ECG is a ubiquitous, inexpensive, and readily available diagnostic tool, which has yet to be utilized for the prediction of angiographic vasospasm. The current article is the first to identify ECG changes that are associated with angiographic vasospasm following SAH. To do so, we used a unique patient database accrued in the CONSCIOUS-1 (Clazosentan to Overcome Neurological ischemia and Infarction OccUrring after Subarachnoid hemorrhage) trial, in which all subjects underwent catheter angiography to determine the extent of angiographic vasospasm. Secondary outcomes, including delayed ischemic neurological deficit (DIND) and clinical outcome as mea- Received April 3, 2012; accepted April 30, From the Division of Neurosurgery, St. Michael s Hospital, Labatt Family Centre of Excellence in Brain Injury and Trauma Research, Keenan Research Centre of the Li Ka Shing Knowledge Institute of St. Michael s Hospital and Department of Surgery, University of Toronto, Toronto, Ontario, Canada. Correspondence to R. Loch Macdonald, MD, PhD, St Michael s Hospital, 30 Bond Street, Toronto, ON, Canada M5B 1W8. MacdonaldLo@smh.ca 2012 American Heart Association, Inc. Stroke is available at DOI: /STROKEAHA

2 Ibrahim and Macdonald ECG Findings and Vasospasm After Aneurysmal SAH 2103 sured by the modified Rankin Scale score, were also correlated with ECG changes. Methods Study Population We conducted a post hoc analysis of 413 subjects enrolled between January 2005 and March 2006 onto CONSCIOUS-1, a prospective, randomized, double-blinded phase IIB trial evaluating the efficacy of clazosentan in preventing angiographic vasospasm. 7 The methods and results have been published. 7 Clinical Assessment All patients with computed-tomography-confirmed SAH were admitted to the respective neurosurgical units and underwent microsurgical clipping or endovascular coiling as deemed appropriate by the treating physician. The severity of the subjects presenting symptoms was classified based on the World Federation of Neurosurgical Societies (WFNS) scale. 8 All subjects were monitored for development of DIND, which was defined as neurological worsening of 2 points on the modified Glasgow Coma Scale, or an increase of 2 points in the abbreviated National Institutes of Health Stroke Scale lasting 2 hours. 9 At 12 weeks postsah, the subjects modified Rankin Scale score was calculated with poor clinical outcome, which was defined as outcomes worse than moderate disability. 10 Electrocardiography An ECG was obtained for all patients within 24 hours of aneurysm rupture and during the time of risk of angiographic vasospasm, defined as 7 to 11 days following SAH. For this analysis, we included also ECG performed 4 to 7 days after SAH. The frequency of abnormal findings on ECG in rate, rhythm, ST-segment, intervals, and T-waves was reported based on these ECG recordings, not on continuous ECG recording. Tachycardia was defined as a heart rate of 100 beats/minute, and bradycardia as 60 beats/minute. STsegment elevation and depression were defined as deviations of 1 mm above and below the baseline, respectively. The frequency of nonspecific ST-changes was also documented. A QT interval, corrected for heart rate 410 ms and a QRS interval 100 ms were considered abnormal, respectively. All ECGs were interpreted by the treating physicians. As serum potassium has been previously reported as an independent risk factor for ECG changes following SAH, 11 we incorporated the electrolyte levels on presentation in the statistical models that included baseline ECG changes. Hypokalemia and hyperkalemia were defined as potassium levels 3.2 mmol/l and 5.5 mmol/l, respectively. The mean time between SAH ictus and bloodwork was