Risk factors

Angiographic vasospasm is seen in 30% to 70% of patients post aSAH ; typically it can be expected to start after postbleed day 3, although hyperacute or early vasospasm has been reported. Symptoms of cerebral ischemia with high risk for debilitating stroke and mortality are experienced by 20% to 30% of patients. The presence and the amount of oxyhemoglobin in the subarachnoid cisterns is believed to be the major trigger of the phenomena that ultimately cause smooth muscle spasm, narrowing of the arterial lumen, and impaired blood flow autoregulation. In their seminal paper, Fisher and colleagues found a strong correlation linking thick cisternal clot with angiographic and clinical vasospasm. The Fisher computed tomographic (CT) rating scale is widely used by neurointensivists and neurosurgeons and has been recently modified to incorporate intraventricular hemorrhage as a significant predictor for vasospasm and also to denote increasing risk as the grade increases. Techniques to remove blood or increase clearance of blood from the basal cisterns with either intracisternal or intrathecal lysis, head shaking, and lumbar drainage have been attempted with variable results. Other potential risk factors include poor clinical grade, early angiographic spasm, history of hypertension, and admission mean arterial pressure (MAP). There have been conflicting reports regarding age as a predictor, with one study identifying age less than 35 years as a risk factor, although this finding was not confirmed by others. One prospective study of 70 patients demonstrated that apart from thick subarachnoid clot, a history of smoking was independently associated with development of symptomatic spasm. Volume status of the patient with aSAH is considered critical, and a large part of critical care in this disease centers on its regulation. Hypovolemia is believed to be a potentially significant contributor to DCI and can be common if not prevented, especially in the presence of natriuresis secondary to cerebral salt-wasting syndrome (CSWS).

Prevention and volume management

Current guidelines advise maintenance of normal circulating blood volume instead of prophylactic hyperdynamic, hypervolemic therapy. Lennihan and colleagues randomized patients with aSAH into hypervolemic versus normovolemic regimens based on the measurements of pulmonary artery diastolic pressures (PADPs) for the first 3 days and central venous pressure (CVP) measurements thereafter, and until day 14, they measured CBF using xenon (Xe) washout. There was no difference between the 2 groups in mean global CBF, rate of symptomatic spasm, or functional outcome. Subsequently, Egge and colleagues published similar results in their randomized prospective trial that compared hypervolemic to normovolemic approaches, finding no difference in the occurrence of vasospasm, TCD ultrasonography recordings, or SPECT (single-photon emission computed tomography) CBF measurements. These studies, despite the small number of patients, suggest that euvolemia should be the goal because extra volume translates neither to an increase in CBF nor to improved outcomes. Importantly, fluid management should also take into account the not-uncommon presence of cardiomyopathy and neurogenic pulmonary edema. Even moderate volume overload can lead to further lung and cerebral edema in these patients, and positive fluid balance has been associated with increased mortality in neurologic and general critical care populations. This discussion raises the question of volume assessment to guide therapy. It is common practice to calculate daily fluid balance (DFB) as a measure of the need for more or less fluid administration, but the correlation of DFB with actual circulating blood volume as measured by integrated pulse spectrophotometry and pulse dye densitometry has been shown to be poor. As a consequence, several institutional protocols for the management of aSAH call for insertion of central venous and/or pulmonary artery catheters for the measurement of CVP, PADP, and pulmonary artery occlusion pressures (PAOPs) as measures of right and left heart preload and also for cardiac output (CO) estimations. The major limitation of this approach, apart from its being invasive, relates to the inaccuracy of extrapolating cardiac filling pressures to volumetric assessments. This inaccuracy is accentuated when cardiac compliance is altered, as may be seen with neurogenic stunned myocardium. The failure of these static pressures to predict volume responsiveness has been demonstrated across the spectrum from healthy volunteers to critically ill mechanically ventilated (MV) patients with sepsis; accordingly, dynamic parameters, such as systolic pressure variation and pulse pressure variation for MV patients, are recommended. An alternative for advanced hemodynamic monitoring is a device that combines single indicator transpulmonary thermodilution technique and pulse contour continuous CO measurements (PiCCO, PULSION Medical Systems AG, Munich, Germany). This device has been used extensively to guide management of different populations of critically ill patients, including those with conditions such as septic and cardiogenic shock and acute respiratory distress syndrome, and it has been used for management in the operating theater. The potential theoretical benefits are direct volumetric measurements of intrathoracic blood volume (ITBV), global end diastolic volume (GEDV), and extravascular lung water (EVLW) volume and also continuous dynamic volume responsiveness parameter (stroke volume variation [SVV]) and CO monitoring. It does require positive pressure MV and minimal spontaneous breathing efforts. The device has been used in patients with aSAH and was found to be a useful tool for volume and hemodynamic augmentation (HA) management. The authors also use PiCCO for selected patients and to target normovolemia goals (GEDV index, 680–800 mL/m ² ; ITBV index, 850–1000 mL/m ² ; SVV ≤ 10%; and EVLW index ≤ 10 mL/kg). It remains to be seen in a prospective trial if this device proves more useful over traditional measures such as DFB and cardiac filling pressures in patients with aSAH. A last comment on the prevention of hypovolemia concerns the occurrence of CSWS. Despite an incomplete understanding of the pathophysiology of the syndrome, it is considered when large urinary output is accompanied by hyponatremia. A similar clinical picture can be seen secondary to iatrogenic reasons such as overzealous fluid administration and the use of natural diuretics, such as hypertonic saline (HTS). Fludrocortisone is often used as an adjunct to volume and sodium replacement in CSWS. It has been evaluated in 2 randomized controlled trials (RCTs) as a means to prevent hyponatremia and volume contraction, with mixed results. An alternative or supplementary fluid management technique is to use colloids, such as 5% albumin, not only as a volume expander but also to potentially prevent sodium and fluid losses associated with CSWS.

