A disturbing report from the front lines of the war on cancer.
In the fall of 1963, not long after Vincent T. DeVita joined the National Cancer Institute as a clinical associate, he and his wife were invited to a co-worker’s party. At the door, one of the institute’s most brilliant researchers, Emil Freireich, presented them with overflowing Martinis. The head of the medical branch, Tom Frei, strode across the room with a lab technician flung over his shoulder, legs kicking and her skirt over her head. DeVita, shocked, tried to hide in a corner. But some time later the N.C.I.’s clinical director, Nathaniel Berlin, frantically waved him over. Freireich, six feet four and built like a lineman, had passed out in the bathtub. Berlin needed help moving him. “Together, we pulled him up, threw his arms over our shoulders, and dragged him out through the party,” DeVita writes, in his memoir, “The Death of Cancer” (Sarah Crichton Books). “Out front, Freireich’s wife, Deanie, sat behind the wheel of their car. We tossed Freireich in the backseat and slammed the door.”
Half a century ago, the N.C.I. was a very different place. It was dingy and underfunded—a fraction of its current size—and home to a raw and unruly medical staff. The orthodoxy of the time was that cancer was a death sentence: the tumor could be treated with surgery or radiation, in order to buy some time, and the patient’s inevitable decline could be eased through medicine, and that was it. At the N.C.I., however, an insurgent group led by Frei and Freireich believed that if cancer drugs were used in extremely large doses, and in multiple combinations and repeated cycles, the cancer could be beaten. “I wasn’t sure if these scientists were maniacs or geniuses,” DeVita writes. But, as he worked with Freireich on the N.C.I.’s childhood-leukemia ward—and saw the fruits of the first experiments using combination chemotherapy—he became a convert.
DeVita decided to try the same strategy on another seemingly hopeless cause, Hodgkin’s lymphoma, a cancer that begins as a solid tumor in the lymph nodes and steadily spreads throughout the body. He teamed up with a fellow-associate named Jack Moxley. Over a few beers one night, at Au Pied de Cochon in Georgetown, the two sketched out a protocol, based loosely on what Frei and Freireich were doing with leukemia. Given the ability of cancer cells to adapt and mutate in the face of threats, they figured they needed four drugs, each effective against Hodgkin’s in its own way, so that whatever cells survived one wave had a chance of being killed by the next. They also had to be careful how frequently they gave the drugs: doses needed to be high enough to wipe out the cancer cells but not so high that they killed the patient. After several months, they settled on a regimen called MOMP: three eleven-day rounds of nitrogen mustard, Oncovin (a brand of vincristine), methotrexate, and prednisone, interspersed with ten-day recovery cycles.
“The side effects were almost immediate,” DeVita writes:
The sound of vomiting could be heard along the hallway. Night after night, Moxley and I paced outside the rooms of our patients, fearful of what might happen. Over the weeks that followed, they lost weight and grew listless, and their platelet counts sank lower and lower to dangerous levels.
Then came the surprise. Twelve of the fourteen patients in the initial trial went into remission—and nine stayed there as the months passed. In most cases, the tumors disappeared entirely, something that had never before been seen in the treatment of solid tumors. In the spring of 1965, DeVita went to Philadelphia to present the results to the annual meeting of the American Association for Cancer Research. He stood up before the crowd and ran triumphantly through the data: “ ‘Our patients were, therefore,’ I said, savoring the dramatic conclusion, ‘in complete remission.’ ”
What happened? An illustrious cancer expert named David Karnofsky made a narrow point about the appropriateness of the term “complete remission.” After that, nothing: “There were a few perfunctory questions about the severity of the side effects. But that was it.” History had been made in the world of cancer treatment, and no one seemed to care.
