Part 1: Inside the FDA...Case Study ,Chasing Cancer

INSIDE THE FDA 
The Business and Politics Behind the Drugs We Take and the Food We Eat 
By Fran Hawthorne  
Introduction

The Holiday Inn in Bethesda, Maryland, is an unassuming stucco building tucked sideways off a slow commercial street, across from a Pizza Hut, a gas station, and a mini-mart. You enter from the side driveway and climb up a wide, curving staircase to reach the Versailles II ballroom on the second floor. 

On a sunny February morning in 2004, as week-old snow lingered in piles at the edge of the sidewalk, it was standing room only in that ballroom. Some three hundred people had come—parents, grandparents, siblings, and friends, bearing posters and white satin ribbons—to talk to the United States Food and Drug Administration about the medicine that had killed someone they loved. 

The long room was decorated in shades of beige and blue, with textured beige wallpaper, beige-and-brown carpeting in a fleur-de-lis motif, and a turquoise ceiling studded with 16 crystal chandeliers. At one end, three tables had been arranged in a large “U” for the two panels of 36 outside experts who had been summoned to advise the FDA, along with a few agency staffers. Facing them were rows of burgundy-and-purple brocade chairs, a battery of TV cameras, and a microphone for the audience. 

They came from Rhode Island and California, from Texas and Colorado, Arizona and Pennsylvania. Most of them were middle-aged, the men in business suits, the women in nice slacks. One mother quoted the Book of Revelations; another wore a button supporting Democratic Senator John Edwards for president. A 10-year-old girl read an Archie comic book, while a boy of about six played with his GameBoy. In the hall outside the ballroom, one blonde woman asked another, “Was your daughter suicidal?” 

They came with tales of the anguish and horror that they and their families had lived through after a teenage son, daughter, grandchild, or friend had started taking an antidepressant medication legally prescribed by their doctor and approved by the FDA. While on the medication, the teenagers had killed themselves, or someone else, or tried to. The families blamed the drugs, and they wanted the FDA to do something to prevent more horror stories. 

Tom and Kathy Woodward. Their 17-year-old daughter Julie had hung herself in the garage six months earlier, after seven days on Zoloft. 

Terri Williams. Her 14-year-old son Jacob had hung himself in the attic with a belt while taking Prozac. A friend held up a picture of Jacob in his football uniform. 

Corey Baadsgaard with his father, Jay. Corey had used first Paxil, then Effexor Then he woke up in a juvenile detention center one morning. Apparently, he had carried a hunting rifle to school and held his class hostage, but he didn’t remember any of that. “These drugs are hell. Look what they’ve done to my son!” Jay Baadsgaard shouted, his voice hoarse. He strode out of the ballroom, slamming the door behind him. 

Glenn McIntosh. His daughter Caitlin hung herself in the girls’ bathroom in her middle school when she was 12; she had been using Paxil and Zoloft. She had been a straight-A student and had hoped to be a veterinarian. 

Eileen and Todd Shivak. Their 11-year-old son Michael had taken Paxil. He was still alive. But he had tried to slash his wrists in class, had run in front of a moving car, and was now afraid of doctors, teachers, and police. “His peers think of him as a freak,” the Shivaks said. 

One after another, more than 60 people spoke. 

The medications had all been approved by the FDA years ago, starting in 1988 for adults. Millions of people said the pills had saved them from unbearable depression, anxiety, compulsive behavior, panic attacks, and stomach pains. Yet the medicines had been controversial almost from the start, because of their ability to alter people’s moods and personality so powerfully. Almost 13 years earlier the FDA had convened a similar meeting of outside experts to discuss whether these pills led to suicidal tendencies in adults; some of the same people now in the audience at the Holiday Inn had been there, too. Back then, emotions had been so intense that the chairman of the advisory panel had worn a bulletproof vest. The Church of Scientology had condemned Prozac. A small study by two Harvard researchers had seemed to show that people on Prozac were prone to suicidal thoughts, and patients and their families had sued Eli Lilly and Company, the manufacturer of the drug. In 1989 a Kentucky printing press operator named Joseph Wesbecker had killed eight co-workers plus himself with an assault rifle and wounded a dozen others a few weeks after he started taking Prozac. The FDA panel back then recommended further research. Still, the FDA had decided that the drugs were beneficial and safe for most people, based on the weight of scientific studies, and should stay on the market. 

For patients under 18, there was the added concern about how these powerful chemicals might affect brains that were still developing. Children’s brain chemistry is different from that of adults. So even if the drugs were completely safe for adults and helped ease their depression, that did not mean they were necessarily safe or helpful for children. Only Prozac had ever been officially authorized as an antidepressant for this age group. Studies on the other drugs (most of them belonging to a class known as selective serotonin reuptake inhibitors, or SSRIs) had not clearly shown that they worked significantly better than a placebo, or fake drug. 

Nevertheless, doctors could legally prescribe any of the medications for any age, and they did: The usage rate for children under 18 jumped more than threefold from the early 1990s to 2001, according to a study by Washington State University; the FDA reported that almost 11 million prescriptions for that age group were written in 2002. 

If there was no sure proof that the SSRIs were effective for youths, neither had any clinical trials on patients clearly and definitively demonstrated that the medications increased the risk of suicide—or at least, that was what the medical community believed. The companies that produced the drugs, anxious not to lose this rich market, insisted that the families’ stories were only anecdotal—though heartbreaking—aberrations. What made things even more difficult to sort out was that the patients taking the pills were unhappy to begin with, by definition, and might have tended toward suicide with or without the medications. It was also hard to define what to consider a “suicide attempt.” Slapping yourself on the head? Stabbing yourself with a pencil during an exam For that matter, even as the use of the antidepressants had been rising, the overall rate of teenage suicide in the United States had dropped in the late 1990s. So maybe the pills were actually helping to reduce the number of suicides. 

The FDA had issued a warning specifically about Paxil in June 2003 after the drug’s manufacturer, GlaxoSmithKline, submitted studies that showed a higher level of what might be suicidal thoughts and incidents among adolescents and younger children taking that drug, compared with patients taking a placebo. (Most of the data about Paxil was not made public, and the New York state attorney general, Eliot Spitzer, sued GlaxoSmithKline a year later for withholding the trial results.) In October came a stronger FDA warning about the whole group of antidepressants. The warnings did not forbid doctors from using these medicines, however. There still seemed to be no definitive proof, either that the drugs led to an increased risk of suicide, or that any drug but Prozac worked in youngsters. The FDA commissioned Columbia University to conduct yet another study. Meanwhile, in December, the British equivalent of the FDA took a stronger step, warning doctors in the United Kingdom to shun all antidepressants but Prozac for children. 

Most of the speakers at the Holiday Inn called for stricter labels on the drugs, and some urged that only trained specialists, not generalists or pediatricians, should be allowed to prescribe them. Some demanded an outright ban. They wanted the FDA to protect their children. Yet many of them were skeptical that the regulators would. 

Dawn Rider exuded an air of competence and confidence; she was a tall woman with a bright red jacket and long, thick, dark hair. Her 14-year old son had died after taking Prozac. Then her husband was given Paxil to help him cope with the death, and his attempt to withdraw from that drug destroyed their marriage, she told the crowd in the ballroom. During the lunch break, I asked her what she hoped the FDA would do. 

“I don’t have a lot of faith in the FDA,” she replied. “There’s too much sway from the pharmaceutical industry.” She pointed particularly to the fact that Mitchell E. Daniels Jr., a former Lilly executive, had been the White House budget director and was running for the Republican nomination for governor of Indiana. (He would later be elected.) And somehow it was only Lilly’s drug Prozac that had been approved for children. “I was sitting there, watching them [on the FDA panel] today. I almost noticed bored expressions.” 

“It’s clear that the FDA is a political entity,” Tom Woodward told the three dozen panelists. “Under the Bush administration, the FDA is putting the drug industry over the interests of the public.” The FDA? The Food and Drug Administration, the agency that was created in 1906 to make sure that Americans were never again poisoned en masse the way Upton Sinclair described in his novel The Jungle? That poll after poll has always shown is one of the most trusted arms of the entire government?

