• Summary
  Click here to print a Quick Reference Guide
• Background
• Clinical Diagnosis

Back to top

This document is a guide and should be used as a tool to assist healthcare workers faced with patients claiming an allergy to penicillin in making safe decisions. If there is doubt, the healthcare workers should err on the side of caution and avoid penicillin-containing antibiotics. If unclear, the healthcare worker should escalate the decision to the senior member of the medical team who can discuss the decision with a Microbiologist.

Allergic reactions to B-Lactam antibiotics are the most frequently encountered immunologically mediated adverse drug reaction ³. Up to 17% of the population report an allergic reaction to penicillin the nature and severity of which may be difficult to determine 5. Studies have shown patients labelled as penicillin allergic are more likely to receive drugs that are more costly, broader spectrum and have the potential for more adverse reactions (including Clostridium difficile infection) and induction of antimicrobial resistance ^{5 7}. Treatment with such agents may be inferior to their B-lactam alternative and associated with greater treatment failure. The in-patient antibiotic cost of managing a patient with an allergy to penicillin is reported to be 63% higher than that of a patient who can tolerate penicillin ⁸. Only 10% of those alleging allergy to penicillin have a genuine allergy ⁹ with 50% of reactions being non-immune modulated or side effects of the drug (for example rash when taking amoxicillin for viral infection, diarrhoea and vomiting etc.) ⁵.
Assessment of the reaction is essential to determining the nature of the reaction and informing future antibiotic choices.

This guideline will instruct on how to make that assessment and discuss the management options available following such an assessment.

Basic Immunology of Penicillin Allergy

Understanding the classification systems of allergic reaction, their underlying mechanisms and clinical presentations aids the healthcare provider in assessing the likelihood and nature of a reaction to a given antibiotic when faced with a patient reporting an allergy. It informs the healthcare provider when making the choice of antibiotic in order to treat the infection most effectively by the safest means.

Allergic adverse drug reactions may be classified by immune mechanism 10 or by time of onset of reaction ¹¹.

Immediate or Type 1 reactions are often associated with systemic manifestations of anaphylaxis (such as diffuse urticaria/wheals, erythema, pruritis, hypotension, cardiac dysrhythmias, bronchospasm, laryngeal oedema, angioedema either alone or in combination). Anaphylactic reactions occur in 4-15 of every 100,000 courses of penicillin and are seen predominantly in the young adult age group (ages 20-49 years). Atopy does not predispose to anaphylaxis but may result in a more severe reaction ¹². There is generally no hereditary mechanism for anaphylaxis to penicillin and as such a family history should not influence the clinician.

Type I reactions result from penicillin or its metabolites binding with preformed specific IgE antibodies bound to tissue mast cells and/or circulating basophils. This complex formation causes the release (de-granulation) of the inflammatory mediators, including histamine, contained within these cells.

Reactions to penicillin occurring between 1-72 hours are known as accelerated reactions and may also be IgE mediated. These reactions are also characterized by urticaria, laryngeal oedema, angioedema and/or bronchospasm but rarely life threatening.

Late reactions occur more than 72 hours after administration of the drug and are rarely life threatening. The mechanism behind such reactions is heterogeneous. The most common reaction is a rash. The combined frequency of rash to penicillin has been estimated at 1-4% of treatment courses with maculo-papular or morbiliform rash being then most common occurring in 5.2-9.5% of treatment courses of ampicillin. These are usually symmetrical, confluent macules or papules. They generally originate on the extremities of ambulatory patients and tend to spare the soles and palms. In non-ambulatory patients the pressure areas are frequently affected.

Late reactions are not assessable by Skin Testing as they are not mediated by IgE antibodies ¹².

The immunochemistry of β-lactam antibiotics is complex and has not been fully elucidated. Penicillins comprise a β-lactam ring, to which it owes its antimicrobial activity, and a 5-membered thiazolidine ring. On degradation the β-lactam ring is opened and binds avidly to protein and the thiazolidine ring is preserved.

In contrast cephalosporin antibiotics have their 5-membered thiazolidine ring replaced with a 6-membered dihydrothiazine ring and undergo fragmentation of both the β-lactam and dihydrothiazine rings on degradation. It has been postulated that these differences account for the differences in immunogenicity ⁶.

Side chain similarity is now thought to relate more closely with cross- reactivity of penicillin and cephalosporins: the closer the side chain homology the greater the probability of cross reaction although both side chain and nuclear haptens may be involved ^{6 13 14}. First generation cephalosporins (such as cephalexin, cephradine) have a much greater probability of cross reaction in patients with a history of penicillin allergy. This is discussed in more detail in the section on Cross Reactivity.

Back to top

Clinical Assessment of a patient giving a History of Penicillin Allergy

The clinician assessing a patient giving a history of allergy to penicillin (or any other antibiotic) should attempt to define the type of reaction ^{12 15-19}. Information should be sought from the patient or patients’ carers, the medical notes and the patients General Practitioner. The aim is to determine the nature of the adverse reaction in order to assess the risk of further exposure to the agent or related drugs. Published tools exist for the assessment of drug hypersensitivity ²⁰ however, as most patients report reactions in the past, a detailed assessment of the reaction can be difficult.

