by Dennis Crouch
The Solicitor General’s new brief in Amgen v. Sanofi shrugs off the case–identifying petitioner’s arguments as meritless:
Petitioners contend that the court of appeals erred by treating enablement as a question of law and by examining the full scope of the claims in assessing whether they are fully enabled. Those arguments lack merit and further review is not warranted.
Gov’t CVSG Brief in Amgen Inc. v. Sanofi, 21-757 (Supreme Court 2022).
The petition argued that genus claims–especially those functionally claimed–are being asked to comply with a heightened enablement standard. The Gov’t agreed that at times it may be difficult to enable an entire genus, but not because of any heightened standard. Rather, “broad claims naturally require more extensive enablement” since the patent’s disclosure “must be commensurate with the scope of its claims. . . . When, as here, a patent claims an entire genus based on its function, the patent must enable that entire genus.” The brief cites to Consolidated Electric Light Co. v. McKeesport Light Co., 159 U.S. 465 (1895) and Holland Furniture Co. v. Perkins Glue Co., 277 U.S. 245 (1928) as representative. In those cases, the Supreme Court used to enforce patents whose scope encompassed, but far exceeded, the examples given in the specification. Of course, the Brief’s bold statements fail to contend with the reality that enabling the “entire genus” will effectively curtail genus claims, and the resulting policy shifts. I expect that a Supreme Court decision in the the case would not go as far as suggested by the Gov’t Brief, but the middle ground might not be enough to flip the result here.
The Gov’t Brief also explains enablement as a mixed question of fact and law. Although the jury sided with the patentee, the district court judge rejected the verdict on JMOL — holding that the claims lacked enablement. That JMOL determination was affirmed on appeal. The Government brief does not delve deeply into the issues raised but simply states:
Petitioners complain about the formulation the court of appeals used in articulating the standard of review, but they do not identify any practical implications flowing from that disagreement.
Id.
Usually, the SG brief is the most important amicus brief at predicting certiorari. However, that might no longer be true for today’s 6-3 conservative majority hearing remarks from a liberal administration.
Dennis, I subscribe to your blog so I read them in my email. This is super convenient, but the font size in your emails is really small and hard to read. Could you update your emails to use a larger font size?
I’ll see what I can do.
Similarly in Holland Furniture Co. v. Perkins Glue Co., 277 U.S. 245 (1928), the Court held that the disclosure of “a particular starch glue” with a certain function did not enable a claim for “all starch glues” with that function, given that “[o]ne attempting to use or avoid the use of [the] discovery as so claimed and described functionally could do so only after elaborate experimentation,” id. at 256-257. The Court explained that “[a] claim so broad” would improperly “extend[ ]” the “patent monopoly” “beyond the discovery” by permitting “the inventor who has discovered that a defined type of starch answers the required purpose to exclude others from all other types of starch.” Id. at 257.
/golfclap
I’m glad someone knows what they’re doing. Nothing so succinctly encompasses the two principle concerns as those lines: unjust enrichment due to an overbroad scope and a lack of notice as to what actual embodiments are cleaved from the public sphere.
In this case, by contrast, the district court determined that a “person of ordinary skill in the art” using the random-generation method and “attempting to obtain a claimed antibody that is not disclosed” “ ‘would have to do essentially the same amount of work as the inventors of the patents-in-suit.’ ” Pet. App. 40a (citation omitted).
Functional claimers better hope that this logic does not catch on. I have yet to see a functional claim where applicants have gotten even halfway there on their scopes.
Random, can you write a bit more. Here in Europe, on enablement, we have a line of case law that can be summarised as “one example is enough”. I think that is often the case, even with functional claims. Suppose the invention is a foldable tennis racquet (and that this subject matter is new and not obvious). How many embodiments do I need? Or suppose I just invented PTCA (percutaneous trans-luminal coronary angioplasty). If I enable just one stent and just one delivery catheter, isn’t that enough to enable the full scope of my claim?
Or is it only in the case of covetous and excessively greedy claims within the unpredictable arts that the mantra “one is enough” does not hold?
For what little my opinion is worth, I think that the one-is-enough cases are all fine as far as they go. The one-is-enough cases are all careful to make clear that a single example may suffice, depending on the nature of the technology. As the EPO’s Examination Guidelines say (F-III.1) “the application is not usually regarded as satisfying the requirements of Art. 83 unless the description gives a number of examples or describes alternative embodiments or variations extending over the area protected by the claims. However, regard must be had to the facts and evidence of the particular case. There are some instances where even a very broad field is sufficiently exemplified by a limited number of examples or even one example… .”
