Published: August 2010
Adverse reactions to drugs are unintended or undesired effects of a drug therapy that may significantly influence management decisions.1 The incidence may be as high as 15% in hospitalized patients.2,3 Budnitz and colleagues, in a study of emergency department visits for outpatient adverse drug reactions, reported that such reactions accounted for 2.5% for all unintentional injury and 0.6% of the estimated visits for all causes.4
Predictable adverse drug reactions are common. These reactions are dose dependent or related to the pharmacology of the drug and include overdose, side effects, secondary or indirect effects, secondary effects related to underlying disease, and drug-drug interactions.1,5 Unpredictable drug reactions are less common and occur in a small subset of patients. These reactions are not related to the dose or the pharmacology of the drug. Unpredictable reactions include drug intolerance, idiosyncratic reactions, pseudoallergic reactions, and immunologic reactions.1,5
Based on the mechanism involved, immunologic reactions can be classified as immunoglobulin E (IgE)-mediated (type 1) and non–IgE-mediated reactions. IgE-mediated reactions usually develop within minutes following the administration of the drug but can occur up to 72 hours later. These reactions include but are not limited to anaphylaxis, urticaria, asthma, angioedema, and hypotension. Non–IgE-mediated reactions can be further classified into antibody-mediated (type 2), immune complex-mediated (type 3), and T-lymphocyte-mediated (type 4) reactions. Non–IgE-mediated reactions include erythema multiforme, serum sickness, hemolytic anemia, drug fever, Stevens-Johnson syndrome, thrombocytopenia, and toxic epidermal necrolysis.6
This chapter summarizes current understanding of drug reactions, with emphasis on IgE-mediated reactions to penicillins and cephalosporins, adverse reactions to local anesthetics, and angioedema due to angiotensin-converting enzyme (ACE) inhibitors and angiotensin receptor blockers (ARBs).
Penicillin is commonly prescribed because of its effectiveness and low toxicity.7 However, adverse reactions to penicillin are common and often complicate medical therapy. Up to 20% of patients admitted to the hospital claim to have penicillin allergy.1 These patients are usually treated with more expensive, more toxic, and sometimes less effective antibiotics. Although the exact prevalence of allergic reactions to penicillin is unknown, allergic reactions are estimated to occur in approximately 2% of patients treated with penicillin.1 Most of these reactions are skin rashes, such as maculopapular or urticarial eruptions.1 However, the penicillins are one of the most common causes of drug-induced anaphylactic reactions,8 and fatalities have been reported.9
Penicillin is a chemical hapten, with a low molecular weight of 300 daltons, that needs to bind to a tissue macromolecule, usually a protein, to become immunogenic.7 The major breakdown product of penicillin is the penicilloyl group (approximately 85%-90%), known as the major determinant.7 The reminder of the breakdown products are minor determinants, so called because they are formed in smaller quantities.7 The minor determinants include penicilloate, penicilloylamine, penilloate, and other simple chemical products of penicillin.7 Immediate reactions following penicillin administration are mediated through IgE antibodies against the major determinants, the minor determinants, or both.
The risk of penicillin sensitization is increased with multiple short courses of antibiotics and can occur with any route of administration.7 Anaphylactic reactions to penicillin occur most commonly in persons between 20 and 49 years old. However, they can occur in children and in the elderly. Race, gender, personal or family history of atopic disease, and allergy to other drugs or to the mold Penicillium are not predisposing factors.7
The presence of IgE antibodies to penicillin can be detected through a skin test to penicillin, a radioallergosorbent test (RAST) to penicillin, or the enzyme-linked immunosorbent assay (ELISA). The skin test for penicillin is the most reliable way to demonstrate the presence or absence of specific IgE antibodies to major and minor penicillin determinants. However, it does not predict the future development of IgE-mediated reactions during subsequent courses of penicillin or the development of non-IgE-mediated reactions caused by other immune mechanisms, such as cytotoxic antibody-mediated reactions, antibody-antigen immune complex-mediated reactions, and delayed-type cell-mediated reactions.10
Penicillin skin testing is performed using benzylpenicilloyl polylysine (Pre-Pen), penicillin G, and minor determinants (if available) to detect the presence of IgE antibodies to the major and minor determinants. Because the minor determinant mixture is not commercially available, penicillin G at a concentration of 10,000 U/mL has been recommended as a partial source of minor determinants. Methods of preparation of the minor determinants have been published elsewhere.11,12 Unfortunately, Pre-Pen, the source of major determinants, has been withdrawn from the market and is not available.
