Since Paxil came on the market in 1992, there have been three separate types of failure to warn lawsuits filed against GlaxoSmithKline over Paxil; birth defects, suicide, and addiction.
Roughly 150 suicide cases were settled for an average of about $2 million, and about 300 cases involving suicide attempts were settled for an average of $300,000, according to a December 14, 2009 report by Bloomberg News. Glaxo paid an average of about $50,000 each to resolve about 3,200 cases linking Paxil to addiction problems. The drug giant has also paid about $400 million to end antitrust, fraud and design claims, Bloomberg reports.
All total, Glaxo has paid out close to $1 billion to resolve Paxil lawsuits since the drug came on the market in1992. The company’s provision for all legal matters and other non-tax disputes as of the end of 2008 was listed as $3.09 billion in its annual report.
The first birth defect trial, in over 600 cases filed, resulted in a verdict for the plaintiffs on October 13, 2009, and an award of of $2.5 million in compensatory damages for the the family of Lyam Kilker, who was born with three cardiac birth defects after his mother took Paxil while pregnant.
In the Kilker trial, Glaxo’s lead attorney was King & Spalding partner, Chilton Varner, and Sean Tracey, from Houston, led the family’s legal team.
Andy Vickery, of the Houston firm of Vickery, Waldner and Mallia, is the lead attorney in several Paxil birth defect cases. The first case set for trial is unique in that it involves Delaney Novak, an infant born with heart defects on April 4, 2002, to Laura and Derek Novak, after Laura was prescribed Paxil off label for migraine headaches.
The Novak case is also unique among the Paxil birth defect cases because Delaney’s parents had their insurance with United Healthcare, and Laura was part of the study that Glaxo contracted for, which resulted in the initial warning letter about birth defects in September 2005.
According to Vickery, Glaxo conducted a study on Wellbutrin (bupropion), another antidepressant, after discovering a possible link to birth defects. “The review found no problems with Wellbutrin, but discovered that a significant number of mothers who had been prescribed Paxil (nearly twice as many as those who had not taken the drug) had children born with heart defects,” he says.
Doctors Ra-id Abdulla, David Healy, Shira Kramer and Suzanne Parisian testified as the experts for the plaintiffs in the first trial. All told the jury they believed Paxil caused Lyam’s heart defects. Doctors Abdulla, Healy and Kramer are also expert witnesses in the Novak case.
Ingenix Study in First Trial
The Ingenix study, with lead researcher, J Alexander Cole, was conducted using data from the Ingenix Research Data Mart, containing insurance information from UnitedHealthcare. The study was not supposed to look at Paxil.
During the testimony of several witnesses in the first trial, the jury was shown a February 7, 2003 e-mail in which Glaxo employee, Graham Cottam, stated that he had informed Anne Bell, the project leader for Paxil, about plans to do the Ingenix pregnancy study on Wellbutrin, and “Anne wanted to be sure that we will not be looking specifically SSRIs or Paxil.”
Doctor Suzanne Parisian, a former FDA scientist, testified that the initial 2002 proposal was “to do a large database study for Bupropion in pregnancy.”
There was “nothing that addresses Paxil,” she told the jury. The “procedure had never been designed to specifically look at Paxil.”
But when the data was broken out for Paxil in the original study, it “showed the increased risk and the pregnancy was changed to Category D,” she explained.
The FDA later requested that Paxil be studied, according to the testimony of Glaxo employees and documentation, she said.
The famous neuropyschopharmacologist and professor of psychiatry from Cardiff University in Wales, Dr. David Healy, explained that Glaxo had hopes that the study would show Wellbutrin as an antidepressant that did not cause birth defects and the company could apply to the FDA to have it classified as a pregnancy category B drug instead of a C.
“It would give the message,” he said, “that this of the drugs we have available to use for women of childbearing years, this would be one of the safer ones.”
Healy told the jury that there was “no reason from the scientific point of view why they would not want to also look at Paxil.”
“And this appeared to be the FDA’s view,” he said, “because FDA said, well, you looked at Bupropion, why don’t you look at Paxil, also,” a couple years later.
When asked whether by the year 2003, he could think of any scientific reason not to do a pregnancy study with Paxil, Healy replied, “No, I can’t.”
In fact, Tracey showed the jury an internal company email written by a Glaxo employee two years later in August 2005, around the time that the results on Paxil from the Ingenix study came out, who asked the question: “Why hasn’t the company gathered data on this until now, 13, I think, years after the product was approved?”
In the case of the study on Wellbutrin, there were only 16 reports of birth defects that indicated there was a signal to do the study, Parisian told the jury. While an internal analysis conducted by Glaxo on Paxil in 2000, showed 79 cases of birth defects.