Neuroprotection

Nimodipine administration from the time of admission and for 21 days is considered the standard of care and is the only recommendation carrying a class I, level A evidence grade in current guidelines. A recent Cochrane review analyzed a total of 12 studies on calcium antagonists (heavily weighted by a single large trial of nimodipine ) and found an outcome improvement with a relative risk reduction of 18% (95% confidence interval [CI], 7%–28%) and an absolute risk reduction of 5.1%. Treatment with nimodipine may prevent 1 poor outcome in every 13 patients with aSAH. However, the medication does not prevent vasospasm and is believed to improve outcome through a neuroprotective mechanism. Alternative explanations have been proposed to explain this beneficial effect, including enhanced fibrinolysis and the observation that nimodipine transforms cortical spreading ischemia back to cortical spreading hyperemia. In recent years, many centers have incorporated the use of HMG-CoA (3-hydroxy-3-methyl-glutaryl-coenzyme A) reductase inhibitors such as “statins” in their standard armamentarium in the treatment of patients with aSAH. A meta-analysis by Sillberg and colleagues included 3 double-blind RCTs of statin versus placebo and found significantly reduced incidence of vasospasm (relative risk [RR] 0.73; 95% CI, 0.54–0.99, number needed to treat [NNT] 6.25), delayed ischemic deficits (RR 0.38; 95% CI, 0.17–0.83, NNT 5), and mortality (RR 0.22; 95% CI, 0.06–0.82, NNT 6.7). All 3 trials have included small numbers of patients, and there is heterogeneity in regards to primary end points. Furthermore, 2 large retrospective studies have reported no benefits from statin use in vasospasm incidence or clinical outcomes. The potentially favorable benefit-risk ratio of statins makes them attractive for wide use in aSAH; the authors hope that future large RCTs such as STASH, which is a multicenter placebo-controlled double-blinded phase 3 trial assessing the clinical benefit of SimvaSTatin in Aneurysmal Subarachnoid Hemorrhage, will provide a definitive answer. The hypothesis is that simvastatin 40 mg given within 96 hours of ictus over 3 weeks reduces the incidence and duration of DCI after aSAH when compared with placebo (Dr Peter Kirkpatrick, chief investigator, University of Cambridge, UK).

