Pemphigoid: Clinical, Histologic, Immunopathologic, and Therapeutic Considerations

A 67-year-old woman noted onset of erythematous urticarial plaques and vesicles around her waist, and the associated pruritus interfered with sleep. Treatment with topical glucocorticoids provided only minimal improvement. Six months later, the patient saw a dermatologist who performed biopsies of lesional and normal-appearing perilesional skin for light and immunofluorescence microscopy studies, respectively. The former revealed a subepidermal blister with an associated eosinophil-rich leukocytic infiltrate; the latter identified continuous linear deposits of C3 in the epidermal basement membrane (BM). These clinical and laboratory findings were consistent with the diagnosis of bullous pemphigoid.

The patient was treated with systemic glucocorticoids (daily morning doses of prednisone ≤40 mg) for the next 3 years. Attempts to reduce or withdraw glucocorticoids were associated with the development of blisters and pruritus. The patient was subsequently referred to a clinic specializing in blistering diseases, where her diagnosis was confirmed by laboratory studies identifying circulating IgG anti-BM autoantibodies against the epidermal side of 1 molar salt-split skin (titer ≥160) and a 230-kd protein in extracts of cultured human keratinocytes. The patient was treated with prednisone, 20 mg every morning; azathioprine, 100 mg/d; calcium carbonate, 1250 mg/d; and cholecalciferol, 200 IU twice daily. Elevation of liver enzymes necessitated a reduction in azathioprine to 75 mg/d. Over the next 6 months the patient's glucocorticoid regimen was reduced to 10 mg/d; repeated attempts at further reduction met with flares of disease. Four and a half years later, her bullous pemphigoid remains active yet controlled.

The stratified squamous epithelium of skin and mucous membranes forms a continuous barrier against the external environment. Though rare, blistering diseases represent dramatic examples of how impairments in this epithelium result in disorders characterized by substantial morbidity and mortality.^{1 - 2} Blistering diseases may be inherited or acquired; of the latter, most are autoimmune. Autoimmune blistering diseases have been the subject of many investigations over the past 50 years. While these disorders were once not distinguished from each other, their autoimmune basis and pathophysiology are now better understood. This understanding is based on the demonstration that these patients have disease-specific autoantibodies directed against antigens in normal human skin. Autoantibodies from these patients have been used to identify these autoantigens and demonstrate (in many instances) that they represent important structural proteins in the epidermis or epidermal BM. In the case of selected disorders, passive transfer of IgG (either patient and/or experimental) against such autoantigens to animals has been shown to induce blisters with the same clinical, histologic, and immunopathologic features as those seen in patients. Among these disorders, bullous pemphigoid represents the prototypic example of an autoimmune subepidermal blistering disease (Table 1).

The purpose of this article is to summarize how the study of patients with bullous pemphigoid and the closely related blistering disorders pemphigoid gestationis (PG) and mucous membrane pemphigoid (MMP) has increased our understanding of pemphigoid pathophysiology and cutaneous biology.

Bullous pemphigoid is a subepidermal blistering disease predominantly seen in older individuals.^{3 - 4} Lesions usually consist of 1- to 3-cm blisters situated on either normal-appearing or inflamed skin (Figure 1). However, bullous pemphigoid is polymorphic and may present as an urticarial and/or a vesicular eruption (Table 2). Lesions tend to predominate on the lower trunk, axilla, groin, or flexor surfaces of the extremities; oral mucosal lesions are present in approximately one third of patients. Pruritus is a common feature of bullous pemphigoid; it may be mild or quite severe. Nontraumatized (eg, nonexcoriated) lesions tend to heal without scarring. Bullous pemphigoid is typically a chronic disease characterized by periods of exacerbation and partial remission; however, some patients experience complete remission after 6 to 10 years of active disease.

Biopsies of newly developed lesions from patients with bullous pemphigoid typically show mast cell degranulation, dermal edema, and an eosinophil-rich leukocytic infiltrate within the papillary dermis and along the epidermal BM; such alterations progress to frank subepidermal blister formation (Figure 1, D). The relative intensity of the leukocytic infiltrate in such lesions generally correlates with their clinical character (ie, lesions on noninflamed skin are relatively "cell-poor"). Approximately 35 years ago, seminal studies by Beutner and Jordon and colleagues^{5 - 7} demonstrated that patients with bullous pemphigoid have in situ deposits of IgG and complement components in their epidermal BMs as well as circulating IgG autoantibodies that bind epidermal BM in normal human skin (Figure 1, E and F). These findings demonstrated that patients with bullous pemphigoid, like those with other autoimmune diseases, have an immune response to constituents of normal tissue (not novel determinants present only in "diseased skin"). Moreover, because such autoantibodies are not present in patients with other blistering diseases, they represent markers for bullous pemphigoid.

