Residency · Residency · Allergy Immunology

Penicillin Allergy Evaluation and Delabeling

The Penicillin Allergy Problem

Scope

Approximately 10 percent of the US population, representing roughly 32 million individuals, carries a penicillin allergy label in their medical record. However, when these patients undergo formal allergologic evaluation, more than 90 to 95 percent are found to tolerate penicillin without adverse effect. This extraordinary discrepancy arises from multiple sources of over-labeling: non-allergic adverse drug reactions such as gastrointestinal upset or headache mischaracterized as allergy, childhood viral exanthems temporally coincident with antibiotic administration that are mistakenly attributed to the drug, family history of penicillin allergy misinterpreted as a personal allergy, and outdated or vague documentation perpetuated across medical encounters. Furthermore, penicillin-specific IgE naturally wanes over time, with approximately 80 percent of patients losing detectable IgE sensitization over a 10-year period and roughly 50 percent within 5 years.

Consequences of Unverified Penicillin Allergy Label

The downstream consequences of an unverified penicillin allergy label are substantial and far-reaching. Clinicians default to broader-spectrum alternative antibiotics, including vancomycin, fluoroquinolones, aztreonam, and clindamycin, which are associated with higher rates of Clostridioides difficile infection (odds ratio 1.26 to 1.69), increased selection pressure for methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant Enterococcus (VRE), and increased surgical site infections (odds ratio 1.5 when cefazolin is avoided for perioperative prophylaxis). These inferior alternatives translate to longer hospital stays, higher healthcare costs (estimated at approximately 64 million dollars per year in attributable costs in one institutional study), and suboptimal clinical outcomes, as alternatives are often less effective than first-line beta-lactam therapy. A prominent example is the use of vancomycin in place of cefazolin for surgical prophylaxis against methicillin-susceptible Staphylococcus aureus, where cefazolin is clearly the superior agent.

Penicillin Structure and Cross-Reactivity

Beta-Lactam Ring Structure

All beta-lactam antibiotics share a common four-membered beta-lactam ring as their core structural element, though they differ in the ring systems fused to this core. Penicillins possess a thiazolidine ring fused to the beta-lactam ring, cephalosporins have a dihydrothiazine ring, carbapenems feature a five-membered ring with a carbon replacing the sulfur, and monobactams contain a monocyclic beta-lactam ring without a fused ring system. Crucially, modern understanding has established that IgE-mediated cross-reactivity between beta-lactam classes is driven primarily by the similarity of their R-group side chains rather than by the shared core beta-lactam ring itself.

Penicillin Allergenic Determinants

When the beta-lactam ring of penicillin opens, the reactive intermediate binds to lysine residues on carrier proteins, forming hapten-carrier conjugates that can stimulate immune responses. The major determinant, penicilloyl (benzylpenicilloyl), accounts for approximately 95 percent of haptenation events and is detected by the commercially available reagent penicilloyl-polylysine (marketed as Pre-Pen). The minor determinants, including penicilloate, penilloate, penicillin G itself, and penaldehyde, are quantitatively less abundant but are more strongly associated with anaphylaxis than the major determinant. A minor determinant mix (MDM) is commercially available in some countries, while in the United States, penicillin G is used as a surrogate for minor determinant testing.

Amoxicillin possesses a unique side chain that can independently cause sensitization, giving rise to selective amoxicillin allergy without cross-reactivity to penicillin G. This phenotype is well described, particularly in Southern European populations, and provides the rationale for including amoxicillin as a separate reagent in the skin testing panel.

Cross-Reactivity Rates (Updated)

Contemporary data have significantly revised the historically cited cross-reactivity rates between beta-lactam classes. The cross-reactivity rate between penicillins and cephalosporins is approximately 2 percent, far lower than the widely quoted 10 percent figure derived from older studies confounded by contamination of early cephalosporin preparations with trace penicillin. This cross-reactivity is driven by similarity of the R1 side chain rather than the core ring structure. First-generation cephalosporins such as cephalothin, cephaloridine, and cephalexin share an R1 side chain similar to penicillin and amoxicillin, resulting in higher cross-reactivity potential. Third- and fourth-generation cephalosporins with structurally dissimilar side chains carry negligible cross-reactivity risk, estimated at less than 0.5 percent.

Cross-reactivity between penicillin and carbapenems is similarly low, at 0.9 to 1 percent, having been substantially overestimated in earlier literature. Aztreonam, the sole clinically available monobactam, exhibits minimal cross-reactivity with penicillin, with one important exception: it shares an identical R1 side chain with ceftazidime, creating cross-reactivity between these two specific agents. The shared R1 side chain between amoxicillin and cefadroxil represents another clinically relevant cross-reactive pair.

