Tuesday, June 1, 2021

“[I]t Is Not Obvious That [Windsurfing / Pozzoli] Has Been Useful”

Apotex Inc v Shire LLC 2021 FCA 52 Rennie JA: de Montigny, Gleason JJA affg 2018 FC 637 Fothergill J

            2,527,646 / lisdexamfetamine [LDX] / VYVANSE / NOC

 In Sanofi 2008 SCC 61 [67], the SCC endorsed the four-step Windsurfing / Pozzoli approach to the obviousness determination. In Bristol-Myers 2017 FCA 76 [64] Pelletier JA remarked that “It is true that the Windsurfing / Pozzoli framework does provide structure but it is not obvious that it has been useful.” In this post I will argue that the trouble lies with the Windsurfing / Pozzoli framework itself, which is unhelpful, confusing, wrong and rarely actually applied as it was originally framed.

In Bristol-Myers 2017 FCA 76 [65] Pelletier JA remarked that the obviousness analysis asks “whether the distance between two points in the development of the art” can be bridged by the skilled person. The problem addressed in Bristol-Myers, Ciba 2017 FCA 225 and Vyvanse, has been how to define this “second point.” The second and third steps of the Windsurfing / Pozzoli framework are aimed at defining this second point, and the question is whether it is a helpful way of doing so. I’ll argue that it is not. It works well enough in easy cases—where it is also unnecessary. In the more difficult cases, it is at best an unhelpful distraction and potentially misleading. For the most part, the FC/FCA have used a sound intuition borne of experience to define the end point of the obviousness inquiry, but the courts' success in this respect has been despite the Windsurfing / Pozzoli framework, not because of it.

The first step the Windsurfing / Pozzoli framework is to identify the notional “person skilled in the art” and the relevant common general knowledge of that person. This has been uncontroversial, but it was of course a well-established step long before Windsurfing [1985] RPC 59 (CA). The heart of the Windsurfing / Pozzoli approach lies in the second and third steps, which are as follows:

            (2)       Identify the inventive concept of the claim in question or if that cannot readily be done, construe it;

            (3)       Identify what, if any, differences exist between the matter cited as forming part of the “state of the art” and the inventive concept of the claim or the claim as construed.

These two steps set up the fourth step, which asks whether those differences would have been obvious to the person skilled in the art. So, steps (2) and (3) of the Windsurfing / Pozzoli approach are a framework for defining the second point of the obviousness inquiry.

1.1.1   How is the term “inventive concept” used in the jurisprudence

Much confusion has arisen because the leading cases, including Windsurfing [1985] RPC 59 (CA), Pozzoli [2007] EWCA Civ 588, and Sanofi, use the term “inventive concept” as though it has some definite meaning which is clearly understood, when in fact it is ambiguous and is used in a variety of ways: Bristol-Myers [64]. In a previous post I noted that the quid pro quo for the patent monopoly is information, and the patent will accordingly be invalid unless the information provided in the specification would not have been obvious to a skilled person. I also argued that this requirement is necessary, but not sufficient, but we can ignore that for the moment. For convenience, I will refer to the new information disclosed in the specification as the “objective inventive concept.” By “objective” I am signaling that this refers to the information that is actually new. This is in contrast to what I will refer to as the “subjective inventive concept,” which is the information that the applicant believed was new.

1.1.1.1            Windsurfing

With these definitions in mind, let us consider how the term “inventive concept” is used in the leading cases, starting with Windsurfing [1985] RPC 59 (CA). My apologies for quoting so extensively in the following discussion, but the concept of the inventive concept has caused such confusion that it is important to take care to see how it was used in Windsurfing.

The EWCA set out the steps in what is now the Windsurfing approach as follows (73–74):

The first is to identify the inventive concept embodied in the patent in suit. . . . The third step is to identify what, if any, differences exist between the matter cited as being ‘known or used’ and the alleged invention.”

The fourth step was for the court to ask “whether. . . those differences constitute steps which would have been obvious to the skilled man or whether they require any degree of invention.” (The order of the steps was rearranged in Pozzoli).

