Are You Confident of the Diagnosis?

What you should be alert for in the history

Determine if the lesions are pruritic. Bullous pemphigoid (BP) is almost always pruritic, and itch can be the first presenting symptom or can herald a relapse.

Ask about aggravating factors or associated conditions, which can include stress, ultraviolet radiation or radiation therapy, menses or pregnancy (also known as herpes gestationis), recent vaccinations, and a history of neurologic disease including dementia, Parkinson’s disease, or cerebrovascular accident. Bullous pemphigoid has been anecdotally associated with several drugs including penicillamine, captopril, furosemide, nonsteroidal anti-inflammatory agents, and antibiotics.

Continue Reading

Determine if the patient has a history of inflammatory skin diseases associated with pemphigoid, including lichen planus or psoriasis.

Ask about a gritty sensation in the eyes, hoarseness, or difficulty swallowing to evaluate for potential mucosal disease. Mucosal involvement in bullous pemphigoid is less common, and when present, usually affects only the oropharyngeal mucosa. If patients have extensive mucosal disease including the conjunctiva, a diagnosis of mucous membrane (cicatricial) pemphigoid should be considered.

Characteristic findings on physical examination

There are two primary morphologies of clinical lesions in BP: tense bullae and urticarial papules and plaques (Figure 1). A classic clinical presentation for bullous pemphigoid is tense bullae on an erythematous base arising on the lower extremities in an elderly patient (Figure 1). However, a wide spectrum of disease exists. Early disease may begin with pruritus, and skin lesions may appear non-specific, with excoriated papules or eczematous patches. A pompholyx variant may present with confluent vesicles on the palms and soles. Localized disease may also occur, most commonly in the pretibial area, sites of irradiation, or around surgical sites.

Figure 1.

Tense bullae, erosions, and urticarial plaques on the legs of an elderly patient, typical of bullous pemphigoid. (Photo courtesy of Misha Rosenbach, MD)

Lesions heal without scarring and usually without milia, although with widespread disease milia can occur after healing.

  • Expected results of diagnostic studies

Skin biopsy for histology: A punch or shave biopsy for histology should be obtained from an urticarial plaque or at the lateral edge of a fresh blister. Biopsy of an old blister can confuse the diagnosis due to necrosis of the roof of the blister and/or re-epithelialization of the base. Histology of BP will show a subepidermal blister, most often with prominent eosinophilic/mixed inflammation in the dermis, epidermis, and blister fluid (Figure 2). Rarely, cell-poor subepidermal bullae are observed. Biopsy of urticarial lesions will demonstrate eosinophilic spongiosis.

Figure 2.

Histology of bullous pemphigoid showing subepidermal blister with eosinophils (arrows).

These routine light microscopy findings are histologically non-specific and may also be observed in allergic contact dermatitis, scabies infestation, and other inflammatory or immunobullous skin conditions. Autoantibodies to the basement membrane zone should be documented by direct immunofluorescence, indirect immunofluorescence (including salt split skin), and/or ELISA.

Skin biopsy for direct immunofluorescence (DIF): A punch or shave biopsy should be obtained from normal-appearing perilesional skin and submitted in Michel’s or other appropriate fixative for direct immunofluorescence analysis. BP will demonstrate linear basement membrane zone staining, classically for C3 greater than IgG. Dermis always has a strong background on DIF testing due to the high concentration of IgG in blood, which extravasates into the dermis during the biopsy, potentially masking the detection of immune reactants at the dermal-epidermal junction. Incubation of the skin sample in sterile saline at 4 degrees C for up to 24 hours prior to transfer to Michel’s fixative will help to wash out unbound dermal IgG and increase the sensitivity of DIF detection.

Even if all skin lesions have healed because of recent therapy, DIF should remain positive for at least several weeks. Immune reactants are typically destroyed within the blister cavity, leading to false negative results if samples for DIF are taken from blistered or urticarial skin.

