By Dennis Crouch
Amgen v. Sanofi (Fed. Cir. 2017). The Delaware district court sided with Patentee Amgen and awarded permanent injunctive relief barring Sanofi from selling its injectable drug product Praulent (alirocaumab) – a human monoclonal antibody used to reduce low-density lipoprotein (LDL) cholesterol – covered by U.S. Patent Nos. 8,829,165 and 8,859,741.
On appeal, the Federal Circuit has vacated the injunction – holding that the district court made reversible errors with regard to written description and enablement. Before walking through the decision, I’ll note here that the Federal Circuit panel here – made up of Chief Judge Prost, Judge Taranto and Judge Hughes – would not be identified as overly pro-patentee.
The drug at issue works by inhibiting the protein PCSK9 from binding with LDL Receptor (LDL-R), thus leading to a reduction of blood LDL. Basically, the claimed antibody binds to a particular residue (amino acid) of the PCSK9 molecule in a way that blocks the PCSK9 from then binding and promoting LDL-R. Representative claim 1 of the ‘165 patent is as follows:
An isolated monoclonal antibody,
wherein, when bound to PCSK9, the monoclonal antibody binds to at least one of the following residues: S153, I154, … or S381 of SEQ ID NO:3, and
wherein the monoclonal antibody blocks binding of PCSK9 to LDL-R.
Looking at the claim – it basically covers all isolated monoclonal antibodies capable of performing the required binding function – and the basic issue in the case is whether that broad claim is sufficiently supported by the specification. On point, the specification describes a number of example antibodies; the screening process that the inventors used to find those examples; and 3D structure of two of the known antibodies, including the antibody that serves as the patentee’s actual drug product (Repatha).
The Written Description and Enablement requirements are both stem from the same sentence in 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same.
Written description focuses on whether the original patent disclosure conveyed evidence that the patentee – at the time of filing – had “possession” of the claimed invention. Ariad. Most written description challenges arise from post-filing amendments to the patent claim scope, however in Ariad the Federal Circuit made clear that originally-filed claims may lack sufficient written description if the claim covers a wide genus of embodiments and the specification fails to disclose “a representative number of species.”
Post-Priority Evidence: The extent of required written description disclosure depends upon the number of potential species. However, the district court refused to allow the accused-infringer attempted to introduce such evidence unless it reflected a pre-priority understanding of the science. On appeal, the Federal Circuit rejected that approach – holding that “the use of post-priority-date evidence to show that a patent does not disclose a representative number of species of a claimed genus is proper.” Note here, the rule here is distinct from the limitation that only pre-priority date evidence is permissible for showing state-of-the-art at filing.
Enablement also focuses on sufficiency of disclosure, but asks the slightly different question of whether, after considering the specification, a skilled practitioner would be enabled to make and use the invention.
Under the enablement requirement, “the specification of a patent must teach those skilled in the art how to make and use the full scope of the claimed invention without undue experimentation.” Genentech, Inc. v. Novo Nordisk A/S, 108 F.3d 1361 (Fed. Cir. 1997).
As with written description, the focus is on enablement at the time the application was filed, however here the court also allows post-priority-date evidence to show that the accused infringers here “engaged in lengthy and potentially undue experimentation to enable the full scope of the claims.” The court writes:
Such evidence could have been relevant to determining if the claims were enabled as of the priority date and should not have been excluded simply because it post-dated the claims’ priority date.
Thus, the district court’s exclusion of such evidence was improper.
= = = = =
In addition to improperly excluding evidence, the Federal Circuit also ruled that the district court antibody-antigen jury instruction relating to written description was also improper. In particular, the jury instructions included the following explanation of the law:
In the case of a claim to antibodies, the correlation between structure and function may also be satisfied by the disclosure of a newly characterized antigen by its structure, formula, chemical name, or physical properties if you find that the level of skill and knowledge in the art of antibodies at the time of filing was such that production of antibodies against such an antigen was conventional or routine.
According to the appellate court, this instruction is wrong because it allows a patentee to satisfy the description-of-use portion of written description by relying upon the level of skill in the art. Rather written description requires an actual written description showing possession of the invention, including express instructions on how to make and use the invention.
The essential problem with the jury instruction given in this case is that it effectively permitted the jury to dispense with the required finding of a written description of the invention. . . . A jury would naturally understand the instruction to permit it to deem any antibody within the claim adequately described merely because the antibody could easily be “produc[ed]” (and, implicitly, used as an antibody).
The decision here repudiates statement to the contrary from Enzo that are here characterized as dicta. On remand, the jury will likely also be asked to consider whether – in this particular case – “the chemical structure of an antigen [provides] structure-identifying information about the corresponding antibodies.
Just out of curiosity, is “the antibody exception to the written description requirement” directly related to the long-standing practice of chemical patents with “product by process” claims and their permissible spec support?
There is a conceptual relationship insofar as both product-by-process claims (now more accurately referred to as “method claims, except weaker”) and the “Antibody-by-function” claims are (were) attempts to skirt around the written description requirement as applied to compositions/articles.