Pemphigoid gestationis (gestational pemphigoid) is a rare subepidermal bullous disease of pregnancy and the puerperium.^{8 - 9} Its prior designation as herpes gestationis relates to the grouped character of lesions typically seen in these patients; the disorder has no relationship to an existing or prior viral infection. Pemphigoid gestationis may begin during any trimester of pregnancy or just after delivery; patients previously affected tend to experience earlier onset in subsequent gestations. Lesions typically involve the abdomen, trunk, and extremities; mucous membrane and facial lesions are uncommon. Like those seen in patients with bullous pemphigoid, lesions in PG are quite polymorphic and may range from urticarial papules and plaques to vesicles or frank tense bullae (Figure 1, B). Lesions in PG are almost always very pruritic. Exacerbations of PG often occur immediately after delivery; brief flares of disease may develop with resumption of menses or on exposure to oral contraceptives in previously affected patients. Although PG was once thought to be linked to an increased incidence of fetal morbidity and mortality, there is now agreement that neonates of affected women are at increased risk only of being slightly premature.^{10 - 11} The hormonal (or other) influences responsible for the development of PG remain to be determined, although its novel link to the gestational state is substantiated by rare cases that have developed in patients with underlying choriocarcinoma or hydatidiform mole.

Biopsies of newly developed lesions from patients with PG typically show teardrop-shaped subepidermal blisters in association with an eosinophil-rich leukocytic infiltrate, findings virtually indistinguishable from those seen in bullous pemphigoid.⁹ Similarly, direct immunofluorescence microscopy studies characteristically show linear deposits of C3 in these patients' epidermal BMs—deposits elicited by low-titer yet avid complement-fixing IgG autoantibodies.^{12 - 13} The demonstration that autoantibodies from patients with PG recognize the same autoantigen bound by IgG from patients with bullous pemphigoid confirmed on a molecular level the similarity of the clinical, histologic, and immunologic features of these 2 autoimmune diseases.^{14 - 17} The pathogenic nature of the autoantibodies in PG was substantiated not only by experimental studies (described below) but also by the development of transient skin lesions in some infants as a consequence of transplacental passage of maternal IgG. Maternal autoantibodies deposit in fetal epidermal BM, activate complement, and elicit blister formation. As maternal IgG is cleared from such infants after birth, lesions resolve correspondingly.

Mucous membrane pemphigoid is a rare autoimmune subepithelial blistering disease characterized by erosive and/or vesiculobullous lesions of mucous membranes.^{18 - 19} Similar lesions develop (to varying degrees) on the skin of some patients; such lesions tend to predominate on the scalp, face, or upper trunk and consist of a few scattered erosions or tense blisters situated on an erythematous and/or urticarial base. Common sites of involvement include the oral mucosa (especially the gingiva and buccal mucosa) and conjunctiva (Figure 1, C). Other sites that may be affected include the nasopharyngeal, laryngeal, esophageal, genital, and rectal mucosae. Scarring of mucous membranes is common, hence the alternate designation of this disorder as cicatricial (ie, scarring) pemphigoid. Serious complications may arise as a consequence of scarring of the ocular, laryngeal, esophageal, and/or genital mucosae. Examples of such complications include symblepharon, ankyloblepharon, entropion, and corneal opacities that may result in blindness or stricture formation. Scarring may also compromise function of the larynx, esophagus, or urethra.

