Hello, and thank you for standing by. My name is Lacey, and I will be your conference operator today. At this time, I would like to welcome everyone to the Natera post-earnings call. All lines have been placed on mute to prevent background noise. After speaker starts, there will be a question- and- answer session. If you would like to ask a question during this time, simply press star followed by the number one on your keypad. If you would like to withdraw your question, press star one again. Thank you. I would now like to turn the conference call over to Michael Brophy. You may begin. Thanks, operator. Good afternoon, good morning. Thank you for joining our conference call to discuss the findings from the ESMO Congress. Today's conference call is being broadcast live via webcast. We will refer to a slide presentation that has been posted to investor.natera.com, and a replay of the call will also be posted to our IR site as soon as it's available. Before diving in with some introductions, I'd like to direct you to the safe harbor statement on slide two. During the course of this conference call, we will make forward-looking statements regarding future events and our anticipated future performance, market size, partnerships, clinical studies, and expected results, opportunities and strategies, and expectations for various current and future products, including product capabilities, expected release dates, reimbursement coverage, and related effects on our financial and operating results. We caution you that such statements reflect our best judgment based on factors currently known to us and that actual events or results could differ materially. Please refer to the documents we file from time to time with the SEC, including our most recent Form 10-K or 10-Q. Those documents identify important risks and other factors that may cause our actual results to differ materially from those contained in or suggested by the forward-looking statements. Forward-looking statements made during the call are being made as of today, October 23rd, 2025. If this call is replayed or reviewed after today, the information presented during the call may not contain current or accurate information, and Natera disclaims any obligation to update or revise any forward-looking statements. Okay. With that, let's jump in. I'm really pleased to introduce today's speakers as noted on slide three. Joining me from Natera, we have Steve Chapman, our CEO, Solomon Moshkevich, and Alexey Aleshin, who all of you know. Alexey is our General Manager of Oncology and Chief Medical Officer. It's also an honor to introduce our distinguished guest and collaborator, Professor Thomas Powles, who was the principal investigator on the IMvigor011 trial that we'll discuss with you today. Professor Powles is also the Chair of the Barts Cancer Centre in London. He leads the Experimental Cancer Medicine Centre and biomedical research cancer grants at Barts QMUL. He has led 23 randomized cancer trials, including studies which resulted in multiple EMA and FDA approvals. He has also led clinical and high-impact translational science projects. Professor Powles has an H-index of 100 and has given plenary presentations at the major cancer meetings. He co-leads the Euromigos Podcast and European Guidelines for GU Cancer, and he is Editor-in-Chief of the Annals of Oncology Journal. In 2023, he was selected to Nature's 10 Global Scientists list, and in 2024, to Time Magazine's 100 Healthcare list. With that, please join me in welcoming Professor Thomas Powles. Thomas, please take it away. It's great to have you. Thank you. Thank you for inviting me. I'm really excited to be here. If you could move to my first slide, that'd be great. My take on this is that we have been really stuck for a long period of time in medical science, relying on cross-sectional imaging and CT scanning to determine direction. It's a very old-fashioned way of looking at the problems that we face. By the time you have visible disease on a CT scan, it represents hundreds of millions of cancer cells. There must be more accurate ways of looking at or monitoring cancer. The answer is yes, there are more accurate ways. It's become apparent over the last five years, or perhaps longer, that we can find and identify small fragments of DNA, of cancer DNA, in the blood. Most of the time, we don't have cancer, or we don't have DNA in the blood. One of the key issues is can we identify this DNA accurately? There have been two approaches to identify this DNA. The first is an informed approach, where we sequence the primary tumor, identify mutations, and track the cancer. That's the Natera Signatera assay approach that I'm going to talk about today. There are other approaches, panel-based approaches, all sorts of different approaches. The accuracy of those varies between different tests. Different tests give different results in the same way as different drugs give different results. I'm going to talk today exclusively about the Signatera assay, which we used in this trial, which shows a high degree of accuracy. We've been testing in urothelial cancer and indeed other cancers for a significant period of time. The slide in front of you shows the study design of IMvigor011. After surgery for metastatic urothelial cancer, about half of the patients go on and relapse, and half are cured by surgery with or without neoadjuvant chemotherapy. After surgery, quickly identifying those at risk and sparing those patients who are not at risk unnecessary and harmful therapy is very attractive. We did a retrospective analysis of a previously negative trial using the Signatera assay, which clearly showed that patients who were ctDNA negative were at significantly lower risk. Those patients who were ctDNA positive had much higher risk of relapse. We then did this prospective study to try and validate that. It's important in that previous trial that we also recognize in an exploratory analysis, we identified those patients who were ctDNA negative didn't seem to benefit from atezolizumab, whereas those who were ctDNA positive did benefit. Validating that was a really important question. It's not actually been done before in bladder cancer or indeed more widely in other tumor types. This is probably the most robust data we have, and that's why I think it's important. That's why