Hematopathology / INFLAMMATORY PSEUDOTUMOR OF THE SPLEEN AND A CLONAL EBV GENOME
Inflammatory Pseudotumor of the Spleen Associated With
a Clonal Epstein-Barr Virus Genome
Case Report and Review of the Literature
Jason T. Lewis, MD,1 Robyn L. Gaffney, MD,1* Mary B. Casey, MD,1 Michael A. Farrell, MD,2
William G. Morice, MD, PhD,1 and William R. Macon, MD1
Key Words: Inflammatory pseudotumor; Spleen; Clonal Epstein-Barr virus
DOI: 10.1309/BUWNMG5RV4D09YYH
Abstract
We report a case of an inflammatory pseudotumor
(IPT) of the spleen occurring in an 81-year-old woman
with a history of a monoclonal gammopathy of
undetermined significance. Eighteen-month follow-up
after splenectomy demonstrated no tumor recurrence
or progression of underlying plasma cell disease.
Histologic examination of the tumor demonstrated a
polymorphic population of inflammatory and
epithelioid and spindle cells. Immunophenotyping
showed large numbers of T cells, B cells, and
polyclonal plasma cells. The epithelioid and spindle
cells were positive for vimentin and CD68 but lacked
expression of follicular dendritic cell markers and
actin. Epstein-Barr virus (EBV) genome was identified
in the epithelioid and spindle cell population by in situ
hybridization using probes specific for EBV-encoded
RNAs (EBER1 and EBER2). Southern blot analysis of
digested DNA extracted from the tumor using an EBVspecific probe (XhoI) demonstrated the presence of a
single high-intensity band, indicative of EBV
monoclonality. While there have been 2 previous
reports of hepatic IPTs containing a monoclonal
population of EBV-infected tumor cells, this is the first
report of such an association occurring in the spleen.
The presence of clonal EBV DNA suggests some splenic
IPTs may be true neoplasms.
56
56
Am J Clin Pathol 2003;120:56-61
DOI: 10.1309/BUWNMG5RV4D09YYH
Inflammatory pseudotumor (IPT) is a mass lesion of
unknown cause found in numerous sites, such as the orbit,1
respiratory tract,2 bladder,3 lymph nodes,4-8 liver,4,5,9-11 and
spleen.4,5,7,12-14 The pathogenesis of IPTs is a topic of debate
and includes infectious,4,9,15 autoimmune,14 and neoplastic
processes.10
Cotelingam and Jaffe12 described the first splenic IPT
and believed it was nonneoplastic with an inflammatory and
reparative microscopic appearance. As in other locations,
IPTs of the liver and spleen are composed of a polymorphic
population of inflammatory cells, including T cells, B cells,
and plasma cells accompanied by plump spindle cells.
However, there are differences between splenic and hepatic
IPTs and those arising in other sites. For example, EpsteinBarr virus (EBV) involves splenic and hepatic IPTs more
commonly than IPTs of lymph nodes, and it is the spindle
cell population that is infected in the former.4,5 The exact
nature of these spindle cells has not been determined, but
vimentin and actin are the most common markers
expressed.4,11,16,17 This has led some investigators to postulate
that the spindle cells are myofibroblastic.16,17 In addition,
some IPTs express follicular dendritic cell (FDC) markers,
such as CD21 and CD35.4,10,18 Interestingly, 2 hepatic cases
with the FDC immunophenotype possessed a monoclonal
EBV genome.10,19 Therefore, some IPTs that arise within the
liver may be EBV-associated FDC neoplasms.18
We describe a patient with a splenic IPT that harbored a
monoclonal EBV genome, apparently localized to the FDC
marker– and actin-negative spindle cell population. This
provides additional evidence that a subset of IPTs, and not
just those that express FDC markers, is a clonal expansion of
an EBV-infected cell.
© American Society for Clinical Pathology
Hematopathology / CASE REPORT
Case Report
An 81-year-old woman, who had a 10-year history of a
stable IgG, λ monoclonal gammopathy of undetermined
significance, sought care because of epigastric pain of
approximately 1 month’s duration. There were no associated
fevers, chills, or weight loss. Her physical examination findings were within normal limits. However, a computed
tomography scan of the chest and upper abdomen,
performed in the workup for her abdominal discomfort,
demonstrated a 5-cm, well-circumscribed, heterogeneous
mass within the spleen that had a central area of low attenuation ❚Image 1❚. Further ultrasonic characterization of the
mass yielded similar findings. A subsequent nuclear medicine technetium 99–labeled RBC study excluded a hemangioma, which is the most common primary benign splenic
tumor.20 An ultrasound-guided biopsy then was performed.
The biopsy cores demonstrated a reactive plasma cell proliferation based on polyclonal immunoglobulin light chain
staining by paraffin immunohistochemical analysis. The
patient decided to undergo follow-up for the mass with serial
computed tomography scans, and these showed a slight
increase in its size during a 7-month interval. Therefore,
open splenectomy was performed without complications.
She has had no recurrence or evidence of a lymphoproliferative disorder in the 18 months after the splenectomy.
Materials and Methods
Histologic Examination and Immunophenotyping
H&E-stained sections (4 µm) of the spleen were evaluated. Initial immunophenotyping of the fresh splenic mass
was performed, as previously described,21 by multicolor flow
cytometry using monoclonal antibodies directed against the
following antigens: CD3, CD5, CD10, CD11c, CD14,
CD16, CD19, CD20, CD22, CD23, CD38, CD45, CD56,
and κ and λ immunoglobulin light chains (Becton
Dickinson, San Jose, CA).
Immunoperoxidase studies then were performed on 4µm sections of paraffin-embedded splenic tissue using the
following antibodies, vendors, pretreatments (if any), and
dilutions: CD20 (clone L26, DAKO, Carpinteria, CA; steam
in citrate buffer; 1:60), CD21 (clone 1F8, DAKO; protease;
1:5), CD35 (clone Ber-MAC-DRC, DAKO; protease; 1:5),
vimentin (clone 3B4, DAKO; protease; 1:500), CD45RO
(clone UCHL-1, DAKO; steam in citrate buffer; 1:60), CD68
(clone KP-1, DAKO; protease; 1:200), S-100 (polyclonal,
DAKO; 1:800), desmin (clone DER11, DAKO; protease;
1:100), smooth muscle actin (SMA; clone 1A4, DAKO;
1:150), lysozyme (polyclonal, DAKO; protease; 1:1,000),
❚Image 1❚ Axial computed tomography image with
intravenous and oral contrast at the level of the upper
abdomen. There is a solid mass in the spleen (arrow) with a
central area of low attenuation, probably representing
necrosis. Incidentally noted is a large, simple hepatic cyst
(arrowhead).
CD23 (clone BU38, The Binding Site, San Diego, CA;
steam in EDTA buffer; 1:20), and clusterin (clone 41D,
Upstate Biotechnology, Lake Placid, NY; steam in EDTA
buffer; 1:200). These studies were performed, as described
previously,22 using the labeled streptavidin-biotin detection
chemistry system on a Ventana ES autostainer (Ventana,
Tucson, AZ) for all antibodies except CD45RO, which
required manual staining. 3-amino, 9-ethyl-carbazole was the
chromogen for all reactions.
In Situ Hybridization
In situ hybridization for the detection of EBV was
performed on formalin-fixed, paraffin-embedded spleen
tissue using the procedure described by Chang et al.23 A
digoxigenin-labeled oligonucleotide probe cocktail directed
against EBV-encoded nuclear RNAs (EBER1 and EBER2)
was detected by sheep antidigoxigenin coupled to alkaline
phosphatase (1:100 dilution, Roche Diagnostics, Indianapolis, IN). The hybridization products were visualized
after developing in 5-bromo-4-chloro-3-indolyl phosphate/
nitroblue tetrazolium plus levamisole solution (10 to 40
minutes), followed by counterstaining with 0.1% nuclear fast
red (3 minutes), dehydration, and coverslipping. Doublelabeling immunohistochemical and in situ hybridization
studies also were performed on paraffin-embedded tissues
using antibodies directed against CD20, CD21, CD23,
CD35, CD45RO, CD68, clusterin, S-100, vimentin, SMA,
and desmin in an attempt to further define the nature of the
Am J Clin Pathol 2003;120:56-61
© American Society for Clinical Pathology
57
DOI: 10.1309/BUWNMG5RV4D09YYH
57
57
Lewis et al / INFLAMMATORY PSEUDOTUMOR OF THE SPLEEN AND A CLONAL EBV GENOME
endonuclease BamHI, resolved by agarose gel electrophoresis, and transferred to a nylon membrane that then
was probed with a radiolabeled DNA probe specific for EBV
terminal repeat sequences (XhoI). Hybridized bands were
detected by autoradiography. Polymerase chain reaction and
Southern blot studies to assess for clonal T-cell receptor and
immunoglobulin heavy chain (IgH) gene rearrangements
were performed as outlined by Lust.25
Results
❚Image 2❚ Cross-section of spleen demonstrating a fleshy
mass that replaces much of the splenic parenchyma. Notice
the central area of hemorrhage and necrosis.