hours. Radiology All patients underwent catheter angiography within 48 hours of aneurysm rupture, and then at 7 to 11 days postsah. Angiographic vasospasm was determined by the percentage of change in arterial diameter between baseline and follow-up angiograms as follows: none (0% 25%), mild (26% 50%), moderate (51% 75%), and severe (76% 100%). Computed tomography scans were performed at baseline, 24 to 48 hours after the aneurysm-securing procedure, 6 weeks postaneurysmal rupture, and again as deemed necessary by the treating physicians. The extent of SAH was quantified using the Hijdra scale, which evaluates the amount of clot in 10 fissures and cisterns. 12 All images were reviewed centrally by 2 independent, blinded reviewers with adjudication of disagreements on angiographic vasospasm. Table 1. Clinical and Radiographic Characteristics of Patients Total (%) N 413 Age (y) (50.8) (49.2) Men 121 (29.3) History of hypertension 175 (42.4) Nicotine use 217 (52.5) WFNS Score (grade) (75.8) (24.2) Rescue therapy 74 (17.9) Anterior circulation 360 (88.9) Posterior circulation 45 (11.1) Aneurysm Size (mm) (42.2) (57.8) Subarachnoid clot burden (Hijdra score) Intracerebral hemorrhage 50 (12.4) Intraventricular hemorrhage 313 (77.5) Hydrocephalus 375 (92.8) WFNS indicates World Federation of Neurosurgical Societies. Error bars represent standard deviation. Statistics Descriptive statistics are reported as frequencies, and relatedsamples are compared using the sign test. Where indicated, variables were dichotomized by their median value. Two analyses were performed, initially using the entire database of patients enrolled onto the trial, and second, only the patients who received placebo (n 96). Binary logistic regression was used to model the presence of vasospasm, DIND, and poor clinical outcome against potential predictive variables. Covariates were chosen for inclusion in the multivariate regression model based on a review of correlative variables in literature, as well as in this database. 3 ECG findings that were previously shown in literature to correlate with outcome were included in the analysis. s with 5 events were excluded from the final models. For all final models, statistical significance was set at P Analysis was performed using SAS 9.3 (SAS Institute, Inc). Results Patients Four hundred thirteen patients with a median age of 51 years were enrolled onto the CONSCIOUS-1 trial (Table 1). Of these, 76% were WFNS Grade 1 3 and the majority of aneurysms were in the anterior circulation (89%). The most prevalent ECG finding within 24 hours of aneurysm rupture (Table 2) was QT prolongation (42%), followed by bradycardia (17%) and nonspecific ST changes (9%). During the vasospasm risk period, QT prolongation remained the most prevalent finding (25%), although the decrease in frequency of QT prolongation over the time interval could not be attributed to chance alone (P 0.001). During the vasospasm risk period, bradycardia and tachycardia were equally prevalent at 10%, although the

3 2104 Stroke August 2012 Table 2. ECG Findings at Baseline and During the Vasospasm Risk Period Baseline ECG (%) Period (%) P Value* Rate Tachycardia (HR 100 beats/min) 9 (4) 24 (10) 0.53 Bradycardia (HR 60 beats/min) 44 (17) 25 (10) 0.07 Rhythm Nodal 12 (6) 14 (6) 0.74 Atrial fibrillation 2 (1) 5 (2) 0.80 ST Segment Nonspecific changes 24 (9) 37 (15) 0.59 Depression 3 (1) 3 (1) 0.61 Elevation 2 (1) 3 (1) 0.67 Intervals Prolonged QT 106 (42) 62 (25) Widened QRS 9 (4) 8 (3) 0.55 Inverted T-waves 12 (5) 10 (4) 0.51 ECG indicates electrocardiogram; HR, heart rate. *As determined by the sign test for discordant pairs. proportion of patients with bradycardia decreased since the baseline ECG (P 0.07). When