HIV-Associated Kaposi's Sarcoma
Primary CNS Lymphoma
Other HIV-Associated Malignancies
Non-HIV Retroviral Malignancies: Adult T-Cell Leukemia/Lymphoma
Although investigators knew before 1980 that retroviruses could cause various forms of leukemia, lymphoma, and solid tumors in animals, not until then was the first human oncogenic retrovirus, human T-cell leukemia/lymphoma virus-I (HTLV-I), isolated . Discovery of the human immunodeficiency virus (HIV) followed in l983 . These viruses are unique in being composed of single-stranded RNA, and they replicate by forming double-stranded DNA that is integrated into the host-cell genome. The enzyme reverse transcriptase is fundamental to the replication process . Once the virus is incorporated into the host cell, a variety of proteins are produced that may activate or transform the normal cell into a malignant one. For example, retroviral genomes contain promoter and enhancer sequences that may activate adjacent host genes and trigger cell division (cis-activation). Expression of viral proteins may also contribute to activation of host genes (trans-activation). In addition, the viral genome may contain oncogenes that directly transform cells upon incorporation into the host cell genome [3–5].
There are three families of retroviruses. Oncoviruses contain both direct transforming viruses and chronic transforming viruses that induce malignancy over long latency periods . HTLV-I, -II, and -V are included in the oncovirus family [4,6]. Lentiviruses include HIV-1 and -2. The third family, spumaviruses, are not yet known to be associated with human diseases [3,4].
HIV induces immune dysfunction in a variety of ways. Immune abnormalities observed in HIV infection include depletion and dysfunction of CD4-positive T-cells; polyclonal activation of B-cells (often associated with hypergammaglobulinemia and autoimmune phemomena); and diminished function of monocytes, macrophages, and natural killer cells (Table 1). Patients exhibit impaired B-cell response to T-cell-dependent antigens, impaired cell-mediated immunity, delayed-type hypersensitivity, and abnormal cytokine expression. Collectively, these immune system defects provide multiple opportunities for malignant transformation on the molecular level and the maintenance of malignant cell growth once established [4,7,8].
|Proposed mechanisms||Pathophysiologic outcome||Symptoms|
|Fusion of gp 120 with CD4||Syncytial formation/cell death||CD4+ T-cell depletion|
|Binding of secreted gp 120 to CD4||Immune-mediated cytolysis||CD4+ T-cell depletion|
|Binding of secreted gp 120 to CD4||T-cells with class II MHC molecules on antigen-presenting cells|
|Binding of anti-CD4 antibodies to CD4||Cell death/defective antigen recognition||CD4+ T-cell depletion or dysfunction|
|Decreased expression of CD4 on the cell surface||Defective antigen recognition||CD4+ T-cell depletion or dysfunction|
|Defective T-cell helper function||Defective humoral immunity||Pyogenic infections|
|Infection with EBV, CMV, and double infection iwth HIV and EBV or HIV and CMV||Polyclonal B-cell proliferation||Hypergammaglobulinemia, autoimmune phenomena|
|gp 120 binding to B-lymphocytes may imitate the binding of interleukin|
|Defective T-cell helper function|
HIV infection of monocytes
|Defective macrophage function|
Production of monokines (IL-1, TNF)
Reservoir of infection in vivo
|Fever, weight loss, interstitial pneumonitis, brain damage|
|Natural killer cells|
|Defective inductive signals||Natural killer cells dysfunction in vivo||Tumor cell growth|
|CMV = cytomegalovirus, EBV = Epstein-Barr virus, HIV = human immunodeficiency virus, IL = interleukin, MHC = major histocompatability complex, TNF = tumor necrosis factor.|
Adapted, with permission, from Lazo PA, Tsichlis PN: Biology and pathogenesis of retroviruses. Semin Oncol 17:269-294, 1990.
By the time the HIV epidemic was underway, investigators had already realized that transplant recipients develop atypical lymphomas involving extranodal sites or the central nervous system (CNS) at 25 to 50 times the expected rate, anogenital cancer at 100 times, Kaposi's sarcoma at 400 to 500 times, and squamous-cell cancers of the skin at 3 to 20 times the expected rate [3,9,10]. These patients present with malignancy within 2 to 8 years of the onset of immunosuppressive therapy and often exhibit particularly virulent tumors . Similarly, HIV-infected patients present with malignancy at a higher than expected rate, at younger ages, and with a more virulent course than the general population.
Malignancy in patients with impaired immunity is the result of multiple factors. Immunosuppression itself can impair immune surveillance, which controls virally mediated cancers. Immunosuppressed patients often demonstrate chronic antigenic stimulation induced by an allograft or by repetitive acquired infections, which may increase the opportunity for random transforming mutations. Finally, dysregulation of the immune system, with a lack of the proper suppressor mechanisms, may result in malignant transformation .
First described in 1983, HIV-1 is now recognized worldwide as a cause of the acquired immunodeficiency syndrome (AIDS). A second serotype, HIV-2, is confined primarily to West Africa. Since the onset of the AIDS epidemic, advances in therapy have prolonged the median life expectancy of patients with HIV infection to more than 12 years . However, a consequence of this improved survival rate is a rising incidence of HIV-related malignancies. Levine has predicted that up to 40% of persons with AIDS will develop cancer ; therapy for these malignancies will thus pose a significant challenge to physicians caring for these patients in the future.
Although many malignancies have been reported in association with HIV infection, only three malignancies are conclusively associated with HIV infection and are considered AIDS-defining conditions. These include Kaposi's sarcoma (KS), intermediate or high-grade non-Hodgkin's B-cell lymphomas (including primary CNS lymphoma), and cervical carcinoma [5,13,14]. Other malignancies (eg, Hodgkin's disease, anorectal carcinoma, pediatric smooth muscle tumors, noncervical gynecologic cancers, and nonmelanoma skin cancers) may be associated with AIDS but are probably underreported .
Prior to the AIDS epidemic, KS was seen mostly as an indolent, pigmented lesion involving the lower extremities in older men of Jewish, Eastern, or Mediterranean descent [15,16]. The disease rarely involved lymph nodes, mucous membranes, or visceral organs. Moreover, up to one third of these patients demonstrated a second primary malignancy, most frequently non-Hodgkin's lymphoma [15,16].
An endemic form of KS also occurs in children and adults in tropical Africa. In adults, it is a benign nodular form or a more aggressive local or visceral form with a median survival of 5 to 8 years [15,16]. In children, the disease involves visceral organs and lymph nodes with lymphedema. It is typically virulent and fatal in 2 to 3 years (Table 2) [15,16].
|Classic||Older men (aged 50-80 yr) of Jewish and Mediterranean heritage||Usually confined to lower extremities, often with venous stasis and lymphedema; M/F ratio 10-15:1||Indolent; survival 10-15 yr; 37% associated with other lymphoid malignancies|
|African||Young adult (aged 25-40 yr) black men in Central Africa||Localized nodular lesions (57%); large aggressive exophytic tumors or invasive to underlying bone (38%); M/F ratio 13:1||Indolent if nodular, otherwise slowly progressive and fatal within 5-8 yr|
|Children (aged 2-13 yr)||Generalized lymphadenopathy (5%); M/F ratio 3:1||Rapidly progressive; fatal within 2-3 yr|
|Renal transplant||Iatrogenically immunosuppressed patients||May be localized to skin or widespread with systemic involvement; M/F ratio 2.3:1||Can be indolent or rapidly progressive; may regress when immunosuppressive therapy is discontinued; fatal in 30%|
|Epidemic||AIDS patients; primarily homosexual men; few intravenous drug users and Africans||Disseminated mucocutaneous lesions often involving lymph nodes and visceral organs, especially gastrointestinal tract and lungs||Fulminant; less than 20% survival at 2 yr if associated with opportunistic infections|
|Adapted, with permission, from Krigel RL, Friedman-Kien AE: Epidemic Kaposi's sarcoma. Semin Oncol 17:350-360, 1990.|
Aggressive KS is the most prevalent cancer among HIV-1-infected patients, with estimated attack rates as high as 30% . The relative risk of KS in HIV-infected adult male homosexuals is 10,000- to 40,000-fold higher than in the general population [3,18]. However, this incidence has decreased over time. From 1981 to 1984, 50% of AIDS patients in San Francisco exhibited KS lesions during the course of their disease. By 1992, the rate had dropped to only 14% [3,13,15,16,18]. Yet despite this decrease, the overall incidence of AIDS-KS continues to rise along with the numbers of HIV-infected patients .
The prevalence of KS varies among different categories of AIDS patients by the route of HIV infection and by geographic location (Figure 1)[8,16]. Kaposi's sarcoma is six times more common in male homosexuals than in other risk groups, and 95% of AIDS-related KS cases in the United States and Europe are diagnosed in homosexual males . Kaposi's sarcoma also occurs more often in San Francisco, New York, and Los Angeles, which have been centers of the AIDS epidemic .
FIGURE 1: Prevalence of AIDS cases with kaposi's sarcoma in diferent groups of HIV-infected patients. Reprinted, with pemission, from Weiss RA: Retroviruses and human cancer. Semin Cancer Biol 3:321-328, 1992.
A history of anal-receptive intercourse or oral-fecal contact is linked to KS development, suggesting that a second infectious agent may be required for the development of KS [12,14,16]. This is supported by an observed decrease in the incidence of KS among homosexual males with changing sexual practices in the late l980s; the occurrence of KS in homosexual males who are HIV seronegative; the nearly equal incidence of KS among males and females in Africa, where HIV is spread by heterosexual contact; and the incidence of KS in women with a history of sexual contact with bisexual men . Despite extensive research, however, no “second agent” has been isolated. Other associations have not been proven to cause KS but instead serve as markers of a population at risk for HIV infection and KS because of lifestyle [12,16].
All forms of KS are characterized by a proliferation of spindle cells in a background network of reticular and collagen(Drug information on collagen) fibers; vascular and lymphatic proliferation, and the presence of mononuclear cells including macrophages, lymphocytes, and plasma cells. Lesions may involve only the reticular dermis (patch stage) or the full thickness of the dermis (plaque or nodular stage). As lesions evolve to the plaque and nodular stages, the number of interstitial cells increases. The abundance of cells is thought to reflect a cytokine-rich environment [3,12,16,20].
The cell of origin of KS lesions is unknown, but the lesions are thought to originate from a pluripotent mesenchymal precursor cell. Endothelial (factor VIIIa) and spindle cell markers and expression of the gene for smooth muscle alpha-actin have been observed in vitro, implicating a vascular or lymphatic endothelial cell or vascular smooth muscle cell as the cell of origin [3,13,16,20,21].
Pathogenesis of Kaposi's Sarcoma
Although KS lesions exhibit malignant behavior, it is unclear whether they are truly monoclonal malignant growths or benign, hyperplastic, polyclonal growths driven by continuous cytokine stimulation. No clonal cytogenetic abnormalities [12,17] and no oncogene rearrangements have been demonstrated in cell lines . It is theorized that early KS lesions are hyperplastic and may progress or regress depending on the patient's immune status and extent of proliferative stimuli. Under continuous stimulation, however, some cells may undergo genetic changes and true malignant transformation .
