Chapter 106. Anaphylactic and Anaphylactoid Reactions

• Mast cells or basophils release mediators of anaphylactic and anaphylactoid reactions as a result of direct drug actions or interaction of antigen with mast cell or basophil immunoglobin E.
• Agents that commonly cause hypersensitivity reactions are antibiotics, blood products, radiocontrast media, muscle relaxants, colloids, preservatives, protamine, and latex.
• Histamine, eosinophilic chemotactic factor of anaphylaxis, slow-reacting substance of anaphylaxis, platelet-activating factor, prostaglandins, and kinins are the six major pharmacologic mediators known to be released in anaphylactic and anaphylactoid reactions.
• Symptoms caused by these mediators are usually immediate but may be delayed by 2 to 15 minutes or, in rare cases, by as long as 2.5 hours after the parenteral injection of antigen.
• Because this reaction includes vasodilation and translocation of fluid from capillaries and postcapillary venules (resulting in loss of fluid and colloid from intravascular spaces), effective plasma volume and systemic vascular resistance is decreased, which can lead to shock.
• The sine qua non of anaphylaxis is severe cardiovascular or respiratory compromise; most conscious patients sense impending doom before the clinical event occurs.
• Tryptase is helpful in distinguishing between anaphylactic and anaphylactoid reactions.
• Although various drugs are used to treat anaphylactic and anaphylactoid reactions, discontinuation of the offending drug, maintenance of the airway, administration of oxygen, blood volume expansion, and administration of titrated doses of epinephrine (very large doses may be needed) are the mainstays of therapy.
• Anaphylactic and anaphylactoid reactions are acute, potentially fatal events, but their morbidity and mortality rates can be decreased by preparation, drills such as those done for cardiopulmonary resuscitation, and prompt recognition and aggressive treatment.

Life-threatening hypersensitivity or pseudoallergic reactions to exogenously derived and administered agents in critical care environments can be, and usually are, treated successfully by a prepared intensivist.1 These reactions are anaphylactic if immune mediated and anaphylactoid if chemically mediated. Data from multicenter studies of perioperative patients show an incidence of life-threatening reactions of 1 in 5000 anesthetics, half of which are immune mediated. Although chemically mediated reactions may be much more common, they usually are less severe.2 Rapid recognition and treatment of such reactions can prevent much of the morbidity and mortality that would otherwise occur.

Anaphylaxis is a life-threatening allergic, immune-mediated reaction. The term allergic applies to immune-mediated reactions, as opposed to those caused by pharmacologic idiosyncrasy, direct toxicity, drug overdose, or drug interactions.3 Anaphylactoid reactions produce the same clinical syndrome but are not immune mediated. Interaction of antigen with mast cell immunoglobin (Ig) E or direct drug actions cause mast cells to release mediators of anaphylactic and anaphylactoid reactions (Fig. 106-1). Vasoactive mediators of these reactions include histamine, eosinophilic chemotactic factor of anaphylaxis (ECF-A), slow-reacting substance of anaphylaxis (SRS-A; a mixture of three leukotrienes, including a potent coronary vasoconstrictor), platelet activating factor (PAF), kinins, and prostaglandins (see Fig. 106-1). Symptoms usually occur within 15 minutes of parenteral injection of the causative agent, although they may be delayed. The net effect of these agents is a marked decrease in systemic vascular resistance and the leakage of fluid from the capillaries and postcapillary venules. This excessive vasodilation and reduction in effective plasma volume can result in shock. Thus, volume replacement and administration of epinephrine are therapeutic mainstays. Common causes of such reactions are antibiotics, blood products, muscle relaxants, colloids, preservatives, latex, protamine, and radiocontrast medium.

Anaphylactic and anaphylactoid reactions to drugs and adjuvants represent an important component of iatrogenic pathology responsible for significant morbidity, mortality, and medical care costs.4,5 However, there is no existing comprehensive epidemiologic study of drug allergy in the general population because many reactions attributed to allergic events are not intensively investigated. Also, many authentic allergic events are not reported. However, several series from different countries have estimated the incidence of anaphylaxis during anesthesia6–17 to be between 1 in 10,000 to 1 in 20,000.9,12 Despite the fact that these reactions are witnessed and occur in a monitored setting, they have a mortality rate ranging from 3.5%18 to 4.7%19 even when appropriately treated.9,12

Skin testing overestimates the incidence of reactions. Although the occurrence of these events is well documented, the prevalence of sensitization to anesthetic drugs in the overall population remains poorly defined. For example, there is a 9.3% prevalence of sensitivity to neuromuscular blocking agents in the general population,20 based on skin test positivity and/or detection of specific IgE to quaternary ammonium ions, although anaphylaxis to neuromuscular blocking agents occurs in only 1 in 6500 anesthetics.9 Similarly, the prevalence of latex sensitization, a major cause of anaphylaxis during anesthesia, depends on the population studied. It ranges from 1% to 6.6% in the general population,21,22 reaching 15.8% in an anesthesiology staff.23 However, this IgE-dependent sensitivity does not necessarily indicate that an allergic reaction will occur in case of exposure to latex.

