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

by Alison Bass

  The day after the FDA's advisory panel voted to recommend the black box warnings, a correspondent for the Boston Globe called Teicher. Carolyn Johnson had been assigned to write a feature about the FDA panel's decision that would appear the following Tuesday in the paper's Health and Science section. Johnson wanted to know what kind of alternative therapies could be offered for pediatric depression now that the SSRIs were probably going to be black-boxed. Teicher told her what he had told other reporters: that the potential for serious side effects with the newer antidepressants didn't mean they shouldn't at times be prescribed-under close medical monitoring. Doctors, he said, are not going back to the older generation of antidepressants, which carry their own severe side effects. The SSRIs, he concluded, will continue to play a role, but they are not for everyone. "In a small population of people, the antidepressants worsen every single symptom of depression," Teicher told Johnson.

  It was virtually the same conclusion he had arrived at fourteen years earlier. Teicher hadn't changed his opinion in the intervening time, but the world around him clearly had.

  BACK IN MANHATTAN, Rose Firestein was also following the steady thrum of news from Washington, but not as closely. She was too busy with another case that had suddenly come to the fore. On September 7, in an unusually fast turnaround time for a New York State investigation, the AG's office reached a legal agreement with Forest Labs, the maker of Celexa. The AG agreed not to sue the New York-based drug company if it would immediately disclose its Phase 3 and Phase 4 clinical trial results on the pediatric use of Celexa and work on posting a public registry of its drug studies, using much the same template that Firestein and her colleagues had agreed upon with GlaxoSmithKline.

  It was perhaps no coincidence that Dr. Lawrence Olanoff, an executive vice president of Forest Labs, was scheduled in two days' time to stand with other pharmaceutical executives before the House subcommittee's hearing on Publication and Disclosure Issues in the Antidepressant Pediatric Clinical Trials. And so when members of the Subcommittee on Oversight and Investigations asked Olanoff on September 9 whether his company would support a comprehensive registry of clinical studies, the doctor was ready with his answer. He told the legislators that in accordance with his company's agreement with the New York attorney general's office, Forest Labs would be posting online summaries of clinical trial results for all its Phase 3 and 4 studies, as well as any pertinent safety information from completed clinical trials dating back to 2000. His timing was superb. The transcript of the hearing reveals that the legislators spent more time grilling executives from the other drug companies than they did the good doctor from Forest Labs.

  The congressional inquisition finally ended at 7:08 p.m. on September 9. But the legislators weren't done. Two weeks later, the same subcommittee held another hearing on Capitol Hill, this time focusing on the FDA's treatment of one of its own medical experts, Andrew Mosholder. In the course of investigating why Mosholder had been prevented from testifying back in February, the legislators discovered that he had been subjected to an internal investigation and threatened with disciplinary action for allowing news of his gagging to be leaked to the press. Mosholder denied leaking the information and mentioned the disciplinary proceedings in his sworn statement to the congressional investigators. But his superiors at the FDA insisted that he delete that information from his statement and submit a cleaned-up version to Congress instead. Mosholder refused. At the September 23 hearing, Rep. Peter Deutsch declared that by refusing to go along, Mosholder had "thwarted a criminal act of obstruction of justice." "As it was," Deutsch said, "the FDA and its lawyers are only guilty of attempting to obstruct justice." (When I asked to talk to Mosholder for this book, a spokeswoman for the FDA said he was not giving interviews.)

  Deutsch and a number of other legislators also blasted the agency for not acting quickly enough to protect vulnerable children. "The FDA is failing to live up to its responsibility to the American people," said Rep. Bart Stupak (D-Mich.). "This hearing illustrates a larger problem where too often ... medical evidence is suppressed and expert opinion is silenced. It illustrates that our system to study the effects of drugs on children is broken."

