Oral argument — Genus Medical Technologies LLC v. FDA

case number 20-5026 Genus Medical Technologies LLC versus United States Food and Drug Administration, a balance. Mr. Winnick for the balance, Mr. Farquhar for the ability. Mr. Winnick, good morning. Please proceed. Good morning, Your Honor. May it please the court, Daniel Winnick for the FDA. The product at issue here falls within the text of the drug definition in the Food, Drug, and Cosmetic Act, and the 23 years. Yet the district court concluded that Congress unambiguously foreclosed the FDA's discretion to treat this product as a drug. That was incorrect for two reasons, one that applies to the FDA's regulation of all products, and one that applies specifically to its regulation of contrast agents. Let me start with the broad reason. The district court's rationale was that because the statutory device definition is narrower than the drug definition, it must be read, it can only be read as impliedly limiting the drug definition. But that assumes Congress could only have meant for the drug and device definitions to be mutually exclusive. And in fact, there's no reason to believe that. The statute used to contain language excluding devices from the drug definition, but Congress removed that language 30 years ago. The current text of the drug definition plainly encompasses the device definition, and the statutory structure and purpose provide no basis to read into the drug definition a limitation not supplied by its text. The canons invoked by the district court require either conflicting statutory provisions or superfluity, and here there is neither under the FDA's interpretation. So it's consistent with the statute's two-track regulatory scheme for the FDA to decide which of the two pathways to apply to a product that satisfies both definitions. Can we talk about- Mr. Witt- Go ahead, Judge Henderson, sorry. Okay, I was just going to ask. Your letter says that it appears that this silk solution is a device. I have a hard time seeing how it's a device. I realize that that's not an issue that you all have raised, but there is plenty of precedent for us to consider an issue antecedent to, and I'm looking now at the Supreme Court's case in 1992, the U.S. National Bank versus independent insurance agents. A court may consider an issue antecedent to and ultimately dispositive of the dispute for it, even if the parties fail to identify and brief the issue. And as I say, I just have a hard time seeing that this oral solution is a device. What does it fall under? It's not an instrument. It's not an implement, a machine, an implant, or in vitro reaction. It's not even an apparatus or a contrivance, whatever that means, and you're left with an article. And I realize that every drug is defined in terms of being an article. So let me address the question first and then speak to its relevance to this case. The clause Your Honor just read, the sort of preparatory clause of the device definition, has been read differently at sort of various times. The FDA's current view expressed in its guidance is that that language doesn't play much of a limiting role, and the reason, frankly, is that there's just such a broad diversity of products that many things that you wouldn't think of as machine contrivance, et cetera, are properly classed as devices. So for example, an example given in the FDA's guidance is a solution that is used to clean surgical instruments or something like that could be a device, even though you wouldn't think of it as a contrivance implant, et cetera. And the FDA has sort of moved past the view it might have held at certain times that that language refers to devices as, quote unquote, things that go clank. So the FDA didn't pass on this question here. It didn't specifically find that this was a device. It merely said, as Your Honor noted, that even if it is a device, the agency would choose to regulate it as a drug. I don't think under the Chenery Doctrine it would be open to the court to reverse on the ground that actually this is a device, since that wasn't the issue passed on by the agency. But I would say as to what- Well, then don't we run into the threat of an advisory opinion if we say, well, we'll make the same assumption that this is both a device and a drug when at least I think it doesn't qualify as a device to begin with. And so in other words, the question before us is, does the FDA have the discretion when it's both a device and a drug to treat it as a drug? And my question to you is, before we even get to your discretion, it's not a device. So without being able to state definitively how the FDA would address the question, is it a device? I can say that the FDA is highly skeptical of the conclusion that this product doesn't fit the device definition. The FDA regards the, you know, it's not- Wait, there were a bunch of negatives in there. What is the FDA? The FDA, if the court were, for example, to remand for the FDA to specifically pass on whether this is a device, I think it's highly likely that the FDA would conclude that this is a device. I don't think- Is it a device? That this does fit the device definition. And I mean, I had the same question to some Henderson, and the best I could do was to think that this is an accessory to a CT scan or an X-ray, that the barium is part of what sets up the scanning technology to actually function. And so it could be plausibly a device, but it also led me to be somewhat surprised by the FDA's position that the drug definition wholly subsumes the device definition, because it seemed to me with some of the other references in the prefatory phrase of the definition of device, there are going to be implements, machines, apparati, that because they are so clearly those things, would be devices and would not be drugs. And that in that, you know, setting aside the district court's application and the Appalachia application of the specific predominating over the general with respect to the exclusion, with respect to the definition of a device, it seems to me more plausible and straightforward that what is more narrowly explicated in the device definition might in fact bring things, be decisive of the categorization for at least a number of devices. But that's not the FDA's position. It is not the FDA's position. And I do think as a textual matter, the relevant word in the drug definition is articles. And the device definition does specify an error or subset, instrument, apparatus, et cetera, or other similar or related article. But that is a subset of articles. And so we return to our broader position here, which is that there is no basis to read the device definition as impliedly limiting the broader drug definition. Right. Unless there is, unless reading, you know, unless