The Emperor of All Maladies: A Biography of Cancer

In the late 1940s, statisticians in England alerted the Ministry of Health that death from cancer had risen fifteenfold in the previous 20 years, and the ministry convened a conference. The result was that the council commissioned a biostatistician named Austin Bradford Hill to devise a study, but the study was not well funded.

In the US, young doctors like Ernst Wynder, on a medical rotation in New York, noticed the association between cancer and smoking. Told by the surgeon general that a study would be “futile” (244), he earned the support of Evarts Graham, a pulmonary surgeon who facetiously commented that nylon stockings had also increased the incidence of lung cancer. When Wynder presented his results, the audience asked no questions and seemed uninterested in the study.

Richard Doll, a medical researcher, joined with Hill, but Doll and Hill’s study also generated little interest. They started with interviews, asking patients with lung cancer about their tobacco use (which Doll believed was unlikely to have caused cancer) as well as other habits. As the results came in, only one factor emerged as associated with lung cancer: cigarette smoking. Wynder and Graham came to the same conclusion, and both teams published their results. The problem with their study is that they had assessed risk after the fact—after patients had already developed lung cancer.

Hill and Doll modeled themselves after Oxford geneticist Edmund Ford’s study—in which he marked moths and followed them prospectively to study their evolution—and “mailed out letters to 59,600 doctors” (249) asking them about their smoking habits. All the patients in their study who died from lung cancer were smokers. 

Doll and Hill published their prospective study results in 1956, the same year that the percentage of adult cigarette smokers in the US “reached its all-time peak at 45 percent” (250). Advertising targeted nearly every segment of the population, and “Americans were consuming nearly four thousand cigarettes per year or about eleven cigarettes per day” (251).

While the medical association was blasé about the link between smoking and cancer, the tobacco industry was concerned, and they began putting tips on cigarettes as a measure that they referred to as making them “safer” (268). They launched an ad campaign in 1954 called “A Frank Statement” that obscured facts and referred to studies in mice, rather than in people, about the link between smoking and cancer. They also stated that the link could apply to other aspects of modern life, and they launched the Tobacco Industry Research Committee and appointed Clarence Cook Little. Little was once head of the American Society for the Control of Cancer before the Laskerites fired him. He endorsed the idea that genetics were behind cancer and that cigarettes were not to blame.

Epidemiologists were exasperated with the traditional understanding of causation in medicine, developed by the microbiologist Robert Koch in the late 1800s, that required association, isolation, and retransmission to back up the idea that an agent caused a disease. This wasn’t possible for cigarette smoking, as researchers could not make mice smoke cigarettes directly. Hill suggested that the definition of causality had to be expanded, and he said that the connection between smoking and lung cancer met several criteria, including its strength, consistency, specificity, etc. Hill came up with 9 categories in all that could help epidemiologists make a connection between a cause and a disease, and this new list of criteria made epidemiology more pragmatic. Amid these ongoing controversies, Graham, himself a 50-year smoker, received a diagnosis of lung cancer and died soon after.

In 1963, a team—composed of Harvard statistician William Cochran, pulmonary doctor Peter Hamill from the Public Health Service, and pathologist Emmanuel Farber—visited the lab of lung pathologist Oscar Auerbach. The men were part of a committee appointed by the US surgeon general and met with Auerbach because his study of autopsies of smokers versus nonsmokers explained carcinogenesis in the lungs.

This process began in 1961, when the American Cancer Society, the American Heart Association, and the National Tuberculosis Association asked President Kennedy to appoint a national commission on the link between smoking and health. His surgeon general, Luther Terry, chose to go forward, and he appointed a ten-member committee. Over time, an inconvertible picture developed: “The relationship between smoking and lung cancer, the committee found, was one of the strongest in the history of cancer epidemiology” (261). They laid to rest any doubt that this link existed.

Terry released his report in 1964 to a group of journalists. However, the reaction in Washington was tepid. Politicians had learned through Prohibition that they could not regulate an industry, and the government did not consider cigarettes a drug that the FDA could regulate. It fell to the Federal Trade Commission, or FTC, to regulate the industry. In the late 1950s, they investigated their claims that filters made smoking safer. Now, they turned to the question of whether cigarette manufacturers had to acknowledge the link between smoking and lung cancer in their advertisements. Passed through Congress, where the tobacco lobby had influence with southerners, the warning became diluted to “Cigarette smoking may be hazardous to your health” (264), removing any mention of cancer. As the author writes, “Congress had turned out to be ‘the best filter yet’” (265).

This outcome galvanized the anti-tobacco lobby. In 1967, a young lawyer named John Banzhaf wrote to the Federal Communications Commission about the “fairness doctrine,” the Congressional mandate that media had to allow the airing of opposing viewpoints. The FCC’s general counsel, Henry Geller, agreed, and Banzhaf filed suit against a TV station. Although he faced an armada of well-paid lawyers, Banzhaf surprisingly won his case. Several anti-tobacco ads appeared on TV, including one from an actor named William Talman, a former smoker dying of lung cancer. On January 1, 1971, TV broadcasted the last cigarette ad, as the tobacco industry decided to take all ads off the air rather than contend with an equal number of anti-smoking ads.

