Thursday, March 18, 2010
The FDA has just set off what could be an earthquake of 9.5 on the Richter scale! This apparently came to pass during discussions between the FDA and a sponsor (I do not know who, but we can all try and guess). They were discussing an end of Phase II proposal and preparing for their pivotal trials to register a new antibiotic for the treatment of serious infections of skin and skin structures. The FDA apparently informed them that they no longer understood how such a trial should be designed because they decided that they did not understand the effect of antibiotics in these serious infections. When I first heard this, I was dumbfounded since the data showing that antibiotics have a dramatically beneficial effect in skin infections was just reviewed with the FDA publicly in 2008 and the issue seemed to be settled with consensus among all involved. The data that was reviewed again came mostly from the pre-antibiotic era and was sparser and less consistent than the data we reviewed for pneumonia. Nevertheless, all present, including the FDA’s advisory committee, seemed convinced that antibiotics worked extremely well in the treatment of these serious infections. So what happened since 2008?
We reviewed pneumonia and decided that the greatest clinical effect of antibiotics in the treatment of pneumonia occurred early in the treatment course – in the first 72 hours or so. After this time, the antibiotic effect diminished. What I surmise from discussions with Ed Cox of the FDA and others in the infectious diseases community is that the FDA then went back to re-review the old data on skin infections. They became concerned that, like pneumonia, the greatest clinical effect occurred early. If so, they reasoned, then the current endpoints looking at cure much later, usually after two weeks or so, were inappropriate. They further worried that the old data, comparing to placebo where the time to response was studied were so limited as to preclude their ability to actually define the extent of this response in a reliable way. So they are now saying that since they cannot define the response to treatment compared to placebo, they certainly cannot define how to run a trial where two antibiotics are compared to each other. What should the acceptable potential difference be between the new agent and the comparator antibiotic they wonder? So they are telling industry that if a company starts a pivotal trial before the FDA can understand the science, they do so at their own risk.
Of course, companies have been on their own for the past several years anyway – so what is new here? What is new is that there is really no way of knowing where the FDA will go here and therefore that an investment of $70 million in two pivotal trials could well be money down the drain to say nothing of exposing so many patients to experimental therapy unnecessarily.
To me, and virtually every infectious diseases clinician that I have spoken to about this, the FDA seems to be living in a different dimension where huge amounts of data and our own experience can no longer be believed. The FDA seems to think that they can stop the world from turning on its axis while they try and understand the science that the rest of us already think we understand well, and that they can do so in the name of science with no ill effect. But this time, given the number of companies small and large that are counting on approval for use of their antibiotic for the treatment of skin infections, this will be the end of the world as we know it. This approach will devastate biotech and undermine large pharma to the point where we seriously risk our ability to have any new antibiotics in the next decade or so. How can the FDA possibly contemplate such a step?