The patent battle between Warner-Lambert (Pfizer) and Actavis and Mylan reached the Supreme Court in February 2018. (See our earlier blog posted shortly before the hearing began). Members of Fieldfisher's patent litigation team were in court to hear the judges' initial reaction to the arguments which can shed light on where the court's reasoned judgment will end up. The issues are of huge significance to users of the patent system. They include whether a patentable invention needs to be made "plausible" at the filing date and, if so, the level of information the patent application must include, and the correct approach to enforcing second medical use claims.
Why are the parties in court?
Pregabalin earned over 3 billion US dollars for Pfizer in 2016. The drug (branded LYRICA) was first patented to treat epilepsy and anxiety. Pfizer obtained a further patent for use of the drug to treat pain. Pfizer relied on this second patent to bring proceedings against certain generic drug manufacturers. In 2016, the UK Court of Appeal upheld a first instance decision that the patent is invalid for insufficiency notwithstanding that the drug is a highly effective treatment for pain and there is no question that the patent enables such use. Both courts also found the patent not infringed by the generics manufacturers who plan to launch generic versions of the drug.
The issues before the Supreme Court
The Supreme Court will rule on the validity of claim 3 of the patent (the use of pregabalin to treat "neuropathic pain") as well as a handful of other use claims for the treatment of specific types of "neuropathic pain". There are four key issues before the court:
- Plausibility - Is there a stand-alone requirement of plausibility, in the context of insufficiency and, if so, what level of information must a patent application include to satisfy this requirement? Can later filed evidence be used to fill any gap?
- Infringement of Second Medical Use Claims - How should the law reward patentees for finding new uses of known drugs whilst at the same time not excluding manufacturers from making and marketing drugs for their original purpose? Will it suffice to avoid infringement if a generic manufacturer uses a "skinny label" on its product indicating only the off-patent uses of the drug?
- Construction - What is the correct approach to the construction of patent claims? If a claim has a dual meaning, should the court apply a "principle of validating construction" by favouring the interpretation which is most likely to render the patent valid?
- Abuse of Process - Is a post-trial application to amend an invalid patent claim to limit it to a part found to be plausible an abuse of process?
A record number of interveners?
Very few patent cases have attracted as many interveners as this one. The Supreme Court is considering written submissions from 10 interested parties (including the European Federation of Pharmaceutical Industries and Associations, the UK BioIndustry Association and the Chartered Institute of Patent Attorneys) as well as oral submissions by The Secretary of State for Health.
The interveners' interests lie, mainly, in drawing the court's attention to the competing policy considerations raised by the plausibility and infringement arguments. The Court of Appeal outlined the tension in the plausibility case, as follows, in the decision under appeal:
“The EPO and domestic law …indicate that the requirement of plausibility is a low, threshold test… which is designed to prohibit speculative claiming. It is not designed to prohibit patents for good faith predictions which have some, albeit manifestly incomplete, basis. Such claims may turn out to be insufficient if the prediction turns out to be untrue. A patent which accurately predicts that an invention will work is, however, not likely to be revoked on the ground that the prediction was based on the slimmest of evidence".
The fear is that barely credible "speculative patents" or "armchair inventions" may deter further, important work on the claimed technology. An example advanced by Pfizer before the Supreme Court is a patent filed for new classes of molecules for every possible disease on the off chance that one of those molecules is effective to treat one such disease. Pfizer say this is different from claims based on an "educated guess" which should be allowed. The question is "how are the courts and patent offices to draw a line between armchair inventions and educated guesses?" The solution currently favoured by the EPO (and more recently, the UK courts) is to require information in the patent application that makes the invention 'plausible'. Pfizer – along with several interveners – has urged the Supreme Court not to adopt any rule of plausibility that would require patentees to include clinical trial data in their patent applications. Such data must be published at the time of the trial which would render any associated invention unpatentable.
Different policy considerations apply to second medical use patents. Drugs which have been on the market for some years have a strong safety record. Finding new indications for those drugs, to solve unmet medical needs, can provide significant long term benefits for society. Second medical use patents are important to incentivise and reward research into new indications for known drugs. Equally, reductions in the price of drugs that follow generic entry onto the market benefit the public, and the governments, who buy those drugs. In practice, the law has struggled to reconcile these competing policy aims. Even if a generic version of a drug bears a "skinny label" showing the non-patented indications only, the risk for the originator is that the generic drug may be prescribed or dispensed for the new patented indications.
Against this backdrop, the Supreme Court is considering the following arguments advanced at the hearing.
At first instance, Arnold J considered the animal data in the patent and concluded that:
“given that plausibility is a relatively low threshold, the data contained in the specification, when read with the common general knowledge, just make it plausible that pregabalin would be effective to treat peripheral neuropathic pain”
This decision, which was approved by the Court of Appeal, invalidated claims 1 (all pain) and 3 (neuropathic pain).
