Paraneoplastic pemphigus (PNP) is an autoimmune vesiculobullous disorder.¹ PNP was first described in 1990 as a distinct clinical entity different from pemphigus vulgaris (PV).¹ PNP is most commonly associated with an underlying hematologic neoplasm such as non-Hodgkin lymphoma, chronic lymphocytic leukemia (CLL), and Castleman disease.² The heterogeneous clinical presentation often delays diagnosis, and constrictive bronchiolitis, a pulmonary manifestation of this disease, is the primary cause of morbidity in PNP patients.^3,4 To more appropriately reflect the diverse clinical findings, including organ involvement, the term “paraneoplastic autoimmune multiorgan syndrome” has been proposed as a more inclusive term than “PNP.”⁵

The worldwide incidence of PNP is unknown; however, it usually affects adults between the ages of 45 and 70⁵ and can also be seen in children between 7 and 18 years of age.⁶ Overall, there is a male predominance. Approximately 80% of PNP cases have been associated with lymphoproliferative diseases.^2,3 These include non-Hodgkin lymphomas, of which CLL is most commonly associated with PNP, and Castleman disease.^2,3 PNP has also been reported, albeit less commonly, with thymoma, carcinomas of various tissue types, sarcomas, malignant melanoma, and Waldenstrom macroglobulinemia.^2,3,5,6,7 When PNP is suspected in pediatric patients, Castleman disease should be considered in the differential diagnosis as the underlying neoplasm.⁶

PNP and PV may appear to be similar; however, studies suggest that the two entities are distinct with regard to immunology and pathogenesis. Both humoral and cell-mediated immunity are involved in the pathogenesis of PNP.³ Of the humoral effectors, it has been found that IgG1 and IgG2 subclasses predominate in PNP sera whereas IgG3 and IgG4 have been found in association with fewer cases.⁸ Cadherins and plakins, which are responsible for keratinocyte cell adhesion, are the protein families targeted in PNP.^9,10 From the cadherin family proteins, autoantibodies are found against the desmosomal proteins desmoglein 3 (Dsg3) and, less commonly, desmoglein 1 (Dsg1).⁹ Anti-Dsg3 autoantibodies disrupt desmosomes, thus resulting in acantholysis, suprabasilar clefting, and blistering.¹⁰

Dsg3 is also the primary autoantigen in PV.³ Thus, the plakin autoantigens more reliably differentiate PNP from PV.³ Plakin proteins are found in desmosomes and hemidesmosomes and mediate attachments between intermediate filaments and transmembrane adhesion molecules.³ The plakin family antigens targeted in PNP include desmoplakin I (250 kDa), desmoplakin II (210 kDa), bullous pemphigoid antigen (230 kDa), envoplakin (210 kDa), periplakin (190 kDa), and plectin (500 kDa).⁹ α2-Macroglobulin-like 1 (A2ML1), a 170-kDa protein, has been recently identified as an additional antigen targeted in PNP.¹¹ Compared to PV, not only are there a greater number of autoantigens recognized in PNP, there is also a broader distribution of epitopes recognized within select autoantigens.^3,8 For example, in PNP, IgG recognizes a wide distribution of epitopes in the Dsg3 extracellular domain. This contrasts with PV in which autoantibodies primarily target the Dsg3 N-terminus.⁸

In addition to the role of humoral mechanisms in PNP pathogenesis, cell-mediated immunity has also been implicated.^3,12 Mononuclear cell inflammatory infiltrates have been observed at the dermoepidermal junction of histologic specimens in PNP tissue samples and have been found to contain cells such as CD8⁺ cytotoxic T lymphocytes, CD56⁺ natural killer cells, and CD68⁺ monocytes/macrophages.¹² Cell-mediated immunity may help explain the various clinical and histologic presentations including the lichenoid-like or erythema multiforme (EM)–like features.^3,12 Furthermore, as demonstrated in neonatal mice, autoantibodies of PNP patients injected into mice produce acantholytic skin lesions, but do not reproduce other internal organ involvement seen in PNP patients.¹⁰

The mechanisms by which hematologic neoplasms induce autoimmunity are not well understood; however, a few speculative hypotheses have been proposed. One theory suggests that the tumor itself produces autoantibodies targeting the skin and oral mucosa. However, this may not be a universal mechanism because PNP triggered by Waldenstrom macroglobulinemia would be unlikely to produce IgG-class pathogenic autoantibodies.³ Additionally, there has been no evidence to suggest that the tumor cells express epithelial proteins, which could conceivably initiate the formation of autoantibody formation.³ The favored hypothesis is that the malignancy induces dysregulation of cytokines, which may drive the development of autoimmunity.³ Serum interleukin 6 (IL-6) may be elevated in hematologic neoplasms associated with PNP including CLL and Castleman disease.¹³ Furthermore, increased production of IL-6 has been associated with multiple clinical disorders including other autoimmune diseases.¹⁴ In murine models, increased IL-6 promoted B-lymphocyte hyperproliferation resulting in hypergammaglobulinemia of the IgG1 subclass.¹⁴ Thus, targeted therapies against IL-6 or, more specifically, its receptor may be considered in specific cases in which IL-6 is upregulated.¹⁴