Use of alpha-agonists for management of anaphylaxis occurring under anaesthesia: case studies and review

Anaphylaxis is a systemic, immediate hypersensitivity reaction caused by rapid, IgE-mediated release of mediators from tissue mast cells and peripheral blood basophils. The manifestations of the reaction are vascular, respiratory and cutaneous, with cardiovascular features being the most common [1]. Death can occur from airway obstruction, bronchospasm or cardiovascular collapse. Most deaths occur among individuals with no history of drug allergies [1].

The clinical manifestations of intra-operative reactions differ from those of anaphylactic reactions outside of anaesthesia. Cutaneous manifestations are far less common and cardiovascular collapse may be more common. The diagnosis can be made more difficult because patients cannot express all symptoms [2].

The incidence of anaphylactic reactions associated with anaesthesia is difficult to predict, but it is estimated that a suspected anaphylactic reaction occurs in 1 in 3500 anaesthetics, with true anaphylaxis seen in 1 in 6000 [3, 4]. In Australia [5, 6] the incidence is reported to be between 1 in 10 000 to 1 in 20 000 anaesthetics. The agents most commonly responsible are neuromuscular blocking agents, latex and antibiotics [2]. Events are probably under-reported.

Management depends on the severity of the event. Severe reactions require early recognition and aggressive resuscitation. Traditionally, the mainstay of treatment is oxygen, fluids and epinephrine, with cardiopulmonary resuscitation and advanced life support instituted as required.

Currently, there are no guidelines or algorithms for management of anaphylaxis that include the use of vasopressors other than epinephrine [7–16]. We present two cases of anaphylaxis occurring during anaesthesia that required the use of an α-agonist for successful management.

Case Reports

Patient 1

A 55-year-old man was scheduled for an elective resection of a right temporoparietal cerebral tumour. His only significant past medical history was sick sinus syndrome; he had had a permanent pacemaker inserted in the previous year. Pacemaker function had been checked recently. He had no known allergies and his current medications were sodium valproate 200 mg, dexamethasone 4 mg twice daily and citalopram 20 mg.

General anaesthesia was induced and maintained with remifentanil (0.05–0.5 μg.kg⁻¹.min⁻¹) and a target controlled infusion of propofol (target levels 3–5 μg.ml⁻¹) following pre-oxygenation. Muscle relaxation was achieved with 50 mg rocuronium prior to intubation. Cephalothin 2 g and hydrocortisone 100 mg were given following induction. Boluses of metaraminol 0.5 mg were given intermittently if the mean arterial blood pressure dropped more than 20% below the baseline measurement.

The patient was stable for the first hour of the procedure, at which time gelofusine was given. Within minutes the patient became hypotensive and quickly deteriorated into electromechanical dissociation. A cardiac arrest was ‘called’ and cardiopulmonary resuscitation measures were commenced according to Advanced Cardiac Life Support guidelines [10]. The gelofusine infusion was stopped and assuming an anaphylactic response, anaesthesia was discontinued and the patient was ventilated with 100% oxygen. Crystalloid was administered rapidly and epinephrine 1 mg intravenously was given three times before starting an infusion at 100 μg.min⁻¹. Despite these measures, effective spontaneous circulation was not achieved. Metaraminol 2 mg was then given, which produced a rapid return of spontaneous circulation. In fact, the epinephrine infusion was rapidly discontinued.

The patient was transferred to the intensive care unit for ongoing respiratory and circulatory monitoring and support. He was eventually discharged from hospital 9 days later. He had a persistent left hemiparesis which was compatible with the residual effects of surgery.

Patient 2

A 66-year-old man was admitted for an elective cysto-prostatectomy for management of his prostatic cancer. His only current medications were inhaled bronchodilators for mild chronic airways disease. He had no known allergies. He had previously undergone uneventful general anaesthesia.

A thoracic epidural was placed before induction under fentanyl (50 μg) and midazolam (1 mg) sedation. Only a test dose of local anaesthetic was administered. General anaesthesia was induced with alfentanil and propofol target controlled infusion (target levels 3–5 μg.ml⁻¹) following pre-oxygenation. Muscle relaxation was achieved with rocuronium 50 mg prior to intubation.

