Roche Aktie

Welcome to the Roche Virtual ADA investor event. My name is Henrik, and I'm the technical operator for today's call. Kindly note that the webinar is being recorded. [Operator Instructions] At this time, it's my pleasure to introduce you to Bruno Eschli, Head of Investor Relations. Bruno, the stage is yours.

Thanks a lot, Henrik. And could I have the first slide, please? So welcome to our fifth IR call in 2026, covering new Phase II results for enicepatide, previously known as CT-388 and petrelintide in obesity which just got presented at ADA in recent days. Today's call is scheduled for 90 minutes. We have planned for roughly 50 minutes for the presentation and then 40 minutes for Q&A.

Let me quickly take you then through the agenda. Our first speaker today, as you can see, will be Morten Lammert, our Global Therapeutic Area Head Cardiovascular, Renal & Metabolism. Morten will provide a quick update on our overall obesity strategy before our second speaker, Manu Chakravarthy, Senior Vice President and Global Head of Product Development, Cardiovascular, Renal and Metabolism, will take us through the latest from the obesity pipeline, including an update on the study start of a Phase II multi-arm fixed-dose combination study called SYNERGY for enicepatide and petrelintide in obesity plus/minus type 2 diabetes.

Our third speaker, as you can see, will be Ildiko Lingvay, MD, MPH, MSCS and Professor of Medicine in Endocrinology and the lead investigator for CT-388 -- for the CT-388-103 study. Dr. Lingvay will take us through the Phase II enicepatide results in obesity. And finally, we'll have a presentation by Louis J. Aronne, MD, previous Director of the Obesity Society, Founder and Chair Emeritus of the American Board of Obesity Medicine and Professor of Metabolic Research. Dr. Aronne is known to the broader scientific community for his foundational role as a pioneer who helped shift the medical perception of obesity from a just behavioral failure to a biological treatable chronic disease.

He serves as a principal investigator and lead clinical trialist for key anti-obesity trials, including petrelintide, and he will take us through the Phase II ZUPREME-1 petrelintide results in obesity. Following these presentations, we will have our 40 minutes Q&A. And in addition to our 4 presenters, we will be joined for the Q&A by Louis André Villeneuve, our lifecycle leader for enicepatide and by Marisel Salzger, our lifecycle leader for petrelintide. And with that, let me hand over to Morten to get us started. Morten, please.

Thank you, Bruno, and good morning from New Orleans. I'm Morten Lammert. I'm the Global TA Head for CVRM, and I thank you for joining us today. I'll not take a lot of time as I know the true highlights of this call is coming after my presentation. But I would like to take this opportunity to provide you with a high-level overview of our strategic blueprint we are executing to become a leader in CVRM. Obesity is a core focus within our CVRM strategy, driving our ambition to deliver innovative and transformative solutions for patients.

Our ambition is clear. We aim to become a leader in obesity. To achieve this, we are focusing on securing a strong market entry before year 2030 and differentiating our opportunities and offerings through the Roche unique capabilities. We have defined strategic pillars to deliver on this commitment. Firstly, we must deliver the near-term portfolio. We are focused on advancing our late-stage assets into Phase III at speed while preparing for launch of our diverse portfolio.

Then we are expanding and differentiating. We will unlock the potential of combination therapies and iterate on validated mode of actions, while over time, also developing novel treatments with transformative potential. And not only treatments need to be transformative, also the care delivery needs transformation. We are thinking beyond the molecule, leveraging our diagnostic capabilities and ensuring we meet patients where they are in their individual journeys.

Next slide, please. The global obesity pandemic represents a monumental challenge to the individual, to the health care systems and to our societies. If prevention and treatment measures do not improve, the global economic impact of obesity could reach USD 3.5 trillion by 2035. By that time, projection shows that more than half of the global population will be living with either overweight or obesity. Despite this health challenge, the treatment rate is still far below that of other chronic diseases. Data suggests that only 10% to 15% of all the eligible patients in the U.S. are currently treated with a branded anti-obesity medication. And treatment is also initiated too late at a point where the BMI is above 35 and many of them have already established comorbidities.

With new and better treatments, we believe we can shift the curve and treat more people and treat them earlier. This would allow us to not only address severe obesity with obesity-related complications, but also focus on health preservation and early disease -- in the early disease stages.

Next slide, please. To successfully engage the patients earlier, we must go beyond one-size-fits-all treatment paradigms. Our strategy is taking point of departure in the patient-centric insights, shifting to tailored solutions that meet the needs for people with obesity. We have studied how patient needs and decision drivers change as a function of their disease states and their journey. If we are zooming in on the far right column, we focus on patients with obesity Class II or Class III and often with established obesity-related complications.

The treatment of choice deliver maximum efficacy and has demonstrated benefit on the relevant comorbidities to achieve -- willing to accept some side effects of the treatment and discomfort that to get that more and higher potency. Left column are people earlier in the disease journey. They know it would be healthy and good for them to lose pounds, but would most likely not consider themselves as living with a chronic disease. Majority of people in this group are looking for 10% to 15% weight loss that can be achieved in the most gentle and the most convenient way. In other words, they make a trade-off between the efficacy, the convenience and in particular, tolerability relative to the first group.

This is why we need patient-centric approaches that is critical to deliver treatment solution matching the different patient needs. Our go-to-market strategy aligned with these changing needs from specialized care for the high-risk patients to primary care and direct-to-patient approaches for early intervention and for health preservation. Next slide, please. Lastly, we are entering the next era of obesity management and expect continued and significant market growth over the coming decade. In addition to pricing and accessibility, we see 3 main drivers of this continued market growth. If we start from the bottom, a continued inflow of AOM naive patients is driving the market expansion today and will continue to contribute to market growth over time.

AOM experienced users will, over time, become a significant part of the future market. This is not the classical switch patients, but represent the cyclical nature of AOM treatment and the market will require new solutions to reengage those patients who, for one or the other reasons, decided to stop their chronic care. Treatment persistence will continue to improve with very low level -- from the very low level today. Education, disease understanding and improved treatments and maintenance strategies will increase adherence and the long-term use.

This market will continue to be hyper dynamic, and we believe our pipeline is very well positioned to offer the right treatments across a heterogeneous patient population from early intervention to late-stage comorbidity management. With this, I would like to hand over to Manu for a pipeline update. Thank you for your attention.

And thank you, Morten, and a warm welcome on my behalf as well, and good morning and good afternoon. Good morning, certainly from New Orleans. So I'll take you through the pipeline update and give you a little bit more of the progress that we have made since our last time together. So if you go to the next slide, please. So -- over the course of the last 6 months, we've had a significant progress across multiple fronts. And I'll walk you through some of those key developments through this slide. It's a little complex, but I'll walk you through this asset by asset. So the first important advance that we've made in -- since the beginning of the year is to advance zilebesiran, which is our once every 6-month angiotensinogen siRNA for uncontrolled hypertension in a cardiovascular outcome study that we call ZENITH.

It's 11,000-person study, and happy to report that we're making really substantive progress in recruiting that study now, and we're making really good headway there. In terms of enicepatide, which you'll hear a lot about from Professor Lingvay in a second. But again, very happy to report not only the presentation of the 103 study at this conference, but more importantly, the advancement of this program into Phase III, so into chronic weight management with 2 pivotal studies that we call ENITH-1 and ENITH-2, which is in people with and without type 2 diabetes and obesity and overweight.

We also have a fairly comprehensive program, which you'll hear at our next update in terms of how we're developing it for glycemic control for a cardiovascular outcome study. So really a comprehensive package to really position enicepatide as a potential best-in-class dual GLP-1/GIP agonist. In terms of additional progress, pegozafermin, which we have now completed the full integration with 89bio. So the company was acquired in October of 2025. And since then, we've completed the integration. And again, happy to report very good forward progress here with the 2 pivotal studies, which we call ENLIGHTEN fibrosis and ENLIGHTEN cirrhosis, so both progressing quite well. And then petrelintide, which you'll also hear a lot about from Professor Aronne in a second. And we also presented data on the ZUPREME-1 study here in ADA. So that has yielded very exciting results for us.

And those results, along with enicepatide data, give us a lot of confidence to then do a combination study that we call SYNERGY, which I'm going to give you a little bit more of a flavor in a couple of slides. So that's a combination study with enicepatide and petrelintide. So that is to be initiated in the second half of 2026. And we've also made the decision to advance petrelintide monotherapy into Phase III in the second half of 2026. Then we have our oral synthetic molecule, which we call CT-996, which is a GLP-1 receptor agonist, again, anticipated for broad indications, both in obesity, but as well as in type 2 diabetes. So we're in the midst of our Phase II program for both obesity and type 2 diabetes.

We're making good progress there. We anticipate that we will be making a Phase III go decision in second half of this year as well. And then we have our ongoing what we call the GYMINDA study, which is the anti-myostatin plus incretin combination study, which is on track to be reading out its Phase II data also at the end of this year. Then we have acmopatide. So we've made the difficult decision, which is entirely strategic to not advance this program for type 1 diabetes into Phase III. So as you may recall, acmopatide is a once-daily dual GLP-1/GIP agonist that we had developed for type 1 diabetes.

At this conference at the ADA, we presented at a symposium the really good efficacy and safety data. And the results certainly speak for itself, but it was a strategic decision not to advance it into Phase III. Just want to emphasize here that while we are making a decision on a molecule, it doesn't mean that we're not committed to diabetes for both type 2 and type 1. So we remain highly committed. And you can see from our portfolio, there's actually some very interesting choices that could potentially be made to really ask ourselves what's the best molecule, what's the best mechanisms that we can bring to bear to really help people with diabetes, both type 1 and type 2.

And then, of course, we have afimkibart for MASH and then NLRP3 in an earlier stage of a program for cardiovascular disease. So if you go to the next slide then, please. I'll give you a little bit of the quick snapshots on 3 programs. Obviously, we won't have time to go through all of them in detail, but I want to highlight 3 programs, and then we go deeper into the data. So enicepatide is our fully biased dual GLP-1/GIP receptor agonist, and I'll explain a little bit better what that really means in a second with the next slide and what does bias really bring. But on the right, just to emphasize a little bit further what I said in terms of where we are. So we've now completed the full Phase I program, Phase II program for obesity is completed.

We're in the midst of almost nearly completing the type 2 diabetes study, which we call the 104 study. Results are expected in second half of '26. And then all of these 4 Phase III programs that are either started or soon to be started are all underway. So in terms of what is special about this molecule, we thought it might be best to explain it with a little bit more of an animation. So if you go to the next slide, please. This introduces the concept of signaling bias which is really the mechanism of action which provides this molecule.

[Presentation]

So, of course, that signal bias, we've seen the benefits of it in preclinical models, several different studies, as you can see published here, and the evidence continues to accumulate, and we have some early hints of our Phase II data that might support its benefits. And of course, Phase III will be the ultimate proof in the pudding, if you will, of how a signal bias molecule may eventually translate to better efficacy and potentially better tolerability. So now let's move to petrelintide. Just as a quick high-level update here. Again, Dr. Aronne will speak a lot about the results, so I'll not go there. But just suffice to say, in terms of its development program, where we are, -- so again, we've completed our Phase I.

We've completed ZUPREME-1 Phase II, and we're awaiting data from ZUPREME-2, which is the overweight obese type 2 population, and that data is expected at the end of this year. And then as I mentioned in the pipeline update, we've advanced the monotherapy program for Phase III and then excited now to present to you the kind of a study design overview for what the SYNERGY trial is going to look like. So if you go to the next slide then, please. So this might seem a little complicated, but I'll kind of walk you through kind of the core principles of how the study is really designed.

So we wanted to do a comprehensive Phase II program, so where we can study both the mono components and the combo components all in one study so that the end goal coming out of the Phase II is to be able to identify the ideal regimen of the best risk-benefit ratio, if you will, of the best efficacy, best tolerability that then we can take into a streamlined Phase III design so that we don't end up having to do monotherapy in Phase III, if possible.

So you'll see in the bottom 3 rows, arms 4, 5 and 6 really represent placebo, petrelintide monotherapy or enicepatide monotherapy. And then the top 3 rows, #1, 2 and 3 are 3 different regimens, 3 different doses, combinations, if you will, of petrelintide/enicepatide combos that we are studying in a fairly comprehensive manner to identify, as I said, the best ideal combination that we can get to with tolerability and efficacy as the key center pieces. So this is a 40-week study. This is going to be in people with obesity or overweight with at least one comorbidity. And this one is -- this one depicted is not in people with type 2 diabetes.

And so the expectation is that emerging out of this trial, we will have identified the right dose for each and that we will take into a fixed dose combination into the Phase III study. This specific design is not a fixed-dose combination. It's a loose combination. So we're giving 1 injection of petrelintide, 1 injection of enicepatide separately. But once we identify what the right combination is, then we can make a fixed dose.

So this is, again, slated to start in the second half of this year. Now my final update on the pipeline per se is on the next slide, which is our oral molecule, CT-996, if you go to that one, please. So you might recall, again, from our last update that we had, we were in the midst of our Phase I program, which is fairly comprehensive. It included people with and without diabetes. And so what you're seeing on the left there is just a quick refresher of the weight loss that we saw over 48 -- sorry, over 4 weeks in people living with obesity, but no diabetes. And so we saw a weight loss of up to 7.3%. Part of that study also included people with diabetes.

I'm happy to report that, that has now been accepted to the EASD, and we can present the type 2 diabetes cohort there and that will have completed our full package for Phase I. And then we are in the midst of 2 Phase II studies, which are also very comprehensive programs to really do a proper full dose ranging across multiple doses in both people living with obesity and obesity and diabetes. So that's what we call the 2 sort of the Phase II studies and then ultimately making a decision based on those data, whether to take this forward into Phase III towards the end of this year.

And then my final slide before I hand it off is just to kind of recap kind of where we are. So you heard from Morten very nicely that there are many unmet needs, and there is a wide heterogeneity of people living with obesity. And then we need to meet people where they are in their journey because obesity is a chronic disease. It's relapsing and it's heterogeneous. So our approach has been a holistic patient-centered approach really anchored on what are the core unmet needs that we are trying to really satisfy. So you heard about tolerability, you heard about the plateauing. You heard about the weight maintenance challenges, the comorbidities, 2 out of 3 people living with obesity have at least 1 comorbidity.

If you lose too much -- too fast body weight, especially in a vulnerable population, elderly population, muscle loss is a significant concern. So that's an important issue. And then 1 out of 5 people do not respond to an incretin. So having more mechanism of actions to treat such a heterogeneous disease is a must-have. So those are our anchoring pillars. And so when we ask those questions of what can we do in our portfolio to satisfy those pillars, you can see every one of our assets that we have carefully and thoughtfully designed and brought together is there to address those unmet needs.

Again, I'm not going to go through it all again, but just simply to provide this visual to say that the way we have constructed our portfolio is really anchored with patients as our North Star and to ask ourselves, where are we going to make the biggest impact to help our patients. So those are the areas. And it's both in monotherapy and in combination therapies that we can bring our portfolio to bear. And then the final piece I will just emphasize are 2 other things. As Roche, one of the advantages that we have is also the benefit of multiple therapeutic areas.

So not just CVRM, but we also have neuroscience, immunology, oncology, et cetera. And so we know that there are over 200 comorbidities of -- in people living with obesity. So we're also looking at what are our other assets within our other parts of our therapeutic areas that we can bring in combination at the right times. And then the final piece is, of course, the diagnostics and the therapeutics, both residing under the same umbrella, where I think we all can appreciate how important it is to not only diagnose, monitor, but also prognosticate how patients are doing.

And so having a diagnostic arm that works in tandem hand-in-hand with our therapeutics arm is another significant advantage that we see and that we want to bring to bear in its totality to really help patients. So with that, it's my true pleasure to hand it off now to Dr. Lingvay to walk us through some of the core data pieces that we presented actually both today, and we also had a late-breaking poster yesterday. So Dr. Lingvay.

Fantastic. Thank you so much. Hello from New Orleans as well. And it's my honor to present the 103 study results. We can get the next slide. And that's on behalf of a number of co-authors that have worked really hard on this study as well as all the study participants and the study staff that were caring for these patients every day. These are my disclosures and these extensive collaborations with industry, along with my firsthand experience with treating patients with metabolic conditions and extensive research experience have really informed the way I look at the field and the way I think about what's most needed in our field.

Let's move on. So enicepatide, as you've heard, it's a dual signal biased, both GLP and GIP receptor agonist. It's a once-weekly subcutaneous injection, and it's being developed for treatment of obesity, type 2 diabetes and weight-related comorbidities. You might remember that the Phase I trial, which was 24 weeks long, had very impressive results. The highest dose in that trial was 22 milligrams of enicepatide, and it produced reductions in weight of up to 18.9% in people without diabetes living with obesity.

It's -- also its safety and tolerability were very reassuring, which means that this compound moved into Phase II. And this is what we're hearing about today. It's the Phase II study, which was intended to evaluate the efficacy, safety and dose response of enicepatide in participants with overweight, obesity, but without diabetes. Let's move on.

There you go. All right. So this study was a double-blind, placebo-controlled multicenter trial conducted exclusively in the U.S. It enrolled participants -- adult participants between the ages of 18 and 75 and they must have a BMI over 27 with at least 1 weight-related comorbidity or BMI over 30 and no diagnosis of diabetes. Participants that were eligible were randomized to 5 different doses of enicepatide ranging from 4 to 24 or placebo. And the primary endpoint was reduction in weight at week 48.

There were a number of secondary endpoints as well. The important one that I'm going to point out is the proportion of participants with normal glycemia among those who had prediabetes at baseline. Let's move on. This is a study design. And you will see that the lowest dose, the 4 milligrams did not have any titration steps. And then the higher the dose, the more titration steps as it should be. And the highest dose was reached after 4 titration -- 5 titration steps. Of note, which is important when we're also looking at the results is that participants had to come in weekly to the study sites to receive the medication.

This obviously influences the practicalities of getting the medicine and participating in the study. And also in this study, there was flexibility on dose titration, but to some degree, people were able to down titrate or slow down the titration if needed for safety, but they must have been on a minimum of 4 milligram of study medication in order to stay on the medication. And also, there was that point at 27 weeks beyond which titration was not allowed. So whatever dose they were able to titrate up to at that point, that's what they stayed on for the rest of the trial to capture that steady-state effect of respective dosages.

Let's move on. All right. So on the left, I'm showing you the baseline demographics, which are fairly typical for participants in obesity studies. Age is 49, predominantly female participants, BMI of 38 and an A1c that's well in the normal range, but there were 50% of participants who had prediabetes at baseline. I also want to point out that there were a good representation of African-Americans, which gives us confidence of the fact that the study results will be representative of a larger population that will be studied in Phase III and the population specifically that we tend to use these medications in the United States.

On the right, I'm showing you the disposition. And you will see that everybody received -- almost everybody received at least one study medication. And then you will see on the second line, the number of -- the percentage of participants who reached the full dose of the study medication. Remember, by week 27, if they were not able to get to their assigned dose, they could not uptitrate after week 27. So in this context, I'm happy to report that those on the 4-milligram dose achieved 99% of participants got to that dose, 8 milligrams 97% and then in the highest dose, 80% of participants were able to get to that 24-milligram dose.

You will see the study completion that is quite good for the obesity studies. If you followed ADA this year, you would see that there have been struggles in the field with keeping participants in the study. You will see that 34% of participants in placebo stopped the -- I mean stopped participating in the trial and stopped the medication. And that's not unexpected for the field. People now have choices. And if they see that they're not getting the responses that they're looking for, they no longer stay in the study. And also, this number is actually very good for the field as it stands currently.

Among people who received enicepatide, anywhere between 14% and 31% stopped the study medication. And I have to point out that I think this is a small study and the 12-milligram dose had an unusually high discontinuation rate, but it stands out, and I would look at the totality of the data and see that in all the other arms, including the highest dose, the 24-milligram arm had only 19% discontinuations. And if you look on the last line, that's the most important one, how many people actually discontinued due to adverse events. And this is not just GI adverse events, any adverse events. And you see that the discontinuation rate because of adverse events is extremely low.

Let's move on to the study results. There you go. This is the most important slide. So people treated with the highest dose, the 24 milligrams of enicepatide reduced their weight down by 22.7% at the 48-week mark. That represents a placebo-substractive difference of 22.5%. And you would appreciate that, that line is quite steep, still going down. So it's anticipated that with ongoing care and ongoing treatment, people will probably lose quite a bit more than 22% of their body weight.

On the right, I'm showing you the treatment estimate. And with the highest dose of 24 milligrams, the placebo-subtracted weight decrease was 18.3%. I also want to point out that even the lowest dose, the 4 milligram, which was required no titration. It's a one-stop shop, achieved quite meaningful weight loss with a placebo -- I'm sorry, with 6.7% body weight loss by the end of the 48 weeks. Let's move on. So these are important secondary endpoints. On the left is the categorical weight loss data. And you will appreciate that the overwhelming majority achieved 5% body weight loss. When you look all the way to the right-hand side, and it's probably unprecedented to see 30% weight loss category in these graphs.

It's something that we just recently added to our graphs in various -- for various studies because not until long ago, that was not even a number that we would dream of. But 26% of the people treated with the highest dose of enicepatide achieved 30% body weight reduction. That's in the range of bariatric surgery, and that's quite impressive, especially coupled with the on-treatment retention that we have seen in this group. On the right, I'm showing you the number -- the percentage of people who achieved normal glycemia if they had prediabetes at baseline. So you will see at the bottom what number -- what denominator we had for each of these group.

Remember, about 50% of participants had prediabetes at baseline and up to 73% of people just within 48 weeks, they reversed from prediabetes to a normal glycemic state. Let's move on. And these are the side effects. And overall, pretty well tolerated for a GLP-1 receptor agonist. Participants who withdraw due to side effects, I mentioned to you before, it's a pretty low number. But let's look at the participants with any GI adverse events. So that means either they had nausea, constipation, diarrhea or vomiting. And you see those numbers compared to placebo at 28%. These range between 47% and 55% across the various enicepatide treatment groups.

Nausea was 19% in placebo and in the enicepatide groups, it ranged from 33% to 46%. Constipation, around 25% across the various groups, diarrhea around 15% with the highest percentage in the highest group. And then vomiting was quite a good number. I know it's never good for anybody to have vomiting. But compared with other GLP-1s, especially GLP-1s that are able to achieve the amount of weight loss that we have seen in this study, these numbers are quite impressive.

So with the lowest dose, 8% and then the highest dose, 21%. Most of these events were mild to moderate, very few Grade 3 events. There were no grade 4 and 5 events reported. Next, please. All right. So let's conclude. This was the first Phase II trial of enicepatide, and it demonstrated clinically meaningful, very statistically significant dose-dependent weight reductions of up to 22.7% at 48 weeks of treatment.

In the highest dose arm, the 24 milligrams, the percentage of participants achieving 10% and 20% weight reductions was 87% and 47.8%, respectively. Among those with prediabetes, 73% had a reversal to normal glycemia. And then I've shown you that fewer participants actually discontinued treatment in the active groups than in the placebo, underscoring the favorable benefit risk profile. Thank you, and I will pass on back to the team.

Hi. I'm Dr. Louis Aronne, and you've heard my introduction before. I was involved in the development of the -- of this trial, ZUPREME-1, which was presented here at the ADA in New Orleans, where I am across the street. I can see the convention center across the street. I'm looking at it longingly because I haven't been there today for the first time in 5 days. I want to point out that in a different universe, amylin analogs would be the primary drugs for the treatment of obesity. When amylin -- the first amylin analog, which was pramlintide and exenatide, first GLP-1 were approved in 2005, almost simultaneously, we found in using them initially for weight loss that pramlintide and the amylin action seemed a little bit better than the GLP-1s because it had fewer side effects.

It was more tolerable and produced very good weight loss. And we published several papers back then looking at the weight loss. But for reasons that I still can't understand, but I'm sure the industry does, it was not developed for obesity at that point in time and was used for a very narrow indication, which was along with insulin in people with type 1 diabetes. But now we're beginning to see that there are amylin analogs. May I have the next slide, please? And what I'm going to be presenting are the results of the Phase II trial of petrelintide, this new human analog. This study was led by Tim Garvey from the University of Alabama. I was on the steering committee for this trial and helped to develop the treatment paradigm.

Next slide, please. And these are Dr. Garvey's disclosures. Next slide. So you're well aware that obesity is a heterogeneous chronic disease and pharmacotherapy we're seeing makes a lot of sense. Despite many advances in the development of obesity medicines, treatment persistence is the big issue, and it remains low for a number of reasons, including side effect profile. Petrelintide, which has a half-life of 10 days is being developed as a weekly injectable for long-term obesity pharmacotherapy. And the objective of this trial was to compare the efficacy, safety and tolerability of various doses of petrelintide versus placebo in people with obesity or overweight and with complications.

Next slide, please. So here is the trial paradigm. It was a randomized, double-blind, placebo-controlled parallel arm to parallel group, 10-arm trial. Each dose arm had a placebo arm, a matched placebo arm. The baseline characteristics you should note are that there are 485 subjects in total, age 47 years. Women comprised only 53% of this group. And you saw that Dr. Lingvay presented about 68% women. And in many trials of GLP-1s, it's 75% and even higher. This is important because what we have seen with the GLP-1s and perhaps now with other trials of amylin analogs is greater weight loss in women. It's kind of the reverse of what we used to see with diet alone.

And so this may have an impact on the results on the magnitude of the weight loss. The mean BMI was 37, body weight was 107 kilograms and subjects were not to have type 2 diabetes, although they could have prediabetes. The dosing was weekly and it escalated every 4 weeks as the subjects tolerated. The primary endpoints were the percentage change in body weight from baseline to week 28 and the secondary or exploratory endpoints included change in body weight at week 42, change in waist circumference at week 42, change in cardiovascular risk factors and treatment-emergent adverse events. So you can see 42 weeks. The titration went up to week 28 and then the dose remained stable, and then there was a following -- a follow-up of 9 weeks.

Next slide, please. Here is the efficacy estimate for the results. And you see in the pooled placebo arm, 1.7% mean weight loss up to week 42. And the various doses, which you can ask the company about, the apparently lowest dose, dose 1, 8.7%, up to dose 5 was 10.2%, but dose 3 in the middle in pink was 10.7%. That seemed to be the greatest efficacy. So placebo subtracted weight loss of 9% to week 42, but the trend continues to be down.

Next slide, please. If we look at cardiovascular risk factors, and this is a small selection of them, we see that waist circumference, which is a very good summary of where the cardiovascular endpoints are going, ranged from 7.9 to 10.8 centimeters, very good reduction in waist circumference and proportional to the amount of weight that was lost. If we look at C-reactive protein, a key measure of inflammation, it went from a 17% up to a 41% reduction in the various doses. And triglycerides similarly went from a 12% to a 21% reduction over the period of 42 weeks of treatment.

Next slide, please. Now here is the most important slide for this trial, and that is pooled petrelintide versus pooled placebo adverse events. And you can see that in the far left column, they're virtually identical. And this is I don't want to say unprecedented, but we have not seen so few adverse events in a treatment group compared to placebo, at least I haven't. And if we look at serious adverse events, you see that they are identical. This is really, really a striking finding. So that while the magnitude of the weight loss may not be what some people would want, it may just take longer, the safety profile here is quite remarkable, really remarkable.

In the middle, we see the mild, moderate and severe side effects. And you see, again, parity and perhaps fewer severe side effects in the severe group. So there were fewer severe side effects in the treated group versus the pooled placebo group. And finally, in the right column, you see adverse events leading to discontinuation and virtually identical. So a very, very good side effect profile that I think if side effects are an issue for a patient, this would be what I would use.

Next slide, please. Here, we look at selected GI adverse events, and most of them were mild. You could see that over 3/4 were mild, and they were reported during dose escalation. That's typically what we see in our trials is as whenever the dose goes up, you may have an increase in the incidence of adverse events. So we see nausea, 19.6% versus 6.2% in the placebo-treated group. Diarrhea, exactly the same. Constipation, slightly greater in the treated group and vomiting lower, 6.2% versus 3%.

So reduced incidence of vomiting in the petrelintide group. So maybe it's a treatment for vomiting as well, just kidding. Next slide, please. Here, we see the adverse events based on the preferred terms, and this is in those -- which were reported in greater than 5% of participants. You can see on the top line, the nausea numbers, which ranged -- so we see in the placebo group, 6.2% in the petrelintide group ranging from 12.3% to 28%. We see some fatigue, which we often see. Again, clinically, I'm not sure how important that is. Diarrhea, a small increase. Injection site reactions, not many, but there and then constipation. And then the usual side effect profiles that we see in our clinical trials.

The one I'd like to point out at the bottom is vomiting. If you look at vomiting, you see that the numbers ranging from 3% -- from 0 in 2 of the doses to 8.5%. And this variability is undoubtedly because of the small size of this trial. But the point is that these are very, very low numbers in the pooled group, 3% versus 6.2% in the placebo group is very, very exciting to see because vomiting is extremely unpleasant in some cases and is a reason why people drop out of our studies. Next slide, please.

So in summary, in this Phase II dose-finding trial, petrelintide demonstrated statistically significant and clinically meaningful weight reduction across all dose cohorts at both week 28 and through week 42 in a gender-balanced population. And again, I think that if you do the math, there should be significantly greater weight loss if women -- if there are more women and if women respond as they do in GLP-1 trials. The treatment was associated with improvement in cardiovascular risk factors. It was well tolerated with rates of GI adverse events and serious adverse events that were similar to placebo.

And in conclusion, these findings support the potential of petrelintide as an effective and well-tolerated obesity medicine with a distinct mechanism of action, could be added to other compounds, for example. And these properties offer the opportunity to improve longer-term medication persistence because I think the reduced side effect profile is going to be very exciting. So thank you very much. And the Phase III studies are scheduled to initiate in the second half of 2026, and we eagerly await them. Thank you.

Thanks, Dr. Aronne. And with that, we will open the Q&A session.

The first questions go to Graham Parry from Citi.

So firstly, a question for Dr. Lingvay. The 12-milligram group, the CT-388 had a very high 14% dropout rate due to adverse events. It was double what we saw in 24 milligram. Just if you can -- is there any additional data around what explained that? Was there any specific sort of AE that was causing the dropout there? And then secondly, a question for Roche. Are these the dose titration schedules that you'd be looking to use in Phase III monotherapy? And then the last question is when you're talking about looking at fixed dose combinations for Phase III for the combination, is that fixed dose in a single device, so unlike what we've seen with CagriSema, for example, where you have a dual chamber pen.

I guess I can kick it off with my question. And this is a Phase II trial with small numbers of participants. So you would imagine that any sort of occurrence of extra 1 or 2 events will influence the numbers quite significantly. And that's why I said that you need to look at the totality of the data across the 5 treatment arms and look sort of at the trend across them. The 12-milligram group was definitely a little bit odd, but the 14% dropout rate due to adverse event was not any -- due to any specific adverse event. It was just a measuring of this and that. I also want to remind you that these participants had to be at the site every single week to receive their injection. And that does influence your willingness to stay in a trial when so many other external factors might influence your ability to come to the site on a weekly basis.

And I'll take the other questions on the FDC and the titration. So the FDC -- so one of the advantages of petrelintide is that it doesn't fibrillate. And therefore, it is compatible with other peptides. So we are able to take advantage of that property and therefore, provide that into one single cartridge. So it is not a dual chamber device. It's a single chamber device. Both will be in the same cartridge. Regarding titration, so this is an every 4-week titration. But as we learn from the field, and both Dr. Aronne and Dr. Lingvay can add to this, we know that patients will require some degree of flexibility. And so in all of our trials, we have been very thoughtful in terms of providing such flexibility.

So while it's every 4 weeks, there are provisions in the protocol that allow some flexibility to take some time to get to the next dose if they need to. There's a flexibility to down titrate. There's a flexibility to stay on the same dose. And so there's multiple avenues provided to participants to get to their assigned top dose. So flexibility is an important component of it, but the titration is on protocol, it's every 4 weeks.

Graham, did this answer all your questions? Do you have any follow-on questions?

So actually, one follow-on. Just in terms of the combination, why would you be varying the petrelintide dose? I mean it looks like you just took dose 3, which looks like it's already optimized with no vomiting and then varied the CT-388 dose, that would be the sensible way to -- why do you need to vary the petrelintide dose in the Phase III?

So we haven't disclosed all of the doses yet. So when we do that, you'll see the full dose ranging there. So we're going to take a full look at all of the factors now that we have the full data set. But the goal here is really to try to find the optimal balance. As you can imagine, we want to maximize tolerability while also really maximizing efficacy. So I think it's a balance that we will look at.

Very good. Then we move on. Next questions go to James Gordon from Barclays. James, please?

