Oral argument — Grunenthal GmbH v. Alkem Laboratories Limited

We have a busy morning in front of us. We have four argued cases. Before we move into arguments this morning, we're going to take care of administrative in-house matter. Kate, will you stand? Kate, it's been a year since you started clerking for me, and now it's over. And I bet that time has flown past pretty fast for you. It has for me because the older I get, the faster time seems to fly. But for you, I bet it flew by pretty fast because of the hard work that you've been doing. And this hard work was exemplified by excellent work. Before me stands an attorney, whom I'm sure will be an extraordinary member of the Bar. And someone that we can look back and continue to thank for her contributions, both to the Court, but especially to my chambers. So with that, my colleagues, I'll turn this over to Judge Taranto as I step aside, as I move for the admission of Catherine Marcon, who's a member of the Bar in good standing with the highest courts of Texas. I have knowledge of her credentials and am satisfied that she possesses the necessary qualifications. Welcome to the Bar of the Court. Thank you very much. Please raise your right hand. Do you solemnly swear or affirm that you will comport yourself as an attorney and counselor of this court, uprightly and according to law, and that you will support the Constitution of the United States of America? I do. Welcome to the Bar of the United States Court of Appeals. The first case this morning is Grudenfall, GMBH v. Alcom, Laboratories Limited, 17-1153. And we had granted additional time for arguments, given the number of issues involved and the number of parties involved. And I received today the breakdown of the arguments. I want to make sure that I have this correct. Mr. Gajarsa, you're going to go first, and you have six minutes, and then you're reserving three for rebuttal. Yes, Your Honor. Is that correct? Mr. Alley, you have five minutes and one minute for rebuttal. Is that correct? Yes, Your Honor. And Mr. Joffrey, you have five minutes. You don't get to rebut today. No, I'm rebutting. Okay. Mr. Stichman, you have 15 minutes and five minutes for your cross-claim. Is that correct? Yes, Your Honor. Thank you. Okay. So you all help me out now, if I don't get the order calling you up. But let's get started. Good morning. May it please the Court. My name is Robert Gajarsa from the Latham Watkins for Hickam and Westford today. In my opening, I'll be discussing the lack of utility for the 364 patent, and if there's time or questions, the invalidity of the 130 patent. The 364 patent lacks an adequate disclosure of utility, and the district court committed legal error by concluding otherwise based on the known analgesic effect of tripentatol hydrochloride and example 16. The utility of the prior art being improved is not the touchstone of the utility inquiry. The improvement itself is that touchstone, and that finds its basis in the statute itself, section 101, where an improvement to a composition of matter is allowable. However, that improvement itself has to be new and useful. And this Court's predecessor, the CCPA, explained that also in Reese Selmy back in 1946. Is Selmy a 101 case or is Selmy a 103 case? I mean, it talks about a lack of invention over the prior art, but I think one could read Selmy as fitting quite neatly into a section 103 context about, you know, overlapping claim ranges with the prior art and then determining whether there's a criticality as to the claimed range. And so to me, I could easily read Selmy as being really more about a modern day 103 case than a 101 lack of invention case. They did use some parlance, Your Honor, that does speak to 103 in today's terms. However, they clearly stated that the invention there, the improvement to the alloy steel, the utility that it had had to be new and materially different than the existing steel. And that statement came and it found its support in the Dow Chemical versus Halliburton case of the Supreme Court in 1945. And in that case, the Supreme Court clearly stated that if you have a known chemical and you simply dilute it, you have to show that the critical improvement there, the advance that you claimed over the prior art, had to be linked to that improvement itself. The dilution has to have some type of function. And indeed, that goes back to the 101 statute itself that shows that the improvement has to be new and useful. And it's very clear here that what we're talking about is an improvement to the dependetol hydrochloride compound. How do you square that theory with the kind of, I think, equally well-established theory that you can get a patent on something new and different but worse than the prior art? We're not saying that it has to be superior to the prior art. It just has to have usefulness. As the Supreme Court in Brenner said— So it does what the prior art does, but just not as well. In that case, you'd have to show that there's usefulness to not doing it as well. I guess I'm struggling to understand how one can frame the question in terms of the word improvement, which suggests something that you can do with this that you couldn't do with the previous thing. Well, indeed, that comes from the statute itself, Your Honor. In Brenner, the Supreme Court said that the first step to determine what the usefulness is is look into the intent of Congress. We believe the intent is very clear here, that it has to be an improvement to the existing composition of matter. So you'd have to link the worse-off principles in that hypothetical. I'm sorry, and how can something that's worse than the prior art be an improvement? That's a very good question. In that context, you'd have to prove that being worse than actually has some function. For example, in the dilution case—or, sorry, in Dow Chemical— Cheaper but less efficient, but then I'm not sure you'd say it's worse than. Right. There's tradeoffs, Your Honor. So in Dow Chemical, one of the aspects of the fracking liquid that they were using, the reason that they said they were diluting it was because it would be slower to dilute the limestone, and there are advantages to that. So it was worse in one aspect, but it was better in another. Are you saying that if a first inventor comes up with a mousetrap, the second inventor that comes up with a different way of trapping mice has to prove that that second mousetrap is functionally somehow better than the first mousetrap? It can't just merely be a different way of doing it that is novel and non-obvious over the first mousetrap? That could be a very complex question, Your Honor. And here, though, the question is whether or not you have an existing mousetrap, for example, and you change one element and claim that's an improvement of