# Oral argument — InreFosamax(Alendronate Sodium)Products Liability Litigation

> Briefs, arguments, decisions, and more.

URL: https://www.frixlaw.com/law-library/documents/argument%3A90985

## Record

- **Collection:** Oral argument
- **Document type:** Oral argument
- **Published:** March 5, 2024

## Text

Alright, we'll call our next case, InreFosamax Products Liability Litigation. Thank you Judge Jordan, and may it please the Court, David Frederick for the appellant. I'd like to reserve four minutes of time for rebuttal, please. That's done. I think it's fair to say we're not going to be cutting you off at 11 minutes, Mr. Frederick. Thank you, Your Honor. There's a lot to talk about here. I'd like to start with the Supreme Court, which in this case reaffirmed three legal principles to govern a remand. First, the Court held that the manufacturer is responsible for its label at all times. Second, the Court held that for the FDA's actions to preempt state law under the Supremacy Clause, it has to act under congressionally authorized directives with the final force of law. And third, because of the changes being affected regulation, the manufacturer can easily change its label prior to FDA review. Well, maybe not so easily, but we'll get into all that, all right? But let's start with the standard of review. You handled that in a paragraph in your opening brief and more fulsomely in your reply. And I guess what I'm wondering is, do you agree with Merck that if that standard of review is a clear error, you lose? No. We think that all of the facts that they identify can be reviewed for clear error because it's speculation by the District Court. But let me go back to what this Court said in Avandia, which is that the standard of review after Albrecht is plenary, which I understand is this Court's articulation of what in other circuits would be called de novo review. Every circuit after Albrecht has treated a preemption question in these circumstances as a de novo question for review by the reviewing court. How could – well, first, it's true that Avandia came a few months after Teva and we did cite it, but did we actually wrestle with this question and have to deal with any subsidiary fact issues there? No, but what the Court in Albrecht had done in its very last paragraph is to say that the error of the Third Circuit the first time this case was on appeal was to treat these questions as fact questions rather than questions of law. And so it would be odd to suppose – Did it say that or did it say it's for the jury – or it's for the judge, not the jury? That's a key point, right? Because the analogy that Teva brings up and Albrecht Supreme Court cites to Teva, one would think they did it for a reason. Teva is a claim construction case and the whole point of claim construction after Markman is to say that this is for the judge to decide, even if there are subsidiary factual issues. It doesn't matter, as the Supreme Court's put it in Markman, it's a mongrel practice. This is a mixed question of law and fact and there could be fact stuff in there, but we think the bench, the judge, is better positioned to do it than a jury. Why doesn't that map pretty precisely, especially given the Albrecht Court's citation to Teva on this analysis? So the last paragraph of Part 3 of the majority opinion says, we consider these factual questions to be subsumed within an already tightly circumscribed legal analysis. And then in the charging paragraph, Paragraph Part 4, the court says, because the Court of Appeals treated the preemption question as one of fact, not law, and because it didn't have the opportunity to fully consider the legal analysis, we're remanding the case to that court. I think the analogy to Patton, if I could push back just a moment, because we are talking about a preemption question that implicates the supremacy clause, we are not giving deference to district court analyses of what federal agencies do without the full review that plenary review or de novo review would institute. And that's why, Judge Jordan, every court has looked at this with the same Teva analogy and has applied de novo review. What's the point of talking about? Help me understand this, Mr. Frederick. What's the logical reason? What reason would there be to cite to Teva, which is all about Rule 5286? There's nothing else going on in Teva except the Rule 5286 analysis and an assertion strongly asserted to the federal circuit, you will give deference, clear error review, to subsidiary fact questions and claim construction, which we know has fact and law questions. What's the point of citing Teva if it's not to acknowledge, yeah, this is overall a question for the bench, but there can be fact questions in there, and if you hit them, you're under 5286. Can you give me a reason why they would cite to Teva if not to send that message? Yes. In the Patton context, it is common for claims to be construed with reference to industry practices that would be extrinsic evidence. We're talking about agency review of a drug label and a drug application. We are talking about the evidence that is supplied directly to the agency, and for judges to evaluate that evidence, certainly there are factual questions here. The question that I think is the predominant factual question is, is a stress fracture an atypical femoral fracture? But the point of the preemption analysis is for a reviewing court to look at those questions, in the appropriate context, without putting a thumb on the scale, like clear error review does for fact findings by a district court. That's telling me that Teva's got nothing to do with it, which is what you actually say in your briefing. You say that's restricted to the Patton context, which makes me wonder, what were they so confused about at the Supreme Court? Because they knew this wasn't a Patton case. Why did they bother talking about it if they weren't trying to send a message about the standard of review? I don't think they were trying to send a question about standard of review, or if they did, every other court of appeals, and there are four of them I think by this point, have not included this court. What's the point? Why talk about Teva, which is only about the standard of review? There's nothing else going on in Teva except 52A6. Why talk about it? Why would Albrecht cite it at all if it were not to give us some idea about standard of review? Because what Teva was really about in this context was having the judge decide and not the jury, and here we fought the fight over who's going to decide these questions, like is a stress fracture an atypical femoral fracture, and the parties have now essentially flipped sides. We thought it was a mixed question of fact in law the first time this appeal came before this court. But I think, Judge Jordan, the important... It's odd because I would have thought if it was really just about the judge not the jury, they would have cited to Markman, because that's what Markman said. But Teva was not about judge not jury. It was about 52A6. But let's leave that to the side for a minute. Judge Jordan, I think that if you were to say that you think this implies a different standard of review, you would be creating an intra-circuit conflict with Avandia and an inter-circuit conflict with, I think, the 7th, the 4th, and the 10th circuits, which have said this is a de novo standard. I understand that Avandia says we're giving this de novo review. And they say it comes up on summary judgment, we're giving a de novo review. But what Avandia actually does is point to some very specific documents in the record, primarily the letter, the complete response letter, and that talks about deficiencies in the information that was provided by the company to the FDA and how they needed to remedy that. So on one hand, it could have been applying clear error review by saying this document, this CRL, is not consistent with the district court's factual findings. And, Judge Freeman, that's why I answered Judge Jordan the question if we lose. No, we don't lose, because I think you can read the complete response letter, which is at A1, 1, 5, 2, and 5, 3. And I don't think you can come to a conclusion that it is final agency action with the force of law that would displace state law under the Constitution. The FDA says on at least four occasions we invite further discussion, we invite a resubmission, we're ready to have a meeting with the Merck officials to talk about this. The Merck officials understood the message. They're debating internally about, well, should we have the meeting or not? At one point, they tentatively request a meeting, and then they decide not to have a meeting with the FDA. And the point is that I just don't think that you can read the complete response letter as anything other than one phase of a dialogue between FDA and Merck. And for the district court to have treated it as a preemptive action I think is wrong as a matter of law. So in your reply brief when you're dealing with the standard of review and you say, quote, Merck seeks to extend TEVA to preemption, but the district court relied on intrinsic evidence, close quote, it looks like you're actually applying a TEVA kind of analysis when you say the district court actually looked to intrinsic evidence. That's the TEVA model, right? Judge Fortin, I think we went under both approaches. I'm trying to be consistent with this court's president in Avandia to say it's de novo review. If you were to say, though, that in these contexts where you're talking about materials that the company is submitting to the FDA about these kinds of fractures, and these atypical femoral fractures are their hideously painful and difficult circumstances for the women who suffered these fractures, they are very different from stress fractures. And there is a whole series of information that is supplied to FDA to provide information about that. And so, Judge Fortin, if you were to say that the guide from TEVA is what you've just indicated, I think that the conclusion you draw is that