Premalignant and malignant conditions complicating HIV infections

Malignancy complicates chronic human immunodeficiency virus (HIV) infection in a disproportionately high number of cases when compared to the general population. This characteristic was recognised early in the course of the HIV epidemic and has changed in pattern particularly after the introduction of combined antiretroviral therapy (cART) in the mid-1990s. The main cause of malignancy is thought to be decreased immune surveillance as a consequence of chronic HIV infection. There are also oncogenic viruses co-infecting with HIV in some cancers, for example human herpes virus-8 (HHV8) in the case of Kaposi sarcoma (KS) and human papilloma virus (HPV) in the case of anal and cervical carcinomas.

Non-Hodgkin lymphoma (NHL), KS and cervical carcinoma are all acquired immune deficiency syndrome (AIDS)-defining conditions. Hodgkin lymphoma, anal carcinomas and lung cancer all occur at higher frequency in people with HIV infection but are not AIDS-defining.

The spectrum and incidence of malignancies in HIV patients has changed since the introduction of cART. Therapeutic strategies for treating malignant disease in HIV patients are mostly the same as those for patients without HIV infection but treatments are often more toxic and less effective, especially if dose reduction or dose intensity of systemic treatments is necessary. The limits of surgery are less clear but it would be rare that a patient with HIV infection would be offered different or less effective surgery than a patient without the infection.

SOME GENERAL RULES: Most cancers in HIV patients should be treated in the same way, type for type and stage for stage, as if the HIV infection were not present. There are some particular points which may guide the inexperienced clinician. Firstly, consult widely with colleagues who are familiar with treating HIV infection and cancer rather than trying to reinvent the wheel. Antiretrovirals may interact with systemic cancer therapies, considerably increasing the toxicity of particular types of chemotherapy in a predictable fashion. Liaising with a well-versed pharmacist is mandatory before administering chemotherapy as there are a number of predictable interactions which can be avoided or minimised either by a dose reduction of the chemotherapy or omission of an antiviral agent for a period or even swapping to a suitable alternative. The many particular well-documented interactions are beyond the scope of this monograph. Experienced pharmacists are available in all tertiary HIV treatment centres in Australia.

Care is needed to distinguish benign HIV lymphadenopathy from lymphoma and Castleman disease. Firstly, not every abnormal lymph node is malignancy and if focal lymphadenopathy persists and a biopsy is necessary, then an excisional biopsy is preferable to a fine needle aspirate which is usually unhelpful. Once the node has been excised, review by an experienced pathologist is mandatory, again usually at a tertiary referral centre but there exist many private experienced pathologists who are willing to be consulted. The interpretation of radiology imaging can present particular problems as well, especially when there is widespread lymphadenopathy complicating untreated and so advanced HIV. Over-reporting and over-treatment can be avoided by a careful follow-up program as discussed at radiology review.

MULTIDISCIPLINARY CARE: In tertiary referral centres most tumour streams will review a contentious case. Referral to a tertiary institution is not usually necessary but obtaining the advice available will, if necessary, allow discussion and even a remote case to be presented by teleconference. The treating clinicians has the opportunity to interact with a central treating group in an easy and helpful way allowing best standard of care for patients irrespective of their location. This approach has been generally accepted to minimise cancer mortality and is an increasing requirement of quality cancer care.

Kaposi sarcoma

KS was the earliest and most visible manifestation of HIV infection when it was first described. In 1981, the Centers for Disease Control and Prevention (CDC) Morbidity and Mortality Weekly Report published an article about the unusual occurrence of KS and Pneumocystis carinii pneumonitis (PCP, later Pneumocystis jirovecii pneumonitis [PJP]), which is generally accepted as the first description of what came to be known as AIDS.^[1] There are four distinct KS variants: the classical or sporadic form, the endemic African form, the iatrogenic immunosuppression-related form, and the epidemic or HIV-related form.

KS is a malignancy of vascular endothelial cells and the result of co-infection with HHV-8^[2] (alternatively referred to as Kaposi sarcoma-associated herpes virus, KSHV).^[3] HHV-8 has also been associated with primary effusion lymphoma and multicentric Castleman disease. There is some evidence that circulating growth factors, such as basic fibroblast growth factor and numerous cytokines (including interleukin 1, tumour necrosis factor and interleukin-6) also play a role in the pathogenesis of HIV-related KS.

The clinical KS variants are now thought to represent different manifestations of the same pathological process. Studies using increasingly specific and sensitive assays of HHV-8 antibodies have shown that the prevalence of HHV-8 infection matches the rate of KS in different populations. Low rates (0.2-10%) have been demonstrated in the general populations of the USA^[4]  Asia^[5], and many parts of Europe. The background rate of HHV-8 prevalence in patients with HIV infection differs considerably according to the mode of transmission of HIV infection. Rates are high for males who have acquired HIV infection through male-to-male sexual transmission and low for those who have acquired HIV through intravenous drug use, blood or factor concentrate infusion, or heterosexual contact.^[6] Current debate still surrounds the exact nature of transmission of this virus. However, there is general agreement that, in endemic areas transmission occurs through saliva similar to the transmission of Epstein Barr virus, whereas in epidemic KS, sexual transmission is thought to be the major mode of transmission.^[7] The exact sexual activity which leads to acquisition of HHV-8 is not yet determined,^[8] although HHV-8 has been detected in both the saliva^[9] and semen^[10] of people with HHV-8 infection. Transmission has also been documented following organ transplantation.^[11] The risk of transmission via blood products is thought to be low, particularly for highly refined plasma products. The development of new KS after autologous bone marrow transplant is thought to originate from the cellular products necessary as clinical support through the transplant process rather than the immunosuppressive effects of the induction process.^[12]

Clinical presentations

Immune-deficient people

Primary infection with HHV-8 appears to be asymptomatic. Immunosuppression is a cofactor for the development of disease in most patients.^[13] KS generally manifests as pigmented lesions involving the skin or the mucous membranes lining the mouth, nose, eye and anus. Skin lesions occur as plaques, patches or nodules. These are reddish or purple in fair-skinned people and bluish or brown in darker-skinned people. Although the skin lesions are usually painless, they can cause painful swelling, particularly when the lower limbs are affected. Internal organs may also be affected, with the gastrointestinal tract and the lung most commonly involved. Lymph node disease is also common. Nausea, vomiting and bleeding may occur with gastrointestinal tract involvement, and shortness of breath with lung involvement.

