Response to Nathan
The following letter was written by me in response to David Nathan’s February 1st Perspective in the NEJM entitled: “Finding New Treatments for Diabetes: How Many, How Fast…How Good?” (Subscription req.) In this essay, Nathan argues that “FDA’s approval process for new antidiabetes medications should take into account their additional and unique contributions, especially when their glucose-lowering efficacy is similar to or less than that of currently available medications.”
In essence Nathan questions the wisdom of allowing approval of new drugs (for chronic diseases, like diabetes) that have not been demonstrated to be better than existing drugs. My response offers a scientific reason why regulators should not have such discretion; unmade economic arguments are at least as compelling.
_______________________________________________________
Dr. Nathan throws light on the subject of hurdles for drug approvability in the U.S. Although his arguments are specific to drugs used to treat Type 2 diabetes, the points raised are germane to a number of chronic disease states, where multiple drug classes are currently in use. Under current U.S. law, FDA lacks authority to restrict approval of new drugs because they might be deemed unnecessary, no matter in whose judgment. It is a mistake to believe that putting such authority into the hands of FDA today is wise, for the simple reason that traditional pre-marketing studies of new drugs, such as those cited by Dr. Nathan, are incapable of determining which individual patients will or will not benefit from use of drugs. Rather, these studies aim to show average effects in populations selected to be representative of users; they ask general safety and efficacy questions and yield general guidance only. Many hope that future pre-marketing approval studies that incorporate highly predictive markers of risk and benefit (i.e. personalized medicine) will allow regulators to tailor approval decisions to specific subsets of patients, but until then it remains for doctors to perform the “N of 1” experiments needed to determine whether a drug benefits an individual or not. A system for capturing results emanating from these large number of experiments is sorely needed. In the meantime, clinicians considering prescribing new drugs for conditions like Type 2 diabetes should demand comprehensive monographs from manufacturers that include all relevant preclinical and clinical information.
