Polyethylene glycol and polysorbate skin testing in the evaluation of coronavirus disease 2019 vaccine reactions
In December 2020, the US Food and Drug Administration (FDA) issued emergency use authorizations for coronavirus disease 2019 (COVID-19) vaccines from Pfizer-BioNTech and Moderna, and widespread vaccination is ongoing. Contraindications to vaccination include a history of immediate allergic reaction to a component or previous dose of an messenger RNA (mRNA) COVID-19 vaccine.1 As of January 18, 2021, anaphylaxis to the Pfizer-BioNTech (Pfizer Inc, New York, New York, BioNTech SE, Mainz, Germany) and Moderna (Moderna, Inc, Cambridge, Massachusetts) vaccines have occurred at rates of 4.7 and 2.5 cases per million doses, respectively.2 The mechanism of allergic reaction is unknown, although inactive vaccine components such as polyethylene glycol (PEG) have been proposed as possible culprit antigens.
PEG is a primary ingredient in osmotic laxatives and a widely used excipient in many medications. It has not been previously used in vaccines, and the molecular weight and structure of the PEG 2000 used in the mRNA COVID-19 vaccines are distinct compared with laxative preparations.3 Allergy to PEG has been described, particularly with higher molecular weight concentrations.4, 5, 6 A review of FDA adverse event reports from 2005 to 2017 revealed an average of 4 cases of anaphylaxis to PEG per year.5 Skin testing has been successfully used to confirm suspected allergy to PEG-containing laxatives and medications, and guidelines for skin testing with nonirritating concentrations of PEG 3350 and polysorbate are available.5 , 7 Recent expert opinion has also provided an algorithm that includes skin testing as part of COVID-19 vaccine reaction evaluation, but the predictive values of PEG and polysorbate skin testing in relation to the risk of hypersensitivity reaction to COVID-19 vaccines are still unknown.8 We report the first 15 cases of PEG and polysorbate skin testing completed in patients who had allergic symptoms after their first dose of the mRNA COVID-19 vaccines or reported a PEG or polysorbate allergy before their first vaccine dose.
Skin testing was performed in patients referred to the allergy divisions of the Mayo Clinics based in Rochester, Minnesota, and Scottsdale, Arizona. The clinical need for skin testing and test selection was provider-determined at the time of evaluation. PEG 3350 (MiraLAX) testing was performed using sequential skin pricks at 1.7 mg/mL, 17 mg/mL, and 170 mg/mL. Methylprednisolone acetate (PEG-containing), methylprednisolone sodium (control), and triamcinolone acetonide (polysorbate 80–containing) testing were performed starting with a skin prick at 40 mg/mL, with subsequent 1:100 and 1:10 intradermal with an additional 1:1 intradermal administration for triamcinolone. Polysorbate 20 testing was performed and given as a 1:1 skin prick followed by 1:100 and 1:10 intradermally with a 0.5 mg/mL concentration with a sterile water diluent. This study was an institutional review board–approved retrospective chart review.
Between January 15, 2021, and February 1, 2021, 15 patients underwent skin testing; 8 had testing because of a reaction to the first dose of COVID-19 vaccine, and 7 had tested before vaccination because of reported PEG or polysorbate allergies (the characteristics and skin testing results of which are presented in Table 1 ). All 8 patients with first vaccine dose reactions had negative PEG 3350 testing, whereas 4 patients had methylprednisolone acetate testing, 3 had triamcinolone acetonide testing, and 2 had polysorbate 20 testing—all of which were negative. One patient had a positive polysorbate 20 reaction (given at 1:10 dilution intradermally), but a sterile water given intradermally resulted in an identical wheal and flare in the patient and one of the authors; thus, this test was interpreted as false-positive owing to irritation. A total of 7 patients successfully received their second COVID-19 vaccine dose without premedication or split-dosing, with the final patient delaying the second dose until vaccine skin testing or split-dosing capabilities are available. In the 7 patients with previous PEG or polysorbate allergy, 1 patient had positive testing to PEG 3350 and methylprednisolone acetate with negative testing to methylprednisolone sodium, triamcinolone acetonide, and polysorbate 20. The 6 other patients all tested negative for PEG. The 3 patients with expanded skin testing also tested negative for polysorbate 20, methylprednisolone acetate, and triamcinolone acetonide. All 6 patients who tested negative received the first dose of the vaccine without any reaction. The patient who tested positive is not yet eligible to receive the vaccine.
