Oral argument — University of Pennsylvania v. Eli Lilly and Company

Haha, Ok, Yeah Ok, the next argued case is number 17-2397 The University of Pennsylvania against Eli Lilly and Company Ms. McLeod Good morning, may it please the court The Penn Claims on Appeal concern an extremely surprising invention that was made over 20 years ago No one would have believed at that time that you could arrest tumor cells from dividing and then hit them with radiation and have a beneficial effect on radio sensitivity Radiation was known then and was still known to work preferentially on dividing cells But the Penn inventors found that when they inhibited tumor cell division using agents that act through ERB-B receptors not only did they get a beneficial effect but the tumor cells responded in an extremely surprising way rather than dying a slow death by accumulated DNA damage to the chromosome which you would typically see in radiation The specification says that disruption of tyrosine kinase activity such as by inhibiting dimer formation results in cytostatic effect on tumor cells In the great brief at 6 and 7, you argue that that passage quote, cannot be reasonably read to mean that disruption of tyrosine kinase activity such as by inhibiting dimerization always results in a cytostatic effect in all tumor cells Why isn't it a reasonable reading? So that passage, if you read the entire paragraph that sentence is actually repeated twice in the patent One was in the context of tumor cells that have heterodimers that have a mutant EGFR receptor And the patent says in those tumor cells you could achieve the cytostatic effect by inhibiting dimerization It also says it in the preceding paragraph that deals with heterodimers consisting of P185 and EGFR So that sentence, while it provides written description support for the claim limitation does not set forth a universal rule This distinction between a cytostatic effect and a cytotoxic effect is at the heart of this appeal Because indeed, a combination of therapies that each have a cytotoxic effect on a tumor cell as a cancer therapy, not particularly surprising But preceding a therapy like radiation with another one that was expected to impede its efficacy that is particularly surprising indeed It is completely lost in the final written decision, nowhere to be found This appeal comes to this court with at least three errors in the context of inherency I hope to have time to cover each one The first was in the context of claim construction Here the board held that the cytostatic effect limitation was merely the inherent result of inhibiting dimerization The immediate effect of that claim construction was to scoop into the scope of the claim what was disclaimed during prosecution And that was the use of antibodies having a cytotoxic effect The board could only reach this claim construction by one ignoring the testimony of three experts, one of which was Lilly's Dr. Reese specifically stated in his deposition that there were circumstances under which an antibody that inhibits dimerization does not have a cytostatic effect on the cell That very sentence shows that the board's claim construction cannot be correct Two, the board could only reach that claim construction by completely ignoring the prosecution history Even if we agree with you on claim construction didn't the board make the alternative finding that the combination of I'm going to botch all of these names, Sally, DiNardo, and Balaban Inherently disclose a cytostatic effect and then you get everything else through the combination They did, and that determination If that's correct and supported by substantial evidence then would you agree that the claim construction doesn't matter? There are actually two questions, two issues of inherency in what they held The first one was whether the cytostatic effect limitation was present in the references And Dr. Reese's testimony alone shows that that isn't true because he stated that C.2.2.5 sometimes doesn't cause a cytostatic effect So inherency of the cytostatic effect limitation was not proven The second issue in the context of inherent obviousness was ignoring unexpected results And this court recently dealt with this issue in the Millennium and Honeywell cases in which you held that unexpected results cannot be ignored in an obviousness analysis The question must always be whether there's a reasonable expectation of success in the claimed invention During prosecution Penn represented that the presently claimed invention uses antibodies that inhibit formation of ERB protein dimers Continuing next sentence, the antibodies used in these methods are therefore also cytostatic Doesn't that prosecution language support a conclusion that inhibition of ERB protein dimers and a cytostatic effect are co-extensive? I think the Penn inventors did make that argument very early on and it was not successful And in fact, after that, they submitted a pretty extensive IDS with references that showed that that wasn't necessarily the case And as a result, they amended the claim to add the cytostatic effect limitation So it was a critical limitation that was added to the claim that then resulted in allowance And it was error for the board to read the claim as if it wasn't there And I will add that the board reached their claim construction by following a claim construction argument that Lilly made for the first time at the oral hearing This court in Dell v. Acceleron held that the board was obligated to dismiss such a late