Tuesday, January 9, 2024

What Is the Evidentiary Threshold for Denying a Permanent Injunction on Public Interest Grounds?

AbbVie Corporation v JAMP Pharma Corporation 2023 FC 1520 McVeigh J

2,504,868 / 2,801,917 / 2,904,458 / adalimumab / HUMIRA / SIMLANDI

As discussed in a previous post, in this decision McVeigh J denied a permanent injunction to AbbVie even though she found that its 458 patent was valid and infringed. This is very unusual. While it is clear that an injunction is a discretionary remedy, it is also well-established that a permanent injunction will only be refused in “very rare circumstances” (Valence v Phostech 2011 FC 174 [240]), as McVeigh J acknowledged [642]. Indeed, I believe this is the only decision apart from Unilever (1993) 47 CPR(3d) 479 (FCTD) to entirely refuse a permanent injunction to a successful patentee. In Jay-Lor 2007 FC 358 [263], Snider J declined to grant a permanent injunction on the basis that it was unnecessary, as the defendant had not manufactured an infringing product for over two years; and in Janssen v Abbvie 2014 FC 489 Hughes J granted a partial injunction, for reasons I will return to. There have also been a few cases in which the permanent injunction was tailored with a short run-off period in which infringing sales or delivery was allowed: Janssen-Ortho v Novopharm 2006 FC 1234 and Weatherford 2010 FC 667; in the UK see similarly Virgin Atlantic v Premier Aircraft [2009] EWCA Civ 1513.

As a matter of law, property rights of all kinds are normally protected by injunctive relief. Why? That question was famously answered by Calabresi & Melamed in one of the most cited law review articles of all time: “Property Rules, Liability Rules, and Inalienability: One View of the Cathedral” (1972) 85 Harv L Rev 1089. Calabresi & Melamed argued that when a right is protected by injunctive relief the owner of the right can set the price at which the right is purchased in a voluntary negotiation, whereas when a right is protected only by damages it is the court that determines the value of the right. The advantage of property rights is that the parties will always know the true value of the right better than the court; the advantage of liability rules is that transaction costs and related problems, such as holdout, may mean that a voluntary negotiation does not reflect the true value of the property.

With that in mind, we can divide cases in which an injunction is properly refused into two broad categories. One is where the grant of an injunction would allow the patentee to extract substantial sunk costs. The classic article is Lemley & Shapiro, “Patent Holdup and Royalty Stacking” (2007) 85 Tex L Rev 1991. This is well-recognized as the main issue in the FRAND context, as well as with patent assertion entities (PAEs) aka ‘patent trolls.’ As a result, injunctions are regularly refused to PAEs in the US. Tailoring of injunctions to allow a runoff period or to allow for redesign, such as in Janssen-Ortho v Novopharm and Weatherford, may be an appropriate partial solution to this problem: see generally Contreras & Husovec (eds), Injunctions in Patent Law: Trans-Atlantic Dialogue on Flexibility and Tailoring (Cambridge University Press, 2022) (and see my Canada chapter in that volume for a discussion of Unilever).

The second is a more amorphous category of cases turning on traditional equitable principles, such as clean hands, laches etc. Jay-Lor is an example of the application of the maxim that ‘equity does not act in vain.’ Sometimes these cases can be explained in sunk cost terms: for example, undue delay in bringing an action may give time for an innocent infringer to invest substantial sunk costs in the infringing goods; while the injunction might be refused on the basis of laches, the result would be the same on a sunk cost analysis. This case, in which the injunction was refused on public interest grounds, falls into the second category. While JAMP no doubt has incurred some sunk costs in developing its biosimilar, costs of that nature are regularly incurred by infringing pharmaceutical companies and are never in themselves grounds for refusing an injunction.