  In hindsight, it is tempting to conclude that the hubbub over the SSRIs was one possible explanation for the bombshell that dropped a week later. On September 30, 2004, Merck announced it was withdrawing its bestselling painkiller, Vioxx, from the market. The news stunned doctors and consumers alike. Until that point, there had been only scattered reports about an increased risk of heart attacks and strokes among people who took Vioxx, a drug that was routinely prescribed for arthritis and had generated sales of $2.5 billion in 2003. The mainstream press might have been excused for not paying much attention to Vioxx's safety issues. The FDA itself was giving out some very mixed signals about the painkiller. Although the agency did order Merck to put a warning label on Vioxx in 2002 -much watered down after Merck protested the FDA then approved the drug for migraines in March 2004 and for use in children a few months later. The FDA's dithering on this and other drug safety issues prompted a columnist for the trade publication Medical Marketing and Media to write in early 2004 that the Bush administration had made the federal agency "an informal `partner' to the industry."

  On August 25, 2004, at a conference in France, FDA investigator David Graham presented an analysis showing that Vioxx was significantly more likely to increase the chance of heart attack and death from cardiac arrest than Celebrex, a painkiller that worked by a similar mechanism: blocking "bad" enzymes known as COX-2 that cause stomach inflammation and pain. Dr. Graham's presentation, however, received little press coverage. The mainstream press did not wake up to this story until September 30, when Merck made its stunning announcement. Merck said it had decided to pull Vioxx off the market after seeing preliminary results from a clinical trial being conducted to determine whether its COX-2 painkiller could prevent the recurrence of colorectal polyps. Initial results from that trial confirmed that people who used the drug had nearly twice as many serious cardiovascular problems as those on placebo. Merck stopped the trial and faced the music.

  Merck may have acted with such dispatch in an effort to blunt the kind of litigation and bad publicity that had dogged the makers of SSRI antidepressants. Or perhaps the drug company did so out of a genuine desire to do the right thing. Whatever the motivation, the drug company only opened up a Pandora's box.

  "Yesterday's announcement [by Merck] rekindled the debate over whether the FDA has been sufficiently aggressive in monitoring drug safety," wrote Washington Post reporter Marc Kaufman, one of many journalists to jump on the story.

  Kaufman quoted Dr. Eric Topol, head of cardiology and chief academic officer of the Cleveland Clinic, as saying Merck's decision to pull the drug was "the right decision about three years too late." Topol was referring to a major study funded by Merck and published in the New England Journal of Medicine in November 2000 that found Vioxx was easier on the stomach than an older painkiller, Aleve. The study had also found a twofold to fourfold increase in the risk of heart attacks among patients taking Vioxx versus those taking Aleve. However, the study's authors did not mention that disturbing statistic in their 2000 paper; they merely noted that the incidence of heart attacks was lower in Aleve than in Vioxx and offered a reassuring explanation about Aleve's extraprotective effects, which steered attention away from Vioxx. (In 2005, the editors of NEJM would accuse Merck and the researchers who did that study of withholding key evidence from the paper they submitted. In fact, the journal editors said they discovered that the authors deleted heart-related safety data from a draft of the Vioxx paper just two days before submitting it to the journal for publication. The journal said it was able to determine this by reexamining a computer disk submitted with the manuscript.)

  By the end of 2004, Merck would be sued on behalf of thousands of people who had had heart attacks after taking Vioxx. Congress would hold yet more hearings to investigate what Merck knew about the heart risk
s of Vioxx and when they knew it. And the federal Department of Justice and the SEC would initiate their own inquiries into the COX-2 painkillers, as would the New York attorney general's office.

  Two weeks later, on October 15, 2004, the FDA dropped its own bombshell of sorts. The agency announced that it was going beyond the recommendations of its own advisory panel and requiring black box warnings on all thirty-two antidepressants currently on the market, old as well as new. Although the agency only had evidence showing elevated risks in nine of the newer antidepressants, FDA officials said they decided to put the warnings on all thirty-two drugs because they did not have sufficient data to clear them of similar side effects. News of this decision was almost buried by the clamor over Vioxx.

  ON THE MORNING of October 28, Rose Firestein rose as usual at dawn, even though she didn't have to go to the office. She had been asked to give a presentation that day to doctors and medical researchers at a meeting of the World Health Organization. The meeting, a preliminary discussion on guidelines for an international clinical trial registry in advance of the WHO's annual meeting in Mexico City, was scheduled to begin at 9 a.m. at the Rockefeller Foundation on the East Side of Manhattan.

