Penicillin allergy: Immediate reactions

INTRODUCTION — Penicillin allergy is the most commonly reported medication allergy. Up to 10 percent of all patients and 15 percent of hospitalized patients report some type of penicillin allergy [1-3]. Upon thorough evaluation, many of these patients are found not to be allergic and are able to safely receive penicillin. However, penicillins are among the leading causes of drug-induced anaphylaxis, including fatalities. Thus, a systematic approach to the evaluation of penicillin allergy is necessary.

This topic reviews immediate allergic reactions to penicillins, which are most often defined as reactions that develop within minutes to one hour of administration. The epidemiology, risk factors, diagnosis, and management of immediate reactions to penicillins will be discussed here. Related topics include the following:

●An algorithmic approach to the use of related antibiotics, with and without access to penicillin skin testing, is presented separately. (See "Choice of antibiotics in penicillin-allergic hospitalized patients".)

●Penicillin skin testing is described elsewhere. (See "Penicillin skin testing".)

●Delayed reactions to penicillins, ranging from maculopapular drug eruptions to severe systemic reactions, are reviewed elsewhere. (See "Penicillin allergy: Delayed hypersensitivity reactions".)

DEFINITION — An allergic drug reaction is defined as a specific immunologic reaction to a drug [4]. The classification and pathogenesis of drug allergies are discussed in detail separately. (See "Drug hypersensitivity: Classification and clinical features" and "Drug allergy: Pathogenesis".)

Immediate versus delayed reactions — The World Allergy Organization has recommended categorizing immunologic drug reactions based upon the timing of the appearance of symptoms [4]. This system defines two broad categories of reactions: immediate and delayed. This approach is intended to distinguish immunoglobulin (Ig)E-mediated (type I reactions), which account for many immediate reactions, from other types, because IgE-mediated reactions carry the risk of life-threatening anaphylaxis if the patient is re-exposed.

●Immediate reactions classically begin within one hour of the last administered dose and may begin within minutes. However, some immediate reactions appear after one hour, particularly if the drug was administered orally or taken with food. For these reasons, some guidelines include reactions beginning up to six hours after the last administered dose in the immediate category [5]. Nevertheless, the period of one hour identifies most immediate reactions. When allergic sensitization first develops, the initial symptoms may appear during the latter days of treatment (ie, not following the first dose of the course) but usually within one hour of the last administered dose and then escalate rapidly with any successive doses.

●Delayed reactions appear after multiple doses of treatment, typically after days or weeks of administration. Most delayed reactions begin after six hours and typically after days of treatment. As an example, delayed reactions to amoxicillin classically start on day 7 to 10 of treatment and may even begin one to three days after cessation of treatment. Delayed reactions are more common than immediate reactions. They may be caused by several different mechanisms, but they are not IgE-mediated. Delayed reactions to penicillins are discussed separately. (See "Penicillin allergy: Delayed hypersensitivity reactions".)

EPIDEMIOLOGY — Penicillin allergy can develop at any age. Some type of penicillin reaction is reported by 5 to 10 percent of patients, although in large scale studies of penicillin skin testing, more than 90 percent of these individuals are found not to have positive skin tests and are able to tolerate penicillins [2,6-8]. In addition, the prevalence of immediate reactions to penicillin appears to have declined in the United States over the last two decades, based on the rate of positive penicillin skin tests [9,10].

Anaphylaxis, the most severe manifestation of immediate penicillin reactions, may be defined as a severe allergic reaction that is rapid in onset and may cause death. Multiple fatalities from penicillin-induced anaphylaxis have been reported [11-13]. Penicillin-induced anaphylaxis occurs with an incidence of between one and four episodes per 10,000 administrations, based on studies that involved parenteral administration of penicillins [11,14,15]. In a later study, the incidence of anaphylaxis with oral amoxicillin was closer to 1 per 200,000 administrations [16]. Amoxicillin and penicillin were the leading causes of severe antibiotic-induced anaphylaxis in one large database, probably because these drugs are so commonly administered [17]. In the United States, it has been estimated that 500 to 1000 deaths per year are due to penicillin-induced anaphylaxis [12]. The diagnosis and treatment of anaphylaxis, from all causes, is reviewed elsewhere. (See "Anaphylaxis: Emergency treatment".)

No studies have prospectively evaluated the rates at which patients become sensitized (ie, develop penicillin-specific IgE) during penicillin treatment.

Risk factors — A number of risk factors have been identified for the development of immediate reactions to penicillin and antibiotics in general:

●Age – Patients between the ages of 20 and 49 appear to be at greatest risk of having immediate allergic (particularly anaphylactic) reactions to penicillin [11]. This observation may reflect age-dependent differences in immune responses to medications or may be an artifact related to usage patterns.

●Frequent, repeated exposure – Frequent, repeated courses of penicillins appear to favor the development of allergic sensitization [18]. This is supported by the observation that patients with cystic fibrosis (who require frequent treatment with intravenous antibiotics) are particularly prone to developing IgE-mediated antibiotic allergies, although prospective studies are lacking, and this patient population also develops non-IgE mediated drug reactions [19].