<image>A chemical structure diagram showing the beta-lactam antibiotic family. Center: core beta-lactam ring structure. Four branches radiating out: (1) Penicillins (thiazolidine ring + beta-lactam): show penicillin G, amoxicillin, ampicillin with their R1 side chains highlighted in red. (2) Cephalosporins (dihydrothiazine ring + beta-lactam): show first-gen (cephalexin - note same R1 as ampicillin), third-gen (ceftriaxone - different R1), and ceftazidime. (3) Carbapenems (five-membered ring with carbon replacing sulfur): meropenem structure. (4) Monobactams (monocyclic beta-lactam only): aztreonam with R1 identical to ceftazidime highlighted. Dotted lines connect drugs sharing identical R1 side chains to indicate cross-reactivity risk. A legend box: "Cross-reactivity driven primarily by R1 side chain similarity, NOT the core beta-lactam ring."</image>

Penicillin Allergy Evaluation

Risk Stratification by History

A detailed clinical history is the essential first step in penicillin allergy evaluation and allows risk stratification into low, moderate, and high categories. Low-risk patients, who account for the majority of those carrying penicillin allergy labels, include those with remote reactions occurring more than 10 years ago, vague or poorly characterized symptoms, family history documented as personal allergy, isolated gastrointestinal symptoms, headache, yeast infection, or an uncharacterized childhood rash occurring during an intercurrent illness. The negative predictive value of a low-risk history alone is estimated at 96 to 99 percent.

Moderate-risk patients include those who describe urticaria, pruritus, or other well-characterized immediate symptoms developing within hours of penicillin exposure, reactions occurring within the past 5 to 10 years, and reactions that required medical treatment. High-risk patients include those with a history of anaphylaxis (hypotension, laryngeal edema, or loss of consciousness), recent reactions within 1 to 5 years, or a previously documented positive skin test. Patients with a history of SJS/TEN or DRESS attributed to penicillin represent an absolute contraindication to re-challenge or skin testing and must permanently avoid penicillins and related agents.

Penicillin Skin Testing (PST)

Penicillin skin testing is a well-validated diagnostic tool employing three principal reagents. The major determinant is tested using commercially available benzylpenicilloyl-polylysine (Pre-Pen). The minor determinant is tested using freshly prepared penicillin G at a concentration of 10,000 units/mL, with the minor determinant mix used where available. Amoxicillin at a concentration of 2 to 25 mg/mL is included as a third reagent because of the possibility of selective amoxicillin sensitization.

The testing technique follows a standardized sequential protocol. Prick/puncture testing is performed first using histamine (1 mg/mL) as the positive control and saline as the negative control. If prick testing is negative, intradermal testing follows, injecting 0.02 to 0.05 mL of each reagent intradermally. Results are read at 15 to 20 minutes, with a positive result defined as a wheal at least 3 mm greater than the negative control. The negative predictive value of a complete penicillin skin testing panel (with subsequent oral amoxicillin challenge) is 97 to 99 percent for immediate reactions. Sensitivity is estimated at 50 to 70 percent, limited by the inability to test all minor determinants and by the natural waning of IgE over time. Testing must be performed by trained personnel with resuscitation equipment immediately available.

Oral Amoxicillin Challenge

Following negative skin testing, an oral amoxicillin challenge is performed to confirm clinical tolerance. The standard approach involves administration of a single full therapeutic dose of amoxicillin (250 to 500 mg) with a 1-hour observation period. An alternative two-step graded challenge protocol administers one-tenth of the dose followed by the full dose, with 30 to 60 minutes between steps. Extended observation may be considered for patients with a history of delayed reactions. The reaction rate following negative penicillin skin testing and oral challenge is less than 2 percent, confirming the high negative predictive value of the combined testing approach.

Direct Oral Challenge (Without Skin Testing)

An increasingly adopted approach for low-risk patients is the direct oral amoxicillin challenge without preceding skin testing, supported by multiple validation studies. The PEN-FAST tool, developed and validated by Trubiano and colleagues, provides a structured scoring system for risk stratification. Points are assigned based on the phenotype of the reported reaction (anaphylaxis or angioedema receives 2 or 3 points, urticaria or other symptoms receives 1 point, and unknown reactions receive 0 points), the time elapsed since the reaction (5 years or less receives 2 points, 6 to 10 years receives 1 point, and more than 10 years receives 0 points), and whether treatment was required (yes receives 1 point, no receives 0 points). A PEN-FAST score below 3 identifies patients who can safely proceed to direct oral amoxicillin challenge. A score of 3 or higher indicates the need for skin testing before challenge. The landmark PALACE study, a multisite randomized controlled trial, demonstrated equivalent safety between direct oral amoxicillin challenge and the traditional pathway of skin testing followed by oral challenge in low-risk patients.

