Outcomes of pediatric patients with suspected allergies to COVID-19 vaccines

Abstract

Background

Adverse effects following immunizations (AEFIs) can contribute to vaccine hesitancy.

Objective

We evaluated clinical outcomes of AEFIs subsequent to administration of the coronavirus disease 2019 (COVID-19) vaccine at 2 pediatric allergy centers.

Methods

Data on pediatric patients referred for COVID-19 AEFI concerns between March 2021 and October 2022 were reviewed. The collected data included patient demographics, clinical characteristics, outcomes of prior COVID-19 vaccination, recommendations after consultation, and outcomes of revaccination.

Results

The 163 patients were separated into 2 groups based on the absence (n = 89 [54.6%]) or presence (n = 74 [45.4%]) of prior COVID-19–related AEFIs. The most common reason for referral without a prior AEFI was another suspected drug allergy (n = 58 [35.6%]). All patients in this group were recommended for COVID-19 vaccination. Of the 163 patients, 82 (92.1%) proceeded with vaccination, with 77 of them (93.9%) tolerating vaccination. Most of those with a prior COVID-19–related AEFI had a delayed cutaneous reaction (n = 60 [37.0%]); 1 patient (0.6%) had suspected anaphylaxis. In this group, 6 (8.1%) were advised to postpone COVID-19 vaccination until their debilitating skin conditions had improved in response to further treatment, whereas 45 (77.6%) tolerated subsequent vaccination to the same or an alternate COVID-19 vaccine type. The most common AEFI on revaccination was urticaria (in 8 of 11 patients [72.7%]). AEFI on revaccination was significantly associated with a history of spontaneous urticaria or angioedema (relative risk = 3.6 [95% CI = 1.30-9.99]; P = .020) and urticaria following COVID-19 vaccination previously (relative risk = 4.12 [95% CI = 1.22-13.87]; P = .017).

Conclusions

Children with a history of urticaria or angioedema related or unrelated to prior COVID-19 vaccination were at higher risk of a COVID-19–related AEFI on revaccination, although most were able to complete the vaccination series under the management of our immunology/allergy service.

Introduction

Coronavirus disease 2019 (COVID-19) has led to more than 700 million infections and 7 million deaths.¹ Post–COVID-19 complications, particularly multisystem inflammatory syndrome in children, can be fatal.² Social distancing reduced transmission but negatively affected mental health, economic stability, and quality of life.³ Development of COVID-19 vaccines, including the mRNA-based BNT162b2 (BioNTech and Pfizer) and inactivated whole-virus CoronaVac (Sinovac Biotech, Beijing, China) vaccines, has ameliorated these issues.⁴ However, incomplete vaccination owing to fears of adverse effects from immunizations (AEFIs), particularly severe allergic reactions, contributes to ongoing risks.^4,5

Among AEFIs, COVID-19 vaccine–associated anaphylactic reactions occur in approximately 7.9 per million individuals (0.008%) and approximately 12.8 per million (0.013%) children receiving mRNA-based vaccines.^6,7 The recurrence rate of severe, immediate allergic reactions following revaccination is low (0.16%), but the rate is higher (13.65%) for nonsevere, immediate reactions.^6,8 Delayed cutaneous reactions, including local reactions and chronic urticaria, are common (1%-2%).⁹ Although generally not life-threatening, these reactions can reduce quality of life and contribute to vaccine hesitancy, delaying the protective benefits of vaccination.¹⁰ A Canadian study involving 101 children with chronic urticaria found no significant risk of allergic reaction after 2 doses of the BNT162b2 vaccine, although further research is required to assess the consistency of these findings in other geographic regions, as genetic variations in specific HLA alleles can influence allergy risks.^11,12

Individuals with suspected vaccine allergies should be evaluated by experienced clinicians.¹³ The diagnostic approach can include skin testing for immediate allergic reactions, whereas individuals with local or mild delayed reactions should proceed with the next dose.¹⁴ Negative skin test results indicate low risk, whereas supervised graded challenge can be offered to patients with positive results.¹³ Few vaccination reactions, aside from anaphylaxis and life-threatening delayed hypersensitivities, are considered absolute contraindications. However, the testing process can be labor-intensive and may induce anxiety, especially in children. The diagnostic accuracy of skin testing for COVID-19 vaccines has not been established, as some reactions have not been reproducible—perhaps because some immediate reactions to mRNA-based vaccines were not IgE-mediated but instead occurred via the Mas-related G protein–coupled receptor member X2 pathway.15, 16, 17 These mechanisms and outcomes of diagnostic investigations remain mostly unknown.

We aimed to investigate clinical outcomes of children referred to 2 tertiary allergy centers for concerns regarding COVID-19 vaccine allergy, with the hypothesis that vaccine-naive individuals would tolerate vaccination without severe reactions, whereas COVID-19–related AEFIs following revaccination consist mostly of flare or new-onset chronic urticaria or eczematous dermatitis.

