Attenuating or preventing cancer begins with understanding how the deviation starts. While molecular biologists uncover the errant signals that subvert cells, epidemiologists close in on environmental triggers--which are, for now, easier to target.

Cancer prevention strategies against long-known culprits are pervasive, yet mostly passive, such as admonitions to avoid smoking, sunning, and obesity. Efforts, however, span several levels, from the World Health Organization's Framework Convention on Tobacco Control, to communities in many nations banning smoking in public places, to numerous Web sites touting diets rich in fruits and vegetables. Add to these efforts early detection screens, such as Pap smears, mammograms, and prostate specific antigen (PSA) tests, and the battle against cancer is in motion. But chemoprevention--pharmaceuticals that counter cancer just before or as it starts--promises to put the struggle into high gear.

GENETIC ROOTS In the rare familial cancer syndromes, an overexpressed oncogene or disabled tumor suppressor gene, present from conception, sets the stage for a second event to trigger cancer, anytime, anywhere. In the far more common nonfamilial cancers, environmentally induced mutations send a somatic cell down a detour pathway to cancer. Other genetic insults accrue, gene expression patterns change, and the disease progresses. But exactly who develops cancer from which environmental triggers lies largely in the genes, and teasing out how multiple genes interact with multiple inputs will be difficult.

Linking environmental exposures to cancer dates at least to 1775, when Sir Percival Pott attributed scrotal cancer in London chimney sweeps to their sooty profession. In the future, analysis of individual cancer risks will include genetic profiling as well as consideration of environmental exposures. Smoking and lung cancer provide a familiar example. "A subtle interplay of polymorphisms makes some individuals more susceptible to smoke-induced lung cancer than others," says Timothy J. Triche, pathologist-in-chief at the Keck School of Medicine at the University of Southern California.

Nutrient deficiencies may facilitate cancer in some individuals. "Everything the experts tell us about diet is aimed at the whole population, and we are not all the same," says Alan Diamond, head of the department of human nutrition at the University of Illinois at Chicago.

		PREVENTING CANCER - TWO VIEWS "Can we prevent cancer? Absolutely, according to epidemiological data on migrant populations. The Japanese historically have a high incidence of gastric cancer and a low incidence of colon cancer. In the U.S., it is the opposite. In the last half of the twentieth century, upon migration to Hawaii, first generation Japanese saw a dramatic switch in the relative incidence of gastric versus colon cancer. By the second generation, the incidence reflected the western diet. One to two generations is not long enough to see a change in genetics, so clearly the environment plays a major role." --Leonard Augenlicht, professor of medicine and cell biology; associate director for translational research, Albert Einstein College of Medicine and Cancer Center, New York "How can we prevent cancer? Don't have your cells divide! Cancer is one of those very hard walls we hit as far as longevity goes. There are lots of things to do to lower your risk of getting cancer--eat well, don't smoke--but to actually guarantee not getting cancer, I don't really know if we can ever do that. Cell division is risky business. The risk of cancer is something we have to learn to live with." --Judith Campisi, senior scientist, Lawrence Berkeley National Laboratory; professor, Buck Institute for Age Research, Nevato, Calif.

ENVIRONMENTAL TRIGGERS Discovering an environmental exposure or deficiency common to a type of cancer might reveal how the disease starts. From an intriguing epidemiological observation, research follows a trajectory to cell culture, animal tests, and rarely to clinical trials.

"Epidemiology is fiendishly difficult to do. There are so many confounding factors that it is hard to get convincing results," says Bruce Ames, codirector of the National Foundation for Cancer Research (NFCR) Center for Genomics and Nutrition at the University of California at Berkeley. So his strategy is to reverse the process-- "to look at human cells in culture, deficient in one nutrient, and see what happens," he adds.

Ames has shown that deficiencies of vitamins B₁₂, B₆, or folic acid cause replicating DNA to incorporate uracil instead of thymine. If two such glitches occur close on opposite strands, repair can yield double-stranded nicks, which can break chromosomes, causing cancer.¹

Selenium is another micronutrient that may protect against cancer, according to hundreds of animal, cell culture, and epidemiological studies. When a clinical trial to test selenium's prevention of skin cancer instead showed lowered incidence of epithelial cancers (those that arise in linings), Diamond decided to dig deeper, and discovered that genetic differences may explain how extra selenium helps some people.² "Selenium substitutes for sulfur in the cysteines of two dozen proteins," he explains. Because variants of two selenoproteins--glutathione peroxidase and Sep15--are associated with increased risk of lung and breast cancers, Diamond suggests that these proteins may be the route through which selenium deficiency lifts a yet unrecognized protection against certain cancers. People with these gene variants, he suggests, might require more dietary selenium than others to avoid the disease. The National Cancer Institute (NCI) Selenium and Vitamin E Cancer Prevention Trial (SELECT) is testing this hypothesis. If the link holds up, only those with an inherited higher requirement for selenium would take supplements, avoiding toxicity from overdose in others.

The selenium clinical trials followed robust preclinical evidence, crucial to not perpetuating spurious links. This was not so for beta carotene, a vitamin A precursor that is an antioxidant and keeps cells specialized. People who eat lots of vegetables rich in this nutrient have less cancer. So strong was anecdotal evidence that although animal studies did not find a protective effect, clinical trials proceeded.

The Alpha-tocopherol, Beta-carotene Cancer Prevention trial gave vitamin E, beta carotene, or placebo to 29,133 male smokers in Finland. Surprisingly, those taking beta carotene developed 28% more lung cancers and had 17% more deaths than those on placebo. These results stalled the next effort, the Beta Carotene and Retinol Efficacy Trial.^3,4 But cancer research is ripe with lemons-to-lemonade tales: vitamin E led to a one-third decrease in incidence of prostate cancer. "We took this as a lead for another clinical trial, to see if vitamin E and selenium decrease occurrence of prostate cancer in 32,000 men," says Peter Greenwald, director of the division of cancer prevention at the NCI, referring to the SELECT trial.

