Merck & Co., Inc. (MRK) Earnings Call Transcript & Summary

Good afternoon, everyone. Welcome to our panel on the use of real-world evidence to address today's health care challenges. My name is Rahul Rakhit. I'm a research analyst here at LifeSci Capital. Today, I'm joined by Ken Tarkoff, CEO of Syapse; and Gideon Blumenthal, Vice President of Oncology, Global Regulatory Affairs at Merck. For the next 45 minutes to an hour, Ken, Gideon and I will explore the ways in which real-world evidence can be used to transform the way we think about the biology and natural history of serious diseases as well as how it can be used to improve clinical decision-making, particularly in the most rare and difficult-to-treat diseases. [Operator Instructions] We'll try to address as many of the questions as time permits. Ken, Gideon, thanks for joining us today. We really appreciate it. I think to start out, it would be good -- or it might be helpful for both of you, just give us a quick background on yourselves and your experience using real-world evidence across the health care spectrum. Ken, would you like to go first?

Yes. Sure. So my name is Ken Tarkoff. I'm the CEO of Syapse. Just by way of background, I've been at Syapse for about 4.5 years. Prior to that, I spent 15 years at McKesson RelayHealth, managing the HIE, pop health and financial decision support businesses. So specifically, my experience in working with real-world data, real-word evidence has been, over the last 15-plus years, dealing with clinical data from complex data sources and aggregating them and normalizing them together. And then most recently, in the last 4 to 5 years here at Syapse, we've been using the real-world data and the real-world evidence to help our health systems and life sciences customers derive insights, specifically in oncology.

Got it. Gideon?

Yes. Hi. Good afternoon and thanks for having me. So my background, I'm a medical oncologist. I trained at the National Cancer Institute, was seeing patients, actually focusing on thoracic malignancies, including lung cancer, at the NCI. And then I spent over a decade at the U.S. Food and Drug Administration, initially as a clinical reviewer in oncology, then as a clinical team leader in thoracic and in head and neck cancers, and then moved into the immediate office under Dr. Rick Pazdur as a Deputy Director in the Oncology Center of Excellence and as Associate Director for Precision Oncology in the Oncology Center of Excellence. We -- in oncology, at my time at FDA, we were very instrumental in sort of helping to think through the strategy and the policy going forward with real-world evidence and how we would deal with it from an oncology standpoint with the mandate from 21st Century Cures and then the framework that was propagated by the agency around real-world evidence. We had a lot of collaborations -- research collaborations with health tech vendors like Syapse and Flatiron and CancerLinQ, et cetera. And we were starting to use RWD and RWE to help formulate our regulatory decision-making and our benefit-risk analysis. So it was very much starting to see an uptick more for supportive information in the regulatory submissions. About 1.5 years ago, I moved over to Merck, and I've been in the oncology department and in global regulatory affairs. And there, I work with clinical development colleagues, commercial colleagues, market access, epidemiology colleagues to figure out how to use RWE to enhance our regulatory submissions in oncology both in the U.S. and also in the EU most extensively. And then also starting to think down the road potentially, how do we use it for Chinese submissions and submission to the PMDA in Japan.

Got it. No, I appreciate it. You kind of walked me through the background. Clearly, a lot of experience in this space, so definitely looking forward to hearing some of your insights.

Maybe just to kick it off, Ken, I know we just got the Syapse presentation, so I apologize for making you repeat yourself. But I guess for the benefit of the people that weren't on that session, maybe you can just give us some background on what is real-world evidence, what is real-world data. How are they differentiated? And how do you define it? And Gideon, please, from your perspective as well, the FDA and as well as some of your previous work, how have you -- how do you see this? And how do you define it as well?

