Direct oral challenges safely reduce the burden of low-risk cephalosporin allergy labels

Approximately 1% to 2% of the population reports a cephalosporin allergy label (CAL).¹ However, many of these labels are misclassified as allergies, and only approximately 8% of CALs correlate with positive skin testing.¹ Furthermore, individuals with true prior IgE-mediated reactions, estimated to be 10%/year or higher, may have loss of hypersensitivity over time and are not truly allergic when tested and rechallenged.² Carrying a label of cephalosporin allergy is consequential and often leads to treatment with alternative agents with inferior outcomes, infections owing to multidrug-resistant organisms, and higher health care–related costs.^3,4 Current literature supports risk stratification of penicillin allergy labels and has described successful delabeling of low-risk labels via direct oral challenge (DOC).⁵ Koo et al⁶ described that a validated history-based risk stratification approach to penicillin allergy labels is applicable to CALs, with the exception that cephalosporins administered intravenously are considered of higher risk owing to higher pretest probability for positive testing for IgE-mediated allergy. Although evidence has supported the safety of delabeling children through DOCs to cephalosporins, evidence to support generalizability to adults is lacking.^7,8 In a group of adult outpatients with low-risk CALs, we evaluated the safety and efficacy of removing CAL using DOCs.

From July 2022 to April 2024, 239 sequential adult patients with a CAL were referred for outpatient drug allergy assessment at Vanderbilt University Medical Center. They underwent history-based risk assessment to exclude moderate to severe allergic reactions and were categorized as low-risk or non–low risk (high or highest risk). Patients were potentially eligible for a DOC if the CAL was low-risk.⁶ The following criteria defined a low-risk history either alone or in combination: (1) isolated urticaria and/or angioedema more than 5 years ago, (2) self-limited cutaneous rash at any point in the past without organ involvement or signs of a delayed rash, (3) only gastrointestinal symptoms, (4) remote childhood reaction with limited details (little information could be recalled about the index reaction), (5) a family history only of cephalosporin allergy, (6) avoidance only from fear of allergy, (7) known tolerance of a cephalosporin since the original reaction occurred, or (8) other non-allergy symptoms.⁶ In terms of risk assessment, if a patient met multiple low-risk criteria, this was still considered low-risk and was counted that way, as listed in Table I. Patients were classified as high risk based on a cephalosporin reaction history including any of the following: anaphylaxis within the past 5 years; delayed urticarial eruptions (>1 hour after drug administration) within the past 5 years; within 1 hour of cephalosporin administration, development of diffuse hives, angioedema, shortness of breath, wheezing and/or coughing, hypotension, loss of consciousness, and diarrhea and/or vomiting. Patients were classified as highest risk and excluded from consideration of DOC based on a cephalosporin reaction history including any of the following: Stevens-Johnson syndrome, toxic epidermal necrolysis, drug reaction with eosinophilia and systemic symptoms, acute generalized exanthematous pustulosis, immune-mediated kidney injury, immune-mediated liver injury, and development of delayed skin/mucosal sloughing or mucosal ulcerations. Patients with allergy labels to intravenous cephalosporins were excluded from low-risk classification and classified as ineligible for direct challenge owing to a higher pretest probability for true allergy observed in validation studies.⁶ Patients with low-risk CALs were presented with one of two approaches: skin testing,¹ followed by oral challenge (OC); or DOC. Skin testing concentrations and challenge doses used were selected according to previously published testing strategies for cephalosporin evaluation cited in the Practice Parameters.¹ All low-risk patients were offered DOC, and the decision to precede OC with skin testing was personalized based on patient preference. Oral challenges (either direct or after negative skin testing) were performed with single-dose oral cephalosporin (cephalexin 250 mg, cefdinir 300 mg, or cefuroxime 250 mg), selected based on structural R1-side chain similarity to the index drug, followed by 90 minutes of observation. After risk stratification and shared patient decision-making, we assessed the proportion of patients accepting and tolerating DOC. This study was done under institutional review board–approved protocols from Vanderbilt University (Nos. 161455, 181180, and 231087).

