Sandoz GmbH v Roche Diagnostics GmbH

In assessing obviousness, the court must identify the inventive concept, determine the common general knowledge of the skilled addressee, identify the differences from the prior art, and ask whether those differences required invention. An experiment is not obvious merely because it would be interesting or technically possible. The skilled addressee must have regarded it as sufficiently likely to produce a valuable result to warrant the trial. In a protein-based pharmaceutical formulation, recognised antimicrobial synergy does not by itself make a combination obvious where the individual preservatives would previously have been rejected because of their apparent incompatibility with proteins. The assessment must avoid hindsight and reflect the selective, conservative approach reasonably adopted at the priority date.

The claimant sought revocation of European Patent (UK) No 0,607,156, concerning preserved multi-dose injectable formulations containing erythropoietin. The defendant amended the claims so that they concerned erythropoietin and combinations of specified preservatives, particularly benzyl alcohol with benzalkonium chloride, at defined concentrations. The claimant alleged lack of inventive step over common general knowledge, the Konings international patent application, and publications concerning synergistic preservative combinations. The central issue was whether, at the priority date, the skilled formulator would have regarded testing those combinations in a protein-based multi-dose parenteral as an obvious experiment.

1. Disposition. The claim for revocation failed. The amended claims were not shown to be obvious and the claim was dismissed. Questions concerning discretion to amend, and the form of order and costs, were left for further consideration.
2. Obviousness framework. The court applied the four-stage approach in Windsurfing International Inc v Tabur Marine (Great Britain) Ltd [1985] RPC 59: identify the inventive concept; identify the normally skilled but unimaginative addressee and common general knowledge; identify the differences from the prior art; and determine whether those differences were obvious without knowledge of the invention.
3. The skilled addressee could be a team, but the relevant team was confined to those involved in production. The protein scientist’s knowledge was pivotal because the formulation had to preserve both antimicrobial efficacy and the integrity of the protein. Regulatory expertise was not part of the production team.
4. Common general knowledge included reliable standard reference works and material generally regarded as a foundation for further work. It did not include the entire contents of encyclopaedic works or scientific articles merely because they were available or widely read, unless their contents had become received wisdom.
5. For an invention to be obvious to try, the skilled addressee must have regarded the proposed trial as holding a sufficient prospect of producing a valuable result to justify the expense and effort. The evidence showed that benzalkonium chloride and chlorobutanol would have been excluded from the initial shortlist because of incompatibility, instability or inadequate bactericidal properties. Knowledge that certain preservatives could act synergistically did not make it obvious to abandon that protein-focused selection strategy and test combinations containing preservatives already rejected for single use.
6. The same reasoning defeated the arguments based on Konings and Denyer. Konings contained an opaque list of preservatives for different uses and did not provide reliable teaching beyond common general knowledge. Denyer, Boehm and Stock concerned general or non-injectable applications and supplied no sufficient teaching that the combinations would preserve a protein formulation without degradation.