[00:00:00] Speaker 00:
Okay, the final case for argument is 23-2074, satuvia bioprocess versus JSR.

[00:00:19] Speaker 00:
Good morning.

[00:00:29] Speaker 02:
Candidly, I have no idea whether it's morning or afternoon.

[00:00:31] Speaker 00:
It's afternoon.

[00:00:33] Speaker 02:
Well, I'm mindful that I'm the last argument on a Friday, but hopefully it'll make it.

[00:00:36] Speaker 03:
It'll be under two hours at all.

[00:00:39] Speaker 02:
May it please the court.

[00:00:40] Speaker 02:
I'd like to start with the issues on the independent claims and our argument on lead compound.

[00:00:49] Speaker 02:
The board here erred by skipping what is the required first step of any obviousness analysis, which is to ask what the person of ordinary skill in the art would have done.

[00:01:01] Speaker 02:
What did they have reason to do?

[00:01:03] Speaker 02:
How would they have confronted the problem before them?

[00:01:06] Speaker 02:
And that principle, which is not limited by any means to the chemical arts, is articulated in the chemical arts in this court's lead compounds.

[00:01:17] Speaker 00:
Even the cases that seem most strident, I don't mean that in the pejorative sense, about lead compounds.

[00:01:25] Speaker 00:
It's not mandatory.

[00:01:27] Speaker 00:
And in this case, I mean, it applies if it makes sense.

[00:01:32] Speaker 00:
And it makes sense if there are thousands of things out there, whatever.

[00:01:35] Speaker 00:
Here we have a finite class.

[00:01:39] Speaker 00:
And I don't know, it just seems to me that in these circumstances, they would have selected those asserted compounds because the prior art

[00:01:47] Speaker 00:
already suggested it.

[00:01:51] Speaker 00:
Doesn't this sound, I mean, we can apply lead compound and say these are the lead compounds, or we can say lead compound doesn't apply here.

[00:02:02] Speaker 02:
So I think my response to that would be to observe that and

[00:02:06] Speaker 02:
With all respect, Your Honor is assuming the conclusion when you say that there's a finite number, because the board... It's just a scenario.

[00:02:15] Speaker 03:
Everybody agrees that in the field, SBA was a widely used ligand for the IGG.

[00:02:28] Speaker 03:
So that's a reason enough to be examining the possibilities within the SPA world, of which there were five wild type possibilities.

[00:02:39] Speaker 02:
Five wild type possibilities?

[00:02:41] Speaker 03:
Right, and then you can start, and only one I think anybody's been discussing about.

[00:02:45] Speaker 02:
And that's the issue.

[00:02:48] Speaker 02:
I think, Your Honor, this would be a different case if the board had found, or if anyone had submitted evidence, that there was reason and advantage to use a wild-type ligand.

[00:02:58] Speaker 02:
That's not what people in the field at the time were doing.

[00:03:01] Speaker 02:
Every single piece of prior art that was asserted in the grounds here started somewhere else.

[00:03:06] Speaker 02:
The notion that there's only five possibilities is wrong.

[00:03:11] Speaker 02:
because the board itself found that Linholt, for example, talks.

[00:03:16] Speaker 03:
Linholt's the one that says there's something weird about all five of the domains within SPA.

[00:03:26] Speaker 02:
So Abramsen refers to all five and says you can make a G29A modification to all five in passing.

[00:03:33] Speaker 03:
Why isn't that enough of a starting point to say, well, we're going to

[00:03:36] Speaker 03:
We're going to test, among others, C. And we might not do B, because we know that at least one bad thing happens.

[00:03:43] Speaker 02:
Because, Your Honor, Abromson isn't solving the same problem.

[00:03:46] Speaker 02:
Abromson is not addressing the problem of a chromatography ligand.

[00:03:49] Speaker 02:
The proteins there were being used in fusion proteins that were attached to other proteins that were being purified.

[00:03:56] Speaker 02:
It was a very different use case.

[00:03:57] Speaker 02:
What was missing from this record, what was missing from the arguments in the petition, and what was missing from the board's decision, was a finding that a person of ordinary skill in the art, designing a new chromatography matrix, or a method of using such a chromatography matrix, which is what these claims are directed to, would have gone back, instead of looking where the art that was asserted actually pointed, which was these different, non-natural, improved versions of Domain B,

[00:04:26] Speaker 02:
would have reset the clock and said, oh no, we're instead going to start with one of the five naturally occurring domains.

[00:04:32] Speaker 00:
Okay, so why, even if you're right, why can't you use the four additional, the natural CPA domains, including domain C, and treat them all as three commas?

[00:04:44] Speaker 02:
Because there needed to have been, there was no such finding here, and there was no such evidence here that that was what the poser would have done.

[00:04:52] Speaker 03:
I thought there was evidence, that is, that the fact finder could infer from the various statements, including in Abrahamus, and about the five compounds, that each would be just

[00:05:10] Speaker 03:
I don't mean to use a contentious word, but each would be obvious to try.

