The Cancer Chronicles Read online

  Some foods appeared to pose small risks. Diets too rich in salt were associated with stomach cancer and red and processed meat with colon cancer, possibly because of nitrosamines, N-nitroso compounds, and other substances. “There was not a very strong association as there is with smoking and lung cancer, where the effect is gigantic,” Riboli said. “We were talking about an increasing risk of one and a half to twofold for some lifestyle habits compared to others.” When a risk is very small to begin with, even doubling it leaves a person with very little chance of getting cancer. But spread across populations of millions, the effects could have a significant impact on public health. Investigating that further, however, would require large epidemiological studies, which can be frustratingly difficult to interpret.

  “The nineteen-eighties was an extremely challenging period,” Riboli recalled. Cancer researchers polarized into two factions. He was reminded of Dante’s Florence, where the warring Guelphs divided into the Neri and the Bianchi, the Blacks and the Whites. “We had two parties, one saying it’s all environmental carcinogens and the other saying cancer can develop without them. I moved from the carcinogenesis party to the lifestyle party.” He became interested not only in the factors that might cause cancer but also in the ones that might prevent it.

  During the decade that followed, he helped with an effort, organized by the World Cancer Research Fund and the American Institute for Cancer Research, to review some four thousand studies on nutrition and cancer and see what patterns emerged. In 1997 the groups issued their report, Food, Nutrition and the Prevention of Cancer: A Global Perspective—inspiration for the 5 A Day program that was all the rage in the years before Nancy’s diagnosis. Based on the best available evidence, fruits and vegetables appeared to have remarkable powers: “Diets containing substantial amounts of a variety of vegetables and fruits may on their own reduce the overall incidence of cancer by over 20 percent.” The number-one recommendation was to eat “predominantly plant-based diets” with five or more servings every day. In her widely read column in The New York Times, Personal Health, Jane Brody gave her summary of the study’s remarkably specific recommendations:

  Foods especially rich in cancer-protective chemicals include the onion family, cabbage-family vegetables (including broccoli, cauliflower, bok choy, kale and brussels sprouts), dried beans and peas, tomatoes, deep yellow-orange vegetables and fruits (like sweet potatoes, cantaloupe and winter squash), citrus fruits, blueberries and dried fruits like prunes and raisins.

  If only it had turned out to be so easy. A decade later, in 2007, came the disappointing follow-up. Riboli was again a key member of the study. As more and better evidence accumulated, the case for fruits and vegetables was unraveling. There was still “limited” to “probable” evidence that some of these foods might slightly lower the risk of certain cancers. But the authors concluded that “in no case now is the evidence of protection judged to be convincing.”

  The problem with the earlier report (and to a lesser extent with its sequel) is that the conclusions were based so heavily on retrospective studies, those where you must rely on people to remember in detail what they had eaten years and even decades earlier—the gestation period for many cancers. “If you ask someone who is seventy years old who has colon cancer what his diet was when he was forty-five or fifty, it’s a tough call,” Riboli said. “For things like smoking or drinking it’s more clear-cut. These are things that are very repetitive and stable.” Things that you remember. “But how often do you eat carrots? How often do you eat pears? Quantify how many pears, how many strawberries, how many eggs—including all the eggs you don’t know about because they are in recipes.” Riboli believed better answers lay in prospective studies, the kind that followed a large population of people as they went about their lives. Then those who developed cancer could be compared with those who did not. “We wouldn’t need to go to someone who is there in a bed with cancer and ask how often he was consuming salads,” Riboli said. “We collect information from people who are living their normal life.”

  While the projects by the World Cancer Research Fund were under way, Riboli had been pushing to form EPIC, the European Prospective Investigation into Cancer and Nutrition. During the 1990s, researchers had begun monitoring the health of 520,000 people in ten countries. Blood samples were drawn periodically and preserved with liquid nitrogen. Heights, weights, and medical history were recorded. Information was collected on diet and physical activity. As, year by year, the database grew, investigators at various universities and government agencies began looking for correlations.