Pfizer's primary case on plausibility is that if the invention has been shown to work across the scope of the claim, it is sufficient. There is no room for a separate objection of plausibility (which is not mentioned in the European Patent Convention or in domestic legislation). Alternatively, if the court concludes there should be any test for plausibility in the sufficiency context it should be a low threshold test. Such a test should not require evidence establishing that every possible condition covered by the claim is plausible. It is enough if the patent is rendered plausible at the same level of generality as the invention is claimed in the patent. Further, if the only way of proving that the patent was plausible at the filing date is by means of later evidence, that evidence ought to be admissible.
Mylan and Actavis are arguing, conversely, that the test for plausibility should be the same as that for obviousness: "was there enough information in the patent to motivate the skilled person to try the invention across the breadth of the claim in the reasonable expectation that it will work?".
Second medical use
Most patients who use pregabalin take it for pain; the market for non-patented indications is comparatively small. Pfizer's concern is that doctors and pharmacists will prescribe and dispense generic pregabalin for pain even if it has a skinny label. Arnold J at first instance thought that the solution to this problem was to get prescribers to prescribe the drug by its brand name LYRICA when it is prescribed for pain. He did not find the generics manufacturers liable for infringement. The Court of Appeal disagreed, ruling that the generics would infringe the patent directly and indirectly if they did not take all reasonable steps to prevent the product being used for pain. A skinny label would not suffice to achieve this. This was because it was reasonably foreseeable to the manufacturer that the product would be used to treat pain, irrespective of the label. The Court of Appeal made it clear that its comments on infringement were not strictly necessary – and would not bind other parties - because it had found the patent invalid
Pfizer did not appeal the Court of Appeal's findings on infringement. Mylan and Actavis applied, successfully, to raise the issue before the Supreme Court. Their case is that the manufacturer should only be liable for infringement if its intention is that the medicine should be used for the patented indication.
Pfizer's case is that infringement should be determined by means of the normal principles under the UK Patents Act 1977. Liability for direct patent infringement is strict and it would be contrary to principle, and the words of the statute, to add a requirement of reasonable foreseeability. Similarly, the Act sets out the knowledge requirement for indirect infringement. If Actavis "knows, or it is obvious to a reasonable person in the circumstances, that [its pregabalin is] suitable for putting, and is intended to put, the invention into effect in the United Kingdom", it will infringe. That should be the end of the matter.
Lord Sumption observed that neither test put forward by the parties offered a 'complete solution' to the problem. He characterised Pfizer's primary case as a 'ridiculous submission' given the reality of prescribing and dispensing practices which would turn all generic pregabalin into infringing products and all subsequent dealings with those products (by doctors and pharmacists) into infringing acts. In the circumstances, his Lordship said "perhaps [Pfizer] must accept that it must put up with less than total security in this situation". Lord Mance highlighted an equivalent injustice in Mylan and Actavis' position; there is a risk that generic pregabalin will be prescribed and/or dispensed for pain, irrespective of the manufacturer's intention. It follows that an intention-based test would seriously undermine the protection afforded by second medical use patents.
The Secretary of State for Health is proposing an alternative (but equally useless!) solution. In his view, there is no way of "squaring the circle". On the Secretary of State's behalf, Michael Silverleaf QC submitted that: " the law is imperfect and with the greatest of respect to your Lordships' powers, you can't fix it. So please don't try too hard".
The Law Lords expressed surprise that the parties had not been able to agree a 'middle ground'. They questioned the advocates with the aim of establishing what reasonable steps a generic manufacturer could take to ensure that generic pregabalin is used for non-patented indications only. A number of steps were contemplated: avoiding oversupply of the generic product (in light of the actual size of the market for the original indication), educating health authorities as to the circumstances in which the generic product can lawfully be prescribed and dispensed, and introducing appropriate notices into prescribing software used by healthcare professionals (eg to require doctors to prescribe by brand name where the product is for the treatment of the patented indications). Whether and how the Supreme Court may feed these suggestions into the law relating to infringement remains to be seen.
If Pfizer wins the arguments on construction, its patent will be held valid. The key point in dispute is what the skilled person would have understood the patentee mean by the words "neuropathic pain" in claim 3. Arnold J held that the term is used quite generally in the patent and so should be construed to mean "central and peripheral neuropathic pain". The Court of Appeal favoured the same construction.
Pfizer's case on appeal is that the term should be construed to mean "peripheral neuropathic pain" (for which the lower courts found the patent plausible). It points to the expert evidence on both sides which was that, at the priority date, some people in the field used "neuropathic pain" to means all types of neuropathic pain while others used it to mean peripheral neuropathic pain. Its meaning will depend on the context in which it is being used. The patent is the relevant context here and Pfizer says it contains a number of important signposts for the skilled person. For instance:
- the examples of neuropathic pain listed in the patent are peripheral;
- neuropathic pain is also described in the patent as pain caused by injury to "peripheral" nerves; and
- stroke and multiple sclerosis are both conspicuously absent from the patent. These are the two best known conditions which involve central neuropathic pain.
