Cancer Chemoprevention Part 1: Retinoids and Carotenoids and Other Classic Antioxidants

In the first of a two part series on cancer chemoprevention, Drs. Singh and Lippman provide a concise, yet complete overview of the principles of chemoprevention. They discuss issues most pertinent to the design and conduct of cancer chemoprevention trials: tolerance of drug toxicities among healthy and high-risk populations, study designs used to test chemopreventive agent toxicity and efficacy, and the mechanism of action of chemopreventive agents (eg, blocking and suppressing agents).

As part of the background discussion of chemoprevention, the authors also carefully consider the use of surrogate end point biomarkers as alternatives to following study participants until invasive cancer develops. Since cancer is a relatively rare disease, studies of agents that may prevent cancer require very large numbers of participants and relatively long follow-up periods, resulting in prohibitive costs. The authors appropriately point to the need for more research on the identification and validation of surrogate end point markers to incorporate into future cancer prevention research.

Beta-Carotene--What Accounts for the Negative Clinical Trials?

Drs. Singh and Lippman focus the first part of their series on retinoids and carotenoids and other classic antioxidants. Although many of the agents discussed in this review have demonstrated efficacy (retinoids) in reducing either invasive cancer or prevalence of surrogate end point biomarkers, other agents, such as beta-carotene, have consistently failed to prevent the major cancers occurring in the United States (eg, lung cancer--the Alpha-Tocopherol Beta-Carotene [ATBC] Trial and Carotene and Retinol Efficacy Trial [CARET]; all cancer sites--the Physician’s Health Study; colon cancer--the Colon Polyp Recurrence Trial; skin cancer--the Basal and Squamous Cell Cancer Prevention Trial; and cervical cancer--the Prevention of Cervix Cancer Trial). Given the exorbitant costs associated with testing beta-carotene in these phase III trials, an evaluation of the problems that may have contributed to these failures is warranted.

Although there has been inconsistency in the epidemiologic literature regarding the role of certain dietary constituents in the prevention of cancer, fruit and vegetable consumption has consistently been associated with a significant decrease in several epithelial cancers.[1] The question that emerged from these findings is, which of the hundreds of nutrient and non-nutrient compounds found in fruits and vegetables are active in preventing cancer?

In the 1980s, research focused on the antioxidant nutrients found in fruits in vegetables, due to the accepted role oxidative damage plays in carcinogenesis. The best described, most readily measurable of these food constituents with antioxidant activity were the carotenoids, compounds with a conjugated polyene structure that allows the molecule to absorb light and to quench, or inactivate, singlet oxygen and free radicals.[2] In fact, most research to date has focused on only one of the more than 600 naturally occurring carotenoids--beta-carotene.

Other Dietary Carotenoids May Exert Effects

As Drs. Singh and Lippman and others have pointed out, overwhelming epidemiologic literature supported an inverse association between dietary consumption of beta-carotene and epithelial cancers, as well as between circulating concentrations of beta-carotene and cancer risk. The problem with these observational studies is that, as we now know, carotenoid levels in blood only reflect overall fruit and vegetable consumption and, thus, may only be a marker of other correlated food constituents that are associated with reduced cancer risk.[3] Unfortunately, phase III trials have been restricted to beta-carotene since it was the only carotenoid readily available in pill form and for which data supported a lack of toxicity to pharmacologic doses in humans.

More recent literature suggests that other carotenoids readily found in the US diet may have more potent antioxidant activity (eg, lycopene). Another factor that may explain the differences between epidemiologic studies and clinical trials is that the protective effect of beta-carotene may be the result of a lifetime consumption pattern of beta-carotene rich foods, which can never truly be tested using a clinical trial design.

Supplement Dose, Nutrient Status of Study Subjects May Be Important

At the time that the beta-carotene trials were planned, there was general agreement that beta-carotene supplementation was safe at high doses. However, results from more recent research indicate that high daily doses of beta-carotene (20 to 30 mg) may actually have adverse rather than protective effects. For example, such high single nutrient doses may inhibit absorption of other protective carotenoids and nutrients and may have a pro-oxidant effect on target tissues rather than the expected antioxidant effect.[2]

Results from the recently published phase III selenium trial indicate that not only dose but also the nutritional adequacy of the study population, relative to the antioxidant nutrient being studied, may influence the response to that nutrient. In a study conducted by Clark and colleagues,[4] selenium supplementation of participants with a history of basal cell or squamous cell carcinoma of the skin and residing in geographic areas of the United States with low soil selenium (marginally sufficient selenium status) had significantly lower incidences of colon, prostate, and lung cancers.

These findings raise the question of whether some antioxidant agents, especially nutrients, may be effective only among those populations with the greatest need for that agent, and may have no benefit in populations with relatively high baseline levels.

Are Multiagent Interventions Needed?

Several carotenoids and other antioxidant constituents in fruits and vegetables may be acting synergistically, rather than in isolation, to provide maximal protection to cells experiencing a high oxidant load. In support of the need for interventions with multiple agents are the results of a large phase III trial conducted in Linxian, China, a region with one of the highest rates of gastric and esophageal cancer in the world. In this trial, supplementation with beta-carotene, vitamin E, and selenium significantly reduced gastric cancer incidence and mortality and nonsignificantly reduced esophageal cancer incidence.

Another method for testing the effects of multiple agents on reducing cancer incidence is to conduct dietary interventions. In the United States, a phase III trial, the National Cancer Institute (NCI) Polyp Prevention Trial, has randomized patients who recently underwent removal of a colon/rectal adenomatous polyp to an intensive dietary counseling group aimed at increasing increase fruit and vegetable intake. The trial's objective is to determine whether this intervention can prevent polyp recurrence. Results of this dietary intervention should be available in the very near future.

Directions for Future Research

Recent research has focused on the antioxidant properties of nonnutritive constituents. As the laboratory work proceeds in this area, future clinical trials will begin to incorporate these new, relatively safe agents. In addition, there is a need to continue studying the more traditional antioxidant nutrients, such as selenium. Studies are needed to clearly define appropriate and safe doses, determine which populations can benefit from supplementation, and evaluate which cancers can be prevented.

References:

1. Steinmetz KA Potter JD: Vegetables, fruits, and cancer: I. Epidemiology. Cancer Causes Control 2:325-357, 1991.

2. Mayne ST. Beta-carotene, carotenoids, and disease prevention in humans. FASEB J. 10:690-701, 1996.

3. Giuliano A, Gapstur, S, Garewal H: Chemoprevention and vitamins: Clinical trials, in Encyclopedia of Cancer, vol. 1. Academic Press, New York, 1997.

4. Clark LC, Combs GF, Turnbull BW, et al: Effects of selenium supplementation for cancer prevention in patients with carcinoma of the skin. JAMA 276:1957-1963, 1996.