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

by Alison Bass

  There were times those first few months, however, when Firestein didn't feel so confident. She knew next to nothing about the byzantine world of health care reimbursements. It took her months to understand the mechanisms by which pharmaceutical companies were inflating their prices. Prosecutors for several other states and the U.S. Department of Justice had already sued a number of pharmaceutical firms for inflating the average wholesale prices of their products. In fact, in 2001, TAP Pharmaceutical Products Inc. paid a record-breaking $ 875 million to settle civil and criminal charges of Medicaid and Medicare fraud in the marketing and pricing of its prostate cancer drug, Lupron.

  But this was new territory for the New York AG's office, and Firestein spent hours reading up on New York case law, working late into the night to understand why and how the whole scheme worked in her state. The AWP case, she would later recall, was "ridiculously complicated, with so many permutations that there were times you just wanted to put your head in the oven"

  Tom Conway sensed her frustration. When she accepted the job, Conway had told her that they were going to have lots of fun together, "a great time." Now, every few weeks, he would stop by and ask her with deadpan earnestness, "Is this the way I promised you it would be? Did we tell you the truth in advertising?"

  Firestein would reply in a tone of mock gruffness, "No! Where's the fun?" And Conway would laugh heartily, delighted that Firestein had a sense of humor about it all.

  FIRESTEIN WAS so preoccupied with figuring out the complexities of the drug reimbursement system that she didn't have much time to dwell on the unhappy fate of Bonnie L. v. Jeb Bush et al. A few months before she had left Children's Rights for the AG's office, U.S. District Court judge Federico Moreno in Miami had dismissed the class action suit on grounds that it was a case that should be heard in the Florida state courts. In doing so, Moreno had ignored years of legal precedent, other judges having consistently found that the federal court had a right, indeed an obligation, to intervene on behalf of abused and neglected foster children. Firestein was disappointed by Moreno's decision, but she tried to comfort herself with the knowledge that other jurists were continuing the fight. Karen Gievers had already launched an appeal of Moreno's decision, and other activist lawyers in Florida were initiating similar actions in the state courts. Firestein hoped the plight of Florida's foster children would not be forgotten, but she knew how stacked the system could be against a handful of well-meaning litigators who were trying to bring about large-scale change through the courts. She had learned from hard experience not to let such legal defeats haunt her. "You can't take it personally," she said. "It's a function of the adversarial system: sometimes you lose. You say some nasty words and then you get on with it."

  FIRESTEIN HAD BEEN toiling on the AWP case for eight or nine months when Conway informed her that Eliot Spitzer was keen on investigating the profit-driven labyrinth of drug research. As Spitzer himself told me, he and Joe Baker, as his health care czar, had been discussing whether some of the financial conflicts his investment division had uncovered on Wall Street existed to some degree in the pharmaceutical industry as well.

  "It began with our being tempted and intrigued by the notion that there was cross-ownership here," Spitzer said, citing news articles about the pharmaceutical companies' having equity in some of the private companies that tested the drugs they had in development. "So we wondered: Is there something that goes amiss here? Do they play games with the numbers?"

  Spitzer's desire to probe the unexamined corners of drug research coincided with a conversation Firestein herself had had with Shirley Stark, her immediate supervisor. Firestein had mentioned her desire to look into the off-label use of psychoactive medications in children. From her work with Children's Rights, Firestein knew that the FDA had not approved such drugs for use in children and adolescents; Prozac was the only antidepressant drug okayed for pediatric use, and that green light had come only a few months earlier-in January 2003. Yet potent medications like Paxil, Zoloft, Celexa, Effexor, and Risperdal were widely prescribed off-label for the under-eighteen age group. In many cases the drugs were helpful in treating symptoms of depression, anxiety, and (in the case of Risperdal) bipolar disorder and psychosis. But in many situations, they were dispensed too readily. Busy primary care doctors and pediatricians were sending pediatric patients home with samples from the drug companies and failing to monitor their reactions to the drugs. Serious side effects were common, as Firestein had discovered in Bonnie L. While she understood that it was legal for doctors to prescribe such off-label use, it seemed to her as though medical professionals were making judgments based on very little information about the drugs' impact on children.

