Medical Examiner

John Bailar’s Righteous Attack on the “War on Cancer”

By putting forth this unpopular opinion, Bailar reminded us of the value of evidence-based decisions.

John Bailar challenged the war on cancer when doing so was almost sacrilege.

Melissa Bailar

When then-President Richard Nixon launched the “war on cancer” in 1971, there was no more admirable cause to support. The dreaded disease was the second leading cause of death that year for Americans, after heart disease, and has maintained that spot for decades. Yet John C. Bailar III, a physician and epidemiologist who died in September at age 83, persistently challenged the war—at a time when doing so was almost sacrilege.

In recent years, others have picked up Bailar’s points, such as the notion that early cancer detection may not save lives. And we’ve moved on to a new metaphor for cancer control: the “moonshot,” championed by Vice President Joe Biden. But, in an era in which even a cancer moonshot is likely to be politicized, it is worth remembering a critic like Bailar, who thoughtfully opposed quick fixes for a complicated disease. Bailar would have been the first to say that it is impossible to separate science and rhetoric, but as a scientist and an advocate, he always tried to focus on what he believed the data showed.

Bailar grew up in Illinois, the son of a chemist; he graduated from the Yale University School of Medicine in 1955. But a passion for mathematics led him to lay down his stethoscope in favor of a career in research. In 1962, after several other jobs, he wound up at the National Cancer Institute (NCI), eventually becoming its chief of demography. Meanwhile, Bailar earned his stripes, obtaining a Ph.D. in statistics from American University. By the mid-1970s, he was a prominent biostatistician and epidemiologist, serving as the NCI’s deputy associate director for cancer control. That meant he had a front row seat to the new war on cancer.

It was fairly shocking, then, when Bailar went public in January 1976 with his misgivings about a new NCI breast-cancer screening program that was technically under his auspices. In 1972, shortly after Nixon’s announcement, the NCI—working with the American Cancer Society (ACS)—had initiated the Breast Cancer Detection Demonstration Project (BCDDP). The goal was to screen 270,000 American women aged 35 to 74 for breast cancer, offering them free annual mammograms for five years. As with many cancer control strategies, mammography relied on the philosophy of early detection—that finding smaller and presumably more curable breast cancers necessarily saved lives.

Bailar’s salvo, “Mammography: A Contrary View,” appeared in the well-respected medical journal the Annals of Internal Medicine. In it, he registered several concerns, many of them drawing from his statistical background. First, he wrote, the benefits of screening mammography “have not been determined.” Although a research study then being run by the Health Insurance Plan of New York suggested its value for women over 50, no such data existed for younger women. “Not every lesion discovered by screening should be considered a success of the program,” Bailar wrote. This conclusion drew on the epidemiologic concepts of lead-time and length-time bias, which falsely elevate actual survival rates by focusing on the date of cancer detection rather than a patient’s actual outcome.

Second, according to Bailar, the risks of mammography “may be greater than are commonly understood.” Among his concerns was the worry that the radiation from repeated mammograms could actually cause breast cancer and that many of the machines being used in the BCDDP were using higher-than-necessary doses of radiation.

Bailar was most assuredly not a showman or publicity hound. When I interviewed him for a book on the history of breast-cancer control and asked him for a photograph, all he had was a wallet-sized image. Was he a whistleblower, intent on providing an insider’s exposé to the public? Some said yes, but in speaking to his daughter, Melissa Bailar, after his death, a different picture emerged. Her father, she said, “saw himself as a scientist, and the scientific questions and data drove his work, not any political motivation.” It helped, she added, that he did not care much for social conventions and what other people thought. Bailar’s wife Barbara, herself a statistician, termed her husband a lifelong “skeptic.”

Regardless of his motivations, the ramifications of what Bailar was charging were explosive. The uproar only increased when, in 1977, another researcher reported that 66 women who had undergone mastectomies as a result of BCDDP screening had not even had cancer. Through Freedom of Information Act requests, a series of journalists obtained internal documents cataloging the concerns of Bailar and other NCI staff. “There is more than a bit to be appalled about in the archives of the Breast Cancer Demonstration Project,” Daniel Greenberg wrote in the New England Journal of Medicine.