Vince DeVita served as the head of the National Cancer Institute from 1980 to 1988. He went on to serve as the physician-in-chief of the Memorial Sloan Kettering Cancer Center, in New York, and then ran the Yale Cancer Center, in New Haven. For the past half century, he has been at the forefront of the fight against one of the world’s most feared diseases, and in “The Death of Cancer” he has written an extraordinary chronicle. DeVita’s book is nothing like Siddhartha Mukherjee’s magisterial “The Emperor of All Maladies.” Mukherjee wrote a social and scientific biography of the disease. DeVita, as befits someone who spent a career at the helm of various medical bureaucracies, has written an institutional history of the war on cancer. His interest is in how the various factions and constituencies involved in that effort work together—and his conclusions are deeply unsettling.
When his first go-round as a clinical associate at the N.C.I. was up, DeVita took a post as a resident at Yale. At what was supposed to be a world-class hospital, he discovered that the standard of care for many cancers was woefully backward. Freireich had taught DeVita to treat Pseudomonas meningitis in leukemia patients by injecting an antibiotic directly into the spinal column—even though the drug’s label warned against that method of administration. That was the only way, Freireich believed, to get the drug past the blood-brain barrier. At Yale, DeVita writes, “you just didn’t do that kind of thing. As a result, I watched leukemic patients die.” Leukemia patients also sometimes came down with lobar pneumonia. Conventional wisdom held that that ought to be treated with antibiotics. But N.C.I. researchers had figured out that the disease was actually a fungal infection, and had to be treated with a different class of drug. “When I saw this condition in patients with leukemia and pointed it out to the chief of infectious diseases at Yale, he didn’t believe me—even when the lab tests proved my point,” DeVita continues. More patients died. Leukemia patients on chemotherapy needed platelets for blood transfusions. But DeVita’s superiors at Yale insisted there was no evidence that transfusions made a difference, despite the fact that Freireich had already proved that they did. “Ergo, at Yale,” DeVita says, “I watched patients bleed to death.”
Later, when DeVita and his fellow N.C.I. researcher George Canellos wanted to test a promising combination-chemotherapy treatment for advanced breast cancer, they had to do their trial overseas, because they couldn’t win the coöperation of surgeons at either of the major American cancer centers, Memorial Sloan Kettering or M. D. Anderson. When the cancer researcher Bernard Fisher did a study showing that there was no difference in outcome between radical mastectomies and the far less invasive lumpectomies, he called DeVita in distress. He couldn’t get the study published. “Breast surgeons made their living doing radical or total mastectomies, and they did not want to hear that that was no longer necessary,” DeVita writes. “Fisher had found it difficult to get patients referred to his study, in fact, because of this resistance.” The surgeons at Memorial Sloan Kettering Cancer Center were so stubborn that they went on disfiguring their patients with radical mastectomies for years after Fisher’s data had shown the procedure to be unnecessary. “The Death of Cancer” is an angry book, in which one of the critical figures in twentieth-century oncology unloads a lifetime of frustration with the obduracy and closed-mindedness of his profession. DeVita concludes, “There are incredibly promising therapies out there. If used to their fullest potential for all patients, I believe we could cure an additional 100,000 patients a year.” He is not the first to point out the shortcomings of clinical practice, of course. What sets “The Death of Cancer” apart is what he proposes to do about it.
After DeVita was rebuffed at the American Association for Cancer Research meeting, he and Moxley went back to the drawing board. They needed to do more than push patients into remission.
Their first step was to alter the combination of drugs in their protocol, replacing methotrexate with a newer compound called procarbazine. Next, they reëxamined the schedule of treatment. Combination chemotherapy is a delicate balancing act. Cancer drugs are typically so toxic that they can be given only in short bursts, so that patients can regain their strength. If the breaks are too long, though, the cancer comes roaring back. In the first trial, they had simply followed the schedule that Freireich used in treating leukemia. Hodgkin’s cells, however, were different. They divided more slowly—and, since cancer cells are most vulnerable when they are dividing, that suggested that the Hodgkin’s schedule needed to be a lot longer.
