11
NOT SO ATYPICAL
This is a field where fads and fancies flourish. Hardly a year passes without some new claim, for example, that the cause or cure of schizophrenia has been found. The early promises of each of these discoveries are uniformly unfulfilled. Successive waves of patients habitually appear to become more resistant to the newest “miracle” cure than was the group on which the first experiments were made.
—Joint Commission on Mental
Illness and Mental Health, 1961
1
ONE OF THE enduring staples in mad medicine has been the rise and fall of cures. Rarely has psychiatry been totally without a remedy advertised as effective. Whether it be whipping the mentally ill, bleeding them, making them vomit, feeding them sheep thyroids, putting them in continuous baths, stunning them with shock therapies, or severing their frontal lobes—all such therapies “worked” at one time, and then, when a new therapy came along, they were suddenly seen in a new light, and their shortcomings revealed. In the 1990s, this repeating theme in mad medicine occurred once again. New “atypical” drugs for schizophrenia were brought to market amid much fanfare, hailed as “breakthrough” treatments, while the old standard neuroleptics were suddenly seen as flawed drugs, indeed.
However, there was something different about this latest chapter in mad medicine.
Prior to the introduction of chlorpromazine, belief in the efficacy of a treatment usually rose in a haphazard way. The inventor of a therapy would typically see it in a rosy light, and then others, eager for a new somatic remedy with which to treat asylum patients, would find it helpful to some degree. And all of the old therapies did undoubtedly “work.” They all served to quiet or weaken patients in some way, and that was a behavioral change that was perceived as good. With chlorpromazine, the belief in efficacy was shaped for the first time by a well-organized company pursuing profits. Yet at that time, the pharmaceutical industry was still in its infancy, and the apparatus for weaving a story of a new wonder drug wasn’t all that well developed. The transformation of chlorpromazine from a drug that induced a chemical lobotomy into a safe, antischizophrenic drug took a decade. But by the late 1980s, the pharmaceutical industry’s storytelling apparatus had evolved into a well-oiled machine. The creation of a tale of a breakthrough medication could be carefully plotted. Such was the case with the atypicals, and behind the public facade of medical achievement is a story of science marred by greed, deaths, and the deliberate deception of the American public.
Recasting the Old
The atypicals were brought to market at a time when Americans had become ever more certain of the therapeutic efficacy of antipsychotic medications. The National Alliance for the Mentally Ill had grown up in the 1980s, and its message was a simple one: Schizophrenia is a biological disorder, one caused by abnormal chemistry in the brain, and medications help normalize that chemistry. That same basic paradigm was used to explain other mental disorders as well, and America—gobbling up antidepressants, anti-anxiety agents, and any other number of psychotropic medications—had in essence accepted it as a way to understand the mind. With this conception of mental illness at work, even patients’ protests against neuroleptics dimmed. They apparently had broken brains and needed the drugs—however unpleasant they might be—to set their minds at least somewhat straight.
And so, as the atypicals arrived, two somewhat curious stories about the therapeutic merits of old neuroleptics were told—one for the ears of other doctors, and one for the ears of the public.
The selling of new drugs necessarily involves telling a story that contrasts the new with the old. The worse the old drugs are perceived to be, the better the new drugs will look, and so as the atypicals moved into the marketplace—which meant that drug firms were hiring well-known psychiatrists to serve as consultants and to run clinical trials—researchers started tallying up the shortcomings of standard neuroleptics. It was an exercise that even seemed to produce a momentary air of liberation within American psychiatry. For so long, investigators had held to the story that Thorazine, Haldol, and the others were effective antipsychotic medications, ultimately good for their patients, and now, at long last, they were being encouraged to see these drugs in an alchemy-free light.
The old drugs, researchers concluded, caused a recognizable pathology, which they dubbed neuroleptic-induced deficit syndrome (NIDS). As would be expected, NIDS was a drug-induced disorder that mimicked natural diseases—like Parkinson’s or encephalitis lethargica—that damaged dopaminergic systems. Two-thirds of all drug-treated patients, researchers calculated, were plagued by “persistent Parkinson’s.” Nearly all patients—some physicians put the figure at 100 percent—suffered from extrapyramidal symptoms (EPS) of some type. (Extrapyramidal symptoms include all of the various motor side effects, such as Parkinson’s, akathisia, and muscle stiffness.) As for tardive dyskinesia, investigators announced that it might be more of a risk than previously thought: It struck up to 8 percent of patients in their
first year of exposure to a potent neuroleptic like haloperidol. The list of adverse effects attributed to neuroleptics, meanwhile, grew to head-spinning length. In addition to Parkinson’s, akathisia, blunted emotions, TD, and neuroleptic malignant syndrome, patients had to worry about blindness, fatal blood clots, arrhythmia, heat stroke, swollen breasts, leaking breasts, impotence, obesity, sexual dysfunction, blood disorders, painful skin rashes, seizures, and, should they have any children, offspring with birth defects. “They have adverse side effect profiles that can affect every physiologic system,” said George Arana, a psychiatrist at the Medical University of South Carolina, at a 1999 forum in Dallas. Nor was it just bodily functions so impaired. “Typical antipsychotic medications,” Duke University’s Richard Keefe told his peers, may “actually prevent adequate learning effects and worsen motor skills, memory function, and executive abilities, such as problem solving and performance assessment.”
2Researchers also began to admit that neuroleptics didn’t control delusions and hallucinations very well. Two-thirds of all medicated patients had persistent psychotic symptoms a year after their first psychotic break. Thirty percent of patients didn’t respond to the drugs at all—a “non-response” rate that up until the 1980s had hardly ever been mentioned. Several studies suggested that even this 30-percent figure might be very low and that as many as two-thirds of all psychotic patients could be said to be “non-responders” to neuroleptics.
3 Perhaps the most revealing confession of all came from NIMH scientists: “Our clinical experience is that while the intensity of thought disorder may decrease with medication treatment, the profile of the thought disorder is not altered.”
4 The drugs, it seemed, might not be “antipsychotic” medications after all.
As for the patients’ quality of life, nearly everyone agreed that neuroleptics had produced a miserable record. More than 80 percent of schizophrenics were chronically unemployed. Their quality of life is “very poor,” wrote New York’s Peter Weiden. Said Arana: “Patients still lie in bed all day. They are suffering.” Long-term outcomes with neuroleptics, commented Philip Harvey, from the Mt. Sinai School of Medicine in New York City, were no better than “when schizophrenia was treated with hydrotherapy.” Said one physician at the Dallas conference: “We will do a great service to our [first-episode] patients by never exposing them to typical antipsychotic drugs.” A 1999 patient survey completed the profile: Ninety percent on neuroleptics said they were depressed, 88 percent said they felt sedated, and 78 percent complained of poor concentration.
5All of this was undoubtedly quite true, and yet it had come at a telling time. New drugs were coming to market and such candor about the old ones served as a powerful foil for making the new ones look good. Psychiatrists who came to the Dallas conference, which was sponsored by Janssen, the manufacturer of the atypical drug risperidone, couldn’t have missed the message: Those who tended to the severely mentally ill would do well to begin prescribing Janssen’s new drug and other atypicals as quickly as possible. The financial forces that helped drive perceptions within psychiatry had changed, and that had led—within the medical community—to a rather stunning reassessment of the old.
