Amgen’s cholesterol drug patent revoked in UPC first

Patent litigation experts at Pinsent Masons said the decision of the Munich Central Division of the UPC provides some guidance as to how the court may assess the validity of patents in future decisions. 

The UPC is a dedicated judicial system for litigating new unitary patents and existing European patents that have not been specifically opted-out of the UPC’s jurisdiction.

The Court’s decision on 16 July is the latest development in the global dispute between Amgen on the one hand and Sanofi and Regeneron on the other, concerning two products – Amgen’s evolocumab, sold under the brand name Repatha, and Sanofi’s and Regeneron’s alirocumab, sold under the brand name Praluent.

The parties agreed to have issues relating to the patent’s validity arising from both actions assessed and decided by the Munich Central Division, with the infringement issues remaining with the Munich Local Division.

Amgen’s patent is a ‘medical use’ patent protecting certain anti-PCSK9 antibodies for the treatment and prevention of various conditions caused by elevated cholesterol levels in the blood. Sanofi and Regeneron claimed that Amgen’s patent was invalid on the basis that there had been no inventive step taken with the apparent innovation. Their argument was based on a disclosure in a scientific publication, Lagace, which they claimed constituted ‘prior art’ rendering the patent as obvious, which expressly suggested that antibodies developed to block PCSK9:LDLR interactions could be explored for the treatment of high blood cholesterol levels.

Amgen defended the patent’s inventive step by arguing that Lagace provided no motivation to the skilled person to develop anti-PCSK9 antibodies that came within the scope of the patent’s claims. In its decision, the Court found in favour of Sanofi and Regeneron and declared the patent to be obvious, and therefore invalid and subject to revocation.

London-based patent litigator Krishna Kakkaiyadi of Pinsent Masons said that the detailed decision provided insightful clues as to how the UPC may approach patent validity assessments in future decisions.

“Such well-structured and reasoned decisions in the early cases will no doubt help drive forward the case law of the UPC in a meaningful manner,” Kakkaiyadi said.

Sarah Taylor of Pinsent Masons noted that some divergence is beginning to emerge over how different divisions of the UPC approach legal issues before them. She highlighted how, in their first decisions on the merits of issues argued at trial, the Dusseldorf Local Division and Paris Local Division appear to have adopted tests which are akin to the problem-solution approach that has been operated by the European Patent Office in the way it approaches patent disputes before it, and said that this differs from the more holistic approach that the Munich Central Division took in this case. Nevertheless, she added that the two approaches were not entirely inconsistent. 

The Munich Central Division said that the first step is to determine a “realistic starting point” in the prior art that would have been of interest to the skilled person considering a similar “underlying problem” as that which the patent claimed to solve. It said this ‘realistic starting point’ may be one of several, and does not need to be the most promising starting point. This approach differs from the problem-solution approach, which requires the closest piece of prior art to be identified. 

The court said it then needs to ask whether it would be obvious for the skilled person to arrive at the claimed solution in the patent from the ‘realistic starting point’. If the skilled person would be motivated to consider and implement the claimed solution in the patent as a next step, then the patent is obvious. While a technical effect achieved by the claimed subject matter compared with the prior art may be an indication of inventive step, a feature selected in an arbitrary way out of several possibilities will not be, it said, adding that the question of whether there has been an inventive step should not be answered “by searching retrospectively” – hindsight, it said, should be avoided.

“While we have seen, to some extent, UPC divisions take different approaches to national courts and the EPO, there is no prescribed approach at this time as to how inventive step should be assessed,” said Taylor. “The coming months may bring more consistency as more decisions on the merits are issued, and as the UPC Court of Appeal is called upon in appeals from these first instance decisions to provide guidance on the correct approach to inventive step assessment.”

Kakkaiyadi added that there are some similarities between the Munich Central Division’s approach with the objective and multifactorial assessments typically carried out by the English courts. He cited some of the references to the written expert evidence made in the court’s decision as an example. 