So far as this goes, this is just common sense. This way of looking at written description tracks the distinction that one frequently sees in the inventive step context between the so-called “predictable arts” and “unpredictable arts.” Some technologies require more explanation than others in order to be fully enabled. There is nothing wrong with acknowledging that reality when assessing enablement, just as one does when assessing inventive step.
I am reminded of an old anti-software patent person (who knew precious little about patent law, or about innovation for that matter), whose ‘claim to fame’ rested on Maths (the philosophy leg) and the equivalency thereof.
The notion being explored here of, “One is enough” is a delightful reminder of how (unintentionally) self-defeating that Anti was.
As I recall, his nom de plume was PoiR. But it has been ages since the TechDirt and Slashdots have provided anything but the most feeble of anti-patent screed.
MaxDrei,
What exactly do you mean by, “within the unpredictable arts”…?
Are those per chance the very same arts that Malcolm crows about as being the “grown up” arts?
With your reference below (EPO) further recognizing that functional language is appropriate in the computing arts, I will take your silence here as agreement in my comment that inherently belittles Malcolm.
Here in Europe, on enablement, we have a line of case law that can be summarised as “one example is enough”.
Surely you don’t. When you invent a mechanism that folds a tennis racket (which is the actual invention) it makes just as much sense to claim “foldable tennis rackets” as it does to claim “all novel, nonobvious inventions as of the filing date.” Surely someone would have filed the omega patent by now if “One embodiment within the scope” was the true rule.
How many embodiments do I need?
Well this is an interesting question that requires some history. It used to be that when a claim was not in line with the embodiments in the spec it was a failure to point out and particularly claim (Gen Elec, Perkin’s Glue). Conversely, if the claim was to the principle it would be a judicial exception (Morse). Following the 1952 act the CPA essentially gutted “point out and particularly claim” (despite the same statutory language being used) and the supreme court cabined means-plus claiming with equivalency. The CAFC eventually recognized that situation was untenable, and started relying upon scope of enablement (which I use to differentiate from “entirely unenabled” akin to your 1 embodiment rule) and written description.
Consequently, the law has varied and the standards are diverse. It’s accurate to say that the claims must (1) inform the scope with reasonable certainty, (2) be significantly more than a law of nature, (3) be enabled across their entire breadth and (4) be possessed across their entire breadth.
The CAFC has enablement/WD cases that make clear that the the embodiments must fill the breadth of the scope. i.e. the number is not relevant so long as they are representative of the full breadth of the claim. They have never really identified how far one embodiment “represents” with respect to filling the entire breadth, rather they have instead used the negative which states that if all the embodiments describe one corner of the breadth, they fail to fill the scope.
I personally think of the example given in Lizardtech with a fuel efficient engine, or the language from Abbvie v. Jannsen:
One factor in considering the question is how large a genus is involved and what species of the genus are described in the patent. If the genus is not large or, even if it is, the specification discloses species representing the genus throughout its scope, the requirement may be met. On the other hand, analogizing the genus to a plot of land, if the disclosed species only abide in a corner of the genus, one has not described the genus sufficiently to show that the inventor invented, or had possession of, the genus. He only described a portion of it. That is the case here.
It is important not to take the analogy of a plot of land too far in thinking of written description issues because, even if one builds a house only in one corner of the plot, one may still own the whole plot. One describes a plot of land by its furthest coordinates, in effect drawing a perimeter fence around it. That may be akin to the function of patent claims to particularly point out and distinctly circumscribe the outer boundaries of a claimed invention. With the written description of a genus, however, merely drawing a fence around a perceived genus is not a description of the genus. One needs to show that one has truly invented the genus, i.e., that one has conceived and described sufficient representative species encompassing the breadth of the genus. Otherwise, one has only a research plan, leaving it to others to explore the unknown contours of the claimed genus.
Personally, although I have no case to say this, I think that an argument can easily be made that a disclosed enabled embodiment meets 112a with respect to its equivalents (achieving substantially similar results by performing substantially the same function in substantially the same way). When claims are limited only by function rather than function + structure, the claims violate 112a as soon as one can prove that there exist nonequivalent results or means. In other words, if you disclose and enable an embodiment, and you use means-plus claiming, you will get a valid scope of equivalence. If you forego means plus claiming and claim using general functional language, that language is invalid once there is proof that nonequivalent embodiments fall within the functional scope.
This only makes logical sense. It seems ridiculous that the same disclosure could support a scope not limited by similar means or results when means-plus usage limits the claim by similar means or results. The very reason attorneys stopped using means-plus claiming is the same reason general functional claiming is improper – it gives dominion over embodiments that the art would find distinct from the embodiments disclosed.
