This is an archive page. The links are no longer being updated.

Testimony on Protecting the Nation's Safety and Health by David A. Kessler, M.D.
Commissioner, Food and Drug Administration
U.S. Department of Health and Human Services

Before the House Committee on Commerce, Subcommittee on Health and Environment
May 1, 1996

Chairman Bilirakis, and members of the Committee, I am pleased to be here today as you consider a very important public health issue -- ensuring that the Food and Drug Administration (FDA) protects the nation's safety and health. Let me say at the outset that we are committed, as I know you are, to ensuring timely access to new drugs and devices that are shown to be safe and effective. We look forward to working closely with you and your colleagues in the House and Senate on bipartisan legislation that accomplishes this goal.

INTRODUCTION

The FDA's primary Mission for 90 years has been to protect and promote the public health. I know you share our view that all Americans deserve assurance that the medicines they take or medical devices they need are safe and really work, and that the foods they eat are safe. We are committed to maintaining high standards so that consumers can continue to have that confidence and assurance. But high standards alone are not enough. Timely access to safe and effective products is important to the public health as well.

To meet the goal of timely approval of safe and effective products, FDA must be open to change. This is a time of unprecedented medical challenges and medical breakthroughs. An important measure of our value as an Agency will be our ability to keep pace with these changes. Thomas Jefferson said it best: "Our laws and institutions must (move forward hand in hand with the progress of the human mind."

As Commissioner, the first way I have sought to effect change is to attract a new generation of leaders to the Agency. Much of the progress FDA has made over the past few years began with our five new Center Directors. The many changes they have made have moved us closer to our goal. Second, we have vigorously pursued and implemented more than thirty FDA reinventing initiatives as part of President Clinton's and Vice President Gore's National Performance Review. Among the changes we have made, are speeding new cancer drugs to patients and targeting our limited resources to higher risk medical devices while no longer requiring FDA premarket review of hundreds of categories of simple devices such as tongue depressors and urinary catheters. Finally, in addition to the many changes we have already made, the administration is committed to supporting bipartisan legislation that does not lower current standards for product safety and efficacy.

If we are to achieve our shared goals of improving the Agency's ability to protect and promote the public health by assuring the public speedy access to safe and effective products, we must base our work on FDA's current performance. Unfortunately, too many of our critics justify the call for "reform" based on how the FDA did its job in the 1980's or earlier. They have missed the substantial progress that the dedicated doctors, nurses, engineers, chemists, microbiologists, biostatisticians, nutritionists and others at the FDA have achieved over the past several years. They would have us ignore the important lessons we have learned about the kind of change that will result in getting safe and effective drugs and devices to the market more quickly. Those who fail to recognize the Agency's performance and achievements threaten to undermine the real progress the Agency has made. While we share the goal of streamlining our operations and speeding safe, life saving drugs to the American public, undermining progress in critical public health and consumer protection is not "reform."

Let me explain what we have done.

THE FOOD AND DRUG ADMINISTRATION IN 1996

A. MEETING AND EXCEEDING THE PRESCRIPTION DRUG USER FEE PERFORMANCE GOALS

According to the General Accounting Office (GAO), the average approval time for new drug applications (NDA) submitted to the Agency in 1987 was 33 months. For NDAs submitted in 1992, the time had been reduced to 19 months. In 1995, approval times were even better: 16.5 months was the median approval time for the 82 NDAs approved in that year. For the 15 priority drugs, the median approval time was 6 months. These latest improvements in approval times have been made possible by shortening the time for completion of most first reviews of applications to only 12 months. How did we do it? In 1992, Congress, the Agency, and the pharmaceutical industry recognized that additional resources were one key to improving FDA's review of drugs and biologicals. Congress, led by a partisan initiative of this Committee, unanimously enacted the Prescription Drug User Fee Act of 1992 (PDUFA). The Agency, in turn, committed to very aggressive application review performance standards, with higher hurdles in each succeeding year until full implementation in fiscal year 1997. These performance standards were negotiated with, and agreed to by, the pharmaceutical and biotech industries.

We already have achieved one of the major 1997 performance goals. We achieved it in fiscal year 1994 -- a full three years ahead of schedule. For the drugs submitted to FDA in fiscal year 1994, we reviewed and acted upon 96 percent of them on time. In most cases, that meant first action within 12 months. 1 1 Under the provisions of the program, if a major amendment is submitted by the manufacturer-late in the process, an additional three months is granted.

We should never forget that the ultimate goal is ensuring that American patients have access to products that work and really make a difference. By that measure as well, the FDA is delivering. The story of AIDS therapies is especially telling.

Of the eight antivirals used to treat AIDS, FDA was first to approve them in seven instances, DDI, DDC, D4T, 3TC, saquinivar, retinovir, and indinavir. In the case of AZT, FDA acted contemporaneously with France and the United Kingdom. On March 1, the Agency approved ritonavir, the second in a new class of AIDS drugs called protease inhibitors, and we were the first country to do so. On March 13, we approved the newest protease inhibitor, indinavar, and this was done in 42 days. But the story goes beyond AIDS. Taxol for ovarian cancer, Fludarabine for lymphocytic leukemia, Pulmozyme for cystic fibrosis, Betaseron for multiple sclerosis, Riluzole for Lou Gehrig's disease, and Cognex for Alzheimer's were all first approved in the United States.

B. FDA IS A WORLD LEADER IN DRUG APPROVAL

One measure of FDA's drug review performance is found in how we compare to the performance of other major regulatory agencies around the world. Three separate studies demonstrate that FDA compares extremely well. First, at the request of this Committee, GAO looked at how FDA was performing, even before PDUFA was enacted. It found that by 1994, FDA review and approval times were faster than those in the United Kingdom -- a country whose regulatory system many critics like to cite as a way of doing things better and faster.

Second, FDA conducted an analysis of the "new molecular entities" (NMEs) introduced throughout the world between 1990 and 1994. We looked at when those drugs were approved here and in Germany, Japan, and the United Kingdom -- four countries that account for 60 percent of global pharmaceutical sales. Let me just give you a brief glimpse at the numbers in the FDA study, "Timely Access to New Drugs in the 1990s: An International Comparison." There were 58 of these NMEs that had been approved in both the United States and the United Kingdom. We approved 30 of them first, the United Kingdom 28. There were 44 of these NMEs approved in both the United States and Germany. We approved 31 of them first, Germany 13. There were 14 of these NMEs approved in both the United States and Japan. We approved 10 of them first, Japan 4.

Moreover, it is not just the numbers that compared well. We also found that there are numerous drugs with important therapeutic values available here, but not in these other countries. On the other hand, virtually all of the drugs available in these other countries but not approved here have closely related drugs (i.e., therapeutic equivalents) in the United States. Moreover, the United States is first to approve a significant proportion of "global" drugs -- those ultimately approved by more than one country.

The third, and most recent analysis was conducted by the Centre for Medicines Research, an industry funded, not-for-profit research group in the United Kingdom. This analysis once again demonstrated that FDA is a world leader in both the quality of its drug reviews and the timeliness of its approval. CMR found that FDA's median approval time for NMEs in calendar year 1994 and 1995 was as short as that in the United Kingdom and shorter than those in France, Spain, Germany, Australia, Japan, Italy, and Canada. Moreover, the FDA has had more first launches of worldwide NMEs than any single European country since 1990. In fact, CMR's analysis of worldwide NMEs launched in the United States and Europe showed that the United States has had a higher percentage of first launches than the top three European countries combined.

What these three studies demonstrate is that if you are an American patient, you have access to therapeutically important new drugs that have been proven safe and effective sooner than any other country's citizens.

C. MAKING NEW DRUGS AVAILABLE SOONER

In addition to shortening review times, the Agency has developed a number of innovative and well accepted mechanisms that give seriously ill and dying patients access to potentially valuable therapies outside of the clinical trials needed for drug development before they are approved for marketing. We recognize that this approach has its risks. But we also recognize that when it comes to getting needed therapies to terminally ill patients, the riskiest thing we can do is to be unwilling to take risks.