As mentioned earlier, DIND seems to be the end result of multiple cooperating mechanisms, and relieving angiographic vessel narrowing does not necessarily translate to clinical improvement. The endothelin receptor-A antagonist (clazosentan) studies may provide another valuable clue in dissociating angiographic vasospasm, clinical outcomes, and DCI. CONSCIOUS-1 was a randomized, double-blind, placebo-controlled phase 2 dose-finding trial of intravenous clazosentan with the aim of preventing vasospasm in patients with aSAH. Clazosentan significantly decreased moderate and severe angiographic vasospasm in a dose-dependent manner; nevertheless, no significant benefit on any morbidity or mortality end points was observed. It is possible that the study was underpowered, and a phase 3 clinical trial (CONSCIOUS-2) is designed to focus on clinical outcomes in patients undergoing aneurysm clipping receiving placebo or 5 mg/h of clazosentan. This lack of a clinical effect has led certain investigators to further challenge conventional notions and question if angiographic vasospasm is no more than an epiphenomenon.

Other medical therapies that have been evaluated for the prevention of vasospasm and poor outcomes include the nonglucocorticoid 21-aminosteroid tirilazad, magnesium, aspirin, low molecular weight heparin, nitroglycerin, and nitric oxide donors. Meta-analysis of the tirilazad mesylate study included 3797 patients and found no effect on clinical outcome despite a decrease in symptomatic vasospasm. Magnesium therapy has been studied in a large placebo-controlled trial of continuous intravenous infusion for 14 days with promising results (Magnesium and Acetylsalicylic acid in Subarachnoid Hemorrhage [MASH]). Van den Bergh and colleagues noticed a reduction in poor outcomes at 3 months by 23%, and the RR of a good outcome was 3.4 (95% CI, 1.3–8.9) for treated patients. A (MASH II) phase 3 clinical trial is currently under way with an aim to include 1200 patients before 2010 to further define the role of intravenous magnesiun infusion in patients with aSAH.

