Background and Definitions

AIDS was first described in 1981, and the first definitions included certain opportunistic infections, Kaposi sarcoma, and central nervous system (CNS) lymphomas. In 1984, a multicenter study described the clinical spectrum of non-Hodgkin lymphomas (NHLs) in the populations at risk of AIDS. The incidence of NHL has increased in a course almost parallel to that of the AIDS epidemic and accounts for 2% to 3% of newly diagnosed AIDS cases. Since the introduction of highly active antiretroviral therapy (HAART) in the mid-1990s, the incidence of lymphomas has decreased, and outcomes have improved. Higher CD4-positive T-lymphocyte (CD4) counts in the HAART era have been associated with a shift in histological diagnoses. The shift is away from primary effusion lymphoma and primary CNS lymphoma, which occur with the lowest CD4 counts, and toward histologies that occur at higher CD4 counts, such as Burkitt lymphoma and Hodgkin lymphoma (HL). In contrast to less-frequent incidences of all the lymphoproliferative disorders in the HAART era, the incidence rate of anal cancer has not changed.

Histology

Pathologically, AIDS-related lymphomas comprise a narrow spectrum of histologic types consisting almost exclusively of B-cell tumors of aggressive type. These include the following:

• Diffuse large B-cell lymphoma (including B-cell immunoblastic lymphoma).
• Small noncleaved lymphoma, either Burkitt or Burkitt-like.

The HIV-associated lymphomas can be categorized into the following:

• Aggressive B-cell lymphoma (see above).
• Primary CNS lymphoma (PCNSL).
• Primary effusion lymphoma.
• Plasmablastic multicentric Castleman disease.
• HL.

HIV-associated Hodgkin lymphoma

Multiple reviews of HL occurring in patients at risk of AIDS have been done; however, HL is still not part of the Centers for Disease Control and Prevention (CDC) definition of AIDS because no clear demonstration of its increased incidence in conjunction with HIV has been shown, as is the case for aggressive NHL. The CDC, in conjunction with the San Francisco Department of Public Health, has reported a cohort study in which HIV-infected men had an excess risk that was attributable to the HIV infection in 19.3 cases of HL per 100,000 person-years and 224.9 cases of NHL per 100,000 person-years. Although this report found an excess incidence of HL in HIV-infected homosexual men, additional epidemiologic studies will be needed before the CDC will reconsider HL as an HIV-associated malignancy.

HIV-associated HL presents in an aggressive fashion, often with extranodal or bone marrow involvement. A distinctive feature of HIV-associated HL is the lesser frequency of mediastinal adenopathy compared with non–HIV-associated HL. Most patients in these series had either mixed cellularity or lymphocyte-depleted HL, expression of Epstein-Barr virus (EBV)-associated proteins in Reed-Sternberg cells, B symptoms, and a median CD4 lymphocyte count of 300/dL or lower. In a retrospective multicenter review of 62 patients, those receiving HAART with chemotherapy had a 74% 2-year overall survival (OS) rate versus a 30% OS rate for those not receiving HAART (P .001). Among 201 patients with classical HL and HIV positivity, the 2- to 5-year OS rate of 88% to 90% after treatment with ABVD (doxorubicin, bleomycin, vinblastine, and dacarbazine), or similar regimens, and HAART, was not significantly different from the OS rate of HIV-negative patients with newly diagnosed HL in two uncontrolled comparisons. These studies confirm that patients with HL who were treated with standard regimens and HAART have outcomes that are similar to those of the uninfected population. Furthermore, immune function recovers over the course of 6 to 9 months after completion of chemotherapy.

Primary effusion lymphoma

Primary effusion lymphoma has been associated with Kaposi sarcoma (KS)-associated herpesvirus (KSHV)/human herpes virus type 8 (HHV8). Primary effusion lymphoma presents as a liquid phase spreading along serous membranes in the absence of masses or adenopathy. In addition to HHV8, many cases are also associated with EBV. Extension of lymphoma from the effusion to underlying tissue may occur. A series of 20 patients, including 19 treated with modified infusional etoposide, vincristine, and doxorubicin plus cyclophosphamide and prednisone (modEPOCH), had a 3-year cancer-specific survival rate of 47% and a median OS of 22 months.

