This episode is audio from the Neglected Hatch-Waxman Act Goal of Data Transparency panel from the Engelberg Center's Hatch-Waxman at 40 and Beyond Symposium. It was recorded on September 26, 2024.
Rebecca Eisenberg, The University of Michigan Law School
Richard F. Kingham, Covington & Burling LLP
Reshma Ramachandran, Yale School of Medicine
Chris Morten, Columbia Law School (moderator)
Announcer 0:00
Welcome to engelberg center live. A collection of audio from events held by the engelberg center on innovation Law and Policy at NYU Law. This episode is audio from the neglected hatch Waxman, goal of Data Transparency panel from the engelberg Center's hatch Waxman at 40 and beyond symposium. It was recorded on September 26 2024
Christopher Morten 0:25
Hi everyone. I am Chris Morton. I am a clinical professor at Columbia Law School. I'm really so delighted to be here. Thank you, everyone. Thank you. Rochelle and Scott. Michael, Katrina Nicole, everyone involved in creating this event. I'm I'm honored to moderate this panel entitled The neglected hatch Waxman act. Goal of data transparency. We have three eminent panelists, Professor Becky Eisenberg of Michigan, law school, practitioner Dick Kingham of Covington and Burling LLP and Professor Reshma Ramachandran of Yale. Medical School. I won't read their incredibly impressive BIOS because they're printed in the program. Michael made the point. The title of this panel is a bit provocative. It's called the neglected how to ask me that goal of data transparency, and Becky dick and Rachel will talk about, first off, what data transparency, of safety and effectiveness data actually means, and why it might matter. And then they'll get into the question of whether this sort of transparency was really ever a goal of the hatch Waxman act. And we'll talk then about kind of more broadly, perhaps, about whether data transparency at the FDA is working well, and whether they have proposals for the agency. Becky Dick Grisham and I have worked out, I think, four prepared questions to spark the conversation. I'm going to ask those four questions and then let each panelist respond in turn a couple minutes a piece, and then we'll save some time for open Q and A at the end, I think, is the provision on screen. I can't actually see it, but yeah, okay, so because this provision is so obscure or perhaps neglected, we thought it might be helpful to actually put it up for folks to see. If you can't see the screens, I think it's also printed in the in the printed materials. This is section 104 of the hatch Waxman Act, which is today codified at 21 USC, section 355, l1 in a few minutes, I'll ask the panelists more about this specific statutory text. I have just one minute of framing that I thought might be helpful to step back from this specific text and look a little bit about the big picture, or how this panel fits into the other panels, the other programming today and tomorrow. This is a panel that's less directly about competition and pricing than some of the others, less about the balance of power between brand name companies and generic or biosimilar companies, and perhaps more about the question of drugs value, their therapeutic their clinical value to individual patients, and their aggregate social value to our economy and to our society. There are some prominent medical researchers, including Reshma, who have raised alarms about certain drug safety and effectiveness, especially as the FDA has, in recent years, and for various reasons, been approving drugs faster and on smaller and smaller evidence bases, of course, we need drugs to be accessible and affordable, but we also need them to be safe and effective for the patients who take them. And we might argue perhaps that a lack of understanding of drug safety and effectiveness creates an economic and pricing problem we have, perhaps not just a crisis of drug pricing, but a crisis of value. We pay a lot for drugs that may have little or no benefit. Figuring out what a fair price for a drug is requires us to know a lot about the drug safety and effectiveness. So I think we'll be talking about that today. I will knock it off then and moderate I'm going to pose My first question to the panel. We'll be rotating the question order for this first question, I think we're just going to go down the row. I'll ask Becky to respond first, and then dick and then Reshma. So question one, what exactly is safety and effectiveness data and information about prescription drugs. How does the FDA use it? Why does this data and information matter? Okay,
Rebecca S. Eisenberg 4:18
so I think of drugs as information products. Basically, a drug combines a chemical with information about what happens when you use those chemicals to treat patients. So data about safety and effectiveness is what turns a chemical into a drug by providing the information that's necessary to use it, advisedly to treat patients, data and information about safety and effectiveness is what distinguishes Coumadin from rat poison, right? The same molecule might be available, but you don't know what. Good it is until you do testing and provide information about its safety and effectiveness. So that's the primary value of information about about safety and effectiveness, and it's considerable value. And some disclosure is necessary in order to realize that value, right? You can't just say, try, you just had a heart attack. Try this rat poison, no, when we take it, you have to say, look, we've done some clinical trials. This is what we found. Doctors and patients need to know what to expect in order to use drugs advisedly, and at least in theory, more data and more information disclosure about safety and effectiveness should allow them to make better choices. Another value of information about data and about another value of information and data about safety and effectiveness is that you need to submit that information in order to get an NDA approved. It has regulatory value and in the world before hatch Waxman, both both innovators and competitors, needed to submit full reports of investigations which have been made to show whether such drug is safe for use and whether such drug is effective in use. And so that gave the data and information about the effects of a drug competitive value in forestalling generic competition. Hatch Waxman fundamentally changed that value in the US market in two ways. First, it created these periods of regulatory exclusivity before generic competitors could free ride on the data submitted in an NDA. And second after that point, they permitted free riding, so they so the data was no longer the key to realizing that competitive value or preserving that that competitive value. In effect, hatch Waxman created a time limited property right in regulatory data, sort of like a patent in regulatory data. I don't want to overstate this analogy, but that makes data secrecy unnecessary to protect this particular bit of competitive value in the US, and presumably that is why Congress enacted the language set forth in the slide directing FDA to make the data available to the public once a generic version of the product has been or could be approved. Now drug companies tell me that data secrecy still has competitive value in some foreign markets. That is the secrecy of data submitted to FDA that might otherwise be used in order to gain regulatory approval to market a competing version of the same product in a foreign market. And maybe that's so. I don't know that there's anything in the statutory language that indicates that that's what, what, what Congress had in mind. But the data have other value as well. That data may be valuable to other firms, including generic competitors, because of what they can learn from it that will help them understand their own products, better, identify new uses or markets for products, avoid having to repeat costly clinical trials to learn things that their predecessors have already figured out. But I think I'm probably over my three minutes now, so I will pause.
Richard F. Kingham 8:27
Okay, well, let me be maybe a little more granular. First of all, I think for our discussion today, it's going to be important to distinguish among different types of data that are submitted to FDA for the approval of a drug. First of all, most people here probably know that drugs proceed in three phases after they've been found to be safe enough to be used in humans, phase one, two and three. And three are the pivotal trials that establish the safety and effectiveness of the drug. There are numerous other studies that are done, for example, special studies in old people or people with dysfunctional livers or kidneys or various things like that. So that's the universe of things that are going on in the clinical side. Come back in a second, because there's a non clinical side as well. Now, what are the documents that FDA has access to and that guide the process? Well, the first one when you start a clinical trial is to have a protocol which describes how you're going to do the study, what the endpoints will be, and so forth. Protocols are actually quite detailed, and they also get amended repeatedly throughout the life of a clinical trial. So you've got that. The second thing is, you eventually have, not usually at the beginning, but toward the end, a detailed statistical analysis plan that tells you how you're going to analyze the data you get from the trial, and what will be a win and what won't be a win, and what things other than the main point we're going to be looking at when we when we apply the statistics now, the actual results of the study then comprise several things. If. First, there's a clinical study report, and there's actually an agreed format for that under international norms, almost every major agency now accepts the same form of clinical study report. It's a narrative description of the outcome of the trial, and it usually has a lot of attachments and annexes dealing with specific subsets of analyzes that were done and so forth. And also, of course, there's a final statistical report as well. So that's the clinical study report, okay, very important document. But underlying it are two other things. First, in the United States, every applicant submits a series of individual patient tabulations that contain every data point collected for that patient in the patient's visits during the conduct of the trial. Okay, this is enormous. Even if you do 5000 people in a trial, you have literally hundreds of 1000s of pieces of data that result. And then there's a further layer, in the United States and elsewhere, the case report form, which is actually the record of every visit to the doctor during the trial, okay, and of course, there were also tests that were done, diagnostics and so forth throughout that process. Now here's the second part of this question was, What does FDA do, or need to do with these data? FDA, of course, looks at the clinical study report. But FDA historically has done something that virtually no other major agency in the world does. It wants those raw data, the tabulations and when they and actually, years ago, they got the case report forms because we didn't have digital things when I started in 73 but today, they get these digital format sets of data with every data point that was collected. Now they may occasionally also ask for a case report form, and they actually routinely get some, but mainly they are actually reviewing, sort of what underlies that clinical study report. Very important function they're performing. I'm not aware of any other major agency in the world that does it the way FDA does okay, and that becomes very important as we discuss what some other agencies in the world are doing entering in terms of the release of their data. Now, just as a footnote, say everything I just described also applies to the non clinical safety studies that were done in order to determine that the drug was safety use in humans, and to gather other information like that. One final point, so I've told you about what FDA does with these data. For companies, of course, their raison d'etre, their reason for existing is their ability to design studies and then perform them in a way that will achieve evidence of safety and effectiveness in accordance with the regulatory agencies expect, and also to conduct those studies according to something called Good Clinical Practice. That's the biggest contribution drug companies make. There are lots of people doing basic research on drugs. There are lots of people doing clinical trials of interest in academic settings and so forth, but their sweet spot is putting together studies that will do the things that are necessary to demonstrate safety and effectiveness under regulatory terms and to do it right so that it will be accepted.
