Section 112 Enablement and Written Description After Amgen: Federal Circuit Doctrine Reshaping Patent Prosecution Practice

Friday, December 5, 2025 

Executive Summary

Instructor(s)
Miku Mehta, Procopio
Matthew Avery, Baker Botts
Anna Hyatt, Dren Bio
Kanda Ishihara, Lyft

Keywords

35 U.S.C. § 112 enablement written description 2025 • Amgen v. Sanofi genus claims enablement standard • Agilent Technologies v. Synthego enabling prior art 102 • US Synthetic Corp. v. ITC Wands factors polycrystalline diamond • Mondis Technology v. LG Electronics written description prosecution amendment • Fintiv v. PayPal means-plus-function nonce words indefiniteness • patent prosecution 112 rejection strategies • written description possession of invention standard • “how does Amgen v. Sanofi affect patent prosecution,” • “what is the difference between enablement and written description under 112”

Legal Analysis 

Enabling Prior Art After Amgen: Why the Enablement Standard Under § 112 Does Not Govern § 102 Anticipation

Agilent Technologies, Inc. v. Synthego Corp. presented the Federal Circuit with an opportunity to reconsider whether the heightened enablement standard articulated in Amgen Inc. v. Sanofi, 598 U.S. 594 (2023), should govern the threshold for prior art to qualify as anticipating under 35 U.S.C. § 102 — and the court declined to extend Amgen in that direction. The patents at issue claimed guide RNAs for CRISPR systems; the invalidating reference was an abandoned patent application that, as Avery emphasized, “was just full of prophetic teachings and the examples in that they had in there in fact did not work and that’s why this company abandoned the application.” The Federal Circuit affirmed the Patent Trial and Appeal Board’s finding that the reference was nonetheless enabling for § 102 purposes, reaffirming longstanding doctrine that prior art need only disclose a single enabled embodiment falling within the scope of the challenged claims — a materially lower threshold than the full-scope enablement required of a patent applicant under § 112. The court thus declined the argument, which Avery characterized as a “clever” one, that Amgen‘s requirement that an applicant enable the “full scope” of its claims should import a symmetrically demanding standard for invalidating prior art. The practical tension this creates is real: as Avery observed, “a junk reference full of prophetic teachings that did not in fact work is good enough to anticipate my claim where I actually proved that it does work.” Mehta raised a collateral implication for information disclosure obligations, questioning whether the divergence between the enabling standard for prior art and the enabling standard for the applicant’s own disclosure requires more aggressive IDS practice where co-inventors have worked in related spaces; Avery suggested that, given the post-2014 high bar for inequitable conduct unenforceability, the practical litigation risk of under-disclosure remains limited, while Hyatt counseled “when in doubt, if you know about it, just cite it.” Avery also predicted, with obvious irony, that the decision may incentivize strategic publication of speculative disclosures — “AI slop that just includes all sorts of prophetic references” — as a defensive prior art generation tactic, a concern that underscores the normative tension at the heart of the holding.

US Synthetic Corp. v. ITC and the Wands Framework: Enablement for Composition Claims in Predictable Arts

US Synthetic Corp. v. International Trade Commission illustrates both the durability of the Wands factors as the governing framework for § 112 enablement analysis and the limits of attempts to extend Amgen‘s functional-claiming critique beyond the antibody context into material science. The claims at issue covered polycrystalline diamond compacts — composition of matter claims defining specific magnetic properties correlated to the absence of a particular manufacturing process — and the intervenors argued by analogy to Amgen that the claims were not commensurate with the disclosure because they purported to cover every non-leaching process. Hyatt explained that the Federal Circuit rejected this argument on the ground that “the spec had really detailed manufacturing guidance in the working examples,” and further reasoned that because the claims were composition claims rather than process claims, the applicant was not required to disclose every possible method of manufacture. The court’s emphasis on the ability to anchor claimed functional properties — coercivity and magnetic saturation — to reproducible structural measurements and process parameters distinguished this case from Amgen‘s unnavigable genus of functional antibodies; Hyatt summarized the holding as requiring that the applicant “teach what the invention is, how to make it, and demonstrate that it worked across the claimed ranges.” Hyatt identified the degree of predictability in the relevant art as the critical variable: “in the more predictable spaces, you might end up in a better spot with less data,” whereas in the antibody space, “I don’t know that it’s entirely clear that we know how much data is enough, or if there is such a thing as enough data to get a broad claim.” Avery raised the prosecution practice question of whether applicants can affirmatively build intrinsic support for the predictability and low-experimentation Wands factors by including qualitative statements in the specification to that effect, though the panel acknowledged the strategy cuts both ways depending on the art. Mehta drew attention to the practical implication for inventor intake: highly skilled inventors often describe inventions at a level that reflects their own deep expertise rather than the ordinary skill standard, making it incumbent on prosecutors to ask “more basic questions to get things in the specification” that would actually satisfy Wands. Irving noted from the audience that § 112 alone — without any § 102 or § 103 prior art grounds — has become an increasingly standalone basis for claim invalidation, pointing to multiple recent Federal Circuit decisions where enablement or written description was the sole dispositive ground.

