WASHINGTON -- The U.S. Food and Drug Administration hasn't done enough to monitor clinical trial data coming from overseas sites and patients, which increasingly dominate the studies used to support drug applications in the United States, according to a report from the Department of Health and Human Services' inspector general.
More than three-quarters of the patients participating in clinical studies submitted for 121 drugs approved by the FDA in fiscal year 2008 were from foreign sites, according to the report, submitted by HHS Inspector General Daniel R. Levinson.
For 10 of these drugs, all the supporting clinical data came from overseas trials, wrote Levinson, whose staff audited all marketing applications approved that year.
FDA Oversight Lacking
Yet despite this heavy reliance on foreign data, less than 1 percent of trial sites outside the U.S. underwent FDA inspection, the audit indicated.
Moreover, the HHS investigators found, "sponsors are increasingly conducting early-phase clinical trials outside the United States without INDs [Investigational New Drug applications]," leaving the FDA unaware that these studies are taking place.
Levinson wrote that the FDA should step up its inspection of foreign trial sites and work with its equivalents in other countries to share inspection findings and future plans.
He also recommended that the FDA begin "taking steps to encourage sponsors to voluntarily consult with FDA on their clinical trial protocols or submit INDs to the agency" when they begin human drug testing overseas.
Finally, the report suggested that the FDA should require that trial data be submitted in a standardized electronic format, after investigators found that the agency was unable to find some trial data because of missing files and incomplete submissions from sponsors.
Indeed, the audit excluded eight applications approved that year because the FDA couldn't locate them.
Dr. Joshua Sharfstein, the FDA's deputy commissioner, in a memo published with the report, agreed with the recommendations and said the agency was addressing each of the problem areas.
Disconnect with North America
Some of the potential problems by siting important trials overseas were highlighted in last year's PLATO study of the antiplatelet drug ticagrelor (Brillinta), which is expected to win FDA approval soon.
PLATO compared ticagrelor to clopidogrel in nearly 19,000 patients, about 90 percent of whom were from outside the U.S. and Canada.
Ticagrelor reduced the risk of cardiovascular events by 11.7 percent relative to clopidogrel overall. But a subgroup analysis found that ticagrelor apparently had no benefit for the study's 1,814 North American participants.
Cardiologists were divided as to whether that finding was anything more than a statistical fluke. But it certainly did nothing to dispel concerns that trials conducted overseas may be sloppier or that genetic, cultural, and environmental differences from Americans mean the data are of questionable relevance to U.S. patients.
As Levinson's report put it, "concerns cited by medical ethicists include the ability of local regulatory bodies and institutional review boards (IRB) to adequately monitor clinical trials to protect the rights and welfare of subjects and to ensure data integrity. Other critics question the extent to which the results from foreign clinical trials conducted in developing countries are generalizable to the U.S. population."
In a statement, Dr. Sidney Wolfe, director of the health arm of the consumer advocacy group Public Citizen and an outspoken critic of the FDA, said the HHS report "confirms some of the worst fears about the serious dangers of the escalating globalization of human experimentation."
"In addition to the increased dangers to human subjects in many countries because they have less adequate protections than in the United States, there are significant threats to the integrity of the data being generated from these experiments because of the decreased ability of FDA to monitor and inspect foreign sites," Wolfe wrote.
Dr. Steve Nissen, a cardiologist at the Cleveland Clinic, told MedPage Today and ABC News in an email that the increasing reliance on foreign trial data is "a significant problem."
"Patients enrolled overseas may differ substantially from U.S. patients. For example, some drugs (rosuvastatin, Crestor) achieve high blood levels in Asians. Other drugs have a different profile of adverse events in non-Caucasians," Nissen said.
He also pointed out that trials may be conducted in countries with high smoking rates and low use of lipid-lowering drugs, which "may make therapies look more efficacious than observed in well-treated non-smoking subjects."
Nissen added that language differences can make documentation such as hospital records harder to interpret.
But Harvard's Dr. Christopher Cannon, a TIMI trials investigator, said overseas data don't pose problems. "The trials we are involved with have equally high standards at all site in all countries," he said. "We do the same oversight of monitoring at sites in all countries."
No Matter How It's Sliced...
Other physicians also said they were less worried by the report's findings.
"The important thing is the quality of the trial, not where it was conducted," said Dr. Larry Norton of Memorial Sloan-Kettering Cancer Center in New York City.
Perhaps the report's most astonishing finding was that 78 percent of the 299,701 participants in trials submitted with the applications approved in 2008 were from outside the U.S.; 54 percent of the 11,944 trial sites were foreign.
The region that appeared most popular for conducting overseas trials was Western Europe, accounting for nearly 60 percent of overseas sites and participants.
But, the HHS investigators found, Central and South America was the favorite location when sponsors sought large numbers of participants per site. The average number of participants per trial site there was about 130, compared with less than 40 in Western Europe.
The report also indicated that FDA inspectors never visited trial sites in seven Latin American countries that accounted for some 40,000 trial participants.
When the agency did conduct inspections, it was usually when the trial site was especially large or when the drug had not yet been approved in the U.S. for any purpose.
Overseas sites were inspected much less frequently than those in the U.S., the investigation disclosed. "Our regression analysis indicated that FDA was 16 times more likely to inspect a clinical investigator at a domestic site than a foreign site," according to the report.
FDA staff told the investigators that logistical challenges hamper their ability to inspect foreign trial sites.
"According to these officials, inspectors are generally allowed one week, including travel time, to conduct these inspections. FDA is unable to easily extend the inspections if significant compliance issues or other problems arise," the report said.
Other obstacles cited by the FDA included the need for work visas and translators, as well as the expense.
Officials told the HHS investigators that each inspection costs about $40,000 and the trend toward multi-site trials with small numbers of participants per site made it more difficult for the agency to justify the costs of inspections.
FDA's Sharfstein indicated that the agency shared the concerns expressed in the report and was taking steps to address them.
"For each [Office of the Inspector General] recommendation, the Agency either has ongoing efforts to address the recommendation or has initiated development of new procedures that will incorporate the recommendation," he wrote.