Leo Pharma (a/s Leo Laboratories Ltd) v Sandoz Ltd

In patent validity appeals, the Court of Appeal should not substitute its own evaluation for the trial judge’s where the findings are supported by material and the overall assessment contains no error of principle. For anticipation by inevitable result, the challenger must prove that following the prior art necessarily produces the claimed product. Experiments must address the prior art as disclosed, since impurities, seeding and processing conditions may affect crystallisation. Obviousness based on accidental discovery requires confidence that the relevant steps were really likely or almost mandated. A screening or crystallisation programme is insufficient unless it was something the skilled team would undertake and would probably reveal the invention.

Sandoz appealed Floyd J’s decision that EP (UK) No. 0 679 154, concerning calcipotriol monohydrate, was valid: [2009] EWHC 996 (Pat). Infringement was common ground if the patent was valid.

Sandoz challenged validity on anticipation by example 4 of an earlier acne-use patent and on four forms of obviousness, involving polymorph screening, product screening, wet milling and routine crystallisation experiments. The central issues were whether the monohydrate was the inevitable result of the prior art and whether the skilled team would have discovered it through the proposed investigations.

1. Appeal dismissed. Lord Justice Jacob held, with Lord Justice Patten agreeing, that an appeal concerning factual findings, evidential evaluations and value judgments succeeds only if the lower court’s decision was wrong under Part 52.11(3) of the Civil Procedure Rules 1998. The Court applied the approach in Merck’s Patents [2003] EWCA Civ 1545 and the Biogen principle [1997] RPC 1. Obviousness required an overall assessment of the technical and non-technical evidence, and there was no error of principle in the judge’s assessment.
2. Anticipation. The challenger bore the burden of proving that carrying out the prior art necessarily produced the claimed monohydrate. The court had to assess the evidence as a whole. Where the starting material and its impurities might affect crystallisation, the burden did not shift to the patentee. Sandoz’s experiment departed from the disclosed recipe, used a different starting material and was seeded with monohydrate. It therefore did not establish inevitability. The theoretical case also failed because crystallisation depended on kinetics, nucleation, processing conditions and possible super-seeding.
3. Obviousness. Where an invention was said to have been found accidentally, the Court required strong confidence that the relevant steps were really likely, almost mandated. A full polymorph screen was not obvious merely because hydrates were known to be possible. Real-world practice, regulatory requirements, motivation, the difficulty of crystallisation and expectation of success all mattered. A limited stability or regulatory investigation might have been routine, but it was not proved that it would reveal the hydrate. Wet milling was an obvious variant, but it was not shown to produce the hydrate reliably. The proposed routine crystallisation experiments were undefined and unsupported by evidence.
4. The trial judge was entitled to reject inconsistent expert evidence and to find that the appearance of the monohydrate was unpredictable. The patent was valid and the appeal was dismissed.

1. Court of Appeal (Civil Division). Sandoz’s appeal was dismissed and the decision below was upheld: [2009] EWCA Civ 1188.
2. High Court of Justice, Chancery Division (Patents Court). Floyd J held the patent valid: [2009] EWHC 996 (Pat).