Castleman's disease

* To our knowledge, five cases of Castleman's disease involving only the central nervous system have been reported previously. We report a sixth case, which occurred in a 47-year-old woman with a 3-month history of headaches and a large superior frontal lobe mass on neuroimaging. Excisional biopsy revealed confluent lymphoid nodular areas with multiple well-developed germinal centers surrounded by concentrically layered proliferations of small B lymphocytes typical of Castleman's disease. Ultrastructural study found 100-nm virallike particles within follicular dendritic cells as well as intercellular spaces. These particles were suggestive of a D-type retrovirus. The patient underwent postoperative radiotherapy and was neurologically normal 3 months after surgery.

(Arch Pathol Lab Med. 1998;122:102S1029)

Since the first description of angiofollicular (giant) lymph node hyperplasia, or Castleman's disease (CD), involving the mediastinum in 1956,1 many cases of CD at various nodal and extranodal sites have been reported.2 In addition, the frequent association of CD with acquired immunodeficiency syndrome and Kaposi's sarcoma (KS)56 and the recent discovery of a virus associated with KS, known as human herpesvirus 8 or KS-associated herpesvirus,79 has prompted a search for this virus within CD lesions.10 Five previous cases of CD confined to the central nervous system have been reported,ll-13 although multifocal CD occasionally involves the meninges.l4 We report a sixth case of unifocal intracranial CD in a 47-year-old immunocompetent woman with a 3-month history of headaches. Ultrastructural study of the present case demonstrated 100-nm nonenveloped viral particles in the cytoplasm of follicular dendritic cells within the tumor that were most consistent with retrovirus.

REPORT OF A CASE

A 47-year-old woman with a 3-month history of progressive frontal headache was found to have a large parafalcine frontal lobe mass by radiography. She had no significant past medical illness or evidence of other tumor sites on physical examination and chest and abdominal radiography, and she had normal peripheral blood findings. Neurologic examination demonstrated no cranial nerve or motor deficits. She was seronegative for HIV and had no evidence of immunodeficiency. A computerized tomographic scan revealed a 5.5 x 5 x 2-cm, left frontal convexity/ parafaldne, densely enhancing lesion (Figure 1). There was surrounding edema and mass effect evidenced by midline shift and ventricular effacement. The imaging characteristics of this lesion are classic for meningioma except in 1 feature: the enhancement extends deep into the sulcal spaces below the lesion, suggesting subarachnoid extension, without displacement of gyri.

A bifrontal craniotomy for partial resection of the tumor was performed, as the mass infiltrated around the faldne vasculature. Postoperative recovery was uneventful, and there were no seizures on full anticonvulsant coverage. Final histologic diagnosis was atypical lymphoid infiltrate consistent with Castleman's disease. Postoperative magnetic resonance imaging revealed dural enhancement consistent with residual tumor, for which the patient was given 5 weeks of radiation therapy (total 4500 cGy). Three months after surgery, the patient had no complaints and no abnormalities on neurologic examination; neuroimaging showed no apparent tumor regrowth. PATHOLOGIC FINDINGS The resected fragmented tumor was granular, graywhite, and firm, and measured up to 2.0 cm in thickness. The tumor was attached to the subdural surface and infiltrated gyral crests superficially. Microscopically, the lesion was typical of Castlemar's disease, with numerous lymphoid follicles consisting of small atrophic germinal centers, with or without penetrating vascular twigs, surrounded by concentric rings of small lymphocytes (Figure 2). Other follicles were composed of concentric proliferations of spindle cells surrounded by layers of small lymphocytes, giving a "targetoid" or "onion-skin" appearance. The interfollicular spaces contained small lymphocytes, plasma cells, plasmacytoid lymphocytes, histiocytes, and immunoblasts. Basophils and macrophages laden with cell debris and hemosiderin were focally numerous in the lesion, and peripheral foci of coagulative necrosis and neutrophilic infiltrate were noted. Immunohistochemical stains on paraffin-embedded sections confirmed that these were extranodal germinal centers containing follicular dendritic cells (expressing CD21 and CD35 antigens) and B lymphocytes (CD20 positive). The plasma cell population was polyclonal and had a K tolambda light chain ratio of 3-4:1. However, flow cytometry study revealed a population of small lymphocytes expressing monoclonal membrane surface K light chain. T cells were present in small numbers.

Electron microscopic study revealed small numbers of 100-nm viral particles in numerous cells. The primary cell type containing virus appeared to be follicular dendritic cells with abundant cytoplasm containing numerous nonbundled intermediate filaments and multiple elongated cell processes. However, no cell junctions were found. Lymphocytes and plasma cells did not appear to harbor viral particles. The virus appeared as uniform spherical particles with surface spikes, representing cytoplasmic capsids without envelopes (Figure 3). The particles resembled retrovirus type D in size (100 run) and configuration. No nuclear membrane budding or nuclear viral particles were found, although occasional intracytoplasmic vesicles contained virus.