Reshma Ramachandran 13:16
All right, so I'm I'll put a disclaimer out there. I am not an attorney. I am a clinician and a primary care clinician that looks at a lot of this data that informs how I prescribe medications to patients. So we talk about what safety and efficacy, effectiveness data information looks like. You know, Dick did a great job kind of overlaying kind of what's submitted to the FDA, and just to kind of put a pin on this a little bit more. You know, there's these pivotal trials that oftentimes are used to support approval for various drugs and also other medical products, and there's also supporting trials that looks in special populations, but also populations that were not adequately enrolled in some of the pivotal trials as well. And that's become of increased importance to the FDA, with its emphasis around clinical trial diversity, to include patients across different demographic populations, and so it gives us as clinicians, real information about the patient in front of us, whether or not this drug that we're going to be prescribing them is truly safe and effective for them all. This is, like critically important, but it's against this backdrop, against this shifting landscape of what we called Total lifecycle management, and this was mentioned some of the previous panel. The previous panels before and also by Chris as well, where, over time, since hatch Waxman has passed, we've seen various sorts of bills and also different pathways that Congress has authorized FDA to use to be able to ensure more timely access to novel products, right? The idea of meeting unmet need or serious conditions, regulatory incentives to expedite development and also review of these new medical products. And with that, there's been a change in terms of the clinical trial landscape. So now we not only have pre market evidence, but we also have post market evidence. So more often than not, we're seeing. Special regulatory programs that are used by companies to actually to expedite development and review in their products, and then having obligations either imposed by the FDA or commitments that they voluntarily agree upon to actually confirm clinical benefit. So things that matter for me as a clinician, how patients feel, function, survive, importantly, and also safety as well. There's various safety requirements that are oftentimes applied on to drugs because trials are getting shorter. So imagine in gene therapy, for instance, something that's given very early on in life, often for children with rare diseases or adults with very serious conditions, and you're giving it one time, and you do a clinical trial for maybe six months, 12 months, maybe up to a year and a half, you're looking at things that's not going to paint the whole picture in terms of long terms of long term safety sort of considerations. And so FDA, oftentimes will impose post marketing requirements that requires the manufacturer to submit additional data around safety and even efficacy downstream. And this, this has huge impacts, as we're seeing now, on payers. We're seeing CMS, and particular Medicaid, taking steps to actually see how they can utilize this data, or collect data in the post market space to adjust reimbursement, but also to provide them real information around clinical outcomes that matters to their population that they're covering. So all that to say is, you know, there's a slew of data out there, and it'll be interesting to have a discussion about what was the intent of actual X Men in terms of addressing pre market data versus post market data, especially in this landscape where we're now heavily reliant on this post market landscape to actually know, do these drugs actually work? Are these drugs actually safe for our patients, especially over a long term period? The only other thing I'll just mention as well, it's kind of interesting to think about this post market landscape in conjunction with entry of generics and also biosimilars, right? Because there's kind of an overlap. We're collecting post market evidence around clinical benefit and safety for these products that are branded, but at the same time, we also have other manufacturers that are hoping to bring a competitor onto the market, right? And so it's a very shift, like quickly shifting landscape in terms of how we apply data disclosure sort of requirements actually ensure knowledge about clinical safety and efficacy, but also ensure a competitive landscape and also an informed manufacturers. This is a good bet to actually put in, you know, put in your R and D dollars, or your, you know, product lines towards right. So if you have a drug, for instance, that's approved in an expedited review pathway, there's not clear evidence of efficacy, you're waiting the post market period you have generic manufacturers are watching this landscape, wondering whether or not they might want to enter the field. But then there comes evidence later on, after they put in some investment around producing a generic competitor, or by a similar competitor, that this drug actually may not be worthwhile from a clinical perspective. So just thinking about that, I think is also interesting in the context of hatch Waxman, because I don't think it anticipated that as well. The last point I'll just mention is, and we'll talk about this a little bit later on, is that we're shifting also outside of the traditional clinical trial landscape. We have things like real world evidence, so the idea of using electronic health records, insurance claims registries, even there's even things like aI trials, for instance, that are talked about, like in silico trials that we're seeing increasingly being used by FDA as confirmatory evidence, in addition to the main pivotal trial that supports approval in the pre and post market setting. And so it will be really interesting to see. You know, we have clinical trials.gov which was not designed to be able to allow public disclosure of that sort of data or even requirements around registration for that data. How are we going to meet that challenge for these new types of data as they come on, more increasing onto the landscape to give us information around efficacy and safety for these products.
Christopher Morten 18:27
Great. In the interest of time, well, I'll say, I heard, I think, significant kind of overlap or agreement, we
Richard F. Kingham 18:36
answered a lot.
Christopher Morten 18:39
Yeah, that's what I was gonna say. I might move us officially to to the question that maybe both, maybe all three panelists touched on, which is, what exactly did hatch Waxman intend? What was the design of hatch Waxman vis a vis the F days governments of this data. So I think again, you have the statutory language on screen. It includes a phrase that says, safety and effectiveness. Data and information which has been submitted in a new drug application and which has not been previously disclosed shall be made, shall be made available to the public upon request, unless extraordinary circumstances are shown, and that's supposed to occur upon a number of triggering events, including approval of a generic version of the drug. So the question to the panelists, what does this provision of the hatch Waxman Act actually mean? What was it intended to do? And also, how has the FDA construed this language in the years since 1984 and on this question, I'll start with Dick, and then I'll ask rashma And then Becky
Richard F. Kingham 19:36
to go, okay, and the fact is, I've lived through this whole process because I started practice in 1973 and on December 24 1974 the FDA published its its final regulation, modernizing the way in which it would impulate the Freedom of Information Act. Previous to that, FDA disclosed almost nothing about the drug approval process and including. Included in that was a provision that's almost verbatim, the same as what was put into the hatch Waxman act. And that was done by my good friend and partner, who is now approaching 9090, years old, Peter Barton Hutt. And his rationale was that the principal value of the data was for the ability to obtain an approval. And once you no longer needed data to receive an approval through an abbreviated process, you you no longer your data were no longer of great value. Now it's a bit ironic, because the only a NDAs in those days were actually for drugs that had been approved prior to 1962 and there were no amdas for drugs that were being supported with all the evidence of efficacy and so forth that was required by the 62 legislation. That didn't happen until 1984 Okay, so he was actually anticipating a problem that didn't even exist at the time that Peter wrote the legislation. I am not aware of any meaningful legislation, say, for example, committee reports and so forth as to what Congress meant. And I think today, in any event, even if there is legislative history that the courts would probably just look at it right on its face, you know. And these days, you know, even though Judge Scalia is gone, I had enough interactions to know that he didn't care much about legislative history, and his acolytes don't either. So I think that, I think that we have to sort of take it on its face. That would be my view. Do you want to talk about implementation now? Or shall I do that now? Or should we do it later? FDA? How FDA has used it? Yeah, I would
Christopher Morten 21:41
say certainly, how FDA has used the language since now. And then. We'll save Question four is like, what should
Richard F. Kingham 21:47
FDA absolutely, absolutely so? So what happened, in fact, was, first of all, if you think about it, if the principal reason why people might want to get your data was because they wanted to use it in some way to get their own product approved. Then, if you make them wait until they can file an application that has no data, what are they going to do? They're not going to ask for anything. So by and large, in, you know, I did a lot of sort of a subsidiary thing that I did was foi cases, defending them on behalf of manufacturers. There were very, very few serious efforts by people to get, quote, all the safety and effectiveness data for small molecule drugs in the time after 1974 and when it did happen, I'm not aware of any case in which FDA actually disclosed all the data. What they usually did was they called upon the innovator, manufacturer and the person requesting the data to work out what would be reasonable. And normally it would turn out it was really only one or two things they wanted to see, and they would be given them, and the issue would would terminate. And as as people may have noticed if, if they looked at the background materials. There was actually a different regulation for Biologics in 74 it was based on Peter's theory that there was no such thing as a generic biologic, so all the data should just be made available under FOIA as soon as the drug was approved. To my knowledge, that never happened either. In other words, I know of no case in which all of the data and the biologics application was ever disclosed, and I was involved in the case that had the most dispute, and that was a case where a competitor was seeking the full bla and we and we actually came forward with affidavits of experts in Europe, saying if they got those data, they could file them in Europe and get an approval. And on that basis, FDA was prepared to take the position that those were extraordinary circumstances. But ultimately, it turned out that the other company only needed a couple of little things anyway, and so we just settled the whole issue. So that's where it's been for all of this time. And I think Peter Hutton now claims that he knew all about the fact that there could be approvals in other countries, but I don't think he did. Because let me just get you this. I want to give you some facts here. Do you know when the drug approval law in Germany went into effect? 1976 okay? The arts nine, replica sets, 1976 when do you think the first British regulatory system that required prior approval went into effect? It was enacted in 68 didn't go into effect till 71 the point was, we were the only country in the world that were doing real drug approvals at that time. Okay, so there was reason not to be worried about other places, because other places in Germany, for example, you just thalidomide just went on the market, there was no government review at all. So it's obviously a different world today, and we'll discuss
