English Court of Appeal upholds validity of patent and SPC covering Astellas' blockbuster treatment for prostate cancer

This dispute concerned revocation actions brought by generic manufacturers Accord, Sandoz and Teva against EP 1 893 196 and the related supplementary protection certification (SPC). Both EP'196 and the SPC cover enzalutamide, which is the active ingredient in Astellas’ prostate cancer drug Xtandi®. The generics conceded that their products would infringe the patent and the SPC, if held to be valid.

The Court of Appeal in Accord, Sandoz and Teva v The Regents of the University of California and Astellas Pharma recently upheld the Patents Court finding that the patent and SPC are valid. The Court of Appeal's decision adds to a line of recent judgments confirming the high standard for overturning a fact-dependent evaluative decision on appeal. A first-instance judge has an advantage over an appellate court as the former has been immersed in the facts of the case and has heard the cross-examination of technical experts. This is particularly so in specialist areas such as patent law. This point was emphasised by Lord Justice Zacaroli, who added a short supplemental judgment on this point, agreeing with Lord Justice Arnold's leading judgment.

Obviousness attack rejected by Patents Court

At first instance, Mr Justice Mellor rejected the generics' assertion that the claimed invention was obvious in the light of each of the two items of prior art, referred to as “the poster” and “the slides”.

Both the poster and the slides reported work done by the inventors, which disclosed a molecule identified as RD162. The only difference between RD162 and the molecule claimed in the patent (RD162', now known as enzalutamide) was the replacement of a cyclobutyl group by a dimethyl group at a position in the molecule referred to by the judge as position "X".

The judge rejected the generics' obviousness attacks on the basis that their expert evidence was "tainted with hindsight", while also acknowledging some of the criticisms made by the generics of one of Astellas' experts. One of the issues identified by Mr Justice Mellor was that the generics' expert approached obviousness by applying the Pozzoli analysis.

The generics' expert did not explain the scenario in which the skilled team decided to take the steps he described as obvious. Instead, he was asked to consider the change from cyclobutyl to dimethyl at position X and asked whether that change would have a material effect. In the judge's view, the manner in which this was put to him was "leading".

The judge agreed with Astellas that there "can be a real danger" in saying that just because a modification is “small” or “conservative” structurally, it is obviously "trivial” and will have no material effect. As he noted, "[c]hemical structures are very different from plate diameters" (drawing a contrast between a small change in a chemical structure with a small change in the diameter of a plate, which would have no technical effect). He also accepted that the skilled team would likely regard the switch from cyclobutyl to dimethyl as "something of a backward step".

The judge clearly found the case quite finely balanced, in particular noting that he had changed his mind on obviousness over the slides more than once. However, he concluded that the skilled team looking for a novel compound (that is, an alternative to RD162 disclosed in the poster and the slides) would not have been motivated to consider close analogues of RD162.

The generics' expert evidence failed to explain a "context" in which the skilled team starting with RD162 would have gone on to test RD162'. The only reason the evidence focused on close analogues of RD162 was driven by the hindsight knowledge of the compound claimed by the patent.

Validity maintained on appeal

The generics contended that the judge erred in principle in two ways. First, they argued that he erred in looking for a “context” or “scenario” to explain their experts' finding when the outcome was technically obvious. Motivation to identify an improved product is not a necessary condition for a finding of obviousness – the skilled person or team is deemed to be interested in alternative ways of achieving essentially the same result as that obtained by the prior art and this, they alleged, in itself, should provide sufficient context.

Second, the generics contended that the judgment contained an internal inconsistency in how the judge reached his conclusion that their expert evidence was tainted with hindsight.

The appeal was dismissed. With respect to the generics' first contended error, the Court of Appeal found that it was legitimate for the judge to have taken into account the fact that the generics' expert did not explain the scenario in which the skilled team decided to take the step which he said was obvious.

The judge had not questioned that it was enough for the skilled team merely to want to develop an alternative to RD162, but the expert had not explained in his evidence that the skilled team would approach the prior art on that premise – he had not explained the team's objective at all.

In particular, the expert did not provide any explanation as to why the skilled team would start by considering a change at position X, nor why the skilled team would consider changing the cyclobutyl group to a dimethyl group at that position when the prior art had shown the cyclobutyl group to be more active. At the very least, it was expected that this would have been discussed within the skilled team.

In relation to the second alleged error, the Court of Appeal re-emphasised that a conclusion on obviousness is a highly fact-dependent evaluative decision where the first-instance judge, who has been immersed in the detail of the case, has a significant advantage over an appellate court.

The Court of Appeal found that Mr Justice Mellor was "entitled to conclude" that the generics' expert evidence was infected with hindsight, bearing in mind (i) the absence of context discussed above, (ii) the sequence in which the generics' expert had considered the documents and (iii) the way in which the generics' expert first report was structured. There was no inconsistency in the judge's reasoning.

Osborne Clarke comment

This judgment follows a recent line of decisions from the appellate courts where the appellate court has made it clear that it is unwilling to overturn a decision simply because it may disagree with the lower court's findings, even in a case such as this where the conclusion was quite finely balanced – see for example Lifestyle Equities CV v Amazon UK Services Ltd [2024] UKSC 8 and Iconix Luxembourg Holdings SARL v Dream Pairs Europe Inc [2025] UKSC 25.  

In order to successfully appeal a first-instance decision, appellant(s) must meet the threshold of establishing that the lower court's decision involved an error of law or principle. This is particularly relevant in the case of obviousness, given that it involves a multi-factorial evaluation by the judge in a specialist area where the judge has heard the cross-examination of technical experts.

The judgment also touches on another important issue that has been a focus of recent patent litigation decisions – the instruction of experts. Here, the Court of Appeal provided some useful guidance on the proper instruction of experts. Lord Justice Arnold, who gave the leading judgment, stated that in his opinion explicitly instructing an expert to follow the Pozzoli approach can create a risk of hindsight. Although the court usually adopts the Pozzoli approach, that does not mean it is necessary to instruct experts, whose role is to provide technical assistance to the court, to follow it in their evidence.

Lord Justice Arnold caveated this by stating that he was not intending to be prescriptive about how experts should be instructed in other cases. Nonetheless, this case serves as another useful warning for parties to consider carefully the manner in which they instruct their experts. In order to avoid hindsight, it may be preferable to ask an open question about what was obvious from the prior art, rather than whether a particular difference was obvious.

The decision also serves as a reminder that, even though motivation to take a particular step is not necessary for a finding of obviousness, experts should explain the context in which a skilled team makes its decisions.