the materials here are all intrinsic to the application to change the label under the PAS. Does that mean we should punt on the standard of review because it doesn't matter in this case? You could do that, but you could also side Avandia where the court said that it is plenary review. And in Avandia, I want to remind the court that it was dealing with an analogous circumstance where there was a approved letter, which was the regulatory precursor to the complete response letter, where the FDA said, your justifications are inadequate. We want you to submit more data. But that's not exactly what happened up here. They said your justifications are inadequate. In Avandia, they said the information provided was inadequate. But it was information, Judge Freeman, to support a particular label change. And that's the key point of parallel between Avandia and here. Whether you call it justifications or data or information, the whole point was that the label is inadequate under state law, and it is not preempted where the agency says, your justifications are inadequate. Please try again. Tell me if you agree with this. It seems to me that the issue in Avandia, there was a big issue with whether complete information was available to the FDA. And there was that particular scientist who had done, who's Nissan, and the FDA said, I need you to address the Nissan study and other studies. Without that, this information is not enough. So that's Merck-Pronghorn, right? Do we have anything similar on Merck-Pronghorn 1 in this case? Yes, I think we do, because we have the difference between what are microscopic cracks in a bone, a stress fracture, and the kind of atypical femoral fracture. And what we have is a statement that says, we do not believe the justifications for warnings about stress fractures are the same. And then the FDA goes on to say, we don't think talking about risk factors for stress fractures is appropriate in this context. So if you're a reasonable drug manufacturer, you get that and you say, well, actually, we should emphasize additional information. We should bring this to the FDA's attention rather than do a big data dump. And so the question of whether the FDA was fully informed of a warning that would be adequate under state laws to test, not whether Merck supplied a whole slew of information and asked the FDA to approve an inadequate warning. Okay, so it sounds to me that your argument is that the language of the precaution Merck proposed was inadequate. But is the language of the precaution that Merck proposes actually the information that is relevant at the first prong of Merck? Yes, that's the test. The test that the Supreme Court in Albrecht said, and this is at 139 Supreme Court at 1678, is that it fully informed the FDA of the justifications for the warning required by state law. But the justifications to me sound like the medical studies. The justifications go to the warning required by state law, which is, are you warning about atypical femoral fractures in a way that advises doctors to make sure that a patient is not going to stay on this drug for a prolonged period of time and thereby run the risk of such significant bone degradation that a woman's femur snaps when she steps across the threshold of her doorway. And so the warning and the justifications have to go together because the purpose of the label is to advise doctors. But a lot of this goes, so the dispute about kind of standard of review and should it be clear error, should it be de novo, I mean doesn't that really center on kind of prong one in terms of the exchange of information from the manufacturer to the FDA? Do we have a similar dispute on kind of the standard of review? Is it clear error or is it de novo on prong two when it kind of comes back to what the FDA then informs the manufacturers of? No, Judge Phipps, that has to be prong two, has to be a de novo question, whether FDA rejected a warning using force of law under congressionally authorized delegations of authority that would trigger the supremacy clause to display state law. I'm just trying to think, how could that be a clear error review? I guess if there's a dispute as to what the FDA said maybe, we didn't receive it, that's a fraud letter. I mean, maybe. You're exactly on the right track and let me emphasize this language from Albrecht. The FDA has to prohibit a warning that would be adequate under state law, prohibit. So if you think it's a jump ball as to what the meaning of an ambiguous FDA letter or action or rejection of a citizen petition, you are obliged to say no preemption. First, there is a presumption against preemption and second, you can't take an ambiguous agency action and say it displaces state law under the Constitution. Well, what are we to do, if anything, with the FDA's position that, no, we did exactly say we're not going to let you do that based on scientific evidence. What you're saying, it sounds like what you're saying is even though the FDA, they haven't done it in this court, but at the district court, at the Supreme Court, they filed amicus briefs and they said, we told them you can't do this, period. You cannot do this consistent with federal law. And we did it on site. It wasn't some interim thing. It was like, you're done. Don't do it. What you're telling us is the official position that the party whose actions we're supposed to be interpreting, their stated position is to be ignored. Well, Judge Jordan, what I would say is that the Supreme Court rejected it by saying the legal standard for evaluating preemption in this context is different than what was offered by the Solicitor General in the context of saying the FDA had gotten plenty of information to know about the risks. I'm not sure. With respect to that, I'm not sure that's answering my question. It sounds like based on what you were saying to Judge Fitz a minute ago, you're really saying that it's a matter of constitutional supremacy law kind of discussion. You can't listen to what the FDA said. Those guys have nothing to say about this. This is a pure question of law. It's the four corners of the document. If the last ten commissioners that the FDA came in swore on the Bible or whatever sacred writ they adopted and said, no, we really meant X, you don't listen to that because you cannot displace this. Albrecht doesn't let you do it. The Constitution doesn't let you do it. Am I over-reading what you're saying to us? A little bit, Judge Jordan, if I could try to clarify it this way. One is that the legal standard advocated by the SG in the Supreme Court was different than what the Supreme Court settled on, and that means that the information and analysis is viewed through a different lens. Secondly, So there could never be a factual issue with respect to prong two. That's your assertion. There could never be. It always has to be a question of law. There could never be a subsidiary issue of fact with respect to prong two. That's why the court says it's so inextricably closely tied to a legal question. And if that's true, what you are saying is you can never listen to what the agency says when it comes forward and says, hey, we're filing an amicus brief because we really want you, the courts, to get this right. When we said this, we want you to know this is what we meant. We meant this. And your position is don't listen to that. That's fact stuff. Stick to the four quarters of the document. Can I offer this suggestion, which is that in Wyeth. You could, but you could answer my question first. I want to answer the question by saying no, because preemption is an inextricably legal question that has got to be resolved by the court. The Supreme Court in Wyeth rejected an FDA attempt to assert preemption in this drug context because it said this is for the courts to decide. But moreover, Ted Shorten, if I could just add. I'm not sure I understood your last question when you said no. Are you saying it is correct that it would never matter what the FDA said because it is purely a question of law? So it just doesn't matter what they say. There's no fact question to be determined. That only thing they could speak to is factual stuff. It's a legal issue, period. The legal effect of an agency action is always a de novo question. Okay, but not the legal effect. I think just to follow up on the Jordan's question. The question of Obrecht-Prompto, whether the FDA informed the drug manufacturer that it would not approve a change to the drug's label to include the proffered warning. Is that a fact question or a legal question? That's a legal question. It could never be a fact question, even if the FDA comes back and says, no, we did inform them of that. Trust us, we did it. This is a question of historical fact. Did we inform them? We informed them. Your assertion is no, that's a question of law because what constitutes informing is purely a legal issue. It doesn't matter what they intended. What they intended is irrelevant. What matters is what they said and the legal interpretation of it. Am I understanding that or am I not understanding? You're understanding this, but can I make this second point, which is that Dr. Sharfstein, who was the FDA official who supervised this complete response letter, has written a brief in this court to say, no, we were not rejecting a warning about atypical femoral fracture. So you have a fact. Because that seems to be saying it is a fact question because we've got the FDA taking an amicus position and we've got Dr. Sharfstein saying that's not true, and that doesn't sound like legal analysis. It sounds like two different knowledgeable people about the FDA arguing about factually what happened. Yes, they did. No, they didn't. And that's why jump ball goes to no preemption because it's a legal question for the court to decide. But on the legal question, if instead of a letter they issued a regulation and then we had an amicus say this is what I thought the regulation meant, and then we had the agency say this is what we think the regulation meant, that wouldn't be a question