The Emperor of All Maladies

by Siddhartha Mukherjee

  Early the next morning, though, Slamon was back on the telephone. He apologized for the intrusion, but her decision had troubled him all night. Of all the variants of Her-2 amplification that he had encountered, hers had been truly extraordinary; Bradfield’s tumor was chock-full of Her-2, almost hypnotically drunk on the oncogene. He begged her once again to join his trial.

  “Survivors look back and see omens, messages they missed,” Joan Didion wrote. For Bradfield, Slamon’s second phone call was an omen that was not missed; something in that conversation pierced through a shield that she had drawn around herself. On a warm August morning in 1992, Bradfield visited Slamon in his clinic at UCLA. He met her in the hallway and led her to a room in the back. Under the microscope, he showed her the breast cancer that had been excised from her body, with its dark ringlets of Her-2 labeled cells. On a whiteboard, he drew a step-by-step picture of an epic scientific journey. He began with the discovery of neu, its rediscovery in Ullrich’s lab, the struggles to produce a drug, culminating in the antibody stitched together so carefully by Shepard and Carter. Bradfield considered the line that stretched from oncogene to drug. She agreed to join Slamon’s trial.

  It was an extraordinarily fortunate decision. In the four months between Slamon’s phone call and the first infusion of Herceptin, Bradfield’s tumor had erupted, spraying sixteen new masses into her lung.

  Fifteen women, including Bradfield, enrolled in Slamon’s trial at UCLA in 1992. (The number would later be expanded to thirty-seven.) The drug was given for nine weeks, in combination with cisplatin, a standard chemotherapy agent used to kill breast cancer cells, both delivered intravenously. As a matter of convenience, Slamon planned to treat all the women on the same day and in the same room. The effect was theatrical; this was a stage occupied by a beleaguered set of actors. Some women had begged and finagled their way into Slamon’s trial through friends and relatives; others, such as Bradfield, had been begged to join it. “All of us knew that we were living on borrowed time,” Bradfield said, “and so we felt twice as alive and lived twice as fiercely.” A Chinese woman in her fifties brought stash after stash of traditional herbs and salves that she swore had kept her alive thus far; she would take oncology’s newest drug, Herceptin, only if she could also take its most ancient drugs with it. A frail, thin woman in her thirties, recently relapsed with breast cancer after a bone marrow transplant, glowered silently and intensely in a corner. Some treated their illness reverentially. Some were bewildered, some too embittered to care. A mother from Boston in her midfifties cracked raunchy jokes about her cancer. The daylong drill of infusions and blood tests was exhausting. In the late evening, after all the tests, the women went their own ways. Bradfield went home and prayed. Another woman soused herself with martinis.

  The lump on Bradfield’s neck—the only tumor in the group that could be physically touched, measured, and watched—became the compass for the trial. On the morning of the first intravenous infusion of the Her-2 antibody, all the women came up to feel the lump, one by one, running their hands across Bradfield’s collarbone. It was a peculiarly intimate ritual that would be repeated every week. Two weeks after the first dose of the antibody, when the group filed past Bradfield, touching the node again, the change was incontrovertible. Bradfield’s tumor had softened and visibly shrunk. “We began to believe that something was happening here,” Bradfield recalled. “Suddenly, the weight of our good fortune hit us.”

  Not everyone was as fortunate as Bradfield. Exhausted and nauseous one evening, the young woman with relapsed metastatic cancer was unable to keep down the fluids needed to hydrate her body. She vomited through the night and then, too tired to keep drinking and too sick to understand the consequences, fell back into sleep. She died of kidney failure the next week.

  Bradfield’s extraordinary response continued. When the CT scans were repeated two months into the trial, the tumor in her neck had virtually disappeared, and the lung metastases had also diminished both in number and size. The responses in many of the thirteen other women were more ambiguous. At the three-month midpoint of the trial, when Slamon reviewed the data with Genentech and the external trial monitors, tough decisions clearly needed to be made. Tumors had remained unchanged in size in some women—not shrunk, but static: was this to be counted as a positive response? Some women with bone metastasis reported diminished bone pain, but pain could not objectively be judged. After a prolonged and bitter debate, the trial coordinators suggested dropping seven women from the study because their responses could not be quantified. One woman discontinued the drug herself. Only five of the original cohort, including Bradfield, continued the trial to its six-month end point. Embittered and disappointed, the others returned to their local oncologists, their hopes for a miracle drug again dashed.

