6. Parke, Davis is to issue a "Dear Doctor" letter by April 21, 1969. The text is to be submitted for approval by FDA no later than April 14, 1969.

This letter will include the following:

(a) A statement concerning the pre-1968 labeling which recommended use intramuscularly.

(b) A statement concerning the labeling changes initiated by FDA in 1968 limiting indications for the product to intravenous use and discouraging the IM and subcuteneous uses.

(c) A statement giving the results of the current study that Parke, Davis performed on Rocky Mountain spotted fever and typhoid fever which showed that the product, by the IM route, is not effective.

(d) The following statement from the new labeling: "chloramphenicol sodium succinate is intended for intravenous use only. It has been demonstrated to be ineffective when given intramuscularly."

(e) The letter shall commence with the prominent statement that chloramphenicol sodium succinate is not effective by the IM route.

(f) A statement that chloramphenicol is not the drug of choice for any of the indicated conditions.

HERBERT L. LEY, Jr., M.D., Commissioner of Food and Drugs.

Mr. FOUNTAIN. Our purpose this morning is to explore the sequence of events which culminated in this exchange of correspondence and to make the record perfectly clear. Our principal witness is Dr. Herbert L. Ley, Jr., Commissioner of Food and Drugs. Dr. Ley, will you come forward, please, along with any others from FDA whom you care to have with you during the course of the hearings. Would you please introduce your colleagues.

STATEMENT OF DR. HERBERT L. LEY, JR., COMMISSIONER OF FOOD AND DRUGS, CONSUMER PROTECTION AND ENVIRONMENTAL HEALTH SERVICE, PUBLIC HEALTH SERVICE, DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE; ACCOMPANIED BY WILLIAM W. GOODRICH, ASSISTANT GENERAL COUNSEL, HEW; J. KENNETH KIRK, ASSOCIATE COMMISSIONER FOR COMPLIANCE; B. HARVEY MINCHEW, M.D., ACTING DIRECTOR, BUREAU OF MEDICINE; JOHN JENNINGS, M.D., ACTING DEPUTY DIRECTOR, BUREAU OF MEDICINE; JEAN LOCKHART, M.D.; EDWIN M. ORTIZ, M.D.; ROBERT C. BRANDENBURG; AND MORTON M. SCHNEIDER, ALL OF FDA

Dr. LEY. On my left is Mr. William Goodrich of the Office of General Counsel, Health, Education, and Welfare. On my right is Dr. B. Harvey Minchew, Acting Director of the Bureau of Medicine of the Food and Drug Administration. And on his right is Mr. J. Kenneth Kirk, Associate Commissioner for Compliance in my office.

Mr. FOUNTAIN. We are delighted to have all of you with us. I believe we have had some of you before, although it seems quite a long period of time since we met in a hearing.

Dr. Ley, I note that your statement is somewhat lengthy. We don't want to rush through this hearing, but we are pressed for time since the subcommittee has another hearing this afternoon on a different matter. And, since the members of the subcommittee did receive copies of your statement well in advance of the hearings and copies are available for the press. would you care to give us a brief summary of

the major points in your statement and let us incorporate the entire text in the record.

I might add that on the basis of the statement which I have received, there appear to be some discrepancies beween the details presented in your statement and the documentation obtained from your files by the subcommittee's staff. So, for this reason, I think it is important that we obtain as clear and accurate an account of what has transpired in this matter as we possibly can.

Dr. LEY. Mr. Fountain, in terms of a brief summary, I believe I would prefer to omit the reading of the testimony to the bottom of page 6 and begin, if I may, with that point. This is a very difficult and complicated subject to summarize; so, I would prefer to delete this portion of the testimony and to remove certain direct verbatim quotes from the telegrams at the end, if this meets with your approval. Mr. FOUNTAIN. Don't leave out anything that you want to say. (The entire text of Dr. Ley's prepared statement follows:)

PREPARED STAtement of DR. HERBERT L. LEY, JR., M.D., COMMISSIONER OF FOOD AND DRUGS, CONSUMER PROTECTION AND ENVIRONMENTAL HEALTH SERVICE, PUBLIC HEALTH SERVICE, U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE

Mr. Chairman, I appreciate the opportunity of appearing before your committee today to discuss the antibiotic, chloramphenicol.