Biopsies of lesions from patients with MMP generally demonstrate subepithelial blister formation in association with a mononuclear cell–predominant leukocytic infiltrate of variable intensity; neutrophils or eosinophils may be present in early lesions as well. Older lesions tend to show a scant leukocytic infiltrate and lamellar fibrosis in the upper dermis (ie, fibrosis characterized by collagen bundles parallel to the surface epithelium). Direct immunofluorescence microscopy of perilesional tissue characteristically demonstrates in situ deposits of IgG, IgA, and/or C3 in the epithelial BMs of these patients' mucosae and skin; circulating anti-BM autoantibodies are found in about half of these patients. Although MMP was once considered a single nosologic entity, recent studies have shown that it is a disease phenotype potentially associated with one of several different anti-BM autoantibodies.¹⁸

That patients with bullous pemphigoid have IgG anti-BM autoantibodies raised the possibility that these patients' sera might be used to identify and characterize autoantigens that represent constituents of epidermal BM and, similarly, determine if patients with this clinical phenotype all target the same autoantigen. Studies by Stanley,²⁰ Diaz,¹⁵ and others have shown that circulating IgG autoantibodies in patients with bullous pemphigoid bind 230-kd and, in approximately 65% of patients, 180-kd proteins that are associated with hemidesmosomes in basal keratinocytes (Figure 2). The first of these antigens to be identified, bullous pemphigoid antigen 1 (BPAG1 or BP230), is now recognized to be a member of the plakin protein family, a group of intracellular proteins that link the intermediate filament cytoskeleton to plasma membrane–associated adhesion plaques such as desmosomes and hemidesmosomes.³ That virtually all patients with bullous pemphigoid have autoantibodies against an intracellular protein in basal keratinocytes was initially considered to be a paradox. Currently, autoantibodies to BPAG1 are thought to develop as a consequence of keratinocyte injury and determinant spreading of the autoimmune response. These autoantibodies are thought to predominate because the corresponding intracellular autoantigen is not "exposed" (ie, accessible to circulating IgG) under ordinary circumstances. Although experimental studies have shown that anti-BPAG1 IgG can enhance inflammatory reactions at sites of injured keratinocytes in vivo,²¹ the autoantigen currently thought to harbor the key pathogenic epitope responsible for initiation and propagation of this disease is bullous pemphigoid antigen 2 (ie, BPAG2 or BP180). BPAG2 is a type II transmembrane molecule regarded as a member of the collagen gene family (specifically, type XVII collagen) due to the presence of 15 interrupted Gly-X-Y collagenous regions in its ectodomain.^{15 ,22 - 24} Ultrastructurally, this protein extends from the cytoplasm of basal keratinocytes to the lamina densa (ie, the BM proper) (Figure 2).^{25 - 26} Epitope mapping studies of bacterial recombinant proteins show that IgG from most patients with bullous pemphigoid binds a determinant within the first noncollagenous segment of proximal ectodomain of BPAG2 (the so-called NC16A domain located adjacent to the plasma membranes of basal keratinocytes).¹⁶ Passive transfer of experimental IgG directed against the murine homologue of this immunodominant determinant to neonatal BALB/c mice produced clinical (ie, blisters), histologic (ie, a leukocyte-rich infiltrate), and immunopathologic (ie, deposits of IgG and C3 in epidermal BM) alterations that mimic those seen in patients with bullous pemphigoid.²⁷ Subsequent studies demonstrated that this experimental animal model of bullous pemphigoid is dependent on the sequential activation of complement, generation of a polymorphonuclear leukocytic infiltrate, and protease activity in the latter.^{27 - 31} While autoantibody targeting of epitopes within the NC16A domain of BPAG2 is thought responsible for the pathogenesis of bullous pemphigoid and PG, the reason a complement-dependent, IgG-mediated animal model requires such epitope specificity is not yet explained. In addition, recent studies have found that the autoantibody response to BPAG2 in patients with bullous pemphigoid is heterogeneous in that some patients' serum reacts with intracellular or C-terminal portions of this protein as well as its NC16A domain.^{32 - 34} A direct correlation has been described between the presence of anti-BPAG2 autoantibodies and disease severity in patients with bullous pemphigoid.^{35 - 37}