it was selected as a presidential session at the ESMO meet and published as a fast track in the New England Journal of Medicine, which is the most important journal in medical science. These patients, as I said before, had urothelial cancer, had recently had surgery, had no evidence of radiographic disease. These patients had been told essentially their cancer probably won't come back. We then did ctDNA testing. We did that at baseline. We tracked it every six weeks for a year. If they were positive at any time or they became positive, they were randomized to atezolizumab or placebo. If they remain negative after that period of time with no evidence of radiology disease, we tracked those patients. In reality, about half of the patients—we screened 800—half the patients became positive, and half the patients were negative, as was validated from our previous work. In the atezolizumab arm, our primary endpoint was investigator-assessed disease-free survival. Secondary endpoint was overall survival. We hit both of those statistically significantly with values of hazard ratios of 0.64 and 0.59, respectively. Both of those were statistically significant. We also followed up the ctDNA negative patients carefully, and I'm going to describe those patients to you right now. If we move to the next slide, you can see disease-free survival, overall survival in subgroups based on the timing of ctDNA testing. One of the key issues in this trial is we looked at patients who were ctDNA positive at baseline. That was 60% of those that became positive. We also tracked for a year, using ctDNA analysis, patients who were negative who became positive. 40% of the patients randomized were negative initially who became positive. This is attractive because a one-off time analysis, while it's relatively accurate, it's also true that about 30% of the patients who are destined to become positive are not positive initially but become positive. By capturing those patients, we can then generate a new group of patients who are persistently negative. The outcomes of the patients who tested initially positive and the outcome of those patients who tested initially negative and became positive were very similar, as you can see from these Kaplan-Meier curves from disease-free survival and an overall survival perspective, with all of the hazard ratios in a similar range: 0.62, 0.66, 0.71, and 0.52 for disease-free survival and overall survival, respectively. Next slide, please. We focus on that third group. I've described the first two groups. Now we're on to that third group, the negatives who remain negative. Here you can see the outcome of these patients is excellent. If you look at the overall survival of other unselected studies in this environment, the median survival of these patients, we hit the median quite quickly, often at about 34 months. Here we see at 24 months, if you're negative and you're tracked negative, 97% of patients are alive at two years. This is very reassuring. Many of those deaths are not cancer-related deaths. The median age of this population is 77. Intervening in this population to try and make that 97%- 98%, that would give a hazard ratio of 0.66 in a trial, would require tens of thousands of patients. 20% of patients get life-changing toxicity associated with systemic therapy. The 3% risk of death is unlikely to be altered by any intervention. That means tracking these patients is really attractive. Next slide, please. As I said before, published in the New England Journal of Medicine. When we move on, are there other data sets showing similar results? The first thing to say is IMvigor011, a positive randomized trial prospective, positive for overall survival, sorry, positive for disease-free survival, positive for overall survival. It was also based off a retrospective analysis of the IMvigor010 trial. At the same meeting at ESMO, we also saw the nivolumab adjuvant data, also with circulating tumor DNA analysis. This analysis was less robust than the prospective study. It was also less robust than the IMvigor010 trial because you can see here from this Kaplan-Meier curve slide that the numbers are relatively modest, 27 in one group, 50 in another group. What I can tell you is these data with these hazard ratios show a very similar thing to what I described before. The hazard ratios in the positive patients for the addition of nivolumab, 0.3, 0.4, a strong association. In those patients who are ctDNA negative, hazard ratios are 0.8 and 0.9. What does this tell us? Again, intervening in these negative patients is not particularly helpful. Two retrospective analyses and one prospective analysis all show really consistent results. This gives us a type of meta-analysis, which is the reassurance that we need that this test is not only accurate but also discriminatory. In conclusion, adjuvant immunotherapy does provide this DFS and OS advantage, specifically with this Signatera assay. We've seen similar results prior in these two retrospective analyses. I'm also confident in saying these negative patients have a much better outcome. This is important because putting these patients in harm's way with immune checkpoint inhibition seems unattractive. Similar approaches should be considered with other histologies. I think expanding this adjuvant population is attractive. One of the things I talked about in the presentation was historically, we've used pathology stage to determine which patients should get adjuvant therapy. In IMvigor010, we put the bar down, and we let a whole load of patients who wouldn't normally be allowed in to be included, the T2 patients, for example, if they were ctDNA positive. This trial showed the subset analysis in these patients, it was also positive. I think it's also important to say that other ctDNA methodologies are likely to give other results. The next slide is a study coming from the NIAGARA study. This is not an adjuvant study. This is a neoadjuvant study, which I presented at ASCO this year. Here, we gave neoadjuvant chemotherapy and surgery, and plus or minus perioperative durvalumab. We showed a 25% reduction in the risk of death associated with perioperative immune checkpoint inhibition. This is the first time we've ever shown a survival advantage in this perioperative setting. We also performed ctDNA analysis. This was performed