Cold Case Files – Paxil Birth Defects
by Evelyn Pringle
Almost like an episode of the TV show, Cold Case Files, the first Paxil birth defect trial was dominated by a story about what happened to the rat pups that died around 1979 and 1980, involved in a study in which Paxil was being tested on pregnant female rats.
The animal studies giving Paxil to rats and rabbits were conducted by a Danish company called Ferrosan before the drug maker, that later became part of GlaxoSmithKline, purchased the drug.
The family’s lead attorney in the case of Kilker v Glaxo, Sean Tracey from Houston, brought in the world-famous neuropsychopharmocology expert from Wales, Dr David Healy, to testify extensively about rat pup study 295.
In summary, Healy told the jury that all the rat pups born to mothers who received Paxil were dead four days after they were born, while eighty-eight percent of the pups not exposed to Paxil were still alive on day four.
In fact, of the 415 rat pups born to mothers who received Paxil, Healy testified that, “One in every ten or actually maybe more like possibly one in every eight or so were born dead.”
As far as he could make make out, all the rats were not autopsied, Healy said, so the question was why the pups died.
“It’s clearly the drug that has caused the death,” he told the jury.
“One of the possible reasons for their death is they’re born with birth defects that lead to them actually dying early in infant life,” he testified. “A responsible approach to data like this is to investigate it further and find out just what the cause is.”
Doctor Suzanne Parisian, a former FDA scientist, also served as an expert for the plaintiffs. She testified that the first safety signals that indicated Paxil could cause birth defects were seen in the animal studies conducted in 1979-1981 period, as well.
Parisian said the studies showed birth defects, embryos that died, and rat pups that did not survive.
Adam Peavy, of the Houston firm of Bailey, Perrin and Bailey, handled her testimony. While testifying, Peavy had Parisian review comments in a memo by a Doctor John Baldwin to Glaxo in 1980, discussing the Ferrosan studies, which stated: “There remains the possibility that this compound could be teratogenic at high-dose levels.”
“We need to ascertain whether Ferrosan have conducted or are conducting or intend to conduct a peri- and postnatal study and a neonatal acute toxicity study,” Baldwin wrote.
Based on her review the documents, Parisian told the jury, Glaxo never ascertained whether Ferrosan did the studies and Glaxo never conducted the studies.
As far as she knew, the company never told the FDA about Baldwin’s statements either, she said, but “they should have.”
On October 13, 2009, the jury in Philadelphia rendered a verdict against Glaxo and awarded the family $2.5 million. More than 600 Paxil birth defect cases are pending in the multi-litigation in Pennsylvania.
Paxil Worse Than Cocaine
In May 2009, a paper was published by a Doctor Sloot and the complete version of the study came out in September 2009. During closing arguments, Tracey told the jury, the “Sloot study is probably the single most problematic document in this case for them because it could have easily been done.”
“They would have known that the drug was a clear teratogen. It was more powerful than cocaine,” he said. “And it was more powerful a teratogen than even the control or as powerful as the control retinol that everybody in this courtroom, everybody that has testified, has said that drug is a teratogen.”
“And had GSK done the studies that Baldwin told them they should do, or a study like Sloot, any time before they started marketing,” he told the jury, “we wouldn’t be here.”
Retinol is an active ingredient in Accutane, a pregnancy Category X drug, meaning it is known to cause birth defects and is not to be used by pregnant women.
During his testimony, Healy was asked to tell the jury what the Sloot study was seeking to do.
“Schering-Plough had acquired a different European company called Organon who had a number of drugs which were serotonin reuptake inhibiting drugs,” he explained.
“One of the things that Schering-Plough wished to find out was could they bring any of these drugs to the market,” he said.
“What they then did was to do reproductive toxicology studies on these drugs that were new to them, plus on, I believe, all of the SSRI drugs that were on the market, plus a number of drugs which also inhibit serotonin reuptake, like cocaine, and they threw into the mix one or two more drugs, one that was known not to cause birth defects and one that was known to cause birth defects closely related to Accutane,” he told the jury.
They had a “particular system to look at the impact of all these drugs on the developing fetus to see were there any indicators of risk,” he said, using rat fetuses.
The study “found that Paxil, of all the SSRIs on the market now, was the most likely to cause birth defects and caused birth defects at the same rate as Accutane did,” Healy said.
He was asked how Paxil did in the study compared to cocaine in terms of potency.