But what to tell the public? Neuroleptics—billed as antipsychotic medications—had been the mainstay treatment for schizophrenia for forty years. Over and over again the public had been told that schizophrenia was a biological disease and that drugs helped alleviate that biological illness. The drugs were like “insulin for diabetes.” What if psychiatry now publicly confessed that the dopamine theory hadn’t panned out, that the drugs induced a disorder called NIDS, and that outcomes were no better than when the mad were plunked into bathtubs for hours on end? At least hydrotherapy hadn’t caused tardive dyskinesia, Parkinson’s, or a host of other side effects. What would the public make of that admission?
A subtler story emerged in public forums. The old drugs were beneficial, but problematic. The new drugs were a wonderful advance on the old. As for the tired dopamine theory, it too proved to have some life left in the public domain.
“Breakthrough” Treatments
From a business perspective, the introduction of a new antipsychotic medication was long overdue when the first atypical drug, clozapine, was brought to the U.S. market in 1990 by Sandoz. By the early 1980s, the market for neuroleptics had devolved into a relatively unprofitable phase. There were more than a dozen neuroleptics on the market, and the leading ones—chlorpromazine and haloperidol—had long lost their patent protection and thus were vulnerable to generic competition. Chlorpromazine was selling for less than $10 per month, and haloperidol for not a great deal more. Sales for all neuroleptics in the United States in the late 1980s totaled less than $400 million, which was much less than what one “breakthrough” medication could hope to generate in a year. The market was ripe for a novel antipsychotic, and it came in the form of a drug that, fifteen years earlier, had been discarded as too dangerous.
Clozapine, marketed by Sandoz as Clozaril, was first tested as an antipsychotic in the 1960s. It was different from other neuroleptics in that it blocked both dopamine and serotonin receptors. When tested, it was found that it didn’t cause the usual high incidence of extrapyramidal symptoms. However, it did cause any number of other neurotoxic effects—seizures, dense sedation, marked drooling, rare sudden death, constipation, urinary incontinence, and weight gain. Respiratory arrest and heart attacks were risks as well. Sandoz introduced it into Europe in the 1970s, but then withdrew it after it was found to also cause agranulocytosis, a potentially fatal depletion of white blood cells, in up to 2 percent of patients.
The return of clozapine was made possible by the fact that, by the mid-1980s, it was no longer possible to ignore the many drawbacks of neuroleptics. Because of the risk of agranulocytosis, the FDA approved it only as a second-line therapy for patients who didn’t respond to standard neuroleptics. Even so, it quickly proved to be a hit in the marketplace. It didn’t appear to cause extrapyramidal symptoms, and at least some patients responded—in terms of the clarity of their thinking—in a robust fashion. Sandoz also initially sold clozapine bundled with weekly blood tests for agranulocytosis, with the test to be done by its affiliate, Caremark, and it put a whopping price of $9,000 a year on the package.
Other drugmakers now had a clear model to emulate. A drug that could block both serotonin and dopamine receptors could hopefully knock down psychosis without causing the usual extrapyramidal symptoms, and it might even improve cognition. Any drug that could do that without causing agranulocytosis could be marketed as a first-line therapy, and generate blockbuster financial returns. In the early 1990s, the medical literature began bubbling with reports of just such a drug, risperidone. Janssen obtained FDA approval in 1993 to sell it, and by the end of 1995, more than twenty reports had appeared in psychiatric journals touting its benefits. It was said to be equal or superior to haloperidol in reducing positive symptoms (psychosis), and superior to haloperidol in improving negative symptoms (lack of emotion). Researchers reported that it reduced hospital stays, improved patients’ ability to function socially, and reduced hostility. Best of all—and this was the sound bite that graced journal advertisements—the incidence of extrapyramidal symptoms with risperidone was said to be “equal to placebo.”
6The media presented risperidone in even more glowing terms. This new drug, the
Washington Post reported, “represents a glimmer of hope for a disease that until recently had been considered hopeless.” Risperidone, it said, did not “cause sedation, blurred vision, impaired memory or muscle stiffness, side effects commonly associated with an earlier generation of antipsychotic drugs.” George Simpson, a physician at the Medical College of Pennsylvania, told the
Post: “The data is very convincing. It is a new hope, and at this moment it appears, like clozapine, to be different from all existing drugs.” The
New York Times, quoting Richard Meibach, Janssen’s clinical research director, reported that “no major side effects” had appeared in any of the 2,000-plus patients who had been in the clinical trials. The
Times also provided its readers with a diagram of how risperidone worked. “Researchers,” it said, think that drugs like risperidone “relieve schizophrenia symptoms by blocking excessive flows of serotonin or dopamine, or both.”
7The dopamine theory, in a slightly amended version, was alive and well. Schizophrenics suffered from not just one neurochemical abnormality, but two, and the new atypicals helped normalize both. As for the older drugs, the
New York Times reported, they “relieve typical symptoms like delusions and hearing voices in about 70 percent of patients. But they are less effective in treating other symptoms of schizophrenia, like withdrawal, lack of energy and motivation, and the inability to experience pleasure.” All of the other papers cast the standard neuroleptics in that same light: They were
less effective (or ineffective) in treating negative symptoms. They did successfully treat positive symptoms in about 70 percent of patients. None of the newspapers told of how the older drugs could impair cognitive function and worsen negative symptoms, nor was it mentioned that they caused a recognizable pathology, known as NIDS, or that, as Philip Harvey had written, it might be that they “had no impact on the overall outcome of schizophrenia.”
8 Instead, in this story told to the public, risperidone’s arrival in the marketplace was successfully placed within the framework of the long-running story of the general
efficacy of neuroleptics. The tale of helpful, antipsychotic drugs was maintained.
It was also a story that Janssen took to the bank. With praise flowing in the scientific literature and in the media, Janssen was able to charge $240 per month for risperidone, more than thirty times the price of chlorpromazine. In 1996, U.S. sales of risperidone topped $500 million, which was greater than revenues for all other neuroleptics combined. That same year, Janssen won the prestigious Prix Galien for its new drug, an award touted as the pharmaceutical industry’s Nobel Prize.
Eli Lilly was the next to bring an atypical to market. However, since Janssen had made it first to the marketplace, Eli Lilly’s challenge was to prove in clinical trials that its new drug, olanzapine (marketed as Zyprexa), was superior to both haloperidol and risperidone. Olanzapine was chemically more similar to clozapine than Janssen’s drug (risperidone blocked D
2 receptors in a more potent manner than did clozapine or olanzapine), and as olanzapine came to market in 1996, reports in the medical journals told just the story that Eli Lilly wanted. Olanzapine, the articles said, worked in a more “comprehensive” manner than either risperidone or haloperidol. It was a well-tolerated agent that led to global improvement—it reduced positive symptoms, caused fewer motor side effects than either risperidone or haloperidol, and improved negative symptoms and cognitive function. It reduced hospital stays, prevented relapse, and was useful for treatment-resistant schizophrenia.
9Apparently, yet another step up the medical ladder had been taken. Olanzapine, the
Wall Street Journal announced, has “substantial advantages” over other current therapies. “Zyprexa is a wonderful drug for psychotic patients,” said John Zajecka, at Rush Medical College in Chicago. Harvard Medical School’s William Glazer told the
Wall Street Journal: “The real world is finding that Zyprexa has fewer extrapyramidal side effects than Risperdal.” Stanford University psychiatrist Alan Schatzberg, meanwhile, confessed to the
New York Times: “It’s a potential breakthrough of tremendous magnitude.” On and on it went, the glowing remarks piling up. Laurie Flynn, executive director of the National Alliance for the Mentally Ill, even put an exclamation point on it all: “These new drugs truly are a breakthrough. They mean we should finally be able to keep people out of the hospital, and it means that the long-term disability of schizophrenia can come to an end.”
