Oral argument — Baxalta Inc. v. Genentech, Inc.

Our final case for today is 2019-1527, Baxalta v. Genentech. Mr. Peterson, please proceed when you're ready. May it please the Court. The District Court erred by construing antibody inconsistently with how the word is used in the claims. Mr. Peterson, I have some questions. In the blue brief at 40, you argue that the District Court characterized the PTO's office action as, quote, really confusing. That's in the last full paragraph on page 40. Isn't it true that the really confusing comment occurred early in the Markman hearing before Genentech's oral argument? Yes, Your Honor, that's correct, and I apologize, I did not mean to mislead the Court there. Certainly, the District Court, in his opinion, reflects what he believed to be a clear understanding of the office action, although we certainly agree with the District Court's earlier comment that it was really confusing. Well, you don't present it that way. I mean, how can a statement by a court about an office action at the start of a Markman hearing be evidence of the Court's understanding at the end of the hearing? Your Honor, I apologize. During the claim construction proceedings, the District Court characterized this office action as really confusing. I don't think that's fair. Your Honor, I apologize for that. In the red brief at 35, Genentech states that you assert for the first time on appeal that because the cited prior art describes IgM and IgA antibodies as having more than two The prior art contradicts the District Court's construction. Is Genentech correct that you're raising this for the first time on appeal? No, Your Honor. Okay, where did you raise that below? Your Honor, I apologize, may I respond to that during rebuttal? It is in the record. I do not have a specific citation. Okay, in the red brief at 62 to 65, Genentech states that you never explained your disclaimer of the claim scope. I'd like you to explain it now. Specifically, what did you disclaim to obtain the issuance of the patent? Your Honor, you'll see this on appendix pages 16040 to 16041, a portion of the transcript of the Markman hearing. What we believe was disclaimed during prosecution related to two amendments that were made, not with respect to the general change from derivatives to fragments, but with respect to modifications relating to original claim five, which originally read, wherein said antibody derivative comprises a complement determining region, peptide. That was amended in issued claim 21. I'm sorry, did you say 16040? Yes, Your Honor. Thank you. Got it. Originally referred to any complementary determining region peptide, CDR peptide. As issued, the claims related only to CDR3 peptides. That was an amendment and a surrender that we acknowledge was made during prosecution. The other surrender that we acknowledge related to an amendment in what was claim seven relating to derivatives comprising amino acids selected from the group with random or variable amino acid sequences. That was modified during the prosecution and in the issued claims to list only specific amino acid sequences. So those are the only two surrenders that we see clearly and unmistakably in the file history. And you see that I'm looking on appendix page 16041, page 37 of the transcript, the court asking Mr. Abernathy, it's lines 15 to 18. Are those the only two surrenders in your view, Mr. Abernathy? Yes, Your Honor. We certainly acknowledge there was an amendment from derivative to fragment generally during the prosecution of the patent, but it is ambiguous as to why that amendment was made. And certainly nothing in file history clearly and unmistakably indicates that the examiner and the patentee understood that amendment to be surrendering all bispecific antibodies, all chimeric antibodies, all humanized antibodies. And in fact, we think the suggestion of a disclaimer of those antibodies is refuted by the claims that actually were allowed. As this court noted in the Intellectual Ventures v. T-Mobile case, a requirement in a dependent claim belies disclaimer in the independent claims from which those claims depend. Here if you look at the face of the claims of the 590 patent, claim 19 refers to an antibody that is a humanized antibody. Claim 4 refers to wherein an antibody or antibody fragment is a bispecific antibody. Just from English grammar, antibody is necessarily used broadly enough in the claims to include humanized antibodies, is used broadly enough to include bispecific antibodies. Under my friend's construction, adopted by the district court, those phrases are oxymorons. An antibody, by definition, cannot be a humanized antibody. An antibody, by definition, cannot be a bispecific antibody. What do you think the Column 5 language and the specification, in your view, was seeking to do? Was it seeking to define antibody or was it seeking to define what antibodies are in the human body? There seems to be, you know, that's one of the most important issues in this case, of course. What's your view on that? Judge Stoll, in our view, it was meant to provide a general description of antibodies in teaching a highly complex art. It was not meant to provide a definition. It was meant to provide... But is it true that that's a description of how they would occur in the human body? It's a description of some naturally occurring antibodies in the human body. Are there naturally occurring antibodies that wouldn't fall within that? Yes, Your Honor. The vast majority of IgM antibodies are pentameters. They have more than two heavy chains and more than two light chains, although those heavy and light chains are identical. And with respect to lexicography and the definition, I'll note