I propose to describe, primarily, events since early 1966, as requested in your letter of April 10, 1969. It has been suggested by your staff, however, that you would be interested in the earlier history of this drug as well, so I will begin with this phase.

Chloramphenicol was first isolated in 1947 from a soil sample collected in Venezuela. It was found to possess marked effectiveness against several gram negative bacteria and also exhibited antirickettsial and antiviral activity. Later clinical trials showed the antibiotic to be of value in treating a variety of infections, including typhus and typhoid fever.

After the chemical structural formula was determined, the antibiotic was prepared synthetically and patented by Parke, Davis & Co. The Parke, Davis new drug application for chloromycetin-that company's brand of chloramphenicolbecame effective on January 12, 1949. Later that year, under new legislation, chloramphenicol was classified as a "certifiable antibiotic," subject to the batch certification provisions of the Federal Food, Drug, and Cosmetic Act.

Beginning in 1950, after the drug was marketed, a few published reports drew attention to the possibility that chloramphenicol might cause serious, sometimes fatal, blood dyscrasias.

A nationwide survey of hospital and clinical case records carried out by FDA produced evidence that there were serious blood disorders including aplastic anemia, hypoplastic anemia, thrombocytopenia, and granulocytopenia-associated with the use of chloramphenicol. In June 1952, certification of chloramphenicol was discontinued pending a review of these data by a committee of outstanding hemotologists and internists established by the National Research Council. On the basis of the committee's report, labeling for the drug was revised to warn of blood dyscrasias associated with the use of chloramphenicol. In addition, the new labeling cautioned against using the drug indiscriminately or for minor infections.

In April 1960, the Council on Drugs of the American Medical Association reported that despite these warnings, physicians continued the use of chloramphenicol for minor infections, including those associated with the common cold. At FDA's request, the National Research Council again evaluated the use of chloramphenicol and recommended stronger warning information in the labeling. In February 1961, a prominent "warning box" was incorporated in the package insert and other labeling, stressing that chloramphenicol must not be used for trivial infections, such as colds, influenza, or throat infections. Parke, Davis was required to mail the new prescribing information to all physicians.

The "warning box" was strengthened further in 1966, but despite these repeated revisions, editorials in the Journal of the American Medical Association, and

warnings in other publications, such as the Medical Letter, the use of chloramphenicol continued to increase. Most of this, we believe, was for medical conditions for which the drug was not indicated or for which it was expressly prohibted, including warts, acne, the common cold, and other simple infections. I would like to submit a copy of the labeling as revised in 1966.1

The "box warning" in the labeling was strongly worded. It told the physicians quite bluntly the dangers of the drug, yet it had not accomplished its intended purpose. Despite the risks associated with chloramphenicol, use of the drug continued to be excessive.

In February 1968, we notified the Parke, Davis Co. that substantial revision in the labeling would be necessary. The new labeling was completed and approved in April 1968.

The revised labeling carried a carefully worded indications section expressed in restrictive terms. It included an estimate of the incidence of fatal aplastic anemia, based on a report made January 1, 1967, by the California Medical Association and State Department of Public Health. This estimate was 1:24,200 to 1:40,500, based on two dosage levels. The labeling also stated the desirability of hospitalizing patients being treated with chloramphenicol to facilitie appropriate studies and observation during therapy. Cautionary information was included regarding its use in pregnancy and lactation. Leukemia was listed as an additional adverse reaction.

On May 7, 1968, following approval of the new labeling, we sent a "Dear Doctor" letter to every physician in the country calling attention to the proper indications for use and the strengthened warning about the hazards of this drug. The letter also was sent to all hospital administrators. The revised labeling was included with the letters, along with a facsimile of our drug experience report form. Physicians were requested to report any adverse reactions associated with the use of chloramphenicol. I'd like to submit for the record a copy of this letter and the enclosures.

A review of data from several sources, including letters from physicians in response to our "Dear Doctor" letter of May 7, 1968, reveals that in about 80 percent of the fatalities which were reported to us, chloramphenicol was prescribed for other than accepted indications.

Let me turn now to the problem of drug equivalency, or biological availability, associated with chloramphenicol.

