A recent, first-instance decision from an Opposition Division (OD) suggests that the European Patent Office (EPO) could tighten their assessment of medical use claims under Article 123(2) EPC (the requirement that a patent or application does not extend beyond the content of the application as originally filed). Here, the OD decided that a medical use claim is “directly and unambiguously” disclosed only if the application as filed provides a “certainty of success” that the therapeutic effect is actually attained for the specifically claimed embodiment.
On the face of it, this could represent a new consideration for the assessment of Article 123(2) EPC. It remains to be seen whether this decision will be upheld by an EPO Board of Appeal or if this approach will be endorsed by other ODs. Nevertheless, the decision provides some useful food for thought.
The take home message for now is, when drafting an application, ensure you disclose the attainment of the therapeutic effect as an unambiguous, unqualified achievement of the specifically claimed invention. Conversely, when attacking medical use claims, consider if there is reason to doubt that the therapeutic effect is definitely obtained for the claimed embodiment and if that can be used as an attack under Article 123(2) EPC.
For more detail, read on.
Background
Novartis’ European patent EP2959894 claims a specific dosage of fingolimod for the treatment of relapsing-remitting multiple sclerosis. Claim 1 reads:
1. A S1P receptor modulator for use in the treatment of relapsing-remitting multiple sclerosis, at a daily dosage of 0.5 mg p.o., wherein said S1P receptor modulator is [fingolimod] in free form or in a pharmaceutically acceptable salt form.
The patent covers Novartis’ blockbuster product Gilenya®, a disease modifying treatment for relapsing-remitting multiple sclerosis (RRMS) that received marketing approval in the US in 2010 and in Europe in 2011.
In addition to being the subject of several (unsuccessful) third party observations, the patent was opposed by 23 parties, all of whom withdrew their oppositions before the first-instance oral proceedings, which took place on 10 February 2025.
The patent does not include any clinical data. However, it does include pre-clinical data in the form of a rat experimental autoimmune encephalomyelitis (EAE) model, which is a known model to study MS, in additional to a clinical trial protocol (without results). During pre-grant proceedings, the proprietor had relied on these EAE data, in combination with post-filed human clinical trial data, to show that the therapeutic effect was sufficiently disclosed in the application as filed and that the requirements of Article 83 EPC were met. Following an initial refusal by the Examining Division, the Board of Appeal in T108/21 agreed with the proprietor, and a patent was granted. The pre-grant appeal decision did not consider Article 123(2) EPC.
Medical use claims in Europe
Medical use claims include the attainment of the therapeutic effect as a functional feature of the claim. In the above example, the successful treatment of RRMS using fingolimod at a daily oral dose of 0.5 mg is a feature of the claim.
For this reason, a piece of prior art that discloses the same drug, for the same purpose, at the same dose, will not be considered novelty destroying if it does not also disclose the actual attainment of the therapeutic effect in an enabling manner, for example in the form of clinical trial results. This is a well-established principle in Europe (covered in a previous article, here).
Amongst the many, many pages of written arguments in the proceedings was an argument from some of the opponents that claim 1 of the patent as granted extended beyond the content of the application as filed (and hence contravened Article 123(2) EPC) because there was no “enabling” or “direct and unambiguous” disclosure of the claimed therapeutic effect. The OD’s consideration of this “convincing” argument begins at section 3.3.19 of its decision.
In brief, the objection was based on the legal principles that (i) the concept of disclosure must be the same for the assessment of novelty, priority and added subject matter, and (ii) a piece of prior art must disclose a therapeutic effect in an enabling manner for it to be considered novelty destroying. As such, the opponents had argued that an enabling disclosure is a criterion for compliance with Article 123(2) EPC. Rather than this being a replacement for the “gold standard” of “direct and unambiguous disclosure”, they argued this was a criterion that must be met in addition to the gold standard.
The OD was not convinced that an enabling disclosure was relevant for Article 123(2) EPC. However, they confirmed:
…the concept of disclosure for the purposes of Articles 54 and 123(2) EPC has to be the same, insofar as in either case it is required that the claimed invention is directly and unambiguously disclosed”
In support of this conclusion (which on its face does not seem too controversial), the OD referred to T2842/18 and T1589/21, which both considered the situation in which an amendment (whose compliance with Article 123(2) EPC was being assessed) was based on indications given in the experimental section of the application as filed. Those decisions both dealt with determining if the therapeutic effect of the claim at issue was “directly and unambiguously” disclosed in the application as filed, and confirmed that this indeed is the correct test under Article 123(2) EPC.
In T2842/18, the proprietor had attempted to rely on a forward looking statement in the experimental section that the attainment of a specific therapeutic effect, which had been added to the claim during opposition proceedings, was “expected”. The Board concluded in that case that “the skilled person would derive from the passage in question that the above mentioned effect might or might not be achieved. The skilled person would not conclude that the effect was definitely achieved”. Therefore, the forward looking statement did not provide basis for the claimed specific therapeutic effect. Had the corresponding statement been in the original claims or in the general description part of the application as filed, in unqualified terms, the outcome of the proceedings may have been different.
Similarly, in T1589/21, the claim was to a single dose, and the language of the claim at issue was discussed only in the examples. The examples shed significant doubt about whether the therapeutic effect could be achieved with a single dose. Hence it was concluded that the therapeutic effect and corresponding language of the claim was not directly and unambiguously derivable. Again, had the effect been mentioned in the context of the claimed invention in the general description in unqualified terms, the conclusion under Article 123(2) EPC might have been different (although the proprietor may have had other issues under Article 83 EPC).