Written Description, Prosecution Amendments, and the Mondis Warning for Continuation Practice

Mondis Technology Ltd. v. LG Electronics, Inc. — involving a display control patent prosecuted around 2007 — demonstrates how a claim term added during prosecution to distinguish prior art can fatally outrun the original specification’s written description, and how that vulnerability is compounded in continuation practice when inventors are no longer engaged in the prosecution process. The specification and drawings in Mondis described a protocol for matching a “received ID number” with a “registered ID number,” but during prosecution — in response to an examiner interview — the applicant proposed and the examiner agreed to amend the claim to refer to “a type of display unit,” a broader formulation that the Federal Circuit held was not “reasonably conveyed” in the original disclosure. Mehta emphasized that this prosecutorial exchange, while effective at the time in overcoming a prior art rejection, illustrates how what once worked — “changing between ‘the’ and ‘an'” as she put it — is now exposed to written description invalidity under current doctrine. The Federal Circuit’s holding reaffirmed the distinct purposes of the two § 112 requirements: “enablement is explaining to the public how to use the invention; written description is showing that the inventor was in possession of the invention — they’re two different things,” as Mehta stated. Ishihara, who had litigated an earlier version of the same patent family, offered a sobering observation about continuation practice: in that case, “nobody was talking to those inventors anymore” and “these patents were all crafted 100% with somebody’s target product in front of them,” with prosecution counsel working entirely from the accumulated file history rather than from original inventor input. Mehta suggested that one prophylactic — reviving the once-common practice of obtaining an updated inventor declaration whenever claims are amended — may be worth revisiting for commercially important patent families, noting that “the inventor is declaring that they invented the claimed invention, so if these are really important patents and you wanna be sure, keep the inventor engaged.” Turning to Fintiv, Inc. v. PayPal, Inc., Ishihara explained that the “Payment Handler” claim term was found to be a nonce word invoking means-plus-function treatment under § 112(f) because it described “hardware that performs specified functions without indicating the how,” and the corresponding specification structure consisted only of identical claim language restated verbatim — a drafting failure that Ishihara characterized as a cautionary illustration of the need to “audit all your terms and make sure that you have enough details in there,” attending to what the term does rather than what it is called.

Generated by AI based on the Interview/Transcript below.

Key Takeaways

• Enabling prior art threshold remains lower than § 112. The Federal Circuit in Agilent confirmed that prior art need only disclose a single enabled embodiment to qualify as anticipating under § 102, regardless of Amgen‘s full-scope enablement requirement for patent applicants — creating an asymmetry that, as Avery observed, means “a junk reference full of prophetic teachings that did not in fact work is good enough to anticipate my claim.”
• Wands factors govern composition claim enablement. US Synthetic reaffirmed that the eight Wands factors remain the controlling framework for enablement outside the antibody context, and that composition claims tying functional properties to reproducible structural measurements can survive enablement challenges even where some experimentation is required.
• Predictability of the art is the dispositive variable. Hyatt predicted that broad functional claims will remain acutely vulnerable to enablement challenges in less predictable arts like large-molecule biologics, while applicants in more predictable fields — material science, software — may sustain broader claims with less data: “in the more predictable spaces, you might end up in a better spot with less data.”
• Prosecution amendments must be checked against the specification. Mondis illustrates that claim language added to distinguish prior art — even at an examiner’s suggestion in an interview — will be invalidated for lack of written description if it exceeds the original disclosure; Mehta cautioned that “when they propose something, you also check your specification.”
• § 112 alone is now a freestanding invalidity ground. Irving observed that Federal Circuit decisions invalidating claims on § 112 grounds alone — without accompanying § 102 or § 103 findings — have accelerated, making enablement and written description analysis an independent priority in any invalidity assessment rather than a fallback to prior art positions.
• Nonce words without algorithmic structure invoke means-plus-function. Fintiv establishes that functional terms like “Payment Handler” — where the specification simply restates the claim language verbatim without disclosing the operative algorithm — are indefinite under § 112(f); Ishihara counseled drafters to “go through and highlight what the term does, not what it is,” and to treat any noun ending in “-er” as a red flag.
• Inventor engagement during continuation prosecution is essential. Ishihara’s observation from Mondis litigation — that the patents were crafted entirely around a target product with inventors who “never saw prosecution counsel after the patent application was filed” — underscores the written description risk created when continuation claims are amended without reference to original inventor knowledge or updated declarations.
• Duty of disclosure remains relevant but litigation risk is limited. Avery noted that following the 2014 tightening of the inequitable conduct unenforceability standard, the practical litigation risk of failing to disclose prior art remains “pretty minimal,” but Hyatt maintained that the prudent practice remains disclosure when in doubt, and Mehta flagged that the Agilent enablement asymmetry may warrant more aggressive IDS practice with respect to related inventor disclosures.

B-CLE Recording (CLE: $100) | Youtube Recording | Resource(s) | Speaker Bio(s) & Contact Info

Download the interview/transcript here!

Interview/Transcript

This interview/transcript was based on a conversation on December 5, 2025 at the 26th Annual Berkeley-Stanford Advanced Patent Law Institute. This transcript features the first prosecution panel, Update on 112–Understanding and navigating the PTO’s application of the law, given by Matthew Avery, Baker Botts; Anna Hyatt, Dren Bio; Kanda Ishihara, Lyft and moderated by Miku Mehta, Procopio.

[WAYNE]
Good morning, everyone.

[AUDIENCE MEMBERS]
Good morning.

[WAYNE]
So it happens when you put a trial lawyer with a very sensitive mic. Well, thank you for coming this morning. Again. Couple of things, we’ll be recording these kind of sessions.

If you have questions, work with Miku on them. If you do have questions, make sure that Willie will get you a mic to, get them on the recording or we’ll repeat them up here for you. But it’s kind of important, we want to capture all of this knowledge for any of these discussions.

And in this type of room today, more nuance, we’re hoping for more discussions when appropriate. So with that in mind, our first session for the day, we start with, with Miku Mehta from Procopio.

