A recent expert consensus panel defined the duration of DIRs as arising from a normal state 2–4 weeks or longer post-injection . Clinically, they noted that DIRs in HA-based filler injections manifest as erythema, painful nodules, induration, and eyelid edema. The authors also acknowledged that certain triggers might be associated with the onset of DIRs, which can include viral infection, active sinusitis, low-quality products, combinations of different products, improper technique, and past and current dental procedures. Lastly, it was mentioned that DIR can occur in the skin at different sites of injection, chiefly dermal and subcutaneous placement, with adipose tissue being the most common placement site .
Based on the historical and temporal events as detailed above, we believe that three of the aforementioned cases demonstrate DIR of HA fillers, triggered by exposure to COVID-19 spike protein. In the one case, the patient exposure to the COVID spike protein was likely achieved following native infection with community-acquired COVID resulted in seroconversion, followed within weeks by DIR. The HA dermal filler involved in this case is classified as monophasic (particles of the same size) 20 mg/ml, non-animal (NASHA) derived, and crosslinked. With a prior history of a recalcitrant eczematous dermatitis, she was being treated with periodic subcutaneous injections of dupliximab at the time of COVID-19 infection. DIR persisted for weeks after COVID infection resolved, prompting evaluation and treatment. Persistent areas of involvement included the previously-injected areas of the tear trough, periorbita, and medial cheeks. The treatment course was difficult, requiring multiple rounds of oral corticosteroids, multiple sessions of intralesional recombinant hyaluronidase, and physical thermal dissolution of the HA by use of a bipolar radiofrequency microneedle device. Although symptomatology greatly improved, evanscent swelling in the peroribtial area still persisted, likely due to residual HA product.
In the second case, the patient was a subject following the Moderna mRNA-1273 trial protocol. Although the trial was conducted as a double blinded, placebo-controlled trial, the patient felt certain that her first vaccine dose to be to be the actual vaccine. Furthermore, she experienced lingering pain in that arm for days following the injection. Saline served as placebo control in this trial because it is highly unlikely to trigger any biologic response.
Eight days following the injection per protocol (without constitutional symptoms or known community COVID exposure), she began to experience the onset of DIR in the periorbital area. This was followed within days by inflammation occurring in previous facial injection sites: firstly the earlobes (performed 18 months prior to the vaccination) and secondly more intense inflammation in the malar cheeks, tear troughs, nasolabial folds, and lips (sites more recently injected 5 weeks prior to the vaccination). Per patient history, she was barred to proceed with the second vaccination dose in the trial due to the concomitant usage of oral corticosteroids, which were used in high doses to qwell the ongoing DIR. A scheduled blood test was carried out per protocol and did not reveal seroconversion following the first dose. Given the scenario, the patient and the primary author felt certain she received the vaccine; however, unblinding of the data revealed that she was given saline placebo. Without confirmation of seropositiviy, and in the absence of any historical trigger factors, the definitive cause of DIR remains unknown. The severity of her inflammatory reaction necessitated multiple treatment measures which although ultimately successful, came at great expense.
It is worth noting that two other cases of DIR reaction to injectible HA fillers occurred (labelled as serious adverse events) in the mRNA-1273 trial . The Moderna FDA-Briefing document details three related cases of hypersensitivity reaction to previous facially-injected dermal filler. In one case, a 46-year-old caucasian female received injections in her cheeks of Juvederm XC in March 2020. In the study, she received her first dose on August 31, 2020, and her second dose on September 29, 2020. Between September 30 and October 5 2020, she developed significant bilateral cheek swelling with no rash, pain, tenderness, oral or respiratory symptoms. Symptoms were considered moderate and the reaction designated as serious. The study participate received diphenhydramine and methylprednisolone and recovered (data on file with Moderna).
In the second Moderna trial case, a 51-year-old caucasian female received Juvederm and Botox injections in her cheeks on October 2, 2020. The study participant received her first vaccine dose on September 12, 2020, and her second dose on October 15, 2020. Between October 17 and 22 she developed bilateral facial swelling with the left side being more pronounced. Swelling was measured at 125 mm on day 3 and 55 mm on day 4 post-vaccination. Symptoms were considered moderate and the reaction designated as serious. The study participant received Prednisolone and recovered (data on file with Moderna). In a third case involving lip swelling of a 29-year-old female that occurred 2 days after vaccine injection, the reaction was termed angioedema and was classified as medical significant. This case had a history of prior dermal filler placement to the lips, of unknown duration prior to the trial .
