Skin Manifestations in COVID-19 Patients: Are They Indicators for Disease Severity? A Systematic Review

Introduction: Until now, there are several reports on cutaneous manifestations in COVID-19 patients. However, the link between skin manifestations and the severity of the disease remains debatable. We conducted a systematic review to evaluate the temporal relationship between different types of skin lesions and the severity of COVID-19. Methods: A systematic search was conducted for relevant studies published between January and July 2020 using Pubmed/Medline, Embase, and Web of knowledge. The following keywords were used: “SARS-CoV-2” or “COVID-19” or “new coronavirus” or “Wuhan Coronavirus” or “coronavirus disease 2019” and “skin disease” or “skin manifestation” or “cutaneous manifestation.” Results: Out of 381 articles, 47 meet the inclusion criteria and a total of 1,847 patients with confirmed COVID-19 were examined. The overall frequency of cutaneous manifestations in COVID-19 patients was 5.95%. The maculopapular rash was the main reported skin involvement (37.3%) commonly occurred in middle-aged females with intermediate severity of the disease. Forty-eight percentage of the patients had a mild, 32% a moderate, and 20% a severe COVID-19 disease. The mild disease was mainly correlated with chilblain-like and urticaria-like lesions and patients with vascular lesions experienced a more severe disease. Seventy-two percentage of patients with chilblain-like lesions improved without any medication. The overall mortality rate was 4.5%. Patients with vascular lesions had the highest mortality rate (18.2%) and patients with urticaria-like lesions had the lowest mortality rate (2.2%). Conclusion: The mere occurrence of skin manifestations in COVID-19 patients is not an indicator for the disease severity, and it highly depends on the type of skin lesions. Chilblain-like and vascular lesions are the ends of a spectrum in which from chilblain-like to vascular lesions, the severity of the disease increases, and the patient's prognosis worsens. Those with vascular lesions should also be considered as high-priority patients for further medical care.


INTRODUCTION
A viral outbreak caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) emerged from Wuhan, China in late December 2019 (1). The disease was named coronavirus disease 2019 (COVID- 19) by World Health Organization (WHO) and was declared as a pandemic on 11 March 2020 (2). After 1 year from the beginning of the pandemic, the full spectrum of COVID-19 presentations and its relationship with disease severity is still unknown. Fever, cough, chills, dyspnea, myalgia, and sore throat are the most common clinical presentations of COVID-19 and as time goes on, different other manifestations have been reported (3). Recently, skin lesions have been described as potential manifestations of COVID-19 (4)(5)(6). The cutaneous changes reported to date include maculopapular rash, vesicular lesions, urticaria-like lesions, and chilblain-like lesions (4)(5)(6)(7)(8). Some of these skin manifestations arise before the signs and symptoms more commonly associated with COVID-19, suggesting that they could be presenting signs of COVID-19 (9). However, the link between skin manifestations and the severity of the disease remains debatable. Due to the great variety of reported dermatologic presentations as well as the inconsistency of data on the association between skin presentations of COVID-19 with poor outcome, we aimed to conduct a comprehensive systematic review on the clinical and histopathological characteristics of skin manifestations in relation to other features of confirmed COVID-19 patients and to evaluate the temporal relationship between different types of skin lesions and the severity of COVID-19.

METHODS
This review conforms to the "Preferred Reporting Items for Systematic Reviews and Meta-Analyses" (PRISMA) statement (10). Registration: PROSPERO (pending registration ID: 215422).

Search Strategy and Selection Criteria
To investigate the prevalence and characteristics of cutaneous manifestations in COVID-19 patients, a systematic search was conducted for relevant studies published between January and July 2020 using Pubmed, Embase, and Web of knowledge.
The following search terms were used (designed using MeSH keywords and Emtree terms): "SARS-CoV-2" or "COVID- 19" or "new coronavirus" or "Wuhan Coronavirus" or "coronavirus disease 2019" and "skin disease" or "skin manifestation" or "cutaneous manifestation." Only studies included if they contained data about the skin manifestation in patients with confirmed COVID-19. There were no language restrictions. We got help from the Google Translate system for non-English papers. Review articles, duplicate publications, and articles with no relevant data were excluded from the analysis. Two authors independently screened the remaining articles. Finally, selected data were extracted from the full-texts of eligible publication by other investigators of the team.