Unknown Speaker 24:52
that later. Great, yeah,
Reshma Ramachandran 24:55
so I'll leave Becky tales. I'm going to discuss more with the legislative history. In any issues around that. I'll speak about this from more of a practical perspective as someone who actively has sought the data as a researcher and also has worked with patients to actually find some of this data for medications, either that are in clinical trials or those that have been withdrawn from the market. Actually, I'll give a very timely example. Folks might have seen the press release yesterday from Pfizer that they withdrawn voluntarily their sickle cell oral treatment from the market, and it was, you know, huge shock, you know, for a number of folks, and especially patients sickle cell patients who were on the medication, an oral medication, that all of a sudden this drug would be withdrawn. And, you know, there's not information about kind of the safety events that they found in their studies or their ongoing surveillance. And so these are the questions that clinicians and patients are asking in terms of, what are the serious safety events that led to this withdrawal voluntarily by the company? Right? This is very different from FDA doing its own investigation of some sort and then issuing, you know, a report of the data as well. So we have, you know, oftentimes used the drugs at FDA database, which has summary reports of FDA reviews and outlines the trials that they consider in making approval decisions, along with other sorts of evidence you'll see on there. What's interesting is that you know, for new drug approvals and for new biologics that are licensed, you can find those sorts of review packages, usually within a timely manner on the drugs at FDA website. If that product is already on the market and gets approved for a new indication, it's actually much more difficult, more often than not, we have to ask the agency, or FOIA, the agency, for that information, and you would think that, hey, you know a cancer drug that's been approved for 20 indications, we should have the data around safety and efficacy and their review considerations from FDA and to be able to understand what information they use to actually make this additional approval right an expanded patient population, potentially millions of people who are now going to be taking this drug for potential indication. So there's that as well. Postmarking requirements, post marketing commitments, that data incredibly difficult to find. In the case of accelerated approval, where there's a required trial after approval, so you get early approval in exchange for a requirement from the FDA that you complete additional study to confirm clinical benefit of the drug within a period of time, there will oftentimes be at least a review document on there. But for other things like required safety studies or other sorts of required studies, even for pediatric populations, it's actually very difficult to find FDA like assessment and their summaries of clinical trial data, clinical trial information on their website and on drugs at FDA. And oftentimes, we have to use the Wayback Machine, because they'll actually remove some of those requirements and summary summaries of some of the studies from their website. So it's not even, you know, kind of an evergreen resource available for researchers, clinicians and patients as well. On top of that, you know, there is also, like, some of the historic data, there's drugs have been approved right since the 1960s 70s, 80s, 1990s the drug approval packages for those are oftentimes not on the FDA website as well. And so we have had to FOIA, the agency, to get that sort of information. So once before the agency, as Dick mentioned, they go to the sponsor to kind of indicate to them what is commercially confidential, what are the things that should be redacted. And there have been instances when researchers, including colleagues that we've worked with and Chris, who's been involved in litigation around this, where this information around a drug that was controversially approved, where there was concerns around safety and efficacy, we asked for that review information, more of the clinical trial data, information from the agency. And there's overreaction left, right and center, that you can't tell heads or tails in terms of what it was the safety and efficacy sort of information that FDA was considering when making these sorts of approval decisions. So, you know, this idea of disclosure is at least best cautious from the FDA and very deferential, honestly to sponsors who are contributing kind of the burden of evidence. And so for me as a clinician, from that perspective in terms of understanding FDA assessment of safety and efficacy, and even post marketing data, especially around whether or not this drug actually works and actually shows safety in a long term sort of perspective, it's been very difficult, I will say, to actually have that sort of data disclosure. So maybe while the intent was there, and that's been discussed some for this, there to be more disclosure around safety and efficacy. In practice, it's actually been a lot more difficult than I
Rebecca S. Eisenberg 29:28
expected. So much. Yeah, so this is, this is very interesting. I mean, FDA has no appetite for data disclosure, and this is just not a priority of theirs. And they read the language, you know, extraordinary circumstances, as if it means ordinary circumstances, right, that they, they, that they, they will just, they will disclose only if it's okay with the with the sponsor. And that's problematic for two reasons. First of all, we don't trust the sponsor, and second of all, we don't trust the FDA. And both of. Those are important reasons. I mean, this approach originated in the FOIA regulations. When FDA was it was implementing its FOIA regulations. FOIA is about providing oversight of government agencies by allowing the public to get access to the basis for decisions and and when in 1974 when, when in that FOIA context, when they were responding to public comments on on these this language about extraordinary circumstances, they said that the statute creates a strong presumption of disclosure, and that extraordinary circumstances means unusual circumstances that arise in rare circumstances that cannot be foreseen at this time and that would justify non disclosure. So, you know, they may have used the same language as as they in their own regulations, their FOIA regulations, but they immediately figured extraordinary means ordinary, and they got some buy in at the time of passage of the statute from the statute sponsors, in the form of last minute jiggering of the legislative history to say, Oh, we mean what FDA says, but what was voted on in past was the language on the slide, which is A lot broader, and I think it's a lot harder for agencies, as Dick was was saying, to get away with that sort of move today than it was 40 years ago. Especially, you know, there's nothing arcane or technical or inaccessible about the statutory language that they are purporting to interpret. And you know, even before the most recent Supreme Court term, when you know, we saw a lot of skepticism towards the administrative state. The disavowed Chevron deference is inconsistent with the Administrative Procedure Act. FDA was losing in judicial review of some of you know, you know, increasingly losing, and its statutory interpretation moves, and it has to be somewhat CR I feel some sympathy for them, and not in this context, but in many others. They really old statute, and they need to sort of constantly be interpreting this, like 1930s statutory language, in light of new problems that weren't what anybody had in mind. But that's not what this is. What's going on here. So, so, and it's especially problematic before courts that don't trust the deep state, right? So you know, and you know there may be good reasons for an agency to withhold regulatory data from public scrutiny over the objections of the of the of the sponsor. But there are also some really bad ones, and courts don't necessarily trust FDA. Reasons. If you suspect that FDA was up to no good in its review of a particular NDA, you might want to see the the underlying data for yourself. And it's not just, you know, left leaning public interest groups that are seeking access to these data. You know, mifepristone is a recent example of a challenge to FDA decision making from the right before a pretty sympathetic District Court. And there will be more challenges, and I think FDA is really quite, quite vulnerable on this issue. So I, you know, I think it's really interesting, the stuff that Reshma was talking about, about the new sources of data that make us think differently about the importance of this, of these RCTs and the gold standard data, there's a lot more sources of information, and there have been some big wins for the use of some of these alternative data sources to discover problems with drugs, but it might also suggest some other possibility for how the FDA might provide the public with access to data. It doesn't necessarily mean that you have to disclose all the cross tabs on your website, you know. So for example, after the after FDA observed these cardiovascular effects of Vioxx in Kaiser Permanente data that with a congressional nudge, they built out the Sentinel system that allows directing queries to data that aren't publicly disclosed, that are kept secret. And it might be possible to do something similar, I don't know. With respect to some of these high quality data, it doesn't have to be just the insurance records in order to provide information to the public in response to queries of data that remain in FDA coffers.
