INTRODUCTION — Infection with HIV predisposes to the development of neoplasms, including lymphoma. AIDS-related lymphoma is generally divided into three types: systemic non-Hodgkin lymphoma (NHL), primary central nervous system lymphoma (PCNSL), and the primary effusion ("body cavity") lymphomas . Systemic NHL accounts for the great majority of lesions, PCNSL for approximately 15 percent, and primary effusion lymphoma (PEL) is much less common, at 4 percent or less .
Primary effusion lymphoma (PEL) will be reviewed here. The other AIDS-related lymphomas are discussed separately. (See "Overview of AIDS-related lymphomas" and "AIDS-related lymphomas: Primary central nervous system lymphoma" and "AIDS-related lymphomas: Systemic lymphoma" and "AIDS-related lymphomas: Treatment of systemic lymphoma".)
EPIDEMIOLOGY — PEL is one of the least common of the AIDS-related lymphomas, comprising 4 percent or less of such lymphomas . The overwhelming majority of cases occur in HIV-infected patients. However, this lesion can occur in the absence of HIV infection and rarely has been seen following solid organ transplantation.
There appears to be an overtly male predominance with men accounting for all 15 cases in one of the original descriptions, and for 10 of 11 cases in a separate single-institution study. This may largely reflect the markedly increased prevalence of HIV infection among men. (See "Overview of AIDS-related lymphomas" and "Overview of AIDS-defining malignancies in HIV infection", section on 'Effect of HAART'.)
Although earlier studies of this rare condition had reported low CD4 counts in patients with PEL , larger and more recent series do not support this observation. Among those infected with HIV, patients with PEL are similar to those with other NHLs in age, race, and HIV transmission category .
As described below, human herpesvirus-8 (HHV-8) plays a role in the pathogenesis of PEL. Those individuals with other disorders related to HHV-8 infection have an increased risk for developing primary effusion lymphoma, and vice-versa. These disorders include Kaposi's sarcoma and multicentric Castleman's disease. (See "Disease associations of human herpesvirus 8 infection" and "Castleman's disease", section on 'Role of interleukin-6'.)
PATHOGENESIS — The malignant cells of PEL are monoclonal B cells (as defined by rearrangement of the immunoglobulin gene) that express cell surface CD38 and contain genomic material from human herpesvirus-8 (HHV-8, also called Kaposi's sarcoma-associated herpesvirus or KSHV) and, in many cases, Epstein-Barr virus (EBV) . The importance of EBV in lymphomatous transformation in this setting is uncertain, as opposed to its primary role in PCNSL . (See "AIDS-related lymphomas: Primary central nervous system lymphoma".)
Cell of origin — The precise B cell subset from which these cells are derived and the biological mechanisms responsible for its unusual growth pattern (ie, limited to body cavities) are uncertain. It has been suggested that the cells represent a preterminal stage of B-cell differentiation . However, others suggest that the development of PEL is not restricted to one stage of B-cell differentiation and may represent transformation of B-cells at different stages of ontogeny .
PELs express a common gene profile that is distinct from that of other AIDS-related NHLs or lymphomas in the immunocompetent population. This profile suggests that, unlike AIDS-related NHLs, the tumor cells are not of germinal center or memory cell origin. Rather, they more likely correspond to a stage of B-cell development intermediate between that of immunoblasts and plasma cells. (See "Overview of gene expression profiling and proteomics in clinical oncology".)
Viral infection — All patients with PEL have HHV-8 infection and many patients also demonstrate evidence of EBV infection. While HHV-8 infection is required for the development of PEL, the mechanisms by which HHV-8 infection might promote tumor growth are uncertain. Loss of the HHV-8 genome results in the death of PEL cells, demonstrating that genes of this virus play a vital role in PEL cell survival .