at baseline with this pie chart you can see here, baseline, pre-surgery, and post-surgery. Baseline BEP is the overall population. You can see here the proportion of patients that are positive in the neoadjuvant setting is higher than the adjuvant setting, 57% rather than 40%. You can see this dramatic drop with neoadjuvant therapy. That drop is higher with durvalumab at 77% compared to the comparator. This underpins sequencing circulating tumor DNA analysis as a better surrogate marker than, let's say, pathological CR or radiological response to these patients. It also post-surgery shows almost all the patients have some form of clearance, but those patients that don't do particularly poorly. This is a strongly prognostic test, post-surgery in the neoadjuvant setting. This is an important test for patients. I think it's fair to say that in the U.S., particularly where this test is now reimbursed, it's being widely used. It's going to have further uptake supported by the IMvigor011 trial, which is positive for disease-free survival and overall survival, but also the NIAGARA study showing that in the neoadjuvant setting, we can get useful information pre-surgery about how treatment has gone. Adjuvant treatment decision-making from the adjuvant setting can benefit from circulating tumor DNA guidance, regardless of the neoadjuvant regime. IMvigor010 showed that with Level 1A evidence and positive meta-analysis data, the duration of ADCED therapy could be tailored by circulating tumor DNA analysis. At the moment, we don't have tools to use and how long we should give treatment. It's possible by getting circulating tumor DNA clearance that we could design trials that would aid us to give an indication of duration of therapies, not just immune checkpoint inhibition, but more importantly, ADC therapy. The frequency of circulating tumor DNA assessment, six weekly, is more attractive than 12 weekly radiology. We are finding cancers earlier. We are intervening more quickly. This is important in the journey to curing our cancer patients. I think the next slide here gives us some indication, just a broad overview of muscle-invasive bladder cancer. Some people say bladder cancer is quite, you know, it's not a prominent cancer. I would quite strongly disagree with that. I think muscle-invasive and non-muscle-invasive bladder cancer is very common. Particularly non-muscle-invasive cancer, the prevalence is huge in the community. Muscle-invasive bladder cancer, there are half a million deaths each year. We're seeing, just in the U.S., we're seeing high incidence with 150,000 cases worth per year. Tom? Yes? I think I was going to take over at this slide. I really want to thank you, Tom. Oh, thank you for keeping going, Alexey. I apologize. I was really keen to keep going. We'll have some more time to have this discussion in the Q&A. First of all. Fabulous. Thank you, Tom. Congratulations to both you and the rest of the collaborators on this amazing work. I personally believe this is a category-defining publication that may have broad implications beyond just muscle-invasive bladder cancer. Before jumping into the muscle-invasive bladder cancer opportunity, let me quickly comment on the flywheel that really drives our progress forward in our oncology business. Our formula is simple and durable. We start with leading-edge technology and constant innovation, layer on unmatched data leadership anchored by peer-reviewed clinical evidence, as we've reviewed today, and excellent customer experience, as well as a broad and talented commercial and medical team that really helps us execute both with customers and academic collaborators like Tom. I think this combination is rare in diagnostics, and it's exactly what enables continued expansion of our category leadership. Now, turning to muscle-invasive bladder cancer, this disease can either be diagnosed as we call de novo or may develop in the setting of prior non-muscle-invasive disease. Taken together, we estimate that there's around 30,000 new cases that are diagnosed in the U.S. alone and over 150,000 cases worldwide. As Tom has mentioned, the clinical journey in muscle-invasive bladder cancer spans neoadjuvant therapy, surgery, adjuvant therapy versus observation, as well as long-term surveillance. Given that we have coverage and strong clinical utility data in each of these settings, we estimate that each muscle-invasive bladder cancer patient qualifies for up to 14 tests over a five-year journey. Today, we estimate that less than 10% of the prevalent testing opportunity is currently penetrated. We expect further acceleration of Signatera adoption in this indication based on both the IMvigor and the CheckMate data. While the majority of muscle-invasive bladder cancer patients have Medicare coverage, for the fraction that do not, we really believe there's a significant opportunity to expand reimbursement based on this data with commercial payers. The last point I think is another important point. Physicians who treat muscle-invasive bladder cancer frequently treat other cancer types, most notably renal cell carcinoma, testicular, as well as prostate cancer. In these indications, we have more and more data that really shows the clinical utility of Signatera. Data like this helps drive a halo effect and broadens adoption, not just in muscle-invasive bladder cancer, but other peripherally treated tumor types. If we go to the next slide, we continue to make significant ongoing investments into generating clinical data in muscle-invasive bladder cancer. We have over 20 studies that are currently running either prospectively or retrospectively, evaluating questions like, can we spare surgery in certain patients? Can we escalate or de-escalate treatment in the neoadjuvant setting? Can we further optimize therapy decisions in the post-surgical setting as well as in the surveillance setting? Two studies of note are the phase II/III MODERN trial that's currently being run by Alliance. The PI for this study is Dr. Matt Golsky. We announced the study in 2024, and the study has now opened in over 300 clinical sites across North America, and it's targeting enrollment of over actually 3,000 patients. The questions this study is asking are actually very relevant. Can we further escalate Signatera positive patients by adding even more immune