We have instituted innovative pre-approval policies, such as the "treatment IND," under which patients for whom there is no satisfactory alternative treatment may get access to a promising drug after clinical testing shows that it may be effective, but before it is approved for marketing. Under FDA's "treatment IND" provision, treatment is given under simplified protocols that collect important safety and, where appropriate, efficacy data.

We also have developed a process to accelerate approval of drugs for serious and life-threatening diseases, by approving such drugs for marketing before the data from traditional clinical trials are complete, if the drug shows an effect on a "surrogate marker." A surrogate marker is a laboratory measurement, such as an increase in the number of CD4 cells in a person with HIV, that is thought to predict actual clinical outcome, such as longer life or a decrease in infections. Recently, we announced a new initiative for cancer drugs that incorporates accelerated approval using tumor shrinkage as the surrogate marker. "Accelerated approval" is a kind of "conditional approval," because the applicant must agree to conduct further studies after marketing to measure the actual clinical outcome, not just the surrogate marker. If the studies do not verify a clinical benefit, FDA can revoke the approval. We have already approved ten drugs under this accelerated approval process since it was. put in place in December 1992. They include six drugs to treat AIDs and AIDs related illnesses, three drugs to treat cancer, and one drug to treat multiple sclerosis.

Finally, we allow other experimental drugs to be made available under an emergency or single patient IND for patients who are in need of a drug but do not meet protocol criteria. As a result of these programs, since 1987 more than 75,000 patients have received access to promising new drugs while they were being studied.

D. IMPROVING MEDICAL DEVICE REVIEW TIMES

We also are making significant strides in strengthening our medical device program. Improvements have been made in the times for the review and approval of medical devices. With respect to the so- called 510(k) process, through which 98 percent of all medical devices evaluated by FDA reach the market, the average review time in fiscal year 1995 was 138 days, down 24 percent from the 182-day average in fiscal year 1994. Moreover, backlogs have been virtually eliminated. Only nine 510(k)s were overdue at the end of FY 95, compared to 500 a year ago and nearly 2000 in December, 1993. The average review time for those devices needing premarket applications (PMAs), which in fiscal year 1995 was 20 months, is still too long, but we are making progress there as well.

It is important to remember, however, where the Agency was in the 1980's. Congress found -- as did the Agency -- that the standards for the review of clinical data on how to use medical devices were lacking. FDA looked at devices the way an engineer would; we did not focus sufficient attention on whether the device would actually make the patient better. Congress told us to do better, and we have. The clinical standards for medical device reviews have been bolstered significantly, and now we are hard at work on reducing review times even further.

E. IMPROVING ANIMAL DRUG AND FOOD ADDITIVE REVIEW TIMES

The American food supply is one of the safest in the world. The premarket determinations that animal drugs are safe and effective, especially when they are intended for use in food producing animals, and that food additives and color additives are safe constitute critical links in the food safety chain.

As is the case with human drugs and devices, the Agency has recognized the need to improve the timeliness of the product review processes in both these areas. In the new animal drug area, approval times that had steadily increased between 1988 and 1994 have begun to decline. This is particularly significant given the changing nature of the types of applications being submitted to the Agency. Five years ago, a large number of the applications were for combination feed drugs, which consist of two or more drugs each already approved for specific indications and species. Today, more applications are for new chemical entities, new species, or new indications. As a result, the complexity of the review has increased considerably. The Center for Veterinary Medicine (CVM) has developed procedures to facilitate the development of better applications and to expedite the review process. By meeting with sponsors prior to the submission of their investigational new animal drug exemption applications and encouraging the phased submission of the technical sections of the New Animal Drug Application, CVM has reengineered the review process to decrease approval times and increase the availability of safe and effective animal drugs.

Similarly, the Center for Food Safety and Applied Nutrition (CFSAN) has developed a plan to increase the timeliness and predictability of Agency action on food and color additive petitions. First, in order to address the backlog of pending petitions, the Agency has allocated additional resources to the program: 23 Agency scientists have been reassigned temporarily to the food additive program; the Agency is awarding contracts for expert review of certain portions of food additive data packages; and one and one-half million dollars was spent for enhanced computing facilities. With these efforts, we have already begun to see a decrease in the petition inventory.

Second, we instituted a threshold of regulation policy under which noncarcinogenic indirect food additives that do not present any substantial safety concerns are exempted from the petition process. The Agency processed 47 submissions under this new policy in 1995.

Finally, we again recognize that one way to decrease review times is to improve the quality of the food additive petitions filed with the Agency. We have been working with the industry to do that. FDA personnel have begun conducting workshops for petitioners to discuss what data and studies are needed for a complete food additive petition.

REGULATORY INITIATIVES AND MANAGEMENT REFORMS

Mr. Chairman, we have improved review times without sacrificing the high standards that give Americans confidence that their drugs are safe and work. Under PDUFA, the Agency was assured of adequate resources to do the job and the job was clearly defined through performance goals. At the same time, how to achieve those goals was left to Agency management, and accountability to Congress was assured through the five year authorization. That was real reform. Reauthorization of this program should be a cornerstone on which we build other improvements.

The progress we have made is the result of much more than the additional resources available under PDUFA and extends beyond the drug review process. Of equal significance has been the commitment of Agency management to streamline and improve the way we regulate. I think we all can agree that strong consistent management, the removal of unnecessary regulatory burdens, and the elimination of wasteful or inefficient practices all serve the critical public health goals to which we are committed. The Agency's contribution to the Administration's Reinventing Government Initiative well demonstrates the depth of our commitment to improving our regulatory processes.

FDA's initiatives, which will speed the availability of new products, streamline FDA procedures, and reduce regulatory burden without sacrificing safety, touch all of the regulated product areas. In biologics, for example, we have eliminated the establishment license application (ELA) and individual lot release requirements for well characterized therapeutic biotech drugs. According to the biotechnology industry, these changes will save their companies millions of dollars, reduce required paperwork by thousands of pages, and cut drug development time by months. In drugs and biologics, we have reduced and are working to continue to reduce the number of manufacturing changes that require pre-approval and have proposed regulations to reduce the number of environmental assessments required to be submitted by industry.

In the device area, not only have we exempted 572 categories of low-risk medical devices from premarket review, we also recently began implementing a pilot program for the third party review of certain low risk medical devices in classes I and II.

In the area of food safety, FDA is working with the industry to build safety controls into their food production system. The Hazard Analysis Critical Control Points System (HACCP), towards which the Agency is moving, is designed to replace the old system of detecting and correcting problems after they occur with a system of preventing them in the first place. It permits companies to devise and implement food safety plans uniquely suited to their needs.

Finally, in the area of animal drugs, we are seeking to replace the current system of requiring submission of medicated feed applications for individual feeds with a system to license facilities that comply with GMPs.

I do not have time today to adequately list or describe all of our initiatives, so I am attaching an appendix to my testimony that sets forth summaries of our efforts and accomplishments. Collectively, these efforts reflect the Administration's commitment to reinventing government while preserving the important health and safety protections that the American people rightly expect for these products.

THE HOUSE REFORM BILLS

Mr. Chairman, I would like to address the three reform bills that are the subject of today's hearing. We understand, from conversations with the majority, that there likely will be some changes to the three bills and we look forward to working with members of the Committee to improve these bills. However, my testimony today is based on the bills as they were introduced and as they stand before the Committee today.

Rather than address each bill and provision separately, I will address the concepts or provisions about which the Agency has the most serious concerns.

First, I will address a concept common to all three bills, and one that the Agency finds to be extremely troublesome the privatization of FDA's review functions.

Second, I will discuss a number of other provisions that we believe would result in unsafe products being sold to the American public.

Third, I will address the provisions that are problematic because they actually lower current efficacy standards.

Fourth, I will discuss the provisions in the food bill that undermine the Nutrition Labeling and Education Act of 1990, one of Congress's most popular public health measures in the last decade.

Finally, I will touch briefly on the bureaucratic requirements these bills impose, which will make it harder, not easier, for FDA to meet its goal of quickly getting quality products on to the market.

