Possible indications of beta-blockers in the perioperative period other than prevention of cardiac ischemia
- First Online:
- Cite this article as:
- Kadoi, Y. & Saito, S. J Anesth (2010) 24: 81. doi:10.1007/s00540-009-0865-x
- 154 Views
According to the guidelines of the American College of Cardiology/American Heart Association 2006 for perioperative cardiovascular evaluation for non-cardiac surgery, beta-blocker therapy should be considered for high-risk individuals undergoing vascular surgery or high- and intermediate-risk patients undergoing non-cardiac surgery. This guideline might induce physicians to increasingly use beta-blockers in the hope of preventing perioperative cardiac complications. However, beta-blockers have potential beneficial effects outside the prevention of cardiac events. In addition to reducing anesthetic and analgesic requirements during the perioperative period, beta-blockers have neuroprotective effects in patients with brain trauma and possible effectiveness in the management of intraoperative awareness-induced post-traumatic stress disorder. Moreover, intrathecal administration of beta-blockers may have antinociceptive effects. Physicians need to bear in mind the benefits of beta-blockers for purposes other than preventing cardiac events when applied in the perioperative period, and they should be familiar with the pharmacodynamics and risk–benefit ratio with their use. This review focuses on possible extracardiac indications of beta-blockers.
Beta-blocker therapy is widely known to reduce perioperative cardiovascular complications [1, 2]. Recent guidelines from the American College of Cardiology/American Heart Association  recommend the use of beta-blockers in high-risk patients undergoing vascular surgery or high- and intermediate-risk patients undergoing non-cardiac surgery. While this guideline might induce physicians to increasingly use beta-blockers during surgery in the hope of preventing perioperative cardiac events, beta-blockers have potential beneficial effects in addition to the prevention of cardiac events. This review focuses on the possible extracardiac indications of beta blocker therapy.
Characteristics of clinically used beta-blockers, Prichard classification
Main metabolic organ
Propranolol = 1
Hepatic = renal
Antioxidant alpha 1-antagonist
Hepatic = renal
Beta-blockers reduce anesthetic and analgesic requirements during the perioperative period
Reduction of anesthetic requirements
Stanley et al.  were the first to report the possible beneficial effects of perioperative use of beta-blockers on the reduction of anesthetic dosage. They compared the dose of sufentanil required to produce unconsciousness during anesthetic induction in 22 patients receiving propranolol (80–240 mg/day, oral doses) and 22 patients not receiving propranolol or any other beta-blocker therapy prior to coronary artery bypass grafting (CABG). The variables compared between the two groups were incidence of intraoperative hypertension, requirements for supplements to treat hypertension, and recovery times. Patients on propranolol therapy required an average of 3.8 ± 0.3 μg/kg of sufentanil to reach unconsciousness, while patients not taking any beta-adrenergic blockers needed 4.9 ± 0.3 μg/kg of sufentanil. Total intraoperative sufentanil dose requirements were 11.1 ± 0.8 μg/kg for patients on propranolol and 15.0 ± 1.0 μg/kg for patients not taking any beta-blockers. In addition, systolic blood pressure increases of >20% the preoperative (control) values during sternotomy and maximal sternal spread occurred significantly more frequently in patients not taking beta-blockers than in those administered propranolol. Of the patients not taking beta-blockers, 18 and 27% required phentolamine to control arterial blood pressure before bypass and during the bypass, respectively. In contrast, only 5% of patients on propranolol therapy needed phentolamine before bypass, none of required it during bypass. Stanley et al. speculated that the possible mechanisms by which propranolol could potentiate the effects of sufentanil include increased occupation of the opiate receptor sites of the central nervous system (CNS), stimulation of CNS opiate receptors, reduction of plasma volume, changes in the distribution volume of sufentanil, decreased sufentanil metabolism, or alteration of CNS membranes, resulting in increased transfer of sufentanil into the brain.
In 1991, Smith et al.  compared esmolol, a short-acting beta-1 blocker, with alfentanil as a supplement for propofol–nitrous oxide–atracurium anesthesia. These researchers showed that esmolol can be used in place of alfentanil to supplement propofol–nitrous oxide–atracurium anesthesia in outpatients undergoing arthroscopic procedures, although esmolol-treated patients required more postoperative pain relief and opioid analgesia than those treated using alfentanil. Johansen et al.  subsequently examined in detail whether esmolol could reduce the propofol concentration required to prevent movement at skin incision. Patients were anesthetized with propofol alone, propofol plus low-dose continuous esmolol infusion (bolus of 0.5 mg/kg, maintenance dose of 50 μg/kg/min), or propofol plus high-dose continuous esmolol infusion (bolus 1 mg/kg, maintenance dose of 250 μg/kg/min) along with 60% nitrous oxide. The esmolol infusion was initiated before the induction of anesthesia and continued until the completion of the study. The researchers were able to show that the high-dose esmolol infusion induced a 26% reduction in the minimum effective plasma concentration (CP50; defined as suppression of movement at incision in 50% of patients) of propofol with nitrous oxide. This led them to conclude that esmolol may have anesthetic-sparing effects in humans under clinically relevant conditions. They also found that the minimum alveolar concentration (MAC) of isoflurane was not altered when esmolol was used alone but that it was significantly decreased while high doses of esmolol was used with alfentanil . These data indicate that esmolol does not have anesthetic effects per se, rather, it has anesthetic-sparing effects when used in conjunction with anesthetic agents. Johansen  examined whether esmolol affected the electroencephalogram (EEG) during propofol/alfentanil anesthesia and found that bispectral index (BIS) was significantly suppressed and that the burst suppression ratio (SR) was increased by esmolol (BIS 37 ± 6–22 ± 6, 40% decrease; SR 5 ± 7 to 67 ± 23, 13.4-fold increase) compared with baseline levels. In addition, discontinuation of esmolol reversed these changes in BIS and SR values.
