Cutaneous Complications of mRNA and AZD1222 COVID-19 Vaccines: A Worldwide Review

Because of the increasing emergence of cutaneous reactions from COVID-19 vaccines worldwide, we investigated the published reports of these complications. We searched the PubMed, Google Scholar, and Scopus databases and the preprint server bioRxiv for articles on cutaneous complications linked to mRNA-1273 (Moderna), BNT162b2 (Pfizer–BioNTech), and AZD1222 (AstraZeneca–Oxford University) vaccines published until 30 September 2021. Eighty studies describing a total of 1415 reactions were included. Cutaneous reactions were more prevalent in females (81.6%). Delayed large local reactions were the most common complication (40.4%), followed by local injection site reactions (16.5%), zoster (9.5%), and urticarial eruptions (9.0%). Injection site and delayed large local reactions were predominantly caused by the mRNA-1273 vaccine (79.5% and 72.0%, respectively). BNT162b2 vaccination was more closely linked to distant reactions (50.1%) than mRNA-1273 (30.0%). Zoster was the most common distant reaction. Of reactions with adequate information for both vaccine doses, 58.3% occurred after the first dose only, 26.9% after the second dose only, and 14.8% after both doses. Overall, a large spectrum of cutaneous reaction patterns occurred following the COVID-19 vaccination. Most were mild and without long-term health implications. Therefore, the occurrence of such dermatologic complications does not contraindicate subsequent vaccination.


Introduction
Coronavirus disease 2019 (COVID- 19) vaccines are an effective tool in reducing the risk of developing COVID-19 and serious adverse outcomes. Cutaneous complications have been associated with COVID-19 vaccination [1,2]. In a US study, cutaneous adverse effects associated with the first dose of messenger RNA (mRNA) vaccines were reported by 1.9% (95% CI, 1.8-2.1%) of health care employees [3]. Of those, 83% reported no recurrent cutaneous reactions. In a prospective observational study from the UK, first-dose and second-dose skin reactions were observed in only 1.1% and 1.7% of patients, respectively, after BNT162b2 (Pfizer-BioNTech) vaccination [4]. While the cutaneous complications of these vaccines may be reported less frequently, they nonetheless impact public perception regarding vaccine safety. The objective of this systematic review was to assess the dermatologic complications of mRNA-1273 (Moderna; mRNA vaccine), BNT162b2, and AZD1222 (AstraZeneca-Oxford University; adenovirus vector vaccine) vaccination.   Most reactions (55.7%) were associated with mRNA-1273 vaccination ( Table 2). Injection site reactions and delayed large local reactions were predominantly caused by the mRNA-1273 vaccine (79.5% and 72.0%, respectively). BNT162b2 vaccination was more closely linked to distant reactions (334/610; 54.8%) than mRNA-1273 (191/610; 31.3%) ( Table 2). Of reactions with adequate information for both vaccine doses (n = 1361), 58.3% occurred after the first dose only, 26.9% after the second dose only, and 14.8% after both doses. Potential mechanisms underlying cutaneous reactions are summarized in Table 3. Table 3. Suggested pathogenetic mechanisms underlying COVID-19 vaccine-related cutaneous reactions.

Skin Reaction Potential Mechanisms
Delayed large local reaction T-cell mediated responses to a vaccine excipient, lipid nanoparticle, or mRNA component [10,11] Urticaria IgE-mediated reactions are more typically associated with the inactive components of the vaccine (i.e., egg proteins, gelatin, and latex) [85] Anaphylaxis Pre-existing antibody recognition of the vaccine excipient polyethylene glycol (PEG); contact system activation by nucleic acid; complement recognition of the vaccine-activating allergic effector cells; direct mast cell activation [85] Morbilliform eruption Immune activation-mediated skin response; prior coronavirus infection may generate a cross-reaction with antigen that mRNA vaccine encodes [86] VZV/HSV reactivation Innate or cell-mediated immune defense failures initiated by the host in response to mRNA COVID-19 vaccines; [22] strong immune response against the S protein from vaccine may distract the cell-mediated control of another, latent virus [2] Pityriasis rosea-like eruption Vaccination leads to a state of altered immunity and may lead to endogenous reactivation of HHV-6 or HHV-7 [40]; T-cell mediated response triggered by molecular mimicry from a viral epitope [41] Pernio, chilblains, and purpuric lesions Vaccine-induced microangiopathy [51]; viral proteins in the endothelial cells of the dermal vessels and accumulation of immune complexes that activate the complement cascade, causing small vessel wall damage [52] DIR to dermal hyaluronic acid fillers COVID-19 spike protein interacts with ACE2 receptors which trigger pro-inflammatory loco-regional TH1 cascade and promote a CD8 and T cell mediated reaction to incipient granulomas [57,58]

