Oral argument — Sanofi-Aventis Deutschland v. Mylan Pharmaceuticals Inc.

A remand is required here at a minimum of a PTAB to correct a critical omission in its final written decision. The PTAB failed to find, as required, that a glargine aggregation problem would have motivated an artisan to modify the FDA-approved and stability-tested glargine formulation by adding a surfactant. And the reason that that failure was critical here is because an artisan would have also understood that glargine depends, indeed requires, a certain amount of aggregation to work, that is, to provide its long-acting effect. Instead of making findings about glargine aggregation or a glargine aggregation problem, the PTAB made findings about aggregation involving naturally-occurring insolence that, unlike glargine, do not depend on beneficial aggregation to do their job. So ultimately, the PTAB never addressed the critical question here about whether an artisan would have added a surfactant to address a glargine aggregation problem, and that's what requires a remand here. And it's clear from this Court's cases and from the precedent from KSR on through Nuvasiv that this Court decided, and Magnum Oil and others, that the PTAB analysis must be thorough in searching, that there's a need for specificity in the PTAB's findings, that it cannot simply make a conclusory analysis, it cannot merely summarize the parties' arguments and reject them without explaining why. But that's exactly what the PTAB did here with respect to this issue. And so... So tell me, I guess, tell me why the following is wrong as a characterization of what the Board did. Sure. That the Board looked at the evidence, including some passages in the specification, but not by any means only that, and said that evidence would have reasonably been read by relevant artisans to say, as a general matter, there's an aggregation risk with insolence. Sanofi came back and said, has evidence that tries to persuade us that glargine is really an exception to that general recognition of a risk, we're more persuaded by the opposite evidence that we don't think that glargine would have been understood by relevant artisans to be outside the category of aggregation risk insolence. And then pretty much similar on the surfactant is a recognized potential solution to that problem. So I have two responses to that, Judge Toronto. The first is, if you look at the PTAB's actual decision at pages 38 and 39, that's where they consider or they purport to consider the issue I just described, which is the failure of an artisan's understanding that a surfactant could interfere with glargine's unique mechanism of action that sets it apart from all the other insolence that it looked at. And if you look at those pages, you don't see that answer there because they didn't address that question. They didn't ever answer the question, why an artisan knowing that one of the key differentiations between glargine and other insolence is the fact that it does precipitate upon injection and that needs that aggregation in order to work, why that would dissuade or would have told an artisan that a surfactant should not be added. They simply didn't make that finding or draw that conclusion. Their analysis, to the extent there was any, was that because surfactants had worked with other insolence that would also work with glargine, but that doesn't answer the question about whether an artisan would have drawn that answer knowing the differences that we pointed out relating to glargine. And the second more general response I would have to that is that the PTAB's own findings in this regard with respect to the propensities of insolence to aggregate was that they're all of a matter of degree in the end, that an insulin of different types aggregates at different degrees. They made that finding. They also made a finding that the container matters, right? That the insolence aggregate more readily in pumps than in other devices, but they never made the finding. And this is, again, the critical point that the amounts or degree of aggregation in glargine would have been sufficient to motivate an artisan to modify the glargine formulation, particularly given all of the other evidence that we know that an artisan would have understood about glargine at the time. Number one, that, like I mentioned, that it depends on aggregation upon injection to work, but at the same time, the Lantus label tells you that it's clear and colorless in its solution. Secondly, they would have known that of all the other vial formulations of insulin on the market, none of them contained a surfactant to address an aggregation problem. They would have known from the prior art that Mylan itself cited that insulin formulations in vials were uniformly stable and didn't require additional surfactants or anything else to stabilize them. I thought the motivation was the idea that it was known that other insulins had a problem with some kind of aggregation, that when you store them in glass vials, that those hydrophobic surfaces have a tendency to induce aggregation, same with the air space that's in the vials, that when your insulin is acidic, there's a higher tendency to have aggregation, and when your insulin is more monomeric, then you're also going to be having a higher It's in a glass vial with some air space. It's acidic. It's got monomers running around in it. So therefore, when you put all that together, when knowing that even other insulins at the neutral pH have had problems with aggregation, and then the Lantus label says, definitely don't take this if it looks cloudy to you. There's enough of a collective concern through all those factors for the