Even with a subepidermal blister and positive DIF result, the diagnosis of BP is not 100% certain since epidermolysis bullosa acquisita and p200 pemphigoid can yield similar findings (although the presence of prominent eosinophils on histology and a classic clinical presentation would be relatively convincing for BP.) To confirm the diagnosis, indirect immunofluorescence on salt split skin or ELISA is beneficial.

Serum sample for indirect immunofluorescence (IIF): Serum is incubated with epithelial substrates (monkey esophagus, normal human skin, or normal human skin pre-incubated with 1 M NaCl). On non-salt split substrates, IIF for BP will classically demonstrate C3 greater than IgG basement membrane zone staining (Figure 3), and will remain positive for at least several weeks after skin lesions have healed. On salt split skin, IIF for BP will demonstrate linear staining of the basement membrane zone on the roof or epidermal side of the blister (sometimes both the epidermal and dermal side are stained).

Figure 3.

Indirect immunofluorescence staining of monkey esophagus with bullous pemphigoid serum demonstrates linear basement membrane zone staining of C3 and IgG

Serum sample for BP ELISA: Antigen-specific ELISAs (BP180 and BP230) are sensitive and specific for BP and are less subjective than immunofluorescence-based tests. However, BP ELISA is not always available from major national reference laboratories. The BP180 ELISA is 84% to 89% sensitive and 98% to 99% specific. The BP230 ELISA is 72% sensitive and 99.5% specific. The combination of the two ELISAs is 97.1% sensitive.

Complete blood count will often show peripheral eosinophilia, although the presence or absence of this laboratory finding does not define or preclude a diagnosis of BP.

  • Diagnosis confirmation

The clinical differential diagnosis for BP is broad and includes epidermolysis bullosa acquisita, porphyria, p200 pemphigoid, linear IgA bullous dermatosis, pemphigus vulgaris, Stevens-Johnson syndrome, toxic epidermal necrolysis, urticaria or urticarial drug reaction, disseminated herpes simplex or zoster, scabies, and arthropod reaction. Localized disease can also be confused with allergic contact dermatitis or dyshidrotic eczema.

Who is at Risk for Developing this Disease?

The incidence of BP ranges from 4.3 per 100,000 person years in the United Kingdom to 6.6 per million in western Europe. The sex ratio of cases in women to men averages 1.5:1, except in Kuwait where a 5:1 sex ratio has been reported. The incidence of BP increases with age, with an average yearly increase in incidence of 17% in the United Kingdom. Rarely, disease can occur in children.

Women may develop BP during or shortly after pregnancy (also known as pemphigoid gestationis or herpes gestationis). The disease is histologically and clinically identical to BP except for its age of onset. Population-based studies have also shown that bullous pemphigoid is associated with neurologic diseases, including dementia, Parkinson’s disease, and cerebrovascular accident. In most cases bullous pemphigoid follows the onset of neurologic disease but in some cases it precedes the neurologic diagnosis. Although the exact mechanism for this association is unclear, both BP180 and BP230 are expressed in brain and skin, leading to speculation that neurologic disease or inflammation may lead to an autoimmune reaction to one or both antigens.

Several medications have been anecdotally associated with BP, including furosemide, penicillamine, captopril, ibuprofen, and antibiotics. Disease may or may not remit with discontinuation of the medication.

The MHC class II allele DQB1*0301 is more often observed in BP patients.

What is the Cause of the Disease?

  • Etiology

The pathologic target of autoantibodies in BP is type XVII collagen, also known as bullous pemphigoid antigen 2 (BPAg2) or BP180, a transmembrane protein of the hemidesmosome. Most autoantibodies bind to the NC16A domain, an extracellular non-collagenous domain. Both IgG and IgE antibodies have been implicated in disease pathology, the latter thought to account for urticarial lesions in BP.

Almost all patients with BP demonstrate antibodies to bullous pemphigoid antigen 1 (BPAg1) or BP230. BP230 is a cytoplasmic plaque protein of the hemidesmosome. Autoantibodies to BP230 likely arise secondary to basal keratinocyte damage and do not play a direct role in blister formation.