But I don’t think there’s any other relationship than that.
The typical broad independent “antibody-by-function” claim would be presumed to cover antibodies against the recited antigen regardless of how those antibodies were produced and that presumption would have held even after the CAFC limited product-by-process claims to the recited process.
Greg: If, however, we see a rapid increase in mAbs coming to market, that would (mostly) indicate that the new law is better for technology.
Greg, once again: the “exception” has been on extremely thin ice for years. It’s not like there’s a ton of cases applying the exception in favor of patentees have suddenly been overturned. The opposite is true, i.e., all the cases over the years attacking the exception and questioning its validity have been affirmed and underscored.
The reason that this case is important (and necessary) is because therapeutic antibodies are a bigger business now than ever. Of course the field is going to keep growing. Why would we want to make it *harder* for companies to compete by granting undeserved patent claims that flout the fundamental principles underlying the patent system??? That would only benefit lawyers.
Now contrast with the s0 ftie w0 ft ies.
Just to echo a comment that Greg made earlier this week (or last week): the idea that a jury of lay people is going to decide “whether – in this particular case – the chemical structure of an antigen [provides] structure-identifying information about the corresponding antibodies” is a tragically absurd quirk of this country’s legal system.
I’ll add this: generally speaking, the quirk works in favor of the richest and most powerful because the jury increases the costs of litigation enormously.
And, yes, that’s why the quirk persists.
On the other hand, that same quirk coupled with substantial evidence review also allows lay juries to stick it to rich and powerful companies in product liability and tort cases where a more educated factfinder may side with the rich and powerful company.
Not everything in law is or should be viewed through a patent lens.
a more educated factfinder may side with the rich and powerful company
Sure, anything is possible. The opposite may also happen.
But is there any evidence that this actually happens systematically? i.e., that juries *in general* tend to be more accurate in their understanding of critical complex scientific facts than “educated factfinders”, or judges? If so, I’d like to see it.
We let judges over-ride juries because we know that juries get it wrong. If the judges keep screwing up, we have mechanisms in place to get rid of them. Juries, on the other hand, don’t get punished for getting the facts wrong. And they don’t get payed squat for the work their supposed to do on top of all that. Did you ever think about that? Patent attorneys arguing about how important it is for innovators to be compensated billions of dollars for their “genius” before a bunch of people whose minds are melting with technical overload while being paid diddly squat?
MM, jurors are not supposed to be experts in technology. The testifying expert is supposed to explain the technology in understandable terms. If the expert fails in this, or he is lying (and jury can tell if he is) then the side producing the expert is going to lose.
The testifying expert is supposed to explain the technology in understandable terms. If the expert fails in this, or he is lying (and jury can tell if he is) then the side producing the expert is going to lose.
Yes, I understand how juries are supposed to work in theory. Is there any evidence that the average layperson is better (or even equal) at evaluating credibility of an expert than a person who is educated on the topic?
All signs point to a resounding “no.” I mean we live in a country where vast numbers of people think the earth is 10,000 years old and who believe that dark skinned people are inherently less intelligent than lighter skinned people. But somehow we expect these people to evaluate the credibility of two scientists arguing about an incredibly arcane issue in the field of monoclonal antibodies? LOLOLOLOLOLOLOLOLOLOLOLOLOLOL
The testifying expert is supposed to explain the technology in understandable terms.
Well, yes, that is the theory. Such a theory cries out for two obvious responses: (1) what is the evidence that this theory works in practice any more often than might be expected from random chance?; and (2) why contrive to make it so hard to achieve justice.
Yes, you can imagine that each side will find experts who are so skilled at communication that they will be able to explain extremely complex information to a nearly random selection of people who have neither the experience nor the motivation to understand it. Every now and again such an outcome might even obtain. But why contrive a system that—in order to function rightly—positively depends on such an inherently unlikely outcome. Would it not make more sense to set up a system where matters that depend on complex questions of fact are decided by people selected for their training and inclination to care about and understand such complexities?
If the expert fails in this, or he is lying (and jury can tell if he is)…
Spare me. This is nonsense on stilts. Juries can tell that a given expert is good looking, or well spoken, or amusing. They think that this means that they can tell whether someone is lying, but really that is simply the conceit that most of us (unworthily) indulge, that we can tell when we are being snowed. That is only true when the snowjob artist is unskilled at her trade. When done well, the viewer has no idea that it is happening.
Greg, juries are supposed to provide completely unbiased triers of fact, not experts in the field who often do have biases and agendas relative to the topic at hand.
Regardless, in actual practice, I have often seen counsel using challenges to remove educated people from the jury. It most often occurs when that party must win its case based upon somewhat shaky scientific evidence. Thus in jury selection, the biases of people are well known and cases are often won or lost based solely on the skill the site uses in jury selection.
Secondly, why do you think both sides like to remove lawyers from the jury?
juries are supposed to provide completely unbiased triers of fact,
Which is absurd. Which is why educated people around the world have dismissed the idea out of hand.