A. ALL THREE BILLS WOULD GIVE PRIVATE PARTIES THE RESPONSIBILITY FOR REVIEW FUNCTIONS THAT DIRECTLY AFFECT THE PUBLIC HEALTH

All three bills include privatization of FDA's review functions. Specifically they would permit third parties to make approval decisions for new drugs, devices, food and color additives, and health claims for foods. The bills would allow the review of these products and claims by reviewers who are paid by the industry to perform the reviews. In some instances, FDA would have a short, not very meaningful, opportunity to review the decision of the third party before that decision becomes final.

However, when a third party decided to initially classify a device into class II, which is the class into which 80-90% of all non- exempt devices fall, FDA would have no opportunity to review the decision; it would automatically become final, even if the third party incorrectly classified a high risk device as class II.

The concept of privatization of product application reviews is problematic for a number of reasons. First, FDA's scientific and clinical experts are charged with exercising independent and unbiased judgment. They comply with stringent financial disclosure and conflict-of-interest requirements designed to protect the decision- making process against bias. It is not clear how or whether this independence can be maintained with the private sector, particularly since the sponsor gets to choose and pay the private party and repeat business may depend on the sponsor's satisfaction with the private party's decision.

Second, FDA's reviewers have extensive knowledge about all of the similar products that are made by different companies around the country. When a reviewer looks at all of the drugs for arthritis and other inflammatory diseases, or all of the heart valves, what that reviewer learns from each review increases his/her understanding of that group of drugs or devices and their effect on the body. As a result, FDA reviewers see problems that reviewers with less information might not see.

This expertise was brought to bear five years ago when the Defense Department (DOD) was getting ready to do battle against Sadam Hussein. To protect our troops from possible chemical warfare, the DOD worked with a manufacturer to come up with a cream that could protect our soldiers, skin from chemical attack. DOD asked FDA to look at the data developed by the company. our reviewer, who had looked at many other similar products, thought that some of the methodology to test the product was oddly out of line as compared to other topical skin protectants, and so he asked our investigators to verify the firm's claims. As it turned out, the consultant hired to develop test methods for the firm submitted fraudulent information and the data did not support the use of the cream against chemical attack. The DOD did not use the cream, and the consultant was prosecuted. You are not likely to find in the private sector the kind of expertise and breadth of knowledge that allowed FDA to uncover the problems with the multishield cream.

The third problem with privatization is the lack of continuity. For example, FDA's reviewers are able to work with the same drug over time -- first reviewing the IND, then reviewing the NDA. By staying involved in a drug's development, reviewers can build on what they already know. Third party reviewers may have little knowledge of the specific development process for the product and/or of the development agreements made during the process.

Mr. Chairman, we believe that contracting out product review to third parties should be done only if there is evidence that it can be done safely. A pilot is essential to determine if that can be done. This is why FDA is conducting its pilot program for low risk medical devices. We are trying to determine whether third parties can accomplish the goal of getting safe and effective products to the American public.

The privatization provisions in the bills raise additional concerns. For example, some seem to be applying different arguably lower -- standards for the products being reviewed. For example, the drug provision indicates that third party reviews shall be conducted under the standards and requirements of the Act, but also states that the approval standard for third party reviews is a "reasonable assurance,, of safety and effectiveness, apparently an intention to lower the standard from the one currently in the FDC Act. For food additives being reviewed by a third party, the standard has been changed (arguably lowered) from "safe" meaning a "reasonable certainty of no harm," to a "reasonable assurance" that the food additive may be safely used.

Moreover, under the privatization provisions, FDA is faced with a nearly impossible burden if it disagrees with a third party that a product should be approved. For example, the drug and device provisions state that when a third party recommends approval, the application is deemed approved unless FDA finds that there is a reasonable probability that the drug or device is not safe or effective. Under the food and color additive provisions, FDA can only deny approval of a food or color additive recommended by a third party if the Agency can demonstrate that there is a "reasonable probability" that the additive is "not safe."

Historically, it has been the responsibility of the drug, device, and additive manufacturers -- those who would profit from the products' uses -- to demonstrate to the Agency that the products they propose to market are in fact safe and, where appropriate, effective. In the case of drugs and devices, FDA's responsibility has been to review the evidence and either approve the drug or device if the Agency concurs that the product has been shown to be safe and effective, or refuse to approve the product if, in the Agency's opinion, safety or effectiveness has not been demonstrated. In the case of food and color additives, FDA's responsibility has been to review the evidence and either approve the additive if the Agency concurs that the product has been shown to be safe or refuse to approve the additive if, in the Agency's opinion, safety has not been demonstrated.

To require the Agency to demonstrate that a product is not safe or, where appropriate, not effective, is to require FDA to demonstrate something for which there may be no data because sufficient studies have not been conducted. Because more often than not there is a lack of evidence of safety or effectiveness, (rather than definitive evidence showing that a product is unsafe or ineffective), this will be an extremely difficult burden for the Agency to meet. The effect of this shift in the burden of proof is to virtually remove FDA from the evaluation of any product that a manufacturer chooses to have reviewed by a private party. Since many of these products literally involve life and death and since there are at best very serious doubts about whether these private organizations can do an adequate job, this provision will seriously jeopardize the public health.

Finally, because there is a risk that only third parties that err on the side of approving applications and petitions will succeed in this market, there is a risk that the incentive will be for third parties to approve applications and petitions -- not to critically review them. If what this bill is trying to accomplish is getting safe and effective products on to the market, broad privatization will not accomplish that end.

B. ALL THREE BILLS WOULD INCREASE THE POSSIBILITY OF UNSAFE/INEFFECTIVE PRODUCTS REACHING THE MARKET

Many of the provisions in the House bills weaken the standards under which FDA has operated. Obviously, we have serious concerns about the provisions that would change the basic efficacy standards, and I plan to spend some time discussing those provisions. But first, I would like to discuss the other measures in these bills that would undermine the Agency's ability to protect the American public from unsafe products. These provisions undercut safety either because they exempt certain activities or products from FDA review or oversight or because they limit the type of information that FDA can ask for or review. Although there is some overlap, I will address the bills separately.

1. H.R. 3199 -- Drugs and Biologics

Removal of Protections for Subjects of Drug Testing. As you know, FDA plays an important role in overseeing the safety of drug experimentation in humans. H.R. 3199, however, eliminates much of our ability to protect human subjects.

For example, the bill shortens the amount of time FDA has to review an investigational new drug application (IND) before clinical trials can begin. It also cuts back on the amount of information FDA can receive when it is evaluating whether to allow an investigational new drug to be administered to human subjects for the first time. The shortened timeframe and the limited information together make it difficult for FDA to stop a potentially dangerous trial. In addition, Phase I and Phase II studies (which are the earliest stages of drug testing in humans) could be conducted without any FDA review of the investigation.

Finally, the bill limits FDA's ability to stop a clinical trial. In fact, the bill would permit trials to continue even though the investigators are not qualified or the sponsor has submitted inaccurate information to the investigator. The effect of the bill is to eliminate important protections that now exist for these investigational studies.

Summary Data. After a sponsor of a new drug has completed its studies, it compiles a new drug application (NDA), which includes a report summarizing the results of the study. In addition to that report, the sponsor will have primary data, which consist of data tabulations (listing of the data points) and case reports (i.e., the forms on which the investigators have listed their observations).

Under current law, when FDA reviews an NDA, it receives the summary report and the data tabulations for all study subjects. FDA initially receives the actual case reports only for subjects who have died or dropped out of the study, unless the Agency asks for additional case report forms.

Section 4 of H.R. 3199 would permit sponsors to submit summary data -- i.e., the study report -- and only those data tabulations and case reports relating to deaths or drop outs due to adverse reactions. If FDA wants to see the additional data tabulations or case report forms, the office director must make a request in writing and must specify the reasons for the request for a particular application.