Christopher Morten 34:40
Becky, can I jump? Because I think you've just flawlessly segued us to the official question three you were just discussing. I think Reshma mentioned clinical trials.gov like in the years since 1984 there are a range of different data sharing initiatives, some public, some private, academic, and so it's not necessarily the case. Case as perhaps it was in 1984 that the FDA is the only place that researchers could go to get this data. So the official question that we hammered out is, what is the state of public access to safety and effectiveness data and information on drugs, vaccines and other medical products today, given that the FDA has not construed section 104 to mean much of anything. How are researchers or competitors or insurers or other uses of this data getting their hands on it? And a specific question is, how does FDA sort of practice of sharing data compare with its overseas counterparts? And on this question, I think raishin was going to go first and then Becky and then dick,
Reshma Ramachandran 35:41
yeah. So we talked about some of the public, publicly available databases, including drugs at FDA, which has the clinical trial or clinical summaries, or review summaries for the various products. And we have also, we talked about clinical trials.gov as well, which includes the protocols and also results information there has been FDA is tasked, despite clinical trials.gov, sitting on an NIH website as being kind of the first mover of enforcement for registration of clinical trials of any sort of drug or device that's meant to be interventional in a patient, and also for reporting and timely reporting of result information. And so you can imagine, for instance, in medical journals that we have something called positive publication bias, so we typically will oftentimes publish articles about drugs and other treatments that work, but when a treatment doesn't work, and your patient trying to figure out kind of what's out there in terms of clinical trials, like, what's the available evidence? Clinical trials.gov can serve as a resource to be able to understand, like, what's in the pipeline, what are the results of things that are that have undergone testing, that may or may not have worked, and to find information that's outside of medical journals. And on top of that, more from like a patient perspective, we have some studies going on about open access. I mean, for patients, medical journals are just not an accessible form of information as well. You pay per article. They're very costly. A lot of like, the top tier medical journals, unfortunately, are not open access and available for patients. So clinical trials.gov has been kind of this resource under the FDA Amendments Act of 2007 and with Chris, we had done a FOIA investigation to look at the state of enforcement, knowing that there was about, you know, a few 1000 trials that had missing results information despite their data being due at a certain time. You know, wondering why the agency which was tasked with enforcement was not being very proactive in terms of going after these different sponsors or these different investigators in terms of timely result information reporting, and we found basically just a lack of enforcement, even in the first step of sending like warning letters to various sponsors, and the big issue that they came back to is just capacity to be able to keep track and to be able to even send out these warning letters that is not even conclusive evidence of non compliance with the statute, but just even a suggestion that they might have not submitted their data in a timely manner. So it's been very interesting to see, like even clinical trials.gov This electronic resource where it should be very easy to upload all this sort of information, there's been huge gaps in terms of at least randomized control trials or clinical trials being available on there these new forms of evidence, though, these observational studies that we kind of alluded to, they're not subject to clinical trials.gov reporting requirements, you know, so they're not this is oftentimes, like, been a huge issue in terms of finding, for instance, registries. They're oftentimes used for, like, rare disease drugs as part of their approvals. We see them oftentimes in post marketing requirements, oftentimes by a CMS in terms of evidence generation to be able to support coverage decisions, as well. Finding information around these registries oftentimes only happens through publications that we see in citations in government documents kind of explain the rationale for why, or why or why not they consider a particular study. So even there, finding that sort of information has been incredibly difficult, just showing again that we might be behind the times in terms of what evidence, generation and availability of data should look like in particular. The other thing I was just going to mention is we, there's been individual initiatives, and part of this is because investigators, anytime you submit a journal article or submit a manuscript onto a journal website, there's a data sharing statement that you have to say yes or no to, and more often than not, everybody kind of checks a box saying, yes, if you contact me, I'm willing to share the data. No problem in there's been several studies have been done afterwards to see if any of those claims, or at least that checkbox, has actually led to data disclosure and kind of data availability from individual researchers and finding often not so basically, kind of a violation of the data sharing statement and really no enforcement to actually ensure that investigators who kind of check off that box actually follow through with making their data available. So issues around that and so as a result of that, we've had individual initiatives, including at Yale. We have the Yale Open Data Access Project, which actually works with both industries. So we have Medtronic and Jensen and Johnson as well a few other companies that volunteer. Provide individual patient level data that's de identified for their products so that it can be used by researchers on a secure portal for further study. So the idea is that they can use this data to do things like comparative effectiveness. Can re stratify the data to look at particular populations to see what efficacy and safety might look like in those populations have a natural history. So if you don't, if it's difficult to have a control arm for a particular study, you can use this data potentially kind of construct that as well. And so there's been voluntary agreement from different manufacturers. And then on top of that, NIH is now doing data sharing as well. In 2023 they passed a policy around this, requiring NIH funded studies to actually put their data into a secure a secure portal to allow for sharing amongst different investigators.
Christopher Morten 40:48
Becky, this complicated rotation
Rebecca S. Eisenberg 40:50
I may have already, sort of, you know, preempted some of my own comments here. So be I'll be be brief. There was, did you include in the materials that everybody received this university's allied for essential medicines on the report is
Christopher Morten 41:05
indeed in
Rebecca S. Eisenberg 41:06
the Salima. It is, yeah, that is online, yeah. It's really quite shocking, but very but it's but it's also very hopeful, because it seems like there's some low cost enforcement mechanisms that FDA is charged with administering, that it's not that it's ignoring, but that it could do to, just like, provide a list of people who are delinquent in their obligations to report results, for example. Apparently that is very effective. People are embarrassed by that, and then they they step up, step up, and post. And it looks really bad, particularly alongside, you know, not sharing the data that are submitted to them that they're required by statute to disclose. It looks like they don't care about their obligations to provide information to the public and disclosure of regulatory data is about good government as much as it is about good science. Although in the case of clinicaltrials.gov it's really more about just, well, it has a more mixed agenda, I suppose, providing information to the public about clinical trials and making sure that you don't correcting for selective disclosure of only the happy ending clinical trials that get published, while burying The ones that might show might be less promising. So, you know, I think FDA should be complying with the law.