Multicentric Castleman disease

The plasmablastic type of multicentric Castleman disease is also associated with a coinfection of KSHV/HHV8 and HIV. Patients typically present with fever, night sweats, weight loss, lymphadenopathy, and hepatosplenomegaly. Patients may progress to primary effusion lymphoma or to plasmablastic or anaplastic large cell lymphoma. Anecdotal responses to the anti-CD20 monoclonal antibody rituximab alone (along with HAART), have been reported. For a prospective cohort of 84 patients treated with rituximab for HIV and HHV8 multicentric Castleman disease, the 5-year rate of relapse-free survival was 82% (95% confidence interval , 72%–92%), and all patients responded again to rituximab at relapse.

Incidence and Prevention

An international database of 48,000 HIV-seropositive individuals from the United States, Europe, and Australia found a 42% decline in the incidence of NHLs from 1997 to 1999 compared with the same incidences in 1992 to 1996, both for PCNSL and for systemic lymphoma. The introduction of HAART is the proposed explanation for this decline. The diagnosis of AIDS precedes the onset of NHL in approximately 50% of the patients; however, in the other half of the patients, the diagnosis of AIDS is made at the time of the diagnosis of NHL and HIV positivity. The geographic distribution of these lymphomas is also similar to the geographic spread of AIDS. Unlike KS, which has a predilection for homosexual men and appears to be on the decline in incidence, all risk groups appear to have an excess number of NHLs; these risk groups include intravenous drug users and children of HIV-positive individuals.

Clinical Presentation

In general, the clinical setting and response to treatment of patients with AIDS-related lymphoma is very different from that of the non-HIV patients with lymphoma. The HIV-infected individual with aggressive lymphoma usually presents with advanced-stage disease that is frequently extranodal.

Common extranodal sites include the following:

• Bone marrow.
• Liver.
• Meninges.
• Gastrointestinal tract.

Very unusual sites are also characteristic and include the following:

• Anus.
• Heart.
• Bile duct.
• Gingiva.
• Muscles.

The clinical course is more aggressive, and the disease is both more extensive and less responsive to chemotherapy. Immunodeficiency and cytopenias, common in these patients at the time of initial presentation, are exacerbated by the administration of chemotherapy. Treatment of the malignancy increases the risk of opportunistic infections, which further compromise the delivery of adequate treatment.

Prognosis and Survival

Prognoses of patients with AIDS-related lymphoma have been associated with the following:

• Stage (i.e., extent of disease, extranodal involvement, lactate dehydrogenase level, and bone marrow involvement).
• Age.
• Severity of the underlying immunodeficiency (measured by CD4 lymphocyte count in peripheral blood).
• Performance status.
• Prior AIDS diagnosis (i.e., history of opportunistic infection or KS).

Patients with AIDS-related PCNSL appear to have more severe underlying HIV-related disease than do patients with systemic lymphoma. In one report, this severity was evidenced by patients with PCNSL who had a higher incidence of previously diagnosed AIDS (73% vs. 37%), lower median number of CD4 lymphocytes (30/dL vs. 189/dL), and a worse median survival time (2.5 months vs. 6.0 months). This report also showed that patients with poor risk factors—defined as Karnofsky Performance Status score lower than 70%, history of previously diagnosed AIDS, and bone marrow involvement—had a median survival time of 4.0 months compared with patients in a good prognosis group who had none of these risk factors, and who had a median survival time of 11.3 months.

In another report (NIAID-ACTG-142), prognostic factors were evaluated in a group of 192 patients with newly diagnosed AIDS-related lymphoma who were randomly assigned to receive either low-dose methotrexate, bleomycin, doxorubicin, cyclophosphamide, vincristine, and dexamethasone (m-BACOD) or standard-dose m-BACOD with granulocyte-macrophage colony-stimulating factor. No differences existed between these two treatments in terms of efficacy for disease-free survival, median survival, or risk ratio for death. On multivariate analysis, factors associated with decreased survival included age older than 35 years, history of intravenous drug use, stage III or stage IV disease, and CD4 counts lower than 100 cells/mm3. The International Prognostic Index may also be predictive for survival. In a multicenter cohort study of 203 patients, in a multivariable Cox model, response to HAART was independently associated with prolonged survival (relative hazard, 0.32; 95% CI, 0.16–0.62).

Pathologically, AIDS-related lymphomas comprise a narrow spectrum of histologic types consisting almost exclusively of B-cell tumors of aggressive type. These include the following:

• Diffuse large B-cell lymphoma (B-cell immunoblastic lymphoma).
• Small noncleaved lymphoma, either Burkitt or Burkitt-like.