Miku’s been with us for many, many years here at the Institute leading these panels and trying to drive discussion. So with that, I’m gonna get out of the way quickly and let Miku guide us through 112.

[MIKU MEHTA]
Thanks, Wayne. And thanks for bringing us all together for this program. We’re very excited today to talk about the maybe least controversial topic from yesterday’s discussions, which is 112.

You notice it wasn’t mentioned much yesterday, when we talked about the courts and the USPTO diverging from each other. Everybody seems to be marching in a similar direction on 112. There wasn’t much discussion of IPRs, ’cause it’s not the subject of IPRs, maybe PGRs.

And it’s getting more important. I hope this presentation is interesting.

And I’m very excited to have on our panel and to be having join us people from the industry and law firms. I think Kanda, Anna, and Matt, if you would like to briefly introduce yourselves, it’ll be great.

[KANDA ISHIHARA]
So hi, everyone. Good morning.

I hope everyone’s awake. I’m barely awake, I feel like. But it’s nice and cold, so, I’m Kanda Ishihara.

I’m at Lyft. I manage the patent portfolio there. I’ve been there for almost five years. We actually, on a day-to-day basis, don’t have very many issues with 112, so I feel like this is, like, the most I’ve thought about 112 in a long time in preparation for this panel.

But I’m, I’m a self-proclaimed prosecution nerd, so happy to have this discussion.

[ANNA HYATT]
Good morning, everyone. I’m Anna Hyatt. I’m currently the head of IP at a biotech company in San Carlos called DrenBio.

Before that, I spent about 10 years in the IP legal group at Genentech. Most of my prosecution experience is in the large molecule antibody space, where 112 is constantly on one’s mind as a prosecutor.

So, looking forward to having a fruitful discussion today.

[MATT AVERY]
Hey, everyone. I’m Matt Avery, partner at Baker Botts in San Francisco. I’ve been there my entire career, since ’09 since I graduated from Hastings where I also now teach patent prosecution. So, I love talking about patent prosecution basically not getting paid for it. So, glad to be on this panel.

[MIKU MEHTA]
Okay. Thanks, team. And uh, so as Wayne mentioned, my name is Miku Mehta.

I’m at Procopio Firm. I’m– I’ve been doing patent prosecution and soft– software and computer hardware and energy area probably for maybe about 25 years, and also mainly doing the drafting and prosecution. So joining the self-proclaimed nerd I’m also one on 112.

So just on the format, we’re going to have a presentation on a series of four cases from today. And we’re gonna go through– we do have slides. Apologies for the slides. We have one for each case.

And talk a little bit about how this impacts prosecution. You should know, and maybe you know, that over the last 15 years, examiners have been receiving lots and lots of training.

The USPTO has hundreds of slides on their website on how to make some of these rejections and implement some of the law that’s come down. They’re– in the last 10 years, going back to that Citrix case, there’s also been Supreme Court cases.

A lot of this is in training. My numbers show that the number of 112 rejections and 112citations has been going up, like, pretty quickly in the last 10 years, and pretty progressively, so we’re getting good clarity of record on that.

So we’ll dive into the cases, and as we get to the end, we’ll have time for discussion. So, I think everybody’s here for the 112. Is there anybody who has never experienced a 112 rejection in office action, who does patent prosecution? If you have, you can share your wisdom with us.

So we’re all in the same boat, and you’re all seeing it here. So I’m gonna try to click. So, look back, and see this is help, and our disclaimer, these are our own views, not the views of our organizations, since this is being recorded and maybe will be listened to again.

(laughing)

So Matt, do you want to lead off with the Agilent Techs. case?

[MATT HENRY]
Sure. So, first case we’re gonna talk about is Agilent case. Which is kind of a funny 112 case ’cause it’s maybe also more about 102 in anticipation.

But enablement was brought in, and it was a very interesting case, and we wanted to talk about this briefly today. I’m glad Professor Lemley’s not in here. He argued this case and lost it, unfortunately. But it would have been a little nerve-wracking if he was in the front row, critiquing my evaluation of his handling of the case here.

So, let me back up and preface, talk about some prior 112 case law, which is Amgen versus Sanofi. Which you’ll need to understand what’s happening in this case.

So if you remember a couple years ago, there was a Supreme Court case, Amgen, that claims were about essentially like functional antibody claims, so genus claims to a genus of antibodies. And the Supreme Court there found those claims to be invalid under 112 enablement, saying that like the spec basically didn’t have enough support for these functional claims. Notwithstanding the fact that the spec had like 26 working examples and also presented, you know, what was characterized as like a roadmap for coming up with more working, you know, functional antibodies. The Supreme Court said, “Well, you know, 26, “that doesn’t seem to be enough because we have “no idea how big this family could be.

“It could be thousands, it could be millions, who knows? So 26 may not be representative.”

[WAYNE]
And then also that, you know, the roadmap that they provided that was just basically like, “Here’s a trial and error kind of way to work through, and maybe you’ll find some more working, you know, functional antibodies. Maybe you won’t.”

And so that’s not good enough, so claims invalid under 112. And so that’s like, you know, where we are with 112 enablement this Supreme Court case from two years ago. So then we get up to this case.

And what happened here is we had two patents and the claims were directed to, sorry. So they were directed to guide RNAs for CRISPR systems. All right?

And the technology isn’t that important. One interesting thing is that they were challenged in IPRs and the lead reference they brought against it was a 102 reference that was an abandoned patent application that taught similar technology. All right?

So also taught guide nucleotides. So at the PTAB the IPRs, the PTAB said, “Okay, Yeah this looks like it’s anticipating art. We’re gonna find these claims invalid under 102.”

