AETIOLOGY

Clinical anaphylaxis in hospital commonly follows injection of drugs, blood products, plasma substitutes, contrast media, or exposure to latex products or chlorhexidine. Outside hospital, ingestion of foods or food additives (especially peanut products) or insect stings may be more common causes than drugs.

Neugut et al.² estimated 1400–1500 deaths per year in the USA and between 3.3 and 40.9 million patients at risk. They estimated radiocontrast media and penicillin to be the greatest cause of death, with food and stings the next groups. In contrast a postmortem study of 56 deaths in the UK³ attributed 19 deaths to venoms, 16 to foods and 19 to drugs and radiocontrast media.

In anaphylaxis, sensitisation occurs following exposure to an allergenic substance, which either alone, or by combination with a protein or hapten, stimulates the synthesis of immunoglobulin E (IgE). Some IgE binds to the surface of mast cells and basophils. Later, re-exposure to antigen produces an antigen–cell surface IgE antibody interaction where two IgE molecules are bridged. This results in mast cell degranulation, and the release of histamine and other mediators, including interleukin, prostaglandins and platelet-activating factor. Histamine is responsible for the early signs and symptoms, but is rapidly cleared from plasma. The overall effects of the mediators are to produce vasodilatation, smooth-muscle contraction, increased glandular secretion and increased capillary permeability. The mediators act both locally and upon distant target organs.

Anaphylactoid reactions may be due to a direct histamine-releasing effect of drugs or other triggers on basophils and mast cells. The symptom complex may also be produced by other mechanisms. Some intravenous drugs and X-ray contrast media may activate the complement system. Plasma protein and human serum albumin reactions may be induced by either albumin aggregates or stabilising agent-modified albumin molecules. Other reactions, including those to dextrans and gelatin preparations, may be activated by non-IgE antibody already present in the plasma or osmotic factors (dextrose, mannitol).

The direct histamine-releasing effects of some drugs may produce reactions due to the effect of histamine alone, and such reactions are related to volume, rate and amount of infusion. Recent work suggests that the site of release of histamine may be important in its clinical effects. Drugs such as morphine and Haemaccel release histamine from skin alone,⁴ and are unlikely to produce symptoms such as bronchospasm, whereas drugs which produce release from lung mast cells (e.g. atracurium, vecuronium and propofol) may be more likely to produce bronchospasm.⁵ Direct histamine release is usually a transient phenomenon, but in some patients severe manifestations may occur, particularly with Haemaccel and vancomycin.

Anaphylactic reactions are usually seen in fit and well patients. It is likely that the adrenal response to stress ‘pretreats’ sick patients, and blocks the release and effects of anaphylactic mediators. The exception to this appears to be patients with asthma, in whom reactions to the additives in steroid and aminophylline preparations may occur, and this may be related to the reduced catecholamine response in asthma.⁶ Patients on β-blockers and with epidural blockade may be more likely to develop adverse responses due to histamine release, and this may also be related to reduced catecholamine responsiveness. Reactions occurring in these groups are more difficult to treat.

CLINICAL PRESENTATION

The latent period between exposure and development of symptoms is variable, but usually occurs within 5 minutes if the provoking agent is given parenterally. Reactions may be transient or protracted (lasting days), and may vary in severity from mild to fatal. Recurrent anaphylaxis is described. Cutaneous, cardiovascular, respiratory or gastrointestinal manifestations may occur singly or in combination.

Cutaneous features include piloerection, erythematous flush, generalised or localised urticaria, angioneurotic oedema, conjunctival injection, pallor and cyanosis. Awake patients may experience an aura, warning of an impending reaction. Cardiovascular system involvement occurs most commonly and may occur as a sole clinical manifestation.⁷ It is characterised by initial bradycardia then sinus tachycardia, hypotension and the development of shock.

In patients reacting due to venom desensitization, bradycardia may be severe and require treatment.⁸ Respiratory manifestations include rhinitis, bronchospasm and laryngeal obstruction. Gastrointestinal symptoms of nausea, vomiting, abdominal cramps and diarrhoea may be present. Other features include apprehension, metallic taste, choking sensation, coughing, paraesthesiae, arthralgia, convulsions, clotting abnormalities and loss of consciousness. Pulmonary oedema is a rare sign. Rarely, some women develop a profuse, watery, vaginal discharge 3–5 days after anaphylaxis. It is self-limiting.

Anaphylaxis is rare in the intensive care unit (ICU), probably because of the protective effects of the adrenal response to stress. However, use of the mast cell tryptase assay (see below) may detect anaphylaxis as an unsuspected cause of shock in intensive care.