Kaposi's sarcoma cells maintained in culture can induce angiogenesis as well as produce cytokines that promote their own growth and the growth of normal endothelial cells, fibroblasts, and other mesenchymal cells . Angiogenic cytokines include interleukin (IL)-1-beta, basic fibroblast growth factor, acidic fibroblast growth factor, and endothelial cell growth factor [17,20]. Other cytokines include IL-6, granulocyte-macrophage colony-stimulating factor (GM-CSF, sargramostim(Drug information on sargramostim) [Leukine]), transforming growth factor beta, and platelet-derived growth factor alpha. Kaposi's sarcoma cells also respond to exogenously produced cytokines originating from HIV-infected T-cells including IL-1, fibroblast growth factor, IL-6, and oncostatin M [3,12,16,17,20,21]. Oncostatin M acts directly to stimulate KS cells and induces KS cells to produce IL-6, which acts to sustain growth in an autocrine fashion.
The HIV virus itself may be responsible for transformation. When Vogel et al transfected fertilized eggs from mice with a recombinant HIV transactivating (TAT) gene with a long terminal repeat sequence, they observed the development of lesions resembling KS in 15% of male offspring . The TAT gene product has also been shown to promote the growth of AIDS-associated KS cells . The process of transformation may be associated with the expression of certain receptors, such as IL-6 and oncostatin M, that distinguish KS cells from their normal counterparts.
Immunosuppression supports the evolution of KS but is not necessarily a prerequisite for KS development as KS can develop in HIV infection with normal or near-normal CD4 counts [13,24]. Other factors are likely involved in the pathogenesis of AIDS-associated KS. The distinct male predominance suggests a role for hormones in the pathogenesis of both classic and AIDS KS. Also, all forms of KS show a strong relationship to cytomegalovirus (CMV) exposure and a history of fecal-oral contact (the “second agent” theory).
AIDS-associated KS may appear at any stage of HIV disease. Immune impairment is usually present: less than one sixth of HIV-infected patients have CD4-positive T-cell counts of less than 500/mm³ .
Typical lesions range from violaceous to brown and may be flat or raised and ulcerated. They are usually multicentric and symmetrical and may be in various stages of development. The lesions do not blanch and are usually not tender. A biopsy must be obtained to exclude bacillary angiomatosis or pyogenic granuloma, which may have a similar presentation in AIDS patients [12,16]. The tumor is often widespread, involving the skin, mucous membranes, gastrointestinal tract, lymph nodes, genitalia, oral cavity, conjunctiva, and/or lungs and airways.
Oral KS is a marker of more advanced HIV infection. Patients exhibit CD4 counts of less than 200/mm³ and associated involvement of the gastrointestinal tract in 50% of cases. Gastrointestinal KS may be manifested by bleeding, diarrhea, or weight loss. Because barium enema may fail to demonstrate flat lesions, these patients should undergo endoscopy instead. Patients with pulmonary KS can present with shortness of breath, fever, cough, hemoptysis, or chest pain or be asymptomatic. Radiographic appearance is nonspecific and may demonstrate infiltrates, poorly defined nodules, or effusions. Effusions are exudative and often bloody [12,25]. Involvement of either the gastrointestinal (GI) tract or lung causes death in 10% to 20% of patients . Patients may also present with disease limited to nodes, in which case lymph node biopsy is required to establish the diagnosis. Significant lymphedema may occur and is cytokine mediated [12,15].
Staging and Prognostic Factors
The development of a universally accepted staging system for KS is complicated by the fact that the usual indicators of tumor burden in other metastatic cancers do not have the same prognostic significance in KS. However, Chachoua et al reported three adverse prognostic factors for survival in a cohort of epidemic KS patients: prior or coexistent opportunistic infection (OI), the presence of B symptoms (weight loss, fever, and night sweats), and an absolute CD4-positive T-cell count of less than 300 cells/mm³ . The most important of these, OI, was associated with a median survival of only 7 months vs 20 months for those without prior OI. Other features, including the ratio of helper to suppressor cells and extent of disease, were not independent predictors of survival in this study.
Based on these findings, the AIDS Clinical Trials Group Oncology Committee of the National Institute for Allergy and Infectious Diseases proposed a KS staging system incorporating extent of disease, severity of immune dysfunction, and the presence of systemic B symptoms (Table 3). They also recommended the following staging methods for these patients: complete physical examination (including rectal and oral examination) biopsy of skin lesions and/or lymph nodes chest x-ray, gastroscopy and colonoscopy (bronchoscopy in patients with abnormal chest x-ray), CT scan of the abdomen, and laboratory studies (complete blood count, common serum chemistries, HIV serology, T4-T8 lymphocyte counts).
|Good risk (all of the following)||Poor risk (any of the following)|
|Tumor (T)||Confined to skin and/or lymph nodes and/or minimal oral diseaseª||Tumor-associated edema or ulceration|
Extensive oral Kaposi's sarcoma
Gastrointestinal Kaposi's sarcoma
Kaposi's sarcoma in other nonnodal viscera
|Immune system (I)||CD4 cells = 200/µL or more||CD4 cells < 200/µL|
|Systemic illness (S)||No history of opportunistic infection or thrush|
No B symptoms*
Performance status = 70 or more (Karnofsky)
|History of opportunistic infection and/or thrush|
B symptoms present
Performance status < 70 (Karnofsky)
|Other HIV-related illness (eg, neurologic disease, lymphoma)|
|AIDS = acquired immunodeficiency virus, HIV = human immunodeficiency virus|
ª Minimal oral disease is nonnodular KS confined to the palate.
* B symptoms are unexplained fever, night sweats, more than 100% involuntary weight loss, or diarrhea persisting more than 2 weeks.
Adapted, with permission, from Krigel RL, Friedman-Kien AE: Epidemic Kaposi's sarcoma. Semin Oncol 17:350-360, 1990.
Therapy for KS patients is palliative and directed toward improving symptoms and overall quality of life since most patients with AIDS-associated KS die of opportunistic infections rather than KS [16,18]. Indications for treatment include cosmetic control, bulky oral lesions, lesions resulting in pain or significant edema, extensive cutaneous disease, or the presence of symptoms referable to viscera, such as bleeding or obstruction (Table 4)[28,29].
|Status of KS||Status of HIV disease||Kaposi's sarcoma treatmentsª|
|Few, small cutaneous lesions, not in exposed areas||CD4 count < 200/µL|
Prior opportunistic infection or B symptoms*
|No treatment |
|CD4 count > 200/µL|
No prior opportunistic infection or B symptoms
Local treatment of IFN + zidovudine(Drug information on zidovudine)
|Cosmetically unacceptable lesions||Any||Local treatments or radiation therapy|
|Extensive cutaneous KS with or without asymptomatic visceral KS||CD4 count < 200/µL|
Prior opportunistic infection or B symptoms
|Single- or multiagent chemotherapy or IFN + zidovudine|
|Localized bulky or painful KS||Any||Radiation therapy|
|Tumor-associated edema||Any||Radiation therapy or chemotherapy|
|Symptomatic visceral KS||Any||Multiagent chemotherapy|
|HIV = human immunodeficiency virus, IFN = interferon, KS = Kaposi's sarcoma|
ª In certain circumstances, different approaches may be combined for patients with multiple KS-related problems, eg, systemic chemotherapy with radiation or bulky disease sites.
* B symptoms: unexplained fever, night sweats, > 10% involuntary weight loss, or diarrhea persisting > 2 weeks.
Adapted, with permission, from Kraun SE et al: Kaposi's sarcoma. Med Clin North Am 76:235-252, 1992.
Criteria for defining the response of KS lesions to therapy are outlined in Table 5 .
|Complete response (CR)|
|No detectable residual disease, including tumor-associated edema or effusion, for at least 4 weeks. If residual pigmented lesions remain, biopsy of at least 1 representative lesion is required to document the absence of malignant KS cells. Restaging of known visceral disease with appropriate endoscopic or radiologic studies should be attempted. If such procedures are medically contraindicated (eg, invasive surgical restaging), the patient may be classified as having a clinical CR.|
|Partial response (PR)|
|A 50% or greater reduction in the number and/or size of previously existing lesions for at lest 4 weeks without:|
Patients with residual tumor-associated edema or effusion who otherwise meet the criteria for a CR will be classified as having a PR.
|Any response not meeting the cirteria for progression or PR|
|The presence of any of the following:|
|Adapted, with permission, from Pluda JM et al: Therapy of AIDS and AIDS-related tumors. Cancer Chemother Biolog Response Mod Annual 12:395-429, 1991.|
Cryotherapy of KS lesions with liquid nitrogen may result in a complete or partial response in more than 85% of cases regardless of anatomic location or activity of underlying HIV infection , although persistent KS can be demonstrated in the deep reticular dermis by biopsy. Patient candidates for cryotherapy include those with indolent KS and macular or papular lesions less than 1 cm in diameter. Two freeze-thaw cycles are recommended per treatment, to be repeated at 2- to 3-week intervals. The liquid nitrogen should be administered with a hand-held spray device rather than a cotton-tipped applicator to prevent transmission of HIV . Advantages of cryotherapy are the short duration of treatment, minimal pain, ease of administration and repeat treatments, and potential combination with other modalities of treatment.
Laser and Surgical Therapies
Argon-laser photocoagulation has been preferentially used to treat vascular lesions owing to the specific uptake of laser energy by oxygenated hemoglobin in the tissues . An argon laser is preferable to an Nd:YAG laser because of the Nd:YAG laser's tendency to cause bleeding. Laser photocoagulation therapy, which can be done on outpatients, can completely shrink smaller lesions, partially resolve larger lesions, lessen bleeding and pain, improve cosmetic appearance, and allow minimal wound care. However, like cryotherapy, it offers limited tissue penetration and is unlikely to resolve large, deep, or exophytic lesions.
Surgery has traditionally been reserved for lesions that cause visceral morbidity (bowel obstruction or bleeding) and for skin lesions that are large, are ulcerated, infiltrate underlying tissues or bone, or occur in areas that can cause morbidity, such as the face. Furthermore, the potential for HIV infection of the surgeon or laser operator remains a concern, and these modalities should be reserved for selected patients only.
Radiation therapy does benefit most patients with KS lesions, as studies have shown. Tappero et al reported that more than two thirds of patients attained at least a partial response to radiation of KS lesions . More recently, Berson et al achieved response rates of more than 80% with appropriate patient selection . Current indications for radiotherapy include bleeding, pain, mass effect (large intraoral lesions, localized painful lymphadenopathy, localized lymphedema of extremities or genitalia), or cosmetically disfiguring lesions at selected sites (facial, ocular, or periorbital lesions or lesions of the feet) that fail to respond to intralesional or cryotherapy .
Toxicities of radiotherapy include residual purple pigmentation, hyperpigmentation, desquamation, or ulceration in treated skin lesions; mucositis of the oral cavity, pharynx, and larynx (often ameliorated by prophylactic systemic antifungal and antiherpetic medication); and dry mouth or altered taste during treatment of oral lesions [34-36]. There is the suggestion that patients with oral lesions or lesions of the feet are especially sensitive to conventional doses of radiotherapy with enhanced toxicities that may limit the effectiveness of such therapy for these lesions [35,36]. Observed responses to radiotherapy vary by site of disease and the overall condition of the patient. Lesions treated to palliate pain or visceral symptoms are less likely to demonstrate objective responses and are more likely to recur locally than lesions treated for cosmetic reasons [33,35]. Lymphedema is also less likely to demonstrate a response: studies have shown only partial resolution in 40% of fields treated .