board to find, yeah, one would be motivated to try to address potential concerns that glaragine, just like all these other insulins, and we don't know of any other insulins that don't have any susceptibility to aggregation, we'd want to try to address this concern with glaragine. So my response to that, and I think there's several pieces of that evidence that I think, for example, the cloudiness issue, the board didn't end up relying on that. So that's not really a basis that this court can use to affirm. But with respect to the remainder of it, I think, again, the question is, what would an artisan have known at the time? And as to many of those factors, we showed that, for example, with the delivery device, there was no other insulin sold in a vial that had surfactant or addressed an aggregation problem. So again, it's a question of degree. With respect to the acidic pH, we showed that, yeah, it was designed to be acidic on purpose, because that allowed for glaragine to aggregate once it was injected. But there was something in the prior art that suggested that acidic insulins have a higher tendency to aggregate. That's correct, but that was about prior art insulins and naturally occurring insulins, not glaragine, which was designed fundamentally differently and was designed to be both clear in its solution and not aggregate in solution, and to aggregate upon injection. And that's what an artisan would have understood about the invention of glaragine at the time. But the PTAB never really found that, regardless of whether there was any propensity or tendency to aggregate based on all the factors that you mentioned, the PTAB understood it was a matter of degree, but never found that that degree would have motivated an artisan to modify glaragine. In particular, given its unique mechanism of action in which a surfactant was thought to disrupt and interrupt that mechanism. And that's a key finding that the board simply didn't make. But what do you make of the sentence right at the top of column three that says, especially, and this is right after, this is a paragraph talking about glaragine, right? Especially at acidic, because it's just said glaragine gets injected in an acidic solution. Especially at acidic pH, insulins, however, show a decreased stability and an increased proneness to aggregation on thermal and physical mechanical stress, which can make itself felt in the form of turbidity and precipitation particle formation. I'm citing something 1997, which is before the priority date here. So two responses. Number one, the article that's cited here doesn't concern glaragine, it concerns bovine insulin, if I'm not mistaken. Right, but the specification is invoking this in talking about a problem that, in talking about insulin glaragine, having just said, we've had some good luck with doing this because of glaragine's properties with using an acidic solution. But now there's this other problem. And one of the ways we know there's this other problem is a 1997 article that says, mix insulin and acids, not so good. Sure. And so my understanding of this is the inventors are describing the problem that they've identified and are solving. And that's the reason why they've connected here in the specification, the acidic... You don't think this portion of the background section is reporting out what this prior art reference branch at all is disclosing? Well, I agree with that it's disclosing what branch is disclosing, or brongae. Brongae, okay. But I've made that mistake many times. That's fine. But again, it wasn't disclosing anything about glaragine. So what the inventors here are doing is linking a disclosure in the prior art relating to other insulins and to glaragine, which has never been done before in any of the prior art cited anywhere in this matter. And so that's what the contributions here and in the next paragraph of the spec are doing. They're the inventors invoking their own knowledge and their own contributions to connect what the prior art said might happen with certain other insulins to glaragine when that was an unexpected result based on all... There seems to be evidence in the record that there were other insulins that had trouble with aggregation. And then there were certain factors that were known to cause those aggregation problems. Is there evidence in the record that there were still other insulins that never had any problem with aggregation? And so, therefore, what we've got on our hands is kind of a mixed record. Okay, we've got some insulins that have an aggregation problem and that was known. And so there's perhaps reason to believe that glaragine as an insulin would also share that aggregation problem. But no, there's actually a lot of other insulins out there that have never had any reported aggregation problem. So my response to that is that the market of insulins that were sold in a vial like glaragine did, none of them had esterfactant to address any problem with aggregation. And so an artisan would have been especially reluctant to add esterfactant to glaragine, especially given what it knew about what glaragine needed in terms of the desirable aggregation to work. And so I don't know the answer that whether there was anything that definitively said there was no aggregation problem, but an artisan looking at the available market said, none of these other vial insulins have done anything about an aggregation problem to add anything that's necessary. And so an artisan would be especially reluctant to modify an already approved, already stability tested formulation to add something additional when it thought that that something additional could be a counterindication for the actual