A clinical review of inhalation anesthesia with sevoflurane: from early research to emerging topics

In clinical practice, sevoflurane and propofol are more frequently used as induction agents, as isoflurane and desflurane are more irritating to the airway [13]. However, a number of clinical studies have demonstrated that all of these drugs are efficacious and safe for induction of general anesthesia (Table 1). In one randomized, double-blind comparison of 8% sevoflurane versus propofol as induction agents, the time to anesthesia induction was longer with sevoflurane [14]. However, a study in 75 adult patients noted that the induction time was not statistically different between sevoflurane and propofol [15]. Additionally, a meta-analysis of five studies by Joo et al. concluded that time to loss of consciousness was similar for propofol and sevoflurane, with a weighted mean difference of 2.84 s (95% CI −12.36, 18.05) in favor of propofol [16].

Anesthesia maintenance and recovery profiles have also been investigated (Table 2). A study comparing the effectiveness of sevoflurane versus propofol in 66 patients revealed shorter times for two recovery parameters. The authors concluded that maintenance of anesthesia with sevoflurane resulted in a more favorable recovery profile, less patient movement, and more favorable hemodynamic responses [1]. In a phase 3 trial, sevoflurane was compared with isoflurane for maintenance of anesthesia. This study indicated that eye opening, response to commands, and request for first postoperative analgesia as a marker of wakefulness were more rapid following discontinuation of sevoflurane than isoflurane [17]. Several studies have compared sevoflurane with desflurane in intermediate recovery and time for discharge in ambulatory patients, and have demonstrated that emergence and/or early recovery from anesthesia are faster with desflurane than sevoflurane [2, 18‐20]. This phenomenon is not unexpected due to the lower solubility of desflurane in blood and tissues [21], but the faster wake-up time has generally failed to translate into early readiness for discharge [2, 19, 20].

Studies in obese patients have compared emergence times between sevoflurane and desflurane (Table 3), which have the lowest fat solubility among the volatile anesthetics. The lower tissue and blood/gas solubility of desflurane accounts for the more rapid rate of recovery versus sevoflurane [22]. Strum et al. noted a nearly 9-min faster time to eye opening and a more than 11-min faster time to extubation with desflurane [23]. McKay et al. and Bilotta et al. also observed faster times to respond to commands and extubation with desflurane [24, 25]. However, Vallejo et al. suggested the difference in emergence time is minimal when the concentration of the inhaled agent is ≤80% of the maintenance dose at the time of discontinuation [26]. Similarly, Arain et al. demonstrated that potential differences in emergence time are not significant when proper titration is employed [22].

The most common adverse events associated with sevoflurane in studies of adult patients include nausea/vomiting, somnolence, coughing, shivering/chills, pain, dizziness, and confusion [17, 20]. Within the population of obese patients, the most commonly reported adverse events are nausea/vomiting and pain [23, 24, 26].

Studies in pediatric populations have compared sevoflurane with desflurane [27‐30], isoflurane [9, 27, 31], or propofol (Tables 1, 4) [6, 32, 33]. Sevoflurane has been described as the agent of choice for mask induction in children due to its lack of airway irritation, hemodynamic characteristics, and lower pungency [3]. In contrast, both desflurane and isoflurane are considered less suitable due to airway irritation [34].

In general, emergence and early recovery are faster in pediatric patients with desflurane than sevoflurane [28‐30], though a recent study by Ghoneim et al. reported no difference between sevoflurane and desflurane, whereas emergence was longer with isoflurane [27]. Other studies have similarly found emergence from anesthesia to be longer with isoflurane than sevoflurane [9, 31]. The faster emergence of desflurane does not translate into a measurable discharge benefit [29, 30].

Relatively few studies have compared sevoflurane with propofol for induction of anesthesia in pediatric patients. Quality of induction, as characterized by intubation conditions [35] and behavior during induction [6], does not appear to differ between the agents. In contrast to findings in adult patients, one study reported a significantly more rapid induction with sevoflurane versus propofol. However, the authors noted that the dose of propofol used was relatively low (3 mg/kg), which may have been responsible for the discrepant results [33]. Consistent with studies in adult patients, measures of recovery from anesthesia (time to extubation [6, 33] and readiness for discharge [6, 32]) appear to favor sevoflurane compared with propofol.

Adverse events most commonly observed in pediatric patients during recovery from sevoflurane anesthesia include nausea/vomiting, coughing, shivering, pain, and agitation/delirium [27, 32, 33, 36, 37].