In September 2005, the conclusions of the Ingenix study were: “The use of paroxetine in the first trimester of pregnancy was associated with an increased risk of congenital malformations compared with other drugs.”
“To your knowledge, prior to 2005 did GSK ever do a single epidemiological study to determine whether or not Paxil caused birth defects?” attorney, Adam Peavey, asked Parisian.
“Not that I have seen,” she said.
Ingenix Downside
During the testimony of Dr. Shira Kramer, an epidemiologist, Tracey put up a slide on the Cole paper that was published in 2007.
The paper was on a study conducted by epidemiologists who were employed by Glaxo to do the research, Kramer explained. It was a continuation of the Ingenix study that looked at Wellbutrin and then Paxil. One of the co-authors was Sara Ephross, an employee of Glaxo.
Kramer was asked to explain the importance of the Cole study. “First of all,” she said, “it was a cohort study comparing … people exposed to Paxil … to people who were exposed to other SSRIs.”
“So one very key thing for you to remember is that here the … unexposed group is not people who were not exposed to SSRIs, they were exposed to SSRIs,” she told the jury.
“That’s a very important point,” she said, “because if SSRIs are a risk factor for cardiac defects, birth defects, then the relative risk that will be generated in this study is going to be lower than it normally would if truly people were unexposed to SSRIs.”
“The other thing that is important to keep in mind,” she told the jury, “is that the information was obtained from an administrative claims database called the Ingenix Research Data Mart. ”
“There was no individual, either examination or interviewing of anyone,” she explained. “The information was extracted from administrative claims data that was available.”
“The other thing that’s important,” she said, “is that initially the population that was studied covered the years 1995 to 2002, and then after the fact an additional two years were added to the study.”
“The published results, based on all of the years that were eventually included in this study, were an odds ratio for all cardiovascular malformations related to Paxil exposure of 1.46, which means that individuals who took Paxil were at 46 percent increased risk of their child having a cardiovascular malformation diagnosed at birth compared to individuals who took other, other SSRIs,” Kramer explained.
“In the second odds ratio of 1.68,” she said, “showing a 68 percent increased risk, now we are comparing women who took Paxil either alone or in combination with another SSRI, compared to the other SSRI group, either alone or in combination with other SSRIs, mono- or polytherapy.”
The published study contained an asterisk that said: “An interim analysis performed by Cole, et al, using births occurring between ’95 and 2002 found an odds ratio of 2.0 for the association between first trimester Paxil use and cardiac birth defects.”
Kramer was asked to explain what that statement was referring to. “Initially, the study was designed to include the years 1995 through 2002 with a sampling ratio of controls to cases of 7 to 1,” she said. “That was the protocol.”
“And when that analysis was done, the odds ratio, instead of being 1.46, which ultimately is what was published, was actually higher, it was 2,” she told the jury.
“That means that the exposed group had a risk of a child with a cardiac malformation two times that of the group not exposed to Paxil,” she added.
The odds ration got smaller when the Glaxo researchers, the authors of the study, “added in two additional years of data with a different sampling ratio,” Kramer explained.
It is not appropriate for an epidemiologist to do that, she said, because “you are changing the rules after you look at the data.”
It “really raises the questions as to, are you trying to influence the data,” she noted.
“I can say very clearly,” she told the jury, “that that is not considered to be appropriate conduct, scientific conduct. ”
“What you are supposed to do is set up a study protocol in advance and follow it, and not change it after you have looked at the results,” Kramer explained.
It is not appropriate to find out the results and change it in the middle, she said, “for obvious reasons, it looks like you are manipulating the data to make it come out looking a certain way.”
“And if you want to do an unbiased, fair study,” she told the jury, “the only appropriate plan of action is to develop a study protocol ahead of time, to follow it, and to analyze the results and not to fool with it, not to fiddle with it and not to change it.”
While testifying, Kramer explained the meaning of “underpowering” a study. You need to have “a sufficient number of people in a study in order to test a certain research question,” she said.
“And if you are going to apply statistical tests to the data that you generate,” she told the jury, “you need to have enough people in that study to have generated enough cases of the outcome and you need to have enough people who are exposed.”
“Now, this case,” she said, “we have got a relatively rare exposure to Paxil, we have a relatively rare outcome, which is congenital cardiac birth defects, so you need to study very, very large populations in order to achieve statistical significance at these levels that we have been discussing.”
She said the “investigators, the research team,” dictates the size of the study.
Kramer went over the reasons why the odds ratio in the Cole study might be attenuated, or lower. It’s “very clear that there are certain characteristics of this study that are making this odds ratio probably lower than it really would be given certain characteristics of the study design,” she told the jury.