The following latent gene products of HHV-8 appear to play significant roles in the development of PEL by promoting proliferation and impairing apoptosis :
Latency-associated nuclear antigen (LANA-1)
Viral cyclin (v-cyclin)
Viral FLICE inhibitory protein (v-FLIP)
Viral interleukin-6
Viral interleukin-8 receptor homolog (vIL8R), also known as viral G-protein coupled receptor homolog (vGPCR)
The transmembrane protein K1
Proposed pathogenic roles for LANA-1 in PEL development include tethering HHV-8 DNA to chromosomes during mitosis to permit segregation of HHV-8 episomes to the progeny cells , activation of EBV promotor regions in coinfected cells , and inhibition of tumor suppressor genes.
V-cyclin binds and phosphorylates the cell cycle inhibitor p27 (KIP1), thereby inhibiting the negative cell cycle control function of p27 and promoting the rapid proliferation of PEL cells . Further, v-cyclin forms a complex with cyclin-dependent kinase (CDK-6) to form an active kinase inhibitor that also promotes cell proliferation.
The HH8 K1 gene is expressed in PEL cells and is up-regulated when these cells enter the lytic phase of the virus life cycle . The product of this gene promotes the survival of infected cells and further dissemination of HHV-8. This occurs through the activation of nF-kappa B, by disrupting cytokine regulation, and by its ability to bind and disrupt other proteins.
HHV-8 infection results in the constitutive activation of nuclear factor-kappa B (NF-kB) in endothelial cells perhaps through the latent protein v-FLIP. Much like in other virally-mediated lymphomas, the survival of PEL cells depends on the unregulated production of NF-kB. Inhibition of NF-kB has been shown to induce apoptosis in PEL cells .
Interleukin (IL)-6 and IL-10 may act as autocrine growth factors to promote lymphomagenesis and the growth of PEL cells . The HHV-8 genome encodes viral IL-6, a cytokine that promotes plasmacytosis and angiogenesis.
Additional studies suggest that constitutive phosphorylation of Signal Transducer and Activator of Transcription-3 (STAT3) in the cells of PELs occurs secondary to IL-10 and viral IL-6. Phosphorylation of STAT-3 directly contributes to the malignant progression of PEL cells by activating the prosurvival (anti-apoptotic) protein survivin .
CLINICAL MANIFESTATIONS — The clinical manifestations of PEL depend upon the extent and distribution of disease. PEL originates on serosal surfaces, including the pleura (60 to 90 percent), pericardium (zero to 30 percent), peritoneum (30 to 60 percent), joint spaces, and, rarely, the meninges . Most affected patients present with a symptomatic serous effusion containing high-grade, malignant lymphocytes, but with no detectable mass lesion . As such, patients usually present with symptoms related to fluid accumulation such as dyspnea (from pleural or pericardial effusions), abdominal distension (from ascites), or joint swelling.
Radiographic testing in primary effusion lymphoma reveals evidence of the local effusions. Chest radiographs and CT reveal the pleural and/or pericardial effusion, slight serosal (pleural, pericardial) thickening, and the absence of parenchymal abnormalities, solid masses, or mediastinal enlargement.
Although PEL is characterized by its predilection for body cavities, these tumor cells may rarely appear as HHV-8 positive solid tumors. These solid tumor variants predominantly occur in the gastrointestinal tract.
With increased screening measures for HIV, many patients are treated with highly active anti-retroviral therapy (HAART) before the development of complications such as PEL. It is unknown how the clinical manifestations of PEL which develops in patients undergoing treatment with HAART differs, if at all, from those of patients who are HAART-naive.
EVALUATION AND DIAGNOSIS — The evaluation of the patient suspected of having PEL involves imaging techniques to detect the effusion followed by fluid examination. While the morphology and immunophenotype can vary, the key diagnostic criterion for PEL is the presence of HHV-8 in the nuclei of the malignant cells.
Fluid examination — Samples of the effusion are almost always positive for malignant cells due to the unique liquid-phase of growth of these tumors .
Morphology — The malignant cells exhibit a range of appearances, from large immunoblastic or plasmablastic cells to those with more anaplastic characteristics; some cells can resemble Reed-Sternberg cells . A perinuclear hof consistent with plasmacytoid differentiation may be seen. The cytoplasm is deeply basophilic with vacuoles in occasional cells.