●Route of administration – Parenteral administration seems to have a higher rate of sensitization compared with oral administration and is more likely to result in anaphylaxis [20]. The vast majority of anaphylactic deaths have occurred in patients treated with intravenous or intramuscular, rather than oral, penicillins [11]. Dose may also be a factor in such cases, since parenteral doses are often higher than those administered orally. Topically-applied penicillin is highly sensitizing and has not been in clinical use for decades for that reason.

●Multiple antibiotic allergy syndrome – The term "multiple antibiotic allergy syndrome" may be assigned to individuals who have experienced allergic reactions (either IgE- or non-IgE-mediated) to two or more non-cross-reacting antibiotics. A number of studies support the theory that patients who develop an allergy to one medication are more likely to react to a second unrelated compound [21-26]. The reasons why certain patients have a heightened propensity to generate allergic reactions to drugs have not been identified but could include genetic as well as exposure factors.

●Hereditary factors – While genetic factors may play a role in penicillin and other antibiotic allergies, first-degree relatives of patients with penicillin allergy do not need to avoid penicillins empirically. Evidence for genetic factors in the development of penicillin allergy include an association between particular human leukocyte antigen (HLA)-DRB genotypes and IgE-mediated penicillin allergy in Chinese populations [27]. Additional evidence for hereditary factors comes from retrospective, survey-based studies [28-30]. For example, children of parents who reported an antibiotic allergy were 15 times more likely to be antibiotic-allergic (by history) [28]. However, the results of survey-based studies are limited by a reliance on patient history (which is known to be a poor predictor of drug allergy) and by lack of confirmatory testing or provocative challenges.

●Other allergic diseases – Patients with other allergic diseases (such as allergic rhinitis, asthma, or food allergies) do not appear to be at higher risk for becoming sensitized (ie, developing drug-specific IgE) to penicillin compared with people without allergic diseases [8,11,31-33]. However, once sensitized, patients with asthma (in particular) are at greater risk for serious reactions and anaphylaxis [11].

●Occupational exposure – Exposure to penicillins in occupational settings (health care, laboratory workers) can result in sensitization, although it is unclear how often this results in clinical allergy. Immediate reactions, as well as asthma, contact urticaria, and contact dermatitis are reported [34-37].

Natural history — Penicillin allergy resolves with time in many (but not all) patients who successfully avoid penicillin. This is discussed in detail below. (See 'Time elapsed since the reaction' below.)

Another clinical question that occasionally arises in patients who have been determined not to be allergic despite a history of a reaction in the past, is whether penicillin allergy can recur if the patient is given penicillins again. This appears to be rare, and it is discussed separately. (See "Penicillin skin testing", section on 'Resensitization'.)

CLINICAL MANIFESTATIONS — Many immediate reactions are type I, IgE-mediated reactions. This type of reaction usually presents with various combinations of the following:

●Pruritus

●Flushing

●Urticaria (intensely pruritic, raised, red plaques that appear and resolve within hours (picture 1))

●Angioedema

●Bronchospasm (wheezing, chest tightness, difficulty breathing, or repetitive dry cough)

●Laryngeal edema (throat tightness or change in voice quality)

●Abdominal distress (cramping, nausea, vomiting, or diarrhea)

●Hypotension

The full spectrum of signs and symptoms possible with immediate reactions, both IgE-mediated and non-IgE-mediated, is shown in the table (table 1).

Some of the features of immediate reaction, such as pruritus, urticaria, and angioedema, can also be seen in delayed reactions to penicillins. However, if a patient reports a pruritic rash or urticaria or angioedema, an evaluation for an immediate reaction is indicated. Mild, skin-limited reactions to amoxicillin in young children are often not IgE-mediated. Studies of this population are reviewed elsewhere. (See "Penicillin allergy: Delayed hypersensitivity reactions", section on 'Delayed urticarial eruptions' and "Penicillin allergy: Delayed hypersensitivity reactions", section on 'Studies in children'.)

If a penicillin reaction is witnessed, it is extremely helpful to record the details of the reaction in the patient's medical record, including the time from the last administered dose to the first symptoms, the observed signs and symptoms, and severity. These details are invaluable to clinicians treating the patient in the future. In addition, if the patient develops a severe reaction that appears to be anaphylaxis, a serum tryptase (collected one to three hours of the onset of the reaction) may reveal an elevation in total tryptase, which indicates massive activation of mast cells and is very specific (although not sensitive) for the diagnosis of anaphylaxis. (See "Anaphylaxis: Acute diagnosis", section on 'Laboratory tests'.)

PATHOGENESIS — Type I reactions require the presence of drug-specific IgE. A small percentage of patients form drug-specific IgE upon exposure to a medication, while most do not. Penicillin in the intact form is not very allergenic, but under physiologic conditions, penicillins spontaneously degrade to a number of reactive intermediates that covalently bind to carrier proteins in the tissue and serum. These complexes of breakdown products and proteins are multivalent and thus more allergenic and can trigger the formation of drug-specific IgE through the coordinated actions of B cells and T helper cells. These interactions, in conjunction with T cell-derived cytokines, enable B cells to mature and secrete drug-specific IgE molecules, which diffuse through the circulation and attach to the surface of mast cells and basophils throughout the body.