Returning to the disclosure of Novartis’ EP2959894, there is disclosure in the general part of the description of various medical uses in paragraphs [0026], [0027] and [0028], and in the Examples section a discussion of an upcoming clinical trial in paragraph [0033] (of the application as published). Claim 1 as originally filed is also a Swiss-form medical use claim that refers to “treating…multiple sclerosis” using a compound “of formulae I to IXb”, and fingolimod is disclosed elsewhere as a preferred compound of formula I.
However, paragraph [0028] states that compounds “may be useful” in the treatment of RRMS, and therefore qualifies the attainment of the therapeutic effect. Other disclosures of RRMS are limited; paragraph [0027] mentions RRMS, but this was considered background information as it discussed the different types of MS in general.
Furthermore, the OD, whilst simultaneously concluding that fingolimod was the preferred compound of the disclosure, and that the daily 0.5 mg dose required only 1 selection from a list (and so was a permissible selection), also concluded that the application as filed did not provide a direct and unambiguous disclosure that fingolimod at a daily dose of 0.5 mg necessarily provided a therapeutic effect in treating RRMS. This is despite paragraph [0026] and original claim 1 referring to various treatments of MS using “a compound of formulae I to IXb”. We note there does not appear to be a disclosure of RRMS as a preferred type of MS to be treated.
Furthermore, paragraph [0033] described the upcoming human clinical trial, without results. Noting the principle of applying the same standard to Article 54 EPC (novelty) as to Article 123(2) EPC, the OD said:
Since the concept of disclosure must be the same for the purposes of Articles 54 and 123 EPC (G1/03 rfd 2.2.2), the same conclusion [that the prior art did not disclose the attainment of the claim therapeutic effect], however, will be arrived at, when assessing the subject-matter claimed in the light of the protocol of the clinical trial disclosed in [0033] of the application as filed.”
We note the prospective clinical trial described in paragraph [0033] is disclosed as specifically investigating fingolimod in the treatment of RRMS at 3 daily oral dosages, including the claimed dose. Two of those daily doses were already known to work, but this was not enough to rescue the third, previously untested (and now claimed) dose as “certainly” obtaining the therapeutic effect and thus meeting the requirements of Article 123(2) EPC in the view of this OD (indeed, the proprietor had been obliged to argue a therapeutic effect for the claimed dose was not expected in view of the known effective doses, for inventive step purposes).
The animal model data in the application did not assist either, because the OD considered that:
the animal data give rise to an expectation, but not to a certainty of success regarding the outcome of the clinical investigation of the 0.5 mg p.o. daily dose”
and hence the “direct and unambiguous” disclosure requirement was not met in the view of the OD.
For all these reasons, the claim was considered to contravene Article 123(2) EPC.
Possible impact
Whilst the OD in this case has relied on previous Board of Appeal case law to reach its conclusion, this first-instance decision seems to go further than the previous decisions, because in this case there was disclosure of treatment in the general part of the description (and original claims), albeit not specifically in the context of the 0.5 mg p.o. daily dose in RRMS. Furthermore, unlike in T1589/21, there is no disclosure in the patent that the therapeutic effect is not actually achieved.
If this decision (which we must stress is a first-instance decision) is upheld in any subsequent appeal and followed in future, it could make medical use claims more difficult to obtain (or easier to attack) where the general description or original claims do not make a “direct and unambiguous” statement that the therapeutic effect is achieved using the claimed invention. A requirement of “certainty of success” is certainly a higher bar than one of “plausibility” and seems to imply an additional requirement over and above what the text on the page actually says, and what the skilled person might glean from the disclosure of the application as a whole.
Paradoxically, ambiguous statements in the Examples such as “the effect may be achieved” that could meet the supposedly more stringent enablement requirement of Article 83 EPC (perhaps in combination with post-filed data that back-up the qualified statements in the application) might not meet the “direct and unambiguous” requirement of Article 123(2) EPC, since the reverse is certainly true (unqualified statements of medical uses in the description that are “direct and unambiguous” within the meaning of Article 123(2) EPC are not de facto also made plausible just because the words have been written on the page).
Practice points
Whilst this (for now) remains a first instance decision, it does bring to mind some possible practice points.
If you are drafting an application that contains medical use claims, in particular those that relate to or arise from clinical trials, it may be important to include unqualified statements of the attainment of the therapeutic effect for each of the envisaged embodiments, especially if you do not have any positive results that map exactly on to the combination of features you are claiming (compound, dose, disease etc). Preferably, these should be included in the original claims (e.g. dependent claims), or at least the general part of the description. It could also be worth including more specific therapeutic effects in the original claims, in case you want to include a specific technical effect in your claim during prosecution (for example to overcome a problematic piece of prior art). We note that statements such as “In some embodiments, compound X may achieve Y technical effect” in the description could be seen as qualified statements (applying the same interpretation as the OD has done in this case), whereas a dependent claim that says an effect “is” achieved is less likely to be interpreted in such a manner. We note again that the disclosure of the clinical trial protocol in paragraph [0033] was not considered enough to provide a “direct and unambiguous” disclosure, and so if you are drafting applications based on clinical trial protocols, make sure you are including unambiguous statements of achievement of technical effects elsewhere in the application.
Conversely, if you are opposing a European patent that claims a medical use, it may be worth considering if there is any reason to cast doubt that the therapeutically effect is “directly and unambiguously” attained for the specifically claimed embodiment (or, to use the words of this OD in this decision, that there is an absence of “certainty” of the attainment of the therapeutic effect). Pointing to this first instance decision might have little weight, but the principles that are set out when applying T2842/18 and T1589/21 could be effective.
At the time of writing it remains to be seen if this decision will be appealed. There are no pending divisional applications in this family.
If you have any questions, please contact Tom Leonard, or your usual Kilburn & Strode advisor.