The end-organ effects of different mediators produce the clinical syndrome of anaphylaxis. In a sensitized individual, the onset of the signs and symptoms caused by these mediators is usually immediate but may be delayed by 2 to 15 minutes or, in rare cases, by as long as 2.5 hours after the parenteral injection of antigen (Fig. 106-2).24 After oral administration, manifestations occur at unpredictable times. True anaphylaxis requires immune-mediated release of some or all of the above vasoactive mediators. In contrast, other mechanisms can liberate the vasoactive substances that produce this clinical syndrome. Patients with anaphylactic or anaphylactoid reactions may manifest some or all of the signs and symptoms described in the following sections.24,25

Respiratory System

Awake patients may complain of nasal stuffiness or itching, difficulty in breathing, or retrosternal tightness. Coughing, wheezing, tachypnea, laryngeal stridor, and cyanosis are other manifestations, as are acute respiratory distress and pharyngeal, epiglottic, or glottic edema. Respiratory symptoms under anesthesia occur but represented 75% of mortality rate in one series.25a

Cardiovascular System

Conscious patients may complain of dizziness. They may also complain of chest tightness, sometimes due to myocardial ischemia. Electrocardiographic changes can vary from tachyarrhythmias and nonspecific changes of the ST segment or T wave, to ventricular fibrillation, electromechanical dissociation, and asystole. In anesthetized patients, hypotension, tachycardia, and cardiovascular collapse are the hallmarks of these reactions.

Skin

Complaints range from itching, warmth, and minor swelling of the face or an extremity to itching of the eyes and nonpitting deep edema in the cutaneous tissue. Patients may manifest the characteristic urticaria or flare of histamine release. However, cutaneous manifestations alone may not be a harbinger of more severe reactions. The myorelaxant atracurium, known to stimulate histamine release, causes a skin rash in about 33% of patients. Antihistamines completely abolish the skin rash. However, the cutaneous manifestation is independent of hemodynamic effects or plasma histamine levels. Thus, the risk-benefit relation of discontinuing the use of a drug that induces a rash should be considered if that drug represents a needed therapy.26 Conversely, even in severe reactions, cutaneous manifestations may not be present in many instances.

Perhaps most impressive in the awake patient is the expression of a sense of doom. Most conscious patients experiencing an anaphylactic or anaphylactoid reaction who were attended by the authors said something like “I feel horrible” or “I'm going to die” before any hemodynamic or pulmonary symptoms or signs were evident.

An anaphylactic or anaphylactoid reaction may involve any combination of these symptoms; however, the sine qua non of anaphylaxis is severe cardiovascular or respiratory compromise. Because this reaction includes vasodilation and translocation of fluid from capillaries and postcapillary venules (resulting in loss of fluid and colloid from intravascular spaces), there is a reduction in effective plasma volume and systemic vascular resistance followed by shock. Thus, fluid resuscitation is a first priority in treating this syndrome.

The most severe life-threatening reactions result from laryngeal edema, bronchospasm, and vascular collapse, which are probably end-organ responses to the release of vasoactive substances. The full variety of vasoactive mediators of anaphylactic and anaphylactoid reactions is not known. However, an understanding of the known pharmacologic mediators responsible for the clinical manifestations provides a rationale for therapy of anaphylaxis.

Histamine

Histamine is a low-molecular-weight amine that is stored predominantly in tissue mast cells and circulating basophils. It dilates capillaries and venules and promotes vascular permeability. Its actions on the histamine H1 and H2 receptors are thought to be responsible for decreased systemic vascular resistance and increased permeability in venules.27 Through H1-mediated effects on vascular smooth muscle, histamine is also a potent coronary vasoconstrictor.28 Vascular responses of vasodilation and local cutaneous reactions (wheal and flare) are mediated by H1 and H2 receptors. H2 receptors mimic β-adrenergic receptors in molecular biology function and role, and mediate chronotropic responses. Histamine has a complex array of effects on the vasculature, depending on the presence or absence of vascular endothelium. It is important to note that histamine is a potent constrictor if the endothelium is damaged.

Eosinophilic Chemotactic Factor of Anaphylaxis

ECF-A, an acidic peptide with a molecular weight of 500 to 600, is stored in mast cell granules. It is chemotactic for eosinophils. The exact role of the eosinophil in the allergic response is unclear. During phagocytosis, eosinophils release histaminase, phospholipase D, and arylsulfatase, enzymes that can inactivate histamine and SRS-A.

Platelet Activating Factor

PAF, acetyl glyceryl ether phosphorylcholine, causes platelet aggregation and bronchoconstriction, increases vascular permeability, and modulates leukocyte function.

Leukotrienes

Leukotrienes are synthesized in response to antigens and are not stored intracellularly. Products of the oxidative metabolism of arachidonic acid through the lipooxygenase pathway, leukotrienes were once known as SRS-A (leukotrienes C4, D4, and E4).29 End-organ effects include potent constriction of bronchial smooth muscle. On a molar basis, leukotrienes are 4000 times more potent than histamine in causing bronchoconstriction in normal human beings. Other systemic effects of leukotrienes include cutaneous inflammation, chemotaxis of polymorphonuclear leukocytes, and promotion of lysosomal enzyme release from leukocytes.

