What kinds of patents does the pharmaceutical industry seek besides patents on the active molecule itself, and how common are these "secondary patents"? Amy Kapczynski (Yale Law), Chan Park (Medicines Patent Pool), and Bhaven Sampat (Columbia Public Health) address these questions in their new PLOS ONE article, Polymorphs and Prodrugs and Salts (Oh My!): An Empirical Analysis of "Secondary" Pharmaceutical Patents. It's a quick read (11 pages), so it is worth downloading for anyone interested in these issues.
The authors examined the 528 new molecular entities (NMEs) approved by the FDA from 1988 to 2005, which map to 1261 distinct patents in the FDA's Orange Book. They coded whether each patent had "primary" claims (i.e., claims for the chemical compound itself) and various types of "secondary" claims (which they categorized as formulation claims; method of use claims; or polymorph, isomer, prodrug, ester, salts (PIPES) claims). 64% of drugs were protected by primary claims (so over a third of drugs don't have a chemical compound claim at all!), 81% by formulation claims, 83% by method of use claims, and 51% by PIPES claims. Many of these secondary patents are "independent" secondary patents—i.e., they have no primary chemical compound claims. Only 2% of patents with primary claims were filed after FDA approval, but about 20% of independent secondary patents were filed after approval, and on average these independent secondary patents added 6-7 years of patent life (as measured from the last expiring chemical compound patent or the expiration of the regulatory exclusivity period). High sales drugs are more likely to have independent secondary patents (e.g., drugs in the top sales category are 18% more likely to have a post-approval independent secondary patent).
As I have discussed on this blog, previous work by Bhaven Sampat and Scott Hemphill (on leave from Columbia Law) has shown that patent challenges by generic manufacturers focus on secondary patents, suggesting that secondary patents tend to be easier to invalidate. Sampat and Hemphill have thus referred to secondary patents as "weak" patents. It may be true that secondary patents are on average easier to invalidate, but referring to them all as "weak" patents (with the associated pejorative connotations) seems to ignore the possibility that their validity is simply more uncertain, or that most are still "strong" (in the sense of being unlikely to be invalidated). It is also not obvious that current patentability doctrines for pharmaceuticals always align validity with social value. Readers, what do you think of the validity or social benefit of secondary pharmaceutical patents? (I'm genuinely curious, so comments here or over email are welcome.)
In any case, this paper is agnostic about the value of secondary patents. The authors state that they "do not attempt here to mediate between those who favor and oppose secondary patents," although they do suggest that developing countries might benefit from not allowing secondary patents (which Kapcyznski has previously argued are not required by TRIPS). But their main take-home message is the empirical point that "secondary patents are of substantial importance to the industry, and that analyses that focus only on chemical compound patents will tend to underreport both the breadth and range (term) of patent coverage in the pharmaceutical sector."