Since December 2019 an outbreak of the novel coronavirus (SARS-CoV-2) disease has quickly spread worldwide. The infectious disease and pneumonia caused by this virus is called 2019 coronavirus diseases (COVID-19). Globally, as of 25 November 2022, there have been 636,440,663 confirmed cases of COVID-19 (7.96% of the world population), including 6,606,624 deaths (1.04% of confirmed cases), reported to the World Health Organization (1). Hospitalization is needed for 3–20% of COVID-19 patients; among these, 10–30% require intensive care (2, 3). Age, obesity, male sex, and comorbidities such as hypertension, heart disease, diabetes, kidney failure, and chronic pulmonary disease are considered as risk factors (4–10). Autoimmune bullous diseases (AIBDs) are a heterogeneous group of disorders including bullous pemphigoid (BP), mucous membrane pemphigoid (MMP), pemphigus, pemphigoid gestationis (PG), linear IgA bullous dermatosis (LABD). These disorders are associated with subepidermal or intraepidermal blistering and their treatment usually involves specific immunosuppressive therapies. Due to skin barrier alterations and chronic immunosuppressive treatments, patients are potentially at greater risk of infections, and they are considered fragile.

This review aims to discuss the pandemic’s burden on AIBD patients in terms of (i) susceptibility to COVID-19 infection; (ii) the relationship between infection and disease course and treatment; (iii) possible effects of vaccination in terms of induction, exacerbation or relapse of autoimmune disease.

To analyze the prevalence of COVID-19 in AIBD patients, and the effect exerted by immunosuppressive treatments in AIBD patients we qualitatively selected studies based on: (i) sample size; (ii) quality of control populations; (iii) standardization by age, gender, and country. To illustrate the spectrum of reported cases of new onset, relapse and/or worsening of AIBDs after COVID-19 vaccination all published cases found in PubMed searches up to November 2022 have been included in the present review.

A recent systematic review analyzed the proportion of AIBD patients with COVID-19 symptoms and positive molecular test (11). This review included 732 subjects affected by AIBDs (409 not specified, 211 pemphigus, 112 pemphigoid). Almost all information was collected via phone/telemedicine visits. COVID-19 symptoms were reported in 70 (9.5%) patients, and the diagnosis was confirmed in 16 (2.2%) patients (11). At present, 11 relevant studies reported prevalence data on SARS-CoV-2 infection in AIBD patients, with a pooled total of 10,060 patients (Table 1). COVID-19 symptoms were observed in 256 patients (2.5%), of whom 162 (1.6%) had a confirmed diagnosis (12–22) (Table 1). In order to compare prevalence data between AIBD patients and the general population, two recent studies performed a standardization by age, gender, country, or region of interest and period involved (12, 13). Kridin et al. carried out a retrospective cohort study on 3,081 patients with BP and pemphigus from the data set of Clalit Health Services in Israel. The adjusted hazard ratio (HR) between confirmed COVID-19 patients and controls was 1.1 [95% confidence interval (CI):0.71–1.71] for BP, 0.79 (95% CI: 0.43–1.46) for pemphigus (13). In parallel, 49 dermatology departments in France enrolled 5,180 AIBD patients and found that 29 (0.6%) had a confirmed COVID-19 diagnosis while 59 (1.1%) had possible, probable, or confirmed COVID-19 diagnosis. The standardized incidence ratio of hospitalized confirmed COVID-19 infection was 0.42 (95% CI: 0.20–0.80) for BP, 1.02 (95% CI: 0.37–2.26) for pemphigus, and 1.18 (95% CI: 0.55–2.23) for MMP (12). Altogether, these studies confirmed that the risk of COVID-19 infection in AIBDs patients was not higher than in the general population.

