Survival and prognostic factors in bullous pemphigoid: A retrospective cohort study

Plain Language Summary

Bullous pemphigoid is a skin disease where spontaneous blisters form due to abnormal antibodies produced by the patient’s own immune system (autoimmune disease). Since it is the most common autoimmune blistering skin disease, which is associated with high mortality and an increasing incidence, it is important to identify predictors for the survival of these patients. Our study has shown that advanced age, neurological diseases, disease of the heart valves, malignancies, use of statins (a group of blood cholesterol-lowering medications), skin infections at the time of diagnosis and widespread skin involvement are associated with lower survival rates. The use of corticosteroid creams/ ointments was linked to increased survival. These identified factors may help dermatologists in the reduction of disease mortality by identifying high-risk patients.

Introduction

Bullous pemphigoid is the most common subepidermal autoimmune blistering skin disease.¹ Over the past two decades, a significant increase in bullous pemphigoid incidence has been observed. This is mainly attributed to an ageing population, certain culprit drugs, as well as enhanced diagnostic techniques.²

Old age, low Karnofsky scores, neurologic and cardiac diseases, prolonged hospitalisation, elevated levels of circulating anti-BP180 antibodies, hypoalbuminemia, polypharmacy and high-dose corticosteroids had all been linked to poor prognosis among bullous pemphigoid patients.^3-10 Yet, many of these risk factors have remained understudied or have not been validated in larger cohorts. In addition, studies regarding the survival and prognostic factors of bullous pemphigoid in Romania have not been published so far.

We sought to evaluate the overall survival and to identify potential risk factors for lethal outcome in bullous pemphigoid patients in a representative cohort from a tertiary referral unit for autoimmune bullous diseases in Romania.

Methods

We performed an observational document-based retrospective study of patients diagnosed with bullous pemphigoid from July 2001 to November 2019 at the Department of Dermatology, Iuliu Hatieganu University of Medicine and Pharmacy, Cluj-Napoca, Romania. This is the only centre in the North-West region of the country which provides an immunopathological diagnosis of autoimmune bullous diseases.

For inclusion, patients with a diagnosis of bullous pemphigoid had to fulfil all of the following three criteria:

1. Consistent clinical findings (generalized tense inflammatory blisters with or without persistent urticarial or eczematous lesions),

2. Direct immunofluorescence microscopy demonstrating linear deposition of immunoglobulin G and/or C3 along the basement membrane, and

3. Enzyme-linked immunosorbent assay with positive immunoglobulin G autoantibodies against BP180 and/or BP230, plus at least one of the following two additional criteria:

   • Histopathological evidence of a subepidermal blister/cleft with eosinophils, or

   • indirect immunofluorescence microscopy demonstrating circulating autoantibodies bound to the epidermal side of salt-split skin.

Patients with isolated or predominant oral disease, those with negative direct immunofluorescence microscopy and negative enzyme-linked immunosorbent assay autoantibodies were excluded from the study. For the detection of specific serum autoantibodies, commercially available enzyme-linked immunosorbent assay kits were used (MBL laboratories, Nagoya, Japan). The cut-off value for positive levels of anti-BP180 and anti-BP230 autoantibodies was ≥9 U/mL. Patients’ sera harvested at diagnosis were frozen and stored at -80 degrees Celsius. Between 2001 and 2008 enzyme-linked immunosorbent assay-based serum autoantibody detection was performed at the Department of Dermatology, University of Lübeck, Lübeck, Germany, and the Department of Dermatology, University of Würzburg, Würzburg, Germany. Since 2008, the enzyme-linked immunosorbent assay probes were analysed in the Department of Dermatology, Iuliu Hatieganu University of Medicine and Pharmacy, Cluj-Napoca, Romania.

The study was approved by the Ethics Committee of the Iuliu Hatieganu University of Medicine and Pharmacy, Cluj-Napoca (decision number 47/2020) and was conducted in accordance with the principles of the Declaration of Helsinki. All patients signed informed consent at initial consultation agreeing to be photographed for medical documentation, their clinical and laboratory data to be recorded, as well as blood and tissue samples to be used for scientific research.

Results

A total of 148 patients (54.1% females) met the enrolment criteria [Table 1]. The median age for bullous pemphigoid diagnosis was 74.5 years (range, 15-92 years). The median time from onset to diagnosis was 61.5 days (interquartile range, 31-151.3 days).

The median autoimmune bullous skin disorder intensity score at diagnosis was 52.5 (interquartile range,26-85.9). Of all patients, 145 (98%) had positive anti-BP180 antibodies and 70 (47.3%) had positive anti-BP230 antibodies, with median titres of 112.6 U/mL (interquartile range, 49.8-164.7 U/mL) and 48 U/mL (interquartile range, 2-99 U/mL) respectively.

