I’ve been thinking a lot about the relationship between the pharmaceutical industry and the Food and Drug Administration recently. Because of this, I’ve decided to author a series of periodic posts to explore this issue and solicit comments from readers on this subject. With this post, I’m creating a new category on this blog, called “FDA and the Pharmaceutical Industry,†and will file articles on these topics under this heading.
The Issue: Non-inferiority vs. Symptom Resolution in Antibiotic Clinical Trials
I attended an interesting talk this week given by Dr. John Powers of the Food and Drug Administration focusing on the development of antimicrobial medications. He pointed out a number of interesting facts about the growing rate of bacterial resistance to antibiotic medications, the lack of new drug development and the need for improved clinical trials. He noted that his talk does not represent the views of the Food and Drug Administration. (Powers delivered his lecture at the CDC’s annual Get Smart: Know When Antibiotics Work conference.)
Over the years I’ve noticed that a common theme of presentations delivered by FDA officials is the design of pharmaceutical industry clinical trials. Powers’ lecture was no exception. He pointed out that manufacturers making new antibiotics must demonstrate that their medications are non-inferior to currently available antimicrobials. For example, a drug maker developing a new macrolide antibiotic (i.e., drugs like Zithromax) does not have to show that it works better than other drugs in the same class.
Powers argued that non-inferiority is an inappropriate criterion for evaluating new antibiotics. Instead, he suggested that pharmaceutical companies should be required to demonstrate that their drug is superior to other medications on the market, specifically when it comes to symptom resolution. This means that a new macrolide would have to reduce the symptoms of a bacterial infections faster than similar medications.
Pharma’s Reaction
After hearing Powers’ presentation, I spoke with a few people I know from the pharmaceutical industry to gauge their reaction to this suggestion. They said:
• Superiority Is Difficult To Prove: They noted that it would be nice to show that an antibiotic positively impacts symptom resolution. However, this type of trial would be difficult to design and mount, particularly because symptom resolution is hard to measure. (I wonder if there are any validated and well-accepted instruments that investigators use to determine how well a drug reduces the symptoms of an infection.)
• The FDA Suggests But Does Not Require Changes In Trial Design: Non-inferiority trials have been the standard of antibiotic study design for decades. Companies might feel more compelled to change the design of antibiotic clinical trials if the FDA required it. However, the FDA has not done so. Companies are unlikely to make major shifts without specific guidance from the FDA about what protocols will be accepted, which instruments are valid and how advisory boards will evaluate data from these studies.
In Summary . . .
Clearly, clinical trial design can and should be improved. However, changes are unlikely to occur if the FDA does not require them. In addition, the FDA would have to work closely with industry on protocol design, as it already does. Finally, in the case of symptom resolution, the FDA would have to help develop validated and acceptable measures.
It will be interesting to see whether and how the FDA adopts Powers’ recommendations.
Next Up: The FDA: Paper Tiger Or Overzealous Regulator?