Novartis AG, LTS Lohmann Therapie-Systeme AG, Novartis Pharmaceuticals UK Limited v Focus Pharmaceuticals Ltd, Actavis Group PTC EHF, Actavis UK Limited, Teva UK Ltd [2015] EWHC 1068 (Pat), Patents Court, England and Wales, 27 April 2015
Journal of Intellectual Property Law & Practice (2015) doi: 10.1093/jiplp/jpv119, first published online: July 9, 2015
The Patents Court has held that Novartis's patent, covering its EXELON rivastigmine transdermal patch to treat Alzheimer's disease, is invalid. The broad claim construction that Novartis proposed in order to establish infringement meant that the patent included added matter over the application as filed and was obvious over a prior US patent.
Legal context
Patentees bringing infringement actions in England and Wales frequently face difficulties putting forward claim constructions broad enough to cover the allegedly infringing activities, but not so broad that the claims are invalid on grounds including added matter, obviousness, and insufficiency. Where patentees do not get this right, or cannot get this right because of the wording of their application as filed, they risk the alleged infringer successfully running a ‘squeeze argument’, as the generics did in this case, essentially saying that ‘if we infringe on that construction, the patent must be invalid’.
Another challenge that Novartis faced here, common to patentees litigating European patents in England and Wales, is the possibility of inconsistent findings when proceedings are heard in different jurisdictions across Europe. Particularly when deciding on infringement, courts in different jurisdictions can take markedly different approaches.
Facts
Novartis's patent EP (UK) No 2 292 219, titled ‘Transdermal therapeutic system for the administration of rivastigmine’, is a divisional patent derived from international patent application No WO 2007/064407. Before the introduction of transdermal therapeutic systems, which most commonly come in the form of patches, rivastigmine was administered orally, which led to side effects such as nausea, vomiting and diarrhoea. Patients would be prescribed a starting dose of oral rivastigmine, which was sub-therapeutic, and then increased to a therapeutically effective dose as they got used to the side effects. According to the patent, the use of a transdermal patch meant that patients could take a much higher starting dose with greater tolerability, so that their treatment was simplified.
Rivastigmine |
Analysis
Skilled team and common general knowledge
Before Arnold J, the parties agreed that the patent was addressed to a skilled team interested in developing a new formulation for rivastigmine. The team would consist of a formulator skilled in transdermal administration, and a clinician and/or neuroscientist in the field of dementia. There was, however, substantial dispute as to the skilled clinician's common general knowledge of how tolerability of rivastigmine could be increased. Arnold J agreed with the generic parties that it was generally accepted at the priority date that rivastigmine should be administered with food, and that it was a reasonable hypothesis that this increased tolerability because it reduced peak plasma concentration of rivastigmine and increased the time at which peak plasma concentration was reached. This was widely stated in reviewed papers put into evidence by the generics.
Novartis's broad claim construction accepted
Arnold J accepted Novartis's construction that the single claim of the patent involved three components:
Because of the wording ‘the starting dose is that of’, the claimed patch could have any structure or composition provided it could deliver the same starting dose as the reference patch. It was not correct to restrict the claim to a rivastigmine patch with the precise structure and composition specified in the final integers of the claim. These features merely described the reference patch.
- rivastigmine for use in treating dementia or Alzheimer's;
- rivastigmine administered via a patch; and
- a ‘starting dose’ of rivastigmine that ‘is that of’ a reference patch with a structure and composition specified in three final integers of the claim.
Invalidity for added matter
The generics' primary objection to validity was that the claim involved added matter in that the patent presented information about the invention which was not directly and unambiguously derivable from the application on which it was based. Although the application disclosed a patch with the same structure and composition as the reference patch in the patent claim, it did not suggest that this should be used as a starting dose. In contrast, the patent explicitly disclosed a rivastigmine patch with a starting dose the same as that of the reference patch. Arnold J accepted that, if the claim was construed as Novartis contended, the amendments to the application made during prosecution amounted to disclosure of a new invention over the application as filed so that the patent was invalid for added matter.
Invalidity for obviousness
In their obviousness attack, the generics relied only on United States patent No 6 335 031, co-owned by Novartis. There was no dispute that the US patent disclosed administration of rivastigmine by a transdermal patch with a structure and composition the same as that of the reference patch in the patent in suit. The essential difference was that the US patent did not disclose a starting dose of that released by a reference patch. The key area of disagreement between the parties was the starting dose the skilled team would select when administering rivastigmine using patches of the kind disclosed in the US patent.
Although it did not disclose a starting dose, the US patent stated that the size of the patch could be determined using tests which involved observing blood levels of rivastigmine ‘after oral administration of a therapeutically effective dose of the compound’. Novartis submitted that this would lead the skilled team to administer as a starting dose a patch of a dose equivalent to that of the starting dose in the oral rivastigmine regimen, that is the sub-therapeutic dose. This would mean the starting dose would be half that released by the reference patch disclosed in the patent. The generics on the other hand said that this statement in the US patent would lead the skilled team to do exactly what the statement said, and administer the lowest therapeutic oral dose of rivastigmine as a starting dose. This would mean the starting dose would be that released by the reference patch in the patent in suit. Arnold J accepted the generics' position and found the patent obvious in light of the US patent.
Patent not insufficient
On the basis of the expert evidence, Arnold J did not agree with the generics that the claim in Novartis's patent was insufficient. In particular:
The claim was not overly broad. Even though the benefit of tolerability was not obtained for any patch within the claim (with the starting dose of the reference patch), a skilled person would know to use a formulation that would have that benefit.The claim was not ambiguous. Even though the specification did not provide a method for determining whether any other patch delivers ‘the same’ starting dose as the reference patch, the skilled team would be able to select an appropriate statistical test.
Generics would have infringed
Arnold J found that, had the patent been valid, the generics would have infringed. There was no dispute that the generics' patches were identical to Novartis's Exelon patches (save for the branding), were made by the same manufacturer, and would be used for the claimed indication to administer rivastigmine with the claimed starting dose.
Practical significance
Arnold J's decision highlights a number of considerations for patentees bringing infringement actions in England and Wales today. First, patentees need to take care that where possible the claim construction they propose to cover the alleged infringement is not so broad as to lead to a finding that the patent is invalid. Patent applicants should also consider the breadth of the claims from an infringement and a validity point of view at the time the patent application is drafted.
Secondly, patentees involved in pan-European litigation seeking to restrain generic entry to the market should have a strategic co-ordination plan in place from the outset, given the possibility of strikingly diverse findings in different jurisdictions. Despite the divergent views on infringement when this case was heard across Europe, the case indicates that courts in different jurisdictions may be more likely to make the same findings as to validity of a respective national designation of a European patent.
Arnold J has granted Novartis permission to appeal, stating that the appeal from the finding of added matter has a real prospect of success, but he is only narrowly persuaded that the same is true of the appeal from the finding obviousness ([2015] EWHC 1553). It remains to be seen whether the Court of Appeal will uphold Arnold J's findings. In the meantime, Novartis's European patent remains under opposition at the EPO by 13 opponents and the Opposition Division has already issued a provisional opinion that the patent is invalid for added matter and insufficiency. Oral proceedings are scheduled for 15–17 December 2015.
No comments:
Post a Comment