Clinical and Laboratory Investigations
Primary cutaneous marginal zone B-cell lymphoma: a clinical,
histopathological, immunophenotypic and molecular genetic
study of 22 cases
O.SERVITJE, F.GALLARDO,* T.ESTRACH,† R.M.PUJOL,* A.BLANCO,‡
A.FERNA´ NDEZ-SEVILLA,§ L.PE´ TRIZ,– J .PEYRI´ AND V.ROMAGOSA**
Departments of Dermatology and **Pathology, Hospital Princeps d’Espanya, Ciutat Sanita`ria i Universita`ria de Bellvitge, C ⁄ Feixa
Llarga s ⁄ n, L’Hospitalet de Llobregat, 08907 Barcelona, Spain
*Department of Dermatology, Hospital del Mar, Barcelona, Spain
†Department of Dermatology, Hospital Clı´nic, Barcelona, Spain
‡Laboratory of Molecular Biology, Department of Cryobiology and Cell Therapy, Cancer Research Institute, Barcelona, Spain
Departments of §Clinical Haematology and –Radiotherapy, Institut Catala´ d’Oncologia, Barcelona, Spain
Accepted for publication 8 April 2002
Summary Background Primary cutaneous marginal zone B-cell lymphoma (MZCL) has recently been
described. Differentiation from follicular centre cell lymphomas and lymphocytomas is often difficult
due to insufficient experience and a lack of large series of patients.
Objectives To characterize primary cutaneous MZCL better, we report clinical, histopathological,
immunophenotypic and molecular genetics features in a series of 22 patients.
Methods All patients were treated and followed up at the same institution. Diagnosis of MZCL was
based on the World Health Organization classification criteria. All samples were routinely tested
with a wide panel of monoclonal antibodies. DNA was extracted from every sample following
standard methods. IgH rearrangement and t(14;18)(q32;q21) studies were performed in all samples.
Results Twenty-two patients (20 men, two women; mean age 50 years, range 24–77) were
included. The mean follow-up was 43 months. Seventy per cent of patients presented with characteristic
skin lesions on the trunk or extremities, consisting of deep red to violaceous infiltrated
plaques, nodules or tumours frequently surrounded by diffuse or annular erythema. Four patients
presented with lesions on the head and neck area. Two patients had disseminated skin lesions. The
main histopathological features were non-epidermotropic, dense lymphocytic infiltrates mainly
distributed in a nodular pattern. Adnexal involvement was usually present, with eventual formation
of lymphoepithelial complexes. Cytologically, the infiltrate was polymorphous with marginal
zone B cells and B-monocytoid cells. Blastoid CD30+ cells were often observed. Colonized reactive
germinal centres and lymphoplasmocytoid differentiation were frequently present. Neoplastic cells
were CD20+, CD79a+, CD5– and CD10–. Monotypic expression of light chains was observed in 18
cases (13 j; five k). Clonal IgH rearrangements were detected in 14 cases. The bcl-2 mutation
t(14;18)(q32;q21) was demonstrated in two cases. Most patients were treated with local radiotherapy.
Complete response rate with this approach was 100%. Six patients (27%) had skin
recurrences from 6 months to 8 years after first treatment. Five patients (23%) had extracutaneous
involvement. Two of them had a large cell transformation and one died of lymphoma. Three of four
patients with head and neck presentation developed extracutaneous disease.
Correspondence: Dr Octavio Servitje.
E-mail: oservitje@csub.scs.es
British Journal of Dermatology 2002; 147: 1147–1158.
2002 British Association of Dermatologists 1147
Conclusions MZCL appears to be a well recognizable entity, clinically, histologically and immunophenotypically.
Although prognosis is generally good, the disease has potential for skin as well as
extracutaneous recurrences. Large cell transformation and head and neck presentation may be
associated with a worse prognosis.
Key words: cutaneous B-cell lymphoma, cutaneous lymphoma, marginal zone B-cell lymphoma
Primary cutaneous marginal zone B-cell lymphoma
(MZCL) has recently been characterized. Its diagnostic
features are so far poorly delineated due to insufficient
experience and lack of consistency in clinical, histopathological,
immunophenotypic and molecular genetics
criteria used in different studies.1–4 Moreover, while some
authors believe that MZCL affects only a few patients,1,5
others consider that this entity accounts for themajority if
not all primary cutaneous B-cell lymphomas.6,7 This
controversy may in part reflect the difficulty in distinguishing
MZCL from follicular centre cell lymphomas of
the skin. Different CD10 expression appears to be a critical
point in the differential diagnosis between these two
entities, especially in MZCL cases with nodular or
follicular differentiation. Unfortunately, data regarding
CD10 expression in these lymphomas are very limited.
The expression of bcl-2 oncoprotein is another point of
discrepancy. Although most MZCLs express the protein, it
appears that only molecular studies searching for
t(14;18)(q32;q21) will be helpful in this differential
diagnosis. However, it has been shown8 that some noncutaneous
MZCLs as well as some cutaneous cases, as
presented here, do show this translocation. Another
major point of discrepancy between groups working in
small cell primary lymphomas of the skin is the separation
of MZCLs from so-called benign lymphocytoma cutis.9 As
primary cutaneous B-cell lymphoma used to be considered
a low-grade malignancy this distinction could be
extremely difficult on histopathological grounds alone. In
this situation evidence for clonality, in terms of restriction
of immunoglobulin light chain expression by the neoplastic
cells or immunoglobulin heavy chain receptor
gene rearrangement, could be of great value. Better to
delineate this entity, we present a series of patients with
clinical, histopathological and immunophenotypic features
of primary cutaneous MZCL with special regard to
CD10 expression and molecular genetic findings.
Patients and methods
Patients
We selected 22 patients with clinical, histopathological
and immunophenotypic features of primary cutaneous
MZCL. Criteria for the diagnosis of primary cutaneous
involvement included primary location on the skin and
absence of nodal or visceral involvement at least
6 months after the initial diagnosis. All patients hadcomplete
staging procedures, including thoracoabdominal
computed tomographic (CT) scan and bone marrow
biopsy and were treated and followed up at the same
institution.
Histopathology and immunohistochemistry
The histological criteria used for the diagnosis of MZCL
were those of the World Health Organization (WHO)
classification of lymphomas.10 All the histological
material was examined together by one experienced
haematopathologist and one dermatopathologist. All
the samples were routinely tested with a wide panel of
monoclonal antibodies including those against CD20,
CD79a, CD3, CD45RO, CD30, CD10, CD5, CD21,
S100, bcl-2, j and k chains.
IgH gene rearrangement
DNA from frozen and formalin-fixed, paraffin-embedded
tissue samples was extracted following standard
methods.11 Oligonucleotide primers were employed
based on the conserved framework III regions of the
VH and the JH genes in a single-step polymerase chain
reaction (PCR) as previously reported, with some
modifications:12 JH: GTGACCAGGGTNCCTTGGCCCCAG;
VH (FR3): ACACGGCYVTGTATTACTGT (Gibco
BRL Custom Primers). A 10-lL volume of DNA solution
(for paraffin samples) or 1 lg DNA in 10 lL water (for
frozen samples) was added to 40 lL of reaction mixture
containing 5 pmol of each primer, 2Æ5 mmol L)1
MgCl2, 10 lmol L)1 of each deoxyribonucleoside triphosphate
(dNTP), 10 mmol L)1 Tris–HCl pH 8Æ3 and
0Æ2 units Thermus Aquaticus Polymerase (Biotaq;
Clontech).
DNA obtained from the peripheral blood of healthy
donors was employed as a polyclonal negative control.
Amplification of DNA obtained from the Namalwa cell
line [American Type Culture Collection (ATCC), Rockville,
MD, U.S.A.] providing a monoclonal pattern was
1148 O.SERVITJE et al.
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
included as a positive control. Double-distilled water,
instead of DNA solution, was amplified simultaneously,
in order to detect any contamination. After an initial
denaturation for 5 min at 95 C, the PCR reaction was
performed with 34 cycles of 94 C denaturation, 62 C
annealing and 72 C extension, each step for 1 min, and
a final extension step for 5 min at 72 C. PCR products
were analysed and photographed after 12% polyacrylamide
gel electrophoresis under ultraviolet (UV)
illumination and staining with ethidium bromide (EtBr).
In order to validate our PCR technique, in selected
cases, gel slices corresponding to amplified bands were
excised, purified and sequenced using the Sequenase kit
(Perkin-Elmer), according to the specifications of the
manufacturer. This sequence analysis allowed us to
confirm the specificity of the band amplified (data not
shown).
The sensitivity achieved with this PCR technique was
determined by using serial dilutions of DNA from the
Namalwa cell line with DNA from peripheral blood of
healthy donors, leading to the identification of sharp
bands at approximate dilutions of 1 in 100 (1%).
t(14;18)(q32;q21)
One microgram of DNA solution or 10 lL volume of
paraffin samples were added to a final reaction volume of
50 lL, containing 0Æ1 mmol L)1 of each dNTP,
2Æ5 mmol L)1 MgCl2, reaction buffer and 25 pmol of
each primer MBR1 ⁄ JH1 (Calbiochem Oncogene
Research Products). Reaction amplifies the major breakpoint
region (MBR) of the bcl-2 gene. DNA from
nucleated cells of peripheral blood of healthy donors
was employed as the negative control, and DNA
obtained from the Karpas cell line (ATCC), carrying the
MBR form of this translocation, was used as the positive
control. As in the IgH gene rearrangement analysis,
purified water was included as an additional sample, to
detect contamination. Single-step PCR was performed
with 35 cycles of denaturation at 94 C for 1 min,
annealing at 65 C for 1 min and extension at 72 C for
1 min 30 s, followed by an extension step at 72 C for
5 min. PCR products were separated by electrophoresis
in a 1% agarose gel, stained with EtBr and visualized
with a UV transilluminator. Gels were blotted on to a
nylon membrane, hybridized with a c-32P-adenosine
triphosphate-labelled specific oligonucleotide and then
autoradiographed. Simultaneously, PCR amplification
of the bcl-2 non-translocated gene was carried out with
specific primers and used as a control of DNA integrity
and amplification reaction efficiency.13
Results
Clinical data
Clinical data are summarized in Table 1. Twenty-two
patients (20 men and two women) were included. The
mean age was 50 years (range 24–77). The mean
period between the appearance of skin lesions and
diagnosis was 27 months (range 1–96). None of the 22
patients had B-symptoms at presentation and lactate
dehydrogenase and b2-microglobulin levels were
normal in all cases.
