FDA’s recent draft of “Guidance for Industry Co-development of Two or More Unmarketed Investigational Drugs for Use in Combination” brought back memories of my early years in the drug industry more than 20 years ago with Sterling Winthrop in Rensselaer, NY. At that time the cardiovascular pharmacology department where I worked was on the cutting edge of research on cyclic nucleotide phosphodiesterase (PDE) inhibitors.
We had discovered numerous potent and selective inhibitors for most of the PDE isozymes that were then known. We knew that other drug companies including Warner Lambert (since swallowed up by Pfizer) and SmithKline Beecham (now GlaxoSmithKline) were working in the same area for some of the same indications.
In the early 90s, probably around the same time as someone at Pfizer had the same idea, we realized that a selective PDE5 inhibitor could be efficacious in the treatment of impotence. Our official target was heart failure, and we couldn’t help but notice that many of the dogs we were treating for that cardiovascular condition were exhibiting the side effect of spontaneous erections. But we couldn’t convince our marketing colleagues that there was a market for that indication. Several years later Pfizer launched sildenafil (Viagra®), coined the term “erectile dysfunction,” and the rest is history. But I digress.
Two things in particular bothered me about drug development in general, at least the way it was practiced at Sterling. First, our goal was to discover the most potent and selective inhibitors (usually) of a selected molecular target (generally an enzyme or a receptor). The theory was that with a very potent drug you could give a lower dose, which should be safer, and a highly selective drug would be less likely to have off-target effects.
The problem, as I saw it, was that we were developing excellent pharmacologic tools, but not necessarily useful drugs. A compound that is very potent at one molecular target might well be just as potent, or even more so, at another. Or, for example, it could cross the blood-brain barrier and cause CNS side effects, whereas another less potent compound might not get into the brain at all. Furthermore, even at that time we knew that many useful drugs were “dirty” in the sense that they hit multiple targets. In many cases, the built-in polypharmacology was necessary for the drug’s efficacy, although we didn’t understand why (still don’t) and couldn’t model it (still can’t).
Second, and this point is more relevant to where I started, we would undoubtedly have made faster progress if we had been able to collaborate with scientists at other companies who were working on the same molecular mechanisms (e.g., PDE inhibition) and/or therapeutic targets (e.g., heart failure). There were too many hurdles and they were too daunting for someone at my level (managing a group of lab technicians) to consider surmounting: commercial (greed), marketing (e.g., is there or isn’t there a market for a drug to treat impotence), legal (liability, co-development responsibilities, etc.), intellectual property (ownership of ideas, composition of matter, etc.), and regulatory.
The FDA draft guidance on co-developing investigational drug combinations addresses and lowers only the last of these hurdles. I suspect that the infamous patent cliff, multiple late-stage drug failures, dry pipelines, and scarcity of innovation may be sufficient to drive drug developers to overcome the other barriers as well. What do you think?