Spotlight commentary: De-labelling the truth: Clearing the fog around antibiotic allergy labels

Abstract

The aim of this spotlight commentary is to review what is new in the field of drug hypersensitivity reactions (DHR) with a focus on hypersensitivity to antibiotics and drug allergy de-labelling by reviewing recent articles published in this journal and situating them within the broader context of current research from other journals.¹

According to Aronson et al, an adverse drug reaction (ADR) is defined as ‘an appreciably harmful or unpleasant reaction resulting from an intervention related to the use of a medicinal product, which predicts a hazard from future administration and warrants prevention or specific treatment, alteration of the dosage regimen, or withdrawal of the product’.² DHRs are classified as type B or idiosyncratic ADRs. DHRs encompass both drug allergies (with proven immune involvement) and non-allergic hypersensitivity (without proven immune mechanisms). Unlike type A ADRs, type B ADRs have previously been considered to be dose independent and unpredictable. However, they require a certain threshold drug/metabolite concentration to trigger a reaction, and some of them can now be estimated by pharmacogenomic and in vitro testing. While type B ADRs are traditionally thought to be ‘off-target’ and unrelated to the pharmacology of the drug, some reactions, such as those to vaccines, biologics and immune checkpoint inhibitors, involve the intended target of the drug. This shows that the classification of ADRs is not completely clear-cut and that the characteristics of type A and B can overlap.^{3, 4}

The pathophysiology of DHRs is not yet well understood. The known pathophysiological models are presented and summarized in the recently published article by Elzagallaai et al.⁴ The authors attempt to explain how a low molecular weight drug molecule can trigger an immune response. In addition to the more well-known pathophysiological mechanisms of DHRs, such as the classical ‘hapten’ hypothesis, the reactive metabolite hypothesis, the pharmacological interaction with the immune system (p-i) concept, the danger/injury hypothesis and the altered peptide repertoire hypothesis, the authors introduce the inflammasome activation hypothesis and the cross-reactivity hypothesis in the article.⁴

DHRs have varied clinical presentations, with the skin being the most commonly affected organ. The prevalence of cutaneous ADRs (CADRs) varies between 1 and 3% in adult patients treated with drugs and up to 10% of patients presenting at a hospital.⁵ CADRs exhibit various clinical forms, of which 29–35 types have been reported. The most common forms include urticaria (with or without angioedema), maculopapular eruptions, fixed drug eruptions, erythema multiforme (EM) and vasculitis. Rarer but more severe forms include erythroderma, drug reaction with eosinophilia and systemic symptoms (DRESS) and Stevens–Johnson syndrome (SJS)/toxic epidermal necrolysis (TEN) spectrum, acute generalized exanthematous pustulosis and serum sickness. The frequency of these patterns varies by population, study setting and diagnostic precision.⁶

The classification, approach and treatment of CARDs are discussed in the article by Del Pozzo-Magaña et al.⁷ A lack of knowledge about CADRs often leads to over- or misdiagnosis. For example, studies show that more than 65% of cases reported as drug-induced EM do not meet diagnostic criteria, and those cases that resemble EM lack clear evidence of drug causality.⁸ Similarly, only a small percentage of patients diagnosed with SJS/TEN actually meet the clinical criteria for these conditions.⁹

Antibiotics, especially beta-lactam antibiotics, are most frequently reported as a cause of severe DHRs in adults and children, as highlighted in two articles by Del Pozzo-Magaña et al⁶ and Ramos et al,¹⁰ respectively. The authors report that beta-lactam antibiotics were responsible for 47% of DRESS cases in a retrospective study of 19 cases.⁷ In the systematic review aimed at investigating the occurrence of moderate and severe ADRs to antimicrobials in hospitalized children, penicillins, cephalosporins and sulfonamides were identified as the main antimicrobials involved.¹⁰ On the other hand, drug allergy labels (ALs) represent a significant burden on the healthcare system. In particular, beta-lactam antibiotics are frequently mislabelled as causing allergies, as they are the most commonly used first-line treatment for various infections.