Within 2 min of intubation the patient was noted to be hypoxic and cyanosed. There was a coincidental fall in blood pressure and marked reduction in the recorded end-tidal carbon dioxide concentration. Tracheal tube position was confirmed; no bronchospasm was audible on auscultation of the chest. Hypotension quickly deteriorated to electromechanical dissociation. A cardiac arrest was ‘called’ and cardiopulmonary resuscitation measures according to Advanced Cardiac Life Support guidelines were commenced [10]. Despite these measures the patient's cardiac rhythm deteriorated to ventricular fibrillation.

Resuscitation was continued for a further 40 min. Little or no response was seen to a combination of rapidly infused fluids (3–4 l of crystalloid and colloid solutions) and six boluses of intravenous epinephrine 1 mg, followed by an infusion of up to 200 μg.min⁻¹. Repeated attempts at defibrillation (15 salvos of three shocks; initially 200, 300 and 360 J and then 360 J thereafter) were unsuccessful and the ventricular fibrillation amplitude and frequency became progressively finer. After repeated attempts, intra-arterial pressure monitoring was established at this point. External chest compressions were producing a systolic blood pressure of approximately 100 mmHg but a diastolic blood pressure of zero. Metaraminol 10 mg was then given in divided doses and the diastolic blood pressure slowly rose over some minutes to about 30–50 mmHg. Shortly after, the patient's rhythm reverted to sinus bradycardia. The rate improved with atropine and gradually a spontaneous output returned.

The patient was stabilised prior to transfer to the intensive care unit for ongoing respiratory and circulatory monitoring and support. He subsequently made a good recovery with mild memory loss and personality change. The diagnosis of anaphylaxis was confirmed by a positive mast cell tryptase (33 units.l⁻¹). The patient declined subsequent skin testing when offered this at follow-up.

Discussion

Anaphylaxis is an uncommon but serious complication of anaesthesia. The estimated incidence of intra-operative anaphylaxis is between 1 : 3500 and 1 : 20 000, with neuromuscular blocking agents and colloids being frequently implicated [3–6]. The pathophysiological effects result from immune-mediated release of mediators including histamine, leukotrienes, bradykinin and platelet activating factor [17, 18]. The result can be profound vasodilation and increased vascular permeability, with transudation of fluid resulting in both oedema and hypovolaemia [19]. The reduction in circulating blood volume has been estimated to be as much as 37%[20, 21]. This culminates in decreased ventricular filling, reduced cardiac output and potential shock. Reduction in diastolic blood pressure results in decreased myocardial perfusion, which in turn can cause ischaemia, cardiac failure and malignant arrhythmias. The same biochemical mediators are also responsible for airway constriction, oedema and obstruction and cutaneous changes such as urticaria.

The management of peri-operative anaphylaxis has been reviewed recently in several journals [11–14], and although updating current knowledge, none of these reviews advocates a change in management significantly different to that proposed by Fisher & Baldo in 1994 [5]. In essence, this involves provision of oxygen and airway support, external cardiac compression as necessary, vigorous fluid loading and administration of parenteral epinephrine as a vasopressor and inotrope to maintain adequate perfusion pressures. Similarly, in guidelines for the management of ‘Suspected anaphylactic reactions associated with anaesthesia’ recently published by the Association of Anaesthetists of Great Britain and Ireland [16], only epinephrine is listed as a vasopressor and/or inotrope to use in the setting of cardiovascular collapse. The potential inadequacies of the Association's guidelines have recently been highlighted by McBrien [22].

There are no clinical trial data providing evidence for treatment of acute anaphylaxis [14, 16]. The recommendations are based on clinical observations, interpretation of pathophysiology and animal models [14, 16]. In the two cases of anaphylaxis under anaesthesia presented here, return of spontaneous circulation was refractory to epinephrine but did occur following the administration of the α-agonist metaraminol.

