The Food and Drug Administration has long been criticized for delaying the approval of new medical drugs and devices. Under outgoing Commissioner David Kessler, the FDA attempted to improve its image. For example, in January 1997, the FDA proudly announced that it had cleared 139 new drugs and biological products in 1996, "a record one-year jump" from 1995. The agency claimed to have approved these products faster than ever. Dr. Kessler was even so confident of his numbers that he has challenged his critics, brashly stating on national television in mid-February, "I ask [my critics] to name one drug, one therapy, that's available somewhere else that works and is not available in this country."
Well, if Dr. Kessler needs the names of drugs he's been sitting on, we've got them right here. A recently released CEI study, Treatment Delayed, Treatment Denied, showcases not only drugs and medical devices available abroad and not in the United States, but also therapies which the FDA delayed in this country.
Taxotere, for example, is a breast cancer drug the FDA approved in May 1996. It was submitted in July 1994 simultaneously in the U.S., Europe and Canada, and its drawn-out approval process illustrates the true meaning of "drug lag." Taxotere was approved in one year by Canada and in only 6 months in Europe. It took nearly two years in this country.
Ironically, the FDA used Taxotere as an example of its new accelerated approval program for cancer drugs (a program originally proposed by the Bush Administration). If this major new drug is considered "accelerated," then it does not bode well for drugs of lesser importance.
Four hundred thousand Americans have angioplasty surgery every year to ameliorate coronary artery disease. A recently approved device, Vasoseal, is used to plug the wound incurred during the procedure. It is faster to apply, it requires only brief pressure, and it reduces the time patients must remain immobile. Vasoseal was submitted to the FDA in 1992, and was not approved until January 1997, a five-year wait. Meanwhile, it had already been available for over two years in Canada, Australia, and the Netherlands.
Finally, there is Ancrod, a stroke treatment. It is made from the venom of the Malayan pit viper, and has a unique property which aids in treating blood clots in the brain. Ancrod is currently in the last round of clinical trials in this country, but it's been available in the United Kingdom since 1974. In fact, Ancrod was singled out by the GAO in Congressional testimony as a drug which ought to be made available as quickly as possible. Strikingly, GAO made this recommendation eighteen years ago, in 1979.
These individual examples illustrate that, contrary to its public pronouncements, the FDA has not improved under Dr. Kessler. His successor, when selected, will come before the Senate Labor Committee. Senators on that panel have a grave responsibility to ensure that we do not get another regime like Kessler's. The agency has marshalled impressive statistics, but they cannot mask the FDA's flaws.
The FDA maintains that there has been a true decline in new drug application (NDA) approval time – the time which a new drug sits in the approval queue at the FDA. In October 1995, the GAO wrote that the average New Drug Approval time for NDAs dropped between 1987 and 1992 from 33 months to 19 months.
As it happens, this is technically correct. There has been a slight improvement in the overall average due to several factors. For example, AIDS drugs are now approved more quickly than others, and so their acceleration improved the average. People with AIDS have celebrity spokespeople, well-organized interest groups, and political savvy. Other groups have not been as vocal, and have suffered for it. It is truly a sad state of affairs when sick people must worry not only about their conditions, but about how best to work the levers of politics. There is nothing wrong in getting AIDS drugs out fast, but there is something wrong when only AIDS drugs are delivered quickly.
The FDA has also seen an 80 percent increase in funding in the past few years and a 20 percent increase in staff under Dr. Kessler. This has primarily been due to user fees, a tax on drug manufacturers. While some of this extra cash has enabled the FDA to speed up reviews on certain drugs, the FDA has spent a larger percentage of this on increased enforcement activities.
Because the improvement in drug approval times is relatively modest, the FDA is obliged to puff up its numbers. Many of the drugs the FDA claims are new are often no more than different versions of the same drug (pill form as opposed to injectable form); others are combinations of older drugs (for example, heart drugs such as Tarka and Lexxel).
The agency is also not averse to asking companies to withdraw applications and resubmit them later. Consider the epilepsy drug Topamax. It is part of an entirely new class of anticonvulsants which could help as many as 600,000 epileptics. When the FDA approved Topamax in December 1996, it had been available in the United Kingdom since October 1995. The FDA's approval came two years after Topamax had been submitted for approval in December 1994. However, Topamax had originally been submitted in August 1994, but the FDA, in asking for more information, requested that the drug company withdraw the NDA and resubmit it later. Add on the four months, and the FDA's record looks even worse.
In short, official NDA time is not an informative indicator of FDA performance. It is only the last stretch of an increasingly longer race. The really interesting numbers are the total drug development time, which has increased from 12 to 15 years in the past decade, and shows no sign of shrinking. This increase in development time is partially due to the FDA's regulatory barriers.
The situation with medical devices is even worse. The United States lags far behind other medically advanced countries in access to state-of-the-art medical devices, and many medical device manufacturers have given up on introducing their products in the U.S. first, opting to start production in Europe.
For some medical devices, known as 501(k)s, Dr. Kessler again tried to obscure his terrible record. Another October 1995 GAO study noted that the median approval time for 501(k) medical devices had dropped since 1993, but also pointed out that the FDA only recently switched to using the median instead of the mean. The mean is the average of a set of numbers, the median is the middle value. As the GAO noted, "When the mean is larger than the median . . . it indicates that a group of applications required lengthy reviews."
The median can be calculated even when some applications are still on hold. The mean will increase, but the median will stay the same. The FDA also knows that many people will mistake the median for the mean, so one year's median might compare favorably with other year's mean times.
For other device approval applications, known as PMAs ("Pre-Marketing Applications"), it was to the FDA's advantage to admit defeat. "Even though we are approving more PMAs, and we have improved the time to first action," intoned Dr. Kessler in a December 1996 speech, the PMA approval time is coming down only slowly. It still takes too long, more than two years, to get a device through this process."
Why is this? The FDA is "ready, willing, and able to bring down the PMA review times, just as we brought down the review times for NDAs, if the resources are made available" [emphasis added]. This is a reference to Dr. Kessler's repeated requests to Congress for the power to offer "user fees" to medical device companies.
The FDA's machinations have not demonstrably improved the lives of American patients. For example, recent CEI polls of cardiologists and oncologists demonstrate extremely negative views of recent FDA performance by these specialists. Indeed, 65 percent of cardiologists and 77 percent of oncologists agreed that "the FDA is too slow in approving new drugs and medical devices." It is a frightening thought coming from doctors operating on the front lines of medicine.
The central question of FDA reform is simple: If the FDA approves a therapy which will start saving lives tomorrow, then how many people died yesterday waiting for the government to act. Dr. Kessler has largely disregarded this question, spending less time worrying about the health and safety of Americans than he has on the health and safety of the FDA. The next FDA commissioner ought to keep in mind that any treatment the FDA delays is a treatment denied to American patients.
P.S. What will happen to Dr. Kessler now that he has left the FDA? His new job is dean of Yale University Medical School. Reportedly Yale Med is having financial problems, and Dr. Kessler's job is to streamline the school's bureaucracy. If Dr. Kessler runs the school the way he has run the FDA, God help this man at Yale.
Julie DeFalco is a CEI policy analyst.