Both percutaneous and intradermal tests are performed using diluted penicillin G at a concentration of 10,000 U/mL, Pre-Pen at full strength, and minor determinant mixture (if available). Histamine and saline skin tests are used as positive and negative controls, respectively. Percutaneous testing is done, and the results are read at 15 minutes. If the percutaneous test findings are negative, intradermal testing is performed. The skin test is positive if it produces a wheal more than 3 mm larger than the wheal produced by the negative saline control (Fig. 1).
Up to 99% of patients tolerate penicillin if skin testing is negative for penicillin using major determinants (benzylpenicilloyl) and a mixture of minor determinants and penicillin G. Approximately 97% of patients tolerate penicillin if skin testing is negative using benzylpenicilloyl and penicillin G (as the sole source of minor determinants).13 However, a few patients who are at risk for anaphylactic reaction are missed with this testing method because penicillin G does not contain all the minor determinants.13 Patients with a history of penicillin allergy and negative penicillin skin test have a low risk of developing an IgE-mediated reaction following administration of penicillin. At our institution, we found a reaction rate of 1.7% in a group of 596 patients with a history of penicillin allergy and negative skin tests.14 Similar to the findings in our study, low rates of adverse reactions also have been reported by other clinicians.15-17
The RAST and the ELISA detect only IgE antibodies to the major penicillin determinant. Therefore, these tests are less sensitive than the skin test.13 If a patient has a positive history and a positive skin test to penicillin, there is a 50% or greater chance of an immediate IgE-mediated reaction if penicillin is received again.13 Similarly, a positive RAST or ELISA demonstrates the presence of IgE antibodies to penicillin, and these patients should be considered at risk for immediate IgE-mediated reactions to penicillin.15 These patients should receive an equally efficacious alternative antibiotic or be desensitized.
The detection of IgE antibodies to penicillin by a skin test is affected by the amount of time between the original allergic drug reaction and the skin test. Many patients with documented IgE antibodies to penicillin by skin test lose the sensitivity with time. It is estimated that up to 80% of patients with a history of immediate reactions to penicillin will have a negative skin test at 10 years.18 However, these patients may be at increased risk of sensitization to penicillin on subsequent administration compared with the rest of the population.17
If done properly, the skin test is safe, with a rate of systemic reaction of less than 1%.14,19 Nevertheless, severe reactions such as anaphylaxis and death have occurred. Serious reactions to the penicillin skin test are usually a result of violations of the skin test protocol, such as administering a dose that is too high or performing intracutaneous testing without prick or puncture testing beforehand.13
Penicillin skin testing can also serve as a way to decrease the necessary dose of broad-spectrum antibiotics such as vancomycin and fluoroquinolones.20,21 These antibiotics, when used extensively, are associated with the emergence of multidrug-resistant pathogens and increased morbidity and mortality.
Most patients labeled allergic to penicillin do not have penicillin-specific IgE antibodies as detected by skin test and can safely be given penicillin. However, patients with a history of penicillin allergy are more likely to experience a reaction on subsequent courses than those without such history.10
The evaluation of adverse drug reactions begins with a detailed history. However, many patients do not clearly recall the drug to which they reacted, the type of reaction that occurred, or the duration of drug exposure. In addition, up to 33% of patients with a vague history of penicillin allergy have a positive penicillin skin test.18 Therefore, it is recommended that even patients who have a vague history of penicillin allergy and who require penicillin have skin testing to determine the presence of penicillin-specific IgE antibodies before it is assumed that the patient will tolerate penicillin.13 Some clinicians suggest that the penicillin skin test needs to be repeated before each course of penicillin, especially for patients with a history of IgE-mediated reaction who received intravenous penicillin, because of the risk of resensitization.1 However, Solensky and colleagues have reported that none of 46 patients with a history of penicillin allergy and a negative penicillin skin test were resensitized after receiving three courses of oral penicillin V.22
Because of the lack of commercially available penicillin minor determinants for skin testing, some authorities recommend a test-dose challenge in patients with a history of penicillin allergy and negative skin tests to major determinants and penicillin G.6 A test-dose challenge might be done using 0.01 of the therapeutic dose (0.001 of the therapeutic dose if the previous reaction was severe), followed by 0.1 of the dose, and then the full therapeutic dose if there is no reaction.6 If a reaction occurs during the test-dose challenge and the drug is essential for treatment, penicillin desensitization is recommended.6
If the penicillin skin test is positive and the treatment with a penicillin antibiotic is essential, desensitization is needed. Oral and intravenous protocols for penicillin desensitization have been published.5,6,8,23 Oral desensitization is safer than parenteral desensitization.23
The basic principle for oral or parenteral (intravenous) desensitization is similar. The initial dose is very small, usually 0.0001 of the recommended dose. The dose is usually doubled every 15 minutes until the full therapeutic dose is achieved.