Pfizer says that the skilled person would realise from these signposts that the patentee was using "neuropathic pain" to mean "peripheral neuropathic pain". This is a construction which accords with the expert evidence on both sides which Arnold J had no good reason to reject.
Mylan and Actavis disagree. They say that the skilled person would have turned to a well-known international study on pain (known as the IASP) to understand the meaning of "neuropathic pain" in the patent. This definition of neuropathic pain in this study encompasses central and peripheral neuropathic pain.
The generics also say that the language of the claim should not change according to Pfizer's tactics in this particular case. Pfizer's case at trial was that claim 3 covers neuropathic pain across the board. Pfizer only introduced this alternative construction when it got into difficulty proving plausibility across the full breath of the claim. Had the patentee intended to use neuropathic pain in a narrow sense, it would have included a definition to that effect in the patent. The generics say it is not permissible to add words in to a patent which are not there. Claim 3 does not say "neuropathic pain" excluding central neuropathic pain. Finally, they dispute the existence of the signposts erected by Pfizer, pointing out, in particular, that the specification refers to a number of conditions such as tumour pain which may involve an element of central neuropathic pain.
If the court rejects Pfizer's preferred construction, the company is urging the Supreme Court to apply a principle of "validating construction" to claim 3, as follows:
- "neuropathic pain" has two possible meanings – a wider meaning and a narrower meaning;
- applying the principle of validating construction, the court should construe the claim to reflect the narrower meaning because that would be more consistent with the patent's specification and the data; and
- where claims are ambiguous in this way, the skilled reader will always understand the patentee to have intended to use the narrower meaning that is plausible (if, as here, the words are capable of bearing that narrower meaning).
Lord Pannick, representing Pfizer, explained that this general principle of interpretation has been applied for many years in contract and public law. This is because a construction that saves rather than defeats a contract will more closely reflect how the agreement would be understood by a reasonable person with a knowledge of the commercial purpose and background of the transaction. Lord Pannick also pointed to strong reasons of public policy for upholding a patent claim that is properly ambiguous if it satisfies all the tests for validity. This is to encourage the investment that is required to produce inventions that benefit the public.
In response, the generic companies are arguing that there is no good reason why contractual principles which govern bilateral agreements reflecting intentions of both parties should be applied to unilateral patent documents. The Protocol on the Interpretation of Article 69 of the European Patent Convention prescribes a specific approach to the construction of patent claims. It should be the start and the end of the construction process.
Lord Mance indicated that he was "struggling" with the principle of "validating construction" in the patent arena and warned that it would need careful "scrutiny". Arguably, Pfizer is asking the Supreme Court to give the rules of construction a patentee friendly gloss. If the court agrees, the scope of protection afforded by patent claims is likely to become far less certain from the perspective of third parties whose activities may be affected by those claims shifting the balance between the patentee's rights to protection and the public's right to clarity.
Abuse of process
After the judgment at first instance, Pfizer made a conditional application to amend claim 3 so as to limit the scope of the claim to peripheral neuropathic pain in order to avoid the insufficiency (plausibility) attack. Arnold J refused that application on the ground that it amounted to an abuse of process. The Court of Appeal confirmed that decision holding, in particular, that Pfizer could have indicated before the start of the trial that it would seek to amend its patent in the event that it was found invalid for insufficiency.
Before the Supreme Court, Pfizer argued that (i) under Article 138 of the European Patent Convention 2000 a patentee is entitled to amend its patent after a ruling by the EPO that it is invalid; (ii) under section 63 of the Patents Act 1977, in the event that a patent is found to be partially invalid, a patentee may amend its patent by deleting the invalid claims; and (iii) the decision that it was too late to seek amendment was a harsh result. Any additional costs arising from the delay in seeking the amendment, (which would have involved a further two day trial), could be compensated by an appropriate costs order.
Actavis' and Mylan's response was (i) Article 138 of the European Patent Convention 2000 does not allow a patentee to amend at any time, in particular, there are very strict rules about amendments at the stage of an appeal to the Technical Board of Appeal of the European Patent Office (ii) Article 138 does not direct member states as to the procedure for allowing amendment, it only requires them to have a procedure and in the UK that procedure is provided by section 63; and (iii) it was open to the patentee to make a conditional amendment application at any stage and, certainly, as soon as it became aware of the problem. The evidence showed that it became aware, at the very latest, during the trial. It was just too late to apply for an amendment after the judgment.
The court's attitude to Pfizer's position was summed up by Lord Hodge's comment that once Pfizer became aware of the problem with the claim as drafted, it chose to go down the construction route rather than seek to amend the patent. That choice may be held against Pfizer in the context of the "abuse of process".
The Supreme Court surprised us all when it recently introduced a doctrine of equivalents into UK patent law. Is it about to throw another cat among the pigeons in the pregabalin case? Watch this space for news of the reported judgment as soon as it is handed down by the Supreme Court.