The production of drug-specific IgE and its binding to mast cells and basophils is called sensitization and is clinically asymptomatic. As with other forms of IgE-mediated allergy, having formed drug-specific IgE puts the patient at risk for an IgE-mediated drug reaction but does not guarantee that one will occur. In other words, drug-specific IgE is necessary but not sufficient for an IgE-mediated reaction. The other factors that must be present for a reaction to occur are not known.

When a sensitized individual is given the penicillin again, the drug again degrades and couples to serum and tissue proteins. This time, if the drug/carrier protein complexes interact with and cross-link enough drug-specific IgE molecules on the surface of mast cells and/or basophils, the cells can activate. Sudden and widespread activation of mast cells and basophils can result in release of an array of vasoactive mediators, which in turn, cause the signs and symptoms of an immediate reaction.

Allergens — In most allergic patients, the principal immune response is not directed against the penicillin molecule in its native state. Instead, the immune response is against new antigenic determinants that are formed once the penicillin molecule degrades and binds to serum and tissue proteins [38-40].

The allergenic components of penicillins are found either in the core ring structure that is common to all penicillins (as well as cephalosporins and carbapenems) or (much less commonly) within side chains that distinguish different penicillins from one another. The portion of the molecule to which the patient has made IgE antibodies determines patterns of cross-reactivity with other related drugs. A detailed understanding of the structure of these drugs is required to interpret skin test results and predict cross-reactivity patterns. This is reviewed briefly here and in greater detail elsewhere. (See "Penicillin skin testing", section on 'Skin testing reagents and concentrations'.)

Allergens arising from the beta-lactam core — The beta-lactam ring of penicillins is inherently unstable, and after administration, it spontaneously degrades into intermediates that form complexes by binding to cell-bound or serum proteins. One such complex is penicilloyl, which is called the "major determinant" because most of the penicillin covalently links with lysine residues in common proteins. A synthetic preparation of the major determinant (penicilloyl polylysine) is commercially available for use in skin testing. The remaining parts of the penicillin molecule undergo a similar process and form other complexes. As a group, these are called the "minor determinants," and only about 10 percent of allergic patients are believed to be reactive to these. One proallergen for the minor determinants is penicillin G itself, which can be allergenic through multiple protein interactions. (See "Penicillin skin testing", section on 'Skin testing reagents and concentrations'.)

Allergens arising from side chains — Some patients form IgE antibodies that recognize the side chains, called R groups, rather than the core ring structure. Clinically, these individuals develop immediate-type reactions to amoxicillin or ampicillin (ie, the aminopenicillins) but are able to tolerate penicillin and are said to be selectively allergic to the aminopenicillins [41,42].

●In North America, selective IgE-mediated allergy to the aminopenicillins appears to be uncommon [43-47]. The largest experience in the United States was of 5006 patients evaluated at an allergy/immunology practice of among whom only 7 (0.14 percent) were selectively allergic to aminopenicillins [48]. The diagnosis of selective allergy to aminopenicillins using skin testing is described separately. (See "Penicillin skin testing", section on 'Aminopenicillins'.)

●The experience from southern Europe is different in that one-third or more of patients are selectively sensitive to amoxicillin or ampicillin based upon skin testing [49,50]. The reasons for this difference between American and European patients are not known, although amoxicillin is available without a prescription in some European countries and accounts for a sizeable majority of all antibiotic use. In addition, rare selective reactions to clavulanate have also been described in European populations [51-53].

●Side chain-specific sensitization to antipseudomonal penicillins (ie, ticarcillin and piperacillin) has been observed among patients with cystic fibrosis who often receive frequent high-dose parenteral courses of semisynthetic penicillins [54]. (See 'Risk factors' above.)

The different reagents used in penicillin skin testing and strategies for detecting selective allergy to aminopenicillins and clavulanate are reviewed separately. (See "Penicillin skin testing".)

EVALUATION AND DIAGNOSIS — The evaluation begins with a careful clinical history. This is initially taken by the generalist, who can sometimes exclude immediate allergy based on history alone. The generalist then decides if the patient should be referred to an allergist for further evaluation and how urgently this should occur. These initial steps in the evaluation are depicted in an algorithm (algorithm 1).

In most situations, the allergist will perform skin testing:

●If skin testing is positive in a history-positive patient, then the allergy is confirmed (ie, evidence of continued skin test reactivity), and the patient must continue to avoid penicillins unless he/she is given the penicillin using a desensitization protocol.

●If skin testing is negative, it is followed by challenge to assure (in a safe environment) that the drug is tolerated and can be used in the future. In vitro tests perform poorly and are rarely useful. When penicillin skin testing is not possible, it may be appropriate to perform a graded challenge, but only in patients believed to have a low probability of true immediate allergy. This type of challenge involves several steps, since the patient could be allergic and have a reaction.