AIDS-related lymphomas, though usually of B-cell origin as demonstrated by immunoglobulin heavy-chain gene rearrangement studies, have also been shown to be oligoclonal, polyclonal, and monoclonal in origin. Although HIV does not appear to have a direct etiologic role, HIV infection does lead to an altered immunologic milieu. HIV generally infects T lymphocytes with the loss of regulation function that leads to hypergammaglobulinemia and polyclonal B-cell hyperplasia. B cells are not the targets of HIV infection. Instead, Epstein-Barr virus (EBV) is thought to be at least a cofactor in the etiology of some of these lymphomas. The EBV genome has been detected in most patients with AIDS-related lymphomas; molecular analysis suggests that the cells were infected before clonal proliferation began. The rare primary effusion lymphoma consistently harbors human herpes virus type 8 and frequently contains EBV. HIV-related T-cell lymphomas have also been identified and appear to be associated with EBV infection.

Although stage is important in selecting the treatment of patients with non-Hodgkin lymphoma (NHL) who do not have AIDS, most patients with AIDS-related lymphomas have far-advanced disease.

Staging Subclassification System

Lugano classification

The American Joint Committee on Cancer (AJCC) has adopted the Lugano classification to evaluate and stage lymphoma. The Lugano classification system replaces the Ann Arbor classification system, which was adopted in 1971 at the Ann Arbor Conference, with some modifications 18 years later from the Cotswolds meeting.

Table 1. Lugano Classification for Hodgkin and Non-Hodgkin Lymphomaa

Occasionally, specialized staging systems are used. The physician should be aware of the system used in a specific report.

The E designation is used when extranodal lymphoid malignancies arise in tissues separate from, but near, the major lymphatic aggregates. Stage IV refers to disease that is diffusely spread throughout an extranodal site, such as the liver. If pathologic proof of involvement of one or more extralymphatic sites has been documented, the symbol for the site of involvement, followed by a plus sign (+), is listed.

Table 2. Notation for Identification of Sites

Current practice assigns a clinical stage (CS) based on the findings of the clinical evaluation and a pathologic stage (PS) based on the findings made as a result of invasive procedures beyond the initial biopsy.

For example, on percutaneous biopsy, a patient with inguinal adenopathy and a positive lymphangiogram without systemic symptoms might be found to have involvement of the liver and bone marrow. The precise stage of such a patient would be CS IIA, PS IVA(H+)(M+).

A number of other factors that are not included in the above staging system are important for the staging and prognosis of patients with NHL. These factors include the following:

• Age.
• Performance status (PS).
• Tumor size.
• Lactate dehydrogenase (LDH) values.
• The number of extranodal sites.

The treatment of patients with AIDS-related lymphomas presents the challenge of integrating therapy appropriate for the stage and histologic subset of malignant lymphoma with the limitations imposed by HIV infection. In addition to antitumor therapy, essential components of an optimal non-Hodgkin lymphoma treatment strategy include the following:

• Highly active antiretroviral therapy.
• Prophylaxis for opportunistic infections.
• Rapid recognition and treatment of intercurrent infections.

Patients with HIV positivity and underlying immunodeficiency have poor bone marrow reserve, which compromises the potential for drug dose intensity. Intercurrent opportunistic infection is a risk that may also lead to a decrease in drug delivery. Furthermore, chemotherapy itself compromises the immune system and increases the likelihood of opportunistic infection.

The treatment of AIDS-related lymphomas involves overcoming several problems. These are all aggressive lymphomas, which by definition are diffuse large cell/immunoblastic lymphoma or small noncleaved cell lymphoma (Burkitt lymphoma). These lymphomas frequently involve the bone marrow and central nervous system (CNS) and, therefore, are usually in an advanced stage. In addition, the immunodeficiency of AIDS and the leukopenia that is commonly seen with HIV infection makes the use of immunosuppressive chemotherapy challenging.

The introduction of highly active antiretroviral therapy (HAART) has led to a marked reduction in opportunistic infections, prolonged survival with HIV infection, and a median overall survival (OS) for patients with AIDS-related lymphoma that is comparable with the outcome in the nonimmunosuppressed population. The use of HAART has also allowed standard-dose and even intensive chemotherapy regimens to be given with reasonable safety to patients with AIDS-related lymphomas, which is comparable with the outcome in non-HIV patients.