ECG abnormalities were analyzed by aneurysm location, there was a trend toward greater frequency of nodal rhythms in patients with anterior circulation aneurysms compared with posterior circulation (P 0.07). No other associations between ECG abnormalities and aneurysm location were identified. Angiographic Vasospasm In the binary logistic regression model, variables that were found to be significantly associated with mild, moderate, or severe angiographic vasospasm (Table 3) were female sex (OR, 2.32; 95% CI, ), high subarachnoid clot burden (OR, 2.41; 95% CI, ), WFNS grade 4 5 (OR, 3.15; 95% CI, ), clipping versus coiling (OR, 2.31; 95% CI, ), as well as a prolonged QT interval, both within 24 hours of aneurysm rupture (OR, 1.86; 95% CI, ) and within the vasospasm risk period (OR, 3.53; 95% CI, ). Tachycardia on baseline ECG was the strongest independent predictor of angiographic vasospasm (OR, 10.83; 95% CI, ). As previously reported, patients on 15 mg/h of clazosentan were less likely to exhibit angiographic vasospasm than were those on placebo (OR, 0.43; 95% CI, ). 7 Furthermore, older patients were less likely to develop angiographic vasospasm (OR, 0.47; 95% CI, ). There were no ECG findings that were associated with severe (versus none, mild, or moderate) angiographic vasospasm in the complete cohort of subjects. However, within the subgroup of patients who only received placebo (n 96), QT prolongation at baseline was significantly associated with severe angiographic vasospasm on multivariate analysis (OR, 14.67; 95% CI, ). As we only included variables with 5 events in the multivariate model, the prevalence of Table 3. Multivariate Analysis of Outcomes Associated With Presence of Mild, Moderate, or Severe Vasospasm (n 203 Versus n 210 With no Angiographic Vasospasm) Baseline ECG OR (95% CI) Period OR (95% CI) Age (y) 51 vs 51) 0.47 ( )* 0.46 ( )* Sex 2.32 ( )* 2.15 ( )* Hypertension 1.44 ( ) 0.89 ( ) High SAH clot burden 2.41 ( )* 2.47 ( )* WFNS 3.15 ( )* 3.39 ( )* Clipping (vs coiling) 2.31 ( )* 2.30 ( )* ACA (vs posterior) 2.14 ( ) 1.18 ( ) ICA (vs posterior) 1.21 ( ) 0.81 ( ) MCA (vs posterior) 2.25 ( ) 1.09 ( ) Treatment With Clazosentan 1 mg/h vs placebo 0.98 ( ) 0.65 ( ) 5 mg/h vs placebo 0.98 ( ) 0.71 ( ) 15 mg/h vs placebo 0.43 ( )* 0.33 ( )* Hypokalemia 0.43 ( ) NA Hyperkalemia Too few events NA ECG Findings Bradycardia (vs normal rate) 0.45 ( ) 0.57 ( ) Tachycardia (vs normal rate) ( )* 1.02 ( ) ST changes 1.50 ( ) 1.3 ( ) Prolonged QT 1.86 ( )* 3.53 ( )* Nodal rhythm ( ) 2.83 ( ) T-wave changes 0.48 ( ) 1.34 ( ) ECG indicates electrocardiogram; SAH, subarachnoid hemorrhage; WFNS, World Federation of Neurosurgical Socieies; ACA, anterior cerebral artery; ICA, internal carotid artery; MCA, middle cerebral artery; N/A, not available. *Denotes significance at P As determined by the Hijdra scale dichotomized by the mean. tachycardia within the placebo-only group (4 patients, 4.2%) precluded multivariate modeling. Delayed Ischemic Neurological Deficit There were no findings on ECGs that independently predicted DIND (Table 4). The most significant predictor of DIND was the presence of mild, moderate, or severe angiographic vasospasm (OR, 69.87; 95% CI, ). Higher WFNS scores (OR, 2.58; 95% CI, ) and higher subarachnoid clot burden (OR, 2.63; 95% CI, ) were also predictive of DIND. Clinical Outcome Factors associated with poor clinical outcome on multivariate analysis (Table 5) included WFNS grades 4 or 5 (OR, 4.80; 95% CI, ), as well as angiographic vasospasm (OR, 2.19; 95% CI, ) and DIND (OR, 4.13; 95% CI, ). Tachycardia and ST changes on ECG during the vasospasm risk period predicted poor outcome (OR, 2.82; 95% CI, ; and OR, 2.90; 95% CI, , respectively).

