EP 3 167 888 B1 relates to the treatment of paroxysmal nocturnal hemoglobinuria (PNH) by an inhibitor of complement. The product is known as eculizumab.
It stems from a third generation divisional of EP 2 359 834 A1 and EP 2 001 490 A1 filed as PCT/US2007/006606 and published under WO 2007/106585
Brief outline of the procedure before the EPO
Proprietor is Alexion Pharmaceuticals, Inc.
EP 3 167 888 A1 has been first refused for added matter. With decision T 1515/20, the board decided that the a patent could be granted on the basis of AR5. In the interpretation of the TBA, feature SEQ ID NO:4 includes 22 extra amino acids.
The TBA examined the claimed subject matter on the basis of this interpretation and concluded it is novel, involves an inventive step and is sufficiently disclosed.
The granted claims read as follows
- An antibody that binds C5 comprising a heavy chain consisting of SEQ ID NO: 2 and a light chain consisting of SEQ ID NO: 4.
- A pharmaceutical composition comprising the antibody of claim 1.
The patent has been opposed by three opponents Samsung Bioepis NL B.V. (Oppo 1), Amgen Technology (Ireland) Unlimited Company (Oppo 2) and Salute HC Ilaç Pazarlama A.S. (Oppo 3)
In reply to the opponents, the proprietor argued that various features of SEQ ID NO:4 would immediately alert the skilled person to the presence of a signal peptide, which would be excluded when making technical sense of the claims, since they would interpret them as being directed to an antibody containing mature light and heavy chains.
The proprietor based its argumentation on the result of the cross-examination carried out during infringement litigation in the UK.
The opposition procedure is still going on at the EPO.
Brief outline of the procedure before the UPC
Procedure before the UPC-LD Hamburg
The proprietor requested a PI against Samsung Bioepis NL B.V. and companies of the Amgen group. The PI was rejected by UPC-LD Hamburg with decision UPC_CFI_124/2024.
The court held that a decision on provisional measures cannot be based solely on the court’s view of the validity of the patent in suit, but also on the likelihood that the OD of the EPO will revoke the patent.
For the UPC-CFI, claim 2 protects a pharmaceutical composition comprising an antibody with a light chain of SEQ ID No. 4 without the first 22 amino acids, which could not be correct, but the correction of would not be allowable under R 139 as it would not be apparent that an error occurred.
The UPC-CFI reached its conclusion on the basis of the evidence of common general knowledge of the skilled person which has been submitted in these proceedings. Decision T 1515/20 was based on different facts than the interpretation of the court, so that no inconsistency can be established.
It was the opinion of the UPC-CFI that it is reasonably likely the EPO will revoke the patent due to lack of sufficient disclosure under Art. 83 EPC.
Accordingly, on the TBA’s claim construction and the applicant’s own submissions and evidence in these proceedings, there is a substantial probability that granted claim 2 will be regarded as non-patentable by the EPO.
Brief outline of the procedure before the UPC-CoA
The rejection of the PI was confirmed by the UPC CoA, with decision UPC_CoA_405/2024 (APL_40553/2024).
The UPC-CoA noted that the SEQ ID NO: 4 sequence shown in par. [0134] of the patent at issue is not identical to the sequence of the light chain disclosed in a corrected CAS database entry for eculizumab in 2009. SEQ ID NO: 4 has 22 extra amino acids at the beginning of the sequence (the N-terminus).
The UPC-CoA considered that the existence of an error and the way to correct it are not sufficiently certain to the average skilled person. For the UPC-CoA, it is not sufficiently certain for the average skilled person that the alleged error must be corrected by deleting exactly 22 amino acids from the sequence of SEQ ID NO: 4.
For the UPC -CoA, claim 2 must be interpreted as meaning that the claimed pharmaceutical composition comprises an antibody comprising a light chain consisting of SEQ ID NO: 4 including the first 22 amino acids. Based on this claim interpretation, it is more likely than not that the subject matter of claim 2 is insufficiently disclosed within the meaning of Art. 83 EPC.
Decision on costs
In the meantime the UPC-LD Hamburg took a decision the costs UPC_CFI_40/2025 (ACT_3177/2025), whereby the value of the litigation was set a 100 Mio € and the ceiling for recoverable costs was set at 2 Mio €.