“It was much more likely to cause birth defects than cocaine,” he said. “Cocaine in this particular system was much safer than Paxil.”
On September 21, 2009, the jury also heard testimony on the Sloot study from Doctor Shira Kramer, an epidemiologist. “Doctor Sloot’s paper demonstrated that Paxil was a clear teratogen,” she said, “that it was not just an effect of developmental or birth-weight related effect, that it was a direct teratogen, and that there was a spectrum of defects observed in rat embryos at low doses, establishing that it was a very potent teratogen, more teratogenic than cocaine and retinol, clear teratogens in their own right.”
Parisian told the jury that the Sloot study, conducted in 2009, could have been conducted in 1981 and it would have answered the question that Dr Baldwin was asking about embryotoxicity in 1980.
“If you use that type of a study, it specifically addresses embryotoxicity,” she told the jury.
Paxil Journey to Japan
Although Paxil was not approved in Japan until 2000, a couple years after it was approved in the US in 1992, Glaxo began looking into marketing the drug in Japan and meeting the requirements for approval by the Minister of Health and Welfare (MHW), the Japanese equivalent to the FDA.
“And the Japanese, they suspected, were not going to accept their dead rat pup studies,” Tracey told the jury in his opening statement on September 15, 2009.
“And so GSK began discussions internally,” he said. “Internally among themselves they said: What are we … going to do if Japan makes us do the studies to find out why the rat pups died?”
While Healy was testifying, Tracey introduced a February 9, 1994 memo to Glaxo employee Charlie Fake, and copied to others, from Jenny Greenhorn titled, “Paroxetine Japanese reprotox requirements,” and asked Healy to explain the meaning of “reprotox.”
Any “agency in the world has a requirement from the pharmaceutical companies to look at the reproductive toxicities of a drug,” Healy said. “This includes the impact of the drug on fertility, as well as the potential for the drug to cause birth defects.”
The first sentence of the memo stated, “we have reviewed the three plans for meeting the Japanese reprotox requirements for their regulatory implications elsewhere, should a valid, significant positive i.e., adverse, result be obtained.”
The next two sentences in the memo stated: “The conclusions reached within regulatory are summarized below.”
“It should be noted that there is little or no regulatory information in this area, and we know of no precedents, neither have we consulted with experts outside of the company due to sensitivity of the issue,” they wrote.
“These are, therefore, only our opinions of the likely consequences of positive findings in such studies,” they pointed out.
“What you want from this kind of study is that there is no harm to the fetus,” Healy explained to the jury.
“A positive outcome means there has been a harm,” he said. “So they’re here trying to plan if things go wrong, how do we handle it.”
Healy testified that, “regulatory implications elsewhere,” refers to the fact that the market in Japan is extremely small and the market in the US is huge. “They really have to work out if it is a worthwhile risk being asked to do particular studies which might cause a problem which we would then have to report back to the FDA,” he said.
“And the implications for the market here in the U.S. may mean that it’s just not worth our while trying to go to Japan at all,” Healy testified.
Indeed, throughout page one they discussed the “consequences” of positive findings in the plans for “such studies” if done for Japan. For instance, they stated at one point: “A change in the pregnancy category from B to C is a possibility.”
“This may have commercial implications as the other SSRIs have a B categorization,” they pointed out.
After reviewing the first page and explaining several statements to the jury, Healy was asked whether it appeared that anybody had considered the safety of Paxil. “It appears to me actually as a scientist that this is the opposite to what one would do,” he said, “if the point of view from the scientific end of things is we want to find out what happens.”
What “seems to be happening here is there is a much more business approach which is, well, we don’t really want to know what happens,” he said. “Because if we know, then there are commercial implications.”
For Plan II, they also discussed the “problem” and stated: “A positive finding in the Japanese style Segment II study would be more of a problem since in this case it is undoubtedly exposure during early pregnancy that is of concern for women.”
“A strengthening of the labeling might be likely, EG, women of childbearing potential should not take the drug unless they know they are not pregnant and are taking adequate contraceptive precautions,” they warned. “Worst case, but just possible, contraindication of women of childbearing potential.”
“What they’re saying here is that an awful lot of women clearly don’t know they’re actually pregnant until a few weeks into the pregnancy,” Healy had explained earlier, regarding what was meant by “contraindication” for women in childbearing years.