The invention at issue was a now-familiar windsurfer— a conventional surfboard fitted with a leeboard, and having a spar mounted to the centre of the surfboard with a universal joint, along with a triangular sail attached to the leading edge of the spar, held taut by two arcuate booms, connected together at each end and connected at their forward ends to the spar (65). The EWCA held that “we respectfully agree with the learned judge that the inventive concept of the patent is the free-sail concept. It is that which constitutes the essential difference between the patent in suit and other conventional vehicles propelled by sail” (74). The free-sail concept in turn (66, my emphasis):

cannot be expressed better than by adopting the language of the learned judge when he said (at page 6 of the transcript) :

"The specification is disclosing the concept and method of constructing a wind-propelled craft carrying a sail which is mounted on a mast free to move in any direction. In contrast with other vehicles with which the reader would be familiar the reader is taught that the mast should not be rigidly secured on the craft but left free to move. It is this freedom of movement which allows the user to control the course of the vehicle in the manner indicated, and indeed in moments of difficulty allows him to let the sail fall so as to take away the propulsive force of the wind. Mast and sail are supported by the user and moved by him so as to steer the craft either when he is running before the wind or when he wants to tack".

That is the essential teaching of the specification.

At the third step, the EWCA compared the invention with the prior art “Darby” article, "Sailboarding: Exciting New Water Sport" which had first appeared in a magazine called Popular Science Monthly.

That article described a sailboard with an unstayed sail attached to a spar held by the user and inserted into a hollow socket in a mast step mounted on the sailboard. The mast step was slotted immediately in front of the spar socket to carry a daggerboard. The basic concept, i.e. the use of an unstayed sail which is used to steer the vehicle and which can be jettisoned in case of trouble, is the same concept as that in the patent in suit. The principal differences are that the sail shown in the publication is a square sail actuated by the user standing on the lee side of it with his back to the spar, and that, instead of the spar being attached to the vehicle by a moveable joint, free movement is achieved by a socket from which the spar can be removed, the suggestion being that the sail should be attached to the sailboard by a lanyard to prevent their drifting apart when the sail is jettisoned. (67, my emphasis)

Thus, as Oliver LJ stated, “[Darby] in fact disclosed, and disclosed to persons knowledgeable in the art, the self-same inventive step claimed by the patent in suit, the only difference of any substance being the use of the kite rig held by the crossed spar and boom instead of a Bermuda rig with a wishbone boom” (74, my emphasis).

At the third stage, identifying the differences between “the matter cited as being ‘known or used’”—namely Darby—and “the alleged invention,” Oliver LJ focused on these differences, namely the use of a triangular sail rather than the square rig in Darby, and an arcuate boom rather than the crossed spar and boom of Darby. Oliver LJ then held that these minor differences would have been obvious to a skilled person.

So, in Windsurfing the EWCA used the term “inventive concept” to mean what I have called the subjective inventive concept; what the inventor thought he had invented. The “inventive concept,” as the EWCA used the term, was the difference between what was disclosed and other “conventional” vehicles—not the difference between what was disclosed and the prior art. Because the focus was on what the inventor thought he had invented, the subjective inventive concept was discerned from an examination of the specification. The Court in Windsurfing did go on to identify the objective inventive concept, at the third step, in identifying the differences between the state of the art and the alleged invention.

So, we can reframe what are now steps (2)–(4) of the Windsurfing test as follows:

(2) Identify the subjective inventive concept

(3) Identify the objective inventive concept

By identifying the difference between the subjective inventive concept and the state of the art

            (4) Determine whether the objective inventive concept is obvious

Is this helpful? The end point of the inquiry, at step (4), is the objective inventive concept. An alternative approach would be to simply identify the objective inventive concept directly, by comparing what is disclosed in the specification with the state of the art and identifying the differences. In comparison, the Windsurfing approach interposes an extra step of first identifying the subjective inventive concept. Perhaps in some cases it is easier to first identify what the inventor thought they had contributed, and then identify what they had actually contributed. But is very difficult to see why it should generally be helpful to add an extra step rather than simply directly comparing the state of the art with the information disclosed in the patent.

I would also suggest that a reason for the confusion as to the meaning of the “inventive concept” has arisen simply from the unusual terminology. It is natural to assume that the inventive concept of the patent is what the inventor has actually contributed, which is, after all, the focus of the fourth step of the inquiry.

Further, the approach breaks down entirely in the more difficult cases where the inventor has approached the problem from the wrong direction entirely. Consider the twist on Bristol-Myers, discussed in my previous post, in which the inventors, instead of pursuing improved bioavailability, were instead motivated by an idiosyncratic pursuit of improved stability. The subjective inventive concept would be a new form of atazanavir with improved stability. This seems to correspond to what the EWCA called the “alleged invention,” given that the patentee will certainly allege that that is what is inventive. The third step then, “is to identify what, if any, differences exist between the matter cited as being ‘known or used’ and the alleged invention” (73–74). But that is the wrong question, as it invites us to focus on whether it would have been obvious to the skilled person that the new form of atazanavir would be a solution to the problem of improved stability, where the proper question is whether it would have been an obvious solution to the problem of bioavailability.

