Perioperative anaphylaxis: pathophysiology, clinical presentation and management

Learning objectives.

By reading this article you should be able to:

• •Describe trends in the incidence and agents that cause perioperative anaphylaxis.
• •Explain the underlying pathophysiology.
• •Distinguish the different patterns of clinical presentation and apply the Ring and Messmer severity grading scale.
• •Implement appropriate management for patients with anaphylaxis.

Key points.

• •Perioperative anaphylaxis is a life-threatening immediate hypersensitivity reaction that is usually IgE-mediated.
• •Antibiotics and neuromuscular blocking agents are the most common triggers.
• •The Ring and Messmer four step (I–IV) grading scale is the most widely accepted tool for describing the clinical severity.
• •The most common clinical presentation includes cardiovascular collapse, tachycardia and cutaneous features (Grade III).
• •The cornerstones of management are adrenaline (epinephrine) and i.v. fluids.
• •Diagnosis is based on the clinical presentation, in conjunction with mast cell tryptase concentrations and the results of skin testing.

Clinical scenario.

A 56-yr-old woman of ASA physical status class 2 was scheduled for laparoscopic cholecystectomy. Her past medical history included hypertension (treated with rilmenidine) and previous uneventful curettage and colonoscopy procedures. Anaesthesia was induced with propofol, sufentanil and rocuronium. Tracheal intubation was performed and mechanical ventilation initiated. Within 2 min of giving rocuronium, the patient developed generalised erythema with cardiovascular collapse (arterial pressure 50/30 mmHg, HR 140 beats min⁻¹, oxygen desaturation to SpO₂ 87%) and hypocapnia (Pe´CO₂ 23 mmHg). She was treated with i.v. crystalloids 1000 ml, ephedrine 12 mg, and adrenaline (epinephrine) 100 μg and this restored cardiovascular stability within 20 min. No further vasopressors were required. Her surgery was postponed to a later date.

Plasma histamine and serum tryptase concentrations were measured 25 and 120 min after the reaction. Histamine concentrations were >100 and 4.6 nmol L⁻¹ (normal values <10 nmol L⁻¹) and tryptase concentrations were 36 and 21 μg L⁻¹ (normal values <11.4 μg L⁻¹). Serum tryptase concentration the next day was 8.4 μg L⁻¹. Skin testing was performed 4 weeks later and showed a strong positive response to rocuronium and cross-sensitisation to suxamethonium. Skin tests to benzylisoquinolines (atracurium, cisatracurium, mivacurium), propofol, and sufentanil were negative.

A diagnosis of immunoglobulin E (IgE)-mediated rocuronium-induced anaphylaxis was made, based on the timing between rocuronium administration and the onset of symptoms, the clinical presentation (Grade III reaction), histamine and tryptase concentrations, and results of skin testing. Rocuronium and suxamethonium were considered to be contraindicated for future anaesthetics. Her subsequent anaesthetic for the rescheduled procedure, with propofol, sufentanil and atracurium, was uneventful.

Perioperative anaphylaxis is a life-threatening immediate hypersensitivity condition that is usually immunoglobulin E (IgE)-mediated. It occurs rarely, but may lead to morbidity or mortality.

The aims of this article are to: i) review the current nomenclature, ii) discuss the key epidemiological points, iii) highlight the key aspects of the underlying pathophysiology, iv) discuss the acute management and therapeutic approach to cardiovascular collapse refractory to adrenaline (epinephrine) and v) discuss briefly the identification of the culprit drug.

This article focuses on perioperative anaphylaxis (Grades III and IV) and not the less severe clinical presentations (Grades I and II). We do not discuss perioperative anaphylaxis in children, which appears to be rare.¹ Anaphylaxis in obstetric anaesthesia has been detailed elsewhere.