Yes. No, it's a good question. And maybe in the effort to try and make it as simplistic as possible, sometimes the words -- the acronyms or the words get interchanged, and they are quite different. Real-world data is the actual data that comes from systems. Traditionally, it was from claims system. Revenue cycle systems was used as the primary source of real-world data. And recently, in the last, let's say, 5 years, there are a lot more systems that are available. So EHR systems, other clinical systems, registries, lab systems and a variety of other systems are the sources for real-world data. So in essence, it is data in the real world, the data that you can capture that has a view of the patient that is foundational to be used to draw insights. Real-world evidence is actually the insight that is drawn from those data sources. It is often that you have to invest in converting the real-world data to be able to -- for it to be useful to be used in a real-world evidence setting. An example I often use is when you actually have treatment data, if you're looking at cancer care, you have to be able to convert that into lines of therapy. It's not usually stored in the clinical systems as lines of therapy, so you actually have to put some logic to be able to convert that into something that's actionable. Another example is mortality. You may have one source of data on mortality, but to have high-quality mortality data, we've learned you have to get multiple sources and combine them together because there are biases in a single source that a multiple source is necessary. So when you do that, you then convert that into the ability to put analytics on it. And then the analytics that you put on that, that drive insights is the category of real-world evidence, which is really powerful that you can then use to do a variety of things. The example with Syapse, we use it with our health system providers to drive better care to their patients; with life sciences, they use it in a variety of places to accelerate the approval of therapy, to be able to do post approval, safety and efficacy, to help with clinical trials and so forth, are all the opportunities for real-world evidence. And again, there are companies that have real-world data, and all real-world data is not the same. And there are companies that do real-world evidence, and all real-world evidence is not the same. And then you actually have to combine the 2 if you want to be able to do some of the stuff that Gideon was referring to as well. So hopefully, that helps with just the simplistic description of the 2 to help with the understanding.

Yes. Absolutely. Gideon, I don't know if you have anything to add on top of that.

No. I mean that was very well said. I think -- in oncology, specifically, I think the care of the patient through the journey of the patient can go from the oncologist's office to the hospital setting, to the infusion clinic, to hospice or, hopefully, we're seeing more and more with better therapies to back to regular -- back to the primary care doctor, back to regular life. So it is -- as Ken mentioned, it's highly complex in the U.S. especially. I think there are other countries with less, shall we say, fragmented care delivery where some of the real-world data may be easier to capture longitudinally. But it can be very, very complex. A lot of the data is unstructured in the electronic medical record, so that's really a huge challenge is to try to compile these disparate streams of data and to aggregate it and then to have humans put it into some sort of structured form so that you can make inferences across data. So just as one example is even something as simple -- Ken mentioned mortality. That would be something that you would think would be fairly straightforward. But in our fragmented system, it can be hard to ascertain that data. Another example would be like ECOG performance status, which is very important in a clinical trial setting as a prognostic marker not always captured in the real world. So there are ways to sort of interpolate ECOG performance status based on other parameters, but it can become very tricky. And I do agree with Ken that there are various levels of data collection and rigor and quality amongst various vendors. And from a regulatory perspective, the regulators are very interested in, obviously, like the highest-quality data, especially if you're using this to make decisions both on the efficacy side and the safety side.

Got it. Just thinking about how both of you referenced the variation in quality of data and the importance behind that. Maybe if we take a step back or just kind of look at where we've been and how we got there, Gideon, you mentioned the Cures Act. Maybe just help us understand, how did that open the doors to get to where we are now and kind of set the standards for real-world evidence and what we need to make real valuable clinical insights?

Yes. So the Cures Act, I mean, Congress really mandated the agency to incorporate real-world data and real-world evidence into regulatory decision-making, incentivize the creation of real-world evidence, task force propagation of guidance documents and kind of was the imprimatur from above that we really have to take this seriously because I think the default position amongst many of the review staff and division -- clinical divisions is that somehow this is inferior evidence than your gold standard clinical trial data, your randomized controlled study, which has a lot of internal validity. But there's obviously a growing recognition that, in many cases, the clinical studies with the rigid inclusion/exclusion criteria, rigid end points, many times -- and particularly in oncology, where everything is fast and furious and we're trying to get the therapies out to patients on the market as quickly as possible, patients are desperate for novel treatments. There's a lot that gets pushed out to the post-marketing study setting like dose finding, dose optimization, studying rare population, studying patients with different comorbidities. So there was a growing recognition that real-world data, real-world evidence can really help supplement the evidence generated in the traditional clinical trial setting. And 21st Century Cures, I think, really helped catalyze all those processes within the agency.

Yes.

