Akebia Therapeutics Inc v Fibrogen, Inc

In a patent dispute concerning HIF-PH inhibitors, the court held that broad claims covering enormous classes of compounds were insufficient where the specification did not make therapeutic efficacy plausible across the claim scope and required a substantial, iterative research programme to identify effective compounds.

A functional claim limitation does not cure excessive structural breadth. A claim is also uncertain where the skilled person cannot identify the test for deciding whether a compound satisfies the functional expression. The court applied the doctrine of equivalents but held that a claim limited to a specific compound did not extend to a structurally different compound where the patent and its prosecution history showed that strict compliance was essential.

Akebia and Otsuka sought revocation of six FibroGen patents concerning HIF-PH inhibitors for treating anaemia and related conditions. Astellas, the exclusive licensee, brought a cross-claim for threatened infringement based on the proposed marketing of vadadustat.

The patents were divided into Family A and Family B. The principal issues were construction, obviousness over Epstein and WO 997, insufficiency for lack of plausibility, undue burden and uncertainty, infringement including equivalents, and threatened indirect infringement of medical-use claims.

1. Family A validity. The use of HIF-PH inhibitors for treating anaemia associated with chronic kidney disease was obvious to try over Epstein, but the route to Compound C was not shown to be a routine search likely to identify it. Claim 17A of EP 531 was therefore not obvious. The broader claims of EP 823 and EP 301 were insufficient.
2. Construction and insufficiency. The passage concerning permitted aryl substitutions applied to all aryl groups in the first section of the definition of R₁, R₂ and R₃. The expression “structural mimetic of 2-oxoglutarate” was uncertain because the skilled medicinal chemist could not identify a reliable criterion for distinguishing compounds within and outside the expression.
3. Plausibility and undue burden. The claims implicitly promised that substantially all compounds within their structural definitions would possess the claimed therapeutic efficacy. The specification demonstrated efficacy for only a small number of compounds and supplied no basis for predicting efficacy across the immense class. Identifying further effective compounds would require a substantial, iterative research programme and could not be achieved across the claim scope without undue burden.
4. Family B validity. The uses claimed in the Family B patents were obvious over WO 997, including treatment of anaemia of chronic disease, refractory anaemia, functional iron deficiency and decreasing hepcidin expression. The claims, other than claim 36A of EP 333, were also insufficient for lack of plausibility and undue burden; claims involving “structural mimetic of 2-oxoglutarate” were uncertain.
5. Infringement. Vadadustat fell within Formula (I), so the relevant Family A and Family B claims would have been infringed if valid. Claim 17A of EP 531, limited to Compound C, was not infringed by equivalents. The claim’s inventive concept was the use of that specific molecule, and the specification, amendment history and prosecution history showed that strict compliance was essential.
6. Threatened infringement. The evidence did not establish that vadadustat would be marketed with the knowledge or intention required for indirect infringement under section 60(2) of the Patents Act 1977. The cross-claim for threatened infringement of the Family B patents therefore failed.