Centeno Responds to the FDA: Part 3

In the last post, we saw that Centeno responded to the FDA by noting that in their own history, the FDA regarded cells as drugs by changing the wording in a statute in a manner that did not conform to standard legislative practice and by making a deal with a drug company that involved assigning a status to that product (fast-track) for which it did not qualify. The next cog in the FDA’s regulatory wheel is that of evidence-based medicine (EBM). This is a tough one for me to visit because EBM represents as large part of my own training in research and biomedical sciences. Therefore, Centeno will be calling me to challenge many of my own presuppositions.

A formal definition of EBM isa method that attempts to apply the best available evidence gained from the scientific method to clinical decision-making. EBM does this by assessing the degree of risks and benefits of treatments or a lack of treatment, and diagnostic tests. The ultimate goal of EBM is to help clinicians understand if a treatment will do more good than harm.

This sounds wonderfully academic and objective, and surely EBM provides an excellent method for developing and evaluating new medicines, procedures and devices. However, EBM asserts that this is the ONLY way to perform medical studies.

Centeno at this point makes a very profound historical point. The history of medicine is filled with discoveries that did not utilize the methods of EBM. He gives the following example: Emergency Medicine. Car crash victims are treated with methods that have not been vetted by EBM. If EBM were the rule in the emergency room, most emergency medicine physicians and nurses would have to stand there with their hands tied while the patient on the table dies before them. This suggests that there is more than one way to acquire medical knowledge and test the quality of it.

Historically, there were two schools of thought as to how one should interpret scientific findings. The Bayesian and Frequentist school differed in their approaches to such data. The Bayesian approach comes from Thomas Bayes who argued that future decisions should be dictated largely by past experiences. Thus, for Bayesians, the best way to make a decision lies in mathematical probability calculations that tell you if A is more likely to work better than B, or something like that. Frequentism comes from the work of thinkers such as Aristotle who said that the probable is “that which, for the most part, happens.” The frequentist looks for those events that occur frequently under the controlled conditions found in a laboratory.

In the 1930s, the debate between frequentism and Bayesianism came to a head when Sir Harold Jeffreys, and theoretical physicist who used mathematical models to study the Earth and its solar system, and Sir Ronald Fischer, a population geneticist. Jeffreys used Bayesian probability to construct forma theories of scientific reasoning, but Fischer used frequentism. Because frequentism is very “on” and “off” with respect to its assertions, it seemed more certain. Because frequentism seemed to make a complex work simpler, it won the day because of its perceived explanatory power and elegance. Bayesianism, was viewed as wish-washy and not worth the time of the researcher.

Frequentism works quite well in the laboratory where the conditions are well controlled. However, once the clinical trials go from the laboratory into inbred animals that all have the same genetic background, to people who can differ wildly in their metabolism, digestion, absorption, or receptor binding, receptor densities, or response to receptor binding, such scientific reasoning does seem to represent reality accurately. In order to deal with the shortcomings of frequentism, some clinical researchers has suggested using so-called “adaptive clinical trials” in which the data gathered as the trial progresses are used to change some aspect of the trial midstream.

The real problem with frequentist thinking in medicine is that it tends to cultivate a one-size-fits-all mentality. Centeno uses the following analogy: The average height for men in the US is 5 feet, 9.5 inches. Therefore, if this information was used to make men’s clothes, it would be an unmitigated disaster.

Centeno thinks that this is the entire problem with EBM – its application to medicine that essentially says that there is no other way to treat a patient, and that if the physician does not follow that recommendation, then you are a quack, instead to tailoring your treatment to the individual patient’s needs.

Frequentist thinking also leads to long, massively expensive drug trials that are often unsuccessful that wastes not only gazillions of dollars, but precious time that very sick patients to not have. Furthermore, since universities receive large sums of money from the National Institutes of Health (NIH), they are married to the frequentist system, as are large pharmaceutical companies. Therefore, these two institutions tend to oppose major reform of the FDA.

One other point at this time is this – socialized medicine is frequentism on steroids. If we take the British National Health Service for example, this institution publishes treatment guidelines for a variety of maladies. If the doctor deviates from the treatment protocol, then the NHS will not pay for it. Also, the NHS also specifies who will receive these treatments. It is a cost-analysis that determines if you get treated and what kind of treatment you get, individual variation in patients be damned.

Many medical breakthroughs that led to things such as in vitro fertilization, modern open-heart surgery, and the surgically-implanted plates and screws that stabilize fractures were all developed outside the frequentist system. Instead of working on a narrow clinical trial that includes a handful of people, why not share information about particular drugs and procedures, which the internet allows, and run clinical trials that make sense the observations of a global village of physicians?

An example might be helpful. Hyaluronic acid (HA) is a complex, charged carbohydrate that acts as a lubricant for joints. Arthritis can be treated by injections of HA into arthritic joints and this “WD-40” for the joints help relieve the pain of arthritis and the inflammation associated with it. Physicians who have used HA have shown that it works well for arthritic hips and shoulders as well as knees. According to Centeno, there are some 104 research studies on the use of HA in the hip. However, the FDA will not approve the use of HA outside the knee unless it goes through the same lengthy, expensive trials in which arthritic patients are given placebos that do not help them. Clearly physician experience has shown that HA works outside the knee. Instead of approving it on a limited basis and then subjecting it to past-marketing research, the FDA is stuck in its frequentist way of thinking and this keeps them from seeing the light when it comes to HA.

There is another factor at play here and this is beautifully covered by an article by the Independent Institute. The FDA is motivated to delay or deny potentially useful drugs because of the exposure of their work. Consider the following scenario: If the FDA approves a drug that turns out to be dangerous, victims are hurt and they appear in the press and television. The FDA repents, fires someone and then pulls the drug. If, however, they do not approve a good drug that can save lives, there are no victims, no sob stories in the press and no terrible news stories on MSNBC. Nevertheless these delays have genuine victims. Robert Goldberg of Brandeis University has stated (how exactly he calculated this remains a mystery to me) that the delays in drug and device approvals have cost around 20,000 lives over the past 30 years.

Thus we have a FDA that works with a system that cannot handle the 21st century and the medical advances it will bring.

Published by


Professor of Biochemistry at Spring Arbor University (SAU) in Spring Arbor, MI. Have been at SAU since 1999. Author of The Stem Cell Epistles. Before that I was a postdoctoral research fellow at the University of Pennsylvania in Philadelphia, PA (1997-1999), and Sussex University, Falmer, UK (1994-1997). I studied Cell and Developmental Biology at UC Irvine (PhD 1994), and Microbiology at UC Davis (MA 1986, BS 1984).