The clinical presentation of KS differs according to the KS variant.^[14] Epidemic KS occurs commonly in advanced immunosuppression, but may occur at any CD4 T-lymphocyte (CD4) cell count. Presentation usually features multiple skin lesions often occurring in the lines of skin cleavage (Langer lines) – a phenomenon not seen in other KS variants. These skin lesions may rapidly multiply and progress to involve internal organs. The classical form of KS presents as skin lesions in localised clusters on the lower extremities and usually runs an indolent course over several years, with those affected usually dying from other causes. There is an association, however, with another primary malignancy, and clinicians should consider the possibility of other malignancies in individuals with KS lesions. Endemic African KS and iatrogenic immunosuppression-related KS usually present as skin lesions. These generally remain localised to the skin, but may spread in some patients and affect other organs. A particularly aggressive lymphadenopathic form of endemic KS occurs in young children and is rapidly fatal.

Kaposi sarcoma-associated immune reconstitution inflammatory syndrome

Control of HIV replication usually leads to stabilisation or regression of KS disease in most patients. A paradoxical immune and inflammatory reaction brought about by the improvement in immune status on cART occurs occasionally, resulting in the development of KS-associated immune reconstitution inflammatory syndrome (KS-IRIS). Such KS may require a short course of systemic chemotherapy.^[15]

Diagnosis

In a patient with HIV or risk factors for HIV, the diagnosis of KS should be considered for any skin lesion. However, diagnosis should rarely be made on appearance alone. Even the most experienced clinician will sometimes be fooled by a lesion mimicking the classical appearance of KS and for this reason the diagnosis is best confirmed by a punch biopsy and histopathological examination of the biopsy. The differential diagnosis of KS includes vascular lesions such as purpura, necrotising vasculitis, haemangiomas, angiokeratoma and venous lakes; inflammatory lesions such as pityriasis rosea, granuloma annulare, erythema multiforme, lichen planus and pyogenic granulomas; naevi, malignant melanomas and other cutaneous tumours such as basal cell carcinomas and mycosis fungoides. Other infective causes, such as secondary syphilis and bacillary angiomatosis, should always be considered.

The histological features of KS parallel the clinical appearance of the lesion. In the nodular form a well-circumscribed, non-encapsulated nodule is seen comprising tightly packed spindle cells. These cells form a vascular-slit pattern with extravasated erythrocytes. In the patch and plaque forms, the spindle cells are less tightly packed and form cleft-like arrangements between collagen bundles at all levels of the dermis. Haemosiderin-laden macrophages are seen in all types. In early lesions, spindle cells may be relatively few, but a striking perivascular infiltrate of plasma cells is often present. The histological appearance of KS may be easily confused with other conditions and should be assessed by an experienced pathologist.

Serological assays confirm past or current infection with HHV-8 and have little clinical utility in the diagnosis of KS. The potential relationship of HHV-8 viral load to the development or progression of KS is poorly understood and requires further investigation.^[16]

Management

Therapies used to treat classical KS have proven effective in both endemic and epidemic forms of the disease. The response in epidemic KS is often less durable. In the vast majority of cases of epidemic KS, control of HIV replication itself through effective cART will lead to quiescence of KS disease, and regression of lesions already present.

There are, however, some circumstances when consideration should be given to immediate initiation of specific KS therapy; for example, extensive cutaneous disease and lesions causing discomfort or significant cosmetic problems. Examples include lesions on the face, over a joint, in the oropharynx or visceral lesions, or skin and nodal lesions in the groin or axilla causing lymphoedema.

A variety of local and systemic treatment options exist for KS. There are few definitive KS treatment guidelines so management is recommended in consultation with an expert.

Local treatments include radiotherapy and intralesional injections of chemotherapy. Intralesional chemotherapy is rarely used because of side-effects and the difficulties of administration. Radiotherapy to one area may be effective but needs to be reserved for urgent cases where systemic therapy is either inappropriate or has failed. The long-term side-effects of radiotherapy to lymph node draining areas in particular the groin need to be considered. Lymphoedema may be avoided or less severe if systemic therapy is used.^[17]

Systemic treatment in addition to cART may be necessary. Cytotoxic chemotherapy forms the mainstay of this approach. Better efficacy and less toxicity have made liposomal doxorubicin or daunorubicin the most useful single agents used to treat HIV-related KS. If required, they are generally given at a dose of 20 mg/m² every 2 weeks with or without granulocyte colony stimulating factor (G-CSF) support. The treatment tends to be well tolerated with the main side-effects related to neutropenic sepsis. Hypersensitivity reactions during the infusion occur in about 5% of people, as with most liposomal compounds. The hand-foot syndrome is relatively common in non-HIV patients treated for metastatic breast or ovarian cancer but much less common in those treated for KS who are HIV positive. Cardiomyopathy, neuropathy and alopecia are rare. There are six randomised trials using single agent liposomal anthracyclines as a comparator versus alternative combinations and the single agent had superior overall response rates of about 80% and duration of response at 30 weeks. Quality of life was also better in the liposomal anthracyclines group. In one small study, the combination of cART and liposomal anthracyclines had a superior response rate to cART alone in a highly selected series of patients.^[18]