Study characteristics of effects of beta-blockers on anesthetic requirements
Stanley et al. 
22 CABG patients
80–240 mg per day of propranolol
Lower requirement of sufentanil during surgery
Smith et al. 
2 mg/kg of esmolol infusion
No difference in hemodynamic variables between esmolol and alfentanil group
Johansen et al. 
Bolus 1 mg/kg, then 250 μg/kg/min of esmolol infusion
Lower propofol requirement
Johansen et al. 
Bolus 1 mg/kg, then 250 μg/kg/min of esmolol infusion
Decrease in MAC of isoflurane
Zaugg et al. 
Intra or pre- and postoperative atenolol treatment
Lower fentanyl requirement during surgery and morphine dosage in the PACU
Coloma et al. 
Esmolol infusion(12.8 ± 13.1 μg/kg/min) during surgery
Use of esmolol as an alternative to remifentanil
Menigaux et al. 
Bolus 1 mg/kg, followed by 250 μg/kg/min of esmolol infusion during induction
Attenuated somatic response to tracheal intubation prevented BIS arousal reactions
Berkenstadt et al. 
Single dose of 80 mg of esmolol
No affect on BIS value
Wilson et al. 
Bolus 1 mg/kg, followed by 250 μg/kg/min of esmolol infusion during induction
Reduction of propofol requirements for anesthetic induction by 25%
Ghosh et al. 
Metoprolol 100 mg for 1 h prior to surgery
Reduction of total dose of propofol required during surgery
Collard et al. 
Bolus of 1 mg/kg, followed by 5–15 μg/kg/min of esmolol infusion during surgery
Reduction of fentanyl requirements in PACU
Tanabe et al. 
Bolus of 0.031 mg/kg, followed by 0.01 mg/kg/min of landiolol infusion during surgery
Reduction of sevoflurane requirement during operation
Tanifuji et al. 
2 and 10 mg/kg intravenous or 200 mg/day for 10 days of propranolol
No effect on halothane MAC
Larson et al. 
Bolus of 1.5 mg/kg, followed by 100 μg/kg/min of esmolol infusion during surgery
No effect of esmolol on the pupillary response
Orme et al. 
Bolus 1 mg/kg, followed by 5–15 μg/kg/min of esmolol infusion during surgery
No effect on anesthetic requirement of propofol
Kurita et al. 
Bolus of 0.125 mg/kg, followed by 0.04 mg/kg/min of landiolol
No effect on isoflurane MAC
In contrast, using dogs, Tanifuji and Eger  examined the effects of propranolol on anesthetic requirements, specifically changes in the MAC of halothane, accompanying acute [2 and 10 mg/kg intravenous (iv)] and chronic (200 mg/day orally for 10 days) propranolol administration and found no effect of acute or chronic propranolol administration on halothane MAC. In addition, they demonstrated that an intravenous beta-agonist, isoproterenol also had no effect on MAC. These researchers suggested that since propranolol readily crosses the blood–brain barrier (BBB), neither acute inhibition of central beta-adrenergic receptors nor chronic receptor blockade with potential changes in central catecholamine levels influence volatile anesthetic MAC. Orme et al.  examined the effects of esmolol on propofol concentration and the prevention of response to command and found no effects of the use of esmolol on anesthetic requirements during propofol anesthesia. Larson et al.  examined whether esmolol infusion (intravenous bolus of 1.5 mg/kg of esmolol, followed by an infusion at 100 μg/kg/min) would blunt pupillary changes in response to noxious stimulation and found that esmolol infusion did not blunt such changes.
Landiolol (ONO 1101) is a newly developed, highly ultra-short-acting beta-1 blocker created by altering the chemical structure of esmolol [21–23]. This agent is approximately ninefold more potent in beta-1 blocking activity and eightfold more cardioselective than esmolol. The suppressive effects of landiolol on cardiovascular performance have been reported to be less potent than those of esmolol at equipotent beta-blocker doses . Landiolol exhibits suppression predominantly on the chronotropic effect, rather than on the inotropic effect. Thus, landiolol is suitable for stabilizing hemodynamics during the perioperative period. There have, however, been a few studies on the anesthetic-sparing effect of landiolol during anesthesia [24–27]. Kurita et al. , using a swine model, examined the effects of landiolol on the MAC of isoflurane required to prevent movement in response to a noxious stimulus in 50% of the experimental animals. A landiolol infusion was administered at a rate of 0.125 mg/kg/min for 1 min and thereafter decreased to 0.04 mg/kg/min. The results showed that landiolol attenuated the increases in heart rate and mean arterial blood pressure that occurred in response to the dewclaw clamp, but they did not alter the MAC of isoflurane. In contrast, Tanabe et al.  showed that a low dose of landiolol (bolus injection of 0.031 mg/kg and continuous infusion at a rate of 0.01 mg/kg/min) reduces intraoperative sevoflurane requirements during sevoflurane/N2O/fentanyl anesthesia in patients undergoing hip surgery.