Current nomenclature

Hypersensitivity reactions are subdivided into non-allergic (immunological mechanism excluded) or allergic (immunological mechanism proved or highly suspected) reactions.² In proved drug allergy, the use of the term ‘immediate’ (IgE-mediated immediate hypersensitivity) or ‘delayed’ (T cell specific-mediated reactions, e.g. contact dermatitis, Lyell's syndrome) describes not only the speed of onset of symptoms (1–6 h vs days to weeks, respectively, after exposure to the allergen), but also indicates the likely underlying mechanism.2, 3 Various definitions have been proposed that describe anaphylaxis either based on the clinical presentation alone, or based on both the clinical presentation and the mechanism involved. Allergic anaphylaxis is usually IgE-mediated. Other less common non-IgE-mediated clinical presentations identical to anaphylaxis are described as non-allergic anaphylaxis. The descriptor ‘anaphylactoid’ is no longer used in Europe.

Current known patterns of incidence and causal agents

Most estimates of perioperative anaphylaxis are based on retrospective data, which may account for the variability in the reported incidence, although methodological differences may also be responsible. Two retrospective studies (including Grades I–IV) estimated the rate of allergic and non-allergic perioperative immediate hypersensitivity to be 1:11,000 anaesthetics in Western Australia (2000–9) and the incidence of allergic reactions only, at 1:10,000 anaesthetics in France (1997–2004).4, 5

Recently, the 6th UK National Audit Project (NAP6) reported upon Grades III and IV reactions (n=256) and deaths (n=10) associated with perioperative anaphylaxis collected prospectively over a 1-yr period (2016).⁶ The calculated incidence of perioperative allergic and non-allergic anaphylaxis was 1:11,752.

The mortality rate associated with perioperative anaphylaxis has been estimated to be 4% and 3.8% in France and the UK, respectively, whilst no perioperative anaphylaxis-related deaths were reported in Western Australia in the decade 2000–9.5, 6 The morbidity rate remains unknown.

Perioperative IgE-mediated anaphylaxis is mainly related to agents given i.v. It usually occurs after induction of anaesthesia, with neuromuscular blocking agents (NMBAs) and antibiotics being the main causal agents. The NAP6 study showed that the drugs most commonly involved are antibiotics (48%) followed by NMBAs (25%). Co-amoxiclav and teicoplanin together accounted for 90% of the antibiotics identified. In more than half of the cases, teicoplanin was used because the patient reported an allergy to penicillins, but the allergy claim was rarely substantiated.⁶ Unlike the UK data, the latest French study (2011–2) showed that NMBAs (61%) were the most common drugs, followed by antibiotics (18%); β-lactam antibiotics accounted for 90% of these.

There are few epidemiological data from North America. As there is no US database from which to generate estimates for the incidence of perioperative anaphylaxis and agents involved, it remains unknown whether the risk of NMBA allergy is lower in the USA than in Europe.

Other agents may trigger perioperative IgE-mediated anaphylaxis. These include dyes (patent blue, methylene blue), antiseptics (chlorhexidine, povidone iodine), colloids (gelatins), iodinated contrast agents, and sugammadex. An increase in the incidence of chlorhexidine-induced anaphylaxis has been reported over the past decade in Denmark (9.6%), the UK (9%), and Belgium (10.3%), whilst it is rare in France. Allergy to propofol, opioids, or amide local anaesthetics is very uncommon. Latex allergy is now rarely reported; no case was reported in the NAP6 survey.⁶

Pathophysiology of perioperative immediate hypersensitivity

IgE-mediated anaphylaxis

Anaphylaxis is usually mediated by IgE bound to high-affinity FcεRI receptors located on the plasma membrane of mast cells and basophils. The phase of sensitisation is clinically silent. On re-exposure, the allergen cross links two specific IgE receptors, creating a bridge between two IgEs, which results in a signal transduction cascade and the release of preformed (e.g. histamine, tryptase) and newly-formed (e.g. leukotrienes, thromboxane A₂, cytokines) inflammatory mediators, leading to the onset of clinical features (Fig. 1).⁷

Fig 1.