Yes. If I could jump in there, Rahul. I completely agree with what Gideon was saying. There's a couple of things that I think are important to call out, which is when most of us get our care, it's not in a totally controlled environment of a clinical trial. It's in the real world. And in the real world, the data is not all stored perfectly, and there are many reasons for that. So for example, in cancer care, much of the elements that we're looking for are not elements that are needed for reimbursement or for payment. So oftentimes, in the clinical systems, the information is not captured in a way that actually provides the insight that's necessary. So the real world is what matters. We want to understand what actually happens in the real world and where none of the choices are perfect. And it's not as if the data is sitting around in a bunch of silos and you just need to link all the data together and magically, you have your answer. It's that you actually have to often go search to find the information to fill in the holes, and you got to be able to have access to that information in context to be able to address it. Gideon used that example of the patient going around and getting different sites of care. They also have different providers depending on the type of cancer. If you're dealing with prostate or bladder cancer, the urologist is very involved in the patient's care, particularly early in the diagnosis on top of the oncologist. It's important to be able to navigate through those experiences, and then sometimes in parallel, to understand what's happening to patients because that's what happens in the real world. Sometimes you actually -- the drug actually gets approved while the patient is already receiving treatment. So sometimes it can get introduced at a different point in time. It's not controlled like a clinical trial, and that's important. And that's information that's useful for us to understand better how to take care of those patients to use those insights to learn more about what's happening. And just because the drug is approved and it's set as a standard of care, there are still lots of pieces of information that are important that all of us can benefit from, and it all happens in the real world. And what the 21st Century Cures Act was, was acknowledged that, that data is valuable. We need to use it, but it's going to take a lot of effort to make it actionable and useful, and not all data sources are the same. So we have to have the right standards. We have to have the right methods to put that in place. And the FDA started doing this when Gideon was there, and they're continuing to do this where they've created partnerships with industry to be able to -- with different players to be able to work with them to begin to answer some of these questions, to develop more expertise in trying to define consistency. So there's actually a bar that you get to where we can all be comfortable, recognizing that real-world data is not perfect. And it's all not equal, and you've got to find ways to get to that level of evidence that we can get comfortable with.

Got it. So that's kind of a great segue to my next question. Gideon, just in the time that you've been in the FDA and you kind of touched on this earlier, the quality of real-world evidence and real-world data and getting comfortable with the idea of using that and having the same comfort level as to have generated from clinical trials, ultimately, how has the perception around this data changed within the FDA over the past few years? And what is it going to take for us to get to a point where we can really start using this data and the insights derived from this data to -- in a regulatory sense but also in clinical settings?

Yes. Great question, Rahul. I mean I think that there have been some use cases. Recently, the one that's often cited is the submission of palbociclib, a CDK4/6 inhibitor for male breast cancer, which is exceedingly rare and it would be very hard to run a traditional clinical trial. So sponsor, Pfizer submitted data. They used electronic medical records from Flatiron and I believe some academic medical centers as well as some claims-based data, I believe, from IQVIA. And they were able to get a labeling extension for this rare population of male breast cancer patients. The FDA did point out some limitations in the data, so I think the FDA was very transparent, some of the limitations in terms of missing data, pre-specification of what the analysis plan was going to be, things like that. And I think as this field develops and evolves and continues to mature, I think, those elements in terms of transparency, being able to publicly disclose what the analysis plan is going to be, so it's not some sort of post-hoc fishing expedition. How are you going to collect the data? How are you going to deal with missing data? What are the key data elements that are going to be necessary? How to make it sort of "so-called regulatory grade" and that's traceable, auditable from a compliance standpoint. I think those will be critical pieces to continue to develop if we want to take it to the next level from supportive analyses to sort of the primary evidence in a regulatory dossier, be it for an expanded indication, a new molecular entity, although that might be further down the road, or using it for -- to satisfy post-marketing requirements.

Got it. Ken, I think -- is there anything to add there, thinking about -- as someone at the helm of a company that's pulling these insights and deriving high-quality -- that's pulling high-quality data, how do we set the standards here? How do we kind of create a framework that the FDA and other regulators and clinicians can get comfortable with?