One important question that emerged during the investigations of these anesthetic-sparing effects of beta-blockers was whether beta-blockers mask the signs of light anesthesia and thus increase intraoperative awareness . Ginsburg , in 1992, initially posed the question, “is IV esmolol an acceptable substitute for an inadequate anesthetic?” Since then, many studies have shown the efficacy of using beta-blockers in reducing anesthetic requirements during the perioperative period without any change in anesthetic depth [4–15]. Studies showing reduced anesthetic requirements by esmolol have mainly been based on BIS. This raises the important issue of whether BIS is a valid predictor of anesthetic depth. “Inadequate” anesthesia can cover several different levels, ranging from implicit memory, explicit memory, and obeying commands during anesthesia without recall, to actual awareness and recall . The proper definition of “inadequate anesthesia” is thus of great importance in addressing this issue.
Beta-blockers seem likely to have no anesthetic effect per se, as reported by Johansen et al. . The mechanisms underlying the effects of esmolol on anesthetic requirements remain unclear. One possible explanation may be that beta-blockers have CNS-modulating activities and exert anxiolytic effects (see following section). Another explanation may lie in an alteration of propofol pharmacokinetics by beta-blockers. Upton et al.  demonstrated that cardiac output is a determinant of the initial concentrations of propofol after the administration of short infusions. A reduction in cardiac output leads to reduced hepatic blood flow, which could affect the metabolism of drugs with a large hepatic extraction ratio, such as fentanyl. Beta-blocker use would thus likely result in the prolongation of the analgesic effects of fentanyl and also elicit a reduction in postoperative opioid consumption. Indeed, Kurita et al.  suggest that the pharmacodynamic changes and pharmacokinetic modifications of anesthetic agents induced by the decrease in cardiac output resulting from the administration of beta blockers should contribute to an anesthetic-sparing effect during anesthesia.
Major questions still exist about the mechanisms and sites of anesthetic action. Classically, it has been widely assumed that the brain, especially the cerebral cortex, is the main site of anesthetic action. Anesthetic effect may result in the several components of general anesthesia: unconsciousness, amnesia, and immobility in response to a painful stimulus. The standard for determining anesthetic requirements is the MAC of an inhaled agent that prevents gross purposeful movement in response to a supramaximal painful stimulus. Antognini et al.  demonstrated, in animal studies, the importance of subcortical structures, such as the spinal cord, in the generation of purposeful movement in response to a painful stimulus under general anesthesia. Laboratory studies [33, 34] have shown that it is of great interest that the mechanisms and sites of action of beta-blockers used for reducing anesthetic requirements during surgery may be closely related to the site of anesthetic action. In addition, the measurement and assessment of anesthetic potency, such as MAC, is also of great interest during investigations of the effect of beta-blockers on the reduction of anesthetic requirements. Rampil et al.  found that anesthetic-induced unresponsiveness to noxious stimuli measured by MAC testing did not depend on cortical forebrain structures in rats. These researchers therefore speculated that beta-blockers administered during the perioperative period may exert their anesthetic-sparing effects, not by effects on the cerebral cortex, but via the spinal cord.
Involvement of the sympathetic system in nociception is well established. Some reports have indicated that beta-blockers may produce antinociceptive effects. Davidson et al.  examined the effects of esmolol infusion on nociceptive behaviors induced by the formalin test in rats. Esmolol infused at low (40 mg/kg/h) and high (150 mg/kg/h) doses did not affect behavioral changes recorded immediately after formalin injection (phase 1; 0–5 min), whereas only the high dose was effective in decreasing limb lifting during the secondary response (phase 2; 10–35 min). Taira et al.  subsequently examined the effects of intrathecal landiolol on nociceptive behaviors in the rat formalin test. Intrathecal injection of landiolol was found to decrease the number of flinches in a dose-dependent manner, with the effect induced by 0.5 mg of landiolol being completely reversed by intrathecal isoproterenol. These researchers concluded that beta-blockers may have antinociceptive effects. Kinjo et al.  subsequently reported almost the same finding. Zhao et al.  recently examined the effects of administering intrathecal landiolol on nociception and spinal c-Fos expression by the mouse formalin test. The intrathecal injection of 750 μg/kg landiolol decreased pain-related behaviors in phase 1 (the initial acute phase: during the first 5 min after formalin injection followed by a prolonged tonic response), while intrathecal landiolol administered at 250, 500 and 750 μg/kg decreased pain-related behaviors in phase 2 (secondary and longer lasting response: beginning about 10 min after formalin injection) during the formalin test. c-Fos protein levels in the spinal dorsal horn were decreased by landiolol administered at 750 μg/kg. In a human study, Zaugg et al.  reported that the intravenous administration of atenolol pre-and postoperatively or during surgery resulted in lower pain scores and reduced the total morphine requirement in the PACU. Chia et al.  reported that patients treated with esmolol infusion during hysterectomy required less patient-controlled intravenous morphine over 3 days than those not administered esmolol, despite similar pain intensity and medication side effects, and they also received significantly lower concentrations of end-tidal isoflurane and fentanyl during anesthesia.