Results and discussion

A total of 163 patients, mostly aged 12 to 18 years (78.5%), with a median age of 14 years (range 2-20 years) were assessed (Table I). Males comprised 58.9% of the study patients and the majority (89.6%) were Chinese. Approximately half of the patients (n = 89 [54.6%]) had no prior COVID-19 vaccine–related AEFI (Fig 1). Of these patients, 85 (52.1%) were vaccine-naive, whereas 4 (2.5%) tolerated COVID-19 vaccines but were referred for being at higher risk of vaccine allergy. The remaining 74 patients (45.4%) were assessed after reactions to the BNT162b2 or CoronaVac vaccine. For those with no prior AEFI to COVID-19 vaccination, the main referral reason was suspected drug allergy (n = 58 [35.6%]) (Fig 2, A). Among the patients referred for COVID-19 vaccination reactions, delayed cutaneous reactions were the most common reaction (n = 60 [36.8%]). Urticaria was the most common symptom (n = 32 [43.2%]), followed by dermatitis (n = 27 [36.5%]) (Fig 2, B). Although cutaneous manifestations were often self-limiting, some severe presentations required aggressive management, including systemic therapies, biologics, and antimicrobial treatments (Fig 3).

Table I.

Demographic characteristics of the 163 study patients

Characterstic	All patients (N = 163)	No prior vaccination reaction (n = 89)	Prior vaccination reaction (n = 74)	Tolerated revaccination (n = 45)	Had symptoms following revaccination (n = 11)
Age at first consultation (y), median (95th interpercentile range)	14 (2-20)	14 (2-20)	14 (3-18)	15 (4-18)	16 (3-17)
Age group at first consultation, no. (%)
6 mo-4 y	6 (3.7%)	3 (3.4%)	3 (4.1%)	1 (2.2%)	1 (9.1%)
5-11 y	28 (17.2%)	7 (7.9%)	21 (28.4%)	11 (24.4%)	2 (18.2%)
12-18 y	128 (78.5%)	78 (87.6%)	50 (67.6%)	33 (73.3%)	8 (72.7%)
>18 y	1 (0.6%)	1 (1.1%)	0 (0%)	0 (0%)	0 (0%)
Sex, no. (%)
Male	96 (58.9%)	55 (61.8%)	41 (55.4%)	28 (62.2%)	5 (45.5%)
Female	67 (41.1%)	34 (38.2%)	33 (44.6%)	17 (37.8%)	6 (54.5%)
Race, no. (%)
Chinese	146 (89.6%)	76 (85.4%)	70 (94.6%)	43 (95.6%)	10 (90.9%)
Non-Chinese	17 (10.4%)	13 (14.6%)	4 (5.4%)	2 (4.4%)	1 (9.1%)
Vaccination status, no. (%)
Unvaccinated	85 (52.1%)	85 (95.5%)	N/A	N/A	N/A
Tolerated prior vaccinations	4 (2.5%)	4 (4.5%)	N/A	N/A	N/A
Adverse reaction after first dose	58 (35.6%)	N/A	58 (78.4%)	40 (88.9%)	9 (81.8%)
BNT162b2	36 (22.1%)	N/A	36 (48.6%)	27 (60.0%)	5 (45.5%)
CoronaVac	22 (13.5%)	N/A	22 (29.7%)	13 (28.9%)	4 (36.4%)
Adverse reaction after second dose	15 (9.2%)	N/A	15 (20.3%)	5 (11.1%)	2 (18.2%)
BNT162b2	7 (4.3%)	N/A	7 (9.5%)	3 (6.7%)	1 (9.1%)
CoronaVac	8 (4.9%)	N/A	8 (10.8%)	2 (4.4%)	1 (9.1%)
Adverse reaction after third dose	1 (0.6%)	N/A	1 (1.4%)	0 (0%)	0 (0%)
BNT162b2	0 (0%)	N/A	0 (0%)	0 (0%)	0 (0%)
CoronaVac	1 (0.6%)	N/A	1 (1.4%)	0 (0%)	0 (0%)
Atopic comorbidity, no. (%)
Atopic dermatitis	56 (34.4%)	29 (32.6%)	27 (36.5%)	16 (35.6%)	5 (45.5%)
Asthma	13 (8.0%)	8 (9.0%)	5 (6.8%)	4 (8.9%)	0 (0%)
Allergic rhinitis	60 (36.8%)	32 (36.0%)	28 (37.8%)	19 (42.2%)	5 (45.5%)
Spontaneous urticaria/angioedema	31 (19.0%)	12 (13.5%)	19 (25.7%)	8 (17.8%)	6 (54.5%)
Food allergy	32 (19.6%)	26 (29.2%)	6 (8.1%)	4 (8.9%)	0 (0%)
Drug allergy	42 (25.8%)	35 (39.3%)	7 (9.5%)	4 (8.9%)	1 (9.1%)
History of anaphylaxis	3 (1.8%)	3 (3.4%)	0 (0%)	0 (0%)	0 (0%)
Multiple allergic comorbidities	69 (42.3%)	41 (46.1%)	28 (37.8%)	16 (35.6%)	6 (54.5%)

N/A, Not applicable.