“Those early few weeks are the period when the harms may happen,“ he said. “And in the case of a drug like this where … there may be a risk of being hooked to the drugs so you cannot get off it if you think you may have become pregnant, then clearly this all becomes extremely risky as it goes on.”
And in this paragraph, they reiterated: “Again, a change in U.S. Pregnancy category from B to C is likely.”
“A positive in the Japanese Segment III would be less of a problem,” they wrote, “since it defines the critical period as the third trimester which authorities would feel more comfortable with.”
“However,” they warned, “they may insist on animal findings being added to the labeling with additional comments and this is certainly the case with the United States, Canada and Australia who have extensive product information.”
In the last paragraph for Plan III, they stated: “This is the same as Plan I, but the lower doses make findings much less likely.”
Again, “this looks like risk management,” Healy told the jury.
“What seems to be happening here,” he said, “is an effort to design a particular kind of experiment that would look like we have been doing things but we’re trying to control the risks, things that could go wrong.”
“So that people will say, well, yes, they have done the experiments, but it didn’t show a particular problem,” he told the jury.
Later in the memo, they stated: “Obviously, conducting no more studies and arguing the case with the MHW would have no regulatory implications elsewhere unless our arguments fail and the MHW requests us to do the type of study we wish to avoid.”
And then it went on to warn: “If they do request a study, there’s a potential problem in that they may direct/insist on our performing a study to their preferred design.”
“At least we are now in the position of being able to do the study we chose and then defending it,” they wrote, with an exclamation point.
After having Healy discuss what was meant by comments throughout the memo, Tracey asked: “Is there anything in this entire document you see, Dr. Healy, that appears as if there is a study being designed by this company to find out the truth about their drug in pregnancy?”
“No, there isn’t,” Healy replied.
While Parisian was testifying, the jury was shown an email dated March 30, 1994, with a discussion about the Japanese government asking for specific information about behavioral changes in babies exposed to Paxil and the protocols for a study that was being designed to respond.
The last sentence read to the jury stated: “Gwyn Morgan will be responsible for the review and the analysis of the protocols to ensure that a likelihood of any potential negative outcome is minimized.”
The goal of pharmacovigilance is not to minimize risk, “particularly when the Japanese government is concerned about the risk,” Parisian told the jury.
The purpose of doing pharmacovigilance and safety surveillance, in terms of toxicology, she said, “is to identify risks so that it can be addressed.”
“And in terms of pharmacovigilance, animal study feeds into the information you have about outcomes of infants,” she told the jury.
In 1997, Glaxo teratologist, Patrick Wier, conducted the rat study, with Segment one comparable to women in the first trimester of pregnancy, and Segments Two and Three comparable to women in the second and third trimesters.
Each “type of study in a rat is trying to simulate what would occur in a woman at those different stages,” Parisian explained to the jury.
Rats are only pregnant for about 22 days, or 7 to 8 days per trimester, she said.
Wier’s study was not designed to find out whether Paxil was a teratogen. It was a behavioral study, “which the ultimate goal was to keep pups alive so that you could look at changes in their behavior,” she pointed out.
“They designed the study to fail,” Tracey told the jury in his opening statement.
A 1997 document stated the doses “have been adjusted accordingly and the risk is now considered as minimal and acceptable,” he said in closing arguments.
But the study did have rats that died with birth defects. Specifically, a ventricular septal defect (VSD), was found in an edematous (swollen) rat; one of the same defects that Lyam Kilker was born with.
The conclusions in the report did not disclose whether the rat with VSD was born to a mother exposed to Paxil. “It was in the text and it was also in appendix 9,” Parisian said.
She testified that the 1980 Ferrosan study also had two edematous rats but they were not autopsied. There is “no information as to why those two rats were edematous,” she stated, “but the company listed them as major malformations.”
“They also had malformations in their rabbit studies,” Parisian reported.
In closing arguments, Tracey pointed out that Patrick Wier had testified that he knew why the rats died because he did a study in 1997.
“Let’s see whether or not that’s true,” he told the jury.
He then reread a memo, gone over during the trial, from July 10, 1998, a year after Wier did his study, which stated in reference to Wier: “Have you worked with Patrick on the concern about the REPROTOX work having been conducted in the seventies? If no, please initiate discussions and set up plans for the MHW discussion. If yes, I would like to know how we intend to justify the validity of these studies and the arguments for rebutting why we should not repeat them.”
“I don’t know how I can take that out of context,” Tracey told the jury.