4 Ibrahim and Macdonald ECG Findings and Vasospasm After Aneurysmal SAH 2105 Table 4. Multivariate Analysis of Outcomes Associated With Delayed Ischemic Neurological Deficit (n 78 vs n 335 With no DIND) Baseline ECG OR (95% CI) Period OR (95% CI) Age (y) 51 vs 51) 0.96 ( ) 1.04 ( [1]) Sex 1.39 ( ) 0.91 ( ) Hypertension 1.66 ( ) 0.82 ( ) High SAH clot burden* 2.63 ( ) 1.94 ( ) WFNS 2.58 ( ) 3.18 ( ) Clipping (vs coiling) 2.09 ( ) 0.87 ( ) ACA (vs posterior) 0.60 ( ) 0.36 ( ) ICA (vs posterior) 0.48 ( ) 0.28 ( ) MCA (vs posterior) 0.63 ( ) 0.43 ( ) Treatment With Clazosentan 1 mg/h vs placebo 1.61 ( ) 1.21 ( ) 5 mg/h vs placebo 0.98 ( ) 0.55 ( ) 15 mg/h vs placebo 1.34 ( ) 0.95 ( ) Angiographic vasospasm ( ) ( ) Hypokalemia 0.01 ( ) NA Hyperkalemia Too few events NA ECG findings Bradycardia (vs normal 0.79 ( ) 0.61 ( ) rate) Tachycardia (vs normal 0.16 ( ) 1.11 ( ) rate) ST changes 1.09 ( ) 2.64 ( ) Prolonged QT 1.92 ( ) 0.71 ( ) Nodal rhythm 0.01 ( 0.01, 999) 0.23 ( ) T-wave changes 0.73 ( ) 6.58 ( ) DIND indicates delayed ischemic neurological deficit; ECG, electrocardiogram; SAH, subarachnoid hemorrhage; WFNS, World Federation of Neurosurgical Societies; ACA, anterior cerebral artery; ICA, internal carotid artery; MCA, middle cerebral artery; NA, not available. *As determined by the Hijdra scale dichotomized by the mean. Denotes significance at P Discussion This exploratory analysis found that QT prolongation both on the ECG performed within 24 hours of aneurysm rupture and that performed during the vasospasm risk period was independently associated with the presence of angiographic vasospasm. Furthermore, tachycardia on the baseline ECG was associated with angiographic vasospasm and poor outcome. ST segment changes on the delayed ECG examination were also associated with poor outcome. The present study adds to studies of ECG changes following SAH, in part related to neurogenic stunned myocardium (Lee et al, 13 and Nguyen and Zaroff 14 for review). ECG changes in association with SAH are thought to be related to increased central sympathetic activity, resulting in a hyperdynamic cardiovascular state. 6 The cerebral circulation is also densely supplied by noradrenergic sympathetic Table 5. Multivariate Analysis of Outcomes Associated With Poor Outcome on the Modified Rankin Score (n 130 vs n 283 With Good Outcome) Baseline ECG OR (95% CI) Period OR (95% CI) Age (y) 51 vs 51) 1.75 ( ) 1.38 ( ) Sex 1.92 ( ) 1.73 ( ) Hypertension 1.75 ( ) 1.54 ( ) High SAH clot burden* 1.54 ( ) 1.42 ( ) WFNS 4.80 ( ) 4.50 ( ) Clipping (vs coiling) 1.24 ( ) 1.15 ( ) ACA (vs posterior) 2.65 ( ) 1.23 ( ) ICA (vs posterior) 1.14 ( ) 0.82 ( ) MCA (vs posterior) 2.00 ( ) 1.85 ( ) Treatment With Clazosentan 1 mg/h vs placebo 0.90 ( ) 0.80 ( ) 5 mg/h vs placebo 0.95 ( ) 0.89 ( ) 15 mg/h vs placebo 2.26 ( ) 1.24 ( ) Angiographic vasospasm 2.19 ( ) 1.81 ( ) DIND 4.13 ( ) 2.17 ( ) Hypokalemia 2.72 ( ) NA Hyperkalemia Too few events NA ECG findings Bradycardia (vs normal rate) 1.59 ( ) 1.44 ( ) Tachycardia (vs normal rate) 4.36 ( ) 2.83 ( ) ST changes 0.94 ( ) 2.90 ( ) Prolonged QT 1.44 ( ) 0.84 ( ) Nodal rhythm 1.22 ( ) 1.96 ( ) T-wave changes 1.32 ( ) 2.31 ( ) ECG indicates electrocardiogram; SAH, subarachnoid hemorrhage; WFNS, World Federation of Neurosurgical Societies; ACA, anterior cerebral artery; ICA, internal carotid artery; MCA, middle cerebral artery; DIND, delayed ischemic neurological deficit; NA, not available. *As determined by the Hijdra scale dichotomized by the mean. Denotes significance at P nerve fibers mainly originating in the superior cervical ganglion. 15,16 Although the pathogenesis of vasospasm is complex and only partially understood, there is evidence to suggest that sympathetic hyperactivity is 1 mechanism contributing to this phenomenon. In 1 study, patients with symptomatic cerebral vasospasm showed improvements in cerebral perfusion following ipsilateral locoregional cervical sympathetic block. 5 Furthermore, Yasargil performed surgical ablation of periadventitial sympathetic fibers, resulting in chronic denervation during craniotomy for ruptured aneurysms, and reported low rates of angiographic vasospasm. 17 It follows, therefore, that ECG manifestation of increased central activity, such as depolarization abnormalities, may serve as surrogate markers of cerebrovascular vasospasm. Interestingly, despite improvements in perfusion in the aforementioned study, proximal vessel caliber did not increase following sympathetic