This provision will seriously compromise FDA's review process. The submission of primary data is important for a number of reasons. First, it is essential to discovering and discouraging the submission of fraudulent data. Second, summary reports sometimes contain errors that can be discovered only by, reviewing primary data. Third, summary data have been screened, processed, and interpreted by the sponsors, who often just do not see the shortcomings of their drugs. Someone who does not have an economic interest in the outcome of the study should be taking a hard look at the underlying data to verify the interpretation of summary data.

Finally, problems associated with a drug may be visible through an examination of the primary data that are not visible from a summary. For example, the sponsor may attribute deaths and discontinuations of study subjects to causes other than the drug, but review of the raw data may show that those adverse reactions are associated with the drug itself. Under the bill, if the sponsor does not think that the drug caused a death or drop out, the sponsor simply would not submit the data tabulations or case report forms.

Even hints of information found in patient report forms can reveal or lead to critical information. For example, when a reviewer in the Center for Drug Evaluation and Research reviewed the patient data for dilevalol, a drug used to treat high blood pressure, she (not the company) noticed that three of the subjects suffered from liver injury while on the experimental drug. Based on her knowledge of this class of drugs, and the fact that other drugs of this type rarely caused this type of side effect, she decided to look further. The company then examined recent data available from Japan and Portugal, where the drug had already been approved. The company found severe liver injury as well as deaths. Based on this careful observation by an FDA reviewer, the drug was not approved in the United States, and it was removed from the market worldwide.

Mr. Chairman, I frankly cannot think of what the objections to submission of primary data might be. The sponsors already have the data. Given the fact that this data may now be submitted on computer disks rather than in hard copy, the "truckloads of paper" argument does not hold. If FDA's drug review times did not meet or exceed the Prescription Drug Use Fee Act goals, which the Agency and the drug industry jointly established, the argument against primary raw data might be that it delays FDA's review. However, FDA is meeting the PDUFA goals. The only possible reason that sponsors can have for not wanting to submit primary data is that they do not want FDA to scrutinize their data. The testing of drugs that we make available to the American public should undergo this type of scrutiny.

Limiting Consideration of Certain Effectiveness Data. H.R. 3199 also limits what FDA can look at in making a determination of effectiveness. For example, Section 5 provides that the determination of effectiveness cannot include the evaluation of relative effectiveness unless the effectiveness of the drug is explicitly compared in the labeling to that of another drug.

Although FDA generally does not base its approval decisions on whether one product works as well as another, there are circumstances where such considerations are important. For illnesses with significant public health implications, such as those that are serious or life-threatening or that involve public health transmittal risks (e.g., sexually transmitted diseases) it generally would be unconscionable to allow marketing of drugs that are less effective than existing therapies.

For example, FDA require a 95 percent efficacy rate for gonococcal eradication for products intended to treat-gonorrhea. This is an efficacy rate products can meet today. Similarly, FDA refused to approve an HIV test kit that demonstrated a lower accuracy rate than one already on the market. It generally would not be in the public interest to have products available that are less effective in treating a public health epidemic or serious or life threatening disease.

Section 5 also provides that the determination of effectiveness cannot include the evaluation of any use not explicitly included in the labeling. This appears to prohibit FDA from requiring, in the labeling, warnings about indications for which a product is dangerous or known not to be effective.

Manufacturing Changes. As you know, FDA's ability to keep our nation's drug supply safe does not end with approval of a new drug. FDA is vigilant to make sure that the subsequent manufacturing of those approved drugs will not adversely affect their safety or effectiveness. Manufacturing changes that are made with the best intentions sometimes result in making a drug more dangerous or less effective. Because of the devastating impact manufacturing changes can have, we have made certain that companies tell us about the changes they are making and we ask that some changes come in for FDA review and approval before they are made. H.R. 3199 would permit drug and device manufacturers to make many risky manufacturing changes without first coming to FDA.

Specifically, section 13 of the drug bill would permit manufacturers of drugs, most biologics, blood, and blood components to make many changes in manufacturing without prior approval as long as there is not a resulting change in formulation or release specifications, the changes are validated, and FDA is notified in an annual report. Other manufacturing changes would require validation, end-product testing (to determine equivalence), and notification to FDA at the time of the change (although not necessarily before a product manufactured under the changed conditions is in distribution).

The concern we have with the approach for drugs, most biologics, blood, and blood components is that even changes that do not affect formulations or release specifications can still change the safety and effectiveness of a product. Similarly, equivalence testing may not always detect problems that can affect safety and efficacy. Moreover, such products may be in distribution before FDA has had time to react to the change.

Perhaps, our concerns are best illustrated by the infamous "Cutter" incident with Salk polio vaccine. Salk polio vaccine is a "killed" vaccine -- i.e., it is free from any live virus. In the 1950's Cutter made a seemingly minor change in the manufacturing process -- it simply changed the type of filtration process used in the manufacture of the vaccine. That change let live virus into the vaccine. None of Cutter's end-product testing was able to detect the deadly change. Eleven people (including children) contracted polio and died, and about two hundred more got polio before the vaccine could be pulled from the market.

Another example relates to a change in the method of growing the cells from which a biotechnology heart drug was made. The new method caused a change in the fine structure of the drug molecule. This change was not detected by final product testing, but it resulted in a less effective dose. Similarly, without submitting the change to FDA for review, a firm changed the particle size of bulk carbamazepine, a seizure medication. Although the product met existing specifications, the change in particle size caused the drug to absorb moisture, and thus, to dissolve more slowly than did the product made from the approved source. The slower dissolution may have resulted in patients absorbing the drug more slowly and thus, getting smaller doses. FDA received reports of seizures in persons taking the drug made from the changed bulk drug.

Similarly, under these new bills, a manufacturer could change the method of viral inactivation in a blood product used to treat hemophilia without first coming to FDA. A change to a new method that was incapable of such inactivation could result in a product that transmits hepatitis or HIV. Because FDA would not learn about the manufacturing changes until the public actually is receiving the product, it could act only after the public has been exposed to a product capable of transmitting hepatitis or HIV.

In sum, we know that even seemingly minor-manufacturing changes can make a significant difference and that the problems wrought by those changes will not always be picked up by end product testing. The bills, provisions that change how manufacturing changes are regulated is going to permit manufacturers to make changes that will harm the public.

GMP Inspections. After a drug has been approved, FDA conducts good manufacturing practice (GMP) inspections to ensure that the company can make the drug the same every time, so that batch after batch of the drug has the same therapeutic profile. Otherwise, the public may receive drugs that do not have the effect they are supposed to have or actually could cause harm. H.R. 3199 permits those inspections to be conducted by third parties. Although FDA has been exploring the use of third party inspections, it objects to section 10 of H.R. 3199 because it essentially eliminates FDA's involvement when a third party performs the inspection.

Pursuant to section 10, a third party performing a GMP inspection generally would not submit a copy of its observations to FDA, even when GMP problems exist. Instead, the third party would oversee corrections and then when the company achieved compliance would just certify to FDA that the company is in compliance. The Agency would never even know that problems had existed. Moreover, FDA would not be able to conduct a GMP inspection for two years after the third party certification, "unless justified by good cause." This is problematic because conditions can change: a change in ownership or internal management may change manufacturing conditions, or personnel may simply drift into bad habits. The latter is more likely to occur if companies know that FDA could not reinspect for two years. Awareness that FDA can inspect at any time reinforces vigilance.

Pharmacy Compounding. One of the most glaring examples of a provision that undermines FDA's authority to protect the public from unsafe and ineffective products is found in Section 18 of the drug bill. Section 18 exempts drugs that are compounded by pharmacists on the order of a licensed physician from the drug and device provisions of the Act, and it exempts bulk drug products or other drugs intended for use by pharmacists for compounding from all of the Act's provisions, except those that relate directly to identity, purity, or quality.

The exemption has no constraints on the volume of compounding. It is likely to encourage large-scale manufacturing under the guise of pharmacy compounding. It would allow bulk drug suppliers or drug manufacturers to circumvent the approval requirements of the Act by shipping bulk drug substances to pharmacies for reconstituting or other processing. A shadow industry of unapproved generic drugs is likely to develop.

