The Emperor of All Maladies

  To quantify “progress,” an admittedly hazy metric, Cairns began by revitalizing a fusty old record that had existed since World War II, the cancer registry, a state-by-state statistical record of cancer-related deaths subclassified by the type of cancer involved. “These registries,” Cairns wrote in an article in Scientific American, “yield a rather precise picture of the natural history of cancer, and that is a necessary starting point for any discussion of treatment.” By poring through that record, he hoped to draw a portrait of cancer over time—not over days or weeks, but over decades.

  Cairns began by using the cancer registry to estimate the number of lives saved by the therapeutic advances in oncology since the 1950s. (Since surgery and radiation therapy preceded the 1950s, these were excluded; Cairns was more interested in advances that had emerged from the brisk expansion in biomedical research since the fifties.) He divided these therapeutic advances into various categories, then made numerical conjectures about their relative effects on cancer mortality.

  The first of these categories was “curative” chemotherapy—the approach championed by Frei and Freireich at the NCI and by Einhorn and his colleagues at Indiana. Assuming relatively generous cure rates of about 80 or 90 percent for the subtypes of cancer curable by chemotherapy, Cairns estimated that between 2,000 and 3,000 lives were being saved overall every year—700 children with acute lymphoblastic leukemia, about 1,000 men and women with Hodgkin’s disease, 300 men with advanced testicular cancer, and 20 to 30 women with choriocarcinoma. (Variants of non-Hodgkin’s lymphomas, which were curable with polychemotherapy by 1986, would have added another 2,000 lives, bringing the total up to about 5,000, but Cairns did not include these cures in his initial metric.)

  “Adjuvant” chemotherapy—chemotherapy given after surgery, as in the Bonadonna and Fisher breast cancer trials—contributed to another 10,000 to 20,000 lives saved annually. Finally, Cairns factored in screening strategies such as Pap smears and mammograms that detected cancer in its early stages. These, he estimated loosely, saved an additional 10,000 to 15,000 cancer-related deaths per year. The grand tally, generously speaking, amounted to about 35,000 to 40,000 lives per year.

  That number was to be contrasted with the annual incidence of cancer in 1985—448 new cancer cases diagnosed for every 100,000 Americans, or about 1 million every year—and the mortality from cancer in 1985—211 deaths for every 100,000, or 500,000 deaths every year. In short, even with relatively liberal estimates about lives saved, less than one in twenty patients diagnosed with cancer in America, and less than one in ten of the total number of patients who would die of cancer, had benefited from the advances in therapy and screening.

  Cairns wasn’t surprised by the modesty of that number; in fact, he claimed, no self-respecting epidemiologist should be. In the history of medicine, no significant disease had ever been eradicated by a treatment-related program alone. If one plotted the decline in deaths from tuberculosis, for instance, the decline predated the arrival of new antibiotics by several decades. Far more potently than any miracle medicine, relatively uncelebrated shifts in civic arrangements—better nutrition, housing, and sanitation, improved sewage systems and ventilation—had driven TB mortality down in Europe and America. Polio and smallpox had also dwindled as a result of vaccinations. Cairns wrote, “The death rates from malaria, cholera, typhus, tuberculosis, scurvy, pellagra and other scourges of the past have dwindled in the US because humankind has learned how to prevent these diseases. . . . To put most of the effort into treatment is to deny all precedent.”

  Cairns’s article was widely influential in policy circles, but it still lacked a statistical punch line. What it needed was some measure of the comparative trends in cancer mortality over the years—whether more or less people were dying of cancer in 1985 as compared to 1975. In May 1986, less than a year after Cairns’s article, two of his colleagues from Harvard, John Bailar and Elaine Smith, provided precisely such an analysis in the New England Journal of Medicine.

  To understand the Bailar-Smith analysis, we need to begin by understanding what it was not. Right from the outset, Bailar rejected the metric most familiar to patients: changes in survival rates over time. A five-year survival rate is a measure of the fraction of patients diagnosed with a particular kind of cancer who are alive at five years after diagnosis. But a crucial pitfall of survival-rate analysis is that it can be sensitive to biases.

  To understand these biases, imagine two neighboring villages that have identical populations and identical death rates from cancer. On average, cancer is diagnosed at age seventy in both villages. Patients survive for ten years after diagnosis and die at age eighty.

  Imagine now that in one of those villages, a new, highly specific test for cancer is introduced—say the level of a protein Preventin in the blood as a marker for cancer. Suppose Preventin is a perfect detection test. Preventin “positive” men and women are thus immediately counted among those who have cancer.

  Preventin, let us further suppose, is an exquisitely sensitive test and reveals very early cancer. Soon after its introduction, the average age of cancer diagnosis in village 1 thus shifts from seventy years to sixty years, because earlier and earlier cancer is being caught by this incredible new test. However, since no therapeutic intervention is available even after the introduction of Preventin tests, the average age of death remains identical in both villages.

  To a naive observer, the scenario might produce a strange effect. In village 1, where Preventin screening is active, cancer is now detected at age sixty and patients die at age eighty—i.e., there is a twenty-year survival. In village 2, without Preventin screening, cancer is detected at age seventy and patients die at age eighty—i.e., a ten-year survival. Yet the “increased” survival cannot be real. How can Preventin, by its mere existence, have increased survival without any therapeutic intervention?