News flash: America is kind of a j0 ke in many respects. That’s increasingly obvious to a lot of people. Many of us have known this to be true since we were, oh I dunno, forteen.
Greg, your elitism is showing. The PTAB proves that the “experts” are too smart by half. Common sense and perspective must be cultivated. Trial by jury is the genius of western law, preserved by the American system, notwithstanding the outcome in Oil States.
Greg, your elitism is showing.
Always. My preference is so strong for the idea that decisions should be made by people with the education and experience necessary to make a good decision, that I have neither the ability nor the inclination to conceal it. My “elitism” (as a preference for competence is sniffingly derided in certain parts) is always showing.
Trial by jury is the genius of western law…
The factual accuracy of this assertion seems rather compromised by the observation that literally no other justice system in the world (“western” or otherwise) uses juries in private law trials.
Trial by jury is the genius of western law
I don’t believe there is any country in “the west” which has followed the US approach and I’m not aware of any evidence that the legal systems in those countries are less “just”.
I am aware of major systematic failings of the US jury system, however.
[T]he idea that a jury of lay people is going to decide “whether – in this particular case – the chemical structure of an antigen [provides] structure-identifying information about the corresponding antibodies” is a tragically absurd quirk of this country’s legal system.
Too true.
I see the concerns to be sure. So how would you – Greg and MM – design an adjudicative system if you could blow it up and start from scratch? Would litigation in your systems operate the same for patent infringement and other disputes?
Would litigation in your systems operate the same for patent infringement and other disputes?
No. I would like to see a special set of district courts set up for handling patent matters and only patent matters. No juries. All bench trials. Judges for these courts would required by statute to have undergraduate degrees in various technical topics (just as is required to stand for the patent bar). That way matters of a technical nature would be decided by people who stand a chance of actually understanding the matters at issue.
“An isolated monoclonal antibody,
wherein, when bound to PCSK9, the monoclonal antibody binds to at least one of the following residues: S153, I154, … or S381 of SEQ ID NO:3, and
wherein the monoclonal antibody blocks binding of PCSK9 to LDL-R.”
What’s the objective physical structure of this antibody?
I thought this was supposed to be the “grown up arts.”
What’s the objective physical structure of this antibody?
LOL Did you even bother to read what this case is about?
And yes I’ve been bringing up this “exception” for you guys over and over again for a solid DECADE because (wait for it) it’s actually always been acknowledged to be an “exception” to the law.
Contrast with the s0f tie w0 ftie types who have somehow managed to convince themselves that patenting “new” logic processing machines in purely functional language was the original vision of the Founders. You guys are a j 0 ke.
So I take it that all of your many biotech cases contain, on filing, independent claims that define proteins by amino acid sequence and full secondary, tertiary, and quarternary structure, as applicable? Or define polynucleotides by nucleotide sequence?
all of your many biotech cases contain, on filing, independent claims that define proteins by amino acid sequence and full secondary, tertiary, and quarternary structure, as applicable? Or define polynucleotides by nucleotide sequence?
Yes, that’s typical. They’re called “sequence listings” and they were made a requirement of filng a long time ago, with full industry approval, to facilitate the examination of biotech patent claims. Obtaining the sequences and preparing and maintaining the listings over the course of prosecution can be a lot of work and stressful at times. But it’s a necessary part of the job. It’s a bit more — shall we say — “rigorous” than picking your nose and dreaming up some “cool” thing that somebody else will program your iPhone to do. And, no, I’m not “jealous.” Just pointing out a fact.
No sweat off my picked nose – I’m a material scientist by training and practice primarily in metallurgy and ceramics.
“They’re called “sequence listings” and they were made a requirement of filng a long time ago, with full industry approval, to facilitate the examination of biotech patent claims.”
and
“And, no, I’m not “jealous.” Just pointing out a fact.”
Too funny.
You are NOT ‘just pointing out a fact,’ as you are directly attempting to compare and contrast two very different arts with one art having committed to something to “make examination easier” in your view.
Since you are doing more than just pointing out a fact, it is indeed fair game to wonder why you are using that ‘fact’ as you are attempting to do.
In all of your volumes of “fluence,” you remain insincere and shallow.
(there’s a word for that – what are the chances that you have accused others of being what describes you? 😉 )
Can you flesh this out more for the non-chemists? I think you are on to something. Structure at bottom is an arrangement of molecules, atoms, isotopes, and so on. I presume that monoclonal amino acid is a structure?
monoclonal antibody.
Is a claim for CDR valid absent a disclosure of all amino acids with that functionality?
Is a claim for an amino acid valid absent a diclosure of all variants?
Every species is also a genus, and susceptible to this written description attack.
“Every species is also a genus, and susceptible to this written description attack.”
This comment says more than most people will recognize.
One of the problems with looking at patents from the “here are my eyeglasses” singular perspective of bio-related is the notion of a SINGLE level between genus and species.
