Friday, 27 October 2017

Mylan v Yeda: the fallout from Lilly v Actavis

As will now be well known to all patent practitioners, at least in the UK, the Supreme Court's judgment in Lilly v Actavis has changed the law quite radically on claim construction. Rather than just applying a general test of what the skilled person would understand the patentee to have meant, as outlined by Lord Hoffmann in Kirin-Amgen, it is now the law in the UK that due account should be taken of equivalents, in line with Article 2 of the Protocol on the Interpretation of Article 69 EPC. This has brought the law in the UK into line with that in other European countries, in particular Germany, where equivalents have been a feature since at least the 1980s.

What was not addressed in Lilly v Actavis was how this change in claim construction would affect how novelty would be assessed. Could it be the case that a claim could be considered to encompass equivalents that were in the prior art? How would the Gillette defence work in this case? Would novelty be assessed in relation to the prior art and its equivalents, or would the equivalents be a matter only for inventive step and not novelty? Some of these questions were considered in the judgment of Mr Justice Arnold in Mylan v Yeda, which issued yesterday (26 October 2017).

The case, relating to EP2949335B1, was about a medical treatment claim, where the claimed invention defined a regimen for treating multiple sclerosis involving three injections of a 40mg dose of glatiramer acetate weekly, with at least one day between each injection. The primary piece of prior art (termed Pinchasi), disclosed treatment of the same condition using the same compound at the same dose, but administered instead every other day. This would, of course, result in the patient getting injections that alternated between three and four each week.

The question was then whether following the treatment disclosed in Pinchasi would inevitably infringe the patent. Mylan argued that "it remained the law that a claim lacked novelty if the prior publication disclosed subject-matter which, if performed, would necessarily infringe the claim. Even if the subject-matter would not fall within the claim on its proper interpretation, it was sufficient that the subject-matter would infringe the claim applying the doctrine of equivalents. Otherwise, a claim could be infringed by a person who did exactly what the prior publication taught, yet the claim would be novel over that prior publication" (paragraph 162). Yeda, however, argued that "it was no longer the law that a claim lacked novelty if the prior publication disclosed subject-matter which, if performed, would necessarily infringe the claim. Rather, the claim would only lack novelty if the prior publication disclosed subject-matter which fell within the claim on its proper interpretation. It was not sufficient that the subject-matter would infringe the claim applying the doctrine of equivalents" (paragraph 163). In deciding which one was correct, Arnold J referred to section I.C.4.5 of the Case Law of the EPO Boards of Appeal, which cites several cases that decided that the disclosure of prior documents did not include equivalents. Equivalents should instead be considered under obviousness. Arnold J then concluded that Yeda was correct, although didn't say whether this meant that anticipation by inherency (as in Merrell Dow v Norton) was no longer the law. I suspect that he did not mean to go this far, but intended it to only mean that equivalents were not to be considered under an assessment of novelty.

Just in case he was wrong on the novelty point, however, Arnold then went on to find that, if it were legally possible for a claim to lack novelty by virtue of the doctrine of equivalents, it would lack novelty over Pinchasi.

What was not considered on the novelty point, but seems to me to be fairly important, was whether the claim language strictly limited the regimen of three, and only three, injections every week. This does not seem to have been argued by either side, nor was it considered in the section on claim construction (paragraphs 140-143). Claim 1, at least on my reading, does not limit the number of injections per week to only three, but defines it as three injections every seven days, with each injection separated by at least one day. This would, in my opinion, also encompass an injection every other day, even though this would result in four injections every other week. Any question of whether the doctrine of equivalents would need to be used to assess novelty would then be irrelevant, as the prior art would clearly fall within the scope of claim 1.

In the end, Arnold J found that the claimed invention was obvious over the prior art anyway, given that one fewer dose every two weeks would probably not make much difference and it was obvious to try a regimen with three per week instead. The patent was therefore found invalid, but perhaps not for the right reasons. If this goes any further, which Arnold J suggested that it might (and I suspect it will, given the very large sums of money involved), perhaps there will be a reassessment of whether the decision on novelty was really properly thought through.


  1. As far as I know, the Yeda line is the established way, on the Continent, to do novelty/claim construction/infringement by equivalent. At a Conference about 15 years ago, I remember Pumfrey J chiding me, that what I supposed to be "a given", namely that if what comes after, and infringes will, if coming before, prejudice novelty, is not at all a Given, on the Continent.

    Presumably Arnold J quotes from the White Book to box in that line. I should now read the Decision though.

    1. If the law of equivalents excluded from equivalents that which was known before the priority date of the patent then the "given" would be maintained as there would be no infringement.

  2. I agree that the claim would appear to read directly on to the prior art regimen.

    The prior art regimen has a day between each dose. It also involves 3 administrations (only) for every 1 out of 2 weeks. Further, the claimed regimen does not seem to exclude 4 administrations per week. Finally, there does not appear to be any requirement for the administrations to take place over more than 7 days.

    So, a straight novelty knock-out, then?

    1. That's what I was thinking. I wonder why Arnold J didn't see it, or at least explain why it wasn't a novelty knock-out?