A general discussion of the diagnosis of drug allergy is found separately. (See "An approach to the patient with drug allergy".)

Clinical history — The purpose of the initial clinical history is to assess the following:

●Was the patient's past reaction allergic in nature? Isolated nausea or diarrhea, yeast vaginitis, or other known side effects of antibiotics are not allergic reactions. No further evaluation is needed if the reaction is clearly not allergic. Another common scenario is the patient who has never actually taken penicillin but is empirically avoiding it because one or more family members has penicillin allergy. Such patients may receive penicillins [55].

●If the reaction was allergic in nature, what type of reaction did the patient likely experience? What precise signs and symptoms does the patient recall? What was the timing of onset of the reaction, both from the precipitating dose, as well as from the initiation of that course of therapy? Do the symptoms and timing of the event suggest an immediate or a delayed reaction? (See 'Immediate versus delayed reactions' above.)

It is difficult to classify a patient's past reaction as immediate or delayed in many cases, and caution is warranted. Patients can sometimes provide a reasonably detailed history if the reaction was in the recent past. The medical record should be reviewed to see if details about the reaction were recorded. However, if the reaction occurred years ago and few details are available, it is prudent not to make assumptions about the nature of it for the reasons discussed below (see 'Accuracy' below). This is a common scenario, particularly in adults. If the history is too vague to categorize, referral for allergy evaluation is indicated (algorithm 1).

●Is there any information in the history to suggest that a severe drug reaction took place? Severe reactions include anaphylaxis and Stevens-Johnson syndrome/toxic epidermal necrolysis. Patients who use specific phrases, such as "anaphylaxis" or "nearly died," report emergency department visits or hospitalizations, or describe blistering or peeling of the skin or mucous membranes or organ damage in association with prior drug reactions, must be taken very seriously and should be referred to an allergist for further evaluation and education about avoidance.

A thorough history of a drug reaction should ascertain the following:

●Each exact agent reportedly associated with a reaction and the indication for its use (to make sure that the agent matches the indication)

●When the reaction occurred (ie, how long ago) (see 'Time elapsed since the reaction' below)

●The dose and route of medication taken (if known)

●Other concurrent medications, especially if they were new or temporary

●Any treatment given and response to that treatment (including the duration of reaction)

It is sometimes helpful to know when the various penicillins became available. Patients who report that they reacted to an injection of "penicillin" typically received benzylpenicillin (ie, penicillin G). Oral penicillin V potassium has been used since the 1960s. Ampicillin was developed in the early 1960s and began to be used widely in the 1970s. Amoxicillin was developed in the early 1970s and widely used beginning in the 1980s. Other semisynthetic penicillins with targeted microbial specificities for Staphylococcus and Pseudomonas were introduced in the 1970s and 1980s.

Historical features that predict true allergy — Three questions have been identified as useful for predicting that an allergy evaluation will show true allergy. In a retrospective study of 410 adults referred to specialty drug allergy clinics for urticarial eruptions occurring in association with a beta-lactam antibiotic (eg, penicillins, cephalosporins) in the past (usually within the past 12 months), the three questions were [56]:

●Did the urticaria develop after the first dose (ie, dose 1) (versus after subsequent doses)?

●Did the urticaria develop within 1 hour of the inciting dose?

●Did the urticaria resolve within 1 day of stopping the beta-lactam in question?

This set of questions was termed the "1-1-1 criterion" by the investigators. The combination of three "yes" answers had a sensitivity, specificity, and positive predictive value (PPV) of 85, 85, and 90 percent respectively, for predicting a positive allergy evaluation. Therefore, patients who answer yes to all three questions should be considered at high risk for immediate allergy and referred to an allergy specialist for a formal evaluation with skin testing, rather than being re-exposed to a beta-lactam through a graded challenge. Recommendations for the management of patients who answered yes to just one or two questions, or couldn't recall the information, were not included in this study. Other potential limitations are the uncertain application to children and the questionable relevance to adults who may have had the urticarial reaction many years earlier. Despite these issues, the adult who can answer yes to these questions with certainty should be evaluated thoroughly before re-exposure is considered.

Accuracy — Despite the importance of the clinical history, several studies have demonstrated that the presence of IgE-mediated penicillin allergy cannot be accurately predicted based upon history alone [33,55,57-59]. Patients with convincing reaction histories to penicillin (eg, pruritus, hives/angioedema, bronchospasm, or anaphylaxis) are more likely to be allergic than patients with vague reaction histories (eg, "rash") [8,33,60,61]. However, vague or incomplete reaction histories cannot be discounted, because some of those patients also have IgE-mediated allergy [8,33,55,60,62]. In a review of the published literature, one-third of patients with a history of penicillin allergy who were skin test-positive had vague reaction histories [58].

There are likely multiple reasons that the clinical history is a poor predictor of penicillin allergy. Patients may not describe cutaneous symptoms accurately or they may have reacted as children and know about the reaction only through parental reports. However, the most important factor is probably the gradual loss of penicillin IgE over time as demonstrated in prospective studies, leading to negative skin tests and safe retreatment in most patients, even in those who had serious systemic reactions in the past.