Yes. I mean it's a great question. I don't have much to add to what Gideon said. I mean there has to be a standard. Not all -- as we said early on, all real-world data is not the same. There has to be a level of completeness and quality. There has to be a level of confidence in the methodology that you're using and how you're sourcing the data, is -- the hard-to-get data is the most important stuff, particularly in cancer care, and you've got to have high-quality processes to do it. And as I mentioned, the FDA is working directly with some key companies in this space to be able to work to set these standards. We at Syapse joined the Real-World Evidence Alliance, where there are 5 companies that got together to try and provide some aggregation of industry to help with an industry response and work together on the things to try and help influence some of these more broad general areas where you can address that's not specific to a company. And so those are ways on how you can move the industry along. I would say that I think the FDA is moving pretty quickly. If you think about how important this is and if you think about how much potential there is, we shouldn't rush this. There shouldn't be a 6-month thing where we make a bunch of mistakes. But there is a ton of opportunity here to add value in the space. And as we continue to have more proof points, we continue to get better standards. We continue to have minimum quality thresholds and minimum bars that we can standardize across in a way that works for the industry. That's when we'll get to a pervasive place. It's coming, you can tell. As long as we're thoughtful about the way we do it and as long as we don't get overly aggressive as an industry and people try some things that are probably not up to par, then we'll continue progressing. I mean the great example, I think post-marketing studies is a perfect place to be able to use real-world evidence because once the drug is approved, the best way to understand what actually happens in the real world is to use real-world data and to use analytics. That's the best way to do it and the best way to have a follow-up, especially if you're trying to rush to get a drug to people who really need it and you want to follow up on things that you weren't able to test in the clinical trials. It can take a very long time to set up a randomized clinical trial to do it; whereas in the real world, you can get those answers a lot more quickly while you're actually getting the care to the patients. So it just makes a lot of sense for it to happen, and I think it's going to continue to accelerate. But it should go at a reasonable pace because we don't want to get ahead of ourselves as an industry. Again -- and I guess the other thing I would say is I completely agree with the comment about the U.S. I mean the U.S. is probably one of the, if not the most complex places to get access to clinical and molecular data to aggregate it together. And so you have to invest a lot to be able to get to that quality. There are some other countries where it's much easier to do it. But the fact that it's complex here, that the FDA has such a great reputation, that if we can prove it here as a standard, it will influence the whole world. And so we want to be able to do it right as an industry because of the potential to impact everybody who needs the benefit of getting better therapies to them so they have better chances on their outcomes.

Got it. Appreciate the color on that and the thing around these post-approval studies. But I guess thinking about the other side of this and ultimately, how can real-world evidence be influential in improving clinical trials, particularly clinical trial designs, end points, the number of patients, how we think about enrollment, et cetera?

I mean I -- well, I'll just jump in. I mean there's a ton of places. I mean you just kind of highlighted it. When you're designing a clinical trial or thinking about a clinical trial or in a clinical trial and you want to have a control arm that you want to be able to -- when you have -- especially for advanced cancer, you want to be able to -- for patients, you actually give them a path. And real-world evidence -- real-world data and real-world evidence is an ideal vehicle at the right quality level to address many and all of those things. It is not designed to replace clinical trials -- randomized clinical trials. There is always a place for it. It will always be a mainstay. But there are many ways that real-world data and real-world evidence can enhance it, maybe accelerate it, maybe have it be more targeted. You can have more information going into the trial. You can get more information faster during the trial, which is good for life sciences. It's good for health systems. It's good for patients to get those therapies to market faster. And there's a lot you can do after the trial, too. So the burden on the clinical trial process post approval can use more real-world evidence. So there's a lot of places you can just enhance the overall process and make it better for everybody. I just think the potential is limitless at this stage.

Yes. No, I agree. I mean I think just thinking about how we use it at Merck to help design trials. I mean, I think the standard of care is changing so rapidly. It's such a dynamic field, and you get dozens of approvals every year, which is a great thing for patients. But it can be hard to understand. If you do a literature review for your control arm and you're making certain assumptions going into your Phase III study with your new drugs, those could be inaccurate. So having a sense of how a standard of care, a control arm might behave in a contemporaneous setting could be very useful for protocol design. I think for other areas that we look into, for example, for a new end point, like if you're looking at -- like, for example, pathologic complete response from -- to neoadjuvant treatment before surgical resection, you can do correlation studies, look at past CR in the real world and correlate that with, say, disease-free survival or real-world mortality. So that can be helpful in potentially establishing surrogacy for an end point that's earlier than some of these longer-term end points like disease-free survival or overall survival. For rare patients -- rare cancer types, we have some experience, for example, with rare cancers like neurofibromatosis, Von Hippel-Lindau syndrome. You can run sort of natural history studies to understand how -- absent a treatment where there are no standards of care, how patients, their cancers behave absent any treatment. And then finally, a lot of these molecular epidemiology studies can be run to understand if you have a new targeted therapy, just to understand -- again, we're sort of creating new disease states as we go. So for some of these target therapies, EGFR mutant non-small cell lung cancer or chronic lymphocytic leukemia, now we have these highly effective targeted therapies in the frontline. Can we learn it in a real-world setting? What sort of resistance patterns emerge that can help us figure out if we have the targeted therapy, the next-gen therapy in hand, how to design this next generation of studies.