  To Firestein, the FDA's general attitude toward off-label drug usage seemed disturbingly laissez-faire. Why this was so, Firestein couldn't really say. She knew, of course, that some consumer groups believed that the FDA had become beholden to the very industry it was supposed to regulate. Consumer advocates, including Dr. Sidney Wolfe and his colleagues at the Public Citizen in Washington, D.C., contended that the agency's increasing reliance on hundreds of millions of dollars in user fees from the drug industry had exerted a corrosive effect on the FDA's ability to monitor the safety of already approved drugs and to review new drugs objectively. FDA medical officers surveyed by the Public Citizen in 1998 cited growing pressure on them to approve new drugs, and they identified twenty-seven new drugs that they believed should not have been approved but were anyway. "Inappropriate pressure from Congress, the drug companies and senior FDA employees create an atmosphere in which the likelihood of drug approval is maximized," concluded the Public Citizen survey.

  The FDA, of course, was not the only beneficiary of the pharmaceutical industry's largesse. Its tentacles reached deep into the corridors of power in Washington. Between 1998 and 2006, the drug companies spent a total of $1.2 billion on lobbying and political contributions in the United States, according to the Center for Responsive Politics, a nonprofit, nonpartisan research group that tracks money in politics. In 2002, the pharmaceutical industry ranked tenth among the top industries in political contributions and seventh in the amount of money it spent on lobbying members of Congress and others.

  From her cramped third-floor office in Lower Manhattan, there wasn't much Firestein could do about the pharmaceutical industry's favored status in Washington. What concerned her, as an assistant attorney general, was whether the companies selling drugs in the state of New York were leveling with physicians who were writing all those off-label prescriptions for children and adolescents.

  "Mind if I look into this once I get my head above water with AWP?" Firestein asked Stark. Stark was only too glad to see her take the initiative. In working with Firestein on the AWP case, Stark had found her to be very bright and creative in her approach to the law. The two women had a lot in common. They were both Jewish and committed to social change, having come of age in the antiestablishment 1960s. Yet both dressed conservatively, favoring suits and pearls. Stark, then forty-nine, was slender, with large green eyes, and teeth that protruded slightly when she smiled. She had dark brown hair cut in a pageboy style similar to the cut of Firestein's more unruly, platinum blond locks. But Stark was more reserved than Firestein, who had strong opinions and was not afraid to get into a verbal dustup with their boss, Tom Conway. On several occasions, Stark, more conciliatory by nature, found herself having to mediate between her two equally stubborn colleagues.

  Stark, of course, had noticed Firestein's vision difficulties from the outset but hadn't said anything about them. A few months after Firestein joined the AG's office, Stark herself had to undergo eye surgery for a weak muscle in one eye. When she returned to work, Firestein confided in her about her own eye problems, and the two women found one more reason to bond.

  In early 2003, Shirley Stark was in the midst of a massive, multistate case against Pfizer for illegally marketing an epilepsy drug called Neurontin for treating conditions for which it had not be
en approved by the FDA. Neurontin had been approved in 1994 to treat epilepsy for a relatively narrow usage: as an add-on when other drugs failed to control seizures. But there wasn't much money in that, and ParkeDavis, which produced Neurontin, wanted to expand the drug's market. (Parke-Davis was a subsidiary of Warner-Lambert, which itself would later be swallowed up by Pfizer.) Since Neurontin's patent was due to expire in 1998 (it was eventually extended to 2000), the company didn't feel it had time to do the kind of extensive clinical trials that might win FDA approval for additional uses. So Parke-Davis, according to a whistleblower's lawsuit filed in 1996, devised a plan to promote the drug for off-label uses, primarily by paying academic experts to put their names on research papers prepared by the drug company. The papers were written with the express purpose of showing that Neurontin worked for common conditions such as pain and anxiety. Although the research was less than authentic, the marketing strategy worked: Neurontin emerged a blockbuster, with sales of $2.7 billion in 2003 alone. But federal prosecutors in Boston eventually filed a brief in support of the whistleblower's suit, and by 2003, forty-eight states, including New York, had launched parallel actions against Pfizer (which by then owned Parke-Davis and Warner-Lambert).