So it goes up to the Federal Circuit what they try what Professor Lemley tries to argue is that, well, one, this was not enabled prior art. It was not enabling prior art, right? And so, you know, those of you that have dealt with 102 rejections, right, there’s this issue like that you could raise, try and get around the 102 rejection and say like, “Oh yeah, well, I mean, it may teach that on its face But it’s not actually enabled.”

And so that’s what they tried to do here of course on appeal that’s a matter of fact. And so, you know, the, the Federal Circuit just said, “Yeah, we’re gonna defer to, you know, the, the PTAB on their factual finding that it is in fact enabling.” So then we get to the clever argument that Professor Lemley tried to make which is that, “Well hey, you remember this Amgen case from two years ago that said that, you know, your spec needs to enable the full scope of your claims here, right?” So what he argued was like, “Well, you know, this abandoned patent application that you’re using is a 102 reference.

That’s not enabling for the full scope of my claims, right? If it was enabled, they would have had this claim.

They wouldn’t have abandoned it, but, you know, their, this prior art reference was abandoned because it was just full of prophetic teachings and the examples in, that they had in there in fact did not work and that’s why this company abandoned the application. And so, you know, why could this now then be enabling prior art to invalidate, you know, my claim where I actually prove that this thing does work? So Lemley kind of focused his argument on, you know, go back and look at Amgen which said you know, that your claim needs to be fully enabled by, you know by the specification and said it should also apply for enabling prior art.

So kind of the, the key holding of this case is that Federal Circuit said like, “No that’s, you’re confounding things. The enablement standard under 112 for enabling a claim is different than what it takes to be enabling prior art for anticipation purposes.” Which is like, okay, great, that’s something that we all probably assume should be the right way, right? That you know, the standard for getting a claim on something and getting your limited monopoly, like that should be higher, right, in terms of the what you have to show in your spec.

The thing that bothers me about this case, and, and what I liked about Lemley’s, you know, clever argument here is like, “Well hey, but this was an abandoned reference, right? It was an abandoned application and they weren’t able to, you know, they wouldn’t have been able to get this claim that we’re going after.

Yet you’re saying that it’s good enough to invalidate my application where I actually showed that this thing works.” And so that seems to be some tension there at least.

You know perhaps arguably in unjust that, you know, kind of this junk reference full of prophetic reference you know, prophetic teachings that did not in fact work is good enough to, you know, and anticipate my claim where I actually prove that it does work. So you know, I think that’s one thing that bothers me about it even though, you know, as a prosecutor, I’m, I’m happy to be able to get my claims easier and don’t have to you know, worry about that as much.

[MATT AVERY]
But, you know, I think it also is problematic in that, maybe it’s opening the door to people making more junk disclosures and like, I don’t know, pumping out AI slop that just includes all sorts of, you know, prophetic disclosures just to, you know, muddy up the waters because apparently that is perhaps good enough to be anticipating, you know, enabled anticipating prior art under 102. So that’s the, the Agilent case.

Miku.

[MIKU MEHTA]
Thanks. Thanks, that’s, you know, that’s a really good point. So, I was just thinking about this.

I don’t know how this impacts our duty of disclosure, if at all in terms of, well, we have our own publications or our own patent applications that are enabled to a certain standard. But then, where there are different inventors, do we have to cite more in where our other inventors have invented in our companies, and that’s prior art, should we be more aggressive in filing information disclosure statements given the difference?

I don’t know. I mean, it’s–

[MATT AVERY]
Good question. I’d like to see an inequitable conduct case where they actually found claims unenforceable because I failed to cite prior art, right? I mean, like–

[MIKU MEHTA]
Yeah, it doesn’t–

[MATT AVERY]
It’s been a long time, right, since the law changed back in what was it? 2014, right? Where they raised the unenforceability standard.

So, I mean, obviously you should disclose everything, right? Disclaimer that you know about. But, I mean, like, the risk there, I think is pretty minimal from a practical standpoint.

[ANNA HYATT]
But maybe when in doubt, if you know about it, just cite it.

[MATT AVERY]
Right. Unless you’re about to go to 51 references for that extra fee. And then you got to think about it a little harder now, right? They’re punishing us for disclosing more, so that’s unfortunate.

[MIKU MEHTA]
So, any other comments from, from a panelist on this case?

[MATT AVERY]
One thing that I was thinking about was maybe, maybe what I should be doing when I’m drafting now is fill up my own applications with AI slop, right?

(laughing)

To thwart future patentees, right? I’ll just like fill it with a bunch of prophetic references to, you know, deter you know, people trying to, you know, get patents in the same space in the future because that’s, that’s good enough to thwart my competitors, perhaps.

Of course, you get into the problem when that app publishes And then it’s, you know, prior art against your own subsequent applications in 18 months, right? So that’s why you make sure you then file all your applications like 17 months and 29 days before, you know, after that AI slop app you filed, right?

[TOM IRVING]
Yeah, dedication disclosure pushed to the limit, right?

(laughs)

So all good. All good. So there’s any questions or comments from audience on this case? Yeah.

[KANDA ISHIHARA]
Good morning? Been holding several cases in the past. So I’m not sure? I’m not saying is just unreasonable but a case that does not have to be England is not a real case.

[MATT AVERY]
Right. So the comment there was like, there’s a lot of case law in this space, essentially along the same lines, that like you know, and which the Federal Circuit repeated here. It’s like, “Yeah, we are standing by our prior case law that You know, for prior art to be enabling, you just have to have like a teaching of a single embodiment that’s enabled, that would fall within the scope of your claims,” right?

