Seminar

Pemphigus
Enno Schmidt, Michael Kasperkiewicz, Pascal Joly

Lancet 2019; 394: 882–94 Pemphigus consists of a group of rare and severe autoimmune blistering diseases mediated by pathogenic autoantibodies
Department of Dermatology mainly directed against two desmosomal adhesion proteins, desmoglein (Dsg)1 and Dsg3 (also known as DG1 and
(Prof E Schmidt MD), Lübeck DG3), which are present in the skin and surface-close mucosae. The binding of autoantibodies to Dsg proteins induces
Institute for Experimental
a separation of neighbouring keratinocytes, in a process known as acantholysis. The two main pemphigus variants are
Dermatology (Prof E Schmidt),
University of Lübeck, Lübeck, pemphigus vulgaris, which often originates with painful oral erosions, and pemphigus foliaceus, which is characterised
Germany; Department of by exclusive skin lesions. Pemphigus is diagnosed on the basis of either IgG or complement component 3 deposits
Dermatology, Keck School of (or both) at the keratinocyte cell membrane, detected by direct immunofluorescence microscopy of a perilesional biopsy,
Medicine, University of
with serum anti-Dsg1 or anti-Dsg3 antibodies (or both) detected by ELISA. Corticosteroids are the therapeutic mainstay,
Southern California, Los
Angeles, CA, USA which have recently been complemented by the anti-CD20 antibody rituximab in moderate and severe disease.
(Prof M Kasperkiewicz MD); Rituximab induces complete remission off therapy in 90% of patients, despite rapid tapering of corticosteroids, thus
Department of Dermatology, allowing for a major corticosteroid-sparing effect and a halved number of adverse events related to corticosteroids.
Rouen University Hospital,
Rouen, France (Prof P Joly MD);
and INSERM Unit 2345, Introduction Two major pemphigus variants can be differentiated,
French Reference Center for Pemphigus diseases are life-threatening, chronic, pemphigus vulgaris and pemphigus foliaceus. Pemphigus
Autoimmune Bullous Diseases, autoim­ mune blistering diseases characterised by the diseases share some clinical characteristics, such as flaccid
Normandy University, Rouen,
France (Prof P Joly)
formation of splits within the epidermis and sur­face- blisters and erosions, and, in contrast to pemphigoid
close epithelia, accompanied by acantholysis, a diseases,1 a positive Nikolsky’s sign (ie, friction of non-
Correspondence to:
Prof Enno Schmidt, Department histopathological term that describes the separation of lesional skin does induce intraepidermal disruption and
of Dermatology, University of keratinocytes from each other (figure 1A and 1B). visible erosion; appendix p 18).
Lübeck, 23538 Lübeck, Germany Autoantibodies are mainly of the IgG isoform and The term pemphigus derives from pemphix, the Greek
enno.schmidt@uksh.de
directed against two structural proteins of epidermal word for blister. Some of the historical aspects of pem­
See Online for appendix
desmosomes, desmoglein (Dsg) 1 and Dsg3 (also known phigus are provided in the appendix (p 1).
as DG1 and DG3; figure 2). Des­mosomes are cell-cell
adhesion structures that connect neighbouring kerati­ Epidemiology
nocytes and are essential for the integrity of various Pemphigus vulgaris and pemphigus foliaceus account for
tissues including the skin. For improved treatment and 90–95% of pemphigus diagnoses. The relative frequencies
prognosis, pemphigus diseases need to be differentiated of pemphigus vulgaris and pemphigus foliaceus within a
from pemphigoid diseases, as the other main group of population vary greatly between countries, with the rela­
autoimmune blistering disorders. Pemphigoid diseases tive frequency of pemphigus vulgaris ranging between
are a heterogeneous group of disorders, characterised 95% in Saudi Arabia and 13% in Mali. In Europe and
by autoantibodies against structural proteins of the North America, pemphigus vulgaris accounts for 65–90%
dermal–epidermal junction and tissue destruction that of pemphigus cases (reviewed by Schmidt and colleagues2).
is mainly based on skin inflammation mediated by the Pemphigus usually manifests between the ages of
Fc receptor, resulting in subepidermal blisters (reviewed 45 and 65 years in most populations, with an as yet
by Schmidt and Zillikens1). Because prognosis and unexplained lower mean age of onset of about 45 years in
treatment vary considerably between pemphigus and South Africa and northeast China (appendix pp 3–4).
pemphigoid diseases, an exact diagnosis is needed. Reports from the UK and France have highlighted that
Diagnosis cannot be made clinically, but warrants the age-adjusted incidence of pemphigus is in­creasing
distinct assessment of both serum antibodies and with age.3,4 In childhood and adolescence, pemphigus
antibodies bound to the skin or mucous membranes. is exceedingly rare, with 1–4% of patients with pemphigus
vulgaris being younger than 18 years (as reviewed
previously5–7). Pemphigus foliaceus in children has been
Search strategy and selection criteria described only in individual cases6,7 outside the endemic
Data for this Seminar were identified by searches of PubMed areas, but in such areas (discussed later) up to 30% of
with the search term “pemphigus” from Jan 1, 2008, patients have been reported to be younger than 20 years.8,9
to Jan 31, 2019. Older literature was cited as selected In most epidemiological studies, a female predominance
milestone articles or indirectly through review articles. has been reported, with ratios of males to females
We also searched the reference lists of articles identified by between 1:1·1 and 1:1·7 in various populations.3,10,11
our search strategy and selected those we judged relevant.
Review articles and book chapters are cited to provide Incidence and prevalence
readers with further information and more references than The incidence of pemphigus differs considerably between
this Seminar can accommodate. populations, ranging from 0·6 per million per year in
Switzerland and 0·8 in Finland, to 8·0 in Greece and

882 www.thelancet.com Vol 394 September 7, 2019

 Seminar

10·0 in Iran. The highest incidences, 16·1 in the USA

A B
and 32·0 in Israel, have been recorded in the Jewish
population (appendix pp 3–4). The variability is most likely
to be related to different genetic backgrounds and trigger
factors (discussed later). This notion is supported by
observations in Israel (2006–15) and Germany (1989–97),
with Israel having a 3·6 times higher incidence of
pemphigus vulgaris in patients with Jewish heritage than
in those with Arabic background, and Germany having an
8·8 times higher incidence of pem­ phigus vulgaris in
central Europeans than in patients originating from Italy
and Turkey.12,13 In in 2014, prevalence was estimated at C D
95 per million inhabitants for pemphigus vulgaris and
10 per million for pemphigus foliaceus.11
Notably, in some rural areas in South America (Brazil,
northern Colombia, and Peru) and Tunisia,14 much
higher incidences of pemphigus foliaceus have been
noted than in other parts of the world. So-called endemic
pemphigus foliaceus was originally described in 1903
and in its most extensive form is known in Brazil as
fogo selvagem (wild fire).8,15 The prevalence of endemic
pemphigus foliaceus in the past century reached 3–5% of
the popu­lation in well studied rural areas of southeastern E F
Brazil and northern Colombia, but has considerably
decreased since then.8,15–17

Associated diseases and genetics and risk factors

Pemphigus vulgaris and pemphigus foliaceus are
associated with various diseases including other autoim­
mune disorders, psoriasis, neurological and psychiatric
disorders, and some malignancies (appendix pp 5–6).10,18,19
The strongest evidence for pemphigus vulgaris risk
alleles among all populations has been provided for G H
the HLA alleles DRB1*04:02 and DQB1*05:03 (reviewed
by Vodo and colleagues20). In fact, most patients with
pemphigus vulgaris express one of these two alleles.
In two meta-analyses of 18 studies on the association
of pemphigus vulgaris with alleles of the HLA-DRB1
and HLA-DQB1 genes, DRB1*04, DRB1*08, DRB1*14,
DQB1*05:03, and DQB1*03:02 were significantly in­
creased and DRB1*03, DRB1*07, DRB1*15, DQB1*05:01,
DQB1*02, DQB1*06:01, and DQB1*03:03 significantly
decreased in patients compared with individuals without Figure 1: Pemphigus vulgaris
the disease.21,22 In 2018, two genome-wide association Lesional histopathologies of oral (A) and skin (B) biopsies reveal suprabasal splitting and loss of contact between
neighbouring keratinocytes (acantholysis). (B) The skin biopsy also shows an inflammatory infiltrate in the serum
studies in Chinese patients with pemphigus vulgaris
crust and in the blister directly below the corneal layer. (C) Extensive erosions and enanthema at the right buccal
identified DRB1*14,23,24 DQB1*05:03,24 and DRB1*0424 as mucosa and palate. (D) Subtle erosions at the lower gingiva close to the teeth. Erosions and crusts on the face (E)
relevant risk alleles. So far, four non-HLA genes have been and extensive denudation of the entire neck and back (F). (G) Distal onychodystrophy and transverse Beau’s line of
associated with pemphi­ gus vulgaris in at least two the thumb nails 8 weeks after massive relapse. (H) Direct immunofluorescence microscopy of a perilesional biopsy
specimen shows intercellular deposits of IgG in the epidermis. All images are from the Department of
populations: DSG3, en­ coding the pemphigus vulgaris Dermatology of the University of Lübeck (Lübeck, Germany) and the Department of Dermatology of Rouen
autoantigen Dsg3; TAP2, encoding an ATP-binding University Hospital (Rouen, France) and have not been shown previously.
cassette transporter involved in antigen presentation; IL6,
encoding the pleiotropic cytokine interleukin-6; and ST18,
encoding a transcription factor involved in inflammation In addition to genetic susceptibility traits, environ­mental
and apoptosis over­expressed in skin tissue in pemphigus factors that trigger the disease have been postu­lated. Low
vulgaris.20,23,25 In both non-endemic and endemic Brazilian concentrations of anti-Dsg auto­antibodies are present in
pemphigus foliaceus, HLA DRB1*04 was the most the sera of about half of clinically unaffected first-degree
frequently reported susceptibilty allele.8,26 relatives of patients with pemphigus,27,28 and in the sera of

www.thelancet.com Vol 394 September 7, 2019 883

 Seminar

by the observation that cholinergic agonists are able to

reduce acantholysis induced by IgG autoantibodies in
Corneal layer pemphigus vulgaris in vitro.33,34
Epidermis
Target antigens
Dermis Dsg1 and Dsg3 are the major target antigens in
pemphigus.35–40 They belong to the cadherin gene family of
Ca²+-dependent transmembrane adhesion molecules
found both within and outside of desmosomes (reviewed
by Waschke and Spindler41). Desmosomes are adherence
structures that connect neighbouring cells, including kera­
tinocytes, and enable the cells to withstand mecha­nical
forces (reviewed by Dubash and Green42). Several of the
structural proteins of desmosomes (figure 2) are recog­
nised by pemphigus autoantibodies. Dsg proteins contain
Nucleus
five extracellular domains (EC1–5), a single transmembrane
domain, and a cytoplasmic domain containing plakoglobin
and plakophilin binding sites. The Dsg molecules interact
Desmosomal plaque Intermediate
filaments
with each other via their N-terminal EC1 and EC2 domains
that are also preferentially targeted by pem­phigus auto­
antibodies.43,44 In most patients, clinical phenotype is
determined by the targeted Dsg molecule. In those with
pemphigus foliaceus, lesions are restricted to the skin and
auto­antibodies solely recognise Dsg1, whereas those with
the mucosal-dominant type of pemphigus vulgaris prefer­
Cell membrane
Actin filaments entially generate anti-Dsg3 autoantibodies. In patients
with the mucocutaneous type of pemphigus vulgaris,
Dsg3
which presents with both mucosal and skin lesions,
autoantibodies against both Dsg1 and Dsg3 are detected.45
Plakoglobin
This remarkable relation of autoantibody specificity to
Epiplakin
clinical phenotype is reflected in the different expres­sions
of Dsg1 and Dsg3 in the epidermis and surface-close
Dsg3 Desmoplakin mucosal epithelia. This relation has also guided the
assumption that both Dsg molecules can compensate for
Dsg1 each other when expressed together in the same cell layer
if the adhesive property is compromised in one of the two,
Envoplakin Plakophilin
as explained by the Dsg compensation theory (appendix
Dsc
p 19).46 In some patients with the mucocutaneous type of
pemphigus vulgaris, autoanti­bodies are restricted to Dsg3,
with skin lesions mainly affecting the face and scalp.47
Periplakin IgG and IgA antibodies against desmocollin (Dsc)1,
Figure 2: Schematic diagram of the epidermal desmosome
Dsc2, or Dsc3 have been detected in the sera of less than
Non-desmosomal Dsg is also shown. Although only homophilic transinteractions between Dsg and Dsc molecules 5% of patients with pemphigus vulgaris and pemphigus
are depicted, heterophilic interactions have also been described. Dsg=desmoglein. Dsc=desmocollin. foliaceus in different populations.48 Increased frequencies
Histopathological image reproduced from Schmidt and Zillikens.1 of anti-Dsc reactivity have been recorded in the rare
variants of pemphigus (ie, IgA pemphigus, pemphigus
inhabitants of high prevalence areas of endemic pemphigus herpetiformis [appendix pp 1–2], and paraneoplastic
foliaceus,15 which suggests more than one triggering factor. pemphigus) and the clinical subtypes of pemphigus
In endemic pemphigus foliaceus, the 43·2 kDa salivary vulgaris, pemphigus vegetans, and atypical pemphigus.48–50
protein LJM11 of the sand fly Lutzomyia longipalpis was Like the Dsg family of proteins, Dsc molecules are Ca²+-
cross-reactive with Dsg1.29 In non-endemic pemphigus, dependent cadherins that form part of the desmosome
reported risk factors include the use of particular drugs (figure 2).41 In addition to antibodies against Dsg1,
such as penicillamine and capto­pril, exposure to pesticides, Dsg3, and Dsc proteins, reactivity against various other
metal vapour, ultraviolet light, and ionising radiation, molecules such as muscarinic and nicotinic ace­tylcho­
sustaining burns, under­going surgery, and stressful life line receptors, pemphaxin, mitochondrial proteins, and
events.30–32 Conversely, nicotine (a cholinergic agonist) has a thyroid peroxidase have been detected in pemphigus by
protective effect in pemphigus, which might be explained various approaches51,52 (as reviewed previously53,54).

884 www.thelancet.com Vol 394 September 7, 2019

 Seminar

Pathophysiology The pathogenic role of anti-Dsc antibodies in pem­phigus

The pathophysiology of pemphigus has been the subject is indicated by the clinical phenotype of severe erosions,
of recent comprehensive reviews.41,52–57 In brief, in the microscopic feature of suprabasal splitting in mice
genetically susceptible individuals, the autoimmune deficient in epidermal Dsc1, and the pathogenic effects of
reaction is driven by autoreactive T lymphocytes and B anti-Dsc3 IgG in vitro.77–79 The contribution of antibodies
lymphocytes. Autoreactive T cells are educated (matured) against non-desmosomal antigens to the pathophysiology
by antigen presenting cells that present specific Dsg of pemphigus is suggested by the correlation of serum
peptides via their HLA class II molecules (encoded by concentrations of anti-muscarinic acetylcholine receptor
the HLA class II risk haplotypes described previously). IgG with disease severity, and the copathogenic effect of
These CD4 autoreactive T lymphocytes are specific anti-mitochondrial antibodies in addition to anti-Dsg IgG
for Dsg mole­ cules and produce IL-10, and drive the in vitro80,81 (reviewed by Amber and colleagues53).
generation of Dsg-specific antibodies by B cells (as
reviewed previously55,56,58). The pathophysiological impor­ Clinical presentation and differential diagnoses
tance of T lymphocytes in pemphigus vulgaris has also Pemphigus vulgaris
been shown in vitro in transgenic mice expressing In most patients with pemphigus vulgaris, the disease
HLA-DRB1*04:02 and human CD4 receptor, which on originates in the oral cavity, presenting as enanthema and
immunisation with human Dsg3, were induced to erosions. Nearly all patients with pemphigus vulgaris will
express human anti-Dsg3 IgG with acan­tholytic activity.59 develop oral lesions at some stage of the disease. Erosions
Ample evidence exists from clinical and experimental are mainly found at the buccal mucosa, palate, tongue,
observations for the pathogenic relevance of auto­ and inner side of the lips, whereas on the gingiva,
antibodies against Dsg1 and Dsg3 in pemphigus, as both enanthema and erosions predominate (figure 1C
presented in figure 3. Nevertheless, not all anti-Dsg and 1D). Pain is highly variable and ranges from minor
antibodies are pathogenic, which might explain the discomfort when chewing hard food to the prevention of
finding that in some patients with pemphigus vulgaris, food intake leading to rapid weight loss. Involvement of
serum concentrations of anti-Dsg3 do not correlate the nasal mucosa gives rise to haemorrhagic crusts. Less
with disease activity (ie, the extent of lesions) and frequently involved mucosal surfaces than the nasal
might persist into remission.54 Pathogenic epitopes on mucosa are the pharynx, larynx (in up to 40% of patients,
Dsg1 and Dsg3 are Ca²+-dependent and con­formational often presenting as hoarseness82), oesophagus, urethra,
dependent, and typically clustered in the EC1 and EC2 glans penis, vulva, and the perianal area. In the
domains, but can be found on all the Dsg extracellular mucocutaneous variant of pemphigus vulgaris, skin
domains (EC1–5).43,44,68–70 lesions arise in parallel with mucosal lesions or develop
By contrast with pemphigoid diseases in which a during the course of the disease. Predilection sites of
cascade of events including complement activation and these skin lesions are the scalp, neck, axillae, and upper
effects mediated by antibody Fc receptors are pivotal to trunk, but any site of the body can be affected. The skin
autoantibodies inducing subepidermal blister forma­ lesions typically present as flaccid blisters, erosions and
tion,1 monovalent fragments of anti-Dsg antibodies that crusts, and large areas can become denuded (figure 1E
lack the Fc portion can also cause acantholysis in vitro and 1F). Nail involvement is also seen in patients with
and in vivo.71,72 The exact sequence of events in acan­ extended disease (figure 1G).83 When mechanical friction
tholysis mediated by anti-Dsg antibodies is yet to be is applied in perilesional skin, an erosion can be induced,
fully elucidated; however, three major mechanisms for which is a characteristic but not pathognomonic effect in
anti-Dsg IgG binding have been identified: (i) direct pemphigus termed Nikolsky’s sign (appendix p 18). Both
interference with Dsg transinteraction (ie, homophilic mucosal and skin lesions heal without scarring, but in
and heterophilic binding of Dsg molecules to Dsg and patients with darker skin, hyperpigmentation at affected
Dsc molecules on neighboring cells), an event termed skin areas might be visible for many months.
steric hindrance; (ii) remodelling of Dsg expres­sion on Pemphigus vegetans, a clinical subtype of pemphigus
the cell surface leading to internali­sation and depletion vulgaris, presents with papillomatous vegetations mainly
of Dsg from the cell membrane; and (iii) signalling located in the intertriginous spaces (eg, the groin, axillae,
events including p38 mitogen-activated protein and intergluteal folds), and with pustules in a grouped or
kinase, epidermal growth factor receptor, Rho GTPase, annular distribution.84
MYC proto-oncogene, and caspase pathways that Differential diagnoses of the mucosal variant of pem­
interfere with cytoskeletal architecture73–75 (as re­viewed phigus vulgaris include mucous membrane pemphigoid,
previously41,55–57). Although these three mechanisms are herpes simplex virus infection, oral lichen planus, apthous
sufficient for acantholysis mediated by anti-Dsg IgG, ulcers, Behçet’s disease, erythema multiforme, and
additional factors, such as non-desmosomal antibodies Stevens-Johnson syndrome. In the mucocutaneous
and soluble Fas ligand (also known as tumor necrosis variant, skin lesions might resemble those seen in
factor ligand superfamily member 6), might contribute pemphigoid diseases, impetigo, varicella zoster virus
to the pemphigus phenotype.76 infection, Grover’s disease, bullous drug eruptions

www.thelancet.com Vol 394 September 7, 2019 885

 Seminar

Normal IgG Pemphigus IgG Normal IgG

erythema, so-called puff pastry-like scaling, and crusting
(figure 4). Intact blisters are rarely seen because of
subcorneal splitting, which allows mechanical destruc­tion
of the blister roof (figure 4E). The face, scalp, and
seborrhoeic areas of the upper trunk are predilection sites
of the lesions (figure 4A–C). The Nikolsky’s sign is positive
(appendix p 18) and, similar to in pemphigus vulgaris,
lesions heal without scarring, but typically result in
Pemphigus IgG postinflam­matory hyperpigmentations. In mild cases, the
C
clinical presentation mimics severe seborrhoeic dermatitis
and Darier’s disease. In contrast to pemphigus vulgaris,
Disease activity

even in severe forms of pemphigus foliaceus, large eroded

areas are not seen and exfoliation might resemble
B Anti-Dsg1 D erythroderma also found in severe atopic dermatitis,
Anti-Dsg IgG psoriasis, mycosis fungoides, and drug eruptions. Erosions
Other are also prone to bacterial and herpes virus superinfection.
Anti-Dsg3
A E In pemphigus erythematosus, a clinical subtype of
Pemphigus IgG pemphigus foliaceus, lesions mainly involve photodis­
tributed areas such as the face, back, and trunk, and
resemble those in cutaneous lupus erythematosus. In
sun damaged skin, the clinical fea­tures and so-called
lupus band-like deposition of immune complexes at the
F
dermal–epidermal junction might be due to tissue
destruction induced by UV irradiation in this zone.85 The
Dsg3 clinical presentation of endemic pemphigus foliaceus is
Spleen
described in the appendix (p 1–2).

Splenocytes Rare pemphigus variants

Dsg3-deficient
deficient Paraneoplastic pemphigus
mice
ice
Paraneoplastic pem­ phigus accounts for about 5% of
pemphigus cases. Clinical presentation is polymorphous,
with flaccid blisters similar to pemphigus vulgaris, pus­
Rag2-deficient tules, tense blisters similar to bullous pemphigoid,
mice erythema-multiforme-like targetoid erythema and lichen­
oid lesions.86,87 The most striking clinical feature is the
Figure 3: Pathogenic effects of pemphigus autoantibodies from clinical and experimental findings severe stomatitis with, unlike in pemphigus vulgaris,
(A) The transplacental transfer of maternal autoantibodies from mothers with pemphigus can cause transient
blistering in newborn babies.60 (B) ELISA values of serum anti-Dsg antibodies parallel disease activity in nearly all
frequent painful ulcers on the tongue and erosions on the
patients with pemphigus foliaceus and most with pemphigus vulgaris.61,62 (C) Degradation of desmosomes (red stain) lips that extend to the vermillion border of the lips and
occurs in cultured keratinocytes after incubation with IgG from patients.63 (D) Acantholysis of cultured keratinocyte resemble those in Stephens-Johnson syndrome (appendix
sheets occurs on treatment with IgG from patients.64 (E) Intraepidermal splitting (arrow) and macroscopic blistering p 20). Other mucosal tissues such as genital, nasal, and
(not shown) occurs in mice injected with sera from patients or monoclonal murine anti-Dsg IgG.65,66 (F) Anti-Dsg3
antibody production and microscopic (arrow) and macroscopic (not shown) blistering are induced in ocular can also be affected.86,88 In almost all patients, a
immunodeficient mice with Rag2 knockout after adoptive transfer of lymphocytes from Dsg3-deficient mice after neoplasm is associated with the pemphigus,86–89 which
immunisation with recombinant Dsg3.67 The actions of other IgGs are debated. The clinical picture in (A) was provided warrants an extensive tumour search. In 70–80% of
courtesy of Susann Ott (Department of Paediatrics, Klinikum Bayreuth, Bayreuth, Germany) and the histology image patients with paraneoplastic pemphigus, a lymphopro­
in (F) was provided courtesy of Hayato Takahashi and Masayuki Amagai (Department of Dermatology, Keio University
School of Medicine, Tokyo, Japan). All other images are from the Department of Dermatology at the University of liferative disorder (mostly non-Hodgkin lymphoma
Lübeck (Lübeck, Germany) and have not been published previously. Dsg=desmoglein. Rag2=recombination and chronic lymphocytic leukaemia) is present, and
activating 2. in the remaining patients, thymoma, malignant solid
tumours, and Castleman tumour are typically diag­
including toxic epidermal necrolysis, and derma­ titis nosed.86–89 In Asia, the most frequent associated neo­
artefacta. When flexures are predominantly affected, plasm is Castleman’s disease90,91 and in some patients,
Hailey-Hailey disease (benign chronic familial pemphigus) the neoplasm is only detected during the course of
should be ruled out. pemphigus.86 In Chinese patients with paraneoplastic
pemphigus, myasthenia gravis has been reported in a
Pemphigus foliaceus third of patients.92 Pulmonary destruction leading to
In contrast to pemphigus vulgaris, in pemphigus foliaceus bronchiolitis has been found in 6% of European patients
mucosal surfaces are spared and lesions exclu­ sively with paraneoplastic pemphigus, but in 20% of Japanese
develop on the skin, predominantly presenting as patients, and 70% of the few cases reported in children,

886 www.thelancet.com Vol 394 September 7, 2019

 Seminar

and appeared to be more frequent with underlying

A B
Castleman’s disease.86,88,93,94 In Japanese patients, the
occurrence of bronchiolitis obliterans was associated with
anti-epiplakin antibodies, and explained experimentally
by the ectopic expression of Dsg3 in the lung.95,96 The
potential lung involvement led some clinicians to use the
term paraneoplastic auto­immune multiorgan syn­drome
C
rather than paraneoplastic pem­phigus.97
Autoantibodies in paraneoplastic pemphigus mainly
target Dsg3 and the plakin proteins envoplakin and
periplakin (figure 2).86,87,98–102 In addition, autoanti­
bodies against Dsg1, desmocollins, the plakin proteins,
desmoplakin, plectin, epiplakin, bullous pemphigoid
antigen (BP)230 (also known as dystonin-e), BP180 (also D E
known as collagen α-1[XVII] chain) and the proteinase
inhibitor α2-macroglobulin-like protein 1 have been
described (appendix pp 7–8 and figure 2). In contrast to
pemphigus vulgaris and pem­phigus foliaceus, cytotoxic
CD4 and CD8 lymphocytes seem important for tissue
destruction in paraneoplastic pemphigus.103,104 This
assumption is in line with the characteristic histo­ Figure 4: Pemphigus foliaceus
Erythema, erosions, and so-called puff pastry-like scaling on the back (A and B)
pathological findings of lympho­cytes in the epidermis, and upper abdomen (C). (D) Direct immunofluorescence microscopy of a
dyskeratotic keratinocytes, and lichenoid and interface lesional skin biopsy showing intercellular deposits of IgG in the epidermis.
dermatitis in patients with paraneoplastic pemphigus.86,89 (E) Histology of a lesional skin biopsy showing a subcorneal split directly below
Additionally, the transfer of Dsg3-specific CD4 cells and the corneal layer with some inflammatory cells in the blister cavity. All images
are from the Department of Dermatology of the University of Lübeck (Lübeck,
B cells from Dsg3-deficient mice into immunodeficient Germany) and have not been shown previously.
mice is able to induce skin lesions with similar
histopathological findings to those in patients with damage).108,109 In a large prospective multicentre study, on
paraneoplastic pemphigus.103 the basis of the quartiles of the disease scores disease
acti­
vity was defined as mild (<15 for PDAI, <17 for
Neonatal pemphigus ABSIS), moderate (15–45 for PDAI, 17–53 for ABSIS),
Neonatal pemphigus is a transient disease in newborn and severe (>45 for PDAI, >53 for ABSIS).108
babies caused by the transplacental passage of auto­ Health-related quality of life in pemphigus has been
antibodies from a mother with pemphigus, giving the established via various generic measures in different
earliest evidence for the direct patho­ genic effect of patient populations and led to heterogeneous, in part
pemphigus autoantibodies (figure 3; reviewed in Zhao contradictory, findings (reviewed by Rencz and
et al105). Skin lesions are present in about half of neonates colleagues110). Although no instrument to specifically
from mothers with pemphigus vulgaris and resolve either assess quality of life in pemphigus is available, a
spontaneously or with mild topical corticosteroids within questionnaire for auto­ immune bullous diseases has
1 to 4 weeks.106 However, pemphigus vulgaris in preg­ been developed and validated in different languages and
nancy seems associated with prematurity and fetal death, populations.111
particularly in mothers with severe clinical disease.
Other rare variants, IgA pemphigus and pemphigus Diagnostic investigations
herpetiformis that together account for <5% of pemphigus Diagnosis of pemphigus vulgaris, pemphigus foliaceus,
cases3 are presented in the appendix (pp 2, 20). and IgA pemphigus is based on a combination of the
clinical presentation, direct immunofluorescence micros­
Outcome measures and quality of life co­py of a perilesional biopsy, serology, and histopatho­logy
An international panel of experts has defined disease of a lesional biopsy (table).112–114 In these two types, direct
endpoints and therapeutic responses.107 The best immunofluorescence microscopy reveals intercellular
validated and most frequently used scores in clinical deposits of IgG (figure 1H and figure 4D), complement
studies are the Autoimmune Bullous Skin Disorder component 3, or both, and in IgA pemphigus, of IgA
Intensity Score (ABSIS; between 0 and 206 according to in the epidermis and mucosal epithelium. The most
the extent of skin lesions, number of involved anatomical sensitive substrate for indirect immunofluorescence
sites in the oral cavity, and discomfort during food microscopy for screening of serum autoantibodies is
intake) and the Pemphigus Disease Area Index (PDAI; monkey oesophagus,114 in which immunoreactants stain
between 0 and 263 based on the number of lesions on the epithelium in an inter­cellular pattern (appendix p 20).
the skin, scalp, and mucosa, and region of skin Highly sensitive and specific ELISA systems for the

www.thelancet.com Vol 394 September 7, 2019 887

 Seminar

General Target antigens* Clinical hallmarks Diagnostic clues

frequency
(% of cases)
Pemphigus 60–90% Dsg3†‡ and Dsg1‡ Nearly always mucosal involvement: mostly Mucosal lesions, positive Nikolsky’s sign, positive direct
vulgaris oropharyngeal, nasal, and genital immunofluorescence microscopy result§, and serum
antibodies against Dsg3
Pemphigus 10–30% Dsg1†‡ Crusted erosions, erythema, scales mainly in the Crusted erosions, no mucosal lesions, positive Nikolsky’s
foliaceus seborrhoeic areas; no mucosal lesions sign, positive direct immunofluorescence microscopy
result§, and serum antibodies against Dsg1
Paraneoplastic 5% Envoplakin†‡, periplakin†, Dsg3†‡, Severe stomatitis involving the tongue and vermilion; Severe stomatitis, neoplasm, dyskeratosis and interface
pemphigus¶ desmoplakin, plectin, epiplakin, BP230‡, erythema multiforme-like and lichenoid skin lesions; dermatitis, and serum antibodies against Dsg3 and
BP180‡, Dsc1, Dsc2, Dsc3, Dsg1‡, and neoplasia, generally leukaemia or lymphoma plakins
α2-macroglobulin-like protein 1†
IgA pemphigus <3% Dsc1†, Dsc2, Dsc3, Dsg3, and Dsg1 Flaccid pustules, annular erosions, crusts, and Pustules, staining of intercellular IgA in the epithelium by
erythematous plaques direct immunofluorescence microscopy

Dsg=desmoglein. BP=bullous pemphigoid antigen. Dsc=desmocollin. *Further target antigens of uncertain pathogenic relevance have been described. †Main target antigens. ‡Commercial detection systems are
available. §Intercellular deposits of IgG, complement component 3, or both. ¶No official consensus for diagnosis has been reached (appendix pp 9–10).

Table: Autoantibody specificities and clinical characteristics of pemphigus diseases

detection of serum IgG against Dsg1 and Dsg3 are the paucity of randomised controlled trials (RCTs;
commercially available.112–114 Alternatively, an indirect appendix pp 11–16). Systemic cortico­steroids remain the
immunofluorescence micros­ copy system based on a mainstay treatment, and the initial objective of treatment
human cell line that expresses either recom­binant Dsg1 is to control disease activity (figure 5).112,113,116,117 Most
or recombinant Dsg3 can be applied (appendix p 20).114,115 clinicians will use oral prednis­one or oral prednisolone
In a recent international prospective multicentre study, as the mainstay corticosteroid treatment, frequently
anti-Dsg1, anti-Dsg3, or both types of IgG autoantibodies combined with oral azathioprine or a mycophenolate
could be detected in 98·5% of more than 300 consecutive compound (mycopheno­ late mofetil or mycophenolate
pemphigus serum samples.48 Histopathology of a lesional sodium). They will also begin to taper the corticosteroid
biopsy might help to reveal subcorneal splitting in at the end of the consolidation phase, when patients have
pemphigus foliaceus (figure 4E) and suprabasal split no new blisters for 2 weeks and healing of about 80% of
formation and acan­ tholysis in pemphigus vulgaris established lesions has occurred.107 The objectives during
(figure 1A and figure 1B). the following maintenance phase are to achieve complete
For the diagnosis of paraneoplastic pemphigus no remission, prevent relapses during prednisolone
official consensus has been reached (suggested diag­ tapering, and avoid as many potential adverse effects as
nostic criteria87,89,98,99 are shown in the appendix pp 9–10). possible.107 However, relapses occur in up to 50% of
The presence of anti-plakin antibodies is essential for the patients, and severe adverse events related to
diagnosis of paraneoplastic pemphigus, in addition to the immunosuppressants in up to 65% of patients, and only
clinical presentation of severe stoma­titis or lichenoid skin about half of patients are able to stop taking corticosteroid
lesions (or both), the presence of a neoplasia, and direct after 3 years.118–120 Treatment principles in pemphigus
immunofluorescence microscopy of a perilesional biopsy have recently been challenged by the first-line use of
showing intercellular IgG deposits in the epithelium that rituximab, which allowed 70% of patients to achieve
might or might not be accom­panied by linear staining of remission off-corticosteroids after 6 months and 90% of
the basement membrane zone (table).87,89,98,99 The serum of patients to achieve remission after 2 years.120
most patients with paraneoplastic pemphigus will also Recommendations for treatment of pemphigus vulgaris
contain antibodies against Dsg3 (appendix pp 9–10).100 and pemphigus foliaceus are outlined in figure 5.
A sensitive screening substrate by indirect immuno­ Oral corticosteroids are indicated in patients with
fluorescence microscopy is rat bladder, in which serum pemphigus vulgaris and in those with pemphigus
autoantibodies bind to the plakin-rich urothelium foliaceus. The initial dosage is usually 0·5–1·0 mg/kg
(appendix p 20).87,98,101 For the detection of anti-plakin daily of oral prednisone or prednisolone in patients
antibodies, only BP230 and envoplakin ELISA systems with moderate pemphigus, and 1·0–1·5 mg/kg daily in
are standardised and widely available.101 Autoantibody those with severe pemphigus.112,113,116 If lesions do not
reactivities in paraneoplastic pemphigus are detailed in improve within 10–15 days, the dose is increased.
the appendix (pp 7–8). Alternatively, intravenous corticosteroid pulses, such as
prednisolone 500–1000 mg daily for 5 consecutive days,
Treatment or dexa­methasone 100 mg daily for 3 consecutive days,
The availability of evidence-based treatment in can be given. Once disease control is achieved, a
pemphigus is hampered by the rarity of the disease and decrease of the prednisone or prednisolone dose is

888 www.thelancet.com Vol 394 September 7, 2019

 Seminar

usually proposed, by 25% every 2 weeks until 20 mg

daily is reached, then by 2·5 mg weekly until 10 mg Initial treatment
daily is reached, with tapering by 1 mg decrements Mild pemphigus: prednisolone, orally (1 mg/kg daily)
Moderate-to-severe pemphigus: prednisolone plus rituximab (2 × 1 g intravenously) or prednisolone plus
thereafter.112,113,116 A more rapid tapering of corticosteroids azathioprine or a mycophenolate compound* (or both orally)
over 3–6 months is pro­posed in patients treated first-
line with a combination of rituximab and prednisone or
prednisolone.120
Many drugs have been applied to reduce the required Disease control achieved Disease control not achieved
dose of corticosteroids and their detrimental adverse Without rituximab: tapering of prednisolone over Mild pemphigus: same as for moderate to severe
12–18 months and azathioprine or a mycophenolate pemphigus
effects. Azathioprine and mycophenolate compounds are compound over 2–3 years Moderate-to-severe pemphigus: increase
frequently recommended.112,113 To our understanding, With rituximab: tapering of prednisolone over prednisolone to 1·5 mg/kg daily, and add rituximab
6 months; repeated rituximab (0·5 or 1·0 g) (1 × 2 g) if not already being given
cyclophosphamide is less frequently used because of or plus IVIGs (2 g/kg over 2–5 days every 4–6 weeks);
after 6 months†
its toxicity, with large regional variations apparent. High- or plus corticosteroid pulses (eg, 500–1000 mg
dose intravenous immunoglobulins (IVIGs) are mainly prednisolone daily for 5 days; 100 mg dexamathasone
daily for 3 days); or plus immunoadsorption (over
applied in severe and recalcitrant pemphigus.112,113,116 The 3–4 days every 4 weeks)‡
importance of these conventional corticosteroid-sparing
drugs in the treatment of pemphigus is likely to change
in the coming years because of the high effectiveness of Treatment of relapse
While tapering prednisolone: back to the last prednisolone dose until disease control
rituximab.116,120 After tapering of prednisolone: restart prednisolone (dose dependent on severity of relapse)
Rituximab, a monoclonal antibody directed against the
CD20 antigen on B-lymphocytes, depletes CD20 B cells
from the circulation for 6–12 months, and was first
introduced in the treatment of paraneoplastic pemphigus Patient taking prednisolone Patient taking prednisolone Patient taking rituximab plus
in 2001 and of pemphigus vulgaris in 2002.121 The clinical alone plus azathioprine or a prednisolone
Plus rituximab (2 × 1 g) or plus mycophenolate compound Plus new cycle of rituximab
effectiveness of rituximab was subsequently highlighted azathioprine or a mycophenolate Plus rituximab or change (2 × 1 g)
in several case series applying different protocols and compound immunosuppressant*
adjuvants.122–124 On the basis of these case series and
two meta-analyses of collectively more than 500 patients Figure 5: Treatment algorithm for pemphigus vulgaris and pemphigus foliaceus
with pemphigus vulgaris or pemphigus foliaceus, Based on recommendations of an international panel of experts116 and revised guidelines of the European Academy
of Dermatology and Venereology. IVIGs=intravenous immunoglobulins. *If rituximab is not available or
researchers found that (1) remission was achieved in
contraindicated. †In particular in patients with initially severe pemphigus and high serum concentrations of
75% of patients after one cycle of rituximab with a mean anti-desmoglein antibodies at month 3; additional rituximab infusions (0·5 g) after 12 months and 18 months
delay of 3–6 months;56,120,121,124–126 (2) 80–90% of patients might be considered. ‡Might also be used as initial adjuvant therapy in patients with severe pemphigus.
reached complete remission on or off therapy, and on
and off therapy with repeated rituximab infusions;120,121,125,126 been approved by the US Food and Drug Administration
(3) relapses were frequent, occurring in 40–60% of and the European Medicines Agency for the treatment of
patients, particularly in those without maintenance moderate-to-severe pem­phigus.
infusions and with increased follow-up time;123 Several arguments favour maintenance infusions of
(4) application of rituximab early in the disease course rituximab to reduce the risk of relapses (eg, 6, 12, and
was associated with a higher proportion of patients in 18 months after the initial dose),120 despite their optimum
complete remission than in late application as second or time and dose being as yet unknown. Patients with high
third line treatment;120 and (5) severe adverse events were serum concentrations of anti-Dsg1 and anti-Dsg3 at
recorded in 4–10% of patients, including fatal outcomes month 3 seem to have an increased risk of relapse
(in 1–2%).120,126 (Joly P, unpublished), and might especially benefit from
In an RCT120 of 90 patients with newly-diagnosed maintenance infusions of rituximab.
pemphigus vulgaris or pemphigus foliaceus showed a In line with the RCTs listed in the appendix (pp 11–16),
large superiority of rituximab over a standard cortico­ a Cochrane review and a meta-analysis of RCTs asses­
steroid regimen. 89% of patients treated with rituximab sing adjuvant therapies, which included ten trials and
(2 × 1 g at baseline and 0·5 g at 12 months and 18 months) 559 participants, concluded that azathio­ prine, myco­
combined with short-term prednisolone (0·5–1·0 mg/kg phenolate compounds, and cyclophosphamide in
daily for 3–6 months) were in complete remission off addition to a corticosteroid were not better at achieving
therapy at 2 years, compared with 34% of those given disease control, clinical remission, or avoiding severe
prednisone alone (1·0–1·5 mg/kg daily for 12–18 months; adverse events including death than corticosteroid
p<0·0001). Additionally, in the rituximab group, the alone.127,128 Although the adjuvant therapies collectively
cumulative prednisone dose was three times lower decreased the risk of relapse by about 30%, analysis of
(p<0·0001) and the number of severe adverse events each adjuvant separately showed no single one that
two times lower (p=0·0084) than in the prednisone was significantly associated with fewer relapses
group.120 On the basis of these data, rituximab has recently than corticosteroid alone.127 However, azathio­prine and

www.thelancet.com Vol 394 September 7, 2019 889

 Seminar

cyclo­phosphamide did show a corticosteroid-sparing pemphigus given various rituximab protocols showed
effect.127 that immu­noadsorption-combined regimens resulted in
Azathioprine at a dose of 2–3 mg/kg daily (with normal the fastest control of disease activity before completion
thiopurine methyltransferase activity) and mycopheno­ of rituximab therapy; thus, immuno­adsorption might be
late compounds (mycophenolate mofetil at a dose of especially helpful in patients with severe disease during
2–3 g daily and mycophenolate sodium at a dose of the 2–4 months until rituximab becomes fully effective.
1·44 mg daily) might be considered as first-line cortico­ Although immunoadsorption is widely available in
steroid-sparing drugs when rituximab is not available or Europe, it is used mainly in Germany and Austria.
contraindicated (figure 5).116 Cyclo­phosphamide at doses Since paraneoplastic pemphigus is most likely to be
of 75–150 mg daily orally or 500–1000 mg monthly triggered by the underlying neoplasm, oncological the­
intravenously is rarely used in the USA and Europe. rapy is paramount. For the autoimmune bullous disease,
Because of its high number of adverse effects and systemic corticosteroids, rituximab, immunoadsorption,
long-term risk of infertility and malignancies, cyclo­phos­ and IVIG have successfully been applied.89,100
phamide might be reserved for patients with refractory Before the initiation of corticosteroid or immunosup­
disease when rituximab is not available.109 Methotrexate, pressive therapy, complete blood count; blood tests for
cyclosporine, and dapsone have very restricted indica­ creatinine, electrolytes, liver enzymes, albumin, fasting
tions in pemphigus and are reserved for individual serum glucose, and hepatitis B and C and HIV serology;
treatment situations. an ocular examination (for glaucoma and cataracts);
The efficacy of IVIGs in pemphigus has been sug­gested and the exclusion of tuberculosis are recommended.116
by several case series129 and one RCT.130 This RCT included Additionally, the vaccination status of the patient
Japanese patients with pemphigus vulgaris or pemphigus needs to be updated, and osteoporosis prophylaxis is
foliaceus who were unresponsive to pred­ nisolone indicated. Treatments such as nursing care and the
(≥20 mg daily). The endpoint time to escape from the treatment of children and pregnant woman are detailed
treatment protocol (ie, the length of time a patient elsewhere.7,112,135 The involvement of patient support
adhered to the protocol without any additional treatment) groups is highly recommended (appendix p 17).
was significantly extended in patients given a single cycle
of IVIG at 2 g/kg.130 In this meta-analysis, IVIG was the Outcomes
only adjuvant that increased the number of patients in The mortality associated with pemphigus vulgaris and
whom disease was controlled.127 IVIGs have a complex pemphigus foliaceus dramatically decreased with the use
mode of action, including saturation of the neonatal Fc of corticosteroids in the early 1950s, from appro­ximately
receptor, inhibition of antigen presentation, and anti- 75% to 30% in most regions.136 Over the past decade,
inflammatory effects of Fc-sialylated IgG (reviewed by studies from the UK, France, Israel, and Taiwan showed
Amber and colleagues131). The main advantages of IVIGs the risk of death to be 1·7 to 3 times higher in pemphigus
are that they act rapidly and do not increase the risk of than in controls.3,4,12,137,138 Infections, in particular pneu­
infection. IVIGs are generally infused at a dose of 2 g/kg monia and septicaemia, were the most frequent causes
over 2–5 days with monthly repetitions. In a case series,132 of death in different study populations, followed by
combining IVIGs with rituximab resulted in long-lasting cardiovascular diseases and peptic ulcer disease.137,139
remission without severe adverse effects. Herpes simplex virus DNA was found in the oral mucosa
Plasmapheresis has shown an unfavourable risk- of 30% of treatment-refractory patients with pemphigus
benefit ratio in pemphigus (appendix pp 11–16). By vulgaris, and might mimic an exacerbation of the disease.140
contrast, immunoadsorption does not require plasma Various comorbidities related to corticosteroid use have
substitution, specifically binds Ig molecules, allows the been reported in pemphigus vulgaris and pemphigus
processing of 2–3 plasma volumes, and rapidly removes foliaceus, including Cushing syndrome, infections (par­
anti-Dsg IgG from the circulation.133,134 Immuno­ ticularly opportunistic herpes virus, and fungal infections),
adsorption is applied on 3–4 consecutive days at intervals osteoporosis, venous thromboembolism, and type 2
of 3–4 weeks. In an RCT of 72 patients with pemphigus diabetes.138,141,142
vulgaris or pemphigus foliaceus, immuno­ adsorption In paraneoplastic pemphigus, mortality is notably
combined with standard therapy (ie, prednisolone at an higher than for pemphigus vulgaris and pemphigus
initial dose of 1 mg/kg daily combined with azathioprine foliaceus, reaching up to 60% in patients with paraneo­
or a mycophenolate compound) led to faster remission plastic pemphigus within 5 years of diagnosis, compared
and required signifi­ cantly fewer corticoste­ roids than with 20–25% in pemphigus vulgaris and pemphigus
standard therapy alone in patients with high disease foliaceus.3,88 The main causes of death in paraneoplastic
activity (DRKS00000566). Although the total number of pemphigus are infections, the underlying neoplasm,
adverse events was lowest in the immunoadsorption and bronchiolitis obliterans.88,93,94
plus standard therapy group, the number of severe Increased serum concentrations of anti-Dsg3, mucosal
adverse events was higher in the stan­dard therapy alone involvement, and younger age at disease onset have been
group (unpub­lished). A meta-analysis126 of patients with described as risk factors for increased disease duration in

890 www.thelancet.com Vol 394 September 7, 2019

 Seminar

pemphigus vulgaris and pemphigus foliaceus.143 For 3 Jelti L, Cordel N, Gillibert A, et al. Incidence and mortality of
patients with pemphigus vulgaris in remission, positive pemphigus in France. J Invest Dermatol 2018; 139: 469–73.
4 Langan SM, Smeeth L, Hubbard R, Fleming KM, Smith CJ, West J.
ELISA values for anti-Dsg3 seem to be predictors of Bullous pemphigoid and pemphigus vulgaris—incidence and
relapse.144 In paraneoplastic pemphigus, erythema multi­ mortality in the UK: population based cohort study. BMJ 2008;
forme-like skin lesions have been described as predictive 337: a180.
5 Mintz EM, Morel KD. Treatment of chronic bullous disease of
of fatal outcome.88 childhood. Dermatol Clin 2011; 29: 699–700.
6 Mintz EM, Morel KD. Clinical features, diagnosis, and pathogenesis
Outlook of chronic bullous disease of childhood. Dermatol Clin 2011;
29: 459–62.
Although the advent of rituximab has revolutionised
7 van Beek N, Schmidt E. Chapter 74: Autoimmune bullous
the treatment of pemphigus, the possibility of severe diseases. In: Höger P, Kinsler V, Yan A, eds. Harper’s textbook of
adverse events, the 10% of patients who do not adequately pediatric dermatology, 4th edn. Chichester: Wiley-Blackwell,
respond to rituximab, and those who require a high in press: 868–97.
8 Aoki V, Rivitti EA, Diaz LA. Update on fogo selvagem, an endemic
number of repeated rituximab infusions120,125,126 indicate form of pemphigus foliaceus. J Dermatol 2015; 42: 18–26.
the capacity for new treatment modalities. Phase 2 trials 9 Diaz LA, Sampaio SA, Rivitti EA, et al. Endemic pemphigus
are underway of ofatumumab (NCT02613910), an foliaceus (fogo selvagem): II. Current and historic epidemiologic
studies. J Invest Dermatol 1989; 92: 4–12.
alternative anti-CD20 anti­body with increased binding
10 Kridin K. Pemphigus group: overview, epidemiology, mortality,
affinity for CD20, and of inhibitors for the neonatal Fc and comorbidities. Immunol Res 2018; 66: 255–70.
receptor (EudraCT number 2017-002333-40), tyrosine- 11 Hubner F, Recke A, Zillikens D, Linder R, Schmidt E.
protein kinase BTK (EudraCT number 2018-002261-19), Prevalence and age distribution of pemphigus and pemphigoid
diseases in Germany. J Invest Dermatol 2016; 136: 2495–98.
and B-cell activating factor (NCT01930175). Furthermore, 12 Kridin K, Zelber-Sagi S, Bergman R. Pemphigus vulgaris and
clinical trials are awaited that are based on Dsg-specific pemphigus foliaceus: differences in epidemiology and mortality.
immunoadsorption70,145 and on chimeric autoantibody Acta Derm Venereol 2017; 97: 1095–99.
13 Hahn-Ristic K, Rzany B, Amagai M, Brocker EB, Zillikens D.
recep­tor technology that engineers human T cells to Increased incidence of pemphigus vulgaris in southern Europeans
express a chimeric autoantibody receptor consisting living in Germany compared with native Germans.
of Dsg3 fused to a CD137-CD3-ζ signalling domain, J Eur Acad Dermatol Venereol 2002; 16: 68–71.
14 Morini JP, Jomaa B, Gorgi Y, et al. Pemphigus foliaceus in young
allowing selective depletion of autoreactive B cells.146 women. An endemic focus in the Sousse area of Tunisia.
Contributors Arch Dermatol 1993; 129: 69–73.
All authors reviewed the literature, collected and interpreted data, 15 Warren SJ, Lin MS, Giudice GJ, et al. The prevalence of antibodies
and wrote the manuscript. ES prepared the figures. against desmoglein 1 in endemic pemphigus foliaceus in Brazil.
N Engl J Med 2000; 343: 23–30.
Declaration of interests
16 Goncalves GA, Brito MM, Salathiel AM, Ferraz TS, Alves D,
ES reports personal fees from Roche, grants from the German Research Roselino AM. Incidence of pemphigus vulgaris exceeds that of
Foundation and Euroimmun, and grants and personal fees from pemphigus foliaceus in a region where pemphigus foliaceus is
Fresenius Medical Care, Biotest, Union Chimique Belge, and argenx, endemic: analysis of a 21-year historical series. An Bras Dermatol
during the conduct of the Review. ES also reports grants and personal 2011; 86: 1109–12.
fees from Novartis, TxCell, True North Therapeutics, Incyte, and Amryt 17 Abreu-Velez AM, Hashimoto T, Bollag WB, et al. A unique form of
Pharma, and personal fees from Genetech, outside of the submitted endemic pemphigus in northern Colombia. J Am Acad Dermatol
work. PJ reports personal fees from Principia Biopharma, argenx, Lilly, 2003; 49: 599–608.
Novartis, Abbvie, Amgen, and Uriage, personal fees and non-financial 18 Kridin K, Zelber-Sagi S, Comaneshter D, Batat E, Cohen AD.
support from Roche, Janssen, Bristol-Myers Squibb, and Pierre Fabre, Pemphigus and hematologic malignancies: a population-based
outside of the submitted work. MK declares no competing interests. study of 11 859 patients. J Am Acad Dermatol 2018; 78: 1084–89.e1.
No authors within the past 3 years had stocks or shares, equity of a 19 Schulze F, Neumann K, Recke A, Zillikens D, Linder R, Schmidt E.
relevant company, a contract of employment or a named position on a Malignancies in pemphigus and pemphigoid diseases.
company board. No organisation other than The Lancet has asked the J Invest Dermatol 2015; 135: 1445–47.
authors to write, be named on, or submit the paper. 20 Vodo D, Sarig O, Sprecher E. The genetics of pemphigus vulgaris.
Front Med 2018; 5: 226.
Acknowledgments 21 Yan L, Wang JM, Zeng K. Association between HLA-DRB1
This Seminar is dedicated to Marcel F Jonkman, Groningen, Netherlands, polymorphisms and pemphigus vulgaris: a meta-analysis.
who contributed enormously to the field of research on pemphigus and Br J Dermatol 2012; 167: 768–77.
who passed away at the height of his career after a short and severe illness. 22 Li S, Zhang Q, Wang P, et al. Association between HLA-DQB1
We miss his vast clinical knowledge, sharp intellect, clear concepts, and polymorphisms and pemphigus vulgaris: a meta-analysis.
strong beliefs in scientific discussions. This work was supported by the Immunol Invest 2018; 47: 101–12.
German Research Foundation under Germany’s Excellence Strategy (grant 23 Gao J, Zhu C, Zhang Y, et al. Association study and fine-mapping
number EXC 2167).We acknowledge researchers who have contributed to major histocompatibility complex analysis of pemphigus
this field whose work was not cited directly but through review articles vulgaris in a Han Chinese population. J Invest Dermatol 2018;
because of space limitation. We thank Stephanie Goletz, Florian Heck, 138: 2307–14.
Sabrina Tofern, and Marina Kongsbak-Reim for their help with the 24 Zhang SY, Zhou XY, Zhou XL, et al. Subtype-specific inherited
immunofluorescence and histopathological photographs, predisposition to pemphigus in the Chinese population.
and Carolin Mahlerwein for the graphical design of figure 2. Br J Dermatol 2018; 180: 828–35.
25 Sarig O, Bercovici S, Zoller L, et al. Population-specific association
References between a polymorphic variant in ST18, encoding a pro-apoptotic
1 Schmidt E, Zillikens D. Pemphigoid diseases. Lancet 2013; molecule, and pemphigus vulgaris. J Invest Dermatol 2012;
381: 320–32. 132: 1798–805.
2 Schmidt E, Borradori L, Joly P. Epidemiology of autoimmune 26 de Sena Nogueira Maehara L, De-Souza-Santana FC, Porro AM, et al.
bullous diseases. In: Murrell D, ed. Blistering diseases, 1st edn. HLA class II alleles of susceptibility and protection in Brazilian and
Heidelberg: Springer, 2015: 251–64. Dutch pemphigus foliaceus. Br J Dermatol 2018; 178: e212–14.

www.thelancet.com Vol 394 September 7, 2019 891

 Seminar

27 Brandsen R, Frusic-Zlotkin M, Lyubimov H, et al. Circulating 51 Sajda T, Hazelton J, Patel M, et al. Multiplexed autoantigen
pemphigus IgG in families of patients with pemphigus: microarrays identify HLA as a key driver of anti-desmoglein and
comparison of indirect immunofluorescence, direct -non-desmoglein reactivities in pemphigus. Proc Natl Acad Sci USA
immunofluorescence, and immunoblotting. J Am Acad Dermatol 2016; 113: 1859–64.
1997; 36: 44–52. 52 Nguyen VT, Ndoye A, Shultz LD, Pittelkow MR, Grando SA.
28 Ahmed AR, Mohimen A, Yunis EJ, et al. Linkage of pemphigus Antibodies against keratinocyte antigens other than desmogleins
vulgaris antibody to the major histocompatibility complex in 1 and 3 can induce pemphigus vulgaris-like lesions. J Clin Invest
healthy relatives of patients. J Exp Med 1993; 177: 419–24. 2000; 106: 1467–79.
29 Qian Y, Jeong JS, Maldonado M, et al. Cutting edge: 53 Amber KT, Valdebran M, Grando SA. Non-desmoglein antibodies in
Brazilian pemphigus foliaceus anti-desmoglein 1 autoantibodies patients with pemphigus vulgaris. Front Immunol 2018; 9: 1190.
cross-react with sand fly salivary LJM11 antigen. J Immunol 2012; 54 Ahmed AR, Carrozzo M, Caux F, et al. Monopathogenic vs
189: 1535–39. multipathogenic explanations of pemphigus pathophysiology.
30 Wolf R, Tamir A, Brenner S. Drug-induced versus drug-triggered Exp Dermatol 2016; 25: 839–46.
pemphigus. Dermatologica 1991; 182: 207–10. 55 Pollmann R, Schmidt T, Eming R, Hertl M. Pemphigus:
31 Daneshpazhooh M, Fatehnejad M, Rahbar Z, et al. Trauma-induced a comprehensive review on pathogenesis, clinical presentation and
pemphigus: a case series of 36 patients. J Dtsch Dermatol Ges 2016; novel therapeutic approaches. Clin Rev Allergy Immunol 2018;
14: 166–71. 54: 1–25.
32 Morell-Dubois S, Carpentier O, Cottencin O, et al. Stressful life 56 Kasperkiewicz M, Ellebrecht CT, Takahashi H, et al. Pemphigus.
events and pemphigus. Dermatology 2008; 216: 104–08. Nat Rev Dis Primers 2017; 3: 17026.
33 Nguyen VT, Arredondo J, Chernyavsky AI, Pittelkow MR, Kitajima Y, 57 Spindler V, Eming R, Schmidt E, et al. Mechanisms causing loss of
Grando SA. Pemphigus vulgaris acantholysis ameliorated by keratinocyte cohesion in pemphigus. J Invest Dermatol 2018;
cholinergic agonists. Arch Dermatol 2004; 140: 327–34. 138: 32–37.
34 Lai O, Recke A, Zillikens D, Kasperkiewicz M. Influence of cigarette 58 Hertl M, Eming R, Veldman C. T cell control in autoimmune
smoking on pemphigus – a systematic review and pooled analysis bullous skin disorders. J Clin Invest 2006; 116: 1159–66.
of the literature. J Eur Acad Dermatol Venereol 2018; 32: 1256–62. 59 Eming R, Hennerici T, Backlund J, et al. Pathogenic IgG antibodies
35 Eyre RW, Stanley JR. Identification of pemphigus vulgaris antigen against desmoglein 3 in pemphigus vulgaris are regulated by
extracted from normal human epidermis and comparison with HLA-DRB1*04:02-restricted T cells. J Immunol 2014; 193: 4391–99.
pemphigus foliaceus antigen. J Clin Invest 1988; 81: 807–12. 60 Wu H, Wang ZH, Yan A, et al. Protection against pemphigus
36 Eyre RW, Stanley JR. Human autoantibodies against a desmosomal foliaceus by desmoglein 3 in neonates. N Engl J Med 2000; 343: 31–35.
protein complex with a calcium-sensitive epitope are characteristic 61 Schmidt E, Dahnrich C, Rosemann A, et al. Novel ELISA systems
of pemphigus foliaceus patients. J Exp Med 1987; 165: 1719–24. for antibodies to desmoglein 1 and 3: correlation of disease activity
37 Amagai M, Klaus-Kovtun V, Stanley JR. Autoantibodies against a with serum autoantibody levels in individual pemphigus patients.
novel epithelial cadherin in pemphigus vulgaris, a disease of cell Exp Dermatol 2010; 19: 458–63.
adhesion. Cell 1991; 67: 869–77. 62 Harman KE, Seed PT, Gratian MJ, Bhogal BS, Challacombe SJ,
38 Hashimoto T, Ogawa MM, Konohana A, Nishikawa T. Detection of Black MM. The severity of cutaneous and oral pemphigus is related
pemphigus vulgaris and pemphigus foliaceus antigens by to desmoglein 1 and 3 antibody levels. Br J Dermatol 2001;
immunoblot analysis using different antigen sources. 144: 775–80.
J Invest Dermatol 1990; 94: 327–31. 63 Farb RM, Dykes R, Lazarus GS. Anti-epidermal-cell-surface
39 Korman NJ, Eyre RW, Klaus-Kovtun V, Stanley JR. Demonstration pemphigus antibody detaches viable epidermal cells from culture
of an adhering-junction molecule (plakoglobin) in the autoantigens plates by activation of proteinase. Proc Natl Acad Sci USA 1978;
of pemphigus foliaceus and pemphigus vulgaris. N Engl J Med 1989; 75: 459–63.
321: 631–35. 64 Ishii K, Harada R, Matsuo I, Shirakata Y, Hashimoto K, Amagai M.
40 Koch PJ, Walsh MJ, Schmelz M, Goldschmidt MD, Zimbelmann R, In vitro keratinocyte dissociation assay for evaluation of the
Franke WW. Identification of desmoglein, a constitutive pathogenicity of anti-desmoglein 3 IgG autoantibodies in
desmosomal glycoprotein, as a member of the cadherin family of pemphigus vulgaris. J Invest Dermatol 2005; 124: 939–46.
cell adhesion molecules. Eur J Cell Biol 1990; 53: 1–12. 65 Anhalt GJ, Labib RS, Voorhees JJ, Beals TF, Diaz LA. Induction of
41 Waschke J, Spindler V. Desmosomes and extradesmosomal adhesive pemphigus in neonatal mice by passive transfer of IgG from
signaling contacts in pemphigus. Med Res Rev 2014; 34: 1127–45. patients with the disease. N Engl J Med 1982; 306: 1189–96.
42 Dubash AD, Green KJ. Desmosomes. Curr Biol 2011; 21: R529–31. 66 Schulze K, Galichet A, Sayar BS, et al. An adult passive transfer
43 Tsunoda K, Ota T, Aoki M, et al. Induction of pemphigus phenotype mouse model to study desmoglein 3 signaling in pemphigus
by a mouse monoclonal antibody against the amino-terminal vulgaris. J Invest Dermatol 2012; 132: 346–55.
adhesive interface of desmoglein 3. J Immunol 2003; 170: 2170–78. 67 Amagai M, Tsunoda K, Suzuki H, Nishifuji K, Koyasu S,
44 Futei Y, Amagai M, Sekiguchi M, Nishifuji K, Fujii Y, Nishikawa T. Nishikawa T. Use of autoantigen-knockout mice in developing an
Use of domain-swapped molecules for conformational epitope active autoimmune disease model for pemphigus. J Clin Invest
mapping of desmoglein 3 in pemphigus vulgaris. J Invest Dermatol 2000; 105: 625–31.
2000; 115: 829–34. 68 Li N, Aoki V, Hans-Filho G, Rivitti EA, Diaz LA. The role of
45 Amagai M, Tsunoda K, Zillikens D, Nagai T, Nishikawa T. intramolecular epitope spreading in the pathogenesis of endemic
The clinical phenotype of pemphigus is defined by the pemphigus foliaceus (fogo selvagem). J Exp Med 2003;
anti-desmoglein autoantibody profile. J Am Acad Dermatol 1999; 197: 1501–10.
40: 167–70. 69 Di Zenzo G, Di Lullo G, Corti D, et al. Pemphigus autoantibodies
46 Stanley JR, Amagai M. Pemphigus, bullous impetigo, and the generated through somatic mutations target the desmoglein-3
staphylococcal scalded-skin syndrome. N Engl J Med 2006; cis-interface. J Clin Invest 2012; 122: 3781–90.
355: 1800–10. 70 Langenhan J, Dworschak J, Saschenbrecker S, et al.
47 Baykal C, Azizlerli G, Thoma-Uszynski S, Hertl M. Specific immunoadsorption of pathogenic autoantibodies in
Pemphigus vulgaris localized to the nose and cheeks. pemphigus requires the entire ectodomains of desmogleins.
J Am Acad Dermatol 2002; 47: 875–80. Exp Dermatol 2014; 23: 253–59.
48 Mindorf S, Dettmann IM, Kruger S, et al. Routine detection of 71 Payne AS, Ishii K, Kacir S, et al. Genetic and functional
serum antidesmocollin autoantibodies is only useful in patients characterization of human pemphigus vulgaris monoclonal
with atypical pemphigus. Exp Dermatol 2017; 26: 1267–70. autoantibodies isolated by phage display. J Clin Invest 2005;
49 Ishii N, Teye K, Fukuda S, et al. Anti-desmocollin autoantibodies in 115: 888–99.
nonclassical pemphigus. Br J Dermatol 2015; 173: 59–68. 72 Rock B, Labib RS, Diaz LA. Monovalent Fab’ immunoglobulin
50 Muller R, Heber B, Hashimoto T, et al. Autoantibodies against fragments from endemic pemphigus foliaceus autoantibodies
desmocollins in European patients with pemphigus. reproduce the human disease in neonatal Balb/c mice. J Clin Invest
Clin Exp Dermatol 2009; 34: 898–903. 1990; 85: 296–99.

892 www.thelancet.com Vol 394 September 7, 2019

 Seminar

73 Diercks GF, Pas HH, Jonkman MF. The ultrastructure of 97 Amber KT, Valdebran M, Grando SA. Paraneoplastic
acantholysis in pemphigus vulgaris. Br J Dermatol 2009; 160: 460–61. autoimmune multiorgan syndrome (PAMS): beyond the single
74 Sokol E, Kramer D, Diercks GFH, et al. Large-scale electron phenotype of paraneoplastic pemphigus. Autoimmun Rev 2018;
microscopy maps of patient skin and mucosa provide insight into 17: 1002–10.
pathogenesis of blistering diseases. J Invest Dermatol 2015; 98 Poot AM, Diercks GF, Kramer D, et al. Laboratory diagnosis of
135: 1763–70. paraneoplastic pemphigus. Br J Dermatol 2013; 169: 1016–24.
75 Jennings JM, Tucker DK, Kottke MD, et al. Desmosome 99 Zimmermann J, Bahmer F, Rose C, Zillikens D, Schmidt E.
disassembly in response to pemphigus vulgaris IgG occurs in Clinical and immunopathological spectrum of paraneoplastic
distinct phases and can be reversed by expression of exogenous pemphigus. J Dtsch Dermatol Ges 2010; 8: 598–605.
Dsg3. J Invest Dermatol 2011; 131: 706–18. 100 Amagai M, Nishikawa T, Nousari HC, Anhalt GJ, Hashimoto T.
76 Lotti R, Shu E, Petrachi T, et al. Soluble Fas ligand is essential for Antibodies against desmoglein 3 (pemphigus vulgaris antigen) are
blister formation in pemphigus. Front Immunol 2018; 9: 370. present in sera from patients with paraneoplastic pemphigus and
77 Chen J, Den Z, Koch PJ. Loss of desmocollin 3 in mice leads to cause acantholysis in vivo in neonatal mice. J Clin Invest 1998;
epidermal blistering. J Cell Sci 2008; 121: 2844–49. 102: 775–82.
78 Rafei D, Muller R, Ishii N, et al. IgG autoantibodies against 101 Probst C, Schlumberger W, Stocker W, et al. Development of ELISA
desmocollin 3 in pemphigus sera induce loss of keratinocyte for the specific determination of autoantibodies against envoplakin
adhesion. Am J Pathol 2011; 178: 718–23. and periplakin in paraneoplastic pemphigus. Clin Chim Acta 2009;
79 Spindler V, Heupel WM, Efthymiadis A, et al. Desmocollin 410: 13–18.
3-mediated binding is crucial for keratinocyte cohesion and is 102 Joly P, Thomine E, Gilbert D, et al. Overlapping distribution of
impaired in pemphigus. J Biol Chem 2009; 284: 30556–64. autoantibody specificities in paraneoplastic pemphigus and
80 Lakshmi MJD, Jaisankar TJ, Rajappa M, et al. Correlation of pemphigus vulgaris. J Invest Dermatol 1994; 103: 65–72.
antimuscarinic acetylcholine receptor antibody titers and 103 Takahashi H, Kouno M, Nagao K, et al. Desmoglein 3-specific CD4+
antidesmoglein antibody titers with the severity of disease in T cells induce pemphigus vulgaris and interface dermatitis in mice.
patients with pemphigus. J Am Acad Dermatol 2017; 76: 895–902. J Clin Invest 2011; 121: 3677–88.
81 Chen Y, Chernyavsky A, Webber RJ, Grando SA, Wang PH. 104 Nousari HC, Deterding R, Wojtczack H, et al. The mechanism of
Critical role of the neonatal Fc receptor (FcRn) in the pathogenic respiratory failure in paraneoplastic pemphigus. N Engl J Med 1999;
action of antimitochondrial autoantibodies synergizing with 340: 1406–10.
anti-desmoglein autoantibodies in pemphigus vulgaris. J Biol Chem 105 Zhao CY, Chiang YZ, Murrell DF. Neonatal autoimmune
2015; 290: 23826–37. blistering disease: a systematic review. Pediatr Dermatol 2016;
82 Mahmoud A, Miziara ID, Costa KC, Santi CG, Maruta CW, Aoki V. 33: 367–74.
Laryngeal involvement in pemphigus vulgaris: a proposed 106 Kardos M, Levine D, Gurcan HM, Ahmed RA. Pemphigus vulgaris
classification. J Laryngol Otol 2012; 126: 1041–44. in pregnancy: analysis of current data on the management and
83 De D, Kumar S, Handa S, Mahajan R. Fingernail involvement in outcomes. Obstet Gynecol Surv 2009; 64: 739–49.
pemphigus and its correlation with disease severity and other 107 Murrell DF, Dick S, Ahmed AR, et al. Consensus statement on
clinicodemographic parameters. Br J Dermatol 2018; 180: 662–63. definitions of disease, end points, and therapeutic response for
84 Zaraa I, Sellami A, Bouguerra C, et al. Pemphigus vegetans: pemphigus. J Am Acad Dermatol 2008; 58: 1043–46.
a clinical, histological, immunopathological and prognostic study. 108 Boulard C, Duvert Lehembre S, Picard-Dahan C, et al.
J Eur Acad Dermatol Venereol 25: 1160–67. Calculation of cut-off values based on the Autoimmune Bullous
85 Oktarina DA, Poot AM, Kramer D, Diercks GF, Jonkman MF, Skin Disorder Intensity Score (ABSIS) and Pemphigus Disease
Pas HH. The IgG “lupus-band” deposition pattern of pemphigus Area Index (PDAI) pemphigus scoring systems for defining
erythematosus: association with the desmoglein 1 ectodomain as moderate, significant and extensive types of pemphigus.
revealed by 3 cases. Arch Dermatol 2012; 148: 1173–78. Br J Dermatol 2016; 175: 142–49.
86 Ohzono A, Sogame R, Li X, et al. Clinical and immunological 109 Hebert V, Boulard C, Houivet E, et al. Large international validation
findings in 104 cases of paraneoplastic pemphigus. Br J Dermatol of ABSIS and PDAI pemphigus severity scores. J Invest Dermatol
2015; 173: 1447–52. 2019; 139: 31–37.
87 Joly P, Richard C, Gilbert D, et al. Sensitivity and specificity of 110 Rencz F, Gulacsi L, Tamasi B, et al. Health-related quality of life
clinical, histologic, and immunologic features in the diagnosis and its determinants in pemphigus: a systematic review and
of paraneoplastic pemphigus. J Am Acad Dermatol 2000; meta-analysis. Br J Dermatol 2015; 173: 1076–80.
43: 619–26. 111 Sebaratnam DF, Hanna AM, Chee SN, et al. Development of a
88 Leger S, Picard D, Ingen-Housz-Oro S, et al. Prognostic factors of quality-of-life instrument for autoimmune bullous disease:
paraneoplastic pemphigus. Arch Dermatol 2012: 148: 1165–72. the autoimmune bullous disease quality of life questionnaire.
89 Anhalt GJ, Kim SC, Stanley JR, et al. Paraneoplastic pemphigus. JAMA Dermatol 2013; 149: 1186–91.
An autoimmune mucocutaneous disease associated with neoplasia. 112 Harman KE, Brown D, Exton LS, et al. British Association of
N Engl J Med 1990; 323: 1729–35. Dermatologists’ guidelines for the management of pemphigus
90 Wang L, Bu D, Yang Y, Chen X, Zhu X. Castleman’s tumours and vulgaris 2017. Br J Dermatol 2017; 177: 1170–201.
production of autoantibody in paraneoplastic pemphigus. Lancet 113 Hertl M, Jedlickova H, Karpati S, et al. Pemphigus. S2 Guideline for
2004; 363: 525–31. diagnosis and treatment – guided by the European Dermatology
91 Liu Q, Bu DF, Li D, Zhu XJ. Genotyping of HLA-I and HLA-II Forum (EDF) in cooperation with the European Academy of
alleles in Chinese patients with paraneoplastic pemphigus. Dermatology and Venereology (EADV). J Eur Acad Dermatol Venereol
Br J Dermatol 2008; 158: 587–91. 2015; 29: 405–14.
92 Wang R, Li J, Wang M, et al. Prevalence of myasthenia gravis and 114 Schmidt E, Goebeler M, Hertl M, et al. S2k guideline for the
associated autoantibodies in paraneoplastic pemphigus and their diagnosis of pemphigus vulgaris/foliaceus and bullous pemphigoid.
correlations with symptoms and prognosis. Br J Dermatol 2015; J Dtsch Dermatol Ges 2015; 13: 713–27.
172: 968–75. 115 van Beek N, Rentzsch K, Probst C, et al. Serological diagnosis of
93 Mimouni D, Anhalt GJ, Lazarova Z, et al. Paraneoplastic pemphigus autoimmune bullous skin diseases: prospective comparison of the
in children and adolescents. Br J Dermatol 2002; 147: 725–32. BIOCHIP mosaic-based indirect immunofluorescence technique
with the conventional multi-step single test strategy.
94 Nikolskaia OV, Nousari CH, Anhalt GJ. Paraneoplastic pemphigus in
Orphanet J Rare Dis 2012; 7: 49.
association with Castleman’s disease. Br J Dermatol 2003; 149: 1143–51.
116 Murrell DF, Pena S, Joly P, et al. Diagnosis and management of
95 Hata T, Nishimoto S, Nagao K, et al. Ectopic expression of
pemphigus: recommendations by an international panel of experts.
epidermal antigens renders the lung a target organ in
J Am Acad Dermatol 2018; published online Feb 10. DOI:10.1016/j.
paraneoplastic pemphigus. J Immunol 2013; 191: 83–90.
jaad.2018.02.021.
96 Tsuchisaka A, Numata S, Teye K, et al. Epiplakin is a paraneoplastic
117 Amagai M, Tanikawa A, Shimizu T, et al. Japanese guidelines for
pemphigus autoantigen and related to bronchiolitis obliterans in
the management of pemphigus. J Dermatol 2014; 41: 471–86.
Japanese patients. J Invest Dermatol 2016; 136: 399–408.

