Evaluating Pediatric Macrolide Allergy Using Direct Graded Oral Challenges

Abstract

This quality improvement study evaluates the use of direct graded oral challenge to confirm suspected macrolide allergy in children.

Macrolide prescriptions account for up to 30% of all oral antimicrobials, often replacing β-lactams in patients with reported allergies.¹ However, some individuals treated with macrolides report allergic reactions.² Most children are not seen by specialists, leading to lifelong avoidance of suspected antibiotics and others from the same drug class.³ Treatment of pediatric patients includes appropriately delabeling antibiotic allergies, ensuring selection of the most appropriate, safe, and cost-effective antibiotic.³

With no standardized skin tests for diagnosing macrolide allergy and challenge tests as the sole recommended confirmation strategy, assessing the safety of direct graded oral challenges (GOCs) is crucial.^1,4 We assessed use of direct GOCs for confirming suspected macrolide allergies in children by evaluating clinical characteristics, management, and risk factors associated with true macrolide allergy.

Methods

The β-Lactam and Other Antibiotics in Children: Tests, Assessment and Management study collected data from November 2012 to November 2023. Patients aged 18 years or younger with suspected macrolide allergy were recruited from allergy clinics at Montreal Children’s Hospital, Montréal, Québec, and Janeway Children’s Health and Rehabilitation Centre, St John’s, Newfoundland and Labrador. Ethics Committees from both hospitals approved this quality improvement study, and oral or written informed consent was obtained from parents or guardians. We followed the SQUIRE reporting guideline.

Data collection consisted of a 2-step, nonblinded direct GOC to the culprit macrolide (a 10% dose, followed by a 90% dose after 20 minutes). Patients were considered tolerant if they had no symptoms during 1-hour observation. Symptomatic patients were identified as immediate reactors and treated accordingly. Families were instructed to contact our allergy team regarding nonimmediate reaction symptoms up to 1 week after the challenge. Data on suspected macrolide reactions, defined as index macrolide reactions, including demographics, clinical reaction characteristics, and management, were collected retrospectively through a standardized questionnaire. GOC outcomes were collected prospectively after informed consent was obtained. Multivariate logistic regression analysis identified factors associated with positive GOC results (ie, reactions). Two-sided P < .05 was considered significant. Data were analyzed using R, version 4.2.0 (R Project for Statistical Computing).

Results

Among 112 patients (66 males [58.9%] and 46 females [41.1%]) with a macrolide allergy, the median (IQR) age was 5.5 years (3.0-9.3) years (Table 1). Implicated macrolides were clarithromycin in 81 patients (72.3%) and azithromycin in 31 (27.7%). Of 112 direct GOCs, most patients (90.2%) tolerated the challenge, whereas 9 (8.0%) had an immediate reaction, and 2 (1.8%) had a nonimmediate reaction. One patient experienced anaphylaxis (flushing, throat tightness, and breathing difficulties) and was treated with intramuscular epinephrine and diphenhydramine 5 minutes after the first dose of clarithromycin. All other positive GOC results were skin-limited rashes with no blisters or mucosa involvement and were classified as mild, benign skin rashes.

Table 1. Demographics of Pediatric Patients With a Suspected Macrolide Allergy and Characteristics of the Index Macrolide Reactions and Direct Graded Oral Challenge.

Variable	Patients, No. (%) (n = 112)
Site
Montréal Children’s Hospital	110 (98.2)
Janeway Children’s Health and Rehabilitation Centre	2 (1.8)
Demographics
Age, median (IQR), y
At index macrolide reaction	2.4 (1.4-6.7)
At graded oral challenge	5.5 (3.0-9.3)
Sex
Female	46 (41.1)
Male	66 (58.9)
Immediate-type index reactions (n = 29)
Urticaria with or without angioedema (skin reaction without systemic anaphylaxis)	28 (96.6)
Anaphylaxis
Total
Milda	1 (3.4)
Moderatea	0
Severea	1 (100)
Treated with epinephrine	1 (<0.01)
Nonimmediate-type index reactions (n = 75)	0
Other morphological variants (eg, maculopapular exanthema)	73 (97.3)
Treated with epinephrine	2 (2.7)
1-8 h After last dose	1 (1.3)
>8 h After last dose	1 (1.3)
Type of macrolide used for graded oral challenge
Clarithromycin	81 (72.3)
Azithromycin	31 (27.7)
Result of graded oral challenge
Positive result (reaction)	11 (9.8)
Immediate	9 (8.0)
Nonimmediate	2 (1.8)
Negative result (tolerance)	101 (90.2)
Immediate-type GOC reactions (n = 9)
Urticaria with/without angioedema (skin reaction without systemic anaphylaxis)	8 (88.9)
Anaphylaxis
Total	1 (11.1)
Milda	0
Moderatea	0
Severea	1 (11.1)
Treated with epinephrine	1 (100)
Nonimmediate-type GOC reactions (n = 2)
Other morphological variants (eg, maculopapular exanthema)	2 (100)
Treated with epinephrine	0

^a Severity of anaphylaxis (mild, moderate, severe) was defined using a modified grading scheme.

Positive GOC results were more likely in children older than 12 years (adjusted odds ratio [AOR], 1.43 (95% CI, 1.26-1.61) and in those with other antibiotic allergies (AOR, 1.22; 95% CI, 1.02-1.46); reactions were less likely in those with a known food allergy comorbidity (AOR 0.84; 95% CI 0.73-0.97) (Table 2). At follow-up visits in January 2024, we found that 12 of 69 patients (17.4%) with negative GOC results used their culprit macrolide since the GOC, with no adverse reactions.

Table 2. Factors Associated With Positive Graded Oral Challenge Reactions to Macrolides.

Variable	Positive graded oral challenge reactions, OR (95% CI)a
	Univariate	P value	Multivariate	P value
Age at challenge (>10 y)	1.33 (1.18-1.50)b	<.001	1.43 (1.26-1.61)b	<.001
Years between index reaction and challenge	1.02 (0.99-1.04)	.20	1.01 (0.99-1.03)	.22
Male sex	1.06 (0.94-1.19)	.33	1.01 (0.92-1.11)	.81
Comorbidity
Antibiotic allergy	1.18 (0.95-1.46)	.14	1.22 (1.02-1.46)b	.03
Asthma	1.02 (0.87-1.19)	.84	1.06 (0.92-1.21)	.42
Eczema	0.98 (0.86-1.13)	.84	1.06 (0.94-1.19)	.35
Allergic rhinitis	0.89 (0.74-1.08)	.23	0.86 (0.71-1.03)	.10
Food allergy	0.89 (0.76-1.05)	.16	0.84 (0.73-0.97)b	.02
Parental drug allergy	0.99 (0.87-1.11)	.97	0.99 (0.9-1.10)	.91

^a Immediate- and nonimmediate-type graded oral challenge reactions.

^b Statistically significant (P < .05).

Discussion

To date, this is the largest pediatric quality improvement study assessing use of direct GOCs for suspected macrolide allergy. Establishing true antibiotic allergy is crucial because labeling for drug allergy may be associated with higher risk of antibiotic resistance, increased health care costs, and poorer concordance with prescribing guidelines.^5,6 A study limitation is recall bias due to our method of documenting index reaction characteristics, but time between index macrolide reaction and patient recruitment during GOC was under a year, which likely limits implications of recall bias. Our findings suggest that direct GOC is a safe, effective strategy for pediatric patients with nonsevere, skin-limited symptoms during macrolide treatment. Future research on adult macrolide GOCs is needed to confirm these findings in children.

Supplement.

Data Sharing Statement