Company accused of altering drug study
Paper co-written by Sick Kids MD understated effects on children, former U.S. regulator claims
A medical journal article co-authored by Sick Kids hospital’s top pediatrician was manipulated by a drug company and understated the risks of a powerful antipsychotic used to treat kids with behavioural problems, the former head of the U.S. drug regulator alleges.
The 2003 article concluded there was no correlation between long-term use of Risperdal and an increased risk of certain side-effects, including the growth of breasts in boys.
The listed authors included respected experts in the pediatric field — Toronto’s Dr. Denis Daneman and two U.S. doctors — and three employees from Janssen, which makes Risperdal.
The medical article was singled out in recent U.S. lawsuits against Janssen as an alleged example of drug company influence on doctors and published research. A court exhibit shows a draft of the study describing an association with certain side-effects — a finding that a Janssen employee internally flagged as “significant.” This information is not in the published version of the article.
“My name is on an article in which there is some data that has been left out. That, to me, crosses a line,” Daneman, pediatrician-in-chief at the Hospital for Sick Children, told the Star.
This week, the Star reported that an internal investigation at Toronto’s University Health Network found falsified data in two scientific articles co-authored by one of its top doctors.
This is a story about the role of a drug company in the production of a scientific journal article.
The alleged manipulation of the medical article is outlined in a report by expert witness Dr. David Kessler, former head of the U.S. Food and Drug Administration and a pediatrician by training.
The report, filed in U.S. court, said Janssen “controlled and influenced” more than 40 manuscripts, including the one co-authored by Daneman. Kessler called the company’s alleged promotion of the drug for unapproved use in vulnerable children “deeply troubling.”
In a statement to the Star, the drug company said patient safety is its foremost priority and it is “committed to the highest standards of scientific integrity.”
“We do not agree with Dr. Kessler’s opinions expressed in his report,” a Janssen spokeswoman said.
An independent biostatistician is now reanalyzing the 2003 study’s data to determine whether the original results stand or if they need to be corrected on the public record, Daneman said.
“If, indeed, these allegations are true, then I would feel used,” he said, adding that his contributions to the paper “were made in good faith and based on the assumption that my colleagues and I had access to all the relevant data.”
He said he was not included in internal Janssen discussions about certain revisions to the article.
“There is this deception there that is intolerable,” he said.
In a 2012 deposition where he was questioned by a U.S. plaintiff’s lawyer, Daneman, a pediatric endocrinologist, agreed the study he participated in had data calculation errors that had the effect of understating the frequency of side-effects.
In an interview with the Star, Daneman said there was no intent on his part to exclude significant results.
Kessler’s analysis was prepared at the request of Philadelphia lawyer Stephen Sheller, who represents more than 400 clients who claim they were misled about Risperdal’s risks and, after taking the drug, grew breasts — a condition known as gynecomastia.
Earlier this year, a Philadelphia jury awarded $2.5 million (U.S.) to an autistic man who grew size 46 DD breasts after being prescribed Risperdal when he was a child. (Janssen has filed a motion to have the verdict overturned.) In another case, the jury ruled the drug did not cause the plaintiff’s gynecomastia, though it did find the company failed to adequately warn “the extent of the risk” to which Risperdal could cause male breast growth.
There are still many more lawsuits waiting to be heard, Sheller said.
In Canada, a class-action lawsuit alleges Janssen manipulated clinical trial results, downplayed the risks and ghostwrote articles about Risperdal — “utilizing hired medical writers, who are not researchers or scientists, to write articles and then submitting them to selected opinion or ‘thought’ leaders to attach their names to them as authors without making any meaningful contribution to the article, to lend false credence to these articles.”
The suit, filed last year, claims the company then got these studies published in journals “to create the false impression of scientific acceptability.” The lawsuit has not been certified and its allegations have not been tested in court. Should the case proceed, Janssen said it “plans to vigorously defend itself.”
Studies have suggested that patients using risperidone, the generic name for Risperdal, more frequently experience elevated levels of prolactin, a hormone that stimulates breast development and lactation, compared to other similar antipsychotics. Elevated levels of this hormone can result in troubling conditions, including infertility and male breast growth, according to FDA researchers.