Although AIBD patients receiving rituximab (RTX) are more prone to viral infections than the normal population (23), only few studies investigated the possible influence of RTX-induced B cell depletion on the SARS-CoV-2 infection rate. A cohort study on 704 AIBD patients reported that among 21 COVID-19 confirmed cases, 14 (66.7%) received RTX within the last 12 months. Interestingly, the relative risk (RR) of COVID-19 infection and hospitalization decreased by 38% (95% CI: 18–57%) and 45% (95% CI: 15–72%), respectively, with each passing month from the last RTX infusion (14). In line with this observation, Joly and French Study Group on Auto Immune Bullous Skin Diseases, and the French Network of Rare Diseases in Dermatology [FIMARAD] reported an incidence ratio of COVID-19 in 23 RTX-treated patients ranging from 3.62 (95% CI: 1.29–8.85) in hospitalized patients with a confirmed diagnosis to 5.37 (95% CI: 3.15–8.96) in patients with a confirmed, probable, or possible diagnosis of COVID-19 (12). On the other hand, Breglio et al. found no difference in the rate of positive SARS-CoV-2 tests between 11 tests collected from RTX-treated patients and 66 tests collected from patients that were either never treated with RTX or received the last infusion at least 1 year before the onset of the pandemic (9.1 vs. 12.1%). However, the number of positive tests does not necessarily match the number of positive patients, since one patient could have been tested multiple times. In addition, the incidence could have been overestimated because patients performing SARS-CoV-2 test are usually symptomatic (24). On the contrary, Özgen et al. reported that among 51 of 247 pemphigus patients with COVID-19, 33% were treated with RTX while among non-COVID-19 patients only 21% were RTX-treated. These findings once more suggested that RTX increases the risk of COVID-19 (21). Of note, among AIBD patients, 45% of SARS-CoV-2 infected and 78% of non-infected were vaccinated; interestingly, only 29% of infected and RTX-treated patients were vaccinated (21), suggesting a role of the vaccine in preventing infection.

Thus, even though these results should be interpreted with some caution due to their relatively small sample size, the risk of SARS-CoV-2 infection seemed higher in RTX-treated patients than in those who did not receive a B cell depletion therapy.

Two recent studies investigated the risk of severe COVID-19 outcomes in patients with immune-mediated inflammatory diseases (25, 26), reporting a similar risk of COVID-19-related death and hospitalization in patients and the general population. In addition, some evidence suggests that immunocompromised patients do not have a higher incidence or complications from COVID-19 than the general population (27, 28). In line with these studies, a systematic review suggested that immunomodulatory treatments of AIBDs do not increase the risks of poor outcomes in COVID-19 patients (11). Later, other studies investigated the course of COVID-19 in AIBD patients. Özgen in Turkey enrolled 51 patients having COVID-19 of whom 40 (78%) had a non-serious disease, 11 (22%) required hospitalization, and one (2%) died (21). In Iran Mahmoudi et al. followed 21 AIBD patients with confirmed COVID-19 diagnosis: 15 (71%) had been hospitalized, 7 (33%) needed intensive care facilities and 3 (14%) died. Of note, two studies found that COVID-19 infection and hospitalization rates were higher with prednisolone dosages > 10 mg/die (14, 29). In parallel, Kridin et al. in Israel followed 36 confirmed COVID-19 cases in a cohort of 3,081 AIBD patients, of whom 17 (47.2%) were hospitalized, and 7 (19.4%) died (13). Similar results were observed in a larger French cohort of 5,180 AIBD patients where Joly and French Study Group on Auto Immune Bullous Skin Diseases, and the French Network of Rare Diseases in Dermatology [FIMARAD] identified 59 possible, probable, or confirmed COVID-19 cases of whom 30 (50.8%) were hospitalized, 7 (11.9%) were admitted to an intensive care unit, and 15 (25.4%) died (12). The latter two studies on the largest cohorts analyzed indicated that the risk of hospitalization was comparable to the general population (13) and that exposure to systemic corticosteroids and immunosuppressive agents was not significantly associated with increased severity of COVID-19 (12, 13). However, both studies reported a higher mortality rate for BP patients. In particular, Kridin et al. reported that COVID-19-associated mortality was higher in BP patients (adjusted HR 2.81) compared to the control population, while in pemphigus patients it was similar (adjusted HR 1.15) to their controls (13). Similar results are stated in Joly and French Study Group on Auto Immune Bullous Skin Diseases, and the French Network of Rare Diseases in Dermatology [FIMARAD]’s study, where mortality risk in confirmed COVID-19 patients was 1.63-fold higher in hospitalized patients with AIBDs than in the aged-matched population (12).