Vesiculobullous lesions
Cross-reactions between SARS-CoV-2 spike protein antibody and tissue proteins such as transglutaminase 2 and 3, collagen, and S100B antigen may play a role in developing these immune-mediated skin lesions [60] Generalized eczematous eruptions Vaccine may act as an environmental trigger in a genetically susceptible individual (i.e., personal/family history of atopy) [61] Radiation recall dermatitis Offending agent upregulates inflammatory cytokines that are already increased in area of irradiation, leading to a local hypersensitivity reaction [87] SDRIFE-like eruption Co-infection by other viruses or uncommon clinical presentation of post-vaccination hyperviscosity [88] Stevens-Johnson syndrome Expression of vaccine antigens on keratinocytes leads to a CD8+ T-cell response against epidermal cells, thus causing apoptosis of keratinocytes and detachment of dermo-epidermal junction in a genetically susceptible individual [66] Psoriasis exacerbation Vaccine increases IL-6 production and recruitment of Th17 cells which are involved in psoriasis; [78] vaccine may activate the plasmacytoid and dermal myeloid dendritic cells, which upregulate type I IFNs that initiate the inflammatory cascade; [81] mRNA vaccines bind to Toll-like receptors that result in increased production of type I IFNs [45] ACE2 angiotensin-converting enzyme 2; HHV, human herpes virus; HSV, herpes simplex virus; IFNs, interferons; mRNA, messenger RNA; SDRIFE, symmetrical drug-related intertriginous and flexural exanthem; Th1, T helper 1; VZV, varicella zoster virus.