  Barbara Bradfield finished eighteen weeks of therapy in 1993. She survives today. A gray-haired woman with crystalline gray-blue eyes, she lives in the small town of Puyallup near Seattle, hikes in the nearby woods, and leads discussion groups for her church. She vividly remembers her days at the Los Angeles clinic—the half-lit room in the back where the nurses dosed the drugs, the strangely intimate touch of the other women feeling the node in her neck. And Slamon, of course. “Dennis is my hero,” she said. “I refused his first phone call, but I have never, ever, refused him anything since that time.” The animation and energy in her voice crackled across the phone line like an electrical current. She quizzed me about my research. I thanked her for her time, but she, in turn, apologized for the distraction. “Get back to work,” she said, laughing. “There are people waiting for discoveries.”

  * Ullrich actually found the human homolog of the mouse neu gene. Two other groups independently discovered the same gene.

  * The drug is also known by its pharmacological name Trastuzumab; the “ab” suffix is used to denote the fact that this is an antibody.

  Drugs, Bodies, and Proof

  Dying people don’t have time or energy. We can’t keep doing this one woman, one drug, one company at a time.

  —Gracia Buffleben

  It seemed as if we had entered a brave new world of precisely targeted, less toxic, more effective combined therapies.

  —Breast Cancer Action Newsletter, 2004

  By the summer of 1993, news of Slamon’s early-phase trial had spread like wildfire through the community of breast cancer patients, fanning out through official and unofficial channels. In waiting rooms, infusion centers, and oncologists’ offices, patients spoke to other patients describing the occasional but unprecedented responses and remissions. Newsletters printed by breast cancer support groups whipped up a frenzy of hype and hope about Herceptin. Inevitably, a tinderbox of expectations was set to explode.

  The issue was “compassionate use.” Her-2 positive breast cancer is one of the most fatal and rapidly progressive variants of the disease, and patients were willing to try any therapy that could produce a clinical benefit. Breast cancer activists pounded on Genentech’s doors to urge the release of the drug to women with Her-2 positive cancer who had failed other therapies. These patients, the activists argued, could not wait for the drug to undergo interminable testing; they wanted a potentially lifesaving medicine now. “True success happens,” as one writer put it in 1995, “only when these new drugs actually enter bodies.”

  For Genentech, though, “true success” was defined by vastly different imperatives. Herceptin had not been approved by the FDA; it was a molecule in its infancy. Genentech wanted carefully executed early-phase trials—not just new drugs entering bodies, but carefully monitored drugs entering carefully monitored bodies in carefully monitored trials. For the next phase of Herceptin trials launched in 1993, Genentech wanted to stay small and focused. The number of women enrolled in these trials had been kept to an absolute minimum: twenty-seven patients at Sloan-Kettering, sixteen at UCSF, and thirty-nine at UCLA, a tiny cohort that the company intended to follow deeply and meticulously over time. “We do not provide . . .
compassionate use programs,” Curd curtly told a journalist. Most doctors involved in the early-phase trials agreed. “If you start making exceptions and deviating from your protocol,” Debu Tripathy, one of the leaders of the UCSF trial, said, “then you get a lot of patients whose results are not going to help you understand whether a drug works or not. All you’re doing is delaying . . . being able to get it out into the public.”

  Outside the cloistered laboratories of Genentech, the controversy ignited a firestorm. San Francisco, of course, was no stranger to this issue of compassionate use versus focused research. In the late 1980s, as AIDS had erupted in the city, filling up Paul Volberding’s haunted Ward 5B with scores of patients, gay men had coalesced into groups such as ACT UP to demand speedier access to drugs, in part through compassionate use programs. Breast cancer activists saw a grim reflection of their own struggle in these early battles. As one newsletter put it, “Why do women dying of breast cancer have such trouble getting experimental drugs that could extend their lives? For years, AIDS activists have been negotiating with drug companies and the FDA to obtain new HIV drugs while the therapies were still in clinical trials. Surely women with metastatic breast cancer for whom standard treatments have failed should know about, and have access to, compassionate use programs for experimental drugs.”