therapy to further improve their outcomes? Also asking the question about further de-escalation in Signatera negative individuals. Furthermore, we also have previously announced the phase III ARCHER study. This is being conducted with NRG, and this is the GU015 study under their designation. This also randomized study opening in over 100 sites across North America. It actually is looking at another important question of whether a shorter course of radiation can improve outcomes compared to standard of care. In this study, Signatera is incorporated as a prospectively defined secondary endpoint, as well as urine tumor DNA, which is another emerging area in this phase that's being evaluated as an exploratory endpoint as well. Next slide. Taken together, this data has a tremendous amount of implications, not just for patient care today, but also for future trial designs. I think the era of MRD-enriched randomized studies is here, and the IMvigor study significantly de-risks this approach. We believe that MRD or treatment on MRD may become a new line of therapy across multiple histologies. Just like we say front-line metastatic or second-line metastatic treatment or adjuvant or neoadjuvant therapy, there may be a time, and I think this time is happening sooner than even we expected, where there's going to be a new line of therapy called treatment on molecular recurrence. This would be defined as Signatera positive patients in the post-surgical setting. Here, there's going to be a lot of new development looking at can we treat these patients with either immune therapy, other targeted therapies, and thus treating these patients earlier, improve outcomes, and give patients a second chance at cure. This is actually something that was shown again in the IMvigor011 study. Furthermore, adjuvant studies, as Tom mentioned, can sometimes be thousands, if not tens of thousands of patients, and take many, many years to run. There's opportunity for MRD-guided studies to be done more efficiently. I think IMvigor itself is a good example where the ctDNA positive patients who receive placebo, their median DFS was just months. The readout actually was able to occur in, I would say, in a record time for an adjuvant study in this space. This can be further accelerated as surrogate endpoints enter and become not just secondary endpoints, but more and more primary or co-primary endpoints. The way this is done is actually looking at Signatera kinetics or clearance as an early indicator of response. Some of this data has been presented both in IMvigor010 and IMvigor011 studies. The more studies that look at this clearance effect and define the optimal time to evaluate clearance, this may then become an FDA-recognized endpoint. In certain tumor types like colorectal cancer, some of this effort is already underway. We believe that with more data, this may become applicable to even more histologies. Lastly, we believe that Signatera-guided treatment is not just relegated to the adjuvant and surveillance settings, but can be really incorporated broadly across the patient's treatment journey. There are studies either being designed or being opened looking at Signatera guiding not just adjuvant and surveillance regimens, but actually guiding perioperative regimens, deciding which patients need ADCs, IO, can the duration of some of these fairly toxic therapies be optimized to maximize benefit and minimize toxicity. Lastly, especially in bladder cancer, asking the question of can some of these patients who have an exceptional molecular response on top of a good clinical response actually be spared surgeries like cystectomies, which are significant surgeries, have high risks of morbidity and mortality. That, I think, would be a significant advancement in the field if we can show with data that these types of surgeries can be safely avoided in select patients. In summary, we believe IMvigor011 establishes a new standard of care both for post-surgical Signatera-guided immune therapy treatment in muscle-invasive bladder cancer. As we've discussed, not all MRD assays are the same, and data from one should not be extrapolated to others. The best data to establish a clinical utility of an assay is a well-designed prospective validation studies like the IMvigor study that we discussed today. We're continuing to invest, and we're not stopping in muscle-invasive bladder cancer. We have a broad pipeline of studies that's looking across the entire care continuum. This data should be reading out in the next few years with a constant drumbeat of publications and presentations. Most importantly, I think this study de-risks the TOMER concept that we've been discussing with investors for the last four or five years, which we believe, again, establishes a new paradigm in oncology and will broadly apply not just in bladder cancer, but hopefully across multiple other cancer types and really revolutionize how we monitor and then how do we intervene on patients who become Signatera positive and, in the end, markedly, hopefully, improve their outcomes. With that, I want to thank everyone for your attention. To close us off, I'll introduce Steve Chapman, our CEO, to make a few closing remarks. Steve? Great. Thanks, Alex. First, thank you to Professor Thomas Powles for joining us today and for leading the IMvigor011 study. IMvigor011 and the New England Journal of Medicine article represent major milestones for Natera and for Signatera. Many of us have had friends or family members that have been impacted by cancer. I'm really proud to be a part of fulfilling the mission of Natera and the investigators. Finally, before we open up for Q&A, I just want to thank the Natera team members that worked incredibly hard on this. Thank you for everything that you do every day. With that, we'll open it up for Q&A. Your first question comes from the line of David Westenberg with Piper Sandler. You may go ahead. Hi. Thanks for taking my question and great presentation here. I would just want to start off with the discussion. Literally had the title, "IMvigor011 provides level one evidence for intervening on positive plasma ctDNA." Is that a fair statement from everyone involved that they agree with that? What else would the NCCN committee need before acting if you had unanimous belief this provides level one evidence? I