the mousetrap. For example, if you had a simple snap mousetrap with a bar, and then you change the form of that mousetrap by simply— You seem to be arguing a novelty, not utility. Well, the utility is separate from novelty, and the Supreme Court in Brenner specifically says that utility forms an essential component of the quid pro quo exchange for the patent system itself. And one of those important aspects of it is showing that you have usefulness. And indeed, the statute itself goes back and says the improvement has to be useful over the prior item. Am I remembering right? And there are so many patents here and issues that may get this wrong. Am I remembering right? This report essentially found two forms of utility. One, the same as the underlying molecule, the pentadol, is that it? Yes. And the second is that this particular crystal structure has greater stability under certain kinds of conditions. Does your argument require saying the first is legally irrelevant and the second is unsupported by the record? No, the second is also a legal error, Your Honor, because the basis for it was the same. And I think it's important at this point to step back if we're talking about example 16 to understand what the term stability is because it's a basic misnomer almost. In classic abstract sense, stability, of course, could be useful in a pharmaceutical composition. You could have extended shelf life. You could have extended resistance to degradation, higher fruit effect, something like that. But here, the stability that they're talking about, that word, isn't talking about that classic sense of stability. It's talking about conversion from one form to the other form, form A to form B and form B to form A. And as the cross-appellants themselves state, it doesn't matter which form you have for the analgesic properties. Indeed, their experts stated, as Grumenthal says on page 55 of his brief and argues the point, it doesn't matter which one you have because they're both going to be the same. So what's the point of having form A? And that's exactly what we asked the district court. And that's a proof we put on below is that it doesn't have any utility. And the district court came back and said as a matter of law, and that's at Apex 147 to 148, that as a matter of law, you don't have to prove that this stability makes it a better pharmaceutical composition because it doesn't have to be better than the prior art. But stepping back, if it's not better than the prior art, what does example 16 show? It simply shows that at room temperature, the known to pentatol hydrochloric compound exists in form A. So if form A is very stable at ambient conditions, then my understanding is it's more useful at that point. That stability makes it more useful to make pharmaceutical compositions. That's exactly the finding we asked the district court to make, and she didn't reach it. And we put on evidence showing that it wasn't more useful. And the district court refused to make the finding that stability in that context, conversion stability, was actually useful. She didn't make that finding. She refused to make that finding based on her interpretation of the law, which was saying you don't have to show it's better than the prior art. But without that finding, there's nothing to show that form A is better than form B. All you're left with is the discovery itself that at room temperature, form A is indeed the form that's prevalent. It's still, as Grumenthal says, the identical chemical compound, regardless of if it's form A or form B, just a crystalline structure that matters. Morning, may it please the court. The one issue that Alkim is addressing is obviousness of polymorphs. Polymorphs should be found obvious where the patent does nothing more than follow a published guide for how to screen for polymorphs, and then claims the result of that published guide. That's what happened here, so the district court's conclusion should be reversed. The reason that we have that obviousness issue here is the prior art already had to pentadol hydrochloride and said that it was a crystallized out. In other words, it was a solid that had a crystal form. The crystal form itself was not disclosed in that patent. However, a guide was published by Stephen Byrne. Dr. Byrne's guide was how to do a polymorph screen. The FDA, as of 1995, wanted anybody developing any drug to do a polymorph screen to determine what the crystal criteria and structure would be. Following that Byrne implementation of the guide, Byrne explicitly identifies nine solvents that one should use to do the crystallization experiment, and then also the evaporation and cooling conditions that a POSA would then implement. The whole point of the Byrne article, Your Honors, was to determine the answer to the question, are polymorphs possible, yes or no? So the amount of information that Byrne intended to provide was to get a POSA to implement the polymorph screen. And at trial, as to that Byrne reference, Dr. Steed, which was Alchem's expert, proposed that, in fact, that is all that anyone needed to do. In fact, when that was done by SSCI, a third-party outsourced company, SSCI was managed by Dr. Stephen Byrne. Did SSCI actually follow the Byrne disclosure every step of the way? My understanding was that it didn't quite use every single solvent disclosed in Byrne, and, in fact, it may have used one that wasn't disclosed in Byrne, and maybe there was a lot of overlap, but it wasn't exactly following Byrne to a T. Is that right? To your response, Your Honor, no. The only unrebutted testimony is by Dr. Steed as to Byrne. No? It didn't follow Byrne to a T? It did, and then it added – yes, it added – it had eight of the nine solvents, Your Honor. It didn't have one of the nine solvents, but it did have the conditions that are spelled out in the Byrne article. Then there are a lot of other also tests not published in Byrne in addition to the reference that SSCI did. But in this case, that distinction really doesn't matter. But I guess what I'm wondering is to what extent can anybody say that there was a very defined set of procedures that one would do to do a polymorph screen? Like was there an answer book already out there that this is what you do, this is the universe of procedures you undertake whenever you do a polymorph screen? No, Your Honor, it wasn't an encyclopedia of the universe. However, Byrne had enough information to say the first step, the low-hanging fruit, the easiest path to take is the nine solvents. The concern I have is the other side put on evidence and testimony that there is some huge number of possibilities that you would do and it would take a lot of time. And then don't we have case law, maybe O'Farrell and Kubin, that say when there are a large number of possibilities