|Patient||Age/sex||Previous allergic disease||Culprit agents||Symptoms||Anaphylaxisb||Time to onset||Treatment||ED||Time to resolution||Skin test performed||Time from reaction to the skin test||Skin test result||Time between vaccine doses||Vaccine outcomec|
|1||24F||None||Pfizer-BioNTech vaccine||Urticaria||No||3 h||Antihistamines||No||4 d||PEG||12 d||Negative||23 d||No reaction|
|2||54F||Drug allergy||Pfizer-BioNTech vaccine||Tachycardia, rhinorrhea||No||10 min||None||No||10 min||PEG, MP acetate||7 d||Negative||18 d||No reaction|
|3||36F||None||Pfizer-BioNTech vaccine||Facial flushing||No||5 min||Antihistamines||No||1 h||PEG, MP acetate, TC acetonide||20 d||Negative||21 d||No reaction|
|4||52M||Venom anaphylaxis||Pfizer-BioNTech vaccine||Oral pruritus, throat fullness||No||Immediate||None||No||5 min||PEG||10 d||Negative||21 d||No reaction|
|5||45F||Food allergy||Pfizer-BioNTech vaccine||Urticaria, throat tightness||No||8 h||None||No||Unknown||PEG||20 d||Negative||29 d||No reaction|
|6||33F||Asthma, venom anaphylaxis||Pfizer-BioNTech vaccine||Urticaria, tachycardia||No||15 min||Antihistamines||No||12-24 h||PEG||20 d||Negative||23 d||No reaction|
|7||20M||Vaccine allergy (flu)||Moderna vaccine||Angioedema||No||3 h||Steroids, antihistamines||Yes||24 h||Expandedd||20 d||Negative||N/A||Not given|
|8||22F||Allergic rhinitis||Moderna vaccine||Angioedema, wheezing, throat pruritus||Level 1||20 min||Antihistamines, steroids||Yes||6 h||Expandedd||21 d||Negative||51 d||Minor lip/tongue tingling|
|9||69M||Drug allergy||Moviprep (PEG)||Rash, flushing||No||During prep||Antihistamines||No||1 h||PEG||2 y||Negative||N/A||No reaction (Pfizer)|
|10||73F||Asthma, drug allergy||Moviprep (PEG)||Headache, nausea||No||Unknown||Unknown||No||Unknown||PEG||5 y||Negative||N/A||No reaction (Moderna)|
|11||46F||Vaccine allergy (flu)||Methylprednisolone acetate (PEG)||Urticaria, dizzy, flushing||No||15 min||Epinephrine, steroids, antihistamines||No||12 h||Expandedd||3 mo||Positive (PEG and MP acetate)e||N/A||Not given|
|12||74M||Drug allergy||Triamcinolone acetonide (polysorbate 80)||Urticaria, wheezing||Level 1||1 h||Steroids, antihistamines||Yes||Unknown||PEG, MP acetate, TC acetonide||3 y||Negative||N/A||No reaction (Pfizer)|
|13||55F||Anaphylaxis||Influenza vaccinesf (polysorbate 20/80)||Flushing, wheezing, cough, throat tightness||Level 1||20 min||Epinephrine, steroids, antihistamines||Yes||Unknown||PEG, Polysorbate 20||7 y||Negative||N/A||No reaction (Pfizer)|
|14||67F||Food allergy, anaphylaxis||MiraLAX (PEG)||Oral urticaria||No||2 h||Unknown||No||Unknown||Expandedd||Unknown||Negative||N/A||No reaction (Moderna)|
|15||60M||Vaccine allergy (Shingrix)||Shingrix (polysorbate 80)||Flushing, urticaria||No||2 h||Antihistamines, steroids||Yes||2 d||Expandedd||3 mo||Negative||N/A||No reaction (Moderna)|
This is one of the earliest reports on the use of skin testing to evaluate both possible COVID-19 vaccine reactions and previous PEG or polysorbate allergies before vaccination. In our cohort, only 1 patient had positive testing. No patients with reactions to their first vaccine dose had positive testing. It may be hypothesized in these cases that PEG is not the culprit antigen or that non–immunoglobulin E-mediated mechanisms such as complement activation–related pseudoallergy are involved. In addition, the short interval between reaction and skin testing may increase the risk of false negatives. A preliminary success has been seen with the safe administration of the second vaccine dose after negative skin testing after a possible first dose reaction. One area of uncertainty is the ability of skin testing to predict COVID-19 vaccine reactions in patients who report previous PEG or polysorbate allergies. Systemic reactions to PEG are dependent on a combination of the molecular weight and absolute amount of PEG in the culprit medication, which can differ substantially between injectable and oral forms of PEG.4 Furthermore, the threshold needed to induce a systemic reaction likely differs among individuals.4 These facts make interpretation of skin tests difficult in patients who have yet to receive a COVID-19 vaccine. An additional area of interest has been the recognition of the innumerable medications and vaccines that contain PEG or polysorbate.8 One of our patients with a COVID-19 vaccine reaction also reported a possible reaction to a polysorbate 80–containing influenza vaccine. The magnitude of risk these previous reactions confer on individuals yet to receive a COVID-19 vaccine and the ability of skin testing to quantify that risk remains unclear.
Although our cohort size precludes any inferences regarding the predictive value of this skin testing, it is clear that allergists will play an essential role in the COVID-19 vaccination effort. As vaccination numbers increase, the absolute number of adverse reactions will also increase, which will provide opportunities to both refine the testing strategy (with no or limited testing potentially being the best strategy) and address vaccination hesitation, with the ultimate goal being accurate risk stratification and safe vaccine administration to the population as a whole.
The authors thank Dr Matthew A. Rank, Dr Gerald W. Volcheck, Dr Alexei Gonzalez-Estrada, and Dr James T. Li for their contributions to this manuscript, including intellectual conception and critical review.
Dr Joshi and Dr Park are co-senior authors.
Disclosures: The authors have no conflicts of interest to report.
Funding: The authors have no funding sources to report.