Pathophysiological mechanisms of perioperative immediate hypersensitivity. COX-1, cyclooxygenase 1.

Alternative mechanisms

Perioperative immediate hypersensitivity may also be triggered by non-allergic mechanisms including either mast cell activation through: i) direct non-specific activation (e.g. atracurium, mivacurium, suxamethonium, propofol), ii) calcium- and phospholipase-dependent mechanism (e.g. vancomycin) or iii) activation of the Mas-related G protein-coupled receptor X2 (MRGPRX2) or a mast cell-independent mechanism such as selective inhibition of cyclooxygenase 1 (e.g. NSAIDs).⁸

Non-allergic immediate hypersensitivity such as histamine release induced by NMBAs (e.g. atracurium, mivacurium, suxamethonium), propofol, or both in young, atopic, stressed, or all three patients is usually mild-to-moderate, while those induced by vancomycin or other antibiotics (e.g. fluoroquinolones), also called ‘red man syndrome’, may mimic anaphylaxis. Activation of the recently identified MRGPRX2 receptor is suspected to be involved in non-allergic reactions induced by NMBAs. Its role remains to be proven in humans. Respiratory symptoms exacerbated by NSAIDs involve only the upper airways or cause severe asthma, angioedema, or both. This pattern is different from that of IgE-mediated drug-allergy (Fig. 1).

Mastocytosis

Mastocytosis is a rare mast cell clonal disorder characterised by an abnormal increase in tissue mast cells. Cutaneous mastocytosis is the most frequent phenotype and is mainly observed during childhood. Systemic mastocytosis primarily concerns adults and does not resolve spontaneously. Various non-specific triggers including histamine-releasing drugs, mechanical (e.g. skin irritation, tourniquet) and physical factors (hypo- or hyperthermia) may elicit non-allergic immediate hypersensitivity. There is no evidence showing a higher prevalence of either non-allergic or IgE-mediated drug-induced immediate hypersensitivity in mastocytosis when compared to the general population.⁸ Perioperative management of mastocytosis has been detailed elsewhere.⁹

Clinical presentation

Perioperative anaphylaxis usually occurs within minutes of induction of anaesthesia. Its presumptive diagnosis is based on clinical signs and their severity, and the timing of the reaction in relation to the drugs administered. The non-intravascular route of administration, such as the injection of patent blue during sentinel node biopsy, typically results in a ‘delayed’ onset of immediate hypersensitivity.

The Ring and Messmer scale

Although the Ring and Messmer scale does not take into account the pathophysiological mechanisms underpinning perioperative immediate hypersensitivity, this four-step (I–IV) grading scale is the tool used most widely to describe clinical phenotypes (Table 1).7, 10, 11 Grades I and II reactions are not life-threatening and are more likely to be non-allergic, although they may still be IgE-mediated. Grades III and IV are life-threatening conditions, also called ‘anaphylaxis’, which are usually IgE-mediated.

Table 1.

Clinical presentations of perioperative immediate hypersensitivity based on the Ring and Messmer four-step (I–IV) grading scale10, 11

Grade	Clinical signs
I	Mucocutaneous signs: generalised erythema, extensive urticaria with or without angioedema
II	Moderate multivisceral signs: mucocutaneous signs, moderate hypotension, tachycardia, or both with or without moderate bronchospasm or gastrointestinal symptoms
III	Life-threatening mono- or multivisceral signs: life-threatening hypotension, tachycardia or bradycardia with or without cardiac arrhythmia, mucocutaneous signs, severe bronchospasm or gastrointestinal symptoms The cutaneous features may be absent before the restoration of haemodynamic stability
IV	Cardiac arrest