Studies of serum samples from patients with MMP found that autoantibodies from these patients generally target 1 of several different autoantigens in epidermal BM.¹⁸ Interestingly, the most common MMP autoantigen identified to date is BPAG2.^{38 - 40} However, in contrast to the reactivity of autoantibodies from patients with bullous pemphigoid or PG, anti-BPAG2 IgG in patients with MMP targets the C-terminal, the most distal extracellular domain of this collagen (with or without associated reactivity to the NC16A domain).^{39 - 40} Autoantibody targeting of the C-terminus of BPAG2 (ie, a site positioned "deeper" within epidermal BM) is thought to account for the tendency of lesions in patients with MMP to scar. Additional studies have shown that a subset of patients with MMP have IgG anti-BM autoantibodies directed against laminin 5 (α3β3γ2), a heterotrimeric protein localized to the interface of the lamina lucida and lamina densa in epidermal BM (ultrastructurally, a site overlapping the localization of the BPAG2 C-terminus).^{41 - 42} Passive transfer of experimental anti–laminin 5 IgG to neonatal BALB/c mice induced noninflammatory subepidermal blisters of skin and mucous membranes independent of complement activation or mast cell degranulation.⁴³ Moreover, experimental as well as patient anti–laminin 5 IgG induced noninflammatory subepidermal blisters in grafts of human skin on immunodeficient mice.⁴⁴ Anti–laminin 5 IgG autoantibodies in most patients with this form of MMP bind the α chain of laminin 5, the subunit of this laminin isoform that is thought to play a key role in promoting adhesion of basal keratinocytes to epidermal BM.^{45 - 46} While the C-terminal portion of BPAG2 and laminin 5 are thought to represent the "major" MMP autoantigens, patients with autoantibodies directed against yet-to-be-defined BM constituents have been reported (as have patients with mucosal-predominant forms of epidermolysis bullosa acquisita, an autoimmune subepidermal blistering disease characterized by IgG anti-BM autoantibodies against collagen VII).^{18 ,47 - 50}

As is true for other autoimmune disorders, certain class II major histocompatibility complex (MHC) haplotypes are overrepresented in patients with blistering diseases. In the case of various forms of pemphigoid, the common class II MHC allele DQβ1*0301 predominates.^{51 - 53} Expression of DQβ1*0301 on antigen-presenting cells is thought to be involved in the presentation of immunodominant epitopes of BPAG2 to autoreactive T cells in patients with bullous pemphigoid. Such immunogenetic considerations are of particular relevance since antibody responses to protein antigens are T-cell dependent, and the latter are potent regulators of immune responses. Accordingly, recent studies have sought to identify T-cell lines and clones from patients with bullous pemphigoid that react to recombinant forms of BPAG2. One such study identified primary in vitro T-cell responses to recombinant forms of the BPAG2 ectodomain in 10 of 12 patients with bullous pemphigoid expressing the DQβ1*0301 allele as well as 8 of 10 DQβ1*0301–positive healthy controls.⁵⁴ While the responsiveness of 3 autoreactive T-cell lines from 2 patients with bullous pemphigoid and 1 healthy control in this study were immunogenetically restricted, peripheral blood mononuclear cells from 14 individuals who were DQβ1*0301–negative showed no response to BPAG2 recombinants. Interestingly, autoreactive BPAG2-specific CD4 T-cell lines and clones from 5 patients with bullous pemphigoid produced both T_H1 and T_H2 cytokines, while autoreactive T-cell lines from 3 DQβ1*0301–positive controls exclusively produced the T_H1 cytokine interferon gamma.⁵⁴ The notion that autoreactive T_H2 cells in patients with bullous pemphigoid play a critical role in autoantibody production was not surprising since IgG4 anti-BM autoantibodies predominate in these patients,^{55 - 56} IgE autoantibodies against BPAG1 are present in patients with severe disease,⁵⁷ BPAG2–reactive controls exhibit only T_H1 responses to this autoantigen,⁵⁴ and recent immunohistochemical and in situ hybridization studies found a predominance of T_H2 cytokines in bullous pemphigoid lesions.⁵⁸

While the studies described above suggest that the DQβ1*0301 allele is associated with the ability of T cells to develop proliferative responses to BPAG2 independent of bullous pemphigoid, the exact mechanisms responsible for the initiation of this autoimmune blistering disease are currently unknown. Organ-specific autoimmune diseases, including various forms of pemphigoid, are not thought to develop from failure of central or peripheral tolerance to autoantigens.⁵⁹ Instead, these disorders appear to result from abrogation of immune ignorance to self as a consequence of immunologic responses to foreign antigens that cross-react with determinants in the normal host. Molecular mimicry is a proposed mechanism by which exogenous agents (eg, a microorganism) may trigger an immune response to self.⁶⁰ In this model, a susceptible host acquires an infection with an agent that has antigens immunologically similar to host determinants yet sufficiently unique to induce an immune response on presentation to T cells. Consequently, "tolerance" to autoantigens is broken and the pathogen-initiated, cross-reactive immune response results in tissue destruction and disease.