  Swamped by the Neurontin case, Stark was only too happy to have Firestein look into other cases of off-label marketing. Eliot Spitzer's conversation with Joe Baker sealed the deal, and by the spring of 2003, the newest member of the AG's consumer protection bureau had turned her attention to the drug industry's widespread practice of making billions of dollars from the off-label marketing of psychotropic drugs.

  SOME MONTHS LATER, in July 2003, Rose Firestein was browsing the official Web site of the Food and Drug Administration when she suddenly reared back in surprise. She blinked and looked again. Buried in the third paragraph of an FDA press release, dated June 19, 2003, was this sentence: "Three well-controlled trials in pediatric patients with MDD [major depressive disorder] failed to show that [Paxil] was more effective than placebo." The reference to these negative findings about Paxil mystified Firestein. Why would doctors so readily prescribe Paxil (making it by 2003 the second most widely prescribed antidepressant for those under eighteen years old) if the results of drug tests were uniformly negative? What was going on here? Was there something that America's physicians weren't being told?

  Up to that point, Firestein had been stuck on the idea that the companies selling psychoactive drugs were somehow misleading doctors with their marketing material. That's what Parke-Davis/Warner-Lambert had been caught doing with Neurontin, and Firestein thought that perhaps a similar phenomenon could explain the widespread prescribing of SSRI antidepressants for children. At one point she believed she had found evidence that GlaxoSmithKline, in its advertisements for Paxil in professional medical journals, omitted the required boilerplate language acknowledging that Paxil had not been approved for use in children and adolescents. She also noticed that a number of drug companies, including Glaxo, claimed on their Web sites that the SSRIs were safe and effective in "people," making no distinction between adults (for whom the antidepressants had been approved) and children (for whom they had not). Firestein knew from her years of working with psychiatrists at Legal Aid and Children's Rights that children and adolescents reacted differently than adults to psychotropic medication. Research had shown, for instance, that children aged ten to fourteen years had a much higher rate than adults of becoming manic after being treated with an SSRI antidepressant. And a review of data from the drug trials would soon reveal that children and adolescents who were treated with SSRIs were two to three times more likely than adults to experience the kind of agitation or hostility that frequently accompany suicidal behavior. Researchers weren't exactly sure why pediatric patients were more sensitive to drugs like Prozac, Zoloft, and Paxil; all they knew was that they were.

  Firestein would later discover that she was wrong about Glaxo's pediatric warning. The required fine print had been folded into the warning for pregnant women on Paxil advertisements, and she had simply missed it. But by then she had stumbled upon the June 2003 FDA press release and was starting to wonder if the drug giant was engaged in something more than marketing shenanigans.

  Reading the FDA Talk Paper on her computer, Firestein learned some surprising news: nine days before the FDA made its own public announcement, the British health authorities had issued a warning that Seroxat (the brand name for Paxil in the United Kingdom) should not be used to treat children and adolescents for depression. The regulatory agency based this advisory on the startling conclusion that "there is an increase in the rate of self harm and potentially suicidal behavior in this age group when paroxetine [the generic name for Paxil] is used for depressive illness." In its Talk Paper, the FDA announced that it too was reviewing reports of a possible increased risk of suicidal thinking and suicide attempts in minors treated with Paxil. "Although [the agency] has not completed its evaluation of the new safety data," the release said, "FDA is recommending that Paxil not be used in children and adolescents for the treatment of major depressive disorder."