So that’s been repeated many times. So really, you know, what is new about this case is holding that up in the context of Amgen from 2023. You know, and what Synthego was trying to do, which I thought was a clever argument, was like, “Hey, maybe it’s changed now because look, you’ve got this new Supreme Court case.”

And the Federal Circuit said like, “Nope, we don’t think that changes anything.” So, yeah.

[TOM IRVING]
Yeah. The other thing I noticed, so many times we look at invalidity defenses or we’re asked to review invalidity, and it used to be where we would say, “Okay, well, if there’s not a prior art invalidity position, then the patent must be pretty strong.” But you can see so many Federal Circuit cases where 112 alone is now the basis for invalidating claims without prior art invalidity, and that’s happening more and more. Last night when I was looking, there were two more Federal Circuit cases from last month and one from last week, where it was just almost 112 alone.

So it’s that, it’s not just that it’s the law, it’s that it’s actually being used to invalidate patents on its own without you know, combination of 102, 103.

[MIKU MEHTA]
So that’s, you know, one of the things we’re seeing, so. The next case we have, and Anna, you’ll have this one, US Synthetic Corp.

[ANNA PIETRETTI]
Yes. To continue on with the theme of enablement, this next case, US Synthetic versus ITC, this is a Federal Circuit decision came out early this year. It’s in the context of composition of matter claims. Really, sort of spoiler, I’d say the case really reinforces the idea that Wands remains the governing framework for enablement analysis. So, and what I found interesting about studying this case was that it’s in the material science space, whereas I generally practice in the large molecule antibody space, as I mentioned, and I got a little bit jealous reading this case because a lot of times this case came down to a lot about being–predictability was one of the Wands factors, and what you’re able to enable and how you’re able to tie structure to function.

Just a little bit of background. So USS is, manufactures these compounds called polycrystalline diamond compacts, or PDCs. They’re essentially like these super hard cutting elements that can be used in drill bits. So the claims at issue were composition of matter claims, and they define specific properties like magnetic properties and they also said, tied these properties to a functional property, which was that they were produced without this process called leaching.

So initially, the administrative law judge and the ITC found that the claims were in fact infringed, and held that they were not invalid under 112, but they found them to be invalid under Section 101. So it went up on appeal and there, the Federal Circuit reversed the 101 ruling and also left the enablement finding intact.

But the intervenors had used that as a fallback position to 101. They– they said, you know, that these claims are invalid under 112, and what they did is they tried to analogize it to the recently decided Amgen v. Sanofi case. And they argued that the claims were not enabled because they covered every process under the sun that did not include leaching, so essentially saying that the claims were not commensurate with the scope of the disclosure.

[ANNA HYATT]
It was an interesting strategy, but the Federal Circuit rejected this argument, and they said that actually the spec had really detailed manufacturing guidance in the working examples, and also pointed out that because the claims themselves were not process claims, that they didn’t have to disclose every possible method of manufacture. And so, what they really emphasized as well is they said, you know, looking at the Amgen case, they said that this is the same statutory enablement test that’s been applied for over 100 years. One of the key holdings was that experimentation is permitted. And that the respondents had failed to prove that there was undue experimentation, I think really going back to the Amgen case, the issue there is they said that, you know, there’s untold numbers of antibodies that could have fallen within the scope of that claim, and all these specifications really said is, “Go find them,” that would have required an undue amount of experimentation.

And so, here they said, “You need to apply the Wands factors.” Probably everybody in this room knows what those are, but looking at what’s the quantity of experimentation needed, how much direction of guidances are in the spec, what do the working examples show, the nature of invention, the state of prior art. And key here is the predictability.

So what they found here was that the specification did describe the scientific basis that linked the functional properties, in this case, things like coercivity and magnetic saturation to the structural properties, which were mean free path and the cobalt content. And then the specification also disclosed process parameters that would produce the required structure, and in addition to that, they also included the measurement standards that you would need to use. And they had working examples where they compared all of their PDCs to the prior art and showed how they worked better.

So, sort of in conclusion, the Federal Circuit said they were enabled because they taught what the invention is, how to make it, and they demonstrated that it worked across the claimed ranges and anchored the claimed properties to reproducible measurements. So, I think some of the key takeaways from this case are that you can require some experimentation and still have the claims be enabled. It helps when it’s in a more predictable art. I think that this is still going to be a significant challenge in the antibody space, for example, where it’s much more difficult currently to tie structure to function.

Maybe in the future with more AI models and things like that it will become more predictable to be able to predict function based in structure. But here, they were able to measure specific properties and correlate those to structural features. So, I think just some sort of takeaway drafting, you know, practice lessons here might be we still can’t rely on fully functional claiming, but as long as you have the data in your specification and you can tie structure and function together, you’re in much better shape.

And keep in mind that the Wands factors reminds remains the governing framework and to draft with them in mind. And there, I think I’ll stop there in case there are questions. Mm-hmm.

[MIKU MEHTA]
Yeah. Thank– Thanks, thanks Anna. That’s really good points on the difference in the level of skill in the art in different technologies. So I keep thinking, well, if we’re working with different inventors, like even within an applicant, you can have different technologies from different inventors. How does that impact your, how you do your invention intake or disclosure? Because I think the inventors are at not at like an ordinary level of skill in the art.

They’re at a very high level of skill in the art and when they tell us something, they may not be thinking about enablement the way that we do. And we have to ask sometimes some basic– more basic questions to get things in the specification. So what, like in view of this case, would we change how we approach inventors to ask them some questions differently based on different technology areas, different levels of skill?

[ANNA HYATT]
Yeah, I think that’s a great question. I mean, I think generally what I have to remember to do is when they bring something to me where, “Hey, we did this, we achieved this great result. We got this structure.