The assessment questions in Box 1 should assist in determining the nature of the reaction.

Management of the patient with a penicillin allergy is dependent on their clinical requirements. The questions below should be answered

• What type of allergy does the patient have (type 1, unknown or non-type 1) (see section “Clinical Assessment of a patient giving a History of Penicillin Allergy”
• Does the patient require an antibiotic at this time?
• If yes, according to Trust Antibiotic Guidelines and the assessment of the patients’ allergy can the patient take the first line antibiotic without risk (See Box 2 and Box 3)?
• Is it essential the patient has this antibiotic or does an alternative with comparable efficacy exist (a discussion with a Microbiologist or Infectious Disease Physician may be required to determine this)?

This risk assessment should be made by the most senior member of the clinical team and should be documented in the patients notes. It is important to note that the balance of risks may change for each infection and should be repeated before every course of antibiotic is initiated.

Cross-reactivity of other classes of antimicrobials in patients with penicillin allergy

Carbapenems (meropenem, imipenem, ertapenem)

Historically, there was believed to be a cross reactivity rate of almost 50% with carbapenems in patients with penicillin allergy. This was based on a single study with few patients ²¹. More recent prospective studies have shown the rate to be much lower: When skin testing patients with confirmed penicillin allergy and a clinical history of type 1 reaction to penicillin, only 1% of patients reacted to imipenem and meropenem ^22-25. Retrospective studies reviewing the case notes of patients reported to have allergy to penicillin found rates of cross-reactivity of up to 9.8% for meropenem and 12.2% for imipenem ^{26 27} when including all reaction types. This implies a cross reaction rate of 1% when using carbapenems in patients who have previously suffered IgE-mediated reactions to penicillins. The risk of a life threatening reaction in patients who have suffered non-IgE mediated reaction to penicillin should be no greater than that of the normal population.

Monobactams (aztreonam)

Patients who have experienced a type 1 hypersensitivity reaction to ceftazidime should avoid aztreonam due to the risk of cross-reactivity due to side chain similarity of the two drugs ^{28 29}.

Studies in the non-Cystic Fibrosis patient group found evidence of cross-reactive skin tests (3.4%) but no evidence of clinical cross-reactivity to aztreonam in the patients with skin test confirmed penicillin allergy when challenged with the drug ^{28 30-32}.

In studies looking at Cystic Fibrosis patients, skin test cross reactivity was demonstrated in 5.3% of patients with clinical evidence of cross reactivity on challenge ^{29 33 34}. The research did not look specifically at penicillin allergy and ceftazidime cross reactivity may account for a proportion of these cases. The findings in the Cystic Fibrosis patient group cannot be extrapolated to the general population as there is greater degree of hypersensitivity and sensitization to antibiotics in this group than in the general population. This degree of sensitization is demonstrated in the studies.

Cephalosporins

As discussed earlier, the commonly held belief that 10% of patients with allergy to penicillin will cross react when given cephalosporins is a myth. This myth is based on historical data and it is hypothesized that an explanation could be that cephalosporins and penicillins were manufactured on the same fermentation production line; hence some cephalosporins contained trace amounts of penicillin.

The true rate of cross reaction is dependent on the generation of cephalosporin. For first generation cephalosporins (cefalexin in this case), patients reporting allergy to penicillins demonstrated a cross-reaction rate of 6.5-7% ^{35 36} however this rate is shown to be higher (14%) when studying patients whose allergy is confirmed by skin testing ³⁷.

Regarding second generation cephalosporins, Pichichero et. al. reviewed the relative risk of giving cefuroxime to patients who had a history of allergy to penicillin or amoxicillin compared to those who had no such history. Eight of 428 patients (1.9%) with a history and 89 of 5410 patients (1.7%) with no history reacted to cefuroxime. When the groups were separated by history of allergy confirmed by skin testing the rates were similar: 2 of 283 patients (1.6%) with and 5 of 397 patients (1.3%) without a history confirmed by skin testing reacted when given the drug. The relative risk was not significantly different for either group (-0.3%; CI -0.7-1.5% for the “history of allergy” group and -0.3%; CI -2.1-0.9% for the “history confirmed by skin testing” group)¹⁴.

Ceftazidime, a third generation cephalosporin with anti-Pseudomonal activity, showed a similar pattern. Of 538 patients giving a history of penicillin allergy, only 3 (0.06%) reacted when administered ceftazidime whereas 57 of the 3427 (1.7%) patients with no history of allergy to penicillin or amoxicillin reacted. Nil patients of a group of 30 who had their allergy confirmed by skin testing reacted to ceftazidime when challenged. Similarly there were no recorded reactions to ceftriaxone of 142 patients with history of penicillin/amoxicillin confirmed with skin testing ¹⁴.

Back to top