5 2106 Stroke August 2012 block. 5 Analogously, stellate ganglion blockade did not seem to affect ECG changes following SAH. 18 The finding that QT prolongation is the most prevalent ECG finding after SAH is in keeping with published series Prospective studies have previously concluded that ECG changes are mainly seen in the early stages 22,23 ; however, we show that the proportion of certain findings (tachycardia and nonspecific ST-changes) actually increase in interval between baseline and the vasospasm risk period. In our multivariate model, we included numerous variables that have been previously linked to ECG changes after SAH following a thorough literature review, including sex 24 and aneurysm location. 25,26 Previous studies have linked various ECG findings to poor outcomes on multivariate analysis, including cardiac arrhythmias (prevalence of approximately 4% in 1 series). 27 Other ECG findings linked to poor outcome include bradycardia, 28 relative tachycardia, and ST- and T- wave abnormalities One observational study identified a tilt toward depressed sympathovagal balance, as determined by heart rate variability, as a contributor to poor outcome after SAH. 34 To our knowledge, only 1 previous study identified the QT interval as an independent predictor of in-hospital mortality 31 ; however, as the investigators did not account for vasospasm or DIND, it is difficult to know whether patients with QT interval were less likely to survive because of symptomatic vasospasm and DIND. In contrast, another previous study corroborated our results that QT prolongation is not associated with outcome after SAH. 21 The association in the current study between prolonged QT interval and angiographic vasospasm may explain the findings of Ichinomiya and colleagues who found that a prolonged QT interval ( 448 ms) at 7 days postaneurysm rupture was a predictor of neurological outcome. 35 This group furthermore found that improvement in QT prolongation was associated with favorable outcome, suggesting a role for angiographic vasospasm as a mechanistic explanation. A significant limitation of previous studies is the lack of assessment of angiographic data to diagnose vasospasm. For example, Brouwers and colleagues found that fast rhythm disturbances correlated with poor outcome, but not cerebral ischemia; but, they did not directly assess arterial luminal narrowing. 30 The current article therefore bridges a gap in knowledge between ECG findings and angiographic vasospasm. It has been postulated that in the presence of known vasospasm, ECG abnormalities may identify patients who will go on to develop cerebral ischemia. 36 This may be related to poor cardiac output as a result of neurogenic stunned myocardium, which may increase the risk of delayed cerebral ischemia from existing vasospasm. Supporting the physiological role for ECG abnormalities as markers of poor cardiac output, it has been previously reported that QT dispersion may be a marker of cardiorespiratory compromise 37 ; and QT prolongation, in addition to T-wave inversion, is a sensitive marker of abnormal wall motion on echocardiography after SAH. 6 However, in a prospective study of 121 patients designed to test the association between cardiac abnormalities and delayed cerebral ischemia, no such relationship was identified. 36 Similarly, we find that no ECG abnormalities were associated with risk of DIND in the current patient cohort, although several ECG findings were associated with the presence of angiographic vasospasm. The main limitation of this study is our inability to account for the administration of cardiotropic drugs, which may influence ECG changes; however, in 1 study, ECG changes were independent of plasma norepinephrine levels. 