TABLE I.

Characteristics of low-risk cephalosporin allergy-labeled patients who underwent direct oral challenge vs oral challenge after negative skin testing

Characteristic	Direct oral challenge (n = 71)	Negative skin test plus oral challenge (n = 62)
Median age, y (interquartile range)	57 (41–71)	55 (32–68)
Sex (%)
Female	56 (78.9)	56 (90.3)
Male	15 (21.1)	6 (9.7)
Race (%)
Asian	1 (1.4)	0
Black	2 (2.8)	1 (1.6)
Unknown	9 (12.7)	6 (9.7)
White	59 (83.1)	55 (88.7)
Non-Hispanic or Latino ethnicity (%)	71 (100)	59 (95.2)
Drug allergy label (%)
Cephalexin	39 (54.9)	26 (41.9)
Cefaclor	12 (16.9)	4 (6.5)
Cefdinir	4 (5.7)	5 (8.1)
Multiple oral cephalosporins	3 (4.2)	7 (11.3)
Cefprozil	2 (2.8)	3 (4.8)
Cefuroxime	0	4 (6.5)
Cefixime	0	2 (3.2)
Unknown cephalosporin	11 (15.5)	11 (17.7)
Drug reaction history (%)
Childhood reaction	21 (29.5)	15 (24.2)
Rash (unspecified)	17 (23.9)	17 (27.4)
Urticaria	13 (18.3)	17 (27.4)
Other (non—IgE mediated symptoms)	13 (18.3)	13 (21.0)
Unknown	11 (15.5)	6 (9.7)
Rash (maculopapular)	10 (14.1)	7 (11.3)
Pruritus (without rash)	6 (8.5)	5 (8.1)
Angioedema	5 (7.0)	7 (11.3)
Nausea or vomiting	3 (4.2)	2 (3.2)
Fever	2 (2.8)	1 (1.6)
Flushing	2 (2.8)	0
Shortness of breath	2 (2.8)	6 (9.7)
Throat tightness (subjective)	2 (2.8)	0
Bronchospasm	1 (1.4)	0
Drug challenged (%)
Cephalexin	68 (95.8)	52 (83.8)
Cefdinir	2 (2.8)	5 (8.1)
Cefuroxime	1 (1.4)	5 (8.1)
Outcome of challenge (%)
Passed	70 (98.6)	62 (100)
Failed	1 (1.4)	0
Has subsequently taken and tolerated cephalosporin treatment after cephalosporin allergy label removal (%)	15 (21.1)	17 (27.4)
Cephalosporin allergy label remained removed from chart (%)	67 (94.4)	56 (90.3)

Of 239 patients with CALs, 152 (63.6%) were risk-stratified as low risk whereas 87 (36.4%) were risk-stratified as non–low risk (high or highest risk) (Figure 1). Of the 87 non–low risk patients, 77 (88.5%) were classified as high risk and 10 (11.5%) were classified as highest risk. Of the 77 high-risk patients, 14 (18.2%) were classified based on drug reaction history, 38 (49.3%) were classified based on intravenous CAL, and 25 (32.5%) were classified based on both drug reaction history and intravenous CAL. Of the 10 highest-risk patients, five received the diagnosis of drug reaction with eosinophilia and systemic symptoms, three had delayed skin sloughing, one had acute generalized exanthematous pustulosis, and one had Stevens-Johnson syndrome.

FIGURE 1.

Stratification of cephalosporin allergy-labeled patients and outcomes of oral challenges.