This shows itself in some of the dialogue (that eventually emerges when the topic comes up) concerning a tactic that is almost universally employed (at least outside of nigh-useless exacting picture claims): the ladders of abstraction.
Those “tainted” by this (lockstep) perspective OFTEN attempt to want exacting singular physical structure in claims.
I have even pointed out that even in the chemical arts (which some may view as the epitome of “single physical structure) will often (with such things in claims as “comprising” and “ranges”) NOT provide actual single physical structures.
Once it is understood that there are MORE THAN just two rungs on a ladder, and then take THAT understanding and look at how nearly most all useful claims are actually written, one may note the similarities between anti-particular-art-unit patent arguments and just general anti-patent arguments.
NB—conformity with the written description requirement is a question of fact. Ariad Pharma. v. Eli Lilly & Co., 598 F.3d 1336, 1351 (Fed. Cir. 2010). I am afraid that exactly how much disclosure is necessary to support any given claim is necessarily going to be too particular to admit of a general, all-purpose answer to your question.
I have no idea how familiar you are with antibody patents, but, yes, it is quite customary to claim antibodies in terms of sets of complementarity determining regions (CDRs), which are (in turn) defined by amino acid sequences.
Greg: I take no position on whether this is good for technology (the real end to which all patent law jurisprudence should be focused), but it is almost certainly good for the clarity and coherence of this branch of patent law.
Except that there is an even bigger and less justifiable exception to “this branch of patent law” floating out there. That’s why I have been discussing the weirdness of this so-called “exception” here for many, many years.
For what it’s worth, this “antibody exception” did precisely *nothing* to promote antibody technology. What it did was slow the field down because it created a situation where reams of junk patents describing new “antigens” (basically anything that might be recognizable by an antibody) tossed in a junky pro forma antibody claim with zero supporting data. That amounted to an effective tax and a cloud hanging over the work of people who were actually in the business of creating and characterizing useful antibodies (i.e., an antibody that could actually be claimed properly, as a composition described in structural terms).
Chem/bio being a grown-up art, there will be very very little whining about this because it was always weird, it was always suspect, and everybody knew it and did their best to plan for it. Contrast that behavior with the behavior of the most entitled art unit in the history of the universe.
I have long wondered why CAFC jurisprudence treated antibodies as some sort of special case in evaluating written description compliance.
Well, you should have just asked. The reason is that when the exception started it was relatively easy to obtain “an antibody” to a new antigen. All you needed to do is inject the antigen into a mouse, basically, and then you let the mouse do its thing. BUT actually determining the structure of that antibody (i.e., the amino acid sequence) was a far more labor intensive proposition. All this was before the CAFC recognized a separate written description requirement so the argument was made that if you can make and use the antibody (easy as pie, as I already explained) then you didn’t need to recite the structure. Never mind that there was no way of knowing whether the broadly claimed antibody read on a zillion other pre-existing antibodies (including naturally occurring ones). Just hand out the claim. What could go wrong? Well, everything went wrong.
Now take a tiny baby step and compare the situation with software/logic patents where, in most cases, it’s always been easy for the developer of the software to describe at least one structure to the PTO (e.g., it’s a medium upon which is printed or encoded a series of digital pits or magnetic impressions with the pattern 1011010100100101000101010 etc <– that's the thing that actually works). Instead of taking even the most modest step of requiring the submission of that kind of data to the PTO, the system went the complete opposite direction and created a pure and absurd fantasy, i.e., algorithms are "the essence of electronic structure." In other words, "no structure" is treated as "structure" because … some rich people like it that way (never mind that most people find it absurd, and that includes most people who actually are skilled in the art). That's all that's going on, Greg.
For what it’s worth, this “antibody exception” did precisely *nothing* to promote antibody technology. What it did was slow the field down because it created… an effective tax and a cloud hanging over the work of people who were actually in the business of creating and characterizing useful antibodies…
With respect, I am entirely sympathetic to this story you are telling, but an honest person would have to admit that there is no way that we can know whether this is true or not. While you (& I) can tell the story above, the other side can equally well tell a story about how prior U.S. law created a nuturing environment in which it made good business sense for companies like AbbVie and Amgen to invest the necessary funds to take a few mAbs through FDA clearance so that they could get to market, where they can do patients some good. Innovation that never leaves the lab, after all, is not really that much of a public benefit.
The reality is that we cannot (yet) know whether the old law or the new is better for innovation. If, however, we see a rapid increase in mAbs coming to market, that would (mostly) indicate that the new law is better for technology. If, on the other hand, we see a slow down in new mAbs coming on the market in the next 15–20 years, that would (mostly) indicate that the old law was better. We do not have enough data as yet to know one way or the other.
an honest person would have to admit that there is no way that we can know whether this is true or not.
Oh give me a break. It’s no different than me dropping a claim into every case I write that’s “A computer, wherein said computer is configured to store, process, or transmit information related to an aspect of [insert new molecule here]”.