Prostaglandins

Prostaglandins are unsaturated fatty acids that are synthesized at the time of the inflammatory stimulus. Like leukotrienes, they are products of the metabolism of arachidonic acid, but through the cyclooxygenase pathway. Their biologic activities are specific to the target organ on which they act. Prostaglandins are potent mediators of the inflammatory response and can increase capillary permeability and result in bronchospasm, pulmonary hypertension, and peripheral vasodilation.

Kinins

Kinins are low-molecular-weight peptides that enhance capillary permeability, dilate certain blood vessels, contract certain smooth muscles, and are perhaps leukotactic. Bradykinin is a potent stimulus for the release of nitric oxide from vascular endothelium. Several processes independent of IgE can generate kinins.

There are multiple processes by which biologically active mediators can be generated to produce an anaphylactoid reaction.

Histamine, and to a lesser extent other vasoactive mediators described above, can be liberated independently of immune reactions to produce an almost identical clinical syndrome. Mast cells (usually derived from connective tissue) and basophils release histamine in response to chemicals or drugs.30 Activation of the blood coagulation and fibrinolytic systems, the kinin-generating sequence, or the complement cascade can produce the same inflammatory substances as are produced in an anaphylactic reaction (see Fig. 106-1).

When given rapidly in large doses, morphine and meperidine can cause histamine release,31 resulting in an anaphylactoid reaction. Such large doses are rarely administered fast enough to cause histamine release except when given to induce anesthesia. Sedative doses of opioids or slowly administered anesthetic doses usually do not cause the release of appreciable quantities of histamine. Radiographic contrast medium, d-tubocurarine, and vancomycin can also trigger histamine liberation and produce anaphylactoid reactions.32 Interestingly, small variations in molecular structure can markedly alter histamine-releasing potential. Cisatracurium, the R-cis,R'-cis isomer of atracurium, is one of 10 isomers present in the commercial form of atracurium. It is only one-eighth as likely to cause histamine release as is the mixture. However, the actual mechanisms for non–immune-mediated histamine release are not completely understood. Why some patients are prone to histamine release in response to drugs is unknown.

Neuromuscular Blocking Agents

Neuromuscular blocking agents are responsible for up to 70% of anaphylactic reactions during general anesthesia.33 In France in 1996, the incidence of anaphylaxis was estimated at 1 in 6500 anesthetics involving a muscle relaxant9 and occurred whether or not the agents cause chemically mediated reactions. In most series, succinylcholine appears to be most frequently involved, although variations in practice patterns account for some differences. In the most recent French series, rocuronium, a steroidal relaxant devoid of chemically mediated histamine release, was the primary drug responsible for reactions.10 Because the main antigenic determinants in the generation of specific IgE antibodies are substituted ammonium ions,34 an 80% cross-reactivity between the different muscle relaxants is observed by skin tests. Cross-reactivity explains why many patients may exhibit allergic symptoms on first exposure to a neuromuscular blocking agent.35 Details for skin testing of these agents are included in a recent article by Laxenaire et al10 (Fig. 106-3).

Antibiotics

Although allergic reactions to β lactams are the most common cause of anaphylactic reactions in the population at large, they occur far less frequently in the operating room (12% of events). These reactions, classifiable as immediate or nonimmediate, can be produced by the four classes of β lactams (penicillins, cephalosporins, carbapenems, and monobactams), which share a common β-lactam ring structure. Cross-reactivity among cephalosporins and between cephalosporins and penicillins can occur.36

Immediate reactions (type I or mediated by IgE), usually within the first hour after drug administration, are characterized by diffuse erythema, pruritus, urticaria, angioedema, rhinitis, bronchospasm, hyperperistalsis, hypotension, cardiac arrhythmias, and anaphylactic shock alone or in combination. Anaphylaxis occurs in about 0.004% to 0.015% of penicillin37 and 0.001% to 0.1% of cephalosporin36 treatments. The fatality rate from shock after administration of penicillin is 0.0015% to 0.002%. Despite the fact that penicillin-induced anaphylaxis is rare, this drug remains the most common cause of anaphylaxis in human beings, accounting for approximately 75% of anaphylactic deaths in the United States each year.38,39

In many cases, a detailed history of a patient's reaction to penicillin may allow clinicians to exclude true penicillin allergy, thus allowing these patients to receive the drug.40 Skin tests are the quickest and most reliable method for demonstrating the presence of β-lactam–specific IgE antibodies. Using the suspected β lactams themselves (in particular cephalosporins) in addition to penicillin major and minor determinants as test agents is highly desirable. Patients with a compelling history of type I penicillin allergy who require penicillin should undergo skin testing. Those with a negative skin test result can take penicillin without serious sequelae.41 Only 10% to 20% of patients reporting a history of penicillin allergy are truly allergic when skin testing is performed. Conversely, many patients with positive skin tests have only vague penicillin allergy histories.42

Although serum-specific IgE assays can be used as complementary tests, negative test results should be interpreted in light of the time elapsed from the last exposure to the suspected β lactam because in vivo and in vitro test sensitivity decreases over time. In some diagnostic workups, patients with a strong history of type I β-lactam allergy and negative skin and in vitro tests undergo a controlled administration of the suspected β lactam.41