10Since its drug was seemingly better than Janssen’s, Eli Lilly was able to put a higher price tag on it. Patients would have to pay nearly $10 per day for this new miracle drug. In 1998, olanzapine sales in the United States alone topped $1 billion. Total U.S. sales of antipsychotic drugs hit $2.3 billion that year—roughly six times what they had been prior to risperidone’s arrival on pharmacy shelves. By that time, AstraZeneca had brought a third atypical to market, quetiapine (marketed as Seroquel), and there was no longer any possible doubt about the superiority of these new drugs. They were,
Parade magazine told its readers, “far safer and more effective in treating negative symptoms, such as difficulty in reasoning and speaking in an organized way.” The
Chicago Tribune echoed the sentiment: The newer drugs “are safer and more effective than older ones. They help people go to work.” Or as the
Los Angeles Times put it: “It used to be that schizophrenics were given no hope of improving. But now, thanks to new drugs and commitment, they’re moving back into society like never before.”
11American science had surely produced a remarkable medical advance. New wonder drugs for madness had arrived.
The Business of Clinical Research
This belief—that the atypicals were superior in safety and efficacy—had a solid scientific pedigree. It was based upon the results of the clinical trials that the pharmaceutical companies had conducted to gain FDA approval for their drugs, which had been published in the best peer-reviewed medical journals. The American Journal of Psychiatry, Journal of Clinical Psychopharmacology, Neuropsychopharmacology —the literature was filled with articles praising the drugs. They were authored by some of the leading lights in American psychiatry, and inevitably the articles included an impressive array of statistics and charts, detailed explanations of methodology, and sober-minded conclusions. What the public couldn’t have known is that this whole arena of science—the clinical testing of drugs—had undergone a profound change in the 1990s, one that lent itself to the creation of fairy tales, and that the FDA, in its review of the same trial data, didn’t buy the companies’ claims of superiority at all.
The refashioning of the clinical testing of commercial drugs can be traced back to the mid-1980s. Up until that point, pharmaceutical firms primarily hired academic physicians to test their drugs. More than 70 percent of all drug trials were conducted in academic settings, and the relationship between the drug companies and the academic doctors was one in which the doctors, in many ways, had the upper hand. The academic physicians often viewed the drug companies with more than a little disdain—grants from the National Institutes of Health were the coveted coin in the academic realm—and the drug companies basically had to come to the physicians as humble supplicants. The academic doctors were known as Thought Leaders, and the fact that they had the upper hand in the relationship ensured that experimental drugs went through at least a measure of independent testing. The academic doctors regularly modified the protocols, even though this often greatly irritated the drug companies.
However, starting in the late 1980s, a for-profit clinical trials industry arose to
serve the pharmaceutical companies. It emerged in bits and pieces. First, community physicians who were feeling financially squeezed by health maintenance organizations turned to clinical trials as a way to supplement their incomes. Some conducted trials as an adjunct to their regular practices, while others opened full-time “dedicated” research centers. Then a group of urologists, from nineteen states, banded together to form Affiliated Research Centers. A pharmaceutical company developing a urology drug could come to Affiliated Research Centers and immediately have community physicians across the country lined up to test it. Doctors in other specialties soon established similar investigator networks. Next came pure business ventures, eager to consolidate services for the pharmaceutical firms. Entrepreneurs raised venture capital with the goal of building nationwide chains of research sites. By 1997, venture capital groups had poured $100 million into such businesses, and two of these venture-funded companies had turned public. It all led Peter Vlasses, director of clinical research for a consortium of university hospitals, to lament: “Everybody under the sun is now a clinical researcher. What used to take place only in academic centers is now everywhere.”
12As this mix of for-profit research sites sprung up, spending by pharmaceutical companies for their services soared, from under $1 billion in 1990 to $3.5 billion in 2000. The role of these for-profit businesses in the research process was very straightforward: Their job was to recruit patients quickly into trials and keep them there until they completed the study protocols. Said one Texas investigator in 1995: “I don’t begrudge [the pharmaceutical companies] viewing me as a vendor. I am providing a technical service, and in that sense, I view it as a business. If I were not turning a profit, I wouldn’t do it. And I don’t think many investigators would.” There certainly was money to be made. In 1997, community physicians experienced at conducting clinical trials reported earning, on average, $331,500 from their research activities. “Dedicated” research centers reported revenues of $1.35 million. A newsletter for neurologists,
Neuropractice, summed up the opportunity in commercial drug trials: “A growing number of neurologists are discovering a gold mine in their clinical practices: their patient population.” A few investigators chalked up even bigger scores. In 1996, pharmacist Jeff Green took his company, Collaborative Clinical Research, public, raising $42 million for his expansion plans. Two Rhode Island psychiatrists, Walter Brown and Michael Rothman, reaped the biggest financial success of all. In 1997, they sold their seven-year-old company, Clinical Studies, which consisted of a chain of research centers along the East Coast, for stock valued at $96 million.
13The commercial testing of experimental drugs had moved out of an academic setting and into a for-profit setting. Struggling to cope with this loss of business, academic centers also began changing their ways. A number of schools opened administrative offices devoted to securing contracts for commercial drug trials. The central “offices of clinical trials” promised the pharmaceutical firms that they would help their physicians start trials quickly and successfully fill them with patients. They too adopted a
service attitude toward the drug firms—that’s what it now took to compete in the clinical-trials business. And with the old disdain toward pharmaceutical money melting away in academia, individual faculty became more eager to work for the drug firms as well. In a 2000 editorial titled “Is Academic Medicine for Sale?”
New England Journal of Medicine editor Marcia Angell catalogued the many ways that drug money flowed to academic doctors:
The ties between clinical researchers and industry include not only grant support, but also a host of other financial arrangements. Researchers also serve as consultants to companies whose products they are studying, join advisory boards and speakers’ bureaus, enter into patent and royalty arrangements, agree to be the listed authors of articles ghostwritten by interested companies, promote drugs and devices at company-sponsored symposiums, and allow themselves to be plied with expensive gifts and trips to luxurious settings. Many also have equity interest in the companies.
14In this new service environment, the drug companies enjoyed the best of all possible worlds. They could utilize for-profit research sites to recruit the bulk of their patients into their large clinical trials. At the same time, they could hire academic doctors to lend intellectual prestige and an aura of independence to the trial results. Together, these services produced the perfect package. The pharmaceutical companies could get their trials done quickly, the public would see the names of the academic physicians on the published articles, and all the while, they would control every aspect of the drug-testing process. They could, for instance, design their protocols without having to worry that academic doctors would insist on changing them, and that meant that it would now be easier for them to set up trials biased toward their own drugs. “A pharmaceutical company,” acknowledged
Journal of Clinical Psychiatry editor Alan Gelenberg in 1999, “goes to great pains to construct studies that are likely to turn out in its favor.”
15 The drug companies also controlled analysis of the data, and that control, the
New England Journal of Medicine wrote, “allows companies to provide the spin on the data that favors them.”
16In short, a dark truth became visible in American medicine in the 1990s. Bias by design and the spinning of results—hallmarks of fraudulent science—had moved front and center into the testing of commercial drugs. While this corruption of the drug-testing process was not unique to psychiatry, it was no accident that the
New England Journal of Medicine, as it sought to illustrate the problem, found the best evidence of it in this specialty. When the journal tried to identify an academic psychiatrist who could write an honest review of antidepressant drugs, it found “very few who did not have financial ties to drug companies.” One author of an article on antidepressant drugs had taken money from drug companies on so many occasions, Angell told an ethics conference in 2000, that to disclose all of them “would have taken up more space than the article.” She concluded: “You are seeing played out in psychiatry the extremes of what is happening elsewhere in medicine.”