  A few early results had found their way into the 2007 report, helping to tip the balance away from a preoccupation with fruits and vegetables. Since then more surprises have appeared. By the time I talked to Riboli, about 63,000 of the half-million people in the study had cancer. There was now only the slightest evidence that eating a lot of fruits and vegetables had made much difference. They did not obviously reduce the overall risk of getting cancer or even of specific cancers like those of the breast, prostate, kidneys, and pancreas. There were suggestions of a small protective effect, especially among smokers, for cancer of the lungs, mouth, pharynx, larynx, and esophagus. But it was too early to make more than tentative conjectures. Besides smoking, a risk factor for many of these cancers was heavy drinking, and people who smoke and drink excessively have been shown, as might be expected, to be less likely to eat fruits and vegetables. A preliminary study found that these foods possibly played a small part in reducing cases of colon cancer but that too remains in dispute.

  In an editorial for the Journal of the National Cancer Institute, Walter C. Willett, a prominent nutritionist (he was the head of the influential Nurses’ Health Study on diet and lifestyle) and one of Riboli’s longtime colleagues, concluded that researchers had been “overly optimistic” and that the EPIC findings only added to the evidence “that any association of intake of fruits and vegetables with risk of cancer is weak at best.” It had become clear with Doll and Peto that synthetic carcinogens were not the smoking gun, and now it appeared that fruits and vegetables were not a magic bullet.

  Diet was not irrelevant. It was EPIC researchers who estimated that for a 50-year-old who ate a lot of red and processed meat (160 grams, or more than a third of a pound a day) the ten-year risk of getting colorectal cancer was 1.71 percent—0.43 percent higher than for someone who ate less than 20 grams. A third of a pound a day is a lot of hamburgers and hotdogs, and again there are complications to bear in mind. The study made adjustments for smoking, drinking, and other confounding factors. But there may be something else about the behavior of carnivores that skewed the results, and other studies have come to conflicting conclusions. There will always be uncertainties with observational epidemiology and the inevitable question of what is cause and what is effect. Getting closer to the answers would require very large randomized trials in which one population would faithfully eat more of some food and the other would eat less. After twenty or thirty years of draconian enforcement, maybe you could say with some confidence whether there was a difference in the risk of cancer. The data EPIC hopes to collect in the coming decades might be the next best thing.

  Moving beyond purely culinary issues, EPIC has strengthened the case against obesity. One study found that older women who had gained 15 to 20 kilograms, or roughly 40 pounds, since they were twenty had an increased breast cancer risk of 50 percent. As in the old animal experiments, fatness itself, whatever its cause, appeared to be the driving force. Along with lack of exercise, it may account for as much as 25 percent of cancer, with dietary specifics falling to as little as 5 percent. This is the message emerging from decades of nutritional and medical research: Understanding cancer lies less in the foods we eat than in how the body stores and uses energy.

  At the center of this metabolic puzzle is the hormone insulin. As we eat and our level of glucose (blood sugar) rises, it is insulin, secreted by the pancreas, that signals our cells to burn
the fuel directly and to store the excess as glycogen (starch) or body fat. As blood sugar falls, the cells draw on their reserves by converting glycogen back into glucose. When still more energy is needed, fat cells release their long-term supplies. Sometimes, however, something goes wrong. The body produces too little insulin or becomes numbed to its effects. When the latter occurs the pancreas responds by producing more insulin. The cells react by becoming even more resistant and so more insulin is secreted. This pathological spiral—a condition called metabolic syndrome—is involved in chronic conditions like hypertension, cardiovascular disease, diabetes, and obesity. It also plays a role in cancer. The reasons are complex. Insulin and closely related hormones called IGFs (insulin-like growth factors) can stimulate a cancer cell, feeding the expansion of tumors and even encouraging angiogenesis. Insulin is also involved in regulating sex hormones. Moreover, a rise in insulin accelerates the accumulation of body fat, and fat cells synthesize estrogen. Insulin, estrogen, obesity, cancer—all are tied into the same metabolic knot.