Immunophenotype — The immunophenotype of the malignant cells in PEL often reflect that of a mature B cell shifting towards terminal plasma cell differentiation. Over 90 percent of cases demonstrate expression of CD45. Other B-cell (ie, CD19, CD20, CD79a) and T-cell-associated antigens are typically negative, but can be seen in a fraction of cases . Activation and plasma cell-related markers such as CD30, CD38, CD71, CD138, and epithelial membrane antigen are usually present.
Genotype — No characteristic genetic abnormalities have been identified in PEL but complex and recurrent cytogenetic abnormalities in the tumor cells have been reported . In addition, immunoglobulin genes and T-cell receptor genes are often rearranged and can be used to demonstrate a monoclonal B cell population by Southern blot analysis.
Viral testing — The key diagnostic criterion for PEL is the presence of HHV-8 in the nuclei of the malignant cells. The most common method for detecting HHV-8 positivity is immunohistochemical staining for the latent viral gene product known as latency-associated nuclear antigen (LANA-1). Despite the usual co-infection with EBV, staining for latent membrane protein (LMP1) is negative.
DIFFERENTIAL DIAGNOSIS — The differential diagnosis for patients who present with an effusion that is subsequently found to contain lymphoma cells includes systemic lymphomas with secondary involvement of the body fluid (ie, secondary effusion), extranodal variants of various subtypes of lymphoma (eg, extranodal large cell lymphoma), and lymphomas that develop as a result of chronic pyothorax (ie, pyothorax associated lymphoma). Primary effusion lymphoma is characteristically distinguished from all of these other types of lymphoma by its HHV-8 positivity .
In addition, there is a broad differential diagnosis for pleural effusions in HIV infected patients which include both infectious and non-infectious causes. These are discussed in detail separately. (See "Pleural effusions in HIV-infected patients".)
Extranodal Burkitt lymphoma — Patients with Burkitt lymphoma can sometimes present with an effusion as the sole marker of disease, most frequently in the context of HIV infection. The defining biological feature of Burkitt lymphoma is c-myc deregulation which is not always technically easy to detect. C-myc deregulation is not found in PEL. Ninety percent of the time the malignant cells of Burkitt lymphoma display a translocation between the long arm of chromosome 8 (the site of the c-myc oncogene) and one of three other chromosomes resulting in t(8;14), t(2;8), or t(8;22). The principal feature that distinguishes Burkitt lymphoma from primary effusion lymphoma is its HHV8 negativity. (See "Clinical manifestations, pathologic features, and diagnosis of Burkitt lymphoma".)
Pyothorax associated lymphoma — Pyothorax-associated lymphoma, reported mostly from Japan, is a rare complication of long-standing pyothorax, most often in association with tuberculosis in the context of Epstein-Barr virus infection. In contrast to primary effusion lymphoma, these lymphomas have no association with immunosuppression and are HHV-8 negative. They also typically present with a tumor mass within the body cavity.
STAGING — The Ann Arbor staging system is not useful in PEL since all patients have stage IV disease by definition . In addition, the International Prognostic Index used in most other types of NHL has not been validated in patients with PEL. However, a pretreatment evaluation can help determine the extent of disease. (See "Evaluation, staging, and prognosis of non-Hodgkin lymphoma".)
Computed tomography (CT) of the chest, abdomen, and pelvis is recommended for all patients with consideration given to nuclear imaging with positron emission tomography (PET) . The use of other imaging modalities, bone marrow biopsy, lumbar puncture, and/or endoscopy is driven by the clinical manifestations.
Evaluation of the newly diagnosed patient should also include a complete blood count with differential, chemistries with liver and renal function and electrolytes, serum lactate dehydrogenase (LDH) level, and HIV serologies.
TREATMENT — Although they have little propensity to disseminate, PELs cause local destruction and have a uniformly poor prognosis without treatment . The median survival after diagnosis without treatment is approximately 2 to 3 months . Even with aggressive chemotherapy, historically the median survival extended on average to only six months . While many cases demonstrate a response to chemotherapy treatment, remissions are often of short duration.