and there's nothing in the art that really suggests going down a particular subset of those large number of choices, then we can't say it's really obvious to just start shooting in the dark or start walking into a dark forest looking for treasure when we don't really have any information as to where in the forest we should start walking down. That would be correct if the Byrne reference was not there. Now I'm asking about Byrne because, as I understand it, there's nothing in the record that says, Byrne, this is the gold standard, this is what you do, you do precisely this, and if you do this, then you're going to get the polymorphs. It doesn't quite talk like that, so that's why I'm trying to now understand what was the understanding in the art at the time of how to do a polymorph screen. Was it something as indeterminate as the other side is saying, that there's just this huge number of choices? Yes, maybe there's a starting point, but in the end, it's a huge number of choices that you would want to do, that you would need to do to satisfy yourself, as opposed to some very limited set of things, which is what I guess you're trying to advocate. Your Honor, there are directions that a person already skilled in the art could take, and Dr. Bernstein and the adversary here is taking the position that, in general, not looking at Byrne, in general, there are a lot of choices that one could make, and the district court made the same finding. In general, there are many choices that one could take, but the other side did not dispute at trial, and the district court did not make a finding specifically as to Byrne that said that that was different, or that was not narrow enough. And in fact, Byrne, within the world of possibilities that a person already skilled in the art would have, which includes different solvents, and as you see in the briefing, it includes anti-solvents or slurring, those are not discussed in the Byrne reference. The Byrne reference is one technique, crystallization, and has nine solvents, and it gives the evaporation conditions. So it's possible... You're making a rebuttal time now. Thank you, Your Honor. I'll just conclude then by saying that this is to show the difference between those polymorphs that are obvious, following the one protocol that's published in the prior art, and giving that targeted approach, versus other polymorphs that may not be when there isn't enough guidance, and that would open the door then to what the other side has explained as a lot of variables that one could look into. Thank you very much. You may proceed. Good morning, and may it please the Court. This case involves three distinct and separate inventions. The first occurred many years ago when Grunenthal synthesized a new class of opioids that had unusual characteristics that could treat nociceptive pain. That's the invention that set forth in the 737 patent. The next occurred when Grunenthal, a separate team at Grunenthal, synthesized a brand new crystalline form of one of those compounds, and that form, which had not existed before, was tepentadol form A, and that's the invention claimed in the 364 patent. The third invention occurred years later. Another team at Grunenthal discovered that tepentadol had the unusual characteristic of being able to treat polyneuropathic pain without destroying the nociceptive pain pathway. That's the invention of the 130 patent, and that's what led FDA to approve tepentadol as the first and only opioid to treat polyneuropathic pain. Now, in analyzing the validity of the 364 patent, the district court undertook a highly factual analysis of the 737 patent. Can you address the obviousness point? It seemed to me that the district court had a couple of different components to the there would be a lot of choices here theory for rejecting obviousness. But the one that the other side focuses on is the idea that there clearly was an interest, supported by the FDA guidance, in figuring out what polymorphs may exist. And there was, at least at the core, a routine set of procedures to follow, if not to get a complete answer, at least to start. And whatever would fall out of that process, which the FDA more or less tells you to follow, gives you an incentive to follow, ought to be obvious. Well, Your Honor, I think the analysis in that capacity starts really back at the question of, is tepentadol even polymorphic? If we assume that you would go down that path, then the question becomes, what tests would you apply and how would you apply them, burn being one of them? But there were other articles and other prior articles out there about all kinds of different tests that need to be done. We don't know whether or not there's hydrates, solvates, polymorphs. We don't know how many forms of polymorphs we have. We don't know if we're going to get a polymorph that is more stable, or if we get one that's more stable, we have no idea if it has the same bioavailability as the crystal that was previously made. Remember, what happens here, which is sort of overlooked by counsel here, is we not only discover a new crystalline form that happens to be stable at room temperature, but it also happens to have the same analgesic property as the original crystal form. There's nothing in the prior art that suggests that's going to happen. In fact, there's lots of science out there that suggests that the more stable form may not be as bioavailable as the original form that's crystalline. So we have a lot of variables. The way I look at it is we've got a lot of levers that are out there to move into different combinations to try to even look for a polymorph to begin with. Burn is helpful, but it's not the answer book. I like that. It's not the answer book on what to follow when out pops polymorphs. That's not what the record shows, and that's certainly not what the evidence was in this case. There's so many variables here. And the district court weighed all of that evidence, heard from all the experts. Did the district court rely on the notion that you articulated as part of your litany that even when you get a bunch of crystals falling out of the process, that you don't know the properties of them, that is, whether they'll be good stuff or not such good stuff? Yes, and what the district court cited in that capacity was the disillusion test. After Form A was identified, they had to analyze Form A compared to Form B. And the disillusion test is what told us that it most likely had the same analgesic property as the original Form B. And that itself wouldn't be routine? It would be, well, but the results, more importantly, the results would not be known. In fact, it's sort of counterintuitive given the science. Yeah, and it is at the core of their theory that you can have and maybe even automatically