“One of the them is that the controls are really not unexposed to the SSRIs,” she said. “They are exposed to drugs in the same class.”
We have “observed in epidemiological literature that other SSRIs are associated with an increased risk of cardiac malformations,” she told the jury. “Therefore, it is likely that since that’s the comparison group, we have got an odds ratio in this study that’s lower than it probably would be.”
The second reason was that the analysis only included live births. “So you are missing fetuses who were miscarried,” she said. “And then there are many miscarriages that are due to birth defects.”
“You are missing fetuses that are aborted, electively aborted, because of known cardiac or other congenital malformations,” she told the jury. “You are missing stillbirths.”
And with a follow up for only nine months, she said, “you are missing congenital cardiac defects that aren’t detected until later.”
“So you have got a fairly substantial population that is not really being captured in this study of exposed fetuses,” she pointed out.
Closing Recap
During closing arguments, Tracey reminded the jury about the e-mail with Anne Bell’s statement to make sure Paxil was not included in the Ingenix study, and said: “This document … two years before this child is born, they are affirmatively saying: We do not want to look at Paxil in pregnancy.”
In her closing argument, Varner told the jury: “Now, Mr. Tracey has talked to you about Anne Bell this morning.”
“He has said that GSK would have done anything to avoid looking at the risk for Paxil,” she pointed out.
“Well, ladies and gentlemen,” Varner said, “GSK funded the study that did look at Paxil for the risk and published preliminary findings in August of 2005.”
As soon “as even a possible link emerged in all of 2005, GSK reacted promptly and proactively to notify both FDA and doctors,” she told the jury.
“It went immediately to FDA. It immediately changed its label,” she said. “And it immediately sent out letters to doctors telling them about the changes.”
In his final summation, Tracey told the jury, “I want to talk to you about the Ingenix study because Ms. Varner said something that is very, very important.”
“She said when they found out what she says is August of 2005, within 21 days, they changed the label. Within 21 days, the doctors got the news,” he recounted.
What “she forgets to tell you is that two years prior to this, Anne Bell said, Don’t study the drug,” Tracey told the jury.
“Had they not listened to Anne Bell,” he said, “had they studied the drug in 2003, Michelle David wouldn’t be sitting here because the warning would have gone out like that.”
“We would have been two years ahead of the game,” he pointed out.
“GlaxoSmithKline did not want to study the drug,” Tracey told the jury. “The FDA made them study Paxil.”
“It was not some sort of voluntary we’re just a good drug company trying to get along,” he said. “It was we don’t want to study it and they’re forcing us to study it.”
Paxil Off-Label Promotion
Paxil is not FDA approved for use by pregnant women, so all mothers who gave birth to infants with heart defects received the drug off label. In Andy Vickery’s first case set for trial, Delaney Novak was born with heart defects after his mother, Laura, was prescribed Paxil for migraine headaches, another unapproved use.
Dr. Dee Mangin is an expert witness in the Novak case. Her research and published work has focused on rational prescribing, and the influence of drug company promotion both to physicians and direct to consumers. She submitted a report on October 13, 2009, which outlines Glaxo’s off label promotion of Paxil around the time of Laura’s pregnancy.
In her report, Mangin defines off-label use as the “practice of prescribing drugs for a purpose outside the scope of a drug’s approved label – often an unproven use or one that has not been widely tested.”
“While it is legal to prescribe off label in the United States, it is illegal for companies to promote off label use,” she notes.
“The risks of off label promotion,” she says, “are that it could lead to exposure of patients to the risks of a medicine for no benefit, and furthermore they may be denied other more effective treatment.”
“GlaxoSmithKline from 2000,” Mangin says, “mounted a multifaceted and targeted national promotional campaign that employed explicit strategies designed to promote sales of Paxil in pregnant women and women of reproductive age.”
An exhibit cited in the report from a “Paxil Tactical Marketing Plan in 2000,” states: “New Paxil data with high media interest, hot flash, postpartum, depression, pregnancy, and lactation will position Paxil as the drug of choice for women.”
“One of the known reasons that physicians change their prescribing behavior is as a response to the volume of evidence containing the same message that the physician is exposed to,” she wrote. “The so-called “Carpet Bombing” technique used in the Paxil campaign feeds directly into this.”
“There are a number of strategies companies can use to highlight use for off label conditions including distribution of individual scientific articles discussing the off label indication and use of the drug as well as mentions of off label use by key opinion leaders in continuing medical education,” the report explains.
“In relation to the off label prescribing for migraine,” Mangin says, “there is no evidence of any effectiveness over placebo for SSRIs in migraine prevention.”