17And all of these extremes were at work as the atypicals came to market.
Eye on the Castle
One of the first academic physicians to tout the benefits of risperidone, in a 1992 article published in the
Psychopharmacology Bulletin , was a psychiatrist at the Medical College of Georgia, Richard Borison. His 1992 report came to be frequently cited in the scientific literature, and over the next five years, he regularly published additional articles related to the merits of the atypicals. In 1994, he traveled to Australia to speak about risperidone, and he also was one of the experts quoted by the newspapers. Risperidone, he told the
New York Times in 1992, was producing results that were “absolutely on the money.”
18It was a quote that revealed more about Borison than it did about risperidone.
Although Borison was popular with the drug companies, he had a shady track record. In 1984, Smith Kline had given him a grant to conduct a test comparing Thorazine to a generic knockoff—in such studies, the drug company hopes to prove that the generic is not really equivalent—and the next year, at the American Psychiatric Association’s annual convention, he reported the results that Smith Kline wanted to hear. Schizophrenics who had been switched from Thorazine to generic chlorpromazine had become agitated and hostile, he told his peers. His findings, which suggested that hospitals and caregivers would be wise to avoid generic chlorpromazine and buy Smith Kline’s Thorazine instead, were widely circulated. However, the FDA chose to investigate his study, which Borison had conducted at Veterans Affairs Medical Center in Augusta in May 1984, and determined that the hospital hadn’t even stocked Thorazine at that time. The patients could not have been switched from Thorazine to generic chlorpromazine at all—they had been on the generics all along. Although he tried to explain this damning finding away, the conclusion was obvious: Borison had simply fabricated the results.
19The FDA publicly rebuked Borison, but since he hadn’t submitted his data to the agency, it lacked authority to formally discipline him. In the wake of the scandal, Borison’s research activities lagged for a year or two, and then all was forgotten. He became a full professor at the Medical College of Georgia in 1988, was made chief of psychiatry at the VA hospital, and soon he and his research partner, Bruce Diamond, a pharmacologist on the faculty, had drug companies giving them one lucrative contract after another. Eli Lilly, Janssen, Zeneca, Sandoz, Glaxo, Abbott, Pfizer, Hoechst Marion Roussel—they all came knocking. The two researchers secured 160 contracts from drug firms over the course of a decade, worth more than $10 million. They received $4 million for schizophrenia drug trials alone. “We knew how to collect the information the way they wanted us to,” one of Borison’s employees told VA hospital officials in 1996. “And we were high en-rollers [of patients into trials], so they loved us.”
20As faculty, Borison and Diamond were supposed to get approval from the medical school to do drug studies. Payments for commercial trials were supposed to be sent directly to the school. But starting in 1989, Borison and Diamond cut the college out of the loop and told the drug firms to send their money directly to them. They opened an office across the street from the medical school and turned it into a full-time research mill, which they called Clinical Therapeutics. In order to keep the school in the dark about their research activities, they used a commercial service to do ethical reviews of their studies. The one thing they let the medical school continue to do was pay some of their expenses—they even placed Clinical Therapeutics’ staff on the school’s payroll.
To run their trials, Borison and Diamond hired attractive young women as study coordinators. When women came to apply for a coordinator’s position, Diamond would wait to “see what they looked like in the waiting room,” employee Angela Touhey told VA officials. “If they were overweight, if they were older, he would refuse to see them. He would ask a coordinator to talk to them and they would be sent home.” There was a financial logic to their hiring preferences. The majority of patients recruited into trials are men, and that is particularly true of schizophrenia trials, which are among the best-paying studies in the business. Borison and Diamond stood to receive $10,000 to $25,000 for every schizophrenic the young women could coax into a drug trial.
With such money waiting to be made, Borison and Diamond gave the coordinators patient-recruitment bonuses that ran into the thousands of dollars. One coordinator was given a new Honda Accord as a bonus. Each time a new contract from a drug firm came in, the coordinators would hit the phones. They would call mentally ill people living in the community and promise them $150 if they would participate in the study. Patients already on locked wards at the hospital would be given cigarettes for participating. Some patients were churned through study after study, as well. “When there is a possibility you’re going to get a car, you’re going to do whatever you can,” Touhey said.
Even though the coordinators lacked medical training, they regularly decided whether patients qualified for the trials. At times, they fudged information about the patients so that they met eligibility criteria. They also drew blood samples and adjusted the patients’ drug dosages. Borison, employees said, rarely bothered to show up at the office. The coordinators would fill in the paper documents and then pass them on to Diamond, who would forge Borison’s signature. At one weekly staff meeting, Touhey told the VA investigators, Diamond made it clear that he wasn’t interested in hearing about the patients. “Bruce said to me, ‘We don’t care about how the patients are doing. We just want to know how many people you have enrolled in the past week or couple of weeks.’” Indeed, Borison and Diamond “had no idea who the patients were,” Touhey said.
The money rolled in. Borison and Diamond stashed more than $5 million in cash and securities in various U.S. banks and Barclay’s Bank in London. Each tooled around town in a new Mercedes Benz, and Diamond liked to show off his $11,000 gold Baume Mercier wristwatch. Borison’s material dreams were even grander. He had an architect draw up plans for an 11,000-square-foot castle, complete with moat and medieval pennants. In anticipation of his new home, he made himself a regular at Sotheby’s auction house, both in New York and London, purchasing such items as fifteenth-century tournament armor ($6,600), bronze doors ($16,000), a stone lion fountain ($32,000), two seven-foot stone entry lions on pedestals ($10,500), a marble statue of Cupid ($6,250), a crystal chandelier ($5,000), a coat of arms ($1,650), and more than 100 other decorative pieces—expensive paintings, marble vases, and antique furniture—that would make a castle fit for a king.
This went on for years. In early 1994, a study coordinator, Terri Davis, threatened to blow the whistle after a patient, who had been improperly admitted to an olanzapine trial, attempted suicide, but Borison and Diamond bribed her to keep quiet. A steady stream of monitors sent by the drug companies to audit their research records came and went. Borison would come in on the days the monitors were there and “set up a mock office,” and the monitors would leave none the wiser. Risperidone was approved by the FDA in 1993, and the staff at Clinical Therapeutics even felt a measure of pride at their contribution—Borison had been a lead investigator in both of the pivotal U.S. studies Janssen had conducted. Finally, in 1996, Angela Touhey left and went to work for David Hess, chief of neurology at the Augusta VA hospital, and that triggered the collapse of Clinical Therapeutics. She told Hess about what had been going on, he investigated, and soon the police had been called. “This whole thing was very dirty,” Hess told the medical school and hospital. “It was basically a numbers game. These patients are purely used for the greed of the researchers. That was very apparent to me what was going on.”
Both Borison and Diamond went to prison, but not for research fraud. Their principal crime was that they had stolen from the college. Diamond was sentenced to five years in prison, fined $125,000, and ordered to pay $1.1 million to the college. Borison got fifteen years, was fined $125,000, and was ordered to pay $4.26 million to the college. As for his public comments about the merits of atypicals, Borison’s last published article on the drugs—his eleventh overall—appeared in early 1997, about the same time that he was indicted. It was titled, “Recent Advances in the Pharmacotherapy of Schizophrenia,” and it took him a full sixteen pages to detail all that he knew about how they had helped his patients get well.
