USE OF ADVISORY COMMITTEES BY THE FOOD AND

DRUG ADMINISTRATION

THURSDAY, MARCH 7, 1974

HOUSE OF REPRESENTATIVES,

INTERGOVERNMENTAL RELATIONS SUBCOMMITTEE

OF THE COMMITTEE ON GOVERNMENT OPERATIONS,

Washington, D.C.

The subcommittee met, pursuant to notice, at 9:45 a.m., in room 2154, Rayburn House Office Building, Hon. L. H. Fountain (chairman of the subcommittee) presiding.

Present: Representatives L. H. Fountain, Don Fuqua, Clarence J. Brown of Ohio, John H. Buchanan, Jr., and Garry Brown of Michigan.

Also present: Dr. Delphis C. Goldberg, professional staff member; Gilbert S. Goldhammer, consultant; and Richard L. Thompson, minority professional staff, Committee on Government Operations.

Mr. FOUNTAIN. Let the subcommittee come to order. Let the record show that a quorum is present.

One of our members is momentarily on the telephone but will be back with us shortly.

We will be continuing our hearings this morning by taking further testimony from FDA Commissioner Schmidt and his staff on the use of advisory committees by the FDA. However, as we informed you yesterday, Dr. Schmidt, we will interrupt our general questioning to focus on two case studies this morning and then, as time permits, we will resume our more general line of questioning on the use of advisory committees.

We will now examine FDA's use of an advisory committee in connection with your approval of a drug, propranolol, for the treatment of angina pectoris.

This drug is manufactured by Averst Laboratories and is also known by the trade name "Inderal." It had been previously approved for marketing in November of 1967 for the treatment of cardiac arrhythmia.

When the subcommittee staff met with you on January 16, 1974, I believe Dr. Goldberg gave you, for your information and study, an analysis which he had made as a member of the staff of the subcommittee, of the new drug application for the use of propranolol for angina. That analysis deals specifically with the role of the Cardiovascular and Renal Advisory Committee in reviewing the evidence for propranolol, and the relationship of the committee's review to FDA's approval of the drug for treating angina.

The subcommittee staff analysis, which is dated January 16, 1974, will be included in the record at this point with the deletion of the names of the advisory committee members.

[The document referred to follows:]

JANUARY 16, 1974.

CHRONOLOGY OF NDA 16-762 FOR PROPRANOLOL ("INDERAL") FOR ANGINA PECTORIS

Original NDA for the angina indication was submitted by Ayerst on April 4, 1968. This drug had previously been approved, on November 13, 1967, with very restrictive labeling for the treatment of cardiac arrhythmias.

This NDA was withdrawn on September 13, 1968, after a meeting between FDA and Ayerst representatives on the completeness of the application, and it was resubmitted on April 2, 1969.

On September 22, 1969, FDA's reviewing medical officer recommended issuance of a "not approvable letter" on both safety and efficacy grounds. The company was advised the NDA was not approvable on October 27, 1969. On November 26, 1969, Ayerst requested that the NDA be filed over protest. By letter of February 13, 1970, John Jennings informed Ayerst that the NDA remained inadequate and that the Bureau of Drugs was recommending that the Commissioner publish in the Federal Register a notice of opportunity for a hearing. On March 25, 1970, Ayerst wrote Commissioner Edwards requesting positive action on the Bureau of Drugs recommendation for a hearing.

On April 17, 1970, a five-man Ad Hoc committee (including four of the five members of FDA's Cardiovascular and Renal Advisory Committee which met April 13, 1973, plus one of the two absent members) unanimously concurred in the recommendation by FDA's medical staff that propranolol should not be approved for angina at that time. The committee based its conclusion on the need for adequate studies to demonstrate efficacy and resolve questions of safety (relating to congestive heart failure).

In a July 1968 summary, FDA's reviewing medical officer had concluded that this drug is "most likely effective" for treating angina but that the studies submitted by the sponsor do not demonstrate it. He stated: "79 uncontrolled studies here and 22 in Canada are submitted. They are practically useless and an open attempt at premarketing spread." He observed also that the data submitted in the NDA do not always jibe with the company's summary.

