How much of an expectation of success do you need?

In a recent decision1, the Court of Appeal has upheld the first instance decision of Birss J2 finding two Genentech formulation patents for Herceptin invalid for obviousness. In doing so, the Court of Appeal has considered important aspects of the “obvious to try” assessment, including the need for motivation, the application of the recent decision in Teva v Leo3 and the use of the "could/would" approach (as often applied in the EPO). It emphasises the multi-factorial nature of the test, and that there is no "one size fits all" test. In addition, the history of the patentee’s development is only helpful if it is the subject of proper evidence.

Background

This case is the second in a trilogy of cases forming Hospira’s strategy to clear the way in respect of Genentech’s patents surrounding Herceptin (also known as trastuzumab, its blockbuster monoclonal antibody cancer drug).

Interestingly, when Hospira commenced these proceedings, it was partnered with Celltrion4 but has since been acquired by Pfizer in a $17bn deal announced in September 2015. Following the deal, Hospira reportedly returned to Celltrion all rights relating to its Herceptin biosimilar (which had already been authorised in South Korea (Herzuma), though no application appears to have been filed in Europe). Presumably the reason for this return was as Pfizer would continue to progress its own Herceptin biosimilar candidate (PF-05280014), which is in late stages of development/Phase III. Pfizer therefore still has an interest in clearing these patents out of the way, subject of course to the development timeline of its own product.

In the UK cases, Hospira has had a full house of victories thus far. It was successful in the first case, also before Birss J (in April 20145), revoking two patents, relating to Herceptin’s dosing regimen and a method of antibody purification respectively. Genentech appealed that decision (only in respect of the dosing regimen patent) but Birss J’s decision was upheld (in February 20156).

Hospira was also successful in the third (and final it seems) case, that time before Arnold J (in June 20157) in which a patent concerning the combination of Herceptin and a taxane for the treatment of HER2-positive breast cancer was found invalid for obviousness. It is not clear whether this decision is under appeal.

In this case, the second action, Birss J had found that two of Genentech’s patents relating to stabilised formulations consisting of lyophilised trastuzumab were invalid on the grounds of obviousness and added matter. The decision considered here is the outcome of Genentech’s appeal, which was again unsuccessful. Floyd LJ, giving the leading judgment, dealt only with obviousness reasonably briefly (and having done so added that “the remaining issues do not arise”, thereby dispensing with added matter).

Obviousness

The obviousness case at first instance succeeded based on prior art called Carter. This became the sole subject of the appeal (Hospira’s other attacks based on another document, Draber, and common general knowledge alone failed and were not appealed).

Carter comprises a pair of documents disclosing that trastuzumab was in phase II clinical trials for breast cancer, which Birss J held would motivate a skilled team (a clinician and a formulator) to produce a lyophilised formulation, thereby reaching the claimed formulation by what was “simply an example of the work product of the necessary screening programme to find a satisfactory combination of excipients”.

In this decision, Floyd LJ began with the overview of the law of obviousness, by describing the multi-factorial approach, as detailed in the leading judgment by Kitchin J in Generics (UK) Ltd v H. Lundbeck8:

“The question of obviousness must be considered on the facts of each case. The court must consider the weight to be attached to any particular factor in the light of all the relevant circumstances. These may include such matters as the motive to find a solution to the problem the patent addresses, the number and extent of the possible avenues of research, the effort involved in pursuing them and the expectation of success.” (emphasis added)

The non-exhaustive examples described by Kitchin J (as he was then), as highlighted above, nicely frame the main submissions and points considered in this decision.

Motivation

Genentech submitted that there were two errors made by Birss J in respect of motivation.

First, while it was accepted that the prior art, Carter, made it at least credible that trastuzumab would be effective, Genentech submitted that there was insufficient motivation for the clinician to go on to ask a formulator to formulate trastuzumab. Genentech relied on the judge’s finding that trastuzumab being in a phase II trial was not part of the common general knowledge (and nor was the existence of any phase III studies). Reviewing the evidence and Birss J’s judgment, the Court of Appeal rejected this argument and was satisfied that it was common general knowledge that the HER2 receptor was a significant potential target for cancer therapies, and clinicians would have been interested in any potential therapies.

