Description of the Evidence
Interventions Associated With Increased Risk of Lung Cancer
Beta-carotene supplementation in smokers
Results of the National Cancer Institute (NCI) Alpha-Tocopherol, Beta-Carotene Cancer Prevention (ATBC) trial were first published in 1994. This trial included 29,133 Finnish male chronic smokers aged 50 to 69 years in a 2 x 2 factorial design of alpha-tocopherol (50 mg/day) and beta-carotene (20 mg/day). Subjects were randomly assigned to one of the following four groups for 5 to 8 years: beta-carotene alone, alpha-tocopherol alone, beta-carotene plus alpha-tocopherol, or placebo. Subjects receiving beta-carotene (alone or with alpha-tocopherol) had a higher incidence of lung cancer (RR = 1.18; 95% CI, 1.03–1.36) and higher total mortality (RR = 1.08; 95% CI, 1.01–1.16). This effect appeared to be associated with heavier smoking (one or more packs/day) and alcohol intake (at least one drink/day). Supplementation with alpha-tocopherol produced no overall effect on lung cancer (RR = 0.99; 95% CI, 0.87–1.13).
In 1996, the results of the U.S. Beta-Carotene and Retinol Efficacy Trial (CARET) were published. This multicenter trial involved 18,314 smokers, former smokers, and asbestos-exposed workers who were randomly assigned to beta-carotene (at a higher dose than the ATBC trial, 30 mg/day) plus retinyl palmitate (25,000 IU/day) or placebo. The primary endpoint was lung cancer incidence. The trial was terminated early by the Data Monitoring Committee and NCI because its results confirmed the ATBC finding of a harmful effect of beta-carotene over that of placebo, which increased lung cancer incidence (RR = 1.28; 95% CI, 1.04–1.57) and total mortality (RR = 1.17; 95% CI, 1.03–1.33). In a follow-up study of CARET participants after the intervention discontinued, these effects attenuated for a period of time. After 6 years of postintervention follow-up, the postintervention RR for lung cancer incidence was 1.12 (95% CI, 0.97–1.31) and for total mortality was 1.08 (95% CI, 0.99–1.71). During the postintervention phase a larger RR among women, rather than men, emerged for both outcomes in subgroup analyses; the reason for this observation, if reliable, is not known.