For almost a century, the FDA has been the Good Housekeeping seal of approval, the Nobel Prize, and Ivory soap (99 and 44⁄100 percent pure) combined. No medicine or medical device can be sold in the United States unless the FDA pronounces that it is safe and that it works. No packaged food can make health claims unless its label is approved by the FDA. Americans count on this agency to make sure that we have a steady stream of wonderful new pills that are potent and perfectly safe at the same time, as well as a supermarket full of goodies that we can gobble up without worrying about food poisoning. We also count on this government agency to be on our side against powerful drug and food companies and to resist political pressure. We trust the FDA so that we do not have to stop and read the label of every can of soup and bottle of aspirin we buy. In fact, we pretty much assume that it will protect us from everything short of nuclear war. 

Undoubtedly, most Americans do not completely understand how this influential government office works. We probably overstate its clout in some categories, like restaurants, and don’t realize how far its power extends into other areas, like microwave ovens and pet food. Some people think it tests every drug that is sold, and or that it inspects all food products. (Neither of these is true.) Still, we know the basics: If the FDA lets us down, we are not just personally disappointed, betrayed, and angry. We could be dead. 

To say you have lost faith in the FDA is like saying motherhood and apple pie have gone rotten—literally, in fact, since the FDA is supposed to ensure that apple pie is safe to eat if you buy it prepackaged from the supermarket. (Not if you eat it in a restaurant, however.) 

So how could this mighty agency that we have relied on for a century mess up so badly? Why didn’t it catch the suicide problem before it ever approved the first SSRI? How can it be legal for doctors to give teenagers drugs that the FDA never approved for kids? Why didn’t the FDA know about the GlaxoSmithKline studies? The parents who came to the Holiday Inn had once trusted the FDA to keep their children safe. And it had failed them. 

Before I started covering health care as a reporter and editor at Institutional Investor magazine in the early 1990s, I probably had more or less the same vague knowledge of the FDA that most Americans do. Luckily, I never had much reason to be concerned with the products it oversees. I come from a healthy, long-lived family, and my husband, my kids, and I have rarely needed a prescription except for the occasional antibiotic. Nor have I had to be a caretaker for my parents or other aging relatives who do take a lot of medication. As for the “F” in FDA, well, I’ve always worried more about the calories in my food than any contaminants. 

Once I began writing about the pharmaceutical industry and health insurance, I got to meet the FDA that the drug companies know. To these companies, it is the all-powerful, arbitrary, nitpicky naysayer that keeps their desperately needed medicines off the market until they run a zillion unnecessary tests to prove things they already proved. The agency is unreliable, one week saying it wants to help manufacturers get their products out to patients quickly, then the next week panicking after too many reports of dangerous side effects. It is mysterious; there is no way of knowing just what a company must do to move its product past the regulatory box-checkers. At best, the FDA is a bunch of bureaucrats who mean well but are scared to be the first to approve something new. Most of all, the agency must be obeyed. It is almost impossible to get through a 10-minute interview with a pharmaceutical executive without hearing at least one complaint or fear about the FDA. 

Of course that is a one-sided view, and the other side can overwhelm you as soon as you walk into an advisory committee hearing, such as the one at the Holiday Inn. There were so many stories at that hearing that I stopped taking notes. It was too much suffering, too many horrible new examples, one after another, without enough time to absorb the horror of the first ones. And it was painful to be there, to picture my own kids’ faces—to be too lucky. The drug companies were wrong; the problem was not that the FDA was keeping good medicines off the market in order to enforce overly stringent rules. The problem was clearly that the FDA had given in too easily to the drug companies’ pressure, had skimped on its due diligence, and had let dangerous products into the marketplace. 

I wondered how it felt to be one of the FDA staffers listening to those stories at the Holiday Inn, knowing that maybe something you had done had caused a family so much pain. A few weeks after the hearing, I asked Dr. Robert J. Temple that question. Heavyset and a bit shorter than average, with longish salt-and-pepper hair that flips over his collar, a thick mustache, round eyeglasses, and thick dark eyebrows, Temple is the FDA’s associate director of medical policy and its resident expert on clinical drug trials. He started working at the agency in 1972, just as it was in the midst of laying out the scientific processes that would define modern drug testing, and he has been in the midst of it ever since. In his job capacity, Temple was one of the three dozen people at the U-shaped table, though he was not a member of the advisory panels and could not vote on any recommendations. He gave a short laugh at my question. “They’re very moving stories,” he replied calmly. “The fundamental problem,” and he leaned forward as though to share a secret, “is you don’t know whether in fact their attribution is correct. Long before there were antidepressants, people committed homicides and suicides. It’s well known that depression is a cause of suicide.” 

In other words, yes, the families’ tales were sad, but heartbreak is not scientific proof. Just because someone takes Pill A and then commits Act B, that does not prove that A caused B. What else was happening in the person’s life that could have led to Act B? What had other people done when they were taking Pill A? The FDA could not base its decisions on emotion. First and foremost, before worrying about drug companies’ profits, before even worrying about consumers’ anxieties or medical needs, the FDA had to look at the science. 

Maybe. But as a reporter at newspapers in California and New Jersey over the years and as the former political reporter for Institutional Investor, I have spent enough time covering politics at the local, state, and federal levels to know that the FDA’s decisions could not always be purely scientific. The FDA is a government agency. Its commissioner is appointed by the president. Its budget and commissioner have to be approved by Congress. Its officials can be hauled before a congressional committee for interrogation at any time. Its major decisions are usually vetted by the Department of Health and Human Services, if not the White House. On top of all that, the FDA regulates the industry—pharmaceuticals—with the most powerful lobbying force in Washington, D.C. Of course all those players try to influence FDA decisions on issues they care about, and of course, the FDA gives in when the pressure is too great. If there are three hundred parents whose children become violent after taking drugs like Paxil and those three hundred parents shout loud enough, Congress, the White House, the pharmaceutical industry, and the FDA will hear. Marion Goff of Rhode Island, one of the parents at the Holiday Inn, knew exactly what she was doing when she brought a friend to the hearing—Stephanie Chafee, wife of Republican senator Lincoln Chafee. 

Chafee stood nearby, silently, while Goff told the FDA experts how one of her twin daughters, then age nine, had taken Zoloft and Paxil. Goff had once found the girl on the window ledge, with one leg already out the window. The girl had also tried to stab herself repeatedly. 

And there is a lot more to the FDA jigsaw puzzle. Now that I was covering health care, I naturally began noticing constant references to the FDA in the news, even in the most unlikely articles. The agency warned pregnant women against using sophisticated ultrasound equipment to take “souvenir” pictures of embryos. Blood banks complained that the FDA was making them ask too many questions of potential donors, about AIDS, West Nile disease, and SARS. A factory in China, certified by the FDA to manufacture ingredients for various medicines, was dumping untreated chemical waste. Cell phone users wanted the FDA to find out if their phones might cause brain cancer. Was there anything the agency didn’t regulate? Indeed, it seemed to have its finger in many of the most controversial and important pies at the American supper table: genetic engineering of plants and animals, abortion, mad cow disease, obesity, drug prices, cloning, Baby Boomer vanity drugs, illegal steroid use by athletes, drug ads on TV. 

How could I fit something this huge into a single book? 

As it turned out, perhaps the grieving parents at the Holiday Inn should not have been so cynical. At the conclusion of the hearing that afternoon, the two scientific advisory committees recommended that the FDA immediately issue stronger warnings to doctors about the risks to children, without waiting for Columbia University to complete its analysis. In its official decision a month and a half later, the agency went even further. First, it asked the manufacturers themselves to place warnings right on package labels, which were more likely to be seen by doctors and patients on an ongoing basis. It also put out a health advisory to physicians and other caregivers to “closely monitor all patients being placed on therapy with these drugs for worsening depression and suicidal thinking,” especially at the beginning of treatment—all patients, not just children. 

This was pretty impressive. The FDA really listens to ordinary people, acts fast, and bucks the big drug companies. The New York Times claimed the new warnings were a break with the FDA’s normal, more cautious procedures, because there was no clear-cut evidence of danger from “well-controlled” human trials. 