Strategies for dealing with penicillin-allergic patients are evolving and are institution specific. Patients who are allergic to penicillin have been reported to have a 5% to 30% incidence of cross-reactivity to cephalosporins. The recent trend in the United States is to use cephalosporins without a convincing history of penicillin allergy36,41 because alternative antibiotics can cause severe anaphylactoid reactions. For example, chemically mediated red-man syndrome manifested as hypotension and flushing3,43 occurs with rapid vancomycin administration.44

Hypnotics

The incidence of anaphylactoid reactions with thiopental is estimated to be 1 in 30,000.33 Most of the generalized reactions are linked to direct leukocyte histamine release. However, there is evidence for IgE-mediated anaphylactic reactions based on skin tests and specific IgE assay.45

Ever since Cremophor EL has been discontinued as a solvent for some non-barbiturate hypnotics, many previously reported anaphylactoid reactions have disappeared. Although less frequent, anaphylaxis to all induction agents has been observed.46,47 However, in recent studies, induction agents caused only 3.7% of all anaphylactic cases during anesthesia.10

Opioids

Although patients often relate anaphylactic reactions to opiates, immunologically mediated reactions to morphine, codeine phosphate, meperidine, fentanyl and its derivatives are quite rare, occurring only once in 100,000 to 200,000 anesthetics.33 Because of the direct histamine-releasing properties and frequent non–histamine-related cutaneous manifestations, there is significant over-reporting of allergy to opiates. Most opioids (with the possible exception of the fentanyl family) may cause direct release of histamine, urticaria along the vein of administration, and vasodilation. Bronchospasm and angioedema have not been reported, even when large doses of these agents were used for anesthesia before cardiac procedures.48 Codeine has been implicated in several cases of anaphylactoid reactions, but in all of the reported instances, more than one drug was used simultaneously. An anaphylactoid reaction to morphine and an inordinate sensitivity to histamine release have been reported, but skin testing in this patient produced negative results.49

Anaphylactoid reactions during infusion of radiocontrast material occur in 1% to 2% of procedures.50 Although the mechanism of such reactions remains unknown, anaphylactic treatment may diminish them. When repeated infusions of radiocontrast materials were given to patients with a history of an immediate anaphylactoid reaction to them, the incidence of repeat reactions was 17% to 35%. Pretreatment with prednisone, diphenhydramine, and ephedrine significantly decreased repeat reactions in these at-risk patients. Steroid pretreatment should be given at least 12 hours before administration of the contrast agent.

True allergy to the para-aminobenzoic ester agents, such as procaine, is well documented, but documented allergic reactions to the amide local anesthetics are rare.51 Methylparaben, a derivative of para-aminobenzoic acid and a preservative used in many multidose vials of local anesthetics, may well be the offending agent in anaphylactic reactions. With a true allergy to one amide local anesthetic agent, a potential for cross-reactivity exists. Although skin testing is used to identify safe local anesthetic agents, there is no support for applying this technique to humans. Clinical history appears to be a good discriminator. Most “reactions” in the dental chair appear to result from anxiety or epinephrine responses. For example, in a series of 71 patients who gave a history of possible allergic reaction to local anesthetics, only 15% had had history of clinical manifestations of a hypersensitivity response (i.e., urticaria, wheezing, or facial swelling).52

Colloid

Anaphylactoid reactions to the clinically used colloids have been studied extensively in Europe (Table 106-1).53,54 Because of kinin contaminants, the purified protein fractions are associated with a higher incidence of reactions than are other colloids. The complex polysaccharides can activate the complement system. These reactions do not depend on prior sensitization. The reported incidence of anaphylactoid reactions after purified protein fractions varies from 0.007% to 0.085%.

Blood Products

Transfusion reactions can be classified as hemolytic or nonhemolytic. Hemolytic reactions result from ABO-antigen–incompatible transfusions in which IgG or IgM antibodies from the donor or recipient react with red blood cells, complement is fixed and activated, cells are destroyed, and complement anaphylatoxins are released.55 Urticaria, hypotension, and bronchospasm may accompany activation of the clotting cascade. In a sedated and/or paralyzed patient, bleeding or severe hypotension may be the only sign of a transfusion reaction. The most common allergic reactions to transfusions are nonhemolytic, febrile reactions from leukoagglutinins. Unlike red blood cells, leukocytes and platelets contain HLA antigens of the human histocompatibility system. Antibodies of the IgG or IgM class to these antigens are known as leukoagglutinins. The precise relation of leukoagglutinins to anaphylactoid reactions remains unclear. Transfusion-related acute lung injury is produced by leukoagglutinin reactions.3 Patients who lack serum IgA will generate antibodies to IgA and may have a severe anaphylactoid reaction with transfusion of IgA-replete normal blood.56 In most cases of transfusion reaction, the magnitude of cell lysis, hypotension, bronchospasm, and increased capillary permeability probably relates directly to the rate and extent of complement activation.

