The English courts have a reputation for being hostile to patents. There exists a perception that if the validity of a patent is challenged – for example, on the classic grounds that the patented invention is not novel, or that it is obvious, or that the patent does not sufficiently disclose the invention – then an English court is more likely to find the patent invalid than would a court on the continent for the equivalent patent in its jurisdiction.

However much the English patent court judges protest at the injustice of this view, it is persistent. Indeed, it may even be self-fulfilling. If a patent is weak, then it is more likely to be challenged. A challenger with a number of equivalent patents in different jurisdictions may choose to litigate first in the jurisdiction where he thinks he is most likely to succeed, which he may believe to be England. Since the patent is weak, the English courts are more likely to find that it is invalid, and so the perception is reinforced.

And yet in a grand total of three appellate decisions in the first half of this year, the English Courts have shown themselves prepared to uphold really quite controversial patents.

In the first, the Court of Appeal affirmed the validity of a patent to an antidepressant, the existence and likely properties of which had been known for many years. In the second, a patent to a new dosage regime for a known treatment for alopecia was upheld, and in the process the court made a significant development to procedural law. In the third, the House of Lords supported the validity of a patent which offered no proof that the claimed invention would actually work.

The first case is H Lundbeck A/S and Generics (UK) Limited and Ors [2008] EWCA Civ 311. The issues revolved around an antidepressant called citalopram. Citalopram is composed of two molecules called enantiomers that are virtual mirror images of each other, rather like a pair of clasped hands. Since enantiomers have very similar physical characteristics, they can be difficult to separate. Nevertheless, Lundbeck achieved separation and then patented one of these enantiomers, known as escitalopram. Escitalopram turned out to be extremely valuable because Lundbeck found that it was more effective as an antidepressant than its parent, citalopram. It has been used by millions of people.

Generics claimed that the patent to escitalopram was merely an improper attempt to extend the life of an earlier expired patent, owned by Lundbeck, to citalopram itself. In that earlier patent, Lundbeck had not only disclosed citalopram, it had disclosed the fact that it was a racemate, made up of enantiomers. In other words, the existence of escitalopram, and its possible use as an antidepressant, were both well-known at the date of the patent. Granted, no one had actually separated escitalopram before, but everyone knew it was there. Generics argued that, since separation of the enantiomers was difficult, Lundbeck might be entitled to a patent to the method of separation that it had discovered. But it was not entitled to a patent to escitalopram itself, which would give Lundbeck a monopoly over the substance itself, regardless of the separation method used.

At first instance, the challenge succeeded. The judge held that simply finding a method for separating escitalopram from citalopram did not entitle Lundbeck to monopolise escitalopram. The scientific contribution that Lundbeck had made was not sufficient for the breadth of the patent monopoly granted.

The Court of Appeal unanimously overturned this decision. It essentially asked whether the patent ticked the boxes for patentability. Was escitalopram new? Yes, since no one had managed to isolate it before. Had it been obtained by a method that was inventive and not obvious? Yes, the first instance judge had found this on the facts. Did the patent describe how to obtain escitalopram sufficiently for the method to be repeated? Yes. The court had some sympathy with the views of the first instance judge about the narrow scope of Lundbeck’s scientific contribution versus the wide extent of its monopoly, but pointed out that this is not relevant when considering the validity of a patent – you do not balance the scientific contribution and the scope of monopoly in golden scales, as it were. So long as the patent ticks the boxes of patentability, a patent to a known substance is valid.

The second case is Actavis UK v Merck & Co Inc [2008] EWC8 Civ 444. The case concerned Merck’s patent to use a substance called finasteride for the treatment of androgenic alopecia, a type of baldness occurring in men and woman. Merck already owned a patent to treat androgenic alopecia with finasteride, with a dosage of 5mg to 2,000mg. The dosage amount claimed in the patent in suit was much lower: 0.05-1 mg per day. Actavis challenged Merck’s low-dosage patent on the basis of this earlier, high-dosage patent, claiming that the low dosage was not new, or alternatively that it was obvious.

Stating the obvious

Quickly dealing with the first point, the court held that the low dosage regime was indeed new: no one had made it or proposed it before. The second point, the challenge to the new dosage regime on the grounds that it was obvious, was a far higher hurdle for Merck to clear. The court commented that dosage regimes would nearly always be obvious, because it is standard practice to investigate appropriate dosage. But in this case, the new regime was not obvious. The court cited two documents, published after the earlier patent but before the patent in suit, which ‘taught away’ from using finasteride to treat androgenic alopecia. It had been thought that finasteride worked by inhibiting a particular enzyme in the scalp which was thought to have a function in androgenic alopecia. The papers showed that this enzyme was not, in fact, present in the scalp, and the court held that the skilled reader would understand these papers to mean that finasteride would not after all be a useful treatment. Merck had flown in the face of this knowledge and trialled finasteride, only to find that it was useful to treat androgenic alopecia, including at the low dosages then claimed.

