From March to September 2016 the team is joined by Guest Kats Emma Perot and Mike Mireles.

From April to September 2016 the team is also joined by InternKats Eleanor Wilson and Nick Smallwood.

Monday, 9 November 2015

Something to be cheerful about? Immunotherapy patent survives pessimistic mindset

One of the great advantages of bifurcated patent litigation [perhaps the only advantage, wonders Merpel] is that bifurcated decisions make for shorter judgments and shorter blogposts to explain them. Merck Sharp & Dohme v Ono Pharmaceutical [2015] EWHC 2973 (Pat), decided on 22 October by Mr Justice Birss in the Patents Court, England and Wales, is a decidedly unbifurcated ruling, being a full-blooded infringement-and-invalidity action running to 243 paragraphs.  Fortunately the IPKat has been able to count on this handy analysis by Laura Thompson (Lexis PSL IP & IT team). Laura asks: what does this latest immunotherapy patent case tell us about the law of insufficiency and obviousness and the increasingly pervasive question of plausibility?

Background

Ono owns a patent relating to immunotherapy -- specifically the use of anti-PD-1 antibodies for treating cancer (and Bristol-Myers Squibb is the exclusive licensee). Ono developed such a monoclonal antibody, nivolumab, branded Optivo. Merck's version is a monoclonal antibody, pembrolizumab (branded as Keytruda). Each company had marketing authorisations for various indications, and is conducting clinical trials for other indications. The patent contained a Swiss form claim (claim 1), and an EPC 2000 claim (claim 3). For the purposes of this case they were treated as the same. Very basically, PD-1 is a type of receptor protein on a cell surface and PD-L1 is a type of ligand--a protein which binds with this receptor to trigger a signal that could lead to a reaction inside the cell. The claims were for use of an anti-PD-1 antibody inhibiting PD-1’s immunosuppressive signal, in a medicament for treating cancer.

Outcome

Birss J found that the patent was valid and would be infringed. The invalidity attacks covered almost the full range of grounds: obviousness, lack of novelty, lack of entitlement to priority, and added matter, as well as insufficiency.

There was no dispute that, if the patent were found valid, Merck's proposed dealings in its product for treating metastasised or non-removable melanoma would infringe. Unusually, Ono said that if it won its infringement case, it would not seek an injunction (in the UK) due to the life-saving nature of these treatments, provided that it could have future damages on the basis of an appropriate royalty, to be agreed or decided by the court.

Common general knowledge ('CGK')

While many CGK propositions were agreed between the parties, there was still much uncertainty surrounding this area of science at the priority date. It is important to cover some of the facts about the CGK because of the effect they had on the skilled team’s mindset and outcome of the invalidity arguments. For example, it had been established that binding the ligands PD-L1 or PD-L2 to receptor PD-1 inhibits T-cell activation, but in some quarters, it had been found this binding stimulates, rather than inhibits, T-cell activation. So there was still controversy over the reaction pathway which would take place once the ligand and receptor had interacted. This issue was important because, for example, the idea of stimulating T-cell immune responses to tumours by activating stimulatory receptors was already known. This idea had been used in tumour vaccines and high dose IL-2 (a different ligand) treatment in metastatic melanoma and renal cancer. It was also known that some cancers, including melanoma, respond to immunotherapy.

CGK takes brains ...
It was known that it is not enough for antigen receptors on a T-cell to recognise an antigen – a further co-stimulatory signal has to be generated in order to fully activate the T-cell to produce a useful immune response. Some papers had reported that ligand PD-L2 binding generated a co-stimulatory signal whereas others reported PD-L1 was doing this. The patent claim, however, is directed to the PD-1 receptor, not to the PD-L1 or PD-L2 ligands. Still other papers suggested the possibility that the PD-1 receptor itself might, once activated, generate either an inhibitory or co-stimulatory signal, depending on the circumstances. Merck said the skilled team would not have been deterred from trying an experiment to inhibit PD-1 so as to turn off the inhibitory signal and see if this would be useful in treating cancer.

Legal effect of uncertainty in the common general knowledge 
Much of the cross-examination and the judgment was devoted to these scientific issues and all the details are not reproduced here. Above is only an attempt to summarise enough facts to illustrate that there was some uncertainty within the CGK. The judge explained that such uncertainty in a scientific area does not stop the knowledge being part of the CGK, so long as the area of debate is well-defined (see paras [24] and [87]). Birss J found that the skilled team would have known in general the PD-1 pathway was inhibitory but it was also CGK that there was a discrepancy because of the possible co-stimulatory effect of the ligands.