The treatment of patients with autoimmune blistering diseases is grounded in clinical experience rather than randomized controlled clinical trials. A broad overview of treatment options for patients with various forms of pemphigoid is outlined herein. Bullous pemphigoid of minimal severity and/or distribution can sometimes be controlled with topical glucocorticoids. However, the mainstay of treatment of this disorder is systemic glucocorticoids, with initial daily morning doses of prednisone typically in the range of 0.75 to 1.0 mg/kg per day.¹⁹ Patients with extensive disease may require the addition of other immunosuppressive agents (eg, azathioprine, mycophenolate mofetil, or cyclophosphamide) to a regimen of daily glucocorticoids for control. Such treatment has reduced the morbidity and mortality due to bullous pemphigoid. Unfortunately, both still occur as a consequence of the disease (and/or its treatment) in elderly and/or debilitated patients. Patients with MMP are treated much like those with bullous pemphigoid except that treatment is guided not only by the extent of disease but also by the particular sites of involvement.¹⁸ More specifically, involvement of ocular, laryngeal, esophageal, and/or genital mucosae merits more aggressive treatment due to the potentially serious complications that can affect these sites. Measures to minimize the adverse effects of chronic treatment with systemic glucocorticoids are warranted.

Most patients with PG also require treatment with moderate (or higher) doses of daily glucocorticoids (eg, 0.5-1.0 mg/kg per day) at some point in the course of their disease.⁹ Despite concern about the use of such medications during gestation, current evidence suggests that there is no difference in the frequency of uncomplicated live births in patients with PG treated with systemic glucocorticoids vs those treated more conservatively.⁶¹ Postpartum flares of PG may require interim treatment with high doses (ie, approximately 1.0 mg/kg per day) of systemic glucocorticoids even in patients whose disease was previously controlled or in remission. If systemic glucocorticoids are administered during pregnancy, newborns are at risk for development of reversible adrenal insufficiency.

The recent production of recombinant forms of autoantigens recognized by B and T cells from patients with various forms of pemphigoid suggests that it may be possible to develop antigen-specific forms of immunotherapy for these diseases. For example, plasmapheresis coupled with antigen-specific depletion of circulating autoantibodies by immunoadsorption substrates bearing recombinant forms of BPAG2 might prove efficacious. Alternatively, production of blocking or toxin-conjugated peptides that either impair or eliminate antigen-specific T cells driving autoantibody production in these patients represents another immunotherapeutic approach. This possibility is of particular interest since bullous pemphigoid, PG, and certain forms of MMP share BPAG2 as a major target autoantigen as well as DQβ1*0301 as a predisposing class II MHC allele.

The study of patients with various forms of pemphigoid has furthered understanding of both disease pathophysiology and cutaneous biology. At present, these disorders can be diagnosed with reasonable certainty based on a combination of their clinical, histologic, and immunopathologic findings. Characterization of the specific reactivity of patient autoantibodies by specialized immunochemical or investigative techniques can elucidate cases that are otherwise difficult to classify. The ability to distinguish these diseases makes it possible to define better the prognosis of patients with various types of pemphigoid.

The 3 forms of pemphigoid discussed herein were once thought to represent autoimmune diseases distinguished by autoantibodies directed against different autoantigens in epidermal BM. Recent studies suggest that autoreactivity in such diseases may be more epitope- than antigen-specific. More specifically, patients with bullous pemphigoid and PG have IgG autoantibodies directed against the NC16A domain of BPAG2, while in many patients with MMP the most C-terminal portion of this same autoantigen (or the α subunit of the closely positioned laminin 5 heterotrimer) is targeted. Adding to this complexity, recent studies suggest that autoantibodies in patients with another autoimmune subepidermal blistering disease (linear IgA dermatosis) are of a different isotype (ie, IgA) and directed against a specific neoepitope on a 120-kd cleavage fragment of BPAG2.⁶² Further elucidation of such B- and T-cell immunodominant epitopes on this autoantigen expands the possibility that new and specific therapies for these autoimmune blistering diseases may soon be developed.