Age-related changes in elderly patients (>65 years of age) affect many organs, but of particular interest to anesthesiologists are the respiratory, cardiovascular, renal, and central nervous systems. These alterations have significant clinical implications for the different stages in anesthesia. This may be illustrated through the inverse relationship between age and minimal alveolar concentration (MAC) that has been observed for halogenated inhaled anesthetics, in which elderly patients require a lower anesthetic MAC to achieve the same effectiveness as adults or children. The MAC for sevoflurane in elderly patients is 1.48%, which is lower than the MAC for children (2.49%) and adults (1.71–2.05%) [38, 39].

Yamaguchi et al. demonstrated that sevoflurane induction is well tolerated [40]. Peduto et al. investigated the recovery from sevoflurane-induced anesthesia in elderly patients compared with isoflurane (Table 5) [41]. The time to extubation, emergence, response to command, and suitability for discharge were shorter for sevoflurane, leading the authors to conclude that sevoflurane provides a more rapid emergence than isoflurane. In a similar comparison of sevoflurane versus desflurane anesthesia in elderly patients, the recovery time was reported to be faster with desflurane [42]. However, the discharge time from the postanesthesia care unit was similar for patients who received desflurane and those who received sevoflurane, even with the previously noted difference in early recovery [42, 43].

A recent meta-analysis of randomized controlled trials in elderly patients showed that recovery times were shorter for desflurane than for sevoflurane. The meta-analysis also addressed cognitive dysfunction and delirium upon emergence from anesthesia and revealed no difference in the incidence of cognitive dysfunction between these agents [7]. This indicates that the shorter recovery time does not translate into a significant effect on cognition; other factors play more important roles.

The most commonly reported adverse events associated with sevoflurane following surgery in elderly include nausea/vomiting, pain, dizziness [42, 44], delirium [45], and short-term cognitive dysfunction [11, 45, 46].

As with other inhaled anesthetics in this class (e.g., isoflurane and desflurane), sevoflurane affects a variety of systems during anesthesia that can result in cardiovascular (hypotension and cardiac depression), central nervous system (CNS) (depressed function), and respiratory (ventilatory depression) side effects, which are well described in the literature [39]. Postoperative adverse events associated with inhaled anesthetics include pain, nausea/vomiting, headache, dizziness, cognitive dysfunction, and delirium [39]. Postoperative nausea/vomiting (PONV) is one of the most frequent adverse events, occurring in approximately 33% of patients. This pro-emetic effect is more pronounced in the early postoperative period (0–2 h postsurgery), and does not appear to differ among the various volatile anesthetics [47]. This increase in PONV has also been observed in pediatric patients [30, 32, 37]. PONV is of significant concern to patients; however, PONV can be successfully managed through anesthesia modification (e.g., use of air rather than N₂O) and/or antiemetic administration [48]. It was recently noted in a meta-analysis that a volatile gas plus one anti-emetic drug is comparable to propofol in terms of rate of PONV [49]. The same meta-analysis also showed that patients receiving TIVA had a significantly higher risk of experiencing PONV in the late postoperative phase (RR 1.41, 95% CI 1.10, 1.79, P = 0.006). Based on these data, it is in the hands of the anesthesiologist to decide which therapeutic approach is most suitable for an individual patient. Other uncommon yet important adverse events identified with use of sevoflurane include hyperkalemia [50], seizures [51, 52], hepatic dysfunction [53, 54], and malignant hyperthermia [55].

In summary, numerous investigations demonstrate the efficacy and safety of sevoflurane for both induction and maintenance of general anesthesia in pediatric, adult, obese, and elderly patient populations. Across the patient populations, evidence suggests that emergence times from sevoflurane anesthesia are generally faster than from isoflurane anesthesia and slower than from desflurane anesthesia, whereas there appears to be no difference in readiness for discharge. Comparisons between sevoflurane and propofol have also shown fairly consistent results; more rapid induction has been noted with propofol and faster emergence and readiness for discharge has been observed with sevoflurane. In addition, the safety profile of sevoflurane has been established in these populations. The most common side effects of sevoflurane are similar to those observed with other inhaled anesthetics.

In a pivotal paper, Kersten et al. showed that the protective effect of isoflurane involved the activation of potassium-activated ATP (K_ATP) channels, a phenomenon known as anesthetic preconditioning [73]. Shortly thereafter, Preckel et al. reported beneficial effects of enflurane, isoflurane, sevoflurane, and desflurane on reperfusion injury following myocardial ischemia in rabbits [74]. Two groups of researchers independently translated these findings into daily clinical practice [75, 76]. They chose cardiac surgery as a model because a reproducible and well-defined episode of myocardial ischemia is part of the surgical procedure in cardiac surgery. Numerous subsequent experimental and clinical studies have addressed the concept of anesthetic-induced cardioprotection. For reasons discussed above, most clinical studies have been performed in the field of cardiac surgery, either with or without CPB. Although the experimental evidence for anesthetic preconditioning and postconditioning is unequivocal and indisputable, the significance of anesthetic-induced cardioprotection in the clinical setting is still controversial.