  This was news to Firestein-and, she suspected, to many doctors and consumers as well. Although the FDA had posted the release on its public Web site, the fact that the agency in charge of regulating America's health was taking a second look at a widely prescribed antidepressant had not yet been reported in the mainstream press or in any medical journals. Most physicians on this side of the Atlantic knew nothing about the British advisory against using Paxil in children and adolescents.

  But what really startled Firestein was the FDA's conclusion that there was currently no evidence that Paxil was effective for children and adolescents with depression. This was very strange. According to what had been reported in the medical literature and at conferences, Paxil not only reduced symptoms of depression and anxiety but was a safe and well-tolerated medication in children and adolescents, and many primary care physicians and psychiatrists swore by the drug. The medical community's confidence in prescribing the SSRI had made it into a blockbuster: in 2002, GlaxoSmithKline earned $55 million in Paxil sales to children and adolescents alone.

  So why was the FDA now challenging the conventional wisdom? Was it possible that the maker of Paxil had in some way misled physicians and consumers about the drug's safety and effectiveness?

  Firestein sensed she was onto something. Feverish with excitement, she began searching the medical Web site PubMed for published studies about Paxil in minors. She soon came across a study titled "Efficacy of Paroxetine in the Treatment of Adolescent Major Depression: A Randomized, Clinical Trial." The study had been published in July 2001 in the respected Journal of the American Academy of Child and Adolescent Psychiatry (JAACAP). It was the only published randomized clinical trial she could find on Paxil in this age group, and its abstract concluded: "Paroxetine is generally well tolerated and effective for major depression in adolescents."

  Its lead author was Martin B. Keller, MD.

  Downloading the study onto her computer, Firestein read the full text of the report and then read it again. And again. She grew increasingly baffled. It appeared as if the positive conclusion in the abstract simply did not correspond to the actual data. The study, for instance, said that there were two primary outcome measures by which the researchers planned to gauge the effectiveness of the drug in treating depression. First, patients in the study had to have a 50 percent reduction in their symptoms on the Hamilton depression (HAM-D) scale or a score of 8 or less on the HAM-D scale for the drug to be considered effective. Second, participants had to show a significant change in their total HAM-D score between the beginning of the study and the end. The protocol for the study also included five other secondary end points against which to measure whether the patients' depression had improved. But according to Firestein's reading of the results, Paxil had not performed significantly better than the sugar pill on either of the two primary measures. Nor had it "separated significantly from placebo" (in scientific parlanc
e) on two out of the five secondary measures listed in the original protocol. How, then, could the authors have concluded that Paxil was significantly more effective than placebo in treating depression in adolescents?

  Firestein was stumped by yet another inconsistency: The abstract and the paper itself had concluded that Paxil was generally safe and well tolerated in adolescents. Yet according to Firestein's reading of the data, Paxil was significantly more likely to cause serious adverse events (SAES) than the placebo or the older antidepressant it was being compared with. In the guidelines for the study, an SAE was defined as "any event that is fatal, life-threatening, disabling, incapacitating, or results in hospitalization." Eleven of the ninety-three patients taking Paxil developed SAEs. That contrasted with only two SAEs in the eightyseven patients taking placebo and five in the ninety-five patients taking imipramine, a tricylic antidepressant that had been around for decades. Very weird, Firestein thought. If anything, there should be more serious adverse events among patients taking the older tricyclic than among those taking the newer SSRI. After all, the drug companies had based their highly successful marketing campaigns on the argument that SSRIs had fewer serious side effects than the older antidepressants. What was going on here?

  Reading further, Firestein discovered that of the eleven patients with serious adverse events in the Paxil group, six were listed as having had "emotional lability (e.g. suicidal ideation/gestures)"; two, as having had "conduct problems or hostility"; two, "worsening of depression"; and one, a bad headache. In all, 10 percent of the ninety-three patients taking Paxil had to be prematurely withdrawn from the study because their reactions seemed to have been so severe. Yet the study itself concluded that "the findings of this study provide evidence of the efficacy and safety of the SSRI paroxetine in the treatment of adolescent depression."