of fact. That would be a pure question of law. And so when they have the same disagreement over what the CRL says, it's your point that that disagreement is just a disagreement of law. That's right, because the CRL is done under a congressionally authorized regulation that promulgated by the FDA to say this is not intended to be final. And so let's just tease that out. Do you think that you've hit on a few times that something has to have force of law, force of federal law, in order to have a preemptive effect? Do you think that CRLs can ever have force of law? Yes, if the agency makes clear that it is intended to be a final statement about a particular label. So the only way it can have force of law is finality? Yes. And that builds in a lot of like APA 704 concepts into agency action. And the APA in 704 has a definition of final agency action, but elsewhere it has many, many, many definitions of agency action without the final modifier. And so I guess my question is why should we read in a finality requirement to a CRL? Because the test is are you applying impossibility preemption? It's not impossible so long as the manufacturer can invoke the CBE regulation and change its label before it gets the FDA approval. So finality must come, so what you're saying is maybe it can qualify as agency action before that, but in order to hit impossibility, the only way we know it's impossible is if you say this is the final word, there's no ifs, ands, or buts about it, you can't do it. If there's any hedging whatsoever, and you tell us this is a question of law, if there's any hedging whatsoever, then your view would be no matter how unlikely it is, no matter how much of a forecast this is, we're supposed to tune that out and say it wasn't fine. We're pretty sure. Everyone who gets this as a regulated party knows what direction this is going. Don't mind that. It didn't bat down every single nail. There's a passage in Albrecht that says because impossibility is such a difficult affirmative defense to show, in light of the CBE regulation, it will be very, very difficult for a manufacturer to say it could not keep its label up to date. And let me hypothesize, Judge Phipps. If the day after this complete response letter had been issued, a Merck scientist said, you know what, we've been thinking about the literature in the wrong way. I think if we connect the dots, we actually have to warn about atypical femoral fractures. Their position is the complete response letter preempts all state law. Our position is if they got some new analysis or some new study the day after the complete response letter, of course state law would have a role to play in testing the adequacy of the manufacturer's label. Well, thanks for bringing that up, because one of the things you said earlier makes me want to dig a little deeper and ensure I understand this. You've asserted in your briefing here at the podium a few times that it's, and I think the word you used was easy, it's easy to get the CBE. If you really think you need to put something out, it's easy to do. You just do it. But in taking a look at the regulations, and I could have misunderstood them, but I thought that the ability to do the CBE needed to be based on some new scientific evidence. That if your PAS is rejected and you've got no new scientific evidence, it would be improper to just go ahead and change your label anyway using the CBE process, because the scientific data is the same and the PAS was rejected. Have I misunderstood the regulatory regime? I think that you've misstated slightly the new evidence requirement, which is both new studies and new analyses of previous studies. Okay. And so that matters, actually. Okay. Okay. You've got to have something new to say it's different now. And your assertion is they had something to say it's different now? Yes. And they could have done that? What's the thing that's different now? The task force a year later is analyzing every bit of information available to Merck, and it came to the conclusion that atypical femoral fractures had to be worked against. The task force report is in the appendix. And I don't think that you can look at the timing sequence and say, and I started with this point, the manufacturer is responsible for its label at all times. That is a direct quote from Albrecht and Wyeth. They have the duty to look at this analysis and look at these studies to keep their label up to date. Well, not to beat the horse that's already dead, Mr. Frederick, but if you're asking us to look at these things that happened over the course of that time period, aren't you specifically directing us to a factual question? In other words, the district court seemed to say the CD really wasn't an option for them because nothing changed. And you're saying, oh, something did change. Now you're directing us to historical fact. Isn't that something that if we look at what the district court is saying, well, that ought to get de novo review, because the district court is looking at the same thing you're looking at. You're just asking us to make a different factual conclusion about what happened in the interim between the PAS and the time you say they should have put in the CD. I'm not saying there are no facts involved in this case. I'm saying that they are subject to de novo review because they are wrapped up in agency action and inferences about how to interpret final agency action. That's a typical APA kind of thing for a reviewing court to do, and there are many APA actions that come directly from district courts where an agency action is challenged, and the reviewing court does not give clear error review to those facts. We're just circling the same point. Your assertion is, yeah, there were factual determinations that the district court made, including the one we were just discussing right here, which underpinned the availability of the CDE, and your assertion is facts have got nothing to do with standard of review here. It's going to be de novo always, citations to Teva notwithstanding. You're not giving that clear error review under 5286. Absolutely not. You're doing your own fact-finding, judge, regardless of what the district court did, right? Yes, that's correct. Unfortunately, this actually gets me to a point I want to make the court understand. The district court went in a very long search, a speculation search of what did the FDA really mean and what did it really intend. I want to urge the court for judicial efficiency and for understanding the proper legal principles to cut that off. The district court literally spent 40 pages speculating about what the FDA would have done if presented with an adequate label. You don't need to do that. You need to say the complete response letter did not address the warning that we know is inadequate because a year and a half later the FDA said that was an inadequate warning. The question is, are these claims preempted? Are you pinning that point back to the point that you made earlier that in order for there to be preemption, something must have force of law? Speculation about what the FDA would do does not yet have force of law. Is that completing the loop? Yes. It's basically the Justice Thomas position. Is that it? I think it's the majority position as elaborated by Justice Thomas on the facts of this case. If you read the majority opinion and it says force of law and it cites two regulations and actions. If it had been the majority position, Mr. Frederick, they'd have said it, but they didn't. It's Justice Thomas's concurrence. He's the one that comes out and says, look, the game is up because this CRL wasn't final. Done. We don't have to be talking about anything else. That seems to be the Thomas concurrence. And that seems to be the position that you want us to take, right? And if you look at the passage around 1678 and 1679 where the court is talking about what are the legal principles that are applying here, it's basically saying the same thing. They're just not applying it. They're leaving it for this court to look at whatever other – Is this the method paragraph? Federal law permits the FDA to communicate its disapproval of a warning by means of notice and comment rulemaking, setting forth the standards, formally rejecting a label saying to the CRL regulation or with other agency action carrying the force of law. What you're saying is, look, the majority gave the three methods for the agency to fully communicate, to fully inform its position. We're going to look to see if any of those made such a communication and whether any of those just constitute clear evidence of the problem, too. Are you saying, and all that Justice Thomas did was kind of apply those in a way that the majority opinion pulled up short because they left that question. They said that was not now before us? That's correct, Judge Fitts. And I would say that in further support of that, if you read the complete response letter where you see the FDA saying we want to meet with you, we're not sure that this justification is adequate, that cannot be preemptive where there's ambiguity at the very least. And the district judge recognized the ambiguity. She then spent 40 pages explaining how she would try to resolve the ambiguity, and that's a rule that we don't think you should be consigning the district judges in this circuit to because it doesn't get to the core legal question, which is what is the legal effect of the complete response letter. So is your position that because there was ambiguity in this CRL, the district court should not have looked at other evidence to determine, you know, what the words in the CRL said? It should have stopped there and said there is no preemption as a matter of law. Yes, but I would also say we think the complete response letter is pretty clear on its face, and Justice Thomas is completely right that it can't be preemptive if you just read the text. And I would just further point to Avandia where the court said, quote, the very text of the letter says it's open question as to what the manufacturer