Application for rehearing
Alexion filed an application for rehearing under R. 245 RoP.
The UPC-CoA held that the application for rehearing is not allowable, because the requirements set forth in Art. 81(1) UPCA and R. 247 RoP have not been complied with. See UPC_CoA_402/2024, APL_40470/2024, App_8340/2025
For the UPC CoA, in relation to the order under review, the existence of a fundamental procedural defect has not been established.
Comments
Correction of an error
The present case shows how it is important to carefully check any gene sequence before filing. The mistake was committed at filing, and in T 1515/20, the board did not accept a correction under R 139.
Position of the UPC on the validity
It is not for me to decide what is right or wrong in the UPC’s or EPO’s decisions.
It is however surprising how the UPC has dismissed the reasons given in T 1515/20 and the justifications given in the decisions of the UPC. How can the UPC be certain that the patent will be revoked for lack of sufficiency?
In spite of what has been said at the UPC, in T 1515/20, Reasons 36, the board concluded that the claimed antibody is sufficiently disclosed in the patent.
Lack of sufficiency or lack of IS-Application of G 1/24
One interesting aspect will be in the decision of the OD to come, and the following BA decision, on how claim 2 will be interpreted.
One important aspect should play a role: will the OD and the TBA find that when interpreting claims according to G 1/24, a limitation can be read into the claim? The limitation relates to ignoring the 22 first amino acids in SEQ ID NO:4. In view of recent case law following G 1/24, the odds are against the proprietor.
This appears a priori more important than to decide upfront that the patent ought to be revoked under Art 83 EPC. When looking at the claims at stake, there is no effect claimed and the right objection should not be lack of sufficiency, but lack of IS, cf. G 1/03, Reasons 2.5.2, last § and T 2001/12, Catchword.
Should the case ever come to the UPC and the validity be at stake, it would be interesting to see whether the UPC will read in the description the limitation suggested by the proprietor.
Application for rehearing
An application for rehearing under Art 81UPCA + R 245 RoP is similar to a Petition for review under Art 112a EPC, whereby the Court of Appeal may
- waive the payment of the fee, rather than reimburse it
- give the application for rehearing a suspensive effect
There is as well an obligation to raise objections, cf. R 248 RoP, similar to R 106EPC.
In principle, an application for rehearing shall not have suspensive effect unless the Court of Appeal decides otherwise, cf. R 252 RoP. The CoA can waive the payment of the fee. There are thus differences with the procedure before the UPC.
Costs
The present case shows that the costs at the UPC are horrendous and for this reason alone is not a forum to litigate for SMEs. The risk of having to face heavy costs, will deter SMEs to litigate before the UPC.
We were however told that the UPC would be wonderful for SMEs….
Parallel case at the UPC
In a parallel case, in which Alexion requested a PI against Amgen, one of the opponents at the EPO, the CoA UPC also confirmed the rejection of the PI, and of the application for rehearing, cf. UPC_CoA_405/2024, APL_40553/2024, App_8339/2025.
Comments
2 replies on “EP 3 167 888 B1-Eculizumab-Procedures before the UPC and the UPC- Application for rehearing at the UPC”
Dear Thomas,
I, too, am surprised by the BoA’s comment on Art. 83, especially considering that the only functional feature of both claim 1 and claim 2 is “binds to C5”. Both claims do not use purpose-limited language, so efficacy in treating PNH is no matter under Art. 83 EPC.
In T 1515/20 the proprietor (in my mind: convincingly) demonstrated that the 22 extra AS attached to the N-terminus are sufficiently far removed from the binding site as to not fundamentally effect the binding sites (proven) ability to bind C5.
However, I, too, am in no position to see into the UPC judge’s minds as to what line of argumentation made them consider the claim as interpreted in light of consulting the description (cf. your comment on G 1/24, i.e. including the 22 extra AS at the N-terminus) to no longer believably bind to C5, instead of denying the claim an effect in the treatment of PNH under Art. 56 EPC. Time will tell, i guess.