Protamine

Many case reports and studies have provided evidence that protamine reversal of the effects of heparin can cause a severe, life-threatening reaction with generation of anaphylatoxins (C5a) and thromboxane.57 Multiple mechanisms have been implicated in protamine reactions, including mechanisms that are immune mediated, mechanisms that are mediated otherwise, and mechanisms that combine immune-mediated and vasoactive-mediated reactions when vasoactive substances are displaced by protamine. Because protamine is a histone derived from salmon sperm, antigenic crossover with fish allergy is possible. Diabetics who have received protamine zinc insulin (25 U contains 0.7 mg protamine) appear to have an increased incidence of this reaction and of antibodies to protamine.57,58 More than one mechanism may be involved; production of C5a anaphylatoxins by any mechanism generates thromboxane A2, which causes pulmonary vasoconstriction, bronchoconstriction, and systemic hypotension. Substitutes for heparin and protamine are being developed, so that patients with known risk factors will be able to undergo safe cardiopulmonary bypass. Until then, dose modification, the use of hexadimethrine (Polybrene) for heparin reversal (obtained by an investigational drug request from the Food and Drug Administration) or ancrod for “anticoagulation” remain options.

Latex

Allergy to latex has been increasingly recognized as an important cause of allergy in anesthesia and critical care.59 Although latex allergy was first described in urologic patients in 1983, the role of latex as an allergen in the perioperative period was not described until 1988. Latex allergy currently represents 12% of life-threatening events, according to one recent French study.60 This true allergy should be distinguished from the cutaneous reactions to the products of vulcanization. It is important to note that, in the same study population, latex allergy was largely unreported in a previous 5-year period. Approximately 1% of the general population has the IgE antibody to latex (vs. 8% of surgeons or nurses), but it is unclear whether the increase in latex allergy is related to the more widespread use of latex since the epidemic of acquired immunodeficiency syndrome or to prior lack of recognition of the allergy.61–63 Clearly, some instances of allergy attributed to drugs were latex allergies. The first documented case of fentanyl allergy was reexamined and found to be an allergy to latex.64

Because of the ubiquity of latex in the operating room and intensive care units, allergies to latex are frequently difficult to distinguish from allergies to drugs. Epidemiologic studies report that patients with multiple latex exposures and a history of atopy are at particular risk. In a population of children with meningomyelocele and other midline disorders, antibodies to latex were detected in more than 35%. In the operative setting, this allergy typically occurs some 30 minutes after induction and is characterized by cutaneous, respiratory, and hemodynamic disturbances. This timing distinguishes latex allergy from reaction to induction agents. Diagnosis of this disorder can be made by radioallergosorbent test or enzyme-linked immunosorbent assay, with approximately 90% sensitivity and with specificity ranging from 60% to 90%.

Protocols for the management of patients with known or suspected latex allergy have been developed, but they are frequently difficult to carry out in the critically ill patient. Aside from the use of nonlatex gloves, masks, and catheters, exceptional vigilance must be paid to percutaneous and inhalational sources of latex.65 In addition, cross-reactivity with ethylene oxide, often used to sterilize instruments, has been reported.

There seems to be good evidence that preoperative preparation and physician training can decrease morbidity and mortality rates from anaphylaxis. The French Society of Anesthesiologists has suggested a comprehensive strategy for such preoperative evaluations.33 An aggressive preoperative evaluation to identify patients at risk for latex allergy significantly decreased the frequency of these reactions in France. An allergic history is important, particularly for latex allergy.

The introduction of chymopapain into the armamentarium for treating herniated nucleus pulposus was highly instructive about prophylaxis and treatment of anaphylactic reactions. Before the introduction of IgE screening, a relatively high incidence (0.3% to 2%) of anaphylactic reactions was associated with chymopapain administration. Although now infrequently used, its introduction prompted a new prospective study of anaphylaxis. From the release of chymopapain by the Food and Drug Administration in November 1982 until September 1983, the incidence of anaphylactic reactions was similar to that in the phase II trial (i.e., approximately 0.4% in men and 1.3% in women for the first 30,000 patients treated).66 Then the incidence decreased significantly (to 0.25% in men and 0.7% in women). The mortality rate from chymopapain anaphylaxis also decreased from 2 of 13 to 3 of 252. We believe that improved survival rate is attributable to awareness of the pathophysiology of such reactions, vigilance during administration of the drug, and pretreatment of the patient with H1 and H2 antihistamines. Although prophylactic administration of antihistamines did not significantly decrease the overall rate of anaphylaxis, it significantly decreased mortality rate attributable to anaphylactic reactions. Alternatively, increased physician awareness and treatment of anaphylaxis may be a significant factor in the decrease in mortality rates.

A subject of some interest is whether the patient with an atopic history is at special risk. It now appears that a history of generalized allergy does not predispose to anaphylactoid or anaphylactic reaction in the perioperative period. Thus, limiting histamine-releasing drugs in this population does not seem justified by current evidence. Populations at risk appear to be better categorized by concurrent illness. Older patients or those with myocardial ischemia present a special problem. They are at greater risk of complications than are other patients from pretreatment (especially vigorous hydration and steroid administration) and therapy for anaphylactic reactions. The possibility of triggering anaphylactic or anaphylactoid reactions in this group should be weighed against the clinical advantage of these drugs.