How did the Court tackle lack of priority?

The judge explained that the insufficiency and lack of priority arguments, as well as the Agrevo obviousness (lack of technical contribution) arguments together covered a lot of the same ground. He then found that the priority document did support the claims and that the skilled addressee would clearly and unambiguously derive a disclosure of the idea of using an anti-PD-1 inhibitory antibody to treat cancer. The fact that other things were also disclosed did not take away from this relevant disclosure (see para [140]).

Mouse test
Important parts of the document were mouse tests in examples 4 and 5. Example 4 was a test on mice, some of which expressed PD-1, while others had the gene which encodes for the PD-1 protein disabled, and had PD-L1 cells transferred into them. Example 5 was a test of an anti-PD-L1 antibody; the myeloma tumour cells in this example showed PD-L1 expression and the anti-PD-L1 antibody caused reduction in tumour volume. The patent therefore stated that inhibition of PD-1 or PD-L1 was effective in treating cancer.
Claim breadth 
As well as deciding there was a clear and unambiguous disclosure, Birss J considered priority from the point of view of plausibility and claim breadth. As to claim breadth, Merck argued that the teaching of the priority document was narrower than what was claimed in the granted patent: the priority document only taught treatment of tumours which expressed PD-L1 because there was no data in the document about tumours which did not express PD-L1.

Plausibility 
On plausibility, Merck's main argument was that, on the basis of the priority document, the skilled team would not have been able to predict that all or substantially all cancers would be susceptible to anti-PD-1 therapy. Since the claims would not be enabled across their breadth by the priority documents, the claim scope was not plausible. The judge disagreed: priority would not be lost due to no disclosure in the priority document of an anti-PD-1 antibody being generated or tested. It was known that some cancer cell lines expressed PD-L1. He decided the mouse tumour model, along with the rest of the teaching in the priority document, made it plausible that an anti-PD-1 antibody would be successful in treating cancer, irrespective of whether tumour cells expressed PD-L1 (or PD-L2) or were otherwise known to be immunogenic (para [137]).

According to Ono:
* Examples 4 and 5 of the priority document contain an in vivo tumour model: the results showed that inhibiting PD-1 function reduces or can eliminate tumour growth (for the types of cancer used in the tests) and expressly teaches that the anti-PD-1 inhibitory antibodies would be expected to have a similar effect. Given this, the fact an anti-PD-1 antibody is not made or tested in the priority document does not matter (para [125]). As mentioned above, Birss J agreed with this point – see para [142]

* the priority document is not limited to treating tumours which express PD-L1: the skilled person would not have believed the expression of PD-L1 on the tumour was required in order to inhibit tumour growth. Birss J also agreed that the priority document is not only making treatment of PD-L1-expressing cancer cells plausible (see para [145] and [155])

* the experiments set out in the priority document are evidence that blocking PD-1 inhibits tumour growth in two different types of cancer. The skilled person would realise these results would have broad application in treating cancer because the blockade of PD-1 treats the immune system, rather than being directed to an attribute of any particular type of cancer. This makes the idea plausible that the invention is effective for treating a variety of cancers (and this is corroborated by post-priority date evidence).
The judge added that the skilled person would not have identified any type of cancer which it was implausible that anti-PD-1 antibodies would treat. It was fair on the basis of the mouse models to predict such an antibody might well work and would be worth testing, and he drew attention to all of the types of cancer for which anti-PD-1 antibodies are being investigated – see para [163]. Merck also constructed an argument that some unfavourable data in other mouse tests had been excluded from the patent and adverse inferences should be drawn. Birss J dismissed this attack (see paras [146]-[148]).

How did the Court tackle insufficiency?

Merck argued:
* the claims are excessively broad, because they cover any and all types of cancer but the specification does not make it plausible that all cancers can be treated by an anti-PD-1 antibody

* the patent is not an enabling disclosure of an anti-PD-1 antibody which is suitable to treat all cancers--the most it does is set an unduly burdensome research project to find one

* the patent does not disclose a principle of general application across all antibodies falling within the scope of the claims, so the technical contribution does not match the breadth of the claims
Plausibility 
Birss J comprehensively reviewed the law on plausibility in relation to insufficiency (as well as the role of plausibility in relation to other grounds of invalidity), in light of various cases. However, at the same time, he seemed to caution against applying those cases too slavishly. For example, compared to Human Genome Sciences Inc v Eli Lilly [2011] UKSC 51, noted by the IPKat here (see paras [136]-[137]), the therapeutic effect in the Ono patent is a functional technical feature of the (Swiss form) claim, whereas in HGS, the court was dealing with a product claim (see para [136]). In addition, he cautioned there is no law of plausibility as such, and just because the concept has proved to be useful in one case does not mean everything said about it in that case necessarily applies to another case.

