Astex Therapeutics Ltd v Astrazeneca AB

In construing a commercial drug-discovery collaboration agreement, the court held that a research Program defined as collaborative work performed during a Collaboration Term ended when that term ended, despite provisions permitting continuation by one party alone. A party therefore earned milestone and royalty payments only for qualifying compounds discovered as a direct result of work performed as part of the Program during that term. On costs, an admissible settlement offer outside Part 36 may be considered under CPR 44.3(4), but comparison with the eventual result cannot alone justify indemnity costs. Extensive disclosure or amendments do not necessarily take conduct outside the norm.

Astex and AstraZeneca entered into a 2003 agreement for a collaborative programme to discover and develop BACE inhibitors for Alzheimer’s disease. The Collaboration Term ended on 20 April 2005. AstraZeneca later nominated two compounds as Candidate Drugs and made milestone payments for one, but subsequently denied that either was a Collaboration Compound.

Arnold J interpreted the Agreement as meaning that the Program ended with the Collaboration Term, found that neither compound qualified, dismissed Astex’s claim and ordered repayment and indemnity costs from 6 December 2016: [2017] EWHC 1442 (Ch). Astex appealed on contractual interpretation, factual derivation and costs. The central issues were the duration of the Program and whether the conduct of the litigation justified indemnity costs.

Floyd LJ gave the judgment, with Henderson and Leggatt LJJ agreeing. The main appeal was dismissed. The independent costs appeal was allowed, the indemnity costs order was set aside, and AstraZeneca’s costs were ordered to be assessed on the standard basis throughout.

1. Duration of the Program. The definition of Program in section 1.32 was construed as referring to the collaborative research programme performed during the Collaboration Term as part of AstraZeneca’s Project. The ordinary grammar, the description of the Program as collaborative, and the need for a provision defining its temporal extent favoured that construction. Sections 3.7 and 3.8 were compelling because they distinguished the ended Program from the continuing Project. The dissolution of the JEC also left no equivalent post-term mechanism for making milestone decisions or supervising the successive stages of the Program.
2. Isolated provisions, including the reference to selection of Leads after the Collaboration Term, did not outweigh the Agreement read as a whole. Nor did the surplus wording in the definition of Results. The court rejected the proposed requirement that compounds be sufficiently referable to collaboration work, since Materials included other materials used in the Program and were not confined to the pre-existing screening libraries. The rival construction would give the Agreement a surprising and potentially limitless reach over AstraZeneca’s pre-existing BACE project.
3. Contractual consequence. Since the Program ended with the Collaboration Term, CD1 and CD2 were not discovered as a direct result of optimisation performed as part of the Program. Astex was therefore not entitled to milestone or royalty payments in respect of them. It was unnecessary to decide the further issues concerning selection, whether optimisation had to start from a specific compound, or the factual derivation of CD2.
4. Indemnity costs. The appellate court should not interfere with a costs discretion unless it is wrong in principle, takes account of irrelevant matters, omits relevant matters, or is plainly unsustainable: F&C Alternative Investments (Holdings) Ltd Barthelemy (No 3) [2013] 1 WLR 548. A non-Part 36 offer had to be considered under rule CPR 44.3(4), but Part 36 could not be applied by analogy so that comparison between the offer and the trial result alone justified indemnity costs. Refusal of a reasonable offer could contribute to such an order only in a special case and where the refusal was unreasonable, a result described as rare in Kiam v MGM Ltd (No 2) [2002] 1 WLR 2810.
5. Extensive disclosure was reasonable because Astex was entitled to test AstraZeneca’s factual account by disclosure and evidence. The later spreadsheet provided some support for Astex’s case. Repeated amendments did not establish conduct outside the norm, particularly since the judge had been able to identify and determine the case on derivation. No feature of Astex’s conduct justified indemnity costs.

• Court of Appeal (Civil Division): [2018] EWCA Civ 2444. The main appeal was dismissed. The costs appeal was allowed and costs were ordered on the standard basis throughout.
• High Court of Justice, Chancery Division: [2017] EWHC 1442 (Ch). Astex’s claim was dismissed, repayment was ordered, and AstraZeneca’s costs from 6 December 2016 were assessed on the indemnity basis.