Richard F. Kingham 42:33
I will respond just to that last thing. No, I read, I read the citizen. I read FDA response. And actually you're it is true that simply having a list of people who are in arrears can be done, and it's fairly easy and and FDA has those for other things, it's actually required to do it, for example, for for 5050, studies, and, more importantly, the post post approval commitments that they make and so forth, and people miss their deadlines regularly, and their names, the company names, just appear. And despite what you said, actually it doesn't. It doesn't matter very much. What the petition actually saw it was formal enforcement actions. And warning letters are formal enforcement actions. 250 letters were suggested a year. That's probably more than the entire Center for Drugs submits in connection with all infractions in a year. And of course, actually initiating civil penalty proceedings involves a major expenditure of resources in what is, in effect, a kind of a trial before Administrative Law Judge FDA, mostly does, uses its civil penalty authority for people who sell tobacco to minors and so forth. There's a civil penalty provision there, but so those were things that, as I think FDA could be expected to say they weren't going to do. But that doesn't mean that compliance shouldn't be better. It should, and I don't disagree with that at all, and naming, naming and shaming, maybe, maybe one small step to do it. Let me talk though about what's going on in other countries. Okay? Because I think there's some stuff that was put into the materials, but I just wanted to flesh it out a little bit, particularly with respect to the EU they're really, I think in the materials, there are two reports. One is about what the EU is doing, and the other is about what Canada is doing. The EU actually has a lot of stuff going on, and I want to make sure people fully understand it. But the one thing I will make clear right now is that the EU is not releasing the so called raw data, that is the tabulations and so forth for a very simple reason. They don't have them and they don't have them because they don't ask for them, and they don't ask for them because they don't review them. They don't actually take those tabulations and manipulate them and so forth. Now they are now considering whether they should do that on a more routine basis. Now my view from my end. Standing of the legislation that's been in effect at least since 2003 that both national agencies and the EMA have the authority to require the tabulations if they want to, they just haven't been asking for them. And the reason they haven't been asking is they don't have the resources to review them. They don't and they don't take the same approach to the review of an application that FDA has done throughout my entire career. A German someone from the German bee farm once said to me, in the United States, you review applications from the bottom up. We review them from the top down. Okay, all right, so, so what people are getting are not the raw data, what they have succeeded in getting through the equivalent of FOIA in Europe, and through a program that the EMA developed in parallel with that is clinical study reports. Now the clinical study reports are tradition, are redacted to a certain extent that redactions are not major. One thing that appears to me, though from some of the statements about the volume of clinical study reports that were received in response to these requests, saying they were 170 574 pages. I've never seen a major pivotal trial clinical study report that was that short. So I think what's happening is the narrative is being provided, and all the annexes and tables are not but I'm not sure about that. In any event, that's been going on now for quite a few years, and on the one hand, those data have not appeared in foreign countries, being used to support approvals and things like that that I know of. And I don't think probably they would be accepted in any real agency, because they would, they would know that this company didn't have all the underlying data in order to in order to support what it was doing. What we don't know is how many of those documents which were requested by competitors were used to gain a competitive advantage in some way. One of the original requests, when we actually had the early litigation 10 years ago or so, was from a company that was selling a product competitive to another company which just got a new indication approved, and they wanted to see how they did it, you know, they just wanted to see how they designed their trial and so forth. Now, you might think that's great. You know, as an academic, we don't think it's so great. We're more in the industry, because one of, one of our, you know, Crown Jewels is how we do our trials and how we design them and so forth. Okay, now, there are other things going on in Europe, and I won't go into all the details, but one of the things is that the pending complete revision of the framework legislation for the approval of drugs in Europe, the directive that goes back all the way to 1965 but was most recently done in 2003 and the regulation that establishes the European Medicines Agency and so forth, they're now being revised, and they're in the council right now. They'll probably take final action sometime next year, and one of the things they are definitely considering is encouraging the national agencies and the EMA to use, to get and use tabulations. Right now, there's only one agency in Europe that is even developing the capability. That's the Danish agency, which I think has about 500 employees. Okay, so I don't think they have a lot of people to throw at, you know, massive evaluations of of of these data. So if anything evolves in that area, it's going to be years. But here's another key point in the recitals to both of the pending pieces of legislation, the direct, the directive and the regulation, there are recitals that explicitly state that in connection with the decision to make greater use by the agencies of the tabulations, they must take great pains to maintain the secrecy of those data because of their concerns under the general the general data protection regulation, which, As you may know, is a much more effective privacy regime than we have in the United States, either federally or in any United States state. And one of the things is that if, in fact, it's possible to take so called anonymized data and someone is capable of connecting them back with the actual patient records, and that's normally the case, then very high levels of protection apply in addition, in addition to informed consent forms in clinical trials in Europe, a separate data privacy form has to be signed as well agreeing to the uses that will be made with the data. And people are not currently agreeing to the disclosure to the general public of their data, even in an anonymized form. These are serious questions in Europe, and they're going to take a long time to resolve. Dick. So
Christopher Morten 49:47
jumping on this, because this again, I think is a nice segue to the last two questions. What's the answer? Yeah, what is the answer? And I hear you say your summary is so helpful of data sharing in other jurisdictions. Is that it seems to me, at least that Health Canada, the European Medicines agencies, are sharing more than the FDA. They are, but they've come up with some new sort of tools to protect data against misuse. There's redaction in certain ways. There may even be screening of certain users to prevent there's
Richard F. Kingham 50:18
one other thing which which, I think if they ever, if they ever move forward more on this, but right now, if you want to draw a study down from the the website, you'd be good getting just the CSR. They actually make you tick a box saying that it's not for commercial use. Now I don't think that's enforceable. And what we were actually negotiating that, and we wanted to have third party contractual rights for us, so we didn't get that. But I'll talk in a moment, or I can do it now. Well, let me, let me maybe give Well, that's respect the holy rotation. There are quite a few things I want to say
Christopher Morten 50:56
about great, yeah. So we have one more prepared question, and then I'm sure you all have questions you'd like to ask, and I realize we're the last panel between you and the happy hour. The last question is forward looking. It is, what recommendations do you all have for the FDA? These would be disclosure of safety and effectiveness data and information that's contained in drug applications, or think more broadly, like, what is the future of law and policy governing the creation, dissemination and use of information on the safety and effectiveness of medical products, and this time, the rotation will go straight down the line. So Becky dick and then Reshma. Thanks, yeah.
Rebecca S. Eisenberg 51:32
Okay, so what is the future? So we've seen a lot of changes right in in as Reshma was was was talking about. So there's no need for me to issue part of my notes, but that, you know, we've seen a lot of progress in being able to use other sorts of data and a lot of creative solutions, creating institutions like the sensible system that I was talking about. And I think maybe, if these concerns are about competitive use of I don't know trial design or of data are real, that's all the more reason, I think, to try to implement something like that, to provide for opportunities to learn from the data without publicly disclosing the data in a way that would otherwise other that would otherwise compromise patient safety or or that would raise competitive concerns. It's a shame, though, to leave all these high quality data inside separate silos at the FDA. It would be great if you could figure out ways of aggregating the data into a resource that could be used to learn more from it, but from by researchers being able to who don't share the perspective of the sponsor or of the FDA, and who might be raising different questions, or might want to, might want to avoid asking questions that have already been resolved before, avoid subjecting new patients to risks when it's unnecessary, because that has been studied. There are data available. It would be great if you could make the data available without revealing them on the front page of The New York Times. I think that there are, there are various institutional mechanisms that might be appropriate. It might require legislation. It might require funding. It wouldn't be as expensive as setting up that Sentinel system, which required tapping into all sorts of disparate and non parallel in many ways, sources of data in order to ask them all the same question, because it's all been, you know, there's going to be some homogeneity to data that are submitted to FDA, which is all that this statutory language on the on the slide requires I talk about what's been submitted to FDA. So it's not going to be the whole world of real world data that has to be made interoperable with other other data sources. So you know that might be a powerful research resource for for doctors and for researchers, and it might respond to the even under the current statutory language without raising setting off the Dick's alarm bells. Okay,
Richard F. Kingham 54:22
so let me, yeah, okay, and we're just doing the regular rotations, yes, yeah. Well, first of all, let me, let me start with something where I have a layman's knowledge, and that's trips, the trade related aspects of intellectual property rights, which I think everybody knows. The the basic treaty was signed in Marrakesh. Was very nice place to negotiate treaties in 1995 and article 39 one of trips, requires all the signatory countries who, the last time I looked, are 150 160 something like that, including China, for example. Which doesn't obviously observe its obligations, and that article specifically requires Step Four among other things, and includes some things like pesticides and so forth. But for pharmaceuticals, there is an obligation to maintain the secrecy of data that are required for support, and that would normally be kept as secret by the applicant, the only exceptions being some major public health reason, or alternatively that they are they are disclosed in the context of something that protects against unfair commercial use. Okay and so. And one thing we have to keep in mind is that provision is largely there because of us. We put it there. I'm not the US, pharmaceutical industry, the United States government, you know, so it would look a little odd if the United States government started disclosing, you know, immediately upon approval, all of the raw data and supported applications. Okay? Statute, No, it wouldn't follow the statute you're talking about right after approval. Oh, I see my point, and by and large, that's been interpreted to connect with the idea of exclusivity periods, okay? But here's another key thing, our exclusivity periods do not depend upon the data being secret. Okay, that's an absolutely correct point that you make, but that's not true in many other countries. There are a lot of countries which say you get six years of protection or five years unless your data are in the public domain. That's specifically in the Australian law. It's the approach the Canadian the Chinese take, and numerous other countries do as well. Some may not have in their law, but it's their practice and so forth. The result is that if we and by the way, the United States and Europe are usually the two first big places where we get drugs approved, there's an obvious reason for that. That's where the money is OK. So the result of that is that if we began disclosing these data immediately upon approval, we would end up finding we have no exclusivity in any country. But here's another thing, even in countries that don't have that problem, I think if these data are legally available, that is, they haven't been stolen or something like that. The government of the United States has chosen to make these data public. Then I don't see any reason why they could not be used in other countries to support approvals. We cut that off in hatch Waxman because the data exclusivity provisions, or the exclusivity provisions in hatch Waxman, are not actually data provisions. They're simply provisions for how long it takes before you can file an abbreviated application, but in other places, and also because of 505 b2, you can't use somebody else's data without satisfying all the same conditions you would if you were filing an A NDA. No other country that I know of has done that. In other words, they have not got a 505 b2 that says you can't have legitimately acquired, not stolen, all the data from somebody's NDA in the United States, and take it there and present it. Okay, there's no 505 b2 concept. So when you start thinking about it that way, it is a really big deal to make actual public disclosure. Now, what you do in terms of allowing legitimate scientific research and so forth to have access in a way that it can't possibly be misused and so forth, maybe that can work. But we also also have to think about whatever disclosures we make in terms of ultimately protecting the privacy of the people who agreed to participate in these trials. And I don't fully understand how it happened, but I've been in several meetings, including one a number of years ago up at Harvard, on on clinical trials, where there were stories told by very bright people who were able to take anonymized, raw data and they could identify who the individuals were. And it had to do in part from biomarkers that were available in other places. We think it right now, all of us who gave our DNA to 23andme It may now be in the public domain or out in the street someplace. Well, I know. So the point is, the point is we're headed toward a bigger problem of re identification, rather than an easier one, because of all the power of the machinery we can bring to bear and all the data that are in the public domain about individuals, DNA and so forth. So this is something we have, from the point of view of innovation and encouragement for innovation, and also from the point of view of medical privacy, we have to think very hard about whatever thing we do. I'm not saying that we're going to say no to everything, but we have to think about these things great.