Up to 10% of the general population and up to 20% of hospital patients in the United Kingdom and the United States are labelled as allergic to penicillin. However, research shows that 90–95% of these labels are incorrect when verified by comprehensive allergy testing.¹¹

The section on allergies in the medical record serves to protect patients by recording adverse reactions to medication and other substances. However, the vague and often incomplete information in these records can lead to confusion as to whether a particular drug reaction is a true immune-mediated allergy that must be strictly avoided in the future or a manageable side effect that does not preclude future use of the drug. Many entries labelled as ‘allergies’ may not involve true immune responses but are simply expected side effects or misunderstandings.

In a study by Catalano et al,¹² documentation of an allergic reaction was defined as adequate if it included information about the time to onset of the reaction after the first dose of the last course, the clinical symptoms and severity of the reaction, and/or the treatment administered, thus allowing physicians to determine the safety of re-administering the drug to the patient. Documentation was adequate in only one-quarter of recorded allergies (20/73, 27.4%). Of the 53 inadequately documented allergies, 30 (41.1%) included information on the severity but not the timing of the reaction in relation to drug exposure, 8 (10.9%) described the timing but not the severity of the reaction, and 15 (20.5%) did not include any of this information. Overall, information on the timing of the reaction was only known for 28/73 (38.5%) reactions—11 of which were immediate, occurring within 1 h of exposure. It is noteworthy that 12 of the 73 documented allergies (16.4%) were classified by the investigators as adverse or ‘on target’ effects and not as immunologically mediated allergic reactions. These included 10 with gastrointestinal symptoms, one with oral thrush. Half of all children who had penicillin, amoxicillin or cephalexin ALs were instead prescribed inappropriate antibiotics.¹²

The process of identifying and removing incorrect ALs is known as de-labelling.

It is well established that incorrect ALs, especially to beta-lactam antibiotics, lead to negative patient outcomes. An estimated 30–40% of hospitalized patients with an AL do not receive first-line antibiotics and have significantly delayed administration of the first antibiotic dose. Some of the reported negative outcomes include higher readmission rates, higher risk of Clostridioides difficile infection and acute kidney injury. ALs have also been associated with guideline deviations, increased use of broad-spectrum and IV antibiotics, poorer patient outcomes and higher antimicrobial resistance.¹³

De-labelling has recently come into focus as it is considered an important tool for antimicrobial stewardship. Several strategies have been proposed for safe and successful de-labelling. The aim is to find the fastest and easiest, but still safe, way to de-label. Gold standard for de-labelling is comprehensive allergy testing, which is most often performed in hospital settings, requiring skin tests and oral drug challenge tests often over several days as per the European Network of Drug Allergy (ENDA) recommendations.¹⁴ Since access to subspecialists to perform this testing is extremely limited and the procedures involved are time consuming, there is a need to find other simpler ways for de-labelling.

The first step is to establish risk stratification methods. Clinical decision tools have proven effective in identifying very-low- and low-risk patients who could be de-labelled alternatively.¹⁵ Some patients without a personal allergy history or those who recently tolerated penicillin may be de-labelled based solely on their history, which is known as direct delabelling, or they could be offered a direct oral challenge (DOC) without prior skin tests which could be safely done in primary care as well.¹⁶

Table 1 summarizes recently published studies focusing on efficacy and safety of alternative de-labelling strategies. The studies summarized in Table 1 demonstrate a consistently low rate of confirmed drug-induced allergies during challenge procedures, supporting the safety and effectiveness of alternative de-labelling strategies in appropriately selected, low-risk patients. Positive challenge rates were notably low, with mild reactions occurring infrequently, reinforcing the overall safety profile of direct oral challenges in clinical practice.

In conclusion, accurate documentation and appropriate management of drug hypersensitivity reactions are crucial for optimizing patient care and minimizing the burden of incorrect allergy labels. Misdiagnosed or poorly documented allergies not only hinder access to first-line therapies but also contribute to antimicrobial resistance, adverse clinical outcomes and increased healthcare costs. The process of de-labelling offers a pathway to address these challenges, with recent advances in risk stratification tools and direct drug challenges paving the way for safer, more efficient practices. By fostering collaboration between primary care providers, allergists and pharmacologists, and by implementing evidence-based strategies for de-labelling, we can ensure that patients receive the most effective and appropriate treatments, ultimately improving outcomes and advancing antimicrobial stewardship.

Both authors contributed equally to writing of the manuscript.

The authors declare no conflicts of interest.