have one more follow-up. Thanks so much. Thank you, first of all, for that question. Just a few comments, and then, Tom, I'd love to hear your thoughts. Level one evidence is defined as prospective randomized data across multiple different sites showing benefit. Here, we're actually showing an overall survival benefit. Based on that definition, this is definitely level one evidence. There's strong precedent for studies that lead to FDA approval with a companion diagnostic to be included in the guidelines, at least in the United States. The exact wording, I think that's something we can't control. Tom, I would love to hear your thoughts, not just for the NCCN, but also more broadly across the world, how you expect this to be incorporated into the guidelines. Firstly, there have been no guidelines discussions around this. I'm not told saying anything that this is what I think. I've been involved in most global guidelines and chaired a number of them over the years. This is strong evidence with OS. It is Level 1A evidence in that respect. There are different ways of defining these at times, but Level 1A is, from a guidelines perspective, it's a randomized phase III. To get Level 1A evidence, you usually need overall survival, but not always. The grade comes back to the strength of the recommendation. This will get a Level 1A evidence in this specific environment. I have very little doubt about that. In this specific environment, of course, I'd be very, very—I don't sit on NCCN, by the way, but I would be shocked if it didn't get NCCN endorsement. That would be very unusual. Got it. Just one more. Can IMvigor011 be broadly applied across other immune therapies, checkpoint inhibitors, including atezolizumab and pembro? Do you believe these results are generalizable across other histology types? Would you rank maybe the GU cancers as maybe the most applicable, or is that really not that applicable? Thank you very much. Again, great presentation. Thank you. Tom, do you want to take that question? I would, but I'd like Alexey to talk about the other tumor types because I suspect he might be better at that than me. Look, I think you've asked a difficult question. I think the likelihood is, in the U.S., that there are many people who will see nivolumab as their go-to drug. While I am a purist and would say I would like to use atezolizumab in this setting with ctDNA, as that is the design of the trial, I suspect the community that's currently doing ctDNA testing, by the way, it's ongoing in the U.S. I see a lot of patients from the U.S. who come to me with a Signatera ctDNA testing. I think that's going to increase. I think there will be clinicians in hospitals who are already using nivolumab, and they will choose to use nivolumab, yes or no, based off the result of this test rather than switch across to atezolizumab. I don't think there's anything I can say to change that. There are people who will say that the nivolumab study was positive in this setting. Of course, because they've done their retrospective analysis, which reinforces the results which I talked about, it therefore makes it translatable. Without being too cynical, I suspect that's why BMS have released the results now rather than doing it two or three years ago because they recognize this as a threat and potentially an opportunity at the same time. I think the fact that they've released the results reinforces that they think this is important. Drug companies historically don't have an obligation to do extensive, complicated biomarker work. Of course, if nivolumab was widely accepted in all comers in unselected patients, I suspect they probably would have said, "Well, we'll keep going. We don't need to do this." I think the fact they've released these results is not just scientifically important, but I think it also shows a change in mindset from the clinicians and the pharmaceutical industry that this test is here to stay and will stand the test of time. Therefore, the BMS folks want nivolumab to be selected based off this as well. That's a very long answer, but I think the short answer is I'm keen to keep going with the trial design. I suspect many of my American colleagues will use this with other drugs. Yeah. Just to follow up on that, I think the data is overwhelmingly suggesting that Signatera has a role in both the post-surgical surveillance, and I would actually argue in the perioperative setting with NIAGARA and some of the other results we've presented. While we can't predict which drug will be adopted going forward, I think it is very clear that Signatera has a role and has a pretty broad role for, I would argue, the majority of bladder cancer patients. I think that's the most important point that we're trying to get across, that Signatera will be adopted. How it's used with different therapies, I think, will depend on the individual physician and which drug and which data set they want to utilize to make treatment recommendations for their individual patient. Your next question comes from the line of Doug Schenkel with Wolfe Research. You may go ahead. Hi. Thank you so much for the question. This is Colleen on for Doug. We have a question for Dr. Powles. With testing in the study done up to seven times in year one, how do you expect real-world testing frequency to settle if atezolizumab is approved for adjuvant use in ctDNA positive muscle-invasive bladder cancer patients? Also, IMvigor011 allowed enrollment for up to 24 weeks after surgery with serial testing done every six weeks. Do therapy benefit outcomes differ for patients who tested at six weeks versus those who enrolled in the study at 24 weeks or later on in the study? With that in mind, is there any signal that earlier initiation of surveillance catches more actionable positives or drives larger benefit from atezolizumab? Thank you. If I could answer the second one first because I'm going to answer it in a slightly shorter way, if I may. I think at the moment, we haven't got clarity around the perfect time for the testing. The trial was obviously somewhat pragmatic in an attempt to enroll patients that allowed up to 26 weeks. My preference would be to do a test as soon after the surgery as possible. I think that makes a lot of sense. I think ctDNA clearance happens quite quickly. I would like, you know, the first test 28 days