But then several newspapers reported that, in fact, even before the hearing at the Holiday Inn, the FDA did have such evidence—and kept it hidden. In studying data from more than 4,000 participants in clinical trials, an FDA drug safety analyst, Dr. Andrew D. Mosholder, said he found that children on antidepressants were almost twice as likely to become suicidal as those on placebos. The agency refused to let him testify about his findings at the hearing and never showed the panelists his report, however. With the incident hitting newspaper headlines across the country, the chairman of the Senate Finance Committee, Charles Grassley of Iowa, launched an investigation that came up with further charges of FDA manipulation. “You don’t just ask someone to clam up,” the senator told the Wall Street Journal. “If there’s any doubt, they ought to put out the caution to the public at large.” All that was on top of the Paxil trial results that GlaxoSmithKline and the FDA had kept from the public. 

So had the FDA actually tilted in favor of the pharmaceutical companies by squelching reports critical of their drugs, even while it seemed to be listening to the patients? 

Well, that was not necessarily the case, either. Bob Temple, the expert on clinical trials, told reporters that Mosholder’s report was “premature” because too much of the underlying data was unreliable—for instance, some of the supposed examples of suicide attempts were vague and might not have been real attempts. He said the FDA did not want to spread unsubstantiated fears, thereby discouraging severely depressed people from getting treatment that might help them. And FDA officials claimed the law did not allow them to reveal GlaxoSmithKline’s proprietary trial results. Even before I had a chance to ask, Dr. Steven Galson, the acting head of the FDA’s Center for Drug Evaluation and Research, insisted in an interview with me that “stories that we’re somehow suppressing people, that’s the farthest from the truth.” 

Later that summer, the Columbia University report did back up Mosholder’s findings, but only after digging into the data more deeply. Finally, another meeting of outside experts in September called for yet stricter warning labels, and the FDA officialdom agreed to implement those changes. In fact, the agency said it would even go back and reanalyze its data on adult suicidal behavior. Temple admitted that all the clinical trials, taken together, seemed to show “an increase in suicidal thinking and action.” 

At a hearing soon afterwards, members of Congress from both parties pounded on the FDA for hiding Mosholder’s report and other information. “The FDA’s lack of cooperation,” declared Congressman Joe L. Barton of Texas, “leaves me wondering whether this is sheer ineptitude or something far worse.” “No agency charged with the public health should have behaved with such indifference to the public safety as is evidenced in this case,” intoned Congressman Peter Deutch of Florida. The House and Senate both launched investigations.[Investigations? Congress has NEVER investigated anything on behalf of the American people,always are they for cover DC] 

Two more possibilities, then. Maybe the brouhaha over the Mosholder report proved that the FDA truly operates the way Temple described it, as an ivory tower of pure science. It is so careful and so insistently scientific that, even under tremendous pressure from consumers, the press, and politicians, it will not issue half-baked announcements until it has all the facts. And if new data cast doubt on its previous findings, it is so scientifically pure that, rather than stonewall, it will pore through all of its research yet again. 

Or maybe, like any institution, it just tried to cover up its own mistakes. Protector of the consumer? Pawn of industry? Pure scientists? Political plaything? 

Now I really needed to write this book. I had to put all the jigsaw pieces together and decide what the FDA is—this sprawling, scientific, political, nitpicky, pioneering, admired, feared, detested, trusted agency.

CHAPTER 1 

Case Study: Chasing Cancer 

Garo Armen, Russ Herndon, Pramod Srivastava, and Renu Gupta started practicing at nine in the morning on the day after Labor Day, 2003. They gathered in a small, green-carpeted conference room just off the seven-floor atrium of the DoubleTree hotel in Rockville, Maryland, half an hour outside Washington, D.C. Across from their room, bathed in the atrium’s soft yellow light, three small waterfalls trickled down an indoor stone wall. 

Okay, what would the reviewers from the Food and Drug Administration be likely to ask when they met that afternoon? 

The four of them worked for a New York City company called Antigenics Inc. one of countless new, small firms trying to use a niche of biotechnology to tackle cancer. Srivastava and Gupta, both born in India and deeply interested in philosophy, were the scientists. Herndon was the businessman, outgoing and boyish-looking. Armen was pretty much everything: CEO, co-founder, fundraiser, public spokesman, elder statesman, and driving force. 

Antigenics’ particular approach was based on work that Srivastava had begun as a graduate student 25 years earlier at the Centre for Cellular and Molecular Biology in Hyderabad, India. That work focused on a kind of protein known as heat-shock proteins, or stress proteins, which are found in all cells of all living organisms, including cancer cells. Under normal 1 circumstances, these proteins play a major role in transporting another kind of protein called antigens within a cell (and thus they have an even more colorful nickname, chaperones). Antigens, for their part, stimulate the body’s immune system to respond to infection or disease. In theory, you could extract and purify the heat-shock proteins that had chaperoned an antigen that stimulated a response to a certain cancer. Then the extracted heat-shock proteins could be made into a vaccine that would contain some trace of that specific antigen and its cancer—the “antigenic fingerprint” of that cancer. If a patient got that vaccine, unique to his or her cancer, the immune system might be reprogrammed to home in on cancer cells bearing the antigenic fingerprint. It would not prevent anyone from getting cancer, but it could stop the cancer from spreading. 

That was the theory, anyway. A number of universities and research institutes in the United States and Europe were also studying the heat-shock protein process, and so far the buzz about Antigenics among scientists and on Wall Street was cautiously positive. The vaccine, which was called Oncophage, had already proved itself in animal experiments, in tests for safety, and even in the first stage of clinical trials on cancer patients. A trial of colorectal cancer patients had just reported some good news about survival rates. Now 650 people with kidney cancer and 350 with skin cancer were participating in further tests at more than 130 sites around the world. 

As soon as doctors removed a patient’s tumor, the specimen was frozen in dry ice and rushed to Antigenics’ labs in Woburn and Lexington, Massachusetts, both near Boston. There, scientists had 24 hours in which to extract the heat-shock proteins—they needed a minimum of seven grams of tumor—and process them into a vaccine. For the next three weeks the vaccines were tested for purity, sterility, and composition. Finally, at least four vials were flown back to each patient and injected—one a week for four weeks, then biweekly. The stuff looked like a small glass of Sprite. 

Of course, these were only tests. Oncophage was still far from being a safe, workable drug, let alone a cure for cancer. The Antigenics scientists figured they would need at least two more years of clinical testing, checking to see if the cancer had spread, before they would be ready to seek official FDA approval. So there really wasn’t much reason to be hanging out at the FDA’s headquarters in Rockville. 

But Antigenics had requested this special meeting because a problem had cropped up. The FDA had recently reorganized. Some 200 reviewers who specialized in protein-based drugs, including staffers who had been working with Antigenics for almost a decade, were about to be shifted to a different branch of the agency. That meant that a whole new crew of scientists would be taking over the review of Oncophage—scientists who did not know Antigenics’ people, its drug, or its history. 

What made the situation even dicier was that Antigenics wasn’t exactly following standard operating procedure. Over the past several years, the company had been negotiating off and on with the FDA in hopes of convincing the regulators that its drug was unique and should be able to bypass some of the normal requirements for quality control. Antigenics was hardly alone; biotech firms right and left were flooding the FDA with revolutionary science, demanding exemptions and challenging traditional testing standards. 

For instance, in order to make sure that volunteers in experimental drug trials—and, ultimately, patients in the general population—are not swallowing something dangerous, the FDA obviously needs data from the manufacturer about the potency and safety of the drug being tested. But the agency also goes a step further, asking manufacturers to explain how they will test their drugs to obtain the potency and safety data. The idea is to reassure doctors that the drug they are prescribing is consistent bottle after bottle and that the method of measuring is accurate. So the manufacturers have to provide details about the tests they use to check a drug—known as assays—even before a human subject can swallow the first pill or be injected with the first dose. 

