Takeda v Roche: "Is it plausible? Is it true?"

The growing clinical and economic significance of antibody-based therapeutics is reflected by the patent cases currently passing before the UK courts, of which Takeda v Roche is the latest example. IPKat is eagerly awaiting the UK Supreme Court hearing on Regeneron v Kymab, relating to Kymab's therapeutic antibody platform technology (IPKat post here and here). Earlier this year, the UK courts considered the validity of Genentech's patent for Taltz, an anti-IL-17 monoclonal antibody (IPKat post here), and last month, the courts considered Pfizer's request for an Arrow declaration relating to Roche's monoclonal antibody drug Avastin (bevacizumab) (IPKat post here).

The latest UK High Court patent decision (Takeda v Roche [2019] EWHC 1911) relates to another pair of pharmaceutical company giants (Takeda and Roche), and another monoclonal antibody therapeutic (Takeda's Vedolizumab, branded as Entyvio).

The decision is a long one. Mr Justice Birss considered issues of claim construction, infringement, novelty, inventive step and sufficiency. Key points of interest in the decision include the interpretation of "antibody" in the context of a patent claim, and the use of plausibility in the test for inventive step.

Case Background

Entyvio is an anti-integrin used to treat ulcerative colitis and Crohn's disease. Entyvio recently outperformed Abbivie's Humira in a ulcerative colitis phase 3b clinical trial and sales of Entyvio reached $1.37 billion in the nine months to December 2018. Roche claimed that Entyvio infringed its European patent EP 2 007 809. The patent had been maintained in amended form by the EPO Boards of Appeal following an opposition by GSK and Novartis (T1784/15). Takeda denied infringement and brought a revocation action in the UK courts against the patent.

The patent related to a modified antibody structure that reduced the capacity of the antibody to cause unwanted cell death. Particularly, Claim 1 of the patent related to a glycosylated human monoclonal antibody, characterised by the fucose content of its sugar chain: at least 99% fucosylation (as measured by LCMS peptide mass analysis). Although not mentioned in the patent in words, Roche argued that the high level of fucosylation abolished the ability of the antibody to cause cell death (as shown in the Figures of the patent).

Infringement - Functional definition of an antibody

Mr Justice Birss first considered whether Vedolizumab (Entyvio) fell within the scope of the claim. A point of contention was whether Vedolizumab had a functional antibody receptor binding region, as was required by the definition of "antibody" in the claim. The Judge construed "antibody" as requiring that the antibody had a functional antibody receptor binding region (as defined in the description of the patent). The antibody receptor binding region had to be structurally capable of binding an antibody receptor and mediating cell death. Fucosylation of the antibody was said to prevent the normally functioning cell death activity of the receptor binding region.

Fucose

Evidence submitted to the court showed that Vedolizumab did bind the antibody receptor, but did not cause cell death. Takeda contended that this was due to a mutation in the antibody receptor binding region of the antibody. In other words, Takeda argued that the antibody receptor binding region was not functional, even without fucosylation of the antibody. Vedolizumab, according to Takeda, therefore had no functional antibody receptor binding region. Roche countered that Vedolizumab was able to bind the antibody receptor. The ability of Vedolizumab to cause cell death was thus not completely eliminated by the mutation in the antibody receptor. The lack of antibody dependent cell death exhibited, Roche argued, was therefore due to the fucosylation of Vedolizumab.

After assessing the strength of the evidence submitted by both parties, Mr Justice Birss, agreed with Roche that Vedolizumab did display some antibody receptor binding activity (however slight). Vedolizumab could therefore be said to have a functional antibody receptor binding region. Vedolizumab was thus found to fall within the scope of the claim (para. 116).

Novelty

Mr Justice Birss found that the patent lacked novelty in view of a number of cited prior art documents. Arguments from Roche that the disclosures in the prior art documents were not enabled were dismissed. The prior art was found to expressly disclose the idea of an antibody with the claimed level of fucosylation. The Judge also found that it would have been routine for a skilled person to arrive at the level of fucosylation by routine experiments (para. 157). Claim 1 was therefore found to lack novelty.

Obviousness: Has the objective problem plausibly been solved?

Mr Justice Birss also went on to consider the issue of inventive step. In the obviousness assessment, the Judge asks the question of whether the claim provided a technical contribution that had been plausibly demonstrated (and was true) across the entire breath of the claim. Mr Justice Birss particularly followed the approach in Generics v Yeda (IPKat post here) and Warner Lambert (IPKat post here).