So MOMP became MOPP: two full doses of nitrogen mustard and vincristine on the first and the eighth days, and daily doses of procarbazine and prednisone for fourteen days, followed by two weeks of rest. Since only twenty per cent of Hodgkin’s cells would divide during the course of that cycle, the regimen would have to be repeated at least six times. A second trial was launched, and the outcome was unequivocal: the regimen had beaten the disease.
When the new results were published, in 1970, the response was better, but there was still considerable resistance. A crucial presentation at Memorial Sloan Kettering was met with “tepid” applause, after which one oncologist after another got up to complain that MOPP didn’t work. DeVita was told that his data must be wrong.
Baffled, he asked one of the hospital’s leading oncologists, Barney Clarkson, to explain exactly how he was administering the MOPP protocol. Clarkson answered that he and his colleagues had decided to swap the nitrogen mustard in DeVita’s formula for a drug called thiotepa. This was a compound they had developed in-house at Memorial Sloan Kettering and felt partial to. So MOPP was now TOPP. DeVita writes:
They’d also cut the dose of procarbazine in half, because it made patients nauseous. And they’d reduced the dose of vincristine drastically because of the risk of nerve damage. They’d also added, at a minimum, an extra two weeks between cycles so that patients would have fully recovered from the toxic effects of the prior dose before they got the next. They gave no thought to the fact that the tumor would have been back on its feet by then, too, apparently.
These alterations had not been tested or formally compared with DeVita’s original formula. They were simply what the oncologists at Memorial Sloan Kettering felt made more sense. After an hour, DeVita had had enough:
“Why in God’s name have you done this?” he asked.
A voice piped up from the audience. “Well, Vince, most of our patients come to us on the subway, and we don’t want them to vomit on the way home.”
Here were physicians at one of the world’s greatest cancer hospitals denying their patients a potentially life-saving treatment because their way felt better. Stories like this are why DeVita believes that a hundred thousand cancer patients in the United States die needlessly every year. The best innovations are sometimes slow to make their way into everyday medical practice. Hence the sustained push, in recent years, toward standardizing treatments. If doctors aren’t following “best practices,” it seems logical that we should write up a script describing what those best practices are and compel them to follow it.
But here “The Death of Cancer” takes an unexpected turn. DeVita doesn’t think his experience with the stubborn physicians at Memorial Sloan Kettering or at Yale justifies greater standardization. He is wary of too many scripts and guidelines. What made the extraordinary progress against cancer at the N.C.I. during the nineteen-sixties and seventies possible, in his view, was the absence of rules. A good illustration was Freireich’s decision to treat Pseudomonas meningitis by injecting an antibiotic directly into the spinal fluid. DeVita writes:
The first time Freireich told me to do it, I held up the vial and showed him the label, thinking that he’d possibly missed something. “It says right on there, ‘Do not use intrathecally,’ ” I said. Freireich glowered at me and pointed a long bony finger in my face. “Do it!” he barked. I did it, though I was terrified. But it worked every time.
Clinical progress against a disease as wily and dimly understood as cancer, DeVita argues, happens when doctors have the freedom to try unorthodox things—and he worries that we have lost sight of that fact. By way of example, he tells the story of a friend of his, Lee, who was diagnosed with advanced prostate cancer at the age of sixty. According to the practice guidelines, the best option for Lee was androgen-deprivation therapy, or A.D.T., which slows down the cancer cells by denying them testosterone. That’s what Lee’s doctor recommended. DeVita understood why: there are strong incentives—like the threat of malpractice suits—for doctors to adhere to treatment protocols. But DeVita judged that Lee’s cancer was so aggressive that A.D.T. would buy him only a short reprieve. The guidelines limited Lee’s treatment options at a moment when he needed maximum flexibility.