[00:05:14] Speaker 02:
The lead compound analysis requires asking the question of when the POSA is addressing a particular problem here, how to make a new and improved chromatography ligand, what are they going to start with?

[00:05:31] Speaker 02:
There was never an argument, there was never evidence, there was never a finding that

[00:05:35] Speaker 02:
those five naturally occurring domains would be a chosen lead.

[00:05:39] Speaker 03:
But what sense would it make in a situation of which maybe this is one, but I'm trying to make it kind of a hypothetical.

[00:05:46] Speaker 03:
If you really had five possibilities, you don't really know what the distinction is to say, you need an affirmative reason to say that possibility two is better than one through five, when you just have no idea.

[00:06:00] Speaker 03:
And the experiment is really cheap and easy.

[00:06:04] Speaker 02:
Taking that hypothetical that when you have only five possibilities, I agree with your honor, but that's not the reality here.

[00:06:12] Speaker 02:
That might be, like for example in KSR, you had the type of simple situation where the question was, you know, where can you put a sensor on a brake pedal and there's only so many possibilities.

[00:06:24] Speaker 02:
This isn't that.

[00:06:24] Speaker 03:
I'm trying to separate the doctrinal demand that I think you're making from the facts here.

[00:06:29] Speaker 03:
I take it you say there has to be a reason to pick

[00:06:34] Speaker 03:
just one above everything else in the world.

[00:06:37] Speaker 03:
And that seems, I don't know, extreme.

[00:06:42] Speaker 02:
And I think that, with respect, is not the argument I'm making.

[00:06:46] Speaker 02:
The argument I am making is that there needs to be, and this is a flexible inquiry, some reason to start where the pose of data.

[00:06:55] Speaker 02:
It might not be the one and only possible starting point,

[00:06:58] Speaker 02:
But there has to be some articulable reason that the fact that the board needs to identify why the person of ordinary skill in the art would, contrary to the references asserted.

[00:07:10] Speaker 03:
I don't remember at this point.

[00:07:12] Speaker 03:
Did the petition at least include, for example, the binding property shown in that trio of bar charts in Janssen?

[00:07:22] Speaker 03:
C seems maybe even a little bit better on some things than

[00:07:28] Speaker 03:
A, B, D, and E?

[00:07:29] Speaker 02:
So the thing that C is a little bit better on in those bar charts, and this is not a ground that the board relied on, was on G29, excuse me, was on fab binding.

[00:07:41] Speaker 02:
But the board found, correctly in our view, that there would not have been an expectation that fab binding would be preserved upon making a G29A modification.

[00:07:52] Speaker 02:
So to the extent there's a reason to use C,

[00:07:55] Speaker 02:
It wouldn't be a reason to use C when making a G29A change.

[00:08:00] Speaker 00:
But before your time runs out, because we can't let this go to the third time today, let me turn to the dependent claims.

[00:08:07] Speaker 00:
And this is a question for both you and the other side, which is you both kind of agree arguably, at least maybe tentatively, that the dependent claim should be treated the same, that there's no distinction between composition and process for purposes of whether or not you apply reasonable expectation, require reasonable expectation of success.

[00:08:28] Speaker 00:
You obviously both come up with absolutely different answers to that question.

[00:08:33] Speaker 00:
But is that fair?

[00:08:36] Speaker 00:
I don't think that's fair.

[00:08:38] Speaker 00:
I apologize.

[00:08:38] Speaker 00:
I thought one of your arguments was that there is no basis for treating this, the one in the first case, the one in the main appeal, differently from the other one in terms of having to apply the executive order.

[00:08:52] Speaker 02:
Well, the reason is not some technical legal reason.

[00:08:55] Speaker 02:
The reason is that the claim language is different.

[00:08:58] Speaker 02:
the composition claims require that they be capable of binding to the fab part of an antibody.

[00:09:03] Speaker 02:
The meta claims require that they actually do so.

[00:09:07] Speaker 02:
And so in order to prevail on the cross appeal, they would have needed to show that, to take their first cross appeal argument, they would need to show that when

[00:09:17] Speaker 02:
the person of skill does what the board found would have been obvious, which is use these to purify whole antibodies where there would not have been any expectation of fab binding, there would have been an expectation of FC binding.

[00:09:32] Speaker 02:
That every time that happens, inherently fab binding occurs.

[00:09:36] Speaker 02:
In other words, they would have needed to prove not just that capability of fab binding sometimes is an inherent property, but that when that method that would have been obvious is performed, there is in fact fab binding to the fab part of a whole antibody.

[00:09:53] Speaker 02:
That's a very different inquiry from does this composition possess the property of being capable of binding at least sometimes to the fab part of an antibody.

[00:10:03] Speaker 02:
But I'd like to direct the court's attention to this.

[00:10:08] Speaker 03:
This is really important, and I'm sorry to.

[00:10:10] Speaker 03:
So a little group of questions.