Second, Genentech had relied on one of its own disclosure documents used in cross examination to argue there was insufficient motivation without the realisation of the problem of degradation in the liquid formulation, which would be tried first. Genentech said that it experienced such issues in its own scaling up process development. However, the Court of Appeal considered that this was not a point which featured largely at the trial and was accordingly considered by Birss J only briefly. The Court of Appeal also pointed to the lack of evidence from Genentech’s employees to mount a case that the liquid formulation was stable before being scaled up, saying that one document in cross-examination was not enough. While Genentech did in fact switch from a liquid to lyophilised formulation, the Court of Appeal did not find the evidence strong enough to overturn Birss J’s decision in this respect.

Expectation of success (and effort involved in pursuing possible avenues)

These factors are considered together and indeed often overlap, as a larger number of possibilities and greater effort usually of course dampens expectation of success.

In particular, the assessment of expectation of success was a central pillar to this appeal and was examined closely in the Floyd LJ’s introduction to the law, including by reference to another Kitchin LJ judgment, this time Medimmune v Novartis Pharmaceuticals9, emphasising that the singular statutory question is whether the invention was obvious at the priority date. Floyd LJ also summarised Kitchin LJ’s passage on expectation of success in Medimmune (at [91]):

“…whether an approach has a reasonable or fair prospect of success is itself another multi-factorial assessment, where this court should again pay proper respect to the evaluation of a trial judge”.

This summary emphasises not only the Court of Appeal’s respect for the trial judge’s closer view of the facts in assessing obviousness, but also the lack of a one-size fits all approach in considering expectation of success under ‘obvious to try’. This became a key theme in this appeal.

Floyd LJ went on to emphasise that the “more or less self-evident approach that what is being tested ought to work” in Saint Gobain10, often cited as a threshold for finding something obvious to try, is not always appropriate, saying furthermore that “is far from being a test of universal application”.

Floyd LJ travelled further through the authorities, and took a similar approach to another often cited passage concerning the "could/would" approach, namely the judgment of Jacob LJ in Actavis v Novartis11. Considering this case, Floyd LJ said that emphasis on ‘would’ should not be taken too far, adding that he did not accept (and did not think that the court in Actavis v Novartis was accepting) that it must be established in every case that “the skilled person would necessarily have arrived at the precise combination claimed” as to do so “would be to place another straitjacket on the law of obviousness”. In doing so, Floyd J emphasised that there are cases where a number of obvious routes exist and even though the claimed route might, statistically speaking, not necessarily be taken, that does not detract from the fact that all of the routes, including the claimed one, are obvious.

The Court of Appeal described Genentech’s submissions on this issue as their major point, and considered whether Genentech were right to say that the judge had erred in principle because he did not sufficiently account for the task of formulating proteins being difficult and unpredictable and one that often encountered dead ends. In particular, Genentech sought to rely on the lack of pointers to the specific combination claimed. Floyd LJ’s introduction to the law had already foreshadowed that this was unnecessarily narrow approach, as he went on to find.

Genentech’s points were considered by the Court of Appeal in three parts: (i) the general expectation of success; (ii) the comparison to Teva v LEO, which Genentech relied heavily on; and (iii) the application of the could/would approach.

Considering the general point, the Court of Appeal were content that the general view of the skilled team would be that they “would expect to be able to produce promising candidates”. Therefore, the Court of Appeal held that the lack of surprise by the skilled team in being successful, as Birss J considered they would be, was simply further confirmation that they had a sufficient expectation of success at the outset. The screening approach needed, using known excipients, was not considered to be inventive and there was no evidence that trastuzumab presented any special problems of its own. Therefore, the Court of Appeal rejected Genentech’s general submission.