  The answer is immediately obvious: the increase in survival is, of course, an artifact. Survival rates seem to increase, although what has really increased is the time from diagnosis to death because of a screening test.

  A simple way to avoid this bias is to not measure survival rates, but overall mortality. (In the example above, mortality remains unchanged, even after the introduction of the test for earlier diagnosis.)

  But here, too, there are profound methodological glitches. “Cancer-related death” is a raw number in a cancer registry, a statistic that arises from the diagnosis entered by a physician when pronouncing a patient dead. The problem with comparing that raw number over long stretches of time is that the American population (like any) is gradually aging overall, and the rate of cancer-related mortality naturally increases with it. Old age inevitably drags cancer with it, like flotsam on a tide. A nation with a larger fraction of older citizens will seem more cancer-ridden than a nation with younger citizens, even if actual cancer mortality has not changed.

  To compare samples over time, some means is needed to normalize two populations to the same standard—in effect, by statistically “shrinking” one into another. This brings us to the crux of the innovation in Bailar’s analysis: to achieve this scaling, he used a particularly effective form of normalization called age-adjustment.

  To understand age-adjustment, imagine two very different populations. One population is markedly skewed toward young men and women. The second population is skewed toward older men and women. If one measures the “raw” cancer deaths, the older-skewed population obviously has more cancer deaths.

  Now imagine normalizing the second population such that this age skew is eliminated. The first population is kept as a reference. The second population is adjusted: the age-skew is eliminated and the death rate shrunk proportionally as well. Both populations now contain identical age-adjusted populations of older and younger men, and the death rate, adjusted accordingly, yields identical cancer-specific death rates. Bailar performed this exercise repeatedly over dozens of years: he divided the population for every year into age cohorts—20–29
years, 30–39 years, 40–49, and so forth—then used the population distribution from 1980 (chosen arbitrarily as a standard) to convert the population distributions for all other years into the same distribution. Cancer rates were adjusted accordingly. Once all the distributions were fitted into the same standard demographic, the populations could be studied and compared over time.

  Bailar and Smith published their article in May 1986—and it shook the world of oncology by its roots. Even the moderately pessimistic Cairns had expected at least a small decrease in cancer-related mortality over time. Bailar and Smith found that even Cairns had been overgenerous: between 1962 and 1985, cancer-related deaths had increased by 8.7 percent. That increase reflected many factors—most potently, an increase in smoking rates in the 1950s that had resulted in an increase in lung cancer.

  One thing was frightfully obvious: cancer mortality was not declining in the United States. There is “no evidence,” Bailar and Smith wrote darkly, “that some thirty-five years of intense and growing efforts to improve the treatment of cancer have had much overall effect on the most fundamental measure of clinical outcome—death.” They continued, “We are losing the war against cancer notwithstanding progress against several uncommon forms of the disease [such as childhood leukemia and Hodgkin’s disease], improvements in palliation and extension of productive years of life. . . . Some thirty-five years of intense effort focused largely on improving treatment must be judged a qualified failure.”

  That phrase, “qualified failure,” with its mincing academic ring, was deliberately chosen. In using it, Bailar was declaring his own war—against the cancer establishment, against the NCI, against a billion-dollar cancer-treatment industry. One reporter described him as “a thorn in the side of the National Cancer Institute.” Doctors railed against Bailar’s analysis, describing him as a naysayer, a hector, a nihilist, a defeatist, a crank.

  Predictably, a torrent of responses appeared in medical journals. One camp of critics contended that the Bailar-Smith analysis appeared dismal not because cancer treatment was ineffective, but because it was not being implemented aggressively enough. Delivering chemotherapy, these critics argued, was a vastly more complex process than Bailar and Smith had surmised—so complex that even most oncologists often blanched at the prospect of full-dose therapy. As evidence, they pointed to a survey from 1985 that had estimated that only one-third of cancer doctors were using the most effective combination regimen for breast cancer. “I estimate that 10,000 lives could be saved by the early aggressive use of polychemotherapy in breast cancer, as compared with the negligible number of lives, perhaps several thousand, now being saved,” one prominent critic wrote.

  In principle, this might have been correct. As the ’85 survey suggested, many doctors were indeed underdosing chemotherapy—at least by the standards advocated by most oncologists, or even by the NCI. But the obverse idea—that maximizing chemotherapy would maximize gains in survival—was also untested. For some forms of cancer (some subtypes of breast cancer, for instance) increasing the intensity of dosage would eventually result in increasing efficacy. But for a vast majority of cancers, more intensive regimens of standard chemotherapeutic drugs did not necessarily mean more survival. “Hit hard and hit early,” a dogma borrowed from the NCI’s experience with childhood leukemia, was not going to be a general solution to all forms of cancer.

