Among malignancies, only Kaposi sarcoma (KS) (see chapters "Human Herpesvirus 8 and HIV-Associated Neoplasms" and "Clinical Characteristics of Kaposi Sarcoma") and non-Hodgkin lymphoma (NHL) have a clear epidemiologic association with HIV infection. HIV infection may alter the clinical course of other malignancies (eg, Hodgkin disease and papillomavirus-associated neoplasms [see chapters "Pathogenesis of HIV-Associated Lymphoma" and "Human Herpesvirus 8 and HIV-Associated Neoplasms"]). In fact, cohort studies have also suggested an increased incidence of papillomavirus-associated anogenital neoplasia in persons with HIV-induced immunodeficiency. As a result, the U.S. Centers for Disease Control and Prevention (CDC) added invasive cervical cancer to the surveillance case definition of AIDS on the basis of the extensiveness and aggressive clinical behavior of this disease when it occurs in HIV-infected women.

KS and NHL are commonly associated with other states of cellular immunodeficiency, particularly in persons iatrogenically immunosuppressed to prevent rejection of organ transplants. Other malignancies occurring more frequently in transplant patients than in control populations include nonmelanoma skin cancers (primarily squamous cell carcinomas) of the skin and lip, as well as soft-tissue sarcomas.(1-3) These diseases tend to occur many years after transplantation, raising the possibility that long-term survivors of HIV infection with severe immunodeficiency may ultimately be at risk for additional malignancies.

Numerous epidemiologic analyses of relevant large and small cohorts have attempted to identify the malignancies most likely to be associated with HIV-induced immunodeficiency. The most statistically robust of these analyses is that of the AIDS-Cancer Match Registry Study Group of the National Cancer Institute. In their most recent comprehensive report, these investigators analyzed linked population-based AIDS and cancer registry data from 11 geographically diverse areas in the United States, including 302,834 adults aged 15 to 69 years with HIV/AIDS.(4) The data were analyzed to determine potential links between various invasive cancers (based on World Health Organization-defined categories) and immunodeficiency by correlating timing of cancer and AIDS diagnoses. With this analysis, only lung cancer, penile cancer, soft tissue malignancies, lip cancer, Hodgkin disease, and testicular seminoma met the investigators' criteria for potential association with immunodeficiency, and only the latter three were felt to have a relationship to immunodeficiency that was biologically plausible and/or not confounded by other risk factors or misidentification. Of note, nonmelanoma skin cancers were not captured in the registry data examined in these analyses.

The more common cancers may also occur in patients who coincidentally have HIV infection, as has been described in numerous case reports and small case series since the advent of the HIV epidemic. Rather than being a consequence of immunodeficiency, these cases probably reflect the background incidence of malignancies in the general population, occurring incidentally in persons with HIV infection. For the majority of common cancers, the analysis of the AIDS-Cancer Match Registry Study Group tends to support this conclusion.(4) Nevertheless, the development of any cancer in the setting of HIV-induced immunodeficiency can create significant difficulties in the care of such patients, who frequently develop intercurrent illnesses and tolerate antineoplastic therapies poorly. This chapter briefly reviews the current state of knowledge regarding common malignancies (other than AIDS-defining malignancies) that develop in patients with HIV infection.

Smooth muscle tumors are rare in children, but leiomyomas and leiomyosarcomas are unusually common in HIV-infected children, and are therefore included in the revised CDC classification of pediatric HIV disease as a Category B symptom.(5-10) Investigators from the Childrens' Cancer Group and the National Cancer Institute reported 11 cases of leiomyomas or leiomyosarcomas (17% of all tumors) in their case series of 64 HIV-infected children with cancer.(10) Unusual localizations, such as spleen, pleural space, adrenal glands, and lungs have been described, although these tumors present most commonly in the gastrointestinal tract. Several cases of intracranial or dural leiomyosarcomas have also been described.

One group of investigators who studied these tumors using molecular techniques found Epstein-Barr virus (EBV) genomes in all specimens tested.(11) Similar tumors from patients without HIV infection were studied and found not to contain the EBV genome. The authors speculated that, in the setting of HIV-induced immunodeficiency, EBV may contribute to the pathogenesis of smooth-muscle tumors in children and otherwise immunocompromised (ie, posttransplant) patients in whose tumors EBV has also been identified.

No specific therapy for HIV-related smooth-muscle tumors has been developed. The course of disease is highly variable. Patients with these malignancies are therefore treated in a manner similar to patients without HIV infection, with consideration given to the physiologic impairment caused by their HIV disease. Indolent tumors (more likely leiomyomas) that probably do not require any medical intervention have been seen in some children, and very aggressive, disseminated tumors have been seen in others. Because smooth muscle tumors are in general not very sensitive to chemotherapy or radiotherapy, local excision, if feasible, is the first line of therapy.(12) Intensive and prolonged chemotherapy as used in non-HIV-infected patients is rarely tolerated by HIV-infected children.

The lungs are the most common site of disease in persons with HIV infection.(13) In the general (non-HIV-infected) population, lung cancer is the most common fatal cancer in men and the second most common visceral cancer in men and women (after prostate and breast cancer, respectively) in the United States. It is therefore not surprising that lung cancer has been diagnosed in a number of persons with HIV infection.

Most epidemiologic studies have not shown the incidence of bronchogenic carcinoma to be increased in persons with HIV infection. However, both an epidemiologic study conducted in Australia and the AIDS-Cancer Match Registry Study Group identified an increased incidence of lung cancer in HIV-infected individuals and, in the latter study, a possible relationship to immunodeficiency.(4,14) Additionally, numerous patients with bronchogenic carcinoma and HIV infection have been described in reports of case series, in individual case reports, and in abstracts presented at various clinical and scientific meetings.(15-32) These reports indicate that HIV-infected patients with bronchogenic carcinoma differ clinically from immunocompetent patients with lung cancer in several respects.