Got it. No, that's really helpful. Some or a lot of the questions we've asked been kind of with your regulatory hat. We got to flip and probably put on your industry hat. Thinking -- from Merck's perspective, I guess, how are they thinking about future targets? How are they thinking about pipeline development, pipeline expansion? How can real-world evidence really be leveraged to strengthen the pipeline and look for new opportunities? And to that, Ken, maybe you can tell us a little bit about how Syapse is supporting that kind -- those kind of efforts and providing those insights that, that partnership can lead to really innovative clinical development.

Yes. So as Ken said, I don't think the traditional clinical trials and the randomized controlled trials are going anywhere. That's the best way to establish causal inference. But I do think that companies like Merck, we're very interested in how we can incorporate real-world evidence to help design better trials, help understand the characteristics of the drug better. So one example would be, and Ken alluded to this earlier, there's a big push from the agency and on the industry side to kind of diversify our clinical trials to understand how drugs perform in underrepresented minorities and elderly patients and patients with various comorbidities like kidney insufficiency and liver insufficiency, brain metastases. And again, because drug development in oncology is so vast, oftentimes, you don't get a perfect sampling of those types of patients in the pre-approval setting. And there's been a big push from the agency to mandate post-marketing requirements to look at some of these populations. So one potential use case down the road is can we design real-world post-marketing studies, which would be a better representation of how the drug actually performs out in the clinic. And then the last thing I could see happening although -- and I actually saw today, I saw a release that actually an older drug, cetuximab, I think there was a real-world study where they -- FDA approved a novel dose for cetuximab based on real-world evidence going -- I think it was going from every 2 -- from every 1 week to every 2 weeks. I can really see a lot of renewed focus on dose optimization in the post-marketing setting. The FDA has announced this Project Optimus, where they really want to better optimize anticancer drug dosing and scheduling. And so the cetuximab case is a good example, where if we could do this in the real-world setting rather than in a clinical trial, I think that could both be very informative and save cost for the industry in terms of running a study versus observing patients who are actually being treated in the real world.

I would agree with what Gideon said. I guess if you talk about it from a company's perspective, first of all, I would say we could fill a couple of hours talking about all the different places that real-world data, real-world evidence can address in large life sciences pharma companies and the opportunities there. I think the challenge for companies like us is to find the areas of focus where we can add the most value and put our energy in that place. And specifically in Syapse, our focus is on the part of the community health system that's dominated by large community health systems. So the community market, about 80% of the care in cancer care in the U.S. is made up about roughly 2/3 health system, 1/3 independent oncologists. Maybe it's more like 60-40. And we focus on the large health system part of the care, where a lot of life sciences don't have direct access to the providers based on the nature of the organizations and how they're set up. And we focus on providing real-world data and real-world evidence in that area. So as we work with our partners, we are focused on the quality and completeness of the data you can get from that segment of the population. You can do things from understanding all different types of research questions that come throughout as the product is coming to market and you want to understand in a major segment of the market what can you actually learn as the product is progressing. There are things you can do with clinical trials, as Gideon was talking about. There's different data sets you could use along the way to be able to help. There are things you can use for regulatory data to actually get a submission so you actually create a specific set of data for a specific purpose. And then there are things you can do post approval as well. The one thing that I think is important that we highlight a lot at Syapse is it's important that real-world evidence is not just used to provide an insight for a publication. It's important that it's used to then take action to impact care. So a big focus for us when we work with our life sciences customers is to make sure when we identify an insight, it's actionable to then bring it back to the health systems and the providers to be able to use that to provide better care to the patients and then measure its impact because sometimes the insight that you find at first actually needs to be iterated as you learn more in the real world about why it's actually being caused. We've learned that a lot in different types of biomarker insights and different drug class insights based on guidelines. You might look at them at a high level and say, that can't be right. They've got to be testing more patients, so they got to be putting those patients on the care. But then you actually find, when you dig in, that there are good reasons based on the real world of what's happening. And that's important for everybody. Obviously, it's important for the life sciences company as they want to get their drug better adopted to provide better care to the patients. It helps them drive adoption of their product. The providers want to do the same. They want to provide better care, and the patients want better outcomes. But you have to be able to iterate through that process to do it, and I think that's really important. I think over the next 5 years, with all these new targeted therapies coming to marketplace -- and as Gideon said, they're getting approved very quickly, and there's not a lot of -- enough evidence to be able to be informed for everything that's going to happen in the real world. And so we're probably counting on industry to be able to answer these questions quickly and take action on them so that we make sure we get the outcome that we all want. Plus, we don't want a bunch of approved therapies without the ability to get them the right person or without being able to learn a nuance about them that's important that actually drives the right result. And that's the opportunity for industry -- so companies like us focus on a major segment of the market to try and optimize that, and that's the play -- what we can do as a player in the space, to try and drive the best outcome for the industry and our customers.

