The first large retrospective series was published on 04/29/20 by Galván-Casas et al,26 widely publicized, containing a dermatological atlas. With support from the Spanish Academy of Dermatology, dermatologists across the country included images of skin lesions from 375 patients with COVID-19 (suspected or confirmed), which were grouped into five main patterns: maculopapular eruptions (47%), urticarial rash (19%), EPL (19%), vesico-bullous (9%), and livedo/necrosis (6%).

The most commonly found pattern was maculopapular eruptions, showing a concomitant onset with general symptoms and an average duration of 8.6 days, similarly to urticarial conditions. Both were considered not useful for the diagnosis of COVID-19 because they are non-specific manifestations common to other common conditions, such as adverse drug reactions and other viral infections.

EPL lesions occurred in younger patients with mild systemic conditions, with a late-onset (mean of 12.7 days from the onset of general symptoms) and prolonged duration. Only one of the 71 cases reported prior erythema pernio (EP). They would be useful as an epidemiological and not a diagnostic marker.

Vesicular lesions appeared before the general symptoms in 15% of the cases in this series, with a mean duration of 10.4 days and relative diagnostic specificity since they are skin manifestations not reported in viral exanthems.

Livedo/necrosis includes a variable range of presentations related to COVID-19 or its complications, such as a hypercoagulable state and vascular damage. They are associated with an older age group and higher mortality.

Although there is great merit regarding the pioneering and significant number of cases, the study limitations are the inclusion of patients who were not confirmed cases of COVID-19 (suspected cases), mostly outpatients, evaluated photographic records, without histopathological analysis or exclusion of the main differential diagnoses.

On 05/04/20, de Masson et al.,27 with support from the French Union of Dermatologists and Venereologists, published a retrospective observational study in letter form that included images of skin manifestations of 277 patients with COVID-19 (252 suspected and 25 confirmed cases) from private clinics throughout France. Surprisingly, EPL lesions corresponded to the vast majority of cases (75%) and were even reported as the single manifestation of the disease. Histopathology study carried out in three cases showed lichenoid interface dermatitis with the presence of microthrombi. Other identified cutaneous patterns were: vesicular (15%), urticarial (9%), morbilliform (9%), petechial (3%), livedo reticularis (1%), and others (15%). Some patients were accounted for in more than one category. It is worth remembering that the French were the first to describe EPL lesions related to COVID-19, which could be a bias for the high frequency shown in this study, non-reproducible in other similar studies. Additionally, the high prevalence of EPL lesions is probably reflected in the large percentage of unconfirmed cases included in the analysis since such lesions predominate in oligo- or asymptomatic COVID-19 cases, which are seldom tested for laboratory confirmation of SARS-CoV-2 infection.

In collaboration with the American Academy of Dermatology and the International League of Dermatological Societies, Freeman et al.28 collected data on 716 cases from 31 countries through an electronic registry and the results were published on 07/02/20. Dermatologists provided data on 54% of the cases, with 89% of all registries originating from the USA. Only 34 Hispanic/Latino and 13 African-American patients were included, impairing the racial diversity required to understand the skin manifestations. Only 171 patients were confirmed COVID-19 cases, with the following frequency of manifestations: morbilliform (22%), EPL (18%), urticarial (16%), macular erythema (13%), vesicular (11%), papulo-desquamative (9.9%) and retiform purpura (6.4%). Most of the skin manifestations started concomitantly with the general symptoms; however, in 12%, the skin lesions were the first symptoms of the disease, showing the possible importance of skin lesions as a marker for the early diagnosis of COVID-19. Regarding the severity, EPL lesions were associated with milder systemic conditions, while 100% of cases with retiform purpura were hospitalized. Histopathology was carried out in 14 confirmed cases. The study limitations include evaluator bias, the impossibility of estimating the frequency/prevalence of skin manifestations, low rate of confirmed COVID-19 cases (25%), and little ethnic diversity, despite the apparent high representativeness of the 31 countries.