declining to do so creates either an which is what the district court believed. And that's simply incorrect. So on, but on your understanding, all devices are drugs, correct? Correct, your honor. And so then let's talk about statutory structure. Statutory structure, as I understand it, is all drugs must go through a very elaborate approval process under one part of the statute that's done by one part of FDA and all devices have to go through a very different approval process that's done by a different part of FDA. And nobody thinks that some particular article should have to go on both tracks, correct? That's correct, your honor. Everybody agrees that the FDA has to identify whether to treat the product as a drug or device. And that's commonplace under the FDCA. So the statutory structure is if drug, then you're on track one, but not track two. If device, you're on track two, but not track one. And the device circle is wholly within the drug circle. And yet you acknowledge that this can't be rationally read to require both. And then you treat the statutory structure, which says the definition has all of these mandatory consequences associated with that product. You read that as saying, well, no, the FDA in fact has unfettered choice to pick whether to treat something as one as opposed to the other. So I disagree in a couple of ways, your honor. First, I don't think any of the mandatory consequences your honor referenced follow from the fact that something falls within the broad drug definition. I think it follows from the FDA's decision that it is to be regulated as a drug as opposed to a device. And again, there's lots of- You don't think the statute is self-executing in saying that no drug may be marketed without FDA approval? I think if something satisfies both the definition of a drug and the definition of a device, and the FDA exercising the role Congress gave it under section 360 BBB2 determines that the product is to be regulated as a device. I don't think under the statute, the drug requirements apply. So if something were a cosmetic and also a drug, the FDA would have discretion to put it through the much later preclearance of or approval process of a cosmetic. I believe that's true, your honor. And it's a great example of the sorts of overlaps that happen all the time under the FDCA. It's not just in the drug device context. There's also, of course, biologics as well. So it is commonplace in this regulatory scheme for products to fall within multiple broad regulatory headings and for the FDA exercising discretion given to it by Congress to decide which pathway to apply. So everything that's a tobacco product is everything that's a modified risk tobacco product is also a tobacco product kind of by definition. So if the FDA wanted to, it could just ignore the modified risk tobacco product, more demanding pathway and treat all modified risk tobacco products as just ordinary tobacco products. It seems quite surprising to me. So we don't think in the context of this case, we don't think it would be remotely open to the FDA to say that it is going to treat as a drug every product that satisfies the device definition. And this goes to the second part of this question, which is, is the FDA's discretion here unfettered? It isn't remotely unfettered. I mean, if the FDA did that, it would clearly be contrary to Congress's decision to create true pathways. What the FDA has done is exercise discretion in a very limited number of cases, which is why the issue hasn't arisen to this in 30 years. But the question is what you can do under your interpretive theory. And if the, on the Venn diagram, if the circle for devices is entirely within the larger circle for drugs, and your theory is that any, any article that satisfies both, which is everything in the smaller circle, you have discretion to treat as either. What follows from that is you have discretion to treat every device as a drug. And that just makes nonsense out of the statutory structure, which is one reticulated scheme for drugs and another for devices. But with respect to honor, we don't think we have discretion to do that. If FDA treated as a drug, every product satisfies the device definition, it would clearly be contrary to Congress's decision to create two statutory schemes. Right. And that's really Mr. Farquhar's view. And I guess, to me, what's, maybe we should move to the argument that's actually made, which is not about the predicate clause in the medical device definition, but about the exclusion. It seems to me there, even if you were more receptive to the questions that we've been asking about the predicate clause, that it in fact specifies some category of things that are devices that can, that don't also fall within the drug definition because of the operation of their additional specificity. So because of the operation of the canon that the specific governs the general, even if you accept that, then it seems to me that the district court applied the same reasoning to the exclusionary clause. And I'm not sure that it necessarily follows. So the exclusionary clause is excluding roughly things that are metabolized or function through a chemical reaction. And it says, those are not devices. And as I understand it, what the district court did was say, okay, if they're not devices, that means they're drugs. And that means that something that is not, that means that that is part of the drug definition and something that does not fall within that exclusion, therefore necessarily is not a drug. And those further implications seem to me not necessarily to follow from the canon that the specific governs the general. And I just wonder, and let me just add one more thing that might make that clearer. So the fact that it's not a device might imply that it's a biological or that it is a food, not that it is necessarily a drug because it's a multipolar statute, not a bipolar statute. And I guess I'm just wondering whether you agree with that. It's a narrower way to look at your position than the one you've marked out where all devices are also drugs. Yes. So under our view, if a product satisfy, if a product falls within the exclusionary clauses of the device definition, but otherwise satisfies the device definition, it is a drug because we regard the drug definition as encompassing the device definition. Now, biologics are also drugs under the statute, which is again, another example of an overlap that is just ubiquitous in the FECA and why it's not so strange to have overlapping regulatory category. We agree, of course, that the district court was incorrect to apply the specific governs the general canon. But I think we take that view for a somewhat different reason, which is that under this court's decisions in the Supreme