Data Extraction
Data about the first author's name, date of publication, country, number of COVID-19 patients, number of cases with skin manifestations, age, gender, location and type of skin manifestations, associated cutaneous symptoms, the onset of skin lesions with systemic symptoms, the median duration of the lesions, treatment strategies and main histological findings of the lesions as well as comorbidities, associated symptoms, drug history, laboratory findings, severity and outcome of the patients were selected for further analysis. All cutaneous presentations related to COVID-19 were categorized into six groups: chilblain-like, vesicular, urticaria-like, maculopapular, vascular, and miscellaneous (lesions that we couldn't subscribe to any of the groups). Petechiae, purpura, livedo, and necrosis were classified into vascular lesions. Two authors (PJ, BH) independently extracted the data from the selected studies. The data was jointly reconciled, and disagreements were discussed and resolved between review authors (PJ, BH, MJN).

Quality Assessment
The critical appraisal checklist for case reports provided by the Joanna Briggs Institute (JBI) was used to perform a quality assessment of the studies (11).

RESULTS
At the first round of review, 381 articles were selected. After removing the duplicates and studies that did not meet the entry criteria, 88 full texts were finally selected for further assessment. Of these, only 47 articles had the characteristics appropriated for systematic review and were entered into the data extraction (Figure 1). Most of the studies were case reports (47%, N: 22) followed by case series (42.4%, N: 20), retrospective hospital/private section-based study (6.4%, N: 3), and crosssectional (4.2%, N: 2). Thirteen articles were originated from Italy, 11 from Spain, 10 from France, 5 from the USA, and others from Belgium, China, Thailand, Kuwait, Indonesia, Russia, Turkey, and Singapore. Information of the 47 analyzed articles can be found in Table 1.
A total of 1,847 patients with confirmed COVID-19 (based on positive RT-PCR or positive antibody tests) were examined in 47 articles, of which 597 patients had different skin manifestations. The overall frequency of cutaneous manifestations in COVID-19 patients was 5.95%.

Characteristics of the Cutaneous Lesions in Confirmed COVID-19 Patients
The maculopapular rash was the main reported skin involvement (37.3%) followed by chilblain-like lesions (18.4%). The prevalence rate of vesicular and urticaria-like lesions was 15% ( Table 4).
The mean age of patients with cutaneous manifestations was 53.3 (ranging from 16 to 92) years. Chilblain-like lesions were more common in younger patients (mean age: 40.7 years) and vascular lesions were more common in the elderly (mean age: 72.3 years). The prevalence of skin lesions was slightly higher in females than males (54 vs. 46%). Urticaria-like, chilblain-like and miscellaneous lesions were more frequent among females ( Table 4). Vascular lesions were more frequent in males (61%). The prevalence of vesicular and maculopapular lesions was almost the same in men and women (51 and 49%).
Trunk, lower limb, and upper limb were the main involved regions. Chilblain-like and vascular lesions were more common in acral areas and except for maculopapular lesions, others were commonly located in the trunk. The maculopapular lesions were more common in extremities. The involvement of palms and soles were rare. Mucous membrane involvement was reported in all types of skin lesions particularly maculopapular and vascular lesions, but it was not reported in chilblain-like lesions ( Table 4). Vesicular rashes could have diffused polymorphic or localized monomorphic patterns (27,37).
In the majority of patients (89.5%), dermatologic manifestations presented after (55%) or at the same time (34.5%) with the onset of systemic symptoms of COVID-19. Urticaria-like lesions appeared usually as a concomitant symptom (47%). In 3.5% of patients particularly with chilblainlike lesions, skin manifestations were the only presentation of COVID-19. In 7% of patients, skin manifestations occurred before the systemic symptoms, particularly in chilblain-like lesions ( Table 4).
The median duration of skin lesions was about 9 days ranging from 1 to 18 days ( Table 4). Urticaria-like lesions had the least duration (5 days) and chilblain-like lesions had the most duration (14 days).
No skin biopsy or histological examination of urticarialike lesions was performed. Therefore, the following results are related to other types of skin lesions.
Perivascular lymphocytic infiltration, spongiotic and interface dermatitis, and vacuolization or keratinocyte necrosis were the common histologic findings in skin biopsies, except for vesicular lesions. In vesicular lesions, the absence of inflammatory infiltrates, atrophic epidermis, and hyperkeratosis was reported. In almost all types of lesions (except maculopapular and vesicular lesions) thrombotic vasculopathy and red blood cell extravasation were present. Langerhans cell aggregations  were seen within the epidermis in maculopapular lesions. Telangiectatic blood vessels were seen within the dermis of vascular and miscellaneous lesions. Virally-induced cytopathic alterations were absent according to reports on the miscellaneous category. Striking vascular and dermal deposits of complement factors (C5b-9, C3d, C4d) and IgM were present in four vascular rashes. Some studies performed an RT-PCR test on skin samples of maculopapular and vesicular lesions and the results were all negative for SARS-CoV-2. More details can be found in Table 2.