Quality of Evidence Assessment
The rating score of the studies included is shown in Table 1. There were only a small number of registry-based studies and cohorts [1,2,7,18,25,30], and the sample size of some outcomes (cutaneous reactions) in registries/cohorts documenting various outcomes (different types of cutaneous reactions) was small (Table 1). Reporting bias applied, as evidenced by limited data for AZD1222 and the fact that most cutaneous reactions were documented in white persons. Studies performed in health care workers confirmed reporting bias, as healthcare workers are more likely to report their reactions [1,7]. A registry-based study may have included a confirmation bias, as providers are more likely to report cases with severe or rare manifestations [1]. Biases relevant to retrospective observational studies, such as selection and information biases (e.g., short follow-up period; course of the reaction determined mainly based on the patient's description), also applied.
Second-dose DLLLs generally occur more quickly (median of 2-3 days) [1,[9][10][11]. Such reactions were fewer in the AAD/ILDS registry but not in a cohort of 103 COVID arm cases associated with BNT162b2 vaccination (54% of reactions) [1,7]. The duration of second dose DLLLs was longer than first dose examples in most studies [1,9,10]. In the American Academy of Dermatology/International League of Dermatologic Societies (AAD/ILDS) registry, the majority of patients who developed a DLLL after both doses of the mRNA-1273 or BNT162b2 vaccines showed a larger reaction after the second dose [1]. ond dose DLLLs was longer than first dose examples in most studies [1,9,10]. In the American Academy of Dermatology/International League of Dermatologic Societies (AAD/ILDS) registry, the majority of patients who developed a DLLL after both doses of the mRNA-1273 or BNT162b2 vaccines showed a larger reaction after the second dose [1].
Histopathology of DLLLs showed perivascular lymphocytic infiltrates with eosinophils and scattered mast cells consistent with a delayed T-cell mediated hypersensitivity reaction [9,89]. The presence of prominently dilated vessels with edematous endothelial layers was a consistent feature [13]. 'COVID arm' typically resolves within one week of treatment with topical corticosteroid, oral antihistamines, and symptomatic therapy. Many cases have been treated with expectant management [2,11].
Anaphylaxis has developed within 150 min post-COVID-19 vaccination. It is uncommon; of 1,893,360 individuals who received the first BNT162b2 vaccine dose, the Food and Drug Administration (FDA) reported 21 patients with an anaphylactic reaction [18]. Of those, 19 were female, 2 were male, and 17 had a history of allergies or allergic reactions. The reaction occurred at a median of 13 min post-vaccination. Of 4,041,396 individuals that received the mRNA-1273 vaccine, 10 females experienced anaphylaxis after the first Histopathology of DLLLs showed perivascular lymphocytic infiltrates with eosinophils and scattered mast cells consistent with a delayed T-cell mediated hypersensitivity reaction [9,89]. The presence of prominently dilated vessels with edematous endothelial layers was a consistent feature [13]. 'COVID arm' typically resolves within one week of treatment with topical corticosteroid, oral antihistamines, and symptomatic therapy. Many cases have been treated with expectant management [2,11].
Anaphylaxis has developed within 150 min post-COVID-19 vaccination. It is uncommon; of 1,893,360 individuals who received the first BNT162b2 vaccine dose, the Food and Drug Administration (FDA) reported 21 patients with an anaphylactic reaction [18]. Of those, 19 were female, 2 were male, and 17 had a history of allergies or allergic reactions. The reaction occurred at a median of 13 Min post-vaccination. Of 4,041,396 individuals that received the mRNA-1273 vaccine, 10 females experienced anaphylaxis after the first dose [25]. Nine of 10 patients had a history of atopic disease, and anaphylaxis occurred a median of 7.5 min post-vaccination. All patients were treated with an emergency intramus-cular or subcutaneous epinephrine injection [18,25]. Mechanisms of anaphylactic reaction are shown in Table 3 [85]. dose [25]. Nine of 10 patients had a history of atopic disease, and anaphylaxis occurred a median of 7.5 min post-vaccination. All patients were treated with an emergency intramuscular or subcutaneous epinephrine injection [18,25]. Mechanisms of anaphylactic reaction are shown in Table 3 [85].
Among the cases that were not associated with anaphylaxis, most eruptions developed within 2 to 3 days post-vaccination and resolved within a week. A generalized eruption (>30% of body surface area covered) in one participant that received the BNT162b2
Among the cases that were not associated with anaphylaxis, most eruptions developed within 2 to 3 days post-vaccination and resolved within a week. A generalized eruption (>30% of body surface area covered) in one participant that received the BNT162b2 vaccine persisted for more than one month [28]. The patient had no significant past medical history or drug allergy. Histopathology showed lymphocytic perivascular infiltrates consistent with maculopapular eruption. A laboratory investigation showed increased liver enzymes and the second vaccine dose was not provided. Tihy et al. indicated that morbilliform eruptions shared histopathologic similarities with drug eruption [17]. Ohsawa and colleagues demonstrated similarities between the immunohistochemical features of morbilliform eruption in one case and those found in COVID-19-associated skin lesions [86]. When treatment is required, morbilliform eruptions respond to topical/systemic corticosteroids and oral antihistamines.
Microorganisms 2022, 10, x FOR PEER REVIEW 21 of 29 vaccine persisted for more than one month [28]. The patient had no significant past medical history or drug allergy. Histopathology showed lymphocytic perivascular infiltrates consistent with maculopapular eruption. A laboratory investigation showed increased liver enzymes and the second vaccine dose was not provided. Tihy et al. indicated that morbilliform eruptions shared histopathologic similarities with drug eruption [17]. Ohsawa and colleagues demonstrated similarities between the immunohistochemical features of morbilliform eruption in one case and those found in COVID-19-associated skin lesions [86]. When treatment is required, morbilliform eruptions respond to topical/systemic corticosteroids and oral antihistamines.

