Since its approval in 1991, the anti-nausea drug Zofran has undergone several major labeling revisions. For a pharmaceutical product, this isn’t uncommon. When a drug is released onto the market, and large amounts of the population begin taking it, we necessarily learn new things about its effects.
After 1991, Zofran quickly became one of America’s leading morning sickness treatments. And now that millions of women have ingested Zofran during pregnancy, researchers are beginning to investigate its effects in this new, and previously unstudied, patient pool.
Through the work of multiple, independent teams, Zofran’s active ingredient has now been associated with an increased relative risk of major birth defects.
At least four large epidemiological studies have linked Zofran to congenital heart defects, cleft palate and kidney malformations, and recent lawsuits have claimed that GlaxoSmithKline has been receiving reports of birth defects with suspected ties to Zofran exposure for more than two decades.
Has The FDA Released A Statement On Zofran’s Link To Birth Defects?
In the face of these apparent warning signs, both GlaxoSmithKline and the FDA have remained silent. Zofran’s labeling has not been changed, and plaintiffs in the recent birth defect lawsuits call this silence a serious failure to disclose vital information about Zofran’s potential risks to the public.
Warning The Public Of New Risks: How Much Evidence Is Enough?
Pharmaceutical companies are empowered by federal law to revise labeling with new warnings themselves. They don’t need the FDA’s permission beforehand. If, on the other hand, a drug company is unwilling to change the label, the FDA can compel them to do so.
But as we’ve said, neither governmental authorities nor GlaxoSmithKline have responded to the significant concerns raised over Zofran’s potential link to major birth defects.
So how does the FDA normally respond to new information about a drug’s potential to cause harm? What kind of evidence, and how much, does it take for a pharmaceutical manufacturer to alter a drug’s labeling?
Causation v. Association: What Kind Of Evidence Is Needed For A Drug’s Label To Change?
In regard to what kind of evidence is necessary, federal law is clear: drug companies must alter a label “to include a warning as soon as there is reasonable evidence of a[n] association with a drug; a causal relationship need not have been definitely established” [emphasis added].
In other words, studies don’t need to prove, beyond a shadow of a doubt, that a drug always causes an effect, that A leads to B in all cases, that exposure to Zofran always causes birth defects in human children.
In controlled clinical trials, researchers can filter out subjects who are predisposed to experience the result that’s under investigation. For example, many major birth defects can be caused by genetic abnormalities. Some are even hereditary; they can be passed down in the form of genetic material from parent to child. If you were going to set up a study to investigate the relationship between Zofran and congenital anomalies, would you include any pregnant women who were genetically predisposed to have children with major birth defects? Of course not.
By carefully selecting the patients that are used in clinical trials, researchers are able to eliminate “false positives,” those cases in which it wouldn’t be clear whether a birth defect was caused by abnormal genetic material or exposure to a potentially harmful drug. In doing so, they can get closer to demonstrating a causal relationship between one drug and major birth defects.
Do We Have Clinical Trials Investigating Zofran’s Effects?
Yes, the FDA requires drug companies to perform multiple rounds of clinical trials before approval. But these controlled studies only look at a drug’s side effects and efficacy among limited numbers of specific patients. Drug manufacturers tell the FDA which conditions they want to market and sell a drug for, and only need to conduct clinical trials involving patients with those conditions.
Which means that we know significantly less about how drugs will affect other portions of the population until after they are approved.
In Zofran’s case, clinical trials provided the FDA with sufficient evidence that the drug was safe and effective for the treatment of severe nausea and vomiting, but only in cancer patients receiving chemo- and radiotherapies. As for its safety during pregnancy? Zofran’s manufacturer never expressed a desire to have the drug approved as a morning sickness treatment, and never performed clinical trials involving pregnant women.
At the time of Zofran’s approval, there was no specific understanding of how the drug could affect pregnant women or alter fetal development.
That would not have been a problem, but in 2012, the federal government charged GlaxoSmithKline for unlawfully promoting Zofran as a safe treatment for morning sickness. If the Justice Department’s allegations are true, GlaxoSmithKline has been marketing Zofran to doctors as “safe and effective” during pregnancy, without any evidence to back it up. The company has always denied this, but regardless of whether or not GSK actually marketed Zofran to physicians as a safe and effective drug for pregnant women, the fact is that doctors throughout the US and Europe are prescribing it “off label” for morning sickness.
