An international evaluation of the cancer preventive potential of carotenoids.

The IARC convened a Working Group of experts in December 1997 to evaluate the cancer-preventive potential of carotenoids and to compile the second volume of the IARC Handbooks of Cancer Prevention. In observational epidemiological studies, beta-carotene is associated with reduced risks for cancer at many but not all sites. It is unclear, however, to what extent beta-carotene itself is responsible for the decreased risks observed. Three large, randomized, placebo-controlled clinical trials indicate, however, that, in substantial doses, supplementation with beta-carotene not only does not prevent lung cancer but may actually increase the risk among individuals initially at high risk of lung cancer. These trials do not provide clear evidence concerning cancers at other specific sites. Thus, the Working Group considered that there is evidence suggesting a lack of cancer-preventive activity for beta-carotene when it is used as a supplement at high doses. At usual dietary levels of beta-carotene, the evidence for cancer-preventive activity was considered inadequate. However, there is sufficient evidence that beta-carotene has cancer-preventive activity in experimental animals, based on models of skin carcinogenesis in mice and buccal pouch carcinogenesis in hamsters. The observational epidemiological data on alpha-carotene, lycopene, and lutein are much less extensive than those for beta-carotene. For canthaxanthin, there are no published data regarding associations with cancer risk. These carotenoids have not been studied in human trials for cancer prevention. In animal models, there is sufficient evidence for canthaxanthin and limited evidence for alpha-carotene, lycopene, and lutein of cancer-preventive activity. Pending further research, supplemental beta-carotene, canthaxanthin, alpha-carotene, lutein, and lycopene should not be recommended for cancer prevention in the general population.