The most frequent clinical presentation seen in 16
cases (70%; patients 1–16) was that of asymptomatic,
localized, deep red to violaceous infiltrated plaques,
grouped nodules or tumours, affecting the trunk
(seven cases) or extremities (nine cases). In three of
these patients skin lesions affected more than one
anatomical site. Frequently, skin lesions were surrounded
by an area of diffuse, infiltrated erythema
with small papules on the surface (Fig. 1). Two
patients presented with arciform infiltrated plaques
on the arm and leg.
In four cases (patients 17–20), skin lesions were
located on the head and neck area. In three of these
cases lesions consisted of violaceous nodules and deep,
infiltrated, erythematous plaques, affecting the face
(two cases) and the scalp (one case) (Fig. 2). In the
other patient a large, single tumour appeared on the
temple. This patient had suffered from lymphocytoma
cutis miliaris some years before and was previously
reported by our group.14
Two cases (patients 21 and 22) presented with
widespread skin involvement by violaceous nodules
and tumours as well as deep subcutaneous masses in
one case (Fig. 3). Both patients had a past history of
Hodgkin’s disease; one of them has been previously
reported.15
In most patients the therapy of choice was local
radiotherapy, or surgery when lesions were very
localized, with initial complete remission in all patients
treated and minimal side-effects. In the two patients
with widespread skin lesions total-skin electron beam
irradiation was administered and achieved rapid control
of the skin lesions with good tolerance of the
therapy.
Relapse ⁄ progression
Data regarding relapses are summarized in Table 1. Six
patients (27%) developed skin-limited recurrences
affecting the initial treated site or distant skin sites.
CUTANEOUS LYMPHOMA: MARGINAL ZONE B-CELL LYMPHOMA 1149
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
The time of presentation of the skin recurrences varied
from 6 months to more than 8 years after the first
manifestation. In one patient (patient 22) multiple skin
relapses were observed during a long follow-up of
more than 10 years. All these patients were again
well controlled with local therapies, such as local
Table 1. Clinical characteristics of 22 patients with primary cutaneous marginal zone B-cell lymphoma
Patient
no.
Age
years Sex
Time
prior to
diagnosis
(months) Skin lesions Location
Initial
therapy
Site of
relapse
(months after
initial disease)
Second
therapy
Status
(months of
follow-up)
1 47 M 6 Red infiltrated plaques Buttock RDT CR (6)
2 77 M 18 Grouped violaceous nodules Chest RDT CR (40)
3 25 M 1 Grouped violaceous nodules Lumbar region RDT CR (38)
4 36 M 26 Arciform infiltrated plaque Arm RDT CR (74)
5 45 M 24 Infiltrated plaque with
violaceous tumours
Arm RDT CR (57)
6 70 M 4 Infiltrated plaque with
violaceous tumours
Lumbar
region
RDT Back (19) RDT CR (37)
7 63 M 18 Arciform infiltrated
plaque
Leg RDT Initial
skin site
(104)
RDT CR (156)
8 40 M 12 Infiltrated plaque with
violaceous tumours
Back RDT CR (16)
9 25 M 84 Grouped tumours and
nodules, infiltrated
plaques
Thigh RDT Contralateral
thigh (18)
Surgery CR (72)
10 24 M 14 Infiltrated plaques with
violaceous nodules
Thigh RDT CR (12)
11 52 M 7 Infiltrated plaque with
large tumours
Back RDT CR (8)
12 41 F 4 Grouped violaceous
nodules
Arm Surgery CR (35)
13 70 M 48 Infiltrated plaques
with nodules
Arm,
lumbar
region
RDT Contralateral
arm (40)
None SD (65)
14 66 M 96 Infiltrated plaques
with nodules
Arm,
thigh
RDT CR (9)
15 25 M 9 Grouped erythematous
nodules
Arm,
trunk
RDT BM (13) None LFU
16 52 M 60 Infiltrated plaques
with nodules
Chest RDT Back (14),
bowel (24)
RDT,
CHOP
CR (20)
17 63 F 2 Large violaceous nodules,
deep infiltrated plaques
Retroauricular
bilateral,
face
Surgery,
RDT
Face (12),
subcutaneous
tissue of the
leg and local
lymph nodes
(18)
RDT,
chlorambucil
DL (48)
18 73 M 23 Deep infiltrated plaques,
violaceous nodules
Supralabial,
face
Surgery Face, parotid
gland, cervical
lymph nodes,
BM (8)
Chlorambucil SD (16)
19 52 M 19 Infiltrated plaques,
violaceous nodules
Scalp,
forehead
None Spleen,
lymph nodes,
BM (10)
CHOP CR (38)
20 75 M 72 Large tumour Temple RDT DUC (12)
21 38 M 8 Large infiltrated plaques
with nodules
Disseminated E-beam Skin (6) SD (15)
22 48 M 36 Grouped nodules
and tumours
Disseminated E-beam Skin
relapses
(many)
RDT,
surgery
SD (126)
RDT, radiotherapy; E-beam, total-skin electron beam radiotherapy; BM, bone marrow; CHOP, cyclophosphamide, adriamycin, vincristine, prednisolone;
CR, complete response; SD, stable disease; DL, died of lymphoma; DUC, died from unrelated causes; LFU, lost to follow-up.
1150 O.SERVITJE et al.
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
Figure 1. Clinical presentation in patients
with primary cutaneous marginal zone B-cell
lymphoma on the trunk or extremities
(patients 5 and 10). Note erythema and
nodules surrounding skin tumours.
Figure 2. Clinical appearance of marginal zone B-cell lymphoma
involving the face and scalp (patient 19).
Figure 3. Disseminated lesions seen in two patients with cutaneous
marginal zone B-cell lymphoma (top, patient 22; bottom, patient 21).
CUTANEOUS LYMPHOMA: MARGINAL ZONE B-CELL LYMPHOMA 1151
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
radiotherapy or surgery, and none developed extracutaneous
lymphoma.
In five patients (23%) extracutaneous involvement
was detected during the follow-up. In one of them
(patient 16), skin relapse was accompanied by a largecell
lymphoma of the bowel. One patient (patient 17)
developed massive involvement of the subcutaneous
tissue of the left leg along with local lymph node and
skin recurrence. One patient (patient 18), with initial
skin lesions limited to the facial area, presented with
parotid gland and local lymph node involvement. In
one patient (patient 19), splenic lymphoma of the
marginal zone type and abdominal lymph node
involvement was diagnosed. Bone marrow infiltration
was detected in three cases (patients 15, 18 and 19).
Two of them also showed involvement at other
extracutaneous sites but in one patient (patient 15)
bone marrow infiltration was an isolated feature
without other signs of cutaneous or extracutaneous
disease. The period for extracutaneous presentation
in our patients ranged from 8 to 24 months after
initial diagnosis. These patients were treated with
chemotherapy or combination (radiotherapy plus chemotherapy)
therapy with good control of the disease
except in patient 17 in whom disease progressed.
Follow-up
The mean follow-up in our patients was 43 months
(range 6 months)13 years). Fifteen patients (68%)
remain in complete remission. Four patients have
stable residual disease. Only two patients died during
follow-up, one of them of lymphoma progression
4 years after initial diagnosis (patient 17) and the
other due to unrelated causes (patient 20).
Histopathology
Data regarding histopathology are summarized in
Table 2. Microscopically, MZCLs were characterized
by dense lymphocytic infiltrates mostly distributed on
the reticular dermis and often extending to the subcutis
(bottom heavy). There was no epidermotropism and a
Grenz zone could be observed in all cases. In seven
cases the infiltrates were limited to the superficial
reticular dermis in a so-called top-heavy pattern. In
most cases (68%) a nodular pattern of infiltration was
observed with the presence of unusual rounded or
elongated structures (Fig. 4A). In seven cases the
pattern of infiltration was rather diffuse. Adnexal
involvement was seen in all but two cases. It consisted
Table 2. Histopathological and molecular genetics findings in 22 patients with primary cutaneous marginal zone B-cell lymphoma
Patient
no.
Distribution
of the
infiltrate
Pattern of
infiltration
Adnexal
involvement
Germinal
centres
Plasma
cells
CD30
blasts T cells
Superficial
light chains
Cytoplasmic
light chains
IgH
rearrangement
Bcl-2
expression t(14;18)
1 TH D E No Yes Yes Yes K K – No –
2 BH N F Yes No No No K NE + No –
3 TH N F Yes No Yes Yes K K + Yes –
4 BH N F + E Yes No Yes Yes NE NE + No –
5 BH N F Yes Yes Yes Yes NE L + Yes –
6 TH D No Yes Yes Yes Yes NE PC + No –
7 TH N F Yes Yes Yes Yes NE K – Yes –
8 BH D F No No No Yes NE K + Yes –
9 BH N No Yes Yes Yes Yes L L + Yes –
10 TH D F Yes Yes Yes Yes L L – Yes –
11 BH D F No Yes Yes Yes K K – Yes –
12 BH N F + E Yes Yes Yes Yes NE PC + Yes –
13 BH D F No Yes Yes Yes NE K + Yes –
14 BH N E No Yes Yes Yes L L + No –
15 TH N F No Yes No Yes NE PC – Yes –
16 BH N E No Yes Yes Yes NE K + Yes –
17 BH N E No Yes Yes Yes NE K + Yes +
18 BH D E No No Yes Yes K NE + Yes –
19 TH N F Yes No Yes Yes K NE – Yes –
20 BH N F Yes No Nv Yes K NE – No –
21 BH N E Yes Yes Yes Yes L L – Yes –
22 BH N F Yes Yes Yes Yes K K + Yes +
TH, top-heavy; BH, bottom-heavy; N, nodular; D, diffuse; E, eccrine; F, follicular; K, kappa; L, lambda; NE, no expression; PC, polyclonal.
1152 O.SERVITJE et al.
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
of infiltration of the eccrine glands and hair follicles by
the neoplastic cells with eventual formation of lymphoepithelial
complexes (Fig. 4B).
MZCLs were characterized cytologically by a polymorphous
infiltrate that included small- to mediumsized
lymphocytes with dense chromatin and pale and
watery cytoplasm (marginal zone cells) with frequent
reniform nuclei (B-monocytoid cells) and a variable
proportion of sparse larger centroblast-like cells (blastoid
cells) intermingled within the infiltrate (Fig. 5B).
In 65% of cases germinal centre-like structures surrounded
by neoplastic cells mimicking incomplete or
aberrant mantle zones were observed. These germinal
centres were considered as reactive and often had
colonization by the neoplastic infiltrate (Fig. 4C). In
most cases there were areas of plasmocytoid differentiation
with the presence of mature plasma cells that
were constantly disposed at the periphery of the
infiltrate and under the dermoepidermal junction
(Fig. 4D). In five cases a moderate degree of tissue
eosinophilia was observed.