The mechanisms of beta-blocker-induced antinociceptive effects remain unclear. Tanahashi et al.  showed that propranolol, esmolol, landiolol, and lidocaine block tetrodotoxin-resistant Na channels in rat spinal dorsal root ganglia in a dose-dependent manner, although very high landiolol concentrations are required to achieve antinociceptive effects.
Noxious stimuli are conducted through the spinal cord, the brainstem reticular formation, and the thalamus to the cerebral cortex, where an electroencephalographic arousal response is evoked [41, 42]. Beta-adrenoceptors are present in various parts of the reticular activating system, particularly the medial septal region of the basal forebrain. In mice and rats, the locus coeruleus-associated noradrenergic system participates in arousal, and beta-blockers within this region reduce forebrain electroencephalographic activity [41, 42]. In a continuation of these studies, the same group of researchers  showed that amphetamine-induced activation of the rat forebrain is clearly inhibited by timolol. Radisavievic et al.  found that norepinephrine modulated excitatory amino acid-induced responses in developing human and adult rat cerebral cortexes. These findings strongly suggest the contribution of beta-adrenergic mechanisms to antinociception via a central site. However, the pharmacokinetics of beta-blockers, particularly esmolol, do not entirely support this mechanism. Esmolol is hydrophilic, as is atenolol. They produce the same plasma/cerebrospinal fluid ratio, and probably do not readily cross the BBB. However, Johansen  found that esmolol promoted electroencephalographic burst suppression during propofol/alfentanil anesthesia. Howie et al.  and Van den Broek et al.  showed that esmolol reduced the duration of seizures induced by electroconvulsive therapy (ECT). These reports suggest that esmolol may cross the BBB. In contrast, another study showed that esmolol had no effect on seizure duration . Given these contradictory results, the question of whether esmolol or atenolol can cross the BBB remains unanswered.
Another potential mechanism of the antinociceptive effects of beta-blockers is via their action on peripheral anti-inflammatory sites [48–52]. The results from a study using human subjects revealed that norepinephrine increases hyperalgesia in response to heat in skin-sensitized by capsaicin . In addition, Khasar et al.  showed that epinephrine produced cutaneous mechanical hyperalgesia and sensitized cultured dorsal root ganglion neurons in the absence of nerve injury via beta-adrenergic receptors and that these effects of epinephrine were mediated by both the protein kinase A and protein kinase C second-messenger pathways. Cunha et al.  reported that carrageenin-evoked hyperalgesia was attenuated by atenolol. Ernberg et al.  reported that injection of propranolol into the human masseter muscle reduces the pain induced by local administration of 5-HT.
Although extensive data from both experimental and clinical studies have shown that beta-blockers may have antinociceptive effects, the body of evidence is as yet insufficient to conclusively state that beta-blocker therapy is effective in blocking nociceptive stimuli. Further extensive experimental and clinical studies are needed to clarify the antinociceptive mechanisms of beta-blockers.
Central nervous system effects
Many studies have demonstrated that beta-blockers affect the CNS [53–67]. Ko et al.  reviewed the association of beta-blockers used in the treatment of myocardial infarction, heart failure, or hypertension with depression, fatigue, and sexual dysfunction, and showed that although the frequency of depressive symptoms was similar between subjects using beta-blockers (20.1%) and placebo (20.5%), the frequency of fatigue and sexual dysfunction symptoms was higher in the those using beta-blockers (fatigue: 33.4 vs. 30.4%, respectively; sexual dysfunction: 21.6 vs. 17.4%, respectively). The side effects of the use of beta-blockers, such as depressive symptoms or increased frequency of fatigue, clearly suggest that beta-blockers affect the CNS. Brismar et al. [54, 62] examined the relationship between beta-blocker-induced CNS side effects and the nightly urinary secretion of melatonin in hypertensive patients. They found that severe CNS side-effects, such as nightmares and sleep disorders, occurred only in patients treated with metoprolol (21%) and that these side-effects were always accompanied by low levels of melatonin. Kostis and Rosen  showed that beta-blockers can affect rapid eye movement (REM) sleep, possibly associated with nightmares. Yamada et al.  indicated that beta-adrenergic receptor occupation is related to sleep disorders. Another study showed that lipophilic beta-blockers have a relatively greater effect on sleep disorders .
The administration of moderate doses of the catecholamines epinephrine and norepinephrine shortly after memory training is known to result in the enhancement of later retention performance [57, 58]. Nielson and Jensen  compared the effects of beta-blockers on arousal-induced modulation of working memory between elderly and young humans. The young subjects, normal elderly subjects, and those taking no beta-blocker medications all showed enhanced long-term recognition performance as a result of arousal manipulation; in contrast, elderly patients with beta-blockers exhibited impaired arousal-induced modulation of working memory.