James Gordon from Barclays. The first one was on eni as a monotherapy. So is that you show superior weight loss versus obesity? And I heard the comments about -- and we heard quite a lot of this at the conference about patients in the placebo arm now end up just going and getting themselves GLP-1 therapies anyway, so it makes it harder to show a difference. So would you actually think about then putting an active comparator so you can say you're better than something else and not have this issue all the placebo patients just start medicating with the GLP-1 anyway? And similarly, in terms of differentiation for eni, including as a petre combo, I think Lilly already doing something similar in terms of doing GLP-1/GIP with the amylin and they've already started the trial there. So the thinking on the differentiation there?

Okay. So I can take that. So regarding the head-to-head, again, we're in the middle of getting our clinical development plans all finalized. So as I mentioned to you, we have, of course, a comprehensive chronic weight management program, but we also have a very comprehensive glycemic control program. So in those studies, we anticipate that we will do head-to-head studies. We're fully aware that retaining patients in a placebo-only arm is getting more and more challenging. And you already saw the hints of the data that Dr. Lingvay showed, 34% of people on placebo dropping out. So fully aware, recognizing that as an issue. So head-to-head trials will be part of the package. We are also going to have long-term extensions as part of the package of keeping patients in trials as a way to provide an incentive.

So after they finish their main, what we call the primary intervention period, they will be invited to participate in a long-term extension. So that's part of our program as well. And then in terms of differentiation, I mean, great question, and we spend a lot of time, obviously, thinking about that every day. I think the way to think about it is on multiple fronts, right? So it cannot just be just can we get an extra 2% weight loss? And would that be a differentiator at the end of the day. So we have to think a little bit more broadly.

So I mentioned already 2 of them. One is a package that includes head-to-head, package that includes long-term extension, a package that includes the flexibility of dosing. So we're working on making the devices highly patient friendly, so multi-dose pens, for example. And then ultimately, when we really think about overall differentiation, we have to also think about where -- what are the patient segments that we really want to go to. So one differentiation -- and again, we'll await the 104 study with enicepatide, which is a Phase II study in type 2 diabetes. we believe that the characteristics of the molecule plus the Phase I data that we have seen and all the vast body of evidence that we have in the preclinical space has the potential that enicepatide can have a differentiation of glycemic control.

So again, the key message here is that we're trying to address the heterogeneic needs of patients, and then we have to figure out exactly where that unmet need will be best served by the characteristics of this molecule.

James, may I quickly double check. Was there an initial question from you about the superiority of weight loss seen versus tirzepatide we should comment on? Or was it just a comment?

Well, I guess the question is, do you think that you could be better? Because I think there was a comment before about how you don't plateau or you plateau...

So how do we see the weight loss?

Yes. So mechanistically, and that's why we showed you the bias signaling, right? So what we know from the science and what we've seen very clearly and consistently and now actually supported by other independent investigators have also shown the importance of bias is that when you have that, you will have a chance for a deeper efficacy. Now the trajectory that Dr. Lingvay showed is very enticing, right? I mean you're seeing a very steep decline, 22.7% weight loss at 48 weeks, no evidence of plateau at all. So you could envision that if you take it to 60, 72 or even 80 weeks, you might get more. I mean -- but again, we have to do the study, right? So we're obviously excited about that profile. We have -- we see that as a potential way to differentiate, but we also see other ways of differentiating.

So yes, to answer your question, that is one of the levers where we see there could be a differentiation in terms of weight loss, but we also see differentiation in terms of A1c or in other areas that I just mentioned.

Very good. James, we answered all your questions.

There's plenty of them I could ask, but I'll let someone else have a go.

Yes. Then we move on. Next, please, Luisa Hector from Berenberg.

So I was wondering if you could tell us anything about the relative proportion of fat versus lean mass loss, whether you've got some data when we might see it? And maybe just to follow up on that question on the running the placebo-controlled trials. I see -- I hear you, Manu, you've got your head-to-head, you're thinking about a long-term extension. But in the trials that have already started, so eni, is there anything more you can do other than the long-term extension? Are you able to kind of check whether there's any unapproved use of GLPs happening along the way or anything you can work within that trial if it does happen?

Yes. So I'll start with the first 2. I would definitely invite Dr. Lingvay to provide her perspectives in her own clinic about how she manages that. I mean we're fully aware that there's always going to be add-ons of GLP-1s, and we have prespecific endpoints to try to manage that. So regarding fat versus lean, I mean, of course, that's going to be a body composition analysis that we will have, and we'll present some of that data at upcoming conferences. We don't have the data as yet right now. But yes, we have measured body composition, of course, in both our ZUPREME-1 study and the Study-103.

What other things can we do on the placebo side? I mean, we are going to do everything that we possibly can to retain patients. And I mentioned a couple of key ways of doing this, right? I mean, head-to-head trials, long-term extension, great relationship with sites and investigators that we want to have to educate people that, look, this is a known phenomenon, but that we have to be able to work with the right sites and the right investigators and the right patients.

So I'm not diminishing that it's a challenge, but we do have methods in place that we feel that we can do that. And plus, we've had prespecified analyses within our protocols to try to ensure that we have a way to look at both the truly GLP-1 naive population as well as those populations that may be using a GLP-1. And we recognize that there will be some diminishment of efficacy just because you're adding it on to another GLP-1. So prespecifying it will be helpful and to really get us a good understanding of what this effect will be. But Dr. Lingvay, if you have anything to add, please feel free.

Yes. So just happy to remind people that this is really not because we want to do studies against placebo. It's a regulatory requirement. And we would very much like not to do them against placebo, but not much of a choice in the matter here. Some of the glycemic studies can be done against active comparator in that space, but in the obesity space really, the placebo is a requirement that drives the design of these studies. That being said, I have to be Roche. It's one of the first companies that has these open-label extensions, which are truly meaningful. I'm hopeful that they will change how patients think about participation in the context of them thinking they might be on placebo.

And it's not just a 6 months extension. It's a meaningful extension for everybody, and everybody can go to the maximum tolerated dose. So I'm really, really thinking that this will be a game changer in the field of obesity when we're still required to use placebo. Of course, all the other measures that Manu mentioned, I think they're doing a great job at selecting the sites right, educating the sites and the site staff and making sure that they're fully engaged with their patients, providing the dietary counseling, which is so essential for people to maintain that interest in the study and make the good changes in the life that will help them and benefit them in the long run and then making sure that they're staying through the end and being great supporters of these patients so they can get to that open-label extension phase.

Then next questions come from Sarita Kapila from Morgan Stanley.

So Roche have previously talked to potentially dosing CT-388 above the 24 milligrams tested in Phase II. How much of a barrier is the 20% to 30% vomiting, diarrhea to pushing the dose higher? And for any of the 3 combo regimens, do they include CT-388 exposure above the 24 milligrams? And then just a quick follow-up on the no plateau comment for CT-388. Is it because the 24-milligram arm reached target dose late? So we noticed there were 6 titration steps before reaching 24 milligrams. So how many weeks of actual exposure did the median patient receive to 24 milligrams by week 48?

Yes. So I'll start with this one and invite others as well as needed. So just the last question first. So it was 20 weeks because they finished their titration at week 28 and the endpoint was at 48. So they were at least able to take up to 20 weeks -- between 16 and 20 weeks. So you're right, it's -- could we have gotten more? This is why we feel very optimistic that we will if they were to maintain it around 24 milligrams over a longer period of time.

The other thing I'll try to remind people here is that when we actually looked at very carefully the performance of people at the 24-milligram dose, the tolerability was actually really well tolerated. The challenge was from 12 to 16. And then once they got to the 24, the background rates of nausea, vomiting, constipation, diarrhea generally tended to be very stable and certainly did not increase. So that gives us actually quite a bit of confidence to actually push the dose to higher doses.

So I think this is a great opportunity to emphasize the other comment that the other person made about how else can we differentiate, we can also differentiate by potentially going at a higher dose. And the reason why we are confident that we can go to a higher dose is, again, as I said, the current data set that we currently have. No plateau there. So it means that you can push to a higher dose. Tolerability is actually quite good at 24. And then the final piece, how to ensure that we can get there, and this is still in evolution in terms of our final design elements. But one way is to just be more flexible, take some more time to get to the top dose.

So I'll just remind people that it's not a race to the bottom to get to the lowest weight loss number in the shortest period of time. We want to be really mindful of taking people through that weight loss journey as comfortably as possible. And our properties of the molecule gives us confidence to say that we can do that safely. So more flexibility, slower titration, great tolerability at 24 already that provides us confidence that we can go higher. And what that higher will be, we'll hopefully come back to you all once we have finalized that design to say what that dose is. But we fully anticipate that we will try to get to a higher dose than 24.

Manu, I think there was also the question whether the Phase II combo had already a higher dose included. Sarita, if I got your question right.

Yes, that was right.

So in the SYNERGY study, are we doing a higher dose?

Yes, exactly.

So this -- so the point of the SYNERGY study is to try to maximize both in the current dose ranges that we have actually studied. We believe that with the current doses that we have had between a really nice dose ranging that ZUPREME-1 did, a really nice dose ranging that the Study-103 did, we have plenty of doses there that we feel really confident to be able to say that we can mix and match the right dose combination to maximize the efficacy. And there, I'll just remind people that combination is really designed for people that are really needing that level of weight loss, right?

So greater than 25% of weight loss, where maybe it's the high BMI category with type 2 diabetes or other comorbidities where getting to that degree of weight loss really will be helpful. But we also remind people that there are plenty of people who need only 10%, 15% of weight loss. And maybe there are other parts of our portfolio that can serve that, like petrelintide, as you heard from Dr. Aronne, right? So we have to really think about kind of where we want to use and where do we want to deploy these medicines. And given our portfolio, we're well positioned to do that.

Sarita, did we answer all your questions?

Yes, that was perfect.

And the next one in the queue is James Quigley from Goldman Sachs. James?

I've got 2 questions, please. So firstly, for Dr. Aronne and Dr. Lingvay. ADA, we saw lots of new data from different therapies for CagriSema yesterday and as well. So what are you thinking in terms of positioning for CT-388 and petrelintide? And we're particularly interested in Dr. Aronne, your view on 388 and Dr. Lingvay, your view on where petrelintide could sit. And second one is a sort of clarification/reminder. So on the lack of dose response in ZUPREME-1, what were the reasons for that lack of dose response? Is it also related to imbalance in gender between groups? Are there other factors that are at play there that you've noticed? And apologies if it was on the slides, but which dose is going forward into Phase III for -- which dose level is going forward for the next Phase III for petrelintide?

Would you like us to respond?

Yes, Dr. Aronne, you might want to start.

So I think that enicepatide, the data is fantastic. It looks like it's at least as good as tirzepatide. I think you have to agree the weight loss hits the tirzepatide point of 22.5% weight loss plus/minus, and it's still going down. Is it going to abruptly plateau? I mean that doesn't usually happen. So it could be that the biased signaling is giving greater efficacy. We won't know until we treat people for a longer period of time, but it's at least theoretically possible. But also in looking at the weight loss curve, it is -- I think it is definitely possible. That would be a big differentiator to have a GIP/GLP-1 that is more effective than tirzepatide. That's been kind of an invisible barrier that people have been trying to break through.

You see CagriSema hitting the same number. And this is -- and a triple agonist obviously goes past that, but this would be the first dual agonist that looks like it could go beyond that 22% to 23% weight loss in this type of trial. As far as the petrelintide, I personally think that, that could be one of the reasons we see the 10.7% maximum weight loss is that this is the lowest percent of women in a trial that I've seen for some time. So it is at least theoretically possible that, that will be a difference in that in the next trial, if you do the math, it should be 14% to 16% if the percent of women is what it is in other trials, around 75% and also if the response is the same as it is with the GLP-1s. So we'll just have to see those.

And maybe also the same question then goes to Dr. Lingvay. I think the question was on how do you see petrelintide positioned within the field currently?

Well, I thought Louis is going to take that, but I'm happy to take that because I'm a big fan of petrelintide. I think its safety profile is so compelling that it's going to be the first-line therapy for a lot of people and especially in the primary care field where they don't have the energy, time or just multitasking ability to deal with all of these side effects and the multiple prescriptions and the multiple dose escalations. This is an ideal drug for primary care patients and for really the way more than half of the population that needs weight loss and it needs weight loss in the range of about 10% or so, give or take, while improving their cardiovascular risk. And we've seen that petrelintide does that very nicely.

It improves blood sugar, improves cardiovascular risk factors like blood pressure and lipids and CRP. And therefore, it's really a nice gentle drug to deliver the type of weight loss that most people need. I'm also excited about the potential of exploring and using an amylin-based therapy like petrelintide in people that are older, that are more frail, that perhaps have more false fractures. Many of those people would benefit from weight loss. But currently, we're a little bit timid about intervening with the agents that we currently have because of the risk of muscle wasting and the risk of additional falls and fractures. If the data pans out as we hope it will, there will be a huge population there in the 75-plus age range that would benefit from an intervention that is gentle, that drives the potential of improved bone health and muscle health, that would be such a welcomed asset in that population.

I agree with Dr. Lingvay completely. I also wanted to point out that if you look at the evolution of our field, as time goes on, people are going to be treated earlier and earlier. Right now, we're dealing with a lot of people with very high BMIs because they haven't had any treatment really available that was accessible to them. So there is a big need for extremely effective compounds. But over time, the right way to manage obesity is to treat people when they get to a BMI of 27 or at the maximum 30, and they'll never get to these catastrophic weights. And at that point, think about it, if you had a BMI of 27, let's say, and you were treated with a medicine that had almost no side effects and you got down to 25 or 24, that would be fabulous. That would be a really great outcome. So I think that is ultimately where we're going to be headed. And so a medicine like this is very well suited for that future.

I think there were 2 questions still open. And maybe, Manu, they go to you. One, the lack of a dose response seen with petrelintide, so the plateauing. Any explanation why this is? And then the second question, which dose to be taken in Phase III development?

Yes. So in terms of the doses to be taken for -- in Phase III, we'll be able to get into a little bit more of that detail in a few weeks' time. We're not able to disclose that right now because we're still finalizing some of those protocols, and we need to get some regulatory feedback on the final design. So we are not able to share it for that reason. But based on the data that we've currently seen, we are confident of what those doses can be.

Regarding the lack of dose response, I mean, this is still obviously in speculation territory, right? I mean I think one of the things we do know about amylins as a class is that unlike GLP-1s, which doesn't seem to have kind of essentially a plateau where you get to, it's largely limited by tolerability. It's not limited by efficacy. Here with amylins, we are starting to see there may be ultimately an efficacy plateau as well. And so the study that was done, which is the right study to do, which is a full dose range, gave us that empirical information that gave us that actual clinical data that we needed to see on the 5 doses that we tested, where does the curve fall.

So I think it's a mechanistic sort of difference. A lot of people are used to seeing incretins where you can keep pushing the dose. And as long as you're not tolerability limited, you will get more -- potentially more efficacy. But with amylin as a different class, a different mechanism, we're starting to see that there is, in fact, a plateau with the efficacy. But as Dr. Aronne mentioned, I mean, with all of the different caveats of this trial, 50% female, maybe in retrospect, if it's 42 weeks, perhaps we should have gone a little bit longer, et cetera.

We feel confident that the data that we have puts us in the right position to test this out in the right settings in Phase III. And obviously, in Phase III, we expect the population to kind of resort to the kind of the norm, if you will, where we'll see 65%, 70% of the population to be female. We'll obviously do a longer-term study, and then we'll really see how these doses play out.

James, all answered? Then next in the queue is Richard Vosser from JPMorgan.

Sorry for the background noise. I've got one question just on flexibility. Manu, you mentioned flexibility and the importance of using it going forward. We've seen other companies have trouble with the right flexibility and the lead -- and that leading to lower weight loss than anticipated. So just wondering how you can guard against making sure that the flexibility doesn't impact efficacy too much. And also linked to that, just thinking about tolerability, the tolerability was pretty good with eni, but not maybe quite as good as tirzepatide at the high dose. So just thinking about what else other than the flexibility you might be able to employ to further improve the tolerability there.

Sure. Okay. So you're right, I mean, you have to have the right balance. If you have too much flexibility, you may not get to the top dose. If you have too much rigidity, you might have too much dropouts. And we've seen, again, empirical evidence from 2 separate compounds showing exactly those 2 things. So we have had a lot of learnings in the field. And so we take those learnings seriously and try to provide that into this protocol where we try to achieve that sweet spot, that balance. I can't get into this every little detail, but there are very specific steps in the protocol where we are allowing for certain flexibility, but not so much that they're not getting to their assigned top dose, whatever dose that they're assigned to.

And again, I come back to what I said before, rightsize, right investigators and really try and motivate the participants in a way that we can achieve this. So the benefit that we have is that we've learned a lot from the field. We try to incorporate those learnings. We'll execute on the Phase III, and we feel confident that we'll get to that sweet spot. Now the tolerability part, I mean, I know that a lot of people love comparing studies. But I just, again, caution people. Just look at the study for what it is. Cross-study comparisons are a little bit challenging.

Oftentimes, people compare a Phase I to a Phase III tolerability profile and Phase II to a Phase III, et cetera. This is a Phase II study. And we have to recognize that there are certain vagaries of this study, as Dr. Lingvay mentioned. They were coming in every week to get their injection. When you come to a clinic every week and you get asked how are you doing? Do you have AEs? You're kind of in some ways, sort of prompting the question to some extent. So I think we have to be a little bit careful of not to try to make an apples-to-apples comparison with the Phase II and Phase III. But what I will say is that, as I said before, with the measures that we are putting in place with the type of investigators that we have in our study and the learnings that we have, especially on the flexibility side, we will manage through the tolerability.

And I think the final piece I'll say and I hand it over to Dr. Lingvay is the data, again, speak for itself. Personally, I was actually really concerned about going to 24 milligrams. But I was reassured by the fact that in the Phase Ib, we were able to get there very fast titration, no safety signals. Yes, the tolerability was higher because you're forcing the titration. Then we slowed the titration down in Phase II to every 4 weeks, then we saw almost half to 2/3 less AEs compared to the Phase I. So which tells us that there's nothing here in the molecule that concerns us. We feel that with the learnings that we have, we can institute the measures to get to better tolerability. And I think the final proof is Phase III to Phase III and ultimately a head-to-head comparison. So I would stay away from a Phase II to Phase III comparison. But let me hand it to Dr. Lingvay.

Yes. So actually, I'm pretty passionate about this topic, so I wanted to speak up because there is a difference in priorities that it's emerging strongly between the regulators who are really looking to make sure that the medicines that we're developing are safe for everybody that might potentially use this medication, investors like you guys, which are really just measuring these drugs against who gets to the bottom faster and harder and bigger and then the patients and those who care for the patients who really want the right therapy for the right person.

So we're designing these Phase III trials and Phase II trials with the regulatory requirements in mind, including people with a BMI of 27 and above. But then we're studying these drugs and pushing people to these highest dose drugs expecting 20%, 30%, 35% body weight loss. And the reality is most people, not only they don't need that, they should not get that, and we're hurting them by pushing them there. So yes, to marry all the interests here, we need flexibility in these studies. We need a lot of flexibility because if you don't get the flexibility, you're hurting people who are participating in the study who don't need that much weight loss, who might have side effects, and there's no reason to push them up if they're doing just fine on the lower dose and achieving the goals that they want.

Again, if you would want the race to the lowest, then we need to do studies that only enroll people with BMIs over 40 or people who truly need those amounts of weight loss that the investors are looking for in order to compare these studies against each other. So please don't measure development -- drugs in development by their amount of weight loss. Look at how much of -- how safe these medications are and in the right patient, do they deliver what they need to deliver and make the people healthier. Lowest weight is not our goal. It's really not our goal.

Then let's move on. Next one is from Steve Scala from Cowen. Steve, please.

Two questions. Does Roche have a strategy for monthly dosing of a GLP-1? And for Dr. Aronne and Lingvay, what percent of patients do you believe will be on a monthly product in 3 to 5 years? Is it a small minority? Is it the vast majority? Or is it somewhere in between? And then secondly, on CT-996, you mentioned go/no-go decision by the end of the year. What do you need to see to advance that molecule?

Dr. Aronne and Dr. Lingvay, please start on the monthly.

Well, as far as what percent of people will be on a monthly, I think we still don't have enough data to really determine how tolerable it will be, how effective it will be. I think that there will be big demand for it. I think it's very appealing. It's very exciting to think that we could have that. Will it work? Will the side effect profile -- obviously, giving something weekly will produce a better side effect profile for the same amount of medication for the average drug. So making that transition, we still have to be certain that that's going to be okay, but I think it will.

I'm not sure 3 to 5 years is going to be the time frame where there will be a massive uptake. But over time, I do think that for chronic disease, this significant number of people may switch to monthly. And it's very appealing as a treatment paradigm. I think for clinics, too, like have the patient come in and give them a shot once a month. I mean that could be very good for a chronic problem like this.

And I think there were also like 2 questions to you, Manu. What do we do in terms of our pipeline in terms of developing a monthly solution? And then 996, what does the go/no-go decision depend upon?

Yes. So regarding the monthly dosing, right? So I think the way we look at it is, again, coming back to the patient needs. As I said before in my opening the way we've sort of designed our portfolio is based on really addressing those kinds of heterogeneous needs. Some people, as I said, need oral, some people need injectables. Some people need a weekly, some people need a monthly or maybe even a quarterly. So at this point in time, we are very open to looking at all of them. And part of our strategy has been to sort of see what are the long-term -- I mean, long extension half-life technologies that we would want to really bring to bear.

And whether it's monthly or quarterly, et cetera, we can come back to refining that. But on a strategic level, I mean, I think that optionality is exactly in line with our strategy that Morten laid out, right? So we're never going to say it's an either/or situation because we just are humbled by the fact that obesity is so heterogeneous and there are needs at different stages for the same person actually. So I think we have to be open-minded about where those needs are. But it will always be guided by the same thing, which is if we were to bring a long-acting, is this going to significantly improve their quality of life above and beyond what they're already getting with their weekly or an oral? Is it really safe? Can we put somebody on for that long and be really comfortable with the fact that the safety and the tolerability profile will be good?

What are some of the conmeds that they may be using during that period of time? Because, again, remember, there is polypharmacy in people living with obesity because they have so many comorbidities. So there may be some advantages to having short-acting compounds for certain people, very long-acting compounds. So it's really not that simplistic when we kind of look at it in that way. Regarding 996, I mean, we'll obviously look at all of the things that we all look at when we made the decision for either enicepatide or petrelintide.

It's going to be a same set of very rigorous set of assessments, safety, tolerability, weight loss, glycemic control, cardiometabolic risk factor modification, the full gamut, right? Because what we're trying to do with 996 is to position it also for a broad set of indications, which includes obesity and glycemic control. So I think we have to look at the totality of the data. I can't get into specific numbers of what we need to see, obviously. But suffice to say, it's a rigorous assessment to really make sure that it meets the bar as a way that we define the bar. So you've heard from define what that bar is. And so it has to meet our bars to advance this forward into our Phase III pipeline.

Then we quickly go on. Next one is Simon Baker from Redburn. Simon?

Apologies again for the background noise here. Two, if I may. Firstly, just a question on petrelintide itself. There's been a lot of debate over the conference about the importance or otherwise of bias between the various -- the 3 amylin receptors and calcitonin. I just wonder if you could give us an idea of the relative affinity of petrelintide for those 4 receptors. And then a question for both Dr. Lingvay and Dr. Aronne and picking up on what Dr. Lingvay was saying. There is a clear disconnect between the market demanding ever higher percentage weight loss and clinicians demanding ever lower GI side effects. And it looks like with petrelintide, you may have a product that fits that profile of about 15% weight loss and very low AEs. So a question to both of you. What proportion of your patients would be adequately treated by a drug with that profile?

So I can take the first one and hand it off to both Dr. Aronne and Dr. Lingvay. So as you know from the data that's been presented already in the public domain, petrelintide is what we call a DACRA. So it's a dual amylin and calcitonin receptor agonist. It basically has receptor selectivity for both the amylin-3 and the calcitonin receptor. So at this point in time, I think the data is still premature to know that amylin-1 is superior to amylin-3 or not. Yes, on the surface, it might look like there's more to be had with amylin-1. But again, as I said before, I would caution us to not come into that trap of cross-study comparisons because ZUPREME-1 had baseline characteristics that were quite different. The demographics were different. Duration of therapy was different and so on and so forth.

So I think the jury, at least in my mind, my own personal mind is that it's still out in terms of what is the perfect ratio to have. But I think what we do know from the field is that having some calcitonin can be beneficial largely because it has potential beneficial effects on bone because we use in calcitonin today in clinical practice to treat osteoporosis. We know from preclinical studies that having some calcitonin actually can enhance insulin sensitivity. Now whether all those things will ultimately translate into the clinic with a DACRA or not in terms of those outcomes, yet to be determined, but that's our plan, right? I mean we want to have a robust clinical development program where we can interrogate the bone, body composition, et cetera, right?

So we do believe that, that mechanism lends itself to that. And that's what we're excited by. But I think to over-index on amylin-1 or amylin-3 purely on a cross-study comparison on weight loss, in my own personal opinion, is a little premature. But Dr. Aronne and Dr. Lingvay, you want to answer the second question?

I can jump in real quick because I agree it's a little bit premature to make this forecast. As I alluded to earlier, I'm very excited about an amylin-based product and especially because in primary care for the vast majority of people with no comorbidities and with a moderate amount of weight loss, targeted, this is a good option. Now as far as additional patient populations are going to depend on what sort of data is amylin going to be able to deliver. Will it deliver the data that it prevents fractures and falls and muscle wasting and all the other potential benefits that we're thinking would position this really favorably for an older population, for example. If that's the case, that's really a quite important segment population that would be favorably positioned to receive petrelintide or a similar drug.

There's also a big question of whether this petrelintide or similar agents will have cardiovascular protection on their own without mixing it with a GLP-1 agonist. If they do, I mean, that's really a game changer because now it can position it across a very wide age range and risk range and pretty much everybody with a weight loss need that is below 15%. It's a potential candidate. So I think as an initial phase, it does have a significant potential, especially in primary care and sort of entry-level treatment with first-line therapy for overweight and obesity. Over time, as data gathers and more information is available, it does have the potential for these additional market segments that are well represented.

I agree with Dr. Lingvay. I see older people would be a great one, primary care because of the side effect profile. I would also add people who aren't responding to the other compounds. So because of mechanisms, I imagine, we're seeing a minority -- a small minority of people, but a significant number who don't respond to GLP-1s or even the GLP-1/GIP combination. And I think that with an additional mechanism added, we may get extra weight loss in those people who are refractory. I mean, think about it if it's just 5% of the population, you're talking about millions of people who potentially could benefit from a drug in this class. So I think there's plenty of market for compounds in this category.

Very good. So we have one final person. Actually, it's Graham Parry from Citi coming back with a final question. And then I think we will close the Q&A session. Parry, the final question goes to you. Graham, the final question goes to you.

Actually just a quick follow-up. So on CT-388, you're saying you'd run head-to-head studies in glycemic studies in type 2 diabetes, but I wasn't sure if you said you would or wouldn't run head-to-head in obesity.

Just to clarify, I mean, we are still fully weeding out all of the Phase III and the Phase IIIb plans for the chronic weight management side. So we're not precluding the possibility that we might run it there. But for glycemic control because the standards of care are already pretty established in terms of what your background therapy is going to be, there, it just -- it's much more obvious. And there, back to what Dr. Lingvay had mentioned before, the FDA is actually well aware of the fact that those background therapies for type 2 diabetes have to be studied, right? So that's the reason why those head-to-head studies are a little bit more firmed up for glycemic control.

But for chronic weight management, there's still some uncertainty around can we not have a placebo from what we have understood from the health authorities is that they want to understand the safety profile of these novel therapies on a "clean background." And so it's not that we want to constantly keep doing placebo-controlled studies. I mean, just to be really clear, it's -- we have to meet certain regulatory requirements. And so exactly how those evolve, when they evolve, we will adapt to those. And at that point in time, we may foresee doing head-to-head even in chronic weight management. But it's a little bit more established in glycemic control, and that's why we have more firmed up plans there.

Yes. I think with that, we close the call. Let me thank again for all our presenters for their time and efforts exploring new treatments for obesity patients. Let me also thank the IR team members who worked on the slides and prepared this event. So I have to call out here, Jan Patrick Scharzon and and also Melanie Wolf for the event organization. I hope the event was helpful and provided a timely update on our CVRM franchise. And if there are any remaining questions, then please do not hesitate and reach out to the Roche IR team. We are happy to further help and come back to you. And with that, we'll close the call, and goodbye.

So welcome to our Annual IR Diagnostics Day 2026 here in London. Let me quickly take you through today's agenda. We have today 10 speakers with us who will provide updates on the Roche Group strategy, the diagnostics strategy, the diagnostics business portfolios and priorities, as well as latest overviews on the pipelines of our 4 customer areas, which are core lab, molecular lab, including Foundation Medicine, pathology lab and near patient care.

This year, we will have a special focus on the upcoming global launch of our truly transformative next-generation DNA sequencing technology, which is called AXELIOS, and on the end-to-end oncology ecosystem, which we have partly built already around it. This ecosystem has recently been strengthened by partnerships with Freenome in early cancer detection. And by the recent acquisition of SAGA Diagnostics, a company known for ultrasensitive MRD testing.

In addition, this emerging ecosystem has also been completed, complemented by the most recent acquisition of PathAI. This is the announcement we made just last Friday, which is a company from Boston developing AI-based digital pathology solutions. And taken together, all these approaches will not only help us to revolutionize cancer detection and care, but also they have the promise to contribute in the future to drug development efforts.

Can I have please the agenda slide? So I will take you through the agenda and also the changes we made last minute. As last year, we'll have 2 sessions, as you can see here, 2 presentation sections of 60 minutes each, followed by 20 minutes of Q&A, where we will have [ always the respective ] speakers of the presentation section on stage.

So for the first session, which is scheduled from 1:00 to 2:10, you see this session will be opened by Thomas Schinecker, our CEO of the Group. Thomas will provide us a quick update on the group's strategy and comment on our longer-term ambition. The second presentation will then really open the IR day. It will be by Matt Sause, our CEO for Roche Diagnostics. Matt will provide an overview on the overall vision and strategy for Diagnostics, and he will talk on the future ambition for the Diagnostics division. Third speaker is Palani Kumaresan, our Global Head of Roche Diagnostic Solutions, who will break down for us the strategic priorities of all our 4 customer areas. He will also provide a quick update on the ongoing mass spectrometry launch, one of our key growth drivers for the future.

The final presentation then of the first session, focusing on Roche Information Solutions, will be given to 2 speakers. It will be held by Moritz Hartmann, our Global Head of Roche Information Solutions; and by Andy Beck. Andy is our surprise guest for today, Founder and CEO of -- sitting over there, of PathAI. Moritz will briefly summarize our overall efforts in digitalization, how we strengthen our product development capabilities and also create new product opportunities, whereas Andy Beck will explain to us how AI-based digital pathology will strengthen our product development capabilities, especially in oncology and immunology and also how this approach could contribute to scientific hypothesis generation and ultimately support future drug development efforts.

Following these presentations, we will have a 20-minute Q&A session before having a quick break of 15 minutes, and then we are back at 2:45 p.m. for the second session. Session 2 will be opened by Josh Lauer, our Global Head of Customer Area Molecular Labs. Josh will provide us with the latest and greatest of our truly revolutionary and game-changing AXELIOS sequencing technology. He will highlight the unrivaled versatility of this technology and its potential to future applications by sharing some of the latest data, which we have generated.

Second speaker will then be Daniel Malarek, CEO of Foundation Medicine, which is Foundation Medicine is part of the molecular labs customer area. Then we'll talk to our increasing product offering of FMI in oncology and the unique opportunity, which we have here to integrate Roche Innovation Solutions, whether it's the AXELIOS sequencing, the VENTANA platform or navify Digital Solutions.

Following molecular labs, Laura Apitz, our Head of Pathology Customer Lab area will then cover the fastest-growing customer segment, Pathology Labs. She will drill down and take us through the strategic priorities of this segment. And fourth speaker, as you can see here, then will be Ildiko Amann-Zalan, our Head of R&D Roche Diagnostic Solutions. Ildiko will cover our near patient care portfolio providing latest updates on the early launch days for our CGM solution and for LumiraDx. And finally, we will close the second session again with Olivier Gillieron, our Lifecycle Leader Cardiometabolic and Neurology with an update on [ collab ] assay highlights, which include novel assays in neurology. So that's the NfL assay, but also the pTau217. You might have seen the news this morning. And also some news on the long-awaited update in infectious diseases on our tuberculosis assays.