Intralesional injections of vinblastine(Drug information on vinblastine), vincristine, or bleomycin(Drug information on bleomycin) reportedly result in objective response rates of 60% to 88% with minimal or no systemic effects. As with cryotherapy, cosmetic response is often better than histologic response with residual disease by biopsy. Median duration of response vary from 4 to 6 months, with 40% of patients experiencing recurrent disease .
Reported concentrations of intralesional vinblastine vary from 0.1 mg/mL to 0.6 mg/mL, with larger doses recommended for oral lesions or larger papulonodular lesions. Total doses administered at a treatment session are limited to 1 to 3 mg [16,37-40]. Patients are allowed a recovery time of 3 to 4 weeks before retreatment. Lesions may require two to three injections for maximal response. Pain at the injection site is the most common side effect, but hyperpigmentation, edema, blistering and ulceration, alopecia, and transient mononeuropathy are also reported. Biological therapeutic agents used intralesionally include alpha interferon (IFN-alpha), tumor-necrosis factor, and platelet factor 4, though all remain investigational at this time [41-47].
Patients with widely disseminated, progressive, or symptomatic disease are candidates for systemic therapy, including biologic-agent therapies, antiretroviral therapies, or single or multiagent chemotherapies. However, systemic therapies may be limited by severe toxicities or the intercurrent development of opportunistic infections.
Single chemotherapeutic agents initially investigated in AIDS-associated KS include bleomycin (Blenoxane) and vincristine (Oncovin), both of which are minimally myelosuppressive. Responses with single agents are primarily partial responses and are measured in weeks [48-53]. The pulmonary toxicity resulting from use of bleomycin in AIDS-associated KS was evaluated by Ireland-Gill et al . They retrospectively reviewed 28 patients treated with bleomycin plus vincristine (BV) or doxorubicin(Drug information on doxorubicin) (Adriamycin, Rubex), bleomycin, and vincristine (ABV) who had undergone repeated pulmonary function testing of breathing capacity, diffusing capacity for carbon monoxide (DLCO), and lung volumes. A decline of more than 20% was considered clinically significant. The median bleomycin dose was 112 U (range, 10 to 313 U). Spirometry and lung volume tests showed no significant changes. A statistically significant difference in DLCO was observed for patients receiving cumulative bleomycin doses of more than 100 U vs those receiving less than 100 U.
Interferon is another single agent that has been tested in AIDS-associated KS. Its testing was prompted by its antiviral, antiproliferative, antiangiogenic, and immunoregulatory activities. Like other biological-therapy agents, interferon offers the advantages of less myelosuppression and fewer systemic toxicities. Responses have been reported to differ depending on extent of disease, prior treatment with chemotherapy, prior or coexistent OI, and CD4-positive T-cell count. IFN-alfa therapy is the only biologic therapy approved by the US Food and Drug Administration (FDA) for AIDS-associated KS .
A number of trials have confirmed the activity of single-agent interferon in AIDS-associated KS. Initial studies were designed to determine the optimal dose and schedule and whether a dose-response relationship exists for interferon . In general, higher doses resulted in an improved response rate, with responses of 20% to 40% reported with doses of more than 20 million units (MU)/m² [55,56]. Responders demonstrated improved survival over nonresponders. The most prominent toxicities included malaise and flu-like syndrome. Studies of combined interferon and cytotoxic chemotherapy demonstrated increased toxicity but not synergistic or additive activity [55,56].
Factors predicting a poor response to interferon include extensive disease, constitutional symptoms, a CD4 count of less than 200/mm³ at initiation of therapy, anemia, and current or prior OI [56,57]. Subsequent studies excluding patients with the worst prognostic indicators have resulted in improved response rates approaching 50%. Patients with CD4 counts of more than 200/mm³ are almost four times more likely to respond than patients with CD4 counts of less than 200/mm³.
It has been suggested that IFN-alfa acts against HIV by suppressing the translation of its mRNA into protein, thus blocking the assembly of viral proteins into intact virions. Zidovudine (Retrovir), although not directly antiviral, acts by blocking the infection of previously uninfected cells . When used in combination, these agents have demonstrated synergistic activity both in vivo and in vitro.
Trials combining interferon and zidovudine have been primarily phase I studies designed to establish the maximum tolerated dose of the combination. In one, Krown et al reported a maximum tolerated interferon dose of 18 MU with 100 mg zidovudine every 4 hours (an alternative schedule of 4.5 MU of interferon with 200 mg of zidovudine was also tolerated). However, the excessive toxicity of lymphoblastoid interferon (IFN-alfa-N1) and zidovudine resulted in early closure of the third study arm.
Dose-limiting toxicities included neutropenia (defined as a cell count of 500 to 750/mm³ and accounting for 10 of 17 dose reductions) as well as severe fatigue, malaise, and elevated transaminase levels. A decrease in hemoglobin concentration to 10 g/dL or lower was seen in all patients treated for 2 or more weeks. Although Krown's interferon-zidovudine studies were not designed specifically to evaluate response, a response rate of 46% was demonstrated. Similar results are reported in studies by Fischl et al  and Kovacs et al . The addition of GM-CSF to the combined regimen of interferon and zidovudine resulted in an improved absolute neutrophil count but not in an increased rate of tumor response, final CD4 count, or improvement in any other hematologic variable .
Lymphoblastoid interferon has also been studied in AIDS-associated KS. Gelmann et al studied three schedules (7.5, 15, and 30 MU/m²/d) for 28 days in 30 patients . Responses were not dose dependent but correlated with high total lymphocyte count, high CD4 counts, absence of OI, and absence of endogenous acid-labile IFN-alfa in the circulation.
Other biologic response modifiers including beta interferon (IFN-beta), gamma interferon (IFN-gamma) and IL-2 (aldesleukin [Proleukin]) have been investigated as therapeutic agents in KS, with disappointing results. A phase II trial of IFN-beta with a serine substitution at position 17 demonstrated a 16% response rate, with 42% of patients having stable disease. Toxicity included injection-site necrosis but rarely significant hematologic toxicity . IFN-gamma has demonstrated minimal activity with a response rate of less than 5% . IL-2 as a single agent has failed to demonstrate significant activity in AIDS-associated KS and in combination with IFN-beta, actually demonstrated disease progression in three of four patients .
Combination chemotherapeutic regimens are associated with higher reported overall response rates and complete response rates. Vincristine-, vinblastine-, and bleomycin-containing regimens are usually used as first-line therapy as they are well tolerated and produce minimal alopecia and bone marrow suppression. Etoposide(Drug information on etoposide) (VePesid)- and doxorubicin-containing regimens are generally used when the disease becomes resistant to the less myelosuppressive agents. Advanced KS with pulmonary or other visceral involvement requires anthracycline-containing regimens .
The optimal duration of chemotherapy is undefined and depends on both patient tolerance and response of KS lesions. Most combinations are administered every 2 weeks, and it is recommended that patients undergo approximately two cycles past best response. Attempts to prolong response with maintenance interferon therapy have resulted in only short responses (8 weeks), but this remains under investigation .
The combination of bleomycin and vincristine has been investigated in AIDS-associated KS [68-70]. Gill et al studied 18 patients with compromised bone marrow function (defined as absolute granulocyte counts of less than 1,500/mm³, hemoglobin levels of less than 10 g/dL, or platelet counts of less than 100,000/mm³ . Patients previously exposed to myelosuppressive agents such as ganciclovir(Drug information on ganciclovir) (Cytovene) or zidovudine or to previous single-agent chemotherapy were included. Patients received 10 mg/m² bleomycin and 1.4 mg/m² (2 mg maximum) vincristine biweekly. Objective responses were observed in 72% of patients. Response lasted 6 to 7 weeks in the two complete responders and a median of 8 weeks in the 11 partial responders.
Gompels et al retrospectively reviewed 46 patients who had received fixed doses of 2 mg vincristine as a bolus and 30 mg bleomycin as an infusion over 18 hours . Vinblastine (2.5 to 5 mg) was administered instead of vincristine if peripheral neuropathy developed. Treatment was repeated every 3 to 4 weeks. Fifty-seven percent achieved a partial response, 34% had stable disease, and none had complete responses.
Rarick et al retrospectively studied the use of zidovudine with the combination of BV in an attempt to reduce the incidence of opportunistic infections . All patients received 10 U/m² bleomycin and 2 mg vincristine in 2 week cycles but on different schedules of zidovudine. Patients were divided into two groups: Eight patients received a full dose of zidovudine (200 mg) every 4 hours; four patients received a half dose of zidovudine (100 mg) every 4 hours. The full-dose patients received a mean of 5.4 cycles of BV, while the half-dose group received a mean of 4.2 cycles of BV. In the full-dose group, 63% of patients achieved a complete response. In the half-dose group, no patients group achieved a complete response and 50% had a partial response. The overall response rate in both groups was 83%, with a median duration of response of 2 months. Only two patients developed infections while on therapy. However, the overall contribution of zidovudine to the response rate remains undetermined as this study was uncontrolled.
Doxorubicin-containing therapies have also been investigated. Gill et al compared the combination of ABV against doxorubicin alone in 61 patients with mucocutaneous KS . The overall response rate for ABV was 88% vs 48% for doxorubicin alone; however, the median survival was the same (9 months) in both groups. A higher dose version of ABV was studied by Laubenstein et al . In that study, the response rate was 84%, but 61% of patients developed OI and 44% required dose reductions to minimize hematologic toxicity . The AIDS Clinical Trials Group is also studying the use of ABV plus zidovudine with growth factor support.
Fumagillin analogs [71,72] and sulfated polysaccharide peptidoglycan compounds produced by bacteria [73,74] are potent inhibitors of angiogenesis and are currently being investigated for activity in AIDS-associated KS. Recombinant platelet factor 4 is also being investigated as an antiangiogenic agent [45,46].
Because cytokines are so fundamental in the development and maintenance of AIDS-associated KS lesions, the use of inhibitors of cytokines or receptor antagonists is being investigated. Interleukin-4 is known to be a potent inhibitor of IL-6 in monocytes, and its activity in AIDS-related KS is now being studied . Pentosan polysulfate, an inhibitor of basic fibroblast growth factor, is also being investigated .
Other strategies under investigation include novel methods of chemotherapeutic drug delivery (liposomal doxorubicin [76-78]), photodynamic therapy , use of differentiating agents such as all-trans-retinoic acid , and the combined use of chemotherapy and biologicals with less myelosuppressive nucleoside analogs such as dideoxyinosine (ddI, didanosine(Drug information on didanosine) [Videx]) and dideoxycytidine (ddC, zalcitabine(Drug information on zalcitabine) [Hivid]). New chemotherapeutic agents including paclitaxel(Drug information on paclitaxel) (Taxol) are being investigated for efficacy and toxicity .
The Surveillance, Epidemiology, and End Results (SEER) program of the National Cancer Institute has demonstrated a more than 50% increase in non-Hodgkin's lymphoma (NHL) cases from 1973 to l987 [82,83]. Although the factors contributing to this increase are not fully understood, the increased incidence of NHL in HIV-infected individuals has contributed in some part to this figure. AIDS-associated NHL is not entirely responsible for the increased incidence, however, as the increase predates the AIDS epidemic. The incidence of B-cell NHL has conclusively been demonstrated to be increased in HIV-infected individuals and as such is an AIDS-defining disease in approximately 3% of individuals newly diagnosed with AIDS [82,84].