Adverse reactions can occur during and after the desensitization procedure.22 Most of these reactions are mild, such as pruritus, rhinitis, wheezing, and urticaria.23 These reactions require symptomatic treatment, and the dose of penicillin should be repeated until tolerated. Severe reactions, such as laryngeal edema, require rapid treatment until the patient is stable and a reduction of the next penicillin dose by one third or more of the previous provoking dose.6 When desensitization is achieved, continuous treatment with penicillin is required to avoid the return of the IgE-sensitive state. A time lapse greater than 12 to 48 hours can allow such sensitivity to return.6
Like penicillins, cephalosporins are commonly used. Cephalosporins share a common beta-lactam ring with the penicillin (Fig. 2). However, they have a dihydrothiazide ring instead of the thiazolide ring of the penicillin molecule. Various adverse reactions to cephalosporin have been described including eosinophilia, serum sickness, febrile reactions, interstitial nephritis, and hemolytic anemia.24 Urticaria, rash, exanthema, and pruritus are estimated to occur in approximately 1% to 2.8%.24 Although anaphylactic reactions after cephalosporin administration are rare,25 death has been reported.26,27
Unlike penicillin, cephalosporin determinants have not been well identified or studied. Therefore, the positive and negative predictive values of cephalosporin skin testing are unknown. A negative skin test does not rule out the presence of IgE antibodies to these drugs. On the other hand, a positive skin test suggests the presence of drug-specific IgE antibodies. These patients should be considered at increased risk for IgE-mediated reaction on administration of the specific drug. They should receive an alternative antibiotic or undergo desensitization. Nonirritant concentrations of commonly used antimicrobial drugs that can be used for skin testing have been published.8,28
If the patient needs a cephalosporin and has a history of cephalosporin allergy, the treating physician can choose one of the following approaches: choose a non–beta-lactam antibiotic, give a graded dose challenge with a cephalosporin with a different side chain from the one to which the patient reacted, or perform a cephalosporin skin test with a nonirritating concentration.24 If the skin test is positive, the patient can be desensitized or tested for another cephalosporin with a different structure.1 Protocols for cephalosporin desensitization are published elsewhere.24 Because the negative predictive value of a cephalosporin skin test is unknown, it is preferable that a negative skin test be followed by a provocative graded-dose challenge, if a challenge is favorable from a risk-to-benefit standpoint.1
The rate of cross-reactivity between penicillins and cephalosporins is unknown. Clinical cross-reactivity between penicillins and cephalosporins appears to be low, especially for second- and third- generation cephalosporins.10 However, patients with a history of penicillin allergy appear to be at increased risk for severe IgE- mediated reactions to cephalosporin, including anaphylaxis.