Both small- to medium-sized cells (marginal zone,
B-monocytoid cells) and large centroblast-like cells
(blastoid cells) were positive for CD20 (Fig. 5A) and
Figure 4. (A) Nodular infiltrate in primary
cutaneous marginal zone B-cell lymphoma
(MZCL). Note periadnexal distribution (patient
4). (B) Detail of (A). Note marginal zone cell
infiltration of eccrine glands with lymphoepithelial
complex formation (arrowhead). (C)
Example of a reactive germinal centre with
colonization by MZCL (patient 5). (D)
Lymphoplasmocytoid differentiation in MZCL.
Note the presence of mature plasma cells at
the periphery of the infiltrate (arrowheads)
(patient 22). Haematoxylin and eosin; original
magnification: A, · 40; B, · 400; C, · 160;
D, · 400.
CUTANEOUS LYMPHOMA: MARGINAL ZONE B-CELL LYMPHOMA 1153
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
CD79a and negative for CD5 and CD10. Only in one
case (patient 17) were small aggregates of CD10+ cells
scattered in the infiltrate. These cells were considered to
be residual germinal centre cells. In most cases a
variable but usually low number of large CD30+ cells
was present in the infiltrate (Fig. 5B). Neoplastic cells
were bcl-2+ in 16 cases (Fig. 5C) and Ki-67 expression
(studied in 14 cases) was limited to large blastoid cells.
In all cases there was a variable but sometimes
significant number of reactive T cells that expressed
the CD3 and CD45RO phenotype. In some cases
reactive T cells were so numerous that they masked
the neoplastic B cell compartment. The combination of
cytoplasmic and surface light chain expression disclosed
a monotypic pattern in 18 of 22 cases (13 j and
five k) (Fig. 5D). In three cases the pattern was
considered polytopic and in one case light chain
expression could not be demonstrated. There was a
constant presence of large number of S100+ cells
intermingled with the neoplastic infiltrate, as well as
along the basement membrane of the epidermis. This
finding probably reflects the role of a large number of
Figure 5. (A) CD20 (L26) staining of
marginal zone cells (patient 10). (B) CD30
(Ki-1) staining of blastoid cells intermingled
within the infiltrate. This was a constant
finding in cutaneous marginal zone B-cell
lymphoma (MZCL) (patient 10). (C) Bcl-2
staining showing positive marginal zone cells
and negative reactive germinal centre cells
(patient 3). (D) An example of monotypic
expression of superficial j light chain in a case
of MZCL. This case was strictly negative for k
chain (patient 11). Original magnification: A,
· 40; B, · 100; C, · 40; D, · 400.
1154 O.SERVITJE et al.
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
antigen-presenting cells (Langerhans and interdigitating
dendritic cells) eventually involved in the process of
lymphoma transformation, with similarities to Helicobacter
pylori gastric MALT lymphoma.
Molecular genetics
The results are summarized in Table 2. Fourteen of 22
cases of MZCL showed discrete bands of 100–120-bp
lengths suggesting a monoclonal expansion. In the
other eight cases, the presence of a background smear
with no discrete bands indicated that the PCR was
successful in amplifying IgH gene rearrangements from
the normal lymphocytes present in these populations,
but did not amplify any malignant clone (Fig. 6).
As far as the t(14;18) translocation is concerned,
only two cases (patients 17 and 22) showed a discrete
band of 150–350 bp in length after PCR amplification,
indicating the presence of t(14;18)(q32;q21). Sequence
analysis was not performed in this study, but the
positive results in these two cases were confirmed by
hybridization of the bcl-2 ⁄ IgH PCR products with a
specific 32P-labelled probe (5¢-GGC CTG TTT CAA CAC
AGA CCC-3¢) according to the manufacturer’s specifications
16 (Fig. 6). With our technique, translocation
involving the minor cluster region of t(14;18) would
not be detected.
Discussion
Primary cutaneous B-cell lymphomas represent a
unique form of extranodal B-cell malignancies. Most
of these lymphomas have been classified so far either as
follicular centre cell lymphomas or cutaneous immunocytomas.
1,17–19 Recently, a new type of primary
cutaneous B-cell lymphoma that shares the histopathological
characteristics of MZCL of the nodal or
mucosal sites has been described.1–4 However, as only
a few studies have been published, there is still
controversy regarding the incidence and diagnostic
criteria of this entity. We present a series of patients
with the clinical and histopathological features of
primary MZCL of the skin and show detailed data on
clinicopathological correlation, immunophenotype and
molecular genetics.
In our patients there was a large preponderance of
men over women and, as in previous reports,2 patients
were mainly adults aged over 40 years. MZCL was
most commonly localized on the trunk or arms. In
agreement with previous studies, lower leg presentation
was rare.2–4 In a high proportion of cases skin
lesions were quite characteristic, consisting of grouped
nodules or tumours that arose over an area of diffuse or
annular erythema. These changes are similar to those
described by Cerroni et al.,2 and enable this entity to be
suspected clinically in a proportion of cases. A small
number of patients presented with lesions on the head
and neck area. Although initially described in association
with follicular centre cell lymphomas, this
location has also been seen by other authors in MZCL
patients.4,20 When affecting this area, lesions tended to
be hard and deeply infiltrated.
A very infrequent presentation, seen in two patients,
was that of extensive, generalized involvement. Interestingly,
these two patients had previous Hodgkin’s
disease that went into complete remission after
standard treatment. The association of primary cutaneous
B-cell lymphomas with Hodgkin’s disease seems
to be extremely rare. Only one previous case of T-cellrich
B-cell lymphoma has been reported so far.21 These
authors demonstrated a common B-cell origin of both
neoplasias. Interestingly, one of our cases presented a
large number of T cells that we interpreted as reactive.
We believe that these cases share some similarities and
may represent a subtype of skin lymphomas associated
with Hodgkin’s disease.
Figure 6. (A) Polymerase chain reaction (PCR) for IgH gene rearrangement
in cutaneous marginal zone B-cell lymphoma. Lanes C3,
C6 and C7 correspond to positive clonal cases showing discrete bands;
negative cases show only smear. pb, peripheral blood from healthy
donors showing polyclonal pattern; Nam, positive control (Namalwa
cell line); w, water. (B) PCR for t(14;18)(q32;q21). Arrow indicate
patients 17 and 22, respectively, carrying the t(14;18)(q32;q21)
translocation; K, DNA extracted from the Karpas cell line used as
positive control. (C) Southern blot analysis with a specific probe
confirmed the positive results.
CUTANEOUS LYMPHOMA: MARGINAL ZONE B-CELL LYMPHOMA 1155
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
Cutaneous MZCLs are characterized histologically
by a polymorphous lymphocyte proliferation mainly
composed of marginal zone B cells. These cells have
specific cytological features and immunophenotype,
being CD20+, CD79a+, CD5– and CD10–. In agreement
with previous reports, we observed that the
lymphocytic infiltrates were arranged mostly in nodular
aggregates that were frequently disposed around
the follicular and eccrine structures.2,20 Furthermore,
some of our cases presented lymphoepithelial complexes
in relationship with eccrine epithelia, similar to
that observed in mucosal MALT lymphomas. As has
been previously reported,2,4,22 we frequently observed
a variable number of blastoid, monocytoid cells
intermingled within the infiltrate. These cells constantly
expressed the CD30 antigen in addition to
B-cell markers. This finding has not been reported so
far, and we believe it supports the hypothesis that
blastoid cells represent activated monocytoid neoplastic
B cells. Another characteristic of MZCL is the presence
of areas of lymphoplasmocytoid differentiation. In
some cases mature plasma cells that always showed
monotypic light chain expression were observed. In
accordance with previous observations,2,20 these plasma
cells tended to be disposed at the periphery of the
infiltrate and under the epidermis. These findings make
differentiation from what has been called immunocytoma
very difficult, and we believe this entity may be
part of the spectrum of MZCL.23 In fact, according to
recently published criteria, the WHO recognizes only
as lymphoplasmocytic lymphoma (immunocytoma)
those cases with usually systemic involvement and
clinical features of Waldenstro¨m macroglobulinaemia.
Most cases with lung, gastrointestinal tract or skin
involvement are better diagnosed as MALT lymphomas,
where a prominent plasmocytoid differentiation is
frequently seen.10
MZCLs of the skin are also remarkable for the
presence of germinal centre cell structures in about
50% of cases. These structures are considered to be
reactive and are often infiltrated by the neoplastic
marginal zone cells.2,4,20,24 Special attention must be
taken in differentiating MZCLs with reactive germinal
centres from follicular centre cell lymphomas. In our
hands, reactive germinal centre cells appear to be
negative for bcl-2 expression (data not shown) and
polytopic or negative for the expression of light chains.2
In addition, follicular centre cell lymphomas usually
express CD10 antigen.24 All our MZCL cases were
negative for CD10. These data are in agreement with
those recently reported by de Leval et al.25 and strongly
suggest that CD10 negativity should be included as a
diagnostic criterion for MZCL of the skin.
A frequent feature of cutaneous MZCL is the presence
of reactive T cells.2,3 These T cells had a mature
phenotype and varied from one biopsy to another in the
same patient, but in some cases they were so numerous
that they masked the neoplastic population.
A major point of controversy is the differentiation of
MZCL from lymphocytoma cutis on histopathological
grounds alone. B-cell rearrangement analysis detected
clonality in 14 of 22 of our cases. It is unclear why this
PCR method failed to detect a predominant clone
in some of our MZCL patients. Somatic mutations in
critical nucleotides of the VH gene are frequent in
marginal zone lymphomas and may prevent annealing
of oligonucleotide primers. Deletion in nucleotides from
the 3¢ end of the VH gene or the 5¢ end of the JH gene
can also prevent primer hybridization and amplification.
Finally, unconventional rearrangements or chromosomal
translocations involving the IgH locus may
also result in failure of PCR amplification.26 However,
with the combination of light chain immunoglobulin
expression and B-cell rearrangement we were able to
detect clonality in all our cases.
We believe that this approach can be extremely
useful in differentiating MZCL from a reactive process.
However, in one of our cases, MZCL developed from
lymphocytoma cutis, suggesting that this entity can
evolve from a reactive polyclonal lymphoproliferation.
We feel that cases showing a morphological and
immunophenotypic pattern of MZCL, even when light
chain restriction cannot be demonstrated, should be
managed as a lymphoma, assuming that only a ‘wait
and see’ procedure will be undertaken and not an
aggressive approach.
Bcl-2 mutation appears to be a highly unusual
finding in cutaneous B-cell lymphomas. In a very
recent report, Child et al.27 suggested that t(14;18) is
only present in those cases with systemic B-cell
lymphoma from the beginning. This mutation was
demonstrated in only two of our patients, one of whom
died as a consequence of lymphoma progression
(patient 17). In this patient extracutaneous involvement
began after 18 months of skin-limited disease
and extensive staging procedures, including CT scan and
bone marrow biopsy, gave normal results. However,
the possibility of occult systemic lymphoma from the
outset cannot be excluded in this case. The other
patient (patient 22) has had multiple skin relapses but
is stable after more than 10 years of follow-up. This
patient had a history of Hodgkin’s disease 7 years
1156 O.SERVITJE et al.
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
before the appearance of skin lymphoma. Hodgkin’s
disease was diagnosed on the basis of unequivocal
clinical, radiological and histological findings and was
classified as clinicopathological stage IIA. After standard
Hodgkin’s disease therapy the patient went into
complete remission and no signs of either Hodgkin’s
disease or systemic lymphoma have been seen so far
after repeated complete staging. Recently, a common Bcell
origin was demonstrated in a patient with cutaneous
B-cell lymphoma and a past history of Hodgkin’s
disease.21 In our opinion, this is very relevant to
patient 22, suggesting a particular clinical situation for
such patients. If our case also represents a common
B-cell origin or not is unclear; extensive microdissection
analysis would be necessary for confirmation.
However, t(14;18) would also be an uncommon
finding in this situation. The t(14;18) translocation is
a very common genetic alteration in follicular centre
cell lymphomas but has also been described in other
lymphoma types, including MZCL of mucosal sites.8
Whether our cases represent some of these exceptions
or true follicular centre cell lymphomas is difficult to
assess. However, both cases totally fulfilled the histopathological
criteria for MZCL, suggesting that more
investigation is needed in order to assess the specificity
of these criteria when applied to cutaneous lymphomas.
The therapeutic approach in our patients was
dependent on the extent of the skin lesions. In most
patients local radiotherapy was the treatment of
choice. The clinical complete response ratio with this
approach was 100%. However, about 35% of patients
had local recurrences at treated or distant skin sites.
These results are in agreement with those reported by
other authors,2 and indicate that, although benign in
prognosis, MZCL has a high potential for local
recurrences. In all our patients the histopathological
features and immunophenotype of skin recurrences
were similar to those of the initial skin lesions (data
not shown). Some of our patients had recurrences a
long time after the initial skin lesions. In particular,
patient 22 has had continuous recurrences over a
10-year period without any sign of extracutaneous
involvement or high-grade transformation. This fact
confirms the low-grade nature of most of these
lymphomas.4,20,22,24
Total-skin electron beam radiotherapy is an
uncommon approach in low-grade cutaneous lymphoma.
However, in two patients we decided to employ
this technique due to the extensive involvement.
Although skin recurrence developed early in both
patients, the disease appeared to follow a more benign
course, with limited involvement in both patients.
Five of 22 patients (23%) developed extracutaneous
lymphoma during the follow-up, affecting subcutaneous
tissue, lymph nodes, salivary glands, gastrointestinal
tract, spleen and bone marrow. In two patients
there was large cell transformation, and one died of
lymphoma. The other patient responded well to
chemotherapy, suggesting the good clinical behaviour
of this entity. One of our patients had isolated bone
marrow involvement without any sign of deterioration.
Unfortunately, this patient was lost to follow-up. Three
of four patients with head and neck presentation
developed extracutaneous disease.
Data on extracutaneous involvement in MZCL are
controversial. Although some authors reported that
these patients never develop systemic disease,2,4,24
other authors have observed an incidence of extracutaneous
involvement similar to that in our series;
Grønbæk et al.22 observed extracutaneous disease in
four of 12 patients with primary cutaneous MZCL.
These patients had lymph node, liver and spleen
involvement. Two of these patients presented large cell
transformation and both died. De la Fouchardie`re et al.3
reported nine cases of MZCL of the skin. Although two
patients did not fulfil the criteria for primary cutaneous
lymphomas, they observed four patients with extracutaneous
disease involving the lymph nodes, parotid
gland, orbit and bone marrow. Bailey et al.20 reported
extracutaneous disease in two of seven patients with
primary skin disease. These patients developed orbit and
breast involvement. Similar to our cases, in these two
studies all patients with extracutaneous disease had
skin lesions on the head and neck area.
In summary, MZCL appears to be a well recognizable
cutaneous B-cell lymphoma. Although of general good
prognosis, some patients appear to have a tendency to
local or extracutaneous disease. Larger series are
needed in order to identify prognostic parameters.
From our data and review of the literature, head and
neck presentation and large cell transformation may be
associated with a worse prognosis.
Acknowledgments
This work has been supported in part by a grant for the
‘Xarxa Tema`tique de Linfomes Cutani’ from the Generalitat
de Catalunya and CIRIT APSQV 1998 SI. We
thank staff of the Audio-Visual Unit of the Ciutat
Sanita`ria i Universita`ria de Bellvitge for their expert
technical assistance.
CUTANEOUS LYMPHOMA: MARGINAL ZONE B-CELL LYMPHOMA 1157
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
References
1 Willemze R, Kerl H, Sterry W et al. EORTC classification for primary
cutaneous lymphomas: a proposal from the Cutaneous
Lymphoma Study Group of the European Organization for
Research and Treatment of Cancer. Blood 1997; 90: 354–71.
2 Cerroni L, Signoretti S, Hofler G et al. Primary cutaneous marginal
zone B-cell lymphoma: a recently described entity of lowgrade
malignant cutaneous B-cell lymphoma. Am J Surg Pathol
1997; 21: 1307–15.
3 De la Fouchardie`re A, Balme B, Chouvet B et al. Primary cutaneous
marginal zone B-cell lymphoma: a report of 9 cases. J Am
Acad Dermatol 1999; 41: 181–8.
4 Tomaszewski MM, Abbondanzo SL, Lupton GP. Extranodal marginal
zone B-cell lymphoma of the skin. a morphologic and immunophenotypic
study of 11 cases. Am J Dermatopathol 2000; 22:
205–11.
5 Willemze R, Rijlaarsdam JU, Meijer CJ. Are most primary cutaneous
B-cell lymphomas ‘marginal cell lymphomas’? Br J
Dermatol 1995; 133: 950–2.
6 Slater DN. MALT and SALT: the clue to cutaneous B-cell
lymphoproliferative disease. Br J Dermatol 1994; 131: 557–61.
7 Slater DN. Primary cutaneous B-cell lymphomas. Arch Dermatol
1997; 133: 1604–5.
8 Garcı´a-Sanz R, Vargas Montero M, Gonzalez Diaz M et al. Detection
of single and associated lesions of the Bcl-1, Bcl-2, Bcl-6,
c-myc, p53 and p16 genes in B-cell non-Hodgkin lymphomas:
value of molecular analysis for a better assignment of the histologic
subtype. Haematologica 1998; 83: 209–16.
9 Schmid U, Eckert F, Griesser H et al. Cutaneous follicular lymphoid
hyperplasia with monotypic plasma cells. A clinicopathologic
study of 18 patients. Am J Surg Pathol 1995; 19: 12–20.
10 Jaffe ES, Harris NL, Stein H, Vardiman JW. World Health Organization
Classification of Tumours. Pathology and Genetics of Tumours
of Haematopoietic and Lymphoid Tissues. Lyon: IRCC Press, 2001.
11 Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning: a
Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor
Laboratory, 1989.
12 Potter M, Steward C, Maitland N, Oakhill A. Detection of clonality
in childhood B-lineage acute lymphoblastic leukemia by the
polymerase chain reaction. Leukemia 1992; 6: 289–94.
13 Miettinen M, Lasota J. Polymerase chain reaction based gene
rearrangement studies in the diagnosis of follicular lymphoma—
performance in formaldehyde-fixed tissue and application
in clinical problem cases. Pathol Res Pract 1997; 193: 9–19.
14 Moreno A, Curco N, Serrano T et al. Disseminated, miliarial type
lymphocytoma cutis. A report of two cases. Acta Derm Venereol
(Stockh) 1991; 71: 334–6.
15 Servitje O, Marti RM, Estrach T et al. Occurrence of Hodgkin’s
disease and cutaneous B-cell lymphoma in the same patient: a
report of two cases. Eur J Dermatol 2000; 10: 43–6.
16 Galoin S, al-Saati T, Schlaifer D et al. Oligonucleotide clonospecific
probes directed against the junctional sequence of t(14;18): a
new tool for the assessment of minimal residual disease in follicular
lymphomas. Br J Haematol 1996; 94: 676–84.
17 Santucci M, Pimpinelli N, Arganini L. Primary cutaneous B-cell
lymphoma: a unique type of low-grade lymphoma. Clinicopathologic
and immunologic study of 83 cases. Cancer 1991; 67:
2311–26.
18 Willemze R, Meijer CJ, Sentis HJ et al. Primary cutaneous large
cell lymphomas of follicular center cell origin. A clinical follow-up
study of nineteen patients. J Am Acad Dermatol 1987; 16: 518–
26.
19 Rijlaarsdam JU, van der Putte SC, Berti E et al. Cutaneous
immunocytomas; a clinicopathologic study of 26 cases. Histopathology
1993; 23: 117–25.
20 Bailey EM, Ferry JA, Harris NL et al. Marginal zone lymphoma
(low-grade B-cell lymphoma of mucosa-associated lymphoid tissue
type) of skin and subcutaneous tissue: a study of 15 patients.
Am J Surg Pathol 1996; 20: 1011–23.
21 Bra¨uninger A, Hansmann ML, Strickler JG et al. Identification of
common germinal-center B-cell precursors in two patients with
both Hodgkin’s disease and non-Hodgkin’s lymphoma. N Engl J
Med 1999; 340: 1239–47.
22 Grønbæk K, Møller PH, Nedergaard T et al. Primary cutaneous
B-cell lymphoma: a clinical, histological, phenotypic and genotypic
study of 21 cases. Br J Dermatol 2000; 142: 913–23.
23 Duncan LM, LeBoit PE. Are primary cutaneous immunocytoma
and marginal zone lymphoma the same disease? Am J Surg Pathol
1997; 21: 1368–71.
24 Yang B, Tubbs RR, Finn W et al. Clinicopathologic reassessment
of primary cutaneous B-cell lymphomas with immunophenotypic
and molecular genetic characterization. Am J Surg Pathol 2000;
24: 694–702.
25 de Leval L, Harris NL, Longtime J et al. Cutaneous B-cell
lymphomas of follicular and marginal zone types. Use of Bcl-6,
CD10, Bcl-2 and CD21 in differential diagnosis and classification.
Am J Surg Pathol 2001; 25: 732–41.
26 Spencer J, Perry M, Dunn-Walters D. Human marginal-zone B
cells. Immunol Today 1998; 19: 421–6.
27 Child FJ, Russell-Jones R, Woolford AJ et al. Absence of t(14;18)
chromosomal translocation in primary cutaneous B-cell lymphoma.
Br J Dermatol 2001; 144: 735–44.
1158 O.SERVITJE et al.
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
Primary cutaneous marginal zone B-cell lymphoma: a clinical,
histopathological, immunophenotypic and molecular genetic
study of 22 cases
O.SERVITJE, F.GALLARDO,* T.ESTRACH,† R.M.PUJOL,* A.BLANCO,‡
A.FERNA´ NDEZ-SEVILLA,§ L.PE´ TRIZ,– J .PEYRI´ AND V.ROMAGOSA**
Departments of Dermatology and **Pathology, Hospital Princeps d’Espanya, Ciutat Sanita`ria i Universita`ria de Bellvitge, C ⁄ Feixa
Llarga s ⁄ n, L’Hospitalet de Llobregat, 08907 Barcelona, Spain
*Department of Dermatology, Hospital del Mar, Barcelona, Spain
†Department of Dermatology, Hospital Clı´nic, Barcelona, Spain
‡Laboratory of Molecular Biology, Department of Cryobiology and Cell Therapy, Cancer Research Institute, Barcelona, Spain
Departments of §Clinical Haematology and –Radiotherapy, Institut Catala´ d’Oncologia, Barcelona, Spain
Accepted for publication 8 April 2002
Summary Background Primary cutaneous marginal zone B-cell lymphoma (MZCL) has recently been
described. Differentiation from follicular centre cell lymphomas and lymphocytomas is often difficult
due to insufficient experience and a lack of large series of patients.
Objectives To characterize primary cutaneous MZCL better, we report clinical, histopathological,
immunophenotypic and molecular genetics features in a series of 22 patients.
Methods All patients were treated and followed up at the same institution. Diagnosis of MZCL was
based on the World Health Organization classification criteria. All samples were routinely tested
with a wide panel of monoclonal antibodies. DNA was extracted from every sample following
standard methods. IgH rearrangement and t(14;18)(q32;q21) studies were performed in all samples.
Results Twenty-two patients (20 men, two women; mean age 50 years, range 24–77) were
included. The mean follow-up was 43 months. Seventy per cent of patients presented with characteristic
skin lesions on the trunk or extremities, consisting of deep red to violaceous infiltrated
plaques, nodules or tumours frequently surrounded by diffuse or annular erythema. Four patients
presented with lesions on the head and neck area. Two patients had disseminated skin lesions. The
main histopathological features were non-epidermotropic, dense lymphocytic infiltrates mainly
distributed in a nodular pattern. Adnexal involvement was usually present, with eventual formation
of lymphoepithelial complexes. Cytologically, the infiltrate was polymorphous with marginal
zone B cells and B-monocytoid cells. Blastoid CD30+ cells were often observed. Colonized reactive
germinal centres and lymphoplasmocytoid differentiation were frequently present. Neoplastic cells
were CD20+, CD79a+, CD5– and CD10–. Monotypic expression of light chains was observed in 18
cases (13 j; five k). Clonal IgH rearrangements were detected in 14 cases. The bcl-2 mutation
t(14;18)(q32;q21) was demonstrated in two cases. Most patients were treated with local radiotherapy.
Complete response rate with this approach was 100%. Six patients (27%) had skin
recurrences from 6 months to 8 years after first treatment. Five patients (23%) had extracutaneous
involvement. Two of them had a large cell transformation and one died of lymphoma. Three of four
patients with head and neck presentation developed extracutaneous disease.
Correspondence: Dr Octavio Servitje.
E-mail: oservitje@csub.scs.es
British Journal of Dermatology 2002; 147: 1147–1158.
2002 British Association of Dermatologists 1147
Conclusions MZCL appears to be a well recognizable entity, clinically, histologically and immunophenotypically.
Although prognosis is generally good, the disease has potential for skin as well as
extracutaneous recurrences. Large cell transformation and head and neck presentation may be
associated with a worse prognosis.
Key words: cutaneous B-cell lymphoma, cutaneous lymphoma, marginal zone B-cell lymphoma
Primary cutaneous marginal zone B-cell lymphoma
(MZCL) has recently been characterized. Its diagnostic
features are so far poorly delineated due to insufficient
experience and lack of consistency in clinical, histopathological,
immunophenotypic and molecular genetics
criteria used in different studies.1–4 Moreover, while some
authors believe that MZCL affects only a few patients,1,5
others consider that this entity accounts for themajority if
not all primary cutaneous B-cell lymphomas.6,7 This
controversy may in part reflect the difficulty in distinguishing
MZCL from follicular centre cell lymphomas of
the skin. Different CD10 expression appears to be a critical
point in the differential diagnosis between these two
entities, especially in MZCL cases with nodular or
follicular differentiation. Unfortunately, data regarding
CD10 expression in these lymphomas are very limited.
The expression of bcl-2 oncoprotein is another point of
discrepancy. Although most MZCLs express the protein, it
appears that only molecular studies searching for
t(14;18)(q32;q21) will be helpful in this differential
diagnosis. However, it has been shown8 that some noncutaneous
MZCLs as well as some cutaneous cases, as
presented here, do show this translocation. Another
major point of discrepancy between groups working in
small cell primary lymphomas of the skin is the separation
of MZCLs from so-called benign lymphocytoma cutis.9 As
primary cutaneous B-cell lymphoma used to be considered
a low-grade malignancy this distinction could be
extremely difficult on histopathological grounds alone. In
this situation evidence for clonality, in terms of restriction
of immunoglobulin light chain expression by the neoplastic
cells or immunoglobulin heavy chain receptor
gene rearrangement, could be of great value. Better to
delineate this entity, we present a series of patients with
clinical, histopathological and immunophenotypic features
of primary cutaneous MZCL with special regard to
CD10 expression and molecular genetic findings.
Patients and methods
Patients
We selected 22 patients with clinical, histopathological
and immunophenotypic features of primary cutaneous
MZCL. Criteria for the diagnosis of primary cutaneous
involvement included primary location on the skin and
absence of nodal or visceral involvement at least
6 months after the initial diagnosis. All patients hadcomplete
staging procedures, including thoracoabdominal
computed tomographic (CT) scan and bone marrow
biopsy and were treated and followed up at the same
institution.
Histopathology and immunohistochemistry
The histological criteria used for the diagnosis of MZCL
were those of the World Health Organization (WHO)
classification of lymphomas.10 All the histological
material was examined together by one experienced
haematopathologist and one dermatopathologist. All
the samples were routinely tested with a wide panel of
monoclonal antibodies including those against CD20,
CD79a, CD3, CD45RO, CD30, CD10, CD5, CD21,
S100, bcl-2, j and k chains.
IgH gene rearrangement
DNA from frozen and formalin-fixed, paraffin-embedded
tissue samples was extracted following standard
methods.11 Oligonucleotide primers were employed
based on the conserved framework III regions of the
VH and the JH genes in a single-step polymerase chain
reaction (PCR) as previously reported, with some
modifications:12 JH: GTGACCAGGGTNCCTTGGCCCCAG;
VH (FR3): ACACGGCYVTGTATTACTGT (Gibco
BRL Custom Primers). A 10-lL volume of DNA solution
(for paraffin samples) or 1 lg DNA in 10 lL water (for
frozen samples) was added to 40 lL of reaction mixture
containing 5 pmol of each primer, 2Æ5 mmol L)1
MgCl2, 10 lmol L)1 of each deoxyribonucleoside triphosphate
(dNTP), 10 mmol L)1 Tris–HCl pH 8Æ3 and
0Æ2 units Thermus Aquaticus Polymerase (Biotaq;
Clontech).
DNA obtained from the peripheral blood of healthy
donors was employed as a polyclonal negative control.
Amplification of DNA obtained from the Namalwa cell
line [American Type Culture Collection (ATCC), Rockville,
MD, U.S.A.] providing a monoclonal pattern was
1148 O.SERVITJE et al.
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
included as a positive control. Double-distilled water,
instead of DNA solution, was amplified simultaneously,
in order to detect any contamination. After an initial
denaturation for 5 min at 95 C, the PCR reaction was
performed with 34 cycles of 94 C denaturation, 62 C
annealing and 72 C extension, each step for 1 min, and
a final extension step for 5 min at 72 C. PCR products
were analysed and photographed after 12% polyacrylamide
gel electrophoresis under ultraviolet (UV)
illumination and staining with ethidium bromide (EtBr).
In order to validate our PCR technique, in selected
cases, gel slices corresponding to amplified bands were
excised, purified and sequenced using the Sequenase kit
(Perkin-Elmer), according to the specifications of the
manufacturer. This sequence analysis allowed us to
confirm the specificity of the band amplified (data not
shown).
The sensitivity achieved with this PCR technique was
determined by using serial dilutions of DNA from the
Namalwa cell line with DNA from peripheral blood of
healthy donors, leading to the identification of sharp
bands at approximate dilutions of 1 in 100 (1%).
t(14;18)(q32;q21)
One microgram of DNA solution or 10 lL volume of
paraffin samples were added to a final reaction volume of
50 lL, containing 0Æ1 mmol L)1 of each dNTP,
2Æ5 mmol L)1 MgCl2, reaction buffer and 25 pmol of
each primer MBR1 ⁄ JH1 (Calbiochem Oncogene
Research Products). Reaction amplifies the major breakpoint
region (MBR) of the bcl-2 gene. DNA from
nucleated cells of peripheral blood of healthy donors
was employed as the negative control, and DNA
obtained from the Karpas cell line (ATCC), carrying the
MBR form of this translocation, was used as the positive
control. As in the IgH gene rearrangement analysis,
purified water was included as an additional sample, to
detect contamination. Single-step PCR was performed
with 35 cycles of denaturation at 94 C for 1 min,
annealing at 65 C for 1 min and extension at 72 C for
1 min 30 s, followed by an extension step at 72 C for
5 min. PCR products were separated by electrophoresis
in a 1% agarose gel, stained with EtBr and visualized
with a UV transilluminator. Gels were blotted on to a
nylon membrane, hybridized with a c-32P-adenosine
triphosphate-labelled specific oligonucleotide and then
autoradiographed. Simultaneously, PCR amplification
of the bcl-2 non-translocated gene was carried out with
specific primers and used as a control of DNA integrity
and amplification reaction efficiency.13
Results
Clinical data
Clinical data are summarized in Table 1. Twenty-two
patients (20 men and two women) were included. The
mean age was 50 years (range 24–77). The mean
period between the appearance of skin lesions and
diagnosis was 27 months (range 1–96). None of the 22
patients had B-symptoms at presentation and lactate
dehydrogenase and b2-microglobulin levels were
normal in all cases.
The most frequent clinical presentation seen in 16
cases (70%; patients 1–16) was that of asymptomatic,
localized, deep red to violaceous infiltrated plaques,
grouped nodules or tumours, affecting the trunk
(seven cases) or extremities (nine cases). In three of
these patients skin lesions affected more than one
anatomical site. Frequently, skin lesions were surrounded
by an area of diffuse, infiltrated erythema
with small papules on the surface (Fig. 1). Two
patients presented with arciform infiltrated plaques
on the arm and leg.