Similar to Spain and France, on 01/18/21, the Italian Society of Dermatology and Sexually Transmitted Disease also coordinated a national study that gathered data from 200 patients with skin manifestations. One of the inclusion criteria for Marzano et al.29 was the presence of COVID-19-related skin lesions confirmed by an experienced dermatologist. A total of 62% had a confirmed diagnosis of COVID-19. The main identified patterns were: maculopapular eruption/morbilliform exanthem (25.7%), EPL (24.6%), papulovesicular eruption (15.5%), urticarial eruption (10.2%), purpura/vasculitis (6.9%), livedo reticularis/racemosa (2.1%) and others (15%), which included pityriasis rosea-like, erythema multiforme-like, erythema nodosum-like, panniculitis and angioedema. Mucosal lesions were not identified. The skin condition appeared on average 14 days after the onset of systemic symptoms and lasted for an average of 12 days. EPL lesions lasted longer than the other patterns, occurred in a younger age group, and represented a lower risk of developing a severe systemic disease, all factors with statistical significance. The mean age of the livedo reticularis/racemosa and maculopapular eruption/morbilliform exanthem groups was also older (p < 0.01). One limitation of the study is the lack of diagnostic confirmation of SARS-CoV-2 infection in 36.5% of the sample due to economic factors that limited the extensive diagnostic testing in Italy and in several other countries.

Finally, the most recent population study, with very significant figures, whose results generated a lot of repercussions, was published on 01/14/21 by Visconti et al.,30 which included data from all over the United Kingdom and had a different design than those of the abovementioned publications.

The first arm of the study collected data from 336,847 users of the “COVID Symptom Study” application, in which they provided epidemiological, laboratory, and clinical information, including on the presence or absence of skin lesions identified by the users themselves. Only 2021 (7.4%) users had laboratory confirmation of COVID-19, and of these, 178 (8.8%) reported skin lesions. The prevalence was slightly higher among women. Additionally, the prevalence of skin lesions was significantly higher among those who tested positive for SARS-CoV-2 when compared to those who tested negative, justifying the inclusion of skin involvement in the list of symptoms in suspected COVID-19 infection due to its diagnostic, predictive factor, regardless of how often it occurs.

The second arm of the study sought to investigate in detail the skin manifestations through a questionnaire addressed to users who reported them and the sharing of photos. Of the 11,544 answered questionnaires, only 694 (6%) had a confirmed diagnosis of COVID-19. In this group, although the majority reported the onset of skin lesions together or soon after the general symptoms, 17% reported that the skin lesions appeared before the systemic symptoms, and in 21%, the skin condition was the only symptom. However, the study did not provide additional detailed information about these findings. There was no description of which skin manifestations occurred prior to the general symptoms and how long before they appeared. Neither did it specify which skin lesions were considered by the individual themselves as the only manifestation of COVID-19, nor did it clarify the reason why laboratory confirmation of the disease was performed in the absence of any suspected symptoms.

In this same arm of the study, 2,328 users shared photos of skin lesions; 260 images were randomly selected for analysis by dermatologists, and of these, 30 were discarded as not attributable to SARS-CoV-2 infection – examples: acne, molluscum contagiosum, herpes zoster, eczema, perioral dermatitis, impetigo, dermatophytosis, which shows the user's own limitations as an evaluator of their skin lesions. The final images were evaluated by four experienced dermatologists. The most frequent dermatological presentations were papular eruptions (41.2% including erythematous-papular and erythematous-vesicular), urticarial (30%), and acral (23.1%) eruptions.

To date, this is the first and only impactful publication stating the frequency and relevance of skin lesions as the only manifestation of SARS-CoV-2 infection. Therefore, caution and further studies on the subject are necessary to understand its actual importance and impact on the diagnosis of COVID-19.

The nature of the study has four main limitations: data based on self-assessment, not representative of the general population, disregarding of rare dermatological manifestations, and the impossibility of ruling out important differential diagnoses, such as adverse drug reactions.

In most studies, it is considered the most frequent skin manifestation of COVID-19. It includes exanthem (Fig. 1), rash, diffuse papules, pityriasis rosea-like, erythema multiforme-like, among others. It usually appears within the first days of the general viral infection symptoms or in up to two weeks, although there are sporadic reports of a late-onset (after one month). It correlates with mild systemic conditions and spontaneously improves within seven to ten days,26 without the need for specific treatment.

Figure 1.

Exanthem in a patient with COVID-19.

(0.06MB).