Yes. Ken, I think you're reading my mind because this is kind of the thing I want to get into next. We spent a bit of time discussing the implications of real-world evidence on -- from a regulatory standpoint, but thinking one from physicians, who are really just trying to improve the quality of care that they provide to their patients; and two, on the industry side. And please, Gideon, provide this perspective as well. And we have all these new drugs coming to market but figuring out how to identify the right patients or how to expand the potential patient population. Maybe can you guys talk about how the use case is there? And even just pull it out of the abstract. Maybe in your experiences, what have been some specific use cases where we've seen this happen?

Well, I guess I would just jump in from our experience. I mean there are so many opportunities to learn things in the real world that can provide better care to patients. I mean we -- just to use a couple of examples. If you just start at the front end, which is are all of the people that are getting the right testing at the right time because you're not going to be able to provide the care if you don't test the right patients at the right time. And then does the provider who gets the result of the test know what to do with the result? And is it actionable? And are all the processes that could get in his or her way in that person's way? And are they informed about what actually is going to happen to the patient? You just start there at a minimum. The reason why I got into Syapse is my father is a Stage IV colon cancer survivor 16 years in as a physician. And when I watched him navigate his way through the health care system and how much he had to educate himself, I was blown away as someone who has been in the industry for a long time saying this is crazy because the level of care you're going to get is based on how actively you're researching for yourself for this information at the time. And that -- there's an opportunity for us to do better. So if you take those examples, just at the front end, are the right people getting tested? Once they're tested, are they getting on the right treatment? Do they have the right information? Are the companies that are offering the products able to provide the information to the right people in a way that's actionable that can actually have an impact right away? If a new drug gets approved and someone is in the course of treatment, what happens? Does the physician actually get the information in a timely way to get the outcome for the patients? Patients have to go do the research on their own. Not being critical to the physicians. They're incredibly busy. There's just no way they can read everything all the time and get all that information. And on top of that, how do they actually make it actionable? So there's a huge -- there are so many examples of what we can do. And ultimately, for any of us who have either personally experienced or know someone has gone through cancer, it's a horrible experience. There's got to be a better way to do this. And with all these new therapies that are coming to market that are showing better outcomes and a better experience, there's a really strong motivation for us to be able to figure out how to make that happen.

Got it. Gideon, you have anything to add?

No. I mean very well put. I mean I think there are -- as Ken said, I mean, I think there are so many insights one could glean from real-world evidence, so much you can learn once the drug hits the market. I think drug development, particularly in cancer, I mean, the initial approval is just the very beginning. I mean take a drug like KEYTRUDA, has many -- I don't even know the latest [ gadgets for ] -- of indications. But -- and there's a lot you can learn both on the efficacy side identifying -- as Ken said, with all the companion diagnostics, are the patients getting the right tests? Because without that, they're not going to get the drug. Are patients -- how are they being treated? Are they being managed appropriately in the real world? What kind of adverse events are they experiencing? What are the medications that they're getting in the real world? What are the populations that should be studied next? What are the unmet needs that need novel combinations or which patients can just get by with a single-agent drug for a very long time? Or when can you stop treatment altogether? So yes, there are a multitude of questions that could be asked in this setting. And I think it's -- from an industry perspective, in addition to the regulatory perspective, I think it can be highly valuable, particularly if predicated on the fact that the data is of high quality.