  • Pathophysiology

As with most autoimmune diseases, the reasons for the loss of immune tolerance in BP are unknown. Because human BP autoantibodies do not cross-react with mouse BP180, it has been more complicated to develop mouse models for study. Our knowledge of disease pathophysiology has resulted from four models: injection of human BP IgG into rabbit cornea, passive transfer of serum IgG from rabbits immunized with mouse BP180 to mice, passive transfer of human BP autoantibodies to immunodeficient mice bearing human skin grafts, and passive transfer of human BP autoantibodies to ‘humanized’ mice (BP180-deficient mice that were subsequently made transgenic for human BP180).

The current working model of disease pathology is as follows: Anti-BP180 IgG binds to the basement membrane zone. Complement fixation and activation causes leukocyte chemotaxis and degranulation of mast cells, which promotes chemotaxis of eosinophils. Release of proteolytic enzymes by inflammatory cells (such as neutrophil elastase and eosinophil gelatinase) causes cleavage of basement membrane zone proteins including BP180, leading to blister formation. Additionally, passive transfer of anti-BP180 IgE recapitulates many aspects of the urticarial phase of disease. Monovalent (Fab) fragments of BP IgG do not cause blisters, and mice deficient in components of complement or neutrophil elastase do not develop blisters, indicating that the effector functions of BP autoantibodies are required for disease.

Systemic Implications and Complications

BP typically affects the skin without systemic involvement. However, one study in the Czech Republic found a significantly higher incidence of neurologic disease (primarily stroke and dementia) in BP patients compared with age- and sex-matched controls. These data are intriguing since an alternatively spliced form of BP230 is expressed in neural tissue, and mice with a genetic deletion of BP230 demonstrate defects in epithelial integrity as well as neurodegeneration.

Secondary infection can occur and may be difficult to identify with widespread urticarial and erosive lesions.

Treatment Options

Optimal Therapeutic Approach for this Disease

Class I steroids such as clobetasol can be applied twice daily to new lesions. Topical tacrolimus ointment may also be considered in areas where chronic steroid therapy is undesirable. A randomized controlled trial of topical steroids (20 grams twice daily) and oral prednisone (1.0 mg/kg/day) in severe cases of BP showed that topical steroids were significantly more effective, had lower 1-year mortality, and were associated with fewer severe complications as compared with oral prednisone.

For mucosal disease, dexamethasone elixir, swish and spit 5cc once to twice daily, is easy to use. Clobetasol ointment or gel can also be applied directly to mucosal erosions. Dental trays (fitted by oral medicine) facilitate occlusion of topical steroids to the gingiva at night.

BP (including widespread disease) may respond to tetracyclines plus niacinamide (tetracycline 2g daily or doxycycline 100 mg twice daily, plus niacinamide 500 mg 3 times daily). This regimen is worth a trial, particularly in elderly patients with multiple comorbid conditions for whom oral steroids and other immunosuppressants would be difficult.

With persistent or widespread disease, oral corticosteroids such as prednisone are indicated. For moderate disease, 0.5 mg/kg/day of prednisone or equivalent may be sufficient. Doses generally do not need to exceed 1 mg/kg/day of prednisone. If patients flare on 1 mg/kg/day of prednisone, the dose can be split to twice daily or three times daily dosing, which increases the therapeutic efficacy without increasing the total daily dose. Before starting high dose steroids, tuberculosis screening should be performed (via tuberculin skin testing or QuantiFeron-TB gold blood assay).

If patients will be on chronic corticosteroids (at least 5 mg daily prednisone equivalent for at least 3 months), osteoporosis counseling and prevention is indicated. Additionally, Pneumocystis prophylaxis should be considered for patients on chronic prednisone, particularly with daily prednisone doses of 15 mg or higher. Patients should remain on high-dose steroids until new lesions cease to form, and then the dose can be gradually tapered to the minimum required to control disease. If patients can be managed with 10 mg (or ideally 5 mg) daily prednisone or less, corticosteroid monotherapy is feasible.