Aside from preoperative assessments, pretreatment with antihistamines remains an option for patients at risk. One might consider giving patients H1- and H2-receptor antagonists for 16 to 24 hours before exposure to a suspected allergen or agent known to cause histamine release. Some experts argue that large doses of steroids (2 g hydrocortisone) should also be administered to women before exposure to agents associated with a high incidence of anaphylactic or anaphylactoid reactions.67

The treatment plan for an anaphylactic or anaphylactoid reaction should be reviewed with other physicians and nurses caring for the patient, and, if anticipated, specific tasks should be assigned in advance (e.g., who starts the second intravenous infusion, who turns the patient supine). The fact that the chance of a serious reaction is slight does not justify lack of planning. If the worst is routinely expected and planned for, the rate of adverse outcomes of anaphylactic and anaphylactoid reactions is likely to decrease.

Although various drugs are used to treat anaphylactic and anaphylactoid reactions, discontinuation of the offending drug, maintenance of the airway, administration of oxygen, blood volume expansion, and administration of epinephrine are the mainstays of therapy (Table 106-2). These procedures are necessary to treat the sudden hypotension and hypoxia that result from vasodilation, increased capillary permeability, and bronchospasm. Establishing a plan beforehand for the successful therapy for these reactions will diminish the risk of unfavorable outcomes. Although a dogmatic treatment protocol is not warranted by the data, data are consistent enough to justify a proposed treatment protocol. Anaphylactic and anaphylactoid reactions are triggered by different mechanisms, but the mediators released and the treatment for these severe reactions are indistinguishable.

Testing for allergic reactions and the diagnosis of allergic disorders begins with a symptom history and physical examination, although many individuals will deny prior exposure. Thus, 20% of patients with allergies to neuromuscular blockers have not received these drugs previously. In addition, a time frame for reaction after the injection of the incriminated drug is very helpful. In chymopapain allergy, 90% of reactions were manifest within 10 minutes on the day of administration. When more than one drug has been injected or absorbed over a short period, a diagnostic strategy based on laboratory tests on samples taken during and shortly after the reaction and on tests carried out days to weeks later may be helpful. Early tests are mainly designed to determine whether or not an allergic mechanism is involved, whereas delayed testing attempts to identify the responsible drug.

Tryptase

Tryptase levels acquired during or immediately after a reaction can be helpful in determining whether the reaction is immunologically mediated. Mast cells activated during an IgE-mediated hypersensitivity reaction release proteases such as tryptase and prestored histamine and newly generated vasoactive mediators. Because tryptase content in basophils is 300- to 700-fold less than in mast cells in humans, tryptase levels in serum are a marker for systemic mast cell activation. Although it can be marginally elevated in different situations,68 an elevated serum tryptase concentration above 25 μg/L strongly suggests an anaphylactic mechanism.69,70

Although a negative tryptase test does not completely rule out anaphylaxis, a 2-year survey of anaphylactoid reaction during anesthesia conducted in France between 1999 and 2000 confirmed its utility in clinical practice. There has been the suggestion that tryptase can distinguish between anaphylactic and anaphylactoid reactions, based on the observation that rapid infusion of vancomycin leads to a 40-fold increase in histamine levels without a change in tryptase.71 The elevated histamine levels persist even when patients are pretreated with oral or intravenous H1/H2 antagonists, although the accompanying hypotension is attenuated, confirming the vancomycin reaction as chemically mediated.72 A contrasting view emerged from a study demonstrating that pharmacologic concentrations of vancomycin can release histamine in isolated mast cell preparations. The most recent epidemiologic evidence demonstrates that an elevated tryptase is a reliable, although imperfect, sign of an immunologic reaction (Fig. 106-4).

The sensitivity of tryptase measurements for the diagnosis of anaphylaxis was estimated at 64%, the specificity at 89.3%, the positive predictive value at 92.6%, and the negative predictive value at 54.3%.73 Peak serum tryptase levels are found between 30 and 60 minutes, and optimum sampling is 1 hour after the start of the reaction. The biological half-life of tryptase is estimated at 2 hours; therefore, increased levels can sometimes be detected for 1 to 6 hours or longer after the onset of anaphylaxis. Increased tryptase levels in sera after death suggest systemic anaphylaxis as a cause of death.74

Skin testing is one of the most frequently used primary confirmatory tests for allergen-specific IgE antibody in the diagnosis of human allergic disease, although its reliability somewhat depends on which class of drug is being tested.

Application of an allergen extract to the skin may be accomplished through a prick puncture or an intradermal injection. Control tests using saline (negative control) and codeine or histamine (positive control) must accompany skin tests to determine whether or not the skin is able to release histamine and react to it. In the prick test, allergen is introduced into the epidermis by means of a puncture. The immediate reaction (wheal and erythema) is read at 15 to 20 minutes. A prick test is considered positive when the diameter of the wheal is at least equal to half of that produced by the positive control and at least 3 mm larger than the negative control. Alternatively, allergen (0.01 to 0.05 mL) can be administered intracutaneously through a 26- to 27-gauge needle. In a skin test titration, the same volume of 10-fold serial dilutions of an allergen extract is injected. The objective of skin test titration is to determine the concentration of an extract that produces a wheal or erythema of a defined diameter (e.g., diameter of the wheal is at least twice that of the injection wheal).