The 2003 study analyzed results from five Janssen clinical trials to investigate prolactin levels in children and teens taking Risperdal, and “explore any relationship with sideeffects hypothetically attributable to prolactin.”
Daneman said his role in the study was commenting on parts of the article and suggesting changes and that somebody else wrote it.
He told the Star he was paid about $1,000 by Janssen for his participation. “I gave the money to charity because of the concerns that had been raised about potential conflicts of interest,” he said.
The article that resulted from the analysis was published in the Journal of Clinical Psychiatry.
The study’s abstract, a summary at the beginning of the article that doctors often rely on for key take-away information that may inform their prescribing, said there was “no direct correlation between prolactin elevation” and side-effects such as gynecomastia.
However, the research had found a statistically significant association between elevated prolactin levels and side-effects during the eightweek to 12-week period of Risperdal treatment, Kessler said. A court exhibit shows a Janssen employee flagged the “significant value” in an earlier draft of the study, asking fellow company officials how they should “handle” it.
The published article does not mention the information. Kessler said the omission of the side-effect information “misled physicians and the scientific community.”
Daneman told the Star the information about the association with side-effects should have been included in the final article. In a deposition, he speculated the information was removed in the editing process, saying “sometimes the final article leaves out details that you would like to put in.” He said the article’s statement that there was “no correlation” was “inaccurate” and “defective,” but said it does not “nullify all the findings in that study.”
Kessler also found the study used an inflated denominator, cutting in half the actual rate at which boys under 10 taking Risperdal in the clinical trials developed breasts. In his deposition, Daneman agreed that the calculation should have been done differently.
Notwithstanding the problems raised in Kessler’s report, Daneman told the Star his medical journal article’s overall finding — prolactin levels taper off after an initial increase and there is no correlation between longterm Risperdal use and gynecomastia — is in line with the findings of other published research.
New research suggests otherwise. An epidemiologic study of data from roughly one million men in the U.S. found men taking risperidone were more likely to develop gynecomastia than men not taking the drug or taking other popular antipsychotics. A previous analysis by the U.S. FDA researchers found a disproportionate number of reported prolactinrelated side-effects for risperidone, and concluded there may be a link between the drug and these kinds of reactions.
Mahyar Etminan, a drug safety researcher at the University of British Columbia and lead author of the new study, said misleading journal articles can lead to serious consequences for doctors and patients.
“When you have this busy clinician who sees the paper is authored by one of the leading people in the field, they are more likely to just read the abstract and believe what they read.
“It can lead to inappropriate prescribing,” said Etminan, who, since conducting his research, has been retained as an expert consultant for lawsuits against Risperdal.
In a statement to the Star, Janssen said Risperdal continues to help countless people throughout the world suffering from debilitating mental illness.
In 2014, the FDA rejected a petition by Philadelphia lawyers champion- ing lawsuits against Risperdal to, in part, have the drug’s approval to be used in children revoked until longterm safety can be demonstrated. The FDA said “there is no evidence that the drug is unsafe.”
The unsuccessful petition, Janssen said, “was based substantially on Dr. Kessler’s report.” The lawyers are appealing the decision.
Daneman told the Star important parts of the study have been overlooked during the controversy, including the finding that numerous patients with suspected side-effects, including gynecomastia, had their symptoms resolve by the end of the roughly yearlong clinical trials.
At the time the study was published, Janssen, along with its parent company, Johnson & Johnson, were illegally promoting unapproved uses of Risperdal in vulnerable seniors and children, the U.S. government alleged in a recent prosecution of the company.
In his expert report, Kessler cited an internal Risperdal business plan, quoting the drug company as saying it aimed to “leverage the data and the business opportunity within the child/adolescent market via medical education.”
In 2013, the U.S. government announced the company would pay $2.2 billion (U.S.) in fines and settlements to resolve federal criminal and civil investigations concerning Risperdal and other drugs. The company has said the civil settlements are not an admission of wrongdoing.
As a result of the concerns raised over the study, Daneman and the lead author, U.S. pediatric psychiatrist Robert Findling, said they sought an independent biostatistician to reanalyze the findings.
“Once we have the results, we will determine whether the paper warrants full retraction, partial correction or whether the original findings stand,” Findling said in a statement.