[00:10:17] Speaker 03:
What do you think capability means in the 765?

[00:10:21] Speaker 02:
So capability means that there is, at least as to some fabs, and fabs are not all identical in sequence.

[00:10:29] Speaker 02:
This wasn't developed here just because of the way the record came together and the arguments JSR made.

[00:10:35] Speaker 02:
But at least in some circumstances, there is some degree of binding to the fab part of an antibody.

[00:10:41] Speaker 03:
And that's the inherent property that would be relevant for the 765 inquiry, not that it always binds to

[00:10:52] Speaker 02:
Correct, but that capability is there.

[00:10:55] Speaker 03:
Now, on the two method claims, it doesn't say capability.

[00:11:00] Speaker 03:
It says both go through these steps and

[00:11:06] Speaker 03:
the ligand will bind to a fab part of an antibody.

[00:11:13] Speaker 03:
Put aside the ligand.

[00:11:15] Speaker 03:
We're going to get to the fragment.

[00:11:18] Speaker 03:
But then you started talking about inherency on that

[00:11:26] Speaker 03:
And that's because there was no establishment of an expectation that that would happen.

[00:11:39] Speaker 03:
The only way that challengers could prove

[00:11:45] Speaker 03:
that element was met is through an inherency analysis.

[00:11:53] Speaker 03:
That would then allow them to say about that what they said about the compound claim, which is that it's carried along with whatever

[00:12:03] Speaker 03:
with all of the actual physical steps.

[00:12:05] Speaker 02:
But JSR didn't argue, and the experts here didn't therefore address, the question of when you purify whole antibodies in one of these matrices.

[00:12:17] Speaker 02:
whether, in fact, inherently there's binding to the fab part of an antibody when there's also an FC part present it could bind to.

[00:12:24] Speaker 02:
And for that matter, how often it happens.

[00:12:26] Speaker 02:
I want to direct the court's attention to Appendix 2210.

[00:12:30] Speaker 02:
This is an excerpt from the argument that my friend on the other side made before the board.

[00:12:36] Speaker 02:
And this was criticizing the data about fab binding in domains C and R. What's the volume of the appendix?

[00:12:43] Speaker 02:
I think three.

[00:12:44] Speaker 02:
Thank you.

[00:12:45] Speaker 02:
Thank you.

[00:12:45] Speaker 02:
And he said, and again, this is talking about the disclosure and the patent, he said, you know, we're not even talking about fab fragments, we're talking about, yes, it can bind to some fab portion of some antibody somewhere in some circumstance.

[00:13:01] Speaker 02:
That's how my friend characterized.

[00:13:04] Speaker 02:
the meaning of this capability.

[00:13:06] Speaker 02:
That is a far cry from the required showing that every time you purify a whole antibody inherently the natural result of doing so or something that happens with certainty every time you do so is binding to that part of an antibody.

[00:13:21] Speaker 02:
There's no development in this record at all of whether

[00:13:25] Speaker 02:
inherently that occurs in a process.

[00:13:29] Speaker 02:
And that's a critical distinction between these planes.

[00:13:31] Speaker 03:
So here's what I guess I'm just, some of this may be a language issue or it may be.

[00:13:37] Speaker 03:
So it sounds to me like the thing that is inherent, the property that is inherent is different in your view between

[00:13:49] Speaker 03:
the analysis of the 765 and the analysis of the method claims, because all you need for, I guess, they need, I guess, for inherency on the 765 is that there is a capability that is inherent.

[00:14:05] Speaker 03:
You say for the method claims, they have to show that there would be

[00:14:14] Speaker 02:
what in fact when the method steps are carried out it binds to the fat part of an antibody because that's what the claim says.

[00:14:25] Speaker 02:
The claim requires that

[00:14:29] Speaker 02:
You know, the process of claim one wherein the ligand binds to the fab part of an antibody.

[00:14:34] Speaker 02:
And that binding, that present tense verb binds.

[00:14:40] Speaker 03:
But say if you ran this process 100 times and it didn't bind 80 times, it did bind 20 times, you would have 20 infringements.

[00:14:49] Speaker 02:
Well, so everyone agrees that it would occur if you purified FRAB fragments that are of interest to purify because there's no FC portion to gone on to.

[00:14:58] Speaker 02:
There's nothing in this record to suggest whether it occurs one time or 20 times or 99 times.

[00:15:05] Speaker 02:
JSR simply didn't argue, if you go back to its petition, which is in the JA, I should have the citation here.

[00:15:17] Speaker 02:
I apologize.

[00:15:18] Speaker 02:
There's six petitions, but the first of them is at Appendix 587-589.

[00:15:23] Speaker 02:
Is this the compound one or one of the method ones?

[00:15:27] Speaker 02:
This is the first of the method ones.

[00:15:28] Speaker 02:
It's the first petition on the 142 in the JA.

[00:15:31] Speaker 02:
And basically the same text is in all four of the method claims.

[00:15:35] Speaker 02:
But they have a single paragraph argument on inherently that talks about this being the property of the ligands, long lines your honor is suggesting.

[00:15:46] Speaker 02:
But what they don't argue is that when you purify a whole antibody, never mind expectations, never mind anything else, inherently it occurs.

[00:15:56] Speaker 02:
There's just no argument or evidence developing the fact that

[00:16:01] Speaker 02:
When you do the thing that would have been obvious, that actually happens.

[00:16:04] Speaker 03:
Why isn't it right to say the capability that's relevant in 765 is that sometimes it will occur, inherently it will sometimes occur?

[00:16:17] Speaker 03:
Same thing is true for the method claims.

[00:16:21] Speaker 03:
And if we're trying to show that

[00:16:24] Speaker 03:
The prior art in combination makes it obvious.

[00:16:28] Speaker 03:
All we need is that this claim will cover the 20 out of 100 instances in which it does occur.

[00:16:37] Speaker 03:
And so since the claim is covering that, there's no difference between the compound and method claims.

[00:16:46] Speaker 02:
And perhaps it does bind.

[00:16:48] Speaker 02:
And I'm sure we'll have a fulsome argument about that.

[00:16:51] Speaker 02:
on infringement if this goes back to the district court.

[00:16:55] Speaker 02:
But in this record, there's been no analysis of whether when you have an antibody that has both an FC region and a FAB region, you pass it through one of these columns where the same ligands are capable of binding to the FC portion of the antibody or to the FAB portion of the antibody.

[00:17:16] Speaker 02:
There's no development of whether, always, sometimes, or never, there's binding to the fab portion of the antibody.

[00:17:23] Speaker 02:
And there's certainly nothing to suggest that the usual standard for inherency is met, which is that it's a natural result or an inevitable outcome of using one of these ligands, matrices, to purify a whole antibody.

[00:17:39] Speaker 02:
It's an argument perhaps they could have made.

[00:17:41] Speaker 02:
Perhaps the experts could have induced opinions

[00:17:46] Speaker 02:
data on whether when you do a whole antibody purification, ignorant of the fact that there might be fab binding going on, the expectation is that fab binding is not going to occur here.

[00:18:02] Speaker 02:
The experts could have opined about whether and to what degree and in what circumstances there is fab binding to that portion of the antibody.

[00:18:14] Speaker 02:
But that's just absent from this record, and I think appeal is too late.

[00:18:19] Speaker 02:
to bring that up as an entirety argument.

[00:18:21] Speaker 02:
And so the district, excuse me, the boards ruling on the method claims is therefore supported by substantial evidence.

[00:18:28] Speaker 02:
You know, the POZO would not have had reason to use these things to purify fab fragments where there would be reason and expectation.

[00:18:34] Speaker 00:
Well, we've spent a lot of time, and that's Amita, on the cross appeal, which is sort of jumping the gun.

[00:18:41] Speaker 00:
But let's hear from the other side and where we serve some time

[00:18:44] Speaker 02:
Sure, sure.

[00:18:46] Speaker 00:
Let's just try to address questions on any other issues.

[00:18:49] Speaker 00:
Okay.

[00:18:49] Speaker 00:
Thank you.

[00:18:57] Speaker 01:
May it please the court?

[00:18:59] Speaker 01:
Eric Diffin for JSR.

[00:19:01] Speaker 01:
I want to start where we just left off with the inherency issue and explain what might be a little bit of a misunderstanding here.

[00:19:06] Speaker 01:
Our inherency argument is based on Sativa's own representations and its patents which he took at face value.

[00:19:13] Speaker 01:
about the inherent ability of these claimed ligands to bind to the fab portion of antibodies, as well as the fab fragments.

[00:19:22] Speaker 01:
Sativa never contested that its patents made that representation, which is permissible under this court's case law, like in Raykow and Haasphere and others.

[00:19:31] Speaker 01:
And not only did Sativa not contest this inherency, he actually embraced it in arguing unexpected properties, both for the composition claims, as well as the method claims.

[00:19:41] Speaker 01:
on appeal.

[00:19:41] Speaker 01:
And I'm happy to give those citations.

[00:19:43] Speaker 01:
On the method claims, it's Appendix 1499.

[00:19:48] Speaker 01:
And for the composition claims, it's Appendix 28, the board noting their argument that this is an inherent property.

[00:19:56] Speaker 03:
And so this is that sometimes FAB binding occurs.

[00:20:02] Speaker 03:
Yes.

[00:20:02] Speaker 03:
I have FAB binding here.

[00:20:06] Speaker 03:
right now not distinguishing between the situation where the fab is still in situ and when the fab is separated.

[00:20:15] Speaker 01:
Yes.

[00:20:15] Speaker 01:
And part of the issue is, as was noted, the claims are agnostic in terms of what antibody we're talking about, what potential fab fragment is being discussed.

[00:20:23] Speaker 01:
And so obviously, if you have a part of a claimant embodiment is invalid for obviousness, for inherency, the whole claim goes down.

[00:20:31] Speaker 01:
It's a separate inquiry whether every single time it's done is there an infringement or not.

[00:20:35] Speaker 01:
But we have the representations in Sativa's own patents that this binding will necessarily occur for certain antibodies.

[00:20:42] Speaker 01:
They said that expressly.

[00:20:43] Speaker 01:
We relied upon that.

[00:20:45] Speaker 01:
In our petitions, Sativa never came back and said, no, that's wrong.

[00:20:48] Speaker 01:
Let me show you some instances where this binding doesn't occur.

[00:20:51] Speaker 01:
They never did that.

[00:20:53] Speaker 01:
And again, they embraced it even more so to say, our claim compositions, our claim methods have unexpected results.

[00:21:01] Speaker 01:
They actually have this binding.

[00:21:04] Speaker 01:
And also wanted to briefly point your honors to Appendix 6288.

[00:21:08] Speaker 03:
Can I just be clear about one thing?

[00:21:11] Speaker 03:
So when he say the patent says capability of binding to FAB, this is about this domain C thing in general, that the only reference to making the

[00:21:28] Speaker 03:
the G29A switch is in that one sentence, and there's nothing more specific about the G29A modified C and its fab binding or what?

[00:21:43] Speaker 01:
Yes, there are representations first in the abstract that represents that the ligands of the present invention are capable of binding fag fragments of antibodies.

[00:21:53] Speaker 01:
Also, column five, lines 32 to 37,

[00:21:56] Speaker 01:
There's a representation of the present invention.

[00:21:58] Speaker 01:
Ligands are capable of binding to the fab part.

[00:22:00] Speaker 03:
Sorry, what was the second one?

[00:22:02] Speaker 01:
I'm sorry, column five, lines 32 to 37.

[00:22:04] Speaker 01:
I'm sorry, I'm trying to be mindful of the time.

[00:22:08] Speaker 01:
And it talks about the present invention.

[00:22:10] Speaker 01:
The ligands are capable of binding to the fab part of antibodies.

[00:22:15] Speaker 01:
So that's OK.

[00:22:17] Speaker 03:
Just so that I guess I'm clear, the present invention

[00:22:23] Speaker 03:
except for the reference to that one embodiment on, I guess, columns 6, lines 49 to 53, doesn't involve G29A switch.

[00:22:34] Speaker 01:
Well, that's the point, is that they're representing what's covered by their invention.

[00:22:38] Speaker 01:
OK.

[00:22:38] Speaker 03:
I just wanted to be clear.

[00:22:40] Speaker 03:
I wasn't missing something more specific.

[00:22:42] Speaker 01:
No, absolutely.

[00:22:43] Speaker 01:
Those three passages, and there's also column 2, line 59 to 63, which, of course, is the foundation that Poe has understood.

[00:22:52] Speaker 01:
that the natural domains also had this ability to bind to fab.

[00:22:56] Speaker 01:
Which is the only thing they tested, right?

[00:22:58] Speaker 01:
What's that?

[00:22:59] Speaker 03:
It's the only thing they tested.

[00:23:00] Speaker 03:
The two figures don't involve the G29A.

[00:23:06] Speaker 03:
That is correct.

[00:23:08] Speaker 01:
So we've taken that face value, what they said in their patents about their purported inventions having this fab binding characteristic, and we've used that

[00:23:17] Speaker 01:
to say that, OK, we believe that's proper proof of inherency under this court's case, like Inrei Kao and other decisions.

[00:23:24] Speaker 01:
And it was incumbent upon Sativa, if they felt that this wasn't an inherent property, that they would dispute that in some form.

[00:23:31] Speaker 01:
They, of course, have the data.

[00:23:32] Speaker 01:
They have all the work that they've done here.

[00:23:35] Speaker 01:
We didn't see any of that.

[00:23:36] Speaker 01:
And again, all we saw was them double down on it by saying it's an unexpected inherent property of whether it's the composition or the method, which I've pointed your honors to.

[00:23:46] Speaker 01:
in the appendix.

[00:23:46] Speaker 01:
But what I do want to think is also important, this notion of the district court applying the wrong claim construction.

[00:23:53] Speaker 01:
Obviously, the focus on the board.

[00:23:59] Speaker 01:
And I apologize, Your Honor.

[00:24:00] Speaker 01:
Thank you for correcting me there.

[00:24:01] Speaker 01:
We're still talking about the cross-appeal.

[00:24:03] Speaker 01:
We're now talking about the cross-appeal, yes.

[00:24:07] Speaker 01:
This is Appendix 104, Note 12, where the board

[00:24:12] Speaker 01:
distinguished between binding to fab fragments versus binding to antibodies as a whole.

[00:24:18] Speaker 01:
It's now not disputed between the parties that the correct construction is that you're not just limited to fab fragment binding, that you can talk about fab portions of the antibodies as a whole.

[00:24:30] Speaker 03:
I think you're certainly material in the board's opinion, suggesting it was thinking about the

[00:24:40] Speaker 03:
the isolation of a separated fab piece.

[00:24:45] Speaker 03:
I'm not quite sure whether the word fab fragment is limited to that or whether the term fab fragment under that definitional portion in the spec actually applies, regardless of whether the fab is in situ or separated.

[00:25:01] Speaker 03:
But my question is, what difference does that make?

[00:25:06] Speaker 03:
It makes it differences that make that the board was not focusing on what the claim clearly is about which is which it certainly includes Binding to the fat portion of the antibody when the fat portion is still part of the antibody in that footnote 12 I disappointed your honor to the board understood that an additional step to sort of isolate this fat fragment from the whole antibody in order to then

[00:25:30] Speaker 01:
perform a separation was required.

[00:25:33] Speaker 01:
And that's why I thought that there was an additional step he needed to show inherency in that circumstance.

[00:25:38] Speaker 01:
And they said, this is different, what you're arguing about, the capability of bonding to the fat portion of the antibody.

[00:25:45] Speaker 03:
And I guess what I'm trying to focus on is, is there a separate isolating step in either of the two method claims?

[00:25:55] Speaker 03:
They're slightly different.

[00:25:56] Speaker 03:
Cleaning is in one, and cleaning is not in the other, right?

[00:25:59] Speaker 01:
They're all directed to a conventional, and there's no dispute about that, separation techniques for a whole antibody.

[00:26:05] Speaker 01:
And that's at least covered in part.

[00:26:07] Speaker 01:
There's also covering the possibility of a fab fragment that's separate isolated.

[00:26:11] Speaker 01:
But that's the point.

[00:26:12] Speaker 01:
The board understood it to be limited to only isolating fab fragments.

[00:26:16] Speaker 03:
Is there any reason to think that just on the binding question, fab binding, it might make a difference whether the fab

[00:26:29] Speaker 01:
chain is attached to the rest of the antibody

[00:26:49] Speaker 01:
binding to the fab portion of the antibody as a whole.

[00:26:52] Speaker 01:
And then he said, in that context, yes, it'll necessarily happened.

[00:26:56] Speaker 01:
That's an inherent property.

[00:26:57] Speaker 01:
And he actually was asked another question.

[00:27:00] Speaker 01:
He said, it would happen.

[00:27:01] Speaker 01:
And clarified, he's talking about binding to the fab part of an antibody.

[00:27:05] Speaker 01:
Again, this is Appendix 6288.

[00:27:08] Speaker 01:
And that is direct evidence of at least, for certain antibodies that are covered by these claims, you're going to have an inherency.

[00:27:15] Speaker 01:
And there's no evidence from the other side.

[00:27:17] Speaker 01:
to contradict that, again, only embracing it for purposes of unexpected results.

[00:27:22] Speaker 01:
What we're talking about here is a binding mechanism.

[00:27:25] Speaker 01:
And the court's recent decision, Inrei Kuvar, said if you're talking about a mechanism of action, that's what this is really.

[00:27:31] Speaker 01:
It's just how do these ligands bind to antibodies to help separate?

[00:27:35] Speaker 01:
It's a known function that's been used in the art for decades, and we're only talking about how that mechanism actually works.

[00:27:43] Speaker 01:
cannot provide a basis for patentability.

[00:27:46] Speaker 01:
That's all that's being claimed here is this capability.

[00:27:50] Speaker 01:
I do want to briefly touch on lead compound just because I started on the cross-appeal and the inherency points.

[00:27:55] Speaker 01:
Unless your honors have more questions on that, I want to at least spend a minute or two on lead compound.

[00:28:02] Speaker 00:
And another minute or two on maybe the two cases that come up in a certain context is Haspera

[00:28:07] Speaker 00:
and Honeywell, and if you could give us your take on those.

[00:28:11] Speaker 01:
Of course, I will definitely do that.

[00:28:13] Speaker 01:
So just very briefly, in terms of the lead compound, this is not a case that involves a modified version of a prior art compound.

[00:28:21] Speaker 01:
Here we're talking about the prior art in multiple references in a single paragraph, say, take the claim mutation and apply it to any one of the five SPA domains, including domain C. Not only that,

[00:28:34] Speaker 01:
The same paragraph tells you to do that for reasons of improving alkali stability.

[00:28:39] Speaker 01:
So a poser would apply these express teachings, which we see in multiple references, as found by the board.

[00:28:45] Speaker 01:
And this is a very straightforward instance of a specific motivation to arrive at the claimed invention, not a new compound.

[00:28:51] Speaker 01:
And so I would suggest that this court's case law lead compound is talking about picking a compound out of a large number, typically for structural similarity reasons, and applying a separate optimization step.

[00:29:04] Speaker 01:
which is what's missing here.

[00:29:05] Speaker 01:
There is no separate optimization step needed here.

[00:29:08] Speaker 01:
It's simply applying straightforward teachings of what to start with, what to change, why to change it in a single paragraph.

[00:29:16] Speaker 01:
If this isn't an extraordinarily compelling case of obviousness, I'm not sure what it is.

[00:29:20] Speaker 01:
So with that, unless there's any questions, I'll shift to the inherency case law question that you raised.

[00:29:27] Speaker 01:
And really, the bottom line is this.

[00:29:29] Speaker 01:
In re cuvaris,

[00:29:31] Speaker 01:
Hasbira, Santeros, Persian, they all stand for the proposition if what you're talking about is an inherent property of the claim subject matter, and especially if it's a mechanism of action, which is what this case is about.

[00:29:44] Speaker 01:
It simply doesn't add patentable weight.

[00:29:46] Speaker 01:
When you talk about the cases that Sativa is relying upon, they're simply saying in those cases an unremarkable proposition that you can't use inherency to either support motivation to combine or to provide an expectation of success.

[00:30:01] Speaker 01:
And it makes sense.

[00:30:02] Speaker 01:
If you think about it, a posa can't be guided or motivated by the unknown.

[00:30:06] Speaker 01:
That's not what this case is about.

[00:30:07] Speaker 01:
We've separately shown, for reasons of alkalized stability, that the claim structure is expressly disclosed all over the prior art.

[00:30:15] Speaker 01:
It was the worst kept secret in the industry.

[00:30:18] Speaker 01:
And to then say that some separate inherent property that's going to be there necessarily, which again is based on Sativa's own patent representations,

[00:30:27] Speaker 01:
as well as their arguments in these proceedings, embracing it for purposes of unexpected results, they can on the one hand say, well, there's this inherent property we want to rely upon to boost up our claims, but if it's in a mechanism of action, and that's all it is, about how this binding works to isolate antibodies, that doesn't get patentable weight under this court's case law.

[00:30:49] Speaker 01:
Unless there are other questions, I wanted to just say 30 seconds on one other issue, and that is,

[00:30:56] Speaker 01:
There was a separate argument, independent on the cross appeal of inherency, about a combination ligand that prior art such as Hover said you could have additional monomer units.

[00:31:06] Speaker 01:
If you wanted fab binding, you could have an additional monomer added to provide that property.

[00:31:13] Speaker 01:
We made this argument, sorry, from our petitions, including the New York.

[00:31:17] Speaker 01:
You didn't reference Hover in your petition, right?

[00:31:20] Speaker 01:
No, we didn't.

[00:31:20] Speaker 01:
Because what we did is we had the argument in every single ground, including the Hover ground.

[00:31:25] Speaker 01:
And we reference back our argument about the additional monomer unit.

[00:31:29] Speaker 01:
But here's the point.

[00:31:30] Speaker 01:
Whether it was related to Hober or any other reference, we made this argument.

[00:31:35] Speaker 01:
We had nothing from the board on it, not even an acknowledgment whatsoever.

[00:31:39] Speaker 01:
So there's that.

[00:31:40] Speaker 01:
But there's also the problem, again, of the board with the improper claim construction.

[00:31:44] Speaker 01:
We submit this case should end on inherency.

[00:31:46] Speaker 01:
But to the extent there's anything left to deal with, the board would have to sort out what to do about this combination ligand theory,

[00:31:53] Speaker 01:
and what to do about the substantial evidence Janssen issue, given that they're applying the wrong construction.

[00:31:59] Speaker 01:
They don't even have the right construction in mind in terms of what these claims cover.

[00:32:04] Speaker 01:
So we submit the case should end on inherency.

[00:32:07] Speaker 01:
But there's certainly multiple issues that would need to be addressed on remand to the extent the court did not believe this case is an inherency case for all claims.

[00:32:24] Speaker 02:
I'd like to begin by addressing this mechanism of action point and its relationship with reasonable expectation of success.

[00:32:44] Speaker 02:
So binding to fab is not the mechanism of action by which prior art ligands worked.

[00:32:50] Speaker 02:
So I think we need to sweep that to one side.

[00:32:53] Speaker 02:
The expectation that the board found is that these ligands, like the prior art commercial product using domain Z, would bind to the FC region of whole antibodies.

[00:33:03] Speaker 02:
The recognition that even with a G29A mutation, which in domain B caused a loss of Fab activity, in domain C it would preserve Fab activity while also preserving the stability benefits of the G29A,

[00:33:20] Speaker 02:
modification.

[00:33:22] Speaker 02:
That is an unexpected beneficial property that is different from the way these would be expected to act and the way these had been, you know, the domain Z had been used in the prior art.

[00:33:32] Speaker 02:
So I think it's just wrong to try to pigeonhole the reasonable expectation case here into cases like Pershing and cases, you know, like Mars where you're talking about just a mechanism of action or some recognition about something that's always happening anyway in your drug.