Although the effects of beta-blockers on the CNS, such as sleep disorders or memory impairment, may not be beneficial to surgical patients, Pitman et al.  undertook a pilot study to determine whether the administration of propranolol shortly after a traumatic event can prevent post-traumatic stress disorder (PTSD), thereby demonstrating that the acute post-traumatic administration of propranolol may have preventive effects on subsequent PTSD. The symptoms of PTSD resemble—and have been suggested to result from—the psychiatric sequelae of some degree of intraoperative awareness , so that the intraoperative use of beta-blockers may also prove effective in the management of intraoperative awareness-induced PTSD. Pitman stated that “The body releases epinephrine and other stress hormones during a trauma, probably as part of a survival mechanism. Epinephrine helps you cope in the traumatic situation, but also makes the memory of the situation stronger. Interrupting beta-adrenergic transmission during or immediately after trauma with propranolol can block the emotional potentiation of memory and forestall PTSD” .
Neuroprotective effects are another possible beneficial effect of beta-blockers on the CNS [70–77]. In animal studies, Lysko et al.  examined the neuroprotective mechanisms underlying the effects of carvedilol on cultured cerebellar neurons and on CA1 hippocampal neurons in gerbils exposed to brain ischemia. These researchers found that carvedilol protected cultured neurons in a dose-dependent manner against both glutamate-mediated excitotoxicity and oxidant stress. They concluded that carvedilol may reduce the risk of cerebral ischemia and stroke by virtue of both antihypertensive actions and antioxidant properties. Savitz et al.  subsequently demonstrated that carvedilol was neuroprotective in focal cerebral ischemia and may protect the ischemic brain by inhibiting apoptosis and attenuating the expression of tumor necrosis factor (TNF)-alpha and interleukin (IL)-1 beta. Little et al.  examined the effects of propranolol on cerebral blood flow and early ischemic changes following middle cerebral artery (MCA) occlusion in cats. Propranolol was infused at a rate of 1 mg/kg/h for 4 h, beginning 1 h before MCA occlusion, and a 4 mg/kg bolus was administered immediately before occlusion. These researchers found that although the changes in cerebral blood flow and EEG in the propranolol-treated cats were not different from those in propranolol-untreated cats, there were significant decreases in infarct volume in propranolol-treated cats relative to propranolol-untreated cats. The same group subsequently showed the beneficial effects of propranolol in acute focal cerebral ischemia . In contrast, Junker et al.  showed that clenbuterol induces neuroprotective effects via an increase in nerve growth factor expression and that this beneficial effect is blocked by propranolol. Goyagi et al. [81, 82] recently examined the neuroprotective effects of several beta-blockers in rats with transient focal cerebral ischemia. The rats received an intravenous infusion of saline, propranolol, carvedilol, esmolol, and landiolol 30 min before MCA occlusion, and the infusion was continued for 42 h. Cerebral infarct volumes in the cortex were lower in rats treated with propranolol (72 ± 33 mm3), carvedilol (64 ± 25 mm3), esmolol (65 ± 18 mm3), and landiolol (44 ± 18 mm3) than in saline-treated rats (205 ± 28 mm3).
A number of clinical studies have demonstrated beta-blocker exposure with reduced mortality in patients with head injury [72–77]. Inaba et al.  retrospectively evaluated 1156 patients with isolated head injury who were admitted to the Intensive Care Unit during a 90-month period. Of these, 203 (18%) received beta-blockers and 953 (82%) did not. Although patients receiving beta-blockers had severe head injury more frequently, had lower Glasgow Coma Scale scores, and more frequently required craniotomy, beta-blocker use was an independent protective factor for mortality (odds ratio 0.54, 95% confidential interval 0.33–0.91). Other investigators have reported almost the same results [74–77]. In addition, Amory et al.  retrospectively examined the perioperative use of beta-blockers on neurological complications in patients undergoing coronary artery bypass graft surgery and showed that the use of beta-blockers was associated with a substantial reduction in the incidence of postoperative neurological complications. These reports indicate that beta-blockers may exert neuroprotective effects when used in the perioperative period. Further extensive investigations are needed to verify these potentially important neuroprotective effects.
Immune response and beta-adrenergic antagonists
Activation of the hypothalamic–pituitary–adrenal axis and the sympathetic nervous system, resulting in a systemic release of adrenal steroid hormones and catecholamines, is a well-known essential component of the response to major surgery, brain injury, or trauma. The balance of pro-and anti-inflammatory cytokines may be altered perioperatively. These changes are thought to be attributable to the stress response of surgery or the response to cardiopulmonary bypass. An important consequence of the pro-inflammatory cytokine activity is increased adhesion of neutrophils. Further, high levels of these stress mediators are known to be immunosuppressive in surgical patients.