Can Clinical Trials Be Performed After A Drug’s Approval?
Yes, but for ethical reasons, they rarely involve pregnant women as patients. More often than not, research to investigate the effect of a drug during pregnancy will take the form of “observational epidemiological studies.” Rather than gather together a group of patients and administer a drug to them, observational research uses birth and other health records to identify women who have been prescribed a drug, and compare their health outcomes to women who have not.
All four studies that have found an association between Zofran and major birth defects were observational epidemiologic studies. Researchers looked back in time through hundreds of thousands of birth records. They split all those records into two groups: one group of women had taken ondansetron and the other group had not. By comparing the rate of major birth defects in children born to these women, the researchers concluded that maternal exposure to ondansetron increased the risk of:
- delivering a baby with cleft palate by 2.37 times
- delivering a baby with kidney malformation by 6 times
- delivering a baby with an atrioventricular septal defect by 4.8 times
- delivering a baby with cardiac septal defects by 62%
Are Associations & Relative Risks Unimportant?
Should we interpret those results to mean that ondansetron necessarily causes birth defects?
No. Researchers have suggested that ondansetron increases the relative risk of birth defects.
Let’s say that 1 out of every 600 women who are not exposed to Zofran’s active ingredient will deliver a baby with cleft palate. If Zofran increases the risk by 2.37 times, then we move from 1 out of 600 to 1 out of 253. We would still expect 252 women to deliver babies without cleft palate. But for a doctor, the situation has changed considerably.
Shouldn’t doctors know that a drug can increase the risk of an adverse health outcome? Especially now, when an FDA approved, safe and effective treatment for morning sickness exists in the US? In light of Zofran’s increased relative risk with major birth defects, many physicians may feel uncomfortable prescribing it, and choose to prescribe Diclegis instead. But in the absence of any change to Zofran’s warning label, plaintiffs in the recent birth defect lawsuits say that GlaxoSmithKline never gave doctors the opportunity to make that choice.
At bottom, these parents are sending a simple message to one of the world’s most profitable drug companies: give America’s physicians the whole story, provide them with the basic information necessary to make the best choices for their patients.
Finding Associations Is Often The Foundation For Causation
After a drug is released onto the market, studies that associate it to an adverse outcome are often the bedrock upon which additional research is based. In many cases, further scrutiny will confirm the association and reveal a causal relationship.
As Adi Jaffe, a Ph.D lecturer in addiction studies at UCLA has noted, repeated findings that demonstrated an association between smoking and cancer led to studies that eventually proved a causal relationship. Check out his editorial in Psychology Today to find a clear description of the difference between causation and association (which he also calls “correlation”). Note that he says associations “are crucial for research and still need to be looked at and studied.”
Regardless, we’ve seen that federal law doesn’t require that a causal relationship be demonstrated for a drug’s labeling to be revised with warnings of a new risk. Federal law only requires an association for such revision to become necessary.
Plaintiffs claim that Zofran’s labeling needed a warning for major birth defects. In the words of one plaintiff, whose case was filed in the United States District Court for the Eastern District of Texas, Texarkana Division, under case number 5:15-34:
“Despite an ability and obligation to add warnings, precautions and adverse reactions to the product labeling for Zofran without prior approval from the FDA, GSK willfully, knowingly, unfairly and deceptively chose not to do so for the sole purpose of maintaining its profits and marketing capabilities.”
Post-Marketing Research: Investigating New Benefits & Risks
We’ve seen that the FDA and GlaxoSmithKline have received significant epidemiological studies linking Zofran to an increased risk of birth defects, and plaintiffs in a wave of lawsuits claim that the company has received hundreds of birth defect reports. But is that enough evidence of a new risk for the FDA to force a labeling revision?
Similar situations in the past have seen the FDA consider epidemiological studies and adverse event reports as warning signs. More often not, the agency will instruct a drug manufacturer to perform new clinical trials designed specifically to investigate the risk that has been demonstrated by post-marketing research. In the meantime, the FDA will release a warning letter, informing both the health community and public of a potential danger.