21
Swept Under the Rug
While the misdeeds of Borison and Diamond do not reveal anything about the merits of the atypicals, they do reveal much about the amount of money that was flowing to investigators who conducted the trials and how an academic physician who spoke well of a drug could expect a steady flow of research contracts, and polish up his CV at the same time. Their scandal provides insight into the
storytelling forces at work as the new atypicals came to market. Those same forces can also be seen in a second behind-the-scenes aspect of the atypical trials, and that is how investigators reported on patient deaths. One in every 145 patients who entered the trials—for risperidone, olanzapine, quetiapine, and a fourth atypical called sertindole—died, and yet those deaths were never mentioned in the scientific literature.
22 Nor did anyone dare confess that the high death rate was due, in large part, to study design.
Pharmaceutical companies developing new drugs always want to get their trials done as quickly as possible. The adage in the industry is that every day delayed in the trial process is a million-dollar loss in potential sales. To get their atypicals approved, Janssen, Eli Lilly, and other companies needed to prove that the drugs reduced psychotic symptoms. Thus, they needed patients who were actively psychotic. To develop this patient pool (and do so quickly), they relied on protocols that required patients to be
abruptly withdrawn from their existing medications. This abrupt withdrawal (also known as a “washout”) could be expected to trigger a return of their hallucinations and delusions. Once the patients were newly sick, they could then be randomized into the trial and treated either with placebo, a standard drug like haloperidol, or the experimental drug. “If you don’t take people who have reestablished active disease, then you don’t know what you are looking at” when you test the drug, explained Robert Temple, director of the FDA’s Office of Drug Evaluation. “That is why you have to have a washout.”
23However, abrupt withdrawal (as opposed to gradual withdrawal) is also known to put patients at risk of severe clinical deterioration. It is contrary to good clinical practice and it increases the risk of suicide, which is precisely how many people died in the trials. At least thirty-six people in the studies of the four drugs killed themselves. Hanging, drowning, gunshots to the head, and death by jumping were some of the ways they chose to go. The overall suicide rate for patients in the trials, on a time-adjusted basis, was two to five times the norm for schizophrenics.
nOne of the thirty-six people who died in this manner was forty-one-year-old Susan Endersbe, from Minneapolis. Her struggles with schizophrenia were of a familiar kind. She’d first begun to grapple with emotional difficulties as a teenager, and then she’d become more seriously ill while a student at the University of Minnesota. For the next twenty years, she went through many ups and downs. At times, she was able to live in her own apartment, with support from social services, but then her symptoms would worsen, and she would check herself into a hospital. The one constant was that she showed a will to live. “She was extremely intelligent and very high functioning for having such a disability, and recognized the illness for what it was,” said her brother, Ed Endersbe. “She wanted very much to live and be a survivor.”
24On May 7, 1994, she checked herself into Fairview Riverside Hospital in Minneapolis. It was a particularly difficult time for her—her mother had been diagnosed as terminally ill with cancer, and now Susan was feeling suicidal. Hospital doctors put her on an antidepressant, and gradually her mood lightened. On May 26, she told nurses that she was feeling much better and would be ready to leave soon. But the very next day, she was referred to psychiatrist Faruk Abuzzahab, and he had a different proposition for her. Would she like to be in a trial for a new drug, sertindole?
25Abuzzahab was a prominent psychiatrist in Minnesota. He’d served a term as president of the Minnesota Psychiatry Society and had chaired its ethics committee. He was also well known in the world of commercial drug research. He’d done a number of studies for pharmaceutical firms and had been a named author on published results. In the spring of 1994, he had a contract with Abbott Laboratories to test sertindole. However, the protocol specifically excluded patients who were suicidal. Nursing notes, according to her brother Ed, also showed that Susan Endersbe had reservations about entering a drug experiment. But no matter. On May 27, the day that Abuzzahab met Endersbe, he enrolled her in the study.
As the protocol stipulated, Abuzzahab immediately withdrew her medications. He also took her off the antidepressant venlaxafine, which had seemed to help her, and very shortly she began to deteriorate. Her emotional despair returned, and to make matters worse, she suffered a flare-up of extrapyramidal symptoms, a common occurrence when antipsychotic drugs are abruptly withdrawn. By June 3, nurses were writing that her suicidal feelings had returned. Devils were now struggling for her mind, her brother said. Even so, Abuzzahab kept her in the study, and on June 8, he randomized her into one of the study arms. She was, Abuzzahab wrote in research documents, experiencing “0” adverse events.
Nursing notes, however, told a different story:
June 8: Passive thoughts of suicide with hopeless/helplessness in coping with changes from study. Patient feels hopeless, has suicidal thoughts of leaving the unit and jumping off the bridge on Franklin Ave.
June 9: Patient states she feels suicidal and has been actively thinking about suicide, stating that she’s different from others because when she attempts, she will succeed. Refuses to divulge method she has planned, however states she is unable to use the method while hospitalized. States she can agree to not harm self while in hospital.
On June 10, Susan Endersbe asked Abuzzahab for a day pass. The protocol prohibited patients from leaving the hospital during the first four weeks of the study, but Abuzzahab shrugged off this rule and granted her a pass for the next day. He didn’t even require that anyone go along.
The next morning, Susan Endersbe prepared to go out. She took the time to dress neatly and to do her hair in a French braid. It was as though she were preparing for an event and wanted to look nice. She went to her apartment, where she watered her plants and gathered up a few keepsakes. As she left, she slipped the key back under the door. She would not be needing it any more—the thoughtful thing to do would be to leave it for the landlord. She then walked directly to the Franklin Avenue Bridge, which spanned the Mississippi River. Just as she had said she would, she clambered over the railing and leaped to her death.
“For nearly 20 years, my sister was managing to win the battle for her survival, and when she went on a drug study there were supposed to be safeguards in place to protect her,” said her brother. “Not only were they not in place, they neglected to have the usual safeguards that she would have had if she stayed on as an inpatient in the hospital. And to wash people out from their medication, to take away any kind of treatment, that to me is inhumane. If they did that to someone with a physical illness, I would think it would be criminal.”
All told, seven people killed themselves in the sertindole trials. At least ten patients did so in the risperidone trials, fifteen in the olanzapine studies, and four in the quetiapine experiments. They were casualties of a drug-testing process that required that “active disease” be reestablished in patients, but when it came time to report the trial results in the scientific journals, this loss of life was conveniently forgotten.
o
Heart of Darkness
Borison’s misdeeds, unreported patient suicides, Abuzzahab’s callous neglect of Endersbe—all of these are dark splotches on the research landscape. They also lead, in stepping-stone fashion, to a much larger story, and that is how the trial process, in the case of the atypicals, was employed not to inform, but to mislead. This story is revealed in FDA documents obtained through Freedom of Information requests.
The scientific background to the clinical trials of the atypical drugs is, in some ways, a confusing one. On the surface, the trials appeared to straightforwardly compare the atypicals to placebo and to haloperidol. But surface appearances can be deceiving. In the first place, there was no true placebo group in the trials. The same “abrupt withdrawal” design that put patients at great risk also produced a placebo group that could be expected to fare poorly. The placebo group consisted of patients going through an event—abrupt withdrawal—that could be expected to make them worse, and then they were left untreated for that withdrawal-induced illness. While that trial design provided companies with a convenient placebo foil for making their drugs look good, it made for poor science. Harvard Medical School’s Ross Baldessarini put it this way: “It could exaggerate drug-placebo differences, and you could get a stronger impression of the benefit of the drug. It may not be a completely fair comparison.”