Grade III reaction

The main feature of a Grade III reaction is cardiovascular compromise associated with systemic vasodilation and hypovolaemia. A typical pattern involves cardiovascular collapse, tachycardia, and cutaneous features (generalised erythema, extensive urticaria, or both). Swelling of the eyelid, lip/tongue, or both may also be associated. Tachycardia may rapidly evolve into bradycardia or cardiac dysrhythmia (e.g. ventricular arrhythmia). Bradycardia may be the presenting feature. In this setting, a sudden decrease in peripheral resistance related to mediator release, combined with decreased venous return caused by interstitial capillary leakage and subsequent massive hypovolaemia, trigger a ‘paradoxical bradycardia’.¹² Three decades ago, investigation of the prehospital management of haemorrhagic shock showed that paradoxical bradycardia, as a result of the Bezold-Jarisch reflex, is related to extreme and rapid blood loss and helps to preserve diastolic ventricular filling despite profound hypovolaemia.¹³

Erythema and urticaria occur because of cutaneous vasodilation, provided that there is preservation of peripheral perfusion despite the cardiovascular disturbances.¹² These cutaneous features may be absent before the restoration of haemodynamic stability.10, 11, 14, 15 Less commonly, sweating, goosebumps, nipple erection, or extreme pallor may be seen as a result of cutaneous vasoconstriction, and may reflect a more severe phenotype than cases associated with peripheral vasodilation.¹² The most likely cause of this is profound stimulation of the sympathetic nervous system (SNS) in response to the severe cardiovascular disturbances, as the sweat glands of the skin (sweating), the piloerector muscles (goosebumps, nipple erection), and the skin blood vessels (vasoconstriction) are controlled by the SNS.

Associated bronchospasm is more common in patients with uncontrolled airway hyperreactivity (e.g. asthma, chronic obstructive pulmonary disease, obesity, and active smoking).6, 16 Asthma may be more difficult to control in patients who are obese because of airway inflammation, comorbidities such as obstructive sleep apnoea and gastroesophageal reflux disease, mechanical factors impairing adequate ventilation, and other undefined factors.

Cardiovascular collapse may evolve into circulatory arrest if not treated in a timely fashion. The diagnosis of anaphylaxis should therefore be considered in cases of unexplained perioperative hypotension where usual vasopressor therapy has had a limited effect.10, 11, 14, 15

Grade IV reaction

A Grade IV reaction is defined as circulatory arrest. This may be the acute presentation of anaphylaxis, manifesting as pulseless electrical activity (PEA), probably caused by severe hypovolemia.6, 12, 15

Management of perioperative anaphylaxis

The Ring and Messmer scale helps to guide acute management based on the clinical presentation (Table 1). The general and therapeutic measures described below should be implemented immediately.

General measures

i) Withdrawal of the culprit drug if known (e.g. discontinue infusion of antibiotic solution), ii) temporary discontinuation or lightening of the depth of anaesthesia may help resuscitation efforts as the cardiovascular disturbances of anaphylaxis are exacerbated by the depressant effects of general anaesthetics—waking up the patient should be considered, depending on the circumstances, iii) use of the Trendelenburg position or leg raising manoeuvres depending on the clinical setting, and iv) call for help early.

Therapeutic measures

The key therapeutic measures include: i) liberal infusion of i.v. fluids, ii) adrenaline i.v., titrated to effect and iii) support of the airway with 100% oxygen.

Airway management

In the case of anaphylaxis after NMBAs, the airway should be promptly secured, if not already done. When anaphylaxis occurs during regional anaesthesia, such as after antibiotics, the best approach is to provide oxygen by a face mask and manage the reaction as the clinical situation dictates. In the event of loss of consciousness, the airway needs to be secured.

Volume resuscitation

Large volumes of i.v. fluids should be infused immediately upon recognition of anaphylaxis to compensate for the peripheral vasodilation and the severe capillary leakage that occurs from the very onset. Up to 73% of the blood volume may extravasate into the interstitial space within 15 min after the onset of the event.¹⁷ International guidelines all recommend the use of crystalloids as first-line therapy. The British guidelines recommend a high infusion rate and indicate that large volumes may be required.¹⁴ A volume of up to 20 ml kg⁻¹, repeated as needed, or up to 30 ml kg⁻¹ is suggested in Australia and France, respectively.11, 15 In Scandinavia, crystalloids or colloids up to 20 ml kg⁻¹ are recommended, and more if indicated.¹⁰ Colloids have been suggested as a second-line agent in France.¹¹ The recent consensus of the International Suspected Perioperative Allergic Reaction (ISPAR) group recommends that crystalloid boluses should be repeated where the clinical response is inadequate, and further fluid resuscitation should be tailored to the severity of anaphylaxis.