With a traditional chemical drug, measuring is fairly routine. However, vaccines are much more variable because they are made from living material, which is inherently inconsistent. And vaccines made to order from the patient’s own tumor are even more variable, a totally new creature for the FDA. “It’s not straightforward, because it’s a personalized cancer vaccine,” Dr. Elma S. Hawkins, a veteran of the industry, explained to me a couple of months after the meeting in Rockville, when she was Antigenics’ vice chairman. Garo Armen had been talking with Dr. Philip Noguchi, acting director of the FDA’s Office of Cellular, Tissue, and Gene Therapies, to get advice on developing the assays. 

Another problem, Hawkins said, is that everything just happened too quickly. Since there are only about 35,000 people in the United States with kidney cancer, Antigenics had been told it would take ten years to recruit its goal of 650 patients. Instead, it filled its ranks in less than three years. “We accrued patients fast into a trial that everybody said was impossible to do. The clinical trial went at lightning speed.” But the paperwork of collecting forms from each trial site did not go as speedily. “Not everything was documented at the FDA the way they would like it to be,” Hawkins said. 

So Antigenics had neither collected all the data that it was supposed to, nor given the FDA the explanation of its assays. Now the new FDA reviewers had sent Antigenics a letter asking for some of that missing information. 

A little before two o’clock, the Antigenics crew headed past the Twin-brook Metro station, some three blocks to the FDA headquarters. The 18-story, dark brown-and-grey monolith stands out in its spare, suburban Maryland neighborhood mainly because of its ugliness and bulk. Row after row of windows and steel look down onto a gently sloping hill marked with scattered stands of skinny trees. In front, the building crams almost right up against the street, with room for just two wooden benches, seven large concrete planters—the kind that are built for security, not beauty—and a single bike rack. Across the street sits a strip mall with a video store, a surplus furniture outlet, and a mailing service. 

As soon as the group from Antigenics got to the meeting, Armen could tell there was a bigger problem than they had practiced for. “When I saw the body language, I knew something was going on,” he recalled later. “I tried to soften them. That backfired. I tried to tell them about the fact that we were doing this because it was supported by an enormous amount of science and that we were doing it because there was a terrific unmet need. They didn’t even look at me.” 

Antigenics couldn’t document how it would test the safety of the drug that it was putting into its subjects? 

Then the FDA, in good conscience, could not allow any more people to be placed at risk. 

As of that moment, the kidney trial was placed on partial clinical hold. No new patients would be permitted to try the vaccine. 

Dr. Garo H. Armen is short and trim, with thinning hair, a light brown goatee generously flecked with grey, and what seems a perpetual small smile of confidence. “Never ever in the last ten years”—the lifetime of Antigenics—“did I ever think about giving up,” he insisted, eight weeks after the clinical hold was issued. 

He was born in 1953 to an Armenian family in Istanbul, Turkey, which meant that his forebears had somehow survived the massacres and mass deportations of Armenians in the Ottoman Empire during the late nineteenth and early twentieth centuries. His father, an auto parts dealer, sent him to the United States in 1970 because the 17-year-old was getting a little too outspoken about Armenian independence. Armen headed for New York City, to the semi-suburban borough of Queens, where he had some distant relatives. Besides, the local public university, Queens College, charged only $200 tuition and offered an English course for students who did not speak the language. Armen blended easily into the borough’s ethnic stew of Italians, Irish, Jews, Greeks, blacks, Poles, and Puerto Ricans. 

Because he was interested in science, Armen studied chemistry at Queens College and earned a PhD in physical chemistry at City University of New York in 1979. At Brookhaven National Laboratories in nearby Long Island, he did research on photosynthesis and energy production. But by then Armen had discovered the thrill of the stock market. 

1981 he took his science background to Wall Street and became a stock analyst specializing in chemicals at E. F. Hutton & Company. Five years later he moved to Dean Witter Reynolds as a senior vice president of research with a specialty in chemical and pharmaceutical companies. (Biotech firms like Antigenics may have a reputation for self-destructing after short-lived bursts of glory, but so far it is Armen’s two Wall Street alma maters that have disappeared. Hutton was acquired by Shearson Lehman Brothers in 1988, and the Dean Witter name was erased in 2001, four years after the company merged with Morgan Stanley Group Inc.) 

Next leap: In 1990 Armen opened his own money management firm, Armen Partners. Instead of just analyzing stocks for others to buy, he did the buying and selling himself, taking a cut of 20 percent of any profits he made. At its peak, Armen Partners was handling $75 million of Armen’s own money plus that of select wealthy individuals. His specialty was biotechnology companies. 

Naturally, he got a lot of hot tips about the newest cures for cancer or obesity. “Most of them didn’t turn out to be anything,” Armen recalled. A few did, however. He made his name launching a cancer business for Immunex Lederle There was also an Irish company named Elan Corporation that was working on an intriguing approach to Alzheimer’s disease. Then, on June 15, 1993—Armen is very precise about this—a scientist named Dr. Pramod K. Srivastava showed up with an idea about how heat-shock proteins could be purified and made into a vaccine for cancer. Another hot tip. But this one seemed more promising than most. 

Like Armen, Srivastava was an immigrant with a passion for science. His background was about as elite as it gets in India: He came from the northern city of Allahabad, one of the most important places in both Hindu mythology and modern Indian history, and from a relatively high-ranking caste of professionals in the Hindu hierarchy. His father was a civil servant and retired Army officer. There is, moreover, hardly a scientific discipline or foreign language that Srivastava hasn’t studied. He has a bachelor’s degree in biology and chemistry, a master’s in botany, a PhD in biochemistry, and at age 47 he enrolled in medical school at the University of Connecticut (where he also ran the Center for Immunotherapy of Cancer and Infections Diseases). Having earned his degrees on three continents, Srivastava has at least a working knowledge of Bengali, English, French, German, Hindi, Japanese, and Urdu. 

At graduate school in Hyderabad in the early 1980s, Srivastava more or less stumbled into cancer research after a friend showed him a cancer cell in a lab. “I couldn’t get over how weird and strange the cancer cells looked, how different from the normal cells,” he later told an interviewer. Scientists had already managed to vaccinate mice against cancer by injecting them with weakened tumor cells, so Srivastava broke that process down to the next level. Using a centrifuge, he separated the tumor cells into various components, then tried vaccinating mice with different sample parts. The one that worked, he found, was the heat-shock protein. However, as he kept experimenting, he realized that the heat-shock proteins had to be bound to short pieces of other proteins called peptides. Then Srivastava put aside his research for a few years to come to the United States for a postdoctoral fellowship in genetics at Yale University. 

After their first meeting in New York, Armen and Srivastava continued to talk periodically for ten months. “Every time we peeled a layer,” Armen said, “it looked better and better.” Armen also had a personal reason for his interest, because his mother had had breast cancer. Although it seemed to go into remission, she died of a stroke when he was 19. 

Finally, in 1994, Armen decided to junk Wall Street, essentially close up his money management firm, and leap to a new career once again. He and Srivastava formed Antigenics to commercialize the heat-shock protein idea. Armen contributed $250,000 of his own money and raised $150,000 from private investors such as a former Dean Witter analyst and the founder of the hedge fund Oracle Investment Management in Greenwich, Connecticut. (He did not tap the investors in Armen Partners because “I thought that would be unethical. This was a very, very early stage development,” far riskier than the kinds of investments his firm typically made for its clients. Ten years later, Armen claimed, those same investors pounced on him for keeping them out of such a good deal. “You can’t win,” he sighed.) The new firm rented a small office on the ninth floor of one of the most famous landmarks in New York, the art deco Rockefeller Center complex on Fifth Avenue. Srivastava and about eight other scientists continued to work in his lab at Fordham University several miles north in the Bronx. 

Armen and Srivastava decided to start with pancreatic cancer, kidney cancer, and a kind of skin cancer known as melanoma. There were a couple of reasons for this approach: People with those particular diseases have few alternative treatments. Also, Antigenics would need a tumor big enough to provide seven grams for processing, and not all varieties of tumors are that large. But Garo Armen had no intention of limiting himself to kidneys, pancreases, and skin. The company’s methodology—its platform, in scientific jargon—could work for all cancers, he believed. In fact, he told me, because it is based on the immune system, the Antigenics approach could have applications for neurological diseases, cardiovascular disease, infectious diseases, and conditions associated with aging. “If we execute well, we have the technology to become the Microsoft of this industry—that level of dominance. We believe that we are the masters of the immune system.” 