Although these are frequent manifestations, they are highly non-specific, since other viral infections are common causes of exanthems, as well as adverse drug reactions. The literature discusses the need to exclude such differential diagnoses before attributing the skin condition solely to COVID-19, even in confirmed cases of the disease. Another confounding factor is the viral reactivation that occurs in cases of drug reactions, such as the activation of HHV-6 in DRESS (drug reaction with eosinophilia and systemic symptoms). One hypothesis would be that the SARS-CoV-2 infection would lead to reactivation of latent viruses, such as HHV-6, and thus predispose to an adverse drug reaction.25,39

It is undeniable that this is the most often reported manifestation in large case series in the literature.40,41 With the arrival of the COVID-19 pandemic in Europe, there was an explosion of cases similar to “classic” erythema pernio, but with peculiar characteristics, different from the classic form: predominance in children and young individuals, at a time of mild temperatures (spring), with no personal history of erythema pernio, collagenosis or other predisposing conditions. Then, EPL associated with COVID-19 was described. Since then, there have been numerous publications on the subject, including those showing the presence of the viral spike in skin lesions,17 even in patients with a negative PCR result (swab). EPL is more frequent in healthy, oligo- or asymptomatic children and young adults with COVID-19, and it may be the only manifestation of the disease. Therefore, in the presence of these lesions, it is suggested that SARS-CoV-2 infection should be investigated. It appears on average after 10 to 14 days of the onset of general symptoms, lasts an average of 14 days, with spontaneous improvement.41 Most patients (in some series, up to 100% of them) have a negative PCR (swab) viral test result. This fact could be explained by the low viral load usually found in oligo- or asymptomatic COVID-19, in children, and in the later stages of the disease (10 to 14 days after symptoms onset). On the other hand, the absence of PCR positivity, therefore, of the laboratory confirmation of SARS-CoV-2 infection, raises the question of whether these patients were really infected. Immunohistochemistry and immunofluorescence studies may contribute to elucidating the relationship between EPL and COVID-19.42

Recently, the relationship between EPL and SARS-CoV-2 infection has been widely questioned in the literature. One of the main controversies is precisely the lack of proof of SARS-CoV-2 infection in these patients. Case series of EPL have been published, which had no epidemiological, clinical, or laboratory evidence to support the diagnosis of COVID-19,43 or even serological evidence, which could contribute to the pathophysiology of EPL.44 Although the presence of the viral spike in EPL lesions has already been demonstrated by immunohistochemical techniques, and even the complete viral structure has been identified by electron microscopy, these were specific reports in the literature - the vast majority of publications on EPL do not have this evidence. Regarding the pathophysiology, at the beginning of the pandemic, EPL, livedo, purpura, and necrosis lesions were interpreted as phases of the same spectrum of skin involvement. It is currently believed that they have distinct pathophysiological mechanisms.45

Infections, including viral ones, are known triggering factors for urticarial lesions. In this sense, SARS-CoV-2 infection could be a new triggering factor. Urticarial conditions associated with COVID-19 have been reported in patients without previous reports of the disease. A study in a private clinic reported an increased incidence of urticaria cases diagnosed during the pandemic, but no investigation was carried out regarding the presence of COVID-19 in these patients. Clinically, the lesions are disseminated, indistinguishable from other causes, and angioedema is rare. It is estimated that in 10% of patients, the skin condition may precede the general symptoms of COVID-19 by a few days, and, therefore, the detection of skin lesions could favor the early diagnosis of the infection. However, for the most part, urticarial lesions appear concomitantly with other symptoms. There is no age group predilection, and it is associated with mild systemic conditions with low mortality. Spontaneous improvement is reported within seven days or with the use of antihistamines and/or low-dose systemic corticosteroids.46

Although rare, vesico-bullous eruptions are considered a specific pattern of skin lesion associated with COVID-19, as they are not commonly seen in viral exanthem and drug reactions (main differential diagnoses). They consist of localized or disseminated vesicles or bullae (Fig. 2), which can affect the palmoplantar region, sparing mucous membranes. Their onset occurs approximately three days after the general symptoms and lasts for seven to 14 days, with spontaneous improvement and are associated with a mild to moderate systemic condition. In a series of cases, an attempt was made to isolate the viral agent from the bullous content using the PCR technique, without success. It is crucial to rule out other viral infections, especially herpetic ones, as a case of disseminated herpes with pneumonia has already been reported in a COVID-19 patient, showing the importance of the differential diagnosis.47

Figure 2.

Vesico-bullous lesions in a patient with COVID-19.

(0.06MB).