So, it seems to me that the Windsurfing framework is unhelpful, confusing, and, in the worst case, wrong. **

Further, the original Windsurfing approach, which first identifies the subjective inventive concept and then uses that to identify the objective inventive concept, is not usually applied, even in the cases that cite Windsurfing. Indeed, it was not even applied in Windsurfing itself. Recall that the third step is to identify the differences that exist between the state of the art and “the alleged invention” (73–74). Does this mean the inventive concept or the invention as claimed? It seems natural to suppose that it means the inventive concept, seeing that step (2), just completed when we arrive at step (3), is to identify the inventive concept. That is why I reframed the test as above.

However, the inventive concept identified at step (2) was the free-sail concept. Darby disclosed the free-sail concept. Thus there was no difference between the inventive concept and the state of the art. At the third step, what the Court actually did was to identify the differences between the state of the art and the invention as claimed; that was the focus of the obviousness inquiry at the fourth and final step. Thus, the identification of the inventive concept actually played no role at all in the Court’s obviousness analysis! Step (2) of the Windsurfing analysis played no role in Windsurfing itself.

It is worth noting that the main issue in Windsurfing was not the nature of the test for obviousness, nor how to identify the inventive concept, nor anything else for which the case is now famous. The patentee recognized that its case was weak if obviousness was to be assessed on the basis of the technical differences over Darby, and its main argument was that a skilled person seeing Darby would have considered it a “beach novelty” which no skilled person would have any interest in improving (73). The patentee argued that the question was not whether the claimed invention was technically obvious to a person desirous of improving the Darby sailboard, but whether it would have been obvious to a skilled person to want to improve the Darby sailboard in the first place (73). The bulk of the decision on obviousness was addressed to this argument (70–73), which the Court of Appeal rejected. The Windsurfing test was expressly a rejection of this argument; the point of focusing on the differences between the state of the art and the invention was that the question was whether those differences would have been technically obvious to a skilled person. Thus, in Windsurfing itself, the real point of Windsurfing framework was to emphasize that the key question is whether the invention was technically obvious, not whether it is commercially obvious.

1.1.1.2           Pozzoli

In Windsurfing, the inventor was mistaken about what he had invented and there was consequently a difference between the subjective and objective inventive concepts—what he thought he had invented and what he had actually invented. This is not an unusual scenario, typically by reason of having overlooked some of the prior art, as in Windsurfing itself. But the inventor is also very commonly correct about her contribution, and the real question is whether it is inventive. That scenario arguably contributes to the confusion as to the nature of the Windsurfing framework. When the inventor is correct about her contribution, there is no difference between the subjective and objective inventive concept and steps (2) and (3) collapse into one another. In such cases, unless we keep in mind the original distinction on the facts in Windsurfing, it can be very difficult to understand why the test even has two steps, as step (3) becomes trivially easy. The problem is compounded because the leading cases have not in fact been rigorous about using step (2) only to identify the subjective inventive concept, often using the claims, the common general knowledge and the prior art to identify the “inventive concept” at step (2). This runs steps (2) and (3) together conceptually, as well as practically, leading to further confusion as to the distinction between the two separate steps. This is illustrated by Pozzoli and Sanofi.

The invention at issue in Pozzoli [2007] EWCA Civ 588was a box for holding music CDs. To simplify somewhat, prior art boxes either carried two CDs side-by-side, which made them easy to remove, but made for an awkward shape with a 2:1 ratio; or stacked the CDs on top of each other, which was a more convenient shape, but more difficult to remove. The claimed invention provided for a box in which the CDs were partially overlapped in a step-like arrangement, which apparently made for a convenient shape while allowing the CDs to be removed easily.

Pozzoli does not offer any clear definition of the inventive concept, no doubt because Jacob LJ did not feel it was important: “In the end what matters is/are the difference(s) between what is claimed and the prior art. It is those differences which form the “step” to be considered at stage (4)” [19]. This is an explicit statement that the distinction between steps (2) and (3) of the Windsurfing test is not important. Indeed, it implies that carrying out those steps is unimportant, so long as the final analysis focuses on the difference between what is claimed and the state of the art. This is consistent with Windsurfing itself, in which, as just noted, the EWCA itself focused on the difference between what was claimed and the state of the art at the fourth step.