No, he didn’t just mean the Microsoft of cancer. He meant the Microsoft of all biotechnology. 

Russell H. Herndon had just finished a speech at a meeting of the Biotechnology Industry Organization, the main trade group for biotech firms, when Garo Armen and Pramod Srivastava came up to him one day in 1994. As the vice president of regulatory affairs at Genzyme Corporation—a relatively big and established company, for a biotech—Herndon handled paperwork and conversations with the Food and Drug Administration. Among other things, he had dealt with the regulators on a type of cell therapy based on the principal of using the patient’s own body to heal itself. An easy conversationalist, with light hair and round, brown eyes, Herndon had earned a bachelor’s degree in biology, taken courses at Harvard Business School, and worked for an eclectic collection of other small firms before Genzyme. 

For their part, the pair from Antigenics knew the science behind their heat-shock proteins, and they knew the business world. They had plans for moving ahead on their research, raising more money, possibly partnering with a big pharmaceutical company, and marketing their vaccine. But they had no idea how to approach the government, how to get the approvals they would need to test their drug in humans, or even what approvals were required. 

“We’ve just formed this company, and we would love to get your advice as to who we should talk to at the FDA, and what sorts of questions they might ask,” they said to Herndon. “What would the product be classified as? What are some of the problems we might encounter?” It was the beginning of Garo Armen’s crash course in the FDA. 

There are four basic steps that any company must take in moving a potential drug from lab to market in the United States: tests on animals (known as preclinical trials), trials on a small group of healthy volunteers to ensure that the drug is safe (known as Phase I clinical trials), then two progressively larger trials on people with the disease to test both safety and effectiveness (known as Phase II and III clinical trials). Animal trials are not regulated by the FDA, but in order to test anything on humans, a drug company must submit what is called an investigational new drug application, or IND—a huge document that summarizes the animal test results, explains the manufacturing process, and outlines the human testing plans in detail. Then, after Phase III, the company files an application to actually start selling the drug. For vaccines like Oncophage, the filing is called a biologics license application, or BLA; for chemical-based drugs, it is a new drug application, or NDA. (There is more on this process in Chapter 5.) 

By 1995 Armen figured he was ready to seek the FDA’s go-ahead to start Phase I testing on humans, and he came up with what he thought was the brilliant idea of having the IND prepared by the medical staff at Memorial Sloan-Kettering Cancer Center in New York, where Srivastava had worked following his stint at Yale. The hospital had experience handling the FDA’s red tape, after all. But that plan did not last long. As Armen put it, in his typical blend of European formality and ironic self-awareness, “After a few months I came to the realization that the movement at Sloan-Kettering was at such a snail’s pace that my temperament didn’t allow me to put up with it.” Antigenics would have to file its own application with the FDA. 

So Armen contacted Mark Boulding, a partner with the Washington, D.C. firm of Fox, Bennett, and Turner who, he had been told, specialized in regulatory law. Boulding met him at Srivastava’s Fordham labs. Armen quickly learned that the FDA’s safety requirements are levels above what he had been accustomed to. For instance, he could not simply hand in his animal toxicity test results. “Our experience with animals suggested no toxicity. That didn’t matter. We had to run [separate] toxicology testing” on human subjects in Phase I. Moreover, the manufacturing process that Antigenics had been using to churn out samples for research purposes at Fordham would have to be upgraded. “We had to have a certain quality of air; certain parts of the manufacturing had to be segmented to guard against cross-contamination. 

“All of a sudden, lightning went over my head,” Armen recalled. “The FDA was not a trivial thing. I said, ‘Holy Moses, we have to go and bring in some talent that really understands this stuff.’” But he said he was not upset at the work that lay ahead. “I was happy, because I now knew what needs to be done.” 

Throughout this time Armen had kept in contact with Russ Herndon, calling him every six months or so with more questions. As far as Herndon was concerned, Antigenics had a fascinating scientific theory that probably was not as good in reality as it sounded. Meanwhile, one of his colleagues at Genzyme, Elma Hawkins, was analyzing Antigenics for another reason. It was her job to scout out possible acquisitions for Genzyme. And she wanted Antigenics. 

Hawkins is a native of South Africa who retains a slight accent and a calendar from that country on her office wall. In a peripatetic career, she spent her adolescence in London, including two years studying at the Royal Ballet, then headed back to Pretoria for college. A professor on sabbatical from the University of Alabama recruited her for her PhD in medicinal chemistry, with a specialization in cancer. From there Hawkins went to Warner-Lambert Company in Michigan, where she shepherded three widely varying drugs through the FDA, and to Boston, where she did research at Tufts University and grew skin at a biotech. After orchestrating Genzyme’s purchase of that biotech, Hawkins created the new company’s molecular oncology unit. She is short and sturdy, with rimless oval glasses and wavy, red-brown hair that brushes her shoulders.

“I was so impressed by the thoroughness of the science. It was a really different way of approaching the treatment of cancer,” she said of her first view of Antigenics. But unless the company could get the FDA to approve its plans for Phases I, II, and III, it was not worth buying. So Hawkins offered to help teach Armen the ropes of working with the regulators. The next thing she knew, she was commuting from her home in Boston to Antigenics’ New York headquarters every weekend—without pay— from February through July of 1996. “I didn’t think he’d take me up on that offer to that degree,” she admitted a little ruefully. 

That spring, Armen also spoke for an hour on the phone with Phil Noguchi, the FDA official. “I said we want to file an IND, but we want to have a pre-IND meeting—I didn’t know what the hell those terms are.” 

A pre-IND meeting occurs when people from a drug company meet with the FDA staff to discuss the specifics of their plans for human trials. The idea is that if the company can find out what the FDA wants in the IND beforehand, then it should be easier to write an application that will be approved. This can be particularly important for a fledgling biotech like Antigenics. It’s probably the scientists’ first application, so they need to understand the process. And their cutting-edge technology may be new to the FDA. As Armen described things, “I thought it was critical for us to sit down with the agency and explain to them the nuances of our technology. The world’s first personalized protein therapeutic—there’s no regulation that governs a personalized protein therapeutic. Unless we acquainted the agency with the kind of details that were critical, we would be at a disadvantage.” He cited one potential stumbling block: Drugs in Phase I are supposed to tested only on healthy volunteers. By definition, however, the Oncophage vaccine could not possibly be tried on healthy people, because it had to be made from the trial subjects’ own tumors, and therefore, Antigenics needed people with cancerous tumors. The FDA agreed to waive the Phase I rule. 

Noguchi, Armen’s main contact at the regulatory agency, is hardly a scary figure. Slightly built, his straight black hair speckled with grey, he speaks calmly and quietly, as careful as Armen is charismatic. He wears the sparkling “summer blue” uniform of a captain in the U.S. Public Health Service Commissioned Corps. (Many people at the FDA are members of this little-known branch of the national uniformed services, which was founded in 1798 to care for sick and injured merchant marines.) 

He, too, has an immigrant’s story, but this one goes back through two generations of racism. Noguchi tells it without any apparent bitterness. His grandparents came from Japan, took up farming near Sacramento, California, and then were interned with other Japanese-Americans during World War II. After the war, his father had to sweep floors for a while despite a degree in architecture and engineering from the University of California at Berkeley; his mother, a nurse, rode in the back of the bus with “coloreds” in Washington, D.C. 

By Phil Noguchi’s generation, at least, things had gotten better. He was born in Sacramento in 1949 and moved to Washington, D.C. for medical school at George Washington University. There he joined a Public Health Service summer internship program, working in the National Institutes of Health’s Division of Biological Standards; when the division was transferred to the FDA in 1972, he followed with it. Then Naguchi heard about a deal where the program would pay his tuition ($2,600 a year at the time) if he promised to spend two years doing research for the government. He signed because of the money. But he stayed, he said, because of the scientific camaraderie and respect. “Everyone’s opinion makes a difference. You’re a junior staff scientist, but if your research has been in a certain area of monoclonal antibodies, and you review a protocol, that factors into how the FDA reviews a particular program.” 

