Side Effects: A Prosecutor, a Whistleblower, and a Bestselling Antidepressant on Trial

by Alison Bass

  Garnier also dismissed the "smoking gun" memo from 1998, which laid out SmithKline's plans not to disclose the negative results about Paxil. He said the memo was contrary to company policy. "Can a company control the millions of memoranda that are written by, in our case, 110,000 people?" the CEO was quoted as saying. "What are the odds that stupid memos were written? What are the odds that memos asking the company to do things against company policy will be written? The odds are 100 percent"

  Garnier's combative comments alarmed Firestein. They signaled that the London-based pharmaceutical company meant to fight the case in court, which could drag it out for years. And while the New York State Psychiatric Institute had joined forces with Spitzer the day the lawsuit was filed, the silence from other medical groups and federal regulators was conspicuous. Firestein couldn't help asking herself, had she and her colleagues so badly misjudged the willingness of the medical community to take a stand for full disclosure?

  THE SAME WEEK, almost two thousand miles away in Pflugerville, Texas, seventeen-year-old Tonya Brooks was rushed into a hospital emergency room, bleeding from that huge, gaping hole in her leg. The self-inflicted wound would require three layers of stitches to suture it shut, and while Tonya lay splayed on the gurney, a parade of people kept coming in and asking her, "Did your father abuse you? Did he beat you up?" No, she kept telling them. No.

  Later that day, she was admitted to Seton Shoal Creek, a mental hospital for children and adolescents in Austin. The doctor diagnosed her as bipolar; according to Tonya, he labeled everyone bipolar. "Ninety percent of the kids I met there, there's nothing wrong with them," she later told me. "They just wanted attention. They act out to get attention." Tonya agreed with everything the doctor said just so she could get out of there. But she liked the nurses; they were very kind. She confided in one nurse that she liked to color, so the nurse gave her a coloring book and told her that whenever she started to feel bad, she could color in pictures to take her mind off things. Tonya Brooks stayed at Shoal Creek for four days, and on June 7, 2004, she went home.

  FOUR DAYS LATER, the Food and Drug Administration reacted to the news of the New York AG's lawsuit in its usual oblique fashion. On Friday, June 11, the federal agency sent an "untitled" reprimand (one step below an official warning) to GlaxoSmithKline, rebuking the company for running "false and misleading" television commercials about Paxil. The FDA letter said that the sixty-second TV ad "broadens the indication for Paxil CR beyond the ... more serious condition of social anxiety disorder to people experiencing more ordinary degrees of anxiety, fear and self-consciousness in social or work situations, while also minimizing the serious risks associated with the drug" The federal agency asked the company to stop showing the ads. The press reports implied that the FDA's action had been prompted by the New York State attorney general's lawsuit, especially since the ads in question had first aired months earlier. It was that same series of TV commercials that had compelled Tonya Brooks in Texas to ask her doctor to prescribe Paxil.

  LATE IN THE AFTERNOON on June 14, 2004, in a clear attempt at damage control, GlaxoSmithKline posted all nine of the pediatric Paxil studies (including the three randomized clinical trials and other research results) on its Web site. The company was doing this "in the interest of transparency," a Glaxo spokesperson told the Wall Street Journal. Spitzer pronounced the posting "a positive first step toward changing a dangerous industry practice." But the AG's office wanted more than the quick posting of a few studies related to one drug; Spitzer and his team had decided early on that as a condition of any settlement, Glaxo would have to launch a comprehensive, publicly available registry that contained the results of all its clinical trials after a certain date, negative as well as positive.

  "They had to understand that we wouldn't settle absent their changing the behavior, their refusal to disclose," Spitzer later explained. He was eager to expand his action against Glaxo into a wake-up call for other drug companies. GlaxoSmithKline, after all, was not the only SSRI maker who was guilty of selective data disclosure. At the end of 2003, British regulators had made it clear that while the published results of randomized trials on several commonly prescribed antidepressants such as Paxil, Zoloft, Effexor, Luvox, and Remeron described these drugs as safe and effective in children, the actual data from their clinical trials showed quite the opposite.