  Or, as another writer put it, “Scientific uncertainty is no excuse for inaction. . . . We cannot wait for ‘proof.’”

  Marti Nelson, for one, certainly could not afford to wait for proof. An outgoing, dark-haired gynecologist in California, Nelson had discovered a malignant mass in her breast in 1987, when she was just thirty-three. She had had a mastectomy and multiple cycles of chemo, then returned to practicing medicine in a San Francisco clinic. The tumor had disappeared. The scars had healed. Nelson thought that she might have been cured.

  In 1993, six years after her initial surgery, Nelson noticed that the scar in her breast had begun to harden. She waved it away. But the hardened line of tissue outlining her breast was relapsed breast cancer, worming its way insidiously along the scar lines and coalescing into small, matted masses in her chest. Nelson, who compulsively followed the clinical literature on breast cancer, had heard of Her-2. Reasoning presciently that her tumor might be Her-2 positive, she tried to have her own specimen tested for the gene.

  But Nelson soon found herself inhabiting a Kafkaesque nightmare. Her HMO insisted that because Herceptin was in investigational trials, testing the tumor for Her-2 was useless. Genentech insisted that without Her-2 status confirmed, giving her access to Herceptin was untenable.

  In the summer of 1993, with Nelson’s cancer advancing daily and spewing out metastases into her lungs and bone marrow, the struggle took an urgent, political turn. Nelson contacted the Breast Cancer Action project, a local San Francisco organization connected with ACT UP, to help her get someone to test her tumor and obtain Herceptin for compassionate use. BCA, working through its activist networks, asked several laboratories in and around San Francisco to test Nelson’s tumor. In October 1994, the tumor was finally tested for Her-2 expression at UCSF. It was strikingly Her-2 positive. She was an ideal candidate for the drug. But the news came too late. Nine days later, still awaiting Herceptin approval from Genentech, Marti Nelson drifted into a coma and died. She was forty-one years old.

  For BCA activists, Nelson’s death was a watershed event. Livid and desperate, a group of women from the BCA stormed through the Genentech campus on December 5, 1994, to hold a fifteen-car “funeral procession” for Nelson with placards showing Nelson in her chemo turban before her death. The women shouted and honked their horns and drove their cars through the manicured lawns. Gracia Buffleben, a nurse with breast cancer and one of the most outspoken leaders of the BCA, parked her car outside one of the main buildings and handcuffed herself to the steering wheel. A furious researcher stumbled out of one of the lab buildings and shouted, “I’m a scientist working on the AIDS cure. Why are you here? You are making too much noise.” It was a statement that epitomized the vast and growing rift between scientists and patients.

  Marti Nelson’s “funeral” woke Genentech up to a new reality. Outrage, rising to a crescendo, threatened to spiral into a public relations disaster. Genentech had a narrow choice: unable to silence the activists, it was forced to join them. Even Curd admitted, if somewhat begrudgingly, that the BCA was “a tough group [and] their activism is not misguided.”

  In 1995, a small delegation of Genentech scientists and executives thus flew to Washington to meet Frances Visco, the chair of the National Breast Cancer Coalition (NBCC), a powerful national coalition of cancer activists, hoping to use the NBCC as a neutral intermediary between the company and the local breast cancer activists in San Francisco. Pragmatic, charismatic, and savvy, Visco, a former attorney, had spent nearly a decade immersed in the turbulent politics of breast cancer. Visco had a proposal for Genentech, but her terms were inflexible: Genentech had to provide an expanded access program for Herceptin. This program would allow oncologists to treat patients outside clinical trials. In return, the National Breast Cancer Coalition would act as a go-between for Genentech and its embittered and alienated community of cancer patients. Visco offered to join the planning committee of the phase III trials of Herceptin, and to help recruit patients for the trial using the NBCC’s extensive network. For Genentech, this was a long-overdue education. Rather than running trials on breast cancer patients, the company learned to run trials with breast cancer patients. (Genentech would eventually outsource the compassionate-access program to a lottery system run by an independent agency. Women applied to the lottery and “won” the right to be treated, thus removing the company from any ethically difficult decision-making.)