Prevention and volume management

Current guidelines advise maintenance of normal circulating blood volume instead of prophylactic hyperdynamic, hypervolemic therapy. Lennihan and colleagues randomized patients with aSAH into hypervolemic versus normovolemic regimens based on the measurements of pulmonary artery diastolic pressures (PADPs) for the first 3 days and central venous pressure (CVP) measurements thereafter, and until day 14, they measured CBF using xenon (Xe) washout. There was no difference between the 2 groups in mean global CBF, rate of symptomatic spasm, or functional outcome. Subsequently, Egge and colleagues published similar results in their randomized prospective trial that compared hypervolemic to normovolemic approaches, finding no difference in the occurrence of vasospasm, TCD ultrasonography recordings, or SPECT (single-photon emission computed tomography) CBF measurements. These studies, despite the small number of patients, suggest that euvolemia should be the goal because extra volume translates neither to an increase in CBF nor to improved outcomes. Importantly, fluid management should also take into account the not-uncommon presence of cardiomyopathy and neurogenic pulmonary edema. Even moderate volume overload can lead to further lung and cerebral edema in these patients, and positive fluid balance has been associated with increased mortality in neurologic and general critical care populations. This discussion raises the question of volume assessment to guide therapy. It is common practice to calculate daily fluid balance (DFB) as a measure of the need for more or less fluid administration, but the correlation of DFB with actual circulating blood volume as measured by integrated pulse spectrophotometry and pulse dye densitometry has been shown to be poor. As a consequence, several institutional protocols for the management of aSAH call for insertion of central venous and/or pulmonary artery catheters for the measurement of CVP, PADP, and pulmonary artery occlusion pressures (PAOPs) as measures of right and left heart preload and also for cardiac output (CO) estimations. The major limitation of this approach, apart from its being invasive, relates to the inaccuracy of extrapolating cardiac filling pressures to volumetric assessments. This inaccuracy is accentuated when cardiac compliance is altered, as may be seen with neurogenic stunned myocardium. The failure of these static pressures to predict volume responsiveness has been demonstrated across the spectrum from healthy volunteers to critically ill mechanically ventilated (MV) patients with sepsis; accordingly, dynamic parameters, such as systolic pressure variation and pulse pressure variation for MV patients, are recommended. An alternative for advanced hemodynamic monitoring is a device that combines single indicator transpulmonary thermodilution technique and pulse contour continuous CO measurements (PiCCO, PULSION Medical Systems AG, Munich, Germany). This device has been used extensively to guide management of different populations of critically ill patients, including those with conditions such as septic and cardiogenic shock and acute respiratory distress syndrome, and it has been used for management in the operating theater. The potential theoretical benefits are direct volumetric measurements of intrathoracic blood volume (ITBV), global end diastolic volume (GEDV), and extravascular lung water (EVLW) volume and also continuous dynamic volume responsiveness parameter (stroke volume variation [SVV]) and CO monitoring. It does require positive pressure MV and minimal spontaneous breathing efforts. The device has been used in patients with aSAH and was found to be a useful tool for volume and hemodynamic augmentation (HA) management. The authors also use PiCCO for selected patients and to target normovolemia goals (GEDV index, 680–800 mL/m ² ; ITBV index, 850–1000 mL/m ² ; SVV ≤ 10%; and EVLW index ≤ 10 mL/kg). It remains to be seen in a prospective trial if this device proves more useful over traditional measures such as DFB and cardiac filling pressures in patients with aSAH. A last comment on the prevention of hypovolemia concerns the occurrence of CSWS. Despite an incomplete understanding of the pathophysiology of the syndrome, it is considered when large urinary output is accompanied by hyponatremia. A similar clinical picture can be seen secondary to iatrogenic reasons such as overzealous fluid administration and the use of natural diuretics, such as hypertonic saline (HTS). Fludrocortisone is often used as an adjunct to volume and sodium replacement in CSWS. It has been evaluated in 2 randomized controlled trials (RCTs) as a means to prevent hyponatremia and volume contraction, with mixed results. An alternative or supplementary fluid management technique is to use colloids, such as 5% albumin, not only as a volume expander but also to potentially prevent sodium and fluid losses associated with CSWS.

Neuroprotection

Nimodipine administration from the time of admission and for 21 days is considered the standard of care and is the only recommendation carrying a class I, level A evidence grade in current guidelines. A recent Cochrane review analyzed a total of 12 studies on calcium antagonists (heavily weighted by a single large trial of nimodipine ) and found an outcome improvement with a relative risk reduction of 18% (95% confidence interval [CI], 7%–28%) and an absolute risk reduction of 5.1%. Treatment with nimodipine may prevent 1 poor outcome in every 13 patients with aSAH. However, the medication does not prevent vasospasm and is believed to improve outcome through a neuroprotective mechanism. Alternative explanations have been proposed to explain this beneficial effect, including enhanced fibrinolysis and the observation that nimodipine transforms cortical spreading ischemia back to cortical spreading hyperemia. In recent years, many centers have incorporated the use of HMG-CoA (3-hydroxy-3-methyl-glutaryl-coenzyme A) reductase inhibitors such as “statins” in their standard armamentarium in the treatment of patients with aSAH. A meta-analysis by Sillberg and colleagues included 3 double-blind RCTs of statin versus placebo and found significantly reduced incidence of vasospasm (relative risk [RR] 0.73; 95% CI, 0.54–0.99, number needed to treat [NNT] 6.25), delayed ischemic deficits (RR 0.38; 95% CI, 0.17–0.83, NNT 5), and mortality (RR 0.22; 95% CI, 0.06–0.82, NNT 6.7). All 3 trials have included small numbers of patients, and there is heterogeneity in regards to primary end points. Furthermore, 2 large retrospective studies have reported no benefits from statin use in vasospasm incidence or clinical outcomes. The potentially favorable benefit-risk ratio of statins makes them attractive for wide use in aSAH; the authors hope that future large RCTs such as STASH, which is a multicenter placebo-controlled double-blinded phase 3 trial assessing the clinical benefit of SimvaSTatin in Aneurysmal Subarachnoid Hemorrhage, will provide a definitive answer. The hypothesis is that simvastatin 40 mg given within 96 hours of ictus over 3 weeks reduces the incidence and duration of DCI after aSAH when compared with placebo (Dr Peter Kirkpatrick, chief investigator, University of Cambridge, UK).