Time elapsed since the reaction — The time elapsed since the last reaction is important because penicillin-specific IgE antibodies decrease over time, and therefore, patients with recent reactions are more likely to be allergic than patients with distant reactions [63-65]:

●Approximately 50 percent of patients with IgE-mediated penicillin allergy have lost the sensitivity 5 years after their last reaction.

●Approximately 80 percent of patients with IgE-mediated penicillin allergy have lost the sensitivity after 10 years.

Thus, because the history is often unreliable and the allergy can be lost over time, most patients who present with past reactions require further evaluation to determine if IgE-mediated penicillin allergy is present.

Exposure to related medications since the initial reaction — Although it is not recommended that patients with immediate reactions to penicillins be exposed subsequently to other penicillins, this sometimes occurs. When re-exposure has occurred, it can provide valuable information (ie, if the patient tolerated the penicillin compound, it would exclude an allergy). However, the clinician should have confidence in the accuracy of the information given and should confirm it using the patient's medical record or pharmacy records when possible.

●Other penicillins – If the patient initially reacted to the aminopenicillins (amoxicillin or ampicillin) but has tolerated penicillin V potassium or G subsequent to the initial reaction, then he/she could have either lost the allergy altogether or have a selective allergy to the side groups present in aminopenicillins. Such patients could receive penicillin V potassium or G again, but they should continue to avoid the aminopenicillins and drugs with related side chains (table 2). If the patient initially reacted to penicillin V potassium or G but tolerated amoxicillin or ampicillin since, then he/she has proven himself/herself not to be allergic to the core beta-lactam structure, as this is shared by all penicillins. The patient should be able to receive penicillins and aminopenicillins safely.

●Cephalosporins – Patients with a history of an immediate penicillin reaction who have tolerated one or more courses of cephalosporins subsequent to the penicillin reaction have proven themselves to not be allergic to cephalosporins, but they could still be allergic to penicillins. It is known that 98 percent of penicillin skin test-positive patients (to the major and/or minor penicillin determinants) are able to tolerate cephalosporins (see "Allergy evaluation for immediate penicillin allergy: Skin test-based diagnostic strategies and cross-reactivity with other beta-lactam antibiotics"). However, if a patient originally reacted to a side chain of amoxicillin or ampicillin and then tolerated a cephalosporin with a dissimilar side chain, he/she could still possibly react to those cephalosporins that share an identical side chain with amoxicillin or ampicillin (table 2) and should avoid those cephalosporins.

When to refer — Referral to an allergist for evaluation and definitive diagnosis is encouraged for patients with possible penicillin allergy who need antibiotics with any regularity [5]. The optimal time for referral is when the patient is well and in a baseline state. Occasionally, the generalist can exclude penicillin allergy based upon the clinical history, although in many patients, the clinical history of the past reaction is too vague to characterize (algorithm 1). (See 'Clinical history' above.)

The urgency with which the patient is evaluated depends upon the setting (outpatient or inpatient) and whether or not the patient has a current need for a penicillin. (See 'Specific patient scenarios' below.)

Impact of penicillin allergy on care — An underappreciated aspect of penicillin allergy is the morbidity, mortality, and economic cost associated with the unnecessary withholding of penicillins and other beta-lactam antibiotics where they are the preferred drugs in patients who are assumed to be allergic on the basis of history alone. Patients labeled penicillin-allergic are more likely to be treated with broad-spectrum antibiotics, such as fluoroquinolones or vancomycin [1,3,66-70]. In a large cohort study of ambulatory adults in the United Kingdom, those labeled as penicillin-allergic were more likely to receive broad-spectrum antibiotics such as fluoroquinolones, as well as macrolides and clindamycin, and were at significantly higher risk for the development of methicillin-resistant staphylococcus aureus (MRSA) and Clostridioides difficile colitis over a mean of six years of follow up than those without this label [71]. These results confirm a similar previous large United States study, which also demonstrated significantly higher rates of MRSA, C. difficile, and vancomycin-resistant enterococcus (VRE) in patients labeled penicillin allergic, compared to matched control subjects without this label [3].

The impact on care in hospital settings and possible complications in patients with significant comorbidities are reviewed in greater detail separately. (See "Choice of antibiotics in penicillin-allergic hospitalized patients", section on 'Impact of penicillin allergy on care'.)

Specific patient scenarios

Outpatient with no current need for a penicillin — The optimal time for referral and skin testing is when the patient is well and not in urgent need of a penicillin. Referral is particularly useful for the following types of patients:

●Those with reactions to two or more classes of antibiotics, since their antimicrobial treatment choices are relatively limited.

●Those who require frequent antibiotic therapy, such as for rhinosinusitis, bronchitis, or urinary tract infections. Patients with cystic fibrosis who require periodic courses of piperacillin and other antipseudomonal penicillins are good candidates for nonurgent evaluation.

