Risk factors for pediatric human immunodeficiency virus-related malignancy.

CONTEXT: Although cancers occur with increased frequency in children with human immunodeficiency virus (HIV) infection, the specific clinical, immunological, and viral risk factors for malignancy have not been identified. OBJECTIVE: To identify risk factors for malignancy among HIV-infected children. DESIGN, SETTING, AND PATIENTS: A multicenter case-control study of children with HIV at 26 institutions participating in the Pediatric Oncology Group. Forty-three case patients with a new malignancy and 74 control patients without a malignancy were matched based on the duration of their infection. Patients were enrolled between January 1992 and July 1998. MAIN OUTCOME MEASURES: Clinical and laboratory factors assessed as putative risk factors included demographic characteristics, HIV characteristics, prior antiretroviral treatment, and CD4 cell count. Coviral infections with Epstein-Barr virus (EBV), cytomegalovirus, and human herpesvirus 6 were assessed by semiquantitative polymerase chain reaction assays and serological testing. RESULTS: Case malignancy diagnoses included 28 non-Hodgkin lymphoma, 4 B-cell acute lymphoblastic leukemia, 1 Hodgkin disease, 8 leiomyosarcoma, 1 hepatoblastoma, and 1 schwannoma. Epstein-Barr virus viral load of more than 50 viral genome copies per 105 peripheral blood mononuclear cells was strongly associated with cancer risk but only for children with CD4 cell counts of at least 200/ microL (odds ratio [OR], 11.33; 95% confidence interval [CI], 2.09-65.66, P<.001). High EBV viral load was not associated with cancer for children with CD4 cell counts of less than 200/ microL (OR, 1.12; 95% CI, 0.13-9.62; P =.99). Zidovudine antiretroviral therapy did not confer a significant protective effect for either the high (OR, 0.81; 95% CI, 0.22-3.09; P =.77) or the low CD4 cell count groups (OR, 0.27; 95% CI, 0.04-1.46; P =.16). The route of HIV infection was not associated with increased cancer risk. CONCLUSIONS: Route of infection, demographic characteristics, and zidovudine use were not associated with the development of malignancy in HIV-infected children. High viral burden with EBV was associated with the development of malignancy in HIV-infected children although the effect was modified by CD4 cell count. The pathogenesis of HIV-related pediatric malignancies remains unclear and other contributing risk factors can be elucidated only through further study.

Human herpesvirus 6 and cytomegalovirus infections in children with human immunodeficiency virus infection and cancer.

AIM: Measure the prevalence of human herpesvirus 6 (HHV-6) and cytomegalovirus (CMV) infections in children and adolescents with HIV infection and malignancy. METHODS: Semiquantitative polymerase chain reaction and serology were used to test for HHV-6 and CMV infections in 31 cases (HIV-infected children with cancer), 56 HIV controls (HIV-infected children without cancer) and 30 cancer controls (HIV-uninfected children with cancer). RESULTS: In cases, HIV controls and cancer controls, HHV-6 DNA was detected in 29, 39 and 34%, respectively, and CMV DNA was detected in 13, 4 and 7%, respectively. Four cases (13%) and no HIV controls or cancer controls harbored HHV-6 subtype A (P = 0.014). In cases, HIV controls and cancer controls, the prevalence of HHV-6 antibodies was 58, 68 and 93%, respectively, and the prevalence of CMV antibodies was 71, 48 and 70%, respectively. HHV-6 seroprevalence was lower in cases than in cancer controls (P = 0.002), even with adjustments for age and CD4 concentrations; however, HHV-6 infection rates (presence of HHV-6 DNA and/or HHV-6 antibodies) were similar in all groups. Stratification showed that CMV infection was more common in younger patients (ages < 8 years) without severe immune suppression (CD4 concentration >200 cells/microl) than in HIV controls (odds ration, 10.343; 95% confidence interval, 1.65, 121.57). Geometric mean titers of serum anti-CMV antibodies, but not anti-HHV-6 antibodies, were higher in cases (1:71) than in HIV controls (1:33) (P = 0.005). CONCLUSIONS: HHV-6 and CMV infections were common among children with HIV infection and cancer. CMV seropositivity also was associated with cancer in younger HIV-infected patients who did not have severe immune suppression. HHV-6A was detected only in HIV-infected children with cancer.