Some reports have shown that beta-blockers may have immunomodulatory effects during cerebral injury [83–89] or myocardial infarction [90, 91]. Prass et al.  investigated the effects of propranolol on the immunoinhibitory alterations induced by stroke. Their study showed that the administration of propranolol to animals with middle cerebral artery occlusion decreased the percentage of apoptotic splenocytes to levels observed in sham-operated mice and prevented decreases in peripheral blood lymphocyte counts. Another study  showed that propranolol prevented an increase in IL-10 plasma levels in a rat model of acute brain injury. Carvedilol increases the production of IL-12 and interferon-alpha  and decreases the production of TNF-alpha and IL-6 . Atenolol reportedly decreases the production of TNF-alpha . Katafuchi et al.  demonstrated that reduction in natural killer cytotoxicity produced by stimulation of the splenic nerve was completely blocked by the intravenous administration of nadolol (a peripherally acting beta-blocker), but not by infusion of prazosin (an alpha-antagonist).
With respect to the effects of beta-blockers on the immunomodulatory system during myocardial infarction, there have been some reports on the effects of beta-blockers on the inflammatory system [90–92]. Prabhu et al.  demonstrated the efficacy of metoprolol treatment in reducing the production of TNF-alpha and IL-1 beta protein in rat myocardial infarction models. Other studies [90, 91, 93] have also shown the effects of metoprolol or carvedilol on cytokine production in patients with myocardial infarction. Loiek et al.  reported that carvedilol prevents the increased production of reactive oxygen species in HL-60 cells.
In contrast, Lang et al.  reported that the continuous infusion of propranolol in septic rats exacerbated the sepsis-induced increase in skeletal muscle IL-6 and TNF-alpha protein. Schmitz et al.  reported that the administration of propranolol (0.5 mg/kg subcutaneous every 12 h) in septic mice increased the splenocyte apoptosis rate, reduced the proliferative capacity of splenocytes, and modulated cellular cytokine release (IL-6 and interferon-gamma), resulting in an increase in sepsis-induced lethality from 47 up to 68%. However, in a clinical study, Jeschke et al.  demonstrated that propranolol not only decreased serum TNF and IL-1 beta levels compared with controls but that it also attenuated the incidence of sepsis in severely burned children. These reports may indicate that the effects of beta-blockers depend on the immunomodulatory condition of the animal before administration of the drug, as suggested by Oberbeck . It is interesting that Benish et al.  found that the combination of a COX-2 inhibitor and a beta-blocker (etodolac and propranolol) attenuated surgery-induced decreases in natural killer cytotoxicity, suggesting a possible reduction in the risk of tumor metastasis.
Although there is as yet no laboratory or clinical evidence supporting the suggestion that the perioperative use of beta-blockers would be able to modulate the immune–endocrine system, when the effects of anesthetic agents on the immune–endocrine system  are considered together with the results of these experimental studies, the perioperative use of beta-blockers may plausibly modulate the immunosuppressive alterations induced by surgical stress.
Effects of beta-blockers on pulmonary function
There have been some reports on the effects of beta-blockers on pulmonary function [101–103]. Although patients with active asthma should not receive beta-blockers, chronic obstructive pulmonary disease is not a contraindication to the use of beta-blockers perioperatively, and even patients with inactive stable asthmatic disease may be given a low dose of highly selective beta-1 blockers, such as landiolol. While the single-dose administration of beta-agonists produces bronchodilation and inhibits airway hyperresponsiveness, chronic repetitive administration may increase airway hyperresponsiveness, airway inflammation, and the risk of death. It is interesting to note that Nquyen et al.  demonstrated that chronic administration of beta-blockers reduced inflammation and mucous metaplasia in a murine model of asthma.
Another possible beneficial effect of beta-blockers on pulmonary function is that beta-blockers enhance hypoxic pulmonary vasoconstriction [102, 103], which may be favorable in patients undergoing one-lung ventilation.
Use of short-acting beta-blockers during intubation or emergence periods
Laryngoscopy and tracheal intubation maneuvers induce marked hemodynamic changes, such as increases in heart rate and systolic blood pressure, which, in combination, result in an unfavorable supply–demand balance of myocardial oxygen. The degree of response to laryngeal stimulation appears to vary with the depth of anesthesia , the duration  and difficulties encountered during tracheal intubation [106, 107], as well as on patient-dependent variables, including age and history of diabetes or cardiovascular disease [108, 109]. Although the hemodynamic changes are transient, drastic hemodynamic changes in patients with pre-existing ischemic coronary disease, hypertension or cerebrovascular disease may increase the risk of myocardial ischemia, arrhythmia, and even infarction and cerebral hemorrhage [110, 111].