There are several other accounts of the ineffectiveness of epinephrine in the setting of anaphylaxis. Waldenhausen et al. [21] reported 49 patients with anaphylaxis who did not respond adequately to epinephrine. Higgins & Gayatri [23] used 10 mg methoxamine to restore circulation in a patient with anaphylaxis to succinylcholine who had been resistant to a total of 5 mg epinephrine. Allen et al. [24] described a patient with anaphylaxis to rocuronium who failed to respond to cardiopulmonary resuscitation, fluids and epinephrine, and eventually required cardiopulmonary bypass. McBrien et al. [25] described two anaphylactic reactions (succinylcholine and cefamandole) and one cardiovascular collapse (secondary to methylmethacrylate cement) in which standard Advanced Cardiac Life Support guidelines did not return spontaneous circulation, but in which there was a good response to methoxamine. Konarzewski & De'Ath [26] reported an anaphylactic reaction which was treated with standard cardiopulmonary resuscitation including cardiac massage, ventilation of the lungs with 100% oxygen, atropine 3 mg and epinephrine 2 mg. Despite vigorous resuscitation attempts over 45 min, the patient died. Gibbs et al. [27] described protracted cardiopulmonary resuscitation with complete recovery in a patient where electromechanical dissociation degenerated to ventricular fibrillation following epinephrine. The patient was successfully defibrillated but required phenylephrine and dopamine to treat persistent hypotension.

There are several potential advantages of epinephrine other than its actions on peripheral vascular resistance and blood pressure. β-agonist actions facilitate bronchodilation [17] and enhance production of cyclic adenosine monophosphate, which may reduce mediator release [28]. Interaction at cardiac β-adrenergic receptors may also augment ventricular contractility. However, epinephrine also has potentially detrimental effects. Intravenous epinephrine has been associated with fatal arrhythmias and myocardial infarction [14, 17]. There is evidence that β-agonist stimulation during ventricular fibrillation may be detrimental. Livesay et al. [29] demonstrated that the β₁-effects of epinephrine augment the vigour of fibrillation, increasing myocardial oxygen demand and intraventricular pressures, which decrease subendocardial perfusion. β₂-stimulation appears to preferentially dilate larger coronary vessels, potentially diverting blood away from the subendocardium, which further jeopardises perfusion. Overall, the combined effects appear to adversely affect the balance of myocardial oxygen supply and demand. β₂-mediated vasodilation may also be deleterious in the setting of anaphylaxis [25].

Guidelines for the management of cardiovascular collapse associated with anaphylaxis are derived from those developed for the management of cardiac arrest. The initial investigations pioneering the use of adjunctive vasopressors in cardiopulmonary resuscitation were published by Crile & Dolley in 1906 [30]. This was expanded upon by Redding & Pearson in the 1960s [31–35], and their work laid the foundations for current resuscitation guidelines. They showed that intravenous administration of 1 mg epinephrine improved the outcome, in terms of restoration of spontaneous circulation from asystole and ventricular fibrillation, in 10 kg dogs [31]. They also showed that there was no significant difference in clinical effectiveness between epinephrine, phenylephrine, metaraminol and methoxamine [34], and claimed that agents with predominantly β-effects were of ‘no value’. Methoxamine was later demonstrated to be twice as effective as epinephrine in resuscitation from ventricular fibrillation [36] and more effective than other drugs for resuscitation from electromechanical dissociation [37]. Redding eventually claimed that: ‘…methoxamine is the drug of choice during attempted resuscitation from cardiac arrest regardless of the ECG pattern.’ These authors, and others, concluded that the efficacy of epinephrine and other vasopressors in management of cardiac arrest lies in their α-adrenergic properties [37, 38]. In fact, it has been demonstrated that pretreatment of dogs with α-blockers (phenoxybenzamine) precludes the ability to obtain return of spontaneous circulation when epinephrine is used for resuscitation [39].