operation of the molecule. I see I'm into my rebuttal time. Let's hear from the other side. We'll save your rebuttal. Mr. Karsten. Thank you, Your Honor. Good morning and may it please the court. What we have here, well, let me first answer the question that my learned friend didn't answer for you, Judge Chen, and that is there is no evidence in the record of an example of an insulin that has no aggregation. Certainly, if there were such evidence in the record, my friend would have been able to point to a chapter and verse readily and was unable to. What Sanofi's position here is, is they don't think that the board below paid enough attention to the fact that they think that glargine is special, that glargine is different. But that's a factual determination. And there was ample record evidence presented below that led to three, I think, fundamental and critical factual findings below that are supported by more than a mere scintilla of evidence. And I think Judge Chen and Judge Strato, both of you identified some of those factual findings in the context of your questions. You're not relying on the proposition that anything that is more than a mere scintilla suffices, are you? No, I think there's certainly all of these. I think that's the legal standard through which... I'm not sure that is the legal standard. Substantial evidence first has to be more than a scintilla, and then it also has to reasonably support. So the scintilla thing is doing no work. It has to reasonably support the conviction. Fair enough. And I think that, Judge Strato, we do have more than ample evidence which absolutely supports each of these critical factual findings. The first is at the final written decision at page 32, where the panel, the board below says, the record supports that an ordinarily skilled artisan would not have suspected insulin glargine to behave differently than other insulins due to differences in amino acids between them when exposed to hydrophobic surfaces. And that's on page... That's at page 32, APPX 32 and page 32 of the final written decision. As it pertains to one of the two, they are essentially identical, but for these two patents. So this notion that the board didn't even consider this idea or this factual presentation by Sanofi that insulin glargine is special is just incorrect. They just disagree with what the board did. Secondly, the board at the final written decision on page 36 went further and said, we find that a person of ordinary skill in the art would have had an additional reason to be concerned about aggregation in the insulin glargine formulations that the Lantus label and Owens disclosed. This is again, final written decision, page 36. They did consider insulin glargine specifically and rendered a factual determination supported by ample record evidence to say, no, no, no, it's not different. And then furthermore, if you look at page 35 of the final written decision, APPX 35, the board considered that to the extent that glargine was different, that that difference actually led to an enhanced risk of potential aggregation. This is on the Jones reference. They say, rather, we consider Jones for what it would have taught the ordinary artisan, that insulin glargine is more monomeric than other insulin preparations. And your honor, Judge Chen, you're absolutely right to point out that that's one of the risk factors identified by the Brange or Brange reference that we talk about. Those two Brange references are exhibits 1014 and 1015. They appear in the record. It's the first page of 6760 and 6797. The Brange 97 reference that that specification reference, column three, pointed to is the Brange reference, exhibit 1015, which starts at page 6797. While these Brange references don't identify glargine specifically, they do talk about the risk factors that relate to aggregation of insulins generally. And if you look further on column three of the specification, at pages, at column three, I think it's lines 32 through 37, and this would be for the 652 patent at APPX1006, you'll see that Sanofi themselves, in the context here, didn't distinguish glargine at all. They said, the present invention was thus based on the object of finding preparations for acid-soluble insulins. Insulins. Not, this is something magic and specific to insulin glargine. They themselves were lumping glargine in together. And if you don't want to look at the specification to render that conclusion, you need look no further than the Lantus label itself, which is one of the primary references that Mylan relied upon below. Did the board rely on it? The board did. I think Mr. Banks said, I think in response to one of Judge Chen's questions, that the Lantus label warned patients not to use vials in which the liquid was cloudy, which makes clear that there's some recognition of a potential for cloudiness, which I assume everybody agrees in this context equates to aggregation. I believe that's correct. But I think he said the board did not rely on that. The board, at various points in its decision, says, based upon the record as a whole, in its analysis. And in the front piece of the board's decision, and here I'm at APPX 19, in which they're describing the import of the various prior art materials that were cited. And this is bridging pages 18 to 19 about the Lantus label itself. Lantus was originally on the market in a formulation without a surfactant for some period of time. And that label acts as prior art here. At page 19, Lantus label instructs that Lantus, quote, must only be used if the solution is clear and colorless with no particles visible, close quotes. Can I ask you, is there something in the record that answers the following