Does that kind of g@ rbage promote progress in the machine logic “arts”? Of course it doesn’t. Will it help the patent owner? Maybe, but you have to weigh it against the cost to the public.
The problem in the machine logic “arts” is the entitlement and investment of a certain class of patent attorneys in maintaining the status quo. That’s it. Get rid of all software patents tomorrow and literally nothing at all would change in terms of progress in “writing software.” This isn’t even debatable.
“This isn’t even debatable.”
LOL – it also isn’t even correct.
But you already knew that, eh?
“This isn’t even debatable.”
LOL – it also isn’t even correct.
Did Alice slow down progress in the “art” of software instruction writing? Or whatever the ridiculous underlying transaction was? You want to make a case for that proposition? Knock yourself out.
Everybody knows that nothing changed. You Big Jeans and Noonan are both so bloated with crow that you can’t even think straight.
Your reply is a non sequitur and does not make your statement to which I replied any less incorrect.
Written description – as in this case – is just not the same as what the anti-software folk try to do with the exceptions to the judicial doctrine of written matter.
This was a very small part of the case, but to the best of my knowledge, yesterday’s opinion represents the first precedential application of Dynamic Drinkware. I was gratified to see that the CAFC applied Dynamic Drinkware in its strong sense to cut the legs out from under the obviousness challenge. I am still waiting, however, for a precedential resolution of my big question post Dynamic Drinkware: does the priority filing need to provide adequate §112 support for any claim in the later published app, or does it need to provide §112 support for all claims in the later published app?
Great question. It must be percolating in the system somewhere.
Here’s hoping.
it basically covers all [algorithms] capable of performing the required [] function
originally-filed claims may lack sufficient written description if the claim covers a wide genus of embodiments and the specification fails to disclose “a representative number of species.”
the specification describes a number of [functions]; the screening process that the inventors used to find those examples; and [algorithms] of the known [functions], including the [algorithm] that serves as the patentee’s actual [program].
The essential problem with the [lawyers] in this case is that it effectively permitted the [office] to dispense with the required finding of a written description of the invention. . . . [The office] would naturally understand the [argument] to permit it to deem any [function] within the claim adequately described merely because the [function] could easily be “produc[ed]” (and, implicitly, used as an [function]).
Rewritten in computer terms to show why virtually all of my allowances are worthless pieces of paper.
Note here, the rule here is distinct from the limitation that only pre-priority date evidence is permissible for showing state-of-the-art at filing.
Evidence is evidence of what it is evidence of. A WD rejection is not a 103 rejection – the entire function of 103 is to show that the actual scope claimed is larger than what was invented, so obviously its proper to introduce later-developed evidence. Think back to Morse – Morse said that his disclosure filled the entire field of printing at a distance. The best way to prove that is not the case is to show a network-connected printer, or a fax machine, or a television. Similarly, the best way to prove that a bow doesn’t fill the field of machines that fire a projectile is to show a gun. If you were limited to pre-dated evidence it wouldn’t be a 112a rejection, it’d be a 103 rejection.
Even for a 103 standpoint, a post-dated document which describes the state of the art prior to the invention would be proper. The best evidence that something is obvious would be the entire field claiming the same invention at around the same point in time. i.e. If one examined an application filed jan 1, and one were to cite 50 applications between jan 2 and say march that all claim the same subject matter, those applications are evidence that it would have been obvious on dec 31 (absent some sort of special reasoning why it wouldn’t) to create the invention.
Honestly, I was surprised at the amount of this case (both the decision and the argument) devoted to the question of post-filing date evidence. I thought that it was fairly uncontroversial that post-filing date evidence of non-enablement was relevant (after all, if it not enabled 3 months after the filing date, how could it have been enabled at the filing date?). See, e.g. In re Wright, 999 F.2d 1557, 1562 (Fed. Cir. 1993) (Article published 5 years after the filing date of the application adequately supported the examiner’s position that the physiological activity of certain viruses was sufficiently unpredictable so that a person skilled in the art would not have believed that the success with one virus and one animal could be extrapolated successfully to all viruses with all living organisms).
The federal circuit seems to want to apply some sort of “best evidence” rule to 102/103, and some lawyers misinterpret that as being directed to the entire statutory scheme.
There is no statutory basis for them to apply a best evidence rule, and even if they did it couldn’t be applied outside of 102/103 as it would write the statute out of existence.
“Prior” art is a suggestion, not a requirement. Graham requires identifying the state and skill of the art at the critical date. It does not limit the manner in which that is proven.
Imagine if you could only prove a crime by talking about what evidence pre-dated the occurrence rather than surmise what took place based upon post-dated evidence.
The “best evidence rule” applies to the admissability of written documents.
It is not some amorphous standard where various types of evidence are considered better than other types.
I agree that the district court’s rejection of the post-filing data to prove lack of written description support was bizarre. Kudos to the CAFC for doing the right thing instead of the absolute wrong thing.