Characteristics of the Confirmed COVID-19 Patients With Skin Manifestations
The overall prevalence of comorbidities among patients with skin manifestations was 17.9% ( Table 4). Hypertension (39%), diabetes (23%), and dermatologic diseases (20%) were the most frequent comorbidities, respectively. Utmost cases with comorbidity were across the patients with maculopapular lesions (40%). Previous dermatologic illnesses were most common in patients with vesicular lesions ( Table 4). Cardiovascular disease, hypertension, and obstructive lung diseases were common comorbidities amongst patients with vascular lesions ( Table 4). Rheumatologic diseases were more frequent in patients with chilblain-like lesions (30%). Diabetes was seen commonly in patients with urticaria-like lesions (46%).
Elevated D-dimer was the main laboratory finding in most of the cases, especially in patients with chilblain-like (100%) and vascular (46%) lesions. Disruption of coagulation condition (increase in PT, INR, and fibrinogen) was reported in patients with chilblain-like and vascular lesions (Tables 3, 4).
Regarding the drug history and medication regimen used for COVID-19, data of 389 out of 597 cases were available, most of which related to maculopapular and urticaria-like lesions. Fifty-two percentage of all cases and 72% of cases with chilblain-like lesions underwent symptomatic treatment with paracetamol, etc., or recovered without any medication. Chloroquine/hydroxychloroquine was the most common medication used in patients (45%). Details in Tables 3, 4.
Most patients had mild disease (48%). The majority of patients with chilblain-like lesions had mild disease (82%) and the majority of patients with vascular lesions had severe disease (68%). Also, most of the patients with maculopapular lesions were moderate (43%) regarding severity (Tables 3, 4).
The overall mortality rate among COVID-19 patients with cutaneous manifestations was 4.5%. Patients with vascular lesions had the highest mortality rate (18.2%) and patients with urticarialike lesions had the lowest mortality rate (2.2%).
Details indicating characteristics of the lesions and the patients are shown in Tables 2-4.