can do. So applying this court's precedence in a very analogous circumstance, you would say no preemption. In the district court, did you object to the presentation of additional evidence to elucidate the words of the CRL? Judge Freeman, there was no proceeding. It was all paper record. But didn't you provide it? Yeah, we provided our paper record, but the court didn't even have a hearing on an oral argument as to how to apply all this. My understanding was that you were relying primarily on the post-task force communications. Some earlier communications, too, where there were some concerns expressed to Merck about the term stress fracture, but then after the task force issued its report, you know, you relied pretty heavily on the FDA's response saying we don't like that stress fracture. To me, that means that you acknowledged that extrinsic evidence was relevant to the interpretation of the CRL language. No, I don't think so. I think what the task force said and what the FDA said the task force did was to make it a little clearer. And then if you tie back to my very first principle, the drug manufacturer is responsible for its label at all times. Merck could have done the analysis the task force did much sooner, years sooner, had it wanted to do something that would dink the huge profits it was making on one of its biggest blockbuster drugs. Am I right? Would it be a fair summary of your position, at least as the pronged to, that because of the presumption against preemption, there will never be any work for fact-finding to do, because either something is clear one way or another, or if there's ambiguity, that ambiguity has to be decided against preemption. I would say- You got it right? Yes, but can I add this caveat? The caveat is that when you say fact-finding, I think in all of these you're dealing with facts, and the question is what weight do you give them in an administrative process? And judges all the time look at facts as they were presented to agencies, as they were analyzed by agencies. They do not give clear error review to those facts. They give de novo review because it is wrapped up in a final agency determination. So can I just tease this out on the CBE point? This is kind of a strange hypothetical starting point, but if there were no further evidence or no new analysis that could be done here, if Merck says, you know, we've done it all. We took this really seriously. We've done everything we can possibly imagine. There's nothing further we can ever do. Does that color the finality of the CRL? Because now they're saying, look, we hit the end of what we can do. We got no shot at a CBE, none, because we're done. Then doesn't the CRL become kind of final because the CBE option is just not viable anymore? Well, the regulations, Judge Phipps, for the CRL say in their present form. So the question is could the manufacturer have altered the explanation of the justification, have altered the wording of the label warning? I don't think that as people who deal with words for a living, we can say there's only one way to express an idea. And that's why when the regulation says in their present form, they're basically telling as part of this dialectic between the agency and the manufacturer. So it's kind of a notion of just like the philosophy of semantics, that words can be expressed differently in different forms, maybe with different meaning? It means that the CRL can never, ever kind of definitionally be final? I think that what Justice Thomas would say is that's correct. What the majority said is we're going to leave open a circumstance in which the wording is so directly on point about the warning that that might actually be a preemptive action. And we don't need to get there in this case because we know it was an inadequate warning. The FDA rejected those very words, struck them out six times when it approved the final label. We've had you up here for a long time, but if I can just hold you for a minute. Of course, I'm here at your disposal. We haven't talked at all about Justice Alito, Justice Kavanaugh, and the Chief Justices concurring in the judgment. And what seemed like a pretty strong concern that they had and emphasized about Section 355, Rule 04A. Can you tell us, particularly where Justice Alito says it's going to be assumed that on remand, the Court of Appeals will consider the effect of that section on the preemption issue in this case? What you think those members of the Supreme Court are assuming we're supposed to do? I can give you what I think you should do. I don't know what you think we should do, but I also want you to – I'm asking you, you're an experienced advocate. You've appeared before these people. You're one of the premier law advocates in the country. So I'm sure you have thought hard about what it is that the Chief and Justice Kavanaugh and Justice Alito might have had in mind in saying that. So I'm asking you to give us your best – you don't have to be carnac the magnificent, but you have to give us an idea about what you think they were talking about when they said, we're assuming you're going to take account of that when you get this back. Judge Jordan, I have often found it perilous to get in the mind of a Supreme Court justice, particularly in a case that I've argued to that court. I will say this, that the 35504 issue was something that came up in a time period after the manufacturer was exclusively responsible for its label. Congress gave the FDA power finally to impose label changes, and it was that provision that FDA finally invoked in 2010. Now, what is supposed to happen in that interim period is, I think, answered by the CVE regulation. And the reason why is because the Part I of that provision is called Rule of Construction. And the Rule of Construction says, quote, this paragraph shall not be construed to affect the responsibility of the responsible person or holder of the approved application, that's the drug manufacturer. I'm sorry, would you run that again? Yes, it's on Addendum 2 of our brief. It says, quote, this paragraph shall not be construed to affect the responsibility of, and if you could just put brackets, the drug manufacturer, quote, to maintain its label in accordance with existing requirements, including Subpart B of Part 201 and Sections 314.70 of Title I and 601.12. But 314.70 is the reference to the changes being affected regulation. And so Congress is actually codifying the changes being affected regulation by saying that even though we're giving FDA this new authority that it has never had before, the manufacturer still can change its label if it satisfies that regulation. And so I think what, if I may, what I think Justice Alito might have been thinking was that there would be some discussion or analysis of how this particular provision would play out. But I want to make that there, from some of the surrounding stuff, it looks like he and the two who joined him had some concern because it looked to them like the FDA knew about this issue for a long, long time, and they were still letting that label stay out there. And when it came to it, they issued a letter saying, a CRL that said, no, you can't change it. So they knew what was going on, and they specifically said, no, don't change it. And that in light of that statutory section, that should mean something to the Court of Appeals. That should be important to the Court of Appeals, all that background information, because maybe, just maybe, if there's all that background information and the CRL says no, it means impossibility. But then first you have to look at the wording of the warning proposed and the reason why the FDA rejected it, which is that we don't think stress fractures are associated with the kinds of atypical fractures we're seeing in the literature. So FDA is conveying, we're not entirely sure stress fracture is the way to describe it. Okay, that's point one. Point two, this statute doesn't come into effect until spring of 2008. Our position is that for years prior to that time, Merck could have used the CBE to provide an adequate label under state law that would have met state law and that would have saved these women from these horrific fractures. Okay. I mean, just to follow up on your question, I think, like, how exactly does, I mean, this concurrence says I assume the Court of Appeals will consider the effect of 35504A on the preemption issues in this case. Just kind of put me through, like, where exactly in the taxonomy? It's probably not a prong one. Is it a prong one thing? Is it a prong two thing? Is it a kind of tail end of prong two when we're talking about the role of the CBE in prong two? Like, where, there's this assumption that we're going to give that a lot of consideration, and I'd just like to know kind of geographically where should that consideration go? I would put it, I would put it in the prong two part because it is really an understanding of why the FDA acted. Part one, prong one is what is the drug manufacturer doing? Prong two is what's the FDA doing? And so I think that geographically, Judge Phipps, I would situate it in prong two, and I would say that in evaluating certain CBEs, there's a role for the FDA to play, which is that the manufacturer submits prior to approval a supplement that says we're changing the label. The FDA can come along and say, no, no, no, we don't want you to have that. Take the label off or take that warning off. That's the Dolan case out of the Seventh Circuit. And so to the extent that what the Justice Alito group of justices was thinking was that there may have been some communications here that were formal communications that would have denied the opportunity for an adequate warning, but that didn't happen actually once you dive into the record, the administrative record. If there are no further questions, I still would like to have rebuttal if the course leaves. We'll still get you four minutes. Thank you, Fred. Thanks very much. We'll go ahead and hear from Mr. Roth. May it please the Court, Yakov Roth on behalf