Background:  HIV appears to be independently associated with an increased risk of lung cancer. Previous work investigating this association has been limited by sample size, lack of appropriate HIV-uninfected controls, or lack of smoking data. We determined lung cancer incidence in a large cohort of HIV-infected and uninfected subjects, adjusting for smoking and other key covariates.

Methods:  Data from the Veterans Aging Cohort Study Virtual Cohort (VC), a large administrative HIV cohort, was merged with the Veterans Large Health Survey, containing detailed self-reported smoking data, yielding a cohort of 3,707 HIV-infected and 9,890 HIV-uninfected patients. Lung cancer was defined using International Classification of Diseases (ICD-9) codes. We calculated incidence rates of lung cancer using Poisson regression. Then, we determined adjusted incidence rate ratios (IRR) adjusting for age, race, smoking exposure, and chronic obstructive pulmonary disease (COPD).

Results:  The overall incidence of lung cancer was 0.26 per 100 person-years among HIV-infected, and 0.16 cases per 100 person-years among HIV uninfected subjects (IRR = 1.50, 95%CI 1.16 to 1.95), with 28,500 person-years of follow-up amongst HIV-infected and 76,800 person-years of follow-up in HIV-uninfected patients. The IRR of lung cancer associated with HIV infection remained significant after adjustment for smoking, age, race, and COPD. (Table 1)

Table 1: Multivariate model including all covariates, lung cancer diagnosis as outcome.

*White race is reference group.

†Never having smoked is reference group.

Conclusions:  In our cohort of demographically similar HIV-infected and HIV-uninfected subjects, incidence of lung cancer was significantly increased in HIV-infected subjects even after adjusting for smoking exposure.