Support 
Birss J made it clear that for a purpose limited medical use claim, to establish plausibility and therefore sufficiency the supporting disclosure likely needs to be more specific than it does for a normal product claim. This is because the supporting material has to be specific enough to cover the claimed medical effect, but also broad enough to support the claim in terms of the therapeutic ingredient claimed (see para [139]).

Principle of general application 
The judge said the evidence had showed that the antibodies claimed work for substantially all cancers. This is true now and such a generalisation was also supported at the priority date. He said it was reasonable to predict anti-PD-1 therapy would work to treat cancer and this is a principle of general application (para [165]). He accepted these antibodies plainly do not (in the current state of science) work on some types of cancer (such as colorectal or prostate) because patients having those cancers do not benefit from PD-1 blockade. However, this does not detract from the generalisation. In addition, he pointed out that these antibodies may yet turn out to be effective against those types of cancers, for example in some form of combination therapy.

Undue burden/research project 
Birss J also considered the undue burden/research project argument. Merck presented the argument on the basis that, even today, researchers are still trying to find an effective PD-1 therapy for some of the major types of cancer. The judge said there is no undue burden in generating a suitable anti-PD-1 antibody, and this is useful for treating many cancers. The work is intensive, expensive and time-consuming, but this does not mean there is no sufficient disclosure in the patent or technical contribution. The patent is not requiring that the antibody must successfully treat colorectal or prostate cancer.

Merck had also relied on a European Patent Office decision, T 1150/09, citing this case because the Technical Board of Appeal (TBA) had rejected the claim as being insufficient and it was a Swiss form claim to the manufacture of a cancer treatment. The TBA decided the claim was insufficient because it was impossible to predict which cancers could be treated using the invention. The Judge refused to take the analogy, saying in effect these cases depend on their facts.

What about anticipation?

Disclosure and plausibility (again) 
Birss J reiterated the requirements for anticipation are a disclosure, which is enabling. He said the law of novelty does not include a requirement of plausibility, but in some cases plausibility (or lack of it) could affect the question of enablement. This discussion had arisen because of the emphasis on Swiss form claims (brought home again in the Warner Lambert v Actavis litigation [aka the Lyrica case, commented on extensively by the IPKat here]) that the therapeutic effect is a functional technical feature of the claim--ie it has to be plausible that the new medical use for the old substance will work to some extent as therapy for the patient (even if it is not a complete cure), since the therapeutic part is the only new part of the claimed invention. Likewise, for a Swiss form claim to fall due to anticipation, the prior art would have to show the therapeutic effect was plausible

The prior art 
The main prior art item was mentioned in the patent--the patent said the prior art discloses manufacture of an anti-PD-1 antibody that inhibits interaction between PD-L1 and PD-1, but that the prior art does not demonstrate in vivo efficacy against tumours (or suggest it). It was also a citation available for anticipation but not obviousness, due to its date – ie a s 2(3) of the Patents Act 1977 citation, but the law on this did not need to be discussed.

Was it enabling? 
The prior art does contain some experiments in vivo, but not tumour models--instead, they were encephalomyelitis tests (used as a model for multiple sclerosis). Birss J decided the prior art disclosed the idea of using an anti-PD-1 agent, to treat various diseases, including cancer. However, the disclosure was not enabling to the skilled person because the prior art document was contradictory and 'hedged its bets' so that it does not clearly and unambiguously direct the reader what to do to treat cancer, so the claims survived the anticipation attack. He explained that while the prior art gave the general idea of using an agent which acts on the PD-1 signal transduction pathway in some medical context, the document does not make it plausible that an an-ti-PD-1 agent would treat cancer. He said this was an example of a piece of prior art which is at the same time neither broad enough nor specific enough.

The second piece of prior art was related to the first, and while it was slightly clearer, it still did not amount to an enabling disclosure.

Inventive step: was there any?

Technical contribution 
Agrevo obviousness (lack of technical contribution) arguments failed and had a lot of overlap with the insufficiency arguments set out earlier (see the 3 bullet points of Merck’s arguments on insufficiency, above).

“Obvious to try” and plausibility (yet again) 
This was also yet another “obvious to try” case. In addition to citing the usual cases on the general law of obviousness, Birss J pointed out in passing the relationship between plausibility and obviousness, which also arose in Generics v Yeda [2013] EWCA Civ 925, noted by the IPKat here. He said the point here is that if a technical effect is not plausible across the full breadth of the claims, it cannot be taken into account in assessing obviousness (see para [208]).