question? And if not, what do you know about it? Is this an absolutely standard thing to put on a label of anything that's being marketed in a vial so that it doesn't actually imply recognition of a problem? I don't know the answer to that question directly. I've not scoured the art to see if each and every vial pharmaceutical product contains that. I know that certainly for protein products and for insulins, which are injectables, it is common to have an instruction not to inject anything into your body that contains particulate matter for obvious reasons. You don't want to inject anything particulate into your body, into your bloodstream, et cetera, that could cause deleterious effects. Having said that, though, the presence here recognizes that there is a potential for aggregation, that there is a potential to have particulate matter develop in a vial of this particular formulation. That's why I guess I asked the question. It seems to me that if this is standard, maybe not universal, but close to universal or standard, then maybe it doesn't actually reflect a recognition of the potential. It's just that you always say that, including on labels of vials where there is no possibility because who knows? Well, I think there's no record evidence that no insulins avoid aggregation altogether. And so I would submit it's a different issue. It's not that this particular vial doesn't have a risk of aggregation because it's just like all the others. I put it differently, and that is this vial does have a risk of aggregation, and so you need to check it just like all the others. I mean, I think that Sanofi has said somewhere here that it's one in 10,000, one in I don't know where exactly the record evidence of this. But sure, maybe it's a minor problem. But that's something about when it started getting reports from patients and was quite surprised about this. They claim they were quite surprised about this. Of course, the board's findings below, the factual findings which are entitled to deference, suggest and tell a very different story. And moreover, to the extent that we saw in Sanofi's reply, et cetera, that we didn't identify or the board didn't identify a motivation, you don't need to have the same motivation to solve the problem that the inventors had. Maybe the inventors were motivated by having a particular vial or a particular report past their desk. Whereas here, we have a generalized understanding by people of skill in the art that this is a risk factor. And in fact, there's fact findings by the board that this risk factor, which applies to insulins generally, specifically implies in an enhanced way here. But you also have a clear statement of enhanced stability with temperature and time and so on with this particular surfactant. And isn't that really the problem to be resolved in this case? I don't believe so. I don't believe that there is this enhanced stability presentation here, Your Honor. Well, the record, the specification, the patents provide data. Are you saying that that's flawed? Well, I'm not quibbling with the data. I'm making an obvious analysis. And so I am not looking at the particularized data to see if the claimed invention solved the problem. Certainly, I think that there's record evidence. We had Professor Langer provide 20 examples of insulins being stabilized by an addition of a surfactant. I think you're reciting at some length the problems that were being experienced with this particular insulin. Well, I agree. I think the question is one of obviousness. At what stage, other than with hindsight, do we say, well, just put in some polysorbate and it's over? Oh, I think there was absolutely no hindsight here at all. And the factual findings of the board below look specifically at primary references. Those two Branga articles, for example, the Lantus label itself. And they all identify, as well as another Sanofi product in a pump, which required the inclusion of a surfactant. And the board made a specific factual finding about that, crediting Mylan's experts over Sanofi's experts here, saying that it was the exposure to the hydrophobic surface that was the cause of concern for aggregation. It was not the particular delivery system. So if this product doesn't solve any special problem, or any old surfactant would do, what is your client's interest? All this pressure to be able to affect the patent structure for this product. So there are two. There's a product on the market. And under FDA guidance and analysis, we are required to essentially duplicate the formulation that is the brand product here. And so we are required to include a surfactant in our formulation. And I'm not quibbling in saying that the surfactant does nothing to stabilize the formulation. In fact, I think the reasonable expectation of success to an artisan is that it would enhance the stability of the formulation. But it's a business decision, ultimately, whether a person of skill in the art decides to add the surfactant or not. In a situation where you've got a problem. A business decision, if it's known or demonstrated, I don't think anyone's quibbled that adding the surfactant, adding the polysorbate has enhanced the stability to the temperature and storage. I believe that's correct, Your Honor. I believe that there is a benefit to adding the surfactant. And that's one of the reasons, among the others, including the FDA requirements here, that my client wants to provide a product. What can you tell me about the timing of the FDA requirements? At what point did the FDA say a Glargine product, the Lantus product, now needs to include a surfactant? So the FDA never, as far as I can tell, based on the record before