As mentioned earlier, DIND seems to be the end result of multiple cooperating mechanisms, and relieving angiographic vessel narrowing does not necessarily translate to clinical improvement. The endothelin receptor-A antagonist (clazosentan) studies may provide another valuable clue in dissociating angiographic vasospasm, clinical outcomes, and DCI. CONSCIOUS-1 was a randomized, double-blind, placebo-controlled phase 2 dose-finding trial of intravenous clazosentan with the aim of preventing vasospasm in patients with aSAH. Clazosentan significantly decreased moderate and severe angiographic vasospasm in a dose-dependent manner; nevertheless, no significant benefit on any morbidity or mortality end points was observed. It is possible that the study was underpowered, and a phase 3 clinical trial (CONSCIOUS-2) is designed to focus on clinical outcomes in patients undergoing aneurysm clipping receiving placebo or 5 mg/h of clazosentan. This lack of a clinical effect has led certain investigators to further challenge conventional notions and question if angiographic vasospasm is no more than an epiphenomenon.

Other medical therapies that have been evaluated for the prevention of vasospasm and poor outcomes include the nonglucocorticoid 21-aminosteroid tirilazad, magnesium, aspirin, low molecular weight heparin, nitroglycerin, and nitric oxide donors. Meta-analysis of the tirilazad mesylate study included 3797 patients and found no effect on clinical outcome despite a decrease in symptomatic vasospasm. Magnesium therapy has been studied in a large placebo-controlled trial of continuous intravenous infusion for 14 days with promising results (Magnesium and Acetylsalicylic acid in Subarachnoid Hemorrhage [MASH]). Van den Bergh and colleagues noticed a reduction in poor outcomes at 3 months by 23%, and the RR of a good outcome was 3.4 (95% CI, 1.3–8.9) for treated patients. A (MASH II) phase 3 clinical trial is currently under way with an aim to include 1200 patients before 2010 to further define the role of intravenous magnesiun infusion in patients with aSAH.

Diagnosis and multimodality neuromonitoring

Before the discussion of HA as the mainstay of medical management, the diagnosis and neuromonitoring of vasospasm and DCI are reviewed. The gold standard for detection of angiographic vessel narrowing is conventional digital-subtraction angiography (DSA). When clinical symptoms correlate with an area of narrowing on DSA, the diagnosis of clinical vasospasm is made. It should be noted that the presence of large vessel narrowing is not necessary for DCI to occur; in fact, Rabinstein and colleagues reported that the presence and location of angiographically demonstrated vasospasm failed to correlate with areas of cerebral infarction in as many as one-third of their cases. Transcranial Doppler (TCD) ultrasonography is commonly used on a daily basis in the neurocritical care unit to follow patients with aSAH and with moderate to high risk for DCI.

The American Academy of Neurology expert committee has given a Type A, Class II level of evidence supporting the use of TCD ultrasonography in diagnosis of severe spasm. A meta-analysis of 7 trials out of 26 reports evaluated the accuracy of TCD ultrasonography as compared with DSA. For the middle cerebral artery (MCA), sensitivity of TCD ultrasonography was 67% and specificity was 99%, with a positive predictive value of 97% and negative predictive value of 78%. The accuracy of TCD ultrasonography was considerably less for detecting spasm in vessels other than the MCA. The noninvasiveness, ease, and wide availability have made TCD ultrasonography the most common neuromonitor for patients with aSAH. As cautioned before and in relation to DSA, Minhas and colleagues observed no correlation between positron emission tomography (PET) and TCD ultrasonography among patients who developed delayed neurologic deficits after aSAH. They concluded that TCD ultrasonography–derived indices correlate poorly with cerebral perfusion values.