We got this product.” And they say, “Here’s how we did it.” And they kind of, you know, mic drop.

But if you want to get some breadth in your claims, it does, it is helpful to include in the specification multiple ways that you could do that, and not just the one way that you did do it. So I think going back and saying, “Hey, you know, I know we did it this way, but are there other ways that would work as well? Let’s make sure we describe those.”

[MIKU MEHTA]
And is that, that’s more or less consistent with what Amgen is, is saying in terms of disclosing more or less every, every species you want to claim?

[MATT AVERY]
Yeah. I think so in terms of like, I think what the takeaway from Amgen is, is you gotta provide more guidance on like how to, you know, if you, you’re claiming a family, like how to interpolate other, other functional members in that family or extrapolate out, you know however, however, you know, what your disclosure is and the direction you’re trying to go in your claims. And certainly, you know, Miku, to your point, like what should we be doing during disclosure meetings is like if the inventors have a few cool examples, but we wanna try and claim broader, like ask them, “Well, how, what if you wanted to go in this direction or that direction? Like what, what would be the guidance that you would provide to someone?

How would you do that?” You know, sometimes the inventors know, sometimes they don’t though.

[ANNA HYATT]
Maybe I can add to that too. In the large molecule antibody space, so sometimes by the time somebody comes with an invention disclosure, they’re really focused on a particular candidate or a lead. And so one thing I’ll look at is, look, how narrowly do we have to claim here? Are we really just claiming this one candidate?

Do we have other molecules that do the same thing? What else have you made and what does the data look like? And so really having to go back and say, “Can, what else do you have?

What else can we include here?” Especially if you wanna get scope, we really need the data to support it. And sometimes we even say, look, if we want to get additional breadth here, say we wanna claim a particular consensus sequence, for example, then we need sometimes to go back and generate more data to show that the full scope of what we’re claiming, you would get the functional properties that you’re claiming.

[MATT AVERY]
So I’m wondering you mentioned something about drafting with the Wands factors in mind. And, you know, one of ’em is quantity of experimentation and predictability, right? What do you think about the idea of like just putting in your spec like, oh yeah, it would be easy to do this experiment like on, you know, on its face, just put in like these qualitative terms about like, oh yeah, it would not be, you know very hard to do this further experiment to get to where I wanna go to.

Or yeah, it would be totally predictable to arrive at that, you know, just so you have intrinsic statements in the spec you could point to, to say like, oh yeah, this, you know, I’m, I’m totally supported by the Wands factors.

[ANNA HYATT]
We definitely try that.

(laughter)

Somebody has a question.

(unintelligible)

[MATT AVERY]
Well, it cuts both ways, right?

[MIKU MEHTA]
A suggestion at least, right?

[ANNA HYATT]
Yeah. I think the tension there, I mean, you know, a lot of in the antibody space specifically, a lot of the screening techniques are routine, right? And so it’s, that’s sort of not the issue, right? It’s actually just if you’re claiming a functional property, but it’s really hard to envision what are all of the antibodies that with, in specific sequences of their binding regions that would actually have that functional property. So I think there’s some tension there.

[MATT AVERY]
So too clever by half.

[MIKU MEHTA]
One thing I will say, when we look at all of these 112 cases is we can see how drastically prosecution practice has changed since a couple of decades ago, where if you had a 101 rejection, all you do is add in, like, non-transitory, or 112 rejection was about changing between the, and a, or an. It’s gotten a lot more away from just examining prior art and giving prior art rejections, which I feel like maybe, you know, 20 years ago, it was almost all prior art in all of these cases were focused on KSR. And with some of these patents that are in the Federal Circuit cases, they’re not as recent as in some of the other Federal Circuit cases.

The one I’m looking at is from, I think it was, like, from 2007 or 2008. You can ask, if you had the time machine and this invention was made, you know, after we had all these cases, would we be working with inventors the same way and drafting the same way? And in hindsight, we can always say the conclusion would’ve been different, but, you know, are the things that we’re doing now as practitioners different enough based on the changes in practice to be ahead of some of these things? That’s a, that’s not a question wait–

[ANNA HYATT]
Yeah.

[MIKU MEHTA]
For an answer, but it’s a kind of a something to think about in the way we communicate with inventors and examiners, right?

[ANNA HYATT]
Yeah, I think across the technology areas, it’s helpful in practice just to really have a conversation with the inventors about what data is there. Have you told me everything that might be relevant here? I think in the more predictable spaces, you might end up in a better spot with less data.

I think in the less predictable spaces, I don’t know that it’s entirely clear that we know how much data is enough, or if there is such a thing as enough data to get a broad claim.

[MIKU MEHTA]
Yeah. Yeah. So, any questions? We’ll have more time at the end, but any–

[KANDA ISHIHARA]
I think, to extending on the how, I feel like we also should extend on, you know, not how you do it right now, but how do you think it could work in the future, right? Yeah. In terms of being wholesome, in terms of including all that.

[MIKU MEHTA]
Right. And in the more predictable areas–

[KANDA ISHIHARA]
Mm-hmm.

[MIKU MEHTA]
The more predictable arts we have to think, because we may get new challenges that we didn’t get before.

[KANDA ISHIHARA]
Right. Right? Right.

[MIKU MEHTA]
So, any other comments or questions on this one?

[KANDA ISHIHARA]
Somebody– You, you, you explained about the antibody cases, and I’m wondering that, like, small molecule, Markush claim, is it more likely to get a 112 rejection model?