That said, the CAFC in other cases has messed things up badly by taking a different tack than it took in this case when it dealt with post-filing data to rebut obviousness. Rather than make it clear that there are just a limited set of facts where post-filing data can be presented to provide evidence of the non-obviousness of a compound/article, they seemed to open the gates wide to allow any post-filing data, regardless of when it was obtained or by whom, to be presented to support any number of what can sometimes be millions of prophetic/precatory statements in the spec. This is a disaster for everyone except speculators.
The EP has pushed back on this and it’s guaranteed that the CAFC will be dialing it back, too, as soon as the defense gets their sh-t together. That’s because the CAFC was simply wrong. Really wrong. And this case highlights the wrongness.
Rather than make it clear that there are just a limited set of facts where post-filing data can be presented to provide evidence of the non-obviousness of a compound/article, they seemed to open the gates wide to allow any post-filing data, regardless of when it was obtained or by whom, to be presented to support any number of what can sometimes be millions of prophetic/precatory statements in the spec. This is a disaster for everyone except speculators.
No, its correct. You just don’t seem to understand that the reverse is also true, i.e. there are “open gates” to show *obviousness* of a compound/article using post-dated data too. Do not confuse having to find what the state of something at one point with evidence that proves that state. If you can create a logical tie between some other point in time and the relevant point in time, the evidence is evidence.
I think you’re misunderstanding my point here. I’m not suggesting that post-filing data to support non-obviousness can never be relevant. On the contrary, post-filing data is going to be relevant and should be admissable as evidence in many, many cases. But the default assumption should not that be that all post-filing data is admissable for proving non-obviousness (which is a conclusion to be drawn as it relates to the claimed invention at the time of filing and that’s by statute).
Again, what the CAFC has done in recent years (and the PTO has followed suit) is to make post-filing (and post-generated) data suffice to overcome prima facie obviousness of a species even where there is no data in the spec showing that particular species was preferred out of a laundry list of thousands of different species.
You can see why this creates a bit of liability issue for everybody, can’t you? Particularly in the age of word processing and the use of computers to create massive wish lists?
But the default assumption should not that be that all post-filing data is admissable for proving non-obviousness (which is a conclusion to be drawn as it relates to the claimed invention at the time of filing and that’s by statute).
But the standard for admissibility is low. All it requires is to move the needle on a relevant issue and not be more prejudicial than probative. For something to be admissible doesn’t mean it has to carry the day.
I’m not entirely sure what you are referring to in this context. Are you talking about secondary considerations? Are you talking about laundry listing species and saying that they constitute support for a genus? If someone provides secondary considerations you can acknowledge them and say that they don’t overcome a strong basis of obviousness. If you are talking about laundry listing, you can point to the list as reasons it doesn’t have support.
In other words, in either case you can call it what it is and maintain your original position. As long as you consider the evidence it is not error to maintain your original position, you’re the jury.
Should be *the entire function of 112a*
Below is a slightly edited repost of my comment yesterday (which happened to be the first published commentary on this decision):
Today’s CAFC opinion in Amgen v. Sanofi represents another big nail pounded forcefully into the the already well-sealed coffin known as “the antibody exception” to the written description requirement.
That’s important for two reasons: (1) there are a ton of claims out there relying heavily on the exception for validity; and (2) antibody-based therapeutics is a bigger business now than ever before.
Everybody following along so far? Now pay attention to what happens next.
First, patent attorneys in this field aren’t going to run around crying about how nothing is patentable and nobody is going to make new antibodies and China is going to be the antibody leader etc.
Second, a skilled attorney is going to recognize that there is now only one major subject matter exception to the written description requirement and that is — as I have been telling everyone for years — (suprise!) software printed on a medium and computers that are claimed in functional terms (as opposed to claiming in structural terms, and distinguishing the art on that basis). And that exception rests on a much thinner foundation than the antibody exception (because antibodies, at least, are made of amino acids and there’s only about 20 of those; software instructions are just information and can be expressed in an infinite number of ways).
Read it and weep, logic patenteers:
By permitting a finding of adequate written description merely from a finding of ability to make and use, the challenged sentence of the jury instruction in this case [ignored] what is perhaps the core ruling of Ariad
In other words, stop conflating enablement with written description. [and “conflation” is the worst thing ever, right, all you very serious people out there?]
We cannot say that this particular context, involving a “newly characterized antigen” and a functional genus claim to corresponding antibodies, is one in which the underlying science establishes that a finding of “make and use” (routine or conventional production) actually does equate to the required description of the claimed products.
Got that? The required description is a description that establishes at least that the claimed structure is different from the prior art structures, and also (for a functionally claimed genus) that sufficient species have been taught. Here it is again:
An adequate written description must contain enough information about the actual makeup of the claimed products —“a precise definition, such as by structure, formula, chemical name, physical properties, or other
properties, of species falling within the genus sufficient to distinguish the genus from other materials,” which may be present in “functional” terminology “when the art has established a correlation between structure and function.”