Or suppose I just invented PTCA (percutaneous trans-luminal coronary angioplasty). If I enable just one stent and just one delivery catheter, isn’t that enough to enable the full scope of my claim?
I dislike speaking factually on subjects I am not an expert on. You should ask yourself, would one of ordinary skill find a material distinction between my embodiments and other undisclosed (and probably unknown to you) embodiments that lie within the claimed scope? If so, you run the risk that a court will seize upon the distinction the art finds to be material as a rationale for invalidating your claim.
To use a more lay-friendly example – Anyone knows the difference in efficiency between a modern-day gun and an ancient bow. They both perform the same broad function (firing a projectile) but use entirely different means (directed chemical explosion versus string tension) and have extremely different results. Someone who discloses bow embodiments certainly *can* draft functional claim language that would cover the gun, but the function of patent law is to encourage the disclosure of the gun, and therefore patent law is served by invalidating any scope to a bow disclosure that reaches the gun. The value of disclosure is only abated when everything that is missing (and obviously you cannot disclose every embodiment of something so something is always missing) are things that would be recognized as being equivalent to that which already was disclosed. That way the same teachings for the same benefits apply, and the public is getting the full value of its exchange and preventing the unjust enrichment of a mediocre inventor with an excellent attorney.
Put succinctly, a functional scope is only enabled when the specification together with the prior knowledge of the art enables *every* distinct manner of achieving the result without undue experimentation. As soon as a nonequivalent embodiment is not enabled, the full scope of enablement is not met and the claims are invalid for lack of enablement.
Thanks for all that, Random, but on the issue whether “one way is enough” (where you doubt what I write) I give you below a Link to the relevant paragraph in the EPO’s MPEP. I am curious whether there is anything in that short paragraph that you would need to dispute or take issue with. Why? Because I like to think that on the question how much disclosure is needed to enable a broad claim there is international consensus.
link to epo.org
MaxDtei,
Thanks (sincerely) for the link, but a minor quibble or two.
The link unfolds into far more than one little paragraph (and Random WILL have difficulties with that fact that legal cases are embedded).
Further, any agreements as to EPO laws and rules are NOT examples of “international consensus” APART from the EPO itself.
There is NO such thing as as patent One World Order, and patent law remains (as it has always been) a Sovereign-Centric type of law.
You don’t need to remind somebody in Europe that independent nation States are sovereign over their own patent law. The patent judges of the EPC Member States convene every other year, to discuss their differences. Patent attorneys in Europe are forever comparing and contrasting these differences. Nevertheless, there is consensus on many issues. Who in the USA apart from Random Guy can seriously dispute anything in the linked text of the EPO’s MPEP?
I do not think that the law as explained in your link (the same that I quoted in 3.1.1 above) describes a standard that is meaningfully different than ours here in the U.S. Sometimes description of a single species will support your genus claim, but more often it will not. That is just as true here as there.
Greg, thanks. Apologies for not noticing that you had already quoted the same para in the EPO Guidelines that I later linked to. I’m in engineering cases rather than pharma, so am not so wedded to the genus/species way of thinking. Hence my examples of a folding tennis racquet or PTCA to tackle a cardio-vascular stenosis. I suspect that it is far more frequently the case in the engineering space that “one embodiment is enough”.
MaxDrei,
Even more so in the computing arts (as per my recollection of the self-defeating rants of P0iR from the Slashdot/TechDirt days).
I am curious whether there is anything in that short paragraph that you would need to dispute or take issue with.
There are some instances where even a very broad field is sufficiently exemplified by a limited number of examples or even one example (see also F‑IV, 6.3). In these latter cases the application must contain, in addition to the examples, sufficient information to allow the person skilled in the art, using common general knowledge, to perform the invention over the whole area claimed without undue burden and without needing inventive skill (see T 727/95).
This is true but uncommon. Rarely do people invent a generic solution to a problem, because rarely are problems as generic as inventors would like to pretend. I agree it can be done, however.
A single example may suffice, but where the claims cover a broad field, the application is not usually regarded as satisfying the requirements of Art. 83 unless the description gives a number of examples or describes alternative embodiments or variations extending over the area protected by the claims.
I would say that it is 80% of the way there, because I believe there is broad agreement that “variations extending over the [breadth of the scope]” are generally necessary in both systems. It has the same ambiguity as in American law, however, where it is unclear how far a given teaching sufficiently extends. This paragraph begs the question of “How many variations and what qualities are satisfactory for a broad field claim?” On that the paragraph provides no answer beyond the relatively conclusory statement that the dearth of examples can’t cause an undue burden. So I can’t necessarily tell you that there would be consensus, but I would say I see nothing that necessarily conflicts with american law.
