The here reported case series of delayed local skin reactions following vaccination with mRNA-1273 mirrors recently published case series of similar findings with both currently approved mRNA-based vaccines mRNA-12733,9 and BNT162b26;7. These delayed skin reactions are rare post-vaccination events of both approved mRNA-based vaccines, with incidences reported to be in the order of 0.8-1.0 % following the first and 0.2–1.1% following the second dose2,6. Delayed skin reactions are typically observed several days after the vaccination, i.e. following a symptom-free interval and hence distinguishable from acute allergic and other immediate local reactions attributable to the vaccination intervention itself. The phenomenon is transient and typically resolves within 3–5 days, frequently without any treatment required. In cases in whom pharmacological intervention is needed, the experience of this case series as well as of earlier reports show that the condition may be expected to be well-responsive to topical glucocorticosteroids and oral histamines. In patients in whom a “COVID arm” occurs following the first of two scheduled vaccinations, the recommendation is to seek the second vaccination as scheduled, potentially administered to the opposite arm if needed8.
An important question unanswered to date relates to the precise molecular and cellular mechanisms underlying the “COVID arm” phenomenon. In response to this question, we noted that the delayed skin reactions observed in our case serious are of two distinct time windows of onset and of two distinct (even though overlapping) clinical phenotypes, indicative of more than a single pathomechanism. Specifically, we observed manifestations of comparatively early onset (days 2–3 post vaccination) characterized by diffuse, poorly demarcated erythema associated with variable degrees of local edema and symptoms of tenderness, local hypersensitivity and itch well-responsive to antihistamines (as exemplified by case 1). These comparatively early skin manifestations appear to be distinct from manifestations of later onset (days 7–10 after the first, days 2–4 after the second vaccination), which are characterized by more sharply demarcated erythema of irregular peri- and intra-lesional morphology. While the former type of skin manifestation clinically reminds of an allergic pathology, we were particularly interested in the pathology and pathomechanisms underlying the latter type of late-onset delayed skin reactions (mirroring events described in earlier reports), with the delayed time window suggesting the involvement of the patient’s adaptive immune response to the vaccination. Even though patient numbers are small, the shorter time to onset following second vaccinations, as observed in our cohort, similarly suggests the involvement of patients’ adaptive immune response, the more so as a similar shortening of time to onset following second vaccination has been observed before3. The exact mechanism of these reactions is not known, however a delayed hypersensitivity reaction has been hypothesized.10
Focusing on these skin reactions of delayed onset, we considered for reference early skin manifestations associated with COVID-19 infection, per se. While COVID-19 associated skin lesions are of a wide clinical spectrum including chilblain-like, urticarial, vesicular and vasculitic lesions, the occurrence of distinct types of skin lesions within distinct windows of onset during the COVID-19 clinical course has been noted11. Histopathological skin biopsy analyses of these COVID-19 associated cases show diverse ranges of morphologies. A consistent histological feature, however, appears to be the presence of prominently dilated blood vessels with swollen endothelial layers and vessels engulfed with red blood cells and perivascular infiltrates12.
The specific mechanisms underlying these endothelial, vascular, and perivascular inflammatory changes associated with COVID-19 infection have not been fully elucidated. A possible lead hypothesis involves direct viral infection of endothelial cells, with preliminary evidence for this derived from existing data of both electron microscopy and polymerase chain reaction (PCR) analyses within skin lesions13,14 15.
Not dissimilar to the cutaneous histomorphology observed in COVID-19 infection, the histology images of the current mRNA-1273 case series and similarly of a previously reported histology analysis of skin lesions following mRNA-based BNT162b2 vaccination6 share prominent features of a superficial and deep perivascular dermatitis with scattered eosinophils and intraluminal neutrophil accumulation (Fig. 9, 10). This histology pattern is commonly referred to as dermal hypersensitivity reaction16. It is not diagnostic for any specific condition or etiology. Additional analyses will be required, therefore, to further dissect the phenomenon itself as well as its molecular triggers.
At least two categorically different molecular pathomechanisms will need to be considered in these future investigations: Firstly, polyethylene glycol (PEG), an excipient contained in both mRNA-based vaccines yet not in DNA vector vaccines1 may trigger allergic post-vaccination reactions as a net consequence of prior sensitization to PEG17 or by direct T cell stimulatory action18. With direct proof lacking, the hypothesis of excipient-mediated pathology may appear to be supported indirectly, at least at first glance, by the fact that delayed local skin reactions have not been reported the same with DNA vector vaccines, i.e. with vaccines also encoding the S protein yet neither containing PEG nor other excipients contained in mRNA-based lipid nanoparticle vaccines1,17,19. An alternative hypothesis still to be considered, however, is the potential direct transfection of endothelial cells with mRNA locally released from lipid nanoparticles, with aberrantly endothelially expressed S antigen responded to by building adaptive cellular immunity. The seeming absence of similar delayed skin reactions with DNA vector vaccines could find its explanation, if this alternative hypothesis would be proven correct, in the differential distribution, dosage, in vivo stability and cargo release characteristics of viral vector DNA vaccines as compared to lipid nanoparticles delivering mRNA.
Therefore, while “COVID arm” is a transient, localized, and clinically benign rare side effect upon vaccination with mRNA-based vaccines encoding the S-protein, the deciphering of the precise molecular and cellular mechanisms underlying this local phenomenon (directly accessible by biopsy) could still become the springboard of a more precise understanding of very rarely occurring endothelitic pathology observed with S-protein encoding vaccines, in general. In-depth analyses of the pathomechanisms underlying delayed skin reactions following mRNA-based vaccinations are called for, in consequence. It could evolve to be of significant clinical relevance to discern whether either (1) excipients exclusive to mRNA-based vaccines (with PEG as the most prominent candidate molecule) or (2) the S-protein encoded by both mRNA-based and DNA vector vaccines and potentially expressed on endothelial cells, or (3) both pathomechanisms in conjunction will constitute the triggering event(s) of very rare and rare adverse events affecting different vascular compartments and endothelial cells, with the spectra of clinical sequelae dependent on the tissue compartments afflicted.