SEER data from 1973 to 1979 (pre-AIDS) revealed a steady rise in incidence of NHL with age, from 3.4/100,000 in men 20 to 24 years old to 68/100,000 in men more than 75 years old. The median age of diagnosis was 63 years for men and 68 years for women. Rates were one third lower for women than men at all ages, and the incidence in blacks was one third lower than in whites of both sexes. By 1987 (during the AIDS epidemic), the nationwide incidence of NHL in the 20- to 49-year age group rose from 6.3 to 11.7. In San Francisco, the incidence rose dramatically from 7.7 to 59.0 . By l985, the incidence of NHL was approximately 60 times greater in persons with AIDS than in the general population (Table 6) [85,86], and Burkitt's lymphoma and primary CNS lymphoma were 1000 times more frequent in persons with AIDS .
|Age (yr)||Observed (n)||Expected (n)ª||Observed/|
|ª Based on the US incidence of non-Hodgkin's lymphoma from 1981-1985.|
Adapted, with permission, from Northfelt DW, Kaplan LD: Clinical aspects of AIDS-related NHL. Curr Opin Oncol 3:872-880, 1991.
The SEER data for NHL also showed a concordance between geographic areas with high AIDS incidence and high NHL incidence. Using marital status as a surrogate marker for homosexual behavior, significant increases in NHL were noted in geographic areas, such as New York, San Francisco, and Los Angeles, with the highest incidence of HIV infection [83,84].
As the AIDS epidemic has evolved, it has become apparent that all HIV risk groups are at risk for the development of NHL. The World Health Organization European Region measured the frequency of NHL in all groups from children infected perinatally to hemophiliacs, intravenous (IV) drug abusers, and homosexual and bisexual men . They noted a higher frequency of NHL in homosexual and bisexual men and hemophiliacs than in IV drug abusers, who exhibited a downward trend in NHL. This same pattern was recorded in the largest US study of AIDS-associated NHL (Table 7). A bimodal distribution of NHL was observed, with a peak in adolescence (ages 10 to 19) and a second peak in middle age (ages 50 to 59). It is suggested that the early age peak is attributable to Burkitt's lymphoma and that the second peak is attributable to immunoblastic and diffuse B-cell lymphomas .
|HIV transmission group||Patients with any |
|Relative risk |
|Hemophiliac or clotting disorder||5.2 (7/14/33)||1.66º (1.07-2.56)|
|Homosexual or bisexual men||3.4 (420/468/1334)||1.13 (0.86-1.49)|
|Transfusion recipient||3.5 (17/15/58)||1.00|
|Children infected perinatally||1.3 (8/4/4)||0.90 (0.41-1.96)|
|Heterosexual contact, except those born in the Caribbean or Africa||1.9 (14/9/32)||0.77 (0.50-1.18)|
|Intravenous drug user||1.6 (64/60/169)||0.60º (0.43-0.85)|
|Heterosexual contact, born in the Caribbean or Africa||1.0 (2/4/9)||0.41º (0.19-0.90)|
|HIV = human immunodeficiency virus, HIV = human immunodeficiency virus|
ª Does not include 79 patients with unknown HIV-transmission group
*Number of cases of primary brain/Burkitt's immunoblastic lymphoma shown in parentheses
^ Relative risk with reference to transfusion recipient, adjusted for age, sex, race, and year of AIDS diagnosis
° Significantly different from 1.0 (P < .05)
Adapted, with permission, from Northfelt DW, Kaplan LD: Clinical aspects of AIDS-related NHL. Curr Opin Oncol 3:872-880, 1991.
Non-Hodgkin's lymphoma can develop in HIV-infected patients at any time during the course of the illness. Northfelt et al examined the degree of immunodeficiency in terms of CD4 cell counts at the time of diagnosis of non-CNS AIDS-associated NHL . Seventy-nine percent of patients had CD4 counts of more than 50, and 58% had counts of more than 100. The median CD4 count at diagnosis was 110. The authors concluded that the degree of immunodeficiency at diagnosis as measured by CD4 cells varies over a wide range and that no specific CD4 count is a useful marker for the development of non-CNS AIDS-associated NHL. These data contrast with those from a study by Pluda et al, which suggested that NHL is a late manifestation of HIV infection . In this study, AIDS-associated NHL developed in 14.5% of patients after a median duration of 23.8 months of antiretroviral therapy. However, 50% of these patients had primary CNS lymphoma. There is evidence to suggest that primary CNS lymphoma is different from AIDS-associated NHL outside of the CNS and is indeed a late finding in HIV infection . However, others have also suggested that lymphoma is a late manifestation of HIV infection since the incubation period between infection and the development of lymphoma in transfusion recipients is approximately 50 months, similar to that for the development of opportunistic infections .
Pathologic Subtypes of Lymphoma
Most AIDS-associated lymphomas (80% to 90%) are high-grade B-cell neoplasms consisting of either immunoblastic or small noncleaved cell lymphomas. This is in contrast to the case in patients without HIV infection, in whom 10% to 15% are diagnosed with high-grade B-cell lymphomas [83,84,92]. Intermediate-grade large-cell, B-cell lymphomas are also observed with greater incidence in HIV-infected patients, although less commonly than high-grade lymphomas. In addition, Ioachim found that nearly twice as many extranodal lymphomas were of high-grade histology, whereas nodal lymphomas were equally divided between high-grade and intermediate-grade histology .
Other B-cell malignancies, including low-grade small-cleaved-cell lymphoma, chronic lymphocytic leukemia , and myeloma , have been reported in HIV patients. T-cell neoplasms, including cutaneous T-cell lymphoma [96,97], precursor T-cell lymphoma , lymphoblastic lymphoma , HTLV-I-associated T-cell leukemia , peripheral T-cell lymphoma , and Ki-1-positive anaplastic T-cell lymphoma , have been described in HIV-infected individuals. These malignancies, however, have not increased in incidence and are not considered part of the AIDS epidemic.
Although the precise mechanisms of lymphoma development in AIDS have not been determined, host factors and not environmental factors are predominantly responsible for such lymphomas. HIV itself is not a transforming virus and thus is not directly involved in malignant B-cell transformation. HIV sequences are not uniformly detected in lymphoma tissue or in the reactive B-cell hyperplasia of persistent generalized lymphadenopathy which precedes the development of lymphoma in 30% of cases . Polymerase chain reaction analysis reveals HIV present only in infiltrating T-cells. HIV infection thus does not appear to be an absolute prerequisite for the development of these lymphomas; in fact, more important than HIV infection itself is the immune dysregulation it causes .
HIV-infected cells produce multiple cytokines, some of which serve as stimuli for B-cell proliferation and differentiation. These include IL-1, 2, 4, 6, 7, 10, IFN-gamma, tumor-necrosis factor, lymphotoxin, and B-cell growth factor. In particular, IL-6 is an autocrine growth factor for B-cell malignancies such as multiple myeloma and chronic lymphocytic leukemia. HIV may also directly stimulate IL-6 from monocytes and macrophages, and elevated levels of IL-6 may predict the development of lymphoma in HIV infection [92,103]. Also, Epstein-Barr Virus (EBV)-positive B-cell lines from AIDS-associated Burkitt's lymphoma patients express large amounts of IL-10, suggesting this cytokine's role in B-cell growth and immortalization. In addition, HIV itself is capable of direct polyclonal activation of B-cells.
Ongoing B-cell proliferation and differentiation may result in an increased incidence of random mutations, which may in turn result in transformation . The normally occurring DNA rearrangements involving the immunoglobulin heavy- and light-chain genes may provide vulnerable sites for mutation or translocations . Molecular events may occur stepwise, with several molecular events required to induce transformation. Patients with reactive lymphadenopathy demonstrate multiple clonal rearrangements of immunoglobulin genes . These lesions may be early precursor lesions, with additional molecular events required for transformation of a single malignant clone. The classic translocations of Burkitt's lymphoma—t(8;14), t(8;22), and t(8;2)—are all demonstrated in AIDS-associated Burkitt's lymphomas as well.
Rearrangements involving the c-myc oncogene have been observed and thus are also implicated, in AIDS-associated lymphomas . Such rearrangements of c-myc imply a molecular mechanism similar to that in sporadic Burkitt's lymphomas. In addition, HIV infection of already immortalized B-cell lines can lead to up regulation of c-myc transcripts. C-myc activation in turn may result in altered phenotypic features that allow cells to escape immune surveillance by cytotoxic T-cells, including absence or low expression of class I major histocompatibilty complex antigens, insufficient expression of adhesion molecules required for effector-target cell interactions, and downregulation of EBV-coded antigens . However, no consistent c-myc rearrangements occur in AIDS-associated lymphoma tissue, suggesting the presence of other operative mechanisms .
The role of EBV in the development of AIDS-associated lymphoma has been suggested but is controversial. Collectively, data suggest that the virus is responsible for at least a portion of cases of AIDS-associated lymphomas and likely involved in all cases of primary CNS lymphoma in AIDS patients. A proportion (less than 50%) of systemic AIDS-associated lymphomas show signs of latent EBV infection and may also express combinations of EBV antigens not previously observed in B-cell lymphomas . Lymphomas that are both EBV- and HIV-positive demonstrate evidence of clonal EBV infection, indicating that EBV integration occurs before clonal B-cell expansion and that EBV may play a role in lymphomagenesis . In this study, approximately one third of patients with reactive generalized lymphadenopathy demonstrate EBV DNA, and the presence of EBV DNA correlates with development of lymphoma at a later time .
Mutation or allelic loss of tumor suppressor genes has also been investigated in AIDS-associated NHL. For example, some smaller studies indicate that p53 mutations occur frequently in NHL (up to 37% in one study), especially in association with c-myc overexpression. In contrast, there is no evidence of retinoblastoma gene inactivation in AIDS-associated NHL . The defect in tumor suppression and the deregulation of oncogene-induced growth may be central to the pathogenesis of these lymphomas. Also, viral proteins may bind to and inactivate p53, leading to dysregulation of cell growth .
In summary, B-cell proliferation induced by HIV or EBV may foster mutations in critical oncogenes, or tumor suppressor genes may occur along with abnormal DNA rearrangements, leading to c-myc activation, clonal selection, and the development of a monoclonal B-cell lymphoma.
Patients often seek medical attention for B symptoms (fever, night sweats, and weight loss of more than 10% of body weight) or for a rapidly growing mass lesion. Though these symptoms occur in approximately 75% of patients with AIDS-associated lymphomas, it is important to exclude other etiologies of these symptoms such as occult opportunistic infections.
Most patients with AIDS-associated NHL have advanced-stage disease at presentation and frequent involvement of extranodal sites. These features have been confirmed in multiple series. Between 64% to 83% of patients present with stage III or IV disease and between 65% and 91% with extranodal disease [110,111]. The most common extranodal sites in all reported series are bone marrow (25%), CNS parenchyma or meninges (32%), liver (12% to 48%), and gastrointestinal tract (26%)[112,113]. Other reported sites include lung , pleura , rectum , testis [116,117], kidney , spleen , and heart . The incidence of rectal NHL described in US series consisting largely of homosexuals is not duplicated in Italian series consisting largely of IV drug abusers .
