Tuesday, October 18, 2011

What To Do about Pay-for-Delay

Dr. Aaron Kesselheim and colleagues provide a useful review/commentary in last week's New England Journal (subscription required) on a problem I addressed in HOOKED--"pay-for-delay" when a brand name drug company and a generic company reach a settlement on the latter's patent challenge to the former, which results in a substantial delay in a cheaper generic drug reaching the market.

In quick summary, here's what happens in a typical case:

  • A lucrative brand name drug is about to go off patent. Typically, however, the company has a number of subsidiary patents on things like coating, etc. that have little to do with the actual active drug ingredient.

  • A generic company bids to come onto the market with the drug--by the Hatch-Waxman Act, the first company to do so gets a plum in the form of 6-month generic exclusivity, during which it can charge almost as much as the brand name price due to lack of competition from other generic makers.

  • The brand name company sues the generic company for patent infringement, often citing one or more of the subsidiary patents.

  • A major and potentiallly protracted court battle ensues.

  • The brand name company offers the generic a cash settlement, in effect paying them to drop their suit and take the cash, on the condition that they delay their product's marketing and preserve the brand name exclusivity for a longer period.
It seems at first that this process hardly serves the public interest. But the authors note that a prolonged court battle could just as easily keep the cheaper generics off the market even longer.

Kesselheim and colleagues make a number of good legal points and discuss some statutory fixes, most of which have little chance of passing the present Congress. But I think their most important argument is about the root cause of all this--silly patents. They propose two measures that could occur either at the Patent Office or at the FDA levels that would allow administrative challenges to minor patents that do not protect truly innovative drug development, leaving the courts free to decide the really tough cases, and industry free to patent and profit from real innovations. These reforms in my view should be the highest priority.

Kesselheim AS, Murtagh L, Mello MM. "Pay for delay" settlements of disputes over pharmaceutical patents. New England Journal of Medicine 365:1439-1445, Oct. 13, 2011.


Marilyn Mann said...

"either at the Patent Office or FDA levels"

You mean either at the Patent Office or FTC levels.

Howard Brody said...

Thanks as always for writing, Marilyn, but actually I DID mean the FDA, not the FTC. In order for a patent to be officially linked to a drug, so that the generic manufacturer has to take it into account, it must be listed in the FDA's "Orange Book." The FDA if it had the backbone could refuse to list a lot of the minor, subsidiary patents in their book right now and that would end a lot of the problem. The authors of the article I review suggest an administrative appeal, whereby a generic manufacturer could challenge a patent's listing in the Orange Book without having to go to court. Cheers, Howard

Controlled Substances said...

This is unfortunately one time where I agree with the FTC. (Normally, I find the FTC obnoxious and ridiculous in what they go after in the patent arena.) But the practice the FTC is going after speaks volumes about the "mess" that Hatch-Waxman is.

Marilyn Mann said...

Hmm... I had thought you were talking about the paragraph where they say that "a commissioner of the FTC reportedly proposed adoption of an FTC rule requiring that the parties to a settlement prove that the agreement is not anticompetitive."

I see now you must be talking about the last paragraph on page 1443, which discusses an administrative review of a patent's validity after a pharmaceutical patent is first listed in the Orange Book. I had read that (in light of the previous paragraph) as still referring to a review before the USPTO. The language is not as clear as it could be, though.