  A more nuanced critique of Bailar and Smith came, unsurprisingly, from Lester Breslow, the UCLA epidemiologist. Breslow reasoned that while age-adjusted mortality was one method of appraising the War on Cancer, it was by no means the only measure of progress or failure. In fact, by highlighting only one measure, Bailar and Smith had created a fallacy of their own: they had oversimplified the measure of progress. “The problem with reliance on a single measure of progress,” Breslow wrote, “is that the impression conveyed can vary dramatically when the measure is changed.”

  To illustrate his point, Breslow proposed an alternative metric. If chemotherapy cured a five-year-old child of ALL, he argued, then it saved a full sixty-five years of potential life (given an overall life expectancy of about seventy). In contrast, the chemotherapeutic cure in a sixty-five-year-old man contributed only five additional years given a life expectancy of seventy. But Bailar and Smith’s chosen metric—age-adjusted mortality—could not detect any difference in the two cases. A young woman cured of lymphoma, with fifty additional years of life, was judged by the same metric as an elderly woman cured of breast cancer, who might succumb to some other cause of death in the next year. If “years of life saved” was used as a measure of progress on cancer, then the numbers turned far more palatable. Now, instead of losing the War on Cancer, it appeared that we were winning it.

  Breslow, pointedly, wasn’t recommending one form of calculus over another; his point was to show that measurement itself was subjective. “Our purpose in making these calculations,” he wrote, “is to indicate how sensitive one’s conclusions are to the choice of measure. In 1980, cancer was responsible for 1.824 million lost years of potential life in the United States to age 65. If, however, the cancer mortality rates of 1950 had prevailed, 2.093 million years of potential life would have been lost.”

  The measurement of illness, Breslow was arguing, is an inherently subjective activity: it inevitably ends up being a measure of ourselves. Objective decisions come to rest on normative ones. Cairns or Bailar could tell us how many absolute lives were being saved or lost by cancer therapeutics. But to decide whether the investment in cancer research was “worth it,” one needed to start by questioning the notion of “worth” itself: was the life extension of a five-year-old “worth” more than the life extension of a sixty-year-old? Even Bailar and Smith’s “most fundamental measure of clinical outcome”—death—was far from fundamental. Death (or at least the social meaning of death) could be counted and recounted with other gauges, often resulting in vastly different conclusions. The appraisal of diseases depends, Breslow argued, on our self-appraisal. Society and illness often encounter each other in parallel mirrors, each holding up a Rorschach test for the other.

  Bailar might have been willing to concede these philosophical points, but he had a more pragmatic agenda. He was using the numbers to prove a principle. As Cairns had already pointed out, the only intervention ever known to reduce the aggregate mortality for a disease—any disease—at a population level was prevention. Even if other measures were chosen to evaluate our progress against cancer, Bailar argued that it was indubitably true that prevention, as a strategy, had been neglected by the NCI in its ever-manic pursuit of cures.

  A vast majority of the institute’s grants, 80 percent, were directed toward treatment strategies for cancer; prevention research received about 20 percent. (By 1992, this number had increased to 30 percent; of the NCI’s $2 billion research budget, $600 million was being spent on prevention research.) In 1974, describing to Mary Lasker the comprehensive activities of the NCI, the director, Frank Rauscher, wrote effusively about its three-pronged approach to cancer: “Treatment, Rehabilitation and Continuing Care.” That there was no mention of either prevention or early detection was symptomatic: the institute did not even consider cancer prevention a core strength.

  A similarly lopsided bias existed in private research institutions. At Memorial Sloan-Kettering in New York, for instance, only one laboratory out of nearly a hundred identified itself as having a prevention research program in the 1970s. When one researcher surveyed a large cohort of doctors in the early 1960s, he was surprised to learn that “not one” was able to suggest an “idea, lead or theory on cancer prevention.” Prevention, he noted drily, was being carried out “on a part-time basis.” *

  This skew of priorities, Bailar argued, was the calculated by-product of 1950s-era science; of books, such as Garb’s Cure for Cancer, that had forecast impossibly lofty goals; of the Laskerites’ near-hypnotic conviction that cancer could be cured within the decade; of the steely, insistent enthusiasm of researchers such a
s Farber. The vision could be traced back to Ehrlich, ensconced in the semiotic sorcery of his favorite phrase: “magic bullet.” Progressive, optimistic, and rationalistic, this vision—of magic bullets and miracle cures—had admittedly swept aside the pessimism around cancer and radically transformed the history of oncology. But the notion of the “cure” as the singular solution to cancer had degenerated into a sclerotic dogma. Bailar and Smith noted, “A shift in research emphasis, from research on treatment to research on prevention, seems necessary if substantial progress against cancer is to be forthcoming. . . . Past disappointments must be dealt with in an objective, straightforward and comprehensive manner before we go much further in pursuit of a cure that always seems just out of reach.”

  * Although this line of questioning may be intrinsically flawed since it does not recognize the interrelatedness of preventive and therapeutic research.

  PART FOUR

  PREVENTION IS

  THE CURE

  It should first be noted, however, that the 1960s and 1970s did not witness so much a difficult birth of approaches to prevention that focused on environmental and lifestyle causes of cancer, as a difficult reinvention of an older tradition of interest in these possible causes.

  —David Cantor

  The idea of preventive medicine is faintly un-American. It means, first, recognizing that the enemy is us.