Reshma Ramachandran 59:56
Sure. I would just say to that last point around concerns, around. Privacy. I know this is something obviously FDA, but also several regulatory agencies are contending with because we have, obviously the advent of personalized medicine. We have car t, which is also very personalized, as well gene therapies. But I think the challenge is not to think about where we exist now, but like, what are the structures and innovation we need to actually create those sort of privacy structures that will enable us to use that sort of data safely, right? So it shouldn't be the idea that we shouldn't have data disclosure in light of our current system. It's how do we innovate our current system to meet that sort of challenge? So that's, that's the thing I'll just kind of put forward. Because, you know, the reality is like we're heading towards that sort of drug innovation landscape regardless, and not having that data available puts power in the hands of not patients and clinicians and even the regulators, but really in the hands of the sponsors who own the data as well. I'll kind of separate this in terms of recommendations for pre market versus post market, and I'll start with a post market and actually kind of piggyback off of what Becky was talking about in terms of the Sentinel system, in terms of how FDA handles safety data. In particular in the post market setting, there's proactive efforts and reactive efforts that FDA engages in to address safety sort of issues, and especially, again, as trials are shorter and shorter and we're looking at the post market setting to actually understand safety considerations for these various products. In the post market setting, it's becoming increasingly important to have availability of that data. So in the post market setting, we have safety reports. So there's a an Online Reporting System that companies are required to input adverse events reporting that around their various products. But for patients and clinicians and for the public, it's kind of whether or not you know the system exists. And if you actually ask most patients and clinicians, they don't know exists, it's called fairs, if anybody's interested, people just don't know about it. So it's very rarely actually used by the public that's actually experiencing the consumption of the particular medical product and might be experiencing those sorts of symptoms after that. FDA then does an investigation of that signal. So they pick up signals through that and also the Sentinel system, which picks up electronic health records across voluntary, volunteer health systems that provide their records to the FDA, and FDA gets it as an investigation. That investigation, though, and like, how FDA resolved those safety signals, is not known. It's not public, so we have to kind of guess. We look through like case reports, we look through clinical trials that are sometimes on clinicaltrials.gov, publications, to kind of understand why FDA resolves this. And we did a study of this over three years to look at safety signals that FDA had collected and had largely resolved, and found that we can only found evidence for about 10% of them. So that tells you like there's still a black box in terms of understanding how FDA even handles like safety information that might be critical for particular patients and also for clinicians as well. So actually, having kind of probably a change in legislation to make that publicly available. I can't see that being a commercially confidential sort of issue. And there could be probably a way to mitigate that sort of concern as well, to actually provide that information to the public, and, more importantly, to clinicians as well. And there's always the concern right like you put out more information around safety sort of issues. It could cause fear mongering. We saw misinformation happening with covid 19 vaccines, with some of those databases as well, but figuring out ways to manage that so we actually provide timely information to clinicians and patients, I think would be important. Then there's these sorts of post marketing, just or studies, just generally, ranging from clinical trials, also these new sorts of evidence that's not required under clinical trials.gov to be registered, much less having results reporting that require probably a statutory change, and the next PDUFA cycle is actually 20 years post fedda. So maybe time for an update to actually meet these sorts of new types of evidence generation that we're seeing increasingly being used by companies and also by the FDA making regulatory decisions around novel medical products. So basically a souped up clinical trials.gov or maybe a new repository for this type of information that's going to be easily accessible and also have more enforcement as well, hopefully on the pre market side. You know, I think I mentioned this. You know, the fact that we just don't even have clinical review summaries around, like supplemental approvals for new indications of drugs that are on the market is just, it's quite strange to me in many ways, but and also on the post marketing requirements and commitment side, understanding kind of how FDA is taking this sort of data around safety and efficacy in the post market period and making regulatory decisions around them is very helpful. Other countries actually do put up information on unimproved drugs, which FDA does not do. And so we're hearing from the companies, usually through a press release about, you know, they got a complete response letter informing them that their drug did not get approved. Here's a justification as they understand it, which can oftentimes be very skewed in a way. And then, as an example, folks might have been following the MDMA based drug from Lycos pharmaceuticals. It was not approved by the FDA. Had a lot of attention. The press release didn't come from the FDA or any sort of information about the fact that it did not get approved. It actually came from the manufacturer, and so the manufacturer painted a narrative about, you know, FDA making a wrong decision about not approving this drug. So that sort of information vacuum, where we don't have FDA being proactive around issuing information, around how they considered safety and efficacy data outside of an advisory committee sort of setting, I think. Is really important as well for patients and clinicians who are trying to understand kind of what's on the landscape or what's in the pipeline too. I'll just stop there. But I think, you know, this is a time right for like innovation to kind of, you know, update hatch Waxman, you know, Bada, you know, all these sorts of data sharing agreements that we have, and also to have interagency sort of coordination. NIH is doing sort of this sort of stuff around data sharing at the individual patient level. You have CMS that's interested in post market evidence generation to make coverage decisions. You have FDA that's also kind of raising these questions about pragmatic clinical trials. So why not have kind of interagency, publicly funded clinical trial sort of apparatus that actually allows for that sort of disclosure without worries about, you know, something that's commercially confidential or trade secret. I
Richard F. Kingham 1:05:45
want to let, could I just, I just wanted to something, just a concrete thing that I wanted to Sure,
Christopher Morten 1:05:49
yeah, I was just gonna say, folks want to start raising their hands, okay, but this is,
Richard F. Kingham 1:05:53
this is just directly on what Reshma was saying, because I agree with you that there, there are ways in which FDA evaluation of post marketing safety issues could be made more transparent. In Europe, there's an organization called the pharmacovigilance Risk Advisory Committee, which is a subunit of the chmp, the principal committee that makes decisions and the prac assigns a rapporteur, and whenever a periodic safety update report is submitted, the rapporteur prepares a report about what if any action should be taken in response to that, and also, of course, if in the intervening period, their expedited reports and so forth, that the rapporteurs report and the final decision of the prac are all public, okay, and and obviously the companies know that It's going to be public in the process and so forth. And I've worked with people on that process. I think it works quite well, and it is considerably more transparent.
Christopher Morten 1:06:53
I think I saw a few hands I see. The first hand I saw was Lisa, and then I think Aaron, and then Sean, and then Artie.