that this Court has never suggested that a to-be verb standing alone is enough to constitute lexicography. And certainly in Abbott Laboratories, this Court rejected lexicography based solely on the use of a to-be verb in specification. You don't see any other indicia here. You don't see, as used herein, you don't see the word antibodies in quotation marks. You don't see antibodies of the present invention. And in fact, if you look just two columns later, column seven carrying over to column eight, what you'll see is a statement that the antibodies of the present invention can be prepared by means of phage display gene libraries or humanizing techniques. We asked their expert about that statement at his deposition, and he admitted that it couldn't be harmonized with treating column five as a definition. He said in his words, quote, I think there is a miswriting there in the specification. That refutes any suggestion that column five can be taken as definitional. Did anybody argue, you know, I'm looking at column five here, the definition comprises multiple sentences in the specification. Did anybody argue that additional sentences should be read in? Like there's another sentence, at least one other sentence in this paragraph that seems to be adding on to the definition that purportedly is provided here. Did anybody, why the cutoff here at line 63? The parties focused on the portions of the definition that were material to the infringement contentions. Here, imicizumab, the accused product, is a humanized bispecific antibody, which would fall outside this narrow definition or narrow description because it has non-identical heavy and light chains and because its synthesis was not induced by an antigen. Sure, your view is that it could have been the whole paragraph, but there was no need to go on because this was enough to distinguish the accused product. Yes, Your Honor. I think in the district court's reasoning, it would have to be the whole paragraph. I point you to appendix page 39, note six, where there was some discussion of the synthesis requirement, which receives very little analysis in the district court's opinion. But the district court included that in the construction, saying because column five is definitional and this is part of column five, it is necessarily part of the construction. Let me take you back, though, because I think the district court went astray. You've heard my argument about the usage of the word antibody in the dependent claims. The district court treated this as a claim differentiation argument, and that's not correct. The argument is not based on the scope of the claims. The argument is not that these dependent claims imply that there is not a limitation present in the independent claim from which they depend. Our argument is just based on the usage of the word antibody in the claims. If you took claim 19, struck out the dependents from claim 1, and just made it an independent claim that recited an antibody that is a humanized antibody, our argument would be exactly the same, that you should not adopt constructions or terms that render claims nonsensical on their face. Would you say that that applies to the specification as well? The specification teaches lots of different adjectives that precede the word antibody. Would you say that in each of those cases, the word antibody should be inclusive of all those different kinds of antibodies? I don't know. An example would be a nail. A nail is a nail, whether it's a rusty nail or a non-rusty nail, a steel nail, or an aluminum nail. Here you've got a lot of different kinds of antibodies, including bispecific, monospecific, different kinds. Would they all still be antibodies? Is that your position? Or am I misunderstanding? They're not all exclusively antibodies. If you look at claim 4, I think here's probably the best thing to take a look at, because we do have or antibody fragments in there. So I realize this is a little bit counterintuitive, but to one of skill in the art, in our view, at least an antibody fragment can be a bispecific antibody. So if you can imagine the antibody as the Y with the ends of the Y that affect the antigens to which it binds, take that bispecific antibody, cut off the bottom piece, and you would still have an antibody fragment that would be bispecific and would bind to multiple antigens. But it wouldn't be an antibody? It would not be an antibody at that point. If it were cut off? Yes, if you cut off the bottom piece of it. It would, in our view, be an antibody fragment. So an antibody fragment is a portion of an antibody, right? Yes, Your Honor. So like a chimeric antibody is not an antibody fragment. It's just a different kind of antibody from the one typically found in human beings. Yes, Your Honor. That's our view. Although, to be clear, we think you could also have an antibody fragment that is a chimeric antibody if you took an antibody that was a chimeric antibody. Then it would be a chimeric antibody, it would be a chimeric antibody fragment, correct? Your Honor, as the terms are used to those of skill in the art, an antibody fragment that was derived from a chimeric antibody that is an antibody would be an antibody fragment that is a chimeric antibody. And I apologize, I realize that's contrary. Look, if you were going to hold simply that chimeric antibodies are antibodies and bispecific antibodies are antibodies, we would have very little quarrel with that. The main thing here is on the face of the claims, not a claim differentiation argument, a usage argument. The word antibody is used in a way that includes humanized antibodies, includes bispecific antibodies. It would be an extraordinary thing to affirm a decision that