As of August 2020, thirty potential vaccines against COVID-19 were in clinical trials with another 139 in pre-clinical development, including both gene-based (mRNA, DNA) and protein-based candidates . Genetic approaches have the benefit of eliciting antibodies and CD4+ helper T cells, they recruit CD8+cytotoxic T cells, through the major histocompatibility class I pathway . COVID-19 prevention efforts represent the first large scale usage of mRNA technology for vaccine development. As such, understanding of complete clinical response to mRNA vaccines is unknown. Published data from the three major completed COVID vaccine trials Phase I,II, and III in the United States/UK/Brazil/South Africa include mRNA-1273 vaccine (Moderna), ChAdOx1 nCoV-19 vaccine (Astrazenica) and RNA vaccine BNT162b1(Pfizer-BioNTech) has mandatorily included reporting of adverse events [13,14,15,16,17,18]. Table 2 summarizes the known cutaneous adverse events from these studies. Transient urticaria and pruritus were noted for the mRNA-1273 trial, while rosacea flaring occurred in the ChAdOx1 nCoV-19 vaccine trial. No incidences of DIR were reported publically for the Pfizer or Astrazenica trials at the time of this submission. Presumably, this is because either they did not occur, or as in the case of mRNA-1273, the numbers did not meet the minimum threshold incidence within the protocol for reporting as serious adverse events.
The Vaccine Adverse Event Reporting System (VAERS) is an online, self-reporting database developed by the CDC and utilized for reporting adverse events from vaccination, performed on the general public . Vaccination cards given after COVID immunization contain the website address and strongly encourage self-reporting of any side effects. A query was performed of VAERS from the inclusive dates of Pfizer and Moderna vaccine public release until Jan 10th 2020, using database terms to identify cutaneous post-immunization reactions. Of the 641 records of patients who self-reported skin-related issues following receiving either the first or both vaccine doses, 152 of those cases were found to include the term (or synonyms for term) “edema” with regards to visible skin swelling (Fig. 10). DIR, though not characterized by extensive soft tissue involvement, can be misdiagnosed as edema, facial edema or angioedema because of the presence of acute onset swelling involving the skin and lips.
DIR as a response to other viral vaccination has been reported sparsely in the literature. One report from Brazil detailed the characteristic induration, edema, and erythema of DIR in a patient in weeks following the administration of the influenza vaccine . The author has personal experience with two patients experiencing DIR with calcium hydroxylapatite (Radiesse, Merz,Germany) filler treatment to the face and hands, respectively, following the first dose of administration of the shingles vaccine (author observation).
Of significance, a variety of HA products were used in the aforementioned cases, and contain different stabilizing/crosslinking formulations, purification, and composition. Artzi and colleagues noted (based on a review of available literature), that the reported rate DIR prior to 1999 were 0.7% and that rate began to decrease after the introduction of highly purified HA products to 0.2%. . In the 400 subjects studied retrospectively, the authors did not report a DIR related to post viral illness or vaccination . However, several studies have shown that the biphasic highly-crosslinked Juvederm vycross filler product line has a higher than expected rate of DIR [3, 5, 22]. It is has been reported that HA-based fillers with a low-molecular weight-degraded products have higher pro-inflammatory activity .
Recent reports in the literature have revealed an incidence of DIRs to Juvederm Volbella of 1.0% per patient and 0.8% per syringe, which is higher than the previously reported incidence of 0.02% and more compatible with the 4.25% incidence of DIRs to Juvederm Volbella previously reported by Artzi et al. [5, 21]. Analysis of the U.S. Food and Drug Administration (FDA) MAUDE database by Ortiz et al. identified the Juvederm family of dermal HA fillers with a high risk profile amongst HA fillers for development of DIR . Comparatively, the incidence rate of hypersensitivity reactions to the monophasic NASHA fillers has been reported at 0.8%, with the rate for delayed reactions reported at 0.3%. Time to onset of delayed nodules ranged from 1 month to 3 years after HA implantation .
As longevity of HA filler increases with newer products, DIR incidence could increase, simply as a function of the presence of a persistent nidus for inflammation. The need for accurate medical recording of anatomic placement and specific filler type (specific HA vs other non-HA fillers) will become more important. A recent small study, utilizing a novel HA-specific Magnetic Resonance (MR) imaging modality, identified filler that had been placed in to subjects 6 years prior . Subjects, which denied filler placement for 2 years prior and denied placement in certain anatomic areas, were indeed found on MR to have residual HA filler in those areas . Therefore, it is reasonable to expect that in future vaccine trials, effort should be dedicated to get an accurate medical history of dermal filler placement as part of the subject screening process.
Cutaneous manifestations of COVID-19 can present with a constellation of cutaneous signs, occurring concurrent/after the onset of the characteristic pyrexia, anosmia/hyposmia, ageusia, and respiratory symptoms . Although chilblains-like eruptions of the digits/toes, petechiae, and acro-ischemia, have been reported, only urticaria has been reported along the spectrum of delay-type hypersensitivity reactions described in the two cases above  (Fig. 11). In a recent meta-analysis of 44 articles with 507 patient cases, the most common cutaneous findings were polymorphic eruptions, erythema, chilblain-like lesions, and urticarial lesions, occurring on an average of 9.92 days (range: 1–30) after the onset of systemic symptoms . In humans, ACE2 is expressed in the skin in multiple cell types. The ARCHS4 database is publicly available and has demonstrated the resident skin cell types associated with high levels of ACE2, including fibroblast and keratinocytes  (Fig. 12). Interestingly, Li et al. showed that adipose tissue has also been shown to contain high levels of ACE2 , wherein may lie a majority of the filler placed anatomically in the face.