The case is significant outside the world of patents because of what the Court of Appeal had to say about its obligation to be bound by its earlier decisions. Actavis referred the court to its earlier decision in Bristol-Myers Squibb v Baker Norton [2001] RPC 1, that a patent to a particular dosage regime for administering a cancer treatment was invalid. Generally, the Court of Appeal is bound by its previous judgments (Young v Bristol Aeroplane Company [1944] KB 718). Actavis said that the effect of Bristol-Myers Squibb was that the court was bound to revoke patents to new dosage regimes.

The Court of Appeal did not agree that Bristol-Myers Squibb said quite what Actavis said it said, but it went on to rule that, in any event, it was not bound by the decision. It pointed to a number of decisions from the European Patent Office (EPO) boards of appeal, to the effect that a previously undisclosed dosage regime would be novel, in the sense that it had never been done before. The UK courts are not bound by these decisions, but pursuant to section 91 of the Patents Act 1977 they have to take ‘judicial notice’ of them. Also, the UK is a party to the European Patent Convention, which seeks to harmonise patent law for all members of the EU, and it was obviously desirable for courts across Europe, so far as possible, to adopt a common approach towards patents. The court therefore held that if it is satisfied that the EPO boards of appeal have formed a settled view of European patent law – in this case, that a new and non-obvious dosage regime is patentable – that is inconsistent with an earlier Court of Appeal decision, then the court may depart from the earlier decision. As an aside, this decision might be laying the ground for a change to the UK’s approach to software patents, on which the Court of Appeal and the European Patent Offices have been at odds (see [2007] Gazette, 29 March, 22).

The third case is Conor Medsystems Inc v Angiotech Pharmaceuticals Inc & Ors [2008] UKHL 49. Angiotech owns a patent to a stent coated with a polymer containing the drug taxol. Stents are tiny metal scaffolds that are used to keep arteries open. A problem with stents is that they can produce an exaggerated healing response, called restenosis, in which smooth muscle cells proliferate and constrict the very artery that the stent was meant to hold open. Angiotech’s ‘big idea’ was to coat the stent with an agent which would inhibit the growth of capillary blood vessels. This would inhibit restenosis because cell tissue cannot grow far without capillary support.

The big idea

Conor objected to Angiotech’s patent on several grounds, but the key issue before the House of Lords was that the patent did not actually show that Angiotech’s ‘big idea’ would work. Pharmaceutical patents often contain experimental data to support the efficacy of the claimed invention. However, Angiotech’s patent offered no such proof, nor did it explain how or why the proposed solution might work. Conor therefore claimed that the patent contained nothing more than a proposition, on a par with many other proposed solutions to restenosis that had not got beyond the ‘idea’ stage. It might be a proposition that was worth a try, but until Angiotech had proved that it would work, it had added nothing to existing knowledge and was not entitled to a patent.

The House of Lords pointed out that patent law requires that a claim be ‘supported by the description’ in the patent (section 14(5)(c) of the Patents Act 1977), which should prevent patents being granted for ideas that are mere speculation. It also cited two cases from the EPO that found against ‘speculative’ claims. In Agrevo (case no. T 0939/92), a claim to a class of compounds alleged to have a common technical effect was rejected because there was nothing to show that every member of the class would have the claimed effect. In John Hopkins University School of Medicine (case no. T 1329/04), a claim to a DNA sequence for a particular use, where the claimed use was only speculative, was not remedied by proof of use generated after the date of the patent. But their Lordships also pointed out that there is no requirement in patent law that a patent should demonstrate by experiment that the invention will work, or for the patent to explain why the invention will work. The specification in Angiotech’s patent contained a plausible proposition that a taxol-coated stent would prevent restenosis. It did not prove the claim as such, but the teaching was clear. The teaching was sufficient for the House of Lords to overturn the first instance and Court of Appeal decisions and uphold Angiotech’s patent.

As with the first two cases, the proposition in Angiotech is striking: that a patent holder does not need to show that, or even to have found out whether, his or her invention will actually work. If monkeys hitting typewriters were to come up with a cure for cancer, and the claim for the cure is supported by the description of the cure, then it is patentable (I would guess by the owner of the monkeys – although I am probably straining the metaphor). The decision may well prompt inventors to file claims early and before their technology has been fully tested, particularly if they know they are in a research race with competitors towards solving a particular problem. Early filers may take a risk, though, that the data in their application is not sufficient to support the claimed invention, or that a later, solid patent claim is undermined by a half-formulated disclosure in an earlier application.

Critics of these judgments might regard the first two as examples of the pharmaceutical industry wrongfully extending their monopolies over old drugs, and the last as an early ‘land-grab’ over an as-yet untested intervention. But the facts of each, at least as they are reported in the judgments, are quite special: escitalopram, the known substance which no one had managed to unlock; finasteride, the drug which no longer appeared to be effective; and taxol-coated stents, which could solve a major problem in angioplasty. As regards the first two, don’t forget as well that new patents cannot stop you from doing that which is disclosed in an old, expired patent. The patent to escitalopram does not prevent anyone from using citalopram, nor does the patented finasteride dosage stop people from using the old dosage. A key justification for granting patent monopolies is that they encourage investment in research, and others would argue that these companies deserve to be rewarded for investing in these developments. They are rewards which, in these instances at least, the English courts have shown themselves prepared to allow.