Although not looking specifically at management of anaphylaxis, there have been a number of human studies comparing α-adrenergic agonists with epinephrine for treatment of cardiac arrest. In a 1985 clinical study, Silfvast et al. [40] found that phenylephrine (1–2 mg) and epinephrine (0.5–1.0 mg) were equally effective in restoring spontaneous circulation in a total of 65 patients treated for ventricular fibrillation, asystole, or electromechanical dissociation. In 1988, Turner et al. [41] administered epinephrine 1 mg or methoxamine 10 mg to 80 patients with electromechanical dissociation. Survival rates following the administration of the two drugs were similar and declined in parallel over time. In 1991, Lindner et al. [42] randomly administered epinephrine (1 mg) or norepinephrine (1 mg) to 50 patients suffering out-of-hospital ventricular fibrillation and failing three consecutive attempts at electrical defibrillation. Return of spontaneous circulation was established in 16 of 25 patients given norepinephrine, vs. only eight of 25 given epinephrine; circulation also returned more promptly after norepinephrine. Six patients given norepinephrine survived to hospital discharge, vs. four of those given epinephrine. In the first hour following return of spontaneous circulation, eight of 16 initial survivors given norepinephrine developed ventricular arrhythmias, compared with six of eight given epinephrine. The possible advantages of pronounced α-adrenergic stimulation in the management of ventricular fibrillation appeared to be confirmed in this clinical setting. In 1992, Callaham et al. [43] compared high-dose epinephrine (15 mg) and norepinephrine (11 mg) in managing out-of-hospital cardiac arrest. Norepinephrine was associated with a greater incidence of survival to hospital discharge, but this difference was not statistically significant. Leschinskiy [44] noted that early use of phenylephrine was beneficial in cases of circulatory arrest secondary to ventricular fibrillation and electromechanical dissociation.

However, not all published work confirms these conclusions. Olson et al. [45] in 1989 reported on the use of repeated doses of either methoxamine 5 mg or epinephrine 0.5 mg in 102 patients suffering prehospital ventricular fibrillation. In this study, epinephrine was significantly superior in promoting defibrillation (49% vs. 27.5%), return of spontaneous circulation (39% vs. 18%), and hospital discharge (20% vs. 8%). Ford et al. [46] stated that only four of 18 patients treated with metaraminol had a good response during an anaphylactic or anaphylactoid reaction. In comparison, 14 out of 17 patients treated with epinephrine improved. However, there is no indication of whether equivalent doses were given and it may well be that the response to epinephrine was secondary to the α-effects of large doses.

The benefit from administration of vasopressors is probably secondary to increased systemic vascular resistance and diastolic blood pressure, especially when shock is distributive in nature. Sanders et al. [47] and Livesay et al. [29] have shown that aortic diastolic pressure was a prognostic indicator of survival in dogs. Paradis et al. [48] showed that return of spontaneous circulation occurred in 24 of 42 patients achieving maximal coronary perfusion pressures (aortic diastolic minus right atrial pressure) of 15 mmHg, compared with none of 58 patients who did not. It is possible that administration of α-agonists gives a more reliable increase in vascular resistance and diastolic blood pressure [25]. In our second case, invasive monitoring of blood pressure demonstrated that the diastolic blood pressure (and probably myocardial blood flow) remained inadequate despite cardiopulmonary resuscitation, fluids and epinephrine given according to the Advanced Cardiac Life Support guidelines. Administration of metaraminol resulted in increased diastolic blood pressure and return of spontaneous circulation which we believe resulted in a successful outcome.

The factors most important to determining outcome from cardiac arrest, whatever the cause, are:

Unfortunately, cardiopulmonary resuscitation alone or with a standard dose of epinephrine (10–20 μg.kg⁻¹) has been shown to only provide 1–2 ml.min⁻¹.100 g⁻¹ of cerebral blood flow, about one tenth of normal [53]. Meaningful cerebral recovery therefore requires prompt restoration of spontaneous circulation. The goals should therefore be the early return of spontaneous circulation and maintenance of adequate coronary and cerebral perfusion. We believe that early administration of an α-agonist (e.g. metaraminol or equivalent) when patients are unresponsive or show limited response to epinephrine, will hasten achievement of these goals and hopefully translate into improved outcomes following intra-operative anaphylaxis.

Acknowledgements

We would like to acknowledge the patients' and their families' consent and support for the publication of these reports.