Effect of HIV on survival in patients with non-small-cell lung cancer in the era of highly active antiretroviral therapy: a population-based study

Abstract

Background: HIV-infected patients with lung cancer have been reported to have poorer survival than uninfected patients. Whether this outcome holds true in the era of highly active antiretroviral therapy (HAART) is unclear. We examined the effect of HIV infection on clinical outcome in patients with lung cancer who are also receiving HAART.

Methods: Patients diagnosed with non-small-cell lung cancer (NSCLC) from Jan 1, 2000, to Dec 31, 2005, with or without HIV infection were identified by querying the Surveillance, Epidemiology, and End Results registry and the Medicare lung cancer database. Survival analysis by stage and treatment delivered comparing the HIV-infected patients with uninfected controls was done with Kaplan-Meier and Cox models with propensity score adjustments.

Findings: 71,976 patients with NSCLC were identified as uninfected controls and 322 patients with NSCLC were identified in the HIV group; median age was 75 years for both groups. Median overall survival for all stages was 7·0 months (95% CI 7·0-7·0) for uninfected controls versus 8·0 months (6·0-10·0) for the HIV group (p=0·16); for those with stage I/II disease it was 37·0 months (36·0-39·0) versus 43·0 months (26·0-58·0; p=0·37); for those with stage IIIA/IIIB disease it was 7·0 months (7·0-7·0) versus 3·0 months (2·0-8·0; p=0·051); and for those with stage IV disease it was 3·0 months for both groups (95% CI 3·0-3·0 for controls; 2·0-5·0 for HIV group; p=0·77). After propensity score adjustment, the survival difference in stage IIIA/IIIB was no longer seen (hazard ratio 0·88; 95% CI 0·71-1·09). The median survival for HIV infected patients with stage I or II NSCLC who underwent surgical resection was 58·0 months (95% CI 57·0-60·0) for uninfected controls versus 50·0 months (42·0 to unestimable) for the HIV group (p=0·88).

Interpretation: We noted no significant difference in clinical outcome between patients with HIV and uninfected controls with NSCLC. Survival after curative surgical resection in early-stage patients was similar in HIV-infected individuals and uninfected controls. These data suggest that HIV status should not affect therapeutic decision making in NSCLC.

Funding: US National Cancer Institute (award number UC2CA148310).