Several factors may contribute to the variability of skin test results, including the individual's biologic responsiveness, the technique used to perform the test, the reagents (stability, vehicle, allergen concentration, and purity), and the method used to evaluate skin reactions. Despite the problems of possible false-positive and false-negative results, prick tests in case of suspected latex allergy and prick and intradermal tests in case of drug allergy are advised.33,75 Recent guidelines concerning the maximal concentration to be used during skin testing for the diagnosis of anaphylaxis during anesthesia have been reported.10 Although some controversy on the validity of skin test persists, the development of such standardized diagnostic protocols should facilitate interpretation.

Allergen-specific IgE antibody is also an important confirmatory test in the diagnosis of type 1 hypersensitivity reactions. The radioallergosorbent test was the first assay reported for the detection of allergen-specific IgE antibodies. This assay is a noncompetitive solid phase immunoradiometric assay. The allergosorbent assay is prepared by covalently coupling allergen onto cyanogen bromide-activated cellulose disks that are then incubated with the patient's serum. The amount of bound immunospecific IgE is determined by incubation with radioiodinated antihuman IgE Fc. Bound radioactivity is proportional to the amount of allergen-specific IgE in the initial patient's serum and is compared with an IgE antibody reference serum. This early allergen-specific IgE antibody assay was followed by second-generation assays with new matrix materials such as the cellulose sponge used in the Pharmacia CAP system, various polyclonal and monoclonal anti-IgE detection antibody combinations, and non-isotopic labels. Automation has improved assay precision and reproducibility.

Despite promising results as in the case of sensitization to neuromuscular blocking agents,76–78 when maximal clinical sensitivity is needed, as in the case of drug allergies, intradermal skin tests are preferred. IgE antibody serologic results are viewed as complementary to skin tests.76

In rare instances, in vivo provocation tests may be performed to confirm the identity of an allergen that may be clinically necessary. These tests place the patient at some risk for a reaction because they involve a direct allergen challenge. Therefore, their indication will ultimately depend on whether one can identify a substitute therapy for the putative allergen.

If a patient must receive a drug that is likely to trigger an anaphylactic or anaphylactoid reaction, several steps should be taken. First, an allergist should be consulted. Second, repeated injections of small amounts of antigen or offending hapten may be given because this procedure decreases the amount of specific IgE antibody produced and may stimulate T-suppressor cells. At the same time, this procedure usually increases the amount of competitive IgG antibody (i.e., blocking antibody). Blocking antibody can compete with IgE for receptors on the surface of basophils and mast cells to decrease the mediators released from these cells. Third, small amounts of drug are injected, for example, 5 mg of lidocaine intravenously. This “test dose” hypothesis is verified by the chymopapain experience. Fourth, these tests are done in the post-anesthesia care unit or other setting where resuscitation equipment is available.

1. Discontinue suspected allergen. This step prevents further recruitment of mast cells and release of mediators.

2. Maintain airway and administer 100% oxygen. Bronchospasm, pulmonary hypertension, and pulmonary capillary leakage may cause severe mismatching of ventilation and perfusion.79 These changes can persist for several hours during anaphylactic reactions, producing hypoxemia. Therefore, the airway should be maintained, and supplemental oxygen (we recommend 100% oxygen) should be administered until the situation improves. Intubation of the trachea should be considered, if it has not already been done.

3. Discontinue all sedative, vasodepressing agents. Agents that may have negative inotropic properties or that may decrease systemic vascular resistance should be discontinued because they may contribute to decreased vascular compensation and often interfere with the body's compensatory response to cardiovascular problems. Thus, we believe that these agents should be discontinued to avoid hypotension.

4. Expand blood volume. Because up to 40% of intravascular volume is rapidly lost into the interstitial spaces in an acute anaphylactic reaction during anesthesia,80 effective therapy consists of rapid replacement of blood volume.81 Rapid fluid loss and successful treatment with fluid replacement have been documented in the case of an anaphylactic reaction observed incidentally by transesophageal echocardiography.82 Evidence does not indicate that colloid is more effective than crystalloid. Rapid administration of 1 to 2 L of lactated Ringer's solution or normal saline is important in the initial therapy for these reactions. Further blood volume expansion may be necessary, if hypotension persists.

5. Administer epinephrine. Epinephrine is a mainstay of therapy in acute anaphylaxis for several reasons. Its α1-adrenergic effects make it useful for treating hypotension. Its β2-adrenergic effects inhibit bronchoconstriction and the release of mediators from stimulated mast cells or basophils by stimulating the production of intracellular cyclic adenosine monophosphate.83 For hypotension, 0.05 to 0.1 mg of epinephrine (0.5 to 1 mL of a 1:10,000 solution) should be given initially as an intravenous bolus and repeated every 1 to 5 minutes up to 0.015 mg/kg (or 10 mL of a 1:10,000 solution) for a 70-kg patient, as needed. In cardiac arrest or with a total loss of blood pressure or pulse, full cardiopulmonary resuscitative doses of epinephrine are indicated (0.01 mg/kg, not to exceed 1.0 mg total) in addition to rapid volume expansion. Larger, even massively larger, doses have been needed when titrated to effect, but we do not recommend them as initial doses, because hypertension and tachycardia resulting in myocardial cell death have been reported in such situations. Intramuscular or subcutaneous administration of epinephrine may be unreliable in a hypotensive patient who requires immediate therapy. Interactions with propranolol are possible, and care must be taken not to overdose the patient with epinephrine to the point of severe hypertension (unopposed α-adrenergic effects) and its consequence. Overdose can occur in the absence of propranolol. The major threat from chronic β-adrenergic blockade appears to be that responses to epinephrine may be insufficient to terminate the reaction, and increased doses of epinephrine may be required.84 We believe that any amount of epinephrine is appropriate when titrated to effect; the doses indicated above provide a starting point.