A penicillin skin test can help to determine which patients are at increased risk for severe IgE-mediated reactions when receiving a cephalosporin. A review of 11 studies concluded that the risk for reaction when receiving cephalosporin in a patient with a positive skin test to penicillin was approximately 4.4%.25 Pumphrey and David reported that 3 of 12 anaphylactic deaths following administration of cephalosporin occurred in patients known to be allergic to penicillin.29
Some patients who react to beta-lactam antibiotics other than penicillin have antibodies directed to side-chain structures rather than to the beta-lactam ring, and cross-reactivity among cephalosporins could be explained by the presence of these antibodies.1 Such antibodies can cause anaphylaxis.13 Figure 3 shows an extensive list of cephalosporins and their structures, including similarities and differences of side-chain structures (indicated by R1 and R2).30 Cefamandole, cefalothin, cepaloridine, cephalotin, and penicillin G have similar side chains. Similar side chains are present in cefadroxil and amoxicillin, ceftazidime and aztreonam, and cephalexin and amoxicillin.24 It can be appreciated that reactions to cephalosporin can occur even with a negative penicillin skin test.
If a patient has a history of penicillin allergy and needs a cephalosporin antibiotic, the following approaches have been recommended10: First, use a non–beta-lactam antibiotic. Next, consider administration of a cephalosporin (preferably second or third generation) if the patient has a history of mild reaction. Although the risk of life-threatening reaction is low, severe adverse reactions can occur. This approach has been discouraged in published guidelines on the diagnosis and management of drug hypersensitivity.10 Finally, perform a penicillin skin test. If the skin test is negative, the patient may receive the cephalosporin with a low risk of developing an allergic reaction.
If a patient in need of a penicillin antibiotic has a history of a cephalosporin allergy, the penicillin skin test can identify those at risk for IgE-mediated reaction to penicillin. If the skin test is negative, penicillin may be given. If the skin test is positive and penicillin is needed, the patient should undergo desensitization.
Carbapenems (e.g., imipenem) should be considered cross- reactive with penicillin and cephalosporin because of the presence of a similar beta-lactam ring.10,13 On the other hand, aztreonam (a monobactam) rarely cross-reacts with penicillin, possibly because of the lack of a second nuclear ring structure.10,13 However, ceftazidime and aztreonam share the same side chains, and clinical cross-reactivity can occur.1
The presence of IgE antibodies to penicillin can be determined by the use of the penicillin skin test. This test can help us to identify patients who have a history of penicillin allergy and can receive penicillin or cephalosporin with a low risk for immediate hypersensitivity reactions, and it can identify patients who should avoid using these drugs. In spite of the importance of penicillin skin testing, Pre-Pen was withdrawn from the market because of manufacturing problems. Without this preparation, a reliable diagnostic test for managing drug hypersensitivity has been lost.
Local anesthetics have been used worldwide in different medical procedures since procaine (Novocaine), the first synthetic local anesthetic agent, was developed.
Local anesthetics reversibly block the generation and conduction of nerve impulses by decreasing the permeability of excitable membranes to sodium through interaction with one or more specific binding sites within sodium channels.31 The duration of anesthesia depends on plasma protein binding and the addition of vasoconstrictors such as epinephrine. Vasoconstrictors decrease the rate of absorption of local anesthetics, localizing the anesthetic to the desired site and allowing its rate of destruction to keep pace with its rate of absorption into the circulation.31
The structure of local anesthetics has three segments: a lipophilic or aromatic group, an intermediate chain linkage, and a hydrophilic or amine group. Based on their intermediate chain linkage, local anesthetics are classified into two major groups: the benzoic acid esters or group 1 and the amides and miscellaneous or group 2. Group 1 anesthetics include benzocaine, butamben picrate, cocaine, procaine, tetracaine, proparacaine. Group 2 anesthetics include bupivacaine, dibucaine, dyclonine, etiodocaine, levobupivacaine, lidocaine, mepivacaine, prilocaine).31,32
Different types of adverse reactions to local anesthetics have been reported. Most of these appear to be vasovagal, anxiety, and toxic reactions.33 Case reports have suggested that adverse reactions to the preservatives in local anesthetics, including methylparaben and sulfites, can also occur.34,35
Following absorption, local anesthetics may be associated with central stimulation manifested by restlessness, tremor, and seizures. Central stimulation is followed by depression, and death is due to respiratory failure.33 Palpitations, tachycardia, diaphoresis, and ventricular arrhythmias have also been described.31,32,36
Hypersensitivity reactions to local anesthetics are uncommon. Contact dermatitis mediated by sensitized lymphocytes is the most common, and it accounts for up to 80% of reactions reported to local anesthetics.37 Patch testing is available for patients with a history of contact dermatitis. Although IgE-mediated reactions to local anesthetics are extremely rare, cases have been reported.37 de Shazo and Nelson evaluated 90 patients with a history of adverse reaction to local anesthetics, and only one patient had a positive test by the intradermal method; the other 89 had a negative challenge.38 Gall and colleagues reported negative skin-test results in 177 patients evaluated for adverse reactions to local anesthetics. Three patients reacted after a provocative dose challenge with the causative drug, two patients had immediate-type reactions to articaine and lidocaine, and one patient had a delayed reaction to mepivacaine.39 In a large series of 236 patients reported by Berkun and colleagues, all subjects had negative skin test results, and only one patient had a positive provocative dose challenge.40
Patients with a history of adverse reactions to local anesthetics are usually advised to avoid them in the future. When surgery is indicated, they must decide whether to undergo surgery without anesthesia, have a procedure with general anesthesia, or forgo benefits of the surgical procedure.