Outpatient with a current need for a penicillin — If a patient has a need for penicillin or a related beta-lactam antibiotic at the present time and can be seen by an allergist promptly, then skin testing can be performed and the allergy either confirmed or excluded. Examples of situations in which urgent allergy evaluation is indicated include pregnant women who require penicillin for the treatment of syphilis. (See "Syphilis in pregnancy", section on 'Maternal treatment'.)

However, timely referral to an allergist may not be feasible, and there are usually other alternative antibiotics that can be used. In these situations, the patient should be treated with an alternative antibiotic.

Inpatient with a current need for a penicillin — In hospitals in which allergists are available for consultation, a referral should be sought if it is determined that a penicillin would be significantly more effective for the patient's infection compared with alternative agents. As an example, a penicillin-allergic patient with native valve endocarditis due to S. aureus and a history of penicillin allergy may sometimes not respond adequately to alternative drugs, such as vancomycin. In this situation, if the patient's past reaction to penicillin is believed to be suggestive of an immediate allergy, then penicillin skin testing may be performed, and if it is negative, the patient may receive the appropriate penicillin compound. If the skin testing is positive, then rapid desensitization, under the supervision of an allergist, may be pursued. (See "Choice of antibiotics in penicillin-allergic hospitalized patients".)

Penicillin skin testing — Penicillin skin testing is the preferred method of evaluation and diagnosis of immediate reactions. Penicillin skin testing is safe, rapid, sensitive, and cost-effective [3]. Results are obtained in less than one hour with minimal patient discomfort, although this type of testing should be performed by allergy specialists, so referral is required. In some centers, pharmacists and infectious disease specialists are being trained to perform and interpret penicillin skin testing as part of antimicrobial stewardship programs [72,73]. (See "Penicillin skin testing".)

Skin testing (sometimes called "immediate" skin testing) is a bioassay performed on the skin that detects the presence of allergen-specific IgE on a patient's mast cells. Patients who have drug-specific IgE detected with skin testing are at high risk for type I, IgE-mediated, immediate allergic reactions, including anaphylaxis.

Indications — Skin testing is indicated if the patient's past reaction had any of the features of an immediate reaction. Signs and symptoms of immediate reactions are shown in the table (table 1).

Skin testing is also recommended if the clinical history is too vague to be meaningful. In this setting, skin testing is used to exclude immediate reactions and thus narrow the differential diagnosis.

Skin testing is optimally performed when the patient is well and in a baseline state. However, it can be performed on ill and hospitalized patients if needed, provided they are not taking H1 antihistamines. The indications, safety, technique, and interpretation of skin testing to penicillins and aminopenicillins are reviewed in detail separately. (See "Penicillin skin testing".)

Skin testing is not indicated as a means of screening patients who have never had a past adverse reaction to penicillin or have never taken a penicillin. For example, it should not be performed on patients with a family history of penicillin allergy, who themselves have never taken a penicillin. Skin testing does not predict allergy in these settings.

Barriers to skin testing — Penicillin skin testing may not be possible if either the reagents for skin testing or a clinician (allergist) trained to perform and interpret skin testing is not available:

●If the skin testing reagent containing the important degradation product of penicillin (the major determinant penicilloyl-polylysine, available as Pre-Pen [brand name] in the United States, Canada, and Mexico and as DAP [brand name] in Spain) is not available, skin testing with penicillin alone is insufficiently sensitive and is not recommended. (See "Penicillin skin testing", section on 'Minor determinants'.)

●Penicillin skin testing requires training and knowledge to perform and interpret correctly and should be done by an allergy specialist whenever possible.

There are also patient-related factors that may make skin testing impossible. These include inability to discontinue interfering medications, such as use of antihistamines and tricyclic antidepressants. These are discussed separately. (See "Overview of skin testing for IgE-mediated allergic disease", section on 'Factors affecting results'.)

If skin testing is not possible or unavailable, graded challenge is the preferred approach to diagnosis. However, graded challenge is only appropriate in patients who are believed unlikely to be allergic to the medication. Patients who have a suggestive history must be assumed to be allergic and could only be given a course of penicillin using a desensitization procedure. (See 'Graded challenge (test dosing)' below and 'Desensitization' below.)

There is great interest in developing protocols for the identification of patients who are at low risk for true IgE-mediated allergy and could be challenged in the primary care setting, without the need for skin testing or allergy referral [74]. Several studies have identified low-risk patients based on the nature of the past reaction to a penicillin and the time that has elapsed since the reaction. To date, these protocols have been largely performed and supervised by allergy specialists, but if an approach can be identified and validated as safe in non-specialty settings, it would greatly expand the capacity to de-label patients as allergic to penicillin. Research in this area is ongoing.

In vitro testing and limitations — In vitro tests are of limited clinical utility in the diagnosis of penicillin allergy, and we do not obtain them in most patients. Enzyme-linked immunosorbent assays for specific IgE to penicillin G, penicillin V, penicilloyl, amoxicillin, and ampicillin are commercially available, but the predictive values are not well-defined.

●The sensitivity of in vitro tests for penicillin-specific IgE was as low as 45 percent compared with skin testing when performed in academic settings [75,76]. A negative in vitro test result is also problematic because it lacks the negative predictive value of skin testing.