We’ll cover two recent examples of this process below.
Post-Approval Trials Made GlaxoSmithKline’s Wellbutrin A Smoking Cessation Aid & Revealed Risks Of Suicidal Thinking
Conducting additional clinical trials after a drug is approved can actually help a drug company. Sometimes physicians realize that drugs prescribed for one purpose are also good at doing something else.
Initially, GlaxoSmithKline’s Wellbutrin was only approved as a treatment for Major Depressive Disorder. But doctors noticed that their patients taking Wellbutrin also had an easier time quitting smoking; their cravings for nicotine weren’t as intense as those in smokers on other antidepressants. GlaxoSmithKline saw potential in this unintended effect, and resubmitted the same active ingredient, bupropion hydrochloride, for approval as a smoking cessation aid. After a new round of rigorous clinical trials, and significant rebranding, bupropion hydrochloride was FDA approved as GlaxoSmithKline’s Zyban, a drug to help smokers quit.
But after several research studies found an increased risk of suicidal thinking in patients taking antidepressants, including bupropion, the FDA released a Public Health Advisory:
“Several recent scientific publications suggest the possibility of an increased risk for suicidal behavior in adults who are being treated with antidepressant medications. Even before these reports became available, the FDA began a complete review of all available data to determine whether there is an increased risk of suicidality (suicidal thinking or behavior) in adults being treated with antidepressant medications. It is expected that this review will take a year or longer to complete. In the meantime, FDA is highlighting that:
Adults being treated with antidepressant medications, particularly those being treated for depression, should be watched closely for worsening of depression and for increased suicidal thinking or behavior. Close watching may be especially important early in treatment, or when the dose is changed, either increased or decreased” [emphasis added].
As you can see, the FDA did not wait for further studies before warning the public about the association between antidepressants and suicidal thinking.
Zofran’s Link To Abnormal Cardiovascular Activity
After becoming aware of a potential link between Zofran and QT prolongation, a change in the heart’s electrical activity that can lead to fatal conditions, the FDA’s response was similar. On September 15, 2011, the FDA released a Drug Safety Communication titled “Abnormal heart rhythms may be associated with use of Zofran (ondansetron).”
Here’s how the FDA explained its decision to “inform[…] the public” of Zofran’s safety review:
“The FDA previously noted cardiovascular safety concerns that suggested Zofran (ondansetron) could cause QT prolongation, which can lead to a serious and sometimes fatal heart rhythm called Torsade de Pointes. Additionally, there are articles published in the medical literature that describe QT interval prolongation with ondansetron.”
The FDA’s public announcement referred to three different “articles,” or studies. Two were clinical trials in which a total of 55 patients were administered ondansetron. The third was a “retrospective cohort” study that looked at how the cardiac rhythms of 76 pediatric patients changed after they were administered ondansetron. Note that a “retrospective cohort” study is a type of observational epidemiological study, and substantially similar to the four studies that have found an association between Zofran and birth defects.
In response to these three studies, along with adverse event reports, the FDA added a new, provisional warning to Zofran’s labeling and instructed GlaxoSmithKline to “conduct a thorough QT study to determine the degree to which Zofran (ondansetron) may cause QT interval prolongation.”
It is essential to note that this labeling change occurred before further research confirmed Zofran’s relationship to QT prolongation. At the time of the FDA’s public announcement, only an association had been demonstrated. GlaxoSmithKline and the FDA had received warning signs, but no conclusive proof. Even so, these warning signs were enough for the FDA to inform the public, revise Zofran’s warning label and compel GlaxoSmithKline to pursue further research.
According to plaintiffs in the recent Zofran birth defect lawsuits, both the FDA and GlaxoSmithKline have now received significant warning signs associating Zofran with major congenital abnormalities. But neither have informed the public or the medical community, and the FDA has yet to instruct GlaxoSmithKline to conduct further research.
In the eyes of these American parents, Zofran’s current labeling instructions are woefully inadequate, leaving doctors without vital information and no way to weigh Zofran’s potential benefits against its potential risks.