DISCUSSION
After 1 year from the beginning of COVID-19 pandemic, the world is still facing a crisis. According to the current literature, more than half of the patients are asymptomatic leading to uncontrolled transmission of the virus (57-60). Recognizing COVID-19 related cutaneous manifestations may assist clinicians in early diagnosis of disease, before the development of respiratory symptoms, and may also be used to identify complications requiring treatment. The current study found that 10.5% of the COVID-19 patients reported skin lesions before the initiation of other symptoms or as their chief complaint. On the other hand, considering cutaneous manifestations is important to make the right diagnosis; as Joob et al. reported a COVID-19 patient with petechiae misdiagnosed with dengue fever (13). Our data demonstrated that 34.5% of cutaneous manifestations occurred at the same time with other symptoms particularly urticaria-like lesions (47%). It may suggest that urticaria-like lesions may be a diagnostic sign for COVID-19. The rest of the skin manifestations appeared later in the course of the disease Vesicular (24) Head (4), anterior trunk (21), posterior trunk (14), arms (8), legs (10), palms-soles (2) 18 disseminated pattern (small papules, vesicles, and pustules with varying sizes of up to 7-8 mm diameter, different stages of the lesions appeared simultaneously), 6 localized pattern (monomorphic lesions, of up to 3-4 mm diameter, at the same stage of evolution, mostly trunk involvement) Pruritus (20), Asymptomatic (4) Before (2), At the same time (3), After  Topical corticosteroids (2) and mainly after the initiation of systemic symptoms (55%) in our review. Galván Casas et al. suggested the chilblain-like and vesicular lesions as epidemiological markers for the disease (24). However, in our study, vesicular lesions (74%) were the most important cutaneous manifestations usually appearing after systemic symptoms of the disease.
Most of the patients with skin manifestations were middleaged females, while, patients with chilblain-like lesions were younger (mean age: 40.7 years) and patients with vascular lesions were older individuals (mean age: 72.3 years). These findings are along with other studies about the chilblain-like lesions (6,19,24,40). Maculopapular lesions were the most common dermatologic presentation of COVID-19 patients that commonly appeared at extremities. It occurred most often in middle-aged patients and was associated with moderate COVID-19 severity.
The overall mortality rate between the COVID-19 patients with skin presentations was 4.5%, with the point that there was the lowest mortality rate among the patients with urticaria-like lesions (2.2%) and contradictory, there was the highest mortality rate among the patients with vascular lesions (18.2%). Previous studies showed a pooled mortality rate of 3.2-6% in patients with COVID-19 (61,62). Thus, the mortality rate of COVID-19 patients with skin manifestations is proportionate to the overall mortality rate of the disease.
Regardless of the type of skin lesions, 80% of COVID-19 patients with cutaneous manifestations experienced a mild and moderate, and 20% a severe COVID-19 disease. A previous study from the Chinese Center for Disease Control and Prevention reported that 81% of COVID-19 patients had a mild, 14% a severe, and 5% a critical disease (63). We don't have any specific data on patients without skin manifestations but comparing the COVID-19 severity in patients with skin manifestations and COVID-19 patients, regardless of their symptoms, demonstrates no obvious difference. Future cohort studies are required to compare the disease severity and outcome of COVID-19 patients with and without skin manifestations.
There is a wide range of cutaneous manifestations related to COVID-19 that in terms of age, associated symptoms, comorbidity, medication, severity, and mortality, chilblainlike lesions, and vascular lesions are the ends of this spectrum. Chilblain-like, urticaria-like, vesicular, maculopapular, miscellaneous, and vascular lesions are associated with an increase in COVID-19 severity and worsening the prognosis, respectively. Vascular lesions were more prevalent in males (61%) compared to females (39%). Considering the more severe disease and higher mortality rate in patients with vascular lesions, we can conclude that COVID-19 is more severe in males compared to females. This finding is compatible with our recent article, in which we assessed the sex-specific risk of mortality in COVID-19 patients (62).
Although skin presentations of COVID-19 are well described, the pathogenesis of skin lesions remains unknown. The direct viral invasion of the skin cells may be one possibility. Angiotensin-converting enzyme 2 (ACE2) is known as a ligand for the Spike protein of SARS-CoV-2 for entering human cells (66). There is a high expression of ACE2 on keratinocytes and sweat gland cells, respectively (67,68). Thus, SARS-CoV-2 can directly infect keratinocytes resulting in necrosis. This hypothesis is consistent with our histologic findings which demonstrated the epidermal and adnexal necrosis in all skin lesions except vesicular rashes. According to Amatore et al., neither viralinduced cytopathic alterations nor intranuclear inclusions were seen in skin biopsies (35). However, SARS-CoV-2 spike and envelope proteins were detected in the endothelial cells of damaged skin in two cases with purpuric rashes (22). RT-PCR for SARS-CoV-2 was performed on skin samples of some patients and was negative in all of them. Since the nasopharyngeal swabs of these patients were positive simultaneously, we assume that it can be a false negative result due to a small viral load or technical problems. Further research is urgently needed.
Skin lesions during SARS-CoV2 infection might be immunerelated phenomena. It has been shown that the presence of virus RNA in blood is related to greater severity of infection (69). Viremia is also associated with the levels of cytokines and growth factors in a dose-dependent manner with markedly higher levels in patients suffering from more severe COVID-19 (69). Recognition of the viral RNA by Toll-free receptors like TLR7 stimulates the intracellular signaling pathways which in turn enhance the cytokine secretion (69).
In a group of patients, with the end of the first week of the infection, a sharp increase in inflammatory cytokines such as interleukin (IL)1, IL2, IL7, IL10, granulocyte colony-stimulating factor (G-CSF), tumor necrosis factor (TNF) α and interferon (IFN)-g occurs. Overactivation of immune responses followed by pro-inflammatory cytokines increase may result in a "cytokine storm" which is an immune pathological condition (69)(70)(71). Increased cytokines allow them to access the skin, where they stimulate various cells, including lymphocytes, dendritic cells, macrophages, neutrophils, monocytes, and Langerhans cells to cause various skin manifestations (22,69). Maybe a hyperviremia state is responsible for vascular lesions in severe COVID-19 patients. We suggest further investigations on the viral load levels among patients with vascular lesions compared with other skin manifestations.
