This browser is not actively supported anymore. For the best passle experience, we strongly recommend you upgrade your browser.
| 8 minutes read

The application of G2/21 by the Court of Appeal

This article is part of our Biotech Review of the Year - Issue 11 publication

On 4 May 2023, the Court of Appeal, with the leading judgment given by Lord Justice Arnold, handed down its decision in the appeal of the High Court’s earlier decision in Sandoz & Teva v Bristol-Myers Squibb1, dismissing BMS’ appeal in its entirety and finding the compound patent protecting apixaban invalid.2


Apixaban is a small molecule compound that inhibits the clotting enzyme factor Xa. Although it is not a biotech case, it is an important decision relevant to the disclosure required in patents which is relevant to the biotechnology field. It is also of wider interest given the approach taken by the Court of Appeal on the question of plausibility and the treatment of the decision of the EPO Enlarged Board of Appeal in case G2/21. This decision (and that of the Supreme Court next year in FibroGen v Akebia) could have a significant impact on all life sciences cases.

Apixaban was developed by Bristol-Myers Squibb and Pfizer, who market apixaban as Eliquis®, and is approved for the prevention of stroke and systemic embolism in adult patients with risk factors, and for the treatment and prevention of deep vein thrombosis and pulmonary embolism. Eliquis came to market in 2012, part of a new class of anti-coagulants termed direct acting anticoagulants (DOACs), in contrast to the existing treatments of warfarin and heparin which had indirect mechanisms of action. In 2020, when Sandoz first started proceedings, Eliquis was the fourth-best selling medicinal product globally with sales in excess of US$9bn.

In the UK, apixaban was protected by EP (UK) 1 427 415, which expired on 16 September 2022. Claim 1 of the patent was to apixaban, or a pharmaceutically acceptable salt thereof. SPC/GB11/042 extended the patent protection for apixaban to 19 May 2026.

Sandoz and Teva started revocation proceedings against the patent and SPC, on the basis that the patent was obvious and/or insufficient for lack of plausibility. Plausibility is not a standalone ground for invalidity, but is addressed by the English courts under inventive step and/or sufficiency, with there being little to no distinction between the two. No separate attack was made against the SPC. BMS counterclaimed for infringement, which was admitted by Sandoz and Teva.

High Court decision

The High Court’s decision was given by Mr Justice Meade. He conducted a thorough review of the case law from the EPO and the UK courts and reached the following key conclusions on the law. First, in light of T 939/92 Agrevo/Triazoles, if a patent discloses new compounds for which the purported utility is not credible, then the only technical contribution is the provision of new compounds, and that in itself cannot be inventive. This is as there is no invention in writing down the structure of a new compound; the invention lies in its effect. Second, the decision of the Supreme Court on the assessment of plausibility in Warner-Lambert v Generics (UK)applied to compound claims in patents just as much as it applied to second medical use claims. Third, again drawing on Warner-Lambert, a patent cannot be plausible purely because it encourages the reader to perform tests for themselves. Finally, it was common ground that the case should be decided by reference to the patent application as originally filed, as if the granted patent was plausible, but the application was not, then the granted patent teaches something new and there must be added matter.

Meade J noted that the patent application in question discloses an enormous number of compounds under a Markush formula. It also identifies and claims 198 individual compounds, and provides 110 examples with synthesis and chemical characterising information (as well as numerous other examples without such information). Apixaban is one of these compounds claimed, with an example disclosing its synthesis and chemical characterisation. The “Utility” section of the patent application describes in vitro and in vivo assays to test compounds, and concludes by stating for the in vitro assay: “a number of compounds of the present invention were found to exhibit Ki’s [a measure of activity against an enzyme] of ≤ 10 µM against factor Xa and another blood factor, thrombin.”

The following findings would be important at the appeal stage. First, Meade J found that the statement above from the “Utility” section could not be applied to any specific compound, including apixaban. Therefore, there was no explicit disclosure that apixaban had factor Xa activity. BMS argued that the fact that apixaban was made in gram quantities, sufficient for at least early stage in vivo animal testing, pointed to it being a compound of interest. However, the Judge held that even if the skilled person noted this, the inference that apixaban might be promising was weaker than a bare assertion, and following Warner-Lambert, a bare assertion does not render a patent plausible. Second, even though the patent application discloses easy tests and, if tested, the skilled person would find that apixaban is extremely effective in those tests, this amounted to no more than the encouragement plus test approach that was rejected in Warner-Lambert. Accordingly, Meade J found that it was not plausible that apixaban could bind to factor Xa, and so the patent was held invalid. 