Very helpful. Thanks, Random.
Again though, we have this word “generic”, or, as Greg uses it “genus”. I suspect that, at least in engineering cases, a great many claims go legitimately through to issue even though there is only one enabled specific illustrated embodiment supporting them in the specification.
You write: “Rarely do people invent a generic solution to a problem, because rarely are problems as generic as inventors would like to pretend”. My reaction is to wonder about scales of “genericity” and how to measure how “generic” any given “problem” is. But that might be because “problem” has such a precisely-defined meaning in the EPO approach to the obviousness enquiry.
MaxDtei,
Your “wondering” here is directly informed by the well-known (perhaps in US only) term of “Ladders of Abstraction.”
This legal term mirrors the technical term so often advanced by Night Writer, and is a term that Random has shown to not be able to breach his fortress on the Peak of Mount S; wherein Random has decreed that all Ladders have at most two rungs).
My reaction is to wonder about scales of “genericity” and how to measure how “generic” any given “problem” is. But that might be because “problem” has such a precisely-defined meaning in the EPO approach to the obviousness enquiry.
That is an important inquiry to be sure, but from my standpoint most of the work is done for me, because people don’t attempt to put forward a specific solution, they attempt to claim virtually all solutions – i.e. the most generic scope you can get given the context. That makes the invalidity argument extremely easy. As Holland furniture points out, finding out that a starch combination with three parts water and alkali is “as good as horse glue” neither supports a claim to “all substances as good as horse glue” nor to “all starch glues made with three parts of water and alkali.” One can similarly analogize that a computer algorithm that performs a function does not support a claim to “all algorithms that generate the functional result” (code for…) nor to “all algorithms having some structural similarity” (which usually manifests as an algorithm “based on” one or two particular parameters).
Perhaps it helps, to look at a claim and ask if it is defining the (new and not obvious) solution to a technical problem or is it merely defining a problem for which the disclosure of the patent application as filed simply fails to make available to its skilled and curious reader any full-width solution?
The former claim is patentable, of course, the latter not, because the quid pro quo of the patent statute (any patent statute) is not satisfied.
Perhaps.
Perhaps not, as I truly doubt that THAT is what Random drones on and on about.
Good reasons for cert denial include the fact intensive and peculiarly patent law nature of claim “enablement” issues, plus the Markman requirement for judicial, not jury, interpretation of claim language.
“the Brief’s bold statements fail to contend with the reality that enabling the “entire genus” will effectively curtail genus claims”
Perhaps in some arts. Software will no doubt press on with broad functional claims and any mention of requiring enablment for the full scope of how those functions are achieved will result in hysteria.
Ben being Ben…
I mean, enabling the entire genus IS the law, and nobody is really disputing that. The question is by what standard do some particular embodiment disclosures enable the entire genus. The mechanism cited in the brief is to quote the incandescent lamp case (“But the patent specification did not identify ‘some general quality, running through the whole fibrous and textile kingdom, which distinguished it from every other, and gave it a peculiar fitness for the particular purpose.'”) for a generic quality.
I don’t know if generic quality is the best requirement, but I certainly agree with the logic that a lack of generic quality coupled with noticably different functional results are a good standard for denial of the scope.
It’s pretty clear where this is going right? Applicants are either going to have to do all the work themselves or they’re going to have to limit themselves to particular embodiments they have discovered work and accept an equivalent means standard. The worst part of the brief for functional claimers is the statement that
In any event, the court of appeals’ enablement determination was reasonable on the record before it. The court explained that “the claims are far broader in functional diversity than the disclosed examples,” noting that “there are three claimed residues to which not one disclosed example binds” and that, “although the claims include antibodies that bind up to sixteen residues, none of [the] examples binds more than nine.” Pet. App. 13a & n.1; see C.A. App. 4283 (listing “competitor antibodies” that bind more and different residues)
There is a zero percent chance that the courts are going to allow functional language to disincentivize additional disclosures. There is no mechanism that would allow a claim to, e.g., binding to sixteen residues when the best disclosure binds to nine. Whatever standard pro-functional people want to come up with, it cannot allow someone who discovers the 9 bind to reach the 16 bind embodiment.
Ben (or somebody else) I had supposed that a “genus” is something in biology or chemistry but not in engineering cases. Is it also to be found in functional claiming in engineering inventions? If so, can you give an example?
As even MaxDrei can grasp, the “Genus/Species” position is bandied about without care.
Very much a result of the unthinking “copy/paste” US patent examiner mode.
I am also reminded of the staggering height of Mount S for those in that mode.
Comments are closed.