There is a paucity of data to guide the treatment of patients with PEL. Since the disease is so uncommon, there are very few retrospective series and no prospective trials in this patient group. In addition, PEL's unique clinical manifestations make trials of other NHL subtypes largely inapplicable.
Treatment approaches that have been used for PEL include highly active anti-retroviral therapy (HAART), cytotoxic chemotherapy, radiation therapy, antiviral therapy, and combinations of these. Our approach to treatment of the patient with PEL depends partly on their HIV status as described below.
HIV infected patients — The optimal treatment of PEL in HIV infected patients is unknown. Most clinicians advocate using a combination of HAART with or without combination chemotherapy as initial therapy.
Initial approach — A key component of the treatment of all HIV infected (HIV+) patients with an AIDS-related NHL is the administration of an effective HAART regimen. This is supported by retrospective studies in other AIDS-related NHL subtypes that have reported a survival benefit with HAART plus chemotherapy compared with chemotherapy alone. (See "AIDS-related lymphomas: Treatment of systemic lymphoma", section on Effect of HAART therapy.)
Retrospective studies evaluating the use of chemotherapy in patients with HIV+ PEL have included patients who did or did not receive HAART. Although data are limited, patients treated with HAART alone appear to have similar outcomes as those administered HAART plus chemotherapy and far superior survival compared with those given chemotherapy alone.
The following is a survey of these retrospective analyses:
A retrospective single-institution study reported outcomes of 10 HIV+ patients with PEL, five of whom had received prior HAART therapy . Five patients were treated with CHOP-like chemotherapy plus HAART, which resulted in two complete remissions (CRs) and a mean overall survival of 16 months. One received HAART alone resulting in a CR and was alive at 14 months of follow-up. Three were treated with chemotherapy but no HAART which resulted in a three month overall survival. One patient received neither chemotherapy nor HAART and was dead in two weeks.
A multi-center retrospective study evaluated the outcomes of 17 HIV+ patients with PEL treated with combination chemotherapy plus HAART . Only two patients were HAART-naive at the time of diagnosis. There were eight CRs. Median survival was six months and the one-year survival rate was 40 percent. This series also reported on two patients who received CHOP chemotherapy without HAART, neither of whom achieved a CR and both died within six weeks.
Another retrospective series of seven HIV+ patients reported on the use of standard doses of CHOP chemotherapy plus high-dose methotrexate with leucovorin rescue [48]. Five patients received chemotherapy plus HAART. Three had a CR and were alive at an average follow-up of 40 months. A fourth died from plasmablastic NHL with PEL in CR at the time of his death. Two patients were treated with chemotherapy without HAART; one had a CR and started HAART one month after chemotherapy and was alive 78 months after diagnosis while the other patient was dead 22 days after the diagnosis.
Case reports have demonstrated that some patients with PEL can obtain CRs with HAART alone . We have also achieved CR with HAART alone in patients referred to us for PEL.
For HIV infected patients with newly diagnosed PEL, we suggest the institution of HAART or modification of an existing HAART regimen rather than administering HAART plus chemotherapy. This reflects the unclear benefits and increased toxicity seen with chemotherapy and reports of prolonged remissions in patients treated with HAART alone. As with all antiviral therapy in HIV+ patients, the goal is to achieve an undetectable viral load. Other experts would initiate both HAART and chemotherapy at the time of diagnosis. Chemotherapy options are discussed in the following section.
When choosing among HAART regimens, it is important to take into consideration overlapping toxicities or interference of anti-retrovirals with chemotherapeutic agents that may be used in the future. As an example, anti-retroviral drugs with excessive myelotoxicity, such as zidovudine, should be avoided in patients who are receiving or may soon receive myelotoxic chemotherapy. The choice of initial HAART regimens in patients with HIV is discussed in more detail separately. (See "Selecting antiretroviral regimens for the treatment naive HIV-infected patient" and "Considerations prior to initiating antiretroviral therapy".)