The cutaneous vascular manifestations associated with COVID-19 comprise a wide spectrum, ranging from transient to fixed livedo reticularis, livedo racemosa, purpura, and necrosis of the extremities (Fig. 3). They are rare manifestations associated with severe systemic conditions, which occur late in elderly patients, those hospitalized in an intensive care environment, with multiple comorbidities and laboratory alterations (for example coagulopathies, elevated D-dimer levels). They are frequently associated with other thrombotic events such as deep vein thrombosis, ischemic stroke (cerebrovascular accident) and disseminated intravascular coagulation, which is responsible for the high mortality in this group. Therefore, skin findings in patients with a potentially fatal disease are very specific and distinct from those found in the outpatient profile. This happens because probably the same microangiopathic and thrombotic mechanisms occur simultaneously in the main target organs, such as the lung and kidneys, and in the skin. Therefore, the treatment of skin lesions consists of treating the underlying disease itself.16,22,26 A summary of the main characteristics of the dermatological manifestations described above is found in Table 4 (Main characteristics of the skin manifestations associated with COVID-19).

Figure 3.

Livedo racemosa, purpura and necrosis of the extremities in a patient with COVID-19.

(0.12MB).

Oral lesions during the course of COVID-19 have also been described and collected in review articles. It is known that SARS-CoV-2 can be detected in saliva using the PCR technique, which is currently used for diagnostic purposes. Moreover, ACE2, a functional viral receptor in the body, is present in the oral mucosa, with greater density on the dorsum of the tongue and minor salivary glands, which would justify the presence of mucosal lesions. Dysgeusia was the first oral symptom recognized in COVID-19, with a widely varying frequency, from five to 88%. Subsequently, numerous oral lesions were also described, with the most frequent being enanthema, petechiae, aphthoid lesions or ulcers, herpetiform lesions, candidiasis, and oral lesions associated with Kawasaki disease. They are usually symptomatic (pain, burning sensation or pruritus) and appear four days before or up to 12 weeks after the general symptoms. The treatment depends on the pattern of the lesion identified. Inadequate oral hygiene, immunosuppression, reactivation of latent viruses (for example, herpes virus), opportunistic infections, stress, trauma secondary to intubation, COVID-19 hyperinflammatory response are predisposing factors. Further studies are required to define which oral manifestations are directly linked to SARS-CoV-2 infection and which are indirectly due to the presence of other mucosal affections that, in these cases, occur in the context of COVID-19. In this sense, research is essential to exclude such conditions. It is suggested that vascular alterations and microthrombus-induced manifestations, such as petechiae and ulcers and enanthema (similar to viral exanthem) may be directly associated with the pathophysiology of SARS-CoV-2.34–48

Widely known in dermatology for a long time, telogen effluvium (TE) can be triggered by metabolic, nutritional changes, drug use, and several systemic conditions, including infections. In COVID-19, it is thought that the inflammatory cytokines released and the use of medications, such as heparin, may be involved in the TE mechanism. There is a report of an increase in the incidence of TE by almost threefold during the pandemic. It is estimated that 10% of patients with COVID-19 develop TE in the weeks and months following the infection, especially in patients with comorbidities, but it can occur even in subclinical conditions. To date, information about the onset, evolution, and prognosis of TE associated with COVID-19 are indistinguishable from TE due to other known causes. There is no peculiarity or specific characteristic in TE induced by SARS-CoV-2 infection. Therefore, the approach, treatment, and follow-up of these patients should be the same adopted for TE in dermatological practice.49,50

In addition to the manifestations most often described and grouped into “patterns” in the large case series, prospective and systematic review studies, there are many isolated cases reports in the literature on several skin conditions possibly associated with COVID-19. Due to the methodological limitation inherent to case report-type publications, the actual role of SARS-CoV-2 infection in these entities cannot be affirmed, that is, whether they are in fact attributable to COVID-19 or not. The isolated conditions associated with COVID-19 that have been published in case reports are shown in Table 5 (Dermatological conditions associated with COVID-19 – Case reports).

The first reports were of cases of urticaria, which even preceded clinical symptoms,51 urticarial vasculitis,52 and angioedema.53 Also, cases of pityriasis rosea or pityriasis rosea-like lesions,54 with one report showing the presence of SARS-CoV-2 by immunohistochemical techniques 55 and another pointing to the possible role of HHV-6 reactivation and other viruses in these cases.56

There have also been reports of small-vessel vasculitis associated with COVID-19,57 leukocytoclastic vasculitis,58 IgA vasculitis,59 granulomatosis with polyangiitis (formerly called Wegener's disease),60 eosinophilic granulomatosis with polyangiitis (formerly called Churg-Strauss syndrome),61 Schamberg purpura62 and purpura fulminans.63