Following the second session, we'll have then, again, our second 20-minutes Q&A before closing the broadcasted part of the event. Participants on site will have the opportunity to meet all senior management at the closing, upper row. Most management will be around for another 40 minutes.

In terms of housekeeping items. Let me just point out that we again have prepared our short survey to collect your feedback. Participants joining online will be invited to join the survey 10 minutes before the final Q&A ends. A pop-up window with the questions will show up, whereas participants on site will receive an invitation to participate in the survey by e-mail roughly 1 hour after the reception took place.

And with that, before handing over to Thomas, let me thank the speakers again for their time and their continuous efforts in improving the life of patients. And also the IR team and the members who contributed to this event. First to name, Birgit Masjost, who led the -- who's leading Diagnostics within the IR team. Birgit had the overall lead for the event, and she will also moderate both Q&A sessions. And then in addition, I have to call out [indiscernible] for speaker support and slide preparation; [ Julio ] [indiscernible] for speaker support and slide preparation and managing the master deck; [ Rafael Pavlovsky ] for speaker support and slide preparation; and last but not least, Eva Losert; Melanie Wolf, Beatrice Hau and [indiscernible] for event organization.

And with that, it's time to hand over to Thomas for providing us an update on the overall group strategy and the longer-term ambition. Please?

Thank you, Bruno, for the introduction, and it's great to see you all here in London today. As mentioned in the Q1 call, in the future, the idea is that I would attend a pharma day and a diagnostic day always in separate years so that I give attention to both divisions. And of course, this time, there are 2 highlights that I thought I cannot miss, and one is the sequence launch. Something that I know for many years, I've always been asked, when are you going to launch, when are you going to launch? And I always said, we are going to launch, and we will have a very competitive solution. So I'm personally also very excited on that. And also, I'm very excited that we made an agreement with PathAI so that Andy could also attend today. Thanks, Andy, for joining us, and welcome to the Roche family.

Now let me go through and give you a brief group strategy update and then also give you an outlook on growth for the group over the next years and into the next decade. On the group strategy side, you've seen the strategy house before the 10-year ambition. So where do we want to be in the next 10 years. But also the different elements of our strategy using AI, digital health, sustainability, people and culture and then diagnostics and pharma. Good thing is not only did we complete the strategy house to really define the direction of the company, we are also making superb progress when it comes to implementation. And let me just go through a couple of things.

First, on the 10-year ambitions, 10-year ambitions on the pharma side, it's about delivering 20 transformative medicines by the end of the decade. And here, we've made superb progress. Not only do we have now 66% of our portfolio with best-in-disease potential. But also, we have increased average peak year sales of our pipeline per project by 67% and a total pipeline value by 63%. So we have a higher pipeline value than ever before in our company. And if you look at the readouts that are going to come over the next couple of years, with up to 19 medicines that could launch by the end of the decade. This is a very strong late-stage pipeline.

But not only have we accelerated what we did in late stage, we also did that in research. If you look at lead identification and optimization, if you look back 3 years ago, we had around 50 of those every year. We have doubled that output. And clearly, AI plays a role in here as well. On the Diagnostics side, it's also about doubling our output on R&D and reaching more patients. Just a couple of numbers that really underline that.

First, if you look at the industry, how many FDA approvals that were in the last 6 years. Roche Diagnostics has more than 50% of those diagnostics approvals. And certainly, we are not spending more than 50% of the industry's budget. So clearly, we have a very productive R&D. And you can also see that we have 3 instrument launches, 3 instrument launches with blockbuster potential, CGM, mass spec, and now coming next generation sequencing. Again, something that I don't think any other diagnostics company has in such a short period of time. Also, we keep filling our menu with additional tests and digital solutions that augments and differentiate us from our competitors. So clearly making a lot of progress here.

Now let me also recap the group strategy and where do we want to go and what are we going to focus on. And you have seen this slide before. So these are the major trends that we see in health care and how we respond to those major trends. Clearly, there are 3 therapeutic areas: oncology, neurology and CVRM that are contributing to more than 50% of the global disease burden. And you'll see the same later on the pharma slide. Especially in situations with health care systems are really in tight budget situations, health care systems will continue to prioritize areas where they can have the biggest impact with the money that they have to spend. So for us, it's important to prioritize the diseases with the highest societal burden. And clearly, have differentiated medicines, transformative medicines. And this is also where then you can address higher patient populations or larger patient populations.

As mentioned already, health care costs continue to grow faster than GDP. So we have to provide solutions that improve outcomes. At the same time, reduce cost of care. Hemlibra being a good example. It improves the life patients with hemophilia A, but also reduces hospitalization with that cost of the system. So we have to write these kind of solutions in order to address this trend.

But what else do health care systems do in order to address the trend? One is decentralization of care, meaning that people that go to the hospital today in the future will be treated in the outpatient setting, meaning primary care or sometimes in at home. So for that, we need decentralized solutions, both in diagnostics and in pharma. Diagnostics, for example, solutions like Lumira where you can do testing in the GP segment as well at the low cost. But also in the diagnostics -- on the pharma side with drug delivery systems. So people can actually take the medicine at home or in a GP setting and don't have to go in the hospital and drive cost in the system. Plus for the patient, it's also a better experience because who wants to be going to the hospital all the time, it's better to do it in your home. Also, what you will see is a trend towards early detection, monitoring and intervention.

Bruno mentioned Freenome for early cancer detection. We also mentioned SAGA for minimal residual disease testing. Again, tests that detect disease early, so you can intervene early and reduce costs in the system and prove the chance for the patients. Of course, access to health care is still a challenge. More than 50% of the world does not even have access to basic care. This is a big opportunity to reach more patients. And then there are transformative technologies, AI being one of them that we have to be at the frontier of because they are going to change the way we either discover medicines or treat patients or even diagnose patients.

That's why we have this group strategy around convention, stopping and curing diseases so that people don't only live longer, they live longer in a healthy way. So not only lifespan, but health span. And the combination of diagnostics and pharma plays a significant role here because you need a diagnosis in order to have the right treatment. You need an early diagnosis to have a high chance for cure, clearly focusing on the diseases that drive the highest societal burden and the highest cost to the health care system and really innovating a long patient journey to improve care for these diseases and meeting where patients are, where they are, and not only in the hospital, but even in the home setting.

So this is around prevention, identifying these patients early for diagnosis, making sure we have an early treatment like with giredestrant in the adjuvant setting where you have a higher chance for cure, stopping the progression of disease or even moving towards cure, as we do, for example, with CAR T cells where you can reset the immune system for certain autoimmune diseases.

Now let me give you an update also on our progress on AI and data. As we implemented our AI strategy in the company, we implemented in 3 levels. First level we call EverydayAI. And And EverydayAI means we've trained 100,000 people on AI, and almost everyone completed it. And if we monitor it now, we can see that 75% of our people now use AI in their everyday work. And this is really using off-the-shelf tools to make them more productive. As we go into the next round of training, we'll also introduce agentic AI, so people continue to get trained, so that we can use these tools to improve the productivity in our company.

The second level of AI is what we call reshape. So basically, we have 3 main processes in the organization. One is service. The other one is marketing and sales, and the third one is R&D. And we can use AI tools across the entire process to shape, reshape and redefine the process and also accelerate the process using AI tools in different steps of the way in order to become more efficient and more productive as a company faster and also create more output. For example, if we generate kind of protocol -- study protocols, we can use AI to generate the protocols. Clearly, this reduces the amount of time that people need to write those protocols, and we can become much more efficient. So we do that along the entire journey.

The third level is really around the big ideas. What are the areas of disruptive potential. And AI can have disruptive potential in research in pharma. What we call Lab-in-a-Loop, basically, you use different AI tools to generate drug targets. You make those and then you test them in the wet lab, you feed back the data and you constantly update your model, you train your models. These are proprietary models that our people have developed in order to accelerate drug discovery to find more targets, become more efficient. For example, we've now embedded AI tools in 100% of large molecule drug discovery programs. And we have more than doubled target entry using AI, and we also can accelerate, as you can see here, we have 3 DevGo accelerations.

Another area that we can see as big ideas is digital pathology and using AI in pathology is one of the areas we'll talk about later as well. Yes, one area, as I mentioned, is digital pathology, and PathAI is clearly taking us into the next era here as well, bringing in the leader in this space to really accelerate our activities. But our activities in AI haven't just started with the acquisition of PathAI, it started much earlier.

Back in 2016, when we did the first collaboration with NVIDIA. And why did we do that in sequencing because we knew as we were developing the sequencer, we need to have chips that are able to handle the amount of data that we produce on the sequencer. The sequencer today, and it's only the start because we're going to improve further on that, the amount of data it needs to handle is 3 high-definition movies per second. There was no computer chip in the past that could be able to do that. That's why we needed to work with NVIDIA. And the speed will even go up further. And this is all computing that actually happens on the chip.

And then as you can see, 2019, for example, it was the first time that an AI algorithm outperformed the radiologists. It was already clear that time that imaging was one of the areas where AI would really play a significant role, and that's why also the acquisition of PathAI. You probably remember also the protein structure prediction by AlphaFold. So you can see a lot of progress that has been made, and we have invested in AI in that time period. And now expanded back in 2023, our collaboration with NVIDIA on the drug discovery.

This has now led to us also investing more into own chips and on-premise infrastructure. With that, we now have the highest compute power in the pharma industry. And we need that as we continue to build our models and co-train our models, feed the models with more data. We need that compute power in order to be able to leverage AI throughout our R&D. Also, we are a founding member of the SAIL Foundation. So this is a shared license agreement with companies like Anthropic, Microsoft or IBM.

Now we already talked about digital health and PathAI clearly playing a role here. There is a strong synergy to both diagnostics and pharma. Clearly on the diagnostics side because the cancer cases keep rising. The complexity of diagnostics keeps rising because there are just more markets that you can look at. And actually, the number of pathologists is actually decreasing. So how do you generate better diagnosis? It's using AI. Clearly, a strong synergy with the Diagnostics business with our advanced staining business with our PHC business in diagnostics. But also on the pharma side, leveraging those tools in research, increasing our speed in research, increasing our PTS in research as well. But Andy and Moritz will talk more about that as well.

Now let me briefly dive into the area of sustainability, also an area where we've made significant progress, and where also Roche is a front-runner. Just to give you a couple of examples here. On the one hand, on environment; the other hand, on access and innovation. In environment, since last year, we have 100% of all our sites globally now on sustainable energy, 100% of our sites. We've reduced water consumption by 10% last year, 26% since 2020. And if you look at greenhouse gas emission alone last year, we reduced it by 34%. We've now reduced it by 78% since 2004. And our goal is not to have a net zero. Our goal is to have a real zero in terms of greenhouse gas emissions.

If you look at access and innovation, just last year, again, we claimed #1 spot in the Global Patient Reputation Survey. This is done by patient organizations, and they rank pharma companies based on how we are to work with. We have global impact in diagnostics, 31 billion tests. That's more than 3 tests, 3 to 4 tests per inhabitant in the world. We've also invested in antimicrobial resistance with the first antibiotic in 50 years. So again, something where we contribute to society. Also in Africa, working together with different foundations to address the AIDS crisis and also expanding that into other areas such as cervical cancer screening. Again, significant progress in the area of sustainability, also making sure we take our responsibility here very seriously.

Now let me talk about people and culture because at the end of the day, we can only do what we do if we have the right people and the right culture in the organization. We have a highly motivated organization. 3 years ago, we talked about a high-performing organization. Now when I see some companies talking about culture in the organization, they also start to talk about high-performing organization. I just want to say, we've been talking about that for 3 years now. And not only did we talk about it, we've done substantial changes in the organization.

For example, the way we give incentives. We make sure that we have a strong differentiation between high performance and low performance in the organization. 3 years ago, we changed our bonus system. What do I mean by that? 3 years ago, we changed that people who are low performers, so greater contribution needed, would not benefit from company multipliers. So they would be capped in their bonus. So there is a significant differentiation to those people who have shown great performance. And what we've done is we created IMPACT Award to spot awards for those people who have really shown tremendous performance, again, shifting money from low performance to high performance.

If I look at our turnover rates, we have pretty much lowest turnover rate in the industry. But then if I look at the turnover rate between the different populations in our people, so between greater contribution needed, valuable and high performance, we have ultra low turnover in high performance and valuable performance. And because of the incentive schemes that we introduced, we have significantly increased the turnover rate in the low performance groups. So in the high performance, in the very, very low single digit. In the low performance, we are 35% to 37% turnover. That's exactly what we want because we want to make sure that we reward high performance in the organization. And with that, you change the culture in the organization to drive this kind of high performance. And we also don't want to make it comfortable for those people who don't deliver the performance.

And part of high performance is also leadership. So we've invested significantly over the last 3 years in leadership training of our people because at the end, people leaders also need to make sure that their teams perform in a very high way. They need to drive accountability. They need to create clear expectations, and they also need to make sure that we focus on efficient decision-making. We've also trained the organization more on different skill development areas because we feel that's also important in order for an organization to be a high-performing organization.

Now what does that translate to? At the same time, as we move into the shift of a high-performing organization, where we are on the journey already for years, we have excellent scores in terms of employee engagement. Just on Monday, we sent out the results of the employee satisfaction survey that we did over the last 3 weeks. And if you look at it, we have 82, and it's plus 3% higher than what we had 3 years ago. At the same time, if you look at the global benchmark, we are 12 points above the global benchmark. And 1 year ago, we were only 8 points above the global benchmark. So we increased the satisfaction of our employees while at the same time in the industry, the satisfaction dropped by 3 points. We are also externally recognized as one of the best employers by LinkedIn, but also by the Science magazine being #8 and #1 spot across all different pharma companies.

And we just put together a couple of quotes. So these are not quotes from our internal employee survey, these are actually quotes from external sites where employees give kind of a sentiment on what's happening in the company. And you clearly can see that we have a strong shift into faster decision-making, strong accountability, stronger link from incentives to achievements. Clearly, in terms of speed, more science rigor, more debates. And this is exactly the cultural shift that we wanted to make 3 years ago. And you can see that it has happened. And this happened while the satisfaction has gone up. And while we still do the changes that are necessary. And you may have seen that in some of our past presentations. We have cut about 1 billion cost out of the system in R&D that we have reallocated in the project, 1 billion, and that's about sourcing. This is about using AI to accelerate.

But it's also, for example, to give you one very concrete example, 3 years ago, Flatiron lost about 250 million a year. This first quarter, Flatiron is positive in profitability. So we've made a significant turnaround in that business, while at the same time, growing high teens. And that's exactly what we need to do. We need to become very, very productive to make sure that our money goes to the areas where it has the highest impact and where it creates speed in the system. And what I like, if I now see a progress in different parts of our R&D, I get e-mails from people in the organization. Oh, we just broke another record in how fast we recruited here. We broke another record how quickly we went to filing. We broke another record, and this is exactly this mentality that I want in the organization that people take pride in achieving the speed that we need in order to be competitive. And this is exactly the direction that we want to go.

Now let me briefly touch on the pharma strategy as well. And you have seen this slide before. This is why we're focusing on these 5 therapeutic areas. The 3 that I mentioned before, cardiovascular metabolism, oncology, neurology, making up about 50% of the global disease burden. If you add immunology, which plays a role in multiple disease areas anyway, and ophthalmology, you are at about 60% of the global disease burden. And these are the areas that actually grow. So we are focusing on the areas of growth and where we can reach as many patients as possible and where we can make a difference. At the end, this translates into the market. This is the size of the market. That's why we're focusing on it.

And if you see the progress that we have made in the pipeline, I already mentioned that we are at an all-time high when it comes to the value of our pipeline, a 63% increase versus 2022. At the same time, we have less volume. Now the volume will continue to increase again, but the point is we can't do everything at the same speed. We need to fast track certain programs so we had to have more money available for programs where we need to go extra fast. Trontinemab being a great example for Alzheimer's, for brain [indiscernible] antibody because it's one of the biggest opportunities we have with a clear differentiation. So we fast track those programs.

At the same time, I already mentioned, we need to become highly productive and efficient, and that's why we work on constant reallocation of R&D resources. Just because we don't announce like other companies, 8,000 or 9,000 people reductions, doesn't mean that we are not working very hard on productivity. We just do it in a very calm way and not in a very, I would say, that outspoken way. What we also introduced is this organizational -- we changed the setup in terms of end-to-end accountability across our entire R&D and also introducing business and scientific Boards. And these are important because these are the challenge Boards.

And what it translates to, you can see here, if you look at this on value and PTS, in dark blue, you can see those projects that have gone into late stage, significantly higher value, significantly higher PTS. In fact, if I look in the past, usually the PTS protection of our people was always correct. So when our people protected PTS is above 50%, usually, that meant that, that trial was successful. If you look at the average PTS right now that we have in the pipeline, it's around 70%. So clearly, a significant progress, not only in terms of value, but also in terms of PTS.

Diagnostics, I will not go into. I will let the Dia team handle that because I know you're going to be in very good hands with them. So I look forward to listening into that as well. Now let me finish with giving you a growth outlook. You've seen this slide before, consistent, strong delivery of growth, 5%, 8%, 5% sales, 8% in core EPS. In fact, if you look at the last 2 years and 3 quarters, in fact, we had an average growth of around 8%. So significant delivery. And I can say we've actually really delivered. And we delivered in this time period when we actually had the cliff of MabThera, Herceptin and Avastin as well. So we continue to grow and we also continue to grow through digesting also the COVID sales.

And if I look forward, between '27 and 2030, we have 19 medicines that could reach patients, 19 Phase III readouts. This is a very full pipeline that we have. And so we have a lot of trials now ongoing in the late stage. And in Diagnostics as well, we have strong pipeline, and we have already launched a number of blockbuster products, CGM, mass spec, very soon, the sequencer, but we also have a number of assays that are going to be very important for our menu. Lp(a)., I know this is also a big target right now in pharma. But also if you look at pTau217, you saw the press release on that for Alzheimer's NfL to look at the progression in multiple sclerosis and the interferon gamma release assay for TB is also one that people are really looking forward to.

Now you've seen this slide before. And what does that translate to in terms of the growth outlook for our company? Diagnostics, as you can see on the bottom, we'll continue to drive growth. It will be growth in the mid- to high single digits over the long term and corp growth ahead of sales. If you look at the pharma portfolio, and here is a big change to what we showed you in September. In September, we said the on-market portfolio and the data we have in hand is going to deliver growth until 2028. And then the on-market portfolio will compensate for any generic version.

Now given all the positive readouts we had in the last 6 months, 4 positive Phase IIIs of Gazyva in a row, the data on giredestrant adjuvant, which is a significant opportunity, adjuvant is more than 70% of the market for [indiscernible]. And also the fenebrutinib data, you see that we will no longer have a situation even with what we have in hand where we are not going to grow. We will continue our good growth momentum, not only until the end of the year, end of the decade, but beyond. We'll continue a strong growth momentum.

And with all the medicines that we have in the pipeline that are going to read out in the next 3 years, this will add on top, BD will add on top. So we can be confident about the growth momentum that we have and that we will continue to have. And if you compare that to the consensus, you can see on the left-hand side, the biosimilar gap that's expected if you take all of the different analyst numbers together, it's about 6.7 billion over a 5-year period. And then you see in the middle, the growth on the on-market portfolio and what's in the pipeline. In many of those areas, we already have the data in hand, giredestrant. Gazyva is a significant opportunity with the different -- in lupus nephritis or SLE, for example.

So if we then go through the list, personally, I would say there is upside potential, and there's definitely more potential for the upside than for the downside. And there are many things that are not uncovered by the models today. So there is the opportunity to drive growth into the future here as well. And let me just say, as I hand over to Matt, we have delivered in terms of what we have said 3 years ago in terms of growth, in terms of pipeline rejuvenation. Our outlook has fundamentally changed given all the positive readouts that we have seen, and we have a very full pipeline in pharma and in diagnostics that are going to drive growth also beyond the decade.

Thank you very much. And now I hand over to Matt.

All right. So welcome, everyone, to Dia Day 2026. So on behalf of myself, and all of our colleagues, we're really looking forward to giving you insight into the diagnostics strategy, our rich pipeline of innovation. You got a bit of a preview there from Thomas, and some of the innovative diagnostics solutions that are going to have a big impact on health care around the world and propel our business to that diagnostics financial ambition, which you just heard from Thomas. And so with that, let's get started.

So anchoring ourselves first with the strategy. The strategy has been unchanged for the last 6 years, and it consists of 3 key pillars. The first, we will continue to innovate and deliver high medical value diagnostic solutions. Second, we will deliver clinical decision support tools, and you're going to hear more about this from Moritz and Andy, to improve clinical decision-making. And then third, we combine these into solutions that span across the entire care journey and give patients and caregivers the tools to better navigate chronic conditions and diseases. And this also creates competitive differentiation from us against other diagnostics companies.

So how do we translate that into ambitious goals for the organization? So we are #1 in vitro diagnostics, and we have a goal to expand our position as the leader in the clinical lab. Second, we have the goal to become a leader in decentralized testing. You heard quite a bit from Thomas about why we feel that's an important area for us to remain focused as an organization. The third is to differentiate ourselves from our competition by offering comprehensive solutions across the highest disease burden areas. And the fourth, to deliver these in a way that help patients and caregivers better navigate the entire disease continuum in a way that delivers better outcomes for patients and society. A great example of this is our ecosystem we're building in oncology and neurology.

So now I'd like to explain a bit about how we are organized as Roche Diagnostics, which is centered around our customers. So going from left to right on the slide, starting with the core lab, this is the biggest business unit within Diagnostics and is focused on routine laboratory operations. The heart of this is our leading serum work area, which includes clinical chemistry and immunochemistry testing. I will also highlight that this includes the new mass spec system, which you're going to hear more about from Palani.

Molecular lab, covers both molecular diagnostics and sequencing. This consists of our mid- to high-throughput PCR systems, syndromic panel testing as well as the forthcoming AXELIOS sequencing solution, and Josh is going to take you through that.

Our pathology lab is centered around cancer diagnosis and companion diagnostics. So this includes our leading advanced and primary staining business as well as our digital pathology and personalized health care portfolio. Laura is here from VENTANA, and she's going to take you through that as well. And also, you'll hear about this from Andy.

Now near patient care provides solutions for decentralized and acute health care settings. This includes diabetes management, which is our largest segment as well as if you look at our LumiraDx, molecular, immunochemistry and clinical chemistry, and I'll cover that a bit later, but you'll hear more in depth from Ildiko, who's our Head of R&D for Diagnostics in her section.

So lastly, our precision oncology subsidiary, Foundation Medicine, which delivers comprehensive tests for oncology including our recent definitive merger agreement for minimal residual disease provider, SAGA Diagnostics. And I would call out that integrated across all of these, we integrate these lab solutions with digital products and solutions across the lab and the clinical settings to enhance their value to health care providers, and I would also call it this improves our competitive position against other diagnostics providers.

So now I'd like to talk a little bit about the size of these customer areas. The first is the core lab, as I mentioned, our biggest business unit. This is a CHF 40 billion segment. Now it's growing at 3%. Again, we hold the #1 position here. But as we expand with solutions like mass spec, we feel we can grow the segment in the future.

The second, the molecular lab, PCR valued at CHF 9 billion and sequencing valued at CHF 7 billion. We are #1 in the laboratory molecular diagnostics based market. This is growing at 4%. And as I mentioned earlier, we are going to very ambitiously enter the sequencing space, which is growing at 7%. Here, we see a lot of opportunity for growth, especially as testing moves more in the direction of the clinical space. In the pathology lab, we are the #1 provider for precision oncology diagnostics, and this is a market growing strongly at 8%.

Lastly, for near patient care, the market is growing at 3%, but what I really want to call it is one of the fastest-growing segments within that, which is continuous glucose monitoring, which is growing at 14%. And again, you'll hear more about this from Ildiko.

So given the growth dynamics in the in vitro diagnostics market, and as you heard, our ambition to grow above the market with mid- to high single-digit growth. Our focus is on the highest growth segments in the diagnostics market to drive our business above market levels. So we will do this by focusing on the diseases with the highest burden on global health. And if you look collectively across our areas of focus, these represent 60% of the global health burden. So I'll discuss some of the drivers that we expect to continue to accelerate testing in these areas.

Starting with oncology. Here, mortality is expected to increase 50% by 2040, and we expect testing to be driven by a significant rise in early blood-based cancer screening, precision oncology testing and the rise in minimal residual disease testing, and this is driven by its incorporation in early cancer regimens as well as the availability of those early-stage treatments.

Second, neurology, where mortality is going to increase by 67% by 2040. And as an example, people living with Alzheimer's is going to reach 50 million people by 2030, the majority of which are never diagnosed. We expect testing to be driven by the approval and access of new disease-modifying therapies that require early detection and therapy initiation, but also combined with the availability of these new early blood-based biomarkers.

So third, cardiometabolic mortality will grow by 40% by 2040. Here, we expect testing to be driven by rising prevalence of type 2 diabetes and decentralization and growth of consumer-based testing.

Lastly, infectious disease mortality will increase 29%. And here, we expect the growth to -- or this mortality and the testing to increase driven by the growth of STI testing as well as syndromic-based panel testing. And here, this is where our strategic investment and acquisition of GenMark is going to be a key part of this, an area where we're going to continue to expand our menu.

So now I'd like to talk how we're going to achieve each of these goals, and I'm going to start with how we plan to expand our #1 position in the clinical laboratory. So if you look at this slide and break it into the left and the right side of the slide, our strategy here is very clear. We will continue to invest in delivering innovative solutions in the areas where we lead. We will not be complacent with our leadership position, and we plan to deliver innovation in our core areas of strength. And the key examples are really on the slide.

So first, cobas ultra. Here, we will combine preanalytics and analytics and deliver high capacity together with intelligent sample handling. Our smart e benchtop immunoassay platform, which we'll launch this year, this is going to be critical to enable our market capture of [ TB IGRA ] as we decentralize out into smaller customers. Our next-generation molecular work area, which will replace our industry-leading 58, 68 and 8800 with the next-generation molecular solution. And in the same theme, we're building an advanced staining fully automated next-generation solution for the pathology lab, which will replace our leading benchmark ULTRA Plus platform.

Now while we continue our growth in the areas where we're strong, we also plan to disrupt the market with innovative solutions that we will launch with the scale that only Roche Diagnostics can bring to the market. And here, I will call out our i 601, the industry's first fully automated clinical mass spec solution. Again, Palani will cover this, but we are seeing really strong excitement from our customers. Some of our customers have even put out press releases when they've acquired the system, and we're seeing very strong interest from the market.

I'd say the same theme with our soon to launch AXELIOS sequencing solution, which will bring unprecedented throughput and flexibility combined with high accuracy. And then I would also say beyond platforms, and I spoke with some of you at the break about this, we'll continue to innovate on our digital solutions as well as our chemistries. So an example here on the slide on the bottom, you'll see our protein to DNA workflows that will enable detection of extremely low amounts of proteins. This is critical for early detection of neurological illnesses as well as cancer. And a second example, well detailed by our definitive merger agreement with PathAI is our importance of AI enhanced digital pathology as a significant area of focus.

So with that, I'd also like to talk about how we plan to become a leader in decentralized testing. So if you look at this slide, you can see the different settings where decentralized testing is utilized from the emergency room and moving out farther away from the hospital to GPs, pharmacies and then even consumer and patient-based testing. And then you can see on the left-hand side, infectious diseases, cardiometabolic and diabetes. These are the conditions where an answer is critical to know right away that are most amenable to these kind of point-of-care testing modalities. And here, what I would call out is that we're going to cover every major point-of-care segment with our solutions in our pipeline and that we have been very intentional about how we built a portfolio that enables us to compete in terms of quality and our cost position, and Ildiko will take you through this as well.

So for the last 2 goals, I will combine them in a single section, how we differentiate ourselves through high medical value solutions and comprehensive disease management. But first, I want to really level set on how we define the different stages in the diagnostics patient journey. So first, on the far left, early detection. This applies to screening tests, which enable health care systems to move to preventive care model, reducing global health burden. Again, as Thomas mentioned, our investment in Freenome, having that option space in early cancer detection. We have a strong level of interest in early cancer screening, but screening more broadly.

Precision diagnosis, this refers to testing that provides precise characterization or disease and enables appropriate patient management. Therapy selection, solutions that enable optimal patient management and therapy selection. A great example is the CGP product delivered by Foundation Medicine. And then disease monitoring. This refers to longitudinal testing solutions and digital tools that enable patients and physicians to manage chronic conditions. An example here is our Diabetes Care business, for example, our CGM. So this is how we divide up how we view the diagnostics patient journey and where we build specific solutions.

So now I'd like to start by walking you through our 4 key disease areas where we see the opportunity for diagnostics as well as the growth potential in each of these segments. And we give an outlook for the segment growth to 2030. You'll hear in detail from the team about our solutions that we're developing for each of these segments.

So I'll start with the segment that has high unmet medical need and where we're investing significantly as Roche to develop best-in-class diagnostics. Early detection for oncology is experiencing significant growth. You see this in the 35% compound annual growth rate for multi-cancer early detection, which is driving this segment to CHF 5.3 billion opportunity by 2030. We feel we have a strong position and optionality in this space given our partnership with Freenome.

Therapy selection for oncology is a CHF 17 billion opportunity, growing at 9%. And here, we see growing demand for targeted companion diagnostics, digital pathology solutions as well as comprehensive genomic profiling, all of which are an area of significant investment for Roche Diagnostics.

Lastly, disease monitoring is also projected to grow at 32% to CHF 5 billion. This is driven by uptake of minimal residual disease testing. Here, our FMI portfolio and AXELIOS solutions will play a key role, especially with the recent acquisition of SAGA Diagnostics.

And so you heard Bruno mention this at the start. But just to be really clear, in oncology, we have built a unique set of assets that span the entire continuum of testing and uniquely position us versus other competitors to lead in this exciting space. So you'll hear detail from the team in their sections, but at a high level, what I want to point out is our industry-leading portfolio of assets.

First, I would highlight, we have the full complement of IVD technologies to enable leadership of lab-based oncology testing. And I would really emphasize that our AXELIOS technology platform is key to drive the future of molecular-based oncology testing together with our digitally enabled IHC VENTANA business, where you see the strong synergy between VENTANA and PathAI. We're also well positioned to compete in the precision laboratory business with our Foundation Medicine organization, especially with the soon-to-be finalized addition of SAGA Diagnostics to deliver on the opportunity in minimal residual disease testing. But we've also, as mentioned previously, built a strong position in option space in early cancer detection with our partnership with Freenome.

All together, we have built a full end-to-end oncology portfolio that will enable us to shape the future of the oncology testing in the market in a way that no other company can. And the team will take you through the details in their presentations. But I'm very excited about how we've pulled this together and built a leading position from end to end in one of the most critical opportunity spaces in diagnostics.

So now you heard about this earlier, but I also want to take the opportunity to really highlight a key part of our oncology strategy, which is the recently announced definitive merger agreement with the PathAI team. So PathAI, as you heard, is the leading provider of digital pathology image management systems. And again, this really enables pathologists to augment existing immunohistochemical or primary staining workflows with AI-powered digital insights to improve patient management. And what I really want to call out here is this business is entirely complementary to our VENTANA business. It also has high levels of complementarity to our pharma services business, both in the pathology space, but also at FMI. It gives us the ability to engage earlier with pharma companies to build companion diagnostics programs and it provides that seamless complementarity again with our VENTANA business. This is going to position us strongly for the future. You're going to hear more about this from their CEO, Andy Beck; as well as our Head of RIS, Moritz Hartmann, but this is a key part of our oncology strategy.

So with that, I'd like to continue to neurology, which is the disease area undergoing rapid growth, again, to our aging society. And I will start with early detection, which represents a CHF 480 million opportunity by 2030. And this is going to be driven again by the availability and proliferation of blood-based biomarkers. So Olivier will talk about this in his section. But here, I would call it that we've just announced today the first CE mark for pTau217 blood-based rule in, rule out with an intended use in primary care.

One of the biggest barriers to diagnosis of Alzheimer's is the referral into secondary care, the availability of imaging and lumbar puncture based CSF test. And by having an intended use in primary care, we have a meaningful competitive differentiation against other test providers.

Now therapy selection is projected to grow at 17%. And here, we mostly focus on optimizing selecting patients for side effect profiles based on their ApoE4 status. But we also see an opportunity, and Thomas mentioned this earlier, for looking at disease progression as disease-modifying therapies enter the clinic, the need to understand progression and optimize patient therapy will increase. So here, I would call out our neurofilament light chain assay, which also received CE Mark in the first quarter, which is able to detect neuro inflammation and be able to help physicians better optimize treatment for patients with multiple sclerosis.