The main prior art discussed on obviousness (which was not part of the anticipation case) dealt with PD-L2 as well as PD-L1 and said the PD-L-PD-1 pathway could be key in autoimmune disease. It made other statements such as that 3 out of 4 of the human breast cancer cell lines examined showed expression of PD-L1, and suggested the pathway could be a attractive therapeutic target, as well as proposing blockade of the PD-1 pathway. The judge found the prior art contained no express or implied disclosure of an anti-PD-1 antibody inhibiting the immunosuppressive signal of PD-1 in a medicine for treating cancer.

What was the expectation level for success? 
There was less dispute about what could be tried on the basis of the prior art disclosure -- eg the possibility of making anti-PD-1 antibodies, and setting up tests using a mouse tumour model. The bigger question was about the attitude to these attempts: was there a sufficient expectation that these would be successful which would make the skilled person want to undertake the tests in the first place? The judge agreed the mouse tumour model experiments would not be difficult to perform, if motivated by a sufficient expectation of success. In short, he found that, if a skilled person decided to carry out the test, that person would be doing it with a hope rather than an expectation of success, so the invention was not obvious.

On the expert evidence, he had found the mouse model results in Examples 4 and 5 of the priority document had been exciting to skilled people and he inferred from this (and other factors) that there would not have been the requisite expectation of success in tests based on the prior art. In contrast, Merck had argued a squeeze: if the patent were entitled to priority and sufficient based on Examples 4 and 5, it must be obvious (see para 218]).

The patent was also not obvious in light of the prior art that had been discussed already in the anticipation attack.

How about added matter?

This point was dealt with shortly. The added matter arguments were similar to the priority arguments. Merck argued that if the prior art did not anticipate the claim, the claim must add matter. Merck argued the patent application as filed contains disclosures which are wider than claim 1. This is due to:
* the choice of targets--PD-1 is claimed, whereas PD-1 as well as PD-L2 or PD-L1 were in the application as filed.

* treatments--cancer is claimed, but cancer, infection and other T-cell disorders were in the application as filed, and

* therapeutic substances--anti-PD-1 antibody is claimed, but many antibodies and other thera-peutic chemical compounds are in the application as filed. For example, anti-PD-L1 antibodies are also mentioned throughout the patent description
Birss J said one of the ideas disclosed in the application as filed, is the use of an inhibitory anti-PD-1 antibody to treat cancer, and a claim to this does not add matter. He was not convinced by the argument about selection from a list either, so the added matter attack failed.

Comments

Post-dated evidence

Birss J's view was that the post-dated evidence corroborated the plausibility at the priority date of anti-PD-1 antibodies being able to treat a wide variety of cancers. See, for example, paras [159] and [163]. Post-dated evidence can be taken into account, if it supports (rather than compensates for the lack of) an earlier established position. The use of post-dated evidence was covered in some detail in Generics t/a Mylan v Yeda [2013] EWCA Civ 925, as well as Eli Lilly v Janssen [2013] EWHC 1737 (Pat), in the context of insufficiency.

Declining to request an injunction

It was unusual that Ono claimed no injunction (in the UK, and as long as it could have damages on the basis of a royalty). Could there be more claims for relief like this in future -- particularly in a consolidating industry, and where collaboration is perhaps more prevalent than at other times in the past. If all other legal factors were equal for this patent and others litigated on a similar basis or the validity decision was on a knife-edge, might such patents have more chance of survival, since the patent's continued existence means (unlike the usual course of events after an infringement finding) that the competitor will not be excluded from the market? The answer at the moment appears to be no, because in para [2] the judge pointed out that Ono's stance on relief does not mean the patent should be judged by the different standard. Wider commercial and other factors are often not revealed in judgments, but for omnivorous patent lawyers who work across different sectors, it is perhaps interesting to go from seeing the long-standing debate about allowability of injunctions in respect of patents essential to a technical standard in the ICT industries, to seeing a voluntary stance on an injunction in respect of a treatment which is essential to prolonging life or the quality of it.

Mindset and expectation of success

This case has been more successful for the patent owner than some others in relation to immunotherapy so far. This seems partly due to the finding that while the attitude of “gloom” was starting to lift slightly, there was still a “history of pessimism and failure” which stayed in the skilled person's mind at the priority date (2002) in relation to immunotherapy for cancer. It was noted that the idea of trying to use or trigger the patient’s immune system to attack cancers was old, but trying to put this idea into practice had often failed. In other areas there had been successful antibody treatments for cancer--not least, drugs like Herceptin—but the judge explained this is a different type of treatment to the one under consideration in this case (see para [32]).