“Over the years, we’ve gained more tools for treating cancer, but the old ability to be flexible and adapt has disappeared,” DeVita writes:
Guidelines are backwards looking. With cancer, things change too rapidly for doctors to be able to rely on yesterday’s guidelines for long. These guidelines need to be updated frequently, and they rarely are, because this takes time and money. . . . Reliance on such standards inhibits doctors from trying something new.
DeVita’s first thought was to get Lee enrolled in a pioneering trial at the Mayo Clinic, where surgeons were removing the prostate along with all surrounding lymph nodes. Fifteen per cent of patients who underwent the procedure survived free of disease. The Mayo doctors wouldn’t operate on Lee, however. His cancer was too advanced. So DeVita found someone who would. “I can be very persuasive,” he writes. Then he managed to get Lee enrolled in an experimental-drug trial for relapsed prostate-cancer patients—only to discover that the study’s protocol called for treatment to end after a fixed number of doses. DeVita felt that Lee needed a much longer course. Lee sought an exemption from the rules of the study, which required a judgment from the hospital’s institutional review board. The lead investigator declined to take it up. DeVita was devastated, though hardly surprised. The system was built to be inflexible.
DeVita’s struggle to keep his friend alive goes on for years. He finagles his way into one experimental trial after another. He improvises. He works his contacts. Finally, with Lee at the end of the line, DeVita hears of an experimental drug called abiraterone. But he can’t get Lee into the trial: the study’s protocol forbids it. DeVita tries to find his way around the rules and fails—and he’s heartbroken when he learns, after Lee finally succumbs to the disease, that abiraterone is so effective against advanced prostate cancer that the trial is stopped in mid-course and the patients in the control group are switched over to the new drug. “I could have told you a story with a happy ending,” DeVita writes, speaking of what he is sure was his friend’s premature death. “I instead chose to tell you one that could have had a happy ending because it illustrates what has been, for me, a source of perennial frustration: at this date, we are not limited by the science; we are limited by our ability to make good use of the information and treatments we already have.”
Here we have a paradox. The breakthroughs made at the N.C.I. in the nineteen-sixties and seventies were the product of a freewheeling intellectual climate. But that same freewheeling climate is what made it possible for the stubborn doctors at Memorial Sloan Kettering to concoct their non-cure. The social conditions that birthed a new idea in one place impeded the spread of that same idea in another. People who push for greater innovation in the marketplace often naïvely assume that what is good for the innovator is also, down the line, good for the diffusion of their ideas. And people worried about diffusion often position themselves as the friends of innovation, as if a system that does well at spreading good ideas necessarily makes it easier to come up with good ideas. The implication of “The Death of Cancer” is, on the contrary, that innovation and diffusion can sometimes conflict.
Practice guidelines would have made the task of curing Hodgkin’s patients with DeVita’s regimen a lot easier. But had those guidelines been in place in the mid-sixties, when DeVita was making the rounds on behalf of his new treatment, they would have imposed a tax on other innovators. The obstacles he encountered in trying to save his friend Lee, similarly, were not capricious or arbitrary. They were there to insure that the results of clinical trials were as clear and persuasive as possible. It’s just that they had a cost—Lee’s death—and in DeVita’s mind that cost was too high.
The angriest chapter of “The Death of Cancer” is devoted to the Food and Drug Administration, because DeVita believes that it has fundamentally misunderstood the trade-off between diffusion and innovation. The agency wants all new drugs to be shown to be safe and efficacious, to be as good as or better than existing therapies (or a placebo) in a randomized experiment involving the largest possible number of patients. For example, the F.D.A. might ask that patients getting an experimental treatment have better long-term survival rates than those receiving drug treatments already in use. The F.D.A. is the country’s diffusion gatekeeper: its primary goal is to make sure that good drugs get a gold star and bad drugs never make it to market.