stage argument If we agree with Lilly, that the PTABs found the challenge claims would have been obvious even under Penn's construction of cytostatic effect and if we affirm the obviousness finding, how are you prejudiced by the alleged APA violation? I would say that in this decision, the board ignored so much contrary evidence Ignoring contrary evidence of record is by definition arbitrary and capricious and the decision is entitled to no deference There is such a thing as harmless error, and you would agree with that Yes, but in this case, the error was not harmless The court completely omitted a reasonable expectation of success analysis in the context of obviousness And even though the final written decision said that they were following Penn's construction, they weren't They completely omitted the fact that a cytostatic effect was distinguished during prosecution from a cytotoxic effect They are not the same. They are not one and the same Going back to the point that you raised about the two different kinds of inherency errors in this inherent obviousness context Here, I think it's pertinent to note that Lilly has conflated these two points in their brief On pages 35 to 36, when they discuss the Honeywell case They say, but Penn doesn't argue that the cytostatic effect limitation was unexpected Therefore, Honeywell is not relevant I think they completely missed the point that there is a sequence of events And I think it was laid out very well in the Parr Farm case Where first you need to determine whether the claimed limitations are in the references And then you consider whether there was a reasonable expectation of success in the claimed invention I don't see any lights on, so I think I'm still within my time I'd like to return to the three buckets of inherency error The third was in the context of the heterodimer claims Here, Lilly had argued that the P185 receptor was inherently present in the cells of Stela When they didn't prove their case, the board substituted its own obviousness rationale Based not on inherency, but based on whether it would have been expected Or would have been understood Let me ask you this, PTAB identified the heterodimer claims as 5, 10, 14, 17, and 20 Do you agree that's all of them? I think there might have been one more I don't have the list in front of me But they were set forth at the beginning of our brief Back to the heterodimer issue They made an argument that it would have been understood that the heterodimers existed at the time Yet they cited Lilly's post-filing date references in support of that argument This was legal error in the context of obviousness of the most egregious sort You agree that presence or absence of a reasonable expectation is a question of fact, right? Well, again, if the court is struggling with the deference to give to this case I would say that ignoring contrary evidence is arbitrary and capricious That's just substantial evidence If they ignore it too much, then their decision lacks substantial evidence Ignoring just means they didn't believe it It can't just be ignoring it It has to be the weight of the evidence is so leaning one way that no reasonable person would come up with the board's conclusion It's not really the arbitrary and capricious standard, it's a substantial evidence standard for facts While you're answering that, how is that decision employing a new legal theory rather than making factual findings? Let's talk about the weight of the decision Penn presented at least five references that led away from the claimed invention Because they showed other instances where agents having a cytostatic effect were administered prior to radiation and had a detrimental effect on radio sensitivity You have all kinds of evidence on your side, but the problem for you on the substantial evidence review is that it doesn't need to be a ton for the board's decisions to be supported and the board cites evidence on all these points Why doesn't that just defeat your claim? The board misapplies the law to those facts How do you mean it misapplies the law? Because I thought we were talking about substantial evidence You're not saying they misapplied the standards, how did they misapply law? You're confusing me by going back and forth between weight of the evidence and a legal error Sorry, I apologize Back to the weight of the evidence Even if this court disagrees that the references that Penn submitted did not rise to the level of teaching away they were at the very least evidence The question isn't whether we disagree with your references and whether they teach away It's whether the board's factual determinations on these key points lack substantial evidence And when we're looking at that, we're not looking exclusively at what you think we should have decided because we're not the trier of fact It's whether the evidence cited by the board is sufficient to reach that determination We could entirely agree with you up here on the factual outcome and still have to affirm the board on the substantial evidence standard I do not think that there's substantial evidence to support the board's decision Why aren't you concentrating on the evidence cited by the board against you and why it's sufficient rather than repeating your arguments about why you should have prevailed in the first place? Because I think that the board ignored... That's what you have to do. You have to show us that the evidence the board relied on is insufficient to support its findings under the substantial evidence standard It's not that yours was good enough to reach the contrary result I will point, since you're asking about the evidence I think the references, the combination of the references already show us that inherency did not exist It's already in the record Let's hear from the other side. I will save you rebuttal time for any further comments Good morning, Your Honor The combination of Saleh, DiNardo, and Balaban teach everything required by the claims of the 558 patent What's your best argument and your response to your friend who says it doesn't, that Saleh shows cytotastic effects and specifically what evidence supports that Because I think that's the key point here Even if you were to take Penn's construction, the board found that there was a cytostatic effect based primarily on the Gulliford reference, which is prior art and shows an inhibition of growth of the A431 cells The Saleh 2 article shows that that inhibition is in the G0 or G1 phase as required by Penn's construction No, I get that. That's where I thought it was But what's your response to your friend's I think what would be their argument that that's not enough to show inherency because the record also shows that there's a cytotoxic effect of the same thing and that if you have both of them going on at the same time, the cytostatic can't be inherent when you have both of them possibly at the same time Despite what Penn argues in their reply brief, which I believe is a new argument in their reply the patent does not talk about cytostatic antibodies or cytotoxic antibodies It talks about effects. You inhibit the dimer formation or you disrupt the tyrosine kinase activity and the result of that is a cytostatic effect In your view, even if the effect of this was 75% cytotoxic and only 25% cytostatic that would still be enough to show that there was a cytostatic effect from administering this Yes, both sides' experts testified that you could have both a cytostatic effect and a cytotoxic effect We've never disputed that cytostatic effects are different from cytotoxic effects but the claim, the scientists both agreed that it was possible to have both at the same time Neither side's experts disputed that. Penn admitted that during the hearing and the claims don't say cytostatic to the exclusion of cytotoxic effects That's nowhere in the claim and so Penn is trying to read in a negative limitation that simply doesn't have any support in the claim or in the specification I'd also like to point out as well that it isn't even part of their construction My friend said that the court misapplied Penn's construction because it didn't deal with whether or not there was a cytotoxic effect But I turn the court's attention to page 31 of their opening brief in which they specifically state what their definition of cytostatic was and it was inhibiting growth by accumulating cells in the G0-G1 phase such that when the agent is removed, growth resumes Those were the three parts of their construction and their construction did not require that a cytostatic effect means that you don't also have a cytotoxic effect In the red brief at 34, you argue, I'll quote for you Legally, it's not necessary to show that C225's cytostatic effect was known at the time because, quote, inherent properties need not have been recognized in the prior art Are you asking for a blanket rule that in inherent obviousness cases an inherent reference needs to antedate the critical date of the claimed invention and keep in mind what the implications of that are? What we're saying here is that, and this was made clear in this court's recent decision in July of this year in endopharmaceuticals, that you can that simply because a property was not known or appreciated by a person of skill in the art at the time of the invention does not mean that it can't be used as part of an inherency analysis even in the context of obviousness Would that undermine the purpose of an obviousness analysis which looks to what a person knew at the time the invention was made? Not in this case, because this is not a situation where we're talking about That's why I said keep in mind the implication Because we're not talking about a situation here where we're combining different compounds or combining different elements together Penn relies, for example, on the Honeywell v. Mexichem case in which you were talking about forming a refrigerant by combining two different chemicals together And although the properties of that combination were inherent a person of skill in the art would have no reason to make that combination in the first place The question that Judge Wallach, the sentence Judge Wallach read absolutely has to be wrong But it really depends on whether you're claiming the old composition which you can't do no matter what you discover or whether you're claiming a new use based on a newly discovered property We all know that when you're dealing with chemicals and biologicals and so on the properties are there So you can all of a sudden say they're inherent and school's out That's not how it works So let's get back to the foundation of what it is that Penn contributed to this art and why the PTAB was correct or theories that they were correct in saying that we're talking about the Saleh reference which teaches administering C225, a known antibody, to A431 cells Whenever you do that, and this much is undisputed Penn has not disputed how the antibody works you disrupt tyrosine kinase activity, you inhibit dimer formation and under the board's claim construction, that's enough to have a cytostatic effect So simply practicing exactly what's taught in Saleh We're not talking about combining different things or adding anything new An automaton who is doing nothing more than repeating the experiments done in Saleh will inhibit dimer formation, will disrupt tyrosine kinase activity It is that sense in which it is inherent because it is the natural, inevitable result of practicing exactly what Saleh teaches You don't need to use any combinations But we disagree with the board's claim construction, that just inhibiting whatever that is is equivalent to cytostatic effects What further do we need to do to support because I do think the board made an alternative finding that Saleh still inherently shows a cytostatic effect apart from its claim construction. What supports that view? There were actually two alternative grounds. One was that it applied Penn's construction and I already talked about that a little bit, and that was relying on the Goldstein reference and Saleh 2, which showed that you met all three portions that you arrested cell growth by accumulating cells in the G-section Saleh 2 specifically shows that there is a cytostatic effect Is that your position? Both Goldstein and Saleh 2, but it talks about if you want to walk through Penn's construction that was the evidence we cited for that. The third ground was that it's undisputed that M225, which is the murine antibody from which C225 was derived, has a cytostatic effect That's, again, never been disputed by Penn and the board found that, relying on the Goldstein reference, that C225 has the same biological effect as M225. The fact that M225 had a cytostatic effect, the effect of C225 is the same as the effect of M225, also shows that C225 has a cytostatic effect. Those were the two alternative grounds that the court relied on independent of its claim construction. In the blue brief, Penn argues that the PTAB aired relying on three references, Waterhouse, Gabbritt, and Guilford. When I look in the red brief, at 59, you argue Guilford is prior art, but you don't mention the other two. Are you conceding? Guilford is the only one of those three that is prior art. The other two were published later but the fact that the board cited additional references that confirm what was already known in the prior art doesn't take away from the substantial evidence supporting the board's finding. I'd like to turn then to the question of the P185 heterodimer claims that were addressed briefly. Isn't there a fundamental problem here that the board used a different theory than was in the petition? That's not a substantial evidence problem. That's a fundamental APA due process issue. I disagree that it's a different theory. The theory was obviousness based on the combination of Salé, DiNardo, and Bauerback. That's exactly what... The board did not actually use the word inherit. That's true. What the board did look is it looked at all of the evidence put forward by both sides. It looked at both Lilly's evidence and Penn's evidence and it credited Lilly's evidence as teaching that A431 cells do in fact express P185. It specifically rejected Penn's evidence as not addressing the issue. There is a very strong factual finding there. Again, that's not a different legal theory. That's a factual finding. When you say express P185, is that equivalent to saying it showed heterodimers? I get tangled up in the list stuff. Gulliford shows both that it expresses P185 and that P185 forms heterodimers with EGFR. That's shown in Goldstein. Your position is that Gulliford expressly shows heterodimers? Yes, it does. It's Figure 5 of Gulliford. It's page 2842 of the appendix. This is explained in the Reese reference, the Reese Declaration as well. Figure 5 on page 2842, Figure 5B, the top left corner of that assay. Figure 5B, the figure on the right hand side. The top left corner shows the fact that it's in both the row and the column. It shows that it's a combination, that it's both EGFR and P185. This was explained in the Reese Declaration as well. Here's the problem with this. Does it use the word heterodimer here anywhere? You're asking me to look at this figure and believe you. Then you're pointing me to the Reese Declaration. Does the Reese Declaration specifically explain why this shows heterodimer? Yes, and I would point the court to the legend beneath the figure as well, which talks about hetero-algo-merization. But the board relied on Reese Declaration. I can't translate from this to the fact that it actually shows heterodimers. This is what the board relied on, however. If this doesn't actually show heterodimers and the board's reliance on it lacks substantial evidence because I think this is the only piece of prior art that shows it. The other two the board used were not prior art. Again, there's never been any dispute that this in fact does show it. Penn has never denied that A431 cells at least sometimes express P185. I'm sorry, I get confused with all this stuff. I think their argument is not the problem that it shows heterodimer, but it doesn't show that the C225 does whatever their claim is to heterodimers. Again, this was a point that was conceded below. They acknowledged in their patent donor response and at trial that C225 disrupts dimer formation. I believe this is page nine of their patent donor response in the footnote. They specifically acknowledged that. Again, that's not an issue that was raised below. That was something that