Background:  Although incidence of AIDS-defining cancers has declined in the HAART era, less is known about incidence of non-AIDS defining cancers among HIV-infected persons. Incidence trends and risk factors for selected cancers among HIV-infected persons were determined and cancer incidence among persons with HIV was compared to incidence in the general population before and after 1996.

Methods:  Data were combined from the HIV Outpatient Study and the Adult/Adolescent Spectrum of HIV Disease project, representing nearly 180,000 person-years of follow-up, for the pre-HAART (1992-1996) and HAART (1997-2002) eras. Data from the Surveillance, Epidemiology, and End Results (SEER) project were used to calculate rates in the general population. SEER rates were standardized to the age, sex, and race distribution of the HIV study population. Standardized incidence ratios (SIR) were calculated to compare cancer incidence within the HIV-infected population and to the general population for pre-HAART and HAART eras. Multivariable Poisson regression was done to assess factors, such as age, race, gender, HIV risk, nadir CD4 count and ART use, associated with incidence of 3 AIDS-defining and 11 non-AIDS-defining cancers.

Results:  Incidence rates of 9 cancers declined significantly in HIV-infected persons in the HAART era compared to the pre-HAART era. Non-AIDS-defining cancers that declined were colon (SIR 0.52, 95%CI 0.40 to 0.69), Hodgkin’s lymphoma (SIR 0.60, 0.48 to 0.75), lung (SIR 0.32, 0.26 to 0.40), melanoma (SIR 0.55, 0.38 to 0.79), oropharyngeal (SIR 0.37, 0.27 to 0.51), and testicular (SIR 0.34, 0.18 to 0.56) cancers.  AIDS-defining cancers that declined were cervical cancer (SIR 0.34, 0.26 to 0.43), Kaposi’s sarcoma (SIR 0.11, 0.10 to 0.12), and non-Hodgkin’s lymphoma (SIR 0.16, 0.14 to 0.17). Differences in cancer incidence between the HIV and SEER populations, as measured by SIR, have declined significantly in the HAART era for 9 cancers (range 40 to 80%).  In Poisson regression, antiretroviral therapy was independently associated with decreased risk for all cancers, except melanoma and prostate.

Conclusions:  Incidence of AIDS- and non-AIDS-defining cancers among HIV-infected persons declined in the HAART era. Furthermore, differences in cancer incidence in the HIV population relative to the general population have lessened. ART may be associated with reduction in cancer risk.