“If they knew the reason,” he asked, “why do they need to defend or justify the validity of the studies?”
“Why do they have to have an argument for not repeating them if they are fine and … everybody knows why the rats died?,” he pointed out.
He noted that Wier’s memo said, “all we really need is a plausible explanation in case the regulators ask us why the rats died.”
One thing is undisputed in this case, Tracey said. “They never did the test at the higher level with the rabbit to figure out whether or not this drug was a teratogen.”
“If what a drug company really wants to know is whether or not their drug is dangerous, they test it. They don’t avoid tests,” he told the jury in closing arguments.
“Because the negative result will cost us money is not a defense to not finding out the truth about your drug,” he pointed out.
During her testimony, Parisian explained that once a drug is approved it is the manufacturer’s job to do pharmacovigilance.
The clinical trials done to get a new drug approved typically do not have women that are pregnant, she testified. Once a drug is marketed, the company will start getting data about mothers who are pregnant and their babies and it’s “the duty of the manufacturer to monitor that,” she said.
In the case of Paxil, the trials excluded both women who were pregnant and women who could become pregnant because the animal date suggested that there was a risk, she noted, citing a 1984 FDA letter.
While testifying, Peavy had Parisian review a 1997 internal Glaxo document titled, “Paroxetine: Pregnancy with Abnormal Outcomes,” discussing the reporting of birth defects in Japan.
The document stated in part: “There have been 17 reports of congenital abnormality reports to the MHW.”
It also stated: “The company should consider it important that this many cases have appeared.”
The document contained a table with a list of adverse event reports collected on birth defects and birth outcomes, based on an internal review of information received about pregnancies, that showed 32 reports of abnormal outcomes with congenital anomalies.
The company had information on a total of 635 pregnancies in general but only knew the outcomes for 315 infants. “So about half they don’t know what happened,” Parisian explained.
You don’t “know what has happened to those babies at all,” she stated.
Parisian testified that Glaxo internally had reports that looked at all birth defects known to the company, which she had reviewed prior to the trial.
One of the reports was titled, “Paroxetine in Pregnancy,” on adverse events and pregnancy outcomes.
This analysis had a cutoff date of May 1996, and would include the knowledge of the company on pregnancies from the date Paxil was approved in 1992, through May of 1996, Parisian said. The table in this report showed 36 cases of abnormal or congenital outcomes.
This report was never submitted to the FDA according to the testimony of Glaxo employees, Parisian said, but it should have been because it “is safety information.”
In 1998, Glaxo prepared another internal report titled, “Paroxetine and Reports of Congential Abnormalities,” with a cutoff date of November 1997, that showed 42 cases of congenital anomalies.
Parisian said this analysis was also never submitted to the FDA, as far as she knew, but under the rules it should have been.
Peavy then had Parisian review the draft version of the report, showing the same cutoff date and 42 congenital anomalies. On page 5, it stated: “The incidence rate of congenital abnormalities as observed in the data reported in this document is 13.3 percent.”
But the draft version contained the following statements that did not appear in the final report: “A figure much higher than that reported by birth-defect monitoring bodies. Taken at face value, this represents an alarming finding.”
Under FDA rules, drug companies are required to report alarming findings to the FDA, Parisian said. “They are also required to report any kind of safety analysis that they have done internally like this to address an issue about safety.”
In fact, Peavy produced a 1984 letter to the company from the FDA, which specifically stated that the agency wanted to be informed about any “alarming findings” in either human or animal studies.
The 13.3 percent incident rate of congenital abnormalities was between six and seven times the 2.5 percent overall birth defect rate you would expect to see for births in the US, Parisian told the jury. The CDC uses a 2.5 percent background rate.
A 13.3 percent rate “would be a type of signal that a manufacturer should address,” she testified.
Another exhibit reviewed for the jury showed the Japanese had received 27 reports of congenital abnormalities since December 1998 and they were requesting more information about safety, and included emails written by Glaxo employees, Ann Bell and Narita San.
An email written by Narita San stated in part: “This is unexpected number of reports, because only 50 cases were reported in the previous position paper. I think we have to take this abrupt increase of reports into consideration to the reply.”
Parisian explained that “the Japanese are saying that they have received 27 reports of congenital anomalies and they want an update on the prior information.”
In 2000, another internal report of the adverse events for pregnancy outcomes was completed by Doctor Stephen Hughes, Ms Easterbrook and Ms Caley, the worldwide clinical safety group, updating the report made in 1998, Parisian said.