So now I'd like to continue with cardiometabolic testing, a segment where we are the market leader. Early detection is a significant market, valued at CHF 6.9 billion by 2030 and one where we are making significant investment, both with our core lab and near patient care portfolios. In early detection, we focus on earlier risk stratification in triage through novel biomarkers and integrated digital and diagnostic solutions.

Acute diagnosis of cardiovascular disease is at CHF 4.2 billion opportunity. And here, we feel that we have a focus on precision diagnosis of cardiovascular diseases with biomarkers such as troponin, which we plan to augment with new and novel biomarkers to improve the diagnosis of myocardial infarction. And here, I would also call, as you heard, Lp(a), while not really for the acute setting, is a test that is really exploding in terms of its utilization. And these are the kind of tests we will also use to augment our cardiovascular portfolio.

So cardiovascular disease monitoring is one of the largest opportunities in med tech overall. And this is primarily driven by glucose monitoring for people with diabetes. This segment is experiencing strong growth, driven by the explosion of continuous glucose monitoring, again, growing at 14%. This is a key market where we plan to expand.

So on that topic, I would like to transition to our Accu-Chek SmartGuide CGM. This is our AI-powered CGM solution that will help people living with diabetes manage their condition by addressing one of the biggest challenges that people living with this disease face, which is uncertainty. The algorithms we provide, such as our glucose predict and night low predict address the concerns people with diabetes have around how their glucose values may affect them, especially during sleep. We are developing future algorithms around therapy initiation as well as diet and exercise. And, and you'll hear about this from Ildiko, a next-generation sensor that will add to this portfolio of tools to enable people with diabetes to live their lives as they choose rather than centered around their disease. We're excited about our progress here in one of the largest and fastest-growing segments in med tech, and you'll hear more about this from Moritz and Ildiko.

So now coming to our final disease area, infectious disease. Here, diagnosis is a very large segment, predicted 2030 market value is CHF 25 billion. And this is really driven by segments such as sexually transmitted infections, where you saw from Thomas slide, are BV/CV, and we are continuing to expand our menu there, and syndromic panel testing as well as screening for diseases such as latent tuberculosis, which we will cover today as well. All of these are areas of significant focus and investment for Roche Diagnostics.

Disease monitoring here is a CHF 1.6 billion opportunity and one where Roche has and will continue to have a significant presence. This is driven by quantitative testing for patient viral load, and this is an area where I would say we've really shaped and built the market over several decades.

So now I'd like to summarize. So we gave really where we are going to be focused, how we're going to win, and I'd like to summarize that with our financial ambition. We are the leader in in vitro diagnostic testing, and we have a clear strategy to stay ahead of our competition and outgrow the market by focusing on the highest growth segments with differentiated solutions. You heard about the AXELIOS solution, which has a differentiated value proposition, and will make a profound impact on the sequencing market for years to come. And this will lead us to a mid- to high single digit -- mid- to high single-digit growth.

For core operating profit, you heard from Thomas, we will grow our profit faster than our sales. This is going to come from the profit contribution of our core business as well as our new launches as well combined with continued operational excellence. So mid- to high single-digit growth with profit growth faster than sales.

And so now to get into our portfolio, I'd like to hand it over to our Head of Roche Diagnostic Solutions, Palani Kumaresan. So thank you very much.

Thank you, Matt. A warm welcome from my side as well to everyone in the room and also those who are joining virtually. I'll try to keep us moving faster because we do want to keep the time for Q&A with you guys. I know many of you have many questions. I'll spend the next few minutes going through the Roche Diagnostic Solutions pipeline. So let me get started. Matt very nicely already laid out the 4 customer areas that are within Roche Diagnostic Solutions. The important point I will mention additionally here is these customer areas very nicely mirror how our customers are set up in the markets.

Now one of the core elements of our Diagnostics business is instrument placements. And so I wanted to start out also showing you how our instrument placements have grown over the course of 2025. These are some key instruments across the 3 lab-based customer areas. You can see very nice double-digit growth. And as we look to the future, this is in one slide gives you a holistic view of some of the key instruments and platforms that are going to be hitting the market in the coming years. '26, obviously, AXELIOS is an important focus, but also cobas smart e. I will spend a little bit of time talking about that for immunoassays. Over the next 3 years, '27 through '29, important additions to part of our automation portfolio, but also our near patient care portfolio. And then as we go into 2030, early part of next decade, Matt already mentioned, next-gen advanced staining, next-gen molecular work area and continue to build on our automation portfolio.

Now going hand in hand with our instruments is a very rich pipeline of tests. And this is, again, a holistic view on one slide. Over 140 new tests that we'll be looking to launch across the 4 disease areas with highest disease burden that match share, but also across some other areas that you see in the bottom. On the right-hand side, some examples, some of them are highlighted in red dotted lines. And the team after the break will get into the details of many of these tests.

So for the remainder of my presentation, what I'll do is I will very briefly, for each of the 4 customer areas, share with you the 5-year strategic priorities and then some key highlights, portfolio highlights. So let's start with Core Lab. Here are 3 important strategic priorities, evolve our next-generation systems portfolio. So I'm here mainly talking about mass spectrometry and also our cobas smart e. Develop integrated lab automation offering. Here, the centerpiece is going to be cobas ULTRA, and I'll say a little bit more about it in a bit. And then continued menu expansion with focus on high medical value assays, particular focus on neurology and infectious diseases, which we will, of course, get into more later on.

So mass spec, I wanted to give you an update here for some of you who might be a little bit new to this segment. CHF 3 billion segment, largely manual, labor-intensive and laboratory developed tests. With the introduction of our fully automated IVD solution a little over a year back, we've had some very positive feedback from customers. Performance, they really like. They absolutely like the end-to-end automation as well as the robustness of the instrument as it has come into the market. We have had placements across influential labs in Asia, in Europe as well as in North America with LabCorp.

Now by the end of this year, we are targeting a total installed base of 100 instruments. We have 39 assays that are launched in CE mark countries, and we look to continue to grow this menu. Now talking about growing the menu, I wanted to quickly share on this slide, how is the menu coming along. Top of the slide, you can see the 39 tests that we have launched in CE mark countries. These are focused on steroids, vitamin D, and therapeutic drug monitoring. So we are talking about immunosuppressive drugs, antibiotics and antiepileptics.

Going into 2026, we are looking at launching 4 more tests in CE mark countries. Now U.S. is the single biggest market for i 601. Here, we have already approval for the instrument along with vitamin D and 6 steroids tests in a CLIA exempt setting. And we are actively working with the FDA to bring the rest of the test that you see in the upper half through a 510(k) process towards the end of this year and then going into next year. Simultaneously in development, we are already working on the next wave of mass spec tests, and that would be primarily drugs of abuse testing.

Let's switch to cobas smart e. This is a low throughput stand-alone immunoanalyzer targeting low-volume labs. So low volume labs for immunoassays is around a CHF 1 billion market, growing at around 5% CAGR. Now in this market, we have had cobas e 411, and we have a very healthy installed base, over 18,000 installations, and this is going to be a successor to e 411. On the left-hand side, you can see some key highlights of this instrument.

One important call out, it will use the same [ reagent cartridge ] as it's used in our other immunoassay platforms, the e 402 as well as the e 801 that is -- that are widely available in the market. And this is something that is really liked by our customers to have the family concept. You can see on the bottom right, some key advantages of this platform. It's a benchtop, built very robust for remote settings with limited maintenance requirements. And we are looking to launch this and see more countries towards the end of this year and then subsequently in other markets, and it will have the full menu of immuno tests that we have on our other platforms like e 402 and e 801.

Now on automation. Here, I mentioned earlier about cobas ultra. And here, we are taking a very systematic approach to bring a series of innovations to the market. This year, we will have control launch of cobas ultra preanalytical system. This is the successor to our highly successful p512 and p612 pre-analytics solutions in the market. Next year, we will bring some important modules to this, like the centrification, which is very important for our customers, but also by directional movement of tubes, cobas ultra dual transport, and mobile robots to allow for automation across islands of analyzers in the lab.

And then working towards our omni transport. This is the ultra-high throughput, 2-dimensional single-tube transport that we are working on with our partners, Hitachi, and we will also bring modular pre and post analytics systems and analytics capabilities.

So what does this all mean? First of all, high level of modularity means our customers only pay for the hardware that they really need. Two, flexibility, space efficiency, it can really work with limited space, lab space, which is an important worry for a lot of our customers. And industry-first ultra-high throughput movement of tubes across different analyzers.

Now before I move on to molecular lab, I did want to very briefly talk about the neurology menu, especially on Alzheimer's. I won't go into the details here because you will later hear from Olivier a very nice data on phosphorylated-tau 217. But what I wanted to emphasize is we are taking a very systematic approach to introducing several different assets leading up to now phosphorylated-tau 217, and we will continue to increase or improve on a build on additional intended users and additional tests coming into the market in the future.

Now moving on to molecular lab, 2 important strategic priorities. On the PCR side, we want to continue to build a menu of tests, primarily looking at sexually transmitted infections, GI, gastrointestinal infections, also hospital-acquired infections on that ecosystem of PCR instruments we have. And on the sequencing side, of course, launched our AXELIOS sequencer.

Now on this slide, you can see that entire ecosystem of PCR platforms that we have in the market. Very briefly, on the left-hand side, we have our high-volume fully automated IVD solutions. 58, 68, 8800. Also the preanalytics solution, cobas prime. For point-of-care testing, this is, of course, liat. On the right-hand side, the open molecular solutions are two, we are talking about digital light cycler and life cycle pro, where we have a broad menu of research use only and IVD test from our TIB Molbiol side of our business, and syndromic panel on ePlex platform from GenMark.

Now Josh will later tell you how that menu is evolving on these platforms. But I also wanted to shine a quick light on donor screening, which is an important part of our PCR portfolio. And this is essentially blood testing for blood donations as well as for plasma fractionation. Together, the serology and nucleic acid testing, NAT is around CHF 1.8 billion market. On the NAT side, we see a good growth based on a few factors. There's adoption of NAT in many NAT naive countries, which still only perform serology testing. We see increasing testing of emerging pathogens. A good example of this is testing for malaria, and increased demand for plasma-derived medicinal products.

While there are many differentiating factors to our portfolio. The most important call out is this end-to-end integrated solution across serology and nucleic acid testing using our automation solutions as well. And towards this, we are introducing some additional solutions. So you see cobas SynergyPlus, which will allow for automated blood pooling, which is a common practice in many countries that are performing NAT testing. And also last month, we launched a multiplex test, MPX-E, which is bringing hepatitis E virus into our already multiplex assay for HIV 1, 2, HBV and HCV. And this allows for essentially workflow simplification for our customers because they don't have to perform a separate [ HEV ] test. So this is a great business for us, very robust, and we continue to maintain a good presence in this market.

Now on AXELIOS. I wanted to mention only a couple of points. Now last year, we introduced the technology at AGBT in February. And since then, we have systematically worked with early access sites like Hartwig in Europe, like Broad Clinical Labs in the U.S. and so on, to systematically demonstrate many of the applications that you see on the left-hand side. And as we get closer to launch in the summer of this year, we will be coming out with application notes that will lay out for our customers, future customers how to realize these applications in their own labs.

Now last month, we brought together a few hundred prospective customers from the U.S. and in Europe. And we had really tremendously positive feedback from the customers, both around speed, throughput, accuracy, the flexibility and of course, the pricing, the $0.06 per million repricing. And Josh will definitely get into a lot of the details here because we know that you're very much looking forward to hearing more about this.

Let me switch gears. Go to pathology lab, 3 important strategic priorities. Strengthening our advanced staining core business, both in histology and cytology. Driving digital pathology adoption. And here, of course, a definitive merger agreement with PathAI will play an important role. And then continue to build our assay menu, the test menu. And here, our pharma partnerships play a very important role, and I'll talk to you a little bit about that.

So how does advanced staining feature? Now here, we have the largest installed base of our advanced staining platform, Benchmark ULTRA and ULTRA PLUS, which is the core of our business. But you see we have primary staining, special staining platforms to go hand in hand as well as our digital pathology scanners. And when it comes to digital solutions that you see in the bottom, we have important solutions for lab automation. And then PathAI's AI site will become an important part of our digital offering for image management solutions.

Now as we look to the future, our next-generation advanced staining platform that we are developing with Hitachi, which will be a successor to Benchmark ULTRA PLUS, that will be important, end-to-end workflow automation here is not only software automation. We're also talking about bringing some clever hardware in the form of mobile robotics, flexible robotics and so on into the pathology lab as well. And then the AI-driven ecosystem, again, PathAI will be a key underpinning for this.

On the assay side, something that is quite remarkable is we have a menu of over 250 immunohistochemistry and in situ hybridization test on our advanced staining platform, and we continue to bring more innovation. We are looking at brightfield multiplexing tests, digital pathology algorithms. But what I wanted to point out actually is on the right-hand side, our personalized health care services business. Stand-alone, this is a great business for us. We have right now over 85 plus pharma partnerships, over 150 ongoing in vitro diagnostics programs. And what this enables us is having -- continuing to build this menu allows us to drive the placement of our instrument. And as we place more instruments in the market, we drive more access to all of these tests. And this is something that our pharma partners really like because they want to have broad access to the companion diagnostic test to make sure that the drugs can be really available to patients all over the world. So not only it's a great stand-alone business for us, it's also synergistically works very nicely with our core business.

Now to close out, I want to show you a couple of slides on near patient care. Key strategic priorities, building a leading portfolio in clinical chemistry and immunochemistry across hospital, decentralized and close to patient setting. Strengthening our position in molecular point of care through systematic menu expansion on liat. Here, the important focus is on sexually transmitted infections. And then deliver a competitive CGM solution.

Now Matt already showed the 6 platforms that form the base for our near patient care portfolio. I wanted to do one more click, share some key launches that are coming up, and then Ildiko will go into the details of some of these.

cobas sense is essentially our benchtop immunoanalyzer bringing lab-like performance in an emergency room setting, particularly for cardiac markers like high-sensitive troponin T. cobas vital is our next-generation blood gas and electrolyte platform. This builds on [ B1, 2 , 3, ] it will bring some unique algorithms like venous to arterial conversion but also significant sustainability and cost benefits. cobas pulse is a handheld hospital blood glucose monitor. We've had very nice growth in installed base since the launch in EMEA, LatAm countries over 24,000 placements. It also brings in a nice integration of third-party apps into this handheld device in a hospital point-of-care setting, and we're looking to launch it in the U.S. later this year.

cobas liat later this year, the lesion panel for Herpes simplex virus, [indiscernible] virus, which causes shingles and syphilis, also with mPOX. And then following that launch of Trichomonas and Candida vaginitis, again, really rounding off our STI menu.

LumiraDx, our focus is on menu that would really drive its adoption in highly decentralized setting. And this is where we are looking at the lipids panel, but also the next generation of HbA1c. And when it comes to Accu-Check SmartGuide with our current generation in the platform getting factory calibration, and pediatric claim is a big focus. And then we are also simultaneously working on the second-generation patch, both of which actually Ildiko will later show you how do these patches look.

With that said, I want to close out my section. And I'll hand it over to Moritz and Andy to talk through the Roche Information Solutions portfolio. And then after the break, we'll come back and we'll double click into some of the key parts of the Roche Diagnostic Solutions portfolio. So thank you for your attention. And Moritz, over to you.

Thanks, Palani. Yes. Thanks, Palani, and welcome as well from my side for being here today. I want to highlight on how our Digital Solutions portfolio and pipeline plays a strategic role in driving value for our customers and for Roche. I will talk specifically about how we're using AI to even elevate that value further. And of course, a good portion of this section will focus on the exciting opportunity with PathAI.

Our ambition is to transform diagnostics with data and digital solutions, moving from our strengths in the central lab towards decentralized care. We have 3 main value propositions across our digital portfolio and pipeline. The foundation for us is the connected lab, where our diagnostics insights portfolio delivers operational excellence to labs. We're making very strong progress in this segment where we have already converted our industry-leading analyzer footprint into the largest connected digital software footprint globally.

Building on this foundation, we're expanding our offering into our second value proposition, the informed clinician. This is where our solutions for clinical workflows and clinical decision-making drive faster and earlier diagnosis and better patient outcomes.

In our third value proposition, we're providing solutions for patients living with chronic diseases such as diabetes to better engage, control and manage their diseases.

Digital solutions contribute in 2 ways to accelerate the overall diagnostics value creation. First, our digital solutions drive important differentiation of our core diagnostics offering. This increases our win rates and actually drives longer customer lifetime value, including through our industry-leading portfolio of data security solutions. With our clinical support solutions and our patient-facing solutions, we're focusing on additional value streams, utilizing our market-leading strength and capabilities. This is why our digital solutions here are either FDA approved or CE marked, which is a very important and much appreciated differentiation factor for our customers. We're also seeing progress with reimbursement and adoption. Recently, the first European country has included an innovative algorithm and reimbursement, and an Asian country has included the same algorithm in their national guidelines.

Let's move on to how we utilize AI to transform the laboratory space. Here, our focus is to turn a fragmented, highly customized, but inflexible software landscape into an AI-native lab operating system environment, essentially doing for lab software what we have done with the [ same ] work area. We're approaching this in 3 phases. Phase I, which we plan to make available at the end of this year. We will build a native AI foundation using operational data. This will allow us to optimize workflows such as better demand forecasting or quality control and normally, detection.

In Phase II, we will build lab clinical intelligence by adding lab results to our trained LLM, and enabling use cases such as the interpretation of test trends or correlation between lab performance and outcomes.

And in the third phase, we will really create a holistic precision engine by now adding selected EMR data to this LLM, which will bring context aware reasoning across the entire patient workflow.

As you can see, we're executing on our digital strategy and our broad digital solutions portfolio and pipeline demonstrates our commitment. In the following presentations, you will hear more about our digital solutions that are very closely linked to other parts of the portfolio. Those highlighted in red here. The navify clinical hub, which complements our FMI offering or the Accu-Chek Care platform, and our leading customer experience, mySugr, which is now part of our CGM solution.

For now, I would like to share more on the exciting news we shared with PathAI. And as Matt already highlighted, the digital pathology market outlook is very strong. While the PathAI transaction is currently subject to regulatory approvals, we're excited by the potential to augment our offering in digital pathology by complementing the scanners and actually the full portfolio of pathology and also our algorithm offering. We will enhance the workflows with a next-generation AI-centric image management system, and we're turning -- and when turning towards pharma and our biopharma customers, we have the potential to build on our already existing successful partnerships in developing companion diagnostics algos. Finally, we will be able to provide a strong portfolio of AI services for biopharma customers with clinical trial efficiency and biomarker discovery.

With that, I'm very happy to welcome Andy back on stage, CEO of PathAI, who is our guest speaker for today and to walk you through more details of that. Andy?

Great. Thanks, Moritz. Thank you. Great. Well, it's my pleasure to introduce you to PathAI. Our mission since founding the company 10 years ago is improving patient outcomes with AI-powered pathology. From the beginning, we've been focused on technical leadership in AI as well as software engineering and medical leadership really focused on developing this technology for applications in pathology. We're also very focused on discovery, science as well as clinical and regulatory validation, and I'll provide a couple of examples of those today.

A key focus of our strategy from the beginning and where we see tremendous synergy with Roche is we're very focused on having one foot in biopharma and one in diagnostics and using those to strengthen each other, as Palani mentioned, with this virtuous flywheel where if we make discoveries in biopharma and we're already widely distributing diagnostics, we can then scale and distribute those insights globally.

And lastly, we're very well positioned early in a market that is just starting to inflect now. Today, the minority of pathology is digital. We expect in 10 years, it will be absolutely routine that pathology is done digital and with the assistance of AI. So we're aiming to build the operating system for precision pathology to power patient diagnosis and drug development at a global scale. And we have these 2 synergistic customer segments, both supported by one underlying AI flywheel.

For clinical diagnosis, we support the pathology lab by building tools, a broad menu of tools to cover the full spectrum of needs in the lab with the overall value proposition that digital and AI will drive increased accuracy, increased efficiency and increased access to precision AI tools to enable patients to get accurate diagnoses and to be recommended the right therapies.

The work in biopharma is very focused on how we can best use AI-powered pathology to serve 2 of the major value drivers within pharma, which is increasing the probability of technical success in clinical development programs, generating new pathologic insights in the mechanisms of response and resistance, and then with this synergistic flywheel, how can we use AI-powered digital pathology to not only increase the probability of success of the trial, but also to accelerate time to peak sales and even increase the amount of peak sales.

So to give you one example of this from our work in translational research, where we aim to unlock insights from whole slide images to drive advances in drug development. I'll focus on our PathExplore and IHC Explore products because I think they very nicely highlight these concepts. These are our products focused on exploratory translational research in oncology. So if you look at these 2 images on the left side of both of these panels, the one on the left is the H&E, which is the standard stain in pathology. It's been used for 150 years. If the pathologists were to look at that, they could easily say this is cancer. But what would be impossible for them to do would be to provide a single cell dissection and enumeration of every cell within the image as well as characteristics of the tumor and the tumor microenvironment.

With our AI-powered system, we can provide not only the -- support the diagnosis, but also provide this rich quantitative feature set, which can then inform studies into insights of pathologic mechanisms of resistance and response to therapies.

We do something very similar with our IHC Explore product, which is deployed on the core biomarker methodology in tissue pathology, IHC, except in addition to identifying every single cell within the image, we can also assess whether that cell is expressing a drug target of interest, which is valuable across oncology, but in particular, in immuno-oncology and to support studies of antibody drug conjugates.

We've been partnering since 2024 with Roche Diagnostics, a leader in tissue-based companion diagnostics. And the partnership is really focused on going from potentially finding a signal using an exploratory tool like IHC Explore, but how do we plug that into Roche's industry-leading platform for turning insights, turning signals into potentially regulated diagnostics not only to support registrational trials, but also to support global commercialization.

And I'd like to provide one case study of work led by our colleagues at Genentech that was presented as an oral presentation at the AACR meeting in 2023 that really, I think, highlights the power of the information that currently is hidden within these H&E images, which are literally created on every single patient in major therapeutic areas like oncology, inflammatory bowel disease and MASH.

So the investigators from Genentech wanted to answer a question. They said, if we applied deep learning power technology to a large set of H&E images from the OAK trial, could we identify a novel patient population who may show the best response to anti-PD-L1 therapy? And then beyond that after finding potentially this population, could this generate new insights into the pathologic mechanisms of response and resistance? And lastly, could there even be a world where identifying this new population in the future could expand the population of patients eligible for immunotherapy?

And what I want to highlight in this Kaplan-Meier curve on the left is this very top curve of the patients who are showing the best response to immunotherapy had a signature, which these investigators characterized as CP1 signature. So this is the population that showed the best responsiveness to anti-PD-L1 therapy. And when they looked into what was driving the signature, there's a very clear biological signal, which was increased lymphocytes and increased plasma cells with decreased fibroblast, was really the sort of cellular [indiscernible] that caused or influenced the environment to make patients most responsive to this therapy. So that's a new biological insight and a new population showing best response to this therapy.

And then they ask the question, is this only in [ all-comers ] or in the overall population as indicated on the left? Or would this even work within the PD-L1 negative population? And they found interestingly that the CP1 low population also showed strong response to anti-PD-L1 therapy even if they were PD-L1 IHC negative. So this is a new biological insight, potentially new prognostic signature, but also potentially a new strategy for increasing market for immunotherapy beyond just PD-L1 IHC.

So that was the biomarker discovery cohort. Because Genentech had digitized so many of their clinical trials, they were then able to validate this finding on a held out data set, which was IMpower150. And they found very similar response where overall strong predictive effect. And even within the PD-L1 negative population, the hazard ratio for the anti-PD-L1 containing regimen was 0.56 compared to about 0.7 in the intent to treat or a biomarker valuable population or much higher within the CP1 high population. So this is just an example of all the data that we have access to, but there are many insights to be discovered.

And this is one example of PathExplore applied to oncology. We also have a rich algorithm menu in liver, IBD, inflammatory bowel disease, as well as a number of different products for the clinical lab. And 2 that I'll quickly highlight is our PathAssist Derm product, which obtained FDA breakthrough device designation in Q1 of this year, and our AIM-MASH AI Assist, which is for use in registrational trials in MASH, which obtained both EMA and FDA biomarker qualification in 2025, which is the first AI algorithm to be qualified by both the EMA and the FDA for use in MASH trials.

On our clinical side of our business, the distribution of our platform is really driven by AISight Dx, which you've heard is a best-in-class digital pathology imaging management system to support diagnosis, collaboration as well as access to AI tools. This received CE-IVDR and FDA clearance last year as well as was the first digital pathology image management system that obtained a predetermined change control plan to enable us to rapidly expand the label and add new scanners and new monitors as needed.

So in summary, we expect that although the minority of pathology today is digital and utilizes AI by 2035, this will be completely routine. This will be driven by overall increased digitization as well as the massive increases of AI that we're already seeing today. But you can only imagine where it will be 10 years from now. Because of that AI-powered CDx, we expect we'll accompany most new approvals.

AI will be run in every single slide, driving increases in both quality and efficiency. This will also enable the generation of massive real-world data sets from billions of slides that can drive R&D decisions, regulatory decisions as well as policy decisions. And we're really -- our mission is really, how do we lead this over the next decade, enabling AI -- native AI slide to become the operating system for pathology, to leverage the tremendous synergies we see with Roche, to generate data at scale from many different areas where we're already synergistic, that you already heard, including FMI, VENTANA, as well as the new NVIDIA AI factory to enable data at scale in frontier AI to accelerate precision medicine. We have ambitions to develop the broadest IVD approved menu in the industry and to continue to focus on pathologists-led innovation to ensure that we're focusing on the products that matter most to physicians as well as to patients.

I just want to end by thanking everyone for their attention. I'm super thankful for the Roche team for their amazing partnership as well as for including me today. And then I truly believe that this partnership will enable us to realize our mission of improving patient outcomes with AI-powered pathology at unprecedented scale and speed.

And with that, I'll hand it over to Birgit to lead Q&A. So thank you.

Hello, everyone, and a very warm welcome from my side as well. So my name is Birgit Masjost, and I'm leading the Diagnostics IR team. And I welcome the speakers from the first session on stage now for the Q&A. I will be moderating the Q&A, and I would very much invite you to raise your hands if you have questions. And for the people who are joining online, please enter your questions in the Q&A tool. Also, I would really like to ask you to restrict your questions on the topics that have been presented today, so mainly focusing on the Diagnostics division.

Yes. And I see Matthew has his hands up. So Matthew, first question goes to you.

It's Matthew Weston from UBS. Thomas, in your opening comments, you talked about your confidence in increasing growth in pharma. But we haven't talked about the financial implications of all of the diagnostics growth opportunities that you've laid out today, Matt.

So I don't know if you can help frame that in terms how you see the midterm growth outlook. Do you see it accelerating, particularly with the new acquisitions? And when we get back to our desks, get our models out, what are the growth numbers putting in? What's the profitability we should be putting in when we take everything we've learned today in terms of real financial performance?

Yes. Great question. And so if you think about our ambition, which is mid- to high single digit, we know that we still have this headwind that we're working through on China. And I think as we get through this really in the next -- we say next year, we expect it to stabilize and then return to growth afterwards, I think then we can really have an assessment of if we want to take that ambition up. But I think at the moment, we want to keep it at that mid- to high single digit as we digest this headwind. But I think absolutely, we expect to see significant tailwind coming from these acquisitions and the new launches. But I think we want to make sure that we have the right baseline to set. But so for the moment, we'd say our ambition is mid- to high single digit.

And again, on the -- you said on the margin, profit will grow faster than sales. And what that means is we will consistently improve the margin over time. That would be how we would frame it.

Okay. Next question goes to Peter Verdult.

Peter Verdult, BNP. Could I ask a question just to you, [indiscernible] on margins? And then hand over to my esteemed med tech colleagues for some deep dive on diagnostics. But for me on the margin, it's just what's the real driver there? Is it mix? Is it SG&A leverage? Is it R&D efficiency? Is it a little bit of everything? But just a little bit more about growing faster than sales, what's the key driver? And then just with oil, and it's topical with oil, looking likely that it's going to be sustained above $100. Some of your peers are talking about 100 basis points of headwind from that. I know it's not a big deal for pharma, Thomas, but maybe a bit more for Diagnostics. Anything you can say about that would be helpful.

Yes. So all of those factors are factored in our financial ambition, but to talk about margins. So again, we are -- we have had and we've had since last year a focused operational improvement program. That's how we worked to stabilize last year. And I would say that's something we're continuing this year. And at the same time, we're also investing in our launches. And a big part of that operational efficiency was to make sure we have the space in our P&L to invest in our launches, which are going to drive future profitability. That's how we're going to get to that profit growing faster than sales. And so we think mid- to high single digit on sales, then you would assume mid- to high to high growth on profit. And that's how we're going to continue to -- that's how we're going to increase the margin over time.

And you mentioned some of these headwinds this year. I mean that's all factored into our ambition this year, which I think we talked about at the start of the year, which is to broadly stabilize the margin this year, and that's where we want to get to, recognizing there are these headwinds but they are something that we factored in when we talk about ambition.

And let me just bring it back to the group level. If you look at group, the last 2 years, we grew profitability by 13% and 14% incorporating profit. So you see we have increased margins on the group level in the last 2 years. And our commitment is to keep margins at least stable going forward. So I think this is a very strong commitment on a group level, and we guide on a group level.

May I follow up with product-specific one on AXELIOS. You noted it as a key part of your midterm sales target for a market that's not necessarily the largest one you're addressing on the diagnostics front. Could you give us a feeling of your market share expectations here? [indiscernible] you're going for, for number one. And secondly, on TB, just a clarification. On the U.S. launch here, should we think it comes along with your full ultra automation upgrade, around [ 2028 plus, ] or is prior to that possible as well?

So maybe I'll take the AXELIOS question, and I think Palani is ideally suited to address the TB IGRA. So I would say we've had -- and I also want to not get ahead of Josh's presentation, you've seen the data, but we've had really great response from the market and from customers. We've seen tremendous interest in the truly differentiated chemistry, which is a move away from sequencing by synthesis to sequencing by expansion, which is fundamentally different. And I think that's been really received a warm welcome from the market.

We're not going to give you our market share ambitions specifically, although we will talk about how we feel the reception from the market is, again, it's encapsulated in our divisional ambition, which is mid- to high single-digit growth, which again is to address Matt's question, we'll be addressing as we get through the China topic. But we do expect this to be an accelerant on our growth and we do expect to be able to capture significant market share and we're very confident, especially with this ecosystem of assets we're building around sequencing to offer full solutions, especially as sequencing becomes more entrenched in the clinical space. We feel we'll be meaningfully differentiated on the technology, but also on the application side and on the channel when you think about things like Foundation Medicine.

I think what we can say is what we had on the slide, it's a blockbuster potential. So that's 1 billion plus. We don't give you exact time line, but I think that's a substantial growth opportunity. And as Josh will talk about, I mean we already have the first orders. So you can see that the excitement is very high.

And I would add that those are orders from around the world. The interest is truly global for this solution.

Quickly on TB. Look, the clinical trials in the U.S. are coming along nicely. We're also having good conversations with the FDA. We see the performance of -- I mean we also have readouts ex U.S. as well, the positive person agreement, negative person agreement at parity with standard of care.

In terms of automation, there are 2 elements to it. One is the front-end automation and then there is the readout. On the readout side, from the get-go, we have a broad installed base of our Elecsys platform. Throughput-wise, super high throughput compared to anything that's in the market today. So 100 tests per hour with a single module of e 801, 400 tests per hour if you combine 4 modules.

Now on the front-end side, where you do the incubation of the tubes. Here, we will start out working with, of course, third-party liquid handler automators. And many of these labs already have those. So we'll make sure that our setup works with that. But we, of course, in our cobas ultra road map that I shared with you we will, of course, also try to bring in elegantly the front-end automation as well in the coming years.

So we feel very good about the automation from the get-go, good automation, bringing some important workflow improvements off the gate on the reading side. And then on the front end, we will, of course, work with third parties and build out in our own automation platform.

Okay. I think next question goes to Dan Brennan from Cowen.

Terrific. Just on maybe on AXELIOS, maybe a 2-parter since you did mention orders, I would just ask you, can you share what you think a successful launch would look like in terms of year 1 placements? And then just on the clinical opportunity there, how will Roche seek to leverage your global diagnostic leadership position to help AXELIOS penetrate diagnostic labs? Kind of where do you see the best opportunity there? And what does early interest look like?

Sure. So I would say for that, if you talk about our global network, we're present in over 100 countries. And what we've seen is some of these early orders, again, like I said, are coming from those countries because we already have that channel in place. And that really gives us a significant advantage in terms of infrastructure to drive the testing.

Now what would a successful first year look like, I think we would ideally want to place 100 of these systems in the first year with thereabouts. But I think what we've seen is a real tremendous positive reception from the market, and we're going to drive that.

Okay. Next question goes to, I think it's Graham, right? Yes, from Citi.

Graham Parry from Citi. So a question on Core Lab 5-year CAGR on Slide 38. That's now showing at 3%, it was 4% this time last year. Just to see if you can help us understand what's driving the change there, is that China or something else?

And then a question for Thomas on pharma margins. You guide to at least stable core operating profit margin. But can we dream a little bit more when we look at things like giredestrant, a wholly owned small molecule product with significant potential there? So what's holding you back from being more optimistic there?

And then thirdly, what -- with PathAI, how inhibited will be PathAI be with third-party contracts with other pharma companies post acquisition?

I'll take the first part. I think definitely, Thomas will take the second. And then maybe Andy. So you hit it. I think you answered your question. China is probably the second biggest global market for immunoassay testing and with the reduction in reimbursement and health care reform, that's dropped the size of that market, and that's the impact on CAGR.

Yes. So to your question, I mean, for the last 3 years, we always said the same. We said that keep margins at least stable. Now if I look back the last 3 years, we always grew margins faster than profit. So profit faster than sales or we increase margins. So I guess what I'm saying is future looking, we will say in the long term that in terms of guidance. If we change the guidance for that year like we did this year. This year, again, if we say we will grow core EPS faster than sales, then we will say that in the beginning of the year. That gives us, a, the flexibility to make the right investments that we need to make, and b, it also gives us the flexibility to be positive in terms of guidance and not already give out a guidance like that right now.

Sure. And on the pharma question, so Roche has a tremendous history with PhDS as well as FMI for building strong relationships with pharma and really driving value for pharma partners. So far, we've been operating as an independent company, but also building a strong portfolio of projects and partners within pharma. In 2024, we signed an AI-powered CDx agreement with VENTANA. So each jsite has tremendous experience supporting pharma, and I would expect that putting us together would even increase the ability of us to drive value with pharma partners.

Okay. Next question goes to Urban Fritsche from ZKB.

Yes. Question on Alzheimer's, pTau217. So how do you envision the implementation of the test in primary care? Will this become a routine test, which patients will do like annually with their health check? Or will it be more on when you suspect some symptoms already?

I mean I can take a -- give a quick answer. But of course, after the break, Olivier will go into more details. Right now, it's still indicated in symptomatic patients. So patients who are showing up with signs and symptoms of dementia. And typically, they show up in a primary care setting before they are then sent over to a secondary care setting. And this is where one of the key strengths of the data that later on, Olivier will be showing is we have shown very good performance for both rule in and rule out, not only in a secondary care setting, which is where many of the others are starting, but also in a primary care setting.

To your point, it's to have an annual kind of test, that's in a more preclinical asymptomatic population. That is not the case at this point in time. No one has a solution for that. That's a pretty tough problem, but that is something that we will absolutely continue to work towards. But right now, it's individuals who are coming in with signs and symptoms of dementia. And here using a simple blood test, we can triage them instead of going through a cerebrospinal fluid test or a PET-CT scan or so on.

Okay. Next question goes to David Westenberg from Piper.

So first, just for Matt, you talked about growing faster than the overall market. First of all, that's hard to do, given the size of the portfolio. Just kind of curious about how you're benchmarking that. There's life science tools growth rate. There's diagnostics growth rate. You're all over the entire diagnostic market. So can you just give us a flavor for that benchmark? Does that -- just one growth rate overall or just versus all the other ones?

And second one for Palani, I'd love to see all that innovation. Can you talk about maybe the trade-off between getting things perfect and getting things out on time. New innovation is always hard. There's always software, other kinds of things that you get in the hands of customers and they go, "Oh, you know what, here's a fix and whatnot." So just how you're thinking about that dynamic?

Yes. So starting with your question, the overall IVD market, I mean IVD, we're not talking about life science tools, is growing at a little bit above 3%. So if you say we're growing mid- to high single digit, that could be as much as double the market rate or 5%, at least [ 66% ] above the market rate. So I think we're significantly outgrowing the market. And I would say, yes, that is challenging, but it's something that we've done, and we will continue to do. But again, I'm referring to the clinical IVD marketplace, not the broader life science tools market.

It's a great question. This is something that we think about all the time. And I was just wondering, maybe I'll give you 2 examples how we try to balance this. So let's take the example of Alzheimer's disease. And here, there are some key assets where we feel we absolutely have to take the lead and generate the clinical evidence. And that's what we are doing with Phospho-Tau 217, Phospho-Tau 181 and so on and so forth.

But we also have the neuro toolkit where we have a range of early biomarkers that haven't been validated yet. These are research use only. And here, we make them available to many of our customers who then can use them to do their own research. They can use that to do research with their biopharma partners and so on and so forth.

So we are striking that balance on certain assets, we feel we absolutely have to take the lead and drive it because that's how we bring medical value. And in other cases, where there is not enough evidence, we work with our customers for them to allow them to generate the evidence.

The second example will be with AXELIOS as well. I mean there are so many different workflows that we can be looking at. And as we are having conversations with customers, they're coming up with their own ideas because now their gears are turning around a completely new chemistry and a new technology. So we will introduce many of the application notes and workflows, but there will be many more that customers will develop. We will hand it over to them. And we will, of course, support them with -- on the library prep side as well as on the bioinformatics side.

So that's how we are trying to strike the balance in many of these cases where we feel like, okay, certain things we have to own and drive to capture the value. And in other cases, we feel there will be other brilliant minds who will be able to essentially generate value.

Okay. Thank you very much. I think we have to close the first Q&A session now. There will be an opportunity to ask more questions in the second Q&A session after the second set of presentations. And for now, we are doing a break. And I would like to ask everybody to reconvene here at about 5 minutes to 3:00. So thank you very much for your attention so far.

[Break]

All right. Welcome back, everyone, from the break. I hope you managed to squeeze in some coffee and bio breaks. My name is Josh Lauer, and I'm the Global Head of the Molecular Labs Customer Area. Today, I'll be very excited to give you a brief overview of our strategic priorities in PCR. But then, of course, I'll spend most of my time expanding on our recent upcoming launch with AXELIOS.

As Palani mentioned, our strength and leadership in PCR really stems from our core position in higher-volume automated platforms that sit in the central lab. This is where our leadership position exists in infectious diseases and in blood screening. But that's only part of our broader PCR ecosystem. Our platform portfolio extends beyond these automated platforms and into syndromic testing, digital PCR and open molecular systems. This broad portfolio helps us cover the entire continuum of testing and enables us to deliver the right assays to the right platform for the right use case.

Speaking of those assays, we are continually investing to expand our menu across this portfolio with over 16 assays planned for launch over the next 3 years. These assays not only generate value for patients, they accrue increasing value to our lab customers who are able to consolidate testing on to single platforms over time.

To highlight a few assays. I would say this year, we will extend our leadership in the blood screening franchise with the launch of a multiplex assay that includes hepatitis E. We will also gain increasing share in the fastest-growing segment of sexually transmitted diseases with the launch of our BV/CV assay. And over the next couple of years, we will launch a variety of products, including completing our syndromic testing menu, completing our STI testing menu, launching panels and gastrointestinal diseases and respiratory and offering new solutions for self-sampling or home sampling, which can expand the market for PCR into decentralized settings.

Now one assay in particular that I'd like to dive a little deeper on is a key launch in the next couple of months will be our hepatitis D assay. Hepatitis D is an important disease that affects over 12 million people around the world. And unfortunately, the chronic forms of hepatitis D lead to liver cirrhosis and liver cancer, which are 2 horrible diseases. We're excited about this assay because it's the first fully automated test to diagnose hepatitis D infections via RNA. And it fits into a very clear diagnostic pathway.

We have 2 forms of testing for this that we envision. The first is, of course, existing guidelines to test patients who have chronic hepatitis B infections for hepatitis D. The second is a growing opportunity to monitor the effectiveness of new therapies for hepatitis D. And so we see this as a growing opportunity over time. We think it's a great example of extending our leadership position in virology to gain increasing share in PCR.

Of course, while we are excited to continue investing in PCR leadership, most of what we want to talk about today is how we're expanding into a new category of sequencing with the AXELIOS platform. As mentioned before, this is a generational technology, and we couldn't be more excited about the positive reception we're hearing from the market about this. We don't have the instruments here today. So we just have them on screens everywhere. If you weren't here last year, I hope you had the opportunity to watch some of the webinars or conferences where we've explained to the SBX technology in more detail. But at a high level, I would say there are 3 components of the system, each which highlight the differentiation of our architecture.

The first is the synthesis instrument, which turns a DNA molecule into an Xpandomer. This makes single molecule sequencing at scale possible for the first time because it improves the signal-to-noise ratio. The second instrument is the sequencing instrument, which runs those Xpandomers through an 8 million well reusable CMOS array, which provides incredible data rates. And the third is the analysis module, which, while we don't talk about it as much, is a critical component of this end-to-end architecture. It is essentially a high-performance compute stack, which is able to keep up with the data rates of the sensor module, processing all this sequencing data in real time.

Before I jump into more on the workflow, we've recently held workshops with hundreds of customers in the U.S. and Europe. And I'd love to show you a brief video to hear what they have to say about AXELIOS. Please queue.

[Presentation]

Okay. I love it. One of the themes you often hear customers talk about and you will hear us talk about is speed. That's not the only attribute of this platform, but it is a compelling attribute. So let's look a little bit at how that speed manifests in an end-to-end workflow just to orient us. I'll walk through some of the steps, but the first thing to be aware of is the speed is actually flexible because customers have the ability to load anywhere from 1 to 4 pools onto this instrument. They can tailor the throughput to what they need, and that unlocks economics for different classes of users. Mid throughput users can actually access the same economics as high throughput users. So the workflows themselves are flexible.

But if we start at the left-hand side, library prep is pretty similar to traditional methods. The important thing here is that we've already shown that you can automate this library prep, not only on Roche platforms but on other leading platforms from [ Beckman, Hamilton, ] et cetera. So this is pretty easy. It makes it impossible for users to walk away and not worry about the steps here.

Where the speed really comes in is when you look at the Xpandomer synthesis and the actual sequencing times. These turnaround times are simply much faster than traditional methods. And if you want a benchmark, you can say that if you wanted to load this thing up fully with 4 queued runs, you can generate 64 genomes in less than 2 days when traditional technologies take significantly more than 2 days, 2 days just for the onboard sequencing time alone.

Obviously, we have shown the ability to go much faster. We have amplification-free workflows, such as the one used in the world [indiscernible] for NICU testing to generate a full VCF file from a sample in less than 4 hours. So the platform is fast no matter how you run it. And as we mentioned before, the unsung hero of this is also the high-performance compute, keeping up with all this processing in real time. This is actually a unique aspect of the architecture as well because we don't have to wait for every flowing cycle of nucleotides to complete a read, our reads are read in real time, and we can immediately start processing those on the compute module so that by the time the sequencing run is done, the data processing is almost complete.

One thing that shocked the market a little bit when it comes to the flexibility of our architecture is the fact that we can reuse this sensor module up to 20 times. This is really impressive because it's part of how we unlock the flexible economics for users. It doesn't matter whether you're running or queuing one run because you only have a certain amount of genomes available or you want to queue up to 4 runs because you're a very high throughput lab. Both labs can access the same economics.

Now obviously, as a leader in clinical diagnostics, we're keenly aware of the importance of carryover whenever you're reusing a component like a sensor module. We've already done experiments that show carryover on the AXELIOS system is actually lower than traditional systems. You might wonder why that is. There are several reasons, but one obvious technical reason is that we're not sequencing DNA. DNA is a very robust molecule. And if it's stuck in fluidic systems, it's very easy for carryover to occur. We, on the other hand, are sequencing Xpandomers. These are more delicate molecules, and they're easily cleaned out via our system cleaning cycle. So carryover is not something we are worried about.

Now one of the top things I hear other than speed, speed, speed, is customers are really impressed that a new technology comes right out of the gate, matching or exceeding the accuracy and cost and performance of traditional platforms. That's a rare thing for a new technology to do. And a lot of that performance data that you're seeing is driven by the duplex workflows that we've described.

But one underappreciated attribute of the platform that we're just beginning to explore with the community, and you saw Michael [indiscernible] talk about that. You saw [ Graham Wiley ] talk about that is people are starting to think about new use cases, things they've never been able to do before. And this essentially is leveraging the fact that SBX has longer reads than traditional short-read platforms. If you want SBX style performance and accuracy, simply fold those in half and you get duplex for high intermolecular consensus quality. But if that's not your focus, what if you want longer reads, we can easily generate 500 to 1,000 base payers at a scale no other platform can do. This is useful for things like novel isoform detection. What if you only want cheap reads? We can tune the system down so that if you're only going for short reads, we can generate output that no one has ever seen before on a sequencing platform. These different use cases will unlock new types of applications, and we're only at the beginning of understanding what the community will do with this new technology.

Some stats that we showed at AGBT, which kind of help customers understand what to expect when we launch this summer. On duplex, you're looking at roughly a couple of terabases. This is sufficient to always generate 16 genomes or more at a very high quality, Q38 or higher. Note that those insert sizes are longer than the traditional [indiscernible] 150 as well. This is part of the quality that we're driving. It's not just Q scores. It's also read length from [indiscernible].

On the simplex side, it's just staggering the amount of data that can come here. On the short read side, over 40 billion reads in 4 hours. Just think about that. That's over 10 billion reads an hour. The sequencing world has never seen output like this. If you want the longer reads, you can still get over 10 billion reads in 4 hours, still unheard of.

These statistics are what we use to communicate pricing expectations to the market. We said that SBX and AXELIOS would be competitive with high throughput platforms, and we delivered that with a list price of $150 per genome. This doesn't require any volume-based discounting. Everyone can achieve this pricing. And while it sounds competitive with high throughput platforms, it's actually way more than competitive if you're running smaller throughput. Remember, lower throughput users don't get the headline number economics the other sequencing platforms enable. They have to pay a lot more per genome. And if they're able to access these economics, it's not just competitive, it's transformative for them.

We already talked about how simplex offers economics that we've just never seen before in sequencing, $0.06 per million reads is multiples lower than the nearest competing technology. And we also gave some examples of how you can turn these reads into different application economics. Now panels all vary, so it's hard to baseline. We can't just come up with dollars per genome or dollars per million reads. But we took a commonly used 1.1 megabase panel sequenced at normal depths to achieve kind of clinical grade accuracy that liquid biopsy requires. And we found that you could do that for $25 a sample. This is multiples lower than what you can achieve with other technologies.

Now that breeds certain questions. After AGBT, we got some questions from the analyst community asking, hold on, how does target enrichment really work. Previously, we only showed data on target enrichment using simplex, and there's a good reason for this. In liquid biopsy, the standard practice is to stack up multiple reads with UMIs, or unique molecular identifiers, all technologies form consensus for these needle in a haystack applications. Here, you want more than 2 reads. You want several reads. And so we used UMIs and our ability to generate cheap simplex reads cheaper than anyone else to show this.

This is also useful because the error profile of SBX is largely dominated by random errors. So as you stack up more reads, you can push beyond Q38. We've shown Q50, and we don't even know yet where that ceiling is, but we're very confident that these simplex reads can deliver the kind of accuracy liquid biopsy requires.

But this is not the only way to do target enrichment. It's just the only way we had presented so far. The answer to the question many of you have asked is, yes, it is possible to do target enrichment with duplex. But because of the way the molecular biology works, it will be a different workflow. Essentially, what this will be is a traditional workflow, which uses some PCR mediated steps at the end to extend [ out a hairpin ] and make a duplex construct. This will still enable high intramolecular consensus accuracy for applications where 2 reads are sufficient. So we'll be interested to see how the market adopts these for different clinical level accuracy workflows.

Stepping back, this is our first-gen RUO platform, and we are very honored that the community has been embracing this platform to demonstrate how it can work across all of the highest volume workflows in sequencing and including some new small ones that we hope will expand to be high volume.

For SBX-D, we've already talked about how high accuracy has been demonstrated in germline. In oncology, we've shown fresh frozen tissue. We've shown FFPE. And our partners at Foundation Medicine has shown MRD. This is all great. But we also showed the unique capabilities of duplex and the speed with the world record, as we mentioned before. We think this can be transformative in NICU settings where ultrafast typically means a few days, and the Broad and Boston Children showed that it's possible to do this same day.

I won't go through all of the other applications. But as I mentioned, simplex is -- there's a long tail of applications that we want to explore with the community. Some of the ones we all hear about, single cell, Perturb seq, [ CRISPR ] screens, that's all great. But there's a lot of things being spatial, proteomics, novel isoform detection that we look forward to expanding on in conferences over time.

A few examples in this vein that we showed at AGBT were the partners with proteomic vendors. Olink has been the largest vendor in this field. We were happy to collaborate with them and show the ability of this platform to generate the kind of data they expect. But this was also done with Pixelgen and Alamar. This is only the beginning of our engaging with the sequencing ecosystem.

So that's part of our strategy with this open platform. It's not only going to be Roche solutions. We showed data with Watchmaker and other vendors. We have a partnership with 10X. We will engage with the NGS ecosystem to make sure AXELIOS is usable across all of the solutions that are predominant in the market. We also see the potential for DNA to be a -- sequencing to be a technology beyond DNA and to power multiomics, not just [ methylation, ] but proteomics, spatial transcriptomics, et cetera. And because it's not only important for research, but for our longer-term ambitions for AXELIOS to be a clinical platform, we will focus on end-to-end workflows and bringing our expertise in automation, library prep, et cetera, into the AXELIOS ecosystem.

But we also said this is a generational technology. So it's not enough to just invest in it being an open platform. We intend to invest in the next 20 years of improvements on SBX in the same way traditional platforms have been optimizing over the last 20 years. I won't go into all of these in great detail, but suffice to say, we have the ability to scale this platform up and down, not only deal with short reads but extend read lengths. We intend to continue optimizing the chemistry for accuracy, to extend reuse of the sensor module and to proliferate AXELIOS into a wide, wide variety of applications.

You know, as a leader in clinical diagnostics that we've been investing in sequencing heavily because we believe sequencing will be a very significant platform in clinical diagnostics. It's already becoming one, but we think this is only the beginning of an ongoing trend. So we were very honored that our partners at Foundation Medicine looked at AXELIOS and demonstrated what it can do in one of the largest high accuracy applications out there in liquid biopsy, minimal residual disease.

This was presented at AACR, but essentially what it shows is that AXELIOS was actually able to detect more MRD positive cases than a competing technology, and it actually showed that it was better at doing this in earlier stage disease as well. As I mentioned earlier, Q38 is not a limit here. We're seeing Q50 and higher, and so our limits of detection should be very competitive in this space.

Obviously, MRD is not the only large liquid biopsy opportunity out there. We've already talked several times in this meeting about the partnership with Freenome, but we're very excited to work with them what AXELIOS can do with their multi-cancer early detection platform and to enable decentralized testing with AXELIOS. So these are examples of our strategic intention to make sure we're partnering with the leading labs around the world to make sure that AXELIOS has the content to be a leading clinical genomics platform over time.

With that, I think it's a great opportunity to hand it over to Daniel Malarek, the CEO of Foundation Medicine, and I look forward to answering your questions at the Q&A.

Thank you, Josh. I'm excited to be here to share how Foundation Medicine is going to continue shaping the oncology market moving forward. Now when we talk about oncology care, what we're essentially talking about is a sequence of often irreversible decisions. So what therapy to give when to change course, how aggressively to treat when to monitor and even how to monitor and went to intervene again. And every one of those decisions depends entirely on how quickly and how confidently we can turn complex biology into insight for oncologists.

And this is exactly where Foundation Medicine operates. And what excites me the most is that actually, we're only at the beginning of this year. So let's zoom out a bit. So Josh just ended speaking about screening and early detection, and I'm going to focus on what comes next. So those critical steps where molecular insights really guides therapy selection, therapy management, monitoring and ongoing care.

But let me start by grounding us in who is Foundation Medicine and what have we built? So Foundation Medicine was founded in 2010 out of the Broad Institute with a core belief that cancer is a disease of the genome. And if we can get those genomic insights into the heads of drug developers, if we can get those genomic insights into the hands of oncologists that ultimately, that will lead to better outcomes for patients. And that's exactly what has happened. And over the past 15 years, we've helped define the field of personalized medicine in oncology. So we were the first to launch a comprehensive genomic profiling assay, pioneers and linking genomic data with clinical outcomes at scale, leveraging advanced machine learning early on in genomics, and we continue to invest heavily in AI. We paved the regulatory path for companion diagnostics, all in parallel to paving the path to reimbursement.

Now these milestones didn't just matter for us. They actually shaped the entire industry. And in the process, on the clinical side, we've delivered more than 1.5 million comprehensive genomic profiles to patients. And one thing I want to highlight is our commercial reach. So we built a platform where, over the last 12 months, over 40% of oncologists in the United States have ordered from us. And that is via our portal and that is via our EMR integrations. And that will be a springboard for all our future launches. And that pioneering work has also resulted in extremely valuable data assets, AI solutions, and industry-leading number of publications.

Now on the biopharma side, we're fully embedded in the biopharma drug development ecosystem with more than 100 approved companion diagnostics, and that represents more than the entire rest of the industry combined. So when we talk about Foundation Medicine, what we're talking about is a scaled clinically embedded biopharma connected platform that has shown to create value across the entire oncology ecosystem.

So how does our business work? Well, our business model actually has 2 aspects to it. So the clinical business and the biopharma business. And in the clinical setting, it's our multimodal portfolio that supports decision-making across the entire patient journey, making us a one-stop shop for oncologists. And our customers range from major academic medical centers all the way to the community oncology practices. And what we consistently see is that adoption is still far from where it needs to be. For example, I was recently speaking with one of the largest community oncology practices in the United States, where they treat around 20% of all cancer patients. And they said that in the metastatic setting, only around 50% of their patients are getting a comprehensive genomic profile, that is unacceptable. We agreed that, that number needs to be north of 90%. So there's huge opportunity there.

On the biopharma side, we work with the top 80 biopharma companies and are currently supporting over 300 active clinical trials to date. And our partners not only value our scientific capabilities, but also our regulatory strength. And what do I mean by regulatory strength? Well, it's our ability to execute complex companion diagnostic programs aligned with regulators and ensure that their launches and commercialization goes smoothly. And importantly, these 2 businesses actually reinforce each other.

So insights from clinical practice, feed in to drug development and advances in drug development then, in turn, feed back into clinical care. And like I said, we're still at the beginning of this journey. So if we look at the market, the oncology testing market is one of the fastest-growing markets in diagnostics. If we look across biopharma, clinical therapy selection and clinical monitoring, we expect the market to essentially double over the next 5 years. And this is due to 3 factors. So the first one is we are seeing an explosion of new targeted therapies and actually new targeted modalities being developed that require genomic insights. So everything from bispecifics, trispecifics, ADCs, cyclic peptides, even molecular glues, individual neoantigen therapies. All of these require those genomic insights. So complexity is exploding for oncologists.

The second reason is CGP adoption is going up. especially as these new therapies are being launched that require genomic insights. And then the third is around molecular monitoring. So as clinical evidence builds showing that molecular monitoring is superior to traditional pet and CT scan monitoring, that market is expected to grow rapidly. And there are approximately 50 million people currently living with cancer, and that number is expected to almost double in the next 10 years.

And for any cancer survivor, there are 2 questions that dominate. The first one is, is my therapy working? The second one is when might my cancer return? And being able to answer those questions by a simple blood test has profound implications for patients and for health care systems. So overall, we're moving from the narrow question of what therapy do I give my patient to how do I manage my patient over time? And this is central to where Foundation Medicine is going.

So 3 core differentiators position Foundation Medicine for the next era of growth. Quality, innovation across the patient journey and being Roche powered. And I'm going to now double-click into each of these.

I'm going to start with quality. I'm going to say that not all cancer tests are made equal. So if assays are not engineered to detect and capture difficult alterations, patients will be missed. -- copy number loss is an example of a difficult to detect alteration. And a powerful example that I would like to use is a copy number loss called MTAP law.

So why is MTAP loss so important? Well, it is actually present in around 10% of all cancers and is a critical biomarker in trials that are evaluating MAT2A and PRMT5 inhibitors, which are showing great promise. And in a recent publication by Sarah Cannon Research Institute, when they looked at their own real-world data, comparing 3 major testing laboratories. They actually found that Foundation Medicine detected MTAP loss at rates consistent with published incidents.

The other labs showed detection rates between 40% to 60% lower. Now that difference matters. So missed calls mean miss trial enrollment, delayed drug development and patients who never received potentially life-changing therapies. And our biopharma partners have also recognized this material difference, and that's why we are the clinical trial enrollment assay currently for 15 different MTAP trials. And with another copy number loss, [ T10 ] Loss, FoundationOne CDx is the only FDA-approved tissue NGS test validated to detect a -- copy number loss. This is why quality matters. It matters for patients. And this is why we believe that every patient doesn't just deserve a CGP test, they deserve a Foundation Medicine CGP test.

The second differentiator is innovation. So we're -- we talked about how molecular monitoring is rapidly expanding. So let me talk about 2 exciting new products that we have. So first, let me talk about F1 monitor. So F1 monitor is a liquid only, so no tissue needed assay used in the metastatic setting, to answer a simple question, is this therapy working? And in studies, we detected progression at roughly half the lead time of traditional CT PET scan imaging.

So that means earlier decisions. That means whether to stop an ineffective therapy or it also means that confidence that the therapy is working. And another differentiator of F1 monitor relative to other common tests is that F1 monitor also provides resistance mutations. So if a therapy is not working, the next question the oncologist will have is why isn't it working and F1 monitor also provides that information.

Now to Pathlight MRD, obviously, you've heard around our acquisition of Saga. And once that closes, we're excited to welcome the Pathlight MRD assay into our portfolio. I'm excited about this. I'm excited about it for a number of reasons. So from a clinical perspective, patient impact perspective. Pathlight has just demonstrated the ability to detect recurrence over 13 months before traditional CT PET scan imaging in early-stage breast. That's the first reason.

The second reason, Pathlight is already Medicare approved. So if Medicare covered for early-stage breast, Stage 2 and 3 across all subtypes over 6 years of a patient's journey and 12 time points.

The third reason, it's a best-in-class assay. By best-in-class, what I mean ultrasensitivity, and disruptive turnaround time. And then just as exciting as the assay itself is our plans to decentralize testing. We have the [ Actelios ] platform where we can do the whole genome sequencing tissue baseline, and we have our digital life cycler platform from which we can do the multi-time point testing. And by decentralizing using our own platforms, we can provide access to patients in all settings. So with this portfolio, Foundation Medicine is exceptionally well positioned to scale and become a leader in monitoring.

So the second thing I wanted to talk about on an innovation standpoint is as our biopharma partners develop increasingly complex therapies, we're expanding how we interrogate disease biology. So we have built a platform to understand disease of operationalize insights at scale. This is our [ Flexomic ] platform. And essentially, just how we were integral in turning oncology into the poster child of precision medicine. Our aim is to take those same capabilities and apply it to other disease areas, neurodegenerative, metabolic and autoimmune diseases.

And this is what our Flexomic lab will bring. So for biopharma, this means better target discovery, better patient selection and also derisked development. Now innovation alone cannot make an impact. you have to have delivery at scale. And this is what being powered by Roche actually means.

So we operate actually 3 different models. So the first model is our centralized testing model, where samples come from over 50 countries into our labs based in the U.S. and based in Germany. And Outside the U.S., it's Roche Diagnostics that are commercializing our products. And in Japan, it's true guide.

The second model that we operate as our installation and services model. So this is where we essentially build our lab at our partner sites. And they run the wet lab workflow, the data goes into the cloud. We collect the data. We run it through our analysis pipeline, we generate the report, which essentially means they get FMI quality in their own lab they get all the latest complex biomarkers that we run all in their lab. We have INS sites in the U.K., where we won 50% of the NHS liquid biopsy, tender. We have at the [ Gustav Rose ] Institute in Paris, where they become a hub for liquid biopsy testing in France. And we also have at the University of Zurich Hospital.

And then the last model that we operate is our completely decentralized model, which is based on the Roche AVENIO kits. And these kids have been codeveloped with us, and they run and are powered by our analysis pipeline. And we believe that the future will be always a combination of centralized and decentralized testing. And so our focus is to hit our solutions, which enables us to win overall.

And with these models, it allows us to scale globally while also not compromising on quality. Now being part of Roche goes beyond just the scalability. We are, as you heard, we're an early access site for the AXELIOS sequencer and implementing the AXELIOS sequencer in our clinical workflows has the potential to shorten our turnaround time and improve our cost basis. And over time, once we embed that in our clinical workflows, our plan is also to onboard our installation and services sites, the sites I just mentioned in the U.K., in France and so forth.

So it's exciting to see how Foundation Medicine and Roche together can accelerate the adoption of the AXELIOS platform in the clinical sequencing market. So secondly, we also leverage VENTANA IHC to aid in treatment selection, and we're going to continue expanding our portfolio to be the one-stop shop for multimodal insights.

And the third area and partnership is with our navify colleagues, where I'm very excited to be launching navify clinical hub decision support applications within our Foundation Medicine portal. We talked about the community oncologists. They're treating 20 to 30 patients every single day. They need to be making decisions and have all the insights at their fingertips. And with the applications around clinical trial matching, dynamic NCCN guidelines, which essentially allows them to see where their patient is in the decision-making tree dynamic publication search and AI patient education, we actually make it easy for them. We bring those insights straight to them so that they can make the right decision for their patients. This is how precision medicine becomes routine practice.

So our product road map reflects a deliberate strategy from single test to platforms from snapshots to longitudinal insight from information to decision support. Each initiative essentially builds on assets that we already have, data, scale, regulatory credibility and extends us into new value pools. And that is our approach to disciplined growth.

So in summary, let me close with 3 points. The oncology testing market is poised for strong growth. Foundation Medicine is evolving from helping decide what therapies to give to helping manage patients over time. And together with Roche, we have an unmatched ability to scale. And Foundation Medicine is going to continue shaping the market by turning complex biology into better decisions and better decisions mean better outcomes for patients, better outcomes for health care systems. Thank you very much.

Thank you, Dan. Good afternoon, everyone. I'm Laura Apitz. I'm the Head of Pathology Lab, and I'm going to be sharing with you the most impactful developments that we have going on in pathology. So Helane already mentioned our 3 focus areas is all about automating our core business. It's about digitization, and it's about growing our high medical value menu.

So we have -- we're a strong leader, #1 leader in pathology lab. And you can see we've had double-digit growth, 14%. It's driven by our advanced gaming business, our personalized health care business, as well as our IT and IS systems are over 13,000 installed base globally. I want to call your attention to the innovation section of the slide because I'm just going to be going through our next-generation systems that we have in development a little bit more in our personalized health care business, digital pathology and how that's helping there. And then also, we cover the cervical cancer disease state in pathology as a holistic disease date.

So first, our vision for transforming the pathology lab is really about bringing end-to-end automation to the lab. Pathology is not very automated today, and it really needs a way to integrate patient workflows and make the lab more efficient. We're going to be doing this with our partners, Hitachi, high-tech to bring out our next-generation advanced staining system.

Also, within the next 18 months, we'll bring out task targeted automation with some AI-driven tools that will help labs go to digitalization overall. And so these are some of the ways we'll be bringing more automation and digitization to the lab. When I turn to our multiplexing portfolio, we've been doing multiplexing for some time. You can see that in our pipeline here. But what we're going to bring out uniquely now next year starting with some duplexes is our new [ transluciccroba ] genes. This allows easy access to doing multiplexing.

And the 2 I want to highlight is P40 and TTF1. This is in the lung cancer space, where we'll differentiate non-small cell lung cancer. And then one of them is SOX-10 and KI-67 where you can tell between a melanoma and a bone 9 lymph node. So this is just the first of many multiplexes that will bring out in the companion diagnostics space in the companion digital pathology space as well as some non-going cytology space as well.

So turning to a little bit more to our personalized health care business. This is a growing business. We've been in it for 25 years. We really have a strong footprint here to grow even more. What's growing even more important is these computational pathology algorithms. And you can see we have over projects in our pipeline for these computational algorithms. And this will be significantly enhanced. You've heard a lot today about Path AI. This will really help us grow this area of digital pathology even more.

But I want to call your attention to our TROP-2 franchise, and it really is going to be a franchise. But it will be the very first computational pathology companion diagnostic that will launch in the lung space next year. And so right now, with the RUO, TROP-2, we're really growing that footprint globally. So it sets us up really nicely for those companion diagnostics algorithms. So digital pathology, I'm hugely excited that we have a complete end-to-end offering here.

So you've heard about the best-in-class image management system that we're going to be bringing into the fold pathology centric collaboration is really what this is all about with some AI-driven tools as well. We also have our DP 200 and 600 scanners and we're going to be bringing out improvements next year to these scanners, faster scan speed as well as improving cytology workflows as well. So you can see we're really rounding out our digital pathology platforms to be able to fuel this growth in the future.

Now I'm going to turn to our cervical cancer solutions. So we're on a mission to eliminate cervical cancer. And you know that HPV infection causes most cervical cancers. So it's really about getting access to the important testing. So the first test I want to highlight is our HPV genotyping test. Detecting 14 individual high-risk genotypes. And this is important for screening programs around the world so they can really customize in an evidence-based way how they want to manage their patient populations all over the world.

And with our large installed base of x 800 molecular platforms, we're really able to bring this test everywhere, and we plan to launch this in CE Mark next year. The other area of serval cancer I want to highlight is really our self-collection. So again, it's all about accessing people for testing and getting closer and closer to their homes, so we can get those samples. And so we're going to do that for HPV. You can see our kit design here that we're going to launch next year.

But what's also really unique is we're adding our FCI testing. So all in one, will have HPV, STI testing, and actually a really seamless workflow to bring it into the laboratories for testing with not a lot of manipulation needed. So we're really excited about both the genotyping and, of course, the self-collection to bring access to this important type of testing. So with that, you see we have a really strong pipeline in pathology, still growing growing stronger. And what we're super passionate about is making sure we're reaching beyond the 44 million patients that we reached last year.

So thank you very much for your time. I'm going to hand it over to [ Elliot ] to talk about near patient testing.

Yes. Thank you very much, Laura. It's always impressive to see the many patients that are positively impacted by our pathology solution. A warm welcome also from my side. My name is [ Eriko Amanda ]. I'm heading the Research and Development Department in RDS, and I'm here today to talk about the highlights in near patient care, our updates of our products and any upcoming launch.

As Palani already said, we are up, and we want to establish new patient care as a leader in this space, and we have 3 priorities to do that. One is building a leading portfolio in clinical chemistry and in immunoassays. We want to strengthen our position in molecular point of care, through systematic menu expansion and we want to deliver a competitive CGM solution. You have heard about that a couple of times.

I would like to dig deeper into our strategy and how we want to achieve that. And first of all, I want to concentrate on the disease areas. We are convinced that testing in proximity of patients that are affected by infectious disease cardiometabolic or diabetes will improve the outcome of these patients. And this is why we do point of care in these 3 prioritized disease areas.

Now we want to serve these patients across the whole spectrum of decentralized setting. And as you appreciate from this slide, these settings are very diverse. We have emergency room and intensive care settings. We have GP settings, we have pharmacies, and we have consumer patients. to understanding the needs in these different settings is of essential importance and we will provide tailor-made solutions for each of these settings in order to become a leader in near patient care.

For my presentation, I would like to start on the very right side on the consumer and patient area and then work backwards to the emergency rule. You have already heard about Accu-Check SmartGuide, our CGM solution a couple of times. In the last year, we have made substantial progress with our controlled launch strategy. We are now available in 45 markets.

Last time when we spoke, we were around 4 to 5 markets. This year, we are available in 45 markets. We have launched across EU, Middle East, APAC and Latin America. In the first half of this year, we have also achieved very important milestones. We continue to ramp up our manufacturing facilities. And we had a full rollout of my sugar as part of the Accu-Check SmartGuide CGM solution. I will talk about both of these points in a minute.

In the next half of the year, we aim for regulatory approval for factory calibration to avoid the first finger prick and first use, and we will broaden our geographical market deployment. Now it's very important that the foundation of our market penetration are our 11th state-of-the-art manufacturing facilities that we have built up over the last couple of months and years.

And these are highly automated, state-of-the-art manufacturing centers. We have built here on our long-lasting know-how that we have in diabetes care for more than 30, 40 years we are in this field. And in order to give you an impression how these manufacturing lines are built up and what they do every day for us, I would like to show you a short video.

[Presentation]

Now with this, we really want to highlight that we are focused on investing in a solid base for our CGN solution and making it available globally. The other highlight and milestone we have achieved over the last couple of months was really the combination of the Accu-Check SmartGuide predict the app, where we have AI-enabled algorithm, for example, for the 2-hour glucose forecasting or the night low prediction, together with our MySugar ecosystem where we have a download of more than 7 million users.

And this combination really provides a scalable digital ecosystem for our diabetes customers. We can, with that accelerate PGM to CGM conversion, and we will also improve clinical outcome. For the future, this means that we integrate data and insights from multiple smart devices and health platforms.

Now moving over to our customers. Up to now, we have focused very much on the multiple daily injection patients. But as a matter of fact, Diabetes Care in Roche has a legacy. We have already more than 30 million BGM users that use our blood glucose monitoring platform. 11 million of them in insulin users and 20 million of them in non-insulin users. And our strategy will be to upgrade those Roche customer base from BGM to CGM by leveraging reimbursement availability. We will also partner with third-party systems and integrate their devices.

For the non-insulin users, we will offer intermittent -- of BGM and CGM via MySugar together and learn user insight. So this is very important because most of the customers that today are not yet on England, but earlier or later end up as an insulin users. And to give them already the opportunity to use the CGM early on will help them to learn how their exercise, how diet is impacting their blood sugars.

And last but not least, we also connect our portfolio with Roche Pharma. And this is really also a nice synergy between the 2 divisions. Now as you already heard, we are constantly improving our device, and we are also listening to our customers. And we have also an evolution in our pipeline from a first generation to next generation. As you can see here, the next generation, the inserter will be much smaller. We will exchange the battery. We will have also less manufacturing step.

But most importantly, if you look at the current size of the CGM solution today, it is like that. And the new one will be like that. And I hope you appreciate the difference in the size that all will actually improve the varying comfort for our customers.

Now with that, I can say that all over, we are absolutely on our way to provide our patients, our strategic intent of a very competitive solution. Now moving on to our next segment, and that is the GP setting. And you heard from Thomas that primary gets more and more important because more of the health care will be decentralized.

And here, in 2024, we made a very important acquisition with LumiraDX. And that platform is really remarkable. We have shown it last year, it's a portable platform that has less than 1 kilogram. And the technology as such, provides multimodality. So you can measure immuno-asset clinical chemistry as well as [ cabilation ] assets and potentially in the future, also molecular assays. It provides multiplexing capabilities. Those 3 assays can be done at the same time. And it also, in addition, comes at room temperature. As you can see, I like to wear this parts here in my pocket because it's always nice to have point of care test with you. It also has finger-pricking capability. So you only need 1 drop of blood for this platform.

In fact, last but not least, is really tailor-made for the GP setting. So since we acquired LumiraDx, we have made commercial available this platform in more than 43 countries, and we have approval in more than 75 countries.

I also want to highlight a couple of commercial wins that we have, for example, in -- and whereas we have a pilot currently running with CRP testing, to fight against AMR, and that is government backed. We have a pharma partnership in India where we test anti-proBNP for cardiac risk. And here, the main differentiator was really the finger pricking capability for anti-proBNP on this device. So really a real platform. We are also investing heavily in alignment with our disease area strategy, we will make it IVD compliant, and we also have an update for our HVAC.

And for 2028 onwards, we will add further tests like the lipid panel and also further tests in infectious disease and inflammation. Moving on to the next big segment. This is a molecular point-of-care segment, which has a market size of EUR 2.65 million. We are participating in this market segment with our [ Pubali ] platform, which is well established and which provides lab-like PCR quality at the point of care.

As you can see here, we have already a pretty broad spectrum and menu there. We are adding now it's actually transmitted disease sales like we have last year with CTMG or the -- panel. What I would like to point out that we address with these platforms, always gaps that are existing in care where this rapid diagnosis is essential. For example, we will now or have already launched the water deli test which is super important for newborns for small babies and elderly.

And here, the decision is an antibiotic is necessary or not is a very important one and needs to be taken rapidly. And again, the lead platform helps in this regard. And talking about lab-like performance, I'm super pleased and happy and privileged to announced the launch of the Cobas Sense instrument and the high-sensitive troponin T in 2027.

We have talked about this platform a couple of times. Today, I would like to provide you the correlation between the Cobas Sense high sensitive troponin T essay and our fifth generation. This is the newest generation of our KOALA platform, the Alexis platform.

And as you will appreciate, this is an almost perfect curve. There is a correlation curve of 0.995 between these 2 assays. This means it's really lead like performance. And it's not only true that this is across the whole curve, but it's, in particular, in this very low range where you need this high sensitivity in order to make this important decision if a patient has a [ myocalli infection ] or not.

Now this supports us energization and enables interoperability between lab and point of care. It also overcome the barriers that we have in the emergency room because up to now, it was always perceived that point of care is not sensitive enough. Now with that, we can help our customers and our positions to reduce the amount of patients in the emergency room because the decisions of if a patient has [ mytocyinfaction ] or not can be done with certainty.

We will also add additional assays additional cardiac assets on the Cobas Sense like NT-proBNP and -- for patients that present in the emergency room with shortness of press or test pain. So a very nice and comprehensive platform that we will bring on the market in 2027 with lab like performance and talking about lab-like performance, it's my pleasure to introduce Olivier to you who will talk about the exciting portfolio that we have in KOALA. Thank you very much for your attention

Yes. I think it's indeed great to see that like performance at the point of care. So my name is Olivier Gillieron. I'm the life cycle leader -- and Neurology, and I'm pleased to be here to provide the overview of the call up assay highlights. In CoLab, our goal is to continuously expand our menu across our 4 key disease states, which are oncology, neurology, cardiotabolic and infectious releases.

Within 24 months, we have actually been able to deliver 10 new solutions to the market, which is peaking literally in this quarter with the introduction of the NFL has post 17 test in neurology as well as our aged solution in infectious diseases. I will focus on these 3 solutions in my presentation today.

So let's start with infectious diseases. As mentioned by Paul earlier, our goal is really to provide a high-quality comprehensive testing menu in infectious diseases. End of last year, we launched our antigen tab, followed by the IgM test earlier this year and the IgG coming later this year.

In 2027, we plan to launch the Lyme IgG and IgM assays with more menu coming further down the road, as you can see on the right side. Now the highlights for this year is indeed our IGRA TB assay. And before I get into specifics, let me give you a bit of disease context.

Tubecolosis is the dealers infectious disease around the world. As you can see on the left side, there are more than 10 million new active cases globally in 2024. At the same time, if we move to the right, up to 25% of the global population may have a latent to the closest infection. This is why it is important to [ test folate ] in TB, not only in high prevalence countries but especially in high-risk populations. And this is where both the tubercolosis skin test and the interferon gamma release assay blood-based are established in clinical routine.

As of today, there are approximately 13 million tests of IGRA TB, which are run on an annual basis globally. Now I'm excited to share that next month, we plan to launch our Alexis Aggregate assay in -- acceptance CE mark. We have validated the test in close to 2,000 patients and we plan to present the results at the ESM Congress in Verona next month.

As there is no gold standard in aggressor testing, we compared our solution to the current standard of care. And I'm happy to share that we have reached our end point and we have a strong performance, which is on par with the standard of care, and you see the PPA of 91.1% and the NPA of 94.6%. We will offer our IGRA TB test on all our second-generation electric platforms as Palani elaborated on it, if you look at our -- which is our high throughput platform, this will enable a throughput of up to 400 IGRA TB samples per hour.

The pets will also be available under [ E42 ] and on the Smart E, which is our solution from our decentralized settings appended by Palani earlier today. We do plan to further differentiate our solution by automating the prenatal steps, and we'll do this on our next-generation Cobas Ultra automation platform.

So with that, we're in a strong position to scale testing into the markets. As you can see on the left, we can leverage more than 48,000 electric instruments as well as our direct presence in more than 150 countries to drive access to latent TB testing. As you can see on the right, the market will evolve into a highly attractive above [ CHF 1 billion ] segment by 2034. And our goal is clearly to capture a significant share and become a leader in IGRA TB testing.

Now let's shift gears and move into neurology. Neurological conditions constitute a significant health and socioeconomic challenge around the world. In Alzheimer's disease, 75% of the patients are undiagnosed. It takes 3.5 years to get a formal diagnosis of the disease. At the same time, 98% of the patients are untreated. If we move into multiple sclerosis, 40% of patients are untreated and [ 35% ] are on low-efficacy treatments despite the availability of highly potent anti-CD20 treatment. So clearly, there is a significant opportunity to leverage diagnostic solutions to shape here pathways and drive access to these novel disease-modifying therapies.

So let me start with multicrosis. I'm happy to share that we have obtained CE Mark for our Alexis neurofilament light test, reflects neuroinflammation in patients with relapsing remitting multiple sclerosis. It is important to highlight that we have established specific NFL healthy reference ranges to support clinical interpretation as NFL values increase with age. In combination with MRI and other clinical data, NFL supports clinicians to monitor disease activity and guide treatment decisions. towards high efficacy treatments, such as [ Ocrevus and fenebrutinib ] moving forward.

Now moving to Alzheimer's disease. Our goal is to drive the early detection of amyloid pathology and therefore, again, access to treatment. In the meantime, we have a total of 7 diagnostic solutions along the patient journey. We have our success ratios, which are used for the -- diagnosis of amyloid pathology.

Last year, we introduced our [ Phosphate ] 11 solution, which is the first blood-based biomarker specifically approved in the primary care setting, which is ruling out amyloid pathology. And early this year, we introduced our up report test, which is a test which is able to discriminate between [ ApoE4 ] carriers and noncarriers.

Now the highlight is clearly our phosphate 17 solution as we have shared earlier today, we have obtained the CE mark, and I will speak to it on our next slide. We are also further expanding our solutions elongation journey. We're looking at digital -- streaming solutions on left side of the patient journey. At the role of pTau217 also in the positively unimpaired population and we are also developing new biomarkers for other pathologies and disease mechanisms as well as treatment monitoring.

Now let's speak about pTau217. I'm really excited about the launch of this test. It has the potential to redefine the standard of care. We have validated the solution in a real-world cohort spanning across care settings, so both primary care and secondary care, but also different clinical stages. Now let me guide you through the results on the right side. So in the middle, you can see that illustration. You can see that we have established 2 cutoffs, and it's important to highlight these cutoffs are consistent and valid in both primary care and secondary care. The 2 -- classified patients into positive, high likely -- pathology, negative, low like use of amyloid pathology and intermediate where patients should be considered for further testing.

Now let's double-click on the performance. I'm really excited about the sensitivities and specificities in both secondary care and primary care. They're all above 93%. If we look at the performance in secondary care, you can see that both NPV and PPV are above 90%. And moving into primary care, the NPV is at 98.6%, and the PPV despite low prevalence is at a remarkable close to 80%. And finally, looking at the patients in the intermediate zone, we have less than 20% in basically that specific zone.

So with this performance, we're actually taking off all performance criteria of the COO guidelines for -- markers, both in secondary care and in primary care. And with that, we really believe that we have a robust best-in-class [ photocatventin ] assay, which will enable clinicians to rapidly unsafely triage and diagnose patients with signs and symptoms of cognitive decline.

Now with the addition of NFL and pTau217, we have the broadest on-market IBD portfolio. We're also further expanding our early pipeline by closely collaborating with academic centers and pharmaceutical companies and by leveraging our newer toolkits. Our goal is to add 5 to 10 new robust assays or researches only tests in specific diseases where we focus on, which are Alzheimer's disease, Parkinson's disease and multiple sclerosis. So let me close with our ambition.

On one hand, we continue to drive the early biomarker innovation. At the same time, we continue to collaborate closely with pharma companies to drive that innovation. But then we're really ready now to scale these tests into the geographies. We have the broadest menu on market, we can leverage our large installed base. And as you can see on the right side, specifically for the Alzheimer's disease market, this market is expected to quadruple in the next 10 years. And our clear ambition is to be the global leader in this field shaping their pathways and making an impact on patients, families and society.

So thanks a lot for your attention. And yes, with this, I would like to hand over to Birgit for the second Q&A and my colleagues on stage.

Okay. Thank you very much, Olivier, and I would like to ask the speakers from the second set of presentations to join me here on stage for the Q&A, including an Palani and Matt.

Okay. Happy to take your questions now. First question goes to Richard Vosser from JPMorgan.

One question on the adoption of molecular testing. There was a about a week or 2 ago, which was trying to adopt earlier molecular testing rather than radiographic progression testing and the regulators pretty much didn't like it. I didn't have enough clinical data of the proof. So pharma companies are probably not going to do the trials that the regulators seem to want. So how do you get adoption of earlier testing, particularly in the metastatic space. So just your thoughts there. And then secondly, on -- testing. What evidence have you seen that the integration of the QTS that are able to differentiate between patients allowing better treatment algorithms, just thoughts there.

I can start on the first one. So the importance of designing your studies in a way that will resonate also with regulators is incredibly important. And I think that's one aspect of it, where we feel that it was a step back in a specific case, but not a setback overall. And molecular guided MRD testing is something that will continue to evolve and that evidence will continue to build.

I think the second point is that pharma companies are also looking for surrogate endpoints to their studies. And this is where ultrasensitive MRD assays will provide that for them. So that evidence is going to build with time. We just have to also work very closely with our pharma partners in developing the advance.

Then I can take the TROP2. Thanks for the question. I think it's very similar to what Dan just said. I mean, evidence is going to build on how these computational pathology algorithms can really help differentiate and detect therapy for TROP2. So obviously, there's evidence, early evidence that it can do that, but we're really waiting for next year, and those important for readouts to see what the true impact is going to be.

Okay. Good. The question goes to Subbu Nambi from Guggenheim I actually have 2 questions on foundation. How has the growth profile of Foundation Medicine evolved relative to peers over the past few years, especially in therapy selection. And it seems like you were lagging for a few years for do you agree? And is that improving as your focus and strategy have evolved. Second, how important is moving foundation content to AXELIOS pursuant to generating above-average growth relative to peers?

Okay. So on the first question, even today, when talking to clinicians, when talking to pharma companies, we are viewed as the best-in-class and that's without doubt. You can go to any ASCO meeting, you can go to ESMO meeting, you can talk to our biopartners, and that's something that we're confident in and they see the quality that we provide. So there, from that perspective, that quality differentiator still stands. I think the second aspect, which is incredibly important is our ability to scale. And that is something that we have unprecedented reach with the Roche partnership. So our reputation has grown, especially outside the U.S. and that's something that we're extremely happy with.

The second question was around, correct me from AXELIOS and inclusion in our clinical workflows. This is something that we're extremely excited about because the potential that AXELIOS provides us is better turnaround times and a lower cost base. And with that, we also feel that the capabilities of the system will also enhance our offerings to our biopharma partners and enable us to further differentiate not only the competition, but also further enhance our own internal developments and what we're able to provide in terms of services for our biopharma partners. So overall, we're very excited about the platform.

Okay. Next question goes to David Evans from Kepler Cheuvreux.

Yes. So maybe for Matt, but just of all of the product launches and innovations and assays we've heard about, which ones do you hope may have the biggest, most visible impact on Rpieoche agnostic sales growth. I mean I assume that AXELIOS would be the biggest single one, but is that like quite a distance. Is there any uptick likely from Gen 2 of Accu-Check or next wave of mass spec launches or equally assays like the Alzheimer's diagnostics seems to be a very big opportunity.

Yes. So I would frame again, back to our financial ambition to mid- to high single digit. I think in order to the kind of growth we expect, which is well above market. We need all of our launches to contribute. So I think they're all extremely exciting. And I think each of them, if I'm being super direct has an extremely, I'd say, exciting aspects to them. This Axelios is really going to be the core of that backbone that we use to drive oncology and it also gives us an opportunity again to really engage in the translational space. So that's exciting. The mass back again, is the first fully automated, fully standardized with an IVD menu mass spec system and we see there is an opportunity to grow our biggest business unit, which is the core lab.

And then if you look at CGM, we had for several years decline in our blood glucose monitoring business. Now we have an opportunity to really revitalize that franchise, bring in a state-of-the-art technology, combine it with our AI capabilities and return that to growth. So all of those together I think, are going to be what drives -- so it would be impossible for me to pick one from that group. But thank you for the question.

Next question goes to Tycho Peterson from Jefferies.

Maybe just a follow-up on the content strategy for AXELIOS. Saga had their own road map, Curious how you're thinking about some of their other indications, lung, head and neck, ovarian, some items I've talked about in their pipeline. And then maybe timing on Freno, converting over in your view? And then lastly, how are you thinking about the willingness of the community hospitals to embrace sequencing in-house? I mean, is this mainly for the -- medical centers for the next few years? Or -- when do you see that transition on the community side?

From the Saga perspective, so they have also launched their -- so we have both indications, breast CRC right now. They have studies in ovarian, and we will further invest in driving clinical evidence in other indications. We're immediately looking into lung and so forth. More than that, I'm not going to comment because the deal hasn't closed, so we need to wait for it to close and then we will put a robust clinical development strategy in place for the Saga acquisition. We're really excited about the capabilities of the assay.

Okay. So I'll just say for Freenome, that's an easy one. As we publicly disclose, they're going to evaluate the AXELIOS sequencer. And so then how that's going to be incorporated in their workflows is going to depend on how that shapes out.

Maybe then the third part, you asked about community settings. Here, I think we have some surrogate for this. Obviously, Foundation Medicine has a lens into serving these different populations. But there is, as Dan described, a lot of demand for some centers to want to bring testing in-house, and that can be for a variety of reasons.

But often, this reimbursement and the ability of some of these community oncology practices to tap into that. The real question is, do you deliver a solution that's capable and economic for them to run and capitalize on that opportunity. We don't see any reason why we can't do that. So sure, AMCs and researchers might drive early momentum, but our view is, over time, as Dan said, these can be both centralized and decentralized solutions.

Okay. Next question goes to Kyle Mikson from Canaccord.

A couple of questions. So on the -- some diagnostic growth drivers in Alzheimer's and neurology. Any interest or appetite to combine markers here like pTau217 and NFL or maybe even more. And the MSD, you kind of referenced this 35% CAGR. Is that international as well as U.S. It's obviously different markets. We have trial kind of read out recently. And then just separately, just taking us back your thoughts on profitability in the NGS AXELIOS kind of segment Obviously, if there is operating profit dilution, it's a smaller segment, not going to really affect the full company, but you could -- pricing headwinds in that industry. How do you think about going the $1 billion? And if it's a blockbuster product, could you -- that building the revenue on $1 billion in expenses? Or could it take more?

So we'll start with the neurology questions. So first of all, I mean, as of today, the stand-alone or pTau217 performs best. But I think the field will indeed move into an approach where multiple biomarkers will be combined and where potentially better evidence is generated. Maybe also for very specific use cases is obviously NFL or also GFAB -- biomarkers, which may add a certain components to the fossil tows -- so I think moving forward and also thinking long term, we will see certainly a combination of biomarkers where, again, we may integrate with more patient-related data into algorithms. So that is probably the future.

On the market potential per se, so it's -- we outlined a global market potential, but you are absolutely right. At the moment, it's certainly in the U.S. market, which is the growing market based on the acceptance of disease modifying therapies. Thinking about the future, we see the growth actually across all geographies. And it's clearly also our goal to have solutions basically for all geographies, including, for example, also China as an important future market.

And I believe you also said MSED. So for them said, yes, that's a global number. But again, it's early days, and I think the potential is that it could be higher, but I think that's a number we feel comfortable with. And then, Josh, I think you can take the profitability

Yes, for sequencing, I see no reason why even blockbuster scale sequencing can't be massively profitable. Molecular diagnostics in general are high-profit diagnostics. I think if you look at the field, you can look at public companies that are out there and their margins. We have every technical entitlement to be just as cost efficient, if not more, than all the solutions that are out there. And I say that based on our technology today, let alone on where we can take it over time, we think 2 things will happen. One is our technology will scale, and we will remain competitive and profitable. And two is the focus of cost is going to shift a bit. We're already seeing potentially diminishing returns on if you get the cost per genome for a certain amount, the costs are elsewhere. They're in library prep, they're in sample collection, they're in other things. So it's not obvious where this will go, but we'll be profitable no matter where it goes.

And what Josh said is assuming we will also continue to maintain the same type of R&D investments as well.

Okay. Next question, we'll go to Doug Schenkel from Wolfe.

You've done a really nice job today demonstrating progress across the portfolio and outlining strategic priorities for the next several years. A lot of this progress has come via organic efforts. It does feel like -- and I haven't gone back and pressure tested this observation, but it does feel like there's more contributions that you're highlighting today from partnerships, investments and acquisitions. Amongst those, I think, of Saga and Foundation and Freenome amongst others. So with that in mind, really a 3-part question. One, over the next 5 to 10 years, what's your expectation for organic versus inorganic development? Two, what's the criteria we should expect you to use? I'm thinking of things like size, growth profile, when and why? And then maybe a little bit redundant, but the third part is, does this criteria differ by diagnostic subsector or even geography based on the geopolitical, regulatory and market backdrop?

Yes. So that's a great question. And what I would maybe start by saying is if you look at what we highlighted today, which is Foundation, which has been part of Roche for a while, but Saga and PathAI, these are 2 particular organizations that fulfilled a very key part of our overarching strategy that we've had. and that being the integration of high medical value diagnostics with clinical decision support tools, very on strategy and in an area where it's very clear that these are going to be growth drivers. And what I would say when we look at and we look at everything, we're the largest IVD company in the world, and that limits a lot of our optionality for inorganic growth.

But in this case, what we saw was, and we showed our key areas of focus, these were key drivers. I mean we know that MRD is going to be one of the biggest drivers of the diagnostics industry in the years to come. We saw an optionality that very much fit with our technology stack and our strategy. So it was a smart bolt-on. And with PathAI, the absolute complementary nature of that to our VENTANA business also made sense and the fact it was very much aligned with our strategy. So that's the way we're going to look at it. But again, we will be somewhat limited by our size and our scope, our scale, excuse me, because we are the biggest IVD company. I hope that answers your question.

Okay. I think one more, and then we need to close the Q&A session. So without question. So...

I mean PAMA has been pushed for about 6 years now. I just want to get a feeling of whether you think PAMA or a version of it through the Results Act gets enforced from next year. And in general, I mean, the U.S. reimbursement environment, do you see it getting worse from here given what seems like a 6-year break?

Yes. So I think, as you know, we're a member of ACLA or associate member of ACLA, we're a member of AdMeds. We work through our industry association organizations to really advocate for reimbursement reform. And that's to see reimbursement levels in diagnostics accurately reflect the payments that are given. And I think that's, as you said, the Results A. The reason PAMA has been pushed out is because I think there's a reasonable level of understanding that this would be challenging for the industry. And I think that's why you see also a focus on reform. It's going to be impossible to say what's going to happen at the end of the year, but I think what you see is a lot of initiative and a lot of, I'd say, groundswell support for seeing some reform. So we're going to continue to advocate for that. But again, as we -- I think we've had the question before, we don't expect that even if it didn't get extended, we wouldn't see this as really material to our business.

Great. Thank you. And with that, I would like to hand it over to Matt for the closing.

Thanks, everyone. Okay. So first off, I just want to thank everyone from all the Roche side, all the speakers and everyone who worked on the presentations in the background, everyone who traveled here. I just want to say a big thank you. It's really been a wonderful team effort. And again, thanks to everyone who's here today and watching us online.

So what I would like to just say, just in summary, and I think you saw this from what we really highlighted today and this really, I'd say, deliberate intentional purpose behind our strategy is we are the IVD leader, and we have a clear right to play and right to win to grow above the market and continue to stay in that position.

The second is I think you heard this from Josh and then the real clear synergy with Dan and the foundation organization, Helios has a differentiated value proposition in the sequencing space, and we feel it is positioned to change this industry. And we have a leading end-to-end oncology portfolio, and this will have benefits from our DI business, but on to pharma. And the question that came specifically around the ODAC, I mean, if you think about it, surrogate endpoints and the ability with something like MRD, that enables a lot of clinical trials that can't be run today. And I think there's -- across the entire Roche group, -- what you see is the synergies of our oncology portfolio to be additive for our growth in the future.

And then finally, you saw from Andy and Morris, we are fundamentally integrating digital and AI into our products that is going to set them apart from the competition and really continue to make us differentiated.

And so with that, I'd like to thank you all. And again, thank the Roche team, and thanks very much.

Good morning. The Roper Technologies conference call will now begin. Today's call is being recorded [Operator Instructions] I would now like to turn the call over to Zack Moxcey, Vice President, Investor Relations. Please go ahead.

Good morning, and thank you all for joining us as we discuss the first quarter 2026 financial results for Roper Technologies. Joining me on the call this morning are Neil Hunn, President and Chief Executive Officer; Jason Conley, Executive Vice President and Chief Financial Officer; Brandon Cross, Vice President and Chief Accounting Officer; and Shannon O'Callaghan, Senior Vice President of Finance. Earlier this morning, we issued a press release announcing our financial results. The press release also includes replay information for today's call.

We have prepared slides to accompany today's call, which are available through the webcast and are also available on our website. And now if you please turn to Page 2. We begin with our safe harbor statement. During the course of today's call, we will make forward-looking statements, which are subject to risks and uncertainties as described on this page in our press release and in our SEC filings. You should listen to today's call in the context of that information.

Now please turn to Page 3. Today, we will discuss our results primarily on an adjusted non-GAAP and continuing operations basis. For the first quarter, the difference between our GAAP results and adjusted results consists of the following items: amortization of acquisition-related intangible assets and financial impacts associated with our minority investment in Indicor. Reconciliations can be found in our press release and in the appendix of this presentation on our website. And now if you please turn to Page 4, I'll hand the call over to Neil. After our prepared remarks, we will take questions from our telephone participants. Neil?

Thank you, Zack, and thanks to everyone for joining our call. As we turn to Page 4, you'll see the topics we will cover today. We'll start by highlighting our Q1 enterprise performance, then Jason will walk through the enterprise financials, our balance sheet and provide an update on our share repurchase program. Then we'll discuss our segment highlights and outlook and introduce our Q2 and increased full year 2026 guidance. Finally, we'll close with a few summary points before opening the call for questions. So let's go ahead and get started.

Next slide, please. As we turn to Page 5, I want to highlight 3 takeaways for today's call. First, we delivered a strong start to 2026 and are raising our full year DEPS guidance. Our Q1 results exceeded expectations across every key metric. Total revenue grew 11%. Organic revenue grew 6%. EBITDA grew 8%, free cash flow grew 11% and DEPS was $5.16. Importantly, enterprise gross retention remains strong, consistently in the mid-90s area.

On that foundation, enterprise software bookings were also strong, core up low double digits on a TTM basis. This continues the momentum from our last call and bolsters our confidence for the balance of the year. On the back of this quarter's performance, we're raising our full year DEPS guidance to a range of $21.80 to $22.05, up $0.50 at the midpoint and more on this later.

Second, we're continuing to accelerate AI velocity across the portfolio. In Q1, AI innovation continued to broaden across our businesses, move deeper into core products and increasingly show up in both product road maps and customer conversations. Businesses like CentralReach, ConstructConnect, Vertafore, iPipeline, Aderant, DAT, Subsplash and SoftWriters all released meaningful new AI-enabled product capabilities during the quarter.

The signal from our own portfolio that AI can be a meaningful growth driver in vertical software keeps getting clearer by the day. On the AI accelerator team at Roper, as a reminder, this is a central strike team that partners directly with our operating companies to accelerate AI product development and capture reusable patterns for deployment across the portfolio. The team is ramping quickly. The team's first partnership was with Vertafore, helping deliver AI agents unveiled at their customer conference last week. This is exactly the kind of portfolio impact we envisioned when we invested in this team, and we expect the pace of partnerships with our operating companies to accelerate throughout the year.

And our third takeaway centers on capital deployment. Since November last year, we have repurchased 6 million shares for $2.2 billion, including 4.9 million shares for $1.7 billion year-to-date in 2026. Importantly, our Board authorized an additional $3 billion of repurchase capacity, giving us $3.8 billion of remaining authorization and north of $5 billion of total capital deployment capacity over the next 12 months. Our approach remains unchanged. We're disciplined and unbiased between acquisitions and opportunistic buybacks, focusing on driving the best risk-adjusted long-term cash flow compounding per share for our shareholders.

Our M&A pipeline today is targeted, focused on high-quality strategic opportunities where we're developing deep relationships and real conviction, and we expect to remain active and disciplined long-term buyers. Before I turn it to Jason, one theme you'll hear throughout today's call, organizational velocity across our portfolio continues to build. The investments we've made over the past 2 years in leadership, in AI and modern engineering practices and operational rigor are working and demonstrating meaningful results. Our businesses are releasing innovation faster, executing sharper and moving with more confidence. And that's what gives us conviction in the balance of the year and beyond. So with that, Jason, let me turn the call over to you.

Thanks, Neil, and good morning, everyone. I'll take you through our first quarter financial performance, starting on Slide 6. As you heard, we delivered a strong first quarter, finishing well above the high end of our DEPS guidance range and ahead of expectations on organic growth. Revenue of $2.1 billion was up 11% with organic growth of 6% and acquisitions contributing 5%. Importantly, recurring software revenue growth across our software segments was again strong at 7%, which continues to be the best indicator of business health and durability.

EBITDA of $797 million was up 8% over prior year. EBITDA margin was 38.1%. Our core EBITDA margin was down 70 basis points in the quarter, driven by lower gross margins in our TEP segment due to mix of more consumables at NDI and Verathon, coupled with higher input costs at Neptune. Core EBITDA margins in our software segments expanded 40 basis points, which includes continued investment in AI. DEPS of $5.16 was above our guidance range of $4.95 to $5 and up 8% over prior year. The upside was driven by the combination of stronger organic growth, a lower tax rate and the benefit of lower share count resulting from our net purchasing activity in Q1. Free cash flow of $562 million was up 11% over prior year.

On a trailing 12-month basis, free cash flow is now $2.5 billion and has compounded at a 19% CAGR over the last 3 years or 15% excluding the impact of Section 174. We continue to view free cash flow per share as the most important metric in evaluating our progress. And on that basis, we were up 15% versus the prior year, given the combination of growing cash flow and a declining share count. Relatedly and for modeling purposes, we exited Q1 with 102.4 million shares outstanding.

Now if you turn with me to Slide 7, I'll walk through our financial position and capital deployment update. We exited Q1 at 3.1x net debt-to-EBITDA, which is up modestly from 2.9x at year-end, given the $1.5 billion we deployed towards share repurchases in the quarter. We have $383 million of cash and $2 billion drawn on our $3.5 billion revolver. Importantly, we closed on a new 5-year $3.5 billion revolving credit facility during the quarter, which provides ample liquidity and improved pricing and terms. This also enhances our cost of capital strategic advantage in the face of an increasingly constrained private credit market that other market participants looking to make acquisitions will be facing.

Even after significant repurchase activity in Q1, we maintained over $5 billion of annualized capacity for capital deployment, which speaks to the strength of Roper's cash generation engine. Neil highlighted the share repurchase activity in the opening. To put it in perspective, our cumulative 6 million of share repurchases is about 6% of shares outstanding and brings us back to a share count we have not seen since 2017. Additionally, our Board approved expanding our share repurchase authorization by another $3 billion, which provides capital deployment flexibility and reflects continued confidence in our vertical market software position, enhanced capabilities and execution velocity to capture the AI opportunities in front of us.

On M&A, the pipeline of high-quality opportunities remains very attractive. As we've discussed, we believe the structural dynamics in the PE-backed software market and a constrained private credit market continue to create a compelling environment for Roper. We remain active and disciplined. With that, I'll turn it back over to Neil to discuss the segment performance and outlook. Neil?

Thanks, Jason. As you turn to Page 9, let's review our Application Software segment. Revenue for the quarter grew 12% in total and organic revenue growth was 5%. EBITDA grew 13%, EBITDA margins were 42% and core margins improved 50 basis points year-over-year. The quality growth here is notable. Recurring and reoccurring revenue, about 85% of the segment grew in the mid-single-digit plus range, while nonrecurring was essentially flat. Stepping back at the segment level, 3 themes stand out for the quarter.

First, enterprise gross retention remained strong, consistently in the mid-90s area. On that foundation, enterprise bookings were also strong in the quarter, consistent with the momentum we described on our January call and supportive of our confidence for the balance of the year. Second, our SaaS transitions continue to advance meaningfully. Several of our larger businesses made real progress on ground-to-cloud conversions and on bringing new cloud-native products to market. And third, AI progress continued to build. The signal of shifting from product investment to product shipping, and you'll see this clearly in the 3 company highlights to follow. First, Aderant delivered a record quarter, strong revenue growth and a new Q1 bookings record.

Strength was broad-based with particularly strong SaaS momentum on Sierra, Onyx and viGlobal. Aderant also launched AI-driven talent evaluation within viGlobal, continued the rollout of a Stridyn AI platform and completed a record number of Sierra cloud migrations in the quarter. Simply put, Aderant is winning in the legal market and doing so from a position of strength.

Second, Vertafore delivered a solid quarter, steady mid-single-digit revenue growth with EBITDA ahead of revenue. Recurring revenue continued to build across agency, MGA and carrier with MGA again leading on double-digit growth driven by strong bookings and high retention. And last week, at their Accelerate user conference in Las Vegas, Vertafore unveiled its new Velocity AI platform, along with a suite of AI agents embedded across the product portfolio from ReferenceConnect and reconciliation to submission processing and e-mail agent automation. AI is a meaningful TAM expansion opportunity for Vertafore, and they're quickly moving to capture it. As I mentioned earlier, this is where the Roper AI accelerator team had its first impact and is exciting to see.

And third, CentralReach continues to execute ahead of our deal model. Recurring software revenue grew well north of 20% with margins expanding, demonstrating the operating leverage in this business as it scales. And most importantly, CentralReach continues to be one of our strongest AI proof points. AI-generated session notes have dropped from 5 to 10 minutes to about 30 seconds, giving clinicians back roughly 8 hours a week to work with autism learners.

BCBAs are saving 140-plus hours a year on report authoring and review and daily claim generation is 6x faster. Customers are responding. AI and AI influenced bookings were 75% of new business in the quarter, up from 0 2 years ago. This is a textbook example of how the AI right to win we believe exists across our portfolio. CentralReach sits inside mission-critical workflows, has proprietary data and is translating that advantage into real growing AI revenue.

Prior to turning to the outlook for this section, I'll provide an update on Deltek and the GovCon market. Importantly, Deltek grew recurring revenue in the mid-single-digit plus range in the quarter, driven by strong private sector demand, partially offset by continued softness in GovCon enterprise. SaaS remains strong with ground-to-cloud conversions trending positively. Consistent with January, we're still waiting for the GovCon inflection. This is not new. We continue to work through the tail of last year's disruption to federal procurement, agency reorganizations and broader budget uncertainty, which is delaying decision-making, particularly on large enterprise perpetual deals.

Longer term, we remain encouraged. The One Big Beautiful Bill is a meaningful positive for defense and government contracting spend, but the benefit reaches us only after our customers win awards and invest in systems, and that takes a bit of time. Consistent with January, we are not baking into our guidance any GovCon inflection or any OBBB benefit and rather will adjust as conditions warrant.

Turning to our outlook for Application Software. We expect organic growth for the balance of the year to be in the mid-single-digit plus range, lower in Q2 on some nonrecurring timing, improving in the back half with CentralReach turning organic and easing nonrecurring comps. Please turn us to Page 10. Total revenue growth in our Network Software segment was 14% and organic revenue grew 5% in the quarter. The quality growth mirrored application software. Organic recurring grew mid-single-digit plus, nonrecurring declined mid-singles as customers move to our cloud offerings and bookings remained strong here.

EBITDA margins were 50.7%, down 460 basis points year-over-year, while core margins held steady, down just 20 basis points. The gap reflects 2 dynamics: our acquisition of Subsplash, a faster growth business with a lower but steadily improving margin profile and our ongoing investment in DAT, particularly Convoy. Stepping back at the segment level, we see similar themes playing out here that we described in application software. First, enterprise bookings were strong and gross retention remained high across our network businesses, together giving us improved visibility into the balance of the year.

And second, AI progress is tangible and shipping to customers today. Let me highlight 3 businesses in this segment. First, DAT is executing well against a mixed freight backdrop. ARPU expansion continues and adoption of our digital freight marketplace solutions remain strong. On the macro, spot rates are up 20% to 30% year-over-year and the carrier side of our ecosystem grew in Q1 for the first time in several years, real green shoots, particularly in the second half of the quarter.

That said, a sharp diesel spike compressed carrier margins late in the quarter, and our guidance continues to assume no meaningful freight market recovery. Our early-stage investment in Convoy inside DAT represents a material TAM expansion opportunity. Today, DAT is a subscription-based 2-sided network. Brokers and carriers pay to access the largest freight marketplace in North America. With Convoy, DAT is evolving into a full end-to-end Agentic and ML-powered marketplace, participating in the workflow and the economics of the transaction itself, a meaningfully larger and more valuable business over time.

The innovation that enables this transformation exists and is working in the market, and we continue to enhance and extend the tech. In the most recent quarter, DAT's RateView AI agent moved into live production, replacing manual rate lookups with instant conversational lane rate guidance. Convoy Load Notes is turning brokers freeform e-mails and chat messages directly into bookable loads, eliminating manual data entry and Loadlink's Voice to Post is enabling hands-free load posting. I work at DAT is not theoretical, it's shipping in production and delivering incredible value to customers today. Turning to ConstructConnect, another strong quarter with recurring revenue up double digits and continued breakout from Boost, their AI-based takeoff solution.

AI AutoCount, which reads construction schedules, launches this quarter. Most importantly, ConstructConnect has now moved its entire product and engineering organization into Agentic coding processes and tools shipping 4x the features versus a year ago. Broadening this across the portfolio to drive multifold product velocity gains is a key priority and an exciting one for enterprise.

And third, Foundry returned to year-over-year revenue growth in Q1 with Nuke closing the quarter at record ARR. Net retention returned above 100% for the first time since the 2023 actors and writers strikes and our recent Griptape acquisition extends Foundry's leadership into AI orchestration across the visual effects and animation pipeline, enabling studios to securely coordinate multiple AI models and agents in their production and post-production workflows.

Finally, and prior to turning to our segment outlook, I'd like to make a couple of quick callouts. SoftWriters launched its AI-enabled order entry product last week, a meaningful workflow enhancement for long-term care pharmacies and Subsplash released Trends AI, giving ministry customers the ability to generate custom data insights through natural language prompts, a key unlock for this customer constituency.

Turning to our outlook for network software. We expect organic growth for the balance of the year to be in the mid-single-digit plus range. A couple of quick call-outs. Subsplash turns organic in Q4, and margins will reflect continued investment in our freight platform acquisitions for the balance of the year. Now please turn to Page 11, and let's review our Technology-Enabled Products segment. Revenue here grew 9% in total and 7% organic, significantly better than expected, driven by strength at NDI and Verathon.

EBITDA margins were 33.6%, down 260 basis points year-over-year, reflecting 2 dynamics: First, input cost pressure at Neptune, principally bronze ingot inflation; and second, a mix shift at both NDI and Verathon towards faster-growing consumables, which carry lower gross margins but more durable reoccurring revenue profiles. Let me start with NDI. Another record quarter driven by exceptional demand for their electromagnetic tracking solutions across cardiac, neurological and orthopedic precision measurement applications.

The EP market, in particular, is a strong multiyear growth vector for NDI. Procedure volumes continue to grow, leading OEMs are introducing new tracking-enabled catheter platforms and NDI has a unique right to win at the sensor layer. Great job by Dave and the entire team at NDI. Turning to Neptune. Revenue declined low single digits in the quarter, which was better than expected, driven by strong execution from Don and the entire team in Tallassee.

The market dynamics were largely as expected with lower mechanical meter volumes partially offset by strong static meter growth. Importantly, Neptune's cloud-based software adoption continues to scale nicely, though off a small base. Consistent with our Q4 commentary, we're not underwriting a Neptune recovery in our 2026 guidance, and we'll continue to monitor underlying demand. Rounding out the segment, Verathon delivered solid growth, supported by strong BFlex and GlideScope demand, and we're optimistic about new product launches planned for the balance of the year.

Turning to our TEP outlook. We expect organic growth for the balance of the year to be in the mid-single-digit range, lower in the second quarter as we face a tougher Q2 comp. We expect net raw material pressure to continue in the second quarter and improving in the back half of the year. With that, please turn us to Page 13. On this slide, we'll cover our Q2 and full year 2026 guidance. Specifically, we're raising our full year 2026 DEPS guidance to $21.80 to $22.05, up from $21.30 to $21.55, a $0.50 increase at the midpoint, which passes through our Q1 beat and the impact of our already executed share buyback.

We're maintaining our full year total revenue growth guidance of approximately 8% and organic revenue growth of 5% to 6%. For the full year, we continue to assume a tax rate in the 21% area and a bit below that in Q2. For Q2, we're establishing our adjusted DEPS guidance of $5.25 to $5.30. To reiterate key assumptions from our segment commentary, full year guidance assumes no meaningful improvement at Deltek's GovCon market or DAT's freight market and modest top line weakness at Neptune versus a year ago.

Finally, on capital deployment, we're entering the balance of 2026 with meaningful optionality. We have $5 billion of firepower available over the next 12 months, a targeted M&A pipeline and $3.8 billion of remaining share repurchase authorization, giving us substantial flexibility to act opportunistically. We will remain disciplined and unbiased between acquisitions and opportunistic buybacks based on what drives the highest and most durable cash flow per share compounding.

Now please turn to Page 14, and then we'll open it up for your questions. We'll conclude with the same 3 takeaways with which we started. First, we delivered a strong start to 2026 with 11% revenue growth, 6% organic revenue and 11% free cash flow growth. Retention and bookings remain strong and position us well heading into the balance of the year. Based on this, we've raised our full year DEPS guidance by $0.50 at the midpoint.

Second, we are accelerating AI innovation across the portfolio. CentralReach, ConstructConnect, Vertafore, DAT, Aderant and others continue to move AI deeper into their products and increasingly into customer activity, and our AI accelerator team continues to build velocity across the portfolio. Finally, on capital deployment, as we discussed earlier, our Board authorization of an additional $3 billion of share repurchase capacity gives us $3.8 billion of remaining authorization. Alongside that, we have $5 billion of capital deployment firepower available over the next 12 months, supporting our targeted M&A pipeline.

We will remain disciplined and unbiased between acquisitions and opportunistic buybacks based on what drives the highest and most durable cash flow per share compounding. As we wrap up, some additional color on the M&A market. A quarter ago, our pipeline was at record levels. Shortly after our call, the broader public software valuation drawdown caused sellers to pause most active processes. We remain active and our pipeline leans more proprietary.

That said, we expect M&A activity to pick back up, timing of which is still to be determined. But when it moves, a large number of opportunities are likely to emerge, and we're in an advantaged position to capitalize on this. We remain very bullish about being a high conviction acquirer of vertical market software businesses with deep proprietary moats where AI accelerates growth. The signal on that thesis from our own portfolio is becoming clearer and clearer.

So in closing, the ingredients for accelerated cash flow per share compounding are coming together. Our portfolio is the strongest it has ever been. Our organizational velocity is accelerating. AI is both TAM expanding and growth enabling, and we're excited to see our product work translate into higher growth. Our capital deployment capacity and flexibility are significant differentiators and our discipline is unchanged. This is how we compete and win and how we continue to compound for our shareholders. With that, operator, please open the line for questions.

[Operator Instructions] Your first question comes from Dylan Becker with William Blair.

Really appreciate it. Nice job here. Maybe, Neil, starting for you. I think it was clear in your commentary, you kind of talked about the accelerating pace of innovation and the right to win and TAM expansion -- kind of TAM expansive nature of AI. But if we think about kind of the embeddability piece and monetization of the platform, I guess, maybe how that layers in incremental conviction as well, too, right? Is that something that can lower friction around adoption? Is that something that can increase kind of the likelihood of success and value alignment with customers, but maybe how the platform positioning and embeddability of agents maybe layers in kind of incremental confidence in that right to win around agents?

Yes. I just want to -- so you're asking about embeddability. I want to make sure you're a little muted on that. I want to make sure I'm answering the right question.

Yes. So the ability to kind of embed it into the existing platform, right, and kind of...

Absolutely. So a few things I'd start with on this. So it really starts with what sort of we talk about internally all the time about the AI, the product magic. We're able to create products now across many, if not all of our software businesses where when the customer sees in early betas and early trials like what the product can do, like their eyes sort of pop out of their head. It's like truly like a magical experience. I didn't know software could do that, right? So that's what gets us like really excited.

We just saw it last week, for instance, at the Vertafore customer conference, just sort of as an example. So in terms of monetization, generally -- so that's one. I'll start there. Second is we believe that the right to win here is sort of on-stack AI embedded natively in workflows is a winning play, a huge incumbent advantage. So it is the second thing. Third thing, monetization, I think, for us is there's not going to be a one size fits all. There are some businesses today that already price on a consumption basis.

I think like SoftWriters and pharmacy automation or what Convoy does at DAT. So I think those will be monetized on a consumption basis. Also those customers' unit economics generally are driven on their own consumption, so it aligns with the customer unit economics. I think more broadly, though, the monetization is going to be one that sort of, as you alluded to your question, balances adoption and long-term monetization. So I think that's going to not be largely consumption-based.

Our customers very much are saying very clearly, they need to be able to plan for and budget what the spend is going to be. So it will likely be some sort of a subscription with an overage based on utilization of the AI tools. I think that aligns nicely with adoption because then the customers are going to be focused on how to realize the magic value, if you will, and not be worried every time they press a button, it costs money. But then when it gets fully adopted and there's like deeply embedded into workflows, we'll be able to sort of grow with that utilization.

And I would just add that our CentralReach business is furthest along in this journey. They've been out in the market with AI products for 1.5 years or 2 years, and all of their AI is incremental. It's based on learners, which you could say is some form of consumption, right? It's not based on practitioners but learners, but that's been -- and customers are seeing real value, as Neil highlighted on the prepared remarks, in terms of workflow efficiency and better revenue realization.

Very helpful. And maybe, Jason, kind of just sticking with you quickly as well, too. Obviously, kind of reiterating the full year revenue guide, 5% to 6% organic. We just did 6% this year. We've got some mechanics kind of layering in and easier comps in the back half as well, too. But maybe kind of just give us a broader sense of how the start of the year kind of layers in conviction and maybe that kind of conservative view that we continue to take to the guidance framework here going forward.

Yes. Yes. Look, it's a strong start to the year, very encouraged by what we've seen. But we're just 1 quarter in. So we want to sort of see how things play out. As Neil talked about, we have a couple of, like you said, mechanical things in the second quarter nonrecurring in AS will be a little bit more impacted than the first quarter. And then we're just -- in TEP, we're comping a high watermark in Q2, but that will -- that will ease off in the second half.

And then as we've talked about, the second half will improve in software with Subsplash and CentralReach turning organic. And then we have just some easing comps in AS. So all that just sort of blends into our sort of holding the range at this point, but we'll see how it plays out.

Your next question comes from Brent Thill with Jefferies.

This is Ria on for Brent Thill. Neil, just curious to hear your thoughts on the private markets given ongoing volatility. Can you just tell us a little bit more about what you're seeing right now and if it's changed your outlook at all?

Talking about private markets on M&A, you're talking about?

Correct.

Yes, sure. So as I mentioned in the prepared remarks, it's definitely been with the public market drawdown. It's been -- it's gone from the busiest we've been in a long time to the least busy. We're still busy. We're still active. As I mentioned, it's more proprietary. It's certainly more targeted. But it's actually -- we think the M&A setup has actually improved a bit for us over the last 90 days in the context that the LP pressure that we've talked about now for a couple of years continues to exist. That has not changed in any capacity.

If anything, it's maybe increased over the course of the next 90 days. The other thing that's happened that's been -- everybody is widely reported, people understand is now we got the private credit dynamic that also is putting pressure on the asset class. So for us, it's -- we think the combination of those 2 will likely service more quality assets in the processes, and we're a very advantaged buyer in that regard. But the timing is still to be determined.

We're modeling out what these maturities look like on the private credit side. There's not a meaningful maturity cliff this year. But if you're a private equity sponsor seller, you want to think about divesting an asset well before maturity. So that's something that Janet and her team are sort of lining up. So we think there's an opportunity here to get -- potentially, I should underscore potentially to acquire AA+ assets at differentiated values given the backdrop and the dynamics here. The timing of this is to be determined, but we'll stay active in process and prosecute the opportunities in front of us.

Yes. And I would just reiterate, we refinanced our 5-year revolver this quarter at a very good cost of capital, sort of tightened up the spread a little bit. So shout out to Shannon and Dave Baker for getting that done this quarter. Just a great job there. And it just positions us well. We have a lot of balance sheet flexibility, and we'll be able to move quickly when the opportunities arise.

Got it. That's helpful. And then just on Deltek's government contracting business, did you see any impact in the quarter at all from the war in the Middle East? And is it having any impact on your outlook for the remainder of the year?

Yes. We asked that very specific question on our call down with Deltek. And the short answer is very little, if any. There certainly is a sliver of the sort of aerospace defense subsector of government contractors that are focused on munitions and sort of war effort. So that's a small sliver of the population of the broader, I'd say, contractor population. So it had, if in effect, a minimally negative impact just in terms of those contractors are focused on the war effort and not on contracting for ERP software, but it wasn't material in the quarter.

Your next question comes from Joseph Vruwink with Baird.

I think all application software is facing this question around whether AI-related spending grabs an outsized wallet share and maybe the incumbents get squeezed along the way. I think the interesting thing about Roper is you have exposure to markets like legal and health care. I think those are the 2 biggest vertical AI adopters so far. And yet I think your respective software exposure there is still doing pretty well. What's your take on this topic? And have you seen any changes year-to-date as we've also seen the big ARR numbers come through from the frontier model providers that make you more concerned in coming quarters?

No. I would say we're -- the punchline on that, the TLDR is no. No impact on sort of the budgetary spend that we sort of compete for. I think the double-click on that is the obvious answer, which is -- and this is a personal opinion that I think a lot of these surveys around IT spend are a little misleading because the whole point of the AI effort is we get to go monetize labor spend. So it's about a whole different bucket of opportunity to capture and provide value to the end market. In the particular -- across the whole platform, we're not seeing sort of an impact to us relative to allocation of budget, especially not in legal and health care, we've had [ fairs ]. We talked about Aderant, which is amazing in the quarter. It's been amazing few years here with Aderant.

Great. That's helpful. And then I heard enterprise bookings up low double digits over the trailing 12 months. Curious what they were in the quarter. And I think your definition excludes price. Maybe can you just comment on pricing power in the aggregate?

Yes. So it was certainly above the double digits. We had an easier Q1 comp last year. I think the TTM is definitely the right way to think about it. Yes. And then in terms of -- it does not include price, you're right. And price has held up very well. I think what we've said historically, we're very thoughtful across the portfolio about pricing, and you have to earn the right and companies are doing that as part of our strategic plan work that we do is understanding that dynamic. And so we've continued to do that methodically over the last half decade or so.

Yes. The only thing I'd add on pricing in addition to what Jason said is we actually think relative to what the market will bear on pricing, we have underutilized that lever in growth. And so it's not like a pan portfolio go raise pricing. That's not how we operate at Roper. But as Jason said, it's like where you have earned the right with your product, your product value and your customer relationships to take a little bit more pricing, then we are doing that. And -- but it's a very strategic. It's a very earned process. And we would hope that we would see a little bit -- I don't know if it's 50 or 100 basis points over the portfolio of software and pricing impact increase over the next couple of 3 years.

Your next question comes from Terry Tillman with Truist.

I wanted to build on the prior question on Legal Tech because, yes, it's in the media reports and some remarkable growth from some of these SaaS natives there. But I'm curious, though, you've called out Aderant a couple of years of amazing. And it does seem like it's like clockwork showing up in the segment level slides every quarter on record this or that. How much more sustainability is there in terms of just the momentum in terms of getting folks to move to SaaS? And just can this train keep going just on the momentum with Aderant. And then I had a follow-up.

Sure. So -- it's -- Chris and the team there have done a great job. I mean I'll give you a little bit a longer answer here. Aderant has been really good for a very long time for us. But what's been happening underneath the hood has evolved to sort of keep it good. It started with how do we just take it to our competitor and outcompete them in the marketplace. And that's how we went from 35%, 40% market share in large law to 60%, 65%.

We just absolutely compete and won, and Chris and his predecessor team did a wonderful job in that sort of era of growth. Well, that era of growth we could see was going to end at some point. So we had to evolve. That's where we sort of said, okay, let's -- we have this installed base of customers. how do we sell them more things. And so we then prosecuted both an organic and inorganic strategy to add the number of bolt-on products that we could or sort of integrated modules that we could sell to this large law customer base that made strategic sense.

So we prosecuted and prosecuting that strategy. Then came along cloud, right? This was a constituent that did not want to move to the cloud, then COVID happened. So we rapidly cloud-enabled the totality of the product set. And then we're now in the still early innings, maybe third or fourth inning, maybe not even that late of moving this customer constituency to the cloud with that lift and shift. And now we have the tailwind of the AI benefit in terms of being able -- -- so there's -- it's a multiple growth driver story, and I think there's quite a long way to go on this. But part of the benefit of owning any business for the long arc of time is you're always looking out Horizon 2 and Horizon 3 for what you have to build either organically or inorganically to sustain or improve growth rates.

Yes. Very helpful. And the follow-up is just what we're seeing though is with particularly not necessarily generative Agentic, I mean, that's a pretty big lift and shift in change management, customers being comfortable having things to go autonomous and even getting it beyond kind of the experimental stage. So are you seeing with some of the businesses, you actually have to hire -- put in forward deployed engineers or kind of change how you go to market or help the customers and it does create some kind of incremental costs or just handholding? Just anything about how you help them consume this Agentic stuff?

Yes. I think that's -- the short answer is yes. I think we mentioned last quarter that this year is going to be just a massive learning year for us across the enterprise on -- I'll put it in like the commercialization bucket of these AI tools, of which FTEs are certainly a component. How do you position it? How do you sell it? How do you price it? How do you get it implemented? How do you get utilization pull-through? How do you drive renewal rates high? I mean that whole customer journey is going to be across the portfolio, a huge set of learnings for us.

We have -- I'll spare the details on inside the portfolios, but we have portfolios where -- businesses where the uptake has just been very natural. We haven't had to have the 4 deployed engineers because when you press the magic button and you get productivity savings that immediately that productivity savings is taken in the customer's operation is something that they can go do tomorrow. In other cases, there's some trepidation. If I press this button, do I lose my job. And you've got to sort of go through the whole change management process of that. I think in almost every case, folks don't -- our customers don't lose -- it's not lose their job.

It's how do you sort of do task replacement, task augmentation, they can go play offense inside their customer to go compete and win. But it's certainly sort of an -- sort of something you have to overcome in that regard. So yes, we do expect across a certain part of the portfolio to do some version of a forward deployed engineer. Yes. Final thing I'd say on that is I think it's kind of from an investment point of view, it's probably more of a reallocation or rebalancing of investment from customer support, customer service to FDE. So I don't know if it's like a huge cost increase. It's just a resource allocation dynamic.

Your next question comes from Joe Giordano with TD Cowen.

Just curious on your talk about like embeddability and subscription plus overage in the future. Like I get the view of like I don't want our customers to think every time they click a button, it costs them money. I fully get that. If these things become embedded and like the efficiencies potentially require less people at your customers, like how do you kind of judge the ROI of the investment necessary to kind of -- I am saying -- I'm saying it to kind of stay in the same place. Like the product is getting better, but you're getting like the same kind of subscriptions and it's like costing you more to maybe achieve that now than it did in the past. So how do you kind of evaluate the ROI on the required spend to kind of get to that place?

Yes. I think it's -- so this is -- these are very hard dollar ROIs. I mean we've said publicly, for instance, at DAT Convoy to manually broker load, it's somewhere between $100 and $200 of labor to do that. You use our load automation, it's somewhere in the $40 range. So it's a demonstrable hard dollar ROI. Similar things can be said at, for instance, at Vertafore, one of the Agentic tools they released last week. It's a reconciliation tool.

The time and motion study is it's like 17 minutes per reconciliation. Our tool does it in 30 seconds. Then you do these like scores of these a day. So you can sort of see the time savings and then you can get to a financial ROI. So these are pretty hard ROI and products and that sales teams are taking that message to the market to the customer base.

Yes. And I would just say that -- sorry, Joe, I would just say we're using local smaller language models, maybe even older versions. So you're not consuming a lot of tokens when you're doing this activity. So it's -- and you can continue to change the prompts to make it more efficient over time. And so we've even at Vertafore, we've taken that cost of goods down meaningfully in a matter of weeks. So I think it's still very accretive from a margin perspective.

Yes, that's kind of -- I'm getting at more of the ROI from Roper standpoint. I get -- I get the ROI from the customers. It's more like if we're spending money to develop new AI tools that are then embedded in the product that we're already offering, like how is the ROI on the increased investment you need in 2026 versus the investment you needed in 2021 to get the same customer and keep the same customer happy.

Well, and I would just say on the development front, I mean, we're seeing demonstrable efficiencies, right, with the frontier models itself. So we're getting a lot more output and a lot more road map consumed. So you talk about just OpEx investment. We're not assuming -- we're assuming productivity, but we're taking that back into the road map. So I don't think it changes fundamentally our P&L structure and our margin profile.

And Joe, apologies for missing the point -- the thrust of your question.

Your next question comes from George Kurosawa with Citi.

On the AI Strike Team, led by Shane and Eddie that you put together, it sounded like they completed their listening tour last quarter and have now been put out into the field. It sounds like some early success of Vertafore. If you could just touch on how they ended up sort of stack ranking the opportunities that they see in front of them and then maybe the scope of their involvement and how much it's led to an improvement in velocity?

Yes. So I'm delighted to double-click into that. So just to remind everybody sort of the 3 objectives of this AI, this Roper sort of accelerator team. One, and first and foremost is to sort of coach and teach, right? This is about enablement of our 21 software companies to do what they've already learned on their own relative to AI and Agentic development and then do it even better. So that's number one.

Second is to partner shoulder to shoulder and build. And then the third one is to, where appropriate, build sort of shared componentry that we can -- where we can share some common run time or routines on the AI front across the Roper companies where it makes sense. So that's sort of the goal and focus of this group. In terms of where we're allocating the team, this is very much an executive leadership team focus. It is basically size of prize and impact is how we're sort of force ranking this.

In terms of Vertafore, it is one of our largest opportunities, if not the largest opportunity we have from an Agentic automation point of view. I think there were 6 agents released last week at their Accelerate conference. That is just the very, very beginning. The model that -- and then we -- this quarter, we'll sort of broaden that from one engagement with one business to be -- it's now 6 as the team grows, and we have now 5 additional businesses that are sort of in the early stages of partnering with.

And the final thing is about speed. I mean I think the unlock here is, at least I think Amy and the team at Vertafore would agree is our team, the Roper team sort of -- very much partnered. So you can imagine leadership resources and our team working hand-in-hand with engineers on the Vertafore team on how to do this AI development, one, because there's a little bit of art to this and not just science.

Number two, there is a speed coefficient that our team brings given their history about sort of modern day like current very contemporary practices of Agentic development and just the pace. And then -- and then there's just good old-fashioned change management. How do you sort of break bottlenecks and barriers to go fast. And we saw literally, I know it's sort of an overused term, but 10x kind of productivity gains partnering with Vertafore on some of this development in terms of speed and quality. So we're super encouraged. It's very early days. I don't want Shane and Eddie to hear this and think they've manifested fully. They've got a lot of work to do, but it could not have gone better, in my view, in the first 6 months.

Okay. That's great to hear. And then I wanted to ask kind of more broadly, when you look across the portfolio, it seems like AI commercialization is in sort of different stages you've got businesses like Aderant, CentralReach that seem to be resounding successes. When you -- others seem to be coming up right behind them. When you look across that landscape, any pattern matching in terms of why some of these businesses seem to be moving a little faster than others? Is it primarily customer-driven? Or what would you attribute the relative successes there to?

I think it is -- Jason, I'll give an opportunity if he wants to add anything. I think if there's a pattern match there, when you have -- there's 21 software companies in the business. And while we want everyone to be going as fast as they possibly can, you have an array of where people are in their maturity. And where we're most advanced, they're the ones that got after and we're able to sort of get the Agentic SKUs into just in development first into the market first.

And sort of now the next wave of this -- we talk a lot about CentralReach and Aderant and Convoy and DAT, they're the tip of the spear. Now we have like 10 or 12 companies, maybe a couple more like just now just getting to market with real Agentic magic SKUs versus like chatbots and embedded sort of GenAI search inside of existing products where the value unlock is. And so we can -- and we also take a little bit offline about sort of more deeper operational pattern recognition, but that's what I would say about the commercialization phase is who sort of had product ready first.

That's right. Yes. And I think the benefit of being part of Roper, we set our President Summit a couple of months ago, and we did an AI sort of showcase for those that are further along. So it just helps with the learning acceleration. But I would agree with Neil that it's those that embrace and saw a true customer problem early on and then got after it a little sooner. But others are coming up the curve very quickly.

Your next question comes from Clarke Jeffries with Piper Sandler.