us, the FDA never said you have to. My understanding is, and perhaps my learned friend representing Sanofi can provide further color to the extent it's not in the record. There were reports that were made to FDA and or Sanofi about some amount of aggregation. They went ahead and reformulated to identify if they could solve the problem of aggregation by including a surfactant. They sought approval with the FDA. The FDA granted approval. And then a year later, after Sanofi had continued selling the old non-reformulated, non-surfactant containing material for over a year after the approval, only then was there a switch to the reformulated surfactant containing product. So I don't know that it was an FDA crackdown that said, no, no, no, you have to do this. Do you remember even roughly what years you're talking about? So I believe that the— The original Lantus came on the market, I think, 2001. Yes. And then what? I believe that the reformulated product was approved in 2004 or 2005, and it entered the market in 2005 or 2006. It was about a year—I remember the year delta between the approval and the introduction to sell material. So, Your Honor, Judge Newman, I do agree that I think that the addition of the surfactant was beneficial. And I think that the evidence of record, the 20-odd references that Professor Langer identified, suggest and make a reasonable expectation that you're going to succeed here. A final point in that, there's this tension or notion, and I've heard Sanofi sort of say, well, there's desired aggregation and non-desired aggregation. Those are very different things, and those are explained pretty clearly and understood by a person of ordinary skill in art. In the vial versus under the skin. Exactly. And that's understood. And the formation of hexamers, the proper—the desirable sort of agglomeration, so the desired form of insulins is a hexameric structure, and that's what you want. It's when these insulin molecules are in monomeric form and can expose a little bit of a hydrophobic surface that they tend to agglomerate and just sort of crash out of solution in an irreversible, non-reversible way. Those are very different mechanisms. Very different mechanisms. I see that I'm nearly out of my time. I understand that one of their—the patent owner's arguments—I'll lean forward here—is that your expert said something about how a surfactant can inhibit aggregation from monomer to hexamer, and it'll act as, I don't know, some instrument of enhancing absorption. And so, therefore, the argument goes that, well, this therefore is some kind of proof that the introduction of a surfactant will actually undo the desired native agglomeration once you inject the monomeric solution into the human tissue. So, what's your answer to that? So, the panel did consider that argument. They recognized it at page 31. They say, among other things, and then they identify a string of things that they use air quotes around the wood could in characterizing some of Sanofi's arguments about why a person of ordinary skill in the arts— It seemed to me the board just summed it up by saying, well, you don't need absolute predictability in an obviousness inquiry. You just need a reasonable expectation of success. And so, therefore, some unpredictability or uncertainty doesn't defeat a reasonable expectation of success. Well, I think that's true, and I think also the board did go ahead and analyze. So, this was essentially Sanofi had presented these arguments, and the board was taking them seriously and considered them all. And for several of those arguments, concluded a significant analysis, including one about potential interference of a surfactant with another additive, metacresol, and found, as a matter of—and found in there, I think, page 38 or 39 of the final written decision that there was no merit to that argument. And then went ahead and issued a conclusory paragraph at, I think, paragraph 39—page 39—saying, we find unpersuasive patent owner's arguments that an ordinary skilled artisan would not have reasonably expected success. So, they did absolutely consider those arguments. In addition to that, Your Honor, I think there is record evidence that suggests and supports the notion that—or the idea that the existence of a surfactant doesn't interfere with formation of hexamers generally in connection with other insulins, not specific to glargine, but with insulins generally. And I think I would cite the board to—the panel to—the article Thoreau 6906 is the appendix site, and there's a pin site at page 6910 in the third full paragraph. It says, if I may read it, since the genopole PF10 content of a neutral insulin solution does not change after removing the insulin by precipitation with zinc ions, we assume there is no significant interaction between insulin and genopole. The genopole is a poloxamer surfactant, and what that describes is they use zinc to form these hexamers of insulin. This is the desired pathway, not the undesirable pathway of aggregation that's irreversible and unwanted, which the surfactant is designed to protect against. And they knew how much surfactant they had in the solution. They precipitated out the insulin in hexameric form, and they found that the same amount of surfactant was still in the solution above. Therefore, they concluded that there's no interference with the surfactant. I am well over time. I apologize. No. You've been able to avoid the blinking light. I've gotten the glaring eyes instead, so… We need to hear what you have to say. Have you any more questions? Any more questions? Thank you very much, Your Honors. I appreciate it. Thank you, Your Honors. I have three quick points I'd just like to make on rebuttal, just picking up where we left off with the desirable versus undesirable aggregation. I think, Judge Chen, you hit it exactly right, which was that Mylan's own expert has testimony that surfactant would have been expected to interfere with the native formation of the hexamers that are vital to the glaragene's mechanism of action once it's injected to form the precipitates necessary to give it its long-acting effect. And again, I think if you look… My understanding of the way the glaragene precipitates once it's injected, it's because the glaragene, while it's sitting in a vial in a solution, it's acidic. But then once you inject it into a human, the tissue is pH-neutral at a 7. And then glaragene, as we know, doesn't handle itself very well at a neutral pH. It ends up becoming unstable, and then it precipitates like that. It precipitates like that, but it's hexamer formations. So, I guess what I'm just trying to figure out, whether the statement that you got out of the other science expert, whether that's in the right context of what's going on in terms of the mechanism of action of glaragene, which is going from an acidic pH solution to a neutral pH human. So, Your Honor, I think that both parties agree on how the mechanism works. And if you look at 6509, for example, in the appendix, that's Dr. Yalkowski, their expert's description of how it works. One of the ways it works is that it does precipitate into these hexamer forms, and that these slowly dissociate over time and then are absorbed into the bloodstream to deliver a long-acting effect. The testimony he gave at his deposition, which he was summarizing an article that he had written, says that a surfactant speeds up that absorption process. So, it inhibits the formation of the aggregates, and it increases their absorption into the bloodstream, both of which would counteract the long-acting intention of the glaragene. Your expert, Dr. Trout, he seemed a little equivocal about the possibilities of a surfactant affecting the so-called native aggregation, agglomeration. He said, well, it could. And so, it sounded somewhat speculative when you actually read the text of his declaration. So, I think this comes into hindsight issue in some ways, because we know that it actually, it works fine in the solution, or now it does, right? So, the question is, what would an artisan have thought at the time? And I think that's the challenge. And we don't have to show, obviously, that it wouldn't have worked, because we know that it does. The challenge is to show there's some likelihood, some reasonable likelihood that it would. And that's what we showed. And I'd like to direct, again, the court's attention to what the board actually said about this. And this is on 39. And I think my colleague described it as conclusory. And that's exactly our point. Because the way that they answered this particular point was to say, because surfactants work for everything else, they work for glaragene. And that simply doesn't contend with the question that was presented here, about whether the mechanism of action would be disrupted based on the surfactant being added. If I may just address a couple of more points that came up in my colleague's argument. As to the warning label, there is evidence that we put into the record saying that the use only when cloudy is a very common thing and addresses other types of problems, including microbial contamination and the like. And that's 14-286. And then if you look at appendix 726, that is the board's institution decision where they pretty much agree with us. What page was that? Sorry, it's 726. And that's the board's institution decision where they essentially agree that Pattenruder's argument regarding Lantus Label's patient warning has merit. And they go on to talk about other evidence. But I think it's a good indication of why they didn't rely on it the second time around. And the passage my colleague read was merely a recitation of what the label says and not really the evidence that they were relying on. And that brings me to the third point I'd like to briefly touch on was my colleague brought up Dr. Langer's declaration and the additional references that he cited. And we have this in our brief. But just to make the point, those were references that Mylan introduced for the first time in connection with its reply brief that we weren't allowed to respond to with evidence. And in fact, the board used that declaration and those references to discredit our expert without giving our expert or Santa Fe an opportunity to respond to it. That was improper. That was a matter of fairness. We should have had the opportunity to put on evidence. Is this the case where the patent owner filed two SIR replies? We did, but they weren't on this issue, Your Honor. The board did not allow us to apply an argument or evidence or anything of the sort on these particular issues or to respond to this particular evidence. We should have had the opportunity to do that. This was evidence that Mylan could have and should have introduced with its petition. It introduced it for the first time on reply. And it was prejudicial because we know from the board's own language, and this is at page 33 of the decision, they used Dr. Langer to discredit our own expert who was replying to their first expert and didn't get a chance to reply to Dr. Langer. And in fact, I think it's quite telling that the tribunal or the panel used the new expert to discredit our expert instead of Mylan's first expert. I see I'm out of time. So, if the panel has any more questions. Anything else, Mr. Banks? Thank you. Thank you both. The case is taken under surveillance.