In the fall of 1966, after the Parke, Davis patent on chloramphenicol expired, other drug manufacturers began to submit applications for the certification of chloramphenicol. These applications were required to show where and how the products were manufactured, to demonstrate that the firms conformed to good manufacturing practices, and that the products met the chemical standards then applicable. No blood level studies were submitted with these applications. I arranged for a discussion of this matter between the Bureau of Medicine and the Bureau of Science. As the result of that meeting, we concluded that the manufacturing processes involved did not present significant control problems, and that satisfactory establishment inspection evidence and compliance with the monograph should be sufficient to insure safety and efficacy. Thus, no additional clinical trials were required; no blood level data to establish comparable absorption were requested. This conclusion-that conformance with the chemical tests and specifications and that the production of the product under good manufacturing practices would assure safety and effectiveness-turned out to be a mistaken one.

After these drugs were certified and marketed, Parke, Davis came to us with the results of blood level studies that had been performed with chloramphenicol capsules produced by other companies. These data indicated that the other capsules, when administered in a single dose, did not produce blood levels as high as those achieved with the Parke, Davis capsules. On October 19, 1967, we suspended certification of all chloramphenicol capsules produced by the new manufacturers. We then arranged for single dose blood level studies on chloramphenicol at Georgetown Medical School. We also carried out an additional single-dose study and, in conjunction with the Department of Defense, a multiple dose study. The results confirmed and extended the information submitted by Parke, Davis; in general, the newly certified products studied did not produce reliable absorption giving acceptable blood levels, either after a single dose or within the first 12 hours of multiple dosing.

1 Due to their length, the attachments to Dr. Ley's statement are not reproduced, but are available in the subcommittee's files.

While there was no evidence available through which we could directly correlate blood levels with clinical response, all of the clinical data to establish the effectiveness of this product had been derived from the use of the Parke, Davis capsules. We, therefore, concluded that capsules produced by other firms would have to achieve comparable blood levels.

On December 20, 1967, we met with representatives of all the producers to give them the requirements they would have to meet to show that their capsules should be certified. The firms were told they could (1) present test results showing that their capsules gave blood levels equivalent to those obtained with the Parke, Davis capsules, or (2) present data showing good clinical response from their capsules.

On January 17, 1968, having received no acceptable data, we decertified all outstanding stocks of these products and offered the firms the alternative of either stopping the sale and quarantining all such chloramphenicol, or recalling it to their possession.

Early in 1968, Rachelle did submit adequate blood level data for its capsules and the firm's one outstanding lot was recertified. In addition, the firm's new production of capsules was accepted for certification.

Rachelle also filed an IND to establish the safety and effectiveness of its chloramphenicol succinate for intravenous administration. Blood level data were submitted comparing Rachelle's and Parke, Davis' chloramphenicol succinate administered intravenously. Initially, it appeared that these data were adequate to justify certification. However, having experienced the variability of blood levels achieved with the use of capsules of different manufacturers, we reviewed information in our files and in the published literature to determine what was known about blood levels produced by parenteral forms of the drug when given by intramuscular, subcutaneous, or intravenous routes.

On March 15, 1968, one of our medical officers summarized the results of this review. He reported that there were no significant blood level data on subcutaneously administered chloramphenicol succinate, and that the blood level data on intramuscular use showed results significantly lower than those achieved by the intravenous route.

This led to a more critical review of the certification files on Parke, Davis' chloramphenicol succinate to determine what blood level and clinical data these contained to support the safety and effectiveness of the parenteral dosage forms. There was little meaningful clinical data in the files. We then concluded that parenteral forms of chloramphenicol, in order to be regarded as effective, would have to show blood levels comparable to those achieved by oral administration. By May 29, 1968, this study of the Parke, Davis data had been completed and the responsible Division in the Bureau of Medicine recommended that certification of chloramphenicol succinate should be discontinued for "all routes of administration." On June 4, 1968, the Office of Certification Services was directed to withhold further certification of chloramphenicol succinate pending resolution of the matter. This was done.

In early July 1968, Parke, Davis submitted two volumes of material, compiled largely from the literature, to substantiate the efficacy of intramuscularly and intravenously administered chloramphenicol succinate. As we have already noted, the blood level data from the literature were not helpful, and the clinical data were inadequate. Thus, the future status of parenteral chloramphenicol succinate had to turn on new blood level studies or the impending report of the NAS-NRC study group.

On July 17, 1968, representatives of Rachelle Laboratories met with staff members of our Bureau of Medicine and Office of Certification Services to discuss their firm's application for chloramphenicol succinate. They were advised of the recent developments in our evaluation of this product and of our conclusion that there had to be continued study of chloramphenicol succinate, without regard to who was manufacturing the drug.