FDA rules prohibit Noguchi from talking specifically about Antigenics or its products, but he is certainly mindful that science is changing rapidly all around the six-floor outpost where the Oncophage reviewers work, some two miles from the main FDA offices. “When there are areas involving new techniques,” he explained, a couple of months after Antigenics was placed on the clinical hold, “what we strive to do is to be open about what we have requested and why. Each product is looked at individually. Together, we can brainstorm and come up with tests that are going to be meaningful. Almost anything that’s being done out there can be accommodated.” However, he also warned that when it comes to vaccines made from a patient’s cancer cells, which a number of companies besides Antigenics are working on, “the mechanism of action isn’t known. We have had a plethora of trials in which the results are highly inconsistent.” To Armen, the FDA as personified by Noguchi seemed flexible and reasonable. “Phil Noguchi said, ‘We will work with you. We’re not going to stop this product from being developed. We’re not going to ask you to do something that is scientifically impossible.’” 

In order to prepare for the pre-IND meeting, Armen, Hawkins, Srivastava, and a few other scientists from Antigenics created a slideshow of over 40 slides, outlining the rationale and the science behind their vaccine, and also detailing the animal toxicity data. It was not difficult, according to Armen: Each person worked on the preparations for two to three days a week over several weeks, “pulling data, analyzing, tabulating so it was compatible with the agency’s standards.” The meeting itself lasted two hours and was relatively low-key. Armen was pleasantly surprised by the sophisticated level of queries from the ten or so FDA staffers. “What impressed me was their questions about taking it to the next level, and the next.” The bottom line: “They said go ahead, prepare the IND.” 

The only problem was, Genzyme had decided not to acquire Antigenics after all. At around $75 million, Elma Hawkins said, the price tag was just too high, considering that “all there was, was Garo and Pramod and a bunch of patents.” But that did not signal the end of Hawkins’ relationship with the company. “Garo in his very persuasive way told me, ‘On Monday you’ll come and work for me.’” 

Armen had set a goal of getting the IND approved by the end of 1996. Since the FDA has 30 days in which to make its decision—if the agency does not reject the submission within that time, it is automatically approved—that essentially meant filing the massive application by around Thanksgiving. The filing would run 1,600 pages in three volumes, with six copies of each. 

“Elma put in 24/7 for a month and a half,” Armen said. As the self-imposed deadline neared, about five other staffers joined her for allnighters. “At three in the morning we’re all standing there, punching holes, assembling volumes and volumes of work,” Hawkins recalled. “I read and reread everything.” Spouses were lassoed to come in and proofread. The firm’s five printers spit out copy after copy. The group ordered in pizza and Chinese food and took turns grabbing naps on the floor and the couch in Armen’s office; amazingly, in all the frenzy, no tomato or soy sauce spilled on the pages. “There were nights I never slept. All I was seeing was the deadline,” said Hawkins. Luckily her old ballet training had taught her how to focus with little sleep. By the last night, Armen said, “eleven of us were working until two-thirty in the morning.” Driving home on the Long Island Expressway, “it was the first time ever that I fell asleep at the wheel”—just for a few seconds. 
26s
But they made it. They filed around Thanksgiving and got the go-ahead by late December. A year afterwards, Hawkins went back to reread that IND. “I found maybe one spelling mistake.” 

The first Phase I trial began in November 1997 with pancreatic cancer patients at Memorial Sloan-Kettering—patients who were given perhaps a year more to live. In other words, they might be spending the last year of their lives helping science. And they didn’t have much to lose.

Ten people ultimately joined the trial, but it was hard to find appropriate subjects. “I’d hear about the patient in the morning and grab my little box of Styrofoam and dry ice,” Hawkins recalled. She would wait in the operating room while doctors removed the tumor, hovering, trying to make sure they saved seven grams for her after doing their biopsy. (If the cancer had spread to the liver, the doctor simply sewed the patient back up; the patient would not live long enough to participate effectively in a trial.) Then Hawkins would grab her precious grams of tumor, pack them in the dry ice, and dash for the Delta Air Lines Shuttle—she was building up frequent flyer miles—to Antigenics’ Massachusetts lab for processing. She repeated this dash every week or two for several months. Then she stopped playing courier, and the samples were more officially delivered by Federal Express. 

Because the FDA, in approving the IND, had essentially approved Antigenics’ blueprint for all three phases of trials, Hawkins and her colleagues did not expect much contact with the regulators for the next few years. Every year they were required to submit a report detailing both the number of patients enrolled in their ongoing tests and the number who had experienced what is known as an adverse event—which means not just side effects from the drug, but any sort of medical complication. In addition, Antigenics would need to file a formal request for a new protocol before beginning any subsequent trial, describing what it intended to study, the names and background of the scientists (known as investigators) who would be conducting the study, the criteria that would be used to determine toxicity, and samples of the informed consent form that patients had signed. The new protocol request is typically about one-inch thick and takes about a month to compile, Hawkins said, but she implied that it is no big deal; it is organizational work that the company would want to do anyway, even if the FDA did not exist. After all, a company would certainly need to line up its investigators and make sure they were qualified, and patients by law have to give their informed consent. So, even while the test on pancreatic tumors at Memorial Sloan-Kettering was under way, Antigenics launched two more—a study of 36 people with melanoma and another involving 42 kidney cancer patients, both at the University of Texas M.D. Anderson Cancer Center in Houston. 

The bigger problem was that more trials demanded more money. In November 1999 Garo Armen went back to his Wall Street and pharmaceutical contacts for an infusion of $39 million. That still was not enough, and so in February 2000, with exquisite timing, Antigenics sold its first shares of stock to the public. This move raised $67.3 million. A month later, the stock market slide began. 

Antigenics plowed ahead. After six years of sporadic conversations, Armen lured Russ Herndon to join him full-time as chief operating officer, over lunch in 2001. (Later, the company would reorganize and Herndon would become president of commercial operations.) As the work expanded, Antigenics began constructing an 80,000-square-foot, two-story, corrugated-metal second home in an office park in Lexington, Massachusetts, a suburb of Boston best known for the battle that launched the American Revolution. By winter 2004, the facility would replace the Woburn labs and house most of Antigenics’ 220 employees, including all of its manufacturing work, with options to double the space and a lease that could be extended to 30 years. The company also moved its New York operations to larger quarters on the 21st and 22nd floors of another building at Rockefeller Center. Connected by an internal spiral staircase, the new digs combine elegant, dark, wood furniture and paneling with industrial blue-grey carpet and basic white walls and cubicles. Armen’s small executive suite, overlooking the Rockefeller Center skating rink, has a foldout bed for his next all-nighter. 

And the trials grew; in addition to the kidney, melanoma, pancreatic, and colorectal studies, the Oncophage vaccine was being tested on lung cancer and lymphoma, and Antigenics also was investigating a genital herpes vaccine and a vaccine for leukemia. Because the target populations for its kidney and melanoma uses are so small, Oncophage was granted so-called “orphan drug” status from the FDA, which provides tax credits and an extra stretch of marketing exclusivity to companies that make medicines for rare diseases. And because Oncophage is aimed at life-threatening conditions that have few other treatments, the FDA also put the drug on its “fast track” designation, which meant the agency would consider the application under somewhat more lenient criteria than normal. “You take advantage of everything you can,” Herndon told me. 

In the midst of all this work, Elan—the Irish company with the Alzheimer’s drug that Garo Armen had invested in back in his Wall Street days—blew up. It was by that point a global, $2 billion company, and Armen was on the board of directors. However, it suddenly halted the Alzheimer’s trials because patients were developing a potentially deadly brain inflammation. Moreover, investors and Wall Street analysts started raising questions about the company’s accounting methodology, a convoluted system that involved 55 separate joint ventures and partnerships and evoked all-too-recent memories of the collapse of Enron Corporation. In July 2002, after the stock had plunged 96 percent and the two top executives resigned, Armen was named chairman. Investors did not blame Armen or the other board members for the mess; indeed, Wall Street was impressed with the fact that Armen personally answered questions at his first press conference as chairman, rather than handing the burden off to Elan’s remaining managers. But the pressure to fix the company was intense. So now, on top of running Antigenics, Armen had to reassure Wall Street about Elan while finding ways to shed its assets, lay off workers, and do something about those joint ventures and partnerships. It took six months before a new chief executive was hired. 