Sure. Got it. I think we've got roughly 10-or-so minutes left. And there are 2 areas broadly -- there are 2 topics I'd like to touch on. First one, I feel like it wouldn't be a panel unless I asked about the implications that you think COVID-19 had -- or that taught us about the potential of real-world data and real-world evidence and how that kind of translated to improvements in clinical trials and kind of forced us to use it a bit of more, rely on it a bit more. I'm very interested to hear both of your perspectives on that. Ken, if you want to go first.

Yes. Happy to go first. Yes, I mean, stating the obvious, it had a massive impact. I mean I think if I had to simplify it, I would say that it wasn't as though everybody didn't know the potential of it. It's just sometimes you don't change unless you're forced to change. COVID-19 actually forced a lot of us to say, okay, we really got to use this information differently. There's got to be a different way to do it. And I think sometimes you need that little catalyst or the push over. The 21st Century Act was a similar thing. Sometimes there's catalyzing events. I mean with COVID-19, there were a lot of things that you couldn't do that you can do now. I mean it's true in everything, right? I mean we are -- today, Syapse is a 100% remote business. If you would have told me before then if we could go remote and still operate effectively, I would have told you no way. But we did learn that you can be just as inarguably even more effective. And I would say, if you look at cancer care -- or real-world data and real-world evidence, with all the news around COVID-19, everyone put their attention on it, realized how important data has been in the decisions that we're making an evidence that you draw off that and that not all data is equal. And I'll just leave it at that and say that the evidence that's driven off of that is critical to making the decisions. I mean we're experiencing it now with vaccines and the impact. And so I would say it's -- my observation is, 2 years ago talking about real-world data and real-world evidence, some people didn't necessarily know what it was. But today, very few people who are paying attention don't know the power of real-world evidence, at least generally, because every day we pick up the news. We're reading about it. We're seeing that information. And so that's had a big impact, and it's also had a big impact on our customers. The other thing I would use is, early on in the pandemic, most of our customers are large health systems on the provider side, and they got impacted heavily based on COVID-19. And many of our -- the actual people in the health systems we work with weren't even able to go into their office because at the time, in many of them, much of the research that was being done on oncology was being put on hold for a couple of months, which is scary. But it was the reality of what was happening for a while. And people started realizing there's got to be ways that we can do this when we're disrupted like that because we can't delay a couple of months getting these therapies to patients. And so there was a lot more appreciation for real-world evidence and appreciation for the power of data and its insights. And I would say it's the same in life science as well, too. I think everyone -- it's pretty rare. If you would talk to someone inside of life science, I'd probably defer to Gideon on this. But it would be our experience that doesn't say we need real-world data or real-world evidence. There might be debates about how to use it. But we're at a stage where people are now, it's like, all right, how are we going to make this actionable, and that's a big change from a couple of years ago. I think it accelerated things, at least from my perspective, in a positive way for everybody.

Got it. Gideon, yes, anything you want to add?

Yes. No, totally agree. I mean Merck's -- we've been able to function quite efficiently from -- in a remote environment. I mean the clinical trials have continued to pace. I mean we were able to pretty much, in most cases, keep patients on study and do our database locks and hit our end points. And some of that, I think there was and there will continue to be some shift in kind of remote monitoring. And hopefully, we'll see a continued push for like remote informed consent, and even people are talking about ways to secure imaging and lab tests more remotely so that patients don't have to schlep to the academic medical center but can be done more where they live. For KEYTRUDA, I think it helped -- actually, we had a Q6 week supplement. We went from every 3 weeks to every 6 weeks. We had a number of accelerated approvals because, especially during the COVID crisis, the value of decreasing interactions with a health care facility for patients sick with advanced cancer was very apparent given the COVID crisis. So the agency really acted with great urgency on the Q6 weeks application. But yes, I mean, I think going forward, we're going to be moving more towards a hybrid model. And yes, I think with some of the lower-quality observations on the therapeutic side with COVID, I think it sort of highlights the necessity for integrated electronic health records, better quality data and greater transparency so that we can vet more rapidly which therapies work and which don't work.

Got it. Understood. Given that there's about 5 minutes left, I do want to pause here and just address the 1 or 2 questions that we kind of got in the Q&A. First, Ken, you mentioned a couple of times, not all data is the same. But we have a question about whether global complementary and alternative medicine data could also be included in these networks? Or is differences within methodologies for collecting this data too great of a barrier? Actually, would be interested to hear both of your perspectives on this.

