Paraneoplastic Pemphigus Associated With Castleman Tumor

OBSERVATION
Paraneoplastic Pemphigus Associated
With Castleman Tumor
A Commonly Reported Subtype of Paraneoplastic Pemphigus in China
Jing Wang, MD; Xuejun Zhu, MD; Ruoyu Li, MD; Ping Tu, MD; Rengui Wang, MD; Lanbo Zhang, MD; Ting Li, MD;
Xixue Chen, MD; Aiping Wang, MD; Shuxia Yang, MD; Yan Wu, MD; Haizhen Yang, MD; Suzhen Ji, MD
Background: Castleman tumor, a rare lymphoproliferative disorder, is one of the associated tumors in paraneoplastic pemphigus. We analyzed the characteristics
of a group of patients with Castleman tumor to clearly
understand and to improve the prognosis of the disease.
compared with our cases. Castleman tumor was a frequently reported neoplasm in association with paraneoplastic pemphigus in China. The disease was found to
be caused by an autoimmune reaction originating from
the B lymphocytes in the Castleman tumor.
Observations: Ten cases of paraneoplastic pemphigus associated with Castleman tumor treated in the Department of Dermatology, Peking University First Hospital, Beijing, China, from May 1, 1999, to March 31, 2004,
were analyzed for clinical aspects, characteristics and histologic features of the tumors, and computed tomographic findings. Literature was reviewed and data were
Conclusions: Castleman tumor in association with paraneoplastic pemphigus is a commonly reported subtype of
paraneoplastic pemphigus in China. Early detection and
removal of the Castleman tumor are crucial for the treatment of this tumor-associated autoimmune disease.
I
Arch Dermatol. 2005;141:1285-1293
N 1990, A NEW AUTOIMMUNE MU-
cocutaneous disease, paraneoplastic pemphigus (PNP), was described by Anhalt et al. 1 The
disease is characterized by distinctive clinical symptoms and signs, such
as severe, painful mucosal erosions, polymorphous skin lesions, histopathologic
hallmarks, and immunologic findings.2,3
The condition typically presents in patients with lymphoproliferative diseases
and, primarily, malignancies.
Castleman tumor is a distinct lymphoproliferative disorder of uncertain origin.4
To better understand PNP associated with
Castleman tumor, and to explore a rational regimen to reduce the high mortality of
this disorder and prevent its accompanying respiratory complications, we analyzed the characteristics of a group of patients in our hospital.
Author Affiliations:
Departments of Dermatology
(Drs J. Wang, Zhu, R. Li, Tu,
Chen, A. Wang, S. Yang, Wu, H.
Yang, and Ji), Medical Images
(Dr R. Wang), Surgery
(Dr Zhang), and Pathology
(Dr T. Li), Peking University
First Hospital, Beijing, China.
METHODS
PATIENTS AND DIAGNOSIS
Peking University First Hospital, one of the largest comprehensive hospitals in Beijing, is both
a public hospital and one that accepts cases for
further diagnosis and treatment from other parts
of China. We diagnosed the first case of PNP
(REPRINTED) ARCH DERMATOL/ VOL 141, OCT 2005
1285
in China in 1999, and 13 cases of PNP were
diagnosed from May 1, 1999, to March 31,
2004. Among them, 10 cases of PNP were associated with Castleman tumors and were analyzed in this study.
The 10 cases were diagnosed by the following criteria1,5-7: (1) clinical features, including
the presence of severe mucosal involvement and
polymorphous cutaneous eruption; (2) histologic features of the skin and mucosal eruption; (3) the presence of Castleman tumor; and
(4) immunologic features. Because direct immunofluorescence for human IgG in the patient’s skin was frequently negative, indirect immunofluorescence with rat bladder epithelium
was used as a specific screening test for the diagnosis of PNP. Immunoblot for the detection
of serum autoantibody against the plakin protein family was used to confirm the diagnosis.
CLINICAL OBSERVATION
AND TREATMENT
Patients’ mucocutaneous lesions, results of routine blood tests, and systemic manifestations
were investigated. A skin or mucosal biopsy
specimen was taken to observe histologic features. If a patient presented with typical mucosal erosion (especially erosive stomatitis),
typical histopathologic manifestations in skin
or mucosa, and immunofluorescence changes,
computed tomography (CT) of the chest, abdomen, and pelvis was carried out to search for
WWW.ARCHDERMATOL.COM
©2005 American Medical Association. All rights reserved.
Table 1. Characteristics of Mucocutaneous Manifestations and Tumors in 10 Patients in the Present Study
Mucosal Lesions
Patient No./
Sex/Age, y
Duration of
Illness, mo
Oral
Ocular
Anogenital
Rash
Location of CD
Type of CD
BO
1/M/39
2/M/17
3/M/30
4/F/25
5/M/29
6/F/17
7/F/22
8/F/48
9/F/41
10/F/24
4
6
5
4
7
4
9
9
8
2
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
−
⫹
⫹
⫹
−
⫹
⫹
⫹
⫹
−
−
⫹
⫹
⫹
−
⫹
⫹
LP
EM
PF
None
PV
LP
SJ
LP
LP
LP
Ret
Ret
Ret
Med
Ret
Med
Ret
Ret
Ret
Ret
HV
HV
Mixed
HV
HV
HV
HV
HV
HV
HV
−
⫹
⫹
⫹
−
⫹
⫹
⫹
⫹
−
Abbreviations: BO, bronchiolitis obliterans; CD, Castleman tumor/disease; EM, erythema multiforme; HV, hyaline vascular; LP, lichen planus; Med, mediastinum;
PF, pemphigus foliaceus; PV, pemphigus vulgaris; Ret, retroperitoneum; SJ, Stevens-Johnson syndrome; ⫹, present; −, absent.
Table 2. Clinical Findings in 18 Patients With Castleman Tumor From Previously Published Articles
Mucosal Lesions
Source
Tagami et al9
Caporale et al10
Redon et al11
Ashinoff et al12
Monpoint et al13
Plewig et al14
Coulson et al15
Fujimoto et al16
Gili et al17
Chin et al18
Hsaio et al19
Kim et al20
Saito et al21
Lemon et al22
Hoffman et al23
Jansen et al24
Lee et al25
Caneppele et al26
Patient Sex/
Age, y
Oral
Ocular
Anogenital
Rash
Location of CD
Type of CD
BO
M/33
F/21
M/21
F/32
F/18
M/45
F/15
M/19
F/21
M/14
M/50
F/19
F/20
M/13
M/52
M/45
M/66
F/49
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
−
⫹
−
⫹
−
⫹
−
⫹
⫹
⫹
−
⫹
⫹
⫹
−
⫹
⫹
⫹
⫹
⫹
⫹
⫹
−
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
⫹
−
⫹
⫹
−
⫹ (PV)
⫹ (LP)
⫹ (LP)
⫹ (PV)
⫹ (LP)
⫹ (LP)
−
−
−
⫹ (LP)
−
−
⫹ (SJ)
−
⫹ (LP)
⫹ (LP, SJ)
⫹ (LP)
Ret
Med
Med
Ret
Ret
NA
Ret
Ret
Med
Med
Ret
Ret
NA
Med
Ret
Ret
Ret
Ret
NA
PC
HV
HV
PC
NA
Mixed
NA
HV
HV
HV
HV
HV
HV
HV
Mixed
HV
NA
−
⫹
−
NA
NA
NA
−
⫹
⫹
⫹
NA
⫹
⫹
−
⫹
−
−
−
Abbreviations: BO, bronchiolitis obliterans; CD, Castleman tumor/disease; HV, hyaline vascular; LP, lichen planus; Med, mediastinum; NA, not available;
PC, plasma cell; PV, pemphigus vulgaris; Ret, retroperitoneum; SJ, Stevens-Johnson syndrome; ⫹, present; −, absent.
the underlying tumor. The tumors found were totally surgically removed. Histopathologic characteristics of the resected
tumors were observed.
For the latest-diagnosed 6 patients who underwent surgery,
10 to 20 g of immunoglobulin (Ronsen Intravenous Immunoglobulin; Chengdu Rongsheng Pharmaceuticals, Chengdu, Sichuan,
China) was given intravenously 1 hour before and during the operation. Among these 6 patients, preoperative treatment with 10
g of immunoglobulin daily for 3 days was given to 1 patent with
severe lesions and respiratory symptoms. In 2 patients with bronchiolitis obliterans, 10 g of immunoglobulin every day or every
other day for 3 to 6 days was given as a postoperative treatment.
IMMUNOLOGIC FEATURES
For indirect immunofluorescence, frozen sections of rat urinary bladder were used as the substrates, and the patients’ serum was used as the primary antibodies. Fluorescein isothiocyanate–conjugated anti–human IgG was the secondary antibody.
After resection of the tumors, we also investigated the changes
in autoantibody titers with indirect immunofluorescence.
(REPRINTED) ARCH DERMATOL/ VOL 141, OCT 2005
1286
For immunoblotting, protein extraction from human keratinocytes was separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred onto nitrocellulose membranes. The membrane stripes were blotted by patients’
serum as the primary antibodies.
An enzyme-linked immunosorbent assay test kit (MESACUP Dsg3 Test Kit; Medical and Biological Laboratories Co Ltd,
Nagoya, Japan) was used to assay anti–desmoglein 3 IgG in patients’ serum. Enzyme-linked immunosorbent assay plates coated
with recombinant desmoglein 3 were incubated for 1 hour with
patients’ serum diluted to 1:100 with the buffer. The secondary antibody was peroxidase-conjugated mouse anti–human IgG
antibody. Developed color was measured at 450 nm.
CULTURE OF TUMOR CELLS
AND DETECTION OF AUTOANTIBODY
FROM THE CULTURE MEDIUM
Tumor cells were isolated after resection of the tumors and cultured as described previously.8 After the isolated tumor cells
were washed 3 times with culture medium and cultured for more
WWW.ARCHDERMATOL.COM
©2005 American Medical Association. All rights reserved.
A
B
C
D
Figure 1. Mucosal manifestations of the patients with paraneoplastic pemphigus. A, Erosions and shallow ulcers in the oral cavity and on the tongue of patient 9.
B, Violet lesion with hemorrhagic crusting on the lips of patient 1. C, Erosions on the penis of patient 2. D, Lesions of the conjunctivae resembling
Stevens-Johnson syndrome in the periorbital area of patient 1.
than 3 passages, the cell culture supernatant was collected, concentrated, and purified by means of centrifugal filters (Centricon Plus-50; Millipore Corp, Bedford, Mass). Positive serum
from patients with PNP, uncultured medium, and supernatant from a culture of Namalwa cells, a cell line from human
B-cell lymphoma, were used as positive and negative controls
in the following tests.
Complementary DNA fragments coding linker regions of envoplakin (located at amino acids 1679-1812, GenBank U53786),
periplakin (located at amino acids 1615-1755, GenBank
NM002705), bullous pemphigoid antigen 1 (located at amino
acids 1762-1891, GenBank M63618), and desmoplakin 1 (located at amino acids 1342-1482, GenBank J05211) were cloned
and expressed as glutathione S-transferase fusion proteins. Glutathione S-transferase protein was used as the negative control,
and the expressed fusion proteins were purified, separated, and
transferred onto nitrocellulose membranes. We blotted the membrane strips with the concentrated cell culture supernatants.
RESULTS
The patients’ clinical data are shown in Table 1, and
data from 18 patients in previously published reports are
listed in Table 2.9-26 The clinical manifestations and the
type and location of the associated tumors were similar
in the 2 patient groups.
(REPRINTED) ARCH DERMATOL/ VOL 141, OCT 2005
1287
In 7 of the 10 cases, the white blood cell counts were
high (12.5-16.7⫻103/µL) because of lung infection and
inflammation in mucocutaneous lesions. Erythrocyte
sedimentation rate was elevated in 7 cases and ranged
from 25 to 77 mm in the first hour. The serum IgG level
was elevated (1800=2140 mg/dL) in 3 cases. Three cases
were found to have a positive antinuclear antibody in
low titers.
CHARACTERISTICS OF MUCOCUTANEOUS
AND RESPIRATORY MANIFESTATIONS
Mucosal Involvement
Oral lesions occurred in all 10 cases as the earliest
complaint and presented as widespread blisters, erosions, ulcerations, and painful stomatitis affecting the
entire oral cavity, tongue, and lips. Conjunctival erosions occurred in 8 of the 10 patients, and anogenital
involvement occurred in 7 patients. Mucosal lesions
may be the only sign, as they were in patient 4. The
lesions in the mucosa were blisters, erosions, or ulcers,
as well as erosive lichen planus–like lesions (Table 1,
Figure 1).
WWW.ARCHDERMATOL.COM
©2005 American Medical Association. All rights reserved.
A
B
C
D
Figure 2. Cutaneous manifestations of patients with paraneoplastic pemphigus. A, Lichen planus–like papules on the trunk of patient 8. B, Erythema
multiforme–like lesions on the hands of patient 1. C, Severe erosions and crusts on the palm and fingers of patient 7. D, Blisters, erosions, and crusts of the skin
of patient 3.
Skin Lesions
Symptoms of Respiratory System
Cutaneous lesions usually appeared after the onset of
oral lesions and were polymorphic. Nine patients had
rashes, including pemphigus-like in 2 cases, lichen
planus–like in 5, and erythema multiforme–like in 1.
Seven cases showed characteristic purple erythema on
the fingers, palms, or soles (Figure 2).
Respiratory tract involvement or bronchiolitis obliterans was found in 7 of the 10 cases (Figure 4). Dry cough
and dyspnea were the common symptoms. Four
patients developed bronchiolitis obliterans several days
after resection of Castleman tumors. Among them, 2 patients showed severe dyspnea with decreased PO2 (86.498.7 mm Hg). Patient 4 underwent a lung biopsy, which
demonstrated dense lymphohistiocytic infiltration as well
as fibrosis around bronchioles (Figure 4).
Histopathologic Features of the Skin
and Mucosal Lesions
Biopsy specimens were taken from cutaneous lesions
in 9 cases and from oral mucosa in 5. Ten specimens
showed acantholysis or blisters in the epidermis. Nine
of them demonstrated suprabasilar acantholysis, and 1
showed acantholysis just below the granular layer. All
of the specimens showed vacuolar interface changes
with lymphohistiocytic infiltration in superficial dermis. In some cases extravasated red blood cells were
found in papillary dermis. Necrotic keratinocytes,
being a characteristic feature of the disease, were scattered in the epidermis of all of the examined specimens (Figure 3).
(REPRINTED) ARCH DERMATOL/ VOL 141, OCT 2005
1288
IMMUNOLOGIC FEATURES
In all of the 9 tested cases, indirect immunofluorescence disclosed a high titer of circulating IgG with intercellular staining of rat bladder epithelium (Figure 5).
The titer ranged from 1:160 to 1:640 at the time of diagnosis. After resection of the Castleman tumors, the autoantibody titers decreased in most patients and became
undetectable in 4 cases within 5 to 9 weeks.
On immunoblotting, IgG autoantibodies in serum from
9 patients recognized human keratinocyte proteins of 210
and 190 kDa, which were usually identified as the memWWW.ARCHDERMATOL.COM
©2005 American Medical Association. All rights reserved.
A
A
B
B
Figure 3. Histopathologic characteristics of the skin in paraneoplastic
pemphigus in patient 7. A, The white arrow indicates suprabasilar cleft in the
epidermis; black arrow, lymphohistiocytic infiltrates in superficial dermis
(hematoxylin-eosin, original magnification ⫻200). B, Histopathologic
changes from the same patient. The white arrow indicates necrotic
keratinocytes or dyskeratosis scattered in the epidermis; black arrow,
extravasated red blood cells in papillary dermis and vacuolar interface
changes in the dermoepidermal junction (hematoxylin-eosin, original
magnification ⫻400).
bers of the characterized antigens in PNP (Figure 5). Serum from some patients also recognized a 250-kDa protein band. However, positive bands of desmoplakin,
desmoglein 3, and bullous pemphigoid antigen 1 were
not detected by immunoblotting. This was probably due
to the ineffective extraction of these antigens from the
keratinocytes. Immunoprecipitation using radiolabeled
keratinocyte extract may be a more sensitive method for
the detection of antibodies against proteins of the plakin
family.5
Of the 9 cases in which an enzyme-linked immunosorbent assay was performed, 6 had positive results, and
in 2 cases results were weakly positive. One case had negative results.
LOCATION AND HISTOPATHOLOGIC FEATURES
OF THE CASTLEMAN TUMORS
All patients remained asymptomatic and were diagnosed as having Castleman tumors after CT scanning. On
CT scanning, 8 cases had tumors in the retroperitoneum and 2 in the mediastinum (Table 1). The diameters ranged from 3.0 to 12.0 cm. The tumors were usually globular or column-shaped, solitary masses with
irregular margins. The average CT value of the masses
(REPRINTED) ARCH DERMATOL/ VOL 141, OCT 2005
1289
Figure 4. Characteristics of bronchiolitis obliterans. A, Computed
tomographic scan of patient 4 shows increased lung markings and
bronchiectasis. B, Histologic section of a lung biopsy specimen from the
same patient demonstrates dense lymphohistiocytic infiltration as well as
fibrosing around the bronchiole (hematoxylin-eosin, original magnification
⫻200).
ranged from 36 to 50 Hounsfield units, and they were
usually isodense without necrosis or liquefaction. However, blotchy, striped, or branch-shaped calcifications
could be seen in the center of the mass in 5 cases in this
group; as calcification was found only in localized Castleman tumor, it was believed to be an important sign for
the diagnosis of this tumor. All of the tumors were enhanced markedly and homogeneously after intravenous
administration of contrast medium. The CT value increased to as high as 126 to 190 Hounsfield units, similar to that of great vessels in this circumstance.
The removed tumors were solid and measured
2⫻ 3⫻ 3 cm to 12⫻ 6 ⫻6 cm. Macroscopically, the tumors were solitary masses of various sizes with relatively intact and smooth capsules. The cut surface showed
gray to yellow color and fleshy appearance. Histopathologically, the normal structure of the lymph node was effaced and replaced by numerous small follicular centers
with prominent central vessels, which exhibited hyalinized walls and prolific endothelial cells. The small follicular centers were surrounded by concentric layers of
follicle center cells. The interfollicular areas showed vascular proliferation and variable numbers of plasma cells.
Of the 10 cases, 9 tumors were hyaline vascular variants
and 1 was a mixed variant27 (Figure 6).
Immunohistochemical staining showed that germinal centers in lymphoid follicles were dominated by
WWW.ARCHDERMATOL.COM
©2005 American Medical Association. All rights reserved.
A
1
2
3
4
5
6
N
A
kDa
250
210
190
170
B
B
Figure 5. Immunologic findings of paraneoplastic pemphigus. A, Western
blot using human keratinocyte protein extract shows that the patients’ serum
recognized multiple antigens of about 250, 210, and 190 kDa. B, Indirect
immunofluorescence using rat bladder shows human IgG staining in the
intercellular spaces of the epithelia (original magnification ⫻400).
CD20⫹ B lymphocytes and CD34⫹ blood vessels located
in the center of the follicles. The blood vessels were surrounded by concentric cuffs of B lymphocytes arranged
in an onion-skin pattern, while the interfollicular areas
were scattered with a few CD45Ro⫹ T lymphocytes
(Figure 7).
COMMENT
FREQUENCY OF CASTLEMAN TUMOR
According to Anhalt et al,1 the most commonly associated neoplasm in PNP is non-Hodgkin lymphoma (42%).
Others are chronic lymphocytic leukemia (29%), Castleman tumor (10%), thymoma (6%), retroperitoneal sarcomas (6%), and Waldenström macroglobulinemia
(6%).1,3,5,6,28 However, in this series, 10 (77%) of 13 cases
of PNP were found to be associated with Castleman tumor. In China, 14 cases of PNP were reported previously29-39; 7 had a confirmed diagnosis of Castleman tumors, and 3 probably had Castleman tumors according
to the manifestations and CT scans. Only 2 neoplasms
were diagnosed as non-Hodgkin lymphoma, and 2 were
other neoplasms. Therefore, 71% of the reported tumors associated with PNP were probably Castleman tumors. It seems that Castleman tumor is the most frequently encountered neoplasm associated with PNP in
cases reported in China.
(REPRINTED) ARCH DERMATOL/ VOL 141, OCT 2005
1290
Figure 6. Paraneoplastic pemphigus–associated Castleman tumor and its
histologic appearance. A, Resected tumor from patient 3, showing a mass
very similar to a huge lymph node with a fishlike appearance. B, Histologic
section of the resected tumor shows a Castleman tumor of hyaline vascular
type: lymphoid tissue with follicles containing multiple capillaries surrounded
by hyaline sheaths (hematoxylin-eosin, original magnification ⫻100).
In this group, 8 tumors were located in the retroperitoneal region of the abdomen and pelvis, and only 2 were
in the mediastinum (Table 1). The location of the tumors in association with PNP was different from that of
tumors without mucocutaneous and respiratory involvement.40 Presumably, direct stimulation by the antigens
coming from the digestive and urinary-reproductive tracts
was also involved in the pathogenesis of PNP in association with Castleman tumors.
CASTLEMAN TUMOR AND ITS ROLE
IN THE PATHOGENESIS OF PNP
The concentrated cultured media from 4 patients recognized recombinant linker regions of envoplakin,
periplakin, desmoplakin 1, and bullous pemphigoid
antigen 1 in the form of glutathione S-transferase fusion
proteins, but did not recognize the 26-kDa glutathione
S-transferase protein. Precultured medium and cultured medium from Namalwa cells showed no bands
(Figure 8).
Castleman tumor represents lymphatic tissue proliferation of uncertain origin or is due to abnormal responses to various stimuli.4,41 Ashinoff et al12 postulated
that possible expression of foreign tumor antigens that
cross-react with epidermal antigens induce the autoreactive clones of T lymphocytes. Anhalt6 speculated that
WWW.ARCHDERMATOL.COM
©2005 American Medical Association. All rights reserved.
A
43 kDa
A
B
43 kDa
B
C
43 kDa
C
D
43 kDa
Figure 7. Immunohistochemical staining of a paraneoplastic
pemphigus–associated Castleman tumor. A, CD20ⴙ B lymphocytes dominate
in a germinal center (arrow) (original magnification ⫻100). B, A few
CD45Roⴙ T lymphocytes scattered in interfollicular areas (arrow) (original
magnification ⫻100). C, Prominent CD34ⴙ blood vessels located in
lymphoid follicles (arrow) (original magnification ⫻100).
tumors associated with PNP may produce plakin proteins that result in initiation of autoimmune responses.
Other investigators believed that the autoimmune reaction was related to the epitope spreading.42,43
Our intensive clinical observation and laboratory investigation found that most of the patients’ lesions improved and that the titers of the circulation autoantibody decreased after resection of the tumors. We studied
the role of B cells in the Castleman tumors in 7 cases.
Results showed that clones of B lymphocytes in the Castleman tumor carrying similar specific rearranged immunoglobulin heavy-chain genes may have the structural
basis for producing autoantibody in some of the cases.8
In one case, the cells isolated from the removed tumor
were cultured, the cell culture medium was collected, and
indirect immunofluorescence and immunoblotting were
performed. Immunoglobulin G was found in the cell culture medium and can recognize the specific antigens in
epithelia.8,44,45 Recently, rearranged immunoglobulin
heavy- and light-chain variable region genes were cloned
and their sequences were analyzed in Castleman tumors associated with PNP. A high incidence of somatic
mutations in complementarity-determining regions and
framework regions was observed in the cloned variable
region of heavy-chain and light-chain genes. These clones
were found to have experienced switch recombination
and have the structural basis to produce autoantibod(REPRINTED) ARCH DERMATOL/ VOL 141, OCT 2005
1291
Figure 8. Western blot shows patients’ serum and cell-cultured medium
recognizing recombinant fusion proteins of linker regions in periplakin (41.7
kDa) (A), desmoplakin 1 (41.7 kDa) (B), bullous pemphigoid antigen 1 (40.4
kDa) (C), and envoplakin (40.9 kDa) (D). 1s through 4s show cultured
medium from 4 patients; 1 through 4, serum from the 4 patients; Na,
cultured medium from Namalwa cells as a negative control; C, precultured
medium; and M, protein molecular standard.
ies.46,47 In this series of patients, IgG antibodies in the cell
culture media were found to recognize the recombinant
envoplakin, periplakin, and desmoplakin fusion proteins, which are believed to be the specific target antigens of PNP. Therefore, we believe that the autoantibodies are directly produced by the cells in the Castleman
tumor. Because the autoantibodies originated from the
Castleman tumor, the early detection and removal of the
tumor is essential for the treatment of this tumorassociated autoimmune syndrome.
USE OF INTRAVENOUS IMMUNOGLOBULIN
IN PREVENTION OF BRONCHIOLITIS
OBLITERANS
Bronchiolitis obliterans is frequently found in PNP and
may cause respiratory failure and death.2,18,21,23 A previous study21 showed that the autoantibodies could be
deposited in bronchial epithelia and that the affected
bronchioles developed severe luminal stenosis (up to
80%-90%). Therefore, autoantibody-mediated injury takes
an important role in the pathogenesis of bronchiolitis
obliterans, although infections and toxic effects induced by chemotherapy and neoplasm may also cause
pulmonary injury.23,48,49 Nikolskaia et al27 reported that
22 of 28 patients with Castleman tumor in association
with PNP died, most of them of bronchiolitis obliterans.
WWW.ARCHDERMATOL.COM
©2005 American Medical Association. All rights reserved.
Intravenous administration of immunoglobulin has
been widely used in the treatment of autoimmune diseases such as pemphigus, bullous pemphigoid, and
systemic lupus erythematosus. The possible mechanism depends on the interaction of the Fc portion of immunoglobulin with Fc receptors and on the modulation
of the activation and effect functions of B lymphocytes.
Intravenous administration of immunoglobulin also neutralizes pathogenic autoantibody and has a strong antiinflammatory effect depending on its interactions with
the complement system, cytokines, and endothelial
cells.50-52 However, few articles have been published about
the effect of intravenous administration of immunoglobulin in the treatment of PNP.51,53
In our study, the first 4 of the 10 patients did not accept intravenous administration of immunoglobulin, and
3 of them developed severe bronchiolitis obliterans within
1 week after the operation. However, in the other 6 patients, respiratory symptoms were controlled or prevented by giving immunoglobulin intravenously before
and during the operation. The response to the intravenous administration of immunoglobulin usually started
within 3 days. One patient who started immunoglobulin treatment before surgery showed improvement of the
skin lesions and bronchiolitis obliterans within 1 week.
However, intravenous administration of immunoglobulin alone cannot substitute for surgical treatment and cure
the PNP. Recently, a recurrent Castleman tumor was
found in the abdomen of a 32-year-old patient (patient
3 in Table 1) when his skin lesions relapsed 2 years after
the surgery. The cutaneous blisters were under control
again after the second surgical removal of the tumor. Unfortunately, his disease relapsed again. Intravenous administration of immunoglobulin can temporarily control the skin lesions, but the patient’s condition worsened
when he did not have the recurrent tumor removed again.
Therefore, we believe that both early removal of the Castleman tumor and intravenous administration of immunoglobulin are important to prevent the development of severe bronchiolitis obliterans.
SURGICAL REMOVAL OF THE TUMOR
AND PROGNOSIS
All of the observed patients showed poor response to treatments such as corticosteroids, immunosuppressive drugs,
and chemotherapeutic drugs before removal of the tumor. However, after total resection of the tumors, 7 of
the 10 patients responded well to general dermatologic
management. Cutaneous lesions disappeared or improved within 6 to 11 weeks, and mucosal lesions significantly improved within 5 to 10 months. Serum autoantibody levels decreased in 6 to 8 weeks. Gili et al17
reviewed 5 cases of PNP associated with Castleman tumor, and lesions in all cases improved after surgical removal of the tumors. Jansen et al24 also reviewed 12 cases
of similar disease, and 8 of the 12 patients markedly improved after removal of Castleman tumors. Early removal of the Castleman tumor has proved to be the most
important management for the disease. If Castleman tumor is detected, total resection of the tumor is the only
way to achieve complete resolution of the disease. As we
(REPRINTED) ARCH DERMATOL/ VOL 141, OCT 2005
1292
have stated, intravenous administration of immunoglobulin is recommended before and during the operation.
During the operation it is important to block the blood
supply to the tumor as soon as possible and avoid squeezing the tumor. Total resection including the connective
tissue envelope is crucial for a good prognosis. Large tumors tend to recur. Early detection and resection are essential for the treatment of the disease.
Accepted for Publication: April 14, 2005.
Correspondence: Xuejun Zhu, MD, Department of Dermatology, Peking University First Hospital, No. 8, Xishiku St,
Beijing 100034, China ([email protected]).
Author Contributions: Study concept and design: J. Wang
and Zhu. Acquisition of data: Tu, Zhang, Chen, A. Wang,
S. Yang, Wu, H. Yang, and Ji. Analysis and interpretation
of data: R. Li, R. Wang, T. Li, and H. Yang. Drafting of
the manuscript: J. Wang, R. Wang, Zhang, T. Li, and Wu.
Critical revision of the manuscript for important intellectual content: Zhu, R. Li, and Tu. Administrative, technical, and material support: J. Wang, Tu, R. Wang, Zhang,
T. Li, Chen, A. Wang, S. Yang, Wu, H. Yang, and Ji. Study
supervision: Zhu and R. Li.
Financial Disclosure: None.
Funding/Support: This study was supported by grant
30371292 from the National Natural Science Foundation of China, Beijing, and grant 20030001021 from the
Research Foundation for Doctoral Subjects of the Chinese Ministry of Education, Beijing.
REFERENCES
1. Anhalt GJ, Kim SC, Stanley JR, et al. Paraneoplastic pemphigus: an autoimmune mucocutaneous disease associated with neoplasia. N Engl J Med. 1990;
323:1729-1735.
2. Pizarro A, Garcia-Tobaruela A, Pinilla J. Bronchiolitis obliterans, Castleman’s disease, and a bullous disease: pemphigus vulgaris or paraneoplastic pemphigus?
Hum Pathol. 1998;29:657-658.
3. Garcia-Rio F, Alvarez-Sala R, Pino JM. The mechanism of respiratory failure in
paraneoplastic pemphigus [comment]. N Engl J Med. 1999;341:848.
4. Castleman B, Iverson L, Menendez VP. Localized mediastinal lymph node hyperplasia resembling thymoma. Cancer. 1956;9:822-830.
5. Anhalt GJ. Paraneoplastic pemphigus. J Investig Dermatol Symp Proc. 2004;9:
29-33.
6. Anhalt GJ. Paraneoplastic pemphigus. Adv Dermatol. 1997;12:77-96.
7. Helou J, Allbritton J, Anhalt GJ. Accuracy of indirect immunofluorescence testing in the diagnosis of paraneoplastic pemphigus. J Am Acad Dermatol. 1995;
32:441-447.
8. Wang L, Bu D, Yang Y, Chen X, Zhu X. Castleman’s tumours and production of
autoantibody in paraneoplastic pemphigus. Lancet. 2004;363:525-531.
9. Tagami H, Sagawa K, Imamura S, Uchiyama T, Takatsuki K. Severe erosive stomatitis and giant lymph node hyperplasia of retroperitoneum (Castleman’s tumor).
Dermatologica. 1978;157:138-145.
10. Caporale A, Giuliani A, Teneriello F, et al. Mediastinal angiofollicular lymph node
hyperplasia, hyaline-vascular type with mucocutaneous manifestations. Ital J Surg
Sci. 1988;18:389-392.
11. Redon J, Sorni G, Gonzalez-Molina A, Caballero M, Baguena J. Pemphigus associated with giant lymph node hyperplasia. BMJ (Clin Res Ed). 1983;287:
1761-1762.
12. Ashinoff R, Cohen R, Lipkin G. Castleman’s tumor and erosive lichen planus: coincidence or association? report of a case. J Am Acad Dermatol. 1989;21:10761080.
13. Monpoint S, Frappier JM, Petibon E, et al. Pemphigus associated with Castleman’s pseudolymphoma. Dermatologica. 1989;178:54-57.
14. Plewig G, Jansen T, Jungblut RM, Roher HD. Castleman tumor, lichen ruber and
pemphigus vulgaris: paraneoplastic association of immunological diseases? [in
German]. Hautarzt. 1990;41:662-670.
WWW.ARCHDERMATOL.COM
©2005 American Medical Association. All rights reserved.
15. Coulson IH, Cook MG, Bruton J, Penfold C. Atypical pemphigus vulgaris associated with angio-follicular lymph node hyperplasia (Castleman’s disease). Clin
Exp Dermatol. 1986;11:656-663.
16. Fujimoto W, Kanehiro A, Kuwamoto-Hara K, et al. Paraneoplastic pemphigus associated with Castleman’s disease and asymptomatic bronchiolitis obliterans.
Eur J Dermatol. 2002;12:355-359.
17. Gili A, Ngan BY, Lester R. Castleman’s disease associated with pemphigus vulgaris.
J Am Acad Dermatol. 1991;25:955-959.
18. Chin AC, Stich D, White FV, Radhakrishnan J, Holterman MJ. Paraneoplastic pemphigus and bronchiolitis obliterans associated with a mediastinal mass: a rare
case of Castleman’s disease with respiratory failure requiring lung transplantation.
J Pediatr Surg. 2001;36:E22.
19. Hsiao CJ, Hsu MM, Lee JY, Chen WC, Hsieh WC. Paraneoplastic pemphigus in
association with a retroperitoneal Castleman’s disease presenting with a lichen
planus pemphigoides-like eruption: a case report and review of literature. Br J
Dermatol. 2001;144:372-376.
20. Kim SC, Chang SN, Lee IJ, et al. Localized mucosal involvement and severe pulmonary involvement in a young patient with paraneoplastic pemphigus associated with Castleman’s tumour. Br J Dermatol. 1998;138:
667-671.
21. Saito K, Morita M, Enomoto K. Bronchiolitis obliterans with pemphigus vulgaris
and Castleman’s disease of hyaline-vascular type: an autopsy case analyzed by
computer-aided 3-D reconstruction of the airway lesions. Hum Pathol. 1997;
28:1310-1312.
22. Lemon M, Weston W, Huff J. Childhood paraneoplastic pemphigus associated
with Castleman’s tumour. Br J Dermatol. 1997;136:115-117.
23. Hoffman MA, Qiao X, Anhalt GJ. CD8⫹ T lymphocytes in bronchiolitis obliterans, paraneoplastic pemphigus, and solitary Castleman’s disease. N Engl J Med.
2003;349:407-408.
24. Jansen T, Plewig G, Anhalt GJ. Paraneoplastic pemphigus with clinical features
of erosive lichen planus associated with Castleman’s tumor. Dermatology. 1995;
190:245-250.
25. Lee IJ, Kim SC, Kim HS, et al. Paraneoplastic pemphigus associated with follicular dendritic cell sarcoma arising from Castleman’s tumor. J Am Acad Dermatol.
1999;40:294-297.
26. Caneppele S, Picart N, Bayle-Lebey P, et al. Paraneoplastic pemphigus associated with Castleman’s tumour. Clin Exp Dermatol. 2000;25:219-221.
27. Frizzera G. Castleman’s disease and related disorders. Semin Diagn Pathol. 1988;
5:346-364.
28. Nikolskaia OV, Nousari CH, Anhalt GJ. Paraneoplastic pemphigus in association
with Castleman’s disease. Br J Dermatol. 2003;149:1143-1151.
29. Meng H, Qingyun K, Long C, et al. Paraneoplastic pemphigus: 1 case report. Chin
J Clin Dermatol. 2002;31:458-460.
30. Qianxi X, Xiuhua H, Jianzhong Z. Paraneoplastic pemphigus: 1 case report. Chin
J Clin Dermatol. 2003;32:213-214.
31. Xiaoying Y, Qinghua S, Chuan M, et al. Paraneoplastic pemphigus: 1 case report.
Chin J Clin Dermatol. 2002;31:710-711.
32. Xiangdong W, Hong S. Paraneoplastic pemphigus: 1 case report. Chin J Clin
Dermatol. 2002;31:790-791.
33. Yukun W, Keyu W, Shulan G, et al. Paraneoplastic pemphigus: 1 case report.
Chin J Clin Dermatol. 2003;32:28-29.
34. Lu Z, Wei H, Hai W, et al. Paraneoplastic pemphigus: a case in association with
dermatofibrosarcoma protuberans. Chin J Lepr Dermatol. 2003;19:101-103.
35. Yan J, Bo D, Mengwu W. Paraneoplastic pemphigus: 2 cases. Chin J Dermatol
Venereol. 2002;16:267-268.
36. Siping Z, Zhenglong Z, Guangying W, et al. Paraneoplastic pemphigus: 1 case
report. Chin J Clin Dermatol. 2005;34:110-111.
37. Huaiyou B, Huanjiang L, Zhiwen L. Clinical characteristics and CT diagnosis of
local Castleman disease in association with paraneoplastic pemphigus (5 case
reports). J Guangxi Med Univ. 2004;21:610-611.
38. Zhimin Y, Hong H. Paraneoplastic pemphigus. Beijing Oral Med J. 2003;11:230-234.
39. Li X, Xuejun C, Xuejun Z. Paraneoplastic pemphigus: 2 case reports and primary
study on the relation to HHV8. Chin J Dermatol Venereol. 2002;16:24-26.
40. Palestro G, Turrini F, Pagano M, Chiusa L. Castleman’s disease. Adv Clin Path.
1999;3:11-22.
41. Resegotti L, Rua S, Dolci C, Grosso B, Pistone M, Testa D. Polyclonal lymphadenopathy presenting as plasma cell leukemia with reversible renal insufficiency.
Acta Haematol. 1983;70:54-58.
42. Bowen GM, Peters NT, Fivenson DP, et al. Lichenoid dermatitis in paraneoplastic pemphigus: a pathogenic trigger of epitope spreading? Arch Dermatol. 2000;
136:652-656.
43. Chan LS. Epitope spreading in paraneoplastic pemphigus: autoimmune induction in antibody-mediated blistering skin diseases. Arch Dermatol. 2000;136:
663-664.
44. Zhu XJ, Chen XX, Tu P, et al. Clinical and experimental study of paraneoplastic
pemphigus with Castleman’s disease. J Chin Dermatol. 2003;32:7-10.
45. Wang LC, Bu DF, Chen XX, et al. The pathogenesis of paraneoplastic pemphigus
associated with Castleman’s disease. Chin J Dermatol. 2004;37:37-39.
46. Wang J, Bu DF, Zhu XJ. Immunoglobulin variable region gene rearrangement
and hypermutation in paraneoplastic pemphigus associated Castleman’s tumor.
Beijing Da Xue Xue Bao. 2004;36:454-461.
47. Wang J, Zhu XJ. The new development on pathogenesis of autoimmune paraneoplastic pemphigus. Beijing Da Xue Xue Bao. 2004;36:540-543.
48. Fullerton SH, Woodley DT, Smoller BR, Anhalt GJ. Paraneoplastic pemphigus
with autoantibody deposition in bronchial epithelium after autologous bone marrow transplantation. JAMA. 1992;267:1500-1502.
49. Mar WA, Glaesser R, Struble K, Stephens-Groff S, Bangert J, Hansen RC. Paraneoplastic pemphigus with bronchiolitis obliterans in a child. Pediatr Dermatol.
2003;20:238-242.
50. Mydlarski PR, Mittmann N, Shear NH. Intravenous immunoglobulin: use in
dermatology. Skin Therapy Lett. 2004;9:1-6.
51. Gelfand EW. Use of IGIV in the treatment of immune-mediated dermatologic
disorders. J Investig Dermatol Symp Proc. 2004;9:92-96.
52. Jolles S, Hughes J, Whittaker S. Dermatological uses of high-dose intravenous
immunoglobulin. Arch Dermatol. 1998;134:80-86.
53. Wang J, Zhu XJ. The role of pathogenic B-cell clones in antibody mediated autoimmune disorders. J Dermatol Sci. 2004;36:141-148.
Announcement
Visit www.archdermatol.com. As an individual subscriber to Archives of Dermatology, you have full-text online access to the journal from 1998 forward. In addition, you can find abstracts to the journal as far back as
1965.
(REPRINTED) ARCH DERMATOL/ VOL 141, OCT 2005
1293
WWW.ARCHDERMATOL.COM
©2005 American Medical Association. All rights reserved.