CHALLENGES OF CLINICAL TRIALS Cancer research traditionally focuses on the end game: malignancy. "We want to be able to give drugs to prevent occurrence or inhibit growth of a precursor lesion," says I. Bernard Weinstein, professor of medicine at Columbia University. This approach is the essence of chemoprevention. Nutrients and phytochemicals are also tested for prevention (see "Candidates for Cancer Preventatives" list below). But clinical trials to assess cancer prevention take many people and many years.

One shortcut to pinpointing when cancer risk becomes reality is to assess biomarkers instead of symptoms. Another is to monitor precancerous tissues, such as intestinal polyps or the oral lesion leukoplakia. A clinical trial must also compare meaningful groups and treatments. It may be invalid, for example, to extrapolate from a population with precancerous lesions to one without them, or from participants taking a nutrient supplement to those who get it in food. Such confusing circumstances arose in examining effects of eating fiber on development of colorectal cancer.

Consuming fiber to prevent colorectal cancer makes sense; fiber escorts toxin-laden digestive leftovers out of the body. But study results have see-sawed. In 2000, the Phoenix Colon Cancer Prevention Physician's Network⁵ and the Polyp Prevention Trial Study Group⁶ demonstrated lack of a protective effect of fiber on polyp recurrence. Yet when most participants did not have polyps and the fiber came from a more typical diet, the hoped-for effect appeared. The NCI's Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial⁷ and the European Prospective Investigation into Cancer and Nutrition⁸ had similar findings.

CHEMOPREVENTION The NCI has 60 ongoing clinical trials of chemopreventives. Michael Sporn, the Oscar M. Cohn professor of pharmacology and toxicology at Dartmouth Medical School, coined the term "chemoprevention" in 1976.⁹ "We had known that cancer is a multistage process for years, from bronchial dysplasia seen in heavy smokers after a long time, and Pap smears that showed that cervical cancer doesn't happen overnight. If a drug can treat cancer when it is invasive, then it is reasonable that a drug can stop it earlier," he says.

		CANDIDATES FOR CANCER PREVENTATIVES Nutrients Calcium Fiber Folate Selenium Vitamin D Vitamin E Phytochemicals Carotenoids Corcumin Diallyl sulfides Genistein Green tea and grape polyphenols Indole-3-carbinol Isoflavones Isothiocyanates L-perillyl alcohol Saponins Terpines Pharmaceuticals Anti-inflammatories NSAIDS (aspirin, ibuprofen, COX2 inhibitors) Chromatin modifiers Nuclear receptor ligands

But the public does not yet envision a daily pill to prevent cancer in the way that it does medications to lower cholesterol or blood pressure. "We are much earlier on in chemoprevention research. Only a handful of large definitive trials have been completed," explains Greenwald.

To gain consumer acceptance, chemopreventives must be relatively safe. "People say, 'how dare we treat healthy people with drugs that subject them to risk'? But how dare we not treat someone who we know has a high risk, when we have something to diminish that risk. We can identify people at high risk using genetic and proteomic markers, and start to think about intervening earlier," maintains Sporn.

Whether benefit outweighs risk depends on the treatment for the cancer, and whether earlier intervention improves survival. For example, even though the Aspirin/Folate Polyp Prevention Study associated daily low-dose aspirin with lowered risk of new colon polyps,¹⁰ current colonoscopy screening protocols save more lives, says Tom Imperiale, professor of medicine at Indiana University School of Medicine. The reason: Most adenomas do not progress to cancer, and prolonged use of "aspirin, if given to enough people, will result in complications, some of which are life-threatening" he says.

Chemoprevention is farther along for the selective estrogen receptor modulators (SERMs) used for breast cancer. When tamoxifen given to women with estrogen receptor-positive disease prevented cancer in the unaffected breast, the National Surgical Adjuvant Breast and Bowel Project P-1 Study was born.¹¹ "This network of physicians found a 49% decrease in breast cancer in those taking tamoxifen compared to controls. Now some women who are at high risk opt for tamoxifen," says Greenwald.

For SERM efficacy, hormone levels and individual reactions to hormones may be key. The Multiple Outcomes of Raloxifene Evaluations Trial¹² showed that the drug lowers cancer risk in healthy women with no history of the disease compared to women on placebo who have high estradiol levels, but not to women on placebo with low estradiol levels. And Ann Hamilton and Thomas Mack, of the Keck School of Medicine, found that among identical twins with breast cancer, the one with the earliest menarche is much more likely to have been the first to develop cancer, suggesting that sensitivity to the pubertal flood of hormones predisposes to the disease.¹³

Despite efforts at prevention, cancer may be a fact of multicellular life, unlikely but perhaps inevitable. "Whether you develop colon cancer is a probabilistic function. For most patients, one tumor may develop over six to seven decades, when there are a trillion cell divisions in the colon mucosa. Only one cell has developed all of the changes necessary for a tumor to grow and not be eliminated by natural mechanisms," says Leonard Augenlicht, professor of medicine and cell biology and associate director for translational research at the Albert Einstein College of Medicine and Cancer Center in New York.

The challenge, then, is to slow down the dance. Says Augenlicht: "Instead of getting one tumor over seven decades, if it becomes eight to nine decades, people die of other things. We don't need a magic bullet; we need something to lower the probability."

Ricki Lewis (rickilewis@nasw.org) is a freelance science writer and textbook author in Scotia, NY.