To date, many studies have shown the effectiveness of esmolol in blunting the hemodynamic changes induced by endotracheal intubation. Figueredo and Garcia-Fuentes  provided an excellent review of the effectiveness of different regimes and doses of esmolol, as well as the grade of hypotension and bradycardia that can be produced when esmolol is used in conjunction with anesthesia-inducing agents. Of the 72 publications they identified, 38 randomized controlled trials containing a total of 2009 patients were ultimately included in their analysis. Following anesthetic induction, systolic blood pressure values decreased (with respect to baseline) by 6.1% in placebo groups and by 13.8% in esmolol-treated patients. Following endotracheal intubation, systolic blood pressure increased by 26.3% in patients in the placebo group compared with an increase of 9.1% in patients treated with the various regimes of esmolol. After induction of anesthesia, minimum heart rates for the placebo group were 7.2% higher than baseline values, while those for the esmolol group were 4.2% lower than baseline values. Endotracheal intubation resulted in a 29.6% increase in the heart rate of patients in the placebo group compared with a 9.3% increase in patients treated with esmolol. These researchers concluded that, with respect to the administration of esmolol “before” or “after” administration of the induction agents, no significant differences were observed between the two alternatives with respect to any of the variables assessed. They also concluded that esmolol is effective in blocking the tachycardia and systolic blood pressure increase induced by airway manipulation. However, the use of esmolol is associated with a dose-dependent risk of hypotension during the induction of anesthesia, thereby precluding its routine use in anesthesia. The use of esmolol in specific risk groups remains controversial, and in groups in whom the risk–benefit ratio is difficult to predict, usage needs to be evaluated on an individual basis. To diminish the incidence and seriousness of the side-effects, a reasonable recommendation is to administer a small loading dose (500 μg/kg) of esmolol over 4 min, followed by a continuous intravenous infusion of 200–300 μg/kg/min.
In contrast to the use of esmolol to prevent hemodynamic changes during endotracheal intubation, few studies have examined the effects of landiolol on hemodynamic changes associated with endotracheal intubation [113–117]. Kitamura et al.  examined the dose-related effects of landiolol on cardiovascular responses to tracheal intubation and found that although the administration of 0.25 and 0.5 mg/kg landiolol decreased the incidence of tachycardia, these doses were insufficient to suppress increases in systolic blood pressure. Hasuo et al.  examined the effects of a continuous infusion of landiolol on hemodynamic responses to endotracheal intubation during induction with isoflurane. Landiolol was infused at a rate of 0.125 mg/kg/min for 1 min as a loading dose, followed by a maintenance dose of 0.04 mg/kg/min. These researchers found that significant increases in heart rate occurred in both the control and landiolol groups in response to isoflurane inhalation and tracheal intubation, although the magnitude of the increase was significantly reduced in the landiolol group. Blood pressure increased after tracheal intubation in the control group but remained unchanged in the landiolol group. Plasma concentrations of norepinephrine increased after induction and intubation in both groups, with no significant differences between groups. Only one study has compared the effects of esmolol and landiolol on hemodynamic responses to endotracheal intubation: Oda et al.  examined the comparative effects of esmolol and landiolol on hemodynamic changes and BIS during anesthesia induction. Esmolol was administered as a bolus of 1.0 mg/kg, followed by infusion at 0.25 mg/kg/min, while landiolol was given as a bolus of 0.125 mg/kg, followed by infusion at 0.04 mg/kg/min. The administration of these agents was started 5 min after the induction of anesthesia, and endotracheal intubation was performed 12 min after anesthetic induction. These researchers observed no significant differences in hemodynamic changes between esmolol and landiolol groups throughout the study periods. Sugiura et al.  examined the optimal dose of landiolol in terms of the hemodynamic alterations induced by laryngoscopy in normotensive and hypertensive patients. Landiolol at a dose of 0.1 mg/kg was the most effective against intubation-induced tachycardia when infused 4 min before intubation in normotensive patients. However, a 0.2 mg/kg dose of landiolol was necessary to prevent tachycardia after intubation in hypertensive patients. They concluded that landiolol had no significant effects on arterial blood pressure at any dose, indicating that landiolol is suitable for stabilizing hemodynamic changes during intubation.
Emergence from anesthesia and tracheal extubation are also associated with marked hemodynamic alterations, such as increases in heart rate and systolic blood pressure. Such changes may induce adverse effects in patients with ischemic heart disease. Physicians therefore need to be cognizant of whether ultra-short-acting drugs, such as esmolol or landiolol, provide hemodynamic stabilization during emergence from anesthesia and tracheal extubation. Some reports have examined the effects of esmolol on hemodynamic changes during the extubation period [120–123]. Dyson et al.  assessed the effects of three doses of esmolol (1.0, 1.5, and 2.0 mg/kg) given as a bolus 2 min after the reversal of neuromuscular blockade, in a double-blinded study, and found that all three doses of esmolol attenuated increases in heart rate but that 1.0 mg/kg was insufficient to control increases in systolic blood pressure associated with emergence from anesthesia and tracheal extubation. They concluded that doses of 1.5 and 2.0 mg/kg controlled both systolic blood pressure and heart rate, but that the larger dose produced significant decreases in systolic blood pressure. Kurian et al.  examined whether esmolol infusion affected the incidence of ST-segment changes during tracheal extubation after coronary artery surgery. Esmolol was infused from the time of the patient's entry into the intensive care unit until 180 min after tracheal extubation, with the aim of maintaining heart rate within the range of 55–75 beats/min. A “sliding scale” of esmolol at 0–300 μg/kg/min was used, depending on heart rate. Patients in the esmolol group displayed a lower incidence of myocardial ischemia than those in the placebo group (3/31 vs. 12/37, p = 0.05). Although esmolol infusion stabilized the hemodynamic changes associated with extubation, seven patients in the esmolol group suffered adverse events related to esmolol infusion that were not found in the placebo group. These researchers therefore concluded that although the use of esmolol reduced the incidence of myocardial ischemia, the incidence of adverse effects makes this agent unsuitable for routine prophylaxis in patients after coronary artery surgery. Kovac et al.  compared the effectiveness of intravenous nicardipine (0.03 mg/kg) and esmolol (1.5 mg/kg) in controlling heart rate and blood pressure responses to emergence and extubation and found that although esmolol was more effective than nicardipine in attenuating heart rate responses to extubation, nicardipine was more effective in controlling blood pressure responses. An interesting study was performed by Grillo et al.  in which a 0.3 mg/kg/min infusion of esmolol was administered from the end of anesthesia to 15 min after extubation to patients undergoing neurosurgery. They found that esmolol blunted increases in cerebral blood flow velocity during emergence, an effect that may be attributable to hyperemia during neurosurgical recovery.