[ANNA HYATT]
So, I don’t practice in the small molecule space, but usually what I hear from my small molecule counterparts is that case law has been more settled there longer. They don’t seem to run into as many issues. That said, I’m really shocked, because I see those Markush groups that seem to encompass untold numbers of species.

So, great question. I don’t know if someone else– anyone else on the panel has comments on that, but–

[MATT AVERY]
No.

[MIKU MEHTA]
No, it’s a good point. Uh, I– I’m not in that space, so I can’t comment on it, but yeah.

[MATT AVERY]
So, I see we got 14 minutes left, Miku.

[MIKU MEHTA]
Oh yeah. So, we’re onto the we’re gonna move from enablement to written description, and we’re gonna talk about a case from about four months ago, but the patent was prosecuted about 18 years ago. So, you’re gonna have to put on your, go into your, the DeLorean and time travel back to 2007 or 2008 when it was just changing, you know, A to THE. To look at this case, the technology is related to display controls.

On your left is the main figure from the specification and on your right is the claim. So the key point is to have control of a display, there’s a protocol for matching the number. It’s like a layer address associated with it’s a registered number associated with a display and determining if that display is matched, and if it’s matched, the process or computer can take control of that display. If it’s not matched, you’re not allowed to take control of that display.

So, it’s just a display control technology. And in the specification, the specification and drawings referred to Step 4, which is comparing a received ID number with a registered ID number. That’s the disclosure in the specification. That’s the key point of matching.

I did look at the file history of this case, and the claims were all rejected based on the prior art. I’ll explain what the applicant and applicant’s representative discussed with the examiner in just a minute here. But then we look at the claims here on your right, and you can see I’ve underlined what was added during prosecution after discussion between the USPTO and the applicant.

The matching was changed to, from identifying set for a number, identifying said display unit to a type of display unit, the unit itself. So this was added during prosecution, right? And you can see there’s no type over here.

There’s some background discretion in the background section of the patent, but it’s not really relevant to this feature. There were, the office action, there were two interviews.

There was a first interview in which the applicant and the examiner agreed, and I made a screenshot of the information. It was from uh, March of 2007.

Examiner pointed out that the reference reads on the art, applicant disagrees. Then about a month later applicant confers that the Office Action is incorrect, and applicant proposes to more clearly specify the identification number as a type of display, which Examiner agrees will read over the previous prior art. So the examiners agreeing to this. This is in 2007.

Since then, there’s been lots of training in the USPTO. We have lots of Federal Circuit cases on this. So what did the Federal Circuit do here?

So it was held that the claim 14 is invalid under 112, due to lack of written description. The Federal Circuit clarified that you don’t need to explicitly state that it’s a type of– it– that it’s– it’s related to a type but you need to reasonably convey it in the specification.

And it wasn’t reasonably conveyed in the specification. This is a written description issue; this is not an enablement issue. Enablement is explaining to the public how to use the invention.

Written description is showing that the inventor was in possession of the invention. They’re two different things, and this is focused on that. And the other point is that there was some amount of expert testimony, and the Federal Circuit reminded us that the specification and drawings can show whether the inventor was in possession of the invention. Maybe different in enablement, or maybe you can provide 132 declaration for other reasons.

But when we’re talking about written description, we’re supposed to look at what was the original disclosure. And for me, this highlights the importance of having inventors, when you meet with them, they don’t necessarily have an education on all of this law, and what is required of written description in the specification. We need to be sure that with enablement and written description, we do a good job of asking the right questions. And where we feel that they’ve given us an invention at a very high level, we’ve asked enough questions about exactly how it works to show the public, or to show the Patent Office that the inventor was in possession of the invention.

As I mentioned there was some discussion with the patent office, and even the patent office initially didn’t check about this issue of the type. This was actually agreed between in an interview.

So, when you’re prosecuting patent application, maybe sometimes you get a proposal for an examiner’s amendment. Be very careful about making sure that when they propose something, you also check your specification. When your applicant proposes something, you check your specification. If it’s an inventor or a member of the patent department, just make sure that everybody is checking and on the same page about this.

And again, I think there’s been a lot more training in the patent office. There’s been a lot more case law in this, we will be more focused on this now, and we are. But just keep this in mind for your drafting and prosecution. So yeah.

[KANDA ISHIHARA]
So, maybe I can add a little flavor to this case. I tried the earlier version of this patent to jury in the Eastern District of Texas. So, I’ve actually deposed these inventors.

But I think something that’s interesting here is you’re talking about having access to the inventors. This is a later version.

Nobody was talking to those inventors anymore. Those inventors never saw prosecution counsel after the patent application was filed, and now it was all through the attorneys from then on.

So you’re talking about the crafting strategy, these patents were all crafted 100% with somebody’s target product in front of them and then the prior art that we’d already put in front of them, so they were trying to get around. So here I think it’s interesting because no one was trying to do the original prosecution well, they weren’t thinking about the original inventor, they were deep into the continuation process.

So–

[MIKU MEHTA]
How product reads and things like that.

[KANDA ISHIHARA]
Yes. So I’d love to hear that description on, you know, how to skirt that line on 112 going forward, because this is a big deal for continuation practice.

[MIKU MEHTA]
Yeah.

[KANDA ISHIHARA]
And by Way, I was on the defense side, not the Mondis side so,

[MIKU MEHTA]
Well, one thing that’s interesting is that you know, there used to be a more common practice of every time the claims are amended, getting a new inventor declaration. I don’t know if anybody’s familiar with that practice. I hadn’t done it, but I’ve seen some firms and some companies say, ‘We’re gonna,’ because they’re, the inventor is declaring that the, you know, that they invented the claimed invention, right?