The exact fluid volume requirements during anaphylaxis remain unknown, but continued fluid administration is likely to be warranted until haemodynamic stability is restored. Fluids are preferably administered through a large bore i.v. catheter, fast-running and not on the same side as the BP cuff.

Adrenaline

The priority of treatment is the restoration of MAP above a level which maintains coronary and cerebral perfusion (usually >60 mmHg) and cardiac output, by counteracting both the hypovolaemic and vasodilatory components of anaphylaxis. In addition to therapy with massive volumes of fluids, adrenaline remains the first-line agent to correct the cardiovascular disturbances. There is no contraindication to the use of adrenaline, but dosing must be titrated according to the severity of the reaction, with close monitoring if the patient. I.V. adrenaline can be administered peripherally by either bolus or continuous infusion.¹⁵

International dosing guidelines for adrenaline in Grade III reactions

There is agreement among the Scandinavian, French, Australian, and Spanish guidelines to recommend an i.v. bolus of adrenaline 100–200 μg, repeated every 1–2 min according to the response.10, 11, 15, 18 An infusion of adrenaline may also be started at a rate of 0.05 μg kg⁻¹ min⁻¹ and up to 0.1 or 0.5 μg kg⁻¹ min⁻¹, either when large doses are needed or after three boluses of adrenaline have been administered. British guidelines indicate an initial dose of 50 μg adrenaline, and that several doses may be required followed by an i.v. infusion, the dosage of which is not specified.¹⁴

The ISPAR group recommends 50 μg for the initial i.v. adrenaline bolus dose in Grade III, followed by 100 μg if there is an inadequate response and a further 200 μg bolus, if necessary. Adrenaline infusion (0.05–0.1 μg kg⁻¹ min⁻¹) is suggested after three boluses of adrenaline and an inadequate response.

In summary, dosing of adrenaline is a difficult issue in a Grade III reaction. Taking into account the narrow therapeutic window of adrenaline, 200 μg is likely a large starting dose while 50 μg might be too low a dose, especially for a very severe acute clinical presentation. An initial dose of adrenaline 50 or 100 μg may be selected according to the severity of cardiovascular disturbances.

Notably, the NAP6 panel expert has recommended that cardiopulmonary resuscitation (CPR) should be started if the systolic BP is <50 mmHg even without cardiac arrest, simultaneously with adrenaline and fluids.¹

Clinical pearls

• i)When cutaneous signs are initially lacking, the appearance of erythema or urticaria during resuscitation is most likely to indicate restored cutaneous perfusion, and suggests that the need for adrenaline should be reviewed.¹²
• ii)When cardiovascular collapse is associated with paradoxical bradycardia, atropine must be avoided as it might directly induce circulatory arrest.12, 13
• iii)A Grade IV reaction most commonly presents as PEA requiring CPR.¹⁵ Atropine is no longer recommended in this setting. In PEA arrest, the European and North American standard adult life support guidelines recommend adrenaline 1 mg, every 3–5 min until return of spontaneous circulation.