●A prospective series of 150 consecutive patients with a history of penicillin allergy illustrates the limitations of in vitro testing. In this study, both skin testing and in vitro ImmunoCAP testing were performed, followed by oral challenge if skin test was negative [77]. Four patients had positive in vitro tests, yet all tolerated the amoxicillin challenge, proving these tests to be false-positive. There were six skin test-positive patients (none of whom were also ImmunoCAP-positive) who were not challenged.

●False-positive ImmunoCAP results were found in 26 percent of patients suspected of having penicillin allergy in one series and were attributed to a clinically irrelevant molecule [78].

The basophil activation test employing flow cytometry is another in vitro test that has been studied for diagnosis of penicillin and other beta-lactam allergies. However, the very limited number of publications available suggests that this test is also inferior to skin testing [79,80]. Furthermore, basophil activation test is not commercially available and only performed in selected specialized centers.

Graded challenge (test dosing) — Graded challenge, also known as test dosing, refers to cautious administration of a drug to a patient who is believed unlikely to be allergic to the medication [81]. Graded challenge does not modify the patient's response to the drug. Therefore, patients who tolerate a drug during graded challenge prove that they are not allergic to it. Conversely, the clinician must be prepared to recognize and treat an allergic reaction if the patient does prove to be allergic. The intent of a graded challenge is to give the drug in such a way that the patient will only be exposed to a small amount, followed by a period of observation, in case he/she does react. Protocols are discussed elsewhere. (See "An approach to the patient with drug allergy", section on 'Graded challenge and drug provocation' and "Choice of antibiotics in penicillin-allergic hospitalized patients", section on 'Test dose procedure (graded challenge)'.)

OPTIONS FOR FUTURE TREATMENT — Patients who have been confirmed to have an immediate reaction to penicillin by skin testing (or because the reaction was witnessed and recorded in detail) have two options: avoidance of all penicillins (although related drugs, such as cephalosporins, may be tolerated) or treatment with penicillin using a desensitization protocol.

Avoidance of all penicillins — Patients with past immediate reactions to penicillins should avoid all penicillins (ie, including amoxicillin and ampicillin and semisynthetic penicillins, such as nafcillin) in the future, unless an evaluation (including skin testing) is performed to define other related medications that can be safely administered to that individual.

Use of related drugs — Cephalosporins, carbapenems, and monobactams share structural similarities with penicillins.

●Cephalosporins are the most closely related to penicillins, and among patients who are skin test-positive to penicillins, approximately 2 percent can be anticipated to react to cephalosporins, with some of these reactions being anaphylaxis. Carbapenems are also related, although the risk of cross-reactivity with penicillins has been estimated at <1 percent. Although these rates are low, reactions can be serious and are avoidable with further evaluation. Thus, we do not advise the use of cephalosporins or carbapenems in patients who report penicillin allergy without at least performing some evaluation to stratify risk. Our approach is reviewed separately. (See "Allergy evaluation for immediate penicillin allergy: Skin test-based diagnostic strategies and cross-reactivity with other beta-lactam antibiotics".)

●Monobactams (ie, aztreonam) share some structural features with penicillins, but they are not clinically relevant. Thus, aztreonam may be safely given to patients with a history of penicillin allergy. (See "Allergy evaluation for immediate penicillin allergy: Skin test-based diagnostic strategies and cross-reactivity with other beta-lactam antibiotics", section on 'Monobactams (aztreonam)'.)

Desensitization — If a patient with a past immediate reaction to penicillin absolutely requires treatment with a penicillin because no other equally effective antibiotic is available, then desensitization may be appropriate. Drug desensitization is a procedure which temporarily induces short-term tolerance, allowing the patient with immediate allergy to receive an uninterrupted course of the medication safely. Drug desensitization renders mast cells unresponsive to the drug in question, although the exact immunologic mechanism is not well understood [81]. Drug desensitization is effective as long as the patient is receiving the drug, although the patient's sensitivity returns shortly after the drug is cleared from the body. It is critical to ensure that the patient understands that the desensitization is temporary and that the allergy is still present. (See 'Post-desensitization care' below.)

Among all antibiotics, penicillin desensitization is supported by the largest body of published evidence [82-87]. Desensitization is reviewed briefly here and in greater detail separately. (See "Rapid drug desensitization for immediate hypersensitivity reactions".)

Desensitization can be performed by oral, intravenous, or subcutaneous routes. The oral route is preferred (if feasible) because it is thought to be safer than parenteral administration. However, desensitizations are often performed in hospitalized patients requiring parenteral medications because it is in this setting that patients with few remaining options for therapy are usually encountered.

Indications — Desensitization to penicillin is indicated in the following patients when there are no equivalent alternate antibiotics:

●Those with a history suggestive of an IgE-mediated reaction to a penicillin and positive penicillin skin testing.

●Those with a history suggestive of an IgE-mediated reaction within the recent past (ie, within the previous 10 years) in whom skin testing cannot be performed in the timeframe required. After desensitization and treatment, such patients should be referred for penicillin skin testing to prove or exclude penicillin allergy with certainty and guide future treatment options.