Moreover, the exemption would allow potentially dangerous compounding. For example, sterile drugs could be compounded (even on a large scale) without regard to current good manufacturing practices (CGMPs) for sterile products. Improperly compounded sterile products could result in serious adverse effects, including death.

2. H.R. 3199 -- Blood, Blood Components, and Human Tissue

Blood and Blood Components. H.R. 3199 deletes blood and blood components from the definitions of "biologic" and "drug." It eliminates the Public Health Service Act provisions relating to blood and blood components and creates a new regulatory scheme for them. We are concerned, however, that the new regulatory structure could endanger the nation's blood supply.

Although the language of the bill is unclear, there appears to be more limited authority to seize and recall certain dangerous blood and blood components. Moreover, the new standard for license suspension would be more difficult to meet. The bill further provides that even revocation or suspension of a license would not prevent the use of products that have left the control of the licensee unless the public health actually had been harmed. Even an extremely high risk of death might not permit FDA to prevent the use of such products through revocation or suspension. Finally, the provision permitting third party GMP inspections would encourage the blood industry to be inspected by the organizations, accredited by FDA, to which they themselves belong. Will the American public have confidence in the safety of the blood supply when the manufacturers can self inspect?

Human Tissue. H.R. 3199 also sets forth a separate regulatory scheme for human tissue. It appears to define human tissue as including some or all cellular and gene therapies and it provides that human tissues may not be regulated as drugs, devices, or biologics. However, cellular and gene therapies should be subject to premarket controls because they require safety and efficacy determinations and more stringent manufacturing controls.

Pursuant to H.R. 3199, FDA may regulate human tissue only if it demonstrates, in writing published in the Federal Register and after a hearing before the Commissioner, that voluntary regulation under generally accepted standards is inadequate to protect the public health with respect to any particular type of human tissue or human tissue generally.

FDA has already made the determination that regulation of human tissue is necessary. The voluntary standards in place prior to 1993 when FDA issued regulations requiring HIV and hepatitis testing for tissues did not adequately protect the public health. Imported tissue from hepatitis infected donors was being brokered to U.S. companies for use in patients. Moreover, HIV has been transmitted to patients through HIV-infected domestic tissue. Requiring FDA to revisit the question of the adequacy of voluntary standards is simply an unjustified waste of the Agency's scarce resources.

3. H.R. 3201 -- Devices

H.R. 3201 is similar to the drug bill in that it severely undermines FDA's authority to ensure that safe and effective devices are being marketed. Like the drug bill, it does this by limiting FDA's authority to review certain devices, limiting FDA's access to information about devices, and limiting FDA's authority to ensure that devices, once approved, remain safe and effective.

Exemptions from Filing a 510(k). H.R. 3201 first undermines device safety by allowing many significantly modified devices to go on the market without any kind of review. It does this by eliminating the requirement for a manufacturer to file a premarket notification (510(k)) under certain conditions.

Section 9, for example, exempts all Class I devices from the requirement of submitting a premarket (section 510(k)) notification, which is the clearance process for most devices. FDA has already exempted 572 categories of class I devices from premarket notification. FDA had public health reasons for not exempting all class I devices. For example, FDA needs to review tests on the technical performance of hearing aids to ensure that they function as intended. Similarly, FDA should have an opportunity to review rebreathing devices, which are used as components within certain anesthesia machines, volume ventilators, and resuscitation devices and which have been directly related to death, serious injury, and serious illness resulting from complications caused by their design, and/or performance. Finally, it is extremely important to the public health that blood warmers not go on the market without any review. After all, the administration of blood that is at the wrong temperature can be fatal. Yet, H.R. 3201 would eliminate these and other important reviews.

Section 9 also eliminates FDA's role in assuring that modifications made to a device do not adversely affect its safety and effectiveness. It does so by eliminating the requirement that a manufacturer file a 510(k) for any modification other than a major change in intended use if the modification can be shown not to adversely affect the safety and effectiveness of the device. Thus, changes to materials, power sources, control mechanisms, and manufacturing processes would be permitted and FDA would not see data. This provision creates an enormous loophole to the general requirement that devices be cleared by FDA prior to marketing.

Change and innovation are important -- over time, many of the devices we now use have improved greatly -- but companies cannot always tell when their design changes are going to have deadly- consequences. The Shiley heart valves illustrate this point. The company had made an artificial valve for diseased aortic and mitral valves of the heart. Picture a metal circle -- like the ring on a key chain -- with a disc inside the ring that would open up to 60 degrees, thereby letting blood flow through the valve. Unfortunately, there were blood clotting and blood flow problems with the 60 degree heart valve.

The company thought that they could fix the blood flow and blood clotting problems by having the valve open up to 70 degrees. At the time it was reported that the 60 degree valve suffered from metal fatigue and could break, causing of catastrophic events, which often resulted in sudden death. Nonetheless, Shiley manufactured the new 70 degree valves and sold them in Europe (though not in the US, because Shiley did not have the engineering and clinical data we asked for).

Unfortunately, the company's theory that the 70 degree valve would be better proved wrong. The 70 degree valve had a failure rate that was 6 times higher than that of the 60 degree valve. Had the 70 degree valve been sold in the US market, many more people would have died. This experience shows clearly the importance of independent review for "improvements" to devices.

It is as important for class II devices that are life sustaining or otherwise critical, as it was for the Shiley heart valve.

GMP/510(k) Linkage. H.R. 3201 not only removes numerous devices from FDA review, it also fails to ensure the safety of the products for which 510(k)s actually do get filed. For example, section 9 prohibits FDA from withholding a 510(k) decision because of a failure to comply with GMPs. This removes the basic assurance that new devices will be well made and work as intended. Under section 9, even a company with known manufacturing problems directly relevant to its new device, might have to be given clearance to market its device. (A similar provision appears in the drug bill, which prevents FDA from ascertaining GMP compliance prior to approving a new drug. This prevents FDA from knowing that a drug manufacturer can reliably and consistently produce high quality drugs.)

Exemptions from PMA Requirements. H.R. 3201 also would weaken the safety standards for the most complicated category of devices-- devices that currently must obtain approval of a premarket application (PMA) prior to being marketed. For example, section 9 virtually eliminates PMA review and approval of devices for specific uses. It would allow a device originally marketed for one clinical indication to later be marketed for other indications, even high risk ones, with only the submission of an abbreviated application (510(k)) that would not document the product's effectiveness for the new indications.

Suppose, for example, that a manufacturer of an ultrasound device, approved for the general diagnostic use of evaluating soft tissue, wants to market its device for the higher risk, specific use of diagnosing cancerous breast lumps. H.R. 3201 could prevent FDA from requiring the manufacturer to produce data showing that the ultrasound device effectively differentiates between a cancerous and benign lesion. Without data clearly demonstrating the effectiveness of ultrasound for this use, many women could have cancerous lumps misdiagnosed as benign, and thus forgo early lifesaving treatment.

Limiting Consideration of Certain Effectiveness Data. For the more complicated devices that actually would remain subject to FDA review under H.R. 3201, the bill includes the same effectiveness determination limitation as the drug bill. The concerns expressed regarding the drug bill's limitations on evaluating comparative efficacy and uses not included in the labeling apply equally to devices. In addition, Section 8 of H.R. 3201 prohibits FDA from evaluating clinical outcomes for devices, unless the device is labeled as having a therapeutic effect. It appears that the Agency would be prohibited from looking at whether devices really help patients -- for example, whether patients with artificial hearts did well or died prematurely -- unless the outcome is part of an explicit labeling claim. This emasculates the present system that requires certain devices be shown to clinically perform as intended and produce useful outcomes for patients and ultimately will impair clinical decision- making by health care providers.