Staging and Prognostic Features
Besides routine chest x-ray and blood studies, patients with AIDS-associated NHL should undergo bone marrow biopsy and lumbar puncture. Computed tomography (CT) of the chest and abdomen, and CT or magnetic resonance imaging (MRI) of the brain or spinal cord, bone scans, and gastrointestinal contrast studies should be done as needed if symptoms suggest disease. Disease should be staged according to the Ann Arbor system (Table 8).
|I||Involvement of a single lymph-node region (I) or a single extralymphatic organ or site (IE).|
|II||Involvement of two or more lymph-node regions on the same side of the diaphragm (II) or localized involvement of an extralymphatic organ or site (IIE).|
|III||Involvement of lymph-node regions on both sides of the diaphragm (III) or localized involvement of an extralymphatic organ or site (IIIE) or spleen (IIIS) or both (IIISE).|
|IV||Diffuse or disseminated involvement of one or more extralymphatic organs with or without associated lymph-node involvement. The organ(s) involved should be identified by a symbol: A, asymptomatic; B, fever, sweats, weight loss > 10% of body weight.|
|Adapted, with permission, from DeVita Jr VT, Hellman S, Rosenberg S (eds): Cancer: Principles and Practice of Oncology, p 1886. JB Lippincott, Philadelphia, 1993.|
Although patients with AIDS-associated NHL are a diverse group, several common factors predicting for short survival have been identified. Levine et al retrospectively studied a group of 49 patients treated for systemic AIDS-associated lymphoma . A Karnofsky performance status of less than 70%, history of AIDS before the diagnosis of lymphoma, and bone marrow involvement were independently associated with poor prognosis. In the absence of all three risk factors, median survival was 11.3 months; for the remaining patients, it was only 4 months. A complete response to therapy was associated with prolonged survival in the good-prognosis group (17.8 months vs 5 months in patients without complete response) but not in the poor-prognosis group. Patients in either group who attained a complete response (CR) to antilymphoma therapy remained at risk for dying of AIDS during lymphoma remission. Thus, attempts at prolonging survival must address both the neoplasm and the underlying HIV infection. The median survival for all patients with AIDS-associated NHL remains approximately 6.5 months .
Therapy for AIDS-Related Non-Hodgkin's Lymphoma
Hematologic Effects of HIV Infection and HIV Therapies: The major factor limiting the use of chemotherapy in AIDS-related NHL either alone or with antiretroviral therapy is hematologic toxicity and poor bone marrow reserve at initiation of therapy. Infection with the HIV virus often results in bone marrow dysplasia, anemia, thrombocytopenia, and leukopenia. Both ineffective hematopoiesis and peripheral destruction of blood cells are responsible for cytopenias. Decreased survival of blood cells may be related to autoimmune phenomena as well as increased turnover driven by multiple infections (Table 9).
|Adapted, with permission, from Zauli G, Dans BR: Role of HIV infection in the hematologic manifestations of HIV seropositive subjects. Crit Rev Oncol Hematol 15:271-283, 1993.|
Between 70% and 95% of AIDS patients are anemic at presentation, with mean hemoglobin levels ranging between 9.1 and 11.7 g/dL [122,123]. Less ill patients with HIV exhibit anemia in 5% to 16% of cases . However, the incidence of anemia and other cytopenias increases with the degree of immunologic dysfunction and with progression from HIV seropositivity to frank AIDS.
Anemia in HIV disease is normochromic and normocytic. Seventy percent of patients receiving zidovudine demonstrate macrocytosis, with a mean corpuscular volume (MCV) of more than 100/fl after 2 weeks of therapy. Anemic HIV patients demonstrate an inappropriately low reticuloctye count, suggesting a hypoproliferative anemia or ineffective hematopoiesis. Patterns revealed by iron studies are consistent with chronic disease marked by an increased serum ferritin and decreased serum iron and total iron binding capacity. Serum ferritin levels parallel disease activity and are higher in AIDS than in HIV seropositivity. Serum ferritin levels may also be elevated owing to the protein's role as an acute-phase reactant .
A positive Coombs' test is observed in 20% of HIV-infected patients with hypergammaglobulinemia. Hemolytic anemia, however, is rarely observed. Nonspecific attachment of other antibodies to red cells is also frequently observed and is usually clinically silent .
The incidence of thrombocytopenia also increases with increasing severity of immune dysregulation, ranging from 5% to 12% in HIV-seropositive patients and reaching as high as 30% in patients with AIDS . Patients can exhibit classic immune thromocytopenic purpura (with increased numbers of megakaryocytes in the bone marrow) or may develop thrombocytopenia as a result of impaired thrombopoiesis, the toxic effect of medications, or a thrombotic thrombocytopenic purpura (TTP)-hemolytic uremia syndrome (HUS). Furthermore, therapies in these patients vary in effect. Therapy with prednisone(Drug information on prednisone) offers a durable response in only 10% to 20% of patients and may further worsen immune status. Zidovudine may increase platelet counts, though how it does so is unknown . The most effective therapy seems to be intravenous gamma globulin (IVIG), which has a response rate of 88% but a median response duration of only 3 weeks .
Leukopenia, encompassing both granulocytopenia and lymphocytopenia, is also observed in up to 75% of AIDS patients. Atypical lymphocytes, hyposegmented granulocytes, and vacuolated monocytes can be seen on peripheral smears. Phagocytosed pathogens may be demonstrated in neutrophils or monocytes [122,123]. Antibodies to granulocytes are also frequent in HIV disease, and though they do not necessarily correlate with the degree of neutropenia, they do correlate with progression to AIDS .
Anemia, leukopenia, and pancytopenia are well-documented effects of treating HIV infection with zidovudine (Table 10). Such treatment is also associated with lower mean hemoglobin levels and an increased need for red cell transfusions . Up to 20% of patients may develop severe neutropenia (less than 500 cells/mm³). However, because zidovudine's toxicity is dose dependent, reducing the dose often ameliorates cytopenias.
|AZT (azidothymidine, zidovudine, 3´-azido-2´,3´ dideoxythymidine)|
ddC (dideoxycytidine, 2´,3´-dideoxycytidine)
ddI (dideoxyinosine, 2´,3´-dideoxyinosine)
d4T (2´,3´-dideoxythymidinene, 2´,3´-dideoxy-2´,3´-didehydrothymidine)
|Adapted, with permission, from Pluda JM et al: Hematologic effects of AIDS therapies. Hematol Oncol Clin North Am 5:229-248, 1991.|
In addition, zidovudine-related myelosuppression and synergistic effects may occur when zidovudine is combined with other drugs commonly administered to treat infections such as ganciclovir, pentamidine, trimethoprim(Drug information on trimethoprim)-sulfamethoxazole, pyrimethamine, sulfadiazine(Drug information on sulfadiazine), dapsone(Drug information on dapsone), and amphotericin B(Drug information on amphotericin b). Other retroviral drugs such as suramin, ribavirin, ddC, ddI, and interferon also produce hematologic toxicity. In particular, studies of the combined use of interferon and zidovudine in Kaposi's sarcoma have demonstrated significant hematologic toxicity.
Chemotherapy: The results of trials of therapeutic regimens for AIDS-associated NHL are summarized in Table 11 [94,111,125-144]. Comparison of these regimens is limited, however, by the fact that various histologies are grouped together in each study, inclusion criteria vary among the studies, and trials have been both retrospective and prospective. However, several conclusions can be drawn from the experience to date in treating AIDS-associated NHL.
|Regimen||CR rate||Median survival||Reference|
|Cyclophosphamide, vincristine, methotrexate(Drug information on methotrexate), and intrathecal methotrexate||7/25||3 months||125|
High-dose methotrexate, high-dose cytarabine, plus cyclophosphamide(Drug information on cyclophosphamide), vincristine, bleomycin, prednisone, and asparaginase(Drug information on asparaginase)
|Lomustine, etoposide, cyclophosphamide, and procarbazine(Drug information on procarbazine)||43%||7.5 months||128|
|Cyclophosphamide, doxorubicin, and etoposide||14/18||17.4 months||129|
|Various regimens including cyclophosphamide, doxorubicin, vincristine, methotrexate, bleomycin, prednisone, and chlorambucil(Drug information on chlorambucil)||9/23||7 months||130|
|Various by histology||94|
|Chemotherapy plus retrovirals|
|Cyclophosphamide, doxorubicin, teniposide(Drug information on teniposide), prednisone, vincristine, bleomycin plus zidovudine concurrent||4/29||4 months||131, 132|
|Modified LNH regimen plus zidovudine in consolidation||88/136||9 months||133|
|12-week modified MACOP-B plus zidovudine in consolidation||10/30||8.1 months||134|
|CHOP plus zidovudine and alpha interferon||18/28||82 days to 641 days||135|
|Etoposide, vincristine, prednisone, cytarabine(Drug information on cytarabine), mitoxantrone(Drug information on mitoxantrone), cyclophosphamide plus antiretrovirals in consolidation||4/29||4.5 months||136|
|Low-dose m-BACOD plus zidovudine in consolidation||16/42||5.6 months||137|
|Chemotherapy plus growth factors|
CHOP plus delayed GM-CSF
CHOP plus early GM-CSF
|Cyclophosphamide, etoposide, doxorubicin, vincristine, methotrexate, plus daily zidovudine and GM-CSF||4/10||—||139|
|m-BACOD plus GM-CSF||8/16||14 months||140|
|LNH 84 or CHVmP/VCR-BLM plus G-CSF vs |
|Cyclophosphamide, doxorubicin, etoposide, plus ddI and G-CSF||4/7||—||142|
|B4-blocked ricin||1/9||6.6 months||143|
|CHOP = cyclophosphamide, doxorubicin, vincristine, prednisone; COMET-A = cyclophosphamide, vincristine, methotrexate, etoposide, and cytarabine; CR = complete response; ddI = dideoxyinosine; G-CSF = granulocyte colony-stimulating factor; GM-CSF = granulocyte-macrophage colony-stimulating factor; LNH = Lymphoma Non-Hodgkin's Study Group; m-BACOD = CHOP plus methotrexate, bleomycin; MGBG = methyl-glycosaminoglycan; ProMACE-MOPP = prednisone, methotrexate, leucovorin, doxorubicin, cyclophosphamide, etoposide; VCR-BLM = vincristine-bleomycin|
ª Overall response rate
(1) The favorable results of dose escalation in non-AIDS-associated high-grade lymphomas do not translate to AIDS-associated lymphomas. This has been illustrated in studies by Gill et al , Kaplan et al , and Levine et al . The study by Gill et al accrued patients sequentially on phase II studies of m-BACOD (methotrexate, leucovorin, bleomycin, doxorubicin, cyclophosphamide, vincristine, dexamethasone(Drug information on dexamethasone)) and a newly developed regimen containing high doses of cytarabine and methotrexate as well as cyclophosphamide, vincristine, prednisone, bleomycin, and asparaginase (Elspar). The regimen was designed to expose patients to high-dose cytarabine and high-dose methotrexate early in the course of therapy to prevent the CNS relapse observed in two thirds of the earlier m-BACOD group. Complete remission was achieved in 54% of the m-BACOD group but in only 33% of the high-dose group, with significantly greater numbers of patients with OI and hematologic toxicity appearing in the high-dose group. Median survival in the m-BACOD group was 11 months vs only 6 months in the high-dose group. The trial was terminated early because of the significant toxicity and high rate of OI.