Dapsone (100 to 200 mg daily) may be effective to lower the daily corticosteroid dose in patients with stable disease. Dapsone can be used in addition to mycophenolate mofetil or azathioprine. As an advantage, dapsone 100 mg daily provides Pneumocystis prophylaxis. Glucose-6-phosphate dehydrogenase (G6PD) activity should ideally be measured before starting therapy, particularly in men of African-American and Middle Eastern descent. Most patients will experience a 1 to 2 gram/dl drop in hemoglobin due to hemolysis, although some patients can experience a severe pancytopenia with or without systemic hypersensitivity reaction. Laboratory monitoring should be performed at least every other week for the first 8 weeks.

In patients requiring greater than 10 mg daily prednisone for control of disease activity, or in patients with contraindications to systemic corticosteroid therapy, other immunosuppressants are necessary to reduce or replace systemic corticosteroids. Mycophenolate mofetil (30 mg/kg/day divided twice daily) is generally well tolerated, although side effects of fatigue, gastrointestinal upset, and tremor are not uncommon, particularly at higher doses, and there is a small long-term risk of lymphoma and fatal infection or reactivation from JC virus with progressive multifocal leukencephalopathy.

Reduction of corticosteroid dose can be initiated as early as 1 month after starting mycophenolate mofetil, although maximal effect of mycophenolate mofetil is not achieved until 2 to 3 months. Azathioprine can be started at 50 mg daily and titrated upward by 50 mg every 1 to 2 weeks until side effect, therapeutic effect, or the target dose of 2.5 mg/kg/day occurs. Measurement of serum thiopurine methyltransferase (TPMT) level prior to start of azathioprine therapy can be performed, although some studies suggest that TPMT levels do not correlate with the incidence of adverse effects or efficacy of azathioprine therapy.

Nevertheless, if serum TPMT levels are very low or very high, azathioprine may not be a good choice for therapy, due to an increased likelihood for adverse effects or lack of effect, respectively. The active metabolites for azathioprine do not significantly accumulate until 6 to 8 weeks after initiation of therapy, leading to a delayed therapeutic effect.

In patients who have severe or persistent disease that is unable to be controlled with corticosteroids and/or other immunosuppressives, other therapies such as rituximab, intravenous immunoglobulin, plasmapheresis, and cyclophosphamide can be considered. B-cell depletion therapy with rituximab (anti-CD20 monoclonal antibody) is anecdotally effective for the treatment of BP and has been proposed by some experts to be considered for first-line therapy for another autoimmune blistering disease, pemphigus vulgaris.

Both lymphoma (375 mg/m
2 IV weekly x 4 weeks) and rheumatoid arthritis (1000 mg IV on days 1 and 15) dosing regimens can be used. In vivo studies from lymphoma patients indicate that peripheral blood B cells disappear from the circulation within days, although antibody production by plasma cells (which are not well targeted by rituximab) can persist for months; therefore, maximal results are typically not observed until 3 to 6 months after infusion, and re-infusion every 6 months may be required. Fatal infection has occurred with rituximab therapy, including bacterial sepsis, hepatitis B reactivation, and progressive multifocal leukencephalopathy from JC virus, although fatal infection is also a potential side effect of first-line therapies for BP, including prednisone and mycophenolate mofetil.

Intravenous immunoglobulin (IVIG, 2 mg/kg, divided over 3 to 5 days) is effective for BP therapy and can be provided by hospital or home infusion. IVIG is thought to induce catabolism of endogenous serum antibodies and offers the advantage of being immunoprotective. A disadvantage of IVIG is that the temporary serum viscosity associated with the infusion can cause stroke or other complications from clotting. The serum half-life of IVIG has been reported to range from 8 to 39 days (average 3 to 4 weeks). Treatment guidelines for autoimmune blistering disease with IVIG suggest an initial frequency of every 4 weeks until disease remits, increasing to 6, 8, 10, 12, 14, then 16 weeks, the latter being the proposed end point for an initial course of therapy.