Fig 1.

Flow diagram of the study. From March 2021 to October 2022, a total of 163 patients (96 males, median age 14 years) were included and separated into 2 groups based on the absence (n = 89 [54.6%]) or presence (n = 74 [45.4%]) of prior COVID-19 vaccination–related AEFIs.

Fig 2.

A, Reasons for referral. For those with no prior reaction, another suspected drug allergy (in 58 of 163 cases [35.6%]) was the most common. Of those referred for an adverse reaction after prior COVID-19 vaccination, the most common such reaction was a delayed cutaneous reaction (in 60 cases [37.0%]). B, Types of cutaneous reactions in those with prior COVID-19 vaccination reactions. Urticaria affected 32 patients (43.2%), followed by dermatitis (36.5%).

Fig 3.

Cutaneous adverse reactions after COVID-19 vaccination. COVID-19 vaccination reactions included large local reaction with swollen plaque at the injection site (A), flare of atopic dermatitis (B), purpuric rash (C), papular dermatitis (D), and folliculitis (E).

The mean waiting period for initial consultation was 86.4 days. Some patients without clear contraindications were contacted and advised to undergo vaccination owing to the urgent need during the pandemic, and all 89 without prior COVID-19 vaccination were advised to proceed with vaccination. Of the 74 patients with a prior COVID-19 vaccine–related AEFI, 68 (91.9%) were advised to proceed with revaccination, whereas 6 (8.1%) were advised to postpone revaccination until their severe active dermatitis had improved subsequent to our clinical management. Of the 9 patients (5.5%) recommended for supervised vaccine challenge, 5 had immediate reactions to prior COVID-19 vaccinations, including suspected anaphylactic reaction and urticarial and nonurticarial dermatitis. Four had no prior COVID-19 vaccination reaction but were advised to undergo supervised vaccine administration owing to a history of Steven-Johnson syndrome in response to multiple drugs, suspected allergy to polyethylene glycol (the excipient in the BNT162b2 vaccine) or history of positive intradermal skin reactivity to polyethylene glycol 3350–containing methylprednisolone acetate, history of drug allergy, or history and non–COVID-19 vaccine–related allergy.

Of the 89 patients with no prior COVID-19 vaccination reaction, 82 (92.1%) proceeded with revaccination. Of those 82 patients, 77 (93.9%) tolerated vaccination, 1 (1.2%) developed immediate-onset urticaria and was treated with oral antihistamines and corticosteroids, and 4 (4.9%) were lost to follow-up. Of the 68 patients with prior AEFIs to whom revaccination was recommended, 58 (85.3%) proceeded, with 45 of them (77.6%) tolerating revaccination (Table I). Eleven patients (19.0%) reported an AEFI on revaccination; all of the reactions were cutaneous, with urticaria being the most common (occurring in 8 patients [72.7%]). Two patients (3.4%) were lost to follow-up; 9 patients (13.2%) switched COVID-19 vaccine (7 from the BNT162b2 vaccine to the CoronaVac vaccine), with 6 of them tolerating revaccination, whereas 3 had an AEFI. Of the 9 patients for whom supervised challenge was recommended, 6 (66.7%) proceeded with challenge, with no AEFIs reported. The remaining 3 patients did not return for follow-up.

Univariate analyses identified history of spontaneous urticaria or angioedema (relative risk [RR] = 3.60 [95% CI = 1.30-9.99]; P = .020) and index reaction of urticaria (RR = 4.12 [95% CI = 1.22-13.87]; P = .017) as significantly being associated with AEFI on revaccination (Fig 4). These 2 factors are clinically related, precluding further analyses. The RRs for the atopic comorbidities of anaphylaxis, asthma, food allergy, index reactions of anaphylaxis, angioedema, and local reaction could not be determined on account of no AEFIs for these characteristics.

Fig 4.

Univariate analyses and forest plot of RRs. A history of spontaneous urticaria or angioedema (RR = 3.60 [95% CI = 1.30-9.99]; P = .020) and index reaction of urticaria (RR = 4.12 [95% CI = 1.22-13.87]; P = .017) was significantly associated with an adverse outcome on revaccination. Hx, History; M, male; Rx, reaction. ∗P < .05.