court's decisions, that canon applies in two and only two scenarios. One is where there would otherwise be irreconcilable conflict in the vehicles in the park. Another says bicycles can go in the park, right? That's a classic example. The other is where you have a broad provision and a general and a specific provision applying side by side so that reading the broad provision as applicable would read out of the statute, the specific provision. So for example, in Forco, you had a broad general venue statute and a narrow patent specific venue statute. And the Supreme court said, well, if you could apply the general venue statute in a patent case, there would be no point to the patent specific venue statute. But again, neither of those is present here. There is nothing that creates conflict, nothing that creates superfluity about the FDA's interpretation, which gives full meaning to the exclusionary clauses, which reads the exclusionary clauses to mean, if you work through chemical action or metabolization, you cannot be a device. Therefore, you can only be regulated as a drug. In which doesn't create conflict. I'm a little bit confused by that. If you operated through metabolization, for example, you could be a non-drug, you could be a food. You could, but you certainly, if you otherwise satisfy the device definition, you would be a drug. So if something satisfies the device definition, the other parts of the device definition, but works, but achieves its primary intended purposes through chemical action within the body or metabolization, then it is a drug. It may also be other things as your honor has noted, but it is a drug because the drug definition fully encompasses the device definition. In other words, it's a drug regardless of whether it satisfies the exclusionary clauses. But if it satisfies the exclusionary clauses, it can only, it's not a device. It ultimately decides the question that you say the FDA otherwise has full discretion over. So the only non-discretionary sorting as between the category of drugs and devices is that which is accomplished by the exclusionary clauses. That's right. And I want to, I do want to make sure I spell out fully what the FDA regards as the limits on its discretion. So it's not just that FDA could not take, you know, all products satisfying the device definition and decide to treat those drugs. I think it's worth appreciating for a second, all the sort of rest of the you know, to read genesis brief, you might think FDA could decide tomorrow, it's going to decide to regulate all band-aids as drugs and not devices. But there are in fact, some 2000 provisions in the code of federal regulations that specify that particular categories of products are to be treated as class one, two, or three devices. The FDA couldn't change any of that without going through notice of comment rulemaking. So again, really the FDA, the authority the FDA is asserting here is extremely limited. But under your view of the statute, you could change those regulations. There would have to be a non-arbitrary or capricious reason for doing so that did not undermine the broader two track scheme. I think if FDA exercised its authority in any of the extremely broad ways that have been posited, it would violate the statutory scheme. What it is here is exercising the authority as to a limited category of products where there are good reasons consistent with the statutory scheme for treating the category of products entirely as drugs. You could make a plausible argument in any case you want that the article satisfies the definition of drugs. Drug regulation is more stringent. You rightly care a great deal about safety and efficacy and you think it's important to have more stringent rather than less stringent regulation for some category of thing about which you have concern. Respectfully, I don't think that would be consistent with the statutory scheme. Congress didn't create these as more and less stringent realms. If FDA were to reason in that way, I think that would be contrary to the statute. What is the limiting principle? You've taken off the table the one that Judge Henderson offered you, but what is the limiting principle in terms of the extent of your authority to reclassify or to categorize within that smaller circle? The limiting principle is that the FDA, the presumption is that something satisfying the device definition is treated as a device. And if FDA wants to regulate a product satisfying the device definition as a drug, it has to have a particular reason for doing so. I'm sorry, the presumption in the statute or just how you choose to exercise discretion? The presumption in agency practice, which I think follows from the statute. And so if FDA is going to override that presumption- If it follows from the statute, that just sounds like an elliptical way of saying that the reticulated scheme associated with the narrower category is exclusive for the range of things to which it applies. No, Your Honor. I think it's a way of saying that Congress's decision to create a separate scheme for devices suggests that as general matter, something satisfying a device definition should be treated as a device, or else, again, if FDA treated all devices as drugs, it would override the two-track pathway. Mr. Winnick, I think this is what's difficult for me, is that it feels like you're at one time reading all the more specific definitions in the lead-in phrase as if they really just reduce article. But then you're circling back, and that's for purposes of your jurisdiction over which track, your authority to assign them to one track or another. But then it seems like you're bringing back in those more specific terms as a guide for your discretion. In other words, could you take crutches, which are an implement, and say, they're a drug because they're an article, and they're meant to be curative or help people in getting over malady? Or can we take x-ray machine and regulate it as a drug? At some level, it's an article. Is that right, that you're on the one hand rejecting that initial language, on the other hand, finding it to still have a bearing on your discretion? Your Honor, the reason crutches, or to use my prior example, band-aids, couldn't willy-nilly be regulated as a drug is not, I think, that it is a thing that goes clank. It's that there is a regulation in the Code of Regulations that specifies that those things are divisive. Nothing about the statute. Those terms that really resonate with common sense about devices should just be taken out. They're a leftover of another time. Instrument, apparatus, implement, machine, contrivance, implant, in vitro reagent, or other similar or related article, including any component, part, or accessory. I mean, I spent a lot of time