While many cases launched against the tobacco industry for liability, by the early 1980s, none won until a clever New Jersey lawyer named Marc Edell found a plaintiff named Rose Cipollone. She began smoking as a teenager, lured by the industry’s promise that smoking was a form of female liberation and that it would steady women’s nerves, especially during the turbulent war years. After a diagnosis of lung cancer, Cipollone died in 1984 at age 58.

Edell’s stroke of genius was not, as previously plaintiffs had done, to suggest that consumers did not know about the dangers of smoking. Instead, Edell asked what the tobacco industry itself knew about the connections between smoking and cancer. The courts granted him access to the tobacco company files of Liggett, Philip Morris, and Lorillard, and he found evidence that they not only knew this link, but they also knew that nicotine made people even more addicted to smoking. The jury found Rose Cipollone 80% responsible, while Liggett (the maker of L&M, the brand Cipollone last smoked before she died) was 20% responsible. The jury did not find the other tobacco companies responsible.

However, Cipollone’s case indicated that the tobacco companies were vulnerable, and several lawsuits followed, battering the industry over time. By 1994, the per capita consumption of cigarettes had fallen for 20 years. After Mississippi sued the tobacco industry in 1994, several states followed, hoping to recoup the cost of healthcare. In 1997, the tobacco industry signed the Master Settlement Agreement (MSA), one of the largest liability settlements in history. Among other things, the industry agreed to restrict advertising. However, the tobacco industry has found new ground in places like China and India, where smoking is now a major preventable form of death. The author still sees patients addicted to smoking who can’t quit, even though they are receiving cancer treatments.

Statistical methods to identify risk factors for cancer are descriptive—they are correlations rather than causes. There is power in these types of case and control studies. For example, in the 1970s, these types of studies identified asbestos as causing a rare form of lung cancer called mesothelioma as well as a hormonal medicine called DES as causing cancer in women who received exposure to it in utero.

In the late 1960s, a bacteriologist named Bruce Ames at Berkeley identified substances called mutagens that increased the mutation rate. These substances, which included dyes, X-rays, and benzene compounds, also tended to be carcinogens. His research identified not just a descriptive approach to cancer but a mechanism by which compounds caused cancer.

Baruch Blumberg, a biologist, found that an inflammation caused by hepatitis could also result in cancer. By collecting blood antigens, he found that a virus—which he called the hepatitis B virus, or HBV—caused hepatitis and was also linked to a form of liver cancer. His discovery embarrassed the NCI, which employed a limited interdisciplinary approach. Blumberg also found a vaccine for HBV.

In 1979, two researchers in Australia, Barry Marshall and Robin Warren, began to investigate the causes of gastritis, which scientists once thought to be stress. Trying to isolate the bacteria they thought caused ulcers, they found an organism they named Helicobacter pylori. Wanting to prove it caused gastritis, Marshall injected himself with the bacteria and became violently ill. He published the case of his illness and proved that the bacteria caused gastric inflammation and was also a likely culprit behind stomach cancer, as subsequent studies showed. The development of an antibiotic that killed H. pylori reduced the incidence of gastric cancer in young people, but not in older people, in whom the chronic inflammation had reached a point of no return. The very diversity of cancer causes did not explain the process of carcinogenesis, and scientists still needed an explanation of how cells become cancerous. 

The understanding of carcinogenesis came from an unlikely source—the work of Greek cytologist George Papanicolaou. After arriving in New York, he found a research position at Cornell studying the menstrual cycle of guinea pigs. He found that the cells changed their shapes and sizes depending on the stage of the menstrual cycle. He developed a technique called the Pap smear, as he found that in every case of cervical cancer, there were abnormal cells. Although other pathologists dismissed the technique, its goal was not to identify cancer but pre-cancer.

A large study from the NCI of women in Tennessee identified women who had precancerous changes but were not symptomatic. This study essentially pushed back the discovery of cervical cancer by 20 years, as precancerous changes could lead to invasive cancers. The identification of precancers allowed women to get cures before it was too late to do so.

Epidemiologists think about prevention in two ways. One is through identifying causes of disease, such as smoking cessation for lung cancer. Another is through screening, such as the Pap smear for cervical cancer. In the 1910s, a surgeon in Berlin named Albert Solomon identified structures that looked like crystals of salt in his images of breasts removed in mastectomies. However, the Nazis interrupted his research, and his technique of taking X-rays of breasts, called mammography, was neglected until the mid-1960s. At that time, mammography became part of American clinics, advocated by radiographer Robert Egan in Houston. Screening tests must pass several tests to be useful. First, they must strike a balance between under and overdiagnosis, between false positives and false negatives. They also must result in decreased mortality rates.