The antigen-antibody complex can lead to complement activation and subsequent mast cell degranulation. This mechanism is suggested particularly for the urticaria-like lesions (43).
A low or delayed interferon response may result in uncontrolled viral replication followed by a subsequent cytokine storm which can lead to severe disease (72). Activation of the host immune system in response to viral antigen deposition may result in vascular damage in COVID-19 infection (73). It seems that high levels of type 1 interferon response, a critical factor in immunity against viral agents, is associated with chilblain-like lesions and mild disease (15,72,74). Activation and aggregation of cytotoxic CD8+ T cells and B cells also lead to lymphocytic thrombophilic arteritis and destruction of keratinocytes (21,22). Nests of Langerhans cells are seen in most of the COVID-19 skin lesion biopsies and have been also reported in another viral-induced skin dermatitis-like pytriasis rosea (75).
Coinfection with other viruses is another potential possibility for COVID-19 related cutaneous manifestations. Some skin lesions in COVID-19 patients are very similar to rashes induced by other viruses like parvovirus18, herpes simplex virus type 1 and 2 (HSV-1, HSV-2), varicella-zoster virus (VZV), and poxviruses, both clinically and histologically. It is probable that because of the attenuation of the immune system, COVID-19 patients are susceptible to coinfection with or relapse of the other viral exanthems. This hypothesis is strongly suggested for vesicular and some miscellaneous lesions (e.g., erythema multiform) due to their unique histologic findings compared to other skin lesions of COVID-19 (24,32,37). A study reported four COVID-19 patients presenting diffuse vesicular lesions which microbiological and serological investigations demonstrated varicella infection (24). Thus, in COVID-19 patients with vesicular lesions, physicians need to investigate other possible etiological factors other than SARS-CoV-2.
Coagulopathy and vasculitis are other possible reasons for skin lesions during COVID-19. Evidence shows that COVID-19 patients are predisposed to coagulopathy and subsequent thrombotic events (76). It seems to be a result of inflammatory cytokine release, hypoxia, and other illness or therapeutic risk factors (76). Microvascular thrombosis of dermal vessels leads to ischemia or vasculitis mainly seen in chilblain-like or vascular lesions. Magro et al. focused on the role of the complement factors activation, especially alternative and lectin pathways, and subsequent thrombotic microvascular injuries (22). Evidence for this hypothesis is the elevated levels of CH50, C3, and C4 in blood samples as well as significant vascular depositions of C5b-9, C3d, and C4d in the dermis of skin specimens (22). According to our histologic findings mentioned in RESULT, vascular thrombosis was reported in almost all skin biopsies (except vesicular lesions). This finding across with the increased level of D-dimer, fibrinogen, and prolonged PT and INR in most patients is in favor of this hypothesis. Another presentation of coagulopathy in COVID-19 patients is hemorrhagic events and subsequent dermatologic manifestations (petechiae, purpura, and livedo). These manifestations are not specific to SARS-CoV-2. Schneider et al. reported a petechial rash associated with coronavirus NL63 (77,78).
Extremely dilated blood vessels were introduced as a diagnostic histological finding for SARS-CoV-2 by Zengarini et al. (28). There are other reports of vasodilation and telangiectatic vessels in the dermis. With this finding, Magro et al. explained a possible pathway in which dysfunction of ACE2 (due to SARS-CoV-2 binding) and subsequent elevated level of angiotensin2 can result in high activation of endothelial nitric oxide synthase (eNOS) and ensuing vasodilation (22).
Drug-induced eruptions may occur during COVID-19. COVID-19 patients usually use a set of medications that potentially can cause cutaneous rashes. The current study found that paracetamol, azithromycin, hydroxychloroquine, lopinavir/ritonavir, and remdesivir were the most common medications used for COVID-19 patients. Paracetamol has been reported to cause asymmetrical drug-related intertriginous and flexural exanthema (STRIFE) (16). However, in Mahé et al. study, despite keeping the drug, skin lesions disappeared; that is very uncommon in drug reactions (16). Najarian et al. mentioned that maculopapular lesions of their patient could be according to azithromycin use or hypersensitivity reaction to azithromycin due to concurrent viral infection (20).
Hydroxychloroquine that has been used in 45% of all the cases (mentioned in Result) is one of the most likely medications to cause different skin rashes. Acute generalized exanthematous pustulosis (AGEP), erythroderma, urticaria, and erythema multiform are some of the skin lesions that have been reported in connection with hydroxychloroquine (79)(80)(81).
However, Robustelli et al. mentioned that the skin lesion developed 3 weeks after discontinuation of the drug (42). As a conclusion, most of our reviewed articles considered the potential possibility of drug-induced exanthems but in almost all cases, dermatologic manifestations preceded the drug intake or the rashes disappeared despite the continuation of drugs (5,7,16,20,23,37,42,43). So it is very unlikely that current COVID-19 medications are responsible for the reported skin lesions.
In our study, the prevalence of comorbidities in COVID-19 patients with skin manifestations is about 17.9% mainly reported in patients with maculopapular lesions. History of serious comorbidities like cardiovascular disease, hypertension, and obstructive lung disease was mostly reported in patients with vascular lesions; suggesting that patients with these skin manifestations are more complicated cases and need more attention. Interestingly, immune disorders were more common in patients with chilblain-like lesions. This finding is not reported yet and we suggest it to be focused on due to the possible relationship with the etiology and pathophysiology of these lesions.
Fever, cough, and dyspnea were more frequent in patients with vascular lesions and less frequent in patients with chilblainlike lesions. Also, 17% of patients with chilblain-like lesions were asymptomatic regarding systemic symptoms. Astonishingly, headache, dysosmia/dysgeusia, nasal congestion/coryza, and irritability/confusion were more common in patients with vesicular lesions. This finding can demonstrate the probable link between vesicular lesions and neurological manifestations. Future investigations are required to clarify the issue.