Manufacturing Changes. As with drugs, FDA's responsibility for safe products does not end when a device is approved by FDA. Manufacturing changes can greatly impact the safety of medical devices. Section 9 would permit manufacturers to make significant manufacturing changes to devices subject to the 510(k) provisions without first coming to FDA. Section 11 of-the device bill would eliminate the need for PMA supplements for changes that "do not actually affect device safety or effectiveness." In effect, this relieves manufacturers from seeking FDA approval for certain changes in manufacturing processes, sites of manufacture, sterilization procedures, etc. The risk is that a manufacturer will determine that a change does not affect safety or effectiveness, when in fact, it does.

Decreased Ability to Monitor Marketed Devices. The Safe Medical Devices Act of 1990 added a number of provisions to the FDC Act designed to enable FDA to monitor devices after they were sold. Specifically, it added a requirement for device tracking to enable FDA or the manufacturer to quickly and efficiently alert patients if a device subsequently proves faulty and presents serious health risks. It also added a post marketing surveillance provision to supplement the approval process because Congress felt that premarket approval could not detect all possible problems that might occur after a device is put into general use. Finally, it added a requirement for user reporting because there were device injuries occurring in hospitals that were never reported to FDA. H.R. 3201, however, eliminates user reporting and significantly curtails FDA's ability to monitor marketed devices under the other two provisions.

For example, section 14 severely restricts FDA's ability to require device tracking. Specifically it changes the criteria for a tracking requirement, and it eliminates FDA's discretion to apply the provision to "any other device." FDA used this "discretionary" authority to require tracking of implantable infusion pumps, silicone breast implants, and other silicone related devices. The legislation will eliminate our ability to track devices the failure of which could have significant public health implications.

Section 15 limits the type of device that can be the subject of postmarket surveillance. The most problematic aspect of this provision is that it only permits FDA to require postmarket surveillance for devices "first introduced or delivered into interstate commerce after January 1, 1991." This would preclude FDA from requiring postmarketing surveillance of devices such as the Shiley heart valves, Landmark catheters, and foam-coated breast implants -- all products shown to subject their users to health risks.

4. H.R. 3200 -- Foods and Animal Drugs

Delaney Clauses. For more than 35 years, consumers have known that manufacturers are not permitted to add to food for human consumption or food and color additives that have been shown to induce cancer. Manufacturers have had the burden of proving that additives or animal drugs used in livestock are safe. The Delaney clauses were premised on a simple public-health principle: Because the public is exposed to many sources of cancer risk (many unavoidable and involuntary), this risk should not be increased by intentionally adding carcinogens to food.

H.R. 3200 would change the long-standing safety standard to which food and color additives that have been held for almost 40 years. H.R. 3200's safety standard of negligible, de minimis, or insignificant risk would result in the intentional addition to the diets of American consumers carcinogenic food and color additives that heretofore have not been permitted. Furthermore, this change, whatever negative public health impact it might have on its own, must be considered in the context of another significant safety-related change proposed in H.R. 3200. Under the bill, when a food or color additive is reviewed by a third party, the burden of proof shifts from the manufacturer, in whom the responsibility for safety demonstration has resided since 1958, to the FDA. In such circumstances, to prevent the marketing of a carcinogen intentionally added to food, the Agency would be required to show that the additive causes a significant cancer risk or that there otherwise is a reasonable probability that the additive is unsafe. It is clear that these provisions ask the American public to take on additional risk from foods without any concomitant benefit.

Although FDA is willing to consider changes to the Delaney clause, our overall focus must remain on assuring that the public is adequately protected. As you know, the Administration has recommended, and I have supported, legislation that would remove pesticide residues from consideration under the Delaney clause. That change was recommended only because it was coupled with other changes to the overall regulatory scheme for pesticides, including the creation of a higher, risk-only, health-based standard for residues along with provisions that would assure a more careful consideration of the effects of pesticide residues on infants and children. In proposing to eliminate the Delaney clause, H.R. 3200 not only fails to provide these extra measures of safety and public assurance, but, as noted, actually weakens the safety standard for food and color additives and makes it more difficult for FDA to deny approval to an additive that, according to the Agency's assessment, has not been shown to be safe. Thus, under the bill, the protection afforded to consumers from unsafe additives falls short of that proposed by the Administration for pesticide residues and that provided by current law.

Preemption. H.R. 3200 prohibits states from establishing requirements (including prohibitions) for a food, drug, or cosmetic that relate to FDA's adulteration, misbranding, or new drug authority if it is not identical to a requirement authorized or required by the FDC Act. It also prohibits states from imposing any notification requirement for a food, drug, or cosmetic that provides for disclosure of the constituents, source, or method of production or processing such product or for a warning concerning the safety of the product or any component or package thereof unless such requirement is identical to a requirement prescribed under the FDC Act.

The effect of this sweeping preemption provision is to take away states, authority to impose requirements to ensure the safety of the food, drug, and cosmetic supply. It basically would place this responsibility totally in the hands of the federal government. Thus, if the state of California wanted to require the maker of Rio hair dye, which caused users' hair to fall out, to warn consumers, or the state of Texas wanted to require a warning on ephedra, they could not unless the Federal government had identical warning requirements. Moreover, at the same time this legislation is transferring all responsibility to FDA, it is weakening FDA's ability to fulfill that responsibility. For example, together the Delaney clause provision and the third party review provision undermine FDA's efforts to ensure the safety of the American food supply. This provision also could have the effect of preempting state personal injury lawsuits involving foods, drugs, and cosmetics. (Compare Medtronic, Inc. v. Lohr v. Medtronic, Inc., 56 F.3d 1335 (11th Cir. 1995), cert. granted, 64 U.S.L.W. 3497, 3500 (U.S. January 19, 1996) (Nos. 95754, 95-886) and cases cited therein.) Finally, a problem with the preemption provision, in the food area in particular, is that there are certain areas (such as grocery stores and restaurants) that, partially due to FDA resource constraints, have traditionally been regulated by states and localities. State and local regulations are critical to protecting the public health. Frequent inspections by state and local authorities ensure that restaurant and nursing home kitchens are sanitary and that food is handled safely. The states are responsible for ensuring that the fast food hamburgers we so often enjoy have been cooked at the appropriate temperature. Yet, under H.R. 3200, they may no longer have that authority. Because of the prohibition in the bill, some areas that significantly impact the public health could be virtually unregulated.

Increased Risks From minor Animal Species Receiving Ineffective Drugs. This legislation would adversely affect the public health, not only because of the bills, impact on the regulation of drugs, devices, and foods, but also because of H.R. 3200's impact on animal drugs. For example, H.R. 3200 would eliminate the requirement to show effectiveness of animal drugs intended for uses in minor species or for minor uses. Elimination of a requirement to show effectiveness takes us back to before 1962. There is no point to having drugs that cannot be shown to be effective. Moreover, the availability of ineffective animal drugs has implications for both animal and human health. The use of ineffective animal drugs in minor species or for minor uses can lead to the unnecessary suffering and death of animals and an increase in the potential for the transmission of diseases from animals to humans. It may also increase the potential for development of bacterial resistance to antibiotics.

Combination Animal Drugs. H.R. 3200 also changes the standards for approving combination animal drugs. It would permit a sponsor to market a combination animal drug containing two or more drugs that are targeted to the same indication without having to show that each drug provides some additional benefit. Moreover, it would not require the sponsor of a combination animal drug to establish the safety of the combination to the animal targeted to receive the combination.

Prior to enactment of the current animal drug laws, manufacturers marketed drugs that contained significant amounts of multiple powerful antibiotics to treat infections in milk cows without showing that each antibiotic added to the effectiveness of the drug. The marketing of similar products today, as the bill would permit, increases the risk of antibiotic resistance, which could result in less effective human antibiotics, and could lure farmers into paying for drugs that are unnecessary and dangerous.

Dangerous Animal Drug Residues. One final provision that would affect the public health by eliminating certain animal drug protections is in the drug and device bills. Both add a new section 909 to the FDC Act that prohibits FDA from limiting or interfering with a health care practitioner's authority to prescribe or administer legally marketed drugs or devices. This provision appears to interfere with FDA's current authority to restrict a veterinarian's use of certain animal drugs in food producing animals on the grounds that such use could adulterate the food supply.