The role of the allergist or immunologist is to rule out IgE-mediated (allergic and anaphylactic) potential to group 1 and group 2 agents. Different protocols for evaluating patients with a history of adverse reactions to local anesthetics have been described.6,32,40,41 Based on the information obtained from patch-test studies, the benzoic acid esters usually cross-react with each other but not with the amides and the other agents in group 2. The amides and the other agents in group 2 do not appear to cross-react with each other.6
One approach to managing these patients begins with identifying the agent that caused the previous untoward reaction. If the drug is a benzoic acid ester (group 1), an amide or a drug from group 2 may be used. If the drug is an amide, another amide may be used and no further evaluation is needed. If the drug is unknown, skin-prick testing with the undiluted local anesthetic to be used can be performed. A negative test can be followed by intradermal testing with 0.1 mL of a 1 : 100 dilution of the same agent. If both tests are negative, the next step is a provocative dose challenge with 1 mL subcutaneous injection of the undiluted agent.
Positive and negative controls should be used with all prick and intradermal testing. Local anesthetics used for skin testing should not contain epinephrine, because false-negative results can occur.41 Skin testing may be performed with preservative-free local anesthetics or with local anesthetics containing preservatives. However, if the result is positive with a local anesthetic containing preservative , skin testing should be repeated with a preservative-free agent. Another positive skin test might indicate the presence of IgE antibodies, and a different local anesthetic should be considered.
True IgE-mediated reactions to local anesthetics are extremely rare. Allergy and immunology evaluation is useful for patients with suspected local anesthetic allergy, because it can identify local anesthetic agents that can be tolerated without elevated risk of IgE- mediated reaction.
Adverse drug reactions to ACE inhibitors are common and include cough in up to 35% of patients and angioedema in approximately 0.1% to 2.2%.42-45 Angioedema has also been reported in association with angiotensin II receptor blockers (ARBs).
Cough associated with ACE inhibitors often develops within the first 2 weeks of starting an ACE inhibitor and generally is a dry, nonproductive cough. Cough may be accompanied by bronchospasm, although this does not occur more often in asthmatics than in nonasthmatics.46 Cough is more common in women and nonsmokers.47
The pathogenesis of ACE inhibitor–associated cough is not completely understood. The mechanism might involve increased levels of prostaglandins, kinins (including bradykinin), and substance P. Bradykinin and substance P are degraded by ACE, and prostaglandin production may be increased by bradykinin. Potential mechanisms involve bradykinin-induced sensitization of airway sensory nerves or activation of bradykinin receptors.48,49
Treatment options include discontinuing the medication and using alternative medications. The American College of Chest Physicians (ACCP) Evidence-Based Clinical Practice Guidelines recommend discontinuing the ACE inhibitor.44 Cough associated with ACE inhibitors usually resolves within a few days of discontinuing the medication but can take up to several weeks.44,50 Cough often recurs with the same or different ACE inhibitor. Cough has not been reported in association with ARBs, and the ACCP guidelines recommend ARBs as a treatment alternative.44 In patients in whom the benefits of treatment with an ACE inhibitor outweigh the risks of a recurrence of the cough, a repeat trial of ACE inhibitor might also be indicated.44
Although the onset of angioedema associated with ACE inhibitors often occurs duriindicatedng the first week or first month after initiating treatment in 25% to 60% of patients, it can occur at any time.42,51 This unpredictable characteristic of ACE inhibitor–associated angioedema can lead to a delay in identifying the association. The newer ARBs, often used as an alternative to ACE inhibitors, have also been associated with angioedema.51-53 Studies estimate that the incidence of ARB-associated angioedema varies from 0% to 32% in patients with a history of ACE inhibitor–associated angioedema.51,54-56
Women, African Americans, and patients with a history of either idiopathic angioedema or C1 inhibitor deficiency are at higher risk for angioedema.42,57 Angioedema associated with either ACE inhibitors or ARBs generally affects the head and neck, and pruritis and urticaria generally do not occur.