The trial by Kaplan et al  employed a new chemotherapy regimen, COMET-A, consisting of cyclophosphamide, vincristine, methotrexate, etoposide, and cytarabine with CNS prophylaxis. This regimen was compared with standard therapies, including CHOP, m-BACOD, ProMACE-MOPP, COMLA, CVP, COMP,* and radiation therapy alone. Although the complete remission rate for COMET-A was 58%, 28% of patients developed OI and the median survival was only 5.2 months. Patients receiving “dose-intensive” regimens, defined as regimens having a cyclophosphamide dose of more than 1 g/m², had significantly shorter survival (median, 4.6 months) than patients receiving less intense regimens (median, 12.2 months).
Levine et al  have reported results for a regimen of low-dose m-BACOD with early CNS prophylaxis and the initiation of zidovudine at completion of chemotherapy. A complete remission rate of 46% was reported, with long-term lymphoma-free survival in 75% of complete responders suggesting that low-dose therapy may be effective. No patient had an isolated CNS relapse; however, despite prophylaxis, OI developed in 20% of patients.
(2) Despite the advent of lower dose regimens designed to minimize toxicity, significant hematologic toxicities are still observed and often result in delays in therapy, dose reductions, or termination of therapy. The use of growth factors has been shown to ameliorate these toxicities and allow improved dose intensity. However, the clinical significance of the effects of growth factors on HIV replication are unknown.
Walsh et al and the AIDS Clinical Trials Group conducted a phase I study of m-BACOD with GM-CSF . Three different doses of m-BACOD were used along with fixed doses of GM-CSF (Table 12). Eight of 16 patients were able to be treated at the highest dose level, and none of these patients experienced dose-limiting hematologic toxicity. However, OI developed in one patient on dose level 1 and in one patient on dose level 2, and the level of p24 antigenemia increased in some patients given GM-CSF.
|Level 0||Level 1||Level 2|
|Bleomycin||4 U/m² d 1|
|Doxorubicin||25 mg/m² d 1||35 mg/m²||45 mg/m²|
|Cyclophosphamide||300 mg/m² d 1||450 mg/m²||600 mg/m²|
|Vincristine||1.4 mg/m² d 1|
|Dexamethasone||6 mg/m² d 1-5|
|Methotrexate||200 mg/m² day 15 followed at 24 h by leucovorin 25 mg q6h for six doses|
|GM-CSF||20 µg/kg SC days 3-13 of the cycle|
|Cytosine arabinoside||50 mg × four doses intrathecally in the first month|
|Adapted, with permission, from Walsh et al Phase I trial of m-BACOD and Granulocyte Macrophage Colony Stimulating Factor in HIV-Associated NHL. Journal of Acquired Immune Deficiency Syndromes 6:265-271, 1993.|
Kaplan et al also studied the use of GM-CSF with the CHOP regimen . Patients receiving GM-CSF all had higher mean nadirs of absolute neutrophil count, shorter mean duration of chemotherapy, fewer chemotherapy cycles complicated by neutropenia and fever, fewer days hospitalized for fever and neutropenia, fewer reductions in chemotherapy doses, and less frequent delays in chemotherapy administration. However, these patients had a significant increase in p24 antigenemia (243%) baseline by week 3 after initiation of chemotherapy, suggesting stimulation of HIV activity.
(3) Opportunistic infections and recurrent lymphoma remain the primary causes of death in AIDS-associated NHL patients. Gisselbrecht et al treated non-CNS AIDS-associated lymphomas with a modified LNH 84 regimen consisting of three cycles of induction therapy with doxorubicin, cyclophosphamide, vindesine(Drug information on vindesine), bleomycin, and prednisone . Patients who achieved a CR received consolidation therapy containing high-dose methotrexate, ifosfamide(Drug information on ifosfamide) (Ifex), etoposide, asparaginase, and cytarabine. All patients received CNS prophylaxis. Sixty-five percent achieved complete remission after induction therapy, 22% had a partial remission or failed, and 15% died during induction therapy. The median survival was 9 months. The authors performed a concurrent prospective analysis of prognostic determinants and confirmed that a performance status more than 1, localized disease (stage I or II), absence of bone marrow involvement, nonimmunoblastic histology, absence of B symptoms, no previous AIDS-defining diagnosis, and a CD4 count of more than 100/mm³ were predictors of improved survival. They also found that 50% of patients having CD4 counts of more than 100, no B symptoms, a performance status of less than 2, and nonimmunoblastic histology had a 2-year survival, suggesting that factors that reflect the patient's underlying immunodeficiency are at least as important and perhaps more important determinants of survival than are factors intrinsic to the lymphoma itself.
Unresolved issues in the therapy for AIDS-associated NHL include the significance and influence on prognosis of pathologic subtype, the selection of therapy based on prognostic indicators, the contribution of antiretroviral therapy to overall survival, the contribution of dose-intensive therapy, and the effect of growth factors on retroviral replication and activity. The AIDS Clinical Trials Group is currently conducting a multi-institutional trial that stratifies patients to either the low-dose m-BACOD regimen or standard-dose therapy to determine if dose intensity improves survival. Growth factor support is being used in both groups to minimize toxicity.
Prior to the AIDS epidemic, primary central nervous system (PCNS) lymphoma was a rare disorder accounting for 1% to 2% of all cases of NHL and fewer than 5% of all cases of primary intracranial neoplasms . Increasing numbers of cases of PCNS lymphoma were reported in the 1970s, paralleling the increasing number of patients with congenital and iatrogenic immunosuppression. The incidence of PCNS lymphoma increased significantly in the l980s with the onset of the AIDS epidemic and was designated an AIDS-defining disease early in the epidemic. As many as 6% of AIDS patients may develop PCNS lymphoma during their illness .
Interestingly, there has been a simultaneous threefold increase in the incidence of PCNS lymphoma in patients who are immunocompetent. The cause of this increase is unknown. If the incidence continues to increase in both immunosuppressed and immunocompetent patients at the present rate, lymphoma will be the most common primary malignant neoplasm of the CNS by the year 2000 .
To learn more about PCNS lymphoma, Fine and Mayer retrospectively analyzed PCNS lymphoma in 792 immunocompetent and 315 AIDS patients (Table 13). The mean age of patients with AIDS was significantly less than that of non-AIDS patients (30.8 vs 55.2), and the ratio of men to women with disease was altered in AIDS lymphomas, with men representing a greater proportion of AIDS PCNS lymphoma patients. Eighty percent of the AIDS patients had a history of OI, and 20% had PCNS lymphoma as their AIDS-defining illness.
|Characteristic||Immunocompetent patients||Patients with AIDS|
|Male:female (ratio)||442:328 (1.35)||118:16 (7.38)|
|Mean age, yr||55.2||30.8|
|History of opportunisitic infection or Kaposi sarcoma, n (%)||Not available||115 of 143 (80)|
|Mean duration of symptoms before diagnosis, mo||2.8||1.8|
|Mental status changes||34.6||53.3|
|Increased intracranial pressure||32.4||14.2|
|Adapted, with permission, from Fine HA, Mayer RJ: Primary central nervous system lymphoma. Ann Intern Med 119:1093-1104, 1993.|
Clinical Presentation and Diagnosis
The most common presenting symptoms of PCNS lymphoma are headache, cranial nerve palsies or focal neurologic deficits, new-onset seizures, hemiparesis, signs of increased intracranial pressure such as nausea and vomiting or papilledema, mental status changes, or subtle cognitive or personality changes (Table 13).
The differential diagnosis of neurologic abnormalities in AIDS is long (Table 14), and diligent evaluation including biopsy may be necessary to obtain the correct diagnosis. PCNS lymphoma is second only to toxoplasmosis as a cause of focal cerebral masses in AIDS. CT imaging reveals lymphomas to involve the cerebrum, brain stem, and cerebellum in decreasing order of frequency [145,148], and lesions are often periventricular in location and necrotizing. Extension to the ventricles or subarachnoid space may allow cytologic diagnosis through the cerebrospinal fluid .
|Primary or indirect effect of HIV-1|
Spinal cord and peripheral nervous system involvement
|AIDS = acquired immunodeficiency syndrome, HIV = human immunodeficiency virus|
Adapted, with permission, from Davenport E et al: Neuroradiology of the immunosuppressed state. Radiol Clin North Am 30:611-637, 1992.
Lymphoma lesions may appear as single or multifocal masses that are isodense, hyperdense, or hypodense, and thus difficult to distinguish from other mass lesions such as toxoplasmosis. However, lymphomas tend to exhibit larger and fewer (more than 3 cm) lesions than toxoplasmosis . Ten percent of lesions may be radiographically occult . In contrast to non-AIDS-associated lymphomas, which almost never exhibit ring enhancement, AIDS-associated lymphomas are twice as frequently multifocal and have ring (50% of cases) or solid enhancement.
PCNS lymphomas usually appear hypointense on T1-weighted MRI images and isointense or hyperintense on T2-weighted images . Although the diagnosis can be suggested by CT or MRI, a definitive diagnosis by brain biopsy is recommended, especially in patients with negative toxoplasmosis titers and whose condition worsens on empiric antitoxoplasmosis therapy [113,148]. Although cerebrospinal fluid (CSF) analysis in these patients is frequently abnormal (pleocytosis, elevated protein, and decreased glucose), cytologic examination is diagnostic in less than one third of cases [112,145]. CSF may also be analyzed for markers of clonality such as B-cell gene rearrangement or kappa or lambda-light chain analysis [145,148].
Immunophenotyping reveals all lymphomas to be of B-cell origin with intermediate- to high-grade histology. Sixty percent of AIDS patients exhibit high-grade histologies of small noncleaved cells or immunoblasts compared with only 22% of non-AIDS patients [145,148].
Upon confirmation of a diagnosis of PCNS lymphoma, a full neurologic evaluation for staging should be performed including lumbar puncture, MRI of the spinal axis, and slit-lamp examination of the eye (7% to 18% of patients have ocular involvement at presentation). The extent of systemic staging required is controversial. However, if physical examination, chest x-ray, and routine laboratory results are unremarkable, further imaging is rarely useful . Other studies that may be done include bone marrow examination, abdomino-pelvic CT, chest CT, or gallium scanning .
The detection of EBV in all cases of AIDS-associated PCNS lymphomas and the fact that these tumors are often polyclonal suggests a primary role for this virus in the pathogenesis of PCNS lymphomas. This is in contrast to the case in lymphomas in immunocompetent individuals and systemic lymphomas in patients with AIDS where EBV expression is variable and occurs in less than 50% of cases. It is theorized that EBV infection causes clonal expansion of B-cells that goes unchecked by the abnormal immunoregulatory mechanisms of a dysfunctional immune system. Oncogene activation (c-myc is implicated) may result in transformation of one clone and a selective growth advantage. In addition, these effects may be enhanced by the decreased immune surveillance normally found in the CNS .
Prognostic Factors and Survival
The overall survival for patients with AIDS-associated PCNS lymphomas is 2 to 3 months vs 12 to 20 months for patients with non-AIDS-associated lymphomas [145,146,150,151]. The reasons for this difference are unclear, but several theories exist. It is possible that the tumors are biologically distinct and thus differ in their inherent responsiveness to chemoradiotherapy. Also, as Fine observed, trials have tended to use lower doses of radiotherapy in AIDS patients: 56% of patients with AIDS received less than 3,500 cGy vs only 12% of patients without AIDS .