figuring out how to write that, how to write it in a broad way, to encompass a lot of things that don't seem to go clank, like maybe an in vitro reagent. They could have just stopped and said article. Let me try to explain why I'm resisting it, because I realize I am, and I realize why the court might be inclined to read that phrase differently. There's a couple of reasons. The first is, if you look at page JA332, this is the part of the FDA's guidance that talks about that phrase, that similar or related article, and explains why there's just a lot of products that naturally, that ought to be regulated as devices, and that aren't sensibly regulated as drugs, that don't obviously fall within one of those headings in the preparatory clause. For example, it talks about gels or powders that are used on the skin as a barrier, or gases used as face fillers, or as I mentioned earlier, liquids used to clean surgical instruments. The other reason is that there are consequences to being able to call something a drug that have to do with parts of the statutory scheme, other than the one we're talking about today. To give an example, if someone tomorrow tried to introduce a little metal box with the claim that it cures cancer, and the FDA wanted to prevent that obviously fraudulent device from being sold to the American public, the typical way in which the FDA would do that under its regulatory authorities is to say that because the box, because the claim is that the box cures cancer, it is a drug, and that therefore selling it without authorization violates the by the specific examples in the preparatory clause of the device definition has implications beyond the context of this case that are troubling to the agency, and I think for good reason. The other thing I just want to make sure I have the opportunity to mention is, even if the court is sort of concerned about the implications of FDA's interpretation across all categories of products, there's a very good rationale for deciding this case in FDA's favor, specifically with respect to contrast agents, which is that in the 2017 amendments to the statute, Congress legislated in a way that pretty clearly accepted as permissible the FDA's then 20-year-old policy of treating all contrast agents as drugs, which is that there was a problem called to Congress's attention by the FDA that there was a lack of specific authority to approve a device designed to use a contrast agent in a manner different from the one in the agent's drug labeling. Congress could have dealt with that problem by telling the FDA to stop treating all contrast agents as drugs, but instead it did the opposite. It wrote the statute in a way that allows FDA to treat the product as drugs, but creates a pathway for approval of the device as a device, and that's a pretty good indication, as this court has recognized in many cases, that Congress approved the FDA's interpretation as permissible. So even if the court is disinclined to reach the broader question, whether FDA can exercise this authority as to other categories of products, it should at the very least reverse the district court's decision as to contrast agents. And again, this was in the FDA's ruling letter as well, this rationale. And tell me about the United States pharmacopoeia and the national formulary and what role their characterization of, for example, contrast agents as drugs plays, if any. So I may be misunderstanding, but I don't think there is any import to any characterization of contrast agents as drugs by those entities. The relevance of the 2017 amendments is not simply that they characterize contrast agents as drugs for the purpose of applying that sort of particular framework for approval created by the 2017 amendments. It's that Congress chose in 2017 to address this problem in a way that accommodated FDA's decision to treat contrast agents as drugs rather than simply telling the FDA, you know, don't do that anymore, which it could easily have done. Right, although it may be that those with an interest in this issue simply weren't at the table. Well, I don't think that's accurate, Your Honor, because, again, Congress wasn't doing this in the vacuum. It was, and this is clear from the committee before we cite, it was addressing this problem in close coordination with the FDA because the FDA had called this problem to Congress's attention. So I think it's pretty improbable that when Congress wrote those provisions in 2017, it didn't know that FDA had for 20 years been treating all contrast agents as drugs. That was the thing that created the problem, at least as to contrast agents that also, you know, satisfied the device definition and could have been regulated as devices. All right. Do my colleagues have any more questions? All set. Thank you. Thank you. All right. We'll hear from Mr. Farquhar then, and we'll give you a couple minutes, Mr. Winnick, and reply. Thank you, Your Honor. Thank you, and good morning, Judge Henderson. Good morning, Judge Pillard and Judge Katsas. Doug Farquhar, representing the Appellee Genus Medical Technologies, may please the court. I'd like to begin actually by addressing the question that Judge Henderson started with, which is, how can this device, how can this medical product be a device? And it will also answer, I think, some of Judge Pillard's concerns about whether something could be a food or a cosmetic and a drug. There is a threshold determination that is made using the intended use portion of the medical device definition and of the drug definition, that if the product is intended to be used, in this case, for the assistance in the diagnosis of a disease, it is a medical product. So if a product is intended to be used to diagnose, treat, cure a disease, or prevent a disease, then it's a drug, and a drug or a medical device, depending on the second step of the analysis. All biologics are drugs, so there's no separation between those. Biologics are a subset of drugs, and that's made clear in the statute and in the regulations. The second step of the process once a product has been determined to be a medical product is, is it a drug or is it a device? The device definition, as Judge Katsas was explaining and explicating, is very, very clear. It is a device if it does not achieve its primary intended purposes through chemical action in or in the body or through metabolic action. If it does achieve its primary intended purposes through metabolic action or through chemical action, it is a drug. That's what's left. There's a universe of medical products. Some are devices, and those that are not devices are drugs. That's the way the statute read before 1990, but Congress got rid of the exclusion from the drug definition of everything that's a device. I understand that