This study, which is one of the largest of COVID-19 vaccine allergy in individuals predominantly younger than 18 years, found that more than 90% of those who were COVID-19 vaccine–naive subsequently tolerated COVID-19 vaccination. No association was found for suspected drug allergy or food allergy, the 2 most common concerns, thus supporting the safety of COVID-19 for these patients.

For those with AEFIs after prior COVID-19 vaccination, most referrals were for cutaneous reactions, including urticaria, maculopapular exanthems, and eczema exacerbation. Urticaria was the most common reason for referral, similar to in previous reports.^18,19 Our patients experienced fewer large local reactions than in a cross-sectional study involving approximately 6000 Japanese adults, which reported a reaction rate of 12.7%; this difference in number of reactions was possibly due to differences in referral practice.^6,20 Our low referral rate for suspected anaphylaxis aligns with the reported low incidence rates for COVID-19 vaccine–associated anaphylaxis.^8,21 An adult vaccine allergy safety clinic in our region reported that 97% of referred patients were recommended for COVID-19 vaccination, with 80% of them proceeding to revaccination.²² Referral concerns for children included broader reasons that required addressing parental hesitation regarding vaccination for these more vulnerable individuals. Most COVID-19 vaccine–naive children assessed were advised to complete their COVID-19 vaccination series in accordance with the international consensus.⁸ Supervised administration for 4 COVID-19 vaccine–naive children alleviated parental concerns and vaccine hesitancy.²³

As normal activities resume, exposure to viral infections is inevitable, necessitating use of COVID-19 boosters and novel vaccines against emerging pathogens. Addressing vaccine hesitancy remains crucial, particularly for children. In our cohort, 85.9% did not experience any AEFIs after vaccination or revaccination. Additionally, all vaccine challenges were uneventful. History of chronic spontaneous urticaria or angioedema and index reaction of urticaria were the identifiable associations. In a recent Australian cohort study of COVID-19 vaccine–related AEFIs in adults, the most common AEFI on revaccination was also urticaria, and index presentation of angioedema was associated with recurrence of AEFIs.¹⁸ The latter was not observed in our cohort, as only 1 patient had an index reaction of angioedema. Of those with a COVID-19 vaccine–associated AEFI, 58.9% had received the BNT162b2 vaccine and 41.1% were administered the CoronaVac vaccine. On revaccination, 10 patients switched from the BNT162b2 vaccine to the CoronaVac vaccine, with AEFIs subsequently reported in only 2 of them. It is apparent that the availability of different vaccine types and choice are paramount in reducing vaccination hesitancy.

Limitations of this study include its relatively small patient sample size and single-region–only design, although the study is one of the largest pediatric studies on COVID-19 vaccine allergy outcomes and risk factors. Referrals were initiated by general practitioners, and the patients who attended were highly concerned about allergic reactions, which can result in selection bias. Because of its retrospective nature, this study is also subject to potential missing information, although medical record documentation and data retrieval were performed by experienced clinicians.

Although the accumulating evidence suggests that mRNA-based vaccines have an overall favorable safety profile, hesitancy persists. Given their high efficacy, RNA-based immunotherapeutics will likely be utilized for vaccinations and other disease treatments in the future. Health care professionals will need to counsel patients on the adverse effects of such immunotherapeutics, particularly those with risk factors that include chronic spontaneous urticaria, as identified in this study. It is vital that health care professionals reassure patients about common side effects and manage the rare but debilitating reactions to optimize clinical outcomes with the aim of completion of the vaccination series. Vaccine hesitancy was ameliorated when patients were reassured and advised to take antihistamines for urticarial flares and seek medical attention if symptoms are not well controlled by standard treatments. Although skin testing and the basophil activation test do not appear prognostic for AEFI following revaccination with mRNA-based vaccines, the utility of newer diagnostic platforms such as enzyme-linked immunospot assay and the mast cell activation test should be explored.^24,25

Ethics approval: This study was approved by the institutional review board of the Hong Kong West Cluster Hospital Authority Research Ethics Committee and the Central Hospital Authority Research Ethics Committee (institutional review board reference no. PAED-2023-003).

Availability of data and material: Data are available on request from the corresponding author.

Clinical implications.

When guided through symptomatic treatments, individuals with past nonanaphylactic reactions to mRNA-based and inactivated COVID-19 vaccines can tolerate subsequent vaccination doses, highlighting the pivotal role of immunologists and allergists in counseling and management.

Disclosure statement

Disclosure of potential conflict of interest: J. S. Rosa Duque received a conference sponsorship from Merck Sharp & Dohme in 2023. Y. L. Lau chairs the Scientific Committee on Vaccine Preventable Diseases of the Government of Hong Kong. The rest of the authors declare that they have no relevant conflicts of interest.