Spiralling into the abyss? Pharma PIs: lessons from Neurim

It would be uncharitable (but sometimes accurate) to suggest that reading patent judgments is a good cure for insomnia. Something that certainly is useful in battling lack of sleep is the drug melatonin, which has been the subject of recent patent judgments in the UK and Sweden on applications for preliminary injunctions (PIs). In this blog post, we examine those judgments and contrast the different approaches of the courts.     

Background

Neurim's Circadin is a drug formulation containing the active ingredient melatonin, designed to improve the restorative quality of sleep for insomnia patients. Neurim has a 'second medical use' patent that covers the full scope of Circadin's marketing approval, and also some uses falling outside of the scope of the approval. The patent was revoked for lack of novelty by the EPO Opposition Division at an oral hearing on 20 November 2019, and an appeal by Neurim against that decision is pending. In the meantime, the effect of the opposition decision is suspended and the patent remains in force with an expiry date in August 2022.

Several generic manufacturers have obtained marketing authorisations for generic versions of Circadin including, in the UK, Teva and Mylan, and in Sweden, Orifarm. Neurim sought PIs against Mylan in the UK (Teva having agreed not to launch early) and against Orifarm in Sweden. In each case Neurim was unsuccessful (in the UK at first instance and on appeal), albeit for different reasons. 

Pharma PIs in the UK

In the UK, pharmaceutical innovators have typically been able to obtain PIs against companies threatening to enter the market with generic products. Patent holders have relied on the risk of an 'irreversible price spiral' if generic products are allowed to launch before the validity of a patent can be tested at trial, though it is recognised that such a price spiral will (in general) only occur once there are at least two generic products in the market.

The strength of the underlying infringement claim and validity of the patent is at the forefront on a PI application; there is no 'mini-trial'. The court instead applies a four-stage legal test set out in the 1975 case of American Cyanamid:

• Stage 1: Is there a serious question to be tried?
• Stage 2: Are damages an adequate remedy for the claimant?
• Stage 3: If not, are damages (or the cross-undertaking in damages) an adequate remedy for the defendant?
• Stage 4: If damages are not an adequate remedy for either side, where does the balance of convenience lie?

Why did Neurim fail in the UK?

Neurim argued that it would suffer significant losses if Mylan's generic product were to enter the market before the result of the EPO appeal was known. It relied on the “irreversible price spiral” argument, ie that entry of generic products would result in a downward spiral in the product price which could not be raised back to former levels, even if Neurim were to be successful at trial. This would cause substantial and unquantifiable loss, such that damages would not be an adequate remedy. Neurim also argued that, as Circadin was effectively Neurim's sole source of revenue, large parts of its operation would have to be shut down if these revenues were lost.

The Court of Appeal found that the first instance judge was correct to examine those arguments with a critical eye. In this case it was highly relevant that there would be only a very short period of generic competition if the PI was not granted (around four months) because trial of the infringement action had been expedited. In this short period, it was unlikely that a second or subsequent generic product(s) would launch after Mylan, given that Teva had agreed not to launch and no other manufacturer had a relevant marketing approval. The Court therefore held that, on the evidence, this was not a “price spiral” case.

The Court also determined that Neurim would be able to cover any revenue reduction over the four-month period to trial with cash reserves, so there would be no consequential loss stemming from the closure of activities such as research and development or educational programmes, contrary to Neurim's submissions.

While acknowledging that the case involved 'extremely unusual facts', the Court decided that damages would be quantifiable and therefore adequate to compensate Neurim for any lost sales or depressed prices because:

• forecasts for Neurim's expected revenue under a monopoly were available up to the point of patent expiry;
• the Court would have access to Neurim's actual sales figures for the period between the hearing and trial to compare to these forecasts; and
• an extrapolation to determine likely sales and prices would be sufficient to calculate any continuing damages for the period between trial and patent expiry.

Neurim's application therefore failed at Stage 2 of the American Cyanamid test.

Notably:

• The Court of Appeal did not consider the strength of the substantive claim. Generic manufacturers often try to argue that the substantive claim is so weak that Stage 1 is not met (and Mylan did so at first instance here) or, alternatively, that the weakness of the case is relevant to the assessment of the 'balance of convenience'. However, previous cases demonstrate that the 'serious question to be tried' hurdle is met for all but the most egregiously bad cases, and unless the substantive case is palpably weak this won't be factored into Stage 4 either. A good example is the case of Novartis v Hospira (concerning zoledronic acid), in which the fact that the High Court had found the patents to be invalid did not prevent a PI being granted pending an appeal.
• The Court of Appeal did not address the fact that Mylan had made no effort to “clear the way” by seeking to revoke the UK patent. This is not a factor relevant at Stage 2 (or 3) – it is not mandatory for generic manufacturers to 'clear the way', and failure to do so would only come into play at Stage 4.

What was different in Sweden?

The Patent and Market Court also dismissed Neurim's request for a PI in Sweden. By contrast with the UK, the court in Sweden did consider arguments relating to the validity of the patent. Swedish courts apply a two-part test:

• Can the claimant show probable cause that the defendant has infringed the patent?
• Can it be reasonably feared that the value of the exclusive right to the patent will be diminished by the defendant's continued infringement?

Normally the Swedish court presumes that a patent is valid and this presumption must be rebutted by the alleged infringer in order to defeat 'probable cause'. In this case however, the court held that there was no presumption of validity as a consequence of the EPO Opposition Division's decision (even though it is under appeal). The burden instead fell on Neurim to show that its patent was likely to be upheld in the final assessment. On considering the prior art found by the Opposition Division to destroy novelty, the court concluded that Neurim had failed to discharge that burden. 

Neurim's application therefore failed at the first stage for lack of probable cause.

Comment

These cases emphasise the different assessment approaches taken in different European countries on PI applications. While there is limited scope to consider the merits of an underlying patent when evaluating whether there is a serious question to be tried under English law, the threshold for this test is very low. In contrast, the Swedish Court will consider the merits of the case and essentially conduct a 'mini-trial'. 

Most civil law jurisdictions in Europe tend to fall nearer to Sweden than to the UK, but there is a high degree of variation in practice. It is very common to have divergent decisions across Europe on essentially the same facts, and patentees need to be attuned to the differences between jurisdictions of interest.

Some commentators view the English Court of Appeal's decision in Neurim as a death-knell for pharma PIs in the UK. While that rather exaggerates the position, the Court has made very clear that assertions as to the damage that will be suffered by a patentee must be properly supported with evidence, in particular as to what other generics might be in a position to launch before a trial takes place. 

While it can be expected that generic manufacturers will be emboldened by the decision, the fundamental legal test remains unchanged and, as with any application, the strength of the evidence before the court is what makes the difference. Patentees should therefore carefully consider the evidence to be led based on the market characteristics of each individual product, as a 'one size fits all' price spiral cannot be assumed.