  "There is probably no other area of medicine in which the academic literature is so at odds with the raw data," declared David Healy in an open letter to Mr. Peter J. Pitts, the FDA's associate commissioner for external relations, earlier that year. The AG's office had arrived at much the same conclusion during its year-long investigation. Much as Spitzer's investment bureau had parlayed a big win against Merrill Lynch into broad-reaching changes on Wall Street, his consumer team now hoped to use the fulcrum of a lawsuit aimed at one drugmaker to reform the behavior of the entire industry.

  That was the game plan, anyway.

  Finally, on June 15, the American Medical Association, the powerful trade organization for U.S. doctors, jumped into the fray. At its annual meeting in Chicago, the AMA declared its support for full disclosure of medical studies and voted to ask the U.S. Department of Health and Human Services to create a public registry of all clinical trials and their outcomes.

  In response to the AMA's declaration, Dr. Jeffrey Drazen, editor of the New England Journal of Medicine, told the Boston Globe that journal editors were considering a similar proposal. By a fortunate coincidence, twelve editors of the most prestigious medical journals in the world had been meeting in Croatia the week the AG's office filed its lawsuit against GlaxoSmithKline. This group, officially known as the International Committee of Medical Journal Editors (ICMJE), had been trying for years to achieve full disclosure of medical research results. These editors were coming at the issue from a slightly different vantage point than Firestein and her colleagues, but with the same end in mind. They wanted scientists to more fully disclose the original objectives and design of their research before they began the actual studies and submitted them for publication. That way, journal editors would know what exactly the original intent and focus of the research was.

  In all too many cases, medical researchers were making last-minute changes in the outcome measures of their studies-changes that could significantly skew the results. A survey published just the previous month in the Journal of the American Medical Association found that at least one primary outcome measure was changed in 62 percent of all clinical trials surveyed, leading to "biased and inconsistent" study outcomes. The JAMA survey also found that 86 percent of respondents didn't report the changes they'd made, making it extremely difficult for journal editors to catch such manipulations.

  "Of the twelve editors sitting around the table in Croatia, we'd all had the experience of someone pulling a fast one on us," said NEJM's Dr. Jeffrey Drazen, one of the ICMJE members leading this effort. "All the editors had to play this hide-and-seek game about what was really being studied."

  Emboldened by the New York lawsuit, Drazen and the other editors meeting in Croatia decided to get tough. There and then, they agreed to require that medical researchers register the key elements of their study protocols, including outcome measures and the size of the study groups, on one of several public research Web sites. If researchers (or their pharmaceutical sponsors) didn't register up front, they couldn't expect to be published by a prestigious medical journal. In three months' time, all twelve journals whose editors belonged to the ICMJE would publish identical editorials laying out their specific demands.

  Back at the AG's office, however, Firestein and her colleagues were beginning to despair of a quick settlement with G1axoSmithKline. They had been rattled by an editorial in the Wall Street Journal that viciously attacked Spitzer as a lawmaker run amok. The June 21 editorial went so far as to accuse the AG's lawsuit of threatening to damage public health. "All in all, there's something dangerous about a prosecutor making health-policy decisions for which he can't b
e held responsible," the Journal's editorial writers declaimed. "Mr. Spitzer has to face the voters every four years, but will he take the blame for depressed kids who suffer because their doctors are scared out of prescribing Paxil?"

  The editorial concluded that, "Whatever good Mr. Spitzer has done by drawing attention to Wall Street's sins, it's clear his good press has gone to his head. He's now firing press releases and lawsuits in so many directions that he's hitting and harming innocent bystanders."

  For months afterward, Spitzer would carry that editorial around with him, showing it at speeches he gave to health care groups. "It says in here that the system is fine the way it is," he would say, and shake his head. "How can you defend a status quo that is so clearly broken?"