Several case reports, case series, and a recent Italian multicenter study proposed a possible, so far unproven, causal relationship between COVID-19 vaccination and AIBD onset (Supplementary Table 1) (41–91). To our knowledge 51 studies reporting new-onset AIBDs, with a total of 121 patients that developed BP (92 cases), anti-p200 pemphigoid (1 case), pemphigus (1 case), PV (15 cases), pemphigus foliaceus (PF) (7 cases), pemphigus vegetans (1 case), and LABD (4 cases) following SARS-CoV-2 vaccination (Supplementary Table 1) were published. Information about age and sex was not reported for 17 patients. The remaining cases included 58 men and 46 women (M/F = 1.3), ranging from 23 to 97 years (median age: 76 years). Information about administered vaccines was not reported for five patients, and latency to lesions was not reported for 17 patients. According to its greater employment, AIBD onset was more frequently associated with the Pfizer vaccine (61.2%). The onset of AIBDs followed the first dose in 44 patients, the second dose in 47 patients, and the third dose in 13 patients; median latency time was 9, 7, and 14 days from the first, second, and third dose, respectively. Information about the diagnosis was not reported for eight patients. The remaining 113 subjects had a diagnosis confirmed by consistent histopathological and/or serological findings. All patients were treated with corticosteroids and/or immunosuppressive drugs, and the vast majority showed a good clinical response (Supplementary Table 1). The largest cohort belongs to an Italian multicenter study that enrolled 21 patients with COVID-19 vaccine-associated BP (65). The authors concluded that BP patients did not differ essentially from idiopathic BP; however, a male predominance and a reduced autoantibody response to BP230 were considered unique features of vaccinated BP cases. Of note, a recent study by Birabaharan et al. on a very large cohort of individuals found no difference in the risk of BP onset among persons receiving the mRNA COVID-19 vaccine compared to an unvaccinated matched control cohort (92), suggesting that in a period with global mass vaccinations, the association between AIBDs occurrence and vaccine could be a random coincidence. A limitation of the study is the only use of International Classification of Diseases (ICD) codes from database sources for BP diagnosis.

A possible speculation that may bring to an agreement among the real-life clinical observations on patients with a new-onset of an AIBD and the convincing data from Birabaharan et al. considers the vaccination as a precipitating factor. Specifically, the vaccine may induce autoimmunity in genetically predisposed individuals by stimulating a pre-existent, and sub-clinical autoreactivity against hemidesmosomal components. This phenomenon could slightly anticipate BP development without significantly modifying the incidence of the disease. Additional investigations on this controversial topic are needed.

