FDA Hit for Poor Clinical Trial Oversight

Just when Food and Drug Administration officials thought they could relax a little after months debating legislation to ensure safe drug use, federal investigators leveled charges that agency oversight of clinical research is weak, disorganized, and thus unable to ensure the safety of clinical research participants. This criticism came from the Office of the Inspector General (OIG) in the Department of Health and Human Services (HHS), which has called repeatedly over the past decade for stronger efforts to monitor clinical research and to take action against rule violators. Sen. Charles Grassley (R-Iowa), who requested the OIG investigation as part of his campaign to uncover FDA malfeasance, promised to keep a close watch over how FDA implements the OIG recommendations.

Jill Wechsler

The OIG complains that FDA cannot identify all ongoing clinical trials, their associated trial sites or Institutional Review Boards (IRBs) overseeing the research. The main reason is that the agency lacks an efficient information system for tracking clinical research activity, site inspections, and resulting corrective actions.

Due to these inefficiencies as well as limited resources, FDA inspects less than 1% of relevant clinical trials. The OIG estimates that from 2000 to 2005, FDA was charged with overseeing some 350,000 trial sites involved in more than 15,000 clinical research programs for drugs and medical devices registered with FDA. However, the agency inspected only 2855 sites during this period, ranging from 633 BIMO inspections in 2001 to 854 inspections in 2003.

Similarly, FDA inspected less than 300 Institutional Review Boards (IRBs) each year, partly because the agency has no complete IRB database and thus cannot identify all review boards evaluating clinical trials for regulated products. FDA is working with the HHS' Office of Human Research Protections (OHRP) to establish a mandatory system for registering all IRBs. The OHRP database already lists more than 3500 review boards, primarily those at research organizations that receive National Institutes of Health (NIH) grants or other federal funding and have to register their IRBs as part of the federal "assurance process." FDA backs IRB registration as a way to improve agency communication with these organizations.

The OIG report [OEI-01-06-00160 available at www.oig.hhs.gov] is the latest in a series of investigations into FDA's bioresearch monitoring (BIMO) program. Back in 1998, the OIG issued reports criticizing FDA's ability to ensure that IRBs appropriately monitor research under their jurisdiction. Another investigation in 2000 documented weaknesses in FDA's clinical trial oversight program. The OIG is currently assessing how efficiently FDA disciplines clinical investigators charged with research misconduct plus how well the agency and sponsors can monitor financial conflicts of interest of clinical investigators.

Coordinating data

FDA established a Human Subject Protection/Bioresearch Monitoring Council in 2006 to better coordinate and streamline the agency's diverse BIMO program, partly in response to such criticism. The agency also has been clarifying research policies through the publication of several new guidances on timely research activities.

But the main problem, according to OIG, is that FDA needs a centralized database of all clinical trials involving drugs, biologics, and medical devices to better coordinate research oversight. FDA currently has six different databases collecting information on BIMO inspections for different Centers, as well as one in the Office of Regional Affairs, which manages the inspection field force. These systems collect different kinds of information, using different formats and definitions, which makes it difficult to coordinate inspections and track responses to citations.

In addition, the OIG criticizes FDA Centers for frequently reclassifying field recommendations for corrective action. The result is that the Centers often decide that an inspection subject may make promised corrective actions instead of receiving a warning letter. And when FDA Centers do send out warning letters, they often cannot determine whether a site takes corrective action.

Another factor undermining FDA BIMO enforcement is the agency's limited oversight authority. FDA rules governing clinical trials apply to sponsors and investigators, but not to clinical trial support staff that now play a large role in implementing studies and dealing with participants. And FDA policies do not extend to foreign study sites, which now produce more than 20% of clinical trial data submitted to the agency. FDA investigators do visit some foreign sites to ensure data integrity, but have little authority over patient protection issues.

In responding to the OIG report, FDA deputy commissioner and chief medical officer Janet Woodcock pointed out that BIMO inspections are just one narrow component of the agency's broader efforts to protect human participants in clinical research. More important is FDA's careful scrutiny of proposed research protocols, a process that usually leads to protocol revisions to ensure human subject safety before the trial begins. And even though FDA visits only a small fraction of sites, "the possibility of an inspection helps keep all parties aware of their responsibilities," Woodcock noted.

Because its resources for site inspections are limited, FDA is looking to apply risk-based approaches to all its field inspection programs, including BIMO activities. The aim is to target FDA oversight to those sites and operations involved in more vital or difficult research activities. This approach also involves establishing quality systems for BIMO activities that provide incentives for sponsors to ensure adherence to good clinical practices.

Registration & disclosure

In addition, the FDA Amendments Act (FDAAA) may address some OIG concerns about better protecting human research participants by expanding information on clinical research activities. The legislation, which was finalized by Congress in late September to reauthorize the prescription drug user fee program (PDUFA), also strengthens FDA authority to require clinical trial registration and disclosure of research results [see View from Washington, September 2007]. And increased user fees should provide additional funding to improve clinical research data banks as part of broader agency efforts to create an electronic platform for managing all regulated product information. FDA envisions the database recommended by OIG as an internal, nonpublic listing of all trials, as opposed to expanded registration of clinical trials on a public Web site required by FDAAA.

The details for expanding the existing clinical trial registration program [ClinicalTrials.gov] were uncertain until FDAAA was finalized and still require considerable explanation to be fully implemented. Registration now extends to all trials for drugs, biologics, and medical devices beyond Phase I intended to support a market application to the agency—and not just those for products that treat serious and life-threatening conditions. Trial listings have to fully describe each study, including name of intervention, start and completion dates, eligibility criteria, target number of subjects, and expected outcome measures. Listings also must provide information on site location, contacts, and updates on recruitment status.

Each listing will have a protocol identification number to facilitate tracking and eventual links to a results database. NIH will begin to establish this data bank in 90 days with information from FDA's application review process. This may include clinical trial data considered by advisory committees or posted in review packages. NIH will link to drug labels posted on its product label database, plus to publications found in its Medline Plus database.

Within a year, the results data bank will expand to include a listing of basic data elements for approved drugs: study participant demographics, trial outcomes, and any agreements between sponsors and investigators limiting results disclosure. During this period, FDA also will work with NIH to post additional information on serious and frequent adverse events of registered test drugs.

The legislation gives FDA, NIH, and sponsors three years to explore ways to provide more complete trial results information to the public. Issues to be decided are whether trial results should be posted for medical products never approved by the FDA; whether there should be nontechnical lay summaries of trials, as well as technical summaries; how much detailed protocol information would be required; time frames for submitting results information (e.g., if results of early trials have to be submitted before filing a market application); and how to register studies supporting new product uses.

FDA will hold a public meeting to discuss these issues and will conduct a pilot project on how best to verify that sponsor information is accurate and not false or promotional. Sponsors want clear rules and an effective information system because they will face fines and legal action for failure to comply with registration requirements.

Jill Wechsler is the Washington editor of Applied Clinical Trials,(301) 656-4634 jwechsler@advanstar.com