In a similar vein, a medical review of the literature dated July 31, 1968, states: "Again, as mirrored in the literature, one gets the impression that Inderal is effective in angina. However, the proof does not come up to our standards of efficacy evaluation."

Ayerst was advised of the NDA approval by Dr. Crout's letter of September 4, 1973.

A "summary of basis of approval," prepared on 7-28-73 by Dr. E. DeVaughn Belton, director of the Division of Cardio-Renal Drug Products, includes the following statements:

The labeling as outlined in this review reflects the literature review and recommendations made by the Cardiopulmonary Advisory Committee in their meeting of April 19, 1973. It also reflects the conclusions drawn from the literature by the Division from the review of the studies from the literature submitted in support of this indication. Although every study submitted did not in each instance meet all criteria, as a group they do provide evidence of a well-controlled quality to support safety and efficacy.

Data submitted previously in the New Drug Application and specifically that of the Gorlin study has been amplified, reanalyzed, and found not to be evidence of a significant safety problem.

With regard to the statutory requirement of substantial evidence of efficacy, note that Belton does not say any individual study meets FDA's criteria. Instead, he says "Although every study submitted did not in each instance meet all criteria, as a group they do provide evidence of a well-controlled quality. . . Concerning safety, Belton says the Gorlin study "has been amplified, reanalyzed, and found not to be evidence of a significant safety problem."

The analysis to which Dr. Belton refers is labeled Appendix #4 (Summary of Gorlin Study) by Dr. O'Neill. This analysis does not appear to eliminate the questions that were earlier attributed to the Gorlin study. It points out a numher of limitations of the Gorlin study including the fact that the subjects in the

Inderal groups were followed for almost one full year less than the control group and that while the controlled study involving 109 patients showed the incidence of congestive heart failure to be higher in the control group, this result was related to prior history of congestive heart failure and myocardial infarction. The analysis also calls attention to the limitations of the study design, reported data, and small sample sizes. It concludes: "It is possible that further analysis and additional information could reinforce or dilute the present conclusions drawn from the Gorlin study."

Concerning the safety question, the following statements by Dr.* appear on page 31 of the Advisory Committee's closed session :

The incidence of new heart failure if that is precipitated by the drug was a serious reservation 3 years ago in our mind and is much less serious in my mind today . . . there is no question patients do develop heart failure because of propranolol. I mean I think that's readily acknowledged. But the incidence is not immense.

In addition to Dr. Belton's Summary of Basis of Approval, the file contains an undated medical officer's summary signed by L. C. Akman, M.D. Four appendixes, including Dr. O'Neill's analysis of the Gorlin study, are attached to this review. Dr. Akman's recommendation for the approval of the NDA contains the following statement:

This Angina Pectoris indication is tentative pending completion of ongoing studies, and subject to the result and conclusions drawn from these studies.

Dr. Akman's medical summary does not contain an analysis of the 33 studies submitted to the advisory committee, nor was any other review and analysis by FDA staff found in the NDA file with respect to these studies.

The minutes for the closed session of the April 13, 1973, meeting of the Cardiovascular and Renal Advisory Committee are incomplete and misleading. The minutes state: "The committee considered the data available to it from studies, from the literature, and from the material presented in the open session. The committee concluded that these data establish that propranolol is safe and effective under certain conditions of use for the treatment of angina pectoris." At no point in the transcript of the closed session is there indication that the committee concluded these data "establish" that propranolol is safe and effective. In fact, the question of whether the studies constitute substantial evidence of effectiveness was never overtly raised. What the committee actually concluded, on the basis of a split vote, is that it would recommend mentioning angina in the package insert in a qualified way. The vote was four to one for adding this indication to the package insert, but subsequently, Dr. * wrote FDA on April 23 saying that his vote was qualified with the stipulation that an appropriate expert on carcinogenesis must approve the drug for long term, chronic therapy in view of the fact that two of the three beta blockers have been implicated as carcinogens. While this qualification was made by letter shortly after the meeting, Dr.'s* position is not reflected in the minutes by a footnote or other means.