26 In the second place, as the FDA reviewers repeatedly pointed out, Janssen and Eli Lilly used biased trial designs to favor their experimental drugs over the standard neuroleptics.
pJanssen’s risperidone was the first of the three atypicals (excluding clozapine) to undergo FDA review. The company conducted three “well-controlled” trials to support its New Drug Application.
27In the first, involving 160 patients at eight U.S. centers, ris - peridone was compared to placebo. Nearly 50 percent of the risperidone patients didn’t complete the six-week trial. Risperidone was superior to placebo in reducing positive symptoms, but neither risperidone nor haloperidol was superior to placebo on the “Clinical Global Impression Scale,” which measures overall improvement.
In the second, which involved 523 patients at twenty-six sites in the United States and Canada, four doses of risperidone were compared to a 20-milligram dose of haloperidol and to placebo. Forty-five percent of the risperidone-treated patients didn’t complete the eight-week trial. Janssen maintained that this study showed that risperidone, at an optimal dose of 6 milligrams, was superior to haloperidol for treating positive and negative symptoms, which were the conclusions published in the medical journals. However, FDA reviewers noted that Janssen had used a single, high dose of haloperidol for comparison, a dose that “may have exceeded the therapeutic window” for some patients, and thus the study was “incapable by virtue of its design of supporting any externally valid conclusion about the relative performance of haloperidol and Risperdal.”
This second risperidone trial conducted by Janssen clearly illustrates how trial design can be used to produce results a company wants. Haloperidol was a drug that had been in widespread use for more than twenty years, and it was well known—as the FDA reviewers pointed out—that high doses were problematic. For instance, Theodore van Putten at UCLA had reported in 1987 that a 20-milligram dose of haloperidol was “psychotoxic” to many patients and that even a 10-milligram dose triggered painful akathisia in 76 percent of patients. Similarly, in 1991, Duke University researchers determined that doses of haloperidol above 10 milligrams daily regularly led “to significant increases in distressing extrapyramidal side effects.”
28 By using a 20-milligram dose, then, Janssen could expect that there would be a high incidence of extrapyramidal side effects in the haloperidol group and thus help create a story of how risperidone, by comparison, was a much safer drug.
In its third study, which involved 1,557 patients in fifteen foreign countries, Janssen compared five doses of risperidone to a 10-milligram dose of haloperidol. Janssen claimed that this study showed that its drug was “more effective than haloperidol in reducing symptoms of psychosis,” but Paul Leber, director of the FDA’s Division of Neuropharmacological Drugs, once again rejected this argument. The study was “incapable” of making any meaningful comparison. The design flaw in this study, Leber noted, was that Janssen had compared multiple doses of its experimental drug to one dose of haloperidol. In order to honestly compare two drugs, an equal number of “equieffective” doses must be tested, as otherwise the study unfairly favors the drug that is given in multiple doses. Such trial design, Leber wrote on December 21, 1993, is “a critical preliminary step to any valid comparison of their properties.”
29In sum, the FDA concluded that Janssen had shown evidence that risperidone was effective in reducing positive symptoms compared to placebo over the short term but had not proven that its new drug was superior to haloperidol (which wasn’t required for approval). As for risperidone’s safety profile, a review of the FDA data shows it was much more problematic than the public had been led to believe. Researchers had proclaimed that the incidence of extrapyramidal symptoms was the “same as placebo.” The
New York Times, quoting a Janssen official, had reported that “no major side effects” had occurred in 2,000-plus patients. Those were results that spoke of a very safe drug. In fact, eighty-four risperidone patients—or about one in every thirty-five—had experienced a “serious adverse event” of some type, which the FDA defined as a life-threatening event, or one that required hospitalization. (Suicides and suicide attempts accounted for more than half of these serious events.) Moreover, in general, the incidence of adverse events in risperidone patients and haloperidol patients was roughly the same. Nine percent of risperidone patients had to drop out because of adverse events, compared to ten percent of haloperidol patients. Seventy-five percent of risperidone patients experienced at least one adverse event, compared to 79 percent of haloperidol patients. Even on a moderate dose of risperidone, 17 percent of risperidone patients suffered extrapyramidal symptoms, and at a high dose, one-third of risperidone patients did—which was about the same incidence of EPS in patients treated with 20 milligrams of haloperidol.
q Wrote FDA scientist Thomas Laughren: “It remains to be seen how risperidone compares with other antipsychotics with regard to EPS, as haloperidol is at the high end of the spectrum.”
In its final letter of approval to Janssen, the FDA made explicit its conclusions about the relative merits of risperidone and haloperidol. Robert Temple, director of the FDA’s Office of Drug Evaluation, told Janssen:
We would consider any advertisement or promotion labeling for RISPERDAL false, misleading, or lacking fair balance under section 502 (a) and 502 (n) of the ACT if there is presentation of data that conveys the impression that risperidone is superior to haloperidol or any other marketed antipsychotic drug product with regard to safety or effectiveness.
30However, while the FDA had the authority to stop Janssen from making false claims in its ads, it had no control over what academic physicians, who had been paid by Janssen to conduct the trials, reported in their medical journals or told the press. They had touted risperidone as superior to haloperidol prior to the FDA’s review of the data, and they continued to do so afterward. In 1997, a group of elite academic psychiatrists revisited the trial data one last time, and in the
Journal of Clinical Psychiatry, they once more told the story of its superiority. They wrote: “Our findings suggest that risperidone has important advantages compared with haloperidol. When administered in an effective dose range, risperidone produced greater improvements on all five dimensions of schizophrenia.”
31In modern American psychiatry, the scientific journals had become a place to make claims that the FDA had explicitly banned from advertisements as false.
The FDA, however, had simply critiqued Janssen’s trials as biased—it didn’t conduct its own studies on the relative merits of risperidone and haloperidol. But once risperidone was on the market, physicians who had not received any money from Janssen could get their hands on the drug and conduct their own studies, and their results revealed, in dramatic fashion, just how egregiously the public had been misled, particularly in regard to the company’s claims that extrapyramidal symptoms were the “same as placebo.”
First, physicians at McMaster University in Hamilton, Ontario, found that in a study of 350 patients never before treated with neuroleptics, a low dose of risperidone caused Parkinsonism in 59 percent of the patients, compared to 52 percent of patients treated with haloperidol. The incidence of akathisia was also higher in the risperidone patients, leading the researchers to conclude that “risperidone may not be a useful alternative to typical antipsychotic drugs.”
32Second, NIMH researchers determined that when risperidone and haloperidol were compared at equivalent therapeutic levels, risperidone induced extrapyramidal symptoms in 42 percent of the patients, compared to 29 percent in the haloperidol group.
33Third, University of Pittsburgh researchers determined that risperidone, when administered to neuroleptically naive patients, caused a disruption in eye movement still present four weeks after treatment was initiated, evidence of a neurological side effect lingering for a much
longer time than it did in patients treated with haloperidol.
34Those studies were just the beginning of reports that, unbeknownst to the public, stripped much of the “breakthrough” luster from risperidone. In 1995, physicians at the University of Pittsburgh Medical Center complained that while the hospital’s spending on antipsychotic medications had soared after risperidone was introduced, it couldn’t find evidence that the drug produced better outcomes. Psychiatrists at the University of California at San Francisco, meanwhile, determined that only 29 percent of patients initially placed on risperidone were still on the drug two years later, with 55 percent quitting the drug because it didn’t work. “Our findings suggest that in routine clinical practice, use of risperidone is plagued by many of the same problems that are well known with older antipsychotic medications,” they wrote. Yet another researcher, Jeffrey Mattes, director of the Psychopharmacology Research Association, concluded in 1997 that “it is possible, based on the available studies, that risperidone is not as effective as standard neuroleptics for typical positive schizophrenia symptoms.” Letters also poured in to medical journals linking risperidone to neuroleptic malignant syndrome, tardive dyskinesia, tardive dystonia, liver toxicity, mania, and an unusual disorder of the mouth called “rabbit syndrome.”