Unknown Speaker 1:07:02
Great to see. Interesting. I've
Christopher Morten 1:07:03
missed anyone. Please keep your hands up
Speaker 1 1:07:04
so you could help me understand more about the kind of political economy around data transparency. And I mean, it seems like the people arguing for transparency and greater disclosure here it's primarily academics and public interest organizations like the great work from Chris's clinic, and like, are there other people who do or could have a financial incentive? And I'm thinking of like, Becky's work with Nicholson on payers and like, if the value of this data is primarily in helping people make better care decisions and understanding the risk better. Like, could it be that some payers, either CMS or Kaiser or or competitors, who the products competing with? Like, are they arguing for greater disclosure? And if not, why not? And are there ways, like align the incentives better to Well,
Richard F. Kingham 1:07:47
I mean, in the sense only in the sense that I've been following this. You know, as long as it's been an issue, I've never seen groups like that weighing in, or anything like that. Basically, the for a long time, it was, you know, it was foi requests from competitors and things like that. But there's not too much of that anymore. It's some in Europe, but I think it's largely been medical advocates, you know, in other words, people in academia and other places who advocate generally in support of transparency, which is a good thing to do against companies that, you know, after all, spend hundreds of millions of dollars to develop these data, and you know, you're going to put their, you know, the baby right out there in front of everybody, and so forth. So I think that's where the friction is. And we've had changing opinions within FDA, you know, when Josh Sharfstein was the deputy commissioner and so forth, they had a lovely, very well written report about how they should be completely transparent in everything. And of course, then it was deep six as soon as they left, and nothing's been done. And it also totally lacked an understanding of the legal problems of just making everything transparent. I mean, it read so nicely because they didn't confront any of the problems when they made the report. But, but I don't see this as some, some other nefarious group that's trying to keep the status quo. I could be wrong. I will just
Reshma Ramachandran 1:09:13
say that there has been increased interest from payers. I mean, part of it is also because of what happened with aducanumab in particular, and kind of the attention around coverage with evidence development. And then you development and the need to have more information for those particular populations, and again, in the post market setting. But when even constructing the registries that happened, CMS said, Well, we're just do it ourselves, because we want to have full access to the data. And when talking to the other entities that were developing their registries, there was no guarantee that they were going to be able to do data linkages between the registries to actually look into the aggregate around safety and efficacy sort of issues. And so they thought, at least we have the most robust registry that takes five minutes to complete, that will guarantee us a robust data set to make coverage decisions. And so we've seen that increasingly. And then now the CMS Innovation Center now with the like gene therapy. Model. So the idea is an outcome based rebate. And so they're negotiating with manufacturers around post market evidence generation and the specific outcomes that will be collected there. And again, this, there's a concern, I will say, that around like, who's going to actually control the outcomes, like the evidence generation apparatus across different states and the capacity to do so, and they've also been in negotiations with, like, some proprietary registries, including from the American Society of Hematology, in terms of, like, how do we actually collect this data, and can we get full control of it and be able to, you know, have access so we can actually have a comparative group compared to what the manufacturers might be providing us in the treated population. So I think there is increased interest driven by folks at CMS were there making these difficult coverage decisions, and also now that we have, like, a commissioner who's talked a lot about pragmatic trials and has written in medical journals around the need for payers to be involved in this, hopefully, more of an impetus for that those sorts of voices to kind of come in to say, like, we can't make determinations around what's reasonable and necessary for A beneficiary population, unless we have access to this data. And I think I'm hopeful at least that we'll see more of that coming down the
Christopher Morten 1:11:08
pipe. We're gathering an incredible queue, actually. So I'm going to ask our distinguished panel to keep responses quick so we can get our sorry. No. Thank you. There's I love this topic, and I'm so delighted to be talking about it. I'm actually going to exercise prerogative and elevate Steve to the next comments. Artist, to surprise you, Steve, that's okay, but to pick, maybe pick up on Lisa's question about who, besides academic researchers, might use this day. I had no idea, actually, what Steve is going to
Speaker 2 1:11:36
have pretty, pretty good guess. This topic really bothers me, because the principle behind why all of the data have to remain confidential is complete bullshit, and everybody knows that, but no one seems to be able to fix it, even inside pharma, working across many different pharma companies, small startups, big ones. When we talk to their value people, they say, well, our lawyers won't let anybody see this. We can't see it the data that you're asking for to help actually understand whether there's a subpopulation that gets more benefit, or whatever it is, to try to help patients, doctors and yes, payers make sense of things. And then, over time, we usually grind the wheels, and eventually they will say, well, we can give you some of it, as long as it's redacted forever and ever. And we say, what if it's redacted until 18 months after you've received approval? Well, our lawyers, Oh, do we have to talk to our lawyers again? Yes, go talk to your lawyers again. This process is nonsense, and it doesn't serve the phase three data. Have no commercially sensitive material about helping competitors launch something new in my book, it's all about preventing the world from figuring out whether a sub population is really where you see the benefit. So we're only going to pay for it for 75% of your FDA label, not 100% of it. And I know this is what happens in Europe, and it's part of what is happening in the US, but they lack the data, and we at ICER lack that data. So I guess my question is, couldn't we get over this around the phase three data, at least, and try to figure out a way? I'm not sure it really has to be the FDA, but it just seems that this is, and I hate to say it's just payers, because it's really not it's patients and doctors, but everybody deserves to know the results of phase three trials in ways that they can manipulate those data and look for things. I just think that that shouldn't be such an unreasonable part of the social contract that we have around people who are getting FDA approval. Sorry, it's not really a question, but responses, response as well. This one Dukes me up.
Richard F. Kingham 1:13:42
No, it's not a question, it's a statement. And I think you elided the problem of foreign governments and so forth, and approvals around the world, if you're going to make real complete transparency, and those data are simply in the public domain that will have a dramatic impact upon how drugs are approved in other countries to the detriment of American and European companies. So I think you have to take some account of that as well, and that's why I was saying there are ways in which the right people can get these data to do the sorts of analyzes that Yoda and others would like to do, and perhaps you but they shouldn't be something that results in Brazil and 15 other countries putting local products on the market, when we did all the research and so forth that. And I think when you explain that to people in the right political roles within the United States, they will understand that.
Reshma Ramachandran 1:14:37
But Steve, just to clarify, you wanted phase three data right? Not like manufacturing or it
Richard F. Kingham 1:14:42
just about the phase three raw data. That's what I'm saying. They can be used to approve a product in another country. If they're legally available. You can simply walk in with a marketing authorization application. But Steve is phase you is can't, can't you see that? Can't you see that you. You know anything about approval of drugs outside the United States? Okay, well, then what? Why do you not know? Why do you think that couldn't happen
if you, if you say we're not going to follow trips, we're going to make all these data public. Okay, you can't argue for trips on the one hand, no, you can argue both ways, because there are two possibilities. One is we observe it, and the other is we don't. If we don't and we release all these raw data, they're lawfully in the public domain, and anybody can use them. And if they show up in China or elsewhere, along with the local studies you have to do and so forth, they can get an approval. In addition, in many countries in the world, the exclusivity periods that have been decided upon are vitiated when the when the when the data get into the public. Okay, so, I mean, you got to think about this a little more. That was a very simplistic statement that you made. I'm sorry, yeah, I'll
Christopher Morten 1:16:08
let Thank you jump in. And
Rebecca S. Eisenberg 1:16:09
then, you know, the you know, most of the profits on drugs are made in the US. Next most is in Europe and Japan
Richard F. Kingham 1:16:17
after IRA. Maybe not.
Rebecca S. Eisenberg 1:16:21
Well anyhow. I mean, I'm not sure where these places are. You can just use them, and I don't know how much, how significant they are is, you know, it's not that, it's not a consideration. It's not that I think it's not bullshit. I wouldn't say that. I just think it's, you know, it's difficult to to measure, difficult to know how far and there are, there are other reasons that are bad reasons for, for for wanting to keep the data secret, and so, you know you you don't want to, yeah. Anyhow,
Christopher Morten 1:16:57
I'm sorry. Let's move on. Yeah. Well, we have a robust queue. The folks that I have are Aaron, Sean, Artie, Melissa and Rachel, and please raise your hand high if I've missed anybody. And maybe, should I ask folks to, should we do pairs, maybe, or, I guess, a pair and a three. So is that okay? I'll ask Aaron and Sean to go back to back, because they're sitting next to each
Speaker 3 1:17:19
other. And then, well, at the risk of moving on to a different topic, I have a question that's somewhat out of left field, which is another part of hatch Waxman that I think has been interpreted in a way to try to inhibit the release of safety data, which relates to the pleva versus Benson case and the fact that generic manufacturers are have been insulated as a result of the Supreme Court's interpretation of hatch Waxman to need not be amenable to product liability suits for safety related information, and there's nothing that inhibits the release of safety information and and data more than insulating a drug company from a product liability so I was just wondering if the folks at the panel had thought about that as Well. I
Richard F. Kingham 1:17:59
have a view.
Christopher Morten 1:18:01
Let me, let Sean get his comment.
S. Sean Tu 1:18:03
I guess my, my question brings us back to our, our panel, which is like, how does this interact with trade secrets, like, how, how, how, and why isn't this part of BP, CIA like, it seems to me that it, why isn't this
Rebecca S. Eisenberg 1:18:20
part of the, I'm sorry, the Yeah,
S. Sean Tu 1:18:23
I don't remember seeing something similar. No, no, there is, yeah.