For example, chloramphenicol is an antibiotic currently approved for use in non-food producing animals to treat respiratory and urinary tract infections. It has not been approved for use in food producing animals because it is associated with severe side effects (e.g., aplastic anemia and death). Under new section 909, however, a veterinarian could prescribe chloramphenicol in food producing animals. Persons who consume the meat from those animals would be exposed to the chloramphenicol and its adverse side effects.

C. ALL THREE BILLS WOULD UNDERCUT CURRENT EFFICACY STANDARDS AND INCREASE THE POSSIBILITY OF INEFFECTIVE AND UNSAFE PRODUCTS REACHING THE MARKET

The privatization provisions that appear in all three bills would lower the approval standards for products subject to third party review. A number of additional provisions in the bills lower the efficacy standards for products regardless of whether they are subject to third party review.

Common Medical Practice. Section 5 of the drug bill would permit the approval of a new use of a previously approved drug on the basis of a demonstration that the new use is common among clinicians experienced in the field and represents reasonable clinical practice based upon reliable clinical experience and confirmatory information. Basically, the bill assumes that if many doctors prescribe a drug for a particular unapproved use, then that unapproved use must work.

A drug may come into widespread use for a condition because common experience suggests that the drug seems to work. Unfortunately, experience shows that testing in clinical trials often demonstrates that the drug is not effective and in fact, may be harmful. Medical history is replete with examples of products and procedures that were based on medical anecdote, not evidence, and were thought for years by most clinicians to be effective, but later turned out to be useless and sometimes dangerous. It may be hard to believe, but clinical experience often provides the wrong answer.

For example, in the late 1980's, physicians began to prescribe the drugs encainide and flecainide, for heart attack victims who were experiencing ventricular premature complexes (VPCs), a type of asymptomatic or minimally symptomatic abnormal rhythm of the heart. This off-label use for these drugs, which were approved for other types of abnormal heart rhythms, was intended to prevent the increased mortality of heart attack victims. The use became so widespread that the National Institutes of Health decided to study it. At the time, everyone was so sure that the drugs worked for the new use that it was thought to be unethical to conduct a controlled clinical trial because that would mean that certain patients would not receive encainide or flecainide. To nearly everyone's surprise, that study demonstrated not only that the drugs were ineffective in reducing the risk of death, but also that the drugs were actually harmful in the patients for whom they were being prescribed off-label. This bill would permit unproven uses such as this to get onto the drug label.

This legislation would move us away from the accepted scientific standard of evidence and back towards evidence based solely on anecdotal experience. Anecdotal reports and poorly controlled observations have proven not to suffice because they can be wrong. We know this because, in 1962, Congress told us to take a hard look at the drugs that entered the market prior to the addition of the efficacy standard to figure out which ones actually worked and which ones did not work, but were still being prescribed by doctors. With the assistance of the National Academy of Sciences, FDA looked at the objective scientific information available on 3,443 drugs. The result of this review found that only one-third of the drugs really worked. The other two-thirds either were not effective, despite the beliefs of physicians, or could not be shown to be effective.

H.R. 3199 would perpetuate the kind of inappropriate and unproven medical practices that existed prior to 1962. It also could have the unintended consequences of creating a disincentive for companies to collect the scientific data necessary to demonstrate whether a product is effective for a particular use.

We must ask ourselves, is it a good idea to go back to the pre- 1962 era and permit drugs to be labeled for uses if some doctors think they work for those uses, or are we better off with the current system where there are incentives-to do the studies to find out scientifically whether the drugs work for new uses? FDA is on the side of getting the scientific information. The public has a right to drugs that actually work as intended.

Approval of Drugs for Serious and Life Threatening Diseases. Section 5 of the drug bill also provides that an application for a new drug for a serious or life-threatening condition shall be considered to have substantial evidence if it could fairly and responsibly be concluded by experts that there is a reasonable likelihood that the drug will be effective in a significant number of patients and that the risk from the drug is no greater than the risk from the condition.

This provision undermines the efficacy standard and shifts the risk analysis away from risk/benefit. FDA strongly supports getting drugs for serious and life threatening diseases to patients as quickly as possible. The Agency has developed a process to accelerate approval of drugs for serious and life-threatening diseases, by approving such drugs for marketing before the data from traditional clinical trials are complete if the drug shows an effect on a "surrogate marker." Thus far, this has been proving to be a successful process. FDA does not and cannot, however, support a provision that would allow approval of a new product that has not been adequately tested.

Product Approval Standards. All three bills contain provisions that explicitly lower the efficacy standard. Section 5 of the so drug bill provides that "one or more" clinical investigations may be needed to show effectiveness. The Agency's current presumption is that, for pharmaceuticals, ordinarily, replicated evidence is expected, but FDA, in particular circumstances (i.e., where a single study offers internal evidence of consistency of the product's effectiveness), may rely on a single adequate and well-controlled trial. The provision, on its face, does not appear to change that presumption. However, we understand that its purpose is to create a presumption that one adequate and well controlled trial will be sufficient for purposes of approving a new drug.

The Agency strongly opposes a change that would eliminate the replication requirement for drug approval. It is a basic tenet of science that data can be considered reliable only if, when an experiment is run for the second time, it produces the same findings and observations that it produced the first time. The American public deserves to receive drugs that have been tested according to well accepted scientific principles.

Section 8 of the device bill changes the current effectiveness standard to include a maximum of one clinical investigation and establishes that no clinical investigation will be required unless the Secretary first consults with a classification panel. Because devices, unlike drugs, have the mechanism of action designed into them, the need for demonstrated replication has not been a part of the way we have reviewed clinical data on devices over the last 20 years. Nevertheless, in those cases where the devices mechanism of action is poorly understood, replication of clinical data is correspondingly important.

Finally, section 201(a) of the animal drug bill redefines the "substantial evidence" criteria that currently must be met to establish efficacy. The provision changes the substantial evidence requirement from adequate and well-controlled studies to scientifically sound studies. The requirement that the studies fairly and reasonably show effectiveness has been changed to a requirement of a reasonable assurance of efficacy. The provision also eliminates the requirement for field investigations unless FDA justifies in writing the need for such investigations to show the effectiveness of animal drugs. This lower standard for effectiveness for new animal drugs is not acceptable.

Off Label Use. The drug and device bills make it legal for a sponsor to disseminate medical texts, journal articles, and government information, and to make presentations at medical and scientific meetings and have displays at trade shows about off label uses.

These provisions assume that the various activities it permits are not promotion and do not encourage off-label uses or uses of investigational products. Their effect, however, is to make it possible for sponsors to promote off-label uses and investigational products.

FDA knows that there are important off-label uses of approved products. In this context, it is important that physicians have access to accurate information about drugs and devices. But we also know that allowing the promotion of these kinds of uses by companies with a financial interest in the products, use can have very serious public health consequences.

The fundamental problem with permitting the promotion of off- label uses is that not all off-label uses help patients; some do not work and some are harmful. The only way to know which ones are safe and effective is to collect and analyze the data supporting a finding of safety and efficacy. Permitting the promotion of off-label uses based on studies reported in journal articles or other texts or unpublished materials, which clearly are an inadequate basis for approval, would undermine the efficacy standard.

One of the most serious consequences of allowing companies to promote off-label uses is that companies would have no incentive to conduct or fund the necessary scientific research and to present data to FDA to verify the safety and efficacy of those off-label uses. In fact, because the Agency might determine that the new use is not supported by the evidence, there would be an incentive to avoid FDA review. For devices, the incentive problem is exacerbated because the investigational device provisions (Section 4) would permit more widespread use of investigational products, which the sponsors can now promote and for which they can be compensated.

If off-label uses could be promoted, manufacturers would have an incentive to do the minimal number of studies necessary to obtain approval for the first, narrowest/easiest indication and then heavily promote the product for other broader -- and possibly more speculative -- uses. For example, interferon-alpha 2b was approved for use in hairy cell leukemia, of which there are approximately 300-400 cases per year. It subsequently was approved to treat chronic hepatitis B and C, of which there are tens of thousands of cases per year. If the manufacturer of interferon alpha 2b could have promoted the product for chronic hepatitis B and C -- the much broader use -- based on preliminary data, we might never have learned whether interferon alpha 2b actually works to treat hepatitis B and C.

