Generics (UK) Ltd & Ors v H Lundbeck A/S

A patent monopoly must correspond to the disclosed technical contribution. Where obtaining and testing an individual enantiomer is an obvious goal, invention in finding a particular route does not justify claims covering independent routes which owe nothing to that teaching. A new and non-obvious useful compound may support a product claim when a way of making it is disclosed.

Claims are construed in the context of the whole specification. A claim to an isolated enantiomer can exclude its presence in a known racemate. Obviousness requires assessment without hindsight, taking account of common general knowledge, research options and prospects of success. Later discoveries of benefits neither described nor foreshadowed in the specification cannot bolster inventiveness.

Generics (UK) Limited, Arrow Generics Limited, Teva UK Limited and Teva Pharmaceuticals Limited sought revocation of a European patent owned by H. Lundbeck A/S, a pharmaceutical research company. The three claims were treated together under an earlier direction concerning their evidence.

The patent concerned escitalopram, the isolated (+) enantiomer of the known antidepressant citalopram. Citalopram was a racemate containing both enantiomers. Claim 1 covered the (+) enantiomer, claim 3 a pharmaceutical composition containing it, and claim 6 a preparation method using a resolved diol intermediate while preserving stereochemistry.

The claimants challenged novelty and obviousness using two earlier US patents and common general knowledge. They also alleged that the product and composition claims exceeded the patent's technical contribution. Lundbeck relied on unexpected therapeutic benefits and commercial success, and conditionally sought amendment. The central questions were whether the claims excluded the racemate, whether the claimed subject matter was obvious, and whether the disclosed preparation method justified claims covering every way of obtaining the enantiomer.

1. Claims 1 and 3 were invalid for insufficiency. The other allegations of invalidity failed, and claim 6 was valid. The court would hear argument on the appropriate form of order if it could not be agreed.

2. Claims were construed according to what the skilled addressee would understand their language to mean in the context of the whole specification. Applying Kirin-Amgen Inc v Hoechst Marion Roussel [2005] UKHL 59, the patent's repeated emphasis on isolation established that claims 1 and 3 concerned the isolated (+) enantiomer and compositions containing it. They excluded the known racemate despite containing no express purity limitation. There was no established convention fixing unspecified purity at 95%, and the precise purity boundary required no determination. Disclosure of a racemate did not itself disclose each individual enantiomer, so the novelty attack failed.

3. The obviousness attacks also failed under the structured approach in Windsurfing International Inc. v Tabur Marine (Great Britain) Ltd [1985] RPC 59. Resolving and testing citalopram's enantiomers was an obvious goal, but the disclosed stereoselective method was inventive. The skilled person would have recognised material geometric differences between saturated and unsaturated systems and would not have expected the required ring closure to be favoured. The poor nucleophilicity of the tertiary alcohol and factors favouring racemisation reinforced that conclusion.

4. Preparative chiral HPLC was outside the skilled team's common general knowledge as a practical means of obtaining sufficient material for testing. Commercial availability of a suitable column did not establish common general knowledge. Column selection and scaling up involved substantial uncertainty. Reconstructing apparently simple steps with knowledge of the successful result introduced hindsight. Lundbeck's contemporary unsuccessful research supported the expert evidence of non-obviousness.

5. As subsidiary observations, unnecessary to the validity conclusions, escitalopram had significant unexpected therapeutic benefits. Those benefits were discovered later and were neither described nor foreshadowed in the specification, so they could not bolster inventiveness. Commercial success attributable to those discoveries likewise provided no assistance. Existing patent protection also prevented an inference that competitors had overlooked the benefits of resolution.

6. Applying Biogen v Medeva [1977] RPC 1, sufficiency required the monopoly to correspond to the technical contribution, even where the disclosed method could produce the claimed product. A genuinely new and non-obvious useful compound could justify a product claim supported by a disclosed preparation method. Here, however, resolving and testing the enantiomers was an obvious objective; testing and formulation were routine. The contribution lay in resolving the diol intermediate and converting its enantiomers while preserving stereochemistry. Claims 1 and 3 covered independent methods, including preparative chiral HPLC, which owed nothing to that teaching or any disclosed principle. Their breadth therefore exceeded Lundbeck's contribution.

These were three first-instance patent revocation claims. An earlier direction provided that the evidence in each claim would stand as evidence in the others, allowing the court to treat them together for practical purposes. No appellate stage was described.