In contrast to esmolol, very few reports have examined the effects of landiolol on the hemodynamic changes of emergence from anesthesia and extubation [124–126]. Nonaka et al.  looked at the effects of continuous landiolol infusion on the hemodynamic changes of emergence. Landiolol was continuously infused at a rate of 0.125 mg/kg/min for 1 min immediately after the injection of neostigmine–atropine. The results demonstrated the efficacy of landiolol administration on hemodynamic stabilization during recovery from general anesthesia. Nakagawa et al.  reported a case demonstrating the usefulness of landiolol in the prevention of myocardial ischemia during extubation and recovery from anesthesia.
Specific points for attention
Metabolic effects of beta-blockers
Several studies have indicated that there is some hyperglycemic effect in patients treated with beta-blockers. Although the mechanism of beta-blocker-induced hyperglycemia is not clearly proven, it is reported that beta-blockers increase fasting glucose by as much as 28 mg/dl and glycosylated hemoglobin by 1% . A recent large study  showed that beta-blockers impaired glucose tolerance and appeared to increase the risk of diabetes on a long-term basis by 28%. Another large study from Bakris et al.  compared the effects of carvedilol and metoprolol on glycemic control, showing that carvedilol stabilized glycosylated hemoglobin and improved insulin resistance compared with metoprolol. However, to date, there have been no reports describing the use of beta-blockers on glucose homeostasis during the perioperative period.
Another criticism of the use of beta-blockers is unfavorable changes in lipid metabolism , although this adverse effect is not of immediate concern in the perioperative period. Significant increases in mean plasma total and very-low-density lipoprotein cholesterol and reductions in high-density lipoprotein cholesterol and free fatty acids concentrations have been reported with the use of atenolol, metoprolol, propranolol, and oxprenolol.
Concurrent administration of beta-blockers with drugs that alter gastrointestinal, hepatic, and renal function may affect the plasma concentrations, duration of action, and efficacy of beta-blockers.
It is widely known that the co-administration of calcium channel blockers with beta-blockers induces suppression of myocardial function, sinus arrest, or atrioventricular block. The review articles by Kjeldsen et al.  and Brouwer et al.  emphasize that although the clinical importance of the combined use of calcium channel blockers and beta-blockers on the treatment of angina pectoris in patients with coronary artery disease has been documented, careful attention should be paid to the concurrent use of calcium channel blockers with beta-blockers in the presence of impaired left ventricular function, bradycardia, or conduction abnormalities.
There have been a number of reports describing the possible interactions of beta-blockers and other agents. Westphal et al.  found an increased bioavailability of digoxin with oral co-administration of talinolol, resulting from competition for intestinal P-glycoprotein. Other reports show the possible interaction of beta-blockers and clonidine [133, 134].
Of great concern to anesthesiologists is a possible interaction between anesthetic agents and beta-blockers. Remifentanil and esmolol are metabolized by non-specific esterases and other tissues, which may alter the pharmacokinetics of remifentanil. However, laboratory studies in rats revealed no pharmacokinetic or pharmacodynamic interaction between remifentanil and esmolol in this animal model [135, 136].
Propofol may alter myocardial beta-adrenoceptor binding and responsiveness. However, Zhou et al.  demonstrated that relatively high concentrations of propofol were needed to antagonize beta-adrenoceptor binding and tissue responsiveness in the rat heart. Based on experimental studies, it has also been reported that volatile anesthetics can modify beta-adrenoceptor stimulations [138–140]. However, as yet, the interaction between propofol or volatile anesthetics and beta-blockers has not been clinically proven.
While beta-blockers have been widely used for hemodynamic stabilization during the perioperative period, they have a number of additional potential beneficial effects other than prevention of cardiac events, such as antinociceptive effects, reduced anesthetic dosage during the perioperative period and, possibly, immunomodulation. Physicians need to bear in mind the benefits of beta-blockers other than for preventing cardiac events when using these agents in the perioperative period, and they should be familiar with their pharmacodynamics and balance of risks and benefits when using them for these extra-cardiac beneficial effects.