Note that there is ZERO established correlation between structure and function in the world of writing instructions that teach computers how to apply logic to information. If there’s any doubt about that, look at the following: 10100101011101010010010010110110110111101010100010101010010101010101001010100101010100001001001001111111000101001010
And tell me what it means (I know exactly what it means — surely there is an “expert” among you who can decipher that using the function/structure relationship that I’ve heard blog t r 0 ll s here insist is a real thing … so go ahead and tell me what it means).
Bottom line again: this is one half of the squeeze play. Logic patenting could be over in a month if “zealous” lawyers start using their br @ ins and stop tying their hands behind their backs.
More good stuff:
Section 112 requires a “written description of the invention.” But this [rejected] test allows patentees to claim antibodies by describing something that is not the invention, i.e., the antigen
More accurately, the rejected test allowed patentees to claim allegedly new antibodies according to their function, i.e., their alleged ability to bind to a newly described antigen.
The test thus contradicts the statutory “quid pro quo” of the patent system <— ooooh, lookie there
where “one describes an invention, and, if the law’s other requirements are met, one obtains a patent. Indeed, we have generally eschewed judicial exceptions to the written description requirement based on the subject matter of the claims. <— hint hint hint See Univ. of Rochester v. G.D. Searle & Co., 358 F.3d 916, 925 (Fed. Cir. 2004) (noting that “the statute applies to all types of inventions”). <— OUCH And Congress has not created a special written description requirement
for antibodies as it has, for example, for plant patents. <– sticking the knife in right there.
Careful readers of this blog will note that I have made every one of these arguments here already. This is going to be an awesome year.
Look, if the SCotUS and CAFC were to have advanced these arguments (i.e., arguments sounding in the §112(a) written description requirement) against software claims, I would never have said “boo” against Alice and its sequelae. The problem is that they insisted on grounding their holdings in §101, instead.
I agree that purely functional claiming is untenable and problematic. The right way to solve the problem, however, is through §112. Trying to deal with these claims through §101 creates as many problems as it solves.
The problem is that they insisted on grounding their holdings in §101,
*sigh*
All these statutes are related, Greg, because they’re all driven by the same underlying policy. Logic isn’t eligible for patenting. Never was. In part, that’s because it doesn’t have any structure in the patent-meaningful sense of the term. That lack of structure is what creates the written description problem. It’s a seamless web.
What happened in our patent system is that rather than recognize that *if* the public thought software patents were a good idea then Congress should come up with a workable system for submitting evidence/art and examination, some judges who were incredibly cl u e le ss decided for themselves that the patent system must be bent around what was plainly understood to be ineligible. And that’s what happened.
The Supreme Court eventually was forced to step in and undo a giant chunk of that judge-made nonsense because it was — in the minds of pretty much everybody EXCEPT for a tiny tiny number of self-interested abusers and their attorneys — the worst nonsense ever to infect any patent system, ever. The mistake the Supremes made in Alice was to not go all the way. I think they assumed that the worst attorneys in the world would get the message and back off but that’s not how these people work.
Your feelings – as wrong and misguided as they are – are noted.
The problem is that they insisted on grounding their holdings in §101, instead.
Their holding is the original reason though. If you look at Morse the theory behind the “101” (and you’ll see why I use quotes in a second) in that case is that there was no basis to believe Morse’s inductive leap that because he had invented a single means of printing at a distance that he had invented all means of printing at a distance.
Instead, they reasoned, that a future inventor may come up with a completely different means of achieving the result that was better than Morse’s, and Morse’s claim would preclude that. Further, Morse himself could improve upon his own means and would have no reason to patent it, as he already had the broader patent for the field. In other words, they found a 112a WD violation because Morse had failed to provide sufficient evidence to disprove the species/genus question.
When Alice talks about “abstractions” and “preemption” its clear that they mean the same thing that WD was invented to fix. *The problem is that the Supreme Court came up with it first, the CAFC ignored it, then realized it was necessary, and reinvented the wheel using a different test.*
The notable distinction between 101 and 112a is that 101 functions without proof (i.e. the fear that you are preempting triggers the invalidation) while 112a requires proof (i.e. one must show that the examples don’t fill the scope). The Supreme Court test is the proper test, because a perfectly functioning patent (i.e. one everyone had knowledge of) would chill the creation of the very evidence necessary to prove its invalidity – nobody invents the gun if a bow patentee has the overbroad scope of “all machines that fire projectiles” because it subjects them to liability of all their profits and they can’t recoup their inventive costs with a valuable patent.
That’s the underlying issue with how lawyers who have a problem with 101 think – the burden of proving your scope is properly sized (and therefore not abstract) actually *lies with the patentee and is not subject to post-issuance deference.* That’s the only major difference between a concrete/abstract dichotomy and a species/genus evidentiary question. A 101 inquiry and a 112a inquiry are largely the same thing.
The fact that the 101 test appears to be bad is because the federal circuit won’t generate the test, and instead keeps analogizing things to other cases. This case could have been taken care of under 101 – the functional result here is a naturally occurring effect of nature, and the patentee doesn’t provide anything significantly more because there’s no limitation to specific structure
Except for the fact that the Court explicitly rejected your theory…
The notable distinction between 101 and 112a is that 101 functions without proof (i.e. the fear that you are preempting triggers the invalidation) while 112a requires proof (i.e. one must show that the examples don’t fill the scope).