Patients who fail HAART alone — Treatment options are limited for patients who fail to respond to the initiation or modification of HAART. The effectiveness of combination chemotherapy with or without HAART in patients with PEL has been evaluated in uncontrolled retrospective analyses, some of which are described in the section above.
There is little to no evidence that conventional chemotherapy regimens used for other aggressive, high-grade NHLs (eg, CHOP) are effective for those with PEL. Such may offer some palliation with response rates of 40 to 50 percent extending median survival from two to three months in those not receiving therapy to a median survival of five to six months.
The following options are offered based upon our clinical experience, case reports, or promising in vitro data:
In our experience, pegylated liposomal anthracyclines given in doses that are used for Kaposi's sarcoma have at least a palliative effect with limited toxicity. These include liposomal daunorubicin at a dose of 40 mg/M2 or pegylated liposomal doxorubicin at a dose of 20 mg/M2. We favor the latter, given its relatively greater water solubility and ease of administration.
Rare cases of PEL are comprised of CD20 positive cells. In such cases, rituximab may offer a therapeutic possibility .
Bortezomib has demonstrated in vitro activity against PEL with efficacy varying depending upon the timing of administration and combination with other agents . It is thought to sensitize PEL cells to chemotherapy-induced apoptosis . Although anecdotal reports on its efficacy are mixed , a phase II trial of bortezomib with combination chemotherapy by the National Cancer Institute has been completed and results are pending.
More intensive regimens, such as high-dose chemotherapy followed by autologous hematopoietic cell transplantation (HCT) or allogeneic HCT have had mixed results with two case reports showing success and another showing no response .
Patients who do not have a remission with institution or modification of a HAART regimen should be enrolled in clinical trials, if available. However, if no such trial is available or if the patient does not wish to participate in a trial, we would consider the use of a liposomal anthracycline alone or liposomal anthracycline plus bortezomib and prednisone (or an equivalent steroid). These agents are administered in combination with continued HAART.
It is standard practice to use granulocyte stimulating growth factors to limit the period of neutropenia in patients with AIDS-related lymphoma receiving chemotherapy . We also routinely administer Pneumocystis jiroveci pneumonia (PCP) prophylaxis, ideally with oral trimethoprim-sulfamethoxazole (TMP-SMX). TMP/SMX can be myelosuppressive and may synergize with chemotherapy to result in a more profound and longer nadir. As such, blood counts must be monitored during therapy. PCP prophylaxis is continued for several months after completion of chemotherapy and until the CD4 counts are stable above 200/mm3. While some clinicians advocate fungal prophylaxis and screening for cytomegalovirus reactivation while on therapy, we do not use these routinely . (See "Prophylaxis against Pneumocystis carinii (P. jirovecii) in HIV-infected patients".)
Palliation therapy — When systemic chemotherapy is not possible, we have had experience with the use of local radiation therapy to the body cavity of origin . This often palliates the patient for a period of time that may approach 12 months. We reserve radiation therapy for the palliation of patients who are either not candidates for other therapeutic options or who have failed other treatment regimens.
HIV negative patients — There is even less evidence to guide the treatment of HIV-negative patients with PEL. This is an extremely rare patient population and the majority are being administered chronic immunosuppressant agents to prevent rejection of a solid organ transplantation. Therapy is based upon extrapolation of the data from HIV infected patients discussed in the sections above. HAART therapy is not indicated in the absence of HIV infection so initial therapy is with chemotherapy. Radiation therapy can also be given to patients who are unable to tolerate or have failed other treatments.
Based upon clinical experience that HIV infected patients with PEL treated with HAART have better outcomes than those who do not receive HAART therapy, we would expect that patients without HIV infection on immunosuppressive therapy would do better if their immunosuppressive therapy were decreased.
For patients with HIV negative PEL, we would consider the use of a liposomal anthracycline alone or liposomal anthracycline plus bortezomib and prednisone (or an equivalent steroid). If the patient is receiving immunosuppressive therapy, we decrease this immunosuppression, if possible.