35 A final blow was delivered in the prestigious medical journal
Lancet. Janssen’s clinical investigators had published results from the same trial multiple times, and critics held up this behavior as illustrative of the “salami science”—characterized by “redundant publication, slippery authorship, and opaque reporting of trial data”—that was poisoning the medical literature. Risperidone, one
Lancet writer snapped, was “a marketing success, if nothing else.”
36But the public heard little of this. The FDA’s criticisms took place behind closed doors, available to the public only through a Freedom of Information request. Researchers who independently assessed risperidone and found that it appeared to cause motor dysfunction just as frequently as haloperidol did (or even more frequently) didn’t have the finances to hire PR firms to publicize their research. Their papers quietly appeared in the medical journals, and the lay public never heard a peep about them. Even the criticism in Lancet didn’t stir any bad newspaper press for Janssen. Besides, Eli Lilly had gained approval to market olanzapine in 1996, and that had spurred the press to burnish the atypicals story anew.
Play It Again, Sam
During the past fifteen years, most pharmaceutical research has focused on developing drugs that act narrowly on targeted receptors, with the thought that such “clean” drugs will have fewer side effects. Olanzapine, while a blockbuster financial success, ironically took antipsychotic drug development in the opposite direction. It, like clozapine, is a “dirty” drug. It acts on a broad range of receptors—dopaminergic, serotonergic, adrenergic, cholinergic, and histaminergic—and blocking any one of those receptors is known to cause an array of side effects. Blocking dopaminergic receptors leads to motor dysfunction. Blocking serotonergic receptors leads to sexual dysfunction, hypotension, and weight gain. Drugs that act on adrenergic receptors may cause hypotension, dizziness, tachycardia, and ejaculatory dysfunction. Anticholinergics may cause blurred vision, dry mouth, constipation, urinary retention, memory problems, drowsiness, fatigue, and erectile dysfunction. Blockade of histaminergic receptors can cause sedation and weight gain. The laundry list of possible side effects from a “dirty” drug like olanzapine is a long one. How this blockade of multiple receptors will play out in the human brain is also anybody’s guess. It’s a scientific crapshoot, but in the tale told to the public, this “dirty” aspect of olanzapine was transformed into a virtue. “Olanzapine,” the Associated Press reported, might be better than risperidone “because it appears to affect even more areas of the brain.”
37As was the case with risperidone, the FDA’s review of the trial data for olanzapine revealed just how far Eli Lilly had spun the trial results.
38 First, Leber and another FDA official, Paul Andreason, found that Eli Lilly’s studies were “biased against haloperidol” in much the same way that Janssen’s had been. Multiple doses of olanzapine were compared to one dose of haloperidol, and the drugs were not compared at “equieffective” doses. In addition, many of the patients in the trials had previously taken haloperidol and presumably had not responded well to it, and including such “bad responders,” the FDA officials noted, made it likely that results for haloperidol would be worse than normal, and thus help make olanzapine look superior by comparison. Concluded Leber: “The sample of patients used is an inappropriate choice” for comparison purposes. Second, he and Andreason determined that of Eli Lilly’s four well-controlled studies, only the smaller two—with a combined total of about 500 patients—provided any useful data related to olanzapine’s effectiveness versus placebo. In one of its two larger trials, involving 431 patients, Eli Lilly had compared three doses of olanzapine to haloperidol and to a low, nontherapeutic dose of olanzapine (which served as a placebo control), and Leber and Andreason concluded it was a “failed” study because there was no significant difference in the reduction of positive symptoms in any of the treatment groups at the end of six weeks. The other large study that the FDA found wanting was Eli Lilly’s large phase III trial, involving 1,996 patients. This was the study that Eli Lilly had used to make claims in the medical journals that olanzapine was superior to haloperidol, and also the one that led to newspaper stories about how olanzapine was a “potential breakthrough of tremendous magnitude.” However, both Leber and Andreason concluded that the study was “biased against haloperidol,” and they detailed specific methods that Eli Lilly had used to favor its drug. Furthermore, since the study didn’t include a placebo arm, it couldn’t show any efficacy data in that regard, either. Concluded Andreason: The study “is fundamentally flawed and provides little useful efficacy data.”
Olanzapine’s safety profile was also not as benign as the newspaper reports suggested. Of the 2,500 patients in the trials who received olanzapine, twenty died. Twelve killed themselves, and two of the remaining eight deaths, both from “aspiration pneumonia,” were seen by FDA reviewers as possibly causally related to olanzapine. Twenty-two percent of the olanzapine patients suffered a “serious” adverse event, compared to 18 percent of the haloperidol patients. Two-thirds of the olanzapine patients didn’t successfully complete the trials. More than one-fourth of the patients complained that the drug made them sleepy. Weight gain was a frequent problem, with olanzapine patients putting on nearly a pound a week in the short-term trials, and twenty-six pounds over the course of a year (for those who participated in the extension trials).
39 Other problems that showed up, with greater or lesser frequency, included Parkinson’s, akathisia, dystonia, hypotension, constipation, tachycardia, diabetic complications, seizures, increases in serum prolactin (which may cause leaking breasts and impotence and which raises the risk of breast cancer), liver abnormalities, and both leukopenia and neutropenia (white blood cell disorders). Leber, in his summation of the safety data, even warned that, given olanzapine’s broad action on multiple receptor types, “no one should be surprised if, upon marketing, events of all kinds and severity not previously identified are reported in association with olanzapine’s use.”
The third atypical to undergo the FDA’s review was AstraZeneca’s quetiapine, and once again, the FDA found plenty to criticize.
40 Four of the eight trials conducted by AstraZeneca were not considered by the FDA to provide any “meaningful” efficacy data. The other four studies, the FDA determined, showed that quetiapine was modestly superior to placebo for reducing positive symptoms but did not prove that quetiapine was superior to haloperidol in this regard. If anything, trial data suggested that haloperidol was more effective. Patients also clearly had difficulty staying on quetiapine. Eighty percent of the 2,162 quetiapine-treated patients dropped out of the trials, compared to 61 percent of the placebo patients and 42 percent of patients treated with standard neuroleptics. Common adverse events included weight gain, sedation, and somnolence; there were also reports of hypotension, tachycardia, seizures, leukopenia, neutropenia, neuroleptic malignant syndrome, liver abnormalities, and bone fractures caused by fainting spells.
Three atypicals reviewed by the FDA, and three times the FDA did not find any convincing evidence that they were superior to the old ones. Instead, FDA reviewers pointed out the ways in which Janssen and Eli Lilly had used biased trial designs to produce results that, when published in the science journals, created a story of superiority (and enabled them to sell their new drugs for ten to thirty times the price of the old neuroleptics). However, such criticism did not require the knowledge of an FDA expert. The methods used by drug companies to make their drugs look good in clinical trials have become so well known that various articles have appeared in medical journals cataloging them. The use of inappropriate doses of the standard drug is a favorite one; so is comparing multiple dosages of the experimental drug to one dose of the standard drug. Yet when the researchers who’d been paid by Janssen and Eli Lilly to conduct the trials reported their results (or put their names on papers written by the companies), they never discussed how the trials were biased by design. They never fessed up, as it were, and their silence spoke volumes about the influence of money.