Survival analysis

During a median follow-up period of 48 months (interquartile range, 11-150 months), 90 (60.8%) patients died, and 38 (42.2%) of them had active disease at the time of death. The cause of death was retrievable in all patients [Figure 1]. Cardiovascular diseases were the leading cause of death, seen in 40 (44.4%) patients. Three (3.3%) patients died of pemphigoid-related causes (sepsis due to skin infection with methicillin-resistant Staphylococcus aureus ) and they showed a lower overall survival as compared to the others (hazard ratio = 7.32 [95% confidence interval, 1.15-46.37], P = 0.035).

Close

Figure 1:

Causes of death by age groups

Export to PPT

The median overall survival was 17.5 months (interquartile range: 0-176). The Kaplan-Meier overall survival rates at 1, 3, 5 and 10 years were 74.2% (95% confidence interval, 67.5-81.6%), 53.4% (45.7-62.2%), 43.6% (35.9-53%) and 31.3% (23.5-41.7%). Only Kaplan-Meier curves for relevant prognostic factors (age ≥75 years, stroke, dementia, Alzheimer’s disease) have been illustrated [Figure 2]. We found no association between autoimmune bullous skin disorder intensity score and survival.

Close

Figure 2a:

Kaplan-Meier survival analysis in bullous pemphigoid patients (n = 148) with age over and under 75 years

Export to PPT

Close

Figure 2b:

Kaplan-Meier survival analysis in bullous pemphigoid patients (n = 148) with and without a history of stroke

Export to PPT

Close

Figure 2c:

Kaplan-Meier survival analysis in bullous pemphigoid patients (n = 148) with and without dementia

Export to PPT

Close

Figure 2d:

Kaplan-Meier survival analysis in bullous pemphigoid patients (n = 148) with and without Alzheimer’s disease

Export to PPT

Discussion

We retrospectively evaluated 148 bullous pemphigoid patients for a median follow-up period of 48 months. This is one of the longest follow-up studies conducted on bullous pemphigoid patients in Europe. So far, only the study of Cortés et al. reported a 5-year survival rate.¹² In line with previously published data, the leading causes of death were cardiovascular and cerebrovascular diseases.^5,6,13 Of note, three (3.3%) patients died of pemphigoid-related causes, showing significantly lower overall survival rates as opposed to those who died of other causes. They developed extensive, refractory forms of bullous pemphigoid and died of sepsis with methicillin-resistant Staphylococcus aureus.

We also sought to identify potential prognostic factors for overall survival in bullous pemphigoid. The median age of 74.5 years in our cohort was comparable with that in previous studies.^8,14,15 Likewise, we identified older age (≥75 years) as a risk factor for overall survival in the multivariate analysis.^5,6,8,14,15

In our study, extensive cutaneous involvement (≥30% body surface area) was significantly associated with lower overall survival. This finding was reported only in two other studies.^8,16 Notwithstanding, a meta-analysis regarding prognostic factors for mortality in bullous pemphigoid patients did not show any significant association.¹⁷ A possible explanation for this discrepancy is the subjectivity and intra- and interobserver variability associated with body surface area-based severity assessment. On the other hand, the autoimmune bullous skin disorder intensity score is also not specific for any particular autoimmune blistering dermatosis.¹¹ Correspondingly, we found no association between autoimmune bullous skin disorder intensity score and survival.^3,6,15,17 Recently, the novel bullous pemphigoid disease area index score was proposed and validated, which is both specific and objective for disease severity assessment in bullous pemphigoid.¹⁸

Particularly noteworthy was the association between skin infections in denuded bullae and lower overall survival. It has been shown that infections are common in bullous pemphigoid, especially among patients with functional impairment and dementia, thus increasing mortality.^19,20 The most common infections in bullous pemphigoid patients are found in the skin and mucosa.²¹ In our cohort, all detected skin infections were staphylococcal and required systemic antibiotics. Of these, three (3.3%) patients died of sepsis with methicillin-resistant Staphylococcus aureus. As far as we know, this is one of the few studies to look at the prognostic role of bacterial skin infections at diagnosis in the survival of bullous pemphigoid patients.

Cardiac and neurological comorbidities were also linked to decreased overall survival. Cardiac diseases such as ischemic heart disease and heart failure were reported as risk factors for death in bullous pemphigoid subjects.^15,20,22,23 We identified cardiomyopathies, valvular heart disease and chronic heart failure as poor prognostic factors for bullous pemphigoid patients in the univariate analysis. However, in the multivariate analysis, the association persisted only for valvular heart disease. We were unable to find any previous reports regarding the role of valvular heart disease in the prognosis of bullous pemphigoid. Interestingly, valvular heart disease as a result of chronic inflammation and scarring of the endothelium was also prevalent in other autoimmune diseases.²⁴ It is therefore tempting to speculate that the many cellular players involved in the inflammatory processes of bullous pemphigoid, might also be involved in the valvular damage, thus increasing the mortality of these patients.²⁵ However, further studies are needed to evaluate this association and corroborate if patients with bullous pemphigoid have indeed a higher risk of developing valvular heart disease as compared to the general population.