Similar to the pathogenesis of ACE inhibitor–induced cough, the pathogenesis of ACE inhibitor–associated angioedema may be related to the accumulation of bradykinin and other vasoactive peptides. ACE catalyzes the conversion of angiotensin I to angiotensin II, which is a vasoconstrictor that can increase blood pressure. ACE also degrades bradykinin, a vasodilator that opposes the effects of angiotensin II. With ACE inhibition, subsequently increased levels of bradykinin can induce angioedema through vasodilation and increased vascular permeability. In some patients with ACE inhibitor–associated angioedema, an active metabolite of bradykinin, des-Arg9-BK, may be abnormally degraded and lead to increased bioavailability of bradykinin.58 Dipeptidyl peptidase IV might also play a role in angioedema. Decreased levels of this enzyme can lead to increased amounts of substance P, which can be associated with increased vascular permeability and leakage of plasma proteins.59
The approach to a patient with a history of an adverse reaction to an ACE inhibitor or ARB should begin with a detailed medical history including a list of all medications, indication for each medication, dose, date of initiation, and duration of therapy. The clinical manifestations of the reaction, including associated symptoms, should also be reviewed. Additional history to obtain is whether the patient had a prior exposure to the same or other ACE inhibitor or ARB medications, the effect of drug discontinuation, and the treatment of the prior reaction. Other factors that may be related to the development of urticaria and angioedema should be considered, including infections; collagen vascular diseases; malignancy; physical factors such as pressure, vibration, or exposure to heat, cold, or sunlight; and family history of angioedema suggesting C1-inhibitor deficiency.
Treatment of angioedema should be tailored to each patient and may include antihistamines, oral or intravenous corticosteroids, or intramuscular or subcutaneous epinephrine for severe airway compromise. No studies support the effectiveness of treatment with antihistamines or corticosteroids. Symptoms generally resolve within 48 hours after treatment is initiated and the ACE inhibitor or ARB is discontinued.
There is no valid skin or blood test to establish a diagnosis of ACE inhibitor–associated or ARB-associated angioedema. Desensitization, a process of administering incremental doses of a drug over hours to days and conversion of a drug allergy to a state of drug tolerance, is not indicated for patients who have angioedema associated with ACE inhibitors or ARB because the reaction is not an IgE-mediated process.
All ACE inhibitors carry the same risk of developing angioedema. Patients who experience angioedema related to one ACE inhibitor are at risk for more severe and frequent episodes with other ACE inhibitors.60 The best approach for a patient with ACE inhibitor–associated angioedema is to avoid other ACE inhibitors and use alternative, equally efficacious, medications.
Although patients with ACE inhibitor–associated angioedema might tolerate ARBs, ARBs should be used with caution because these patients can also develop angioedema associated with ARB. Studies estimate that the incidence of experiencing ARB-associated angioedema varies from 0% to 32% in patients with a history of ACE inhibitor–associated angioedema.51,54-56 The largest of these studies involved more than 2000 patients with congestive heart failure. Either an ARB (candesartan) or placebo was given to patients with a history of adverse reaction to ACE inhibitors. In the treatment group, 39 of 1013 patients previously experienced angioedema or anaphylaxis from an ACE inhibitor. Of these 39 patients, only 3 (7.6%) developed angioedema associated with candesartan. More studies are needed to evaluate the incidence of developing ARB-associated angioedema in patients with ACE inhibitor–associated angioedema. Any patient3–associated angioedema prescribed an ARB should be counseled regarding the possible risk of developing angioedema.