UK Drug Regulator Destroys All Trial Data After 15 Years

Cochrane Investigator Says His Attempts to Obtain Data on Fluoxetine Were Thwarted

Caroline Helwick

July 11, 2011

July 11, 2011 — Clinical researchers in search of drug data in the United Kingdom may come up empty-handed if the data are more than 15 years old.

After storing data for 15 years, the UK drug regulator, the Medicines and Healthcare products Regulatory Agency (MHRA), routinely destroys all the evidence about the benefits and harms of drugs it approves, according to the director of the Nordic Cochrane Centre in Copenhagen, Denmark.

In a letter to the British Medical Journal published online July 7, Peter C. Gøtzsche, MD, described his efforts to obtain data on fluoxetine for an analysis of antidepressant medications he is performing with Anders W. Jørgensen, a PhD student. Dr. Gøtzsche revealed that access to unpublished trial reports in some cases was denied because the data had been "destroyed."

The researchers recently applied to the European Medicines Agency (EMA) for access to unpublished trial reports for 8 antidepressant drugs approved for use in the European Union (EU).

Because some drugs, including fluoxetine, have not been approved centrally, the EMA advised them to contact the relevant national drug agencies. For fluoxetine, the United Kingdom acts as the Reference Member State, according to the Mutual Recognition Procedure in the EU. The response from the MHRA, however, was less than helpful. In fact, it was shockingly negative.

"The MHRA informed us that it no longer holds the requested reports," Dr. Gøtzsche told Medscape Medical News.

Under the MHRA record management policy, all application files and data for licenses are held for 15 years. After this point, files are destroyed unless they are needed for "legal, regulatory or business" reasons, or unless they are considered to be of "lasting historic interest," he said.

"Legal or historic interest? How ironic," the investigators write in their letter to the BMJ.

This was not the first time Dr. Gøtzsche has had trouble gaining access to clinical trial data. In May 2011, he and Mr. Jørgensen published an account of their difficulties in accessing data regarding 2 antiobesity drugs, rimonabant and orlistat (BMJ. 2011;342:d2686).

In that publication, the investigators noted that clinical trial results are "reported selectively" and that comparisons of published trials with unpublished data from drug regulatory agencies show that the benefits of drugs are "much over-rated and the harms under-rated." They further maintained that "selective reporting can have disastrous consequences."

Objectivity Needed

In the field of psychiatry, objectivity is particularly needed, said Dr. Gøtzsche.

"Court cases have shown serious scientific misconduct in placebo-controlled industry-sponsored trials of antidepressant drugs, including fluoxetine — for example, re-coding suicidal events as 'emotional lability,' 'hospital admission,' 'lack of effect,' or ‘drop-out' while patients were taking the drugs, and adding suicides to the placebo group that had not occurred while patients were taking placebo," he notes in the current letter.

This practice underlies his current research, which aims "to look at the details and study the adverse effects before the drug companies have interpreted and misrepresented them."

"For adverse effects, I’m thinking of suicidality, which has been documented for children though not adults. We are very interested in looking at these data," he told Medscape Medical News.

The action of the MHRA may make it impossible for independent researchers to correct the “seriously flawed publication record” on fluoxetine, he maintained, which is the only drug approved for use in childhood depression.

"So much for drug safety," he added.

The destruction of data has far-reaching implications for all research, he further suggested.

"I cannot think of anything more important in clinical research than to get access to full clinical study reports that drug companies make on their drugs," he said.

Previous Trials Inform Future Research

John Powers, MD, formerly a lead medical officer with the US Food and Drug Administration (FDA) and now associate clinical professor of medicine at George Washington University School of Medicine, agreed with Dr. Gøtzsche on this point.

"I worked at the FDA for a decade and was not aware that FDA discarded any previous applications. I remember looking at prior New Drug Applications that had been stored for decades," he said.

Dr. Powers said comparative trials designed to show the similarity of new agents vs older agents are meaningless unless all the study designs are similar.

"You would not be able to conduct appropriate studies if all the information from previous trials were discarded," he said, adding that the same holds true for meta-analyses. "It doesn’t make sense to be discarding data that is the basis for current and future trials."

Furthermore, he added, clinical hypotheses are built upon a foundation of prior work that informs the new hypothesis.

"A lot of information is in the historical evidence. By discarding this work, you weaken your current work as well."

FDA Response

"The FDA cannot comment specifically on any other agency’s or country’s plans for records retention," said Sandy Walsh from the FDA’s Office of Public Affairs. "For our own records, we develop retention plans in conjunction with the National Archives and Records Administration (NARA)."

Once approved by NARA, the Records Control Schedule determines how long records are to be retained and when and how they are to be disposed of. For drugs that are approved, all supporting materials are kept until they are "withdrawn by the commissioner."

Once withdrawn, or if the drug was never approved, the application may be deemed "historically significant" by the Records Control Schedule and retained by the FDA for 20 years, then transferred to NARA for long-term, indefinite preservation. If the record does not meet the "historically significant" criteria, it becomes "temporary" and is not transferred to NARA but is retained by the FDA for 30 years, then destroyed, according to Ms. Walsh.

Data from the Adverse Event Reporting System are retained for 30 years after the case is closed or "when no longer needed for trend analysis or reference purposes, whichever is latest," Ms. Walsh said. "Hence, even though data are 'temporary records' we use the latter option to maintain it 'forever.' "

Physical copies used for data input and scanning are destroyed once the data are entered and no longer needed.

Important Precedent

After their 3-year struggle to acquire documents on the antiobesity drugs, Dr. Gøtzsche and colleagues succeeded in effecting changes that at least hastened the EMA’s responsiveness. "Our case set an important precedent," he said.

Dr. Gøtzsche has now asked the United Kingdom Department of Health and the EMA for help in obtaining the data the MHRA has destroyed, and he has recommended that legislation be introduced to prevent further damage. Because documents can be scanned, he noted, "lack of space is no excuse."

He has also discussed the problem with the European Commission and the European Parliament. "They were shocked to learn about this," he said.

Meanwhile, he reported that his analysis of antidepressants "is in full swing," and he and his colleague are attempting to obtain the fluoxetine data through other sources. "We can’t get it from Eli Lilly, of course," he added. "No drug company will give you what you ask for."

Dr. Gøtzsche, Dr. Powers, and Ms. Walsh have disclosed no relevant financial relationships.

BMJ. 2011;343;d4203. Abstract


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.