<image>A clinical decision algorithm for penicillin allergy evaluation. Starting box: "Patient with penicillin allergy label." Step 1: History assessment with three branches. LOW RISK (vague history, >10 years ago, GI symptoms only, PEN-FAST <3): proceed directly to single-dose oral amoxicillin challenge with 1h observation. MODERATE RISK (urticaria, well-described immediate symptoms, PEN-FAST >=3): perform penicillin skin testing (Pre-Pen + penicillin G + amoxicillin), then if negative proceed to oral amoxicillin challenge. HIGH RISK (anaphylaxis, recent reaction, positive skin test): options include avoidance with alternatives, or penicillin desensitization if drug essential. RED STOP SIGN: History of SJS/TEN, DRESS, AGEP, interstitial nephritis, hemolytic anemia to penicillin - DO NOT test or challenge; permanently avoid and use non-cross-reactive alternatives. Final outcome box: "Negative evaluation - REMOVE penicillin allergy label from medical record, prescribe penicillin/amoxicillin to fill."</image>

De-Labeling Process

Documentation

Successful de-labeling requires meticulous documentation and reinforcement. The allergy section of the electronic medical record should be updated to reflect the evaluation outcome, changing the entry from "penicillin allergy" to a notation such as "penicillin allergy evaluated - tolerated amoxicillin [date], allergy label removed." Prescribing a penicillin-class antibiotic for the patient to fill, such as amoxicillin, serves to reinforce tolerance and anchor the negative evaluation in the dispensing record. Patients should receive written documentation of their negative evaluation, and the primary care provider should be notified of the de-labeling.

Institutional Programs

Institutional penicillin allergy de-labeling programs have demonstrated significant impact, reducing broad-spectrum antibiotic use by 25 to 40 percent. These programs have been effectively implemented through pharmacist-led and nurse-led protocols in inpatient settings, extending the reach of de-labeling beyond the allergy clinic. Point-of-care penicillin skin testing in preoperative clinics allows timely de-labeling before surgery, enabling the use of cefazolin for perioperative prophylaxis. Integration of penicillin allergy status review into routine antimicrobial stewardship activities represents a high-value, systems-level intervention.

Cephalosporin Allergy Evaluation

For patients with a history of penicillin allergy who have undergone successful evaluation and de-labeling, any cephalosporin can be administered without concern. For patients with a reported cephalosporin allergy, evaluation should focus on R1 side chain cross-reactivity patterns. If the index cephalosporin (the one that caused the reaction) shares an R1 side chain with the desired cephalosporin, the cross-reactivity risk is higher, and skin testing followed by graded challenge is appropriate. If the R1 side chains differ, a graded challenge is typically sufficient. Cephalosporin skin testing employs prick followed by intradermal testing, though standardization is limited and non-irritating concentrations vary by agent. The standard graded challenge protocol administers one-tenth followed by the full dose, separated by 30 to 60 minutes.

Carbapenem and Monobactam Use

Carbapenems, including meropenem, imipenem, ertapenem, and doripenem, can be administered safely in the vast majority of penicillin-allergic patients, as the true cross-reactivity rate is less than 1 percent. Skin testing and/or graded challenge may be considered when the reaction history is particularly concerning. Aztreonam, the only available monobactam, has no clinically significant cross-reactivity with penicillin and can be administered safely to penicillin-allergic patients without prior testing. However, aztreonam does cross-react with ceftazidime due to their identical R1 side chains, and this specific cross-reactive pair must be recognized.

Key Clinical Pearls

• Over 90% of patients labeled penicillin-allergic tolerate the drug after proper evaluation
• Penicillin allergy labels increase C. difficile, MRSA, VRE, surgical site infections, and healthcare costs
• Cross-reactivity between penicillins and cephalosporins is driven by R1 side chain similarity (~2% overall, NOT the outdated 10%)
• PEN-FAST score <3 identifies low-risk patients who can safely undergo direct oral amoxicillin challenge without prior skin testing
• History of SJS/TEN, DRESS, AGEP, or organ-specific reactions to penicillin is an absolute contraindication to skin testing or oral challenge
• Aztreonam is safe in penicillin allergy (no cross-reactivity) but cross-reacts with ceftazidime (same R1 side chain)
• After successful de-labeling, update the EMR, prescribe a penicillin-class antibiotic, and provide written documentation to the patient
• Penicillin allergy de-labeling is a high-value antimicrobial stewardship intervention

References

1. Shenoy ES, et al. Evaluation and management of penicillin allergy: a review. JAMA. 2019;321(2):188-199.
2. Blumenthal KG, et al. Impact of a clinical guideline for prescribing antibiotics to inpatients reporting penicillin or cephalosporin allergy. Ann Allergy Asthma Immunol. 2015;115(4):294-300.
3. Trubiano JA, et al. Development and validation of a penicillin allergy clinical decision rule. JAMA Intern Med. 2020;180(5):745-752.
4. Stone CA Jr, et al. Penicillin allergy testing and drug challenge outcomes in a large health system. J Allergy Clin Immunol Pract. 2021;9(7):2765-2772.
5. Joint Task Force on Practice Parameters (JTFPP). Drug allergy: an updated practice parameter. Ann Allergy Asthma Immunol. 2010;105(4):259-273.

Read this lecture as Markdown