Plasmapheresis allows for the rapid removal of antibodies from the circulation, but must always be used in conjunction with adjunctive immunosuppressants to prevent new antibody production.

Cyclophosphamide (50 to 200 mg daily) is among the fastest agents for treating disease. However, its risk of blood count and liver test abnormalities, infertility, and hemorrhagic cystitis with bladder carcinoma, together with the advent of other effective therapies such as rituximab, have led to the decreased use of cyclophosphamide in the management of severe BP.

Case reports and one small case series have shown therapeutic benefit in bullous pemphigoid patients with omalizumab, a monoclonal antibody that binds and suppresses IgE.

Treatment options are summarized in Table I.

Topical corticosteroids
Tetracyclines plus niacinamide
Oral corticosteroids
Mycophenolate mofetil
Intravenous immunoglobulin

Patient Management

The goal of treatment is to obtain a complete remission off therapy, although many patients may only achieve a partial remission off therapy, or a complete remission on minimal therapy. When starting patients on therapy, risks of medications should be discussed. There is no systemic medication that is 100% safe. However, the 1-year mortality of BP has been reported to range from 6% in the United States to 19% in the United Kingdom to 41% in France; therefore, the risk-benefit ratio often favors treatment.

Open erosions can become superinfected with Staphylococcus aureus or herpesviruses; culture of refractory or worsening lesions should be considered. Remind patients that their skin during active disease is fragile, so crusted blisters should not be scrubbed.

Patients should receive regularly scheduled dental cleaning from a hygienist experienced with oral blistering disease. Patients often avoid cleanings due to painful mucosal disease; however, over time plaque buildup leads to chronic gingival inflammation that aggravates disease. If necessary, patients can take a short (1 to 2 week) course of prednisone surrounding their cleanings to prevent flares of disease.

Whether or not serial BP180 ELISA monitoring is of clinical value to follow disease activity is unclear. Studies have shown that 65% of remission sera will still maintain a positive BP180 ELISA index value. However, other studies have suggested that BP180 ELISA titers parallel disease activity, and one study found that a BP180 ELISA value greater than 27-IU/mL at the time of clinical remission was most predictive of disease relapse. Patients should ideally be maintained in complete remission for at least 1 year before all immunosuppressive therapy is discontinued. Often patients want to taper off their medications quickly, but then flare and have to go back on high doses of medications for disease control, which may reduce the chance for disease remission.

Typically, corticosteroids are tapered off first, then adjunctive immunosuppressants are slowly tapered over the course of 1 year. However, the tapering regimen should be tailored for each patient depending on side effects and response to therapy. It is common for patients to have a small flare with each dose taper; as long as lesions heal within 1 week and no further lesions form, the taper can be continued. Recurrence of pruritus often precedes the development of frank urticarial or bullous lesions and may herald disease relapse.

The Centers for Disease Control recommends that all patients on immunosuppressive therapy receive influenza and other regularly scheduled vaccinations. While on immunosuppressive therapy, patients should be reminded that they should not receive live vaccines (e.g. nasal influenza or zoster).

Unusual Clinical Scenarios to Consider in Patient Management

Pregnant patients who develop BP (also known as pemphigoid gestationis) should be referred for high-risk obstetrical care. Pregnant patients are usually treated with prednisone monotherapy, as most other immunosuppressive agents are pregnancy category D, including mycophenolate mofetil, azathioprine, tetracyclines, and cyclophosphamide. Dapsone is pregnancy category C. There are sparse data regarding the safety of immunosuppressives in men whose pregnant wives can be exposed to drug in seminal fluid. Rare cases of birth defects with azathioprine use by fathers have been reported. An ongoing registry of male transplant patients receiving mycophenolate mofetil has shown no significant increase in birth defects.