[00:33:50] Speaker 02:
This is a new novel, again nobody had been using before, that has an advantage over the other G29A modifications that as a class were found to be obvious under a KSR style analysis.

[00:34:06] Speaker 02:
And it's error to just short circuit reasonable expectation of success in those circumstances and say that the limitation requiring capability of fab binding is of no patentable weight.

[00:34:18] Speaker 02:
It's also error to not even address unexpected beneficial properties as a secondary indicia on the grounds that that's inherent anyway.

[00:34:28] Speaker 02:
Most unexpected properties of a compound are inherent to that compound.

[00:34:32] Speaker 02:
The question is whether there is some novel discovery that warrants patent protection, which was

[00:34:39] Speaker 02:
The case here is not the case when you just have the pharmacokinetics that explain why a drug always works.

[00:34:45] Speaker 02:
So I think it's very important to separate those out.

[00:34:49] Speaker 03:
So you don't think that it is correct doctrinally to say if you have wholly independent of the property that the result turns out to have a sufficient motivation to create it

[00:35:09] Speaker 03:
and a reasonable expectation of success in creating it, and maybe even also success in meeting the goals, which here would be both stability after cleaning and binding.

[00:35:27] Speaker 03:
Forget about fab at the moment.

[00:35:32] Speaker 03:
At that point, any unexpected property is just immaterial.

[00:35:38] Speaker 02:
Not where the unexpected property lets you do something you couldn't do before, which is here purify fab fragments that don't have the FC region that would have been known to bind.

[00:35:46] Speaker 02:
And where the requirement of a capability of fab binding is not a property of this finite universe of known solutions if the board

[00:35:56] Speaker 02:
finite universe of predictable solutions to accept the board's holding on the independent claims.

[00:36:03] Speaker 02:
But he's a property of one particular ligand that we selected from that putatively obvious genus.

[00:36:12] Speaker 02:
I do think it's very different from a situation where a patentee in an attempt to gain some patentability on what is otherwise an obvious compound, single compound,

[00:36:24] Speaker 02:
starts reciting properties of that compound in the claim and then argues that they wouldn't have been known.

[00:36:30] Speaker 02:
This is not only unknown, but contrary to the expectations of the POSA and independently useful.

[00:36:38] Speaker 02:
And independently useful to purify fragments.

[00:36:40] Speaker 02:
That's why we have such trouble with the inherency argument on the methods.

[00:36:47] Speaker 03:
And when you say fragment, you mean separated.

[00:36:51] Speaker 02:
separated frag frag.

[00:36:53] Speaker 03:
Even though the claim doesn't require it.

[00:36:55] Speaker 02:
We agree that the claim language encompasses both fragments and whole antibodies, the sad portion of whole antibodies.

[00:37:02] Speaker 02:
But when you purify whole antibodies, all the data in the patent, all the data in the art, are about

[00:37:09] Speaker 02:
separated fragments, and there's no real advantage anyone's pointed to to binding to the fab portion of a whole antibody in addition to the FC portion.

[00:37:19] Speaker 02:
Because you've got the whole thing.

[00:37:20] Speaker 02:
You can grab something else.

[00:37:22] Speaker 02:
And so that's why the board was correct in what the board's finding was supported by substantial evidence.

[00:37:28] Speaker 02:
that it would not have been obvious to purify fab fragments where you know there's going to be fab binding.

[00:37:35] Speaker 02:
And there was nothing else in the record for them to glom onto about whether inherently in purifying whole antibodies there's binding to a fab portion of the antibody.

[00:37:46] Speaker 00:
Thank you.

[00:37:46] Speaker 00:
And since you've raised the cross appeal, your friend has about a minute to respond.

[00:37:51] Speaker 00:
Thank you.

[00:37:54] Speaker 01:
Thank you, Your Honor.

[00:37:56] Speaker 01:
I just want to respond on the first point about this notion that the FC portion of the whole antibody makes the big difference, and that's where the binding occurs.

[00:38:04] Speaker 01:
There is zero evidence of that presented by CETIVA in these proceedings.

[00:38:08] Speaker 01:
My colleague didn't point you at any.

[00:38:10] Speaker 01:
In fact, that's contrary to what their patent says once again.

[00:38:13] Speaker 01:
Column 5, lines 32 to 37, referring to the present invention, all of what they allege to have the invention talks about the capability to bind to the fat portion of antibodies, whole antibodies.

[00:38:25] Speaker 01:
And once again, they embraced this property in arguing unexpected results, Appendix 1499, for the Fab Method claims.

[00:38:34] Speaker 01:
They're stuck with these representations.

[00:38:35] Speaker 01:
That's the strategy they chose to pursue.

[00:38:38] Speaker 01:
And it results in invalidation.

[00:38:40] Speaker 01:
Thank you, Your Honors.

[00:38:41] Speaker 00:
We thank both sides.

[00:38:42] Speaker 00:
The case is submitted.

[00:38:43] Speaker 00:
That concludes our proceeding for this morning.