have, but at least can have obviousness when the tests that you conduct are ones as to which you have no idea what the answer will be, but whatever the answer is, you're going to get those answers. So the unknown, this may be, in fact, crossing a doctrinal bridge that we have never crossed, and it certainly goes beyond this language of identified and predictable in the obviousness-to-try paragraph of KSR, but why are they wrong? Well, I think— If everybody would at least starkly burn, run the tests, you either find nothing or you find something, why isn't whatever you find obvious? Well, I think it goes back to the reasonable expectation of success. It's what we're doing in that capacity is taking the result of the test and saying we would have had a reasonable expectation looking backwards. When we start the test, let's—I'm assuming you're hypothetical. Let me assume the extreme of your hypothetical, that everybody knew exactly what the test was going to be. And everybody would know to start with example 25. That's a separate point, which I'm putting on one side. Then it's still there's a real big question out there as to what the reasonable expectation is in coming up with the invention at the time of the invention or time of the work. We have no idea before we send this thing through the polymorph machine, if that's what the court is, before we send this thing through the polymorph machine, we have no idea where we're going to get on the other side. We have no idea if we're going to get polymorphs. We have no idea if we're going to get hydrates. We have no idea we're going to get solvates. Why wouldn't it be that getting the information of whether polymorphs exist and that if you had a reasonable expectation that you would get that information would be good enough? So let me try to crystallize down this hypo even more. Let's just say that it was well understood in the art and the FDA said so and commanded that when you develop a new compound, you should go look for polymorphs. And that it's well defined that the way, the way to get polymorphs is to add water and keep the liquid up to 300 degrees Fahrenheit. And then you will get whatever polymorphs there are. They either exist or they don't exist. And if they do exist, you'll be able to define them. Are you saying that those resulting polymorphs would in fact be a non-obvious invention that merits a patent under those circumstances? Under those circumstances, in the hypothetical that you gave, which doesn't have any of the variability that I think exists in this case, I think it would be a very difficult case to make of non-obviousness. Because I understand a lot of your argument to be based on the theory that, well, even under my hypothetical, we would have no idea what structures there would be if any would exist and what they would look like and what kind of properties they would have. And so therefore, the inventor ought to get credit for that. But under my hypothetical, it seems like it's just extremely routine, automaton-like activities that you would undertake, that you would necessarily undertake. Yeah, and as I understood your hypothetical, it really reduced it to a binary analysis, right? You send it in and get information. The one issue I think I might take with your hypothetical as we apply to this is don't forget the case law here. We've got to have a reasonable expectation of success of the invention, not the information, not the process, not the machine, not the levers, not anything else. The inventor has to have a reasonable expectation success of getting the invention. The invention here is a brand-new crystalline form. And by the way— And by the way, without—am I remembering right? This is the patent for which there's no findings of unexpected results. There's none beyond—no, there's no separate finding in that respect beyond just the utility. And I do take issue with counsel's comment, by the way, in his opening. This is not an improvement patent. There's not an improvement here. This is a brand-new crystalline form that didn't exist before the inventors of the 364 patent made it and identified it. It didn't exist. The example 25 in the earlier patent made form B. And the appellants put on no evidence, absolutely no evidence. The district court saw no evidence that if you put anything into a polymorph screen other than form A that you will get form A. What counsel forgot to tell you in the record is every test they ran, they ran on form A. So it's not surprising that if you put form A in, you get form A out. What they didn't put on any evidence of is if you put in form B, like from example 25, that you would get form A. And that makes this novel and non-obvious. It seems to me when I look at the graph on appendix 57373, that's the burn chart. The text there pretty well describes the entire process, and it describes processes that I think that you're saying that's where the variability resides. But it lists out the entire process. It says it outlines investigations in the form of polymorphs, the analytical tests that are available for identifying polymorphs, studies of the physical properties and controls that are needed to ensure the integrity of drug substance containing either a single morph or a mixture. What's left? I mean, that's pretty thorough. Again, and we don't deny that those tests exist. By the way, burn is one of many in the art there. They're describing what we call at one test called a DSC with lots of variable. There's other tests out there as well. But assuming you were to pick out burn and assuming you were to follow that graph that you were looking at, Your Honor, again, there's a bunch of if-thens. I mean, this is sort of like an if-then claiming, right? If this, then that. If this, then that. If this, then that. But that's not what this patent is. This patent is not on a polymorph screen. This patent is on a brand new crystalline form that had not been made and had not been known prior to the inventor's work. That form was discovered or made known through this polymorph testing. Right. It was made known and identified. But don't forget what we did. We made it. We had to make that new crystal and discover it before we could claim it. And again, I can't emphasize enough, the district court had all of this evidence, including the burn article, including the flow chart Your Honor's looking at before it, and heard testimony from some of the top polymorph experts in the world on this issue, and made a very detailed factual analysis and concluded that in this case, pentadole hydrochloride produced or could not have been predicted. It was non-obvious in this case. I guess it's sort of that point that from the beginning of what I think is my understanding of the case has intrigued me most doctrinally. This does seem to be a case in which I think it's actually undisputed, maybe even expressly conceded by the other side, that what would come out of the test was not predictable. And so the theory is, as long as the test was utterly routine, whatever comes out is obvious, unless there's unexpected results, which there aren't here. What is wrong with that theory as a matter of obviousness law? It feels newish. It feels potentially dangerous. But I'm not getting much of a sense of how to think about whether that is a significant doctrinal step and whether it's a bad step or a good step. I would urge the court that I think it's a bad step. That assumes that, number one, there is a set number of parameters that are clearly identified as the polymorph test. And I understand that's the hypothetical the court's working with. It also assumes that there's something in the prior art, something about Tepentadol, something about its molecular structure, something that suggests that this may be polymorphic and there may be more than one form. Don't forget, when we put this through polymorph screens, we don't know if there's three forms, four forms, five. We don't know how many forms there would even be, even if it was polymorphic. This is a completely unexpected situation when we're talking about, even if there is support to do a polymorph screen, we have no idea and the inventors of the 364 patent have no idea what they're going to see or get. There's no level of predictability. In fact, does it lead you through that? It's like you said, it's an if. If you get this and you go here, and if you get that, you go on to the next step. But whether or not the test is easy or laid out to get to the invention, as Section 103 says, doesn't create an obviousness issue. It doesn't matter how the invention was arrived at. What matters is whether the invention itself is non-obvious in light of the prior art. Well, it goes to a reasonable expectation of success, doesn't it? It does, but if the success is defined by the invention, which the case law of this court says, the invention here is form A. It's not doing a polymorph screen. It is having crystalline form A, which is a specialized crystal. And in this instance, it does have utility. It is stable at room temperature and has the same pharmacological effect as Section B. What about our case law on result-effective variables? So we have case law that if there's a claimed invention that says you can make a certain composition with a temperature range of this, pressure range of that, and we have case law that says, well, even if those temperature ranges and pressure ranges are different from what was disclosed in the prior art, if it's recognized that those variables are what we've called result-effective variables, then even you don't necessarily have to prove an expectation of success for those particular claimed ranges. We can just recognize the fact that it would be obvious to ultimately land on those claimed ranges. And so there's a potential parallel to this situation. If we recognize that polymorph screening is something one would do and then therefore whatever comes out of that polymorph screen, why wouldn't that be obvious like a result-effective variable? I think a significant difference is in those cases, as I generally refer to them in the result variable, we're talking about a new characteristic of something that already exists, right? There's a new benefit. There's a new characteristic. We put it through a test. We found out that there is some new aspect of this. It's heat-resistant. It's whatever the result is of the test itself dealing with the characteristic of something that's already known. We're not dealing with the same thing here. I'm not sure that really captures the universe of what Judge Shen was talking about. You can have result-effective variables that will affect the creation of different things with different percentages of X or percentages of Y. So it can actually be a new thing, not just a… Maybe I'm just misunderstanding, but the invention, the creation of some brand new crystalline form, it's not an improvement. It's not a change. It's not something that is in the prior art. It is a new creation of a new crystal that is going through this variability, and there may be some predictable results if I understand the hypothetical. But the crystalline structure is a brand new crystalline structure. I don't think the court has case law that says when you create a brand new compound or a brand new crystalline structure… Okay, you're into your reserved time, but I'm going to let Judge Shen… Okay, we haven't explored very much the utility part of the argument here. The other side is saying form A, form B, there's no difference in terms of utility, and it's not a big deal because they behave precisely the same. So even if we get to the question of whether one is more stable than the other, it doesn't matter if one turns into the other at any point in time because they both have the exact same properties. It doesn't matter. They're interchangeable. Could you give us a little more on your final answer on utility? Sure. I guess I'd start with the first part of yours, which is actually itself important. The fact that form A turns out to be pharmacologically equivalent to form B is not just happenstance. It's important significance because the idea that we've discovered a new polymorph that has that same pharmacological activity and is stable is itself unexpected, and it is itself useful in the sense that we now can shift to form A, keep it stable at room temperature, and know that we're not giving up any of the pharmacological activity that we originally discovered. The fact that it has pharmacological activity, the fact that it is stable at room temperature, and as the court looked at example 25, we demonstrated that usefulness and utility. After all, this is a drug that the defendants anxiously want to copy and put on the market. It is useful in treating both nociceptive pain and polyneuropathic pain. Unfortunately, I think I may have run out of time, but I'd like to get to infringement maybe in my rebuttal if that would be all right with the court. I'd like to bring it up in your rebuttal. We'll restore you to four minutes on rebuttal time. Thank you, Your Honor. So who's next here? Is it Mr. Daufrey? Ms. Garrison. Thank you, Your Honors. I understand that— Put him back to three minutes. I understand that infringement might be addressed in the rebuttal. Would you like me to address those issues now, or is there a chance to respond? You better address them now. Okay. Well, the infringement analysis itself, what I anticipate he's going to say is that it's pure legal error, but it's not. It's factual findings straight up, and there's nothing that's anything different. Indeed, it's a making of their own case because the carve-out tracks their invalidity argument, and Dr. Brown, their expert, admitted that. And that was at 11.210, 11.208 of the record, that the anticipation argument tracks their third theory for inducement, which is that the meaning of pain is broad enough. And that, however, is supplemented by the district court's clear factual findings, and there's a whole host of them, that DPN is not covered by Indication 1, that CLBP's study is nociceptive or at most mononeuropathic. Those facts alone require affirmance. In addition to the extensive other findings, that the treatment of polyneuropathic pain is not on label, and therefore that the label itself would not encourage the treatment of polyneuropathic pain, that it cannot be reasonably understood to be an instruction to engage in an infringing act, and to the extent that doctors prescribe it, they would be motivated. By the way, you're making the argument at this point both for you and for – who was the loser on the inducement? One of the generics was a loser, right? That was for a different patent, Your Honor. That was the 130 patent. I'm sorry, that's the same patent, but we're not making that argument for ALCO. Okay. We have a different label. Right, I know, but are you – are they going to stand up and talk about the same point in – on – I don't believe they addressed – So are you – I mean, they adopted – We're speaking for Hickman and Westward. They adopted your brief? I believe they adopted it for the invalidity, I think. Their label is different, right? Yes, there's no carbide. Do you have anything to say about – is theirs the label – no, I'm sorry, you have two labels. You have two clients, right? I'm trying to get this straight. We have two clients, one label. One label? I thought one of them had this reference to the Janssen product. That was taken out, Your Honor. It's been taken out. It's been taken out. It's not an issue in the case. Because that was a kind of wink-wink, nod-nod reference in that label. That's gone? That is not an issue in the case. It's not an issue that's been appealed by the plaintiffs. Okay. It's no longer there. So that carve-out, those factual findings are not clear. They're plausible. They're based on the classic battle of the experts. But just going back to utility real quick, Judge Reinert, going back to your point about whether or not the district court found it more stable was useful. She did not. She found that there was a legal bar to even addressing that factual finding. But you heard my friend from Cross Appellants today state that you actually do need dissolution testing to determine whether or not what's the most, quote, stable, the form that exists at room temperature, whether or not it's useful. And that agrees with Dr. Bernstein's, their expert's testimony at 10-9-11 of the record where he said in order to determine what is most suitable, you can't just know what's the most stable form. You actually have to do the testing. And at minimum, we would need remand on that issue for her to make that finding. But at the end, just relying on the prior art for utility is not enough because even as this Court said in Juicy Whip, a case that is relied on by Cross Appellants, the invention of the product is what matters. The improvement is what matters. And the improvement here is that you have Form A instead of Form B. And the fact that Form A has the same properties as the Supreme Court said in granted a patent based on utility because you have to show why those properties, why existing at room temperature in Form A instead of Form B matters. And it's undisputed here that for pharmacological activity, it doesn't matter. They have the same. As their expert even stated, it doesn't matter whether you have Form A or Form B. So while they're trying to save their patent on obviousness by saying that it's unexpected and you need the dissolution testing, it wasn't there. And that's what makes this case unusual for a polymorph patent. As a polymorph patent, this is not the typical polymorph patent. Every polymorph patent I've ever seen that claims a new polymorph explains why it's useful, says it has a food effect, not just that it exists at room temperature. And at its heart, that's all the testing here does say, Form 16, that at room temperature, you have Form A. So what? If you had Form B, it would convert to Form A. That's a natural law. As the Supreme Court said in Funk Brothers, you have to take that discovery and apply it. And they haven't done that here to show that it's useful. And under Brenner, there is no utility to that because it's still just an object of research. They left it for someone else to discover. What about it's used to treat a different form of pain? A different form of pain, Your Honor? Yes. I'm not sure I understand. Okay. It's admitted by Grumethal in their argument is that it's the identical chemical structure. It's the same thing. It's just arranged differently. And it has the same identical pharmacological effect. That's what the patent says at column 4, lines 13 to 16, that it has the same. Grumethal says that at page 55 of their brief, that it has the identical same structure and analgesic effect, regardless if you have Form A or Form B. So then the question still comes back to, why does having Form A matter? What did they give to the public to assure them that they had a useful invention? And the answer is, what makes this case even more unusual, is that they had the dissolution testing. They had that testing. At Apex 9804 of the record, that testing was disclosed as DTX 1158, which can be found at 49209 of the record. They had the dissolution testing, but they didn't disclose it. So as the Supreme Court said in Brenner, this is not the case where they've disclosed something useful that is good for the public, that is part of the quid pro quo, but it's still an object of research. And it doesn't matter if you have Form A or Form B. Okay. Thank you. We'll put you back to your original minute. Thank you, Your Honors. As to the polymorph obviousness, the district court's error is essentially on pages Appendix 143 to 145. It's confusing reasonable expectation of success with predictability and requiring predictability, even though the court had already found the procedures are routine to do the polymorph screen and that a person who already has skin in the air would know how to do that screen. What case or cases would you cite as stating or holding, if any, do, that you can have a reasonable expectation of success as to something you cannot even identify in advance? Cuban would be the closest case, Your Honor, where there's a sequence for that and nobody knows the sequence ahead of time, but the key quote is the enabling methodology is disclosed and therefore combining, and that's an obvious to try case, combining the prior art which had the direction to find out sequences and had prior compound, but didn't know the result ahead of time. So we have one in which among the things you cannot even predict is whether the thing existed. That's not true in Cuban because somehow or other there are a bunch of codon sequences that are going to be the gene, right? Yes, Your Honor, and here, whether there's polymorphs or not, that's a question where it's answered in the 737 that there is at least a crystal form and in determining even how to characterize that form, which the FDA would require, that's the step that the polymorph would do. So at that point, while one does not know how many forms there are, that's the routineness of doing the polymorph screen to find out the result to say what is the polymorph landscape, and that's exactly what that Berne article was intended to address. So in conclusion, the distinction was between using the routine steps following the prior art versus what the court addressed, which was predictability, and that's also the expert's response specifically when asked about SSCI is that Appendix 10907 was asked specifically about SSCI as a reference, and if one followed that, does that still make it non-obviousness? And the expert's answer was it's not predictable. So the only distinction here that the judge and their experts relied on is not knowing the answer ahead of time, and it's not addressed in the opinion whether the pathway would be obvious specifically in view of Berne as opposed to the general landscape of polymorph patents. Thank you very much. Michael Joffrey for activists. Obviously, I'm the appellee, and we have not had the benefit of hearing Mr. Sitzman's argument, but I will address the issue of infringement. May it please the court, a fundamental flaw in DepoMed's cross-appeal is their assertion that polyneuropathic pain is an unlabeled use under the Indication Act of AFC-Little, and that's just not true. And this was an issue that was disputed at trial, and the district court... I'm sorry about this, but why does the label, on-label, off-label matter as opposed to whether physicians would understand something in a certain way and would understand that it is being encouraged? The touchstone is encouragement, Your Honor, that's true, but because this is an ANDA case, we don't have any benefit of knowing what doctors actually do or that there's any other type of encouragement of record. All that exists, because this is an ANDA case, is the ANDA filing itself, including the label. So if we're going to look for... By the way, is that troublesome little bit about a reference to Janssen, is that out of the case? Yes, that was an issue with Westward. That is out of the case, in my understanding. As to activists' label, it was never an issue. And in any event, the point being is the only thing to look at, the only evidence to look at, is what is in the label. In an activist's label, what happened was there was a specific carve-out of anything dealing with polyneuropathic pain. So there was nothing in the indication, and there was no testing within the label showing anything about polyneuropathic pain. And that matters a lot, because as DepoMed admits in its brief at 10, before 2007, and even continuing today, there was widespread controversy surrounding the use and efficacy of opioids for treating neuropathic pain. So if you have a doctor that's looking at a label on an opioid, and it doesn't say anything about polyneuropathic pain, they would naturally not think that it has any ability to treat polyneuropathic pain. It would solely be about... Are you saying opioids were not being used as a second line of treatment? I'm saying they were being used... You're not saying that. No, no, no, but I am saying... That they were being used. Yes, but there was, as they admit, a controversy about the use of opioids. And that continues even to this day about using it to treat polyneuropathic pain. So given that fact, given that general reticence in the art about using an opioid to treat polyneuropathic pain, in a label that's absolutely devoid of talking about polyneuropathic pain, a doctor would not read this label and say, oh, this must be for treating polyneuropathic pain. And so there's therefore no possibility of encouraging the infringing use. And in fact, Janssen knew that. So Janssen did not think when it originally got the indication for the branded product that this covered neuropathic pain. They originally, like activists, had a single indication, and that indication was for severe pain. Internal Janssen documents describe that the indication was not covering neuropathic pain. So Janssen then went out, spent $50 million on testing in order to get a second indication that specifically went to a type of neuropathic pain. That indication that is now on their label is not an activist's label. So even Janssen knew that polyneuropathic pain could not exist within the scope of activists' label. And because of that, there can't be any encouragement to infringe. Is there any difference between your label and your co-counsel's client's label? So our label, the answer is there's slight different wording, but I think that there's been no dispute that they're essentially co-extensive. Right. Westward, not Elkham. Indication 2 has been carved out. All references to a DPN trial study has been carved out. That's us and Westward. That would not be the same with Truth for Elkham. And to what extent would doctors appreciate the fact that these carve-outs occurred? Would they somehow understand in a larger context, okay, for the patent owner's label, they see and know that there's indication 1, there's indication 2, and underneath indication 1, there's a discussion of the DPN study to support indication 1. But now I have these ANDA people's labels and I see, ah, they've removed indication 2 and they've removed any discussion of the DPN study to support their indication 1. Would the doctors somehow see and read and understand the overall context of what's going on? Doctors would understand the context, but they would also see specifically what activists did. And the only thing activists did was include indication 1 in their label. They didn't include the studies or anything else. So the thing that you can't do is... I guess what I'm wondering is maybe would they understand, would they have a baseline understanding of what the other side's label actually says and does? The NDA. So there may be doctors that know that. The problem, of course, is that you can't combine what some doctor knows about something else and do some sort of literature hunt where you combine some other label with what activist does and then say, oh, from that we can infer that activist is doing an infringing conduct. In fact, what you have to look at is... No, my inference would be the converse of that. They'd be assuming, oh, activist is trying to run away from using their particular generic for polyneuropathic things. And we did. We specifically carved out and then we filed a section 8 in our NDA notice letter saying we don't infringe this. And so, therefore, altogether that would lead to the assumption that not only were we not acting to encourage infringement, we were doing everything we could to avoid infringement. Could the answer to the question of what your label is tends to suggest you intend to encourage change over time as the world of research changes? So that, for example, here, assume that nothing happens till 2025. We're talking here now about the difference between 25 and 28, right? 2025 and 28, just on that assumption. So by seven years from now, there's new research and everybody now understands that chronic form of pain covered by the only explicitly stated indication is understood by everybody to include polyneuropathic and the diabetic peripheral pain. If you kept the label this way, might you newly be accused of inducing? I think that the issue, though, is what is the act that is actually being performed? Marketing with this label. At some point, if marketing comes along, perhaps there would be some issue of intent that could clearly be inferred. For example, if contrary to law, activists were to market off-label uses, which they can't really do. It wouldn't happen for long. Theoretically, that kind of activity could be used to show an intent to actually encourage infringement despite the label. But here we have a label that was produced a little while ago, and you also have the mission. I guess the question is, legally, is the inquiry fixed as of what time? Isn't it as of the marketing, whether you will be inducing when you market? Well, there has to be an intent and an act of encouragement at a given time. There's only one action, and that one action was filing an ANDA at this point in time. Now, later on in time, there will be marketing, etc. But at this point in time, you're somewhat fixed because you only have an intent and one action, which is the filing of the ANDA, which would trigger 271. But you're going to sit there for seven years on my assumption and do nothing. And then in 2025, you're going to say, okay, there's only the one thing left, the 130, and now I'm going to market, assuming that. And at that point, why doesn't your intent count with all the knowledge existing in the world? Because then it becomes a pure speculation about what we might do in ten years, which seems both problematic just as a legal matter, but also it is also the problem is that there's no evidence in the record, obviously, of what we're going to do in ten years. So by sort of necessity, you have to be fixed at the time and look at the only action that activists took. And the only action is the filing of the ANDA. Thank you. Before I jump right into infringement, I wanted to respond to Judge Taranto. You asked for a case site. I wanted to give you one. On the 364, Pfizer versus Teva 555, Fed Appendix 961 at 970. Activists and Westward's labels induce infringement of the 130 patent. All the experts agreed in this case, all three medical experts in this case, agreed that the indicated use on defendant's proposed labels includes polyneuropathic pain. Counsel did not tell you that the medical expert that they hired agreed with all of the other medical experts that indication one that's on their label includes polyneuropathic pain. In fact, as the district court found, 95% of the time will be used for the treatment of polyneuropathic pain. Now, that evidence was undisputed in the record. The instruction here appears in the indication section, which is where FDA requires parties to put down their intended use. This has induced infringement. The district court committed legal error when it used non-infringing uses to negate the finding of intent and induced infringement. And that's contrary to this court's precedent. On at least three occasions in the last 18... Are we talking 271B or 271C? 271B. Okay. On at least three different occasions in the last 18 months, this court has said using non-infringing uses or substantial non-infringing uses to negate a finding of inducement is improper. As the court said in Sanofi, there's no legal or logical basis for the suggested limitation on inducement. Section 271B on inducement does not contain substantial non-infringing use restriction of 271C on contributory infringement. I'm going to assume, maybe hope, but assume that Sanofi doesn't say that that evidence is irrelevant, only that it's not required. Why would it not in the real world be relevant to assessing intent as manifested in some sort of encouragement, whether there actually is anything else that anybody could use the thing for? Of course it's relevant. Well, it may be relevant, but... Well, that's the question, right? The district judge made a factual finding about intent as manifested in words of encouragement. But the district court cannot use that finding to negate the induced infringement as per AstraZeneca. This court said that as long as some users will be instructed to use this in an infringing manner, that non-infringing uses could not or would not negate the finding of induced infringement. It was, in fact, irrelevant in that case whether or not there were non-infringing uses on label or off label. What do you do about the significance of the sequenced way in which, was it Janssen, I guess, went to the FDA and said, and in fact got approval for what extended release in 2011, was it, and then later extended release in 2012? I may have the dates wrong. First with indication one, second with indication two. Doesn't that tell you that the world of providers would not automatically assume that indication one is an encouragement of indication two? Otherwise, what was the need? Well, as the record reflects, the need was a marketing one by FDA. FDA required that second DPN study in order to say that this has been approved for DPN. But the record is unequivocal on what doctors will do. We had three medical experts, including the experts from the defendants, who testified very clearly that indication one includes polyneuropathic pain. There was no equivocation on that. And the district court made a factual finding that some users will infringe. The problem is she committed legal error when the district court used non-infringing alternatives, non-infringing uses, to take that away. And that is what was improper. That's the error here. Unless the court has any further questions, I'll leave it at that. Thank you. Thank all the parties for their arguments.