Because the incentive to conduct research on uses of drugs and devices will decrease, the end result will be that the dissemination of off-label information will actually reduce the amount of good scientific information that health care providers receive about drugs and devices.

Widespread promotion of unapproved uses also raises significant safety concerns. Even under current law, which prohibits the promotion of off-label uses, we know of a number of instances where physicians have used drugs for off-label uses that have resulted in disastrous consequences. The off-label use of encainide and flecainide, discussed earlier, was supported by published peer- reviewed journal articles. If the unapproved uses had been heavily promoted by drug companies, it is estimated that thousands more unnecessary deaths would have occurred.

The current law governing promotion requires balance. Changing the law to allow the distribution of journal articles and other similar materials that promote off-label uses will allow drug sales representatives to provide materials that describe favorable study results of their product for a particular use, without providing copies of materials that demonstrate the other view. What makes this situation even more troubling is that when FDA has evidence that a particular use is unsafe or ineffective, federal confidentiality laws frequently inhibit our ability to disseminate that information. Thus, there are off-label uses about which positive studies appear in the literature and negative data are contained in our files. Depending on its source, FDA may be unable to use the negative information to ensure that the medical community and the public have all of the available facts on which to base treatment decisions. At the very least, it is irresponsible to open up promotion of off label uses without giving the Agency explicit authority to rebut that promotion.

FDA recognizes that there are important off-label uses. We believe, however, that the best way to address concerns that information about those uses is not reaching medical practitioners is to get those uses for which the drug actually works in the labeling. As you know, a subsequent indication for a new drugs is added via a supplemental new drug application, which often needs to present only efficacy information to support that new use.

FDA has been developing ideas for encouraging and expediting efficacy supplements for unapproved uses. These include: taking steps to help expedite the review of efficacy supplements; clearly explaining what data is needed for efficacy supplements for new indications; and seeking out the most appropriate off-label uses and getting them onto the label.

D. H.R. 3200 ROLLS BACK THE PROGRESS MADE BY THE NUTRITION LABELING AND EDUCATION ACT OF 1990

In 1990, Congress passed the Nutrition Labeling and Education Act (NLEA). The driving force behind the NLEA was the desire to end consumer deception and to help consumers in choosing a healthful diet. Among other things, the NLEA sought to ensure that claims about the nutrient content of foods (e.g., "high in vitamin C") would make sense to consumers and that consumers would receive accurate information about the health benefits of foods. The NLEA accomplished these goals by defining nutrient content claims and by requiring that FDA review proposed diet/disease claims to ensure that there is an adequate scientific basis for the claim. H.R. 3200 not only significantly rolls back the progress that NLEA made on both of these fronts, it also undermines our ability to provide information to consumers by limiting the type of information we can require on the food label. For example, section 104 limits our ability to require disclosure of a method of production or an ingredient.

Synonyms. The NLEA sought to make sense out of the panoply of nutrient content claims on food labels, by standardizing the terms that could be used. only those descriptors defined by FDA (through notice and comment rulemaking) (e.g., "low", "good source of" or "high") could be used. H.R. 3200 amends that provision to permit synonyms of those defined terms, provided only that they are not false or misleading.

The bill allows food manufacturers to determine appropriate synonyms. There is no standard for companies to meet before using a new descriptor and the burden would be on the Agency to prove that a term is not a synonym for a defined term or that the term is false or misleading. This provision could result in a large increase of nutrient content descriptors. The uniformity and clarity that the NLEA created would be lost. The resulting plethora of uncontrolled terms would confuse consumers by diminishing the educational value of the defined terms. For example, is "lots of fiber" a synonym for "excellent source of fiber" or "good source of fiber"? The use of such vague terms would permit marketers to expand the market by blurring the distinctions between the nutritive value of different products.

FDA's current regulations permit a company to petition for the use of synonyms. In addition, FDA recently proposed to amend its nutrient content claim regulations to permit manufacturers to use synonyms of defined nutrient descriptors as long as such synonyms are not false and misleading and the synonym is "anchored" to the defined term -- i.e., the defined term appears prominently on the label. The Agency's proposal properly balances the concern of unbridled uses of synonyms that can lead to consumer confusion with industry's need for greater flexibility to market the healthful aspects of their products.

Health Claims. H.R. 3200 would allow any materials prepared or distributed by any federal health agency or by the National Academy of Sciences to serve as the basis for a health claim -i.e., to fulfill the significant scientific agreement requirement.

The provision is vague and overly broad. The bill would appear to allow ANY information prepared for whatever reason (i.e., not specifically for the purpose of evaluating a diet/disease claim relationship) to form the basis of a health claim. For example, materials relating to a particular diet/disease relationship prepared and distributed as background for a hearing by the NAS or a briefing by NCI could be used as the basis for a health claim. Moreover, information contained in pamphlets written by a private organization and distributed by a Federal health agency or the NAS could form the basis for a health claim under this provision. Information could be taken out of context. Finally, there appears to be no remedy for FDA if it disagrees with a claim evaluation made by another agency. FDA would be unable to require a private or public debate about the meaning of the data or information that is relied upon to make the diet/disease claim.

This provision fundamentally alters the criteria for health claims that were so carefully worked out in the Nutrition Labeling and Education Act of 1990. Under the bill, there would be no single body to make decisions on health claims. Because there may be inconsistent decisions by different organizations using different standards, we would see the type of consumer confusion and misinformation that was rampant before passage of the NLEA. Even worse, we may see cases where claims that come from information prepared or distributed by a federal health agency or the NAS are just wrong. For example, although reports and recommendations of certain federal health agencies and the National Academy of Sciences indicated that beta- carotene prevents lung cancer, subsequent studies found not only that beta-carotene was not effective in reducing the risk of cancer, but that for some persons (e.g., smokers) it actually increased their risk of cancer and death.

E. THE REFORM BILLS WOULD INCREASE BUREAUCRATIC REQUIREMENTS WITHOUT PROVIDING ADDITIONAL RESOURCES

All three bills impose significant, new administrative burdens. For example, they require the Agency to hold numerous meetings throughout the review process and to provide detailed written responses to sponsors on virtually any issue that may arise throughout the review process. They also require the Agency to establish a dispute resolution process, which duplicates existing mechanisms, and they subject virtually any FDA decision about drugs to judicial review (thus, threatening to paralyze the Agency with useless litigation).

The excessive burden imposed by the bills is perhaps best illustrated by a provision in the device bill. Section 11 first creates thirteen new deadlines, many of which are unreasonably short. And then, having created these deadlines, which we probably cannot meet, it requires a report be submitted to the Commissioner and the Secretary every time a deadline (including one of the eight interim deadlines) in the bill is missed. The Commissioner then has 10 days to provide an explanation to the applicant of the reasons for failure to meet the deadline. The Secretary must submit an annual report to Congress that summarizes each of these missed deadlines. This provision takes deadlines and reporting to the extreme.

CONCLUSION

Mr. Chairman, I have spent the past five and a half years pushing very hard to get safe and effective products to patients as quickly as possible. We have a strong foundation on which to build: the Prescription Drug User Fee Program, the efforts that have been made under the Administration's reinventing government initiative, and our successes in expediting development and approval of drugs for life- threatening illnesses.

The problem we see with the three bills now being considered is that too often they try to solve yesterday's problems with untested solutions. our concern is that these measures, if enacted into law, would not only do little to actually help FDA get products to patients faster they would undermine our ability to protect American citizen's from unsafe and ineffective products.

Be assured, that the FDA and the Administration will continue to push for change, but the right kind of change. We look forward to continuing to explore new ideas on how to best assess new product effectiveness, how to get products labeled to reflect the best data on their use, how to facilitate clinical development and how to get the Agency's independent review of new products done expeditiously so that important scientific advances move to the bedside more quickly.