●Those with a history suggestive of an IgE-mediated reaction more than 10 years ago in whom skin testing cannot be performed in the timeframe required. However, patients with distant reactions are unlikely to be allergic and may be candidates for graded challenge instead of desensitization. Patients who pass a graded challenge are proven not to be allergic to penicillin and do not require subsequent skin testing.

Contraindications — Desensitization should not be attempted in patients with histories of blistering reactions, such as Stevens-Johnson syndrome or toxic epidermal necrolysis, because even small doses of the drug may induce severe progressive reactions. Desensitization is also not appropriate for patients with histories of hemolytic anemia or interstitial nephritis, because these reactions depend upon IgG antibodies and cannot be "desensitized" by this method.

Safety issues — It is strongly recommended that desensitizations be performed by allergy specialists because experience is essential in managing breakthrough symptoms and adjusting the protocol appropriately. The main risk during a desensitization procedure is that of a recurrent immediate reaction. Most reactions are mild and less severe than the patient's initial reaction, although significant reactions, including anaphylaxis, do occur.

In experienced hands, desensitization is safe and effective. Desensitizations are usually performed in the hospital and often in the intensive care setting because it is critical to have the staffing and equipment available to manage anaphylaxis. However, oral desensitization to penicillin has been performed by allergy specialists in the outpatient setting as well, with one-on-one nursing and intravenous access. Penicillin desensitization has also been safely performed in pregnant women [86,87]. We know of no published fatalities resulting from failed penicillin desensitization procedures. (See "Rapid drug desensitization for immediate hypersensitivity reactions", section on 'Safety' and "Syphilis in pregnancy", section on 'Desensitization'.)

Post-desensitization care — The effects of drug desensitization for immediate reactions are temporary. After the drug is cleared from the body, the patient's allergy returns, and continued avoidance is essential. If a penicillin is needed again, it must be given by desensitization (if it is needed in the near future) or if years elapse before it is needed, the patient should be re-evaluated to determine if the allergy is still present. It is critical to ensure that the patient understands this.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Drug allergy and hypersensitivity".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Beyond the Basics topic (see "Patient education: Allergy to penicillin and related antibiotics (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●Allergy to penicillin is the most commonly reported medication allergy, although more than 90 percent of patients who report allergic reactions to penicillins either were never allergic or had an earlier allergy that resolved. (See 'Introduction' above.)

●Allergic reactions to penicillins may be categorized based on the time of onset of symptoms. Reactions that begin within minutes to one hour of administration are classified as "immediate." Although there are exceptions, this definition is intended to capture the majority of type I, immunoglobulin (Ig)E-mediated reactions. (See 'Definition' above.)

●Signs and symptoms of immediate reactions include flushing, pruritus, urticaria, angioedema, bronchospasm, laryngeal edema, gastrointestinal symptoms, and hypotension. An array of symptoms may be seen (table 1). (See 'Clinical manifestations' above.)

●The allergenic components of penicillins are found either in the core ring structure that is common to all penicillins (as well as cephalosporins and carbapenems) or (much less commonly) within side chains (called R groups) that distinguish different penicillins from one another. The portion of the molecule to which the patient has made IgE antibodies determines patterns of cross-reactivity with other related drugs. A detailed understanding of the structure of these drugs is required to interpret skin test results and predict cross-reactivity patterns. (See 'Pathogenesis' above.)

●The diagnosis of immediate allergic sensitivity to penicillin is based on clinical history, skin testing when available, and sometimes graded challenge (algorithm 1) (see 'Evaluation and diagnosis' above):

•The time elapsed since the patient's last reaction is important in assessing the risk of a recurrent reaction, as patients with IgE-mediated allergy may lose the sensitivity over time if they are not re-exposed. Only 20 percent of patients with confirmed IgE-mediated penicillin allergy remain allergic after 10 years of avoidance. (See 'Clinical history' above.)

•Penicillin skin testing is the preferred method of evaluating patients with possible immediate reactions because it is rapid, sensitive, and associated with minimal patient discomfort, although referral is needed because this type of testing is usually performed by allergy specialists. (See 'Penicillin skin testing' above.)

•In vitro tests, such as immunoassays for penicillin-specific IgE or basophil activation tests, are inferior to skin testing and are of limited value. (See 'In vitro testing and limitations' above.)

•If skin testing is not available and an equivalent alternative antimicrobial does not exist, patients with a low likelihood of allergy based upon clinical history, time elapsed since the reaction, and possible in vitro testing may be diagnosed via graded challenge. Patients who tolerate graded challenge prove that they are not acutely allergic. Challenges should be performed in a setting equipped with the staff and resources to treat possible allergic reactions, including anaphylaxis. (See 'Graded challenge (test dosing)' above.)

●For patients with past immediate reactions who are found to remain sensitive, either by skin testing or challenge, the options for future treatment include use of unrelated antibiotics (ie, avoidance of penicillins) or treatment with penicillins via desensitization. (See 'Options for future treatment' above.)