  But Firestein couldn't help wondering how much the Wall Street Journal's salvo had shaken Spitzer privately. The Journal, after all, was the most prestigious business newspaper in the country, perhaps the world. Such an ad hominem attack would certainly have stung her. While the attorney general got the lion's share of credit for his staff's victories, he was also taking all the hits first from Wall Street and now from the pharmaceutical industry. Firestein had to admit she was kind of relieved that she wasn't the one on the firing line. She had a feeling her skin would not be as thick as Eliot Spitzer's.

  ose Firestein was perspiring as she waited to pass through security. It was a humid, overcast day in late July and she was wearing one of the two tissue-wool suits she owned. She needed to dress in layers at work even in the summer because her office was freezing. Cold enough to make ice cream, she often said. But the weather outside was a different story. By the time Firestein walked the eleven blocks to the new Daniel Patrick Moynihan United States Courthouse on Pearl Street, she felt as if she had been baked in a 350-degree oven.

  This was Firestein's first visit to the new courthouse, which had been built directly behind the old New York State Supreme Court building on Foley Square. Although the white-marbled interior of the Moynihan courthouse was airy and filled with light, Firestein couldn't help thinking that the new edifice, as costly as it must have been to build, did not have the moral majesty of the old courthouse, with its coffered ceiling beautifully sculpted with plaster rosettes.

  Firestein was here because the attorneys for GlaxoSmithKline had preemptively removed the attorney general's consumer fraud case against the drug giant to federal court. One day, Joseph Sedwick Sollers (Wick) had told her the company was interested in talking about settlement; then he and his cocounsel had gone ahead and moved the case out of the New York State courts. The case was now slated to be heard by Judge Miriam Goldman Cedarbaum of the Southern District Court of New York. The judge had called a preliminary conference for this afternoon.

  Firestein and the two GlaxoSmithKline attorneys-Wick and Dwight Davis-arrived early and cooled their heels outside the courtroom until the lunch recess was over and the chamber was unlocked. While they waited, Wick mentioned that Glaxo might be ready to talk again about settling the case. But Firestein had heard that song before. She wasn't about to get her hopes up. This was not the first time she had tangled with GlaxoSmithKline -they were one of the litigants in the wholesale drug pricing case-and she knew the company, which had annual earnings of $9.32 billion, could be a formidable opponent. In the pricing case, the AG's office had accused three pharmaceuticals of wildly inflating the average wholesale price of their drugs. Millions of dollars in restitution and fines were on the table, and Glaxo and the two other defendants -Pharmacia (later acquired by Pfizer) and Aventis (which later merged with Sanofi)-were aggressively contesting the complaint. It was now in its second year, with no end in sight.

  JUDGE MIRIAM CEDARBAUM presided over a wood-paneled courtroom on the fourteenth floor of the new courthouse, its royal blue window drapes decorated in stars and wreaths. The carpeting was also blue, with a matching pattern of stars and wreaths. Behind the judge's chair hung a black metallic etching of the great seal, the familiar bald eagle holding the olive branch in one talon and spears in the other. The black-robed judge herself wore an expressive face and long salt-andpepper hair parted in the middle. When listening to an oral argument, she would rest her right hand under her chin with an index finger extended just below her lip. Her eyes were stern, and one eyebrow seemed to be in a perpetual upside-down V.

  The first thing the judge wanted to know was how the AG's of fice planned to prove intent. How did Ms. Firestein know that GlaxoSmithKline intended to deceive consumers when it withheld two studies showing that Paxil was no more effective than placebo in pediatric depression?

  Firestein had spent considerable time studying New York's consumer fraud statute, and not just for this case. A few weeks after her office filed the lawsuit against GlaxoSmithKline, the New York Times and the Wall Street Journal reported that another large pharmaceutical company, Forest Labs, had also neglected to publish two negative studies about its SSRI, Celexa. Yet the same month-June 2004 -the American Journal of Psychiatry had published a study concluding that Celexa did reduce symptoms of depression in children. Was this another example of selective disclosure? Firestein's sources seemed to think so. She had wasted no time firing off a letter to Forest Labs requesting "any information that may exist with respect to off-label clinical trials or promotion" of Celexa, which was then the fourth most commonly prescribed antidepressant for children and adolescents.