One more comment on G 1/24: I disagree with your analysis insofar, as there would be a “limitation” read into the claim. The claim the proprietor intended (directed to eculizumab) and the claims as granted pertain to two different chemical entities. As such this is not a “limitation” in the strictest sense of the word, but rather an interpretation as a different compound as verbatim claimed. As such, I cannot help but to be reminded of the German Federal Court decision BGH – Rotorelemente (X ZR 43/13) in which an “obvious” error of the claim was “inherently corrected by the reader” when consulting the description and especially the figures. However, in my personal opinion, even when considering BGH – Rotorelemente, the proprietors chances are bad: In BGH – Rotorelemente the claim as granted was nigh non-sensical, while in the present case the claim is directed to an existing and sensible chemical compound – albeit just not the compound the proprietor wanted to protect. An interpretation of the claim as granted as reading on eculizumab (i.e. without the 22 extra AS) would – in my personal opinion – be completely contrary to the principle of legal certainty for the public. Alexion wrote “A” but meant “B”. If it were really interpreted that “A” means “B”, what happens when a competitor infringes on “A”? Would the claim verbatim reading on “A” no longer cover “A”? That appears nonsensical to me.
As question: If the claims as granted will be interpreted as “A” (i.e. with the 22 AS), shouldnt the ED have insisted on the adaption of the description to exclude all language directed to protecting eculizumab and labelled all eculizumab examples as “comparative experiments”? A demand like this would have very quickly made this unfortunate mistake aware to the applicant at a point in time where this might still have been salvagable.
The present case really is model example how even a easily understandable careless mistake can lead to mind-boggling damages in the field of pharmaceutical chemistry. I do not want to trade places with the poor sod that has to take responsibility for this legal saga.
(And regarding the cost-issue: While I see your point and do not completely disagree, I would be reserved to point to a case between massive pharmaceutical companies regarding a blockbuster medicament to reach the conclusion that litigation before the UPC is prohibitively expensive for SMEs.)
@ L.T.,
Thanks for your extensive comments.
Art 83 was not a problem raised by the BoA, but by the CFI and the CoA UPC. I am also at a loss why Art 83 should be a problem as binding to C5 is a neutral formulation.
I appreciated your say: Alexion wrote “A” but meant “B”. I could not have said it better.
In my comment, I spoke of a limitation, being well aware it is not a limitation as such, for want of a better formulation. The patent might still be valid as it covers A, and this is why the board agreed on AR5 as being patentable, but it should have been directed to B. I saw the limitation in the fact that the 22AS should actually be ignored. However, in view of the situation, switching from A to B as correction of an error under R 139 was doomed to fail.
As far as infringement is concerned, it will be interesting to see what could be decided, should the patent be maintained by the EPO. As the proprietor convincingly argued that the 22 extra AS, would not hinder the binding on C5, which is at the core of the treatment of PNH, one possibility would be to envisage infringement by equivalence. Even with the 22 extra AS, the treatment appears still possible. As you said, time will tell.
The BGH is well known as a “patent repair shop”. Just look at its decisions on added matter. For this reason, I take lots of its decisions with a pinch of salt, especially when it comes to the patent being its own dictionary. Whether accepting that A can also treating PNH as B does will be a key issue. In the pemetrexed disodium case, at grant, the applicant had no proof that other salts would work as well, and yet it got infringement by equivalence before the UKSC. There the description should have been amended, but the applicant was stroppy and the examiner gave up.
In the present case, adapting the description would, IMHO, not have brought more than making the applicant aware early of its mishap. It would however have required the applicant to come up with experiments showing that the 22 extra AS do not hinder the treatment of PNH.
It will be interesting to watch the development of the present case before the EPO.
I agree with you that big pharmaceutical companies have pockets deep enough to absorb such costs, should they loose before the UP.
I had more in mind the Vusion/Hanshow case, where Vusion had to fork out ¼ of a million € in cists to Hanshow, after having lost twice before the UPC and being refused a PI. Adding Vusion’s own legal costs, the amount should not be too far away from ½ Million €. As it is the UPC which decides the value of the litigation, it might not chose a low value for it.
I agree with you that I would not want to be in the shoes of the person who either gave the incorrect sequence to the patent attorney, or included the incorrect sequence in the description.