In hospitalized patients with severe disease, blood should be cultured so that bacteremia can be identified and treated. A thin layer of triamcinolone 0.1% ointment can be spread on sterile linens and wrapped around patients twice daily. Unfortunately, there is no single consensus regimen among experts about how to treat severe pemphigoid, owing to the sparsity of randomized controlled trials in this rare disease. Historically, these patients were treated with corticosteroids, cyclophosphamide, and plasmapheresis due to their rapid therapeutic effect.

Several other regimens can be considered for the hospitalized patient, including:

  • intravenous corticosteroids (as high as 60 mg methylprednisolone 4 times daily), adjunctive immunosuppressant such as mycophenolate mofetil, and IVIG (which is immunoprotective)

  • intravenous corticosteroids (+/- adjunctive immunosuppressant) and rituximab for long-term control

  • intravenous corticosteroids (+/- adjunctive immunosuppressant), plasmapheresis to immediately remove serum antibodies, followed by rituximab for long-term control

The timing of certain combinations of medications should be considered. For example, rituximab infusion should not be administered immediately before plasmapheresis or IVIG, as the former would clear the rituximab and the latter may induce its catabolism. However, rituximab is thought to rapidly bind and deplete circulating B cells (within days), and peripheral B cell counts begin to recover as early as 8 days after infusion.

Theoretically therefore, plasmapheresis and IVIG could be considered as early as 1 to 4 weeks after completion of rituximab infusion, although studies indicate that rituximab half-life progressively increases with subsequent weekly infusions (suggesting saturation of in vivo binding sites), and that a higher serum concentration is associated with better treatment outcome in B cell lymphomas. Conversely, plasmapheresis can be performed immediately prior to rituximab, but the ideal timing of rituximab infusion after a course of IVIG is unknown, with recommendations ranging from 1 to 6 weeks based on the serum half-life of IVIG.

What is the Evidence?

Stanley, JR, Wolff, K, Goldsmith, LA, Katz, SI, Gilchrest, B, Paller, AS, Leffell, DJ. “Pemphigoid”. 2011. (A more complete review of the clinical presentation and management of BP.)

Vodegel, RM, de Jong, MCJM, Meijer, HJ, Weytingh, MB, Pas, HH, Jonkman, MF. “Enhanced diagnostic immunofluorescence using biopsies transported in saline”. BMC Dermatol. vol. 4. 2004. pp. 10(An informative visual comparison of Michel's fixative, snap freeze, and saline for diagnostic immunofluorescence testing with pemphigus and pemphigoid. Saline is the preferred medium for pemphigoid, although samples must be processed or otherwise fixed within 24 hours.)

Sakuma-Oyama, Y, Powell, AM, Oyama, N, Albert, S, Bhogal, BS, Black, MM. “Evaluation of a BP180-NC16a enzyme-linked immunosorbent assay in the initial diagnosis of bullous pemphigoid”. Br J Dermatol. vol. 151. 2004. pp. 126-31. (Describes the validation of the BP180 ELISA for pemphigoid.)

Yoshida, M, Hamada, T, Amagai, M, Hashimoto, K, Uehara, R, Yamaguchi, K. “Enzyme-linked immunosorbent assay using bacterial recombinant proteins of human BP230 as a diagnostic tool for bullous pemphigoid”. J Dermatol Sci. vol. 41. 2006. pp. 21-30. (Describes the validation of the BP230 ELISA for pemphigoid, including the increased sensitivity of testing when combining the BP180 and BP230 ELISAs.)

Joly, P, Roujeau, JC, Benichou, J, Picard, C, Dreno, B, Delaporte, E. “A comparison of oral and topical corticosteroids in patients with bullous pemphigoid”. N Engl J Med. vol. 346. 2002. pp. 321-7. (In this study of 341 BP patients randomized to topical clobetasol [20 grams twice daily] or oral prednisone [0.5 to 1.0 mg/kg/day depending on disease severity], no difference in survival, efficacy, or complications was observed with moderate disease, but in severe disease topical corticosteroids were significantly more effective than oral prednisone [99% versus 91%], had increased 1-year survival [76% versus 58%] and had fewer complications [29% versus 54%].)