Or the claims, that they’re inventor of the claims, that’s all you can be the inventor of. So, you know, if these are really important patents, and you wanna be sure, keep the inventor engaged, and if that’s your practice, then have them submitting to a declaration I’m not commenting on the merits of that, but that’s a possibility, so. Yeah.

[ANNA PIETRETTI]
The administrator–

[MIKU MEHTA]
Yes.

[AUDIENCE MEMBER ONE]
This documents discontinuation practice, especially in the context that you were was saying of litigation targeting or even pre-litigation targeting

(unintelligible)

targets? These inventors may be long gone from the company. It could be many years later. And–

[ANNA HYATT]
There’s your mic.

[MIKU MEHTA]
Oh, I’m sorry.

[AUDIENCE MEMBER ONE]
Yeah, sorry about that. Thank you. So to this point, did I turn it off?

[ANNA PIETRETTI]
It’s on.

[MIKU MEHTA]
It’s on.

[AUDIENCE MEMBER ONE]
To this point, when you prosecute these cases long after, the reality is that the inventors are often long gone, and as you were saying there used to be this practice of getting a declaration. That actually would be a good policy, I think, simply because you’re adding this specific thing to the independent claim, and I think in honesty, many of us don’t go back and figure out, like, ‘Do we actually have all the right inventors named in this application?’ You know, if we change the fundamental nature of what we’re trying to say we invented based on this application in a way that means that you should go back to the inventors and figure out who actually should be named.

[MIKU MEHTA]
Yeah, do an inventorship check. Yeah. That’s a good point. Probably move on to the last one here. So, yeah. So, yeah, kinda we’re onto–

[KANDA ISHIHARA]
Sure.

[MIKU MEHTA]
The 112 FAB.

[KANDA ISHIHARA]
Yeah, so we have about five minutes left. I’m gonna talk about Fintiv. It’s probably good we only have five minutes. I think, feel like it’s the least complicated one.

A little bit of background, so Fintiv is a small tech company that owns patents related to mobile wallet and e-payments. You might recognize the name from a case that they filed a few years ago against Apple, where the Apple’s challenge in the PTAB led us with the Fintiv rules on when to deny patent review if there’s a, a pending parallel district court case. Obviously, we’re gonna talk about the 112 issues that came out of Fintiv’s lawsuit against PayPal.

So, Fintiv sued PayPal based on four patents.

[KEERTI REDDY]
During claim construction, the District Court held that a term called Payment Handler in the claims were invalid due to indefiniteness. After they completed the two-step 112

(f)

analysis, where the first step is whether or not this is a means plus function, and the second step is identifying the corresponding structure in the specification. So the District Court determined that Payment Handler terms are means plus function, so 112

(f)

does apply. And if you replace the term Means with Payment Handler then that invokes 112

(f)

. And the language from the claims, there’s like two different four cases, but two different limitations.

It’s a Payment Handler that exposes a common API for interacting with different payment processors and a Payment Handler service operable to use APIs of different payment processors, including one or more APIs of banks’ credit and debit card processors for bill payment processors.” That’s the claim language. And if you look at the spec, it’s exactly identical in the spec. And so the District Court for for the second step concluded that the spec doesn’t disclose any adequate corresponding structure, and in this case the algorithm, like the how of that limitation.

So they concluded that this is indefinite on 112

(f)

. And on appeal Fintiv argued that 112

(f)

is not invoked because there is structure. So just going through a little bit of some of the issues that came out in the last two minutes of this case is the first one is like, when is 112

(f)

is invoked. The legal framework is that if there’s a rebuttal presumption that the claims turned, do not have means then that’s not means plus function, but the and Fintiv argued that Handler and Payment Handler identify structure. The Court disagreed with that, and they failed to identify structure, and that Payment doesn’t impart structure into that term.

So terms like– And this is kind of the where the nonce word issue comes in the term Payment Handler. Payment Handler–

[KANDA ISHIHARA]
–so that, um, the Court’s analysis on that was that these are nonce words, like module or mechanism and that these are, the, this is just hardware that performs specified functions without indicating the how. So I think the real important this, importance in this is like not really focusing on what the term is, but what the term does. Um, I think that’s what Fintiv was missing in the claim language here. They also tried to argue that the connecting words, like operable to or configured to help them, but that didn’t work in that case.

Um, so I think that I’ll, I’ll just skip to some of this uh, sort of prosecution steps, the do’s and don’ts, ’cause I think we have one minute left, and I’m trying to speak s– fast from New York, so I can do that. Let’s see.

So do I think– I feel like these are somewhat obvious, but maybe I think what is important, too, is in your claims, to audit all your terms and making sure that you have enough details in there. So go through and highlight what the terms does, like I said, and not really what it, like what it is. Um, red flags might be just adding E-Rs to Words handler, right?

And then make sure, include details in your algorithm. I think, and just with, for 112 generally, it sounds like, you know, the more the merrier maybe in terms of details. And then don’t just repeat spec language and put into your claims.

Like if you look at the limitation in the claims of the case, and you look right to the spec, it’s exactly the same, and that’s all that appears. So, I think they sort of failed in that respect. I think we’re running out of time if there’s any comments or questions, but I feel like this was somewhat of the most least complicated one, in terms of just making sure that you have, you know, specific claim language and details in your spec to support it.

[MIKU MEHTA]
But it’s one of the most common rejections we’re getting now, so you hit it right on the head, and everybody should be really aware of that. So, we can go afterwards, right, Wayne, if there’s, if there’s questios. Yeah, yep. So thank you very much, and thank you, panelists, for a great work.

(audience applauding)