DeVita reminds us, though, that this gatekeeping can hinder progress. A given tumor, for instance, can rarely be stopped with a single drug. Cancer is like a door with three locks, each of which requires a different key. Suppose you came up with a drug that painlessly opened the first of those three locks. That drug would be a breakthrough. But it can’t cure anything on its own. So how do you get it through a trial that requires proof of efficacy—especially if you don’t yet know what the right keys for the two remaining locks are? Since cancer comes in a dizzying variety of types and subtypes, each with its own molecular profile, we want researchers to be free to experiment with different combinations of keys. Instead, DeVita argues, the F.D.A. has spent the past two decades pushing cancer medicine in the opposite direction. He continues:
Drugs are now approved not for a specific cancer or for general use in a variety of cancers but for a specific stage of a specific cancer and specifically after and only after patients have had all current treatments, which are listed drug by drug, and the treatments have all failed. Doctors risk F.D.A. censure if they use an approved drug under any other circumstances, and patients are penalized because insurance companies won’t pay for treatments not approved by the F.D.A.The vital insight gained by using an approved drug in a different way for a different tumor has been lost.
There’s a second problem with the “efficacy” requirement. Suppose Drug A, the existing treatment for a certain type of cancer, wipes out all but a billion cells in the typical patient’s tumor. Drug B, your alternative, wipes out all but a handful. DeVita points out two curious facts. First, a typical tumor has so many billions of cells that even a drug that leaves a billion cells untouched will look good after an initial treatment cycle. More important, after five years the patients on both Drugs A and B may have identical survival rates. That’s because of something called the Norton-Simon effect: smaller populations of cancer cells grow back faster than larger populations. But, in reality, Drugs A and B aren’t identical. If you are designing a combination of drugs to cure a cancer, DeVita writes, “the treatment that reduced the population to a few cells is the one you want to go forward with.” How many researchers and companies sit on promising therapies because they don’t want to spend several hundred million dollars on a clinical trial, only to fall short of the F.D.A.’s high bar?
DeVita would have the F.D.A. take a step sideways—away from worrying exclusively about standards and safety, and closer to the innovation end of the continuum. In this respect, his position echoes that of Peter Huber, who in his 2013 book, “The Cure in the Code,” called on the F.D.A. to stop evaluating drugs as cures and start evaluating them as tools—“molecular scalpels, clamps, sutures, or dressings, to be picked off the shelf and used carefully but flexibly down at the molecular level.”
What critics like DeVita want, in other words, is a return to the world of Freireich’s N.C.I., where clinicians had the freedom to tinker and improvise, and DeVita’s portrait of the way things were gives us a glimpse of what the future may look like. Discretion means more MOPPS. But it also, inevitably, means more TOPPS. Discretion means Freireich, the great genius, growling “Do it.” But surely Barney Clarkson growled “Do it” as well, when some fresh-faced clinical associate questioned the wisdom of substituting thiotepa for nitrogen mustard. Modern medicine is intent on addressing “practice variation”—on bringing bad doctors up to the level of the good ones. Going back to the days of the old N.C.I. makes that problem worse, not better. If you think that there are more Freireichs than Barney Clarksons out there, that is a trade worth making. But DeVita does not acknowledge how difficult that change might prove to be.
When DeVita faced the naysayers at Memorial Sloan Kettering, who worried about their Hodgkin’s patients on the subway ride home, he informed them curtly, “If you told those patients that the choice was between being cured and vomiting, or not vomiting and dying, don’t you think they might have opted to take a cab?” This is how diffusion happens in a world without a diffusion gatekeeper. But how many doctors are capable of that kind of hand-to-hand combat? Life on the innovation end of the continuum is volatile, fractious, and personal—less a genteel cocktail party, governed benignly by bureaucratic fiat, than the raucous bender where your boss passes out in a bathtub. When DeVita returned to Memorial Sloan Kettering years later, as the physician-in-chief, the hospital got better. But DeVita didn’t last, which will scarcely come as a surprise to anyone who has read his book. “The problem with Vince,” the hospital’s president reportedly said, in announcing his departure, “is that he wants to cure cancer.”