  It makes sense that connections like these would have evolved. A woman must be well nourished in order to produce healthy babies. In times of famine, there is no excess energy to store, and the metabolic machinery reacts by lowering the availability of estrogen. It is not a good time to conceive. As more food becomes available, fat accumulates—energy the mother will need during pregnancy and nursing—and more estrogen is released, stimulating ovulation and, after conception, the production of breast milk. Here is the basis of the “mysterious sympathy” Ramazzini wondered about more than three centuries ago. But in a civilization where food becomes abundant, and overly so, the sympathy is upset. The age of menarche decreases, adding to the number of estrogen cycles and raising the risk of breast cancer. Increased nutrition may also unleash the hormones that produce greater body height—another risk factor for cancer. “What this shows,” Riboli said, “is how something which is just a modulation of a normal physiological process—which remains normal and doesn’t cause any disease—has a major impact later in life with cancer. This is not chemical or physical or viral carcinogenesis. It is metabolic carcinogenesis.” The ancient idea of cancer as a disposition of the whole body has returned in a more sophisticated form.

  The amount of fat in storage also affects the immune system in ways that might promote malignancy. In addition to fat cells, fatty tissue contains gobs of macrophages—cells that flock to infected trouble spots to ingest invaders and that can also be diverted to aid in a cancerous attack. And the fat cells themselves secrete other agents that promote inflammation—a healing mechanism that involves the rapid creation of new tissues. There is a thin line between that and tumorous growth. More than a century ago Rudolf Virchow suggested that chronic inflammation, with its power to accelerate cellular proliferation, was a cause of cancer. (That might explain why aspirin and other anti-inflammatory drugs appear in some studies to lower cancer risk.) Obesity has been described as a kind of “low grade chronic inflammatory state” and tumors as “wounds that do not heal.” Chemokines, integrins, proteases … neutrophils, monocytes, eosinophils—there is so much invisible apparatus behind the crude feel of a throbbing joint or a hot, pus-filled wound. Inflammation has also been tied to metabolic syndrome and diabetes. Cancer, obesity, diabetes—the strength of these connections is hinted at in studies of grossly overweight people who undergo gastric bypass surgery in a last-ditch attempt for relief. As their body mass decreases their diabetes recedes, and there is evidence that they get less cancer.

  The deeper you look the more convoluted this all becomes. Cortisol, the stress hormone, and melatonin, which regulates sleep, are also hooked into the metabolic loops involving energy, estrogen flux, and inflammation. Epidemiological studies have suggested that women who work at night may have a higher risk of breast cancer. Considering that and other evidence of the effect of sunlight and sleep cycles on the body, the World Health Organization added “shiftwork that involves circadian disruption” to its list of probable carcinogens—one more avenue that may warrant exploration. All these phenomena are joined at the cellular roots and understanding cancer will require sorting them all out. The overall incidence of cancer has leveled off in recent decades. Are our bodies learning to adjust to the new rhythms? We can never know for sure how cancer rates in the twenty-first century compare with those hundreds of years past. If over the long run there has been an increase, then part of the story could be the modern changes shaking our metabolic core.

  By the time I caught up with Riboli, he and his colleagues were talking less about broccoli, cauliflower, bok choy, kale, and brussels sprouts and more about the body’s energy balance and how the fulcrum has shifted since ancient times. I’d read the debates about the so-called paleo diet—was it richer in fruits and vegetables or in meat and fat? In any case it was low in refined carbohydrates and sugar—energy-packed foods that hit the blood so quickly, causing spikes in insulin and potentially disrupting so many biochemical cascades. Toward the end of our interview Riboli pulled from his bookcase a binder of charts. “At the end of 1800 the usual consumption of sugar in most European countries was two to three kilograms per person per year,” he said. “Now it is between fifty and sixty kilograms.” I pictured a hundred-pound pile of sugar and eating it over the course of twelve months. I was reminded of the journalist Gary Taubes, who argues that carbs and sugar, rather than dietary fat and overeating, drive the modern obesity epidemic and the damage it causes, including cancer, by skewing how the body uses energy.