When should an injunction be refused on public interest grounds? The premise of the patent system is that the return to the patentee is beneficial to society by inducing innovation. So, on the one hand, the return to the patentee must be protected. Harm to the public interest may be taken into account on the other side of the balance. This does not include harm in the form of higher prices, which is inherent in the patent exclusivity: the premise of the patent system is that the harm from higher prices for a limited term is more than outweighed by the concomitant incentive to innovate. The harm to the public that might warrant refusing an injunction is typically more idiosyncratic, turning on the facts of the case. In the US, harm to medical patients for whom the patentee’s product is not an adequate substitute for the infringer’s product is one of the more common reasons for refusing a permanent injunction, particularly in the context of medical devices: see Seaman “Permanent Injunctions in Patent Litigation After eBay: An Empirical Study,” (2016) 101 Iowa Law Review 1949, 1991.

In Canada, the point is illustrated by Hughes J’s decision in Janssen v Abbvie 2014 FC 489 (discussed here), which also related to HUMIRA. Janssen’s infringing product was STELARA, a biologic used to treat psoriasis, which, in its severe form, can be disabling [17]. While HUMIRA is used to treat psoriasis, and has significant market share, it is not a perfect substitute for STELARA: STELARA operates by inhibiting IL-12, while HUMIRA targets TNF-α [21]. There were also two other TNF-α drugs on the market. A common treatment scenario was for the physician to switch the patient among the TNF-α drugs before going to the sole IL-12 drug (STELARA) [23]. A permanent injunction would therefore have allowed AbbVie to prevent the sale of a drug which AbbVie itself does not supply, and which for some patients is the only effective treatment for a disabling condition. On these unusual facts, AbbVie did not even seek a complete permanent injunction. Instead, AbbVie sought a tailored injunction with an exception for existing patients and restrictions on new patients. The basic idea was to ensure that any patient who could effectively be treated by HUMIRA would be, and STELARA could only be used for patients who responded better to it. This is a classic example of the kind of case in which a permanent injunction might properly be denied for public interest reasons.

In this case, there were two key reasons why McVeigh J refused the permanent injunction. First, AbbVie has licensing agreements with seven other pharmaceutical companies in Canada that offer adalimumab biosimilars [632], [638]; second, JAMP’s SIMLANDI is the only product available as a low-citrate high-concentration (80 mg/0.8 mL) formulation [633]–[636].

The first point goes to the need to preserve the incentive to innovate. In lieu of an injunction, McVeigh J granted a AbbVie a reasonable, running royalty on future sales of SIMLANDI, saying “[t]his rate should easily be determined given the licensing agreements it has with seven other biosimilar pharmaceutical companies” [643].

With respect, the matter is not quite so simple.

There are broadly two reasons that AbbVie might have licensed. First, it might not have wished to enter the Canadian market directly. It is very common for innovative patentees to exploit their inventions by licensing, especially in secondary markets. Royalty rates paid by a voluntary licensee in such a situation are not directly comparable to the reasonable royalty assessed by a court. A voluntary licensee normally performs some kind of value-added services for the patentee, such as product development, reverse tech transfer, clinical trials, marketing, manufacturing or distribution, and may also receive additional value from the licensor, such as supporting IP such as trademarks or trade secrets relevant to the patented technology, or ongoing technical support. The royalty paid by the licensee reflects the value of those services in both directions, as well as the value of the patented technology itself. A compulsory licensee under a running royalty will not generally perform all the same services, though it may perform some. This means the royalty paid by the voluntary licensee cannot simply be applied to the compulsory licensee. Instead, the royalty should be adjusted to account for differences in the services provided by the infringer as compared with those provided by the voluntary licensee. See generally my article with Professor Cotter, A New Framework for Determining Reasonable Royalties in Patent Litigation, (2016) 68 Florida Law Review 929, esp 954ff and Ch 1 Reasonable Royalties § 1.3.6 Comparable Licenses in Biddle, Contreras, Love & Siebrasse (eds), Patent Remedies and Complex Products: Towards a Global Consensus (Cambridge University Press, 2019).