  "We don't have to prove intent," Firestein told the judge, feeling very much on the hot seat as she stood alone behind the long plaintiff's table. "As you're well aware, under common law, you do have to show intent to defraud. But under the State of New York's consumer fraud statute, you don't. You might want to look at People v. General Electric it has a very clean explanation of the statute"

  Judge Cedarbaum looked over at the two Glaxo attorneys. They nodded in agreement. This was clearly not the issue on which they planned to stake their ground today. The judge then turned to the question they did care about: why had they removed the case to federal court?

  Standing behind the defendant's table, Dwight Davis, an elegantly suited Wick by his side, explained that the AG's complaint was in fact preempted by federal law, indeed by Congress, when it had established the U.S. Food and Drug Administration as the nation's regulatory agency in charge of drug safety and approval. Under federal law, he said, only the FDA could determine whether the antidepressant Paxil was safe and effective for use in children.

  "Congress has fully occupied the field here, and there's nothing left for the state legislature to do," Davis said.

  Judge Cedarbaum quirked an eyebrow at Firestein, who had remained standing.

  "We have a number of responses to that, Your Honor, among them being the fact that the FDA does not regulate the practice of medicine," Firestein began. "Within New York, as in other states, the regulation of the practice of medicine is solely the responsibility of the state"

  Firestein argued that by selectively publishing only one study purporting to show Paxil's efficacy and concealing two other studies that showed no efficacy, GlaxoSmithKline had prevented the state's physicians from "properly and independently exercising their professional judgment on behalf of their child and adolescent patients."

  She drew a quick breath and continued.

  "But Your Honor, the main point I'd like to argue here is that the defendant seems to be asserting preemption as a defense," she said. The Glaxo attorneys were essentially claiming that the FDA's jurisdiction in regulating drugs preempted any regulatory role by New York State. Firestein argued that they were using this preemptive argument as a defense against the AG's lawsuit. If so, they had made a procedural mistake. There were many reasons a case could be removed to federal court, but a defense wasn't one of them.

  "As Your Honor knows, a defense based on federal law is not a basis for removal," Firestein said. She cited a recent Supreme Court case affirming that a defense could not be used as a basis to remove a case from state to federal court.

  Judge Cedarbau
m nodded.

  "You're right," she said. "A defense is not the basis for a change in jurisdiction."

  The judge turned to the Glaxo attorneys. Firestein could hear the two men exchange words; they did not sound happy. Dwight began to argue that he was not using a defense as the basis for removing the AG's lawsuit to federal court, but Judge Cedarbaum waved her hand as if to say it didn't matter.

  "Nobody's filed a remand yet," she said. She meant that since the AG's office hadn't yet filed a motion to return the case to state court, the attorneys on both sides would have plenty of time to argue the grounds for removing the case to federal court.

  A LITTLE LATER, it was Firestein's turn to squirm. She had already made it clear to the judge that she intended to file a motion to remand the case to state court. But under the court's regulations, she had only thirty days from the date GlaxoSmithKline had removed the case in which to file the motion, and she wasn't sure she could meet that deadline, especially if Glaxo decided they wanted to talk about settling the case.

  "Will I waive my objections to removal if I don't file within thirty days?" she asked the judge.

  "No, of course not," Judge Cedarbaum said impatiently. "You never waive objection to the lack of federal jurisdiction."

  "But sometimes you can waive your objection to procedural defects in the notice," Firestein persisted. "I just want to make sure I'm not waiving my position."

  Judge Cedarbaum narrowed her eyes and frowned.

  "That's ridiculous!" she said. "If I tell you you can file your papers beyond a certain date, then that's when you can file your papers!"

  Firestein flushed and looked down at the table. This was not the first time she had incurred a magistrate's wrath. She knew that she came across as overly aggressive sometimes, but she couldn't help herself. Even before she lost much of her sight, a handicap that prevented her from reading other people's body language, she'd had this tendency to push people's buttons. It came, she thought, from being the third child in a family of four children with two very busy parents. If she hadn't been persistent, no one in that tumultuous household would ever have heard her. But sometimes, she knew, she went too far.