38 Another limitation is that clazosentan was associated with a higher incidence of hypotension than was placebo in CONSCIOUS-1, and this might influence ECG changes. Conversely, analysis only of the placebo patients had similar findings in this data. Clazosentan is not known to have any direct effects on ECG. Our strengths include the availability of catheter angiography on all subjects for the evaluation of angiographic vasospasm and systematic and consistent collection of outcome variables. Furthermore, we present one of the largest series evaluating ECG changes in SAH patients, and the only such series evaluating the association between ECG change and vasospasm. Conclusion Abnormalities on ECG, such as QT prolongation and tachycardia, are independently associated with angiographic vasospasm. Tachycardia and nonspecific ST-segment changes are also significantly associated with poor outcome. None. Sources of Funding Disclosures Actelion Pharmaceuticals Ltd was the sponsor of the CONSCIOUS-1 trial; the company provided the authors with the trial data set, but had no role in this exploratory analysis nor in the development of the article. The data analysis and writing are the work of the authors. Dr Macdonald is a consultant for Actelion and is chief scientific officer of Edge Therapeutics, Inc. References 1. Rinkel GJ. Medical management of patients with aneurysmal subarachnoid haemorrhage. Int J Stroke. 2008;3: Fisher CM, Kistler JP, Davis JM. Relation of cerebral vasospasm to subarachnoid hemorrhage visualized by computerized tomographic scanning. Neurosurgery. 1980;6: Ibrahim GM, Vachhrajani S, Ilodigwe D, Stat M, Kassell NF, Mayer SA, et al. Method of aneurysm treatment does not affect clot clearance after aneurysmal subarachnoid hemorrhage. Neurosurgery. 2012;70: Bunc G, Kovacic S, Strnad S. Sympathetic nervous system exclusion following experimental subarachnoid haemorrhage prevents vasospasm in rabbits. Wien Klin Wochenschr. 2000;112: Treggiari MM, Romand JA, Martin JB, Reverdin A, Rufenacht DA, de Tribolet N. Cervical sympathetic block to reverse delayed ischemic neurological deficits after aneurysmal subarachnoid hemorrhage. Stroke. 2003;34: Mayer SA, Lin J, Homma S, Solomon RA, Lennihan L, Sherman D, et al. Myocardial injury and left ventricular performance after subarachnoid hemorrhage. Stroke. 1999;30: Macdonald RL, Kassell NF, Mayer S, Ruefenacht D, Schmiedek P, Weidauer S, et al. Clazosentan to overcome neurological ischemia and infarction occurring after subarachnoid hemorrhage (CONSCIOUS-1): randomized, double-blind, placebo-controlled phase 2 dose-finding trial. Stroke. 2008;39: Teasdale GM, Drake CG, Hunt W, Kassell N, Sano K, Pertuiset B, et al. A universal subarachnoid hemorrhage scale: report of a committee of the World Federation of Neurosurgical Societies. J Neurol Neurosurg Psychiatry. 1988; 51:1457.

6 Ibrahim and Macdonald ECG Findings and Vasospasm After Aneurysmal SAH Brott T, Adams HP Jr, Olinger CP, Marler JR, Barsan WG, Biller J, et al. Measurements of acute cerebral infarction: a clinical examination scale. Stroke. 1989;20: Farrell B, Godwin J, Richards S, Warlow C. The United Kingdom transient ischaemic attack (UK-TIA) aspirin trial: final results. J Neurol Neurosurg Psychiatry. 1991;54: Fukui S, Katoh H, Tsuzuki N, Ishihara S, Otani N, Ooigawa H, et al. Multivariate analysis of risk factors for QT prolongation following subarachnoid hemorrhage. Crit Care. 2003;7:R7 R Hijdra A, Brouwers PJ, Vermeulen M, van Gijn J. Grading the amount of blood on computed tomograms after subarachnoid hemorrhage. Stroke. 1990;21: Lee VH, Oh JK, Mulvagh SL, Wijdicks EF. Mechanisms in neurogenic stress cardiomyopathy after aneurysmal subarachnoid hemorrhage. Neurocrit Care. 2006;5: Nguyen H, Zaroff JG. Neurogenic stunned myocardium. Curr Neurol Neurosci Rep. 2009;9: Tuor UI. Local distribution of the effects of sympathetic stimulation on cerebral blood flow in the rat. Brain Res. 1990;529: Kurth CD, Wagerle LC, Delivoria-Papadopoulos M. Sympathetic regulation of cerebral blood flow during seizures in newborn lambs. Am J Physiol. 