Pozzoli of course modified step (2) of the test to add claim construction as an alternative to identification of the inventive concept: “Identify the inventive concept of the claim in question or if that cannot readily be done, construe it: [23]. However, when Jacob LJ actually got to that stage, it appears that he did identify the inventive concept: he began by saying that they had heard argument on ascertaining the inventive concept first [38], and it appears that that was what he undertook to do [39]. He ended the discussion, which included consideration of the prior art, the specification, and the claims, by identifying something he called a “précis” and the “take-home message from the claim” which was “really no more than ‘overlap the discs, hold them in the known way via their centres yet space them via a step-like arrangement so they can be got out’” [49]. (This was evidently intended as a statement of the inventive concept, as it was introduced with the comment “What then is the inventive concept of claim 1?”)

It is not at all clear how Jacob LJ decided that was the inventive concept. In outlining the test, he stated that the second stage, identification of the inventive concept, “is not the place where one takes into account the prior art. You are not at this point asking what was new” [21]. This implies that he is using the “inventive concept” to mean the subjective inventive concept, as in Windsurfing. However, when he actually came to apply step (2) to the facts, the discussion did actually compare the claimed invention with the prior art [40], [43], without making it clear whether this was based entirely on what the specification said about the prior art. The trial judge expressly considered the prior art in identifying the prior art with “What was new about the patent” ([47], quoting the trial judge at [87]–[88]), and while Jacob LJ refined this, he seems to have largely accepted it. I don’t know how he would otherwise have identified the key aspects of the invention. Jacob LJ’s discussion was otherwise largely descriptive of the invention as described in the specification. The claims at issue did indeed claim a number of other features, eg the disks being axially retained. It is clear from the facts that axial retention was part of the prior art, and for that reason it seems clear that axial retention is not part of the inventive concept, and indeed he described the rosettes as “conventional” [43]. But if Jacob LJ was being serious in saying that “you are not at this point asking what is new,” the fact that axial retention was not prior art is not in itself sufficient reason to exclude it from the inventive concept. In theory, Jacob LJ might have excluded it based on his parsing of the specification, but he doesn’t tell us how, or which passages he relied on to that effect.

Being realistic, it is perfectly clear that Jacob LJ did consider the prior art in deciding what the inventive concept was, even though he said one shouldn’t. In any event, Pozzoli certainly has no clear methodology for identifying the inventive concept that any other court could follow.

Recall that Jacob LJ articulated step (3) as “Identify what, if any, differences exist between the matter cited as forming part of the ‘state of the art’ and the inventive concept of the claim or the claim as construed” [23]. When he got to step (3), on the facts he referred to it as “Differences over the prior art” [Heading to 51], which he described as “the idea of overlapping yet spaced apart rather than side by side (the 2-to1) or concentric (the Brilliant Box)” [51]. The idea of “overlapping yet spaced apart” is essentially how he described the inventive concept.

There are two ways to look at step (2) and (3) in Pozzoli. One way is that at step (2), Jacob LJ identified the inventive concept by comparing the invention to the prior art to see what was new. This corresponds to what I have called the inventive concept. That being the case, it is no wonder that step (3) was “easy”—if the inventive concept is the difference between the prior art and the invention as disclosed, then pointing out that difference between the inventive concept and the state of the art simply requires restating the inventive concept. My impression is that is actually what was going on.

The other way of looking at it is that in step (2), Jacob LJ actually identified the subjective inventive concept by looking only to the specification without regard to the prior art—or at least only taking into account the prior art to the extent that it is referenced in the specification and taking at face value everything that the specification says about the prior art. That must be a difficult task, and certainly Jacob LJ was not scrupulous in “blinding” himself to the prior art. In any event, if the patentee was right about the state of the art, as appears to have been the case, then the subjective and objective inventive concept are the same and so we would get to the same place by step (4).

Ultimately, Jacob LJ focused on the differences between what was claimed and the prior art and decided the differences were obvious. His discussion of the “inventive concept” strikes me as ultimately being a very convoluted way of identifying those differences. The approach in Pozzoli amounted to “Identify the differences between the claimed invention and the prior art. Would those differences have been obvious to a skilled person?” It is not clear to me what Jacob LJ meant by the inventive concept, or what methodology he used to identify it, and so I cannot extract any guidance on either point. In my view, the Windsurfing framework was not helpful in Windsurfing, and the Pozzoli framework was not helpful in Pozzoli.