Which type of data? I'm sorry.

They're asking about global complementary and alternative medicine data.

I mean I think the way -- it's a complicated question because I would say there's like layers of data that you can get access to. And maybe each time you go to a layer, there's more insights that you can get, and you have to be able to bring all those data sources together. In the U.S., one of the fundamental challenges, just the core journey of the patient and understanding that core journey is so -- in many cases, is complicated to be able to bring together. And there are many different players in the space who have a slice of that data from a particular view, and you've got to be able to bring that together. As you go further out, there are other sources of data that continue to enhance that. A big source of data that's valuable is patient-generated data. So when you go into a provider system, you typically have what I would call provider-generated data, the EHR and all the surrounding clinical systems and genomic systems. And then you've got the patient-generated data, which is from the patient, which is a different set of data but also very valuable. You can go to other alternative sites of care. You can get other -- all the other types of players. And every time we're adding those layers, we have to find ways to bring that together. What I would say is what -- from my perspective, what's most fundamental is you have to understand that core journey at a very high-quality level before you can bring in other sets of data because those other sets of data don't add enough value if you don't understand at a high-enough level of quality the core. And so that would be the way I -- if -- so for example, the way we look at it at Syapse is we -- there are other sources of data that we can go after. But we sometimes deprioritize those that are interesting and add value. But at the end of the day, if you don't get the core right, they don't provide the actionable value that's necessary. And that would be how I would answer that.

Got it. Gideon, I don't know if you have anything to add, but happy to just jump to the next one, which I actually think is a good question to kind of tie this up in, is how do you see the role of real-world evidence evolving over the next 5 years. I know we have a couple of minutes, and that's a loaded question but...

Yes. Good question. I think Ken answered it well. To the previous question, I think you have to sort of build the foundation. I think there's been a lot of progress, but there's still a ways to go for the regulators to really have that confidence to take it to the next level. A lot of the topics, we talked about structuring the data, making sure that it's traceable, auditable, having pre-specification of analysis plans, having that transparency. All of those elements are going to be critical. And then thinking down the road, blue sky, I think we'll likely see more automation, more AI, machine learning, to derive patterns from the data. I think we'll incorporate other -- once we build that foundation with the fundamentals, we'll be able to incorporate other streams like wearables and other patient-generated data, which will really be very complementary to the provider-generated data.

Got it. Ken, anything to add?

Yes. I guess -- well, I mean, I would say the next 5 years is going to be very exciting in the space because I think, as the foundation gets developed more and more, I completely agree, as Gideon highlighted, is there's so much potential in this space. And as we get those standards in place and as we get the foundation in, you can do a lot of the really cool stuff on top of it. I mean, I guess, if you would say now with the state that we're in is just learning how to develop it and how to use it in a way that adds value. And then the opportunity is, I would argue, unlimited on what you can do. I think we're going to see more and more talk out -- I mean a lot -- most of the talk is inside of cancer. I think we'll see other specialties start to develop more and more use. I mean, obviously, there is use of real-world data and real-world evidence in other specialties. It doesn't get as much attention as oncology does for obvious reasons, but I think we'll see that as well too. And I think we'll see more international exchanges, where you can actually get more insights that you can share across geographies. I think we'll see more of that in the next 5 years. And so I mean, that will -- assuming all those things happen, which I'm pretty confident that they will over the next 5 years, it's going to change the way that medicine is practiced. If you think about that you can have -- you can use technology to get access, you go back to the example of my father's experience, no one will have to go through that again. The provider will be able to get access to a lot better sources of data with a lot more actionable insight. And that's just a better way for the industry to take care of patients. And so I'm optimistic with that as we use technology to change it. And I mean I think that's why many of us work in health care is because we want to ultimately have that impact. And I think that I can see, in 5 years, we'll really notice the difference, and that's pretty exciting.

Absolutely. Now I do think it's a great place for us to pause on and kind of call it from here. But once again, Gideon, Ken, really appreciate you guys taking the time to be here and providing these insights. They're incredibly helpful and incredibly thought provoking. So I really appreciate that. And to everyone in the audience who joined us, thank you for participating in the symposium. Thank you for joining this panel, and I hope you guys enjoy the rest of your day.

All right. Thank you.

Yes.

Thanks, Rahul.

Yes. See you, guys.