Another complication is that the value of a license will generally depend on the degree of exclusivity. The per unit rate the licensor charges to a single licensee will be higher than the rate it can ask if it licenses to two licensees who compete with each other in the same market. If the licensor does license to multiple licensees, it will typically provide some kind of territorial or product exclusivity to each. One way or the other, unless the market is already perfectly competitive, the entry of a new unrestricted competitor into the market will affect the rate that can be charged to voluntary licensees. So, even if the AbbVie now demands a royalty of $1/mg from its current licensees (to pick an arbitrary number), the entry of JAMP may reduce the value of the licence to those licensees, so future renewals may be at a lower rate due to the reduction in market power. This is a loss to AbbVie that must be compensated in the reasonable royalty paid by JAMP to AbbVie if the incentive to innovate is to be preserved. Further, it is not obvious that the current licenses will reflect some simple rate, such as $1/mg. The licenses may reflect individual bargaining power, or idiosyncratic advantages of the licensees, so that even if the voluntary licensees provide the same services, their royalties may be different.

Another possibility is that the current licenses were negotiated under the threat of litigation. In that case, the royalty rate would be discounted by the parties’ estimate that the patentee would prevail against the potential infringer / licensee in an infringement action. But after litigation, such as in this case, we know with certainty that the patent is valid and infringed, and the reasonable royalty must therefore be higher than the negotiated royalty to avoid the problem of double discounting. As we explained in Ch 1 of Patent Remedies and Complex Products at 22-23:

it is well established in U.S. law that the parties to the hypothetical negotiation are assumed to have known that the patent was valid and infringed, even though actual parties would not. This rule is required to achieve just compensation, because the opposite view – that the parties should be assumed to discount the royalty to allow for the probabilistic nature of the patent (as would presumably be done by parties to an actual negotiation) – would result in so-called double discounting; not only would the court-approved royalty derived from the hypothetical negotiation include a discount for the risk of nonliability, but then pre-litigation negotiations in which royalties were based on the expectation of such a court award occurring with a less than 100 percent probability would include a further discount for risk of non-liability.

Further, the rates in the various licenses might be quite idiosyncratic — there is no particular reason that each ‘generic’ would arrive independently at the same estimate of its probability of prevailing. And these royalties would also reflect the pricing issues with truly voluntary licenses, discussed above. It is also entirely possible that the current licensing situation is a combination, where AbbVie has entered into some licenses on a truly voluntary basis and others in contemplation of litigation.

There are some cases in which it is indeed easy to calculate an accurate reasonable royalty, such as when the patentee offers a licence to all comers on demand at a standard rate. This is clearly not such a case. With that said, it is possible that it will be easy to calculate an accurate reasonable royalty; not all of the difficulties described above will arise in every cases. But the mere fact that there are seven other licensees of similar products certainly does not give any guarantee that the reasonable royalty calculation will be easy.

The other side of the coin is the effect on patients of an injunction. McVeigh J relied on two points.

One is the potential that “non-medical switching” could negatively impact patients, through the “nocebo effect” which could result in a perception of increased injection site pain. It seems to me that if we give the nocebo effect any weight at all, a permanent injunction would never be granted against any infringing pharmaceutical that has actually been launched, as the effect, by definition, does not turn on any real difference in the product, but on the mere knowledge that the product is different [636].

The other point is that JAMP’s formulation is the only high-concentration / low volume, citrate-free product on the market. (It’s not clear to me whether YUFLYMA is on the market, but I’ll assume it’s not.) It is possible that the greater volume and / or the citrate could cause increased injection site pain for some patients, and “[t]hough the evidence is scant for those few patents, it could be very harmful” [635].

This raises two points of law. One is the threshold for the harm to the public interest. Though the point is not normally made explicitly, it is clear enough that we are willing to tolerate some degree of harm to the public beyond the increase in prices. For example, in Valence v Phostech 2011 FC 174, Gauthier JA refused to even tailor a permanent injunction by allowing a two year grace period for the completion of a new factory designed to use a non-infringing process. Presumably, requiring the infringing factory to shut down immediately would entail significant disruption to the workers, such as temporary or even permanent unemployment. (Moreover, in that case a strong argument can be made that tailoring would have been appropriate on substantive grounds, as the infringement appears to not have been intentional and the infringer had incurred very substantial sunk costs.) The question then is how much non-price harm to the public we are willing to allow.