In four cases (patients 17–20), skin lesions were
located on the head and neck area. In three of these
cases lesions consisted of violaceous nodules and deep,
infiltrated, erythematous plaques, affecting the face
(two cases) and the scalp (one case) (Fig. 2). In the
other patient a large, single tumour appeared on the
temple. This patient had suffered from lymphocytoma
cutis miliaris some years before and was previously
reported by our group.14
Two cases (patients 21 and 22) presented with
widespread skin involvement by violaceous nodules
and tumours as well as deep subcutaneous masses in
one case (Fig. 3). Both patients had a past history of
Hodgkin’s disease; one of them has been previously
reported.15
In most patients the therapy of choice was local
radiotherapy, or surgery when lesions were very
localized, with initial complete remission in all patients
treated and minimal side-effects. In the two patients
with widespread skin lesions total-skin electron beam
irradiation was administered and achieved rapid control
of the skin lesions with good tolerance of the
therapy.
Relapse ⁄ progression
Data regarding relapses are summarized in Table 1. Six
patients (27%) developed skin-limited recurrences
affecting the initial treated site or distant skin sites.
CUTANEOUS LYMPHOMA: MARGINAL ZONE B-CELL LYMPHOMA 1149
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
The time of presentation of the skin recurrences varied
from 6 months to more than 8 years after the first
manifestation. In one patient (patient 22) multiple skin
relapses were observed during a long follow-up of
more than 10 years. All these patients were again
well controlled with local therapies, such as local
Table 1. Clinical characteristics of 22 patients with primary cutaneous marginal zone B-cell lymphoma
Patient
no.
Age
years Sex
Time
prior to
diagnosis
(months) Skin lesions Location
Initial
therapy
Site of
relapse
(months after
initial disease)
Second
therapy
Status
(months of
follow-up)
1 47 M 6 Red infiltrated plaques Buttock RDT CR (6)
2 77 M 18 Grouped violaceous nodules Chest RDT CR (40)
3 25 M 1 Grouped violaceous nodules Lumbar region RDT CR (38)
4 36 M 26 Arciform infiltrated plaque Arm RDT CR (74)
5 45 M 24 Infiltrated plaque with
violaceous tumours
Arm RDT CR (57)
6 70 M 4 Infiltrated plaque with
violaceous tumours
Lumbar
region
RDT Back (19) RDT CR (37)
7 63 M 18 Arciform infiltrated
plaque
Leg RDT Initial
skin site
(104)
RDT CR (156)
8 40 M 12 Infiltrated plaque with
violaceous tumours
Back RDT CR (16)
9 25 M 84 Grouped tumours and
nodules, infiltrated
plaques
Thigh RDT Contralateral
thigh (18)
Surgery CR (72)
10 24 M 14 Infiltrated plaques with
violaceous nodules
Thigh RDT CR (12)
11 52 M 7 Infiltrated plaque with
large tumours
Back RDT CR (8)
12 41 F 4 Grouped violaceous
nodules
Arm Surgery CR (35)
13 70 M 48 Infiltrated plaques
with nodules
Arm,
lumbar
region
RDT Contralateral
arm (40)
None SD (65)
14 66 M 96 Infiltrated plaques
with nodules
Arm,
thigh
RDT CR (9)
15 25 M 9 Grouped erythematous
nodules
Arm,
trunk
RDT BM (13) None LFU
16 52 M 60 Infiltrated plaques
with nodules
Chest RDT Back (14),
bowel (24)
RDT,
CHOP
CR (20)
17 63 F 2 Large violaceous nodules,
deep infiltrated plaques
Retroauricular
bilateral,
face
Surgery,
RDT
Face (12),
subcutaneous
tissue of the
leg and local
lymph nodes
(18)
RDT,
chlorambucil
DL (48)
18 73 M 23 Deep infiltrated plaques,
violaceous nodules
Supralabial,
face
Surgery Face, parotid
gland, cervical
lymph nodes,
BM (8)
Chlorambucil SD (16)
19 52 M 19 Infiltrated plaques,
violaceous nodules
Scalp,
forehead
None Spleen,
lymph nodes,
BM (10)
CHOP CR (38)
20 75 M 72 Large tumour Temple RDT DUC (12)
21 38 M 8 Large infiltrated plaques
with nodules
Disseminated E-beam Skin (6) SD (15)
22 48 M 36 Grouped nodules
and tumours
Disseminated E-beam Skin
relapses
(many)
RDT,
surgery
SD (126)
RDT, radiotherapy; E-beam, total-skin electron beam radiotherapy; BM, bone marrow; CHOP, cyclophosphamide, adriamycin, vincristine, prednisolone;
CR, complete response; SD, stable disease; DL, died of lymphoma; DUC, died from unrelated causes; LFU, lost to follow-up.
1150 O.SERVITJE et al.
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
Figure 1. Clinical presentation in patients
with primary cutaneous marginal zone B-cell
lymphoma on the trunk or extremities
(patients 5 and 10). Note erythema and
nodules surrounding skin tumours.
Figure 2. Clinical appearance of marginal zone B-cell lymphoma
involving the face and scalp (patient 19).
Figure 3. Disseminated lesions seen in two patients with cutaneous
marginal zone B-cell lymphoma (top, patient 22; bottom, patient 21).
CUTANEOUS LYMPHOMA: MARGINAL ZONE B-CELL LYMPHOMA 1151
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
radiotherapy or surgery, and none developed extracutaneous
lymphoma.
In five patients (23%) extracutaneous involvement
was detected during the follow-up. In one of them
(patient 16), skin relapse was accompanied by a largecell
lymphoma of the bowel. One patient (patient 17)
developed massive involvement of the subcutaneous
tissue of the left leg along with local lymph node and
skin recurrence. One patient (patient 18), with initial
skin lesions limited to the facial area, presented with
parotid gland and local lymph node involvement. In
one patient (patient 19), splenic lymphoma of the
marginal zone type and abdominal lymph node
involvement was diagnosed. Bone marrow infiltration
was detected in three cases (patients 15, 18 and 19).
Two of them also showed involvement at other
extracutaneous sites but in one patient (patient 15)
bone marrow infiltration was an isolated feature
without other signs of cutaneous or extracutaneous
disease. The period for extracutaneous presentation
in our patients ranged from 8 to 24 months after
initial diagnosis. These patients were treated with
chemotherapy or combination (radiotherapy plus chemotherapy)
therapy with good control of the disease
except in patient 17 in whom disease progressed.
Follow-up
The mean follow-up in our patients was 43 months
(range 6 months)13 years). Fifteen patients (68%)
remain in complete remission. Four patients have
stable residual disease. Only two patients died during
follow-up, one of them of lymphoma progression
4 years after initial diagnosis (patient 17) and the
other due to unrelated causes (patient 20).
Histopathology
Data regarding histopathology are summarized in
Table 2. Microscopically, MZCLs were characterized
by dense lymphocytic infiltrates mostly distributed on
the reticular dermis and often extending to the subcutis
(bottom heavy). There was no epidermotropism and a
Grenz zone could be observed in all cases. In seven
cases the infiltrates were limited to the superficial
reticular dermis in a so-called top-heavy pattern. In
most cases (68%) a nodular pattern of infiltration was
observed with the presence of unusual rounded or
elongated structures (Fig. 4A). In seven cases the
pattern of infiltration was rather diffuse. Adnexal
involvement was seen in all but two cases. It consisted
Table 2. Histopathological and molecular genetics findings in 22 patients with primary cutaneous marginal zone B-cell lymphoma
Patient
no.
Distribution
of the
infiltrate
Pattern of
infiltration
Adnexal
involvement
Germinal
centres
Plasma
cells
CD30
blasts T cells
Superficial
light chains
Cytoplasmic
light chains
IgH
rearrangement
Bcl-2
expression t(14;18)
1 TH D E No Yes Yes Yes K K – No –
2 BH N F Yes No No No K NE + No –
3 TH N F Yes No Yes Yes K K + Yes –
4 BH N F + E Yes No Yes Yes NE NE + No –
5 BH N F Yes Yes Yes Yes NE L + Yes –
6 TH D No Yes Yes Yes Yes NE PC + No –
7 TH N F Yes Yes Yes Yes NE K – Yes –
8 BH D F No No No Yes NE K + Yes –
9 BH N No Yes Yes Yes Yes L L + Yes –
10 TH D F Yes Yes Yes Yes L L – Yes –
11 BH D F No Yes Yes Yes K K – Yes –
12 BH N F + E Yes Yes Yes Yes NE PC + Yes –
13 BH D F No Yes Yes Yes NE K + Yes –
14 BH N E No Yes Yes Yes L L + No –
15 TH N F No Yes No Yes NE PC – Yes –
16 BH N E No Yes Yes Yes NE K + Yes –
17 BH N E No Yes Yes Yes NE K + Yes +
18 BH D E No No Yes Yes K NE + Yes –
19 TH N F Yes No Yes Yes K NE – Yes –
20 BH N F Yes No Nv Yes K NE – No –
21 BH N E Yes Yes Yes Yes L L – Yes –
22 BH N F Yes Yes Yes Yes K K + Yes +
TH, top-heavy; BH, bottom-heavy; N, nodular; D, diffuse; E, eccrine; F, follicular; K, kappa; L, lambda; NE, no expression; PC, polyclonal.
1152 O.SERVITJE et al.
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
of infiltration of the eccrine glands and hair follicles by
the neoplastic cells with eventual formation of lymphoepithelial
complexes (Fig. 4B).
MZCLs were characterized cytologically by a polymorphous
infiltrate that included small- to mediumsized
lymphocytes with dense chromatin and pale and
watery cytoplasm (marginal zone cells) with frequent
reniform nuclei (B-monocytoid cells) and a variable
proportion of sparse larger centroblast-like cells (blastoid
cells) intermingled within the infiltrate (Fig. 5B).
In 65% of cases germinal centre-like structures surrounded
by neoplastic cells mimicking incomplete or
aberrant mantle zones were observed. These germinal
centres were considered as reactive and often had
colonization by the neoplastic infiltrate (Fig. 4C). In
most cases there were areas of plasmocytoid differentiation
with the presence of mature plasma cells that
were constantly disposed at the periphery of the
infiltrate and under the dermoepidermal junction
(Fig. 4D). In five cases a moderate degree of tissue
eosinophilia was observed.
Both small- to medium-sized cells (marginal zone,
B-monocytoid cells) and large centroblast-like cells
(blastoid cells) were positive for CD20 (Fig. 5A) and
Figure 4. (A) Nodular infiltrate in primary
cutaneous marginal zone B-cell lymphoma
(MZCL). Note periadnexal distribution (patient
4). (B) Detail of (A). Note marginal zone cell
infiltration of eccrine glands with lymphoepithelial
complex formation (arrowhead). (C)
Example of a reactive germinal centre with
colonization by MZCL (patient 5). (D)
Lymphoplasmocytoid differentiation in MZCL.
Note the presence of mature plasma cells at
the periphery of the infiltrate (arrowheads)
(patient 22). Haematoxylin and eosin; original
magnification: A, · 40; B, · 400; C, · 160;
D, · 400.
CUTANEOUS LYMPHOMA: MARGINAL ZONE B-CELL LYMPHOMA 1153
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
CD79a and negative for CD5 and CD10. Only in one
case (patient 17) were small aggregates of CD10+ cells
scattered in the infiltrate. These cells were considered to
be residual germinal centre cells. In most cases a
variable but usually low number of large CD30+ cells
was present in the infiltrate (Fig. 5B). Neoplastic cells
were bcl-2+ in 16 cases (Fig. 5C) and Ki-67 expression
(studied in 14 cases) was limited to large blastoid cells.
In all cases there was a variable but sometimes
significant number of reactive T cells that expressed
the CD3 and CD45RO phenotype. In some cases
reactive T cells were so numerous that they masked
the neoplastic B cell compartment. The combination of
cytoplasmic and surface light chain expression disclosed
a monotypic pattern in 18 of 22 cases (13 j and
five k) (Fig. 5D). In three cases the pattern was
considered polytopic and in one case light chain
expression could not be demonstrated. There was a
constant presence of large number of S100+ cells
intermingled with the neoplastic infiltrate, as well as
along the basement membrane of the epidermis. This
finding probably reflects the role of a large number of
Figure 5. (A) CD20 (L26) staining of
marginal zone cells (patient 10). (B) CD30
(Ki-1) staining of blastoid cells intermingled
within the infiltrate. This was a constant
finding in cutaneous marginal zone B-cell
lymphoma (MZCL) (patient 10). (C) Bcl-2
staining showing positive marginal zone cells
and negative reactive germinal centre cells
(patient 3). (D) An example of monotypic
expression of superficial j light chain in a case
of MZCL. This case was strictly negative for k
chain (patient 11). Original magnification: A,
· 40; B, · 100; C, · 40; D, · 400.
1154 O.SERVITJE et al.
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
antigen-presenting cells (Langerhans and interdigitating
dendritic cells) eventually involved in the process of
lymphoma transformation, with similarities to Helicobacter
pylori gastric MALT lymphoma.
Molecular genetics
The results are summarized in Table 2. Fourteen of 22
cases of MZCL showed discrete bands of 100–120-bp
lengths suggesting a monoclonal expansion. In the
other eight cases, the presence of a background smear
with no discrete bands indicated that the PCR was
successful in amplifying IgH gene rearrangements from
the normal lymphocytes present in these populations,
but did not amplify any malignant clone (Fig. 6).
As far as the t(14;18) translocation is concerned,
only two cases (patients 17 and 22) showed a discrete
band of 150–350 bp in length after PCR amplification,
indicating the presence of t(14;18)(q32;q21). Sequence
analysis was not performed in this study, but the
positive results in these two cases were confirmed by
hybridization of the bcl-2 ⁄ IgH PCR products with a
specific 32P-labelled probe (5¢-GGC CTG TTT CAA CAC
AGA CCC-3¢) according to the manufacturer’s specifications
16 (Fig. 6). With our technique, translocation
involving the minor cluster region of t(14;18) would
not be detected.
Discussion
Primary cutaneous B-cell lymphomas represent a
unique form of extranodal B-cell malignancies. Most
of these lymphomas have been classified so far either as
follicular centre cell lymphomas or cutaneous immunocytomas.
1,17–19 Recently, a new type of primary
cutaneous B-cell lymphoma that shares the histopathological
characteristics of MZCL of the nodal or
mucosal sites has been described.1–4 However, as only
a few studies have been published, there is still
controversy regarding the incidence and diagnostic
criteria of this entity. We present a series of patients
with the clinical and histopathological features of
primary MZCL of the skin and show detailed data on
clinicopathological correlation, immunophenotype and
molecular genetics.
In our patients there was a large preponderance of
men over women and, as in previous reports,2 patients
were mainly adults aged over 40 years. MZCL was
most commonly localized on the trunk or arms. In
agreement with previous studies, lower leg presentation
was rare.2–4 In a high proportion of cases skin
lesions were quite characteristic, consisting of grouped
nodules or tumours that arose over an area of diffuse or
annular erythema. These changes are similar to those
described by Cerroni et al.,2 and enable this entity to be
suspected clinically in a proportion of cases. A small
number of patients presented with lesions on the head
and neck area. Although initially described in association
with follicular centre cell lymphomas, this
location has also been seen by other authors in MZCL
patients.4,20 When affecting this area, lesions tended to
be hard and deeply infiltrated.
A very infrequent presentation, seen in two patients,
was that of extensive, generalized involvement. Interestingly,
these two patients had previous Hodgkin’s
disease that went into complete remission after
standard treatment. The association of primary cutaneous
B-cell lymphomas with Hodgkin’s disease seems
to be extremely rare. Only one previous case of T-cellrich
B-cell lymphoma has been reported so far.21 These
authors demonstrated a common B-cell origin of both
neoplasias. Interestingly, one of our cases presented a
large number of T cells that we interpreted as reactive.
We believe that these cases share some similarities and
may represent a subtype of skin lymphomas associated
with Hodgkin’s disease.
Figure 6. (A) Polymerase chain reaction (PCR) for IgH gene rearrangement
in cutaneous marginal zone B-cell lymphoma. Lanes C3,
C6 and C7 correspond to positive clonal cases showing discrete bands;
negative cases show only smear. pb, peripheral blood from healthy
donors showing polyclonal pattern; Nam, positive control (Namalwa
cell line); w, water. (B) PCR for t(14;18)(q32;q21). Arrow indicate
patients 17 and 22, respectively, carrying the t(14;18)(q32;q21)
translocation; K, DNA extracted from the Karpas cell line used as
positive control. (C) Southern blot analysis with a specific probe
confirmed the positive results.
CUTANEOUS LYMPHOMA: MARGINAL ZONE B-CELL LYMPHOMA 1155
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
Cutaneous MZCLs are characterized histologically
by a polymorphous lymphocyte proliferation mainly
composed of marginal zone B cells. These cells have
specific cytological features and immunophenotype,
being CD20+, CD79a+, CD5– and CD10–. In agreement
with previous reports, we observed that the
lymphocytic infiltrates were arranged mostly in nodular
aggregates that were frequently disposed around
the follicular and eccrine structures.2,20 Furthermore,
some of our cases presented lymphoepithelial complexes
in relationship with eccrine epithelia, similar to
that observed in mucosal MALT lymphomas. As has
been previously reported,2,4,22 we frequently observed
a variable number of blastoid, monocytoid cells
intermingled within the infiltrate. These cells constantly
expressed the CD30 antigen in addition to
B-cell markers. This finding has not been reported so
far, and we believe it supports the hypothesis that
blastoid cells represent activated monocytoid neoplastic
B cells. Another characteristic of MZCL is the presence
of areas of lymphoplasmocytoid differentiation. In
some cases mature plasma cells that always showed
monotypic light chain expression were observed. In
accordance with previous observations,2,20 these plasma
cells tended to be disposed at the periphery of the
infiltrate and under the epidermis. These findings make
differentiation from what has been called immunocytoma
very difficult, and we believe this entity may be
part of the spectrum of MZCL.23 In fact, according to
recently published criteria, the WHO recognizes only
as lymphoplasmocytic lymphoma (immunocytoma)
those cases with usually systemic involvement and
clinical features of Waldenstro¨m macroglobulinaemia.
Most cases with lung, gastrointestinal tract or skin
involvement are better diagnosed as MALT lymphomas,
where a prominent plasmocytoid differentiation is
frequently seen.10
MZCLs of the skin are also remarkable for the
presence of germinal centre cell structures in about
50% of cases. These structures are considered to be
reactive and are often infiltrated by the neoplastic
marginal zone cells.2,4,20,24 Special attention must be
taken in differentiating MZCLs with reactive germinal
centres from follicular centre cell lymphomas. In our
hands, reactive germinal centre cells appear to be
negative for bcl-2 expression (data not shown) and
polytopic or negative for the expression of light chains.2
In addition, follicular centre cell lymphomas usually
express CD10 antigen.24 All our MZCL cases were
negative for CD10. These data are in agreement with
those recently reported by de Leval et al.25 and strongly
suggest that CD10 negativity should be included as a
diagnostic criterion for MZCL of the skin.
A frequent feature of cutaneous MZCL is the presence
of reactive T cells.2,3 These T cells had a mature
phenotype and varied from one biopsy to another in the
same patient, but in some cases they were so numerous
that they masked the neoplastic population.
A major point of controversy is the differentiation of
MZCL from lymphocytoma cutis on histopathological
grounds alone. B-cell rearrangement analysis detected
clonality in 14 of 22 of our cases. It is unclear why this
PCR method failed to detect a predominant clone
in some of our MZCL patients. Somatic mutations in
critical nucleotides of the VH gene are frequent in
marginal zone lymphomas and may prevent annealing
of oligonucleotide primers. Deletion in nucleotides from
the 3¢ end of the VH gene or the 5¢ end of the JH gene
can also prevent primer hybridization and amplification.
Finally, unconventional rearrangements or chromosomal
translocations involving the IgH locus may
also result in failure of PCR amplification.26 However,
with the combination of light chain immunoglobulin
expression and B-cell rearrangement we were able to
detect clonality in all our cases.
We believe that this approach can be extremely
useful in differentiating MZCL from a reactive process.
However, in one of our cases, MZCL developed from
lymphocytoma cutis, suggesting that this entity can
evolve from a reactive polyclonal lymphoproliferation.
We feel that cases showing a morphological and
immunophenotypic pattern of MZCL, even when light
chain restriction cannot be demonstrated, should be
managed as a lymphoma, assuming that only a ‘wait
and see’ procedure will be undertaken and not an
aggressive approach.
Bcl-2 mutation appears to be a highly unusual
finding in cutaneous B-cell lymphomas. In a very
recent report, Child et al.27 suggested that t(14;18) is
only present in those cases with systemic B-cell
lymphoma from the beginning. This mutation was
demonstrated in only two of our patients, one of whom
died as a consequence of lymphoma progression
(patient 17). In this patient extracutaneous involvement
began after 18 months of skin-limited disease
and extensive staging procedures, including CT scan and
bone marrow biopsy, gave normal results. However,
the possibility of occult systemic lymphoma from the
outset cannot be excluded in this case. The other
patient (patient 22) has had multiple skin relapses but
is stable after more than 10 years of follow-up. This
patient had a history of Hodgkin’s disease 7 years
1156 O.SERVITJE et al.
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
before the appearance of skin lymphoma. Hodgkin’s
disease was diagnosed on the basis of unequivocal
clinical, radiological and histological findings and was
classified as clinicopathological stage IIA. After standard
Hodgkin’s disease therapy the patient went into
complete remission and no signs of either Hodgkin’s
disease or systemic lymphoma have been seen so far
after repeated complete staging. Recently, a common Bcell
origin was demonstrated in a patient with cutaneous
B-cell lymphoma and a past history of Hodgkin’s
disease.21 In our opinion, this is very relevant to
patient 22, suggesting a particular clinical situation for
such patients. If our case also represents a common
B-cell origin or not is unclear; extensive microdissection
analysis would be necessary for confirmation.
However, t(14;18) would also be an uncommon
finding in this situation. The t(14;18) translocation is
a very common genetic alteration in follicular centre
cell lymphomas but has also been described in other
lymphoma types, including MZCL of mucosal sites.8
Whether our cases represent some of these exceptions
or true follicular centre cell lymphomas is difficult to
assess. However, both cases totally fulfilled the histopathological
criteria for MZCL, suggesting that more
investigation is needed in order to assess the specificity
of these criteria when applied to cutaneous lymphomas.
The therapeutic approach in our patients was
dependent on the extent of the skin lesions. In most
patients local radiotherapy was the treatment of
choice. The clinical complete response ratio with this
approach was 100%. However, about 35% of patients
had local recurrences at treated or distant skin sites.
These results are in agreement with those reported by
other authors,2 and indicate that, although benign in
prognosis, MZCL has a high potential for local
recurrences. In all our patients the histopathological
features and immunophenotype of skin recurrences
were similar to those of the initial skin lesions (data
not shown). Some of our patients had recurrences a
long time after the initial skin lesions. In particular,
patient 22 has had continuous recurrences over a
10-year period without any sign of extracutaneous
involvement or high-grade transformation. This fact
confirms the low-grade nature of most of these
lymphomas.4,20,22,24
Total-skin electron beam radiotherapy is an
uncommon approach in low-grade cutaneous lymphoma.
However, in two patients we decided to employ
this technique due to the extensive involvement.
Although skin recurrence developed early in both
patients, the disease appeared to follow a more benign
course, with limited involvement in both patients.
Five of 22 patients (23%) developed extracutaneous
lymphoma during the follow-up, affecting subcutaneous
tissue, lymph nodes, salivary glands, gastrointestinal
tract, spleen and bone marrow. In two patients
there was large cell transformation, and one died of
lymphoma. The other patient responded well to
chemotherapy, suggesting the good clinical behaviour
of this entity. One of our patients had isolated bone
marrow involvement without any sign of deterioration.
Unfortunately, this patient was lost to follow-up. Three
of four patients with head and neck presentation
developed extracutaneous disease.
Data on extracutaneous involvement in MZCL are
controversial. Although some authors reported that
these patients never develop systemic disease,2,4,24
other authors have observed an incidence of extracutaneous
involvement similar to that in our series;
Grønbæk et al.22 observed extracutaneous disease in
four of 12 patients with primary cutaneous MZCL.
These patients had lymph node, liver and spleen
involvement. Two of these patients presented large cell
transformation and both died. De la Fouchardie`re et al.3
reported nine cases of MZCL of the skin. Although two
patients did not fulfil the criteria for primary cutaneous
lymphomas, they observed four patients with extracutaneous
disease involving the lymph nodes, parotid
gland, orbit and bone marrow. Bailey et al.20 reported
extracutaneous disease in two of seven patients with
primary skin disease. These patients developed orbit and
breast involvement. Similar to our cases, in these two
studies all patients with extracutaneous disease had
skin lesions on the head and neck area.
In summary, MZCL appears to be a well recognizable
cutaneous B-cell lymphoma. Although of general good
prognosis, some patients appear to have a tendency to
local or extracutaneous disease. Larger series are
needed in order to identify prognostic parameters.
From our data and review of the literature, head and
neck presentation and large cell transformation may be
associated with a worse prognosis.
Acknowledgments
This work has been supported in part by a grant for the
‘Xarxa Tema`tique de Linfomes Cutani’ from the Generalitat
de Catalunya and CIRIT APSQV 1998 SI. We
thank staff of the Audio-Visual Unit of the Ciutat
Sanita`ria i Universita`ria de Bellvitge for their expert
technical assistance.
CUTANEOUS LYMPHOMA: MARGINAL ZONE B-CELL LYMPHOMA 1157
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
References
1 Willemze R, Kerl H, Sterry W et al. EORTC classification for primary
cutaneous lymphomas: a proposal from the Cutaneous
Lymphoma Study Group of the European Organization for
Research and Treatment of Cancer. Blood 1997; 90: 354–71.
2 Cerroni L, Signoretti S, Hofler G et al. Primary cutaneous marginal
zone B-cell lymphoma: a recently described entity of lowgrade
malignant cutaneous B-cell lymphoma. Am J Surg Pathol
1997; 21: 1307–15.
3 De la Fouchardie`re A, Balme B, Chouvet B et al. Primary cutaneous
marginal zone B-cell lymphoma: a report of 9 cases. J Am
Acad Dermatol 1999; 41: 181–8.
4 Tomaszewski MM, Abbondanzo SL, Lupton GP. Extranodal marginal
zone B-cell lymphoma of the skin. a morphologic and immunophenotypic
study of 11 cases. Am J Dermatopathol 2000; 22:
205–11.
5 Willemze R, Rijlaarsdam JU, Meijer CJ. Are most primary cutaneous
B-cell lymphomas ‘marginal cell lymphomas’? Br J
Dermatol 1995; 133: 950–2.
6 Slater DN. MALT and SALT: the clue to cutaneous B-cell
lymphoproliferative disease. Br J Dermatol 1994; 131: 557–61.
7 Slater DN. Primary cutaneous B-cell lymphomas. Arch Dermatol
1997; 133: 1604–5.
8 Garcı´a-Sanz R, Vargas Montero M, Gonzalez Diaz M et al. Detection
of single and associated lesions of the Bcl-1, Bcl-2, Bcl-6,
c-myc, p53 and p16 genes in B-cell non-Hodgkin lymphomas:
value of molecular analysis for a better assignment of the histologic
subtype. Haematologica 1998; 83: 209–16.
9 Schmid U, Eckert F, Griesser H et al. Cutaneous follicular lymphoid
hyperplasia with monotypic plasma cells. A clinicopathologic
study of 18 patients. Am J Surg Pathol 1995; 19: 12–20.
10 Jaffe ES, Harris NL, Stein H, Vardiman JW. World Health Organization
Classification of Tumours. Pathology and Genetics of Tumours
of Haematopoietic and Lymphoid Tissues. Lyon: IRCC Press, 2001.
11 Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning: a
Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor
Laboratory, 1989.
12 Potter M, Steward C, Maitland N, Oakhill A. Detection of clonality
in childhood B-lineage acute lymphoblastic leukemia by the
polymerase chain reaction. Leukemia 1992; 6: 289–94.
13 Miettinen M, Lasota J. Polymerase chain reaction based gene
rearrangement studies in the diagnosis of follicular lymphoma—
performance in formaldehyde-fixed tissue and application
in clinical problem cases. Pathol Res Pract 1997; 193: 9–19.
14 Moreno A, Curco N, Serrano T et al. Disseminated, miliarial type
lymphocytoma cutis. A report of two cases. Acta Derm Venereol
(Stockh) 1991; 71: 334–6.
15 Servitje O, Marti RM, Estrach T et al. Occurrence of Hodgkin’s
disease and cutaneous B-cell lymphoma in the same patient: a
report of two cases. Eur J Dermatol 2000; 10: 43–6.
16 Galoin S, al-Saati T, Schlaifer D et al. Oligonucleotide clonospecific
probes directed against the junctional sequence of t(14;18): a
new tool for the assessment of minimal residual disease in follicular
lymphomas. Br J Haematol 1996; 94: 676–84.
17 Santucci M, Pimpinelli N, Arganini L. Primary cutaneous B-cell
lymphoma: a unique type of low-grade lymphoma. Clinicopathologic
and immunologic study of 83 cases. Cancer 1991; 67:
2311–26.
18 Willemze R, Meijer CJ, Sentis HJ et al. Primary cutaneous large
cell lymphomas of follicular center cell origin. A clinical follow-up
study of nineteen patients. J Am Acad Dermatol 1987; 16: 518–
26.
19 Rijlaarsdam JU, van der Putte SC, Berti E et al. Cutaneous
immunocytomas; a clinicopathologic study of 26 cases. Histopathology
1993; 23: 117–25.
20 Bailey EM, Ferry JA, Harris NL et al. Marginal zone lymphoma
(low-grade B-cell lymphoma of mucosa-associated lymphoid tissue
type) of skin and subcutaneous tissue: a study of 15 patients.
Am J Surg Pathol 1996; 20: 1011–23.
21 Bra¨uninger A, Hansmann ML, Strickler JG et al. Identification of
common germinal-center B-cell precursors in two patients with
both Hodgkin’s disease and non-Hodgkin’s lymphoma. N Engl J
Med 1999; 340: 1239–47.
22 Grønbæk K, Møller PH, Nedergaard T et al. Primary cutaneous
B-cell lymphoma: a clinical, histological, phenotypic and genotypic
study of 21 cases. Br J Dermatol 2000; 142: 913–23.
23 Duncan LM, LeBoit PE. Are primary cutaneous immunocytoma
and marginal zone lymphoma the same disease? Am J Surg Pathol
1997; 21: 1368–71.
24 Yang B, Tubbs RR, Finn W et al. Clinicopathologic reassessment
of primary cutaneous B-cell lymphomas with immunophenotypic
and molecular genetic characterization. Am J Surg Pathol 2000;
24: 694–702.
25 de Leval L, Harris NL, Longtime J et al. Cutaneous B-cell
lymphomas of follicular and marginal zone types. Use of Bcl-6,
CD10, Bcl-2 and CD21 in differential diagnosis and classification.
Am J Surg Pathol 2001; 25: 732–41.
26 Spencer J, Perry M, Dunn-Walters D. Human marginal-zone B
cells. Immunol Today 1998; 19: 421–6.
27 Child FJ, Russell-Jones R, Woolford AJ et al. Absence of t(14;18)
chromosomal translocation in primary cutaneous B-cell lymphoma.
Br J Dermatol 2001; 144: 735–44.
1158 O.SERVITJE et al.
2002 British Association of Dermatologists, British Journal of Dermatology, 147, 1147–1158
No comments:
Post a Comment