41Every once in a while, a researcher has stepped forward to poke holes in the atypicals story. A team of English scientists, led by John Geddes at the University of Oxford, reviewed results from fifty-two studies, involving 12,649 patients, and concluded in 2000, “there is no clear evidence that atypical antipsychotics are more effective or are better tolerated than conventional antipsychotics.” The most common ruse that had been employed to make the drugs look better, Geddes found, was the use of “excessive doses of the comparator drug.”
42 An embarrassing, yet revealing, squabble also briefly erupted in the medical journals over the relative merits of risperidone and olanzapine, with Janssen complaining that Eli Lilly’s studies were biased in ways that—surprise, surprise—favored olanzapine. Then Janssen funded a comparative trial, and that trial concluded risperidone was superior to olanzapine. It all made for a tawdry spectacle, and finally a truce of sorts was called. Several studies concluded that it was impossible to say one or the other was better; they were different drugs, with different risk-benefit profiles, and perhaps it was best to leave it at that.
43Indeed, why would either company want to stir the pot? Both risperidone and olanzapine had quickly become astonishing financial successes. Total annual U.S. sales of antipsychotic medications roared past $2.5 billion in 2000. Worldwide sales of olanzapine were projected to hit $3 billion in 2001. As
Forbes.comcrowed on January 25, 2000: “Zyprexa (olanzapine) and its main competitor, Risperdal, can split a lot of market between them. Since they are both expensive drugs, they will fill company coffers.” Praise from newspaper and magazine writers continued to flow as well. American science, the
Washington Post told its readers on July 29, 1998, had developed several “breakthrough” medications that “have proven to be much more effective than older medications in helping schizophrenics lead functional lives and with far fewer side effects.”
44 Money, glowing press—this was a good-news story all around, and finally the National Alliance for the Mentally Ill put it together into its full mythic glory. In 1999, it published a book titled
Breakthroughs in Antipsychotic Medications and inside the front cover were framed, color photos of the new wonder pills. The NAMI authors wrote: “Conventional antipsychotics all do about the same job in the brain. They all correct brain chemistry by working on the dopamine systems in the brain . . . the newer medications seem to do a better job of balancing all of the brain chemicals, including dopamine and serotonin . . . give the new medication plenty of time to do a good job!”
45Like tonics once pitched from the backs of wooden wagons, atypicals could apparently transform the troubled mind into one awash with chemicals operating in perfect harmony.
r
A State of Confusion
One of the saddest aspects of this “research” process, and the storytelling that accompanied it, is how it has left everyone in the dark about the real merits of the atypicals. There are certainly many anecdotal accounts of patients who are doing well on them, and so perhaps in some ways they truly are superior to the old drugs. Yet anecdotes do not make for good science, and the testing process was such that little can be known for certain. Are the atypicals, for instance, even any better than placebo at knocking down psychosis? If patients suffering a first episode of psychosis were separated into two groups and one group were given a placebo and the other olanzapine, what would the results be at the end of six weeks? Or perhaps more to the point, if a sedative were compared to olanzapine or risperidone, what would be the results? No one knows. As for their comparative merits versus standard neuroleptics—again, who knows? In fact, it is actually quite easy to envision a scenario in which haloperidol would be the drug being hailed today as the new wonder medication and olanzapine would be the drug being carted off to the trash heap. All one has to do is imagine their coming to market in reverse order, such that in 1995 olanzapine had been the “old” drug and haloperidol the “experimental” drug. In that case, multiple doses of haloperidol would have been compared to a single, high dose of olanzapine—in other words, the trials would have been designed to favor haloperidol—and researchers would likely have been able to announce that haloperidol appeared superior in several ways and didn’t cause the troublesome side effects associated with olanzapine, like weight gain and sleepiness. The researchers would even have been able to offer a good explanation for why haloperidol had a superior side-effect profile. Whereas olanzapine was a “dirty” drug that acted on multiple neurotransmitters, haloperidol was a clean drug that more precisely honed in on a very specific receptor, the D2 receptor. Modern science had simply produced a more refined drug.
The biggest question, of course, is how the new drugs will affect patients’ lives over longer periods of time. The old drugs—as was shown by the WHO studies—led to an increase in chronic illness and limited the possibility of recovery. They were harmful over the long run. Will the new drugs be equally harmful? Less harmful? Or, in fact, helpful over the long term? No one knows. However, there are already plenty of reasons to worry about their long-term effects. The atypicals—just like standard neuroleptics—cause an abnormal increase in D
2 receptors.
46 And while certain side effects, such as the risk of tardive dyskinesia, may be reduced with the atypicals, they also bring their own set of new problems. For instance, there have been reports that olanzapine can trigger obsessive compulsive disorder, with researchers speculating that this may be due to the drug’s hindrance of serotonin activity. Then there are the metabolic problems associated with olanzapine: Just how great is the increased risk of poor health with this drug because of weight gain? Some patients are putting on sixty, seventy, eighty pounds. Reports are also filtering into the medical literature about how olanzapine can dramatically increase triglyceride and blood sugar levels, which are risk factors for cardiovascular disease and diabetes. Is this a drug that will lead to early death for many?
What makes this question all the more pressing is that there remains today great uncertainty over what schizophrenia is, or isn’t. The public has been led to think of schizophrenia as a discrete disorder, one characterized by abnormal brain chemistry. In truth, the biological underpinnings of madness remain as mysterious as ever. In fact, schizophrenia is a diagnosis applied to people who behave or think strangely in a variety of
different ways. Some people so diagnosed are withdrawn. Some are manic. Some act very “silly.” Others are paranoid. In some people, the crazy behaviors appear gradually. In others, psychosis descends abruptly. Any well-reasoned concept of “madness” would require teasing apart all these different types and would also require an understanding of how outcomes for the different types—in the absence of neuroleptics—might differ. Yet there is little research in American circles devoted to seeing this more complex picture. It is a shortcoming so pronounced that it caused Nancy Andreasen, editor of the
American Journal of Pyschiatry , to burst forth in 1998 with a remarkable confession: “Someday in the twenty-first century, after the human genome and the human brain have been mapped, someone may need to organize a reverse Marshall Plan so that the Europeans can save American science by helping us figure out who really has schizophrenia or what schizophrenia really is.”
47Two hundred years after Benjamin Rush founded American psychiatry, and still the problem remains as confounding as ever. What is madness? Where do you draw the line separating the normal mind from the crazy one? The drawing of that line is a profound event for a society, and a life-altering event for those diagnosed as ill. And it is here that one can see, once again, how the storytelling that brought the atypicals to market is exacting a great cost. With the new drugs presented to the public as wonderfully safe, American psychiatrists are inviting an ever-greater number of patients into the madness tent. They are prescribing atypicals for a wide range of emotional and behavioral disorders, and even for disruptive children, including—as the
Miami Herald reported—toddlers only two years old. Yale University psychiatrists are even giving olanzapine to teenagers who are not even ill but simply said to be at risk of developing schizophrenia, either because they have siblings diagnosed with the disorder or have begun behaving in troubling ways.
s Researchers, the
Wall Street Journal reported, “hope the new drugs will intervene in the brain-damaging process that leads to schizophrenia, even though they don’t know for sure what that process is.”
48That is the story in American mad medicine today: The line between the sane and the not-so-sane is now being drawn in such a way that two-year-olds can be put on “antipsychotic” medications, and some researchers are busily speculating that their wonderful new drugs can stop an unknown brain-damaging process in people who aren’t yet ill. Madness is clearly afoot in American psychiatry, and bad science—as so often has been the case in mad medicine—has helped it on its way.