Several laboratory investigations have revealed that ischemic preconditioning and anesthetic-induced preconditioning share a common phenotype. Based on current evidence, two main intracellular signal transduction pathways are thought to convey cardioprotection by transducing signals from cellular receptors to ion channels located in the mitochondria. Briefly, both the reperfusion injury salvage kinases pathway, via G-protein-coupled receptors, and the survivor-activating factor enhancement pathway are most likely involved [77]. The final pathway is the inhibition of the opening of the mitochondrial permeability transition pore (mPTP) and facilitated opening of K_ATP channels. Protective actions of sevoflurane on endothelial cells have also been shown in human volunteers, which may yield additional beneficial effects [78]. Finally, sevoflurane administration may reduce inflammatory markers. In a recent meta-analysis of patients undergoing coronary artery bypass grafting (CABG) surgery, sevoflurane pretreatment significantly reduced concentrations of the pro-inflammatory cytokines interleukin (IL)-6 and IL-8 compared with controls [79].

After publication of the first promising results from clinical studies, the issue of anesthetic-induced preconditioning gained a great deal of interest. De Hert and others demonstrated that sevoflurane reduced troponin I release following cardiac surgery with CPB and after aortic valve replacement or mitral valve surgery [80‐83]. In one of these studies, which compared sevoflurane with propofol, the greatest effect on troponin release was observed when sevoflurane was applied before, during, and after CPB.

In the largest trial published thus far, enrolling 414 patients, De Hert et al. observed a significant difference between total intravenous anesthesia (TIVA) and sevoflurane in 1-year mortality (12.3 versus 3.3%, respectively; P = 0.034) [84]. Also, with respect to long-term outcome, the incidence of late cardiac events 1 year after cardiac surgery with CPB was significantly less in a randomized, controlled, double-blind study of patients who received 2 MAC sevoflurane or placebo (3 versus 17%; P = 0.038) on CPB for 10 min before aortic cross-clamping [85]. In addition, evidence from retrospective “real-life” database analyses of patients undergoing CABG surgery in Italy (N = 34,310 patients) and in Denmark (N = 10,535 patients) suggested that the use of volatile anesthetics is associated with a reduced 30-day mortality [86, 87]. Bignami et al. reported significantly reduced 30-day mortality rates with volatile anesthetics (P = 0.035) and with increased duration of volatile anesthetic use during surgery (P = 0.022) [86], whereas Jakobsen et al. found significant reductions in 30-day mortality with sevoflurane versus propofol in patients without a previous myocardial infarction (2.49 versus 3.55%, P = 0.025) or without unstable angina (2.28 versus 3.14%, P = 0.015) [87].

Despite this relatively robust evidence and American College of Cardiology Foundation/American Heart Association guidelines recommending the use of inhaled anesthetics in patients undergoing cardiac surgery with CPB (class IIa, evidence level A) [88], there are still doubts among clinicians about the efficacy of anesthetic-induced cardioprotection in daily practice. These doubts have been fueled by several trials that failed to show beneficial effects of sevoflurane in cardiac or major vascular surgery. A French dual-center study failed to demonstrate a significant cardioprotective effect after 15 min of sevoflurane administration prior to CPB [89]. In the most recent randomized study, enrolling 200 patients, Landoni et al. did not find a difference between patients anesthetized with sevoflurane or propofol in postoperative cardiac troponin release, 1-year all-cause mortality, rehospitalizations, and adverse cardiac events [90]. In major vascular surgery, Lurati-Buse et al., Lindholm et al., and Zangrillo et al. all failed to show beneficial effects of a sevoflurane-based anesthesia regimen compared with TIVA [91‐93].

In the most recent meta-analysis of sevoflurane-induced cardioprotection in patients undergoing cardiac surgery, which included 15 studies with a total of 1646 patients, the authors found improved postoperative cardiac output, reduced postoperative 24-h cardiac troponin I concentrations, reduced postoperative usage of inotropes and pressors, a shorter ICU stay, and a decreased incidence of atrial fibrillation [94]; mortality and major morbidity, however, were not different between groups.

Pharmacologic neuroprotection from cerebral ischemia and reperfusion continues to be an area of interest. Animal studies have suggested that ischemic neuronal injury results from early excitotoxic cell death mediated by excessive release of glutamate, and by oxidative stress caused by reperfusion injury together with delayed cell death as a result of apoptosis [95, 96]. However, early neuroprotective effects of volatile anesthetics in animal models indicate a beneficial effect in reducing excitotoxicity and only minimal effects on delayed cell death caused by apoptosis. Isoflurane has been demonstrated to reduce excitotoxic cell death in vitro [97] and in vivo [98] in models of focal [99, 100], hemispheric [101], and global [102, 103] ischemia. Desflurane and halothane were observed to have neuroprotective properties if used at 1.5 MAC during the ischemic event when compared with awake rats subjected to an ischemic insult [102].