It is clear from a reading of the transcript of the closed session that three of the five members present (a majority) were not aware of any adequately controlled clinical study showing the effectiveness of propranolol for angina.

This position was taken by Dr.* (page 11), Dr.* (pages 13-14) and Dr.* (page 21). Moreover, neither of the other two committee members present expressed the view that there are any adequately controlled clinical studies demonstrating the drug's effectiveness. Not only is this position of the committee not reported in the minutes, but the latter appears to convey the opposite impression. The actual basis for the advisory committee's recommendation to include angina in the labeling for propranolol is suggested in the minutes by the statement: "However, the extent to which the drug is now used by physicians in angina leads the committee to accept its use under well-defined labeling conditions." FDA's "dilemma" and the need to take some action because propranolol was being widely used for an unapproved purpose was emphasized by Dr. Belton in the closed session (page 8) as follows:

I think this is where we stand right now-and in light of the fact that we know down the road a year or two from now we will have acceptable studies. We can't predict what those studies will be. But this is the best information we have at this point, and this information we really feel we should tell the practicing physician out there who is using the drug, or we should tell him something. We're sort of in a dilemma.

*Names deleted to preserve confidentiality.

It is quite clear from the closed session transcript that committee members wh favored labeling the drug for angina were motivated not by evidence of effec tiveness but, rather, by their desire to provide information to the practicing physician. This is borne out by the following statements on pages 25-27:

Dr. *: What I'm worried about is the detail men have been telling the doctors to use it, and the doctors have been using it, and now we, the FDA is telling the drug company that it's all right for the detail men to have been telling the doctors? Is this the round robin? Or am I misinterpreting it? It sounds this way.

Dr. * It's a little bit that way. My feeling would be this way though : That you're right the detail men-I think they have had to be pretty cir cumspect, but, in effect, the doctor down the street or somebody else tells them to use it, and there is no evidence of instruction on the worries that we heard here today. And I would think-I would rather, if I have my choice, get some good statement in the package insert, and that would require the detail men to bring up these qualifications of not using it when there is a history of heart failure, and so forth.

Dr. * It doesn't seem to me we are being unscientific, Dr. * when we say this is being used, that the evidence is still incomplete, and that studies are underway, but when it is used the following precautions are in order. Mr. FOUNTAIN. As I understand it, Dr. Schmidt, you agreed at the conclusion of the January 16 meeting with our staff to provide your comments on that analysis after you had an opportunity to study it. Then on February 19 you wrote me that you were in the course of preparing a reply when you were notified that these hearings would be held and therefore chose to discuss the staff analysis as well as other staff documents, as part of your testimony today.

A discussion of this matter is contained in attachment VII to your prepared statement, but I don't believe I find that particular attachment responsive to the issues which are raised in our staff analysis. It doesn't appear to contest any of the statements or conclusions contained in Dr. Goldberg's analysis.

Does this indicate that you agree or that you accept the the staff's January 16 analysis?

accuracy

of

STATEMENT OF ALEXANDER M. SCHMIDT, M.D., COMMISSIONER, FOOD AND DRUG ADMINISTRATION; ACCOMPANIED BY J. RICHARD CROUT, M.D., DIRECTOR, BUREAU OF DRUGS; AND PETER BARTON HUTT, ASSISTANT GENERAL COUNSEL-Resumed

Dr. SCHMIDT. No, sir, it just seemed to me that a number of questions had been raised and since we were preparing for these hearings and since I was led to believe that these hearings would explore in-depth some of these issues, it seemed to me more productive to have a giveand-take discussion of these and bring the issues out at a time when you could question our responses and we could determine more accurately what was meant by the document. So that really, in essence, is why I deferred answering until this hearing.

Mr. FOUNTAIN. All right, would you take advantage of this opportunity, then, to give us your response and to specify those parts of the staff document with which you agree, and those parts with which you don't agree; the parts that are correct, or the parts that are incorrect, as you see them?

Dr. SCHMIDT. Well, in essence, I believe the document raises a couple of issues. One is a procedural issue in how the FDA staff does *Names deleted to preserve confidentiality.