Perioperative management of aneurysmal subarachnoid hemorrhage: Part 2. Postoperative management.

Abstract

Although aneurysm clipping removes the risk of aneurysmal rebleeding as a cause of neurologic deterioration after subarachnoid hemorrhage (SAH), during the subsequent 2 to 3 wk patients remain at risk for cerebral vasospasm, hydrocephalus, seizures, and other complications that pose a threat to neurologic and medical recovery. Cerebral Vasospasm Pathophysiology Delayed cerebral ischemia with subsequent infarction is the leading cause of death or major disability among patients surviving to hospitalization. The most common cause of delayed ischemia is cerebral arterial vasospasm [1]. Vasospasm after SAH has been identified by angiography in up to 60% of patients [2]. Patients suffering from vasospasm are more likely to incur death or disability, and patients dying of SAH are more likely to have severe vasospasm than are survivors [3]. For these reasons, the prevention and treatment of vasospasm have emerged as major goals in the management of patients with ruptured intracranial aneurysms. A global reduction in cerebral blood flow (CBF) occurs among patients with vasospasm with the greatest reduction in areas of the brain responsible for neurologic deficits. Once CBF falls below a critical threshold, ischemia occurs with resultant neurologic deficits determined in part by the neuronal territory supplied by the affected vessel(s) [4]. CBF values less than 20 mL centered dot 100 g-1 centered dot min-1 as measured by positron emission tomography scanning have been found in the affected hemispheres of patients with symptomatic vasospasm; values less than 12 mL centered dot min-1 centered dot 100 g have been associated with irreversible deficits [5]. Some evidence suggests that a primary defect in cerebral oxygen utilization follows SAH [6]. CBF autoregulation is impaired, and the degree of abnormality correlates with the severity of spasm [4]. Loss of autoregulation causes CBF in the territory supplied by the affected vessels to become perfusion pressure dependent, which may explain the utility of hypertensive therapy to reverse ischemic deficits in some patients [7,8]. Decreased CO2 reactivity is also found in patients with SAH, and the magnitude of the reduction has also been correlated with ischemic deficits [9]. Despite the reductions in blood flow, cerebral blood "volume" is increased in patients with spasm, perhaps due to small vessel dilation distal to the areas of spasm [10]. In addition to functional abnormalities, structural alterations have been found in cerebral blood vessels affected by spasm. Leukocytes, red cells, and macrophages can be found in arterial walls. Mediators, such as eicosanoids, interleukin-1, and immune complexes, are increased, indicating an inflammatory component. Platelet aggregation occurs in the vascular endothelium near the site of aneurysm rupture [11]. Degenerative changes have been described in the tunica intima and media along with smooth muscle proliferation and collagen deposition, which increases vessel wall thickness [12]. The structural abnormalities induced by an inflammatory reaction may be responsible for obstruction to blood flow and the associated physiologic alterations. The functional and structural abnormalities of cerebral arterial vasospasm may also result from the action of mediators produced by blood and its breakdown products [13]. The amount and location of subarachnoid blood seen by computed tomography (CT) is predictive of the incidence and severity of vasospasm. Thick, focal blood collections at the site of aneurysm rupture are most predictive of ischemic stroke, while patients with thin, diffuse collections of blood are unlikely to develop vasospasm [14]. Aggressive removal of the blood clot from the subarachnoid space may reduce the incidence and severity of spasm [15-17], and antifibrinolytic therapy, which prevents clot lysis, may increase the risk for vasospasm [18,19]. Injection of blood into the basal cisterns or direct application to cerebral arteries produces spasm in experimental models. Hemoglobin, and especially oxyhemoglobin, a metabolite released by hemolysis of a subarachnoid clot, can produce spasm of cerebral vessels in vitro and in vivo [20]. Oxygenated hemoglobin produces superoxide free anion radicals, which could induce vasospasm by a number of mechanisms [21]. Superoxide radicals decrease nitric oxide production in endothelial cells and smooth muscle [22]. They also oxidize nitric oxide to