Delayed Inflammatory Reaction (DIR) to Dermal Hyaluronic Acid Filler
DIR to hyaluronic acid dermal filler presents clinically as edema with inflammatory, erythematous nodules at the site of prior dermal filler injections. The AAD/ILDS registry reported one DIR after BNT162b2 and eight after mRNA-1273 vaccination [1]. Munavalli and colleagues reported three DIRs after BNT162b2 and four after mRNA-1273 vaccination [57,58]. The reactions occurred within 10 days after vaccination. Marked improvements were noted within 5 days of lisinopril 5-10 mg administration in all patients. In a patient who developed DIR after the first mRNA-1273 dose, preventive lisinopril treatment was successful before the second dose [58]. Angiotensin-converting enzyme 2 inhibitors (ACE-I), such as lisinopril, can block ACE2 receptor targeting by the SARS-CoV-2 spike protein that releases a proinflammatory cascade. This observation may explain the efficacy of lisinopril treatment in the above DIRs. A case was treated with hyaluronidase injection [59]. The American Society for Dermatologic Surgery released guidance in which it was outlined that patients with dermal fillers do not have any contraindication to receiving any COVID-19 vaccine, and that those who already received the vaccine remain candidates for the future receipt of dermal filler [90].

Eczematous Eruption
A pruritic generalized eczematous eruption was described in three patients within 14 days post-BNT162b2 vaccination [22,61]. Two patients had a history of atopic dermatitis and another dyshidrotic eczema. Cases of localized eczematous dermatitis and hematogenous contact dermatitis have been reported [23,24].

Discussion
DLLLs were the most common post-vaccination skin complication, followed by local injection site reactions, urticarial eruptions, zoster, and morbilliform eruptions. Most local reactions were associated with the mRNA-1273 vaccine and most distant reactions with BNT162b2. Zoster was the most common distant reaction. To our knowledge, this finding has not been reported. There is considerable geographic variation because most participants in the studies included were from Europe and the USA. Most patients (81.6%) that developed cutaneous reactions were female [1,2,7]. Female predominance was observed not only in US studies that included the health care workforce (consisting of 76% females [91]), which might reflect a reporting bias [1], but in European studies as well [2,7]. Some authors propose that women's immune systems may be more reactive to coronavirus proteins, leading to a lower susceptibility to the disease and a higher reactogenicity to vaccines [2].
Most reactions were effectively managed with minimal to no long-term morbidity, and the completion of the vaccination course was recommended [1]. Anaphylactic reactions are rare with COVID-19 vaccines [18], and the incidence has been similar to what is noted with other virus-based vaccines [92]. Fatalities were not reported. CDC recommends that vaccination be contraindicated in patients who have had a severe or immediate allergic reaction to the COVID-vaccine or any of its components and that clinicians consider a referral to an allergist-immunologist in such cases [18]. As most people that experienced anaphylaxis had allergy histories [18,25,93], it is very important that clinicians screen for a history of anaphylaxis or angioedema or a proclivity to allergic reactions, e.g., a history of atopy or allergic reactions to vaccine components. Individuals with histories of allergic reactions to one or several of the COVID-19 vaccine ingredients should not receive vaccination [94]. Receiving a different COVID-19 vaccine for the second dose is appropriate for patients with a proclivity to allergy experiencing first-dose reactions. Studies have shown that heterologous prime-boost vaccines are effective and may provide higher immunogenicity than using the same vaccine for booster doses [95].
Some patients experienced reactions to mRNA vaccines, such as pernio/chilblains and erythromelalgia, which mimicked COVID-19 infection [1]. This finding suggests that the vaccine replicates the host immune response to the virus, and some components of such cutaneous reactions result from an immune response to the virus rather than direct viral effects. It is important that clinicians distinguish cutaneous reactions to vaccines from signs of COVID-19 occurring post-vaccination. However, in some cases, the devel-