In 1963, three researchers—surgeon Louis Venet, statistician Sam Shapiro, and internist Philip Strax (who had lost his wife to breast cancer)—set out to test whether mammography reduced death from breast cancer. They used the over 80,000 women enrolled in HIP, a precursor to an HMO in New York. Some women were randomly assigned to be followed (those who had not received mammography), while others were screened, in part in a mobile X-ray clinic in midtown Manhattan. The reduction in mortality—40%—from screening was remarkable, and the study led to the American Cancer Society starting the Breast Cancer Detection and Demonstration Project (BCDDP) to enroll 250,000 women in screening. However, the study conducted through HIP had erred in pulling out too many women with pre-existing cancers from the control group, thereby perhaps skewing the results. A follow-up study in Canada had erred in assigning too many women with pre-existing cancers to the treatment group.

Finally, a study launched in Malmo, Sweden, in the 1970s showed that mammography had a clear benefit for older women—women over 55—but not for younger women. A later exhaustive analysis confirmed that among women ages 55 to 70, mammography reduced death by 20-30%, but showed little benefit among younger women. The reason that the medical community took so long to accept this test is that among younger women, there are many false positives. However, among older women, there are enough cases of breast cancer that even a rudimentary tool can find an early tumor and therefore be of benefit. In addition, the size of a tumor does not always predict its behavior, as even small tumors can metastasize. Therefore, mammography is in some ways a crude tool. However, finding the mechanism that produced cancer remained elusive.

In February of 2004, the author struggled with several cancer cases in which the patients lingered near death. He met a doctor named Thomas Lynch who possessed the ability to move his patients beyond despair with consideration to the treatment’s process, not just the cure.

In the 1980s, doctors seemed to have reached a time that mixed hope with despair. The autologous bone marrow transplant, or ABMT, came to prominence, a procedure that involved harvesting one’s bone marrow and planting it back in that patient or in another patient. It was a tricky and potentially deadly procedure, but it could be curative. In an autologous treatment, doctors harvested a patient’s own bone marrow and treated it for cancer, using high levels of drugs, before being implanting the marrow back into the patient. Frei launched a protocol called the Solid Tumor Autologous Marrow Program (STAMP) with a young doctor from New York named William Peters. Over time, several patients entered remission, which would not have been possible with standard chemotherapy.

In 1981, researchers published in Lancet the cases of young men in New York who had a rare form of cancer called Kaposi’s sarcoma. These were the first identified cases of AIDS, which Susan Sontag described as having the same sense of guilt and shame that cancer confers. Many of the first doctors to treat AIDS were oncologists, including Paul Volberding in a San Francisco clinic called 5A. The Gay Men’s Health Crisis, a New York organization, borrowed tactics from cancer activist groups to lobby for a cure. Robert Gallo at the NCI, as well as a team from France, found the virus that causes AIDS. Pushing for a cure, the GMHC split into an activist group called ACT-UP. Its leader, Larry Kramer, was convinced that the FDA’s process for approving lifesaving drugs was too slow. At the same time, Peters pushed for a Phase III trial to test the STAMP regimen, a megadose of chemotherapy, for breast cancer.

In the meantime, the treatment of breast cancer through marrow treatment ballooned. One of its foremost practitioners was Werner Bezwoda of South Africa. In 1991, a California schoolteacher and mother of three named Nelene Fox received a diagnosis of advanced breast cancer and asked her insurance company, Health Net, to pay for bone marrow treatment. They turned down her claim, but news of her case spread, and she was able to raise her own money for the treatment. Bezwoda, at the same time, revealed that 90% of the women treated at his Johannesburg clinic had responded to his treatment.

Fox, unfortunately, died at age 40, and her family sued and won a case against Health Net for $89 million. Many insurers chose to support bone marrow treatment, and in some states such as Massachusetts, the law mandated health insurance companies to cover the treatment. In the 1990s, about 40,000 women underwent bone marrow treatment for breast cancer at a cost of $2 billion to $4 billion.

Although Bezwoda presented triumphantly at a cancer meeting in 1999, the other three trials, including that of Peters, found no benefit to bone marrow treatment. Investigators sent to Johannesburg to look at Bezwoda’s records found them fabricated and in disarray—his treatment was a sham, and the era of STAMP was over.

In 1997, Bailar launched another statistical study of the war on cancer: He found that while the death rate had plateaued, mortality had increased slightly. In people over 55, cancer deaths had increased, while they had decreased for younger people. Those cancers with screening tests, such as cervical cancer and colon cancer, had falling death rates, as did cancers among children. However, lung cancer remained the deadliest form of cancer, responsible for one quarter of cancers. Cancer had gone through its adolescence, and now scientists still faced the question of what caused cancerous cells.