  Riboli and his colleagues suspect that all energy-rich foods are a problem. Although they are high in calories they can leave us unsated and wanting more. “If I go and buy a burger or a sandwich, most often it contains between five hundred fifty and six hundred kilocalories,” he said. “If I prepare a nice pasta dish, Italian style—with some sauce, pimento, some vegetables—I barely reach five hundred kilocalories. But I have something so voluminous that I feel full. I eat a sandwich and have the impression that I haven’t eaten anything, but I’ve had more kilocalories—more energy.” That empty feeling might spur the desire for a candy bar. Maybe that is reason enough to eat more fruits, vegetables, and fiber. They fill your stomach, reducing your energy intake and therefore your insulin load.

  The other side of the energy equation is physical exercise, and in modern times people are able to lead more sedentary lives. “You and I are having a very pleasant conversation sitting here,” Riboli said. “At another time in another place we might be having this conversation walking in a field. We are moving less and eating more.” Exercise is not, however, a simple matter of burning off pounds. Exertion makes you hungry and you may respond by consuming at least as many calories as you expend. More important may be the effects of exercise on keeping insulin and other hormones under control. Lower your weight and exercise more. “Twenty years ago these were just ideas,” Riboli said. Now EPIC is seeking scientific support. The work is only beginning. An official statement from EPIC promises to explore the complex interactions between genetic, metabolic, hormonal, inflammatory, and dietary factors. More knots to untangle.

  I told Riboli I was feeling even better about having walked to his office all the way across Hyde Park. He laughed and as I put away my notebook he took me on a brisk walk down the hall, out of the building, and beyond the gate of the hospital grounds, until we were standing on the side of Praed Street. He pointed up to a window in the old hospital building, the one to Alexander Fleming’s office. He told me a story that has become part of the legend—how Fleming had accidentally left the window open, allowing the spores of penicillin fungus to contaminate the agar plate. That detail may be apocryphal, but it is an encouraging reminder that a great medical discovery can come suddenly through an act of serendipity.

  As I walked toward the Tube station—I’d exercised enough for the day—I thought of how it can never be so easy with cancer. The infectious diseases we have defeated were each caused by a single agent—an ide
ntifiable enemy that could be killed or vaccinated against. With cancer we would have to seize control of a whole slew of factors, including the mishmash of symptoms arising from imbalances in energy metabolism. And the biggest risks will always lie beyond our grip: old age and entropy. Cancer is not a disease. It is a phenomenon.

  What left me feeling more optimistic is what EPIC might find in the future. In coming years as more people in the study come down with cancer, researchers will be able to analyze their blood in minute detail to see what it was like years or even decades before they got sick. With technologies like nuclear magnetic resonance, they will be able to scrutinize thousands of blood chemicals, looking for signs that might portend the later onset of cancer. This is a very different way of doing medical research. A scientist traditionally begins by posing a hypothesis—based on an observation or a statistical study or a simple hunch. Maybe a high level of a vitamin increases or lowers the risk of some cancer. Then you go looking for evidence. With resources like those at EPIC, connections may emerge that no single mind would have come to suspect. The result could be reliable tests that give early warning for a malignancy the way high cholesterol warns of heart disease. Maybe by then there will be something we can do about it.

  Chapter 11

  Gambling with Radiation

  One surefire carcinogen Riboli and I didn’t talk about is radioactivity. Here the mechanism is straightforward: The unstable nucleus of an element like radium shoots out particles and rays with so much energy that they can tear through molecules, break chemical bonds, and wreak all kinds of cellular hell. Emanations this forceful are called ionizing radiation (atoms stripped of electrons are ions). If the radioactive particles don’t strike a gene head-on, inducing a mutation, they might leave a wake of corrosive free radicals in the cell’s cytoplasm—a condition called oxidative stress than can damage the genome indirectly. Shifting into panic mode, the mangled cell might send signals to neighboring cells, inducing more stress and genomic shock. Most of the exposure we receive from this carcinogen comes from natural sources. The greatest contributor is said to be radon rising from the soil below.