  It was an uneasy triangle of forces—academic researchers, the pharmaceutical industry, and patient advocates—united by a deadly disease. Genentech’s next phase of trials involved large-scale, randomized studies on thousands of women with metastatic Her-2 positive cancer, comparing Herceptin treatment against placebo treatment. Visco sent out newsletters from the NBCC to patients using the coalition’s enormous Listservs. Kay Dickersin, a coalition member and an epidemiologist, joined the Data Safety and Monitoring board of the trial, underscoring the new partnership between Genentech and the NBCC, between academic medicine and activism. And an all-star team of breast oncologists was assembled to run the trial: Larry Norton from Sloan-Kettering, Karen Antman from Columbia, Daniel Hayes from Harvard, and, of course, Slamon from UCLA.

  In 1995, empowered by the very forces that it had resisted for so long, Genentech launched three independent phase III trials to test Herceptin. The most pivotal of the three was a trial labeled 648, randomizing women newly diagnosed with metastatic breast cancer to standard chemotherapy alone versus chemotherapy with Herceptin added. Trial 648 was launched in 150 breast cancer clinics around the world. The trial would enroll 469 women and cost Genentech $15 million to run.

  In May 1998, eighteen thousand cancer specialists flocked to Los Angeles to attend the thirty-fourth meeting of the American Society of Clinical Oncology, where Genentech would unveil the data from the Herceptin trials, including trial 648. On Sunday, May 17, the third day of the meeting, an expectant audience of thousands piled into the stuffy central amphitheater at the convention center to attend a special session dedicated to Her-2/neu in breast cancer. Slamon was slated to be the last speaker. A coil of nervous energy, with the characteristic twitch in his mustache, he stood up at the podium.

  Clinical presentations at ASCO are typically sanitized and polished, with blue-and-white PowerPoint slides depicting the bottom-line message using survival curves and statistical analyses. But Slamon began—relishing this pivotal moment—not with numbers and statistics, but with forty-nine smudgy bands on a gel run by one of his undergraduate students in 1987. Oncologists slowed down their scribbling. Reporters squinted their eyes to see the bands on the gel.

  That gel, he reminded his audience, had identified a gene wi
th no pedigree—no history, no function, no mechanism. It was nothing more than an isolated, amplified signal in a fraction of breast cancer cases. Slamon had gambled the most important years of his scientific life on those bands. Others had joined the gamble: Ullrich, Shepard, Carter, Botstein and Levinson, Visco and the activists, pharma executives and clinicians and Genentech. The trial results to be announced that afternoon represented the result of that gamble. But Slamon wouldn’t—he couldn’t—rush to the end point of the journey without reminding everyone in the room of the fitful, unsanitized history of the drug.

  Slamon paused for a theatrical moment before revealing the results of the trial. In the pivotal 648 study, 469 women had received standard cytotoxic chemotherapy (either Adriamycin and Cytoxan in combination, or Taxol) and were randomized to receive either Herceptin or a placebo. In every conceivable index of response, women treated with the addition of Herceptin had shown a clear and measurable benefit. Response rates to standard chemotherapy had moved up 150 percent. Tumors had shrunk in half the women treated with Herceptin compared to a third of women in the control arm. The progression of breast cancer had been delayed from four to seven and a half months. In patients with tumors heavily resistant to the standard Adriamycin and Cytoxan regimen, the benefit had been the most marked: the combination of Herceptin and Taxol had increased response rates to nearly 50 percent—a rate unheard of in recent clinical experience. The survival rate would also follow this trend. Women treated with Herceptin lived four or five months longer than women in the control group.