By judgment of 15 August 2023 in PI proceedings brought by Bristol-Myers Squibb (“BMS”), the Court of Appeal of The Hague prohibited Sandoz, Stada and Teva from marketing generic apixaban. In this judgment, the Court of Appeal set aside the decision of the PI judge of 17 May 2023. The Court of Appeal applies a different interpretation of the Enlarged Board of Appeal’s ruling G2/21 than the Provisions Judge had done at first instance.
BMS is the proprietor of EP 1 427 415 B1 (“EP 415”), which expired on 16 September 2022, and Supplementary Protection Certificate (No 300500) based on this patent (the “SPC”). which is in force until 19 May 2026.
In 2022, the Provisions Judge of the District Court of The Hague ruled that there was a reasonable chance that EP 415 would be invalidated in proceedings on the merits, since the technical effect on which BMS had based the alleged inventive step, had not been rendered plausible in EP 415. The judge therefore refused the preliminary injunction, paving the way for the market entry of generic apixaban. When the G2/21 decision was handed down by EPO’s Enlarged Board of Appeal, BMS seized the opportunity to commence a new PI case to provoke a different ruling from the PI judge, on the interpretation of G2/21.
PI decision in first instance
In his judgment of 17 May 2023, the PI judge, judge Brinkman, assessed whether the patent application disclosed that apixaban had an improved effect (stronger inhibition of factor Xa) compared to the prior art. He quoted key paragraphs from G2/21 and referred at length to the UK appeal decision in which Lord Justice Arnold, taking into account G2/21, ruled that it was not plausible from the application that apixaban is a factor Xa inhibitor, let alone a factor Xa inhibitor of sufficient potency to be useful in therapy, holding the patent invalid.
The PI judge held that there is a reasonable chance that the patent will be annulled in Dutch proceedings on the merits. To this end, the PI judge considered that the average person skilled in the art would not infer from the application that apixaban has an improved effect. According to the PI judge, the technical teaching in the application was only that a useful compound should be found, not that this compound would also be more useful and thus be more effective. The application only described the general desire to find compounds with an improved effect, not that these compounds had actually been found. Therefore, the effect of improved inhibition of factor Xa claimed by BMS could not be taken into account when assessing inventive step. BMS’s infringement claim was thus rejected.
PI decision in appeal
BMS appealed the PI judgment. The Court of Appeal also quoted G2/21, but came to a different conclusion. The Court went along with BMS’s reasoning and considered that there was no reasonable chance that EP 415 would be declared invalid in Dutch proceedings on the merits. Applying G2/21, the Court of Appeal considered it significant that the relevant effect (improved factor-Xa inhibition) had been “expressly and specifically” mentioned in the original application as the primary objective of the patent. Whereas the PI judge considered this statement to be “a mere wish”, in the Court of Appeal’s view it is an important circumstance for assessing the technical teaching that the skilled person would derive from the original patent application.
The Court of Appeal defined the technical teaching of the patent as “that which is taught to the average person skilled in the art about how the technical problem can be solved by technical means”. In a succinctly motivated reasoning the Court expressed its view that the average skilled person would read in the application that the primary objective of the patent is to find an improved factor-Xa inhibitor, that the application discloses within which preferred range the Ki-value should fall, and that the application discloses a test by which the skilled person can determine the Ki-value of the compounds. The skilled person would also read in the application that apixaban had been synthesised and recrystallised on a much larger scale than the other compounds mentioned in the application and that apixaban was specifically claimed in claim 8 of the original application. All this, according to the Court of Appeal, makes it possible for the average skilled person with his common general knowledge on the priority date to infer from the application that apixaban is the most promising candidate. The skilled person would thus derive that the technical teaching implies that the claimed technical effect can be realised.
The Court of Appeal also referred to the parallel French and Norwegian decisions, in which the patent had been upheld. In these judgments, it was held that, based on the test method disclosed in the application and the greater quantity of apixaban synthesised, the average person skilled in the art would find it credible / plausible that the patentee saw apixaban as a promising compound. Furthermore, the Court of Appeal expressly distinguished its decision from the UK appeal decision of Lord Justice Arnold, since, according to the Court, the English decision is based on the stricter test for assessing sufficiency of disclosure (as determined in the Warner-Lambert v Generics case).
Since the Court of Appeal found that there is no reasonable chance that the patent be found invalid in proceedings on the merits, it granted the preliminary injunction claimed by BMS.
Concluding comments
It is worth noting that the Court explicitly states that its explanation of G 2/21 does not open the door to speculative patents, since it “requires that the technical effect is already encompassed by the technical teaching of the application and embodied in the same invention”. However, the Court’s reasoning does seem to leave room for patents on ‘armchair inventions’, since the Court seems to attribute an important role to the purpose of the invention stated in the patent application plus has been willing to accept a rather indirect indication that the claimed compound is ‘a promising candidate’ to achieve this purpose as sufficient for the skilled person to derive from the application the technical effect has been achieved (i.e. the invention has been made). In our view, the reasoning applied in the apixaban appeal case will require some further refinement in order to truly close the door for speculative patents.
Furthermore, it is interesting to note that in the French and Norwegian decisions, the courts persist in using the prevalent concept of credibility/plausibility, whereas the The Hague Court avoided using these terms and sticked to the G 2/21 phrasing. Even though the outcome was the same as in the Netherlands, the test applied French and Norwegian decisions appears to be based on a somewhat different interpretation of G2/21.
The Court’s rejection of applicability of the England and Wales Court of Appeal’s reasoning in the apixaban case is also rather tersely motivated. What is not mentioned in the Dutch appeal judgment, is that Lord Justice Arnold did not rely solely on the Warner-Lambert precedent, but also elaborately discussed G2/21. In the end Lord Justice Arnold’s assessment of inventive step focused on the question whether the technical effect was derivable from the application at the filing date for the skilled person with the common general knowledge in mind, which seems to be in line with G2/21. Arnold interprets applies the ‘derivability’ criterion from the G2/21 decision differently from the Dutch Court of Appeal, and in line with the Warner-Lambert decision. It is remarkable and arguably also undesirable courts in different EPC states both apply the G2/21 decision to the same set of facts but arrive at different outcomes.
All in all, it remains to be seen in which direction the interpretation of G2/21 in Dutch patent litigation will go in future cases. The test applied by the Court in the case at hand may well be subject of an appeal to the Dutch Supreme Court. Moreover, in future (pharmaceutical) patent cases, it is to be expected that G2/21 will be discussed further and the criteria applied by the Court of Appeal in the apixaban case may well be further refined.
Finally, it is interesting to note that two of the three judges sitting on the Court of Appeal panel in the apixaban case (R. Kalden and P.H. Blok) are also judges in the UPC Court of Appeal. This judgment perhaps gives some insight into the interpretation of G2/21 that the UPC may be expected to apply.
For further inquiries, please contact Jaap Bremer, Jens Brugman or Jarieke Timmerman.