  Sitting around the large conference table amid distinguished doctors, journal editors, and top health-policy makers from around the globe, Firestein felt like a fish out of water. What could she, an attorney with no medical background, add to the conversation? She kept her remarks brief, talking about what the AG's office had hoped to accomplish with its case against GlaxoSmithKline and what the limitations of its author ity were. When she finished, she felt redundant and slightly foolish, like a footnote buried at the bottom of a page.

  But then a doctor from Southeast Asia-Firestein couldn't read her name tag but she sounded as if she was from India or Pakistan turned to her and said warmly, "We've been working on this issue for a decade and were never able to get it moving. But your lawsuit brought public attention to the issue. What you've done really moved the ball."

  Several other people chimed in, saying they too sensed there was now a new opportunity to push drug companies to disclose the results of all their clinical drug trials. Firestein smiled but said nothing. As the discussion flowed around her, she basked quietly in the realization that top medical experts-many of them important officials who were making health policy in their own countries were saying that a case she had poured her heart into had precipitated significant change, not only here but globally. It dawned on Firestein that she had finally done what she set out to do some thirty years earlier: make a difference in people's lives. A sense of pure contentment enveloped her. Tomorrow, this feeling would surely dissipate, lost in the maze of endless work, annoying commuters, and the constant heartache of being blind in a vast and oblivious city. For now, though, it would do. It would do just fine.

  ON MARCH 22, 2006, Tonya Brooks and her parents sued GlaxoSmithKline for failing to warn them that Paxil can make adolescents suicidal. Representing them in that lawsuit was Baum, Hedlund, Aristei, Goldman, and Menzies, the Los Angeles law firm that brought suit on behalf of dozens of other children who had killed themselves or attempted to while taking Paxil. The very same month, researchers at GlaxoSmithKline published a paper acknowledging that they had discovered ten previously unreported cases of suicidal thoughts and behaviors among the children and adolescents who participated in the original Glaxo-funded trials of Paxil. The 2006 paper, published in the Journal of Child and Adolescent Psychopharmacology, was prompted by the FDA-mandated review of SSRI data. The researchers for Glaxo reexamined five double-blind, placebo-controlled trials of Paxil, including the three studies of adolescents with depression (studies 329, 377, and 701), along with two other studies of youngsters with obsessivecompulsive disorder and social anxiety disorder. They concluded that overall, "suicide-related events" occurred almost four times more often in patients taking Paxil than in those taking a sugar pill (the odds ratio was 3.86 and the relative risk was 3.76). That came to an almost 50 percent greater risk than the rate of suicidal behaviors found in the FDA's reanalysis two years earlier.

  In the Keller study alone, the Glaxo reviewers identified four additional cases of possible suicidal behaviors among teenagers taking Paxil. In an e-mail to Keller and his coauthors, Regan Fong, one of the Glaxo reviewers, wrote, "Four of the 10 additional events potentially suggestive of intentional self injury, suicidal ideation or suicide attempt ... occurred in study 329, all in the paroxetine group."

  This e-mail was among the documents uncovered in the Baum Hedlund legal proceedings against G1axoSmithKline. Those documents, together with the 2006 Glaxo paper, provide strong support for Donna Howard's assertion that the researchers for study 329 miscoded and underreported data on Paxil's impact in children and adolescents.

  Keller and his coauthors, for example, reported in their 2001 paper that six patients in the paroxetine group suffered "emotional lability"the code term for suicidal thoughts or gestures-as compared to three in the tricyclic antidepressant group and one in the placebo group. However, when the FDA insisted that GlaxoSmithKline (and other drug companies) comb through their trial data again, the number of patients in the Keller study who were suicidal jumped from six to ten (out of ninety-three participants in the Paxil arm).

  Keller declined to be interviewed for this book. When asked about these discrepancies, Mary Anne Rhyne, a spokeswoman for GlaxoSmithKline, said that "it is not productive to have this kind of patientby-patient discussion here" She would say only that the additional cases unearthed by reviewers for Glaxo "do not change the interpretation of the results of 329."

