The most convincing evidence for a preventive role for any modality is obviously demonstration of incidence reduction produced by that modality. However,
cancer prevention trials with
cancer incidence as an endpoint have logistic problems rendering them essentially impossible to conduct for most
malignancies. Hence a workable strategy often involves analysis of other, indirect lines of evidence to reach conclusions. For
oral cancer, dietary epidemiologic evidence points to a protective role for foods rich in
carotenoids. Other
anti-oxidants, such as
vitamin C, are also implicated. Similarly, laboratory evidence points to a
carcinogenesis inhibitory role for both
retinoids and
carotenoids. Clinical studies have targeted premalignant lesions, i.e.,
oral leukoplakia. For over two decades the efficacy of
retinoids, natural and synthetic, has been known. Nevertheless, it has been difficult to translate this into a recommendation for prevention because of the toxicity of
retinoids. The synthetic
retinoid most often used in these trials is
13-cis-retinoic acid. This compound is toxic even at very low doses (0.1 mg/kg/day), particularly when given over several weeks to months. Hence, although effective, it cannot be advocated for prevention or oral cavity
cancer. Studies with nontoxic
antioxidants, such as
beta-carotene, are much more recent. Early results are promising in that
beta-carotene, alone or in combination with other nutrients, can reverse
oral leukoplakia without toxicity in short-term trials. Studies currently under way will demonstrate whether durable remissions can be obtained using this strategy. It should be emphasized that such long-term trials are problematic to conduct with the toxic
retinoids because the risks of prolonged exposure to them outweighs the chance of
cancer development in the usual
leukoplakia lesion.