Paclitaxel, at a low dose of 100-135 mg/m² every 2 to 3 weeks, is associated with response rates of around 71%. In patients who have failed other KS therapy, the response rate to paclitaxel is reduced to about 60%. Unfortunately the studies were not controlled for the use of cART. Side-effects include myelosuppression and neuropathy. Hypersensitivity is usually abrogated by the use of dexamethasone before infusion.^[19] Combination chemotherapy therapy regimes using doxorubicin, bleomycin and vincristine (ABV) are generally only appropriate when the two first-line agents are not available. Interferon alpha or beta have activity against KS, but can be toxic so are essentially outmoded therapies in the Australian setting.^[20]

New and experimental therapy

A variety of promising KS-specific agents such as retinoic acid were in development before the widespread introduction of cART. Retinoic acid inhibits interleukin (IL-6) production, an important cofactor in the pathogenesis of KS. There is evidence that anti-angiogenic agents such as thalidomide, matrix metalloproteinase inhibitors and imatinib have activity against HIV-related KS.^[21]

Antiviral agents such as ganciclovir and foscarnet may prevent the development of, or treat, KS in patients receiving them for cytomegalovirus infection.^[22] Trials of many of these agents have been abandoned because of the dramatic decline in the incidence of KS, leading to difficulties in trial recruitment and therefore evaluation of the efficacy of treatment.

Prognosis

Studies examining both progression and survival in epidemic KS in the pre-cART era found that progression of KS is best determined by markers of KS severity, whereas survival is best determined by markers of immunodeficiency (previous and concomitant opportunistic infections and a low CD4 cell count).^[23] Even in the pre-cART era, fewer than 10% of AIDS patients with KS died as a direct consequence of their KS. In the post-cART era, it is generally understood that cART itself improves the outlook of those with KS-complicating HIV infection. If KS is rapidly progressive or symptomatic, systemic chemotherapy with, or followed immediately by, cART is indicated. The mainstay of systemic therapy is liposomal anthracyclines followed by paclitaxel as salvage. Other therapies are not often necessary but could include interferon, not often used because of its toxicities, or angiogenesis inhibitors such as thalidomide as occasionally referred to in case reports.

KS is not a curable condition. It tends to run a waxing and waning course reflecting the status of the immune system in the natural history of HIV disease progression and treatment. It is not uncommon for patients to require intermittent systemic treatment for years. They tend to re-respond to the same treatments although resistance to treatment can develop necessitating second and subsequent lines of therapy.

There is a paucity of randomised trial evidence evaluating the use of systemic therapies and cART. It is thought that protease inhibitor-containing regimens are not more superior to non-protease inhibitor-containing regimens.^[24]

Prophylaxis

The best prophylaxis for KS is treatment that restores immune function such as cART. Chemotherapy is not used prophylactically.

Non-Hodgkin lymphoma

Epidemiology

The risk of developing NHL is dramatically increased in people with HIV infection, with a 134-fold increased incidence above that seen in the general population.^[25],^[26] Histologically, AIDS-related NHL (AIDS-NHL) is B-cell-derived and characterised by cellular pleomorphism and diffuse growth patterns. Approximately 70% of AIDS-NHL are systemic (as opposed to primary central nervous system lymphoma [PCNSL]). They are predominantly of the Burkitt or Burkitt-like lymphoma type (also termed small non-cleaved-cell lymphoma), and the diffuse large B-cell lymphoma type. Pre cART about 30% of AIDS-NHL were PCNSL, the majority of which were diffuse large B-cell lymphomas.^[27] Two additional forms of lymphoma which are nested within the systemic NHL grouping have been described in association with HIV infection. Primary effusion lymphoma, also termed body-cavity-based lymphoma; and a specific form of diffuse large B-cell lymphoma involving the oral cavity, termed plasmablastic lymphoma. Primary effusion lymphoma is a lymphoma of the serous membranes which produces lymphomatous effusions without exhibiting significant bulk tumour growth.^[28] Other types of NHL, such as low-grade B- and T-cell NHLs may occur sporadically in the context of HIV infection, but these are not considered to be HIV related.1 The incidence of AIDS-NHL has changed since the introduction of potent antiretroviral therapy, with a significant decrease in the incidence of PCNSL and a smaller decrease in diffuse large B-cell lymphoma; however, small non-cleaved-cell NHL has shown no decline.^{[29] [30] [31]}

Since the widespread use of cART, the prognosis of HIV-NHL h as improved with better tolerance of chemotherapy and better response rates. Both the rate of complete remission and the duration of complete remission have increased. The treatment-related deaths due to complications of HIV are also declining.^[32], ^[33] American research published in September 2016 confirms that people with well treated HIV continue to have a much higher incidence of KS and NHL than the general population. This risk persists after virologic suppression and immune recovery and may be increasing despite the benefits of ART. On this basis it is suggested that a better biological understanding of the current KS and NHL burden among those with HIV may lead to different therapeutic approaches. It is acknowledged that the clinical context of KS and NHL has changed with the overall incidence decreasing over calendar time. However new diagnoses occurring whilst on treatment and in routine clinical care were at higher median CD4 counts and often good HIV viral suppression. This could partly be accounted for by the higher proportion of people with HIV now engaged in clinical care. Research into the biologic differences between these malignancies in those with HIV needs to be done to possibly prevent and better treat them.^[34]