the immediate task there was to open the way for combination products, but that clearly could have been done in a way that excluded devices that are not combination products. Well, Your Honor, it's more than part, the FDA characterized that as being part of the reason that that clause was excluded, but if you read the legislative history on that point, it is quite clear in the legislative history. First of all, the Congress viewed this as being a slight alteration in the definition of drug. That is in the Senate report. I can give you the site if you'd like, but the Senate report on the drug. It is just before the part, the first time that Congress says this is a slight alteration of the drug device, which is intended to permit combination products, and combination products have a component of a drug or a device, to be regulated as a drug if the primary mode of action is that of a drug, that is, it achieves its primary intended purposes through chemical or metabolic activity. It just seems puzzling why Congress would have removed that definition. I mean, you've reminded me that because the choice as between the device track and the exclusion even needed if there's a third category that Congress creates and says, if it's combination, we don't automatically include it in device just because it meets that definition, nor do we automatically include a drug because it meets that definition. Instead, we call on FDA to determine its primary mode. No need to take away the limitation on drugs to those that operate through chemical or metabolic functions. Your Honor, the language that was removed from the drug definition in 1990 by the Safe Medical Devices Act had been in the statute since 1938. At that point, it was unclear until 1976 after the Bacto-Unidis case, it was unclear exactly where the line was drawn between drugs and devices. Quite explicitly, the original Medical Device Act in 1976 quite explicitly said, we need to, as Congress, we need to set up a clear line between drugs and devices. They left in that earlier language, but it really almost became surplusage because they said, here's the fence. If you're chemical action, metabolic action, you're a drug. If you don't achieve your primary intended purposes through that mode of action, you are a device. But Mr. Farquhar, the first paragraph of the MDA has the language that you seem to think doesn't have to be complied with. That is the instrument, apparatus, implement, machine. What purpose does that serve? Just to try to describe without being exhaustive some of the more common devices? Your Honor, originally when that language was in the statute, that was the only definition of what about how could this product be a device? I would turn to, first of all, the FDA in its current language, the language that's in the briefs, talk about how it appears to meet the definition of a device. In the district court, the FDA actually admitted that it does fall within the definition of a device. It said in the joint appendix, let's see, I'm trying to find it. It's 77 and joint appendix 104. But even if you go further back, if you go into the citizen petition response in 1997, that was part of the administrative record, the FDA actually addressed this very because Judge Henderson, one of the parties in that matter, put into their citizen petition, how could a contrast agent possibly be a device? And FDA correctly, in our view, said it is because it is an accessory to a diagnostic procedure. It is a device. Barium sulfate, like those ultrasound products, is an accessory to the diagnostic procedure of taking x-rays or CT scans. Now, to get back to your question now, why is that language there? The language is vestigial, but it is quite broad. If you read it, yes, it talks about device contrive and so on and so forth. But it also talks about similar or related articles. Article is actually the same way that drugs are described. They're articles. So it's really what the distinction is once you have a medical product. In other words, once the use is intended to treat, diagnose, cure disease, it is a medical product. And then the distinction is either does it through chemical action, it's a drug. If not, it's a device. And if I could, Mr. Winnick actually, I don't think he did this on purpose, but he said FDA could switch what's a device and what's not through regulation. That's not true. If you look at 21 USC 360C, it specifically says medical devices will be class 1, 2, or 3. And these are how we make you make the decision, FDA, as to whether it's class 1, 2, or 3. Well, Mr. Winnick might not say, well, that doesn't say all medical devices. But if you look back at the legislative history of the Medical Device Act in 1976 that resulted in this classification system for medical devices, the legislative history, as Judge Boasberg said, says all medical devices fit into one of these three classes. So you can't excuse a medical device from the class by regulation. The statute says if it's a medical device, which it makes perfect sense, is because it's a medical device, because it meets the definition of a medical device, has to be classed in this particular manner. So- I mean, my understanding of FDA's response to that is once they choose to treat it as the statute permits them to as a drug, then it no longer has to fit into 1, 2, or 3. It becomes a drug and only a drug. That's correct. I wanted to make sure, though, that it was clear that it's the statute that requires it. What FDA claims is we have the discretion to regulate, to classify and regulate a medical product as a drug because it meets the intended use portion of both the definitions, the drug and device definition. But I want to turn now to the reasons that they don't have that discretion. First of all, they can point to nothing in the statute that says, where Congress said, we want FDA to have that discretion. The only thing that they point to is the removal of the clause, but does not include devices or their components, accessories, parts, et cetera. That's a pretty big deal. On his face, it seems like a pretty big deal, but they did not touch the structure of the definition of what a device is or how a device is required to be regulated and classified. Also, so if it's ambiguous about what they intended by that, you're permitted to look at the legislative history. And the legislative history discusses this particular change in two different places in the legislative history. Both times it says it is intended to permit combination products to be regulated as drugs under appropriate circumstances, combination products. It's focused on combination products and our product is not a combination product. Then also they, again, they say just before that point in the legislative history, this