of Merck. Your Honors, Judge Wolfson carefully and methodically applied the two-step framework that the Supreme Court articulated in this case. And she reached the same conclusion that the late Judge Pisano had reached when he looked at the record and the same conclusion that a unanimous panel of this Court said was well within the range of reasonable inferences from the record. It also aligns with what the FDA itself represented in its briefing to the Supreme Court. Let's get you right to the standard of review, Mr. Roth. Sure. How, if at all, does PEBA play in here? And if it does play in here, is it really the case that it plays in at both prong one and prong two? Here's how we see it, Your Honor. The Supreme Court said, when this case went up, preemption is properly considered a legal question and therefore the super standard of review that we apply to the ultimate question is going to be de novo. The Court quickly went on to say, of course we understand and acknowledge that in answering that question, there will sometimes be contested facts that will affect the outcome. And in conducting the inquiry, the District Court is supposed to resolve those factual disputes within the scope of this overall legal inquiry. And the Court analogized that to PEBA and claim construction, which has the same framework. It's a legal question, but sometimes there are going to be factual issues that are subsumed within it. And as PEBA made clear, Mr. Frederick says, every court that's looked at this says it's just de novo. And you yourself said it was just de novo in a bandia. So you would be creating an intramural and an intramural problem for yourselves if you went down that path. I don't think that's right, Your Honor, because while it's true that the ultimate standard on preemption is de novo, the legal question, I'm not aware of any court that has looked beneath that into questions of underlying factual dispute, those subsidiary factual questions. And is that because they didn't need to or because they thought they couldn't deal with District Court stuff because it was all de novo review anyway? Your Honor, I'm not aware of a case where it mattered, in other words, where the underlying question that affected the outcome turned on a factual dispute. By the way, there aren't a whole lot of cases since – court of appeals cases since Albrecht that have actually looked at this. There's a handful. But a bandia is one of them. Yeah, and a bandia said correctly, this is a legal question. It's de novo review. And it didn't get into the issue of, well, what happens if in the course of answering that legal question, there's some historical factual debate that the District Court resolved after poring through a 1,000-page record? Are we actually going to look at that de novo when that's never what we do under Rule 52? It's not what we do in the claim construction context under TEBA, which is what the Supreme Court cited in suggesting that this was a parallel type of inquiry. Is prong one of Albrecht a factual issue or a legal issue? I think it could be either, depending on what the nature of the dispute here. Here, I think it's largely a factual dispute. We said we presented the material. They said, no, you didn't. Is there something you didn't present? And the District Court went through their complaints and said, no, they did present this. Merck did provide this study. Plaintiffs haven't substantiated any of their complaints. So I think – So in Avandia, the court said you didn't address the Nissen study. You didn't address other studies. Therefore, you didn't present all the information to the FDA. And do you agree that they were treating that as a factual issue? Well, Your Honor, in Avandia, that was in the letter. So the letter that the FDA sent to GSK said, in order to evaluate this, we need the following four pieces of information that you didn't give us. Please give us A, B, C, and D. So I don't know if there was really a factual dispute about it. The FDA had said, this is what we need from you in order to make a decision on this issue. And then GSK provided it. And then the FDA said, yeah, go ahead and add the warning. And, in fact, add a more stringent warning that you would even propose because this is a more serious issue than we realize. Okay. So in this case, prong one is a fact issue. And do you agree that the language of the proposed precaution is among the points of information that needs to be provided to the FDA? No, Your Honor. The way I understand the inquiry, the first prong is about, did you provide the underlying scientific basis for the warning that plaintiffs say was required? So here, did you provide the data, the studies, the case reports, the analysis of why there is a need to add a warning, a reasonable basis to believe there's a causal association between the phosphonate use and this risk of fractures? Did you provide that data to the agency so it could evaluate the question for itself? I don't think it has anything to do with the language of the warning. Let's take a hypothetical scenario where the manufacturer provided all of the data available. There's no question that they gave it all to the FDA. They put them every study under the sun. And then they provide a proposed precaution that is entirely unrelated to the information they provided. And the agency says, we're not going to approve that. So have they satisfied Obrecht-Prong 1? Yes. I think in that situation they'll have satisfied Prong 1. Maybe they wouldn't be able to satisfy Prong 2, because maybe that denial can't properly be read as a rejection of an appropriate warning. But if they provided the scientific basis, I think we're into the question of, all right, what did the FDA say back? What did they do with this? Okay. So we do have to assess whether the warning was appropriate. Well, we have to assess it in the context of we need to understand what was the agency saying when it said no. What is the effect and scope and meaning of that rejection? And so for that purpose we have to look at what was submitted, what did the agency do, and what is the statutory and regulatory context in which the agency takes that action. If we then assume that there's factual components to Prong 1, what work is there to do in the way of fact-finding at Prong 2 if, as your learning colleagues on the other side tell us, and there's a lot of backing up, that the presumption, the legal presumption, is against preemption. So you're in a world, now I'm not saying it as well as they did and maybe not even the way they would say it, but if you're in a world where something's clear and unambiguous, you don't need fact-finding because it's clear and unambiguous on its face. But if you're in a world where that legal document, that CRL, is ambiguous, then the presumption is doing the work and there's no need for fact-finding there either. So Prong 2 is always going to be a question of law. What's the answer to that? Let me answer it this way, Your Honor. I think most often Prong 2 won't require resolving factual disputes, and I think this case can easily be resolved based on the language of the complete response letter in light of the statutory and regulatory background without getting to factual disputes. But let me give you a hypothetical where I think it would be appropriate. So imagine that a manufacturer proposes a warning and the FDA response says, Denied for reason code A-22. And we don't know what code A-22 is not defined in the regulations, but the FDA has some internal guidebook under the desk that says, Oh, A-22 means the science isn't there. And would that really be an ambiguity? Because if there's code A-22, it's not like nobody knows what code A-22 is. You'd have a party say, Code A-22 means this, and the other side would say, Oh, it does not. Right. Not fact-finding at that point, really. I think it could be. We could have a witness who says, Oh, yeah, I was at the FDA. A-22 means this. Here's the document. And the other side could say, Not authenticated or you're not a credible witness. I mean, there's all sorts of factual disputes that could go into that. Now, that's a bit of an extreme over there. I don't think we need to get there for this case. The reason I think this case should be simple to resolve is because we know from the FDA's regulations that if there's a proposal and they don't like the wording, they tell you how to fix it. That's what the FDA said in its amicus brief. That's what the regulations say. They don't reject warnings for editorial concerns. And, of course, that makes perfect sense because their whole job and their statutory duty is to make sure patients and doctors have the information they need. I mean, you're the ones who argued vociferously for the no review of everything. So, apparently, you thought fact-finding didn't make a difference at prong two, but maybe I'm wrong about that. Is fact-finding needed at prong two or not? I don't think it's needed, but to the extent the court gets past that question of the CRL in light of the statutory and regulatory background and still thinks there's ambiguity, then I think it is appropriate to look at the extrinsic evidence of what is the meaning and intent of the FDA action. Why wouldn't it be, at that juncture, appropriate to say, you know, we're not going to go down, looking at emails and text messages and transcripts of voicemails to try to figure out what was going on, and the district court shouldn't do that either. That's what the presumption is for. The presumption is there specifically to keep us from running down that particular path. Your Honor, I don't think that's how the Supreme Court has treated the presumption in this context. Gwyeth was based on the presumption. The presumption was the reason why the court said, in order to get preemption, you need to show that the FDA would have rejected label change. And I think we do that in this case by pointing to what the FDA actually did in the CRL, understanding that properly in light of the regulatory context. But if, as in my A22 