As with systemic AIDS-associated NHL, most patients die of sequelae of advanced HIV disease rather than lymphoma [145,151]. CNS lymphoma is a late manifestation of HIV disease, with most patients having CD4 counts less than 50/mm³ and a well-established AIDS diagnosis manifested by prior or concurrent opportunistic infections. Twenty-five to 100% of autopsied AIDS patients with CNS lymphoma have coexisting CNS infections including HIV encephalitis, toxoplasmosis, cryptococcal meningitis, or cytomegalovirus encephalitis . Patients who present with PCNS lymphoma as their AIDS-defining disease, however, live longer than patients with a prior history of opportunistic infections and represent the few cases, in the literature, of long-term survival .
Standard therapy consists of whole-brain radiation [145,151]. The optimal dose and fraction schedule are undefined, but doses of 2,200 to 6,000 cGy have been reported [145,151,152]. Radiation therapy can substantially improve the quality of life even if survival is not prolonged. Though there are minimal data regarding the use of chemotherapy, they do suggest that responses can occur and that survival is similar to radiotherapy . Intracarotid chemotherapy has also been reported .
Although several studies suggest an increased incidence of Hodgkin's disease in HIV patients , it is still not considered an AIDS-defining malignancy . However, there is evidence to suggest that concurrent HIV infection alters the natural history of Hodgkin's disease  and that Hodgkin's disease in individuals at risk for AIDS is a predictor of HIV positivity.
HIV-positive patients with Hodgkin's disease have been reported to have a high frequency of stage III and IV disease, B symptoms, and atypical patterns of disease spread with extranodal presentations [155,157]. In one series, two thirds of patients had extranodal disease and 48% had bone marrow involvement at diagnosis.
Histologically, patients present with a higher frequency of poor prognosis, mixed cellularity, and lymphocyte-depleted subtypes [157-159]. The survival of patients with concurrent HIV and Hodgkin's disease is much shorter than for those with Hodgkin's disease alone but better than for those with AIDS-associated non-Hodgkin's lymphomas . A large series reviewed by Ames revealed a survival of only 30% at 1 year with a median survival of only 8 months . As with AIDS-related NHL, patient deaths are primarily a consequence of advanced HIV infection and not Hodgkin's disease; in the Ames series, 70% of deaths were due to OI . Standard therapies are employed but again are limited by severe cytopenias, which are often exacerbated by involvement of bone marrow by Hodgkin's disease.
It is currently recommended that all patients diagnosed with Hodgkin's disease be tested for HIV infection. Those found to be HIV positive may require modified treatment as it is cytopenias and OI, rather than uncontrolled Hodgkin's disease that worsens prognosis. A study designed to determine the natural history of Hodgkin's disease in HIV-positive individuals and to determine the efficacy and toxicity of ABVD alone (doxorubicin, bleomycin, vinblastine, dacarbazine(Drug information on dacarbazine)) or with growth-factor support is currently planned by the AIDS Clinical Trials Group .
HIV infection in women is the most rapidly increasing subtype of the disease, and women now account for 40% of all persons infected worldwide and 12% of all infected persons in the United States. The primary mode of transmission of the virus is now through heterosexual contact. An association between cervical cancer and AIDS is anticipated on the basis of common sexual risk factors. Also, immunosuppressed women such as transplant recipients have long been known to be at increased risk for lower genital tract neoplasia . Furthermore, both transplant patients and AIDS patients are at increased risk for human papillomavirus (HPV) infection, long known to be involved in the pathogenesis of cervical cancer .
The association between HPV infection and cervical dysplasia and carcinoma is well established; in fact HPV subtypes are divided by their ability to cause genital tract neoplasia. The risk of genital tract neoplasia by HPV subtype is low for subtypes 6, 11, and 42; intermediate for subtypes 31, 33, 35, and 51 and high for subtypes 16, 18, 45, and 56 . Once contracted, the infection is lifelong and places the patient at continued risk of cervical dysplasia. Immunosuppression induced by HIV infection can result in reactivation of HPV infection. Coinfection with multiple HIV subtypes may be demonstrated in up to 13% of histologic specimens . It is unknown, however, whether other viral infections such as genital herpes modify the incidence or clinical course of cervical neoplasia in HIV-infected women .
The increased incidence of cervical neoplasia in women with HIV infection was only recently recognized as an AIDS-defining illness (January l993) since few women have been included in clinical trials (only 6.7% of participants in AIDS Clinical Trials Group are women ) and no gender-specific trials have been done. The relationship between HIV infection and other gynecologic cancers (ie, ovarian or vulvar) remains to be determined. Vulvar cancer is reported with a rapidly progressive course .
The incidence of HIV infection in women under age 50 presenting with cervical cancer is approximately 19% . Most women with coexistent HIV and cervical cancer are asymptomatic for HIV disease and die of advanced cervical cancer rather than AIDS. Consequently, the Gynecologic Oncology Group is currently conducting a study involving HIV testing among women with cervical cancer to better define the incidence and pathogenesis of the disease .
Gynecologic Manifestations of HIV Disease
Owing to the varying manifestations of gynecologic diseases in HIV-positive women, interpreting Papanicolaou (Pap) smears for malignancy is often difficult. HIV-infected women tend to have recurrent and refractory vaginal candida infections, which tend to occur earlier in the course of HIV infection than oral or systemic fungal infections. HIV-infected women also demonstrate increased rates of sexually transmitted diseases including syphilis, trichomoniasis, gonorrhea, pelvic inflammatory disease, and genital ulcer disease. HIV-infected women show more consistent cytologic evidence of HPV infection than HIV-negative women [162,164]. For instance, in a study by Maiman et al 97% of HIV-positive patients versus only 50% of HIV-negative patients had evidence of HPV infection . Other studies confirm these results . HIV-infected women have as much as an 18-fold increased risk for the development of genital condylomata, which are resistant to primary therapy with failure rates as high as 40% . Furthermore, HIV prevalence rates in women presenting for prenatal care or for treatment of sexually transmitted disease vary from 2% to 13% . In light of such data, women presenting for gynecologic illness should be considered for HIV testing .
HIV-infected women demonstrate a wide range of cytologic abnormalities including inflammatory changes, hyperkeratosis, parakeratosis, trichomoniasis, herpetic changes, HPV-related changes, and varing degrees of cervical neoplasia. HIV-positive women show cytologic abnormalities in 30% to 60% of Pap smears and cervical dysplasia in 15% to 40% of smears [161,162,167]. The prevalence of these abnormalities increases as the immunodeficiency becomes more severe (Figure 2), with patients with cervical intraepithelial neoplasia (CIN) demonstrating lower absolute CD4 counts and CD4:CD8 ratios.
FIGURE 2: Graph of T-cell studies by cervical histology.
CIN = cervical intraepithelial enoplasia.
Reprinted, with permission, from Maiman M et al: Colposcopic evaluation of human immundeficiency virus-seropositive women. Obstet Gynecol 78:84-88, 1991.
A disturbing feature of cervical dysplasia in HIV-infected women is the inaccuracy of cytologic examination (Pap smear) in predicting CIN on biopsy [162,168]. A study by Maiman et al compared results of Pap smears with those of colposcopy plus biopsy and revealed that cytology alone correctly predicted CIN in only one of 13 patients ; conversely, 12 (39%) of 31 patients with normal smears had CIN on biopsy. In a more recent study, Wright et al  reported a 19% false-negative rate for Pap tests in HIV-seropositive patients, yet, even though this rate fell within the range for such tests in the general population (10% to 40%), the authors still recommended repeat Pap screening to reduce the risk of missing CIN on Pap smears. In addition, higher rates of concomitant vaginal infections may further obscure smears. As a result of these studies, it is recommended that HIV-positive women undergo more frequent cytologic screening (every 6 months) and undergo baseline colposcopy regardless of previous cytology [162,169].
Cervical dysplasia in HIV-infected women is more often of higher grade (CIN II to III) with extensive cervical involvement as well as involvement of other sites (vulva, vagina, perianus) and the endocervix [162,164]. In addition, there is no association between CIN grade and age in HIV-positive women as there is in young HIV-negative women with high-grade lesions.
HIV-positive patients with invasive cervical cancer more frequently present with higher stage disease, as demonstrated in a study by Maiman et al in which 71% of HIV-positive patients had clinical stage II or higher disease versus 37% of HIV-negative patients . Disease stage is also more likely to be revised upward based on surgical findings. In the study by Maiman et al, the mean CD4 cell count in HIV-infected patients with invasive cancer was 362/mm³ vs 775/mm³ in HIV-negative patients . One hundred percent of the treated HIV-positive patients had persistent or recurrent disease vs 58% in the HIV-negative patients. The median time to recurrence was 1 month in the HIV-positive patients and 9 months in the HIV-negative patients, with a median time to death of 10 months in HIV-positive patients and 23 months in HIV-negative patients.
Preinvasive Lesions: HIV-infected patients with intraepithelial neoplasia have been treated with therapies including laser therapy, cryotherapy, cone biopsy, hysterectomy, and topical therapies . Yet overall results have been disappointing, with high recurrence rates of 40% to 60% [161,162]. Recurrence has been directly linked to the extent of immune deficiency: rates of more than 50% occur in patients with CD4 counts of less than 500/mm³. Although the actual rate of progression of these lesions to invasive cancer is unknown, it is known that untreated lesions are more likely to progress in HIV-positive women . Patients require repetitive treatments and close surveillance to avoid progression of lesions. The AIDS Clinical Trials Group is currently investigating topical fluorouracil plus ablative therapy as prophylaxis against recurrent CIN. In that study, patients are randomized to receive ablative therapy alone vs ablative therapy plus 2 g of fluorouracil vaginally every 2 weeks for 6 months.
Invasive Cervical Carcinoma
Management of HIV-positive patients with invasive cervical cancer is complicated by the fact that patients present with more advanced stages of disease, present with metastatic disease at uncommon sites, and demonstrate a significantly poorer prognosis. Unlike patients with other AIDS-related malignancies, most patients die from advanced cervial carcinoma rather than from HIV disease . Again, patients with CD4 counts of more than 500/mm³ fare better than patients with lower CD4 counts.
Therapies for invasive cervical carcinoma include surgery, radiation, and chemotherapy . The usual indications for surgery are early disease with curative intent or complications of advanced disease such as bowel obstruction. Pelvic irradiation is often complicated by significant lymphopenia and worsening of overall immune status. Chemotherapy regimens include cisplatin, bleomycin, and vincristine and are complicated by high rates of hematologic toxicity. Antiretroviral agents may be used in combination with these modalities but often exacerbate hematologic toxicities. The effect of zidovudine on the development or recurrence of cervical dysplasia or neoplasia is unknown .
Epidemiologic data confirm an increasing incidence of anal cancer in both men and women that predates the AIDS epidemic . Suspected reasons for this increase include changing sexual habits and, more probably, increased exposure to HPV. Men with a history of homosexual activity are at highest risk. This is supported by a study of single, never-married men aged 20 to 49 in San Francisco between 1973 and 1989 , which showed a sevenfold increase in cases of squamous-cell carcinoma of the anus reported in that group. Other studies have reported incidences of anal cancer among homosexual men up to 40 times that expected in the general population .