Moving on to the second heading, having found the above, it was clear the Court of Appeal viewed this case differently to that of Teva v LEO, in particular Floyd LJ considered that this case did not have such a general perception by the skilled team that they would not be able to succeed in formulating the combination, using known excipients or otherwise.

Lastly, in considering could/would, Floyd LJ emphasised that you do not always need a “would be done”. It was common ground that an arbitrary selection only needed a “could be done” to be obvious. In this judgment, the Court of Appeal arguably extended that principle to empirical fields if the circumstances apply. In doing so, the Court of Appeal restricted the application of the could/would test in this case, commenting that in some cases it will be seldom be possible to predict that any individual experiment will work. In this case, it appears that the overall degree of confidence in the skilled team was sufficient that it carried them through non-inventive work, without needing to consider specifically whether they would reach the specific combination claimed.

Comment

Overall, Floyd LJ’s judgment has important implications to both procedural and legal aspects of obvious to try.

On the procedural side, the Court of Appeal’s message to patentees is that if you wish to rely on facts from the original development surrounding an invention, you need to lead evidence of what actually happened; using disclosure documents in cross examination is not enough. A similar message was given recently in a another case before the Patents Court, by Carr J in his judgment in Cubist v Hospira12. In that case, the judge considered a defence to obviousness submitted by the patentee arguing that if the relevant steps were obvious, they would have been taken by the inventor’s predecessor (Eli Lilly and Company). However, they were not and in fact the patentee (in consultation with Eli Lilly) did not find the development (of daptomycin) at all straightforward. Carr J referred to this as a “powerful point”, but rejected it and in doing so indicated that he had “borne in mind that relevant witnesses with personal knowledge of the development, both from Lilly and Perseptive [another involved party], were not called by Cubist” adding that it had accordingly “not been possible to gain a real understanding of the decisions made and the reasons for making them”.

On the legal side, the message could not be clearer that the UK Courts continue to resist the application of any one size fit all tests for obviousness. The only approach which continues to pervade nearly every assessment is the stepwise approach in Windsufing/Pozzoli, as Birss J followed at first instance in this case. This is likely in part due to the flexibility allowed within the framework of its approach.

In this respect, perhaps the most pertinent comment was Floyd LJ’s reference to the following quote:

“Cases, so far as regards the law, are most useful, but when they are applied to particular facts, they, as a rule, are of little service. Each case depends on its own particular facts and the facts of almost every case differ.”13

In some ways, the above would suggest that reliance on detailed judgments on obviousness and how they are applied to facts, no matter how similar, should be resisted and may ultimately be futile, as was the case here for Genentech’s reliance in this case on Teva v LEO. Accordingly, trends on obviousness may continue to sway in an unpredictable manner, dictated, as they must be, by the facts specific to each case.

1_Hospira v Genentech_ [2016] EWCA Civ 780, 27 July 2016.
2_Hospira v Genentech_ [2014] EWHC 3857 (Pat), 21 November 2014
3[2015] EWCA Civ 779
4A partnership which led to the launch of the first Remicade (infliximab) biosimilar drug in Europe.
5[2014] EWHC 1094 (Pat)
6[2015] EWCA Civ 57
7[2015] EWHC 1796 (Pat); [2016] R.P.C. 1
8[2007] EW HC 1040 (Pat)
9[2012] EWCA Civ 1234, [2013] RPC 27
10_Saint Gobain PAM SA v Fusion-Provida Ltd_ [2005] EWCA Civ 177
11_Teva UK Ltd v Leo Pharma A/S_ [2015] EWCA Civ 779; [2016] RPC 5
122016-EWHC 1285(PAT) / 2016 WL 03180966 - in particular see at [301].
13Quoting what Lopes LJ said in Savage v Harris & Sons (1896) 13 RPC 364 cited by Sir Donald Nicholls VC in Molnlycke AB and another v Procter & Gamble Ltd and others [1994] RPC 49 at 114.