It’s not true that 101 functions without proof. It *can* function without proof in the most extreme situations (e.g., where you claim logic as logic). But proof is certainly required in the usual cases, specifically the proof that the non-abstract elements (as recited in the claim, not separately) are in the prior art. That’s what Mayo and Alice were all about. Similarly, proof that a product is naturally occurring is certainly required to find ineligibility under a product of nature theory.
The far bigger “evidentiary” distinction is that 101 doesn’t care about clarity and possession but the written description requirement plainly does.
If you look at Morse the theory behind the “101” (and you’ll see why I use quotes in a second) in that case is that there was no basis to believe Morse’s inductive leap that because he had invented a single means of printing at a distance that he had invented all means of printing at a distance.
I am with you this far. I agree that both (1) our modern §112 Ariad inquiry and (2) our modern Alice test for “abstract ideas” find their roots in the same Morse decision.
The problem is that the Supreme Court came up with it first, the CAFC ignored it, then realized it was necessary, and reinvented the wheel using a different test.
Here I start to part ways with you. I cannot agree that the CCPA/CAFC ever “ignored” the insights of Morse or found that they needed to “reinvent” the old Morse test. It is simply the case that when Morse was later codified, the CCPA found it more workable to locate the principle that Morse articulated—to ground its actions in statute—in §112 rather than in §101.
The Supreme Court test is the proper test…
Now you totally lose me. The CCPA’s application of Morse (“has the patentee claimed more than s/he described in the specification?”) is infinitely easier and more reliable of application than the circular and largely incoherent Alice two step. The mere fact that Alice, on its own terms, only kicks in when the claim is “directed to” an exception category—where no one can give a coherent account of what it means for a claim to be “directed to” something (because this phrase certainly does not mean merely “recites”)—tells you everything you need to know about the care and thought that went into crafting this test. Pure nonsense.
As explained above, the use of post-priority-date evidence to show that a patent does not disclose a representative number of species of a claimed genus is proper.
This approach to enablement puts a large swath of patent claims in limbo. If the invention discloses 2 of the 3 known species, a claim to the genus is held valid. But if 15 years later 7 more species are discovered, the patent all of a sudden is invalid?
This reasoning comes down to a disagreement with the Patent Act. Many simply believe that 20 years of exclusive rights to certain discoveries is improper, particularly when the patentee does not fully develop those rights.
It is like cancelling a claim held by a silver miner because he was not mining for copper.
This approach to enablement puts a large swath of patent claims in limbo.
And deservedly so.
If the invention discloses 2 of the 3 known species, a claim to the genus is held valid. But if 15 years later 7 more species are discovered, the patent all of a sudden is invalid?
Look, if you discover two species of antibody, there is nothing to stop you from claiming those two antibodies. If, however, you will insist on claiming instead “all antibodies that bind to antigen L,” then you have no one to blame but yourself when seven more species of antibody that bind antigen L are discovered (post filing date) that have no obvious structural relation to your two, and these discoveries are cited to show that your (unwarrantedly broad) claim lacks written description.
Our public policy is and (rightfully so) always has been to encourage design-arounds. Our law has long regarded the prohibition against practicing this invention as a spur to clever people to think up new inventions that accomplish the same end by different means. If one is allowed—instead of claiming the antibody that one has discovered—to claim all antibodies that achieve the same effect, one saps this spur to further invention of any actual motivating force.
This approach to enablement puts a large swath of patent claims in limbo. If the invention discloses 2 of the 3 known species, a claim to the genus is held valid. But if 15 years later 7 more species are discovered, the patent all of a sudden is invalid?
That’s because nobody actually invents the genus of a functionality. If a single structure caused a result, the claim would be to the structure. The claim to the functionality is just to get an overbroad scope from what was invented.
The claims are not in limbo, the claims were invalid when drafted. If you invent a bow, you claim the structure of your bow, you don’t claim “a machine that fires projectiles” and hope nobody invents a gun in the interim (and you definitely shouldn’t be dumb enough to sue the gun guy after he does)
So you do not believe in functional claiming?
you do not believe in functional claiming?
Functional claiming at the point of novelty is illegal. It’s not a matter of “belief”. You can’t claim a new composition by distinguishing it from an old composition purely on the basis of its function. Or, more accurately, you can claim it that way but the claim is invalid and if you try to enforce against a good attorney you will lose.
“at point of” and “purely” are two very different things.
Reference here Prof. Crouch’s coined term of Vast Middle Ground…
(move that goalpost back please)
Of course I believe in functional claiming. I also believe in invalid claims. I have seen both with my own eyes.
Greg also believes in being a pedant too (and in more than one way).
Try this. Write your claims for a bow using only structure. Give me an hour and I will copy your invention without infringing your claims.