hypothetical, there were really some lack of clarity about what it said or two interpretations and the extrinsic evidence resolved that in a clear way, as the district court found here clear and convincing, she said, I don't see why it would be inappropriate to look at that. It's interesting because when I read kind of this, when I read the clear evidence test, it's actually a legal question, which is kind of a fun, you know, people practice a lot of clear evidence, it's just a question. Well, it's kind of a fun one, really. But my way of thinking about it is that it involves kind of a two-phase exchange of information. And tell me if you disagree with this, because I'll come around to my question. The first one seems to be the manufacturer providing its justifications for the warning. And then the second one seems to be the FDA pushing back and saying, hey, this is why your warning's a no-go. And so I'm just trying to figure out, like, where do we fit actually providing the warning into either one of these prongs, because is it built into prong one, you know, the manufacturer's proposed warning? Is it built into prong two? Is it just something that we just kind of say, hey, we know that that has to happen. But if we read prong one as justifications only, that means no warning. It doesn't require warning. And if we read prong two as just reasons for rejecting the warning, that's kind of no warning. And so I'm kind of just, is there ever an obligation? You see, I'm just kind of struggling on that, on where does it fit. Yeah, I mean, I do think it's a two, there's the back and forth. I agree with what your Honor has laid it out. You don't get the FDA's response unless you propose the warning, and there's going to be a proposal. Is this like Chevron where, like, step zero, good warning, you know? I don't know, but, like, it doesn't seem to fit clearly into one or two. If we're playing discrete prong, I'm not sure where it fits. I think it's a prong two issue, although there are lots of preemption cases where the manufacturer does not propose a warning. The FDA takes action in other ways that conveys that it would not. The need for, like, the denial of a citizen petition happens all the time. Somebody else says, you should demand a warning for this. The FDA says, we looked at it, no, we don't think so. The manufacturer then doesn't have to propose anything because it has already been conveyed through agency action that the agency won't approve that. Here, given the way this was teed up, the agency action was initiated by Merck's proposal, and the whole debate at prong two is, what did the rejection mean? Did it mean, no, because we don't buy it? Or did it mean, no, because we don't like that you used the word stress fractures in the warning? And I think their interpretation is clearly implausible as a legal matter because the regulations provide that if the agency has an editorial concern with language, it will suggest appropriate language that would solve the problem. And that is what the agency did with respect to the adverse reactions part of our proposal. They didn't say, go ahead and do it. They said, you can do it, but make this change. It's also what happened in 2010. So after the task force report came out, after the agency came around on the science, there was a back and forth. And we said, we want to put in this language about stress fractures. And the FDA said, no, here's a red line. Take it out. If that was the only issue in 2008, that's what the agency would have done in the complete response letter. So is that kind of a guide for interpreting complete response letters to the extent that they say, look, if we have a complete response letter that doesn't make a counterproposal, then it's safest and best to view that as final? I'm not sure how to use the term final. But binding force of law. Should we say force of law? Should we say force of law? Such that it can be federal law for purposes of preemption? Yes, I agree with that, and that's the majority opinion. Sorry? That's the majority opinion. They lay out three ways. Right. And one of them, I've actually pulled up the page, because one of the three that they identify is formally rejecting a warning label that would have been adequate under state law, see example, the regulation that provides for complete response letters. So they are clearly saying there, a complete response letter is a way. It can get the job done. Yeah, it can get the job done. It can get the job done. But it doesn't necessarily get the job done. That's right. And if it had said, we don't – if the CRL had said, look, this doesn't work. You don't talk about stress fractures. We don't believe that that's appropriate. Here's how you should fix it. Here's some alternative language. And then we dropped it. Then we wouldn't have a case. I'm sorry to interrupt. You're making good points. Please circle back to them. But I guess my question is, what you're saying is, if the CRL does not contain an explanation, then everyone should know, as to a modification for the warning label, then everyone should know that that's kind of a for all times, for all places rejection. It's never for all times, because there can always be new things. For purposes of what we've got here. And so I guess, just to tease that out, that does bank a little bit on, you know, giving a lot of prominence to the role of agency silence, right? It's the silence of proposing alternative that makes the CRL have force of law, right? And that kind of, it feels a little weird to me to sit there and say, from silence we will now infer a force, from rejection plus silence of alternative, we will now infer a force of law. Is that too much? Am I pushing too hard on that, you think? I think it's too much to say that's silence. I'd ask you, Your Honor, look at the CRL, because what it says is, you've completed a review of your application as determined, and determined that we cannot approve these applications in their present form. Describe below our reasons for this action and our recommendation to address this issue. Paragraph one then is the reasons, the three reasons for why they're saying no. Paragraph two is the recommendation to address the issue. And they say, we recommend that you add low energy, femoral shaft and subtrochanteric fractures in the adverse reactions post-marketing experience subsection of the respective package insert, which is exactly our point. The agency looked at it and said, we think this can get into adverse reactions. That's a lower standard of scientific proof. So that's what you should do. And they tweaked the language, because that language that they proposed was not exactly what we had suggested. So the silence on precautions should be read as impossibility. Yes, it should be read as rejection. The rejection establishes impossibility. There is more information on 1152 of the appendix. In paragraph one, they say, your justification for the proposed precaution section language is inadequate. Identification of stress factors may not be clearly related to the atypical sub, you know, to the fractures that have been reported in the literature. So the stress fraction language may not actually be relevant to the literature. Isn't that like part of the explanation of why this is? Yes, but, Your Honor, the difference is, we read the three sentences as three reasons for the rejection, because they say reasons for our action, and they're reading it as all one that you've bundled together. So why is it inadequate? It's because you use this language of stress fractures that may not be relevant. So this is classic legal analysis, right? Yes, and the district court. What would facts have to do with that? Your Honor, I don't need to rely on any facts to establish that, because our reading is not only the reading that is correct under the regulations, it's also the reading that the FDA said was correct in light of the regulations, which is, again, all part of legal analysis. When you say the FDA said, at what point in time? In its amicus briefs. In the amicus briefs. Yes, the FDA represented in its amicus briefs in the Supreme Court that this is the right way to understand what they're doing. I don't want to read into their nuts showing up here. I don't think so, Your Honor. They were invited to appear by the Supreme Court, so they filed a brief and said, we looked at it, this is how you read the regs. As I read the amicus brief, they're saying, this is how you, the Supreme Court, should interpret the record. So they're not actually saying, this is our expert opinion, these are our qualifications, this is what we think. They're saying, if we were the judge, this is what we would do. Your Honor, I think a large part of their brief was explaining the way the regulatory scheme operates, and you need to understand that backdrop, the regulations, how does the FDA approach these letters, what do they do if they have a problem with the wording. That's spelled out in regulations and the FDA walks through that and then said, in light of that, the proper reading of this complete response letter is that we said, our position was, the science isn't there yet. What do we do with the Bates case, right? It's cited in the public law scholars amicus brief. They emphasize it pretty heavily where the Supreme Court said, we have a duty to accept the reading that disfavors preemption. If we're duty-bound to look at this and argue back and forth what this means, because that's what's happening in this courtroom right now, why aren't we duty-bound, to use the Supreme Court's words, to accept the reading that disfavors preemption? So, Your Honor, that's not a drug preemption case. I'm talking about, understood. Does it have no bearing here? I think it doesn't, because I think in this context, it's been the presumption in that framework and the way we understand the supremacy clause is all built into the Wyeth-Merck test. That's what gave us this inquiry. There's a different sort of a presumption in a drug case than in another case, even though in