It is not clear what McVeigh J meant by “very” harmful and “few” patients, but this seems to be a far cry from Janssen v AbbVie in which it was uncontested that there were a significant number of patients whose potentially debilitating psoriasis could only be effectively treated by the infringing product. In this case, as I understand it, it is not disputed that the non-infringing alternatives will treat the disorder just as effectively as SIMLANDI, and the harm is increased injection site pain. (Since there are several citrate-free alternatives [638], it seems that the main issue is the increased volume.) In principle, the issue is not the pain from the alternative as such that is important, but the incremental pain of the alternative over SIMLANDI. So, SIMLANDI is available in a 40 mg/0.4 mL pre-filled syringe where most of the alternatives are a 40 mg/0.8 mL — so, double the volume for a 40mg dose. As I understand the evidence (as well as from people I know), many people do not find the 0.8 mL dose to be at all painful or even uncomfortable, so presumably those who do are unusually sensitive to dose volume and might find a 0.4 mL dose to be painful as well, though presumably less so. The harm if the injunction is granted is not the pain to those patients from the 0.8 mL dose, but the additional pain as compared with the 0.4 mL dose. It is not clear to me from McVeigh J’s reasons whether the “very harmful” pain refers to the pain of the larger dose, or the incremental pain. (Perhaps the main harm is to those who are particularly sensitive to both citrate and injection volume, which would imply a smaller subset of affected patients.) Now, perhaps the harm is more significant than intuition suggests. It may be that there are some patients for whom the additional pain is so substantial that they would skip injections and impair their treatment rather than endure the injections with HUMIRA or one of the licensed alternatives. That would align this case more closely with Janssen.

Of course, the degree of harm is a matter for evidence. My point here is that it seems reasonable that there should be some substantive threshold for the nature of the harm suffered in order to warrant denying an injunction on public interest grounds. If evidence established that 10 patients in Canada would suffer arm soreness for two minutes after an injection rather than for one minute, I would suggest that should not be a sufficient harm to warrant refusing a permanent injunction on public interest grounds. From McVeigh J’s brief description, the harm in this case appears to be substantially less, both in its nature and its extent, than the harm at issue in Janssen v AbbVie.

As noted, the extent of the harm is not clear from McVeigh J’s reasons. This raises the second point: what is the proper evidentiary threshold for refusing injunctive relief on public interest grounds? In this case, we do not know how many patients would be affected and how serious the harm is, because the evidence is “scant” and “limited” [635]. All we know is that it “could” be very harmful. When evidence is scant, any degree of harm is possible. Suppose an expert witness was asked whether the injection site pain might dissuade some of patients from following the correct dosage schedule, and the answer was “Anything’s possible.” Would that be enough to warrant refusing the permanent injunction on public interest grounds? More generally, can a permanent injunction be denied on the basis of a speculative harm? Surely the presumption in favour of granting injunctive relief to a successful patentee is strong enough that more than mere speculation should be required to overcome it. The more difficult question is whether the standard should be something like the balance of probabilities, or a lower threshold, such as evidence which is clear and not speculative.

The substantive threshold and the evidentiary threshold are distinct issues. It might be uncontested that 1% of the patients would suffer soreness for two minutes instead of one minute; or one expert might provide unsupported opinion evidence that large numbers of patients would fail to comply to their injection schedule dosage. Given McVeigh J’s explicit acknowledgment of the limited nature of the evidence, it might be easier for the FCA to address evidentiary threshold. It will be difficult to address the question of whether the substantive threshold has been met when the extent of the harm is not actually established in the evidence.

On the whole, the reasons given by McVeigh J strike me as very tenuous grounds for denying a permanent injunction. Further, even if some kind of significant harm to some patients could be established on a non-speculative basis, this does not strike me as a good case for denying a permanent injunction entirely. A tailored injunction, along the lines of that granted in Janssen v AbbVie, which would allow the use of SIMLANDI by the specific patients who cannot tolerate HUMIRA or one of the other licensed biosimilars, would seems to me to strike a better balance.

No comments:

Post a Comment