Beissert, S, Werfel, T, Fieling, U, Bohm, M, Sticherling, M, Stadler, R. “A comparison of oral methylprednisolone plus azathioprine or mycophenolate mofetil for the treatment of bullous pemphigoid”. Arch Dermatol. vol. 143. 2007. pp. 1536-42. (Seventy-three [73] patients with BP were randomized to treatment with oral methylprednisolone [0.5 mg/kg/day] plus azathioprine [2 mg/kg/day] or oral methylprednisolone [0.5 mg/kg/day] plus mycophenolate mofetil [2 g/day]. Similar efficacy and cumulative corticosteroid doses were observed in both groups, although mycophenolate mofetil had significantly lower rates of liver toxicity compared with azathioprine.)

Fairley, JA, Baum, CL, Brandt, DS, Messingham, KA. “Pathogenicity of IgE in autoimmunity: successful treatment of bullous pemphigoid with omalizumab”. J Allergy Clin immunol. vol. 123. 2009. pp. 704-5. (First published case report on omalizumab treatment of a BP patient refractory to prior corticosteroids, azathioprine, and minocycline. After a 16-week course, body surface involvement declined from 50% to 5% with resolution of tense bullae, although small erosions remained. This report supports an in vivo role for IgE autoantibodies in BP pathogenesis.)

Ahmed, AR, Dahl, MV. “Consensus statement on the use of intravenous immunoglobulin therapy in the treatment of autoimmune mucocutaneous blistering diseases”. Arch Dermatol. vol. 139. 2003. pp. 1051-9. (Provides guidelines for the use of IVIG, including indications, pre-screening, premedications, dose, frequency, monitoring, and therapeutic end points.)

Bernard, P, Reguiai, Z, Tancrède-Bohin, E, Cordel, N, Plantin, P, Pauwels, C. “Risk factors for relapse in patients with bullous pemphigoid in clinical remission: a multicenter, prospective, cohort study”. Arch Dermatol. vol. 145. 2009. pp. 537-42. (Patients [114] with BP in clinical remission for at least 3 months on less than 10g clobetasol cream per week or 7 mg daily prednisone equivalent discontinued all topical and systemic therapy and were followed for 1 year. The only factor significantly predicting disease relapse was a BP180 ELISA titer greater than 27-IU/mL.)

Fischel, F, Barbe, C, Joly, P. “Clinical and immunologic factors associated with bullous pemphigoid relapse during the first year of treatment: a multicenter, prospective study”. JAMA Dermatol. vol. 150. 2014. pp. 25-33.

Taghipour, K, Chi, CC, Vincent, A, Groves, RW, Venning, V, Wojnarowska, F. “The association of bullous pemphigoid with cerebrovascular disease and dementia: a case-control study”. Arch Dermatol. vol. 146. 2010 Nov. pp. 1251-4. (A study that showed bullous pemphigoid is significantly associated with cerebrovascular disease and dementia.)

Brick, KE, Weaver, CH, Savica, R, Lohse, CM, Pittelkow, MR, Boeve, BF. “A population-based study of the association between bullous pemphigoid and neurologic disorders”. J Am Acad Dermatol. vol. 71. 2014 Dec. pp. 1191-7. (Retrospective study of one Minnesota county that confirmed there is an association of bullous pemphigoid with neurologic disorders, especially dementia and Parkinson disease.)

Yu, KK, Crew, AB, Messingham, KA, Fairley, JA, Woodley, DT. “Omalizumab therapy for bullous pemphigoid”. J Am Acad Dermatol. 2014. vol. 71. Sep. pp. 468-74. (Article that discusses the use of omalizumab in 6 patients that demonstrated the fact that omalizumab neutralizes the activity of IgE in patients with bullous pemphigoid and improved 5 of 6 patients' disease activity.)