is a slight alteration. It would make no sense for FDA to make such a massive change in a scheme, a dual regulatory scheme, as Judge Boasberg described it, that is so important that in 1976, the Congress set forth these very specific rules about how you have to define medical devices and how you have to represented the holding in RADLACS and the language in RADLACS. In the final brief that FDA filed, it said the narrow versus general or specific versus general canon is restricted to these particular circumstances. That's actually not what RADLACS said. What RADLACS said, what Judge Scalia said in that decision is it is perhaps most frequently applied in those kinds situations. But if you look at the other language in RADLACS, it fits quite specifically the situation we have here in terms of narrow versus specific. In particular, it quotes the Chase decision from 1890. Even in 1890, this is what the Supreme Court said about it. It is an old and familiar rule that where there is in the same statute a particular enactment and a general one, which in the most comprehensive sense would include what was embraced in the former, the particular enactment must be operative. This language is still valid. It was recently cited, I think it was in the Eagle case, if I remember right, or in the Eddie Rondak case. It is still operated that when, especially when, and this is what Judge Scalia said and what has been repeated in more recent cases, where, especially in the same enactment, here, G, subsection G, subsection H1, same enactment, right next to each other, where there is a specific definition or specific provision, and then there's a general provision, when the conduct at issue falls within the scope of both provisions, the specific presumptively governs whether or not the specific provision also applies to some conduct that falls outside the general. So it's clear that the narrow or specific versus general canon is not restricted to the circumstances that Mr. Winnick has argued. I would like to go back to, in terms of the device definition and why it is such a clear line that's drawn by the statute, a line that still remains. And the reason is because, as I say, it's a universe of products that are medical products. Medical devices are specifically defined. What is not a medical device, by default, is a drug. Makes perfect sense. It's implicit. But as was held in the Eagle decision, implicit, if I can say, just quote that case, if the text clearly requires a particular outcome, then the mere fact that it does so implicitly rather than expressly does not mean that it is Chevron sense. I see my time is up. I did want to mention just a couple of other things, if I might, Your Honors. Thank you, Judge Henderson. One of the things that you may need to get to if you don't believe that this case is resolvable on Chevron step one is whether the decision that Mr. Winnick is advocating for here, the decision that our product can be regulated as a drug, was an arbitrary and capricious decision. And I would like to point out that that decision relies on two things. The other case is the Bracco decision that came out of the district court. That response to the citizen petitions, as well as the government's view of that case now, distorts what the holding in that case actually was. Of course, it wouldn't be controlling on this court anyway. But what the holding in that case was, there were two holdings, really. The first is you have to treat similar products in a similar fashion. That is a holding I've actually argued on. It's a correct holding. I've cited that case many times for that proposition. But then what it said is, these appear similar to me, Judge Friedman said. It is likely that FDA has the discretion to regulate these products either as a drug or device. And that is not a decision. What he actually decided was, and instructed FDA to do, was either treat these products the same or explain, give me a rational, give me the reason why you're not treating them the same. So it's not the holding that there was this discretion. I don't think any court has ever that there's discretion for FDA to classify a product that meets the definition of a device as a drug. And it would be contrary to the structure, the text, the purpose, and the history of the statute. I'd be happy to answer any other questions. I have a question on H1. And how is a device recognized in either the formulary or the pharmacopeia? What does the recognition involve? Yes, Your Honor. I will try to answer that. There actually is, the United States Pharmacopeia historically is a non-governmental institution that has monographs for particular drugs and particular devices. And what FDA has, or what the statute has said, and this is because this has been historically in the statute forever. Because back in 1938, the U.S. Pharmacopeia and the National Formulary, which is part of USP, were the Bible in terms of what was a drug and what was a device and whether it had to meet these particular requirements and specifications in order to qualify for listing in the USP or for using the USP label. These products, that is not an issue here. This particular product, barium sulfate, used as a contrast agent for radiological health would be, in fact, superseded by the statutory definition anyway. But I'm confident that if it had been listed in USP as one or the other, one of the parties who's before you would have recognized them and brought it to your attention. So is it recognized or not? I'm not sure, Your Honor. I don't know the answer to that question. Well, but it's supposed to be. No, Your Honor, I'm sorry. Oh, it's an alternative. Well, now, wait a minute. It says, one, recognized, or two, intended for use, or intended to affect. So how do you read the or comes after H2? That's correct. But if you look at the way that the statute is indented and so forth, I'm looking at the version that's in the government's original version. So they would? The way it's indented, it's one, two, or three. In other words, you don't have to put the or after one. I see. I see. Okay. Thank you. Any more questions? All right. Mr. Winnick, why don't you take two minutes? Oh, go ahead, Judge Pillard. I just, I think you answered this, Mr. Farquhar, but I just wanted to confirm it. I heard Mr. Winnick to say that he's not attracted to the suggestion that the prefatory language that you described as somewhat vestigial is actually playing any role in distinguishing devices from drugs. And I thought I heard you to say more or less the same thing. That's correct. You wouldn't see that as operative in addition to, in your view, or in contradistinction to, in Mr. Winnick's view, the metabolic and or chemical action exclusion. You both think that the lead-in language