after surgery. That would be my preference. We allowed longer periods in because by the time the patient's been referred to a hospital and they've signed a consent form, it can be longer than that. That's why we were inclusive up to 26 weeks. My preference is to go earlier. The reality is it doesn't matter when you do the test. If you're radiologically negative and you're ctDNA positive in bladder cancer, you need to do something. If you wait until the cancer is visible on radiology, there are about half the patients that's going to turn out to be too late. The sooner you do the test post-surgery, the better as far as I'm concerned. My preference is if you're radiologically negative and you haven't had a test, you should have one, in my opinion. In terms of the data on whether or not there are better outcomes associated with different timings and the implications of that, I'm going to give you two broad things to think about. Number, well, three. Number one is we need to do more work on that issue. We only found the results out a relatively few number of weeks ago. This is an area of ongoing research. I think we could look at ctDNA levels, not just the binary black and white, positive and negative, to see. I suspect the highest levels will be directly post-surgery. I suspect when patients relapse much later, the chances are the levels are going to be relatively lower at defining positivity. I think that's an interesting thing, a piece of research for the future. I also feel that we showed that not only the positive patients benefited at baseline, but the patients who were negative who became positive. For me, that's really important because that says to us, you know, this isn't just baseline results. This is tracking. The tracking piece is important to me because, as I said before, there are some negative patients who do become positive. We salvage those patients. To me, that allows us to reassure those individuals who feel that just having a binary test at baseline is not yet accurate enough. We need to track those patients. I agree with that. I think that switching population from negative to positive, that benefit's important. That's the first question. It was a bit longer than I wanted. I apologize. The other thing you asked about is how frequently one should do the testing. Look, I'm comfortable with eight weeks. Six weeks is what I think we should do. It's not impossible to give patients a series of blood tests. If you remember, you only have to do the sequencing at the beginning. All the rest is just blood tests. What you can do is you can give patients blood tests every six weeks. They don't have to see a doctor every six weeks. They just get a blood test. They get a date. They're having the blood test done. You can look at the results on the computer and monitor it. It's not that to come in and be seen every six weeks. What I would do is I would personally see patients every three months for the first year. I think that's about right after an operation. Then they would get these sequential results done every six weeks. I would stick to the six weeks. If someone came to me and said, "Look, I want to do it every eight weeks," I'd be relaxed about that. If they said they were doing it every two weeks, I think, honestly, that's too much. Of course, we discussed this for a long time within the study group and the benefits and risks and all the bits and pieces and the pragmatism. The pragmatic answer is six weeks feels about right. 12 weekly feels for me too long. That's what we were doing with imaging. That feels too long for me. All right. Thank you so much for that thoughtful response. Just one question on patients who did not test positive on Signatera until later in the two-year window. For those patients who became ctDNA positive later, should they be considered as progressing and perhaps offered a combination therapy? Based on the IMvigor011 results, would you be comfortable with atezolizumab monotherapy used in this population? That's a really good question. You would have seen some atezolizumab data at ASCO GU presented by Professor Go. Brilliant presentation from China. I'd love to talk about that as well. There's also the EV/pembro data, which has superseded platinum-based chemotherapy. The way I look at this, for what it's worth, and by the way, I think it's a brilliant question. I think you can either be a purist or a pragmatist. The purist will say single agent atezolizumab, great outcome, great result, brilliantly, you know, survival, you know, 32 months. That's a really long great outcome for these patients. Frontline EV/pembro, median survival 32 months or now 34 months. I mean, the original cut was 32. And DV toripalimab 31 months. Atezolizumab is generating similar results to those two combination trials. Number one, and number two is neither of those drugs has ever been combinations that have ever been tested in this space. Number three, really important, we need to think about minimizing toxicity in these patients. Therefore, monotherapy with immune checkpoint inhibition is associated with great outcome. If you can clear their circulating tumor DNA, those patients go on and do extremely well. You might say, "Look, let's give monotherapy atezolizumab. If they get circulating tumor DNA clearance, we'll go after it. If they don't get clearance, we're going to give combination therapy." That, for me, sounds like a purist type approach. There's a pragmatic approach where people will say there's been a recent study called 901, sorry, 905. In that trial, we gave perioperative EV/pembro. It outperformed anything we've seen before, including single agent immune therapy. That suggests to us that earlier immune checkpoint inhibition antibody-drug conjugate combinations are associated with superior outcomes. The circulating tumor DNA patients are clearly in harm's way. They count as that 905 population. Under those circumstances, they should get EV/pembro. I can see both sides of that debate. I think the guidelines committees and the NCCN committees will stick to clinical trials. I think that the pragmatists in the U.S. trying to cure bladder cancer will be giving more aggressive therapy upfront. I think that hybrid system where I said if you don't get clearance and then you go in with the combination, then I think that sounds attractive. It's