the drug cases and in Wyeth, we're told there's a presumption against preemption, and here it says we're duty-bound to accept the reading that disfavors preemption. You're saying those are two different standards. So, what the Supreme Court did in Wyeth was look at the particulars of this regulatory scheme. It's got a presumption. It established a presumption, right? It didn't establish a presumption. It rested on the presumption and said, in light of the presumption, we're not going to assume you can't add a warning. We're going to assume you can add a warning, unless you come in with evidence that the FDA would have said no. And if you come in with that evidence, clear evidence that the FDA would have said no, then you've overcome the presumption. And here, you can't get better evidence than the FDA saying no when we ask for permission to do the thing that they say we were required to do, which is add a warning. You have noted the clear evidence standard a couple times. I want to read something from the district court and have you tell me whether the district court actually played that standard. This is from page 106 of the appendix. And it's the court's sixth opinion saying, the more likely scenario is that the FDA's actions taken in this case convey doubts that the agency had about the underlying science, a deficiency no revision or edits could solve, hence the agency did not propose any agency in action point. When the court says, the more likely scenario, is it applying a clear evidence standard or is it just applying a more likely than not standard and that's not good enough? Your Honor, the Supreme Court said clear evidence isn't a heightened standard beyond preponderance. So when this went up to the Supreme Court, there was a question, what is the standard of proof? It said it's not the same as clear and convincing, but did it say it's not heightened? What does the word clear mean if it's not saying something different than preponderance? Well, that was the fight that got us the last, I think the last case in the Supreme Court opinion. I'm not sure they exactly spelled it out, other than to say what it means in this context is you've got to show these two things. Clear evidence in this context means you've got to show you gave them the basis for the warning and you've got to show that they said no. That's just the same, that's nothing saying more than preponderance. I don't understand, unless we assume the Supreme Court just uses language sloppily to say you have to establish this by clear evidence, has to do some work and if it's not doing any other work than saying standard preponderance of evidence stuff, it's nonsense. Well, perhaps, but I think they pretty clearly said when it went up the second time, we weren't referring to the standard of proof. That's not what we meant. It's a legal question, so standard of proof is sort of inapposite entirely on this question. It's just sort of what Justice Alito called a flourish in describing the two steps. Can I get in a little bit on the, I guess I'll join the standard review party. I guess I'll ask a question about standard review too. If the standard, I'm going to the methods paragraph. So we've got this methods paragraph. I see they're basically the majority opinion saying there's three methods that we will accept as having federal law value for purposes of preemption. So the first one they say is notice and comment rulemaking. So I guess what I'm thinking is what sort of factual review, if it is a factual question and there are subsidiary factual questions, what sort of subsidiary factual questions are presented in notice and comment rulemaking? And if it's a null set, I don't think that means that we can't have clear error review. It just means that under that method, we don't really have to get, I'll tip my hand at least as I see that playing out. Well, let me give you an example, Your Honor. One of the labeling standards is in order to add a warning through the CBE process, you need newly acquired information that has to meet the requisite level of proof. So many preemption cases come up this way where the manufacturer says, we didn't have any, there wasn't any. It didn't satisfy that regulation standard. So we couldn't have added a warning because there was no newly acquired information. And the district court then looks at it and looks at the information and says, yeah, you're right, this isn't newly acquired or you're wrong, this was newly acquired. So you can have factual. That's classic. In your view, you'd say that's classic subsidiary fact. Subsidiary. Where the ultimate question is de novo. We're just in 52A. In the course of applying that regulation, there may be facts that matter to the inquiry. But at least for the notice and comment rulemaking one, I was kind of pinching you on a little bit. Like probably we're going to be close to the null set on facts. That was the notice and comment rulemaking. The rule says, here's what you need. Then we have to apply it. We have to apply that to a particular fact pattern to say, all right, in light of that notice and comment rule, could you have done something or were your hands tied? And in order to answer that question, there's going to be some factual issues that come into play. What's your response to the, I mean, multiple amicus came in here, and I'm interested in particular to your thoughts on what the 23 states that weighed in said in arguing that, and now I'm paraphrasing. I may be doing a bad job. But it seemed to me their message was, in essence, if you give agency inaction, the kind of sweeping preemptive effect that Merck is arguing for, you will greatly undermine state sovereignty and you will, in fact, prevent state court law from doing the work Congress wanted it to do. Congress could have sometime in the last 70 years of the existence of the FDCA have said, you know what, express preemption. FDA speaks. Everybody go away. They haven't done that. The Supreme Court's recognized they haven't done that. So in order for the state tort law to have the role it ought to play here, you cannot let agency inaction do the kinds of things Merck is asserting it's doing here. Go ahead and take a swing at that. Your Honor, we're not asking for preemption on the basis of agency inaction. We are asking for preemption on the basis of the agency telling us no. Okay. So your assertion when the district court says, as the district court did, because I read you the way the district court put it. The inference, yeah. The FDA's subsequent inaction, it did not mandate a label change despite substantial ongoing review, confirms its then existing perspective. I mean, the district court itself calls out agency inaction. So the court looked at, I think appropriately looked at, what the agency said and did and didn't say and didn't do in order to construe the meaning of the action that the agency actually took. The court was very clear. The thing that has preemptive effect is what the agency did. And now I have to understand what the agency did and what it meant. And in order to do that, I look at the regulations and the context. And as part of that, I understand that if it were true that the agency was convinced there was a basis to add this warning that was warranted by the concern was with the language, here's what they would have said about that. Well, there's the thing. Here's what they would have said. That's where you take us into the world of hypotheticals. But what the agency actually did in the CRL is, as quoted to you by Judge Freeman, said your justification for the proposed precautionary section language is inadequate. It doesn't say you couldn't frame adequate language. It says what you gave us is inadequate. That's right. And you want us to say, no, that was final. Because what it really meant was nothing you could put in this label would be adequate, right? In other words, you wouldn't require us to draw in order for your position to win. Yes. You need to construe the CRL that way. And that is the correct way to, legally correct way to construe the CRL in light of the regulations. Because, again, the regulations say, and the agency said in its amicus briefs, that when the concern, and they say this in the regulations, when we have editorial concerns with language, we will propose, we will suggest a way to remedy them. We won't just say, no, we don't like this. We're going to sit on our hands for 18 months, ignore our statutory obligation to make sure patients have the information they need, publicly announce to doctors around the country, you should continue to adhere to the warning as it stands today, because we looked at the data and we haven't found a connection. We're not going to do that. The FDA amicus, you're saying that should have real weight with us. I mean, the FDA amicus in the Supreme Court. We should be paying close attention. I think it helps us understand the regulatory context. What do we make of the amicus brief filed here by some former leading authorities at the FDA saying, yeah, it's just not right. Well, Your Honor, I think the agency is entitled to some deference in how it construes its own regulations. Would that agency be speaking the same way? Now, this is why I asked, is there something to be drawn from their not being here? I mean, there's been a change in administration, and sometimes that means that the words you get out of agencies aren't the same words you got out of them before. Maybe, but, Your Honor, they didn't participate last time in this court. They participated because the Supreme Court called for the views of the Solicitor General. The Solicitor General then consulted, as is clear from the brief, repeatedly says, consultation with the FDA, here's what we understand to be the regulatory practice and procedure that governs here. They made those statements, and I think it's fully appropriate to rely on them, especially when this is a case where the agency has no dog in the fight. They're not a party to this, right? They were asked, what are your views? And they came