vasoinactive compounds. A decrease in nitric oxide increases the activity of protein kinase C within the smooth muscle cell. Protein kinase C increases the release of calcium from intracellular stores and causes myofilament activation by an unknown mechanism [23]. Superoxide radicals also increase the formation of eicosanoids. Alterations in eicosanoid compounds have been implicated with decreased levels of prostaglandin I2, a dilator, and increased levels of the constrictor prostaglandin E2 found in experimental models of SAH [24]. Eicosanoids are thought to increase myosin light-chain phosphorylation through a calcium/calmodulin-dependent mechanism. Lipid peroxides are produced by superoxide free radicals through an increase in the activity of iron-dependent lipid peroxidases. Lipid peroxides increase the activity of protein kinase C by decreasing nitric oxide and by increasing the formation of diacyl-glycerol. Endothelial cell modulation of cerebral vascular tone may be partially caused by a balance between endothelium-derived releasing factor and endothelin release from the vascular endothelium [25]. Evidence also suggests that endothelin may play a role in the vascular remodeling associated with delayed cerebral vasospasm [26]. Endothelin, a potent vasoconstrictor, is found in increased amounts in the cerebrospinal fluid and plasma of patients with SAH [27]. Endothelin antagonists may attenuate the functional and structural vascular changes associated with cerebral vasospasm [28,29]. Other substances may play a role in cerebral vasospasm. Bilirubin levels are increased in the cerebrospinal fluid in a time course consistent with vasospasm; the application of bilirubin to cerebral arteries produces constriction [30]. Degeneration of cerebral arterial adrenergic nerve endings and increased catecholamine levels have been described in experimental models suggesting a role for denervation supersensitivity [31]. The search for the exact spasmogen(s) responsible for vasospasm is an important undertaking, since it may lead to the production of effective pharmacologic therapy. Diagnosis Clinical. The manifestations of clinical vasospasm typically include alterations in consciousness such as drowsiness and disorientation, or transient focal neurologic deficits [32]. The neurologic deficits depend in part on the location of the spasm and the state of collateral circulation. For example, bilateral anterior cerebral arterial spasm can produce akinesia, mutism, and diplegia. Middle cerebral artery spasm in the dominant hemisphere is associated with hemiparesis and aphasia. Spasm involving posterior fossa structures can lead to respiratory and hemodynamic abnormalities [33]. Clinical vasospasm may be accompanied by increasing headache, increasing meningismus, fever, and tachycardia. The timing of clinical vasospasm is characteristic in that it rarely occurs within the first 3 days after aneurysm rupture, peaks in 7-10 days, and typically resolves over 10-14 days. Clinical vasospasm is unlikely to develop after Day 12, although it has been reported to persist for weeks after SAH [34]. Delayed neurologic deterioration after rupture of an intracranial aneurysm can be caused by rebleeding, edema, hydrocephalus, seizure, hyponatremia, drug intoxication, or other medical complications in addition to clinical vasospasm. CT, while not capable of diagnosing vasospasm itself, is an important tool to rule out many of the nonspasm intracranial abnormalities [35]. Neurologic deterioration should be investigated with a CT scan to rule out hydrocephalus, subdural hematoma, or rebleeding. Blood should be drawn to rule out hyponatremia, hyperglycemia, hypoxemia, hypercarbia, or acidosis. Electroencephalographic studies can diagnose seizures that are not clinically obvious. Transcranial Doppler (TCD) studies may be useful in assessing the risk and severity of vasospasm. Angiography is used to confirm the diagnosis of vasospasm and assess the patient for angioplasty. Very often, however, the diagnosis of vasospasm is presumptive once the other more readily identifiable causes of delayed neurologic deficits have been excluded. Angiography. Cerebral angiography remains the "gold standard" for diagnosis of cerebral vasospasm. The angiographic appearance of vasospasm is typically that of smooth, luminal narrowing of involved arteries that may be confined to the area of aneurysm rupture, localized to a remote area of the brain, or represented diffusely [35]. Occasionally segmental spasm is seen with interposed skip areas. The diffuse pattern has been correlated with the worst prognosis [36]. The indications for angiography are complex because angiographic vasospasm can be identified in about 60% of patients with ruptured aneurysms, but only about half of these will develop clinical neurologic deficits [2]. The use of angiography has been reduced by noninvasive methods such as TCD but it has been increased by the advent of cerebral angioplasty. TCD. TCD has been used with varying success as a safe, repeatable, noninvasive method to identify and quantify vasospasm. Doppler signals from basal cerebral arteries can be obtained by application of an ultrasound probe to the skull. Velocity profiles detected by Doppler sonography increase as the diameter of the affected vessel decreases [37]. Velocities greater than 200 cm/s have been associated with a high risk of infarction after SAH, and patients with velocities less than 100 cm/s are unlikely to have clinical vasospasm [38]. The ability of transcranial Doppler to predict vasospasm has been questioned. Recent studies have attributed poor correlation between TCD and clinical or angiographic findings to the influence of collateral circulation and the effects of nimodipine [39]. Changes in measured velocities over time may be more reliable than absolute values in predicting symptomatic spasm. Serial measurements may be the best means of using TCD to predict the development of spasm. Confirmatory angiography may be avoided if the TCD study is positive, but additional studies may be necessary if the clinical picture is suspicious and the TCD study is negative [40]. Prevention Calcium Channel Blockers. Orally administered nimodipine has been shown to improve neurologic outcome and mortality from vasospasm after aneurysmal SAH and has now become standard prophylactic therapy. Among patients with good neurologic grade within 96 h of SAH [41], nimodipine-treated patients had a significantly better outcome and a lower incidence of severe arterial narrowing, but a similar overall frequency of vasospasm. Similar results were reported in a second controlled study [42]. Nimodipine improved neurologic deficits among the patients with a poor grade without improving mortality [43]. A British study of 554 good and poor grade patients reported a significant improvement in neurologic grade, and reduction in cerebral infarction rate, among patients receiving nimodipine [44]. The overall incidence of angiographic arterial narrowing was not reduced by nimodipine despite the improvement in outcome. One possible explanation is that the drug exerts a vasodilator effect on distal vessels not well imaged by angiography. Alternatively, neuroprotective effects of the calcium channel blocker may explain its beneficial effects [43]. Nicardipine, a dihydropyridine calcium channel blocker that is available for intravenous administration, reduced the incidence of clinical and angiographic vasospasm in a group of patients with SAH without improving overall outcome [45]. Patients in the placebo group more often received hypertensive/hypervolemic therapy, however. The therapeutic utility of hypervolemia and hypertension in treating ischemic deficits in the placebo group may explain the similar outcomes in this study. The most common complication of therapy with calcium channel blockers is hypotension, which occurs in about 0%-8% [43,46]. The hypotensive effect of calcium channel blockers is due to decreased systemic vascular resistance, and may make it difficult to achieve the goals of hypertension for hypertensive/hypervolemic therapy. A potential for decreased gastric motility, especially when combined with vasoconstrictor therapy, also exists [47]. Removal or Blockade of Spasmogenic Substances. Subarachnoid blood appears to be the proximate cause of vasospasm. Therefore, attempts have been made to remove thrombi from the subarachnoid space as soon as possible to prevent the development of spasm or lessen its severity. Experimental and clinical evidence suggests that early surgical removal of a cisternal blood clot prevents the development of vasospasm [48,49]. Intrathecal instillation of thrombolytic agents, such as urokinase or tissue plasminogen activator, allows the clot to be removed without the risk of damage induced by manual evacuation [50]. Concerns about bleeding complications and brain trauma have prevented the widespread adoption of subarachnoid clot removal until confirmatory clinical trials are performed. Pharmacologic agents have been used to block inflammatory responses that may be contributing to the development of vasospasm. For example, high-dose glucocorticoids improved outcome in an uncontrolled series of high-risk patients [51]. Ibuprofen has been shown experimentally to prevent vasospasm, presumably by blocking the effects of prostaglandins and thromboxanes [52]. Enthusiasm has been generated by studies of the 21-aminosteroid U-7400 6F, an inhibitor of iron-dependent membrane lipid peroxidation. This nonglucocorticoid compound has been shown to reverse vasospasm in experimental models of SAH [52,53]. Treatment The treatment for cerebral arterial vasospasm is multifactorial and includes hypertensive/hypervolemic therapy, angiographic dilation of spastic vessel segments, and pharmacologic dilation of affected blood vessels. The goal of treatment is the augmentation of blood flow through the areas of the brain supplied by the narrowed artery in an effort to limit cellular damage produced by ischemia. Hypervolemic/Hypertensive Therapy. Hypervolemic/hypertensive therapy has developed into a mainstay of treatment for cerebral ischemia associated with SAH-induced vasospasm [54,55]. The initial hemodynamic goals are to increase cardiac output and blood pressure with aggressive intravascular volume expansion. Hemodilution usually results from hypervolemic therapy. Target hematocrit values are typically between 30% and 35% [56]. Vasoactive drug infusions are added when necessary to raise arterial blood pressure if intravascular volume expansion alone is inadequate. The end-point of hypertensive/hypervolemic therapy is reached when neurologic deficits resolve or when complications of therapy ensue. Although central venous pressure monitoring is a reasonable guide to volume expansion, pulmonary artery catheterization may be a more useful guide because central venous pressure values do not correlate well with pulmonary capillary wedge pressure measurements in patients with SAH [45]. The pulmonary artery catheter can be used to titrate volume to a pulmonary capillary wedge pressure that maximizes left ventricular stroke volume and minimizes the risk of developing hydrostatic pulmonary edema [56,57]. The fluids used for volume expansion should be isosmolar and contain adequate amounts of sodium to counteract the tendency toward hyponatremia in these patients [58]. Hormonal compounds, such as vasopressin and fludrocortisone, have also been used to counteract excess volume losses and natriuresis [59]. Vasoactive drugs are used to raise arterial blood pressure further if neurologic deficits do not respond to volume therapy and adequate hypertension has not been achieved with intravascular volume expansion alone. Dopamine, dobutamine, phenylephrine, isoproterenol, norepinephrine, and metaraminol have been used for this purpose [55,60]. There is no consensus on the levels of blood pressure that must be achieved to treat vasospasm. Goals that have been advocated include the following: 20 mm Hg above premorbid levels or systolic blood pressure of 185 mm Hg, blood pressure >10 mm Hg above premorbid value, systolic blood pressure of 240 mm Hg and mean blood pressure of 150 mm Hg, systolic blood pressure of 160-200 mm Hg for clipped aneurysms and 120-150 mm Hg for unclipped aneurysms, systolic blood pressure of 150-175 mm Hg for clipped aneurysms and 130-150 mm Hg for unclipped aneurysms [55,56,61-63]. It is not clear which components of hypertensive/hypervolemic hemodilutional therapy are necessary or sufficient to treat vasospasm. Successful treatment of ischemic deficits has occurred with hypervolemia and hypertension without hemodilution [55]. Experimentally, volume expansion without hemodilution has failed to increase CBF [64]. Some evidence points to hypervolemia as the key component while other authors feel that hypertension is more important than hypervolemia [65-67]. The primary goal of therapy is the reversal of ischemic neurologic deficits. Once this goal is achieved, it is not clear that any escalation in therapy to achieve particular physiologic parameters is beneficial. Hypertensive/hypervolemic hemodilution therapy has been shown to increase CBF in patients with vasospasm after SAH [61]. While there has been no large randomized trial demonstrating efficacy, there is strong evidence from uncontrolled series to support the concept that this approach improves symptomatic ischemia and reduces morbidity and mortality [55,56,61,66,68]. There is less evidence to support the use of prophylactic hypervolemic/hypertensive therapy to