Other agents

Sugammadex and rocuronium-related anaphylaxis

In the setting of rocuronium-related anaphylaxis, McDonnell and colleagues were the first to suggest that sugammadex should be administered as a therapeutic manoeuvre, as the cyclodextrin molecules encapsulate and bind rocuronium molecules.¹⁹ In their report of a single case, the authors acknowledged that the haemodynamic stability observed after sugammadex may have possibly been coincidental. Instead, it may have been secondary to the adrenaline and fluids therapy that had been instituted, in addition to the reversal of neuromuscular block, which, by increasing muscle tone, could have helped to restore venous return and cardiac output. In a subsequent case series, sugammadex did not modify the course of anaphylaxis; furthermore, sugammadex may independently trigger IgE-mediated allergy.20, 21

Antihistamines and steroids

There is no evidence for the use of H₁-, H₂-, or both antihistamines, steroids, or both in acute anaphylaxis.10, 11, 14, 15, 18 H₁-antihistamines are useful for the symptomatic treatment of urticaria, angioedema, and pruritus in awake patients.¹⁵

Management of associated bronchospasm

Adrenaline, administered to restore cardiovascular stability, is usually effective in relieving the bronchoconstriction through its β₂-agonist effect. In the setting of persistent bronchospasm, short-acting β₂-selective agents (terbutaline or salbutamol) are key drugs and may be administered via a nebuliser (four puffs, repeated every 15–30 min if necessary, and up to 10 puffs) or, if available, with a metered-dose inhaler (5–10 mg h⁻¹) connected to the inspiratory limb of the ventilator circuit.¹⁶ Their onset of action occurs within 5 min, the peak effect is within 60 min, and the duration of action is 4–6 h. I.V. glucocorticoids (e.g. methylprednisolone, 1–2 mg kg⁻¹) remain a key component of management, as their potent anti-inflammatory effects decrease airway inflammation. Their beneficial effects on airway mechanics are delayed and can take 4–6 h in acute bronchospasm.

Refractory anaphylactic hypotension

This clinical entity remains undefined and might be the result of factors including: i) underdosing of adrenaline, or overdosing of adrenaline, which has resulted in receptor desensitisation, ii) delayed administration of adrenaline, iii) inadequate fluid resuscitation, iv) failure to discontinue the culprit drug (e.g. chlorhexidine-coated central venous catheter), and v) failure to lower or discontinue maintenance anaesthetic concentrations.

In cases of refractory hypotension, alternative vasopressors (i.e. noradrenaline, metaraminol, and vasopressin) or other agents, such as glucagon, should be considered, but only after adequate doses of adrenaline and volumes of fluids have been given.10, 11, 15, 18 The recommended rate of noradrenaline infusion varies from 0.05 or 0.1 μg kg⁻¹ min⁻¹ to 0.1 or 0.5 μg kg⁻¹ min⁻¹. Metaraminol is not available everywhere. Arginine vasopressin (AVP) has been suggested as a rescue therapy. In prehospital or perioperative settings, cardiovascular stability was restored after AVP administered 10–40 min after the onset of anaphylaxis, despite a lack of response to catecholamines, phenylephrine, or both.²² Where desensitisation of adrenergic receptors is suspected to be contributing to persistent anaphylactic hypotension, the vasoconstrictive effects of AVP are mediated by non-adrenergic vascular V₁-receptors. The Australian and Spanish guidelines recommend a bolus of vasopressin (1–2 IU; 0.03 IU kg⁻¹) followed by an infusion of 2 IU h⁻¹.15, 18 Doses should not exceed 0.04 IU min⁻¹. The role of AVP in anaphylaxis still needs to be clarified.

Glucagon is usually recommended in patients regularly taking β-blocker medications.1, 10, 11, 15, 18 Its use was reported two to three decades ago, in two uninvestigated cases of iodinated contrast-induced anaphylaxis. In one case, fluid replacement was lacking and in the second, high doses of adrenaline may have contributed to refractory hypotension, leading to the need for an intra-aortic balloon pump.23, 24 As evidence supporting the use of glucagon is of very limited quality, this recommendation needs to be re-evaluated.