Several vaccinations such as tetanus and influenza have been reported to exacerbate AIBDs (93, 94). The mechanism mentioned above for AIBD onset after vaccination could also be considered for vaccine-associated relapse or worsening of a pre-existing disease. In fact, even in patients who achieve disease control or remission, clinical or sub-clinical autoreactivity against hemidesmosomal components is probably present. Kasperkiewicz and Woodley analyzed 30 published papers on AIBD during the pandemic and reported that 10% of patients had a flare or worsening following vaccination (95). To date, relapse/worsening after COVID-19 vaccine has been described in 185 patients (21, 42, 55, 61, 69, 75, 79, 82, 83, 96–100) (Table 2). Considering cases with known gender and age, the M/F ratio was 0.8 (8 M, 10 F). Ages ranged from 28 to 83 years, and the median age was 62 years. Similar to new-onsets, relapse and worsening were more frequently associated with the Pfizer vaccine (44.4%), as expected. The relapse/worsening of AIBDs followed the first dose in 27 patients, the second dose in 10 patients and the third dose in three patients; mean latency time was 7.5, 11.2, and 15.1 days from the first, second, and third dose, respectively (Table 2). Most patients were treated with corticosteroids and/or immunosuppressive drugs and showed a good clinical response (Table 2). Of note, it should be considered that exacerbation of the disease after the first dose does not preclude the administration of a second dose. In fact, very few patients presented disease exacerbation after both doses (83, 96) (Table 2). In general, disease relapse/worsening events can be reduced if the AIBD is stable at the time of vaccination (83). Of note, the risk of exacerbation should not dissuade from getting the vaccine.

Kasperkiewicz and Woodley reported that vaccination did not negatively affect the clinical course of AIBDs in most patients. Only about 6% of post-SARS-CoV-2-vaccinal cases presented clinically with de novo AIBDs, and 10% had a flare or worsening of pre-existing AIBDs usually well controlled with standard immune-suppressive treatment (95). However, the fact that vaccination could be less effective during immunosuppressive therapy should be considered. Despite the ongoing treatment with oral corticosteroid, all five AIBD patients analyzed by Damiani et al. developed IgG antibodies against SARS-CoV-2 (> 150 UI) 1 month after receiving the second dose (96). In this context, to achieve maximum efficacy and protection by vaccination, patients should be preferably vaccinated during disease remission or at least under low immunosuppression condition (101, 102). Nonetheless, lowering the corticosteroid dosage before or during the vaccination is not advisable due to the risk of exacerbations (15). Different considerations should be made in case of RTX therapy, especially in unvaccinated patients. B cell-depleting treatments have the most prominent effect on anti-SARS-CoV-2 IgG neutralization titers (103). Specifically, RTX-treated patients have failed to mount any humoral anti-SARS-CoV-2 response, even with a 4–12-month interval between RTX infusion and vaccination (104). Thus, a panel of experts advises that in these patients, vaccination should be performed either before therapy initiation or 4–6 months following RTX infusion (105).

In COVID-19 and AIBD research area further well-designed investigations are needed. In fact, there are only two studies on large cohorts with standardized findings on prevalence and disease course and very few studies on RTX treatment and COVID-19 disease. In particular, most of the studies do not consider the time between the last RTX treatment and COVID-19, that is the most important parameter to evaluate B-cell population at the time of SARS-CoV-2 infection. Moreover, few studies compare the impact of COVID-19 infection in vaccinated and not-vaccinated patients so far. However, despite of these limitations some important advances can be underlined.

The risks of COVID-19 infection and its severe course were comparable in AIBD patients and in the general population, except for RTX-treated patients. Of note, among COVID-19-infected BP patients the mortality rate was higher than in general population, while for pemphigus data were inconclusive. Thus, AIBD patients during the COVID-19 pandemic, especially if unvaccinated, should be subjected to risk-benefit assessments when considering RTX as a therapeutic option. The causal relationship between vaccination and AIBD onset, relapse or exacerbation has not been demonstrated so far. It could be speculated that vaccination may induce an autoimmune response in genetically predisposed persons by stimulation of pre-existent and sub-clinical autoreactivity. Additional investigations on this controversial topic should be performed. Of course, the risk of these rare adverse events should not discourage vaccination itself, although it should be performed during disease remission or at least while under low immunosuppression.

Finally, the impact of the pandemic on patients and especially on fragile ones has been heavy, but the precious acquired knowledge could also be used in possible, albeit undesirable, future pandemic scenarios.

AP and GD designed the manuscript. AP, JS, FrM, FeM, and GD wrote the manuscript. GD supervised the work. All authors contributed to the preparation of the manuscript and approved the submitted version.