Rachelle's representatives contended that FDA should either immediately certify their chloramphenicol succinate for intravenous use or recall the lots Parke, Davis had marketed prior to the suspension of certification.

It was FDA's position, however, that we should not certify additional batches of chloramphenicol succinate for any manufacturer without clarifying information on parenteral administration. But neither did we want to simply sweep chloramphenicol succinate off the market while these questions were being resolved. Chloramphenicol is intended for use only in extremely serious conditions, as you know, Mr. Chairman. Critically ill patients, particularly children, can

not always take a drug orally. This is when the physician must have an injectable form.

There are two parenteral forms of chloramphenicol available. One the aqueous solution of chloramphenicol base, has the disadvantage of pain on intramuscular injection or the risk of damaging veins on intravenous use. The labeling for this product carries the statement, "not recommended for pediatric use." The second parenteral form, chloramphenicol succinate, is much better tolerated and is the product customarily used in pediatric practice.

There was an obvious need to move as quickly as possible to resolve the issues involved. On July 22, 1968, Dr. Minchew summarized the situation for me, pointing out the need to expedite the report of the NAS-NRC on parenteral chlor amphenicol. We asked that the report be expedited and it was.

On August 9, 1968, we received the report on chloramphenicol. I submit copies for the record. The panel endorsed the stronger warnings that we had required in May, emphasized the hazards in the use of this antibiotic, and recommended the use of less hazardous agents where they could be expected to accomplish the desired therapeutic effect.

We reexamined the labeling that had been developed in May for the various chloramphenicol preparations and concluded that the labeling for the parenteral forms required further revision in the light of the blood level data that were known to us. The academy review panel had noted the higher and preferable blood levels obtained by intravenous use as compared with intramuscular administration. It recommended that the patient be shifted to oral chloramphenicol as soon as possible since these preparations gave better blood levels. On August 15, 1968, FDA publicly reported the academy panel's findings and announced that certification of chloramphenicol succinate would be resumed after labeling was revised. The news release on this decision emphasized that this drug was for intravenous use and cautioned against intramuscular use.

New labeling reflecting these recommendations was approved September 3, 1968, and subsequently was the subject of a Federal Register publication (October 19, 1968). I submit for the record, a copy of the revised labeling.

Particularly noteworthy here, as we indicated in the news release, is the fact that the product was designated chloramphenicol succinate for intravenous injection. The indications were to use the product for intravenous injection. The labeling stated that while intramuscular injection had been used, it was not recommended. This distinction was also incorporated into supplement E of the 1968 Physicians Desk Reference which was issued in mid-November 1968.

Both Parke, Davis and Rachelle were notified in September that the new labeling had been approved and that the FDA would certify chloramphenicol succinate on the basis of the revised labeling. Since September 1968, no chloramphenicol succinate has been certified unless accompanied by the new labeling. Rachelle had not previously marketed chloramphenicol in this form. Parke, Davis did have chloramphenicol succinate in distribution which had been certified prior to June 4, 1968.

In your letter of April 4, Mr. Chairman, you asked FDA to explain why no action had been taken to assure the removal from the market of these lots of chloramphenicol succinate which had been distributed by Parke, Davis with the old labeling.

On July 26, 1968, we had been advised that the company had a 4-month supply of single dose vials of chloramphenicol succinate and a 13-month supply of multiple-pack vials for hospital use. We were told then that 6 weeks would be needed, after the approval of new labeling, to prepare the new insert, package the product, and make the distribution to warehouses.

Therefore, when the revised labeling for chloramphenicol succinate for intravenous injection was approved in September 1968, we understood that a small supply of single-dose vials, and a larger supply of institutional packages, remained in distribution. We did not consider a recall of this material essential because we anticipated that the new production would move out promptly with the revised prescribing information.

But in November 1968, Rachelle complained that the Parke, Davis detail force was using the claim that their firm's product permitted intramuscular use, while Rachelle's did not. We made inquiries and, while we were unable to obtain evidence that Parke, Davis was promoting its product for other than intravenous use, we did find that the Parke, Davis product apparently was purchased, in some cases, because of the indications for intramuscular use.

We have noted that at about this same time, in mid-November, the Physicians Desk Reference mailed supplement E of its 1968 edition to all physicians in the