Discussion
DLLLs were the most common post-vaccination skin complication, followed by local injection site reactions, urticarial eruptions, zoster, and morbilliform eruptions. Most local reactions were associated with the mRNA-1273 vaccine and most distant reactions with BNT162b2. Zoster was the most common distant reaction. To our knowledge, this finding has not been reported. There is considerable geographic variation because most participants in the studies included were from Europe and the USA. Most patients (81.6%) that developed cutaneous reactions were female [1,2,7]. Female predominance was observed not only in US studies that included the health care workforce (consisting of 76% females [91]), which might reflect a reporting bias [1], but in European studies as well [2,7]. Some authors propose that women's immune systems may be more reactive to coronavirus proteins, leading to a lower susceptibility to the disease and a higher reactogenicity to vaccines [2].
Most reactions were effectively managed with minimal to no long-term morbidity, and the completion of the vaccination course was recommended [1]. Anaphylactic reactions are rare with COVID-19 vaccines [18], and the incidence has been similar to what is noted with other virus-based vaccines [92]. Fatalities were not reported. CDC recommends that vaccination be contraindicated in patients who have had a severe or immediate allergic reaction to the COVID-vaccine or any of its components and that clinicians consider a referral to an allergist-immunologist in such cases [18]. As most people that experienced anaphylaxis had allergy histories [18,25,93], it is very important that clinicians screen for a history of anaphylaxis or angioedema or a proclivity to allergic reactions, e.g., a history of atopy or allergic reactions to vaccine components. Individuals with histories of allergic reactions to one or several of the COVID-19 vaccine ingredients should not receive vaccination [94]. Receiving a different COVID-19 vaccine for the second dose is appropriate for patients with a proclivity to allergy experiencing first-dose reactions. Studies have shown that heterologous prime-boost vaccines are effective and may provide higher immunogenicity than using the same vaccine for booster doses [95].
Some patients experienced reactions to mRNA vaccines, such as pernio/chilblains and erythromelalgia, which mimicked COVID-19 infection [1]. This finding suggests that the vaccine replicates the host immune response to the virus, and some components of such cutaneous reactions result from an immune response to the virus rather than direct viral effects. It is important that clinicians distinguish cutaneous reactions to vaccines from signs of COVID-19 occurring post-vaccination. However, in some cases, the development of COVID-19 after immunization cannot be excluded as a plausible cause of cutaneous reactions. Still, available data suggest that prior COVID-19 does not predetermine cutaneous reactions, or reactions of a greater severity, after vaccination [2]. This review has several limitations. The search for articles was restricted to those written in English. Many of the included studies were case reports and studies with small sample sizes that may confer publication bias. Additionally, most studies included participants from the USA and Europe, and there is a lack of data from other parts of the globe. Also, there are limited data for the AZD1222 vaccine. The above may reflect underreporting and limit the generalizability of the results. The short duration of participant selection, including the follow-up period, in large studies is an additional limitation because providers entered data at one point in time [1] and/or the study was conducted within a short period of time [2,7]. Lastly, most vaccine reactions were documented in white individuals, and this raises concerns about disparities in vaccine access, health care access after experiencing an adverse effect, the differential likelihood of reporting to registries, and/or the recognition of such reactions in patients of color [1].
Dermatologists should contribute to the improved documentation of cutaneous reactions and safety monitoring by reporting their observations to VAERS. Appropriate patient counseling regarding cutaneous reactions to COVID-19 vaccines is crucial and prevents generating concerns disproportionate to potential complications. General practitioners should be aware of such reactions and can play an important role in patient counseling. Lastly, the appropriate identification and management of vaccine reactions often requires a multispecialty approach involving dermatology, allergy, and infectious disease specialists [96].
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/microorganisms10030624/s1, Figure S1. Flow diagram of study selection performed according to Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines.

Conflicts of Interest:
The authors declare no conflict of interest. Eleftherios Mylonakis was involved in clinical trials on COVID-19. These trials were supported by Regeneron, NIH, and SciClone Pharmaceuticals, Inc. All funds were given to the institution, and Eleftherios Mylonakis received no direct funds.