If the above treatment does not resolve symptoms and signs, additional treatment with antihistamines, aminophylline, catecholamines, corticosteroids, calcium, and endotracheal intubation may be indicated.

Antihistamines

Histamine is one of the major mediators of the acute manifestations of anaphylactic and anaphylactoid reactions. Because the vasodilatory effects are mediated by H1 and H2 receptors, both types of receptor site must be blocked if all potentially harmful cardiovascular effects of histamine are to be antagonized. Studies in which patients were pretreated before histamine release demonstrate that adverse cardiopulmonary responses can be blunted or prevented when H1- and H2-receptor antagonists are used.85 Although histamine is only one of the mediators released in anaphylactic and anaphylactoid reactions, it may account for many of the initial adverse manifestations. No clinical evidence indicates that administration of antihistamines is effective in treating anaphylaxis once mediators have been released. Administration of antihistamines is therefore recommended only as secondary treatment in acute reactions. (Suggested doses are 1 mg/kg diphenhydramine or 0.1 mg/kg chlorpheniramine as the H1 blocker and 4 mg/kg cimetidine as the H2 blocker.)

Catecholamines

Inhaled β2 Agonist

If bronchospasm persists (and hemodynamic function is stable), terbutaline or albuterol may be administered as a bronchodilator.

Intravenous Isoproterenol

If bronchospasm persists, isoproterenol may be useful as a pure β-adrenergic agonist and bronchodilator. This drug may also prove useful when combined with epinephrine in the patient taking β-adrenergic receptor blocking drugs. The β2 adrenergic effects of isoproterenol cause vasodilation and possible hypotension, especially in patients already experiencing vasodilation or depletion of blood volume. Tachycardia is another possible unwanted effect. Because isoproterenol dilates the pulmonary artery, it may be useful for treating the increased pulmonary vascular resistance of severe anaphylactic reactions when oxygenation is a problem or right ventricular dysfunction occurs (but see Chap. 26 for cautions in this regard). (Starting doses for persistent bronchospasm are 0.5 to 1 μg/min per 70 kg.)

Intravenous Epinephrine

When hypotension and bronchospasm persist, an intravenous epinephrine drip may be given after blood volume has been expanded and epinephrine boluses have been administered. The starting dose of epinephrine is 1 to 2 μg/min per 70 kg and should be titrated to the desired effect.

Intravenous Norepinephrine (Levophed)

Blood pressure can be maintained in a hypotensive patient with norepinephrine until adequate volume expansion has been achieved. Although α-adrenergic drugs may theoretically potentiate the release of mediators, hypotension is deleterious to cerebral and coronary perfusions and must be treated aggressively.

Steroids

Corticosteroids should be given in severe reactions, such as shock with refractory bronchospasm and hypotension; there is no evidence to suggest an appropriate dose or preparation.68,86 We believe 2 g of hydrocortisone phosphate (or its equivalent) is appropriate for severe cardiopulmonary dysfunction. Large doses of methylprednisolone (35 mg/kg) have been shown to inhibit complement-induced polymorphonuclear cell aggregation and lysosomal enzyme release in vitro. Corticosteroids may decrease release of vasoactive mediators and metabolites of arachidonic acid in anaphylaxis by stabilizing membrane phospholipids or by generating macrocortin, which inhibits phospholipid turnover.

Airway Evaluation

If laryngeal edema occurs in anaphylactic reactions, swelling can persist.87 In such instances, the larynx can be examined on extubation.

Calcium

Because release of mediators is a calcium-dependent process, theory and experimental evidence indicate that administration of calcium may worsen an anaphylactic condition.88

Further study is necessary to confirm the effects of trauma and sedative-hypnotics on immunity and the mechanism of this alteration. For now, we know that the perioperative period interferes with innate immunity (e.g., by mucous and skin membrane transgressions), but the effect of alteration of acquired immunity is not clear.

An integral part of therapy is to communicate effectively to patients when an agent is responsible for a true allergic event, and to identify the responsible agent. Although anaphylactic reactions in the perioperative period remain uncommon, there has been a significant tendency to attribute perioperative hemodynamic and respiratory compromises to an anaphylactic origin without appropriate testing. In our preoperative clinics, there is a 200-fold over-reporting by patients of anaphylaxis to opiates or local anesthetics. It is important, medicolegally and for patient safety reasons, that patients be fully aware which agents are implicated.