Sevoflurane has shown anesthetic preconditioning (APC) potential [104]. Wang et al. reported that repeated sevoflurane APC reduced infarct size in rats after focal ischemia, and further investigated whether the inhibition of apoptotic signaling cascades contributes to sevoflurane APC-induced neuroprotection. Male rats were exposed to ambient air or 2.4% sevoflurane for 30 min per day for four consecutive days and then subjected to occlusion of the middle cerebral artery for 60 min at 24 h after the last sevoflurane intervention. APC with sevoflurane markedly decreased apoptotic cell death, which suggests that suppression of apoptotic cell death contributes to the neuroprotection against ischemic brain injury [105].

Despite some evidence of neuroprotective effects of inhaled anesthetics, conflicting data have also been reported. Extensive studies in neonate rodents have shown that postnatal exposure to high doses of commonly used anesthetic drugs, in isolation or in combination for a period of several hours, can induce apoptotic cell death with potentially long-term functional consequences [106]. Postexposure environmental factors also appear to play a substantial role in the expression of neurotoxicity [107]. General anesthetics can be a potential cause of neurologic sequelae in animal models, such as deficits in learning and memory [108, 109]. In addition, it should be noted that animal studies demonstrating neuroprotection by inhaled anesthetics have failed to translate to the clinical trial setting and, to date, it remains unclear if a quantitative and qualitative correlation can be made between the effects on rodents and those in humans. Therefore, despite research efforts, clinical evidence for ischemic brain protection by inhaled anesthetics remains controversial [110].

In summary, evidence demonstrates cardioprotective effects of inhaled anesthesia during cardiac surgery and, despite some contradictory findings, it is conceivable that sevoflurane also exerts cardioprotective effects outside cardiac surgery. The potentially beneficial effects may be diminished or even abolished by advanced age, diabetes, and myocardial remodeling. Therefore, it is of paramount importance to tailor the cardioprotective technique to the individual patient (e.g., by enhancing stimulus intensity through repeated sevoflurane wash-in and washout) and to avoid the use of drugs that are capable of abolishing cardioprotection in patients at risk for perioperative cardiovascular complications. Preliminary evidence also suggests that sevoflurane may possess neuroprotective properties in hypoxia/ischemia; however, a substantial amount of research is necessary to confirm the neuroprotective effect in the clinic.

Upon discontinuation of general anesthesia, patients typically remain unconscious for several minutes and then progress through a series of stages concluding with full consciousness. In general, the stages of emergence include the return of spontaneous respiration, patient behaviors prompting extubation, eye opening, and the ability to respond to questions [147]. Recently, novel approaches aimed at accelerating emergence and improving patient outcomes have been proposed and are under investigation.

CNS stimulants, which promote wakefulness and exhibit pro-cognitive effects in humans, may be useful for speeding up the emergence process. These drugs include amphetamine and methylphenidate, which are approved for the treatment of ADHD, and modafinil and armodafinil, which are approved for the treatment of narcolepsy. Animal studies have revealed decreased emergence times with methylphenidate [148] and dopamine receptor agonists [149] in rats anesthetized with isoflurane. A clinical study of methylphenidate investigating its effects on emergence time and short-term cognitive dysfunction is underway [150]. Demonstrating accelerated emergence and improved cognitive function with methylphenidate could provide physicians with an important new tool for the postanesthesia care of patients.

Beyond reducing the time to emergence from anesthesia, the potential to ameliorate both POD and POCD following surgery is of significant medical interest. A short-term treatment with pro-cognitive agents could potentially reduce the incidence of POCD. As mentioned, methylphenidate is being studied in this regard. In addition, amantadine, a dopamine agonist/weak NMDA antagonist, has been found in an animal model to reduce cognitive impairment following propofol anesthesia [151]. It is interesting to note that memantine, an NMDA antagonist, is approved for the treatment of dementia in Alzheimer’s disease. Consequently, it is tempting to speculate that memantine or other treatments for Alzheimer’s disease such as donepezil or rivastigmine, which exert their therapeutic effects through the inhibition of acetylcholinesterase, may be beneficial too. The potential also exists for the use of nonpharmacologic methods in the prevention of POCD. A recent study by Saleh et al. randomized elderly patients undergoing surgery to receive preoperative cognitive training during three 1-h sessions or no cognitive intervention. Neuropsychological testing 1 week after surgery revealed a significantly reduced incidence of POCD in the intervention group (15.9%) compared with the control group (36.1%) [152].