The Zhao study also evaluated the role of ACE2 constituitive expression in the cutaneous manifestations of COVID-19 . They found higher than expected levels of ACE2 expression in keratinocytes of COVID-19 patients who also had detectible viremia and skin lesions .
One possible explaination for COVID spike protein related to DIR with HA fillers could be explained by the genesis of the HA filler granuloma, a seminal imflammatory event. Failure of effective phagocytosis of longstanding HA, coupled with biofilm formation, and direct T cell activation common of bio-implants can lead to fibrosis and granuloma formation, favoring a TH1/CD8+T cell presence [28, 29]. In tissue, including the skin, a relative higher concentration of ACE2 are present to maintain immune homestasis, by regulating the production of pro-inflammatory angiotensin II relative to levels of the metabolites, angiotensin 1–7 (anti-inflammatory). However, ACE2 is a ligand for the COVID spike protein and active infection can bind available ACE2, further tipping the balance in favor of angtiotensin II and development of a pro-inflammatory response. Additional studies have detailed the presence of ACE2 in the cutaneous endothelium of mid/deep dermal and subcutaneous microvasculature . Complement-mediated destruction of cutaneous microvasculature, secondary to docking by SARS-CoV-2 spike protein, has also been demonstrated . The implantation of dermal HA filler is known to stimulate angiogenesis and retain neovascularity in human and murine models . Therefore, the possibility exists that higher (than baseline) dermal and subcutaneous levels of ACE2 are localized in the vicinity of acute and chronic injected cutaneous HA filler boluses, setting the stage for an peri-bolus filler DIR triggered by occlusion occurring post-vaccine or during/after active COVID-19 infection.
Based on this concept, in the third case with evolving DIR post Moderna vaccination, we investigated a hypothesis that ACE-I could potentially modulate DIR as a treatment option. Since ACE2 is bound by spike protein, it is essentially otherwise engaged and unable to convert angiotensin II to angiotensin 1–7. By reducing the levels of angiotensin II with ACE-I, this burden on ACE2 may be lessened, mitigating the effects of rising levels of spike protein. In this single case ancedote, treatment with an oral ACE-I may have had a positive effect in rapidly reducing the edema and inflammatory response secondary to rising angiotensin II levels, in a non-immunosuppresive way, unlike CS which could potentially blunt antibody response to the virus. As an additional mechanism of reducing interstitial edema, lisinopril decreases angiotensin II-induced aldosterone secretion by the adrenal cortex, which leads to an increase in sodium excretion and subsequently increases water outflow.
The literature details that a viral vaccine or illness can serve as a DIR triggering event. Published works have documented cases after viral illness [3, 6, 32, 33]. Turkmani et al. described 14 cases of HA DIR following an influenza-like illness . In all cases, subjects had no allergies or other predisposing autoimmune or infectious conditions. HA filler was present 2–8 months prior to developing the DIR, and patients experienced DIR within 3–5 days of developing the influenza-like illness. A variety of different branded HA fillers were used in these patients, and the most common anatomic areas affected were the cheeks, tear trough, and lips. In this series, most of the patients responded to oral corticosteroids only, while a few required addition intervention with intralesional hyaluronidase injections.
Bhojani-Lynch also described a series of 5 patients with DIR, 4 of which reported an influenza-like illness a few days prior to the onset of inflammation . The author used a similar management approach of oral CS to suppress inflammation quickly, utilizing the hyaluronidase for persistent or non responsive areas of inflammation . In the COVID cases reported in the USA, both oral and IV prednisone clinically improved edema and pain. Whereas the SARS-CoV-2 targets the ACE2 receptor for cell entry, Xiang et al. described CS as one part of a multi-step approach for COVID-19 respiratory symptom management through ACE2 agonistic activity . Of potential clinical significance, Artzi et al. made mention of smaller nodules being more frequently observed in conjunction with viral illness or post dental procedures, which tended to heal spontaneously .
One final clinical observation in our three confirmed COVID-related cases is with regards to time of onset of the symptoms. There is a significant difference in the time of onset of DIR in the first case, which started weeks after COVID seroconversion and the vaccine-related cases, which started days after vaccination. Case 4 interesting did not show evidence of DIR until the second dose. One potential explaination may be due to the more rapid rise in serum (and subsequent dermal concentration) of the spike protein in the vaccination cases, because of the rapid production due to the nature of the mRNA vaccine. Native COVID exposure and seroconversion may result in a longer rise in serum protein levels, as the virus takes more time to evade the host immunosurveillance, replicate, and produce appreciable levels of spike protein.