Yet a paper titled, “Paroxetine in the Treatment of Chronic Daily Headache,” by Carol Foster, MD, and Jacklyn Bafaloukos, RN, that was distributed to doctors, specifically states: “The dramatic improvement in the patients in our study suggests that paroxetine appears to be a safe and effective drug for the treatment of chronic daily headache.”
“The strategies outlined where reprints about treatment of migraine with paroxetine, large numbers of form letters containing summaries of studies of use in headache were sent to physicians and detailing and providing free samples to physicians likely to treat women with migraine were therefore encouraging use of Paxil and exposure to its risks when in reality it is no more effective in this situation than a sugar pill,” Mangin reports.
Encouragement “of unapproved use for migraine further attempted to expand the market beyond that which was medically justified and likely to lead to unnecessary exposure to the risks of Paxil,” she advises.
In an August 11, 2009 deposition, Laura’s doctor testified that Glaxo sales representatives would commonly leave reprints of articles on off label uses and salespeople did discuss the literature on the off label use of Paxil for migraines with him. One of the sales representatives visiting the doctor at the time stated in a deposition that it was his habit to distribute all such articles.
But most importantly, the doctor said he would not have prescribed Paxil to Laura had Glaxo told him back in 2000, or early 2001, that there was an association between Paxil and birth defects. He further noted that there was no benefit from Paxil that would outweigh the risks of birth defects and that he had not used Paxil in his practice since the Dear Doctor letter warned about birth defects.
Delaney suffers from a septal heart defect. “None of the information from the medical records of the family or their statements on potential genetic, environmental and pharmaceutical causes of heart defects indicates any other factor more likely to have caused her condition than the Paxil exposure,” Mangin points out.”It is clear that if the prescribing doctor had been informed of the risk of heart defects, Laura Novak would not have been exposed to Paxil,” she notes.
In the report’s conclusion, Mangin states: “It is my opinion that this promotional campaign for Paxil was inappropriate given the scientific knowledge and what was known by the company at the time.”
“The degree of comfort with the use of this medication in the reproductive years and pregnancy is likely to be influenced by GSK’s misleading promotional campaign where concerns were minimized, efficacy was overstated, the idea of off label prescribing was seeded for migraine, and lastly the marketing specifically targeted a group at higher risk in terms of safety concerns – pregnant women and women in the reproductive age group,” she reports.
“This Paxil promotional campaign was irresponsible, and potentially disastrous from a public health perspective as it was likely to expose a much greater proportion of the population to these potential harms,” she concludes.
Glaxo’s Phone Book
During closing arguments on October 8, 2009, Tracey told the jury regarding Glaxo: “They have a telephone book full of doctors.”
Referring to an exhibit introduced during the trial, he said: “This is all the doctors that they pay to give speeches on their behalf to push their drug, to sell it, to convince other doctors to prescribe their drug.”
While Healy was testifying, Tracey had him go over some of names of doctors in the book that included Lori Altshuler, Vivian Burt, Lee Cohen, Charles Nemeroff, Jeffrey Newport, Zachary Stowe, Katherine Wisner and Kimberly Yonkers. None of these doctors appeared to testify on Glaxo’s behalf in the trial.
What “they did was aggressively market this drug to women,” Tracey told the jury.
All “these names of people that they ghost-wrote articles for to get the doctors … to sell the drug,” he noted.
Doctor Healy told you that “they altered the prescribing practices in this country,” he recounted. “What they set out to do, they succeeded in doing. They got doctors to prescribe the drug to women.”
“And they did it,” Tracey said, “by having seminars where they would put these doctors, experts in the field, on their payroll, that the doctors would go and listen to, unwittingly knowing what they are really hearing is a marketing campaign.”
In reference to another exhibit viewed during the trial, Tracy said: “This document describes that it worked. When the doctors came out, these are the comments they made after attending these seminars: Will prescribe Paxil to pregnant women. My comfort in treating depression in pregnancy has increased. Treating pregnant patients with confidence. Will feel more comfortable giving Paxil to pregnant women.”
In citing $765 million in the US alone, between 1997 and 2005, Tracey told the jury: “This is the number for over a nine-year period this company spent to convince doctors to sell their drug, to prescribe their drug to women of childbearing years.”
“And they got a heck of a return on it,” he said. “Net. After expenses. Almost 14 billion dollars for a nine-year period.”
“Out of the 700 million dollars they spent trying to sell this drug to people,” Tracey stated, “there is not one shred of evidence in the record about how much money they spent to try to figure out whether it was going to induce birth defects.”
“And as far as I can tell in the record,” he said, “after they bought it, they did one animal study and they didn’t spend another penny.”