EBV-infected cells. In situ hybridization studies were
performed first, without counterstain, followed by
immunoperoxidase studies (as described in the preceding
section).
Molecular Analysis
Southern blot analysis was performed using a previously
described technique.24 Briefly, following extraction and
purification, lesional DNA was digested with the restriction
A
The resected 200-gm spleen contained a solitary, wellcircumscribed 6.5-cm mass. The cut surface of the tumor
was tan and fleshy with areas of hemorrhage and central
necrosis ❚Image 2❚ . The remainder of the spleen was
compressed but otherwise grossly unremarkable. Microscopically, incomplete fibrous bands separated portions of the
lesion from the residual normal spleen ❚Image 3A❚. Most of
the tumor was hypercellular but punctuated in areas by zones
of sclerosis and necrosis. The cellular regions were
composed of a polymorphic population of inflammatory
cells, including plasma cells, lymphocytes, polymorphonuclear leukocytes, and histiocytes, some forming nonnecrotizing granulomas. There were other cells present that varied
from epithelioid to spindle shaped with large, ovoid nuclei
❚Image 3B❚.
Flow cytometric immunophenotyping of the tumor
revealed large numbers of T cells associated with a smaller
population of polyclonal B cells. There also was a large
population of polyclonal plasma cells, confirming the
B
❚Image 3❚ Spleen, inflammatory pseudotumor (IPT). A, The normal splenic tissue (left) is well demarcated from the IPT (right) by
dense fibrous tissue (H&E, ×100). B, Occasional spindle cells (arrow) are present within a background of numerous plasma
cells, small lymphocytes, and histiocytes (H&E, ×200).
58
58
Am J Clin Pathol 2003;120:56-61
DOI: 10.1309/BUWNMG5RV4D09YYH
© American Society for Clinical Pathology
Hematopathology / CASE REPORT
❚Image 4❚ In situ hybridization for Epstein-Barr virus–encoded
nuclear RNA1 and RNA2 shows strong and diffuse staining
in the epithelioid and spindle cells (×200).
previous paraffin immunohistochemical results on the
splenic core biopsy specimen. Paraffin immunoperoxidase
stains confirmed a predominance of CD45RO+ T cells in the
splenic tumor accompanied by scattered CD20+ B cells. The
epithelioid and spindle cells were positive for vimentin and
demonstrated weak staining for CD68. There was equivocal
lysozyme positivity in these cells and no staining with the
remaining antibodies used.
In situ hybridization revealed EBV RNA (EBER1 and
EBER2) within numerous epithelioid and spindle cells
throughout the tumor ❚Image 4❚. The inflammatory cells
were consistently EBER-negative. The EBV RNA–positive
cells were focally positive for vimentin and CD68 but negative for CD20, CD21, CD23, CD35, CD45RO, clusterin, S100, SMA, and desmin. A single EBV-specific band was
detected by Southern blot analysis, indicating the presence of
a monoclonal population of EBV-infected cells ❚Image 5❚.
Additional molecular genetic studies (polymerase chain
reaction and Southern blot analysis) failed to reveal evidence
of clonal T-cell receptor or IgH gene rearrangements.
includes lymphocytes, plasma cells, and histiocytes. The
majority of lymphocytes are T cells, with fewer numbers of
B cells.4,10,14 In addition, the plasma cells have been shown
to be polyclonal.3,4,6,10,11,13
Despite histologic similarities, studies have demonstrated site-associated differences among IPTs. For example,
while EBV genomes have been identified in IPTs of the
lymph nodes, spleen, and liver,4,5,10,16 there is a much greater
incidence of EBV positivity in splenic and hepatic IPTs
compared with those occurring in lymph nodes.4,5 Not only
does the frequency of EBV infection vary depending on site,
but differences in the infected cell types also are apparent.
Among EBV-positive hepatic and splenic IPTs, such as our
case, it is the spindle cell population that is infected, whereas
the virus resides in the lymphocytes of EBV-positive nodal
cases.4,5,10,18 Furthermore, as in our case, the concentration
of EBV within IPTs is much higher than in surrounding
tissue, providing indirect evidence that EBV infection has an
important role in the pathogenesis of some IPTs.4,5,10,18
The nature of the spindle cell population in IPTs has
been debated. Menke et al26 determined that many of the
spindle-shaped fibroblast-like cells in nodal IPTs were probably monocyte derived because of their expression of CD68
or HAM-56. Neuhauser et al16 examined 12 splenic IPTs and
found that SMA was the most commonly expressed marker
(10/11 cases), followed by vimentin (7/10 cases). They also
identified 2 of 10 cases that were positive for CD21, a
Discussion
1
IPT historically has been considered a benign, tumefactive process of unknown origin. It has been found in multiple
sites, including the orbit,1 respiratory tract,2 bladder,3 lymph
nodes,4-8 liver,4,5,9-11 and spleen.4,5,7,12-14 Regardless of location, virtually all IPTs share certain histologic features. They
are composed, in varying proportions, of plump spindle cells
and a polymorphic inflammatory cell infiltrate, which
2
3
4
❚Image 5❚ Epstein-Barr virus (EBV) Southern blot. Lane 1
contains DNA from a negative control lacking EBV DNA.
Lanes 2 and 4 contain positive control DNA from a clonal
EBV-infected neoplasm diluted 1 to 10 and 1 to 1 with
negative control DNA, respectively. Lane 3 contains the
patient’s DNA. Note the presence of a single high-intensity
band (arrow) indicative of EBV monoclonality.
Am J Clin Pathol 2003;120:56-61
© American Society for Clinical Pathology
59
DOI: 10.1309/BUWNMG5RV4D09YYH
59
59
Lewis et al / INFLAMMATORY PSEUDOTUMOR OF THE SPLEEN AND A CLONAL EBV GENOME
follicular dendritic cell marker.16 This supported the findings
of Arber et al4 who, by using double-labeling studies, were
the first to report a splenic IPT that contained EBER1-positive spindle cells that also expressed FDC markers. Selves et
al10 subsequently reported a hepatic IPT that contained
spindle cells positive for EBV and FDC markers.
The possible role of follicular dendritic cells in IPTs was
further suggested by Cheuk et al.18 They described 11 FDC
tumors that had the histologic appearance of an IPT, such as a
spindle cell background and a predominant inflammatory
component.18 All tumors showed positive staining for at least
1 FDC marker, and all showed labeling for EBV by in situ
hybridization. They postulated that some IPTs arising in the
liver or spleen are actually EBV-associated FDC neoplasms
and suggested that the term inflammatory pseudotumor–like
follicular dendritic cell tumor might be a more appropriate
designation for this group.27 However, they noted that some
IPTs lacked EBV and FDC markers.27 Other investigators
also have described splenic and hepatic IPTs that were negative for FDC markers, whether or not positive for EBV.4,5,16 In
agreement with this apparent phenotypic variability, the EBVinfected epithelioid and spindle cells in our case were positive
for vimentin and CD68 but negative for multiple other
markers, including CD21, CD23, CD35, SMA, and desmin.
It has been theorized that follicular dendritic cells are of
mesenchymal (stromal) cell origin,28 as are fibroblasts and
myofibroblasts.29 Because of the variable immunophenotypic patterns seen in hepatic and splenic IPTs, it is possible
that they arise from a common mesenchymal cell that is
capable of differentiating along different pathways ❚Figure
1❚. The majority would develop a myofibroblastic phenotype
and be positive for SMA and vimentin. Others would acquire
FDC characteristics and express CD21 and CD35. The final
category, represented by our case, would remain less differentiated and express only mesenchymal and histiocytic
markers, such as vimentin, CD68, and lysozyme.
Mesenchymal (stromal) cell
+/–
Epstein-Barr
virus
Regardless of the exact nature of the spindle cells, we
regard the presence of clonal EBV DNA as the most important finding in the IPT case reported herein. There have been
2 previous reports of clonal EBV infections in hepatic
tumors with the histologic appearance of IPTs.10,19 However,
both of those expressed the FDC markers CD21 and CD35.
Our case, therefore, is the first splenic IPT with clonal EBV
that lacked FDC markers. The presence of a clonal EBV
genome suggests that the epithelioid and spindle cells within
the tumor arose from a single infected cell, with concurrent
expansion of that cell’s EBV genome.24 This phenomenon
has been described in other neoplastic processes, such as
Burkitt lymphoma30 and nasopharyngeal carcinoma.24 Thus,
the finding of clonal EBV DNA in a splenic IPT adds further
support for their being true neoplasms, possibly of
mesenchymal origin, rather than reactive or reparative
processes. A similar concept was proposed for the FDCs in a
case of hyaline vascular Castleman disease by the recent
demonstration of clonal high mobility group protein I-C
(HMGIC) gene rearrangements within CD21+ cells.31
From the Departments of 1Pathology and 2Radiology, Mayo
Clinic, Rochester, MN.
Address reprint requests to Dr Macon: Dept of Pathology,
Mayo Clinic, 200 First St SW, Rochester, MN 55905.
* Dr Gaffney has relocated to the Department of Pathology,
University of New Mexico Health Sciences Center, Albuquerque.
References
1. Blodi C, Gass JDM. Inflammatory pseudotumor of the orbit.
Br J Ophthalmol. 1968;52:79-93.
2. Bahadori M, Liebow AA. Plasma cell granulomas of the lung.
Cancer. 1973;31:191-208.
3. Jones EC, Clement PB, Young RH. Inflammatory
pseudotumor of the urinary bladder: a clinicopathological,
immunohistochemical, ultrastructural, and flow cytometric
study of 13 cases. Am J Surg Pathol. 1993;17:264-274.
4. Arber DA, Onsi WK, Van de Rijn M, et al. Frequent presence
of the Epstein-Barr virus in inflammatory pseudotumor. Hum
Pathol. 1995;26:1093-1098.
5. Arber DA, Weiss LM, Chang KL. Detection of Epstein-Barr
virus in inflammatory pseudotumor. Semin Diagn Pathol.
1998;15:155-160.
6. Facchetti F, De Wolf-Peeters C, De Wever I, et al.
Inflammatory pseudotumor of lymph nodes: immunohistochemical evidence of its fibrohistiocytic nature. Am J Pathol.
1990;137:281-289.
IPT with no specific
cellular differentiation
IPT with myofibroblastic
features (actin+)
IPT with follicular dendritic
cell features (CD21+, CD35+)
❚Figure 1❚ Proposed origin and differentiation pathways for
hepatic and splenic inflammatory pseudotumors (IPTs).
60
60
Am J Clin Pathol 2003;120:56-61
DOI: 10.1309/BUWNMG5RV4D09YYH
7. Kutok JL, Pinkus GS, Dorfman DM, et al. Inflammatory
pseudotumor of lymph node and spleen: an entity biologically
distinct from inflammatory myofibroblastic tumor. Hum
Pathol. 2001;32:1382-1387.
8. Perrone T, De Wolf-Peeters C, Frizzera G. Inflammatory
pseudotumor of lymph nodes: a distinctive pattern of nodal
reaction. Am J Surg Pathol. 1988;12:351-361.
© American Society for Clinical Pathology
Hematopathology / CASE REPORT
9. Chen KT. Inflammatory pseudotumor of the liver. Hum
Pathol. 1984;15:694-696.
10. Selves J, Meggetto F, Brousset P, et al. Inflammatory
pseudotumor of the liver: evidence for follicular dendritic
reticulum cell proliferation associated with clonal EpsteinBarr virus. Am J Surg Pathol. 1996;20:747-753.
11. Shek TWH, Ng IOL, Chan KW. Inflammatory pseudotumor
of the liver: report of four cases and review of the literature.
Am J Surg Pathol. 1993;17:231-238.
12. Cotelingam JD, Jaffe ES. Inflammatory pseudotumor of the
spleen. Am J Surg Pathol. 1984;8:375-380.
13. Monforte-Munoz H, Ro JY, Manning JT, et al. Inflammatory
pseudotumor of the spleen: report of two cases with a review
of the literature. Am J Clin Pathol. 1991;96:491-495.
14. Thomas RM, Jaffe ES, Zarate-Osorno A, et al. Inflammatory
pseudotumor of the spleen: a clinicopathologic and
immunophenotypic study of eight cases. Arch Pathol Lab Med.
1993;117:921-926.
15. Lupovitch A, Chen R, Mishra S. Inflammatory pseudotumor
of the liver: report of the fine needle aspiration cytologic
findings in a case initially misdiagnosed as malignant. Acta
Cytol. 1989;33:259-262.
16. Neuhauser TS, Derringer GA, Thompson LDR, et al. Splenic
inflammatory myofibroblastic tumor (inflammatory
pseudotumor): a clinicopathologic and immunophenotypic
study of 12 cases. Arch Pathol Lab Med. 2001;125:379-385.
17. Coffin CM, Watterson J, Priest JR, et al. Extrapulmonary
inflammatory myofibroblastic tumor (inflammatory
pseudotumor): a clinicopathologic and immunohistochemical
study of 84 cases. Am J Surg Pathol. 1995;19:859-872.
18. Cheuk W, Chan JKC, Shek TWH, et al. Inflammatory
pseudotumor–like follicular dendritic cell tumor: a distinctive
low-grade malignant intra-abdominal neoplasm with
consistent Epstein-Barr virus association. Am J Surg Pathol.
2001;25:721-731.
19. Shek TWH, Ho FCS, Ng IOL, et al. Follicular dendritic cell
tumor of the liver: evidence for an Epstein-Barr virus–related
clonal proliferation of follicular dendritic cells. Am J Surg
Pathol. 1996;20:313-324.
20. Butler JJ. Pathology of the spleen in benign and malignant
conditions. Histopathology. 1983;7:453-474.
21. Hanson CA, Kurtin PJ, Katzmann JA, et al. Immunophenotypic analysis of peripheral blood and bone marrow in the
staging of B-cell malignant lymphoma. Blood. 1999;94:38893896.
22. Kurtin PJ, Hobday KS, Ziesmer S, et al. Demonstration of
distinct antigenic profiles of small B-cell lymphomas by
paraffin section immunohistochemistry. Am J Clin Pathol.
1999;112:319-329.
23. Chang KL, Chen YY, Shibata D, et al. Description of an in
situ hybridization methodology for detection of Epstein-Barr
virus RNA in paraffin-embedded tissues, with a survey of
normal and neoplastic tissues. Diagn Mol Pathol. 1992;1:246255.
24. Raab-Traub N, Flynn K. The structure of the termini of the
Epstein-Barr virus as a marker of clonal cellular proliferation.
Cell. 1986;47:883-889.
25. Lust J. Molecular genetics in lymphoproliferative disorders.
J Clin Lab Anal. 1997;10:359-367.
26. Menke DM, Griesser H, Araujo I, et al. Inflammatory
pseudotumors of lymph node origin show macrophage-derived
spindle cells and lymphocyte-derived cytokine transcripts
without evidence of T-cell receptor gene rearrangements. Am
J Clin Pathol. 1996;105:430-439.
27. Cheuk W, Chan JKC. Authors’ reply [letter]. Am J Surg
Pathol. 2001;25:1559.
28. van Nierop K, de Groot C. Human follicular dendritic cells:
function, origin, and development. Semin Immunol.
2002;14:251-257.
29. Gabbiani G. The cellular derivation and the life span of the
myofibroblast. Pathol Res Pract. 1996;192:708-711.
30. Neri A, Barriga F, Inghirami G, et al. Epstein-Barr virus
infection precedes clonal expansion in Burkitt’s and acquired
immunodeficiency syndrome–associated lymphoma. Blood.
1991;77:1092-1095.
31. Cokelaere K, Debiec-Rychter M, De Wolf-Peeters C, et al.
Hyaline vascular Castleman’s disease with HMGIC
rearrangement in follicular dendritic cells: molecular evidence
of mesenchymal tumorigenesis. Am J Surg Pathol.
2002;26:662-669.
Am J Clin Pathol 2003;120:56-61
© American Society for Clinical Pathology
61
DOI: 10.1309/BUWNMG5RV4D09YYH
61
61