Although AIDS-NHL occurs in association with HIV and the Epstein Barr virus, there is no proven causal relationship between HIV replication and NHL. In the context of Epstein Barr virus -associated NHL, HIV causes lymphoid hyperplasia accompanied by a reactivation of Epstein Barr virus infection which gives rise to the production of polyclonal, immortalised B cells. These activated B cells may then accumulate chromosome abnormalities and rearrangements during repeated cycles of replication.^[35] This is thought to occur because of lack of specific anti- Epstein Barr virus CD4 cell function as a direct result of HIV affecting quality and function of the CD4 cells.^[36] Epstein Barr virus genes are incorporated into 100% of PCNSL in people with HIV infection and detection of Epstein Barr virus by PCR may provide an important tool assisting diagnosis. Epstein Barr virus is always associated with the endemic form of Burkitt lymphoma, but is not consistently present in Burkitt-like NHL (small non-cleaved-cell lymphomas). Plasmablastic lymphoma also has a strong association with Epstein Barr virus, distinguishing it from plasmacytomas, with which it is frequently histologically confused.^[37] Primary effusion lymphoma is causally related to co-infection with HHV-8 and HIV.^[38]

Clinical presentation

The clinical presentation of AIDS-NHL varies according to the histological type. In systemic AIDS-NHL, the small non-cleaved cell type presents at a higher CD4 cell count (median 150 cells/μL) compared with the diffuse large B-cell lymphoma, which occurs at more advanced levels of immunosuppression (median CD4 count 50 cells/μL). Extranodal involvement occurs in the vast majority of AIDS-NHL cases (75-90%),^{[39] [40]} with the most frequently involved sites being bone marrow, gastrointestinal tract, meninges and liver. The common finding of B-symptoms (fever, night sweats and weight loss exceeding 10% of body weight in 6 months) may lead the clinician to suspect an opportunistic infection, therefore delaying diagnosis.^[41]

Investigations

Patients with asymmetrical or rapidly progressive lymph node enlargement (with or without constitutional B-symptoms) should preferably undergo excisional lymph node biopsy. Fine-needle aspirates and core biopsies often do not provide adequate samples for histological evaluation and diagnosis. Immunophenotyping should be performed on the biopsy to aid the plan for treatment. This is particularly important given the increasing number of monoclonal antibody adjunctive treatments now available. For instance, CD20 positivity mandates consideration of the use of rituximab (monoclonal anti-CD20) in each case, although the risks of treatment need to be weighed up against the benefits for each patient.^[42]

Staging evaluations to be performed include:

• computed tomography (CT) scans of the brain, abdomen, pelvis and thorax
• positron emission tomography (PET) scan
• magnetic resonance imaging (MRI) scan of the brain if clinical symptoms or signs suggest a problem
• bone marrow biopsy with aspirate and trephine for histopathology, immunophenotyping, and T-cell receptor gene rearrangement studies if indicated to exclude any infective cause for symptoms
• lumbar puncture for cerebrospinal fluid analysis: biochemistry, microscopy and culture, cytopathology and flow cytometry. At least 5 mL should be sent allowing for the adequate exclusion of CNS disease and for special tests including detection of Epstein Barr virus DNA by a polymerase chain reaction assay
• assessment of tumour burden includes serum lactic dehydrogenase and uric acid levels.

Approximately 30% of people will have bone marrow involvement and 10-20% will have leptomeningeal involvement.

Baseline estimations of left ventricular function (ultrasound or nuclear medicine) should be considered in all patients. The risk of cardiac disease must be assessed with extra caution if an anthracycline is to be used in the chemotherapy regimen.

Management

In patients with systemic AIDS-NHL, combination chemotherapy is the standard. However, in patients with substantial tumour bulk, radiotherapy may be used in addition to chemotherapy. A number of chemotherapy regimens have been tried in AIDS-NHL, including standard and modified CHOP (cyclophosphamide, doxorubicin, vincristine and prednisolone), M-BACOD (methotrexate with folinic acid rescue, bleomycin, doxorubicin, cyclophosphamide, vincristine and dexamethasone) and infusional CDE (cyclophosphamide, doxorubicin and etoposide). More recently, as patients present on cART with higher mean CD4 cell counts, high-dose chemotherapy followed by infusion of autologous stem cells can be used in situations where the type and stage of disease is unlikely to be adequately treated by standard dose chemotherapy. While this approach involves more toxic treatment than standard dose chemotherapy, consideration of its use is necessary. Previously it was generally thought to be too toxic for use in the HIV patient with NHL, but the situation is changing with advances in antiretroviral therapy.

Early on in the HIV epidemic, results from chemotherapy were poor, because treatment was complicated by toxicity, due in part, to substantial bone marrow involvement in many patients and advanced immunosuppression.^[43] More recently, infusional regimens such as CDE or EPOCH (etoposide, prednisolone, oncovin [vincristine], cyclophosphamide and hydroxydaunorubicin [doxorubicin/adriamycin]) have been associated with superior results compared with standard dose CHOP chemotherapy. The role for dose-dense treatment (every 10 to 14 days instead of the usual 21 days standard dose therapy in HIV-NHL) awaits full evaluation. The addition of rituximab to CHOP chemotherapy (R-CHOP) has been the subject of three small studies. There has been some benefit shown in selected studies. The routine use of rituximab cannot yet be made given the potential for an increased risk of infections complicating its use.^[44] [45] However the addition of rituximab to the infusional regimens is also showing promising results particularly with G-CSF support.^{[46] [47]}

The AIDS Malignancy Consortium published data examining the use of EPOCH followed by rituximab or EPOCH given with rituximab, which demonstrated that the combined treatments showed better initial responses.^[48] Patient numbers were limited but the safety and use of rituximab with combination chemotherapy is gaining acceptance. The median CD4 cell count at presentation is higher than a decade ago and vastly higher than before cART became available. This better immune constitution probably allows safer rituximab administration – which depletes CD20-positive lymphocytes and predisposes particularly to PJP even in non-HIV hosts.