Finally, I’ll note that Jacob LJ remarked that in some cases “it is not even practical to try to identify a concept—a chemical class claim would often be a good example of this” [20]. Given that Rennie JA in Vyvanse was of the view that a chemical compound claim is an example where it is essential to identify the inventive concept, and Jacob LJ in Pozzoli thinks that is a case where it is impossible to identify an inventive concept, it seems clear enough that they are not using the term in the same way. My own intuition is that Rennie JA is right—the inventive concept of a chemical compound, or class of compound, is the discovery of the use for the compound, or perhaps the method of synthesis if they have a known use. Unfortunately, Jacob LJ’s remark was made in passing, and he did not elaborate, but whatever he might have had in mind escapes me. This does leave a puzzle as to what Jacob LJ meant by the term “inventive concept.”

1.1.1.3           Sanofi

In Sanofi, the invention claimed in the 777 patent was the dextro-rotatory isomer of a racemic compound and the question was whether it was obvious over the prior art 875 patent which disclosed the racemic compound. At step (2) the SCC stated that it was “apparent” that the inventive concept of the claims in the 777 patent is “a compound useful in inhibiting platelet aggregation which has greater therapeutic effect and less toxicity than the other compounds of the ‘875 patent and the methods for obtaining that compound” [78]. The SCC did not provide any express definition of the inventive concept, nor did it tell us how it arrived at this conclusion; did it take a subjective approach looking only to the specification, or an objective approach, comparing the disclosed invention to the prior art? We don’t know.

At the third stage, the differences, as in Pozzoli, were essentially the differences between the claimed invention and the prior art [80]. As in Pozzoli, identification of the differences was easy. Again, we don’t know whether the SCC followed Windsurfing and took a subjective approach at step (2) and an objective approach at step (3), or whether it took an objective approach at step (2), making step (3) essentially redundant.

1.1.1.4           Summary

The Windsurfing test, at least as set out in Windsurfing, applied a two-step approach to identifying the objective inventive concept. At step (2) is the subjective inventive concept—what the inventor thought it had invented. Step (3) identified the objective inventive concept by comparing what the inventor thought it had invented with the state of the art, in order to identify what it had actually invented. Step (4) then asks whether that objective inventive concept is obvious. An alternative to this two-step process would be to simply compare the invention with the prior art, thereby determining the objective inventive concept in a single step, before going on to ask whether it is obvious.

There are clear disadvantages to the two-step approach. First, the two steps are confusing. Second, it will be difficult for a court to ignore the prior art—having already heard evidence on it, as a practical matter—while undertaking a subjective analysis of the specification. Most importantly, there is always a possibility of error in undertaking a parsing of the specification to see what the inventor intended. That is true even if the inventor was right about what was invented—the court might not always parse the specification correctly. The problem is worse if the inventor was wrong about what was invented. In either case, an identification of the subjective inventive concept might parse away elements that actually are inventive, either because the inventor did not realize they were the truly inventive aspect, or because the court misconstrued the specification.

I cannot think of any advantages to the two-step approach. Pozzoli and Sanofi do not illustrate any particular advantage to the two-step approach. Indeed, we can’t really even tell if the court applied a two-step approach or a one-step approach that was artificially separated into two steps. In both cases, the effect of the Windsurfing framework was to make the analysis more confusing by introducing a step of identifying an “inventive concept” without defining what that means or a methodology for identifying it. In both Pozzoli and Sanofi, the ultimate analysis would have been the same if the court would have asked whether the new information disclosed in the specification and embodied in the claim would have been obvious to a skilled person.

Finally, apart from the unnecessary difficulties created by the Windsurfing / Pozzoli approach even in the most straightforward cases, it breaks down entirely when the inventor did not address the same objective problem that would have been addressed by a skilled person.

I conclude that the Windsurfing / Pozzoli approach should be abandoned. It is never actually helpful. Even in the easiest cases it can cause confusion. And in the more difficult cases it will lead to error. As discussed in my last post, I would suggest that the best candidate for a replacement is a form of the objective problem and solution approach. But even if we retain a focus on the inventive concept, the better approach would be to simply identify the objective inventive concept in a single step.

No comments:

Post a Comment