On 23 March 2023, shortly before the hearing at the Court of Appeal, the Enlarged Board of Appeal (EBA) handed down its decision in G2/21. The Technical Board of Appeal (TBA) in T 116/18 Sumitomo/Insecticide compositions noted the two conflicting strands of case law in the previous TBA cases: in determining whether to allow the patentee to rely on post-filed evidence to support a technical effect, some cases appeared to apply ab initio plausibility, where a skilled person needs to consider whether the technical effect in the patent is plausible. Others applied ab initio implausibility, where a patent is plausible if the skilled person would not consider the technical effect implausible. It therefore referred questions of law to the EBA. The EBA ruled that: “A patent applicant or proprietor may rely upon a technical effect for inventive step if the skilled person, having the common general knowledge in mind, and based on the application as originally filed, would derive said effect as being encompassed by the technical teaching and embodied by the same originally disclosed invention.

This appeal presented the Court of Appeal with the first opportunity to consider G2/21. In his review of the applicable law, Arnold LJ explained that the EBA found the two lines of case law (ab initio plausibility and implausibility) as reconcilable with the core question as to whether the technical effect relied upon was derivable from the application. He also noted that the EBA had reviewed the case law from EPC contracting states, including the UK Supreme Court in Warner-Lambert, and concluded that the national courts had approached the issue of plausibility in the same manner as the TBA, and therefore in line with the core question. However, Arnold LJ commented that he did not agree with the EBA that the two lines of case law were reconcilable. In Warner-Lambert, the majority had followed the ab initio plausibility line, and found the relevant claim invalid, whereas the minority, applying ab initio implausibility, considered the claim valid. Given this dichotomy, Arnold LJ considered the test set out by the EBA to be closer to ab initio plausibility than implausibility. However, given that he was bound by Warner-Lambert in any event, this is commentary rather than firm law.

Finally, he noted that although the reference was only concerned with the assessment under inventive step, the EBA had explained that second medical use claims would still need be assessed under sufficiency, and would face a higher bar to relying on post-filed evidence than with the test under inventive step. 

Court of Appeal decision

BMS appealed on the following grounds:

i. In the case of a claim to a single chemical compound, there is no requirement that the specification makes it plausible that the compound is useful.

ii. The standard of plausibility laid down by the majority in Warner-Lambert is too high and in this case the standard should be that of the minority in the Supreme Court, or the standard of the majority should be applied less strictly.

iii. The ‘encouragement plus tests’ approach can render a patent plausible.

iv. Standing back, the claimed invention fulfilled the patent bargain. 

v. The application contained an implicit disclosure that apixaban had a nanomolar Ki (far more effective than the 10 µM value given in the “Utility” section for some compounds of the invention) against factor Xa, which made it suitable for therapeutic use.

vi. Taking the statement about the compounds of the invention, and the disclosure of Example 18 (which is apixaban), together rendered it plausible that apixaban was a factor Xa inhibitor. 

Arnold LJ agreed with Meade J that Warner-Lambert applied to compound, as well as second medical use, claims. This was because: i) Warner-Lambert relied upon EPO case law applying the concept of plausibility to compound claims; and ii) most importantly, the fundamental principle that the scope of the patent monopoly should be justified by the patentee’s contribution to the art applied no matter the type of claim. Given this finding, Arnold LJ was bound to follow the majority in Warner-Lambert, which as explained above, was seen as consistent with G2/21 in any event. He also viewed Warner-Lambert as clearly rejecting the ‘encouragement plus tests’ approach, and that there was no obligation to stand back at the end of an evaluation and consider whether a patent fulfilled the patent bargain. Grounds i) to iv) therefore failed.

Ground v) was considered a new case, not run by BMS below at first instance, and so was also rejected. Ground vi) was an assertion that the judge had erred in his assessment of plausibility. Such challenges often struggle at the appeal stage, as the Court of Appeal can only interfere if the judge has made an error of law or principle. Arnold LJ did not consider Meade J to have made such an error, and moreover agreed with his assessment. 

With Lord Justices Nugee and Warby agreeing with Arnold LJ, BMS’ appeal was therefore dismissed. BMS sought permission to appeal from the Supreme Court, but this has been refused on the basis that the appeal raises no arguable point of law. 


Although the Court of Appeal has been unanimous in upholding the first instance judgment, the law on plausibility remains far from settled. In the UK the Supreme Court is set to hear the appeal in proceedings between FibroGen and Akebia in March 2024. This also concerns plausibility, although with different issues to those raised in the Sandoz case. Further, the UK currently may be at odds with the EPO and other EPC contracting states. Since G2/21, the original referring EPO board in T 116/18 has held the patent valid and allowed the patentee to rely on post-filed data. Further, the French court of first instance, and the Dutch appeal court in preliminary injunction proceedings, have both held the French and Dutch designations of the apixaban compound patent valid when applying G2/21. Whether and how the Supreme Court addresses G2/21 and this divergence in the FibroGen v Akebia proceedings remains to be seen.


biotechreview11_changing, biotechreviewoftheyear_11