Then the FDA reorganized the way it reviewed protein-based drugs. And the Antigenics crew set off for their meeting in Rockville. 

Armen said he asked just two questions after the FDA reviewers read aloud their prepared statement putting Oncophage on clinical hold. “What does this mean?” and “How am I going to explain this?” 

As he recalled, one of the FDA people told him, “I don’t care what you tell investors.” Armen replied: “I don’t mean investors. What am I going to tell doctors and patients who have no treatment options—‘I am going to take a potential hope away from you’?” In fact, the clinical hold did not mean that Antigenics would have to pull its treatment away from anyone. Patients already in the Phase III trial could continue. New patients could even be admitted to other trials. The only change was that no new patients could begin the Phase III kidney and melanoma trials. 

But it was also a fact that Armen did have to worry about investors. A ruling like this is what Wall Street calls a “material event”—something that could have a significant effect on a company’s finances. After all, if Antigenics never managed to supply enough information to get the hold lifted, the FDA would be unlikely to ever approve the drug. No drug, no revenue. The four Antigenics executives knew that they would have to make a public announcement by the next day, which would not be fun. 

Because everything must be precise when it comes to science, government regulations, and financial statements, Armen and his colleagues wanted a copy of the statement that the FDA staff had read aloud to them. That way, they could use the exact wording in their own public announcement. There would be no misunderstandings or misinterpretations. Sorry, the FDA reviewers told them, but the statement could not be given out until it had gone through the agency’s formal review process—certainly not today. 

In the end, the FDA staffers read the letter aloud again, slowly, several times, while the four from Antigenics scribbled it down like stenographers. Then Armen and his colleagues headed back to the DoubleTree hotel. 

“Holy crap,” Herndon remembered thinking. “We’ve got to figure out what it is they want and get it to them as fast as possible.” Armen later described his own reaction as “disappointed,” though not angry. Srivastava saw things differently. “My friend,” he said to Armen, “obviously the outcome of this meeting is not very heartening. But we will be better off a couple months down the road.” After all, Srivastava was sure Antigenics had the patient data the FDA wanted, just scattered at dozens of trial sites and not in the right format. And the company had been working on the assay details, planning to submit an explanation in a couple of months anyway. So all that was needed was to gather the data and speed things up a bit. Once that was done, Srivastava predicted, “we would no longer be one of a thousand companies out there. Our standing amongst the FDA would rise by the way we addressed the issue.” 

Noguchi, the FDA official, would not, of course, talk about how serious Antigenics’ particular situation was, but he pointed out in one of our interviews that a clinical hold can be issued for any range of reasons, from minor to major, from missing paperwork to fatalities. It could be that “you didn’t tell us in any detail who you are going to test, you didn’t tell us enough about how you are going to make the product.” It could be that the FDA does not think the investigators have sufficient credentials for this kind of work. A month after Antigenics was put on hold, the same thing happened to a New Jersey company called DOV Pharmaceutical, Inc., because the FDA wanted more safety information about its anti anxiety drug. When 18-year-old Jesse Gelsinger died during a clinical trial of an experimental gene therapy technique at the University of Pennsylvania in 1999, followed by two patients developing leukemia during a trial in France, the FDA threw a temporary hold on some trials using a similar type of gene therapy (which involves trying to insert a therapeutic gene into a patient’s cells, typically by having a benign virus deliver the gene). 

“It’s simply an administrative mechanism that we have. We have to have some reassurance about the quality of things. We’re dealing with human subjects, human safety, and the rights of human subjects. We tend to be conservative, and rightly so,” Noguchi explained. Besides—and this is where things can get tough for companies like Antigenics, that make their products out of biological material rather than synthesized chemicals—“you have to demonstrate that this is a product that you can make consistently, that you have some idea of the quality and potency. This is all to help the physician. We are not looking to have something that works sometimes but not all the time.”

When a clinical hold is announced, Noguchi added, “Wall Street may go completely bananas, or they may not do anything at all. It’s given a mystique by the investment community.” Back at the DoubleTree, the group from Antigenics rented a small conference room on the fifth floor for an hour to strategize. Their first priority was to draft a press release for investors plus a statement for the staff, both of which would have to be ready early the next day. They would also have to start working on a plan to collect the data that the FDA wanted. Armen, Gupta, and Srivastava then took the train back to New York while Herndon flew to Lexington. From the train, Armen called his vice president of corporate communications, Sunny Uberoi, to warn him to prepare for a marathon. In fact, Armen, Uberoi, and then chief financial officer Jeff D. Clark stayed at the Rockefeller Center headquarters until two in the morning, went home to grab a couple of hours sleep, then returned at 5:30 to get the press release out by 7 AM. 

Like any PR, the public statements tried to emphasize the good news— that the FDA had no concerns about the drug’s safety, that the other trials were not impacted. The three-paragraph press release took until the third sentence to mention the clinical hold. In standard, soothing corporatese, it quoted Armen promising that “we expect to provide the FDA with the required information within the next six to eight weeks.” In fact, Srivastava had thought the job could be done in four weeks, but the rest of the gang decided to give themselves some breathing room. 

The company held a conference call with investors at 9 AM, and two hours later the staff assembled—the New Yorkers in the corporate boardroom, the far larger Massachusetts crowd via teleconference. The stock dropped by a dollar, to around $13 per share, which was a pretty considerable chunk, though not as bad as other companies on hold have suffered. (Dyax Corporation, another small firm working on a cancer treatment, plunged from $14.41 to $10.26 a couple of years later, for instance.) Armen fielded queries from his worried staff for an hour. And after that, for the next seven weeks, it was almost like the days when the company had been putting together the IND, with people working 24/7 in the labs in Massachusetts. Patient forms had to be collected from the trial sites, some of them going back years. The assays they had been using had to be analyzed and described. “We had to find a way of characterizing the product more precisely, in a way that the FDA is satisfied with,” Elma Hawkins explained. The company gathered the data that it thought would satisfy the government. Then it waited. 

Of course, during this time other work had to go on. There were some promising results from trials that were not on hold, perhaps buttressing Garo Armen’s vision that his product could be applicable for a wide range of cancers. The ten patients in a Phase I trial for pancreatic cancer at Memorial Sloan-Kettering reported a median survival rate of two and a half years, almost twice as long as patients in standard therapy. And the Journal of Immunology published an analysis showing that Oncophage produced a significant reaction in patients with skin cancer and colorectal cancer—an increase in the production of T-cells that could attack the cells specific to those types of cancer. Staff began moving into the new Lexington, Massachusetts, facility. But the clinical hold overwhelmed everything, and the stock price, after a short pickup in late September, slid lower and lower, even below $10. 

Armen came to the realization that nearly ten years of lessons on living with the FDA had not been enough. Antigenics needed more people, full-time, who knew how the agency thinks. “The clinical hold was a reality check,” is the way Uberoi, the communications vice president, described it. So Dr. Renu Gupta, who had been only a consultant to Antigenics when the hold was first issued, was hired full-time as senior vice president of development. Following time-honored industry tradition, the firm also brought in an FDA alumna, Taylor Burtis, as senior director of regulatory affairs. 

Gupta, petite and precise, could not have found a more perfect job. Her father, an Air Force officer named Amar Mapb Verma, had died of kidney cancer in 1997 at age 70. Since then, “I have been on a quest to make a greater effort toward the research of agents for the treatment of kidney cancer,” the daughter said. 

On November 20, Taylor Burtis asked Russ Herndon to step into her small, white-walled office, just a few doors down the hall in the Lexington building. He had to stay calm, Burtis advised, so that nobody passing by and looking through the window would notice any emotion. The FDA had just called to say they were off clinical hold. 