doesn't really play an operative role anymore. I'm glad you asked that question, Your Honor. And if I might give just a couple of examples, I think it would help the court understand. We both agree that that language in terms of articles in the drug definition and instrument, apparatus, implement, machine, contrivance, implant, blah, blah, blah, is very, very broad language. It is the intended use of the product that makes it a medical product, and it is the chemical action that we contend, in this case, makes it a device. Now, it seems sort of counterintuitive that something that you swallow would not be a drug. And that's precisely why this explanation or why this definition is in here on chemical action setting up the fence. So, for example, we mentioned in our brief, there is a device that you swallow that is an alternative to a colonoscopy that passes through your system and takes pictures of your GI tract as it's working its way through. The fact that our product is swallowed, the fact that it's liquid, does not make it necessarily a drug. It is a device because it is not achieving its primary intended purposes through chemical action within or on the body of man or being metabolized. It is inert. It does not chemically interact with anything. It passes through your system. It's excreted. If I might take just a quick second, I actually, my daughter is a pediatrician, and I was talking to her about this argument the other day, and she said, oh, we just used barium sulfate today. We had a child, they wanted to check motility in the child's digestive tract, so they gave the child scrambled eggs with barium sulfate in it, and then they watched through the x-ray as it passes through the system. So, that's not a drug-like effect. That is a device-like effect showing, as it goes through the tissue, showing whether there's deformities in the tissue and showing how it tracks through the system. I appreciate your allowing me to divert into that little anecdote. I don't see why it isn't a drug-like effect if both of them are used for diagnosis. Barium sulfate, what you just described, was diagnostic. That's correct, and devices specifically... Which is what a drug has to be. No, Your Honor. No, Your Honor. I would disagree with that because it is, if you look at the number two, it's intended for use in the diagnosis of disease or other conditions. This is, I'm sorry, H2 under device. Devices can be used and frequently are used, mostly are used probably, in terms of helping to diagnose diseases. But, in fact, the same term about diagnosis applies in both the drug and device definition. That's the more fundamental. Right. That's what I was saying. Just because it's diagnostic doesn't mean it's not a drug. That's correct. It's the fact that it does not achieve its primary intended purpose. Thank you very much. Which relies on the exclusionary clause, which isn't included in the definition of drug. That's right. But, again, the corollary of that clause is that the things that do achieve chemical action, their primary intended purposes through chemical action, and that are medical products, they are drugs. That's what's left. Thank you. We would ask that you affirm the district court decision. All right. Mr. Winnick. Thank you, Your Honor. Five points, which I promise are quick. First, as I noted at the beginning, Chevron deference applies under Mylan. So the FDA wins unless its interpretation is unambiguously foreclosed or unreasonable, and it's neither of those things. But I thought you relied only on clear language and didn't rely on any exercise of discretion in this case. So we think the FDA properly exercises its discretion to treat this product as a drug. But in the reading of the statute, you think it's clear? We think the statute leaves it open to the FDA to treat a product satisfying the device definition of drugs that Congress hasn't clearly required the FDA to do otherwise. Second, as to RADLAX, we don't mischaracterize that decision. The decision says the canon is most frequently applied to avoid a contradiction, but it is also properly applied to avoid superfluity. Our point is that there is neither a contradiction nor superfluity. Third, the question before the court isn't what Congress had in mind in 1990. The question is whether there is a basis to read into the text of the drug definition a limitation that isn't in its text. And the relevance of what happened in 1990 is simply that it would be all the more incorrect to do that where Congress expressly removed the limitation 30 years ago. I think this is a proof—oh, go ahead. But you're fired out, and I'll ask you the question. I'll remember it. Fourth, aside—even if this court weren't to rule on a basis specific to contrast agents under the 2017 amendments, it should have no concern that if the FDA were in the future to apply the authority it's claiming here more broadly than it has done in 30 years, that the court won't be able to confront that. Manufacturers can easily bring APA challenges to any future decisions that go beyond the narrow terms the FDA is defending today. Finally, I don't think it's open under Chenery for the court to decide this case based on the preparatory clause, but the FDA would regard that as highly disruptive to its regulatory authorities. So to sum up, the district court's decision is wrong because it reads into the statute exactly the language Congress removed 30 years ago without any compelling rationale in the statutory structure or purpose and in a way that upsets 23 years of settled industry expectations. I'm sorry, Judge Pillard. I said I would remember. You said there's no concern here about superfluity, but I think the questioning that you heard from the bench in your argument in chief was why is your reading not rendering the device definition superfluous insofar as everything that's a device is also a drug? And our answer is that the FDA might do that if it were just deciding to treat every device as a drug, but it's not doing that or even close, and it's not asserting the authority to do that. I guess the question is, what's the frame for superfluity? Is it in the scheme that Congress is mapping, or is it in the world as the FDA exercises its power? And I thought that it was the former, that the superfluity should be identifiable from just reading the statute without knowing how the FDA is going to act. But maybe your view is, no, we have an option, and so it's not superfluous because it's giving us an option. Yeah, I think superfluity is a fairly formal analysis. Is the language in the statute given full effect? And our point is that the exclusionary clauses of the device definition and the device definition as a whole is given full effect in FDA's interpretation because it specifies