2024 Week 36
Just two weeks ago we reviewed UPC case law on preliminary injunctions (PIs), giving an overview of the big issues that sway the outcome of a PI application at the UPC. But looking at the output from the UPC in the last fortnight, there are already three new significant decisions on PIs.
So, although it is very soon to be writing again on the same topic, these decisions merit it.
Herbicide case – PI granted
In Syngenta v Sumi Agro, Syngenta applied for a PI against Sumi Agro based on EP 2152073 B1, claiming a herbicidal composition. The UPC Local Division Munich granted the PI, restraining Sumi Agro’s 2023 version of its product KAGURA.
The court went to some effort to interpret the scope of protection provided by the claim 1 (defining the composition). Of interest in particular was the fatty acid content – Syngenta provided testing data relevant to this. The court drew an adverse inference from the fact that Sumi Agro did not try to show their own testing data on the same product.
Sumi Agro argued that claim 1 lacked novelty or inventive step but the court disagreed, commenting that to reach a composition inside the scope of claim 1 from a single piece of prior art would require a selection of a rapeseed oil having the required free fatty acid content. For inventive step, the court in effect followed a straightforward problem-solution approach. More interestingly, Sumi Agro also argued that claim 1 is so compositionally broad compared with the examples that it must either be insufficient or obvious across the scope of the claim. The court dismissed these arguments as being obsolete when considering current EPO Board of Appeal case law.
The UPC Local Division Munich affirmed its previous guidance that there is a safe harbour of about 2 months for the patentee to bring a PI application on an inter partes basis. They acknowledged that this is different to the 1 month guidance provided by the UPC Local Division Düsseldorf, but stuck to their guns.
The court did not order any security for damages from the patentee, on the basis that it is a large UK-based company. It should be straightforward for the defendant to pursue a claim for damages in the event that the patent is ultimately found to be invalid or not infringed in the main action. The court also made no order for costs, this being deferred to the main proceedings.
Barcode scanner – PI granted
Hand Held Products v Scandit is another decision from UPC Local Division Munich, following the same approach as Syngenta v Sumi Agro (and issued on the same day). The court repeated their point about a 2 month safe harbour for urgency.
It is interesting to note that the patent EP 3866051 B1 is a unitary patent and so the scope of the PI is necessarily all of the UPC states at the time of registering of the unitary patent. No carve-outs necessary or possible [see the next case!].
The court decided that the alleged infringement was not a direct infringement. However, they decided that it was an indirect infringement – means related to an essential element of the invention that it intended to put the invention into effect in the UPC territory. On the facts of the case, this was enough to allow the court to order the PI.
The court had some interesting things to say about validity, which were identically expressed in the Syngenta v Sumi Agro. Although a PI application is a summary procedure, the court considered that it is not possible to take shortcuts on patent validity. So, to reduce the workload, the court strongly suggests that only the best 3 validity attacks from the defendant should be considered. Then each of these will be considered fully.
The court decided that the patent is, more likely than not, novel and inventive over the cited prior art. As in Syngenta v Sumi Agro, the court acknowledged the problem-solution approach, but with the proviso, acknowledged in other recent cases, that it is not necessary to agonise over the selection of the closest prior art – there can be several realistic starting points.
Unlike Syngenta v Sumi Agro, here the court did require security for damages from the patentee. This was on the basis of the defendant’s argument that the patentee is based in the US and pursuing a claim for damages could be difficult and expensive. Possibly simply because the patentee had not argued against this, the court ordered the patentee to pay a security deposit of Euro 0.5m to the court, which is a bit of a sting in the tail.
Omalizumab biosimilar antibodies – PI refused
In Novartis and Genentech v Celltrion, Novartis and Genentech are the proprietors of EP 3805248 B1, covering their product XOLAIR®. They filed two PI applications, one against Celltrion and another against a group of Celltrion subsidiaries in view of their biosimilar OMLYCLO®.
Celltrion argued that the UPC Local Division Düsseldorf did not have competence to hear the case. The court rejected this on the grounds that the defendants have a commercial relationship and the action relates to the same alleged infringement. The defendant in the first action is the producer and distributes products to defendants in the second action, and at least one of these is in Germany. This was enough for the local division to have jurisdiction over the case, including all of the defendants.
Novartis and Genentech were asking for a PI that excluded the Netherlands from its scope. Why was this? Because validity and infringement of the patent are in issue in Dutch national proceedings. Celltrion still argued that the lis pendens rules should apply so that the PI action is stayed pending the outcome of the Netherlands action. The court disagreed. In their view, the PI application was a different cause of action due to the carve-out of the Netherlands. Should there be a stay due to these being related actions? The court acknowledged that it has discretion to order a stay in this case but decided not to in view of the urgency of the PI application.
The court was at pains to interpret a particular feature of claim 1, relating to the amount of histidine in the claimed formulation. They were critical of the defendant’s position on this point, which simply referred to their expert’s opinion without really distilling the technical argument. In the end, the court was satisfied that the challenged embodiment would fall inside the scope of claim 1.
You can see that at this point, the patentees have a good number of their ducks lined up: the court has jurisdiction, the product falls inside the scope of protection, there is no stay of proceedings. So, how did the court reach a decision not to grant the PI? Validity? Nope – read on.
Celltrion had obtained EMA marketing authorisation for their biosimilar in May 2024 and soon after announced its plans to expand its market share. The UPC decided that the marketing authorisation and the marketing-type announcements were not quite enough to satisfy the requirement for a PI that infringement of the patent is “imminent”. For example, pricing had not been confirmed. The court stated:
In order for an infringement to be imminent, in the present case means that all prelaunch preparations must have been completed in such a way that an offer [to supply the pharmaceutical formulation to a customer] can be made at any time. Rather than looking at individual events in isolation, it is necessary to make an overall assessment of the activities.
Celltrion requested security for costs against the patentees, which was denied. However, the court ordered the patentees to pay substantial interim costs for the PI proceedings.
Because the court considered that there was no imminent infringement, there was no decision on the likely validity of the patent. This is slightly disappointing because in the event that the “imminent infringement” point is overturned on appeal, it will probably be necessary for the case to be sent back to the Local Division to consider validity before a PI is granted, further delaying proceedings.