Afterwards, Antigenics officials would claim that they had not been particularly worried about being able to provide the right information or get ting the hold lifted. They did admit to being pleasantly surprised that the FDA moved so fast, and with no further questions (although it would take until the following July for the last red tape to be sealed). The stock spiked briefly. “It was—I don’t want to say vindication, that’s too adversarial. It was a great relief. We did it, we were on the right path; if you work with the FDA you get there,” Elma Hawkins summed up. 

“Look,” said Armen, “they have a whole bunch of boxes that need to be checked. It is our obligation to make sure the boxes are checked.” 

Even with that vote of confidence, Oncophage was hardly ready to go on pharmacy shelves, and Antigenics was hardly in a position to pick a fight with the FDA. That May, the company’s regulatory crew hoped to troop back down to Rockville for another meeting, this time to ask the FDA to approve Oncophage after just one successful Phase III trial, rather than the traditional two. That rule-bending has become increasingly common for urgent cases like cancer. Also, the company would go over the next trials’ “housekeeping details,” as Armen put it. “How we are getting data, how we are bringing outside reviews of data. We will ask the FDA if they have any suggestions to tweak any part of it. Maybe they want five independent radiologists instead of one or two assessing the test results. So that our application won’t be rejected If the data are good but because of some stupid reason they get thrown back, that doesn’t help anyone.” 

In preparation, Renu Gupta was on the phone with FDA staffers about once a month. Antigenics was also amassing an information package that the FDA had to receive at least two weeks before any meeting could be held. The package would run about 300 pages, with the data so far from the Phase I, Phase II, and animal trials, Antigenics’ analysis of the data, and detailed plans and timetables for the next tests. It also had to include, according to FDA regulations, “a list of the specific objectives/outcomes expected from the meeting,” “a proposed agenda, including estimated amounts of times needed for each agenda item and designated speakers,” and “a list of specific questions grouped by discipline.” Interviewed early into the process, Gupta figured it would take eight to ten people working less than full-time three months to prepare the package. 

During that rush of preparations, Antigenics dropped its idea of asking the FDA to let it get by with just one trial—at least for now. Instead, it would start a second trial while it continued to gather results from the first one. According to a formula understood only by statisticians and the FDA 214 of the 650 patients in the first trial would have to show signs of a recurrence of their cancer, including patients getting Oncophage as well as those getting a conventional treatment like chemotherapy. A sample that size would allow a meaningful comparison of the rate of recurrence among Oncophage versus non-Oncophage patients. If the ones with Oncophage had a significantly lower rate, then that would be the time to go back to the FDA, show off the results, and ask if this was strong enough proof of the vaccine’s effectiveness to cut short the second trial and file for approval. Antigenics expected to have the data by winter. “Before the results are out, it would be very unproductive to have this kind of discussion. That’s a hard sale. We don’t want to force them to fix into a position they couldn’t reverse later on,” Armen explained. Besides, he sensed a change in attitude at the FDA—more of an insistence on going by the book—tied in with the staff transfer. However, the second trial, if it ran its full course, would cost the company $15 to $20 million and delay things an extra year and a half, to mid-2006 or even later. 

As it turned out, once Gupta submitted the briefing material, the FDA wanted some additional information, and the meeting had to be put off two months until late July. (Armen was right on the money in one of his predictions: the FDA did, indeed, ask for a third radiologist to assess trial results, as a tie-breaker.) In addition, the agency requested yet more background on the assays that Antigenics would be using to test the drug’s potency; that was another 300-page submission. On top of all that, virtually the entire FDA review staff changed yet again. Even Phil Noguchi transferred to a different part of the agency. So now about ten people from Antigenics’ regulatory and clinical staff were spending an hour or two per week on the phone and e-mail trying to bring the newcomers up to date, plus sending copies of material they had already submitted. Armen guessed it would take the new FDA people a few months to catch up. 

It was a heck of a maiden voyage for a company trying to cure cancer. To be fair, it was also far from the norm. FDA staff “have been gracious enough to admit,” Armen told me with a wry smile, “that this has not been an optimal way to deal with our situation.” At least the 90-minute July meeting went well, according to Armen—in fact, the agency even asked Antigenics whether its method of electronic data collection could be used as an oncology standard. Then it was on to the next step.

About 25 staffers from all parts of the company were already beginning to write the 400 to 900 volumes—that’s 400 to 900 volumes, not pages; each volume has several hundred pages itself—that Antigenics would be submitting to the FDA when it finally sought approval for Oncophage. The huge section on clinical trials would have to wait, of course, but they could get a head start on the rest by gathering the data and drafting the portions about animal tests, safety, and manufacturing. Also, Antigenics was preparing to file its second IND application, this one for its herpes vaccine, which Armen figured to be a lot easier than Oncophage’s. In November, Gupta checked the protocol for the second kidney trial with the FDA’s medical reviewer. Even though this protocol was “quite similar” to the one for the first trial, Gupta said she would not start the trial or mail the protocol to the investigators until “all the i’s are dotted and all the t’s are crossed. I want to know from the FDA if we’re ready to go.” 

To pay for these plans, Antigenics had amassed a nest egg of $135 million from its various stock sales and private offerings, which Armen claimed would see the company at least up to the conclusion of the first trial. (Like most biotechs, it had yet to actually make a profit.) At that point, “if the data is positive, our ability to raise money will not be an issue,” Armen said dismissively. And if not? “We’ll have to scramble like any biotech with bad results.” In addition, executives were talking with a dozen or so bigger drug companies about some sort of partnership in which the larger company’s sales force would also pitch Oncophage. Meanwhile, the chief financial officer went back home to Texas, and Antigenics hired a new one from Ireland. “For investors,” wrote Bloomberg Markets Magazine earlier that year, “the next 12 months will provide the best indication of whether betting on Antigenics was worth the wait.” 

In the main lab on the first floor of the Lexington building, on a wintry Wednesday in 2004, a 26-year-old research associate named Ben Roscoe was working on the future of Oncophage by watching liquid drip. 

The lab stretches over 2,000 square feet, the size of a small restaurant, one alcove after another. Each alcove, or bay, covers about 670 square feet, consisting of a row of beige, glass-doored cabinets above a black countertop, then a narrow aisle, then a black countertop with computer monitors sitting on it, then a second black countertop with computer monitors, then another narrow aisle, then another black counter underneath beige, glass doored cabinets. 

What Roscoe was aiming to do that Wednesday afternoon was to purify a synthesized protein that would be used, in turn, to purify one particular patient’s Oncophage vaccine. The protein had been engineered by Antigenics scientists to contain an amino acid sequence that binds to nickel. That would allow Roscoe to use a purification technique called immobilized metal affinity chromatography, which takes two hours and involves several steps. First, a gel-like solid containing nickel ions was put in a tall, clear plastic column, and a solution containing the special protein was poured in. Next, a clear liquid buffer was added to wash the nickel-protein combination, in order to control acidity and alkalinity and to remove contaminants. After that Roscoe had used a dropper to contribute yet another clear liquid, a chemical called imidazole, which would bind the nickel to the protein. The column was then attached with a large metal clamp to the top of a beaker sitting on the countertop. So now Roscoe, sitting on a tall stool at one of the center counters in his bay, in white lab coat and plastic goggles, was watching as the leftover buffer slowly dripped from the bottom of the column into the beaker. Eventually, that excess would be spilled out. What remained was the target protein. 

After the protein did its job purifying the Oncophage sample, the patient would get his or her vaccine. Antigenics scientists would track the disease’s progress in that patient, and in dozens of others—and in more trials. Some patients, hopefully, would stave off a recurrence of their cancer for a few more months, or even years. Once it had enough evidence, Antigenics would be back to the FDA with hundreds more pages of data. Then it would launch further trials. It would take more data to the FDA: more all-nighters with the printer spitting out copies, and more meetings where the FDA might send it back to redo something or might give a thumbs-up. And this process would go on and on. 

Meticulous, yes. Nitpicky, exciting, risky, potentially life-saving, potentially life-threatening, and no doubt wasteful of paper. Also probably difficult to streamline much. 

And that was only if Antigenics was lucky enough to have a drug that worked.  

next
Beyond Science