Investigational therapies — Active research on the use of parenteral or intracavitary antiviral agents is ongoing. Case reports have been published on the use of intracavitary cidofovir and parenteral zidovudine in HIV negative patients . The use of parenteral or intracavitary antiviral treatment requires more investigation before it can be recommended outside of the setting of a clinical trial.
FOLLOW-UP — For patients initiating HAART or changing a HAART regimen, we reevaluate their HIV status four weeks after starting the regimen with a measurement of viral load. We expect to see a drop in the viral load by at least one log(10) at this time. By 12 weeks the viral load should become undetectable. If it is still detectable at 12 weeks, a change in regimen should be considered.
Response of the PEL to therapy is evaluated six to eight weeks after attainment of an undetectable viral load or completion of chemotherapy with a history, physical examination, laboratory studies (complete blood count, lactate dehydrogenase, and biochemical profile), and a PET/CT scan, which provides information on the size and activity of residual masses and allows for the distinction between active disease and fibrosis . Disease response is then determined by the International Workshop Criteria (IWC) as detailed in the table.
While many patients will demonstrate a response to treatment, these remissions are often short-lived. Following documentation of a complete remission, patients are seen at periodic intervals to monitor for treatment complications and assess for possible relapse. The frequency and extent of these visits depends upon the comfort of both the patient and physician.
We generally follow our patients every three months. At these visits, we perform a history and physical examination, complete blood count, chemistries, LDH, and chest x-ray. Abnormalities found on these studies are further evaluated with CT scan. A biopsy should always be obtained to document relapsed disease before proceeding to salvage therapy.
Causes of death include not only progression of lymphoma, but also opportunistic infections and other HIV-related complications . It is critical for these patients to be closely followed by their HIV specialist in order to control their HIV infection and prevent opportunistic infections.
SUMMARY AND RECOMENDATIONS
Primary Effusion Lymphoma (PEL) is one of the least common of the AIDS-related lymphomas. Although most cases occur in HIV-infected patients, this lesion can occur in the absence of HIV infection most often in patients with other causes of immunocompromise (eg, solid organ transplantation recipients). Latent gene products of human herpesvirus-8 (HHV-8) appear to play a role in the pathogenesis.
PEL originates on serosal surfaces and most affected patients present with a symptomatic serous effusion containing high-grade, malignant lymphocytes, but with no detectable mass lesion.
The diagnosis of PEL is usually made based upon a cytologic examination of the effusion which almost always contains malignant cells. The key diagnostic criterion is the presence of HHV-8 in the nuclei of the malignant cells.
Patients with PEL have a poor prognosis. Mean overall survival is less than three months without treatment and approximately six months with chemotherapy. However, with the use of highly active anti-retroviral therapy (HAART), some patients can have more extended survival.
There is limited data to guide the treatment of patients with PEL and recommendations are based on expert opinion, case reports, and small retrospective case series. Treatment is modified based on whether the patient is HIV infected.
For HIV infected patients with newly diagnosed PEL, we suggest the institution of HAART or modification of an existing HAART regimen rather than administering HAART plus chemotherapy (Grade 2C). Care must be taken to choose a HAART regimen with minimal overlap of chemotherapy-related toxicities.
HIV infected patients who fail HAART therapy alone should be enrolled in a clinical trial. However, if no such trial is available or if the patient does not wish to participate in a trial, we suggest the use of a liposomal anthracycline alone or a liposomal anthracycline plus bortezomib and prednisone (or an equivalent steroid) (Grade 2C). These agents are administered in combination with continued HAART.
For patients with HIV negative PEL, we suggest the use of a liposomal anthracycline alone or a liposomal anthracycline plus bortezomib and prednisone (or an equivalent steroid) (Grade 2C). If possible, immunosuppressive therapy is decreased.
For patients who are unable to tolerate combination chemotherapy, the administration of local radiation therapy to the body cavity of origin is an option.
by John P Doweiko ,MD & al