HIV-infected patients with lung cancer are relatively young; the majority of patients reported in the literature were younger than 50 years old. Adenocarcinoma was the histologic diagnosis in the majority of these patients. In contrast, non-HIV-infected persons with lung cancer are usually much older,(33) and adenocarcinoma is the histologic type in only 30% of cases.(34) A case-control study substantiated that HIV-infected lung cancer patients were younger than controls.(22) Anecdotal observations by authors of the case reports cited previously suggest that HIV-infected patients tend to experience more rapidly progressive cancer than do comparable non-HIV-infected patients. Experience with HIV-infected lung cancer patients at San Francisco General Hospital supports this conclusion.

Despite these apparent differences, the actual clinical presentation of lung cancer in patients infected with HIV does not differ significantly from that in non-HIV-infected patients. Symptoms may include cough, dyspnea, chest pain, and hemoptysis. Typical radiographic findings include mediastinal adenopathy, hilar masses, parenchymal pulmonary masses, and pleural effusions.(15) Diagnosis may be complicated by the similarity of these features to those associated with HIV-related opportunistic infections, KS, or NHL.

An HIV-infected patient will require evaluation for possible lung cancer only after the more common HIV-related infections and malignancies have been ruled out. Sputum or pleural fluid cytology--histologic and cytologic evaluation of samples obtained via bronchoscopy or by percutaneous fine-needle aspiration--constitute the most common means of establishing the diagnosis.

Published information about therapy for lung cancer in the setting of HIV infection is limited to anecdotal experience described in the case series reports referenced above. Unfortunately, anecdotal experience suggests that standard chemotherapy produces little or no benefit and is poorly tolerated in patients already debilitated by HIV-induced immunodeficiency.(16,22,31,32) Palliative radiotherapy may be useful in some situations.

Germ cell cancer is the most common malignancy in men less than 40 years of age, the age range in which HIV infection most commonly occurs in men in the United States. It is therefore not surprising that several case reports and reports of case series have described germ cell cancers in a number of men with or at risk for HIV infection.(18,35-48) Germ cell malignancies have no recognized association with immunodeficiency, although one recent epidemiologic cohort study suggested an increased incidence of seminoma in HIV-infected men as compared with a control population.(49)

There have been no prospective studies of treatment for germ cell malignancy in HIV-infected persons. Our early report described a retrospective review of 15 patients with documented HIV infection who were treated for germ cell cancers in San Francisco.(45) The clinical presentations of these patients were comparable to those of 279 patients with testicular cancer but not at risk for HIV infection who were treated at the same hospitals. The two patient populations were remarkably similar with respect to disease stage and histologic subtype. The majority of the HIV-infected patients with germ cell cancers received standard therapy for their tumor stage and histology, and they tolerated treatment well. Five of the seven patients who received chemotherapy achieved complete remission, and none of the 15 had died of testicular cancer. In this study, there were six deaths, all caused by progressive HIV disease at a median of 20 months following the germ cell tumor diagnosis. The median follow-up duration of surviving patients was 42 months, with six patients remaining tumor free, two remaining free of progressive disease with stable residual masses, and one patient with recurrent disease. Median survival was 40 months for patient with CD4 T-lymphocyte counts >200 cells/µL, and 26 months for patients whose counts were <200 cells/µL.

Subsequent reports from Germany, Italy, France, and Great Britain have described larger series of patients with HIV infection and germ cell malignancies.(44,46-48) The proportion of patients with seminoma was somewhat greater in all of these series, consistent with epidemiologic data suggesting a greater incidence of this histologic type in persons with HIV infection.(35) However, in all other important respects (including treatment outcome and mortality) the findings have been remarkably similar. Taken together, these observations suggest that germ cell tumors in persons with HIV infection should be treated with standard therapy according to disease stage and histology.(44-48) Patients generally tolerate combination chemotherapy well in this setting, although patients with more advanced HIV disease may experience progression of their underlying illness.(40)

There is a wealth of anecdotal information related to incidence and clinical behavior of nonmelanoma skin cancer in persons with HIV infection. To cite a recent example, Nguyen et al (50) reported on a consecutive sample of 10 patients infected with HIV who had "aggressive" squamous skin cancer based on the following criteria: diameter >1.5 cm, rapid growth rate, local recurrence, and/or evidence of metastasis. Five of these patients died of metastatic squamous skin cancer within 7 years of their initial diagnosis despite treatment. Stage of HIV infection and degree of immunodeficiency were not associated with increased morbidity and mortality. Patients initially undergoing combination surgery and radiation therapy or radical neck dissection had the best outcomes. The authors concluded that patients infected with HIV can develop rapidly growing squamous skin cancers at a young age, with a high risk of local recurrence and metastasis. They recommended that high-risk squamous skin cancers should be managed aggressively and not palliatively in patients infected with HIV.

However, this disease has not been subjected to systematic study. Evidence-based approaches to clinical care of patients with nonmelanoma skin cancers in the setting of HIV infection are therefore lacking. Risk factors (fair skin, excessive sun exposure, and related skin damage) are probably similar to those in non-HIV-infected persons, and standard management strategies for prevention and treatment, as for non-HIV-infected persons, are appropriate for patients with HIV infection.

As with the other cancers described in this chapter, there is considerable anecdotal information regarding plasma cell neoplasms in persons with HIV infection.(51-53) Plasma cell neoplasms are also similar, as they also have not been subjected to systematic study in the setting of HIV disease, and evidence-based approaches to clinical care of these disorders are therefore lacking. Based on anecdotal experience, standard treatment strategies applied to non-HIV-infected persons are generally recommended as appropriate for patients with HIV infection.