1988;255:H563 H Yasargil MG. Microneurosurgery, vol. 2: clinical considerations, surgery of the intracranial aneurysms and results (microsurgical anatomy of basal cisterns & vessels of brain). New York, NY: Thieme, Estanol B, Aguilar F, Badui E, Cesarman E. Stellate ganglion blockade and atropine do not affect the electrocardiographic changes of subarachnoid hemorrhage. Arch Invest Med (Mex). 1984;15: Frangiskakis JM, Hravnak M, Crago EA, Tanabe M, Kip KE, Gorcsan J 3rd, et al. Ventricular arrhythmia risk after subarachnoid hemorrhage. Neurocrit Care. 2009;10: Hravnak M, Frangiskakis JM, Crago EA, Chang Y, Tanabe M, Gorcsan J 3rd, et al. Elevated cardiac troponin I and relationship to persistence of electrocardiographic and echocardiographic abnormalities after aneurysmal subarachnoid hemorrhage. Stroke. 2009;40: Goldstein DS. The electrocardiogram in stroke: relationship to pathophysiological type and comparison with prior tracings. Stroke. 1979;10: Di Pasquale G, Lusa AM, Manini GL, Dominici P, Andreoli A, Limoni P, et al. [Cardiac arrhythmias associated with subarachnoid hemorrhage. Prospective study with dynamic electrocardiography]. G Ital Cardiol. 1984;14: Di Pasquale G, Pinelli G, Andreoli A, Manini G, Grazi P, Tognetti F. Holter detection of cardiac arrhythmias in intracranial subarachnoid hemorrhage. Am J Cardiol. 1987;59: Fukui S, Otani N, Katoh H, Tsuzuki N, Ishihara S, Ohnuki A, et al. Female gender as a risk factor for hypokalemia and QT prolongation after subarachnoid hemorrhage. Neurology. 2002;59: Andreoli A, di Pasquale G, Pinelli G, Grazi P, Tognetti F, Testa C. Subarachnoid hemorrhage: frequency and severity of cardiac arrhythmias. A survey of 70 cases studied in the acute phase. Stroke. 1987;18: Hirashima Y, Takashima S, Matsumura N, Kurimoto M, Origasa H, Endo S. Right sylvian fissure subarachnoid hemorrhage has electrocardiographic consequences. Stroke. 2001;32: Frontera JA, Parra A, Shimbo D, Fernandez A, Schmidt JM, Peter P, et al. Cardiac arrhythmias after subarachnoid hemorrhage: risk factors and impact on outcome. Cerebrovasc Dis. 2008;26: Coghlan LA, Hindman BJ, Bayman EO, Banki NM, Gelb AW, Todd MM, et al. Independent associations between electrocardiographic abnormalities and outcomes in patients with aneurysmal subarachnoid hemorrhage: findings from the Intraoperative Hypothermia Aneurysm Surgery Trial. Stroke. 2009;40: Melin J, Fogelholm R. Electrocardiographic findings in subarachnoid hemorrhage. A population study. Acta Med Scand. 1983;213: Brouwers PJ, Wijdicks EF, Hasan D, Vermeulen M, Wever EF, Frericks H, et al. Serial electrocardiographic recording in aneurysmal subarachnoid hemorrhage. Stroke. 1989;20: Huang CC, Huang CH, Kuo HY, Chan CM, Chen JH, Chen WL. The 12-lead electrocardiogram in patients with subarachnoid hemorrhage: early risk prognostication. Am J Emerg Med Jun 3. [Epub ahead of print]. 32. Kawasaki T, Azuma A, Sawada T, Sugihara H, Kuribayashi T, Satoh M, et al. Electrocardiographic score as a predictor of mortality after subarachnoid hemorrhage. Circ J. 2002;66: Sakr YL, Lim N, Amaral AC, Ghosn I, Carvalho FB, Renard M, et al. Relation of ECG changes to neurological outcome in patients with aneurysmal subarachnoid hemorrhage. Int J Cardiol. 2004;96: Chiu TF, Huang CC, Chen JH, Chen WL. Depressed sympathovagal balance predicts mortality in patients with subarachnoid hemorrhage. Am J Emerg Med May 11. [Epub ahead of print]. 35. Ichinomiya T, Terao Y, Miura K, Higashijima U, Tanise T, Fukusaki M, et al. QTc interval and neurological outcomes in aneurysmal subarachnoid hemorrhage. Neurocrit Care. 2010;13: Schuiling WJ, Algra A, de Weerd AW, Leemans P, Rinkel GJ. ECG abnormalities in predicting secondary cerebral ischemia after subarachnoid haemorrhage. Acta Neurochir (Wien). 2006;148: ; discussion 858. Epub 2006 Jun Macmillan CS, Andrews PJ, Struthers AD. QTc dispersion as a marker for medical complications after severe subarachnoid haemorrhage. Eur J Anaesthesiol. 2003;20: Grad A, Kiauta T, Osredkar J. Effect of elevated plasma norepinephrine on electrocardiographic changes in subarachnoid hemorrhage. Stroke. 1991;22: