| |||||||||||||||||||||||||||||||||||||||
Like HIV-infected patients, recipients of solid-organ allografts are at increased risk for a variety of tumors, presumably because of the immunosuppressive regimens used to prevent graft rejection. Kaposi's sarcoma develops in up to 1 percent of organ-transplant recipients and is highly associated with HHV-8 infection.12,13,14 In contrast, HHV-8 infection is unusual in transplant recipients with other tumors, including non-Hodgkin's lymphoma.15,16,17,18,19,20 HHV-8 infection is also rarely associated with non-Hodgkin's lymphoma in HIV-negative patients, even in southern Italy, where HHV-8 seropositivity is relatively common.3,11,21,22,23,24,25 Only a few HHV-8–associated lymphomas have been reported in HIV-negative patients, nearly all of them primary-effusion lymphomas.26,27
We describe an unusual case of primary-effusion lymphoma associated with HHV-8 infection in an HIV-negative recipient of a cardiac transplant. The primary-effusion lymphoma was preceded by Kaposi's sarcoma and persistent polyclonal plasmacytic infiltrates involving multiple organs, including the cardiac allograft.
Case Report
The patient was a 59-year-old Haitian man who underwent orthotopic heart transplantation in December 1989 because of end-stage congestive heart disease. Laboratory studies performed before transplantation showed polyclonal hypergammaglobulinemia (serum gamma globulin concentration, 3010 mg per deciliter; normal range, 518 to 1200). The results of an enzyme-linked immunosorbent assay for HIV antibodies and a Western blot assay for HIV type 1 antigens, both performed at another laboratory, were reported to be weakly positive in serum samples obtained at the time of transplantation. However, studies performed at our institution, including enzyme-linked immunosorbent assays for serum HIV antibodies, a polymerase-chain-reaction (PCR) assay for HIV sequences in serum, and viral cultures, were all negative. On the basis of these findings, the patient was considered to be HIV-negative.
The patient had no identifiable risk factors for HIV infection, such as homosexual activity, intravenous drug use, or prior blood transfusion. He had married once and was the father of eight children. He had frequently visited Haiti after his emigration to the United States at the age of 32 years.
Gross and microscopical examination of the explanted heart showed findings consistent with the presence of an idiopathic dilated cardiomyopathy. There were prominent perivascular and interstitial infiltrates of eosinophils and lymphocytes — findings suggestive of drug-induced hypersensitivity myocarditis.
In May 1990, a biopsy confirmed the presence of Kaposi's sarcoma in an inguinal lymph node. At this time, the immunosuppressive regimen consisted of azathioprine (25 to 50 mg per day), cyclosporine (a dose resulting in a blood concentration of approximately 200 µg per deciliter), and prednisone (8 mg per day). Kaposi's sarcoma appeared in the skin in October 1991 and remitted with a reduction in the immunosuppressive regimen (blood cyclosporine concentration, approximately 50 µg per deciliter). The absolute CD4 count in October 1990 was 576 cells per cubic millimeter, with follow-up CD4 counts of 480 cells per cubic millimeter in March 1993 and 510 cells per cubic millimeter in April 1995.
Beginning in August 1991 and continuing through 1997, endomyocardial-biopsy specimens revealed persistent interstitial polyclonal lymphoplasmacytic infiltrates of mild-to-moderate intensity, which were considered atypical of the usual findings in endomyocardial-biopsy specimens obtained after transplantation, such as evidence of rejection, endocardial lymphoid infiltrates, evidence of healing of previous biopsy site, or ischemic injury. Immunohistochemical stains showed the presence of approximately equal numbers of plasma cells expressing kappa and lambda immunoglobulin light chains, a finding consistent with the presence of a polyclonal reactive process. Serum gamma globulin concentrations increased further and were persistently elevated throughout the post-transplantation course (range, 3800 to 4300 mg per deciliter). Serum electrophoresis showed minor biclonal kappa M spikes in February 1994 (total IgG, 910 mg per deciliter) and again in October 1997 (total IgG, 790 mg per deciliter), in each instance superimposed on a broad polyclonal IgG peak. Bone marrow biopsy in February 1994 showed only a mild polyclonal plasmacytosis, and the results of urinary protein electrophoresis were unremarkable.
Other conditions that developed in the post-transplantation period included refractory hypertension, chronic renal insufficiency that was presumably due to cyclosporine therapy, and atrial fibrillation.
The patient presented in October 1997 with abdominal pain and dyspnea of several months' duration that was thought to be due to worsening congestive heart failure. The immunosuppressive regimen at this time consisted of cyclosporine (blood concentration, 100 to 200 µg per deciliter) and prednisone (8 mg per day). A chest radiograph showed a prominent left-sided pleural effusion. Thoracentesis revealed a cell-rich effusion (18,550 cells per cubic millimeter) containing a high concentration of lactate dehydrogenase (13,700 IU per liter). The concomitant serum lactate dehydrogenase concentration was also markedly elevated (1526 IU per liter; normal range, 107 to 231). On cytologic examination of the pleural fluid, more than 95 percent of nucleated cells were found to be atypical lymphoid cells with a plasmacytoid appearance. Whole-body computed tomographic scanning confirmed the presence of new bilateral pleural effusions and cardiomegaly that was unchanged from previous examinations, but there was no evidence of tumor masses or lymph-node enlargement. The results of a skeletal survey were normal. The patient declined a bone marrow examination.
On the basis of the cytologic findings and the results of immunophenotypic, genotypic, and ultrastructural studies (described below), a diagnosis of primary-effusion lymphoma was made. Initial treatment consisted of a reduction in the blood cyclosporine concentration to 50 to 100 µg per deciliter, therapeutic thoracentesis, and one cycle of cyclophosphamide, vincristine, and prednisone. Because of rapid reaccumulation of pleural fluid, pleurodesis with bleomycin was performed, followed by a course of cyclophosphamide, doxorubicin, vincristine, and prednisone and two cycles of ifosfamide and etoposide. A partial reduction in the serum lactate dehydrogenase concentration and partial resolution of the effusions were noted, but cytologic examinations of pleural fluid in February and March 1998 showed persistent primary-effusion lymphoma. The patient died of multisystem organ failure and respiratory compromise, with persistent pleural effusions, in April 1998, six months after the diagnosis of lymphoma had been made. An autopsy was declined.
Methods
Cytologic, Immunophenotypic, and Ultrastructural Studies
Cytologic studies were performed with Wright–Giemsa staining. Immunoperoxidase studies were performed on formalin-fixed, paraffin-embedded cell blocks with the use of antibodies to CD30 (Ber-H2), epithelial membrane antigen, CD45 (leukocyte common antigen), the B-cell markers CD20 (L26) and CD79a, the T-cell markers CD45RO (UCHL1) and CD3 (all from Dako, Carpinteria, Calif.), and Epstein–Barr virus (EBV) latent membrane protein 1. In situ hybridization was performed with a ribonucleotide probe for EBV-encoded transcripts (BioGenex, San Ramon, Calif.), according to the manufacturer's instructions. Analysis of cell-surface markers was performed by flow cytometry on fresh samples of pleural fluid, as previously described, with monoclonal antibodies directed against CD3, CD5, CD10, CD13, CD14, CD19, CD20, CD33, CD38, and CD45 (all from Becton Dickinson, San Jose, Calif.) and CDIIb (Immunotech, Marseilles, France).28 For ultrastructural analysis, cells from fresh samples of pleural fluid were fixed first in paraformaldehyde–glutaraldehyde and then in osmium tetroxide, embedded in Epon, sectioned, and stained with uranyl acetate. Structural features were compared with those of primary-effusion-lymphoma–derived cells that have been shown to support the replication of HHV-8.29
For immunofluorescent staining, tumor cells from pleural-effusion fluid were fixed on glass slides in a 1:1 solution of acetone and methanol at –20°C. The cells were incubated with or without immune serum from an HHV-8–infected patient with Kaposi's sarcoma, stained with fluorescein isothiocyanate–conjugated antihuman IgG antibody, and then examined with the use of confocal microscopy. The lymphoma cell lines BC-130 and BCBL-1,31 which are positive for HHV-8, were used as positive controls, and the EBV-infected cell lines IB4 and B95-8, which are negative for HHV-8, were used as negative controls.
Molecular and Karyotypic Analyses
DNA was purified from pleural-effusion samples by proteinase K digestion and phenol extraction. DNA was isolated from sections prepared from archival formalin-fixed, paraffin-embedded tissue by proteinase K digestion followed by purification on QIAamp silica columns (Qiagen, Chatsworth, Calif.). A 233-bp HHV-8 genomic sequence (KS330Bam) was amplified from 1 µg of genomic DNA by a PCR assay with the use of a previously described pair of primers.32 Control primers from the human BCL2 locus were used to confirm that the template DNA could be amplified. To assess the configuration of immunoglobulin heavy-chain genes, Southern blotting was performed with the use of 32P-labeled probes to the JH region (Dako) after the digestion of DNA with BglII or BamHI (New England Biolabs, Beverly, Mass.). The presence of EBV was detected on Southern blots with the use of a BamHI W probe specific for EBV genomic DNA. Cytogenetic analysis was performed on cells in metaphase harvested directly from the pleural fluid. Harvesting of cells, slide making, and trypsin–Giemsa banding were performed as described previously.33
Results
Cytopathological, Genotypic, and Karyotypic Findings
The patient's pleural fluid contained numerous large plasmacytoid cells, most of which had one to three prominent nucleoli and moderate amounts of deeply basophilic cytoplasm (Figure 1). Multinucleated cells and scattered mitotic figures were also seen. Immunohistochemical and flow-cytometric analyses revealed that these cells bound antibodies against the activation markers CD30, epithelial membrane antigen, and CD38 and failed to react with antibodies against B-cell markers (CD10, CD19, CD20, and CD79a), T-cell markers (CD3, CD5, and CD45RO), myelomonocytic markers (CD11b, CD13, CD14, and CD33), CD45, cytokeratin, or EBV latent membrane protein 1. The tumor cells also failed to express EBV-encoded RNA transcripts, as assessed by in situ hybridization.
|
Detection of HHV-8 Latency-Associated Nuclear Antigen
To detect the presence of HHV-8–specific antigens in tumor cells from the patient's pleural fluid, indirect immunofluorescent staining was performed with immune serum from an HHV-8–infected patient. Antibodies from most HHV-8–infected patients react with the latency-associated nuclear-antigen complex,34 which includes an HHV-8–encoded polypeptide of unknown function. Punctate nuclear staining in a pattern consistent with the distribution of latency-associated nuclear antigen was seen with immune serum in more than 90 percent of the tumor cells (Figure 2A). A small population of cells in the effusion fluid demonstrated diffuse immunofluorescent staining independently of the presence or absence of immune serum (Figure 2A and Figure 2B), a finding consistent with nonspecific binding of secondary antibody to admixed macrophages, but no punctate nuclear staining was seen when immune serum was omitted (Figure 2B). Immune serum produced an identical pattern of punctate nuclear staining in the HHV-8–infected BC-1 cell line (Figure 2C and Figure 2D) and the BCBL-1 cell line, whereas no staining was seen in the EBV-infected B-cell lines IB4 and B95-8. Supernatants from the patient's pleural fluid also produced strong punctate nuclear staining in the BC-1 and BCBL-1 cell lines, indicating the presence of HHV-8 antibodies.
|
Electron-microscopical examination of the pleural fluid revealed large tumor cells (10 to 20 µm in diameter) with prominent nucleoli, abundant rough endoplasmic reticulum and free ribosomes, and short, blunt surface projections (Figure 3). A number of tumor cells contained herpesvirus nucleocapsids in various stages of maturation, including complete virions, approximately 100 to 150 nm in diameter, with round, electron-dense cores, as well as incomplete virions without cores. Viral particles were most numerous in cells with peripheralization of nuclear chromatin and disruption of the nuclear membrane. Occasional enveloped forms were seen budding from the nuclear membrane (Figure 3, inset). A few viral particles were also seen within admixed benign macrophages.
|
To determine the duration of HHV-8 infection in the patient, PCR was used to screen archival tissue specimens for HHV-8 genomic sequences. With the use of a primer set that is specific for HHV-8, abundant product of the expected size (233 bp) was noted after only 15 cycles of amplification of tumor DNA (Figure 4), a result that can be explained by the presence of a large number of HHV-8 copies in the tumor-cell population due to active viral replication. Smaller amounts of HHV-8–specific product were also amplified from template DNA prepared from a lymph-node–biopsy specimen involved by Kaposi's sarcoma (May 1990), one of the post-transplantation endomyocardial-biopsy specimens with prominent plasmacytic infiltrates (October 1997), and tissue from the explanted heart (December 1989). HHV-8 sequences were not detected in an endomyocardial-biopsy specimen that had no plasmacytic infiltrates (March 1990) or in a colon-biopsy specimen obtained for evaluation of gastrointestinal bleeding (July 1997), even when DNA was subjected to 40 cycles of amplification.
|
Primary-effusion lymphoma, a rare, rapidly fatal neoplasm, was initially described in patients with late-stage AIDS.3,7,35,36 The tumor has a distinctive presentation, with malignant peritoneal, pericardial, or pleural effusions in the absence of an identifiable tumor mass or nodal involvement. It consists of large nucleolated cells with an immunoblastic or anaplastic appearance, which are positive for activation markers such as CD30 and negative for B-cell markers. An association with both EBV and HHV-8 infection has been established. Genotypic analysis invariably demonstrates immunoglobulin-gene rearrangements, confirming that these tumors originate from B cells.3,7,36 Karyotypes have been described for only two other primary-effusion lymphomas, both of which also had complex chromosomal aberrations without rearrangements specific for other types of lymphoma.36 The tumor described here resembled those previously reported, except that it had a more plasmacytic appearance, did not express CD45, and was not associated with either EBV or HIV infection.
The dual infection of tumor cells with EBV and HHV-8 in most cases of primary-effusion lymphoma has suggested that these herpesviruses have a synergistic role and has raised questions about the relative contribution of each virus to the transformed phenotype. The absence of EBV in tumor cells from our patient suggests that HHV-8 is the critical infectious agent in the pathogenesis of primary-effusion lymphoma. The recent description of five other EBV-negative cases of primary-effusion lymphoma and the establishment of several EBV-negative cell lines from these tumors further support the key role of HHV-8.26,27,37,38,39,40 Most cases of EBV-negative primary-effusion lymphoma have occurred in patients who were also HIV-negative, whereas primary-effusion lymphomas in HIV-positive patients are usually EBV-positive. The reason for these virologic differences is unknown. The status of HHV-8 infection in primary-effusion-lymphoma cell lines varies, although the infection is usually latent, with only a small number of cells undergoing lytic replication.31,41,42,43,44 We detected viral replication in freshly isolated tumor cells, as has been found in primary-effusion-lymphoma cell lines passaged in mice.29
Detection of HHV-8 DNA sequences in the explanted heart of our patient conclusively demonstrates that infection preceded cardiac transplantation. Whether viral infection contributed in any way to the development of dilated cardiomyopathy is uncertain. No link has been established between HHV-8 infection and cardiac disease, and it is possible that the positive PCR result reflected the presence of HHV-8–infected lymphocytes or monocytes in the tissue. The source of HHV-8 infection in our patient is unknown, but he was from Haiti, where the prevalence of antibodies against HHV-8 is relatively high. Recent studies indicate that sexual transmission is the predominant mode of spread of HHV-8.45,46 In one series in Italy, a region where HHV-8 infection is relatively common, Kaposi's sarcoma appeared to result from an infection acquired before solid-organ transplantation in 10 of 11 patients; as in our patient, lesions often arose within six months after transplantation.47 Together, these observations suggest that prior viral infection is the principal cause of HHV-8–associated neoplasia in recipients of solid-organ transplants.
Primary-effusion lymphoma is typically highly aggressive and refractory to therapy. It would be useful to identify patients who are at high risk for this tumor, since therapy with antiviral agents might suppress lymphomagenesis.31 The HHV-8 genome encodes genes homologous to several cellular proto-oncogenes implicated in B-cell lymphomagenesis, including Bcl-2 and cyclin D, as well as growth factors such as interleukin-6, all of which may contribute to the proliferation and survival of infected B cells.48,49 The presence of persistent hypergammaglobulinemia and polyclonal plasmacytic infiltrates in the myocardium and other tissues in our patient suggests that widespread B-cell hyperplasia preceded the development of lymphoma.
Supported by grants from the National Cancer Institute (CA66849, CA67380, and CA73580) and the Cancer Research Institute.
We are indebted to Dr. Geraldine Pinkus for assistance with the immunohistochemical and in situ hybridization studies; to Dr. Janina Longtine, Dr. Frank Kuo, and Genya Milman for assistance with the Southern blot studies; to Drs. Richard Benjamin and David Dorfman for assistance with the flow-cytometric studies; and to Thomas Morgan for assistance with the karyotypic analysis.
Source Information
From the Department of Pathology (D.J., J.M.G., G.L.W., J.F., J.C.A.) and the Division of Infectious Diseases (M.E.B., K.M.K.), Brigham and Women's Hospital and Harvard Medical School; and the AIDS Research Center, Massachusetts General Hospital and Harvard Medical School (D.T.S.) — all in Boston. Address reprints requests to Dr. Aster at the Department of Pathology, Brigham and Women's Hospital, 75 Francis St., Boston, MA 02115.
References
| |||||||||||||||||||||||||||||||||||||||
Related Letters:
Appearance of Human Herpesvirus 8 on Electron Microscopy
Orenstein J. M., Herndier B., Aster J. C., Jones D.
Extract |
Full Text
N Engl J Med 1999;
340:62-64, Jan 7, 1999.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