Regarding conditions that resemble drug reactions, patients with COVID-19 and erythema multiforme have been described,64 as well as case series of acute generalized exanthematous pustulosis (AGEP),65 reports of toxic epidermal necrolysis (TEN)66 and SDRIFE (symmetrical drug-related intertriginous and flexural exanthem).67

There are also case reports in specific topographies – ungual involvement, with a description of the “red half-moon sign”68; palmoplantar involvement with erythrodysesthesia69; follicular eruption,70 unilateral exanthem,71 unilateral livedo reticularis72 and pressure ulcers in unusual locations (such as the sternum, knees) from prolonged pronation.73 Dermoscopic descriptions of purpuric papular eruption74 and EPL lesions have been published.75

There are also reports of herpes zoster76 and isolated reports of Grover's disease-like conditions,77 Melkersson-Rosenthal syndrome,78 erythema nodosum,79 sarcoid lesion,80 Gianotti-Crosti syndrome,81 seborrheic dermatitis,82 annular lichen planus,83 erythema annulare centrifugum,84 granuloma annulare85 and rapidly progressive alopecia areata.86

Superficial perivascular dermatitis with mild lymphocytosis, thrombosed vessels with neutrophilic and eosinophilic debris; interface dermatitis, superficial and deep perivascular lymphocytic infiltrate, with or without vasculitis (Fig. 4).

Figure 4.

Histopathological findings of exanthem in a patient with COVID-19. (A), Leukocytoclastic vasculitis in the capillary vessels of the superficial dermis and erythrodiapedesis (Hematoxylin & eosin, ×100). (B), Capillary vessels containing intact and degenerated neutrophils partially destroying the wall, extravasated red blood cells, and the presence of eosinophilic amorphous material in the vascular lumen, suggestive of fibrin thrombus (Hematoxylin & eosin, ×400).

(0.24MB).

In the epidermis, presence of parakeratosis and hyperkeratosis, with apoptotic keratinocytes and, eventually, epidermal necrosis. Dense lymphocytic infiltrate in the superficial and deep dermis, also in the hypodermis, with a perivascular pattern and signs of endothelial activation. Sometimes, eccrine lymphocytic infiltrate is similar to that seen in eccrine lymphocytic hidradenitis. Other common findings are erythrodiapedesis, edema in the papillary dermis with the formation of a subepidermal bulla and increased interstitial mucin deposition. The presence of vascular microthrombi in the capillaries of the superficial dermis is rare, a well as in dermal venules.

Perivascular lymphocytic infiltrate with some eosinophils and edema in the superficial dermis, eventually with vacuolar interface dermatitis and keratinocyte necrosis, with a pattern similar to that of erythema multiforme. In urticarial vasculitis, there is extravasation of red blood cells with perivascular neutrophilic infiltrate, macrophages, necrosis, and fibrin deposits.

Acanthosis, vesicles due to vacuolar degeneration of the basal layer with dyskeratotic, multinucleated cells (suggestive of viral inclusion) without inflammatory infiltrate. In these cases, the differential diagnosis with herpetic dermatoviruses is essential.

Thrombotic vasculopathy with extensive deposition of C5b-9 and C4d (immunohistochemical technique) in the microvasculature.

Throughout the pandemic, SARS-CoV-2 infection was observed to affect more men than women. Men are twice as likely to be admitted to intensive care units. The association between the androgen pathway and SARS-CoV-2 infectivity has been described. Therefore, such predisposition to androgen-mediated infection could justify the high rates of hospitalization and mortality concentrated in men. Therefore, androgen expression in the form of alopecia (androgenetic alopecia – AGA) could, by itself, be a risk factor for COVID-19, and a high frequency of AGA has been reported among hospitalized men; and also the use of anti-androgens could be a protective and/or therapeutic factor in the evolution of these patients. In this sense, a published study showed that men with AGA or benign prostatic hyperplasia using a 5-alpha-reductase inhibitor who contracted COVID-19 showed a significant reduction in symptoms in the outpatient setting. In a case-control study in a hospital setting, men using 5-alpha-reductase inhibitors had a lower chance of being admitted to the ICU (Intensive Care Unit), suggesting this protective effect. However, it is still uncertain whether the introduction of anti-androgens after SARS-CoV-2 infection would provide any benefit. Apparently, dermatologists should encourage their AGA patients to maintain anti-androgen use during the pandemic. This is undoubtedly a very promising field of research.88–90