Ultimately, the court found that the disclosure in Lagace was a realistic starting point. It considered that the skilled person, working on the problem of treating or preventing ailments caused by high cholesterol, would have pursued the route of developing anti-PCSK9 antibodies that blocked PCSK9:LDLR interactions. The court was convinced that at the relevant date, PCSK9 was a well-established target based on the disclosure in Lagace and the evidence that a considerable number of pharmaceutical companies were already pursuing various options to target and inhibit PCSK9.

Amgen’s attempts to discredit the teaching in Lagace or cast them in a more ambiguous light also met with failure. The court noted that the skilled person would be inclined to accept research results and reasonable conclusions in a prior art document at face value unless there were any apparent errors or omissions.

In addition, the court said that the absence of a reasonable expectation of success does not follow from the mere fact that there are other ways of solving the problem. The court characterised Amgen’s arguments as “uncertainties” that would still not have prevented the skilled person from taking the obvious next step. Therefore, none of Amgen’s arguments convinced the court to hold that the skilled person would have no ‘reasonable expectation of success’ to find anti-PCSK9 antibodies as claimed by the patent.

“The court helpfully clarified that it was not going to decide in this case whether or not a ‘reasonable expectation of success’ is required as a necessary precondition to finding that a claimed subject matter lacks inventive step, in circumstances where there is a motivation towards the claimed subject matter and the next steps to be taken would involve no more than routine experimentation,” said Kakkaiyadi. “This potentially draws up the frontlines for another case in the future around whether or not certain types of factual circumstances require assessments of expectations of success.”

Amgen’s argument that the skilled person’s next steps would not have generated anti-PCSK9 antibodies that came within the scope of the patented claims also failed. The court emphasised that the test for obviousness was not whether the skilled person would inevitably – i.e. each and every time they performed their obvious next steps – come within the scope of the claim, but only whether the skilled person would arrive at a result covered by the claim without inventive contribution. Earlier in its decision, the court needed to construe the scope of the independent claim. Unhelpfully for Amgen, this resulted in a claim construction that was broad. This then meant that it was easier for the court to find that routine experimentation would result in antibodies that fell within the scope of this broad claim.

According to Kakkaiyadi and Taylor, there is a growing body of UPC case law – including this decision – which approves the approach to claim construction set out by the UPC Court of Appeal in the 10x Genomics v NanoString decision, which is based on Article 69 of the European Patent Convention (EPC) and the Protocol of Interpretation.

In this regard, the Munich Central Division specifically noted that the patent specification may represent the patent’s own “lexicon”, and where terms used in the patent differ from general usage, “it may therefore be that ultimately the meaning of the terms resulting from the patent specification is authoritative.” The court noted the Court of Appeal’s confirmation that these principles also apply to validity assessments.

However, Kakkaiyadi and Taylor said this relatively settled part of UPC law could be influenced by recent legal developments relating to claim construction and the interpretation of Article 69 EPC at the European Patent Office (EPO).They highlighted that a referral (Case G1/24) requesting clarification on how and when the description and the drawings in a patent can aid the interpretation of its claims was made to the Enlarged Board of Appeal (EBA) at the EPO earlier this year. They said the guidance the EBA provides in the case could potentially influence the UPC’s own decision-making.

Amgen now has a two-month period from the date of service of the decision to lodge an appeal at the UPC Court of Appeal in Luxembourg. Kakkaiyadi and Taylor said recent developments in the infringement action suggest that Amgen will do so. On 29 July, the Munich Local Division, with the agreement of the parties, paused the infringement action pending the outcome of an appeal in the revocation action.

Kakkaiyadi and Taylor said the nature and extent of the Court of Appeal’s appraisal of first instance decisions on their merits will be closely watched by practitioners. They said that at a recent meeting organised by the International Association for the Protection of Intellectual Property (AIPPI) in London, UPC practitioners opined that appeals were likely in most UPC cases. It was also suggested that such appeals could result in findings being overturned on points such as claim construction and inventive step in a good number of cases.

“If a sizeable proportion of UPC cases end up getting overturned on appeal, then commercial certainty for businesses may be achieved only after the conclusion of the Court of Appeal proceedings,” Kakkaiyadi said. “This would mean that litigation strategies and timelines will need to account for not one, but two sets of proceedings at the UPC.”