There is limited information regarding autologous or homologous bone marrow transplantation in patients with HIV and relapsed or refractory NHL. Most data pertain to autologous bone marrow transplantation. In small series, stem cell collection capacity and toxicity of treatment were similar to that of autologous bone marrow transplantation in those without HIV infection; however numbers and experience are limited.^{[49] [50]} Autologous bone marrow transplantation should be considered if HIV is well controlled, and the lymphoma disease is refractory to standard treatments but remains chemosensitive. It should only be considered, performed and managed by a specialist unit experienced in transplantation.

Patients with CD4 cell counts greater than 100 cells/μL have remission rates of up to 94% and median survival exceeding 23 months when treated with combination chemotherapy.^[51] Some factors associated with a poor prognosis include: CD4 cell count below 100 cells/μL; prior AIDS-defining illness; age greater than 35 years; intravenous drug use; and low performance status (Karnofsky performance score less than 70).^[52] Referral of all cases of AIDS-NHL to a medical oncologist or haematologist experienced in the current management of HIV-associated malignancies is strongly recommended. The addition of cART to chemotherapy is associated with an improved outcome. Constantly evolving new therapeutic agents and treatment regimens mandate advice from experienced experts regularly involved in the treatment of such diseases.

Primary central nervous system lymphoma

Clinical presentation

Primary central nervous system lymphoma (PCNSL) is now a rare complication of HIV infection. Patients diagnosed with PCNSL have usually been diagnosed with an AIDS-defining condition and are typically profoundly immunosuppressed (CD4 cell count below 50 cells/μL).^{[53] [54]} Many cases of PCNSL are diagnosed at autopsy.^[55] The most common features at clinical presentation are headache, nausea, lethargy and fatigue, personality changes, and focal neurological deficits. Seizures may occur in up to a third of patients. Fever is rare and the onset of symptoms occurs over several weeks.

Investigations

Characteristic changes shown on a brain CT or MRI scan include the presence of a single lesion with diffuse enhancement after administration of contrast, and surrounding oedema.^[56] The lesions are commonly periventricular in location, but can be found in the cerebrum, cerebellum, basal ganglia and brainstem.^[57] Unfortunately, radiological findings may not always discriminate between lesions caused by PCNSL and lesions caused by other processes such as toxoplasmosis. Failure to respond to toxoplasmosis therapy should raise the possibility of PCNSL. Brain biopsy usually establishes the diagnosis but may also cause significant morbidity and mortality. Lumbar puncture should be considered because malignant cells are demonstrated in 10-20% of patients. The detection of Epstein Barr virus-DNA by PCR in the cerebrospinal fluid occurs in over 75% of patients presenting with PCNSL and may obviate the need for a formal brain biopsy.^[58]

Management

Although PCNSL is extremely sensitive to radiotherapy, treatment outcome has, until recently, been extremely poor with early studies showing a median survival of 72 days.^[59] The incidence of PCNSL has decreased and the overall survival improved since the widespread use of cART.^[60] The lack of CD4 cell-specific EBV immunity is associated with an increased risk of developing PCNSL.^[61] Recently, the addition of chemotherapy (usually adjuvant CHOP) to radiotherapy has been associated with improvements in response rate and survival, and improvements in immune status induced by potent antiretroviral therapy.

High-dose methotrexate is reported to have similar response rates to radiotherapy and standard dose chemotherapy, although good comparative studies have yet to be done in the HIV population.^[62]

Primary effusion lymphoma

Primary effusion lymphoma is uncommon, accounting for less than 3% of AIDS-NHL.^[63] It presents as a lymphomatous effusion arising in serous cavities (pleural, pericardial and peritoneal), without a tumour mass. It is a high-grade B-cell NHL associated with HHV-8 infection. In a cohort of 28 patients, a poor performance status and no treatment with cART were poor prognostic factors. In this group, 50% achieved a complete remission in response to chemotherapy^[64] and 32% were alive at four years.^{[65] [66]} This remains a rare but serious complication of HIV infection.

Hodgkin disease

Epidemiology

Prior to cART, Hodgkin disease (HD) was the most common non AIDS-defining cancer to occur in people with HIV infection. The relative risk of HD in this group is about 13 times the risk in the general population.^{[67] [68] [69]} The risk of HD has increased in the post-cART era. It is the only malignancy thus far to be reported as such.^[70] Studies have shown that HD occurs at increasing frequency with advancing immunosuppression, with the relative risk of developing HD being greatest in the 6 months before, and 3 months after, a diagnosis of an AIDS-defining illness.^[71] [72]

Clinical features and treatment options

The clinical presentation and pathological features of HD in people with HIV infection are often different to those of HD in the general population^{[73] [74] [75]} The disease tends to present with lymphadenopathy outside the usual head and neck area and to have more aggressive histopathological subtypes.

Table 17.1 Differences in clinical presentation of Hodgkin disease in people with HIV and without HIV ^{[76] [77] [78]}

Response to therapy for HD in people with HIV infection was poor in the pre-cART era.^[79] In the post-cART era, the rates of response and duration of response to treatment approached those of HD in the general population. The optimum therapy includes radiotherapy and/or chemotherapy. The chemotherapy regimen is usually ABVD (adriamycin, bleomycin, vinblastine, dacarbazine) or a MOPP (mechlorethamine, oncovin, procarbazine, prednisolone) variant.^[80] Radiotherapy to bulky disease or early stage disease contributes significantly to ideal disease control.

Significant drug toxicity may be seen in people with HIV infection with HD. Dose reductions are often required because of neutropenia and thrombocytopenia and other more specific HIV-related phenomena such as neuropathy or opportunistic infections such as cytomegalovirus disease.^[81] The use of Stanford V regimen (a 12-week course of doxorubicin, vinblastine, mechlorethamine, vincristine, bleomycin, etoposide, and prednisolone, and adjuvant radiotherapy in combination with cART) has been used in HD in patients with HIV infection.^[82] The Stanford V regimen was associated with less toxicity than other treatment strategies, and 55% of patients did not require any dose reduction of chemotherapy or delay in chemotherapy cycles.