Takotsubo syndrome

Takotsubo syndrome (TTS) has been reported after anaphylaxis.²⁵ TTS is characterised by an acute reversible left ventricular systolic dysfunction (hypokinesia, akinesia, or dyskinesia) sharing common features with acute coronary syndrome. Four major variants have been reported. TTS may be triggered by an acute release of endogenous catecholamines (e.g. phaeochromocytoma) or therapy with catecholamines. Adrenaline-induced TTS is characterised by the rapid onset of symptoms, apical or basal patterns of TTS localisation, and a good prognosis. The dose of adrenaline used (>1 mg) is one of the most important risk factors.²⁶ In summary, echocardiogram, ECG, and troponin measurement may be useful in patients who have been given high doses of adrenaline.

Extracorporeal life support

Extracorporeal life support has been successfully used in a few cases of refractory anaphylaxis. The French and Australian guidelines suggest that cardiac bypass or extracorporeal membrane oxygenation may be considered to re-establish adequate perfusion, where it is available.11, 15

Outcome after perioperative anaphylaxis

Proceed with or abandon surgery?

There are no defined criteria regarding the decision to abandon or proceed with surgery after anaphylaxis. This decision has to be evaluated on a case-by-case basis taking into consideration the urgency of surgery, severity of reaction, response to treatment, and underlying comorbidities.1, 11, 15 In a recent study, no difference in major hypersensitivity-related complications was found for cases with Grade III reactions, whether surgery had been abandoned or continued, once resuscitation had been achieved. Surgery was frequently postponed in Grade IV reactions, which were associated with a high rate of complications.²⁷

Biphasic anaphylaxis

The incidence of recurrent clinical manifestations of anaphylaxis after treatment, also called ‘biphasic anaphylaxis’, remains unknown. No recurrence was reported in the NAP6 study.¹

Morbidity and mortality

Overdosing of adrenaline has been associated with pulmonary oedema, ventricular dysrhythmias, myocardial ischaemia, and death while late or absent treatment with adrenaline has been associated with fatal outcomes. This emphasises the need for careful titration of doses.10, 15

Follow up in the anaesthetic allergy clinic

Evaluation after perioperative anaphylaxis is essential. The patient should be informed of the perioperative event and given a referral for follow-up. Written information with detailed information about drug exposures to be avoided until investigations have been performed should be provided. This may be aided by the use of informational materials and brochures designed to assist in either treatment or in initiating investigation and referral according to local guidelines.1, 10, 11, 15, 18 In the UK, it is the responsibility of the anaesthetist to refer the patient to the allergy clinic.¹

This evaluation should be conducted by physicians experienced in the field. Conclusions are based on an assessment of a detailed report of the clinical event provided by the anaesthetist in charge, and review of the anaesthetic chart, tryptase, skin testing, and eventually blood test results.10, 11, 14, 18, 28 The evaluation serves to: i) establish whether anaphylaxis has occurred, or determine an alternative diagnosis, ii) identify the underlying mechanism and culprit drug if anaphylaxis is confirmed, iii) identify safe alternatives to the culprit drug where IgE-mediated anaphylaxis is confirmed, and iv) provide explanations and reassurance to the patient and a detailed medical report, including advice for future anaesthetics.

Histamine and tryptase measurements

Histamine is a preformed mediator produced by basophils and mast cells, which is measured in the plasma (N<10 nmol L⁻¹). The peak is immediate and the elimination half-life is 15–20 min. Histamine concentrations are preferably measured within 30 min after anaphylaxis onset, and up to 1–2 h after, if not done earlier.¹¹

Tryptase is a neutral serine protease contained predominantly within mast cells. Total tryptase concentration is measured in the plasma or serum (N<11–13 μg L⁻¹). It reaches a peak between 15 min and 2 h after anaphylaxis onset; the elimination half-life is around 90–120 min. There is no universally agreed upon timing for tryptase concentration measurements. A time frame within 2 or 6 h after onset of anaphylaxis has been suggested, while some experts recommend that the ideal timing is 1–4 h.10, 11, 14, 18, 28 Acute tryptase concentrations need to be compared with baseline tryptase measured the next day or during follow-up. A level above 1.2×baseline+2 μg L⁻¹ is considered to be indicative of mast cell activation.²⁹