LIMITATIONS
There were limited articles that mentioned complete data about all the items including the disease severity and outcome of the COVID-19 patients with dermatologic presentations. Another limitation was the absence of data about the COVID-19 patients without skin manifestations. Future cohort studies are required to compare the severity and prognosis of the disease in patients with and without skin manifestations, considering other related characteristics. Such studies help to better understand the prognostic value of the cutaneous manifestations in COVID-19 patients.

CONCLUSIONS
Cutaneous lesions occur most often in middle age individuals at the same time or after the systemic symptoms of COVID-19. Urticaria-like lesions commonly (47%) occurred at the same time with other symptoms. It may suggest that urticaria-like lesions may be a diagnostic sign for COVID-19. A maculopapular rash is the main reported skin involvement in COVID-19 patients and is associated with intermediate severity of the disease. The mere occurrence of skin manifestations in COVID-19 patients is not an indicator for the disease severity, and it highly depends on the type of skin lesions. Chilblain-like and vascular lesions are the ends of a spectrum in which from chilblain-like to vascular lesions, the severity of the disease increases, and the patient's prognosis worsens. We highly suggest emergency and general practitioners to evaluate the suspected COVID-19 patients for any cutaneous manifestations. Those with vascular lesions should also be considered as high-priority patients for further medical care.

AUTHOR CONTRIBUTIONS
PJ, MN, and MM designed the study. PJ and BH performed the review literatures, collected the data, and wrote the first draft of the manuscript. PJ, HV, and MD helped in manuscript preparation. MM critically reviewed the manuscript.