Early stage HD is increasingly treated with abbreviated courses of ABVD chemotherapy before lower volume more targeted radiotherapy with good survival results and less long-term toxicity. Longer-term toxicity may include second malignancies which occur less frequently when the disease is managed this way (German Hodgkin Disease Study Group).^[83]

Cervical cancer

Epidemiology

Invasive cervical cancer has been included among AIDS-defining conditions since 1993. There is now evidence of an increased incidence of invasive cervical cancer as a consequence of HIV infection. Studies from Europe have shown significant increases in rates of invasive cervical cancer^{[84] [85]} as have studies from the USA and Africa.^{[86] [87] [88] [89] [90]} Decreased survival early in the epidemic may have masked the risk for invasive cervical cancer in data used from this period, while late diagnosis of HIV may negate the potential gains from the increased cervical screening recommended in this group.^[91] In one of the US studies to show an increased risk of invasive cervical cancer, the diagnosis of cervical cancer preceded the diagnosis of HIV infection and was the trigger for HIV screening in 71% of cases.^[92] The mode of HIV transmission also appears to influence risk, with injecting drug users having three times the rate of invasive cervical cancer compared to those who contracted the infection by heterosexual contact.^[93] It has been suggested that the increased incidence of invasive cervical cancer in this group may reflect a component of immunosuppression induced by injecting drug use,44 but epidemiological studies have not shown an increased risk in HIV-negative injecting drug users.^[94] A more likely explanation is a behavioural one, whereby such activities as sex work may expose people to frequent HPV infections, and more infrequent use of cervical screening programs by women who are injecting drug users with HIV infection than other women with HIV infection.^[95] Research from Nigeria demonstrates that the prevalence of cervical dysplasia in women who contracted HIV heterosexually is high at 10.9% versus 2.6 % in an HIV-negative control group.^[96] This finding supports the screening and active intervention programs in women with HIV, to detect the disease at an early stage when therapy is effective.

Clinical features and management

There are few studies evaluating the clinical presentation and prognosis of invasive cervical cancer in populations with HIV. Information has come predominantly from case series which lacked HIV-negative control groups. Women in these studies tended to present with more advanced cervical cancer, frequently with metastases at unusual sites at the time of diagnosis.^[97] They experienced high recurrence rates after standard treatment (88% at one year) and a high death rate.^[98] Invasive cervical cancer was not associated with advanced immunosuppression; the mean CD4 cell count at diagnosis was 312 cells/μL, and cervical cancer was the AIDS-defining illness in 93% of the patients.^[99] In two studies, women with HIV infection were younger than HIV-negative women at the time of invasive cervical cancer diagnosis and they had a poorer response to therapy.^{[100] [101] [102]}

Surgical excision is the standard of care for early invasive cervical cancer, but, as many women with HIV present with advanced disease, this option is rarely adequate. For more advanced disease, consideration should be given to the same therapy options currently offered to HIV-negative women: that is, radiation combined with chemotherapy.^[103] The prognosis for women with HIV infection appears poor, with invasive cervical cancer being the cause of death in 95% of women with both diagnoses.^[104]

Screening

Although the association between increased rates of cervical dysplasia and cervical intraepithelial neoplasia (CIN) and HIV infection is clear, the effect of HIV on the development and progression of these precursor lesions to invasive cervical cancer is complex and incompletely understood. HPV infection appears to be critical to the development of CIN, while HIV influences both the susceptibility to, and nature of HPV infection. Women with HIV are also likely to have oncogenic HPV subtypes and multiple subtypes infection and to demonstrate persistent HPV infection.^{[105] [106] [107]} Additionally, CIN in women with HIV infection is likely to involve multiple sites, and have high rates of recurrence and rapid progression after standard therapy.^[108] Given these findings, cervical screening programs advocate more frequent Pap smears in women with HIV infection than in HIV-negative women. Current clinical guidelines for women with HIV infection recommend annual Pap smears after two successive normal smears 6 months apart.^[109] However, screening cervical cytology in women with HIV infection has been called into question, as a number of studies show histologically-proven dysplasia despite normal cervical cytology in 18-30% of cases.^{[110] [111]}

A vaccine against HPV 6, 11, 16 and 18 has been licensed. The safety and efficacy of these vaccines in preventing new infection with oncogenic subtypes, or in boosting immunity to current infection with one of the oncogenic subtypes in HPV-HIV co-infection needs to be investigated. The benefits of an assiduous screening program should also be studied.^[112] Given the potential problems with screening, as well as the often extensive and multisite dysplastic involvement and the rapid progression of these lesions in a young population, another approach is semi-annual screening with colposcopy.

The use of adjunctive 5-fluorouracil combined with standard excisional/ablative therapy was shown in a randomised trial to lead to fewer recurrences of CIN, a longer time to recurrence and a lower grade of dysplasia at recurrence.^[113]

Anal cancer

Epidemiology

The first case of anal cancer in a patient with HIV infection was reported in 1989,^[114] and the first case series followed soon after.^[115] Anal cancer in the period before 1970 was a rare malignancy with a reported incidence of 0.7 per 100 000 in males and slightly more in females (0.9 per 100 000).^[116] Dramatic changes have been observed in these rates in the general population, with incidence rates in males having risen to 1.2 per 100 000 and in females to 1.3 per 100 000 between 1995 and 1999.^[117] The reasons for the increase are not entirely clear. However, there is speculation that increased rates of infection with oncogenic HPV subtypes associated with the development of anal cancer (e.g. HPV 16) may have occurred over the same period.^[118] Homosexual men with a history of receptive anal intercourse and men with HIV infection regardless of transmission category are at additional increased risk of anal cancer compared with the general population, with the highest rates seen in homosexual men with HIV infection. The relative risk of anal cancer in men with HIV infection increases as HIV disease progresses, from 13.9 in the 2 to 5 years before AIDS, to 27.4 in the 2 years before AIDS, and 84.1 at or after an AIDS diagnosis.^[119]