Combined histamine and tryptase measurements are recommended in France and in the USA, while only tryptase measurement is advocated in Scandinavia, the UK, Australia, New Zealand, and Spain.10, 11, 14, 18, 28

Skin testing

Skin testing allows identification of the culprit agent, the pathophysiological (allergic vs non-allergic) mechanism, and safe alternatives (especially in IgE-mediated allergy).10, 11, 14, 18, 28 Life-threatening immediate hypersensitivity associated with increased histamine, tryptase, or both concentrations (compared with baseline) and positive skin tests to the culprit drug prove an IgE-mediated mechanism. The absence of tryptase increase does not preclude the diagnosis. The culprit and cross-reactive drugs identified must be avoided while negative skin-tested drugs can be used subsequently. Drug challenge testing can be useful in cases of negative skin testing (e.g. antibiotics) while some guidelines mention its use in the investigation of perioperative allergy.¹⁸

Mild or moderate immediate hypersensitivity, associated with negative skin tests, and with or without increased histamine, tryptase, or both concentrations suggest a non-allergic mechanism, such as histamine release. Finally, a clinical history associated with tryptase increase, persistently raised baseline tryptase, and negative skin tests suggests mastocytosis. These patients require haematological investigation.

Serum-specific IgE measurement

Serum-specific IgE assays are available for a few drugs, including suxamethonium, antibiotics (e.g. amoxicillin), morphine, chlorhexidine, protamine and latex and as the quaternary ammonium ion (QAI) group of NMBAs.10, 11, 14, 18, 28 These tests are not commercially available everywhere. The sensitivity of latex- or chlorhexidine-IgE assays is excellent, while that of the other assays is low (approximately 30–60%). An increased specific IgE serum concentration after a drug does not necessarily prove that the drug induced the reaction.¹⁴ In addition, IgE concentration to a QAI may be increased in up to 10% of a general population.¹⁸ Therefore, these results must be interpreted in light of the relevant clinical history and skin tests results.

Basophil activation test

The basophil activation test (BAT) is based on the upregulation of two markers (CD63, CD203c), and assesses the degree of blood basophil activation by the suspected drug. Its role in the investigation of immediate hypersensitivity to drugs, including NMBAs, needs to be better defined. BAT is not universally available.

Conclusion

Perioperative anaphylaxis is usually IgE-mediated and is primarily linked to antibiotics and NMBAs. The clinical phenotypes may vary and the diagnosis can be missed when the onset is very rapid. Management is guided by the clinical presentation. The cornerstones of treatment are adrenaline and i.v. fluids. Tryptase concentrations and skin testing are useful to retrospectively establish the aetiology of the reaction; their results must be correlated with the clinical history. A comprehensive assessment provides clarity about the culprit drug and safe alternatives, thus ensuring patient safety during future anaesthetics.

Declaration of Interest

The authors declare that they have no conflicts of interest.

MCQs

The associated MCQs (to support CME/CPD activity) will be accessible at www.bjaed.org/cme/home by subscribers to BJA Education.

Biographies

Pascale Dewachter MD PhD is a consultant anaesthetist and the lead consultant in allergo-anaesthesia in the Groupe Hospitalier Paris-Seine-Saint-Denis. Her research focuses on perioperative anaphylaxis. She has authored many articles on perioperative immediate hypersensitivity and anaphylaxis, allergy to contrast agents and mastocytosis. She is involved in several organisations that provide national and international guidelines.

Louise Savic MCRP FRCA is a consultant anaesthetist with an interest in drug allergy in the Leeds Teaching Hospitals Trust. Her research interest is on perioperative immediate hypersensitivity and anaphylaxis including antibiotic allergy and delabelling.

Both authors belong to the International Suspected Perioperative Allergic Reaction (ISPAR) group.

Matrix codes: 1B01, 2A06, 3I00

Supplementary data

The following is the Supplementary data to this article: