Anesthesia for craniotomy in adults

INTRODUCTION — Craniotomy is performed for a variety of indications, including tumor resection, intracranial vascular procedures, evacuation of hematoma, and trauma.

This topic will discuss overall anesthetic management for craniotomy. Anesthetic management for some specific types of craniotomy is discussed separately.

●(See "Anesthesia for posterior fossa craniotomy".)

●(See "Anesthesia for intracranial neurovascular procedures in adults".)

●(See "Anesthesia for awake craniotomy".)

PREOPERATIVE EVALUATION — For the purpose of preoperative evaluation, craniotomies may be considered moderate- to high-risk procedures, depending on the procedure performed [1,2]. Preoperative evaluation and testing prior to anesthesia are discussed in detail separately. (See "Preoperative evaluation for anesthesia for noncardiac surgery".)

The following are issues of particular concern for craniotomy:

●Neurologic status – In addition to the usual preanesthesia evaluation, we focus on preoperative neurologic status, including current and prior deficits, signs and symptoms of increased intracranial pressure (ICP), and history of seizures.

●Cardiopulmonary status – Preoperative cardiac evaluation may be of particular concern for patients who are at high risk of venous air embolism (VAE; eg, surgery in the sitting position, surgery near venous sinuses). Evaluation and risk assessment in this setting are discussed separately. (See "Intraoperative venous air embolism during neurosurgery", section on 'Risk of paradoxical embolism'.)

●Medications

•Antiseizure medications – Many patients who undergo craniotomy are taking antiseizure medications, which can affect metabolism of a variety of drugs, including neuromuscular blocking agents (NMBAs). Patients should be instructed to take the usual dose on the morning of surgery, unless the patient is to undergo seizure focus mapping. For more detailed information on potential drug-drug interactions, refer to the drug interactions program within UpToDate.

•Glucocorticoids – For patients taking glucocorticoids preoperatively, stress doses may be required before induction of anesthesia and blood glucose may be elevated.

•Anticoagulants – In most cases, aspirin, nonsteroidal anti-inflammatory drugs (NSAIDS), and other medications that affect coagulation should be discontinued in advance of craniotomy. (See "Perioperative medication management", section on 'Medications affecting hemostasis'.)

●Laboratory evaluation – Preoperative blood testing should be based on patient factors and the extent of the planned procedure. We usually perform a complete blood count, measure blood glucose and electrolytes, and send a blood type and screen, in addition to any testing indicated by patient comorbidities. We crossmatch for blood products if there is a high risk of hemorrhage (eg, pathology involving cerebral sinuses, highly vascular intracranial tumors). (See "Preoperative evaluation for anesthesia for noncardiac surgery", section on 'Preoperative testing'.)

PREMEDICATION — Premedication for craniotomy should be individualized based on the patient's level of anxiety, baseline neurologic status, and comorbidities. Patients with intracranial pathology may be especially sensitive to sedatives and opioids; premedication should be titrated to effect using small doses of medication (eg, midazolam 1 to 2 mg intravenous [IV], administered in 0.5 mg increments). For patients with increased intracranial pressure (ICP), we withhold sedation until the patient is fully monitored in a setting that would allow immediate airway management.

ANESTHETIC MANAGEMENT — General endotracheal anesthesia is the preferred technique for craniotomy, though for specific indications, the procedure can be performed awake. (See "Anesthesia for awake craniotomy".)

General concerns — The following issues that affect anesthetic management should be addressed before any craniotomy, in consultation with the surgeon:

●Need for brain relaxation, including dose and timing of the necessary medications (see 'Planned brain relaxation' below)

●Patient positioning (see 'Positioning' below)

●Risk of venous air embolism (VAE) (see "Intraoperative venous air embolism during neurosurgery")

●Blood pressure (BP) goals and anticipated need for changes during surgery (see 'Goal for intraoperative blood pressure' below)

●Need for and doses of antiseizure medications (see 'Antiseizure drugs' below)

●Use of neuromonitoring (see "Neuromonitoring in surgery and anesthesia")

Variable surgical stimulus — Craniotomy involves periods of intense, painful stimulus separated by relatively long periods of low-level stimulation. Skull pinning is a brief, intensely painful stimulus. Skin incision, raising the bone flap, and opening the dura are also particularly painful, and may require adjustment of anesthetic depth. These painful events are often separated by periods of low stimulation during positioning and performing intracranial brain dissection.

Although not as stimulating as incision and opening, wound closure is also painful and may take as long as an hour. Opioids, acetaminophen, and/or antihypertensive medication may be required as the anesthetic is lightened for emergence.

Skull pinning — Options for minimizing hypertension and tachycardia during skull pinning include administration of intravenous (IV) medications (eg, fentanyl or remifentanil, propofol, esmolol, and lidocaine), use of local anesthetic infiltration at the pin sites, or use of scalp blocks.

●IV pretreatment – Timing of pinning should be coordinated with the surgeon to allow effective pretreatment. We usually administer esmolol 0.5 to 1 mg/kg IV and opioid (eg, fentanyl 100 mcg IV or remifentanil 50 mcg IV) approximately one minute prior to pinning. A small dose of propofol (eg, 20 to 50 mg IV) may be added depending on the patient's hemodynamic status and risk factors for subsequent hypotension.

Typical doses of IV drugs used in anticipation of pinning include fentanyl 50 to 100 mcg or remifentanil 25 to 50 mcg, propofol 20 to 50 mg, esmolol 0.25 to 0.5 mg/kg, lidocaine 1 mg/kg, or nicardipine 0.5 to 1 mg IV [3-6]. Higher doses of opioids can result in hypotension after the stimulus of pinning is over, particularly if administered with propofol.

●Scalp blocks – In some institutions, scalp blocks are placed prior to skull pinning to provide analgesia for pinning, skin incision, and closure, as well as postoperative analgesia [7,8]. The blocks can be placed before or after induction of anesthesia and may ameliorate the hemodynamic response to skull pinning. In some institutions they are placed at the end of surgery before emergence from anesthesia.

One single-institution randomized trial evaluated scalp blocks versus standard care in 60 patients who underwent supratentorial craniotomy [9]. Scalp blocks resulted in lower mean arterial pressure (MAP) and heart rate during skull pinning and skin incision, reduced propofol and remifentanil doses during the procedure, reduced postoperative pain scores, and modestly reduced opioid consumption in the first two postoperative days. (See 'Postoperative pain control' below.)

Monitoring — Standard physiologic monitors may be sufficient for select minimally invasive craniotomies (eg, burr holes for subdural hematoma evacuation). In most cases, additional monitoring is indicated. (See "Basic patient monitoring during anesthesia".)

●Arterial catheter – An arterial catheter is placed for most craniotomies for continuous BP monitoring and to facilitate blood sampling for blood gases and electrolytes.

●Monitoring for venous air embolism – Monitors for VAE are discussed separately. (See "Intraoperative venous air embolism during neurosurgery", section on 'Monitoring for venous air embolism'.)

●Central venous catheter (CVC) – CVCs are not routinely placed for craniotomy unless required for venous access or for possible air aspiration during VAE.

●Neuromonitoring – Intraoperative electroencephalography (EEG) and evoked potential monitoring during craniotomy have implications for the choice of anesthetic medication. These monitoring modalities are discussed separately. (See "Neuromonitoring in surgery and anesthesia".)

Intraoperative intracranial pressure (ICP) monitoring is discussed separately. (See "Anesthesia for patients with acute traumatic brain injury", section on 'Monitoring'.)

Induction of anesthesia — Goals for induction of anesthesia for craniotomy include:

●Hemodynamic stability – Drugs and doses should be selected to maintain cerebral perfusion pressure (CPP) while avoiding hypertension and the risk of intracranial hemorrhage.

●Avoiding increase in ICP – Medication and ventilation should be managed to avoid increase in ICP, especially for patients with preoperative increased ICP.

Choice of induction agents — Induction agents should be chosen based on patient factors and the effects of the agents on cerebral physiology. With the exception of ketamine, IV induction agents cause reductions in both cerebral metabolic rate (CMR) and cerebral blood flow (CBF), resulting in no change or decrease in ICP (table 1). Autoregulation and responsiveness to carbon dioxide (CO₂) are preserved. (See "Evaluation and management of elevated intracranial pressure in adults", section on 'Physiology'.)

●Propofol – CMR, CBF, cerebral blood volume (CBV), and ICP are reduced with induction doses of propofol [10,11], and autoregulation and CO₂ responsiveness are preserved [12]. Propofol can induce an isoelectric EEG. The induction dose should be adjusted for patient factors. (See "General anesthesia: Intravenous induction agents", section on 'Propofol'.)

●Barbiturates – Barbiturates cause a dose-dependent reduction in CMR, CBF, and ICP, while autoregulation and CO₂ responsiveness are maintained [13-15]. Thiopental (available outside the United States) can produce an isoelectric EEG; by contrast, methohexital activates seizure foci. Barbiturates can cause hypotension, though (at an equivalent dose) to a lesser extent than propofol. (See "General anesthesia: Intravenous induction agents", section on 'Methohexital'.)

●Etomidate – Etomidate does not usually decrease BP or cardiac output; an induction dose decreases CBF and CMR and reduces ICP without adversely impacting CPP [16-18], while preserving CO₂ responsiveness [16]. However, even with a single induction dose, etomidate inhibits corticosteroid production in the adrenal gland for up to 24 hours [19,20]. (See "General anesthesia: Intravenous induction agents", section on 'Etomidate'.)

The effects of etomidate on cerebral vasculature are complex. In animal models, etomidate is a cerebral vasoconstrictor, possibly mediated by mitochondrial dysfunction and inhibition of nitric oxide synthase [21-23]. A study of human brain tissue oxygenation during intracranial aneurysm surgery reported that administration of etomidate at a dose that produced EEG burst suppression resulted in a 30 percent reduction in tissue partial pressure of oxygen (PO₂), with a further 32 percent reduction of PO₂ during temporary artery clipping [24]. Because of these studies, we avoid administration of etomidate for induction of anesthesia for patients with cerebral vasospasm or other conditions associated with cerebral ischemia.

Etomidate is associated with a higher rate of postoperative nausea and vomiting (PONV) than other induction agents [25,26], which is a disadvantage in patients undergoing craniotomy.

●Ketamine – Data regarding the effect of ketamine on cerebral physiology are conflicting. Some human and animal studies have reported that ketamine increases CBF, CMR, and ICP [27-31]. By contrast, other studies have reported no change or a decrease in these parameters, particularly when ketamine is administered with other anesthetics [32-36]. Given the uncertainty, we believe that ketamine should be used with caution for craniotomy, especially for patients with increased ICP. (See "General anesthesia: Intravenous induction agents", section on 'Ketamine'.)

Opioids — Opioids cause minimal effects on cerebral physiologic parameters as long as MAP is maintained [37-39]. (See "General anesthesia: Intravenous induction agents", section on 'Opioids'.)

Lidocaine — Lidocaine 1 to 1.5 mg/kg IV may be administered during induction of anesthesia to suppress the cough reflex during laryngoscopy and to blunt, but not eliminate, the hemodynamic response to intubation [40,41]. (See "General anesthesia: Intravenous induction agents", section on 'Lidocaine'.)

Neuromuscular blocking agents — Nondepolarizing neuromuscular blocking agents (NMBAs) have no direct effects on cerebral physiology. Less commonly used NMBAs that are associated with histamine release at higher or rapid doses (ie, atracurium, mivacurium [outside the United States]) can theoretically cause a reduction in CPP because of a decrease in MAP, cerebral vasodilation, and an increase in ICP [42]. (See "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Atracurium' and "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Mivacurium'.)

Succinylcholine can produce a transient increase in ICP, possibly the result of an increase in CBF related to the arousal response to muscle fasciculations [43]. However, the brief increase in ICP with succinylcholine can be attenuated by administration of a defasciculating dose of a nondepolarizing NMBA [44].

For rapid sequence induction and intubation (RSII) in patients in whom ICP is a concern, we administer a defasciculating dose of nondepolarizing NMBA (eg, rocuronium 2 mg IV, cisatracurium 1.5 mg IV, or vecuronium 0.3 mg IV) followed by succinylcholine 1.5 to 2 mg/kg IV. For RSII when succinylcholine is contraindicated (eg, burns, denervation injury), we administer rocuronium 1 mg/kg IV or high-dose remifentanil (propofol 2 to 2.5 mg/kg IV followed by ephedrine 10 mg IV and remifentanil 3 to 5 mcg/kg IV) [45]. (See "Rapid sequence induction and intubation (RSII) for anesthesia", section on 'Neuromuscular blocking agents (NMBAs)'.)

Positioning — Positioning for craniotomy requires meticulous attention to detail to avoid nerve damage, skin pressure injury, cervical spine injury, and damage to the eyes. These procedures can be performed in a variety of positions, including supine, prone, lateral or semilateral (Park Bench), and sitting or semisitting. The head is usually held in skull pins attached to a head frame and is often turned to the side, sometimes with the neck flexed.

General concerns related to positioning, including potential nerve or skin damage and physiologic effects of various positions, are discussed in detail separately. (See "Patient positioning for surgery and anesthesia in adults".)

Some issues specific to craniotomy include the following:

●Cervical spine injury – If the patient will be positioned with the head rotated or the neck flexed, the patient's neck range of motion should be examined preoperatively.

●Excessive neck flexion – Neck flexion can kink the endotracheal tube (ETT) during positioning and/or later during the case when the ETT warms or with minor head movement. A wire-reinforced (armored) ETT can be used if kinking is a particular concern. Additionally, excessive neck flexion can impair venous drainage from the brain, contributing to increased surgical site bleeding or brain edema. Tongue edema can also occur as a result of excessive neck flexion. Reducing the degree of neck flexion can help mitigate these adverse effects.

●VAE – VAE is a risk whenever the operative site is above the level of the heart, and this risk is higher during surgery around venous sinuses. For craniotomy, most patients are positioned somewhat head-up to facilitate venous drainage and brain relaxation. (See "Intraoperative venous air embolism during neurosurgery".)

●Pneumocephalus – Supratentorial pneumocephalus (STP) can develop during procedures performed in the sitting position as cerebrospinal fluid (CSF) drains out of the cranial cavity at the durotomy site. In a series of 106 consecutive patients that underwent sitting craniotomy, 42 percent had postoperative STP detected on a computed tomography (CT) scan, with volumes ranging from 6 to 280 mL [46]. Six patients in this series had intraoperative somatosensory evoked potential (SSEP) changes (ie, reduction in signal amplitude) that were attributed to STP, all of which measured greater than 90 mL in volume on postoperative CT scan.

If symptomatic or associated with loss of intraoperative SSEP signals, STP is usually referred to as "tension pneumocephalus." Tension pneumocephalus can result in delayed emergence from anesthesia, locked-in syndrome, and lateral rectus muscle palsy [47-52], and requires emergent evacuation.

Maintenance of anesthesia

Choice of anesthetics — The choice of medications for maintenance of anesthesia for craniotomy should be based on the degree of preexisting intracranial hypertension, the need for brain relaxation during surgery, the use of neuromonitoring, and the patient's medical issues.

Anesthetics can affect cerebral physiology through changes in cerebral metabolism and blood flow either directly or indirectly by changing ICP and CPP (table 1). In many cases, a balanced anesthetic including relatively low doses of the potent inhalation anesthetics (ie, isoflurane, sevoflurane, desflurane, or halothane [where available]), with or without nitrous oxide (N₂O), and opioids is appropriate; for patients with elevated ICP, we suggest using a predominantly IV technique [11]. A strategy for anesthesia during neuromonitoring is discussed separately. (See "Neuromonitoring in surgery and anesthesia", section on 'Anesthetic strategy'.)

There is no consensus on the ideal anesthetic regimen (ie, total intravenous anesthesia [TIVA] versus inhalation anesthesia) for elective craniotomy in the absence of elevated ICP. A meta-analysis including 14 studies with over 1800 patients who underwent craniotomy reported various outcome measures for TIVA compared with inhalation anesthesia [53]. ICP was approximately 5 mmHg lower and CPP approximately 16 mmHg higher with TIVA than with inhalation anesthesia, with no difference in operative conditions after dural opening, recovery profiles, postoperative complications, or neurologic outcome.

Potent inhalation agents — The potent, halogenated inhalation anesthetics (ie, isoflurane, sevoflurane, desflurane, halothane [where available]) are all dose-dependent cerebral vasodilators.

●At doses below 1 minimum alveolar concentration (MAC), they reduce CMR and do not abolish coupling of CMR and CBF, such that without other contributing factors there is a net modest decrease in CBF.

●At doses above 1 MAC, these agents disrupt cerebral autoregulation and vasodilation predominates, leading to a gradual increase in CBF and CBV with increasing doses [54-59].

The degree to which the potent inhalation agents decrease CMR varies. At MAC-equivalent doses, reduction in CMR and therefore CBF in order of decreasing potency is as follows: sevoflurane > desflurane and isoflurane > halothane [54-58,60,61]. In one small study in healthy volunteers, sevoflurane reduced CMR to a similar degree as propofol at equal depth of anesthesia; however, propofol reduced CBF more than sevoflurane [61]. These results support a preference for TIVA rather than inhalation anesthetics (particularly at high doses) for patients with increased ICP or for whom significant brain relaxation will be required. However, if inhalation anesthesia is used, sevoflurane would be our preferred agent based on its more favorable profile on CMR and CBF. (See 'Choice of anesthetics' above.)

Responsiveness to CO₂ is maintained during administration of volatile anesthetics [62].

Nitrous oxide — N₂O can cause increases in CBF, CMR, and ICP. Autoregulation in response to changes in CO₂ appears to be preserved when N₂O is administered [63,64]. The magnitude of changes in cerebral physiology with N₂O is affected by the administration of other anesthetic drugs and by ventilation, as follows:

●N₂O alone – N₂O alone can cause substantial increases in ICP and CBF in patients without intracranial pathology [65] and in patients with intracranial tumors [66].

●N₂O with IV anesthetics – Concomitant administration of IV anesthetic medications can blunt the increase in CBF that occurs with N₂O alone. Studies of CBF when N₂O was added to anesthesia with barbiturates [67,68], opioids [69], benzodiazepines, and propofol [70] have reported minimal or no increase in CBF.

●N₂O with volatile anesthetics – When added to anesthesia with a volatile inhalation anesthetic (ie, isoflurane, sevoflurane, desflurane, or halothane), N₂O can result in a substantial increase in CBF. As an example, one study that compared CBF during 1.5 MAC isoflurane anesthesia with 0.75 MAC isoflurane and 65 percent N₂O reported 43 percent greater CBF with the anesthetic that included N₂O [64].

However, whereas N₂O increases CBF, it does not increase CBV when used along with a volatile anesthetic. As CBV is a greater determinant of ICP than CBF, N₂O is not necessarily contraindicated during craniotomy in patients with concerns for significant elevations in ICP [71].

●N₂O with hyperventilation – Since autoregulation is preserved, hyperventilation can prevent an increase in CBF during N₂O anesthesia [64].

N₂O may expand the gas content of air-filled spaces. If N₂O was not used during anesthesia, it should not be started following dural closure because of potential expansion of intracranial air. If N₂O has been used for the duration of the anesthetic, N₂O is unlikely to expand intracranial gas after dural closure, as the residual gas would already include N₂O.

We avoid using N₂O in patients who have recently had a craniotomy, due to concerns for potential expansion of any residual intracranial gas. It is unclear how long N₂O should be avoided after craniotomy, though residual air may last several weeks [72]. One contributor to this topic takes a conservative approach and waits eight weeks.

Intravenous anesthesia — IV anesthetics are commonly used as part of a balanced anesthetic that includes inhalation agents, or as TIVA. IV anesthetics are discussed in detail separately. Issues specific to craniotomy are discussed here. (See "Maintenance of general anesthesia: Overview" and "Maintenance of general anesthesia: Overview", section on 'Total intravenous anesthesia'.)

●Propofol infusion – Propofol infusion causes reduction in CMR, CBF, CBV, and ICP [61,73,74], while CO₂ responsiveness and autoregulation are maintained [12,75].

●Opioids – When administered as part of IV anesthesia with controlled ventilation, opioids have minimal, clinically irrelevant effects on cerebral physiology [69,76,77]. Morphine may cause histamine release in some patients, which could increase CBF.

●Dexmedetomidine – Dexmedetomidine is a highly selective alpha-2 agonist with sedative, sympatholytic, and analgesic properties that may be administered as an adjuvant for general anesthesia or for conscious sedation for awake craniotomy and other neurosurgical procedures [78].

Both animal and human studies have shown that dexmedetomidine is a cerebral vasoconstrictor that causes a dose-dependent reduction in CBF [79-83]. The other effects of this drug on cerebral physiology are less clear and may be species dependent. In dogs, dexmedetomidine has been consistently shown to have no effect on CMR; this, coupled with a reduction in CBF, could lead to cerebral ischemia. In humans, dexmedetomidine may reduce CMR along with CBF, similar to other IV anesthetics. A study in which healthy human volunteers received dexmedetomidine sedation reported parallel reductions in CMR and CBF, unchanged during hyperventilation [80].

The vasoconstrictive property of dexmedetomidine may be of theoretical concern in patients at risk for regional cerebral ischemia or compromised flow metabolism coupling (eg, traumatic brain injury [TBI], subarachnoid hemorrhage, intracranial lesions). However, data regarding this issue are limited. A small, retrospective study of patients with acute neurologic injury related to vascular lesions reported no reduction in brain tissue PO₂ with dexmedetomidine administration during craniotomy [84].

Neuromuscular blocking agents — Patients are typically paralyzed during anesthesia for craniotomy unless neuromonitoring precludes the administration of NMBAs. If the anesthetic is lightened during less stimulating periods of surgery, maintenance of neuromuscular block can reduce the chance of coughing or movement.

●Severe cough can result in straining, increased ICP, and brain herniation through the craniotomy.

●Movement while skull pins are in place can lead to slipping at the pin site, bleeding, and possible cervical spine injury.

For patients with upper motor nerve lesions and weakness or paralysis, the twitch monitor used to guide NMBA dosing and reversal should be placed on the unaffected side. (See "Monitoring neuromuscular blockade", section on 'Paralyzed limb'.)

Neuromuscular monitoring should always be used to guide administration of NMBAs, preferably with a quantitative monitor. We maintain relatively deep neuromuscular block (ie, one to two twitches on train of four stimulation) until the head frame is released from the operating table. Cough and movement while the skull is still fixed in place can result in cervical spine injury. Importantly, required doses of NMBAs may be increased in patients who chronically take antiseizure drugs, whereas acute administration of antiseizure drugs (eg, for prophylaxis during surgery) may potentiate neuromuscular block [85].

Hemodynamic management

Goal for intraoperative blood pressure — MAP should be controlled during craniotomy to maintain acceptable CPP (ie, MAP - ICP, or MAP - central venous pressure [CVP] if CVP > ICP). We suggest aiming for a CPP of 65 to 80 mmHg. Assuming a normal ICP (or CVP) range of 5 to 10 mmHg, MAP of 75 to 90 mmHg is a reasonable target range for an uncomplicated patient.

With some individual variation, autoregulation of CBF typically occurs within a MAP range of 60 to 150 mmHg [86]. Outside of this range, the brain is unable to compensate for changes in perfusion pressure, and the CBF increases or decreases passively with corresponding changes in pressure, resulting in the risk of ischemia at low pressures and edema or hemorrhage at high pressures. We aim for a MAP above the lower limit of autoregulation, with a margin for error.

The following considerations should determine the goal BP during craniotomy:

●Patient factors – Patient comorbidities may require modification of the goal BP during craniotomy. Normal cerebral autoregulation may be disrupted in patients with ischemic stroke [87], TBI [88], diabetes [89], and hypertension. A study of awake patients with drug-induced hypotension reported that the lower limit of cerebral autoregulation was increased in patients with chronic hypertension compared with normotensive controls (113 versus 73 mmHg) [90].

For patients with hypertension, we usually aim for a MAP close to baseline.

Reviewing cerebral imaging, estimating the patient's preoperative BP, and obtaining information about possible fluctuations in the neurologic examination that might be related to changes in BP preoperatively are warranted for decision making.

●Intracranial pathology

•Cerebral autoregulation might be blunted or abolished in certain conditions, either regionally (eg, brain tumors) [91] or globally (eg, TBI) [92,93]. In areas of intracranial pathology, CBF will be partially or entirely dependent on BP. In many patients with elevated ICP, MAP should be increased to maintain CPP. However, in patients with TBI, ICP should be reduced first, rather than increasing BP, as augmenting BP may worsen cerebral edema in these patients and further elevate ICP. This issue is discussed separately. (See "Anesthesia for patients with acute traumatic brain injury", section on 'Goal intracranial pressure and cerebral perfusion pressure'.)

There is no widely available real-time monitor for cerebral autoregulatory dysfunction.

•Occlusive disease of the cerebral arteries may lead to reliance on collateral arterial blood flow; higher CPP would be required to perfuse the ipsilateral brain. BP management during endovascular therapy for ischemic stroke is discussed separately. (See "Anesthesia for endovascular therapy for acute ischemic stroke in adults", section on 'Hemodynamic management'.)

●Procedure-related factors – Surgical maneuvers (eg, application of a temporary clip on a major cerebral artery) may require a higher CPP to assure collateral circulation [94]. By contrast, a lower BP may be required during specific portions of intracranial vascular procedures.

Vasoactive drugs — Vasoactive drugs are commonly administered during anesthesia for craniotomy to achieve BP goals. The effects of these drugs on cerebral physiology are complex and reflect the baseline BP, the status of autoregulatory mechanisms, the mechanism of the drug effect, and the magnitude of BP change:

●Vasopressors – Vasoconstrictors are often administered to counteract the vasodilation that is typically caused by anesthetic agents. Small boluses of short-acting agents are often administered during induction of anesthesia (eg, ephedrine 5 to 10 mg IV, phenylephrine 40 to 80 mcg IV). An infusion of a vasopressor may be required during maintenance of anesthesia, especially during periods of reduced surgical stimulation.

•Phenylephrine – Phenylephrine, a direct alpha-1 adrenergic agonist, is usually the first-choice agent in this setting. The effect of phenylephrine on CBF is controversial. Pure alpha agonists are thought to be systemic, but not cerebral, vasoconstrictors. Therefore, in most circumstances, when BP is increased with phenylephrine, CBF increases [95-98]. However, several studies in anesthetized patients [99,100], and others in awake volunteers [101,102], have suggested that under at least some circumstances, phenylephrine may be associated with a decrease in cerebral oxygenation. Methodologic concerns have been raised over the use of near-infrared spectroscopy (NIRS) for these studies [103], but the possibility of decreased cerebral perfusion with phenylephrine in patients at particular risk should be considered.

•Other vasopressors – Effects of infusions of vasopressors may be different, even at the same BP. Two small studies found improved CBF with ephedrine infusion versus phenylephrine [104]. Ephedrine is not typically administered as an infusion, and further study is required before recommending ephedrine infusion rather than phenylephrine during craniotomy.

-In a randomized trial of patients with supratentorial brain tumors who were assessed with positron emission tomography (PET) after induction of anesthesia, ephedrine infusion was associated with improved CBF and regional cerebral oxygen saturation in the normal brain hemisphere but not in the diseased hemisphere, compared with phenylephrine, at similar BPs [104].

-In another randomized trial of 24 patients who underwent craniotomy for brain tumor and were assessed with magnetic resonance imaging (MRI) and NIRS after induction of anesthesia, ephedrine infusion improved CBF and tissue oxygenation in both cerebral hemispheres [105]. By contrast, phenylephrine infusion reduced capillary transit time heterogeneity (a measure of cerebral oxygenation) in the hemisphere without tumor. Both groups had similar mean arterial BPs.

Other vasopressors may be indicated, depending on the patient's cardiac function and other comorbidities. A beta-1 adrenergic receptor agonist (eg, dobutamine, an inotrope) or an agent with both alpha and beta agonist properties (eg, norepinephrine, dopamine) may be required. Low-dose vasopressin (eg, 0.01 to 0.04 units per minute) may be useful as a supplementary vasopressor for hypotension refractory to other treatment. Vasopressin at higher doses is usually avoided because it can cause cerebral vasoconstriction [106]. (See "Use of vasopressors and inotropes".)

Similar to phenylephrine, a study of awake volunteers reported a reduction in cerebral oxygenation when norepinephrine was administered [107].

●Vasodilators – Vasodilators or beta blockers may be required during maintenance of and emergence from anesthesia. (See 'Emergence from anesthesia' below.)

Vasodilators (ie, nitroprusside, nitroglycerin, hydralazine, and calcium channel blockers) dilate the cerebral circulation and can increase CBF if adequate MAP is maintained. Therefore, vasodilators should be used with caution in patients with increased ICP, as these drugs may increase CBV and exacerbate intracranial hypertension.

●Beta blockers – Beta blockers either reduce or have no effect on CBF and CMR [108].

Our approach to hemodynamic management — Our approach to the maintenance of adequate mean BP is as follows:

●Optimize intravascular volume. (See 'Fluid management' below.)

●Titrate anesthetic agents to match the level of surgical stimulus in order to minimize hypotensive anesthetic effects and hypertensive responses to stimulation.

●For treatment of hypotension, treat initially with ephedrine 5 to 10 mg IV, particularly if hypotension is likely temporary, as occurs before an expected surgical stimulus (eg, incision).

●Thereafter, administer phenylephrine by infusion at the lowest dose necessary to achieve adequate CPP, titrated to effect.

●Treat hypertension by deepening the anesthetic and, if necessary, by administering titrated boluses of short-acting vasodilators (eg, labetalol 5 to 10 mg IV, esmolol 20 to 50 mg bolus) or, if necessary, by vasodilator infusion (eg, nitroglycerin 10 to 400 mcg/minute IV, or nicardipine 2.5 to 15 mg/hour).

Antiseizure drugs — Seizure can occur in 15 to 20 percent of patients without history of epilepsy following a nontraumatic supratentorial craniotomy. We administer a single dose of levetiracetam prior to incision for supratentorial craniotomy. Levetiracetam, phenytoin, and fosphenytoin are equally effective for seizure prophylaxis. We administer levetiracetam because in contrast with the other two drugs, it is not associated with hypotension during administration, has more reliable pharmacokinetics, and does not require serum monitoring. Also, unlike phenytoin, there is no concern over tissue injury with extravasation of levetiracetam.

Options for antiseizure drugs in this setting include the following:

●Levetiracetam – 500 to 1000 mg IV, up to 3000 mg IV as indicated

●Fosphenytoin – 10 to 20 mg phenytoin equivalents (PE)/kg over 30 minutes, maximum rate 150 mg PE/minute

●Phenytoin – 15 mg/kg IV, ≤50 mg/minute to avoid hypotension and bradycardia

Phenytoin is rarely administered where fosphenytoin is available. Extravasation during administration of IV phenytoin may cause severe tissue necrosis, and inadvertent intra-arterial administration of phenytoin has been associated with gangrene.

Patients who are taking antiseizure medications preoperatively should be maintained on their regular doses throughout the perioperative setting to avoid seizures.

Fluid management — We suggest maintaining normovolemia to achieve adequate cerebral perfusion and to avoid cerebral edema. IV fluid should be administered at a rate and volume that achieves even fluid balance. The following general principles apply:

●Choice of crystalloid solution – Isotonic (eg, plasmalyte), slightly hypotonic (eg, Ringer's lactate), or slightly hypertonic (eg, 0.9 percent sodium chloride [NaCl]) crystalloid solutions can be administered as maintenance fluid during craniotomy.

•Hypotonic fluid – Hypotonic fluids can increase brain interstitial fluid even in the healthy state [109,110]. When used in moderation, this effect is likely not clinically significant.

•Isotonic fluid – Isotonic crystalloid solutions do not increase the interstitial fluid content of the brain with an intact blood-brain barrier.

•Hypertonic fluid – Hypertonic solutions decrease the interstitial fluid content of the brain with an intact blood-brain barrier, pulling water across the cerebral capillary endothelium down its osmotic gradient. Hypertonic saline (HTS) can increase the volume of the brain with impaired blood-brain barrier function [111].

Large volumes of normal saline (0.9 percent NaCl) can cause hyperchloremic acidosis [112]. As an alternative, plasmalyte, or Ringer's lactate alternating with saline, can be used to avoid hyperchloremic acidosis. (See "Intraoperative fluid management".)

●Colloid solutions – Colloid administration during craniotomy is controversial, especially for patients with TBI (see "Anesthesia for patients with acute traumatic brain injury", section on 'Intraoperative fluid management'). A post hoc analysis of resuscitation with albumin compared with saline in patients with severe TBI reported worse long-term outcome with albumin [113]. However, this finding may not apply to intraoperative fluid management with cerebral edema from other causes. In a hypovolemic and hypotensive patient, 5 or 25 percent albumin can be used to restore intravascular volume status quickly.

Starch solutions should be avoided for craniotomy because they can interfere with platelet function and the factor VII clotting complex [114] and could result in bleeding. (See "Intraoperative fluid management", section on 'Hydroxyethyl starches'.)

●Fluid balance – We aim for an even fluid balance during craniotomy. Positive fluid balance can create or worsen cerebral edema [109], while hypovolemia may reduce CPP.

Urine output can be large when mannitol and diuretics are administered for brain relaxation. In this setting, we match urine output with fluid administration, modified as required to replace blood loss, and we use 0.9 percent saline. (See "Intraoperative fluid management", section on 'Monitoring intravascular volume status'.)

Ventilation — The goal partial pressure of CO₂ (PaCO₂) should be discussed with the surgeon preoperatively. Hyperventilation should be used only when indicated. Hyperventilation and the resulting reduction in CBF may be required to reduce ICP or to improve surgical exposure by relaxing the brain. (See 'Planned brain relaxation' below.)

General considerations include the following:

●Hypercarbia should always be avoided during craniotomy. Elevations in PaCO₂ result in increased CBF and may increase ICP. Unless hyperventilation is required, we maintain PaCO₂ at 35 to 38 mmHg.

●Therapeutic hyperventilation should be guided by blood gases rather than by end-tidal CO₂ (ETCO₂). While ETCO₂ generally correlates well with PaCO₂, a number of factors (eg, age, lung disease, surgical positioning) can result in significant discrepancy [115,116]. In our clinical experience, PaCO₂ is commonly 8 to 10 mmHg higher than ETCO₂ in patients with chronic obstructive pulmonary disease (COPD), though this difference is variable.

●The vasoconstriction that accompanies hyperventilation may result in ischemia, particularly for at-risk brain tissue (eg, after TBI, after subarachnoid hemorrhage, or under surgical retractors during craniotomy). Hyperventilation to a PaCO₂ of 25 to 30 mmHg can improve surgical conditions during supratentorial craniotomy [117]. However, multiple human and animal studies using a variety of methodologies have reported evidence of brain ischemia in injured brains with hyperventilation to a PaCO₂ of 25 to 30 mmHg [118].

As part of a multimodal approach to brain relaxation for surgical exposure, we hyperventilate as briefly as possible to achieve a PaCO₂ of 30 to 35 mmHg. Ventilation should be returned to normal gradually to avoid rebound cerebral vasodilation.

Hyperventilation for acute cerebral edema is discussed below. (See 'Intraoperative cerebral edema' below.)

Brain relaxation — Brain relaxation, or brain shrinkage, may be part of the surgical plan or may be required in response to unexpected brain swelling or tightness during the procedure.

Brain relaxation or shrinkage may be required to improve surgical exposure and to avoid ischemia related to pressure from retractors placed during surgery. Techniques for this purpose include administration of diuretics to reduce intravascular volume, mannitol or HTS for osmotherapy, glucocorticoids to reduce swelling, hyperventilation for vasoconstriction, and elevation of the patient's head to facilitate venous drainage. For some procedures, a lumbar or ventricular CSF drain is placed preoperatively to reduce brain bulk.

Osmotherapy works by creating an osmotic gradient that draws water out of brain tissue to reduce brain bulk. Effective osmotherapy with mannitol or HTS requires an intact blood-brain barrier; areas of brain with a disrupted blood-brain barrier may swell more with osmotherapy [111].

Planned brain relaxation — When brain relaxation is planned as part of the procedure, in consultation with the surgeon, we use the following regimen after induction of anesthesia:

●Furosemide 10 to 20 mg IV

●Dexamethasone 10 mg IV

●Osmotherapy with mannitol 0.5 to 1 g/kg IV administered slowly, over 10 to 15 minutes to avoid hypotension, or 3 percent HTS 3 to 5 mL/kg [119] (see 'Intraoperative cerebral edema' below)

●Hyperventilation to ETCO₂ 27 to 32 mmHg (aiming for PaCO₂ of 30 to 35 mmHg) (see 'Ventilation' above)

Intraoperative cerebral edema — If the surgeon encounters cerebral edema (ie, the "tight brain"), management requires a quick review of the physiologic principles of ICP dynamics and optimization. We use the following checklist to manage cerebral edema:

●Optimize cerebral venous drainage

•Elevate the head – This should be the first maneuver used to improve venous drainage from the brain. If possible, adjust the head rotation and flexion.

•Reduce intrathoracic pressure – Positive intrathoracic pressure impedes venous return to the right heart, including from the brain. If oxygenation permits, reduce or discontinue positive end-expiratory pressure (PEEP) and adjust the ventilator settings to decrease mean airway pressure.

●Reverse any cerebral vasodilation

•Optimize ventilation – Increase in minute ventilation may reduce cerebral vasodilation, depending on the starting ETCO₂ and PaCO₂. Send an arterial blood gas to guide therapy.

•Convert to IV anesthesia – Discontinue volatile anesthetics and N₂O.

•Treat elevated CMR – Maintain normothermia and adequate depth of anesthesia.

●Drain CSF if possible – If an external ventricular drain (EVD) or lumbar drain is in place, evacuation of CSF can rapidly reduce ICP. CSF should be drained in increments (eg, 5 to 10 mL at a time) with ongoing assessment of brain conditions.

●Start osmotherapy – If the steps above are ineffective, osmotherapy is indicated. For osmotherapy to treat intraoperative cerebral edema, we administer 0.5 to 1 g/kg of mannitol, 3 to 5 mL/kg of 3 percent NaCl, or 0.4 mL/kg of 23.4 percent NaCl over 15 to 20 minutes. For emergent therapy (eg, active brainstem herniation), we administer 23.4 percent NaCl, 30 mL over ≤3 minutes.

Mannitol (usually 20 percent solution) and various concentrations (3, 7.5, and 23.4 percent) of HTS have comparable efficacy in equiosmolar doses [120]; 20 percent mannitol and 3 percent NaCl are roughly equi-osmolar. The following considerations apply during osmotherapy [121,122]:

•Mannitol – Twenty percent mannitol can be administered via a peripheral IV and should be infused over 15 to 20 minutes to avoid hypotension. Mannitol causes an initial increase in intravascular volume followed by osmotic diuresis and net negative fluid balance. Mannitol should be avoided in patients who may not tolerate the initial increase in intravascular volume (eg, patients with congestive heart failure) or who cannot eliminate mannitol (eg, patients with renal dysfunction).

•HTS – Three percent NaCl can be administered via a peripheral IV, while more concentrated solutions (ie, 7.5 and 23.4 percent) must be administered through a CVC; except for emergent situations, HTS should be infused over 15 to 20 minutes. HTS causes a sustained increase in plasma volume. Concerns related to HTS therapy include the following:

-Rapid administration of HTS can result in hypervolemia and hypertension due to rapid expansion of blood volume; this is rarely of clinical concern in the operating room (OR).

-Rapid administration of HTS can cause acute elevation of serum sodium. As an example, administration of approximately 5 mEq/kg of sodium (approximately 45 mL of 23.4 percent NaCl) over two minutes can increase plasma sodium from 140 to 152 mEq/L. While osmotic demyelination syndrome (ODS) has occurred with rapid elevation in patients with severe hyponatremia, there are no reports of ODS occurring in the setting of osmotherapy for brain relaxation in normonatremic patients. (See "Osmotic demyelination syndrome (ODS) and overly rapid correction of hyponatremia".)

-Rapid administration of concentrated HTS (ie, 30 mL of 23.4 percent NaCl over less than one minute) can cause osmotic disruption of the blood-brain barrier, which can result in seizures, cerebral edema, and intracranial hypertension. If seizure occurs and if the dura is open, first-line treatment should be irrigation of the brain with iced saline. If necessary, further treatment includes propofol 10 to 20 mg IV or midazolam 1 to 2 mg IV to terminate the seizure.

●Treat refractory intracranial hypertension – High-dose barbiturate therapy (eg, pentobarbital 5 to 20 mg/kg IV bolus, followed by 1 to 4 mg/kg per hour, titrated to EEG burst suppression) may be used to control elevated ICP refractory to maximum standard medical treatment. It should be noted that pentobarbital, even after administration for a few hours at this dose range, will have a very long context-sensitive half-life. Vasopressor therapy may be required to maintain adequate CPP if high-dose barbiturate causes hypotension.

Alternatively, propofol can be used to help control ICP, but caution is required as high-dose propofol is associated with hypotension and propofol infusion syndrome. (See "Management of acute moderate and severe traumatic brain injury", section on 'Sedation and analgesia'.)

Brain protection — Management of temperature and glucose is important for patients with brain injury and ischemia (eg, during aneurysm clipping). In addition, in these settings, medications are often administered for neuroprotection (eg, barbiturates, propofol). Neuroprotection during craniotomy is discussed separately. (See "Anesthesia for intracranial neurovascular procedures in adults" and "Anesthesia for patients with acute traumatic brain injury", section on 'Neuroprotection'.)

Glycemic control — We manage glucose and insulin administration (IV bolus or infusion of regular insulin) to achieve blood glucose of 110 to 150 mg/dL and treat values above 180 mg/dL. Neither the glucose level above which neuronal damage occurs nor an ideal target plasma glucose concentration has been established, but tight glucose control (ie, target blood glucose of 80 to 110 mg/dL) is associated with increased risk of hypoglycemia. (See "Glycemic control in critically ill adult and pediatric patients".)

Hypoglycemia causes and exacerbates neuronal damage and should be avoided during craniotomy [123]. However, hyperglycemia is associated with increased morbidity and mortality after TBI [124-126] and decreased survival after brain tumor resection [127,128]. (See "Anesthesia for patients with acute traumatic brain injury", section on 'Glucose management'.)

Intraoperative hyperglycemia has also been shown to be associated with increased risk of postoperative infection [129,130].

Emergence from anesthesia — Most patients are awoken and extubated in the OR after craniotomy.

Management of emergence — Important aspects of emergence after craniotomy include the following:

●Readiness for extubation – Extubation may be delayed for patients who undergo infratentorial surgery if there are concerns for lower cranial nerve dysfunction that might impact airway reflexes, and for patients who undergo long procedures in the prone position [131-133]. (See "Anesthesia for posterior fossa craniotomy", section on 'Emergence and extubation' and "Extubation following anesthesia", section on 'Higher-risk extubation'.)

●Minimal residual anesthesia – Ideally, the patient should be awake enough for an adequate neurologic examination (eg, responding to commands, moving all extremities on command, adequate vision assessment) before leaving the OR.

●Pain control during emergence – In contrast with other surgical procedures, opioids should not be administered at a dose that is expected to prevent anticipated pain, but rather titrated as needed following extubation and neurologic examination. Liberal opioid dosing based on anticipated pain or guided by hemodynamic endpoints (eg, to keep systolic BP below 160 mmHg) often results in somnolence, a delay in a satisfactory neurologic examination, and occasionally unnecessary head imaging.

For patients without preoperative altered mental status and for those who were taking opioids preoperatively, we usually administer morphine 3 to 5 mg IV or hydromorphone 0.5 mg IV 30 minutes prior to emergence, and titrate further opioid postoperatively after neurologic examination.

Patients who receive scalp blocks may awaken from anesthesia with little or no pain.

●Hemodynamic management – Following a routine tumor resection for a patient without chronic hypertension or with controlled chronic hypertension, we aim to maintain a systolic BP of <160 mmHg, as systolic pressures >160 mmHg are associated with increased risk of postoperative intracranial hemorrhage [134,135].

Hypertension is common on emergence from anesthesia for craniotomy and should be treated quickly. In addition to an association with intracranial hemorrhage [134], hypertension can worsen cerebral edema in those areas where the blood-brain barrier is disrupted. Treatment should be titrated to avoid hypotension, cerebral hypoperfusion, and enlargement of an area of cerebral ischemia [136].

Medications commonly administered to control hypertension on emergence after craniotomy include labetalol, esmolol, and nicardipine. We typically start with IV labetalol because it is easy to dose and, compared with esmolol, has a duration of action long enough to permit safe transport to the postanesthesia care unit (PACU) or intensive care unit (ICU). If labetalol is contraindicated or not effective at doses up to 300 mg per hour, nicardipine infusion would be warranted and provides ongoing hemodynamic control during transfer of care.

A prospective study of preemptive BP control in 42 patients who underwent craniotomy found that when added to enalaprilat, nicardipine (2 mg boluses IV) was as effective as labetalol (5 mg boluses IV) at controlling postoperative BP [137].

Perioperative use of these medications is discussed separately (table 2). (See "Intraoperative use of vasoactive agents", section on 'Antihypertensive agents'.)

The duration and severity of postoperative hypertension vary. Some patients require only a few intermittent boluses of antihypertensive medication, while others may need a continuous infusion to control BP. Preexisting and uncontrolled hypertension may predict the need for more intensive and prolonged treatment.

Once the patient is transferred to the PACU or ICU, it is important to communicate the BP goals and doses of medication administered.

●ICP control – During emergence, every effort should be made to avoid coughing, straining, retching, and vomiting, all of which can increase CVP and ICP. Prophylactic antiemetics should be administered routinely, and airway suctioning should be performed before the depth of anesthesia is lightened.

During emergence, ventilation should be assisted until the patient maintains adequate minute ventilation. Hypoventilation on emergence can result in increased PaCO₂ and can cause cerebral vasodilation and increased ICP.

Delayed emergence — When the patient is slow to emerge from anesthesia, the cause may be related to surgical, anesthetic, preexisting, or physiologic factors. If the patient has a significant preoperative neurologic deficit, emergence may be delayed and extubation may need to be deferred. In patients with preoperative deficits and in whom there is a high likelihood that the surgical procedure may improve the patient's mental status (eg, evacuation of a large epidural hematoma), neurologic examination may be attempted and extubation considered.

As some potential causes require emergent action, assessment of the patient who fails to emerge from anesthesia should include the surgeon and should proceed in a systematic fashion. Causes, evaluation, and management of delayed emergence are discussed separately (table 3). (See "Delayed emergence and emergence delirium in adults", section on 'Delayed emergence'.)

After craniotomy, cerebral edema, intracerebral hematoma, tension pneumocephalus, and ischemia (total occlusion of an artery or hypoperfusion) are potential surgical causes for a delayed emergence. Pupils should be examined along with response to pain and reflexes.

When no cause for delayed emergence can be identified, an emergent CT scan should be performed to assess for intracranial hemorrhage, brain edema, pneumocephalus, or other pathology.

Differential emergence or awakening — Transient focal neurologic deficits can occur following sedation or general anesthesia in patients with a prior history of a neurologic deficit related to stroke, brain tumor, and carotid disease. In the context of anesthesia, this phenomenon has been called differential awakening; patients typically exhibit a focal motor deficit immediately upon emergence that improves over 30 minutes to several hours. The mechanism remains to be elucidated, but the time course suggests a pharmacologic cause.

Transient focal neurologic deficits have been elicited with administration of both sedatives and opioids to patients with a history of neurologic deficits [138-141], and in patients with brain mass lesions without known deficits [142,143]. Reversal of opioid-induced deficits with naloxone and of midazolam-induced deficits with flumazenil [143] have been reported.

When a focal neurologic deficit is evident on emergence from anesthesia, multidisciplinary evaluation should be performed, with differential emergence included among potential etiologies.

POSTOPERATIVE CARE — Although debated, even an uneventful craniotomy remains an indication for admission to the intensive care unit (ICU) [144]. The primary indication for intensive care is to allow serial (usually hourly) neurologic examinations and rapid response to abnormalities. In addition, postoperative intensive care allows continuous blood pressure (BP) monitoring and control, increased intracranial pressure (ICP) monitoring when required, and treatment of pain and postoperative nausea and vomiting (PONV).

Some institutions have created postoperative pathways for transfer of selected patients from the operating room (OR) to the postanesthesia care unit (PACU), and then to a stepdown unit rather than the ICU. In a single-institution observational study of 324 patients who underwent neurosurgical procedures, 94 of whom went through a protocol that bypassed the ICU, length of hospital stay and costs were less in patients who bypassed the ICU, and there were no significant complications or adverse outcomes in those patients [145]. Patient eligibility criteria for bypassing the ICU included age ≤65 years; no comorbidity requiring ICU-level care; if craniotomy for tumor, supratentorial and tumor <3 cm; no intraoperative adverse events; duration of surgery <5 hours; estimated blood loss <50 mL; and routine extubation.

Patients who undergo awake craniotomy are commonly sent to the PACU and then an inpatient floor, rather than the ICU. This is discussed separately. (See "Anesthesia for awake craniotomy", section on 'Postoperative care'.)

Postoperative pain control — Up to 75 percent of patients experience moderate to severe pain in the first few days after craniotomy [146-148]. Acute pain is often more severe after infratentorial compared with other types of craniotomy [148], possibly due to retraction and incision of muscles in the region. A significant portion of patients continue to have headache for months or years after craniotomy [149,150]. Although severe acute postoperative pain is associated with increased risk of persistent pain, it is unclear whether effective control of acute pain reduces the incidence of chronic pain.

Similar to other types of surgery, multimodal opioid-sparing analgesic strategies are increasingly used for craniotomy. Pain control may include local anesthetic infiltration of the incision, scalp blocks, acetaminophen, and opioids. Nonsteroidal anti-inflammatory drugs (NSAIDs) are avoided due to potential for bleeding.

Scalp blocks may be used for early postoperative pain control for craniotomy and are part of some enhanced recovery after surgery (ERAS) protocols for craniotomy [151]. Scalp blocks may provide some analgesic benefits and can be quickly performed before anesthesia, after induction, or at the end of the case prior to emergence. The author uses scalp blocks for awake craniotomy, but not routinely for craniotomy with general anesthesia.

The evidence in support of scalp blocks is inconsistent with respect to pain scores and postoperative analgesic use.

●In a 2019 meta-analysis of randomized trials that evaluated scalp block for craniotomy, scalp block produced a modest decrease in postoperative pain scores at 0 to 6 hours (10 studies, 44 patients) and 12 hours (8 studies, 294 patients) postoperatively, mean difference 0.98 and 0.95 on a 0 to 10 scale, respectively [152]. Scalp blocks reduced postoperative opioid requirement (7 studies, 324 patients), 149 versus 165 morphine milligram equivalents. For this outcome, there was very high statistical heterogeneity. The quality of evidence was low or very low for all outcomes.

●In a 2020 randomized trial of 89 patients who underwent supratentorial craniotomy, bilateral scalp blocks did not reduce postoperative pain scores or opioid consumption during the first 24 hours after surgery [153].

●A single-center prospective observational study evaluated postoperative pain outcomes in 134 patients who underwent supratentorial craniotomy with multimodal analgesia , 46 of whom also received bilateral scalp blocks [154]. Multimodal analgesia consisted of intraoperative dexamethasone, tramadol and acetaminophen, and postoperative acetaminophen and as-needed opioids. Scalp block was associated with reduced need for postoperative rescue analgesia in the first 48 hours (39 versus 65 percent of patients). There was no difference in postoperative pain scores in patients who received scalp blocks.

SUMMARY AND RECOMMENDATIONS

●Preoperative planning – The anesthesiologist and surgeon should discuss preexisting increased intracranial pressure (ICP), positioning for surgery, the risk of venous air embolism (VAE), goals for blood pressure (BP) and ventilation (goal partial pressure of carbon dioxide [PaCO₂]), and whether neurophysiologic monitoring will be used. (See 'General concerns' above.)

●Choice of anesthetic technique – General endotracheal anesthesia is the preferred technique, though for specific indications, the craniotomy can be performed awake. (See 'Anesthetic management' above.)

●Choice of anesthetic medications

•Total intravenous versus inhalation anesthesia – The optimal anesthetic regimen for elective craniotomy is controversial. In many cases, we use a balanced anesthetic including low doses of a potent inhalation anesthetic (ie, isoflurane, sevoflurane, desflurane, and halothane), with or without nitrous oxide (N₂O), and opioids. For patients with increased ICP, we suggest using a predominantly intravenous (IV) technique (Grade 2C). (See 'Maintenance of anesthesia' above.)

•Effects on cerebral physiology – Anesthetics have a variety of effects on cerebral physiology (table 1).

-IV induction agents – With the exception of ketamine, IV induction agents (ie, propofol, barbiturates, etomidate) cause reductions in both cerebral metabolic rate (CMR) and cerebral blood flow (CBF), resulting in no change or a decrease in ICP, while responsiveness to carbon dioxide (CO₂) is maintained. (See 'Choice of induction agents' above.)

-Volatile anesthetics – Isoflurane, sevoflurane, desflurane, and halothane are dose-dependent cerebral vasodilators. While they reduce CMR, they can blunt cerebral autoregulation by uncoupling CBF and metabolism, and they can increase CBF and ICP. Below 1 minimum alveolar concentration (MAC), there is a modest decrease in CBF. Above 1 MAC, CBF increases. Responsiveness to CO₂ is maintained. (See 'Potent inhalation agents' above.)

-Nitrous oxide – N₂O can increase CBF, CMR, and ICP, with preserved CO₂ responsiveness. The magnitude of changes in cerebral physiology with N₂O is affected by the administration of other anesthetic drugs and by ventilation. (See 'Nitrous oxide' above.)

-Propofol infusion – Propofol infusion causes reduction in CMR, CBF, and ICP, while CO₂ responsiveness is maintained. (See 'Intravenous anesthesia' above.)

-Opioids – When administered with controlled ventilation, opioids have minimal effects on cerebral physiology. (See 'Intravenous anesthesia' above.)

●Positioning – Positioning for craniotomy requires meticulous attention to detail to avoid nerve injury, skin pressure injuries, ocular injury, and airway compromise. Supine, prone, lateral, or sitting positions may be used. The sitting position is associated with a higher risk of VAE, hypotension, and pneumocephalus and is relatively contraindicated in those with a potential right-to-left intracardiac shunt. (See 'Positioning' above.)

●Blood pressure goal – BP should be controlled during craniotomy to maintain adequate cerebral perfusion pressure (CPP). We suggest aiming for a CPP of 65 to 80 mmHg (Grade 2C). For patients without an ICP monitor, this translates to a mean arterial pressure (MAP) goal of 75 to 90 mmHg for an uncomplicated patient. For patients with an ICP monitor in place, the goal BP is determined by the existing ICP, using the formula CPP = MAP - ICP. For patients with hypertension, we aim for MAP close to baseline. (See 'Hemodynamic management' above.)

●Fluid management – We suggest administering IV fluid at a rate and volume that achieves even fluid balance to achieve adequate cerebral perfusion and to avoid cerebral edema (Grade 2C). For patients who receive mannitol and/or diuretics for brain relaxation, we match urine output with IV crystalloid. We use 0.9 percent saline. (See 'Fluid management' above.)

●Ventilation – Hypercarbia should always be avoided during craniotomy. Hyperventilation reduces CBF and may be required to reduce ICP or to improve surgical exposure by relaxing the brain. Hyperventilation should not be used routinely, as the vasoconstriction that accompanies hyperventilation may result in brain ischemia.

•When indicated, hyperventilation should be guided by blood gases rather than by end-tidal CO₂ (ETCO₂).

•When hyperventilation is used as part of multimodal approach to brain relaxation for surgical exposure, we suggest hyperventilating as briefly as possible to achieve a PaCO₂ of 30 to 35 mmHg (Grade 2C). (See 'Ventilation' above.)

●Brain relaxation – Brain relaxation may be part of the surgical plan, or it may be required in response to unexpected brain swelling. Planned brain relaxation usually includes administration of diuretic, glucocorticoid, osmotherapy with mannitol or hypertonic saline (HTS), and hyperventilation. (See 'Planned brain relaxation' above.)

Management of intraoperative cerebral edema is described above. (See 'Intraoperative cerebral edema' above.)

●Emergence

•Emergence from anesthesia for craniotomy should be rapid and smooth to allow a postoperative neurologic examination.

•Opioids should be titrated as needed for pain after emergence.

•Hypertension is common on emergence and should be treated rapidly to avoid intracranial hemorrhage. (See 'Emergence from anesthesia' above.)