in and said, here's how the regulatory scheme works. In light of that, our action is appropriately understood as a real no, rather than we just don't like your words. Okay. And that is sufficient to preempt. I do resist the word final that Your Honor suggested. The word final is not in the Supreme Court's majority opinion as a necessary step. And I think in some sense it's never, nothing the agency does is final because you can always have new information that comes up later that would warrant going back. So it's not final in a- A force of law. Yeah, a final force of law. And the force of law kind of implies, doesn't it, that the CRL shouldn't be tentative. It should be a definitive statement because of the evidence that exists at that time. That's right. I agree with that. And I agree with force of law. I just don't, I think final puts too much of a premium on it. And the Seventh Circuit in Dolan specifically said, once the agency says that in that authoritative way, we don't say, well, the manufacturer could have asked for a further review and maybe changed their mind. That's getting into possibility of possibility, which isn't sufficient. We've kept you up there for a long time. I want you to tackle, if you would, at the end here, the same thing I asked Mr. Frederick, which is what are we to make of Justice Alito's statement that we're supposed to be paying particular attention to that 355 rule O section? Do you agree with sort of where the discussion between Mr. Frederick and Judge Phipps came down, that that's a prong to issue, and we should be thinking about it in that context? I do agree with his answer to that question. I think it's appropriately a part two question because I think it's part of understanding the agency's action and what the agency's action meant legally, is to have a background appreciation of the agency's statutory obligation that if it becomes aware of safety information that warrants a label change, it is supposed to do something about it. It doesn't take away from our duty, but I think it tells us how we properly read what the agency has done and what the agency has done. I don't want to hear the CRL that said no. Okay. Anything else? All right. Thanks so much, Mr. Roth. Mr. Frederick, that's your rebuttal. My friend never acknowledged the requirement that manufacturers are responsible for their labels at all times. And I think that you can begin and end the analysis there. He misstated the regulation, which we set out at page seven of our opening brief, that requires in a complete response letter the FDA to, quote, describe all of the specific deficiencies that the agency has identified in an application. It's not the FDA's duty is to rewrite the manufacturer's label. It's to identify, as Judge Freeman, you pointed out, we don't think this is an adequate wording. He admits that he doesn't have any case that says there's clear error review to be done at the Court of Appeals level. So I think, Judge Jordan, the point that I was making about creating a conflict stands. Judge Phipps, your questions about the warning, where does it fit in, it fits into both prongs. If you read that passage from Albrecht carefully, it's talking about information and justifications for the warning that would be adequate under state law, and the FDA's rejection or prohibition of that warning that would be adequate under state law. So you have to look at both. So you look at both of those, but is there any prong in the Clear Evidence Center that says you have to provide the warning? It just seems to be implicit. Well, what the manufacturer is charged with doing under state law is ensuring that it is providing warnings that would prevent the injury that would be caused by the adverse reactions. So the state law requires this failure to warn as a classic tort. The warning that the manufacturer has been charged with under FDA practice and long-standing state tort law is to provide warnings that would prevent the injuries that would be caused by the adverse reactions. Ultimately, what they're trying to do is to shift the burden to the FDA, but why is Albrecht in opinion saying no? If that's what they're trying to do, why is it a fair reading of this to say what they're trying to do is get an accurate assessment of the record that, in fact, as Justice Alito was arguing in his concurrence, the way the majority framed it left out a whole lot of stuff about how Merck was trying over a period of time to get this thing straight and feed the FDA was saying, yeah, no. And then when they got a no that looked like a definitive no, they had a definitive no. And it would be a curious world in which you give the regulatory authorities all the information that they're asking for, get told you can't change the label as you propose, and then face state tort liability because you didn't change the label. I don't want to think like you, don't want to try to get inside the mind of the Supreme Court justices, but it seemed like those who were concurring the judgment were a little troubled that it seemed like Merck was trying to be a decent corporate citizen here and at the end of the day, skillful plaintiffs' lawyers will do as they do, come in and say it wasn't enough, you're still liable. When they're really getting whipsawed. You know, we can cut through all the regulatory stuff, and that's what's really being argued here is we're getting whipsawed. We did what we could with the federal government, they said no, and now plaintiffs' lawyers wielding state law will come in and beat us over the head with what's a nice drug company to do? And that's not a bad question to ask. What are they supposed to do, Mr. Frederick? Let me answer the question. If the word stress fracture that were stricken six times when Merck resubmitted its label after the task force had been stricken and the words atypical femoral fractures in the complete response letter put in their place, we probably wouldn't be here. There was a legitimate scientific uncertainty at the FDA about the femoral fractures. Answer Mr. Rock's assertion that if that was all that was going on, if they just said, hey, say this, we'd have said it. But they didn't do that. All they said was not good enough. Like how often do you have to bid against yourself before you think we're done? Like they told us no. How often do you have to go back and be told, you know, no, no, before you think, no, I think they really mean no? Judge Jordan, it's the reason why the manufacturer is charged with responsibility for its label is that it has supreme information of any body actor, any actor out there, any scientist, any FDA official. And that's why Merck was in full capacity of assessing all of the science that had been developing to them. And we point out on page 12 of our reply brief, in fact, to answer Justice Alito, they didn't provide everything. The task force identified two key points in which Merck had not submitted information for the record. And so I don't think that as you sit here today, the exercise should be, can we try to reengineer every last word of the complete response letter when the day the complete response letter made it to Merck, the scientists got on the phone and said to each other, yeah, FDA is rejecting stress fracture warnings. And so, you know, we can debate this. That sounds like underlying facts. Well, it is an underlying fact in the sense that we're not here asking for clear error review of a phone call because we know from your decision in Avandia that a phone call between the FDA and the drug company is not entitled to any special deference and we can still have preemption in spite of a phone call. Okay. I use the word clear and so let me ask you a question that I put to Mr. Monk. What are we to make of the clear evidence statement? Is that really just kind of, is it just a flourish? I said nonsense, which is probably not a good thing to use when you're talking about a higher court. But why say it? What's the point of saying it at all if all it means is, yeah, preponderance of the evidence? I think that the, I actually don't know what to make of all that. Okay. What I would say, though, Judge Jordan, is that if you go back and you put a red pencil on every place where the district court speculated, you quoted one of them more likely than not. There were many, the decision is actually kind of shockingly full of speculations about what the FDA was thinking and what it did. And that's not the kind of exercise we should do. The Supreme Court said you have to find out whether the FDA would have, et cetera, et cetera, you know? That sort of sounds like we want you to go back and think about, yeah, the words are would have. Doesn't that ask a court to try to do a little bit of prognosticating? Isn't it a direct invitation to do that? Not any saving pages that took a year to write kind of prognosticating, Judge Jordan, for an affirmative defense. Okay. And the key point that I want to, you know, stress here, when the states, a diverse group of states, comes into this court to say, do not have what would be ambiguous agency action or inaction, as you might properly describe this, encroach on state law that is designed to protect citizens from horrific injuries, where the manufacturer has the responsibility to do this. I think that, you know, there is a step, many steps that have been taken too far. You know, we've spent a lot of time this morning and afternoon debating the meaning of the complete response letter. If it is at a draw at this point, although I think it is pretty clear that the FDA is inviting more interaction, you have to say that there is no preemption because the manufacturer has the duty to keep its label appropriately up to date to prevent inadequate warnings. Okay. Thank you, Mr. Frederick. Thanks again, Mr. Roth. It's well argued on both sides. Briefing was very helpful and we appreciate it. We got the matter under advisement. Thank you. Okay.

---

Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/argument%3A90985. Public record. Not legal advice.