In Warner Lambert, the Supreme Court considered concept of sufficiency as applied to Swiss-style second medical use patents. The Court confirmed that, to be sufficiently disclosed, a second medical use claim must be plausibly demonstrated by the patent over its entire scope. Plausibility objections are also often encountered at the EPO in the context of inventive step. The EPO requires an applicant to plausibly demonstrate that the objective technical problem to be solved by the claimed invention has indeed been solved (Case Law of the Boards of Appeal, I.D.4.1). The effect must also be plausibly demonstrated, e.g. by empirical data, over the entire scope of the claim (T 409/91).

In Takeda v Roche Mr Justice Birss asked whether the patent made a technical contribution to the art (para. 207). This question could be phrased, the Judge argued, in terms of whether the patent plausibly demonstrated a technical contribution to the art (para. 203). He noted:

In relation to each disclosure there are five questions to answer: Is it disclosed in the patent? Is it plausible? Is it true? Is it a technical advance? Does it support claims of the breadth they are?

The relevant art included Shield et al. (2002) (as common general knowledge). Shield showed that 98% fucosylation reduced antibody dependent cell death ("specific lysis") at low antibody concentrations. At high antibody concentrations (>10 ng/ml), the effect of fucosylation on antibody cell death was abolished (para. 37).

EP2007809, Figure 1

Roche argued that the technical contribution made by the patent was that 99% fucosylation of an antibody eliminates the antibody's cell death activity, even at high antibody concentrations. The effect of 99% fucosylation was compared to the effect of lower fucosylation of Shield et al. (2002).  The patent itself did not expressly mention the technical effect of 99% fucosylation in so many words. However, Roche argued that the technical effect was clearly shown in Figure 1. Figure 1 shows the antibody-dependent cell death activity of the fucosylated antibodies  described in the Examples (clones 1-7) compared to a control, non-fucosylated antibody (positive control).

Mr Justice Birss found that it was plausible from Figure 1 that fucosylation eliminated the cell death caused by the antibody (at the antibody concentration used in Figure 1, presumed to be 25 ng/ml). However, the Mr Justice Birss found that Figure 1 did not plausibly demonstrate whether antibody mediated cell death would be eliminated at a higher antibody concentration. The Judge was also not convinced that it was even true that antibody cell death would be abolished at higher concentrations of antibody (para. 223). It was concluded that the data provided in the patent did not support the breadth of claim 1. The claim was therefore found to be obvious in view of the prior art (para. 225):

Figure 1 makes plausible the idea that the antibody tested reduces ADCC [i.e. antibody dependent cell death]. It also makes plausible the idea that the antibody exhibits no ADCC at the concentration (presumably) tested. However it does not make plausible a wider proposition about the effect at higher concentrations since it simply does not address it. Rather than eliminating ADCC at any relevant concentration, for all the skilled person knows what has happened is simply that for the antibody tested, the concentration at which ADCC occurs has been increased above the level in the Figure. Nor am I convinced the proposition is true. 

Mr Justice Birss also found (in contrast to the Board of Appeal of the EPO (T1784/15, r. 76), that the feature of  >99% fucosylation could not support an inventive step. In particular, the Judge found that this feature had no technical effect (para. 208), was an arbitrary selection (para. 209) and was limited to the production of the antibodies by a particular cell type (para. 212):

[T]he contribution is limited to CHO cells. That does not support claims of the width of the relevant claims in this case, because they are product claims not limited to products made in CHO cells and because fucosylation is well known to depend on cell type.

It may just be this GuestKat, but Mr Justice Birss' application of the test that a claim provides a solution to a technical problem that has been plausibly solved, over the entire scope of the claim, seems potentially stricter than the EPO approach (at least with respect to the point on cell type)?

Insufficiency

Takeda also argued that the claim was insufficient because the scope of the claim was ambiguous. The Judge agreed with Takeda's arguments that it would not be clear to a skilled person whether an antibody fell within the scope of the claim. In particular, whether or not an antibody's fucose content falls within the scope of the claim will depend on the LCMS peptide mapping machine used to analyse the antibody. It would also not be clear to a skilled person which LCMS machine to use. Thus, the Judge concluded "[t]he claim is truly ambiguous and invalid" (para. 254). The patent was therefore also found invalid for insufficiency.

Roche's patent was therefore found to be infringed but to lack novelty, inventive step and sufficiency of disclosure. Roche is expected to appeal the decision.