www.thelancet.com Vol 394 September 7, 2019 893

 Seminar

118 Almugairen N, Hospital V, Bedane C, et al. Assessment of the rate 134 Kasperkiewicz M, Shimanovich I, Meier M, et al. Treatment of
of long-term complete remission off therapy in patients with severe pemphigus with a combination of immunoadsorption,
pemphigus treated with different regimens including medium- and rituximab, pulsed dexamethasone and azathioprine/mycophenolate
high-dose corticosteroids. J Am Acad Dermatol 2013; 69: 583–88. mofetil: a pilot study of 23 patients. Br J Dermatol 2012; 166: 154–60.
119 Herbst A, Bystryn JC. Patterns of remission in pemphigus vulgaris. 135 Kushner CJ, Concha JSS, Werth VP. Treatment of autoimmune
J Am Acad Dermatol 2000; 42: 422–27.
 bullous disorders in pregnancy. Am J Clin Dermatol 2018;
120 Joly P, Maho-Vaillant M, Prost-Squarcioni C, et al. First-line rituximab 19: 391–403.
combined with short-term prednisone versus prednisone alone for 136 Bystryn JC, Steinman NM. The adjuvant therapy of pemphigus.
the treatment of pemphigus (Ritux 3): a prospective, multicentre, An update. Arch Dermatol 1996; 132: 203–12.
parallel-group, open-label randomised trial. Lancet 2017; 389: 2031–40. 137 Huang YH, Kuo CF, Chen YH, Yang YW. Incidence, mortality,
121 Schmidt E, Goebeler M, Zillikens D. Rituximab in severe and causes of death of patients with pemphigus in Taiwan:
pemphigus. Ann NY Acad Sci 2009; 1173: 683–91. a nationwide population-based study. J Invest Dermatol 2012;
122 Ahmed AR, Spigelman Z, Cavacini LA, Posner MR. Treatment of 132: 92–97.
pemphigus vulgaris with rituximab and intravenous immune 138 Hsu DY, Brieva J, Sinha AA, Langan S, Silverberg JI. Comorbidities
globulin. N Engl J Med 2006; 355: 1772–79. and inpatient mortality for pemphigus in the United States.
123 Colliou N, Picard D, Caillot F, et al. Long-term remissions of severe Br J Dermatol 2016; 174: 1290–98.
pemphigus after rituximab therapy are associated with prolonged 139 Chams-Davatchi C, Valikhani M, Daneshpazhooh M, et al.
failure of desmoglein B cell response. Sci Transl Med 2013; 5: 175ra30. Pemphigus: analysis of 1209 cases. Int J Dermatol 2005; 44: 470–76.
124 Joly P, Mouquet H, Roujeau JC, et al. A single cycle of rituximab for 140 Kumar S, De D, Handa S, et al. Identification of factors associated
the treatment of severe pemphigus. N Engl J Med 2007; 357: 545–52. with treatment refractoriness of oral lesions in pemphigus vulgaris.
125 Ahmed AR, Shetty S. A comprehensive analysis of treatment Br J Dermatol 2017; 177: 1583–89.
outcomes in patients with pemphigus vulgaris treated with 141 Wormser D, Chen DM, Brunetta PG, et al. Cumulative oral
rituximab. Autoimmun Rev 2015; 14: 323–31. corticosteroid use increases risk of glucocorticoid-related adverse
126 Wang HH, Liu CW, Li YC, Huang YC. Efficacy of rituximab for events in patients with newly diagnosed pemphigus.
pemphigus: a systematic review and meta-analysis of different J Am Acad Dermatol 2017; 77: 379–81.
regimens. Acta Derm Venereol 2015; 95: 928–32. 142 Leshem YA, Atzmony L, Dudkiewicz I, Hodak E, Mimouni D.
127 Atzmony L, Hodak E, Leshem YA, et al. The role of adjuvant therapy Venous thromboembolism in patients with pemphigus: a cohort
in pemphigus: a systematic review and meta-analysis. study. J Am Acad Dermatol 2017; 77: 256–60.
J Am Acad Dermatol 2015; 73: 264–71. 143 Saha M, Bhogal B, Black MM, Cooper D, Vaughan RW, Groves RW.
128 Martin LK, Werth V, Villanueva E, Segall J, Murrell DF. Prognostic factors in pemphigus vulgaris and pemphigus foliaceus.
Interventions for pemphigus vulgaris and pemphigus foliaceus. Br J Dermatol 2014; 170: 116–22.
Cochrane Database Syst Rev 2009; 1: CD006263. 144 Daneshpazhooh M, Zafarmand Sedigh V, Balighi K, et al.
129 Czernik A, Toosi S, Bystryn JC, Grando SA. Intravenous Immunologic prediction of relapse in patients with pemphigus
immunoglobulin in the treatment of autoimmune bullous vulgaris (PV) in clinical remission. J Am Acad Dermatol 2016;
dermatoses: an update. Autoimmunity 2012; 45: 111–18. 74: 1160–65.
130 Amagai M, Ikeda S, Shimizu H, et al. A randomized double-blind 145 Hofrichter M, Dworschak J, Emtenani S, et al. Immunoadsorption
trial of intravenous immunoglobulin for pemphigus. of desmoglein-3-specific IgG abolishes the blister-inducing capacity
J Am Acad Dermatol 2009; 60: 595–603. of pemphigus vulgaris IgG in neonatal mice. Front Immunol 2018;
131 Amber KT, Maglie R, Solimani F, Eming R, Hertl M. 9: 1935.
Targeted therapies for autoimmune bullous diseases: current status. 146 Ellebrecht CT, Bhoj VG, Nace A, et al. Reengineering chimeric
Drugs 2018; 78: 1527–48. antigen receptor T cells for targeted therapy of autoimmune
132 Ahmed AR, Kaveri S, Spigelman Z. Long-term remissions in disease. Science 2016; 353: 179–84.
recalcitrant pemphigus vulgaris. N Engl J Med 2015; 373: 2693–94.
133 Behzad M, Mobs C, Kneisel A, et al. Combined treatment with © 2019 Elsevier Ltd. All rights reserved.
immunoadsorption and rituximab leads to fast and prolonged
clinical remission in difficult-to-treat pemphigus vulgaris.
Br J Dermatol 2012; 166: 844–52.

894 www.thelancet.com Vol 394 September 7, 2019