Consistent with previously published data, we found a strong association between neurological disorders and decreased survival. Neurological diseases assessed collectively, and a history of stroke were identified as poor prognostic factors. Similar to other studies, dementia and Alzheimer’s disease were associated with decreased overall survival only in the univariate analysis.^20,26 The association between bullous pemphigoid and neurological disorders is not yet fully understood. A proposed hypothesis is that BP180 and/or BP230 autoantibodies cross-reactively target the namesake neuronal isoforms, which are expressed in the central nervous system.¹ Interestingly, a previous study showed that patients with bullous pemphigoid have a 3.6-fold increased risk for neurological disorders prior to the diagnosis of bullous pemphigoid, raising the issue of the exposure of neuronal isoforms as triggers for bullous pemphigoid development.²⁷

In our study, malignancy was found as a significant prognostic factor in the prediction of survival for bullous pemphigoid. The relationship between bullous pemphigoid and cancer still remains controversial. In a 10-year population-based study, no increased risk of cancer in bullous pemphigoid patients has been shown.²⁸ Conversely, there is evidence that suggests an increase in hematologic malignancies among bullous pemphigoid patients.²⁹ A possible explanation for an association with cancer is the cross-reactivity of antibodies directed against tumour-specific antigens of malignant cells with basement membrane bullous pemphigoid antigens.²⁸ However, no cancer screening is currently recommended for bullous pemphigoid patients at diagnosis.

Comparable to other published reports, we did not find diabetes, lung, kidney and thyroid disease as risk factors for lethal outcomes among bullous pemphigoid patients.^6,8,15 Furthermore, we found that elevated erythrocyte sedimentation rate was significantly associated with lower overall survival, but this did not persist in the multivariate analysis, analogous to other studies.^5,8,16,20

Remarkably, we found a positive correlation between the use of topical corticosteroids, either alone or combined with oral corticosteroids and overall survival among bullous pemphigoid patients. However, we did not find oral corticosteroids or other immunosuppressive agents, as predictive factors for survival in bullous pemphigoid. Conversely, a large prospective trial by Joly et al. showed a greater mortality rate and higher corticosteroid-associated side effects among patients treated with oral prednisone, compared with those treated with topical corticosteroids.³⁰ This study changed the traditional paradigm of therapy, with topical corticosteroids being currently recommended as the first line of treatment for bullous pemphigoid.³¹ Nonetheless, the protective effects of topical corticosteroids as monotherapy could also be due to their use in milder forms of bullous pemphigoid. In our study, almost every patient received either dual or oral corticosteroid treatment, with only four (2.7%) subjects receiving topical treatment alone.

Contrary to the study of Rozenblat et al., we found that the use of statins was associated with lower overall survival rates among bullous pemphigoid patients.²⁶ Interestingly, a more recent well-matched population-based cohort study with appropriate follow-up time did not show any significant association between the use of statins and the risk of developing bullous pemphigoid.³² It is noteworthy that in our study 131 (88.5%) patients had cardiovascular diseases, which were also the leading cause of death, followed by stroke. Therefore, many of the patients were taking statins at the time of diagnosis. A similar explanation might be feasible for acetylsalicylic acid, which was also associated with lower survival. Moreover, its daily use in healthy older adults was linked to higher all-cause mortality, mainly attributed to cancer-related deaths, but also to gastrointestinal side effects.³³ However, in our cohort this correlation did not persist after adjustment for confounders.

The strengths of this study are: (1) complete diagnostics were performed in all patients, and only those with positive direct immunofluorescence microscopy and positive enzyme-linked immunosorbent assay BP180 and/or BP230 immunoglobulin G autoantibodies were included; (2) the long observation period of up to 18 years in the same geographical area; (3) extensive data regarding demographic, bullous pemphigoid characteristics, comorbidity profiles and types of treatments were recorded; (4) follow-up data were included for every patient; missing data about the date and cause of death was obtained from national registers and (5) we adjusted the analyses for important known confounders.

This study lacks a control cohort to validate the obtained results, since the majority of bullous pemphigoid patients are elderly and already have many associated comorbidities. Other study limitations include the retrospective document-based design and that it was based in a single institution, though this was a referral centre for autoimmune bullous diseases. In addition, indirect immunofluorescence microscopy was not performed in all patients.

Conclusion

This is the first study assessing survival and prognostic factors of bullous pemphigoid in Romania. Advanced age, extent of skin involvement, cardiac and neurological diseases and malignancy were identified as predictors of decreased survival in bullous pemphigoid. We found for the first time that valvular heart disease and skin infections at diagnosis were independent risk factors for decreased survival in bullous pemphigoid. We also observed that the use of topical corticosteroids was associated with increased overall survival, whereas the use of statins was linked to lower survival rates. Our findings may aid dermatologists in the reduction of bullous pemphigoid mortality by identifying high-risk patients and managing them accordingly.