a good question. I don't know how this is going to unfold, but I do know that circulating tumor DNA is going to have an important role in decision-making. Your next question comes from the line of Daniel Brennan with TD Cowen. You may go ahead. Great. Thank you. Thanks for doing the call, Professor Powles. Maybe the first question, just kind of a simple one, but a high-level one. How would you answer the idea that does this study kind of remove the debate around the survival benefit derived from these MRD tests? I think it does. I mean, I think that achieving overall survival with hazard ratios in the 0.5s is pretty compelling. I really enjoyed a number of presentations at the plenary session at ESMO, but many of those never achieved OS. None of the breast cancer presentations achieved that. I'm not suggesting for a second the breast cancer work isn't brilliant and amazing. Certainly, if you came in with a hazard ratio of 0.59 in breast cancer in the perioperative setting, people would fall off their seat and be jumping around the room. Absolutely, if you ask that question. Maybe going back to an earlier question that was asked, maybe just one more. You had the page in the deck on the NIAGARA protocol, which you presented at ASCO. Just a little bit more on this idea. What's been the reaction from that presentation? I think some experts we've spoken with think EV/pembro will become the therapy of choice in muscle invasives. Do you think that, again, back to this idea, do you think ctDNA testing will be applied to EV/pembro? Is it being applied in the NIAGARA protocol? Do you think you're going to need more prospective randomized studies that will really establish the use case in these other therapies? This is what I think. I will say this in every room I go in. It's not for this particular meeting. I think the following. I think that this trial has demonstrated that c tDNA is a very useful test in the postoperative setting in this environment. I think it's discriminatory, it's predictive, and prognostic. It's also fair to say from the NIAGARA study, where we've tracked ctDNA before and after the operation, we've shown at every time point ctDNA is strongly prognostic. Indeed, postoperatively, being ctDNA positive is a disaster. With the hazard ratio, I think it was a hazard ratio of 15 or something crazy. The discriminatory feature and the prognostic component of ctDNA applies in the adjuvant setting, but it also applies very strongly in the context of NIAGARA in the neoadjuvant setting. I don't currently feel that we have done the trials to demonstrate whether or not ctDNA tells us when we can stop therapy. What ctDNA currently tells us is when we should start therapy, and it tells us that patients are in harm's way if they remain positive. It also tells us that becomes increasingly apparent with patients who are on therapy. If you're initially positive, that's bad, but if we start therapy and you're still positive, you haven't cleared, that's even worse. This information is really important. In the MODERN trial, they're looking at ctDNA clearance as a primary endpoint. What I see is I see this field evolving quite quickly. I see people looking at this test as being accurate and discriminatory. It's been validated with Level 1A evidence in the adjuvant setting, but it also has strong evidence in the perioperative setting. I can see it being widely adopted across the board to determine outcome and also tailor-made treatment. Your next question comes from the line of Puneet Souda with Leerink Partners. You may go ahead. Yeah. Hi, Dr. Powles and the entire team. Thanks for the excellent presentation here. If I could ask on the halo effect that Alex talked about, maybe Dr. Powles, I would love to get your view. To what extent, as you talked about a number of drugs that, you know, as a purist, you would use atezolizumab, but Americans would use nivo or pembro. When it comes to this halo effect of using this in other tumor types, to what extent are these data sets actually transferable to those tumor types? How do you think your colleagues would look at these results and find comfort or raise questions based on the tumor types they study and prescribe for? I'll start, but I'd like Alexey to talk about this too. At my presentation, there were a series of ctDNA presentations. It was a plenary session, but there were some colorectal presentations too. I got the impression that these tests were performing differently. They performed different ctDNA tests, and they were getting slightly different results in different cancers. I think the sensitivity and specificity, which is high in urothelial cancer, needs to be shown in other tumor types before it's widely adopted. I think the type of tests that we use is important. I also know because the meeting organizers told me that the reason they wanted to put this in a general session and not as a bladder cancer plenary session with the DV toripalimab in the EV data was they felt that this data has much wider implications beyond urothelial cancer as the first study with this design to prospectively evaluate this question. I think the answer is, in the end, this is going to be broadly applicable. I don't think I have a particular huge bias either to Natera or indeed the field to say that doing X-rays every six months and hoping to identify cancers early is how we'll be doing cancer medicine in 2030. That feels foreign to me. I think there's going to be a very big shift towards accurate ctDNA testing, accurate identification of minimal residual disease. I also believe the global mindset will change away from telling patients, "We're going to do an operation. We're going to cut the cancer out." Then they get a chest X-ray, and they say, or a CT scan, and they're given the all-clear. I think that's a thing from the 1980s. In the future, you'll do the operation postoperatively. If you're ctDNA positive, essentially, you just have early metastatic disease, irrespective of what the chest X-ray shows. Thanks, Tom. Alexey, do you want to? Yeah. Just to add to that, Puneet, I think you can view this as both immediate and kind of intermediate long-term impacts. I think in the immediate setting, a lot of oncologists are using Signatera broadly across multiple cancer types. More and more, people are looking at data from different