TCD ultrasonography, apart from measurement of flow velocities, can also be used to characterize the state of pressure autoregulation, which has been shown to be deranged in patients with aSAH. Soehle and colleagues calculated and followed the moving correlation coefficient between slow changes of arterial blood pressure (ABP) and mean (Mx) or systolic flow velocity. The investigators demonstrated an increase in Mx during vasospasm reflecting a derangement of cerebral pressure autoregulation. The authors have mentioned earlier the potentially beneficial effect of statins in preventing DCI and improving outcomes. A plausible explanation of this effect was published by the Cambridge group and it relates to an improvement in the state of pressure-flow autoregulation as measured by the transient hyperemic response test (TCD ultrasonography derived).

Using brain tissue oxygen (PtiO ₂ ) as a surrogate for CBF, Jaeger and colleagues found deranged CBF-autoregulation that does not improve after aSAH-ictus to be closely associated to the development of DCI. They calculated ORx (oxygen reactivity index), which is the moving linear (Pearson) correlation coefficient between the values of cerebral perfusion pressure (CPP) and PtiO ₂ and varies between −1 and +1. The more positive, the more it indicates a passive relationship between CBF and MAP/CPP, meaning a pressure-passive nonreactive vascular bed. Of note, PtiO ₂ alone was not different between the DCI and non-DCI groups. The investigation of vascular reactivity and pressure autoregulation indices in patients with aSAH is fascinating and potentially it may yield an early marker for detection of DCI before clinical symptoms ensue. The investigation also provides alternative mechanisms and therapeutic targets for DCI, placing the focus from the proximal segments of the circle of Willis to the microcirculation responsible for CBF regulation.

Microdialysis (MD) is increasingly used in the neuromonitoring of patients with severe TBI and aSAH. A consensus meeting on MD based on the available literature noted that glutamate was found to be the earliest marker of the onset of vasospasm followed over time by lactate, the lactate/pyruvate (L/P) ratio, and glycerol. Sarrafzadeh and colleagues compared MD with PET in 15 patients with aSAH and found glutamate to have the closest correlation with regional CBF (rCBF). Lactate, L/P ratio, and glycerol were significantly higher in symptomatic patients. It is also of interest to note that in this same study and in most symptomatic patients the measured PET-rCBF values were higher than the accepted critical thresholds of ischemia. In an earlier comparison of MD with TCD ultrasonography and DSA by the same group of investigators, MD was shown to be more specific but less sensitive as a diagnostic tool for DIND. Brain tissue oxygenation is actively being researched, especially in TBI, and several centers use it routinely to prevent, detect, and treat secondary brain insults. Retrospective data from a prospective database of patients with aSAH were reported. The investigators observed an association of lower PtiO ₂ with mortality. More specifically, low PtiO ₂ on the first day of monitoring, lower mean daily PtiO ₂ , lower mean minimum PtiO ₂ , and longer cumulative duration of compromised PtiO ₂ tend to be associated with an increased mortality rate at 1 month after aSAH in this cohort. Finally, perfusion imaging in patients with aSAH using PET, SPECT, MRI, and CT methodologies is actively investigated. CT can provide expediently combined computed tomography angiogram and dynamic computed tomographic perfusion (CTP) scans and is becoming increasingly used for the diagnosis of DCI. CTP is based on the central volume principle, which states that the CBF value is the ratio of the blood volume within all blood vessels in a given volume of tissue (cerebral blood volume [CBV], which is measured in milliliters per gram) to the mean transit time (MTT, measured in seconds) of the contrast agent, from the arterial input to the venous drainage, within the volume being evaluated (CBF = CBV/MTT). Recent articles are finding MTT to be an early predictor for DCI and angiographic vasospasm in animal models and human subjects. In addition, relative CBF and MTT values have correlated well with estimated rCBF as measured by SPECT in patients with vasospasm after aSAH.