Aetiology

It is thought that anal dysplasia represents the precursor lesion for anal cancer. Anal dysplasia and squamous intraepithelial lesions (SIL) or anal intraepithelial neoplasia (AIN) are described variously by different groups. In the AIN nomenclature, lesions are described using a similar staging system to that used for CIN, whereas in the SIL nomenclature, lesions are described in terms of low-grade SIL (equivalent to CIN I/II) and high-grade SIL (equivalent to CIN III/IV). As with cervical dysplasia, many low-grade SIL will regress, whereas few high-grade SIL regress. However, in contrast to cancer of the cervix, the time to development of anal cancer after development of high-grade SIL is not currently known. Studies looking at the rate of anal dysplasia in men and women with HIV have shown remarkably high rates in these groups. Rates of up to 65% have been detected in homosexual men with HIV infection,^[120] while 26% of women with HIV infection and 8% of HIV-negative women had some grade of anal SIL.63 HPV subtype appears to cause different histopathological severity, with HPV 16 almost exclusively associated with high-grade SIL and anal cancer, while HPV 6 and 11 are associated with low-grade SIL.^[121] The list of viral subtypes has been expanded, and subtypes are now classified into two groups: group A, associated with low-grade SIL (HPV subtypes 6, 11, 42, 43, 44); and group B, associated with high-grade SIL (HPV subtypes 6, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58).^[122] As with cervical dysplasia in women, infection with multiple HPV subtypes is more common in men with HIV than HIV-negative men (73% versus 23%), and infection with group B HPV subtypes is seen more commonly in those with advanced HIV disease.^[123]

Screening

Anal cytology has become the most common method for detection of anal dysplasia; however, anal cytology does not always correspond to histological stage, and a low clinical threshold for biopsy to determine the grade of dysplasia is recommended.^[124] Given the high rates of anal cancer and anal dysplasia in people with HIV infection, some advocates have recommended that screening programs, similar to those for cancer of the cervix in women, be undertaken.^[125] Although screening may detect early anal cancers, the impact of such screening programs on the development of anal cancer is not clear. In contrast to the situation with cervical cancer, the treatment of SIL has yet to be shown to decrease either SIL or the later development of anal cancer.^[126] In one study, 23 of 29 patients treated surgically by either excision or cauterisation had persistence of anal SIL or recurrence in less than one year.^[127]

Initial studies demonstrated minimal impact of cART on either anal SIL or HPV infection; however, these studies only examined the first 6 months after initiation of therapy.^{[128] [129]} Regression or decreased development of anal dysplasia and an increased clearance of HPV infection may become evident with longer follow-up.

Management

The management of AIN or SIL once detected is to remove them if possible and then perform regular anal mapping which consists of a formal series of biopsies to exclude invasive disease every 6 to 12 months. Topical fluorouracil may relieve symptoms but has not been demonstrated to reduce the rate of progression from in situ to invasive disease.^[130]

Anal cancer is highly sensitive to the effects of both chemotherapy and radiation. In treated patients, survival at 5 years in the general population is 80%. Combined chemotherapy and radiation with 5-fluorouracil and mitomycin C has become the standard therapy because sphincter function can be maintained in the majority of patients.66 Until recently, it was thought that therapy-related toxicities (haematological, skin and gastrointestinal) were increased in people with HIV infection, and that dose reductions were often required.^[131] However, recent studies have shown that increased risk of toxicity with severe mucositis is related to CD4 cell count and is usually only severe in patients with pre-treatment CD4 cell counts of less than 200 cells/μL.^{[132] [133]} In the post-cART era, the outcome for HIV patients has approached that for HIV-negative patients. In the largest reported series of 32 patients, the 5-year, cancer-free survival was 75% after sphincter-preserving chemoradiotherapy: close to the 80% survival reported in an HIV-negative series. Overall survival was predictably associated with a higher CD4 cell count and lower HIV viral load.^[134]

Lung cancer

Epidemiology

The relationship between HIV and the risk of development of lung cancer remains controversial, as the epidemiological data are both conflicting and inconclusive.^[135]

Studies using data from early in the HIV/AIDS epidemic, before the widespread use of prophylaxis for opportunistic infections and the use of cART, were not able to demonstrate any increased risk of lung cancer because most patients did not survive long enough for lung cancer to develop or be diagnosed. One of the largest studies conducted to date linked HIV/AIDS and cancer registry data from both early and late in the HIV epidemic, and involved 302 834 people. This study has suggested that lung cancer occurs more frequently in the setting of HIV-related immunosuppression compared to the general population, with a relative risk of 4.5,^[136] but this risk declined between 1990 and 2002 with the relative risk changing from 3.3 to 2.6 in this time period.^[137]

It is currently not clear whether this excess rate of lung cancer is related solely to immunosuppression or whether other contributing factors, such as higher rates of tobacco consumption or chronic respiratory infections, are also involved. Few epidemiological studies using linkage data have been able to include smoking rates in the analysis, as this information is not recorded in the databases. By assuming a smoking rate of 100% in their HIV cohort, and comparing this to the expected rate of lung cancer in male smokers in the corresponding age group, Parker et al. were able to demonstrate that the rate of lung cancer in the HIV cohort was greater than could be explained by smoking alone.^[138]