Among those stratified as low risk, 71 of 152 (46.7%) agreed to DOC whereas 81 of 152 (53.3%) had skin testing before considering OC (Figure 1). Of the 81 patients who underwent skin testing, 62 (76.5%) proceeded with OC whereas 19 (23.5%) did not undergo OC owing to a loss to follow-up (Figure 1). From the cohort of 71 patients who underwent DOC, 56 (78.9%) were female, 59 (83.1%) self-reported as White, and median age was 57 years (Table I). The most common drug allergy labels were cephalexin (39 of 71 [54.9%]), cefaclor (12 of 71 [16.9%]), and an unknown cephalosporin (11 of 71 [15.5%]) (Table I). The most common reported adverse drug reactions were childhood reaction (21 of 71 [29.5%]), unspecified rash (17 of 71 [23.9%]), and urticaria (13 of 71 [18.3%]) (Table I). Moreover, of the 71 patients, 68 (95.8%) were challenged with cephalexin, two (2.8%) with cefdinir, and one (1.4%) with cefuroxime (Table I). Of the 71 patients who underwent DOC, 70 (98.6%) passed and were successfully delabeled. Subsequently, 15 of 71 (21.1%) patients tolerated a known cephalosporin treatment after CAL removal (see Table E1 in this article’s Online Repository at www.jaci-inpractice.org) and 67 of 71 (94.4%) maintained CAL removal from the medical record (Table I) with a median follow-up of 15.9 months (interquartile range, 11.8–19.7). In terms of adverse events, one of the 71 patients (1.4%) developed pruritus of the neck and eyes approximately 40 minutes after ingestion of cefuroxime, which resolved with antihistamines. This reaction was similar to the index reaction and the patient was not delabeled. Compared with low-risk patients who preferred skin testing before OC, we observed no positive skin tests, and all 62 patients (100%) tolerated OC (Table I). In terms of safety, no cases of cephalosporin anaphylaxis or other adverse drug reactions related to cephalosporins were observed in low-risk patients with CAL during challenge or subsequent treatments.

A point-of-care DOC to an oral cephalosporin with structural similarity to the index drug is a safe and effective delabeling strategy in those with low-risk CALs. Delabeling low-risk CALs without the need for skin testing is cost-effective and less time-consuming.⁹ Furthermore, because common CALs are associated with cephalosporins with R1 side chains that are not available in a parenteral form for skin testing (eg, cephalexin) this approach makes mechanistic sense. Additionally, cephalosporin antibiotics are widely used in clinical practice, and delabeling unnecessary CALs avoids the need for treatment with alternative agents. Although there is a possibility of selection bias in low-risk patients who were not opposed to going straight to DOC when offered, it was reassuring to us that the safety profile of both testing approaches, DOC versus skin testing followed by OC, was essentially the same. Future studies should consider a randomized controlled trial design.

Limitations of our study included that data collection for subsequent cephalosporin treatments was obtained from the Vanderbilt University Medical Center electronic health record from July 6, 2022 to December 5, 2024. Therefore, we cannot comment on additional cephalosporin treatments received outside our health care system. This study included small sample sizes, and there is the possibility that large sample sizes might detect subtle differences between the described groups. Approximately one fourth of the low-risk patients who had negative skin testing were lost to follow-up and did not undergo cephalosporin OCs. Therefore, there may have been a possible selection bias in those who underwent OCs. There is also the possibility of selection bias in patients who chose to undergo DOC versus skin testing followed by OC, because the choice was patient preference. We used a history-based risk stratification approach to categorize low-risk CALs, and therefore there may be limited external validity if a different approach is used. Our study was performed in an outpatient drug allergy clinic and did not investigate challenging low-risk CALs in the inpatient setting. Additionally, our study was performed in North America, which has different use patterns of oral versus parenteral aminopenicillins and thus may have different cephalosporin sensitization patterns compared with Europe or Australia.¹ Direct oral challenges are a safe and efficacious strategy for delabeling low-risk CALs in the outpatient setting and were acceptable for around half of patients when it was offered in this small cohort. Our study supports the possibility of future prospective and randomized studies investigating delabeling low-risk CALs in other practice settings to optimize personal and public health outcomes.

Clinical Implications.

Direct oral challenges are a safe and effective strategy for delabeling low-risk cephalosporin allergy labels in adults in the outpatient setting.