As happened in Regeneron v Genentech [2013] EWCA Civ 93, noted by the IPKat here, the difference between hope and expectation of success could be a fine line; but, back to the present case, it seems that establishing a mood of pessimism among skilled people at the priority date could certainly help towards turning a case in favour of the patent owner.

In a different context (solvents for active ingredients for treating psoriasis), in Leo v Teva [2014] EWHC 3096 (Pat), noted by the IPKat here, which he brought up in the current case, Birss J had found that some non-aqueous solvents were well worth investigating, but the Court of Appeal drew attention to the fact this did not equate to a fair expectation of success, and indicated that the Judge had decided that including a solvent as part of a research project amounted to obviousness (see para [32] of Teva v Leo [2015] EWCA Civ 779, noted by the IPKat here). In the current case, Birss J made the hurdle of fair expectation higher, but in part this is probably due to the nature and complexity of the testing and the broader science.

Amendment 
There was an application to amend the claims to narrow them down to treatment for melanoma. Merck accepted such amended claims would be novel, entitled to priority, and sufficient, but said this invention would still be obvious. The application was conditional and so did not need to be carried out since Ono had won on all grounds in the Patents Court.

5 comments:

MaxDrei said...

Mindful that other jurisdictions take a more liberal line than the EPO, on permitting an amendment during prosecution, to insert an intermediate generalisation, in the novelty section of this Decision I pick out the sentence:

"He said this was an example of a piece of prior art which is at the same time neither broad enough nor specific enough."

and ask whether this is a vindication of the EPO's hard line.

THE US anon said...

MaxDrei,

I do not think that you are asking the right question.

"Vindication" sounds as if one sovereign's choice is "wrong," while another's is "correct."

There is no such thing as a "correct" One World Order choice - each sovereign is correct unto itself.

Dave said...

I can't find that sentence anywhere in the judgement. Are you referring to para. 202? This states:

Dana Farber 557 does not contain a cancer model studied using an agent which can be unambiguously identified as an anti-PD-1 agent. Merck relied on example 15 and on the common general knowledge itself, but I am not satisfied that either of these, or both of them, is enough to make up for the absence of such a model. Overall in my judgment [sic] the content of Dana Farber 557 is not specific enough in one way and not broad enough in another. It is not specific enough to render plausible the use of an anti-PD-1 agent for the relevant disease. On the other hand, while the content may be sufficiently broad to render plausible the idea of using an agent which acts on the PD-1 pathway in medicine generally, and in that sense is similar to the situation in HGS; the content is not broad enough in its application or, looked at another way, not specific enough for cancer, to render plausible the use of that agent in the treatment of cancer.

Misquoting a judgement out of context makes it look like you are setting up a straw man...

Do you think an EPO BoA would have judged this issue differently, and if so why?

MaxDrei said...

Dave, readers, my apologies. I quoted a sentence from Jeremy's report above, not from the Decision itself, but in my over-hasty posting gave readers the impression that my quoted words can be found in the Decision itself.

Sorry Dave, I am not a pharma specialist so cannot opine on whether DG3 would decide novelty different from Birss J. Perhaps other readers can.

Paradoxical said...

I felt a bit sorry for Birss here as I think Jacob's Court of Appeal judgment in Teva v Leo is stifling a possible debate on how we look at reasonble expectation of success. Here there was controversy in the art about inhibitory/stimulatory effect of PD-1. It could have easily been resolved by injecting a cancer mouse model with anti-PD-1, i.e it was binary, but unpredictable. The problem to be solved was not treating cancer, but in reality resolving the controversy. How should we look at reasonable expectation of success in this scenario? It's Jacob's fault Birss could not do anything more complex as the Teva v Leo judgment does not set up the framework for that, which it should have done. In Teva v Leo we have a 2D sort of choice because it is about choosing a solvent and then predicting whether it will work. In that scenario I think reasonable expectation of success should be a high hurdle, but here we are not really choosing the therapy, we are merly asking whether it works, which is a 1 dimensional choice.

I also feel sorry for Birss because he has to follow what the EPO is doing on plausibility, and so he found that unless the prior art has plausible teaching it is not citable for novelty even though there was an explicit statement to administer anti-PD-1 to treat cancer. I think that is controversial step in the case law, but it is inevitable now that the EPO is thinking like this and what the UK courts think of it will clearly be irrelevant.

Subscribe to the IPKat's posts by email here

Just pop your email address into the box and click 'Subscribe':