what the FDA cannot regulate as a device. If something falls within the exclusionary clauses, it must be regulated as a drug and not a device. What's your best answer to Mr. Farquhar's point that this was something that Congress thought of as a small change? And indeed, if it had intended to throw a lot of discretion into FDA's court, wouldn't you expect that there would be some discussion of what guides that discretion? So I think a few points. One, it has been a relatively small change. Again, this hasn't come to the court in 30 years because the FDA asserts this authority very rarely. Second, the Congress did put in the statute a means for the FDA to exercise discretion in classifying products as to which it's unclear whether they should be treated as drugs or devices. That's section 360 BBB2. And third, I find it instructive to sort of pretend for a second that 1990 had never happened, pretend the statute had always read exactly as it does today. In that world, it would be just as true as in this world that it would be improper to read into the drug definition a limitation that isn't supplied by its text. So the only additional point that 1990 gets us is that it would be even more unwarranted to do that where Congress removed the same limitation when it was expressed in the statute. So, you know, to all of Mr. Farquhar's points about what the legislative history does or doesn't suggest, even if you think, you know, Congress might not have meant to do this, you know, the only effect that would have is that you put less weight on that additional reason we win. It doesn't go to any of our more fundamental points about the new propriety of reading into the drug definition a limitation that simply isn't in its text and that is not necessarily implied by the structure or purpose of the statute. You mentioned that the, I mean, I find, I do find it notable that it's been 30 years and this is the first time this issue has come before us. Makes it seem like it isn't a dramatic change. But then when you say the FDA asserts this authority very rarely, I almost find that to be an incomprehensible sentence in view of the way you read the authority because every time it regulates a device as a device, it exercises this authority in your view. Well, except that in practical terms. There's no statutory meaningful distinction other than the biochemical or the chemical and metabolic. Carve out. So the way I see it, Your Honor, is this. From Mr. Farquhar's telling, you would think that every time a manufacturer goes to create a new product, a band-aid or a crutch or whatever else, they are, you know, sitting there in a state of uncertainty about what FDA is going to do. But the reality is that for the overwhelming majority of products, you know, it is specified in the CFR what, you know, what device class the product falls into if it's a device. So the authority FDA is defending comes into play only in a narrow, limited number of cases where there's not some regulation on point where, you know, it's a historical matter. There's been some ambiguity in how the product is to be treated. We've identified sunscreens, contrast agents is one. But this is not some large list of things. And, you know, even as to products in this category, again, it's it's it's it's I think rings hollow to suggest that even contrast agent manufacturers, you know, are sort of hurt by some uncertainty about this. When, as Gizgidimikas brief points out, it's been perfectly clear to contrast agent manufacturers since 1997 how they were going to be regulated. And what Genus is seeking here is an exception to that consistent, longstanding and perfectly clear policy, which, for the reasons we argue, is sensible and should be upheld. For those and the you resist the. So I had thought that part of that there was a distinction between what's included in the device definition, which distinguishes it from the drug definition in terms of of machines and implements versus what's excluded from the device definition, which isn't necessarily included in the drug definition. But I think I had been operating under what may be an incorrect view of a multipolar FDA regulatory regime that something could be a drug or a biological. But now that I've been corrected that all biologicals are drugs and where you're talking about a cosmetic, the statute's explicit about how to deal with the overlap. And in fact, so so if it's if something that's excluded from the device definition isn't thereby defined as a drug, why not as a practical matter? Does that make sense? I know that's a little bit of a jumbled question. No, that's right. I mean, so so I think if you were to regard the device definition in the abstract, the thing that makes it a drug is not that it is excluded from the device definition. The thing that makes it a drug is it falls in the drug definition. Right. And whereas it might otherwise overlap with the device definition, there is no such overlap if it's in the exclusionary clauses. Does that does that make any sense? Yeah. But doesn't that just reinforce the district court's reasoning? I mean, you're saying no, because everything that's in the device definition is in the drug definition, and therefore the fact that it is not excluded doesn't mean it is exclusively a device, right? Yeah, that's right. I mean, the core of the district court's reasoning, to boil it down, is Congress could only have intended for these to be mutually exclusive categories. They have to be mutually exclusive categories. And our point is, that's not right. It's certainly true that if FDA treated every device as a drug, it would eviscerate the two track scheme and that would be impermissible. But there's nothing at all inconsistent with the statute, or frankly, unusual in the context of the FDCA, for the FDA occasionally to decide that something that satisfies both definitions should be treated under one track or the other. Can I just try one more time? Tell me what is the legal principle in the statute that limits your asserted power to reclassify such that you could not treat every device as a drug in the statute? The FDA has to have a good reason for treating as a drug a product satisfying the device definition. And it has to be consistent with the statutory scheme. And it would be inconsistent with the statutory scheme if the FDA exercised that authority as to every product satisfying the definition of a device, because as Rana has noted, Congress clearly envisioned that there would be two distinct tracks. And if any manufacturer in a future case thinks that FDA has exercised that authority impermissibly, it can easily bring an APA challenge and this court can set limits on the FDA's discretion. Thank you. All right. Your case is submitted. Thank you.