Clinical features and management

Lung cancer in people with HIV infection differs both clinically and pathologically from that which occurs in the general population.^[139] People with HIV infection are generally younger at the time of diagnosis of primary lung cancer, with a median age between 38 and 49 years,^{[140] [141]} compared with the median age of 68 years in the general population.^[142] The median CD4 cell count at diagnosis ranges from 150 to 230 cells/μL.^{[143] [144]} Although all histological subtypes may be seen in patients with HIV infection, adenocarcinoma predominates (occurring in 48% of patients), while only 13% have small cell carcinoma.^[145] The prognosis for people with HIV infection diagnosed with primary lung cancers tends to be poor, with a median survival of 4-20 weeks,^[146] [147] and a one-year survival rate of 0-10%.^{[148] [149]} The reason for this markedly decreased survival in comparison with other populations is not clear. Although opportunistic infections have been described in patients with HIV with lung cancer treated with chemotherapy, three-quarters of patients die from lung cancer itself rather than a secondary disease.^[150]

The evolution of biological-targeted agents continues and laboratory estimation of relevant targets needs to be considered in each case. For example EGFR-expressing tumours may respond to oral erlotinib (a tyrosine kinase inhibitor that acts on EGFR).The list of other such biological agents is developing and discussion is beyond the scope of this monograph.

Other cancers

As people with HIV infection are living longer, other neoplasms such as skin cancers, renal cancer, testicular cancer and lip cancer are occurring at increasing rates.^[151] Increased incidences of skin neoplasms including basal cell carcinoma, squamous cell carcinoma ^[152] and melanomas have been reported in people with HIV infection.^[153] Basal cell carcinomas occur at a younger age in this group, with 54% of cases occurring in patients aged less than 40 years compared with only 5% of this age group in the general population. In people with HIV infection, basal cell carcinomas are more likely to occur on the trunk, with only 29% occurring on the head and neck area, in comparison with 85% in the general population.^[154] Squamous cell carcinomas also occur at a younger age in people with HIV infection, and behave more aggressively. Recurrence occurs in up to 20% of cases following therapy and there is a tendency to early metastasis in patients with HIV infection.^[155] With these unfavourable characteristics, it is important to refer any suspicious lesions for prompt assessment and biopsy. Although there is an increased risk of its occurrence in other forms of immunosuppression,^[156] malignant melanoma remains a rare cancer in people with HIV infection.^[157]

Data from the US Multicenter AIDS Cohort Study indicate that there is a significant increase in the incidence of testicular cancer (standardised incidence rate 3.9) in homosexual men with HIV infection,^[158] but the natural history of this malignancy, as well as cure and survival rates, are similar to those in the general population.^[159] [160] Increased rates of lip cancer in people with HIV infection have been noted in Australia.^[161] However, no clinical series has yet been reported, so the natural history of this malignancy in HIV infection is unknown.

Castleman disease

Multicentric Castleman disease is an unusual and rare type of lymphoproliferative disorder. It is characterised by persistent slowly progressive lymphadenopathy. In the general patient population, the disease usually progresses slowly and has not been regarded as a malignancy given it is polyclonal in origin and may exist for years or decades without requiring treatment. In patients with HIV infection, the situation is usually markedly different: the disease may progress quickly and require treatment similar to lymphoma. Steroids generally control fever but bulky disease requires other chemotherapeutic agents.

The histological picture in lymph nodes is of angiofollicular hyperplasia and plasma cell infiltration forming sheets of cells in the mantle and the interfollicular zone of lymphoid organs. These plasmablastic cells are derived from naive B cells and are polyclonal rather than monoclonal in origin.^[162] Castleman disease shows a strong association with HHV-8, which is demonstrated in 100% of patients with HIV infection who develop this condition.^{[163], [164]} HIV patients tend to present with multicentric Castleman disease as opposed to general patients who tend to present in a unicentric fashion. The two main histological variants (hyaline vascular and plasma cell as described above) are also represented differently in the HIV population.

Increased rates of malignancies associated with HHV-8, such as Kaposi sarcoma and primary effusion lymphoma, have also been noted in patients with HIV infection and multicentric Castleman disease.^[165] The clinical features include constitutional symptoms of fever, fatigue and weight loss, with splenomegaly and peripheral lymphadenopathy.^[166] There may be a cyclical nature to the symptoms, at least initially. Table 2 indicates common signs and symptoms. Diagnosis is made by a lymph node biopsy, which shows changes characteristic of hyalinised germinal centres, vascular hyperplasia and plasma cell infiltrates.^[167]

A number of therapeutic interventions have been tried in patients with HIV infection and multicentric Castleman disease, including corticosteroids, chemotherapy, chemotherapy plus corticosteroids and splenectomy.^[168] [169] There have been two case reports describing durable remissions using oral etoposide.^[170] Most descriptions of therapy of multicentric Castleman disease in the setting of HIV infection are isolated case reports and small case series. More recently, the use of rituximab (humanised monoclonal anti-CD20) has been reported with some degree of success, however, given the persistent B-cell lymphopenia that usually persists after administration, this drug should be carefully considered as an option in each patient individually. The evidence for its use is the subject of only small case series, but is gaining worldwide acceptance.^[171]

Therapy tends to lead to a rapid but short-lived clinical response, with disease progression in most patients. Ongoing corticosteroids may control symptoms well and the dose should be maintained at a minimum to control fever and other systemic symptoms. Mortality is high, with 40% of those with Castleman disease dying within 12 months of diagnosis.^[172] Resolution of multicentric Castleman disease following cART has been described,96 however, other investigators have described no effect of cART on multicentric Castleman disease.^[173] Attempts to use HHV-8-specific antivirals have also been unsuccessful. A particularly aggressive form of multicentric Castleman disease has been described in patients soon after initiation of cART, probably representing a form of immune reconstitution inflammatory syndrome.^[174]

Table 2. Clinical presentation of multicentric Castleman disease

Adapted from: Oksenhendler E, Duarte M, Soulier J, et al. Multicentric Castleman’s disease in HIV infection: a clinical and pathological study of 20 patients. AIDS 1996;10:61-7.