prevent the development of symptomatic vasospasm [61,69]. Hypertensive/hypervolemic therapy carries risk. The most common complication is hydrostatic pulmonary edema, which occurs in up to 26% of patients [69]. Patients with coronary artery disease may develop myocardial ischemia from the combination of increased preload and afterload; both increase myocardial oxygen demand. Complications associated with placement of invasive hemodynamic monitors including hemothorax, pneumothorax, dysrythmias, and sepsis. Preoperative rebleeding or postoperative rupture of a second, unclipped aneurysm may occur. Many protocols limit the degree of hypertension achieved in this group of patients, although some evidence suggests that the risk of aneurysm rupture with hypertensive therapy may not be as great as commonly believed [55,70]. Angioplasty. Balloon angioplasty can reverse or improve spasm-induced neurologic deficits Figure 1 and is advocated for patients in whom symptoms persist despite maximum hypervolemic/hypertensive therapy [71,72]. About 70% of patients will improve clinically. Early angioplasty may be beneficial because the favorable physical characteristics of affected vessel walls may be lost once morphologic changes begin to appear. Patients treated within 6-12 h after the development of ischemic symptoms have better results than those treated after 72 h [71]. Although experience is still limited, general criteria for the procedure include new onset of a neurologic deficit not attributable to causes other than spasm, unresponsiveness to hypervolemic/hypertensive therapy, angiographic evidence of vasospasm in an area amenable to angioplasty, and absence of recent infarction. The risks of angioplasty include aneurysm rupture, intimal dissection, vessel rupture, ischemia, and infarction [72].Figure 1: The response of vasospasm to angioplasty. A: Vasospastic right middle cerebral artery. B: Same arterial segment immediately after balloon angioplasty. Note improvement in diameter of affected segment and distal flow. (Courtesy of Tony P. Smith, MD).Pharmacologic Dilation. Many drugs have been used in attempts to reverse vasospasm after SAH. These include sympathetic and parasympathetic adrenergic drugs, serotonin antagonists, nitrates, phosphodiesterase inhibitors, prostaglandin analogs and inhibitors, adenosine, oxygen free radical scavengers, and local anesthetics--all without proven clinical benefit [54]. The prophylactic administration of the dihydropyridine calcium channel antagonists has been the only drug therapy proven successful in improving outcome from SAH, although the mode of action may not be through cerebral vasodilatation. Data are limited but at least one multicenter, placebo-controlled trial demonstrated reducted mortality and severe morbidity among patients receiving nimodipine as treatment for established ischemic deficits [46]. In the future, many novel drug therapies can be expected to be tested in attempts to block the action of vasospastic mediators as more information regarding the pathophysiology of vasospasm emerges. For example, topical application of an endothelin receptor antagonist [73] and intravenous administration of magnesium sulfate [74] have partially reversed delayed cerebral vasospasm in experimental models. Hydrocephalus Hydrocephalus, occurring in about 25% of patients surviving the initial hemorrhage, is another cause of neurologic dysfunction in the perioperative period [75,76]. Blood in the ventricular system obstructs ventricular drainage pathways and CSF absorption sites (arachnoid villi) producing both obstructive and communicating hydrocephalus. Factors associated with the development of hydrocephalus include advanced age, hypertension, intraventricular hemorrhage, thick focal blood accumulation, posterior aneurysm, and poor neurologic grade [77]. Hydrocephalus is readily diagnosed by CT scan. Ventricular drainage is usually successful in improving neurologic symptoms due to hydrocephalus, although some patients ultimately require permanent shunt procedures [76]. Ventriculostomy performed before ablation of the aneurysm carries a risk of aneurysm rupture produced by a sudden reduction in intracranial pressure. The risk of rebleeding may be minimized by avoidance of excessive CSF drainage [78]. Seizures Seizures occur in up to 13% of patients after SAH and in 40% of those with a neurologic deficit [79]. Seizure activity may indicate rebleeding or rarely may be produced by vasospasm [80]. Regardless of cause, generalized seizures can