Christopher Morten 1:18:27
This language, to my knowledge, is only in the hatchbacksman Act and applies to new drug applications on small molecule
S. Sean Tu 1:18:33
drugs, right? And then what is the legal basis for arguing that a trade secret should kind of protect this information like, well,
Richard F. Kingham 1:18:41
there's a there's a long legal discussion about the distinction between confidential commercial information and trade secrets. And you know, the classic trade secret is a way that you make something or the secret formula for Coca Cola or something like that. And that's why, for example, there's a much tougher fight on manufacturing information than there is on the things like the safety and effectiveness data and so on, and I so the question then is, are the safety and effectiveness data confidential commercial information that would be protected? The other thing is, of course, that the way the Freedom of Information Act works. It doesn't prohibit an agency from disclosing information under before the Trade Secrets Act does, but if the agency has a set of regulations that protect confidential commercial information, then it has to follow them, okay? And then you have to worry about what CCI is, okay? Now there actually, you know, there were a number of cases over the years that narrowed the concept of what was CCI, and then there was a Supreme Court case that actually greatly broadened it. So you can actually argue now that CCI is anything that somebody thinks is confidential and gives to the government. I'm not saying that's the right answer, but the point is, you know that it's gone back and forth. Forth,
Christopher Morten 1:20:00
just to make in case Rachel or Becky, want to respond, yeah,
Richard F. Kingham 1:20:04
sure. But I want to get back to, I want to get to the pleva case, though, if you want, go ahead, Becky, yeah,
Rebecca S. Eisenberg 1:20:11
no, I was thinking about the pleva cases. You know, one way of reading this language on the slide is that it's about providing information to generics about the products that they're selling, and that that would be important information for them to have. If you're, you know, selling a product that you're you know that you're responsible for in various ways that you would want to know as much as you could about that product, and you might want to take various measures to make it safer or to protect yourself from liability or something if you knew about those problems, and you don't know about those problems, and now if there's the further benefit that it might protect you from, from from liability. So I take your point. I think that's a huge gap in incentives that's introduced by this sort of formalistic approach in pleva versus Mensing to product liability, that is regrettable. I don't know that that was contemplated in
Richard F. Kingham 1:21:10
the no, definitely, but that
Rebecca S. Eisenberg 1:21:15
was, that was what
Christopher Morten 1:21:16
I wanted to do. I can't resist. I have to make two so quick, such quick points in response to Sean one that there's a European Medicines Agency policy of disclosing clinical trial data. And in connection with that, there was a challenge that went up to, I always screw up the European court system, but very high court determined that there was no, you know, in European Parliament's taking of trade secrets when this information is disclosed. Interesting precedent. I also have written a paper, I have to say, sorry, plug arguing that the Trade Secrets Act does not prevent the FDA from disclosing clinical trial. No, I
Richard F. Kingham 1:21:48
think the we can talk more about consensus that this is not a trade secret issue. It's a confidential commercial information.
Christopher Morten 1:21:54
Yes, but that also doesn't prevent the FDA from disclosure. Just allows the
Richard F. Kingham 1:21:58
FDA. It does. If the FDA regulation to the yes, the writer, you're right, yes, yes,
Christopher Morten 1:22:02
the regulation, not the
Rebecca S. Eisenberg 1:22:04
statute. One more agenda, but of tort liability here is disclosing all of this debt, these data would be better. It would provide better protection from tort liability, right? It would be. The information would be out there. We're not hiding anything. Selective disclosure of just the good news and not the bad news is what sets a firm up for tort like, well,
Richard F. Kingham 1:22:24
the plaintiff's lawyers get our MBAs. Okay, so I'm not sure about that. Let's, can I just say something about pleva? Because I think that or not, we have three, we have three questions
Christopher Morten 1:22:34
still, so let's see what those are. We have Artie and then, sorry, Michael and Mickey ran around. I think, let's go, oh, okay, sorry, sorry. And then Rachel, and then Melissa. Okay, Melissa,
Unknown Speaker 1:22:46
everybody wants a drink, we'll
Christopher Morten 1:22:50
give Melissa. Sorry, different Melissa, sorry,
Speaker 4 1:22:58
it's okay. I don't know who I am. I'm sorry for being the third person who didn't pass so you guys don't get to drink it. But I do want to talk about trips compliance really briefly, and mostly my question is about conjunctions. So article 39 the relevant sentence here is, members shall protect such data against disclosure, except where necessary to protect the public, or unless steps are taken to ensure that the data are protected against unfair commercial use. That's an or so either you have to make sure the data protected against unfair use, I agree, or it has to be necessary to protect the public. So the question is, is the disclosure of clinical trial data plausibly protecting the public? And the second question is, while we're on the topic of trips, Article 66.2 is one of the only articles that gives an affirmative obligation of states to promote tech transfer. So the kind of like, Oh, no. What if China made generics, like, trips says you're supposed to do that. Trips? Is it trips? Trips is supporting of
Richard F. Kingham 1:23:47
that, okay, first of all, you're right. It is a conjunction. And I, my read has always been that it would, you would make such a disclosure when you had, you know, a significant public health issue, like a covid pandemic or something like that. If your view is that we can simply say that we'd like to know about it, because doctors would like the information, so now we don't do it. That sounds a little far fetched to me. I understand, but, but you but, but if it's something as simple as doctors would like to have it, then in effect, you've just eliminated the whole provision, haven't you? No, I mean, no, again, this was a
Speaker 4 1:24:24
contested treaty, right? Like this. This provision was forced in many countries didn't want it, and so this was the middle ground, which honestly is quite donors. You just have to protect the public. It's the same thing as compulsory licensing, but there's a call that has to be catastrophe. No, there doesn't have to be catastrophe. There just has to be public use. But there's always this kind of, like slide, this kind of, I interpretive imaginary, that it can only mean exceptional circumstances, when, in fact, it was not for exceptional circumstances.
Richard F. Kingham 1:24:48
And there is, there is a very clear contrary position as well. That's all I'm saying. And you say what you're saying, but other people would say something different. So
Christopher Morten 1:24:56
last thoughts from Rachel Becky, before we close.
Reshma Ramachandran 1:24:59
Yeah, no, the only thing I'll just, just say, you know, I think, kind of to hammer on the point, there's been various studies to be done, and we'll see even, you know, even there's also been some estimates, like post IRA, with some of the price points have come out in terms of revenue that's actually gartered in the US versus other countries, including in China and India. I mean, the question is like, how much are, you know, the 2% of revenue versus having access to information on safety and efficacy. How do you weigh that from kind of a clinician, patient, moral perspective? Just to put that out there, you know, if that's like the kind of the real worry here, and I'll just also contend that, you know, I'm like, just struck like, be, you know, the idea I would imagine you would need a lot more data for new drug application, having seen several of these, besides just clinical trial data to actually get approval. So I just want to if there is an example of when the phase three data that was publicly available, including from public health agencies that have been conducting it, I'd be curious to
Richard F. Kingham 1:25:55
see, well, there are three. There are three main components. There's quality, safety and effectiveness, as the terms are used under the eCTD and so forth. And the safety and effectiveness data are the non clinical and clinical data that are in those modules. The you know. The other big part of the of the of the submission is how you're going to make the product. That's something you know, because you're making the product. And all you have to do, if, if, if need be, is make sure that you can convince the agency that the thing you're making is the thing that was the subject of this, this research, okay, in other words, that you have a product that's bio equivalent to it, and so forth. So I think, I think it would be an issue. And I also remind people that that's one half of the coin. The other half is that if we simply disclose this data, that many countries will then say that the exclusivity periods they've granted are vitiated. And I think they would. Australia law specifically says that China does it okay, and it's in their FTA, with Switzerland and so forth. So look, this is a much more serious issue than some people are suggesting. I don't doubt that we will get to some place where better use can be made of the data that are FDA, and I would think that the industry should cooperate in that. But I think the idea that we simply put all these data in the public domain and forget about the potential adverse consequences. Is naive or something worse.
Christopher Morten 1:27:29
I think that's the time we have. Thank you all so much. Reception is just to the side here. Thank you so much dick and bakey applause.
Announcer 1:27:48
The engelberg center live podcast is a production of the engelberg center on innovation Law and Policy at NYU Law, and is released under a Creative Commons Attribution, 4.0 International license. Our theme music is by Jessica Batke and is licensed under a Creative Commons Attribution, 4.0 International license.