histologies to form an opinion of is Signatera useful for their patients. I think a lot of people have always said, "Hey, we want to see Level 1A evidence, right? Does the test have predictive ability?" I think this study answers that and reassures a lot of folks who may be using the test on breast cancer or colon cancer that the assay performs in a rigorously well-controlled prospective randomized study. I think that's immediate. I think intermediate and long term, in the last, I would say, week, we have had more meetings with collaborators with the pharmaceutical industry that basically want to replicate this study in pretty much every cancer imaginable. People see this opportunity. There is a huge benefit for being first to do a study in the treatment on molecular recurrence setting. If you think about it, after IMvigor011, if anybody wanted to repeat a randomized Signatera-guided adjuvant surveillance study, they would now have to use not placebo as control, but actually atezolizumab because that's now the new standard of care. There is kind of this race now to see if this can be replicated more broadly. I think you'll see in the next few months more and more announcements happening about additional studies that are being launched on top of the many, many studies already ongoing testing this significant question. Your final question comes from the line of Catherine Schulte with Baird. You may go ahead. Hey, guys. Thanks for the question. I mean, just assuming FDA approval, you know, this would be the first use case of using MRD as a formal companion diagnostic for a drug. Over the last decade, we've seen companion diagnostics looking at PD-L1 expression levels become pretty standard in the immunotherapy world, including for Tecentriq. How do you think the IMvigor results will influence future trial designs for immunotherapies? Do you think the focus will be in trials where PD-L1 status has already been shown to not be predictive, or do you think MRD will really be a frontline approach? Just on that vein, do you think MRD has a similar trajectory as IHC in terms of companion diagnostic potential? My only question, Alexey, then. Let me start there, and then definitely want to hear Tom's thoughts. I think the question or the answer is yes. We do think that there is going to be tremendous interest at looking at this as a companion and in certain settings as a complementary diagnostics to various novel existing therapies, as well as various treatment approaches. I think radiation, I think local therapy. Just like PD-L1 staining kind of opened up a new field across multiple different histologies, we see the same happening with the Signatera result, where selecting patients becomes increasingly important. The exciting thing that we've seen is that ctDNA status typically outperforms many existing kind of traditional companion biomarkers. Take CALGB/SWOG 80702 in colon cancer. There we showed that even in, let's say, PIK3CA mutated patients, ctDNA still identified patients who benefited regardless of the mutational status, both in positive and negatives. In this result as well, we see that there's a benefit regardless of PD-L1 status. Just like Signatera outperforms, as Tom mentioned, all clinical pathological risk factors for prognosticating patients, I think there's more and more data that for predictive claims, Signatera may outperform many existing biomarkers that are being utilized. I don't think that's going to be universal, but it's exciting to kind of see just how powerful the test has been in multiple randomized studies in terms of its predictive ability. Tom, your thoughts? Yeah. Alexey, I don't have much to add. I agree with you. Some of the work, if you look in the original Nature publication, which we published with IMvigor010, which identified this as a provocative and exploratory analysis, we looked at RNA and DNA of the ctDNA positive and negative patients. While the patient's characteristics were not that different from one another in terms of being ctDNA positive, essentially, almost all cancers have the potential to become metastatic. When they do, they have their ctDNA positive. I think it's true to say two things. The first is that not all tumors shed ctDNA in the same way. We know, for example, renal cancer has a lower level of shedding than bladder cancer. We also know that within tumor groups, the shedding is not identical. The luminal type tumors might have a slightly lower shedding of ctDNA than the basal tumors. Finally, I think it's also worthwhile noting that in urothelial cancer, perhaps not in other cancers, there may be a link between tumor biology and ctDNA positivity in the knowledge that not all tumors secrete ctDNA at an identical concentration. There may be a link with immune biology. That might be why it works well in the ctDNA positive patients and the hazard ratio in the ctDNA negative patients is 1.1. I actually genuinely see a world. It's not here today. I don't love the PD-L1 biomarker, by the way. I know it's used in breast cancer. I gave another talk at ESMO Congress about ADCs, breast cancer. My feeling is that the assumption that PD-L1 negative patients don't respond to immune therapy and PD-L1 positive do is an oversimplification. We know innate and adapted immunity both have an important role to play in predicting response. In the future, I can see platforms or multiple platforms where one looks at tumor expression and from the tumor itself of DNA or RNA alterations and link that with or without ctDNA positivity. I think in the end, I hope biomarker work will involve both. Great. Maybe last one for me. I believe the BTI assay was used for patients in China. Were there any differences in performance for that assay versus Signatera? I know the answer to that question. The answer is that although the subgroups were somewhat smaller from China, the trends were in the same direction. I think we showed the, I can't remember if it was a PFS or the OS. I'm sure it'll be in the New England Journal of Medicine paper too. Certainly, we addressed this in the New England Journal of Medicine paper. The answer is that we showed trending going in the same direction. There's nothing there that doesn't suggest it doesn't work in a very similar way. This concludes the Q&A session and today's conference call. You may disconnect.