raise intracranial and systemic pressures, as well as increase cerebral oxygen demand and lactate production, increasing the risk of aneurysm rupture or cerebral ischemia. Seizure activity is treated acutely with benzodiazepines or barbiturates as well as with phenytoin for longer term control. Because of the potential risk of rebleeding with a seizure, the administration of a prophylactic anticonvulsant is recommended in the immediate posthemorrhage period [58]. Hyponatremia Hyponatremia is surprisingly common after SAH (10%-34%), usually develops several days after the hemorrhage, and often parallels the time course of vasospasm [58]. Traditionally, hyponatremia had been attributed to the syndrome of inappropriate antidiuretic hormone, and fluid restriction was the typical therapy. However, in SAH there is evidence for cerebral "salt wasting" in many patients that depletes sodium and blood volume [81-83]. Dehydration, reduced circulating blood volume, and hypotension are risk factors for the development of vasospasm [84,85]. Prolonged and excessive use of mannitol can also provoke dehydration and hyponatremia. The decline in sodium concentrations typically occurs 3-15 days after SAH and may last for more than 2 wk. The decrease in circulating blood volume after SAH has been attributed to many causes including hypothalamic dysfunction and secretion of atrial natriuretic peptide [81,86]. Fluid restriction as a therapy for hyponatremia exacerbates the hypovolemic state, promotes symptomatic vasospasm, and promotes cerebral infarction [59,83]. Therefore, at least during the interval of vasospasm risk, therapy for hyponatremia should include efforts to maintain or increase intravascular volume and increase the serum sodium concentration with isotonic salt-containing fluid. Fluid restriction should not be instituted to treat hyponatremia [58]. Other Medical Complications Medical complications after SAH are determined in part by the severity and extent of neurologic insults, the therapies utilized, and the underlying medical condition of each patient. Many complications are common to hospitalized patients, especially those treated in intensive care units, and include pneumonia, sepsis (including invasive line sepsis), gastrointestinal bleeding, deep venous thrombosis, and pulmonary embolism. Pneumonia occurs in 7%-12% of patients with SAH. Neurogenic pulmonary edema, which is postulated to result from pulmonary capillary disruption caused by sympathetic hyperactivity [87], occurs in 1%-2% of patients with SAH. Fever is common due to the presence of blood in the subarachnoid space, and makes the diagnosis of significant infections more difficult. Significant gastrointestinal hemorrhage occurs perioperatively in 2%-4% of patients after rupture of a cerebral aneurysm, and the incidence can be reduced by prophylactic administration of H2 blockers or other prophylactic measures [88]. Deep venous thrombosis occurs in 1%-5% of patients treated for a ruptured intracranial aneurysm, and the incidence of pulmonary embolism is 0.8%-2.2% [89]. Intermittent pneumatic calf compression stockings have been shown to reduce the incidence of deep venous thrombosis in neurosurgical patients [90]. Prophylactic low-dose subcutaneous heparin is seldom used prior to or immediately after aneurysm clipping but is appropriate in patients confined to bed for a prolonged period after surgery [91]. Patients with neurologic injury are generally hypermetablolic [92,93]. Specific requirements for nutritional support for patients after SAH have not been reported. In patients with decreased levels of consciousness, enteral alimentation may be the preferred route for nutritional support. The risk of aspiration of gastric contents is decreased by the use of continuous rather than bolus feedings, strict elevation of the head, attention to gastric residuals, and the high pH of enteral formulas [94]. In patients with an unprotected airway or with decreased gag reflexes, jejunal feedings may further reduce the risk of aspiration. Conclusions Aneurysmal SAH is a multisystem illness that requires coordination of a wide variety of medical resources and practitioners. The anesthesiologist's role in the care of aneurysm patients may be wide ranging and include preoperative and postoperative care as well as intraoperative management. Many new therapies for the management of SAH have emerged recently, and more can be expected. It is important that the anesthetic management of SAH incorporate this new interdisciplinary approach.