The analysis, with a cutoff date of March 2000, showed Glaxo knew the outcome of 656 pregnancies, out of a total of 1,189 known Paxil pregnancy exposures. By then, there were 79 reports of congenital abnormalities, meaning 12 percent of pregnancies with known outcomes resulted in birth defects.
It was possible that some of the babies in the pregnancies with unknown outcomes also had birth defects, Parisian said. “They are unknown what they had.”
While testifying, Parisian was asked to explain the general rule of “underreporting.”
It is general “knowledge that adverse events that the FDA receives are underreported,” she said. “The FDA’s numbers range from 1 to 10 percent.”
So the FDA looks at the adverse event reports they receive as the tip of the iceberg, “maybe 10% of the reports that are actually out there,” she pointed out.
If talking about underreporting figures for 79 cases of congenital defects, she said, “that would be 1 percent maybe to 10 percent, which would be 7,900.”
“That would be babies that actually lived to getting born,” she told the jury. “That’s nothing to do with the miscarriages and what is not known.”
The reports discussed during her testimony, “were not your classic safety analysis,” Parisian explained. They were a “reporting or a tabulation of what information the company had.”
They were being prepared for the Japanese because the Japanese government wanted safety information on pregnancy outcomes and they were also concerned about “late pregnancy, what the behavior would be of the baby when the baby was born if the mother was taking Paxil,” she told the jury.
But under the rules, Glaxo was required to turn over any internal safety reviews, reports, and analysis to the FDA, Parisian told the jury, “particularly with pregnancy, which the FDA has been asking for information since the product was approved.”
None of the analyses were turned over to the FDA, according to the testimony of Glaxo employees, Parisian said.
Under cross-examination, Glaxo attorney, Joe O’Neil, asked Parisian to identify the FDA regulation she was referring to in stating the reports should have been submitted to the FDA. “I would specifically use (21 CFR) 314.81,” she said, “FDA talks about issues with children, information about children that they are to receive that information.”
“Also, as a postmarket study asked (for) by the Japanese which is the 2000 report,” she stated. “That would be under (21 CFR) 314.80.”
While cross-examining Parisian, O’Neil put up a power point slide listing the FDA approvals of Paxil and Paxil CR, the controlled release version, for additional indications between the initial approval for major depression disorder in 1992 through 2004, including panic disorder, obsessive compulsive disorder, general anxiety disorder, and PSTD.
Several times, Parisian noted that Paxil was not approved for use by pregnant women. O’Neil pointed out that for each new indication, FDA approved the label and Paxil was never contraindicated for use by pregnant women on the label.
Under redirect examination, Parisian testified that when Glaxo applied for approval of the new indications, the FDA was not given the internal analysis of birth defects available at the time or told about Dr Baldwin’s comments on the early animal studies.
Although the 2000 report with 79 cases of birth defects was supposedly done for the Japanese government, it was never given to the Japanese, Tracey told the jury in closing arguments. “It wasn’t given to the people it was prepared for,” he said.
And after the report in 2000, Glaxo stopped doing the pregnancy analysis, he said. “For the next five years we have nothing,” he told the jury. “We don’t have any reports. We don’t know how many birth defects are reported to them.”
FDA Stamped Paxil Label
Glaxo’s lead attorney in the Kilker trial was King & Spalding partner, Chilton Varner.
In her closing argument, Varner stated: “FDA repeatedly reviewed and approved the safety of Paxil, including the scientific data and the animal tests, 13 different times between 1992 and 2004.”
In his final summation, Tracy told the jury: “Ms. Varner has said time and time and time again the FDA rubber-stamped” the label.
“What we know is the FDA didn’t have all their information,” he said.
“The FDA didn’t know what they thought about the animal studies,” he stated. “The FDA didn’t know the analysis they had done.”
“And one thing you need to understand is to this day, the FDA still hasn’t seen these documents,” Tracey told the jury.
“To this day, only my experts have seen the internal GSK documents,” he said. “To this day, you are the only people outside of this courthouse that have laid eyes on these documents.”
“These documents were under seal until this trial began,” he pointed out.
(The Paxil Birth Defect Litigation Update Series is sponsored by the Houston law firm of Vickery, Waldner and Mallia at http://www.justiceseekers.com )
(Evelyn Pringle is an investigative Journalist and Researcher focused on exposing corruption in government and corporate America)