Astex Therapeutics Ltd v Astrazeneca AB

A contractual definition of a collaboration compound must be construed in the context of the agreement as a whole. Where eligibility depends on chemical structure modification performed with a specified aim, the enquiry is historical and objective, focusing on what the scientists did, their starting point, their aim and the result. A compound must have a sufficiently close connection with the optimisation work to be its direct result; a long or indirect development path will not suffice. The agreement may also make selection or nomination a condition of status. An agreement expressed to continue while specified activities continue may expire when those activities cease, even without formal termination.

Astex and AstraZeneca entered into a research collaboration concerning BACE inhibitors. The agreement provided for milestone payments and royalties in respect of “Collaboration Compounds”. After the collaboration term ended, AstraZeneca developed two candidate drugs, CD1 and CD2, and initially paid milestone sums concerning CD1.

AstraZeneca later contended that neither compound was a Collaboration Compound. Astex claimed contractual and related relief. AstraZeneca counterclaimed for restitution of the two payments made concerning CD1. The court determined the proper construction of the agreement, whether CD1 and CD2 qualified, whether the agreement could expire, and whether the payments were recoverable.

1. The claim was dismissed insofar as it depended on CD1 or CD2 being Collaboration Compounds. On the proper construction of the agreement, the Program ended when the Collaboration Term ended on 20 April 2005. Work performed thereafter was not work “performed as part of the Program”.

2. Hits and Leads required a positive selection or nomination under the agreement. There was no obligation on AstraZeneca to nominate Hits or Leads after the Collaboration Term, and its failure to do so did not confer that status on later compounds. A CD could be selected after the Collaboration Term, but that did not resolve the separate requirements of the definition of Collaboration Compound.

3. “Chemical structure modification” meant modifications actually performed as part of the Program. The relevant enquiry was historical, based on the work done by the scientists, the specific starting point, the aim and the result, without hindsight. The aim was the scientists’ objectively assessed intention, including any immediate purpose and broader aim.

4. The “direct result” requirement imposed a close-connection test and was intended to provide a reasonably bright-line rule. A result could be direct despite an intervening step, but the more steps involved, the less likely that conclusion became. Obviousness, inventiveness, breakthroughs and major advances were irrelevant.

5. CD1 was not a Collaboration Compound. Its development did not begin from qualifying AFFITs, Hits or Leads, and the path from any qualifying optimisation work was too long and indirect. CD2 likewise was not a Collaboration Compound. The AiZ core was inspired by, but was not chemically modified from, the DHIZ core, and no AiZ had been selected or nominated as a Lead under the agreement.

6. AstraZeneca was mistaken when the payments concerning CD1 were made, and that mistake caused the payments. It was therefore entitled to restitution of the two sums of US$1 million.

7. The agreement was capable of expiring. It would expire if AstraZeneca ceased pursuing pre-clinical research referable to the Results. The court granted a declaration to that effect.