Several prospective nonrandomized trials and pooled individual data from 19 prospective trials that included 1,546 patients show that the addition of rituximab to combination chemotherapy improves the complete response rate, progression-free survival, and OS. Several other prospective nonrandomized trials and the pooled individual data from the same 1,546 patients also show that infusional EPOCH (infusional etoposide, infusional vincristine, infusional doxorubicin, cyclophosphamide, and prednisone) produced better outcomes than did CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone) (OS, hazard ratio, 0.33; P = .03). Concurrent use of HAART with the infusional EPOCH regimen is controversial; one group advocated for HAART after completion of chemotherapy, while others allowed concurrent therapy.

For patients with Burkitt lymphoma, dose-modified regimens such as R-CODOX (cyclophosphamide, doxorubicin, vincristine, methotrexate, cytarabine and rituximab)-M/IVAC (ifosfamide, etoposide, and high-dose cytarabine) or R-EPOCH (etoposide, prednisone, vincristine, and doxorubicin in combination with rituximab) have shown good results with concurrent or sequential HAART.

Patients at risk of subsequent CNS involvement include those with bone marrow involvement or those with EBV identified in the primary tumor or in the cerebrospinal fluid (i.e., by polymerase chain reaction). Intrathecal chemotherapy is usually considered for patients who are at higher risk of CNS involvement.

Highly selected patients with resistant or relapsed lymphoma after first-line chemotherapy and with continued responsiveness to HAART underwent second-line chemotherapy followed by high-dose therapy and autologous peripheral stem cell transplantation. Long-term survivors have been reported anecdotally for these highly selected patients who relapsed.

As with other AIDS-related lymphomas, primary central nervous system (CNS) lymphoma (PCNSL) is an aggressive B-cell neoplasm, either diffuse large B-cell or diffuse immunoblastic non-Hodgkin lymphoma (a subtype of diffuse large B-cell lymphoma). AIDS-related PCNSL has been reported to have a 100% association with Epstein-Barr virus (EBV). These patients usually have evidence of low CD4-positive T lymphocyte counts, high HIV viral load, severe debilitation, and focal neurologic symptoms such as seizures, changes in mental status, and paralysis.

Computed tomographic scans show contrast-enhancing mass lesions that may not always be distinguished from other CNS diseases, such as toxoplasmosis, that occur in patients with AIDS. Magnetic resonance imaging studies using gadolinium contrast may be a more useful initial diagnostic tool in differentiating lymphoma from cerebral toxoplasmosis or progressive multifocal leukoencephalopathy. Lymphoma tends to present with large lesions, which are enhanced by gadolinium. In cerebral toxoplasmosis, ring enhancement is very common, lesions tend to be smaller, and multiple lesions are seen. Use of positron emission scanning has demonstrated an improved ability to distinguish PCNSL from toxoplasmosis. PCNSL has an increased uptake while toxoplasmosis lesions are metabolically inactive. Antibodies against toxoplasmosis may also be very useful because most cerebral toxoplasmosis occurs as a consequence of reactivity of a previous infection. If the immunoglobulin G titer is less than 1:4, the disease is unlikely to be toxoplasmotic. A lumbar puncture may be useful to detect as many as 23% of patients with malignant cells in their cerebrospinal fluid (CSF). Evaluating the CSF for EBV DNA may be a useful lymphoma-specific tool because EBV is present in all patients with PCNSL. Despite the many evaluations, however, most patients with PCNSL require a pathological diagnosis. Diagnosis is made by biopsy. Sometimes, a biopsy is attempted only after failure of antibiotics for toxoplasmosis, which will produce clinical and radiographic improvement within 1 to 3 weeks in patients with cerebral toxoplasmosis.

Radiation therapy alone has usually been used in this group of patients. With doses in the range of 35 Gy to 40 Gy, median duration of survival has been only 72 to 119 days. Survival is longer in younger patients with better performance status and the absence of opportunistic infection. In the highly active antiretroviral therapy (HAART) era, a median survival of 18 months has been seen with radiation therapy alone. An anecdotal report using HAART and high-dose methotrexate for patients with AIDS-related PCNSL showed a median survival that had not been reached with a median follow-up of 27 months. Most patients respond to treatment by showing partial improvement in neurologic symptoms. Autopsies have revealed that these patients die of opportunistic infections as well as tumor progression. Treatment of these patients is also complicated by other AIDS-related CNS infections, including subacute AIDS encephalitis, cytomegalovirus encephalitis, and toxoplasmosis encephalitis. Spontaneous remissions have been reported after HAART.