As in cervical cancer, HPV is also strongly associated with the development of anal cancer. The same subtypes of HPV are implicated in malignant transformation in anal cancer . However, other factors probably contribute as well to cervical and anal cancer as not all cancers are HPV positive. Many HPV-negative anal and cervical cancers have been shown to contain p53 mutations and it is known that HPV E6 protein transforms epithelial cells by inactivating p53. It is likely that p53 mutation or inactivation represents the final common pathway of malignant transformation for these epithelial cells. In addition, c-myc activation occurs in approximately 30% of HPV-16-associated anal cancers and in premalignant anal lesions .
Like HIV-infected cervical cancer patients, HIV-infected anorectal cancer patients demonstrate a higher incidence of precancerous lesions, a higher incidence of high-grade lesions, lower CD4 counts, and a higher overall incidence of HPV infection, often with multiple subtypes present simultaneously. Patients with lower CD4 counts (and frank AIDS) demonstrate more severe disease than do patients with asymptomatic HIV infection [170-172]. Futhermore, the natural history is one of rapid progression to invasive and morbid lesions.
Screening for anal intraepithelial neoplasia or anal cancer involves routine Pap smears and anoscopy with biopsy of any suspicious lesions. Anal Pap smears have a reported sensitivity of approximately 70% (equivalent to that of cervical smears) and, as in cervical cancer, tend to underestimate the grade and incidence of neoplasia . There are currently no standard recommendations regarding the optimal type and frequency of screening tests for anal cancer . In any case, screening should be vigilant in patients who complain of abnormal discharge, bleeding, pruritus, bowel irregularity, or rectal or pelvic pain and in patients with previous preinvasive lesions or abnormal Pap smears. Other patients who should be considered for screening include HIV-negative men with a history of anal-receptive intercourse, HIV-positive men and women with CD4 counts less than 500/mm³, and HIV-positive or HIV-negative women with a history of high-grade cervical intraepithelial neoplasia .
Patients with high-grade intraepithelial anorectal lesions (grade II or III) should be considered for ablative therapy (Table 15). Patients with grade I lesions can be followed up every 6 months as these lesions may spontaneously regress or progress slowly enough to allow detection before invasive cancer develops [170,171]. Invasive lesions are treated with combined modality therapy with chemotherapy and external beam radiotherapy. Anecdotal experience suggests that HIV-infected patients have a decreased tolerance to full pelvic radiotherapy with increased myelotoxicity and mucositis, thus limiting the size of treatment fields . No studies comparing the efficacy of different therapeutic modalities or the time course and incidence of recurrence have been published.
|For anal intraepithelial neoplasia (AIN-2, AIN-3)|
Ablative therapy after four-quadarant biopsy to rule out invasive cancer; electrocautery, cryoablation, or laser ablation may be performed at the disecretion of the treating physician
|For invasive canal cancer|
Surgical excision ± local radiotherapy may be considered for small, localized cancers with minimal depth of invasion, or
Combined modality therapy (external beam radiotherapy + fluorouracil/mitomycin or fluorouracil/cisplatin chemotherapy
|Adapted, with permission, from Northfelt DW: Cervical and anal neoplasia and HPV infection in persons with HIV infection. Oncology 8:33-40, 1994.|
Many other solid tumors reportedly occur in conjunction with HIV infection. However, information about such tumors is found only in small series or case reports, and not enough epidemiologic evidence exists to demonstrate conclusively the increased incidence of these tumors. Collectively, however, these tumors do have some traits in common: often atypical and aggressive presentations, onset at an earlier than expected age, and absence of commonly defined risk factors. Other reported tumors include head and neck malignancies, skin cancers, lung cancers, gastrointestinal malignancies, testicular germ cell tumors, melanomas, thymomas, gliomas, and leiomyosarcomas .
Adult T-cell leukemia is caused by the HTLV-I discovered by Poiesz in l980. As with all retroviruses, HTLV-I replicates in lymphocytes by integrating its DNA into the host genome. Infection causes an antibody response, which can be detected by screening with commercially available enzyme-linked immunoassay kits. Infection should be confirmed by western blot analysis, radioimmunoprecipitation, or assays to document antibody to core or envelope antigens. These tests, however, do not reliably distinguish between HTLV-I and HTLV-II and can result in false-positives . HTLV-II, although isolated from a cell line from a patient with hairy-cell leukemia and detected with increased frequency in IV drug abusers, is not known to be responsible for malignant disease . Polymerase chain reaction (PCR), although more costly and difficult to perform, can distinguish between HTLV-I and HTLV-II and is becoming the diagnostic test of choice [175,176]. Another member of this retroviral family, HTLV-V, is suspected to be the etiologic agent of mycosis fungoides .
The epidemiologic patterns of HTLV-I and adult T- cell leukemia/lymphoma have distinct geographic distributions. The virus is endemic to Japan, the Caribbean, equatorial Africa, and Central and South America. In the United States, seropositive individuals are concentrated in the Southeast. Within these regions, the incidence of HTLV-I infection also clusters in particular ethnic subpopulations, notably persons of African descent in the Caribbean and Central and South American regions  and persons of black, Hispanic, or Native American descent in the United States .
The virus can be transmitted horizontally through sexual transmission (both heterosexual and homosexual), parenterally through infected blood products or IV drug abuse, and vertically from mother to child through infected breast milk. Breast feeding especially can result in a significant rate of seroconversion: in one Japanese study, 20% of breast-fed infants seroconverted vs only 1% of bottle-fed infants .
The risk of infection after receiving seropositive blood products is strong, approximately 63% of patients become infected after transfusion. In contrast to HIV, however, HTLV-I cannot be transmitted by plasma derivatives. This is supported by the fact that hemophiliacs do not have an increased risk of HTLV-1 infection.
Programs for screening of HTLV-I antibodies are in use in Japan and the United States. In Japan the program has significantly reduced the rate of HTLV-I transmission. However, the impact of screening in the United States, where the prevalence of the disease in the population is low, is at present unknown .
The patterns of transmission are manifest in the population as an increase in seropositivity with age. Breast feeding is responsible for seropositivity in children, and the incidence of seropositivity rises dramatically with the onset of sexual activity in adolescent years. Once contracted, infection is lifelong. Approximately 1% to 5% of infected individuals will develop one form of adult T-cell leukemia-lymphoma.
Although HTLV-I preferentially infects CD4-positive cells, it does not have to bind to the receptor to enter the cell. Infection of circulating cells does require, however, cell-to-cell contact with an infected T-cell, though the exact mechanism of infection is unknown. Once infected, cells are activated, as manifested by their increased expression of the IL-2 receptor and major histocompatibility complex class II antigens. The tax gene of HTLV-I then trans-activates cellular genes for IL-2, the alpha chain of the IL-2 receptor, and the c-fox proto-oncogene, which serve as the preliminary events of T-cell transformation. The infected cells can then be maintained in continuous long-term culture.
Although the exact events of transformation have not been elucidated, a stepwise model of transformation begins with early tax-induced polyclonal T-cell proliferation mediated by deregulated expression of IL-2 and the IL-2 receptor. Additional molecular events are then required for transformation and selective monoclonal growth. Though the transforming events are unknown, the tax gene is capable of impairing DNA repair by suppressing the expression of beta-polymerase. Consequently, karyotypic abnormalities (trisomy of chromosomes 3, 7, or 21) and chromosomal abnormalities (deletions of 6q21, 10p13, and 14q11; translocations; and loss of the X chromosome) are described, although no consistent genetic abnormalities are noted. The virus does not carry oncogenes and is not known to activate cellular oncogenes by insertional mutation .
How the transformed phenotype is maintained is unknown since HTLV-I infection does not involve chronic viremia or expression of viral genes . The role of cytokines in maintenance of the transformed clone is also unknown, although HTLV-I-transformed T-cells do produce a variety of cytokines including IL-1-alpha, IL-1-beta, IL-2, IL-3, IL-5, IL-6, platelet-derived growth factor, interferon-gamma, lymphotoxin, tumor-necrosis factor, and transforming growth factor beta .
Clinical Features and Diagnosis
Adult T-cell leukemia/lymphoma has a long incubation period (20 to 30 years) from the time of infection to development of clinical disease . In that time, patients may exhibit a premalignant form of the disease, which may or may not progress into a true malignant clone. This premalignant phase is characterized by the presence of atypical lymphocytes with lobulated nuclei and lymphadenopathy with HTLV-I seropositivity and monoclonal integration of HTLV-I. Clinically, patients show involvement of the skin, an absence of visceral involvement, normal white blood cell counts, and a minority of circulating leukemic cells. Approximately 50% of patients will progress to overt malignancy with either the chronic or acute forms of the disease. Structural genes in the integrated viral DNA may be abnormal, but no abnormalities in the transforming region have been noted .
Patients whose lymph node biopsy shows abnormalities may exhibit lobular, atypical, Ki-1-positive cells, which may mimic the appearance of Reed-Sternberg cells. Consequently, patients have been misdiagnosed with atypical Hodgkin's disease in the past. But PCR analysis now demonstrates integration of HTLV-I DNA, and serial follow-up of these patients demonstrates transformation to overt T-cell leukemia/lymphoma over time . Therefore a diagnosis of adult T-cell leukemia/lymphoma should be considered in all cases of atypical Hodgkin's disease.
The median age of onset of T-cell leukemia/lymphoma is 56 years, with a male-to-female ratio of 1.5:1. Approximately 25% of seropositive patients will present with lymphoma and no peripheral blood involvement. The acute form of the disease is characterized by lymphadenopathy, hepatospenomegaly, and skin lesions. Less commonly (in 50% of cases), patients may demonstrate hypercalcemia and lytic bone lesions. This is considered a paraneoplastic manifestation and is caused by activation of the gene coding for parathyroid hormone—related protein by the viral tax protein . Patients also demonstrate increased lactate dehydrogenase and hyperbilirubinemia levels.
The lymphadenopathy involves all peripheral areas, the retroperitoneum, and the pulmonary hila but notably spares the mediastinum. The leukemia is characterized by circulating lobulated T-cells, mild or no anemia or thrombocytopenia, and mild involvement of the bone marrow. Skin involvement can vary from patches and papules to nodules and tumor formation.
Patients with the acute disease also demonstrate abnormalities of the immune system, particularly an increased incidence of OI, which cause death in approximately 50% of patients. The chronic form of the disease is characterized by visceral involvement, lymphadenopathy, and elevated leukocyte count but no hypercalcemia or hyperbilirubinemia.
In adult T-cell leukemia/lymphoma immunophenotyping of peripheral blood cells or biopsy material from lymph nodes reveals terminal deoxynucleotide transferase (TdT)-negative T-helper cells (CD4 positive). It also reveals the expression of a mature phenotype (CD2, CD3, and CD5 positive) and the p55 chain of the IL-2 receptor, signs of cellular activation. Furthermore, the T-cells demonstrate rearrangement of the T-cell receptor V-beta gene and monoclonal integration of HTLV-I DNA .
Prognostic Factors, Treatment, and Survival
Several factors predict poor response to therapy and short survival—the presence of leukemia, poor performance status, and elevated serum lactate dehydrogenase levels. So far, however, no specific therapy has been shown to consistently improve survival in these patients with adult T-cell leukemia/lymphoma. Standard chemotherapies produce poor responses and median survivals ranging from months to less than 1 year. Corticosteroids have been shown to benefit patients with myelopathy, but similar results have not been observed in leukemia/lymphoma patients. Nevertheless, promising new therapies include pentostatin and anti-IL-2-receptor (anti-Tac) antibodies .