adopts a construction that renders these claims nonsense on their face. If the panel has no further questions, I'll reserve my remaining time for rebuttal. Very good. Thank you. Mr. Stone? May it please the Court. I would like to begin with… Let me ask you a question. Regarding the disclaimer issue, if IgM and IgA remain in the claim, then they're not disclaimed, right? It is not our contention that IgMs or IgAs were disclaimed. No, Your Honor, derivatives were disclaimed. in column 5 that an antibody consists of two identical heavy chains and two identical light chains. I think there's a little gamesmanship going on there. IgMs and IgAs are created in our B cells as monomers that have two identical heavy chains and two identical light chains. Some of them, then, for IgA associate into pairs of two, and some of them for IgM associate into groups of five of those antibodies. Whether the pentameric form of an IgM and the dimeric form of an IgA are within the scope of that paragraph, I'm confident that if they ever accused this patent against someone who made a pentameric IgM, they would say you have five infringing products, but certainly the monomeric form is within the scope of the definition in column 5, and it is not our position that those were disclaimed, derivatives were disclaimed. I would like to return to Your Honor's question about the disclaimer, if I may. What we say in the red brief is that they've never plausibly explained the disclaimer. They started by saying there was no— Let's just start with whether there is a disclaimer. Absolutely. A disclaimer has to be clear and unmistakable. Yes, Your Honor. They changed the word derivative to the word fragment. Yes, Your Honor. I will be honest. I'm having trouble completely understanding what that change impacted. Sure. But I'm really struggling with the idea that whatever it impacted mounts to a clear and unmistakable disclaimer. If I can't figure out and find it confusing, how is it clear and unmistakable? May I answer that question, Your Honor? I want to start with, I hear two questions. One, was there a disclaimer? And two, how do we know what it was? The district court began the Markman hearing by asking the following question, coincidentally in this very courtroom, at page A16034 of the record. Why would you point that out? For no reason other. I apologize, Your Honor. I intended no significance to it. I apologize. It was an entirely inappropriate thing to say. I'm terribly sorry, Your Honor. I did not intend it that way at all. We'll talk about that. I'm terribly sorry, Your Honor. I apologize. The district court asked a question, and I think there's also agreement that there was a disclaimer in the prosecution as to derivatives, but there's no agreement as to what a derivative is. Is that a fair statement? The district court said, I take it that there's agreement that there was a disclaimer as to derivatives, but there's no agreement as to what a derivative is. Is that a fair statement? And both counsel answered, yes, Your Honor. And the district court was entitled to rely. I apologize, Your Honor. I truly do. I did not intend to suggest anything by that. And I ask that I apologize. I do not believe you. But go ahead. I truly didn't, Your Honor. Then you wouldn't have said it. What is your point? The point is this, Your Honor. I'm terribly sorry. The district court asked the parties, I think there was also agreement that there was a disclaimer in the prosecution as to derivatives, but there's no agreement as to what a derivative is. Is that a fair statement? And both parties agreed, yes. And the district, yes, Your Honor, from Bexalta. Yes, Your Honor, from Genentech. The parties then, and the district court then, endeavored to figure out what is a derivative, because the parties both agreed that derivatives were disclaimed. That's the beginning of the Markman transcript. So do I understand correctly that what happened was that then the court turned to the Column 5 language and then looked at the language at the top of Column 6 and said, because Column 6 uses the word derivative and says it includes bispecific antibodies, and the definition of antibodies in Column 5 doesn't include bispecific antibodies, it must be that bispecific antibodies are in fact not antibodies, but in fact are fragments? I don't, well, are derivatives, but I don't think that's quite right, Your Honor. I think that what the district court said was first, when the patent was filed, it referred to antibodies and antibody derivatives. And in each of the instances in Columns 5, 6, and 7, with the exception of the one that I will come to at the bottom of Column 7 into 8, in every other instance it refers to antibodies and antibody derivatives as one term. And in fact in Column 6 refers to the term in the singular, factor IX antibodies and antibody derivatives. The court said, how do I figure out what's an antibody, what's an antibody derivative? The patent, the court said, tells me what an antibody is, because in Column 5 it says antibodies are immunoglobulins, et cetera. And then every instance thereafter where it refers to things that are inconsistent with that definition, chimeric antibodies, which are arguably inconsistent, by specific antibodies, which are definitely inconsistent because they don't have two identical heavy chains and two identical light chains. In every instance where those embodiments are referred to, they are preceded by the phrase antibodies or antibody derivatives, including in the claim as initially filed. And so the question is, when Bexalta surrendered antibody derivatives, which they conceded they did, they disclaimed antibody derivatives, and replaced it with fragments. The question is, what is a derivative? And so the court puts- So he said this by specific language in Column 5, for example, is one of those things that you're saying must be an antibody derivative. Well, in Column 7 where it defines a by specific antibody, Your Honor, it says by specific antibodies are macromolecular or heterobifunctional crosslinkers. The district court found that that was within the scope of derivatives as initially filed, as did the examiner. And the examiner's rejection says you don't have support for antibody derivatives, including by specific antibodies. And it's worth pointing out that the way that one makes a by specific antibody is by taking two other antibodies, cutting them in half, combining them, and deriving a by specific antibody from them. And so the record in this case is that a by specific antibody is an artificially engineered product. In fact, the district court's construction of by specific antibody, which Bexalta has not challenged on appeal, is that a by specific antibody is an artificially engineered macromolecular heterobifunctional crosslinker. There's no contesting in this courtroom that by specific antibodies are artificially engineered. That's the construction below and Bexalta hasn't challenged it on appeal. The question then for this court is what's a derivative? And what the district court did is exactly what Teva tells the district court to do. The district court looked first at the specification, found no definition there, then looked at the claims, found no clarity there, looked at the prosecution history, and then turned to expert testimony. Can I ask you something? Of course. How is that consistent, what you've said, with claim four that doesn't have derivative in it at all? Your Honor, there is no explanation. It essentially defines or includes a by specific antibody. Your Honor, I think that's the core question on the appeal. There is no set of facts or no explanation for either side that explains all of the facts. There is something missing from each side. For us, it is claim four and 19. Claims four contain embodiments that the examiner rejected and the district court concluded that Baxalta failed to conform those claims. That is the explanation that the district court concluded reviewing. The examiner didn't catch that. That's correct, Your Honor. There's nothing in the prosecution history that suggests that the examiner changed his mind. There's no moment in the notice of allowance, for example, where he says. You don't think the issuance of these claims suggests anything? I don't think they can, Your Honor, because the patent office issues claims. But you're presuming that we think the word derivative included these terms, wherein the word fragment does not. You're starting from the premise that the prosecution history amounts to a relinquishment of by specific, humanized, chimeric, and all of these other forms of antibodies. Right? Yes, Your Honor. And the district court made a factual finding that in this court is reviewed for clear error under TEVA that a derivative is any significant change to a column five antibody. With that factual finding, which Baxalta. Why is that a fact finding? Because the district court relied on expert testimony. What the TEVA case specifically said. So it's not a fact finding. I'm sorry? You're saying that expert testimony is a fact finding. No, I believe that the subsidiary fact finding based on expert testimony is a fact finding under TEVA. TEVA specifically, actually, for what it's worth, the Supreme Court gave us two examples. This happens to be one. In some cases, the district court will need to look beyond the patent's intrinsic evidence and to consult extrinsic evidence in order to understand, for example, the meaning of a term in the relevant art during the relevant time period. That's exactly what the district court did here, was ask experts from both sides, what did the term antibody derivative mean? Because the parties agreed there was a disclaimer of antibody derivative, an agreement that is binding on both parties in this court. Antibody derivatives were disclaimed. The court then wanted to figure out what's an antibody derivative. The court looked at the intrinsic evidence and the extrinsic evidence. And came to a conclusion, right? I'm sorry? And developed a conclusion. But you just said that the court looked at the intrinsic evidence and the extrinsic evidence. Yes. And then came up with a construction. I think that's kind of like obviousness. So the ultimate construction is legal, but there are underlying fact findings, including the crediting of the expert. I think that's 100% right. And I think the conclusion that— Okay. I just want to make sure we're on the same page. Absolutely right. The ultimate construction is reviewed de novo. The subsidiary facts are reviewed for clear error if but only if they rest on extrinsic evidence. This one does. What is a derivative rested on extrinsic evidence? The court looked to the experts and said, how was that term used in the art? Did it have a meaning? That's exactly what Teva then goes on in the next sentence to say is reviewed for clear error. So where we arrive in this courtroom is with an agreement that the district court— the district court elicited an agreement from the parties that there was a disclaimer of derivative. And then set out to figure out what is a derivative. And the conclusion that the court reached, which is a subsidiary fact finding to its claim construction, is that as used in the art, the term antibody derivative meant any significant change to a Column 5 antibody. Under that definition, it is undisputed that chimeric antibodies, humanized antibodies, and bispecific antibodies are antibody derivatives because one makes each of them by making a material change to a Column 5 antibody. That is completely consistent with every part of the prosecution history, but the part that Your Honors are struggling with, which is why are they still in the claim? And so the district court asked the question, why are they still in the claim? And concluded, which I believe is reviewed de novo to be clear, that Bexalta failed to conform the claims and that the examiner allowed them. And the question is, does the fact that the examiner rejected them, yet they're in the claim, yet they match what Bexalta concedes they disclaimed, allow them to recover the scope? And an unbroken line of cases... It doesn't match what Bexalta concedes they disclaimed. Bexalta does not concede they disclaimed the stuff that's in Claim 4 or the stuff in Claim 19. Bexalta concedes they disclaimed derivatives and doesn't challenge the finding of what a derivative is. I agree that they have not said, we disclaimed bispecific antibodies, we wouldn't be standing in front of Your Honor. But they have agreed they disclaimed derivatives and asked the district court, as did we, to figure out what is a derivative and haven't challenged the fact finding of what is a derivative. So the district court found that a derivative is a significant change to an antibody from Column 5. I think... And that's not challenged on appeal, you say? There is certainly no place where they say... I mean, it's based on their own experts' testimony. I'm not sure how they would mount a clear error challenge to it. But there's certainly nowhere where they say that finding was wrong. I don't, I mean... Can you just point to me, just point to me in the record where that is, that a derivative is anything that's changed from an antibody in Column 5? Absolutely. On page A45 in the district court's opinion. The district court finds at the bottom, thus I find that the term antibody derivative was used in the patent to denote antibodies within the Column 5 definition that had been altered in some significant way. As initially drafted, there was no inconsistency between the dependent claims and the Column 5 definition of antibody. That is a finding of what an antibody derivative is based on expert testimony. It's reviewed for clear error in this court. It's also based on the specification, right? It's not exclusively. I apologize, Your Honor. It's not exclusively based on expert testimony, but I believe it becomes reviewed for clear error because of the inclusion of expert testimony. I agree with you that it is not purely based on that. It's also based on the structure of the patent. Well, but if we don't need expert testimony in order to review it, if the intrinsic record produces a result, then we don't have to look at the experts, right? I agree with that, Your Honor. If the court finds that the definition of antibody derivative were clear from the patent, it would not then need to look at the expert evidence on which the district court relied, absolutely. And with that, I would actually like to talk about the language that, because I'm running out of time, on what the language that flows from Column 7 to Column 8 where Mr. Peterson pointed out correctly that the carryover sentence says, the antibodies of the present invention can be prepared by methods known from the art, by conventional hybridoma techniques or means of phage display, which he says that sentence is inconsistent with the Column 5 definition. I would point out that the next sentence says, the production of the inventive antibodies and antibody derivatives may, for instance. I'm sorry to interrupt. Of course, please. I know you have a point you want to make, but I have something that's bothering me, and I want to make sure I ask you about it. I'd love to be helpful. I'm a little troubled because the definition of antibody derivative seems to presume that the antibodies in the specification are limited to that, which is in Column 5. And so it feels a little circular. I don't think it presumes it, Your Honor. I think it starts by saying we have a term that originally was both antibodies and antibody derivatives, and we have a definition of what's an antibody in Column 5. So, in other words, if I didn't agree that what was in Column 5 is definitional, that I'm not limited to think that derivative means something that's based on that definition, right? Just so I understand. I think that's correct, Your Honor. If I may, the language in Column 5 has all the indicia of a definition beyond merely the word is, and I just want to make the point that in the cases like Abbott, where you're applying the higher standard for lexicography, something that came up two arguments ago about more than one plain and ordinary meaning in the art and O2 micro, the parties agree that there's more than one definition of antibody known in the art. Here, the patentee specified one of them. No case has ever held that that's lexicography. Rather, what this Court says is lexicography is defining this to mean a wedding ring. And so when looking at the language in Column 5, in addition to saying antibodies are, and Bexalta would like to read it to say antibodies include or some antibodies are, which it doesn't say. It says antibodies are. There is in Abbott and the other cases rejecting is as sufficient. There are other definitions in the patent that say as used herein or adopt some different definitional format. There's nothing like that here. There is something else, though. Just again, just to make sure I'm airing everything out. If you're going to say that anybody has different meanings in different contexts, why wouldn't you look at the entire specification, which has many different examples of antibodies, or the claims even, which are I think the original claims as they were originally presented, so they're actually part of the original specification, right? So why wouldn't you look at that and say the word antibody, in the context of this entire specification, has a broad meaning? Because in the original claims that claim those embodiments, it said the antibody or antibody derivative of claim one that includes monospecific antibodies, chimeric antibodies, humanized. It never said the antibody that is a bispecific antibody. What about the claims that said, I don't know if it's claim 19. Sure. There are some claims that say the antibody, antibody derivative wherein the antibody is blah, blah, blah. There is one of them. It's claim 19. You're right. It's the antibody or antibody fragment wherein the antibody is humanized. And as we said at the district court, claims four and 19 are the two that are inconsistent. That one is inconsistent. Interestingly, in its initial iteration when filed, it said where the antibody derivative is humanized. Over the course of the prosecution, including after derivatives dropped out, that changed to where the antibody is a humanized antibody. And I can see that that claim and claim four are inconsistent with the notion that column five is a definition of antibody. That's the fact that doesn't fit from our end. The fact that doesn't fit from their end is what did they disclaim. Because what you just heard is that in changing the meaning of the word antiderivative to fragment in claim one, somehow the only thing that did was change claims seven, eight, and 21 and 22. But it somehow had no substantive effect in the claim in which it was made. And that is what the district court found was not a plausible explanation. That Bexalta, having been the party that prosecuted this, still can't say what the change from derivative to fragment meant. And so each of us stares at a part of it that we don't have a good answer for. My explanation for claim four and claim 19 is that those were rejected, and as the district court concluded, Bexalta failed to conform those claims. If the court starts with whatever is in claim four has to be within antibody, then the court is quickly going to conclude that column five is not definitional. I concede that. However, this court's cases teach that you can't do that. You don't look at the plain claim language first? First. But then the same case says, and then you look at the specification. And this court has said, and forgive me for one moment just so I can get the exact case, in North American vaccines that the dependent claims are merely, quote, an aid to interpretation, in dacocytamation that the presumption that they're included is a rebuttable presumption. All of these things operate together. Yeah, but setting aside whether it's a claim dependency argument or not, or whether it's a look at the language of the claims to try to understand what the patent owner or inventor meant by that claim language, then how does that affect whether we're allowed to look at dependent claims to see how the inventor used the words? I think you are allowed to look at dependent claims to see how the inventor used the words. I think you have to look at all of the claims for that purpose. I agree with that. But I don't think that that means that, I mean, to take the North American vaccines case, you have an independent claim to a conjugation of a sugar to a protein without significant cross-linking. And the question is, does that mean cross-linking on the same protein or between two proteins? And you have a dependent claim where everyone agrees you would have cross-linking between two proteins. And yet the court says that's not what claim one means. I'm not sure I have the number right. But that's not what the independent claim means. And so we have dependent claims that give us evidence about what the term means. We also have a rather complicated history in which those embodiments were rejected. And there's no moment when the examiner says, you know what, you're right. I think you do have support for those. In fact, Baxalta's argument after the rejection of them wasn't about bispecific antibodies at all. And so we don't have an answer to why they're there except that they stood rejected and somehow got allowed. And that happens. And this court's cases teach that sometimes, you know, this court invalidates patents that the examiner has granted every day. Sometimes mistakes get made. And the answer cannot be that having failed to conform the dependent claim, they get to recover that scope. But I think that is the central issue. If you start there, you end with column five is not definitional. But what about column six? Do you think column six is at all definitional? Forgive me, Your Honor, which part of it? I'm happy to answer. The inventive antibodies and antibody derivatives, which include bispecific. I think that is—I'm not sure I think that that's definitional, but I think that that's completely consistent with our argument. Because what that says is the inventive antibodies and antibody derivatives and organic compounds derived from— and then there's a list of things. Our contention is and has always been bispecific antibodies are antibody derivatives. They're also arguably compounds derived therefrom. Well, why then, though, when the patent actually calls out antibodies in some places and antibody derivatives, why then when referring to humanized antibody, chimeric antibody, bispecific antibody, do they not say derivatives? Because the term in the art for an antibody that contains the DNA sequence of more than one species is a chimeric antibody. The term in the art for an antibody that contains the binding sequences from one sequence and the rest is human is a humanized antibody. Those are antibody derivatives in that you derive them from other antibodies. But the term in the art for that is a chimeric antibody. In the same way that dry ice isn't ice. In this patent, a chimeric antibody is an antibody derivative but not an antibody. What about the language, and I think this was referred to below, and I know it was referred to in the briefs, at the bottom of column seven. Yes. It says the antibodies of the present invention can be prepared by methods known in part by conventional blah, blah, blah techniques. And I think that these techniques, including fake display gene, we are told that what this is referring to is something where you would probably classify it as a derivative. I think that's right, Your Honor, and that's actually the point I was trying to make earlier. I'm glad we came back to it. If Your Honor reads the next sentence, the next sentence says, the production of the inventive antibodies and antibody derivatives may, for instance. That paragraph is talking about both. The for instance in the second sentence makes that clear. I agree that the word derivative is not in the first sentence of the paragraph, but the whole paragraph is talking about both. That is the one moment in the patent when something that is inconsistent with column five being definitional is said, and the next sentence takes it back. That sentence should read, the antibodies and derivatives of the present invention, as all of the ones that precede it and follow it do. I hesitate to say that that's a typo, but the next sentence makes clear, by virtue of the for instance, that that sentence is talking about how to make both antibodies and antibody derivatives. And I simply don't think that the absence of the word derivatives from the first sentence of the paragraph. Thank you, Your Honor. I appreciate it. Again, Your Honor, I apologize. Don't worry about the time. We went nine minutes over with him, so you talk as long as you feel like you need to, and I'll cut you off when it ceases to be helpful. Thank you, Your Honor. Judge Wallach, I owe you some citations. The discussion of IgM antibodies as being included or excluded by the construction was first raised at the preliminary injunction hearing. You find that on appendix pages 13918 to 13919. And the discussion was repeated at the Markman hearing. That is appendix page 16075. And you also see this in the district court's opinion, noting the dispute on appendix page 53, note 10, discussing the dispute between the parties over whether IgM antibodies would be included by the district court's construction. Thank you. You're welcome, Your Honor. In our view, this claim construction is resolved by the intrinsic evidence standing alone and the strongest possible intrinsic evidence, the language of the claims themselves. What this court said in Becton-Dickinson v. Tyco is that a construction that renders asserted claims facially nonsensical cannot be correct. What do you think the meaning of derivatives is? Sorry. Go ahead. What do you think the meaning of derivatives is? Your Honor, I believe it's ambiguous from the file history, but I believe it is likely coextensive with fragments. And the examiner suggested replacing a term that is not commonly used in the art with the term that is. What my friend didn't suggest was taking a look at how the word derivatives is used standing alone in the patent. And I do think that's helpful to look at. Column 20 is probably the best place to look to. And if you look, column 20, I believe it's lines 34 to 35. Therefore, an antibody, open parentheses, or an antibody derivative, e.g. SCFV, that's a single chain variable fragment, FAB, et cetera. So a FAB is where you've taken just one branch of this Y and broken it off of an antibody. An SCFV is where you've taken a variable region from a heavy chain and a variable region from a light chain and stuck them together using an artificial linker sequence. You see a similar reference to antibody derivatives. So in our view, of course, both of those are antibody fragments. They are pieces of antibody. You see a similar example in column... Well, you actually say column 6 actually says that expressly. So you don't have to say in your view. Column 6 says or antibody fragments, which partially or completely lack the constant region, e.g. FV, FAB, basically all the same thing. Yes, Your Honor. Not all the same, because I guess S, to be clear, SCFV, whatever the heck that is, ain't there, but the rest of them are. And so it defines antibody derivative in column 20 at line 35 and then has almost the exact same, almost the exact same definition with the similar e.g. at column 6 for antibody fragment. Yes, Your Honor. You see something similar in column 30, actually. If you look at lines 15 through 18, again, that's a similar... Though antibody derivatives such as FAB, FAB prime 2, I'm told is how that is said, SCFV, etc. Again, these are fragments. What you don't see is any reference in either the file history or in the specification to antibody derivatives such as... So nowhere in here does it say antibody derivatives, for example, a bispecific antibody. That's correct. Nowhere does it say antibody derivatives, for example, a chimeric antibody. That's right. And nowhere in the specification... Or even humanized. Like it doesn't... Antibody derivative dash humanized. That's correct, Your Honor. And nowhere does it say, either in the file history or in specification, that all bispecific antibodies are antibody derivatives. My friend says these claims were rejected. We