Penn admitted before the patent trial and appeal board. Again, I just want to make the point that this was not a new legal theory. This was a factual argument that Penn raised for the first time in their patent donor response as to the extent to which A431 cells express P185. I don't know. It sounds like a new legal theory to me. I thought you were arguing that heterodimer was inherent in Sally and they didn't agree with that. They had to go to Guilford to get there. That sounds like a new combination to me that they didn't have a chance to respond to. This court has said in Genzyme and Ariosa that the board may properly look to other references as to what a person of skill in the art would have understood in reading Sally at the time. That's exactly what the board did here. We specifically talked about Guilford, Waterhouse and Gaborit in the petition. The board relied on the exact same references, the exact same teachings within those references. This is not a situation where there was some new art or some new legal theory that was put forward for the first time. It's exactly the same legal theory, obviousness based on the combination of Sally, DiNardo and Balaban. All the board did was it relied on an admission by Penn that A431 cells at least sometimes express P185. Under this court's decision in Ravalmo v. Bowler-Edelstein, the board is allowed to do that. The board is allowed to rely on admissions from the patent donor in making its determination. Penn cites the Edel case and some others, but in Ravalmo, this court specifically said that those cases don't limit the board's ability to make its own legal conclusions based on a fact that has been admitted by the patent donor. That's exactly what's happened in this situation. Thank you, Your Honors. I'd like to respond to several points that my friend here made. He cited Endo B. Custafarm. What this case held was that a party cannot take advantage of inherency to reach an undisclosed, unknown limitation that is not necessarily present in the prior art. Their own expert testified that there are circumstances under which C225 does not cause a cytostatic effect. That is the answer. There is no inherency of that limitation in the prior art. I'd like to respond to the point that they made about Salie. My friend said administering C225 when you do this, it's enough to have a cytostatic effect. This was disproven in Salie 2, same group of authors, same system in which they disclosed that C225 was cytotoxic. And not only that, we already know from that reference what happened when they administered C225 in front of radiation. It was not radiosensitizing, which is not a surprise because it was not cytostatic. I'd like to respond to another point that they made about Genzyme. There is a distinction between new evidence and new arguments. Switching from an inherency argument where you're arguing that something necessarily exists to an obviousness argument, that is a new rationale. Are you talking about the heterodimer issue now? Yes, I am. There is a distinction between arguing that they were inherent and whether it would have been obvious to treat cells that express such heterodimers because those types of tumors were extremely aggressive. Can you address that one on the merits too? Why Guilford, even if they properly raised the obviousness argument based upon Salie and Guilford, that they don't show whatever the heterodimer issue? Gulliford showed that in that system there was a low level of P185 expression. We submitted, talking about the weight of the evidence, we submitted I think it was at least six references showing systems where A431 cells do not express P185. And Gulliford technically was not prior art because Penn established a date of conception that predated that reference. What I really wanted you to address was the very specific part of that reference he showed me and talked to God and said it meant something. Do you agree with what it means and that the problem is that it doesn't, even though it might express heterodimers, it still doesn't teach using whatever the other thing is too? Oh, absolutely. You raised the point about the Lee reference. It wasn't until the Lee reference that we even knew that C225 inhibited dimerization. Even if Guilford shows heterodimers, it doesn't itself teach? It doesn't itself teach that one could take C225 and inhibit those heterodimers without the teaching of Lee. You're still relying on an inherency argument. In that context, the fact that A431 cells sometimes express heterodimers is not enough. I want to use my last minute and a half to close here. In sum, there is not a single reference of record or combination of references that establishes a reasonable expectation of success in treating a tumor with an ERB-B antibody or any agent for that matter that has a cytostatic effect on the tumor cell prior radiation. Rather, the art suggested that the opposite, that this order of administration would have been expressed and expected to have an adverse effect. Yet the inventors found that the nature of the response of the cells to radiation was radically and beneficially changed. This is the very essence of non-obviousness, and it was legally improper for the board to dismiss these results as merely inherent after combining references that teach neither the claimed limitations nor the unexpected effects. For this reason, we respectfully ask that this court reverse the decision of the board. I still have 20 seconds if there are any remaining questions.