  According to several experts, that simply isn't true. Even without the addition of the mysterious patient 70, the 2006 paper by the Glaxo reviewers concluded that the adolescents taking Paxil in the Keller study were almost four times more likely to develop suicidal thoughts or behaviors than those taking placebo. That is a significant difference, and had it been published in 2001, it would have alerted physicians and the media to a dangerous side effect of Paxil.

  Keller and his coauthors also chose not to do a proper statistical comparison of the patients in the Paxil group who suffered serious adverse events versus those in the placebo or imipramine group. That statistical analysis would have shown that Paxil users had five times the number of serious psychiatric effects as placebo takers and more than twice the number of those taking the older tricyclic antidepressant. Again, even without the additional unreported cases of suicidal behavior reported in 2006, such an analysis would have alerted doctors to Paxil's significant adverse effects in adolescents.

  "This kind of analysis would have shown that the psychiatric issues are significantly greater in the paroxetine [Paxil] group than in the placebo group and it's a statistically significant difference," says Dr. Roger Grimson, a former biostatistician at the State University of New York (SUNY) who has advised plaintiffs in a lawsuit against Glaxo. "They didn't do that statistical analysis, and they should have"

  So why weren't these comparisons made? The contention that Paxil is "generally well tolerated in this adolescent population" was one of the key findings of the 2001 Keller paper. If Keller and his coauthors had accurately reported all of the suicide-related events in their study or done a statistical comparison of serious adverse events, they would not have been able to reach that conclusion. They would not have been able to tout Paxil as a safe and well-tolerated drug. And that would have made their research sponsor, G1axoSmithKline, very unhappy.

  IN THE THREE YEARS since the New York State AG's office shone a spotlight into the black hole of drug research, there has been a growing outcry about the enormous influence the pharmaceutical industry wields over the practice of medicine. The impact of corporate money on the way doctors prescribe drugs and treat patients has become one of the most contentious issues in health care today. As articles in the nation's leading newspapers and medical journals and newsletters reveal, the pharmaceutical industry pays millions of dollars in sp
eaking and consulting fees to physicians throughout the nation. Take, for example, a series of stories by the New York Times in the spring of 2007 about payments from drug companies to physicians in Minnesota (the first state to require such full disclosure). According to the New York Times, pharmaceutical companies paid Minnesota doctors, nurses, and other health care providers at least $57 million between 1997 (when the law was passed) and 2005. More than 25 percent of the licensed doctors in Minnesota received money from drug companies. More than one hundred doctors received more than $100,000 each from the industry, among them physicians who sit on FDA advisory panels and help craft the guidelines that other doctors rely on when prescribing drugs. It might come as no surprise to the readers of this book that psychiatry was the specialty with the highest number of doctors receiving payments from drug companies.

  But the medical profession's ties to Big Pharma go beyond payments to individual doctors. In the past decade, industry money has become the single largest source of funding for the medical courses that practicing physicians need to take in order to remain accredited, according to Dr. Daniel Carlat, a professor at Tufts University School of Medicine and the editor in chief of the Carlat Psychiatry Report. Moreover, drugmakers routinely offer free junkets, lunches, gifts, and product samples to physicians in order to promote their wares.

  Such generosity works: research shows that doctors who receive consulting or other personal income from drugmakers are more likely to report positive findings about a particular drug than researchers who don't receive money from the industry. Doctors with such financial conflicts of interest are also more likely to prescribe newer and more expensive drugs than doctors who don't have such conflicts. A New York Times article in May 2007, for instance, showed a distinct correlation between the financial relationships that Minnesota psychiatrists had with the makers of antipsychotic drugs like Risperdal, Zyprexa, and Seroquel and the growing use of such drugs among children in that state. According to the Times analysis, during the very same period (2000 to 2005) that drug company payments to Minnesota psychiatrists rose more than sixfold-to $1.6 million the prescriptions of antipsychotics to children in Minnesota's Medicaid program climbed more than ninefold. Those who received the most consulting speaking fees from the makers of these antipsychotics tended to prescribe the drugs most often. On average, the Minnesota psychiatrists who received at least $5,000 in personal income from the makers of these drugs appear to have written three times as many prescriptions for them as psychiatrists who received less or no money, according to the Times analysis.