Bailar’s concerns about whether mammography, especially in women under 50, causes more harm than good, inaugurated a debate that persists today. But he had bigger fish to fry. In 1986, on the 15th anniversary of Nixon’s war, Bailar and Elaine M. Smith coauthored an article in the prestigious New England Journal of Medicine titled “Progress against Cancer?” Citing data that overall cancer mortality had only declined by 5 percent since 1971, and that breast-cancer mortality had not budged at all, Bailar and Smith called the war a “qualified failure.” Indeed, data from the Centers for Disease Control indicated that death rates from cancer actually increased until 1991, when they finally peaked. As would be expected, the cancer establishment immediately challenged Bailar and Smith’s findings, calling them “erroneous” and even “reprehensible.” But other skeptics, such as San Diego oncologist Michael Shimkin, called the article a “public service.”

In 1997, Bailar was at it again, writing again in the New England Journal of Medicine with medical student Heather L. Gornik that “blind faith” in screening and treatment had left cancer “undefeated.” Lost in all the rhetoric and reliance on expensive and ineffective surgery, radiation, and chemotherapy—the authors argued—was a strategy for controlling cancer that was smarter than catching it early: preventing it in the first place.

To physicians, cancer survivors, activists, bureaucrats, and legislators waging the war, Bailar was a constant nuisance. In response to his criticisms of the BCDDP, the ACS and NCI had doubled down, arguing that women with conditions such as cystic breasts, early menstrual histories, no past pregnancies, and even simply a high fear of breast cancer should be considered “high risk.” In sum, this meant 80 percent of women were at high risk—a concept that the statistician Bailar labeled “mathematically absurd.”

Bailar particularly raised the ire of Arthur I. Holleb, the ACS medical director at the time of the BCDDP. As a clinician, Holleb remained wedded to the idea that finding breast cancers earlier necessarily led to a better prognosis, something we now know to be untrue. Playing the doctor card, Holleb memorably termed Bailar’s epidemiologic research “the practice of medicine without the tears.” Others were less veiled, suggesting that by interfering with screening, Bailar was in essence “murdering” women.

Despite this harsh criticism, Bailar’s career thrived. After his years at the NCI, Bailar moved to Harvard, McGill, and then the University of Chicago. He retired in 2000. Although he was too sick near the end of his life to comment on the moonshot, he remained a fervent proponent of cancer prevention, according to Melissa Bailar. Bailar’s work received validation from both inside and outside the medical profession. In 1990, he was named a MacArthur Fellow. Three years later, he was named to the prestigious Institute of Medicine.

So was Bailar “right”? It depends whom one asked. Critics reasonably argued that his use of cancer mortality as the primary marker for the war’s success was misleading. And they pointed to prevention programs, most notably with regard to cigarette smoking, that had produced genuine change. But it was hard to challenge his basic premise, which was that the war on cancer had been grossly oversold. The rhetoric that accompanied Nixon’s war had not been subtle: Nixon stated that the effort once used to split the atom and send a man to the moon now needed to be turned to cancer. One oncologist even predicted that the disease would be eradicated by the time of the American bicentennial in 1976. Yet overall mortality from the disease has declined by only about 10 percent since 1971. And while survival from certain cancers, such as breast and prostate, has gone up, deaths from liver, kidney, and pancreatic cancer have actually increased.*

To some degree, the language surrounding the moonshot appears less fanciful than that surrounding the war on cancer. Thanks in part to the funding generated after 1971, we now know much more about cancer, particularly the role played by genetic mutations. We also know that it makes less sense to focus on the organs in which tumors originate (lung cancer, ovarian cancer, etc.), and more sense to see cancer as many different diseases that differ from person to person. Biden himself has cautioned that the goal of new cancer treatments may not be so much to cure the disease but to control it.

Still, proponents of the moonshot, signed into law by President Obama on Dec. 13, 2016, are still using flowery language that may not jibe with the scientific reality. Biden, for example, has called for a “national commitment to end cancer as we know it,” and Obama hopes that the moonshot “make[s] America the country that cures cancer once and for all.” Scientists hope to have an “effective vaccine-based immunotherapy to combat cancer by 2020.”

Such promises may not be realistic. For one thing, the fact that cancer is so many diseases will make it harder to find cures that work for everyone. Second, a series of recent studies have documented high rates of serious side effects to the exact types of immunotherapies being touted as curative. Third, as Norwegian physician Jarle Breivik has argued, cancer is a disease of aging—and we have to die of something.

These are not reasons to stop trying to cure cancer. But they are reminders, like those of John Bailar, to focus on the actual science.

*Correction, Jan. 12, 2017: Due to an editing error, this story originally stated that survival rates from breast and prostate cancers have gone down. They have gone up. (Return.)