June 21, 2007

Petition to Include an FDA Staff Presentation at Certain Advisory Committee Meetings

Andrew von Eschenbach, M.D., Commissioner
U.S. Food and Drug Administration
5600 Fishers Lane
Rockville, MD 20857

Dear Dr. von Eschenbach:

Public Citizen, representing 100,000 consumers nationwide, hereby petitions the Food and Drug Administration (FDA) for a policy statement requiring that all future advisory committee meetings discussing specific products include an oral scientific presentation by FDA staff on the products reviewed, including a thorough safety and efficacy assessment. In a study that we recently reported in the medical journal The Lancet,[1] we found that 49 public meetings of the FDA’s human drug advisory committees reviewing drug products since 1997 did not include an oral presentation on the product by the FDA (see Appendix). Because the sponsor can be relied upon to present—often in misleading terms—its own case, this leaves a seriously one-sided record upon which advisory committee members must make important public health recommendations. There is no excuse for such an omission, and it undermines the protective role that these committees play in helping ensure the effectiveness and safety of our nation’s medicines. 

The absence of an FDA presentation at an advisory committee meeting means the committee members are only presented, in person, with the data, analysis and opinions of the drug product’s sponsor.  The sponsor clearly has a strong incentive to cast a favorable light on its product. In advance of the meetings, committee members do receive briefing materials from both the drug sponsor and the agency, but there is no guarantee that members actually review the hundreds of pages of material in detail. 

It is quite possible that oral presentations at the committee meeting are significantly more effective at communicating the analysis and views of the FDA experts to committee members—and influencing their votes—than previously reviewed briefing materials. The disproportionate salience people generally attribute to recent experiences is a well-recognized cognitive bias, known as the “recency effect” in psychology.[2] And while some concerns of the FDA reviewer may arise in committee discussion even without an FDA presentation, these issues gain far greater force when they are part of an overall oral presentation by the agency.

In studying this issue, Public Citizen’s Health Research Group (HRG) used an existing database from a previous study on FDA advisory committees covering the period 2001-2004.[3] The database was expanded using similar methods to generate a database covering meetings between January 1, 1997, and June 30, 2006. Data were collected from meeting agendas and transcripts posted on the FDA’s Center for Drug Evaluation and Research (CDER) Web site.[4] We used the data to determine whether the FDA made an oral presentation at these meetings.   Oral FDA presentations were defined as scientific presentations of the FDA’s findings and views regarding a specific drug product (i.e., not reviews of applicable regulations or presentations of the questions to the committee). We did not include “general matters” meetings.

We found that the FDA did not make an oral presentation at 18% (49/275) of drug-specific advisory committee meetings between 1997 and mid-2006, but there was no trend over time. The proportion of meetings at which the FDA failed to give a presentation varied significantly over the nine-and-a-half-year period studied, but there was only one year (2004) in which the FDA presented at all relevant meetings. In 1997 and 2003, the FDA failed to give a presentation at one-third of the meetings. The drug sponsor presented at all but one of the meetings during this period.  This was a meeting of the Non-prescription Drugs Advisory Committee on June 12, 2003 regarding the generic over-the-counter medication syrup of ipecac.  The FDA did present at that meeting.

Advisory Committee Meetings on Drugs and the Occurrence of an FDA Presentation, 1997-2006*

Year	Advisory Committee Meetings	Meetings Without FDA Presentation	Percentage of Meetings Without FDA Presentations
1997	35	10	29%
1998	33	4	12%
1999	40	4	10%
2000	26	2	8%
2001	31	7	23%
2002	23	6	26%
2003	39	13	33%
2004	15	0	0%
2005	24	2	8%
2006*	9	1	11%
1997-2006*	275	49	18%
* Based on meeting data posted on CDER Web site by June 30, 2006.

It is unclear why the agency would fail to make a presentation at certain drug advisory committee meetings. One hypothesis is that the agency decided whether to present based on the anticipated outcome of the meeting.  We found a trend toward statistical significance between whether an FDA presentation was made and whether the committee voted favorably toward the drug product (RR: 0.76; 95% CI: 0.56-1.04). (This means that meetings with an FDA presentation were somewhat less likely to result in favorable votes for the drug.)  Conceivably, the FDA tended to present more often in meetings that had expected (and actual) outcomes unfavorable to the drug.  Alternatively, the presence of an FDA presentation might have made it less likely that the committee would vote in favor of the drug.  We found no relationship between FDA presentations and the “closeness”[5] of committee votes (RR: 1.05; 95% CI: 0.41-2.72).

The lack of an FDA presentation has the potential to influence the committees’ proceedings.  On May 17, 2006, for example, the Peripheral and Central Nervous System Drug Advisory Committee held a meeting to review an application by Novartis to add an indication to the cholinesterase inhibitor Exelon (rivastigmine) to treat the dementia associated with Parkinson’s disease. In its briefing materials for rivastigmine, the FDA had identified serious failures in the Novartis application: The company failed to demonstrate that dementia associated with Parkinson’s disease was clearly defined or could be clinically evaluated as a distinct form of dementia, and Novartis could not show that rivastigmine had any clinically meaningful benefit in well-designed and well-conducted studies, let alone that such findings could be replicated. 

Had the agency made a presentation on rivastigmine and countered the sponsor presentation, advisory committee members might have given more than the limited attention they gave to the problems identified in the FDA briefing materials. Proceeding without an FDA presentation, the meeting concluded with a committee vote favorable to the new indication for Parkinson’s. The agency, apparently overruling the reservations articulated by its reviewer, soon acted to approve the ill-supported indication.

In passing the Federal Advisory Committee Act in 1972, Congress ordered agencies to ensure “the advice and recommendations of the advisory committee will not be inappropriately influenced by…any special interest, but will instead be the result of the advisory committee’s independent judgment.”[6] An advisory committee cannot make an “independent judgment” if committee members are more familiar with the drug sponsor’s point of view than that of the FDA. That committee members hear from the FDA experts is critical, for it is the FDA, not self-interested drug companies, that is entrusted with protecting the public.

In the interest of improving drug safety and ensuring that advisory committee advice is informed not only by the perspectives of drug sponsors but also by the analysis and views of FDA experts, we strongly urge you to issue a policy statement requiring that all advisory committee meetings addressing specific products include an in-depth presentation to the committee by FDA experts on the products.

Sincerely, 

Asa Tapley
Research Associate

Peter Lurie, M.D., M.P.H.
Deputy Director 

Sidney M. Wolfe, M.D.
Director
Public Citizen’s Health Research Group

 

Appendix

Advisory Committee Meetings on Drugs without an FDA Presentation, 1997-2006† *

 

---