Paraffin sections of B5-fixed tumor were subjected to immunohistochemical viral localization of herpes simplex virus I and II, cytomegalovirus, Epstein-Barr virus (monoclonal antibodies from Dako Corporation, Carpinteria, Calif), HIV (DuPont p24), and human herpesvirus 6 (Virotech, Rockville, Md) with appropriate positive and negative controls by a standard PAP technique.15Analysis with Epstein-Barr episomal RNA for localization of Epstein-Barr virus RNA was negative by a previously published technique.lb Polymerase chain reaction analysis did not detect KS-associated herpesvirus (human herpesvirus 8) in DNA extracted from paraffin-embedded material. This analysis involved use of commercially obtained probes (GIBCO-BRL, Grand Island, NY) and detection using pulse-field gel electrophoresis.10

COMMENT

Clinically, this case has many features similar to those in previously reported cases of intracranial CD (Table). A majority of cases (5/6, including ours) have been reported in women, whereas no gender predominance has been found in CD of the mediastinum and lymph nodes.7 Four of these 6 patients presented with new onset seizures. All demonstrated a mass by radiography that was usually (5 / 6 cases) thought to be consistent with meningioma. This probably reflects the leptomeningeal origin of this extranodal CD process. The age at presentation ranged from 25 to 82 years (median, 55 years; mean, 52 years); a history of trauma, autoimmune disorders, or immunodeficiency syndrome was either not mentioned or was not present in any of the reported cases.

Various pathogenic mechanisms leading to the hyperplastic/ dysplastic /neoplastic lymphoproliferative process of CD, seen as massive reduplication of "normal" lymph node germinal centers, have been proposed. These mechanisms include immunodeficiency, autoimmunity, or response to a chronic low-grade infection.l7 The first 2 mechanisms are suggested by the increased incidence of CD in persons with immunodeficiency states, such as KS and acquired immunodeficiency syndrome, and associated autoimmune phenomena, such as autoimmune cytopenias, temporal arteritis, and the POEMS syndrome (polyneuropathy, organomegaly, endocrinopathy, myeloma, and skin changes).17 Increased expression of CD5^sup -^ B cells in CD compared with normal lymphoid tissue, as reported by Hayakawa and Hardy,ls also favors the autoimmune theory since CD5^sup +^ B cells are implicated in generating autoimmune responses. However, as in the present case, CD often occurs in patients without demonstrated immunodeficiency or autoimmune disorder.

Causation of CD by low-grade chronic infectious processes, on the other hand, is suggested by the histologic characteristics common to the lesions in all patients, regardless of immune competency. The majority of CD specimens resemble reactive changes, with a polyclonal proliferation of lymphocytes and variable numbers of plasma cells. While occasionally these proliferating populations apparently give rise to monoclonal malignant disorders (eg, non-Hodgkin's lymphoma and KS), most patients with CD have a benign course with eventual resolution without therapy. However, multifocal CD frequently has a more aggressive course that is associated with poor outcomes.17 Isolated central nervous system disease behaves in a manner similar to its extracranial unifocal counterparts, with good outcomes seen over the short follow-up periods reported.ll-13

The recent finding of a KS-associated herpesvirus, or human herpesvirus 8, combined with the observation of increased incidence of CD in patients with acquired immunodeficiency virus and KS has prompted a revealing search for that virus in patients with CD. Kaposi's Sarcoma-associated herpesvirus has been identified by polymerase chain reaction in both HIV-related and HIV-negative multicentric extracranial CD, both with and without associated KS.9 The present case is, to the best of our knowledge, the first reported case of CD with ultrastructural evidence of viral particles in association with a central nervous system lesion. No herpesvirus was evident on ultrastructural study or by polymerase chain reaction and class I major histocompatibility antigen study. However, numerous follicular dendritic cells contained 100-nm cytoplasmic viral capsids with surface protrusions and lacking envelope structures, consistent with type D retrovirus. Human herpesvirus 8, human herpesvirus 6, HIV, Epstein-Barr virus, cytomegalovirus, and herpes simplex virus I or II were excluded as infectious components of the CD lesion by in situ hybridization, polymerase chain reaction, and immunohistochemical techniques. Our patient was HIV negative serologically and had normal CD4 levels and no evidence of hematologic abnormalities or history of opportunistic infections.

These findings suggest that a viral infection other than human herpesvirus 8 may stimulate the extranodal reactive lymphoproliferation recognized as CD. The number of infected cells and the productive nature of the infection in this case support the hypothesis that the virus is integral to the lymphoid response present. While previous viral studies on CD focused on multifocal CD in lymph nodes, this patient had a single extranodal site of CD involvement. Additional studies of viral profiles in unifocal CD will be of considerable interest in understanding the mechanisms involved.

References

1. Castleman B, Iverson L, Menendez VP. Localized mediastinal lymph-node hyperplasia resembling thymoma. Cancer 1956;9:822-830.

2. Keller AR, Hochholzer L, Castleman B. Hyaline-vascular and plasma-cell types of giant lymph node hyperplasia of the mediastinum and other locations. Cancer 1972;29:670-683.

3. Weisenburger DD, Nathwani BN, Winberg CD, Rappaport H. Multicentric angio-follicular lymph node hyperplasia: a clinicopathologic study of 16 cases. Hum Pathol.1985;16:162-172.

4. Frizzera G, Banks PM, Massarelli G, Rosai J. A systemic lymphoproliferative disorder with the morphologic features of Castleman's disease: pathological findings in 15 patients. Am I Surg Pathol.1983;7:211-231.

5. Lachant NA, Sun NC, Leong LA, Oseas RS, Prince HE. Multicentric angiofollicular lymph node hyperplasia (Castleman's disease) followed by Kaposi's sarcoma in two homosexual males with the acquired immunodeficiency syndrome (AIDS). Am J Clin Pathol. 1985;83:27-33.

6. Lowenthal DA, Filippa DA, Richardson ME, Bertoni M, Straus Dl. Generalized lymphadenopathy with morphologic features of Castleman's disease in an HIV-positive man. Cancer.1987;60:2454-2458.

7. Chang Y, Cesarman E, Pessin MS, et al. Identification of herpesvirus-like DNA sequences in AIDS-associated Kaposi's sarcoma. Science.1994;266:18651869.

8. Huang YQ, Li II, Kaplan MH, et al. Herpesvirus-like nucleic acid in various forms of Kaposi's sarcoma. Lancet. 1995;345:759-761.

9. Moore PS, Chang Y. Detection of herpesvirus-like DNA sequences in Kaposi's sarcoma in patients with and those without HIV infection. N Engl J Med. 1995;332:1181-1185.

10. Soulier J, Grollet L, Oksenhendler E, et al. Kaposi's sarcoma-associated herpesvirus-like DNA sequences in multicentric Castleman's disease. Blood. 1995;86:1276-1280.

11. Lacombe MI, Poirier J, Caron JP. Intracranial lesion resembling giant lymph node hyperplasia. Clin Pathol.1983;80:721-723. 12. Severson GS, Harrington DS, Weisenburger DD, et al. Castleman's disease of the leptomeninges: report of three cases. J Neurosurg.1988;69:283-286.

13. Gianaris PG, Leestma JE, Cerullo LJ, Butler A. Castleman's disease manifesting in the central nervous system: case report with immunological studies. Neurosurgery.1989;24:608-613.

14. Stanley MW, Frizzera G, Dehner LP. Castleman's disease, plasma-cell type: diagnosis of central nervous system involvement by cerebrospinal fluid cytology. Acta Cytol. 1986;30:481-486.

15. Cornford ME, Said JW, Vinters HV. Immunohistochemical localization of human immunodeficiency virus (HIV) in central nervous system lymphoproliferative disorders of patients with AIDS. Mod Pathol. 1991;4:232-238.

16. Chang KL, Chen YY, Shibata D, Weiss LM. Description of an in situ hybridization methodology for detection of Epstein-Barr virus RNA in paraffin-embedded tissue, with a survey of normal and negative tissues. Diagn Molec Pathol. 1992;1:246-255.

17. Shahidi H, Myers JL, Kvale PA. Castleman's disease. Mayo Clin Proc.1995; 70:969-977.

18. Hayakawa K, Hardy RR. Normal, autoimmune and malignant CDS B cells: the Ly-1 B lineage? Ann Rev Immunol. 1988;6:197-218.

Accepted for publication June 21, 1998.

From the Departments of Pathology (Drs Gulati, Sun, and Cornford) and Radiology (Dr Herman), University of California at Los AngelesHarbor Medical Center, Torrance, Calif, and the Department of Pathology, University of California, Los Angeles (Dr Said).

Reprints: Marcia E. Cornford, MD, PhD, Department of Pathology, Box 12, 1000 W Carson St, Torrance, CA 90509.

Copyright College of American Pathologists Nov 1998
Provided by ProQuest Information and Learning Company. All rights Reserved