Generics UK (Mylan) v Novartis: EWHC 2403 Rivastigmine
The High Court has recently handed down judgment resulting in the revocation of the Novartis patent and SPC relating to rivastigmine (Exelon®). This is a further action where the Court was faced with a patent relating to the resolved enantiomer of a prior art racemic compound. Unlike the previous occasions, the evidence presented was supportive of the conclusion that it would have been obvious for the skilled person to have resolved the prior art racemic compound.
Rivastigmine is an acetylcholine esterase approved for the treatment of dementia/Alzheimer’s Disease. It is the (-) enantiomer of a compound disclosed in EP0193926A (Weinstock) and a subsequent paper by the same group. Novartis developed the racemic compound by resolution to identify the preferred enantiomer and took this compound through clinical development. The patent in suit claimed priority from an application filed in March 1987.
The parties acknowledged the previous cases decided by the UK courts on similar issues and the judge, Floyd LJ, correctly stated that the case must be decided on its own facts. The judgment therefore includes no analysis of those earlier cases.
Weinstock (the patent disclosure and paper) described a genus of phenyl carbamate compounds that, other than in one possible embodiment, all possessed a chiral dialkylaminoalkyl substituent. Of the compounds identified as being preferred, all possessed a chiral center, there was no data presented on the one achiral derivative.
There was a difference of opinion between the expert medicinal chemists on whether the skilled person would have directly resolved the most active compound (RA7) or first conducted a more detailed structure/activity analysis of the generic class disclosed involving the synthesis of further derivatives including the achiral compound.
The judge concluded that while conducting a more detailed analysis was always an option, it did not exclude the resolution of RA7 from the realms of being an obvious step to take. The Weinstock disclosures were taken as providing sufficient promise to render the further development of RA7 as obvious. It was not denied that the synthesis of further compounds could also have unearthed improved derivatives but this would not render the further development of RA7 as inventive.
The court then considered the rationale for resolution and whether there was an expectation of improved properties. Here the difference in the experts' approach was that GUK’s expert, taking a practical approach, described resolution in such circumstances as simply being “good science”, whereas Novartis’s expert preferred to undertake a theoretical study for potential benefits before proceeding with the resolution. In particular, as he could not have foreseen any potential for improvement through resolution, he would not have resolved.
The court established that there was reason to resolve, if not for improved potency, which in this case was acknowledged as not resulting in an improved therapeutic window, but for at least the 'practical benefits' of avoiding potential side effects arising on drug metabolism.
In reaching this conclusion the judge noted that Novartis had not argued that the resolution itself was potentially problematic and therefore a barrier to resolution.
This case provides an excellent example of how case-specific facts influence the overall decision. In contrast to the decisions in escitalopram (Generics v Lundbeck  UKHL 12, difficulty to resolve and relevant dates) and levofloxacin (Generics v Daiichi  EWCA Civ 646, acknowledged facts in a specific area of antibiotic chemical development), the judge found no reason to conclude that the resolution of RA7 would not have been undertaken with an expectation of success in the sense of improved properties.