disagree vehemently. The examiner never said, I reject antibodies that are bispecific. It said, I reject antibody derivatives that are bispecific. Some antibody derivatives that are bispecific. And again, it's undisputed... How do you interpret that? You interpret it as saying you have no enablement for antibody fragments that are made of bispecific antibodies. Is that how you're interpreting it? That's how we're interpreting it, and it's not all bispecific antibodies. It's a rejection for not all antibody derivatives. It's some bispecific antibodies with reference to different artificial micro-sequences. Your Honor, I'll confess, the prosecution history is difficult to follow. I think the key, though, I go to appendix page 16021. Because my friend's argument is that this amendment from derivative to fragment was the key that unlocked patentability. That by surrendering this vast swath of antibodies known to science, except for this basic description in column five, that that was what finally overcame this enablement rejection. I'll point out, this is the response to the enablement rejection following the amendment. What language do you want us to focus on here on page 16021? What I want you to focus on is the absence of language. Because this is... So if you look just a couple of pages back, you'll see this is where we had a phone call with the examiner. The examiner suggested the amendment. And in my friend's view, the examiner suggested the amendment in order to overcome the enablement rejection. I think if that's correct, what this would say would be, as discussed on the call, we have given up the broad claim to antibody derivatives. And now, as you suggested, claim only antibody fragments. That overcomes the enablement problem. This is where, if you were looking for a clear and unmistakable disclaimer that said the change from derivative to fragment was a narrowing amendment to overcome the enablement rejection, this is where you would see it. If that was the impetus, the enablement rejection for the amendment, surely it would be referred to here. And I'll point out, even if it was a narrowing amendment, there is nothing in the file history that shows that all bispecific antibodies were understood to be antibody derivatives. Your Honor, I believe Judge Stoll, you are absolutely correct. The district court's reasoning on this point is a bit circular. When the opinion is discussing antibody derivatives, both on appendix page 44 and on appendix page 45, the district court is relying on its already determined construction of antibody. So on appendix page 44, since bispecific antibodies are not within the definition of antibodies, they must be within the definition of antibody derivatives on appendix page 45. But the specification makes clear that an SCFV is not an antibody fragment using the definition of antibody from the specification. So the whole reading of the prosecution history is premised on the idea that this definition of antibodies is correct. As we see it, it's quite circular reasoning, and it's unclear to us why the district court continued on with the prosecution history when the reading of the prosecution history was premised on a construction that the prosecution history was then being used to support. So, yes, my friend and I do stand in different places here. What do we do with appendix page 15987, which amounts to the examiner's rejection? I think the one that preceded the page that you pointed us to. And in number 12, it seems that the examiner, at least, wherein said antibody derivative is chimeric antibodies, humanized antibodies, single chain antibodies, bispecific, et cetera, et cetera. So is that not the examiner defining an antibody derivative as those types of antibodies? I don't believe so, Your Honor, because of the wherein. Wherein the antibody derivative is those. And, again, in our view, we've always accepted that an antibody fragment, which everyone agrees is at least a subset of derivative, can be a chimeric antibody, can be a humanized antibody, can be a bispecific antibody. But our view is that the examiner didn't reject antibodies that were chimeric antibodies, humanized antibodies, and bispecific antibodies. My friend and I draw different inferences from that. He suggests that because there's no reference to the examiner rejecting antibodies that were bispecific antibodies, that means they weren't claimed. We draw the inference that they were claimed and not rejected as not enabled. The prosecution history doesn't answer that except for the fact that the claims were actually allowed. Yes, we have different arguments. I'll admit the prosecution history is ambiguous. He can make arguments based on it. But, first, disclaimer is a high bar. It requires clear and unmistakable disclaimer. Our arguments on the face of the claims themselves. Phillips tells you that you look to the claims, and sometimes just how a word is used in the claims can be a strong basis for construing them. It's not a claim differentiation argument, my friend suggested. It's just a usage on the face of the claims argument. And here, the district court found the construction we proposed was an ordinary meaning of antibody. It is an ordinary meaning that renders the claims sensible on their face, where their construction renders the claims nonsense on their face. Neither lexicography nor prosecution history disclaimer justifies adopting a construction that renders these claims nonsense on their face, and the claims should trump the prosecution history. There's a hierarchy here, and you cannot simply say these claims were issued in error. Unless the court has questions. Thank you very much. This case is taken under submission. Thank you.