I just wanted to follow up on the comments around ground to cloud conversions advancing meaningfully. I'd love to understand the impact of SaaS transitions broadly in the application software segment. Is that contributing points of growth today? You made the comment around 85% of the segment is in the mid-single-digit plus range in growth, while nonrecurring was essentially flat. So I would just wanted to know if it's something that would be of increasing benefit or already playing out in that segment? And then one follow-up.

Yes, happy to take the question. So just as you think about the percentage of products that are cloud-enabled, it's 2/3 or so today. So we have about $1 billion of maintenance. And we think that, that will convert over say, the next 5 to 10 years or so, and that should convert at 2 to 2.5x lift from maintenance to SaaS. And so today, we're kind of -- if you think about the percentage that we have to go, we're sort of in the first or second inning of that journey. And so it does add, call it, 50 to 100 basis points of growth a year and should for the next 5 to 10 years.

The only thing I'd add is when we talked about this in the past, Clarke, we've also said we are very much pacing this ground to cloud conversion at our customers' pacing. We're not like forcing it to them. I'll say with the advent of AI, I should have mentioned earlier on the monetization, another monetization method for AI is embedding the AI sort of features in the cloud product, and that is a very compelling pull to make this transition go a little bit faster.

So instead of 8 to 10 years, maybe it's 4 to 6. I don't know what the right number is, but we would expect to see that go a little bit faster. The other thing is we made a tremendous amount of investment over the last 3 years getting product enabled. That was because we're growing our customer pacing, there was an urgency to get products enabled. And now we are extraordinarily product enabled. So basically feature parity, if not more so in the cloud product than on-prem. So I think the setup here is a little bit better than it was a few years ago.

Yes. And I would just say it's mostly -- and it's going to be Aderant a little further along, as you know, PowerPlan's in the early innings, but definitely much more cloud-enabled today than they were. And then when you think about those that are a little bit further behind, it's more health care, but that's our CliniSys business and labs. That's just kind of the nature of that end market. And so that -- so we see the areas of Deltek, Aderant and PowerPlan being those that will be more near term in terms of cloud migration.

Perfect. That all makes sense. And then one thing that kind of stood out to me was the margin impact in the Application Software segment. The margin impact of businesses owned for less than 4 quarters was actually positive year-over-year. Just wanted to unpack that. Is that -- is the takeaway here that even the earlier-stage acquisitions last year are getting to margin parity quickly?

So in application software, it's our CentralReach business, and that business is -- it's a very -- the business has ample R&D investment. I think R&D as a percent of revenue is like 20%, but they just have extremely strong incrementals. They're very a cloud-native platform. And so as they expand, they have very good incrementals there. When we talk about the acquisitions in our Network Software segment, we've talked about our -- the business Convoy that we added on to DAT. It's a technology investment.

We're super committed to that investment to automate the spot freight market over time. So that actually has a drag on margins. That plus our Subsplash business, which is a lower margin, faster-growing business that as they grow, they will scale margins. But you can see in our Network Software segment, it does have a pretty meaningful drag on margins. Now over time, as Convoy continues to grow, that should be a tailwind as we go into the out years. But this year, it is a little bit of a drag on margin.

Your next question comes from Josh Tilton with Wolfe Research.

Congrats on a really strong start to the year. I will keep it to one given the hour. But my question is just basically, you're very clear that the guidance still doesn't assume a recovery at Deltek and DAT for the rest of the year. Can you just remind us the confidence that you have in the rest of the application and network software business and kind of offsetting that weakness throughout the year?

I'd say just if we go through the segments. So with Application Software, we feel good about sort of what's going to happen in the second half. We just talked about CentralReach as just having a sort of a really strong start under our ownership and a lot of that's recurring. And so that's just going to flow through in the second half. We've talked about being 80 basis points or so of accretion in the second half for that segment, and we still feel good about that.

And then in network, DAT is looking good in the first quarter. We'll see sort of how things play out. Foundry will continue to be sort of getting better throughout the year. They had a great start to the year. And then Subsplash turns organic in the fourth quarter, and that's for sure, accretive to the segment. So yes, I feel good about the rest of the segment or the rest of the business.

Your next question comes from Ken Wong with Oppenheimer.

Great. Just one for me. It sounds like the kind of the downtick in 2Q is just purely due to tough comps, but just wanted to kind of make sure and clarify any geopolitical macro dynamics that you guys baked into that assumption as well given kind of the current situation that arose?

No, not at all. I mean it's just like timing really in the AS segment. It's our nonrecurring perpetual activity. And so that's squarely what it is. We have clear visibility to that. And then on TEP, no, I think we're comping 9% quarter is a high watermark last year. So it's just sort of a comp in the second quarter in TEP. It will get better in the second half. So nothing geopolitical at all. We're mostly U.S., as you know. So we don't see anything in the Middle East.

Your next question comes from Julian Mitchell with Barclays.

Maybe first off, I just wanted to try and put a finer point on the full year guidance. So is it fair to say that the sort of core EBITDA guide is essentially unchanged, and it's really a kind of share count-driven guide? And maybe help us understand what the share count assumption is now at the sort of guidance midpoint. And I think the guide embeds no extra buybacks beyond today. Just wanted to check that.

That's correct. Yes. So we had about a couple of hundred million of share repurchase between the end of the quarter and today. And so I think as I mentioned, but the ending share count for Q1 is 102.4 million, and then you've got some obviously dilution to add on top of that. So that's what we're assuming. But yes, you're right. It's -- we've mainly flown through the first quarter beat and then the buyback activity for the balance of the year.

Yes. And the first quarter beat for us, Julian, was partially from our model operating and partially buyback.

That's helpful. And within the network business, DAT has had a very tough sort of demand or macro backdrop, and it's been executing well within that. Finally, the last 6 months, there's better signals in the freight markets in the U.S. Maybe sort of flesh out a little bit more what you're seeing in that business and sort of what's dialed in for that transport linked business in the U.S. for the balance of the year, please?

Yes. So as we mentioned in the call, we're not -- in the guide, there's not an assumption for improvement. Also, I'll just double-click a little bit on the prepared comments. So for the first time in look, Jason, in a couple of 3 years, we've had carrier -- the carrier count side of the network increased, which is certainly a green shoot that we've been waiting quite a time.

Now we've had some head fakes intra-quarter on that number in the past. And so we're going to remain cautious. Also the input costs or diesel costs are certainly not helpful. So carrier margins or profitability would be a little bit challenging. And so -- but we're cautiously optimistic that there might be a freight recovery and rejection rates got better, the rates got better as we talked about, 20% or 30% better. So we'll see how it plays out, but we've underwritten no improvement in the outlook.

Your next question comes from Deane Dray with RBC Capital Markets.

This is Kenny Sim on for Deane. I wanted to ask about Neptune business. So one of your peers has some meaningful project delays disruption in the quarter for their water meter business. Have you seen anything similar in terms of the industry dynamic or even any market share changes during the quarter?

Yes, I appreciate the question. So for us, in our Neptune business, we would say largely no. We've not seen a project -- any project delays. Now the backdrop on that is slightly different than the competitor you described. And Neptune plays in the segments that are on the smaller municipalities. So we have never had a large amount of project-based work, generally speaking. So it's really not an apples-to-apples sort of question.

The other part of this is we had pretty decent sort of short-cycle demand in the quarter. And I think that's largely because we -- and I'm not commenting about our competitor because we don't know their business the way they do. But we -- Neptune did a good job managing channel inventory in 2025. And so the hope or expectation is we'll be able to ship closer to retail in 2026 on the short-cycle side. I think we saw that play out at least early in the year in Q1.

If I just have a follow-up, just if you could unpack the cost pressure dynamics for the Neptune business or even at the overall TEP segment level, either in the magnitude or the time line to offsetting those, that would be helpful as we kind of think about the segment incremental margins moving forward.

Sure. Let me -- I'll take a crack at this, but I definitely want to ask Jason to sort of correct and sort of amplify anything. So on Neptune, it's really the ingot cost. And what we decided to do, I think Don and team did a very sort of wise thing here. We did -- if you remember, 3Q, really July of last year, we pushed the sort of, I'll call it, tariff or a raw material sort of surcharge into the market. It really had a negative demand impact in the short run. The signal from the customer was, hey, we certainly appreciate we've got to onboard sort of global price inflation, but I'd rather do it through regular weight pricing versus surcharging.

And so we will sort of -- we expect, by the way, cost -- the baseline assumption we have is ingot cost is going to stay high. I mean this is with all the data centers and just the demand for copper, but this is a derivative impact of that. So our baseline assumption is this input cost is going to stay high for a while. So it will just be corrected or the margin will be captured through regular way pricing, which takes a couple of quarters to sort of work through backlog and get into the market. So we're taking a longer view on that.

In terms of the balance of the segment, it's really -- it's both Northern Digital and it's Verathon. These are businesses that are per our strategy, per the market opportunity are becoming more reoccurring in nature, reoccurring consumables, which is a great thing about the predictability of growth and the absolute levels of growth in the businesses, but the consumables come with a lower GP percentage. So GP dollars are going up, but GP percentages may be a little pressured on those 2 businesses. Now they also do a very good job managing below GP to EBITDA or OP, where we don't think there'll be a lot of OP compression over the long arc of time because they do have natural leverage in the business. Those are the dynamics at play. Jason, anything you want to amplify there?

No, I think you covered it.

This concludes our question-and-answer session. We will now return back to Zack Moxcey for any closing remarks.

Thanks, everyone, for joining us today. We look forward to speaking with you during our next earnings call.

The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.

My name is Henrik, and I'm the technical operator for today's call. Kindly note that the webinar is being recorded. [Operator Instructions] One last remark, if you would like to follow the presented slides on out as well, please feel free to go to roche.com/investors to download the presentation. At this time, it's my pleasure to introduce you to Thomas Schinecker, CEO Roche Group.

Thank you very much, and good morning and good afternoon to everyone. I'm very much excited to share our Q1 2026 results with you. Overall, a very strong performance in sales in the first quarter. And including there are some clinical and regulatory milestones that we have achieved.

Now let's start on the performance side. Q1 was strong with group sales at plus 6%, driven by pharma with plus 7%. Diagnostics grew plus 3% despite headwinds from the China health care pricing reforms, which we are still experiencing in the first half year. Excluding China, Diagnostics was at plus 5%.

In addition, a weaker and earlier flu season impacted the sales of Xofluza and on the Diagnostics side, the respiratory sales in point of care, but also -- or patient care and the Molecular business. This is across both divisions, approximately CHF 170 million, which more or less also explains the difference to the consensus.

Now moving to key milestones. Q1 was very busy with significant pipeline progress. On the pharma side, we filed giredestrant in adjuvant ER-positive for 2 negative breast cancer in the U.S. using a priority review voucher. And we do expect approval across two indications in the adjuvant and in the second line plus before the end of this year. Lunsumio was filed in the U.S. Gazyva, we had a series of filings in the U.S. and EU across SLE and INS. There's still a number of important readouts or there have been a number of important readouts, for example, the positive FENhance 1 results in RMS. This is the last one from fenebrutinib Phase III studies that -- where you also saw the results presented at AAN, and there was an IR call yesterday as well on this topic.

Now together with FENhance 2 and FENtrepid, this is the first and only BTK inhibitor with positive Phase III across both RMS and PPMS, and we do expect a filing in the next couple of months. I know that Teresa will go into more details here.

Gazyva, I think, is very much a success story. We've had 4 positive readouts in a row. The latest one was in MAJESTY in membranous nephropathy. And it follows positive readouts in [ LN ], so the lupus nephritis, SLE and INS.

In obesity, we announced positive Phase II results for petrelintide. In the study, we've shown excellent tolerability. In fact, we had more dropouts in the placebo than in the control arm. And this tolerability makes it an ideal candidate to combine with our very strong CT-388, which the GLP-1 GIP, which actually showed at week 48 similar weight loss than a leading other player after week 72. So we have a very strong GLP-1 GIP. We have a highly tolerable amylin, and we find this is a good combination.

Now moving to business development. I just want to highlight two deals. On the pharma side, we extended our collaboration with C4 Therapeutics on the greater antibody conjugates. And this is a promising novel platform, which could potentially overcome certain therapeutic windows and resistance issues that other drug mortalities have.

On the Diagnostics side, we acquired SAGA or we're in the process of acquiring SAGA Diagnostics. And they provide a minimal residual disease test to really monitor disease progression and so that we can pick out those patients and where we need to then change the therapy.

On the regulatory side, we've had CE marks for a new test for blood screening, which includes hepatitis E. We had the Elecsys APOE4p and the Elecsys Nfl for neurology portfolio.

Looking ahead, we still have a couple of important Phase III readouts this year and also some launches in diagnostics. There are two more NME readouts. One is divarasib in non-small lung cancer and sefaxersen in igAN. We have two readouts for Itovebi coming in breast cancer and also Lunsumio in second-line follicular lymphoma.

In Diagnostics, we are very excited that AXELIOS sequencer is coming this summer. And I hope that we can see you at the Diagnostics IR Day because we will certainly highlight this there. We'll also launch the Elecsys pTau217, the Elecsys IGRA tuberculosis test, cobas HDV, so hepatitis D, the ePlex gastrointestinal panel and so on. So we have quite a lot of launches and exciting things ahead for this year as well.

Now this slide I've shown over the last couple of years. And what you can see is that we have shown consistent strong growth over the last quarters. We have post-pandemic average growth of 8% and even in '23 and '24, if you strip out the pandemic sales, we did also have an underlying growth of 8%. So we continue to have a good performance on the sales side. and we will continue to drive this as we go into next quarters as well.

Now let's look at the numbers in detail. At the group. I mentioned that we grew at 7%. This was driven by Pharma with 7%, Diagnostics grew 3%. And as mentioned, this is due to the health care pricing reforms in China and excluding China Diagnostics was at plus 5%. Overall, the impact on the flu season is about 1%. It's about CHF 170 million that I was mentioning before.

Now let's take a closer look behind the drivers as well. First, all segments delivered growth. And now looking at it one by one, I will start with the oncology strong Phesgo conversion from Perjeta continues. The HER2 franchise grew 2.5% in the first quarter. And we do expect the total two franchise to peak this year with a strong tail remaining. Tecentriq returned to single-digit growth, and this was really driven by new indications. Alecensa had an exceptionally strong Q1 with double-digit growth. We are expecting overall low single-digit growth for full year 2026.

In hematology, Polivy sees continued strong uptake in first-line DLBCL and Columvi and Lunsumio continue to grow in later lines of DLBCL and follicular lymphoma. Hemlibra saw a very strong Q1 with market share gains and volumes.

On the neurology side, Ocrevus keeps growing nicely by the strong uptake of the subcutaneous formulation. I know Teresa will go into more details here as well. And Evrysdi is expanding its global leadership position in SMA globally.

On the immunology side, Xolair continues with a strong uptake in food allergy. We're expecting strong growth for the full year 2026 despite the first biosimilar launch in half year -- in the second half year. This will be the first, and there will be no other biosimilars that will come in for the next 2 years after that. Gazyva, we are still very much excited about all the positive trials we have. I mentioned 4 positive Phase III in a row. And we're excited about the launches in lupus nephritis, which is ongoing.

On the ophthalmology side, we saw an acceleration and continued market share gains here as well. And then now Teresa will go into details here as well. Diagnostics, strong growth in pathology, but also Core Lab growing despite China health pricing reforms, I think I mentioned all of this already.

Now let me briefly talk about our expanding NVIDIA collaboration with this large-scale AI factory. So we've had a very long-standing collaboration with NVIDIA, over a number of years, and this collaboration really started in 2016 together with NVIDIA, specifically on the DNA sequencing side because given the amount of data that we're going to produce, and specifically also with the chemistry from Stratos that we brought in, demand data needs to be digested and converted and you need really high-performing chips in order to be able to do that. There's no sequencer in the world that actually produces as much data as our sequencer. So we really needed to go now steps. And here, the partnership with in NVIDIA was really critical. And since then, we've also made a number of acquisitions in this space like [ present, ] for example, but also entered into strategic collaborations for drug discovery as well. And we've significantly built up our internal capabilities when it comes to AI as well.

Now with the recent agreement, we are expanding our GPU compute infrastructure. And with that, we have the largest AI factory in the pharma industry. And also, what's important earlier in this month, we joined forces with Anthropic, IBM, Meta and Microsoft to launch a collaborative licensing so on for the development of AI foundational models called Shared AI License Foundations or SAIL. Now Roche Gentech is the only pharma company among the founding members. And you can see we're really pushing forward on the AI side. And why are we doing this? Because we believe that AI will have a disruptive potential across the entire value chain in our company from drug discovery through manufacturing, diagnostics imaging and we're really pressing forward on this with all the great people we have in this space in our company. Now this will allow us to accelerate the development of therapeutics and diagnostics.

Now let's turn to the outlook. We've made substantial progress across all the therapeutic areas at the start of the year. You see a lot of green ticks. There's strong momentum from a very strong year of pipeline readouts in 2025, now also going into 2026 with a number of key milestones that have been achieved.

On the regulatory side, we have just submitted giredestrant in adjuvant breast cancer. We used a priority review voucher and with that, we're expecting a launch later in this year, so still in this year. We've also filed a Gazyva and SLE in the U.S.

On the key readouts in Q1, I would like to highlight on the one hand, giredestrant in the first line, your positive breast cancer persevERA, the study field, but we saw a numerical benefit in curves separating will be presented at ASCO. Again, I want to reiterate that more than 70% of the opportunity is in the adjuvant setting. This is really a smaller patient population. And we still have another trial in the setting ongoing that could still allow us to enter in the first-line setting as well.

Now with fenebrutinib, we have the first and only BTK inhibitor with positive Phase III studies across RMS and PPMS and we will be filing in the coming months. Petrelintide, excited about the clean placebo-like safety profile with double-digit weight loss to be achieved. This makes petrelintide a good candidate to combine it with a very strong GLP-1, the CT-388. The Phase II in this combination will be initiated. And just to say, we are making very rapid progress with the CT-388 molecule. We've just also initiated another 2 Phase III studies here.

Looking forward, we still have six pivotal Phase III studies reading out, including two NMEs, divarasib and sefaxersen.

We have a very exciting pipeline. So I do believe that given all the readouts we had in the last couple of years, this will give us a much better view in terms of growth in the next coming years. And then between '27 and 2030, we have up to 19 NMEs that could launch. And as you can see, many of these NMEs have quite significant sales potential attached to them. And we wouldn't stop there.

We will continue to do BD activities. We'll continue to try to speed up more molecules so that we can even increase this number. So I think this will really lay the foundation than also for growth in the next decade, which makes us confident that we can continue to have a good growth momentum.

Now let me reiterate a couple giredestrant, you will see already on the market at the end of the year. There are two more NMEs that we'll read out this year. In addition with divarasib and sefaxersen. In addition, we expect two NMEs to be filed in the coming months. One is fenebrutinib and the other one is vamikibart. And again, the implication is that we don't only have a good short to midterm outlook, but also a very solid base to deliver long-lasting future growth as well.

On the Diagnostics side, we have outlined several platforms. On the left-hand side, these are all CHF 1 billion-plus opportunities. So in diagnostic terms, blockbuster opportunities. And very, very excited about sequencer coming just -- and is just around the corner. And we will show it to you at DIA day. Here with the sequencer, we have clear differentiation on speed and accuracy, specifically also on cost because you don't have to have a full plate to achieve lower costs if you compare that with some other leading companies in this space.

On the test side, and this is just a small selection of the test that we're launching. But you can also see the complementary to the pharma portfolio with Elecsys NfL in combination with multiple sclerosis, pTau217 with Alzheimer's. We really have a good strategy to combine and bring together Diagnostics and Pharma. And I know Matt will cover more on what's coming your way on these slides as well.

Now let me wrap up with the guidance. The guidance is confirmed at mid-single-digit sales growth and high single-digit core EPS growth and further increase in dividend in Swiss francs.

With that, very happy to hand over to Alan.

Yes. Thanks, Thomas. Sales call today. So just a brief comment on sales itself, currency and then once again confirming the guidance. When we look at sales, yes, let me start really with the big frame. I start on the left-hand side, you compare to the right-hand side, the sales on Q1 2026, I think minus 5% when you look at Swiss francs. You see on the right-hand side, this big red bar with minus CHF 1.6 billion, that is really the currency impact from the conversion. Then you shift a little bit to the left and then you see the plus 6% in CER, and I will go into currencies on the next slide. But let me talk very quickly about the plus 6% at CER. You see it split down on Pharma with a plus CHF 842 million, excluding the loss of exclusivity product and then loss of exclusivity, minus CHF 63 million. When you put the two bars together, you get to the plus 7% in CER and the minus CHF 63 million, I think, is a small number. You've heard we are expecting a lot of exclusivity impact on sales of roughly CHF 1 billion negative. We stick to that. Certainly, it was a good start. Actemra helped us, MabThera helped us, but we think these two products will slow down really for the rest of the year and that I think makes the CHF 1 billion still doable, if you like, and still realistic.

When we go to the right-hand side, Diagnostics side, Thomas framed it well, and I'm sure Matt will talk a lot about it. Diagnostics was a plus CHF 148 million excluding China. That is representing the 5% in CER. And then you see the China health care pricing reform is minus [ CHF 60 million ]. When you put the two together, you get to the plus 3% in CER, and there has been also an effect from respiratory.

Good. Now as promised, let's talk about currency, and that's the next slide. And here you see it, and it's very clear that the Swiss franc has appreciated basically against all currencies. So you see on the left-hand side, the CER growth was plus 6%. And you see on the right-hand side, the minus 5%, respectively, the minus 4.7% in Swiss francs. You see really, I think that the major driver here is the U.S. dollar. You might have noticed that we have used the U.S. dollar now carefully to bring that in. when you look at the plus 6% in CER and the minus 5% in Swiss francs, if we were converting to U.S. dollar, we would have grown with 9% which I think supports the constant currency growth, clearly.

Good. With that, let's go to the outlook here. And that is certainly always challenging because we're always assuming that currency rates remain stable until the end of the year, and that makes this outlook, how should I say, not very probable. Nevertheless, we want to give you an orientation. So first of all, as shown on the last slide, significant impact in Q1 with the minus 11 percentage points on sales. When we assume that the currency rates at the end of March remains stable until the end of the year, then I think really you see what you had to see at half year with the minus 7 percentage points on sales and the minus 10 percentage points on cooperating profit and core EPS. When you look really at the full year, with a minus 4 percentage points on sales and minus 6% on core operating profit and core EPS, respectively. If you were taking today's rates, I would argue that's broadly in line with what you see in the full year projection.

So let me remind everybody, I think last year in Q1, we had a positive impact from currencies. So I think the volatility is enormous. Why do I say this? Because the projections for full year, we will see what we really end up with. I also respect the point. It's very hard to follow our numbers with this volatility that we're going to see knowing that we have roughly 50% of our sales in the U.S. dollar.

I don't want to dwell too much about the guidance. I think guidance is confirmed. We are very well on track here to achieve the guidance let me reiterate once again that the loss of exclusivity impact of roughly CHF 1 billion is still expected for the full year 2026. And that gives me the pleasure to hand over to Teresa.

Great. Thank you very much, Alan. Okay. Let's get started. So as you have now heard a couple of times, pharma sales grew by 7% at constant exchange rates, reaching CHF 11.5 billion in Q1. We saw a strong performance in the U.S., international and Japan with the latter two achieving double-digit performance growth. The EU was impacted by a number of pricing and one-off effects for specific products, which I'll mention more about in the following slides. But overall, pharma volumes were up by an impressive 17%.

My standard comment on this graph, this graph has all absolute values and year-over-year growth rates presented at Constant Exchange rates in the first quarter of the year. Our top brand, Xolair, Phesgo, Hemlibra, Vabysmo, Ocrevus and Polivy generated roughly CHF 800 million in new sales. We are going to cover the growth dynamics for all of these brands in following slides. But let me pause here for just a quick moment and talk a little bit about Xofluza. So as a reminder, we had a very strong flu season and the for very strong Xofluza sales in Q1 and Q4 of 2025. As I mentioned at our full year call earlier this year, we did anticipate a decrease in flu-related sales in China for Q1 relative to what we saw in Q1 of last year. And that is exactly what we saw materialize and you can see that here in the graph with decline in Xofluza. So that was fully attributable to just a weaker flu season.

So now let's continue by having a closer look at our key TAs. And as always, we will start with oncology. So oncology sales increased by 3% to CHF 3.7 billion. Phesgo continues to deliver strong growth with the global conversion rate climbing to 55%. As previously shared, we are aiming for at least a 60% conversion rate at peak. For Kadcyla, we see increasing competitive pressure in both the U.S. and EU in line with our expectations. Additionally, Q1 performance in the U.S. was negatively impacted by some purchasing patterns. As Itovebi launch momentum continues to be strong. And before we move off of breast cancer, let's turn to giredestrant.

So a lot happened with giredestrant in Q1. So we'll take a little time here to review the program in some detail. So unfortunately, the persevERA study in the first line read out negatively, though as Thomas mentioned, we did see a numerical improvement in PFS. We will present those results at ASCO in June, including at a separate IR event.

At the positive side, we successfully submitted the U.S. filing for lidERA in adjuvant ER-positive HER2-negative breast cancer. And again, as you heard Thomas mentioned, for this filing, we used a priority review voucher which should speed up the regulatory process significantly, enabling approval by the end of the year.

We previously shared that we had filed evERA, our positive Phase III in post-CDK ER-positive HER2-negative metastatic breast cancer in the U.S. and the PDUFA has now been set for the 18th of December.

Let me also remind you of our perspective on the commercial opportunity for giredestrant because it seems that post persevERA, there has been some confusion here. So let's start by taking a big step back and reminding ourselves of the composition of the overall breast cancer market.

So hormone receptor positive HER2 negative breast cancer accounts for about 70% of breast cancer patients. So just to put that in a little bit of context, HER2 accounts for only 15% of patients. So we're already talking about a much larger patient population. Let's translate that into the actual commercial opportunity. We believe that the total SERD opportunity across all lines of therapy is somewhere between USD 20 million and USD 30 billion. Within that ER-positive population, adjuvant ER-positive breast cancer is by far the largest piece with more drug traded products than patients than in first-line metastatic and a much longer treatment duration, so around 5 years. And again, I think this has been really sort of misrepresented in some of the coverage post persevERA. And of course, while we had hoped for a positive persevERA trial, it is really important to remember that the first-line metastatic breast cancer setting only represented around 10% of the overall giredestrant opportunity. Adjuvant is by far the larger market, and we believe there, our data is both strong and compelling.

With the positive lidERA and evERA trials, we believe we have captured around 80% of the overall SERD opportunity with lidERA accounting for roughly 70% and evERA for another 10%.

Let me also remind you that we have one additional Phase III and first-line endocrine-resistant patients, pionERA, which reads out next year. Now this represents 40% of the first-line population. So still a significant portion of that first-line population. Giredestrant is combined with the physician's choice of CDK4/6. And in this study, we expect 40% of patients are going to carry an ESR1 resistance mutation compared to just the 5% that we saw in persevERA.

So overall, we see ourselves very well positioned to capture a meaningful share of this significant market giving us confidence in the overall commercial potential on giredestrant. You saw in Thomas' slide peak sales well north of CHF 3 billion. And I think, in fact, you've heard us say before, this could be our largest selling product, and that certainly does seem to still appear to be very possible. And with that, we invite you to join us at ASCO to learn more.

So now let's shift to Alecensa. Alecensa had a very strong Q1, partly driven by buying patterns in the international region. We did signal at our last call that due to increasing competitive pressure, especially in the U.S. and EU, we do expect low to single-digit or low single-digit growth for the full year for Alecensa, and that continues to be the case.

Moving on to Tecentriq, growth is driven by our new indications, in particular, IMforte in small cell. Additional new indications like IMvigor011 in muscle invasive bladder cancer and ATOMIC in and dMMR colon cancer are expected to help carry this growth momentum forward.

Let me mention here that we saw one-off buying pattern that negatively affected EU performance in Q1, and we do expect that, that growth is going to gradually recover over the next quarters.

With our performance in this quarter, I also want to confirm our full year 2026 outlook for low single-digit growth for Tecentriq.

And finally, looking forward, as Thomas mentioned, we expect Phase III readouts for Itovebi as well as the first Phase III readout for KRASG12C inhibitor divrasib later this year.

Now let's move on to hematology. The hematology franchise delivered strong growth of 18% at Constant Exchange rates, achieving CHF 2.2 billion in sales. Hemlibra continued strong global growth momentum from last year, driven by increasing adoption in non-inhibitor patients. Of note, the U.S. performance benefited from a base effect due to buying patterns, which negatively impacted Q1 sales last year. As we have previously shared, we expect single low-digit growth for 2025, and this is driven by the anticipated headwinds from competitor launches later this year.

Let's stay briefly with hemophilia A and take a look at the latest NXT007 developments. We have initiated the Phase III trials in ZEBRHA 1 and 2, which compared NXT007 2 Factor VIII in Hemlibra, respectively. And for both trials, we are using an auto-injector from day 1 for enhanced administration convenience for patients.

Moving on to malignant heme, Polivy's uptake in first-line DLBCL continues to grow steadily. We're now at 39% of U.S. patient share with more than 95,000 patients treated globally. Please note that the EU Q1 performance here was negatively impacted by a base effect due to the release of sales accruals in Germany in Q1 of last year.

Overall, Gazyva delivered solid growth in Q1, but its performance does bear a little bit of explanation given our launch in immunology and the delay in the data that allows us to tease out performance between the two therapeutic areas. So what is it that we see happening. In oncology, because Gazyva sales in the U.S. and EU were negatively impacted by the launch of Venclexta and acalabrutinib combination in first-line CLL, which increased competitive pressure on the Venclexta Gazyva regimen. We believe this dynamic is masking the uptake that we see in immunology, which I'll discuss in just a little bit. We are eager to show Gazyva sales by TA, we're actively working on getting that data, and we'll be able to have more visibility into the immunology uptake in the future. So again, more to come in a few slides.

And finally, rounding out heme, our CD20, CD30 bispecifics, Columvi and Lunsumio continue to seek good uptake in their respective indications. Importantly, we completed the U.S. filing of Lunsumio in second-line plus DLBCL based on the positive SUNMO results, and we expect the readout for the Phase III CELESTIMO study of Lunsumio in second line plus follicular later this year.

So now let's move on to neurology. Neurology started the year strong with 10% growth at Constant exchange rates, reaching CHF 2.4 billion in sales. For Ocrevus, we see continued good growth driven by the subcutaneous formulation, which is known as Zunovo in the U.S. As anticipated, subcu furthered its acceleration, and you can see in the global patient count, which has increased roughly 24,000, this represents an increase of 7,000 patients compared to last quarter, and that's almost 2,000 more patients than the increase we saw in Q4. So that acceleration is starting to happen.

As previously shared, the U.S. Zunovo uptake is a very good indicator of how we're expanding the addressable market for MS, roughly 60% of the Zunovo volume is coming from community practices and roughly half of patient starts are now new to the Ocrevus brand. One more comment on the U.S. performance. Q1 sales were negatively impacted by some payer dynamics, which we often see in Q1, largely around recertifications and lower sales days relative to Q1 2025. All of that having been said, for 2026, we continue to expect high single-digit to low double-digit growth for Ocrevus as a whole.

So a reminder that we upgraded our peak sales expectations for Ocrevus franchise for CHF 9 billion by 2029, and that includes CHF 2 billion incremental sales from Ocrevus subcu.

So more key updates for the MS franchise. We are very pleased to report a positive outcome for the Phase III FENhance 1 study of fenebrutinib in RMS. And just a few days ago, the full results for FENhance 1 and 2. We're presented at AAN, and I will share those highlights briefly on the next slide, but let's round out quickly with the other products in our neurology portfolio.

Starting with Evrysdi off to a very strong start to the year, partially boosted by tender-related buying in international. And we continue to see Evrysdi expanding its strong global position, thanks to the global rollout of the tablet formulation.

Moving on to Elevidys and DMD. We continue to believe strongly in the positive risk-benefit profile in the ambulatory population. And therefore, we have made the decision to initiate a new Phase III trial to enable regulatory resubmission in the EU and other regions globally. This new trial ONCORE includes a placebo-controlled arm and just focused on early ambulatory patients. We're confident that this is the most viable path to achieve EU approval and that it will allow us to bring this new treatment option to patients.

And another highlight of the quarter is the positive III Enspryng data in MOGAD, which was presented just a few days ago at AAN, and I'll cover that in more detail.

So now let's move on to fenebrutinib. So with the positive readout of FENhance 1, we now have 3 positive Phase III trials for fenebrutinib in MS, FENhance 1 and 2 and RMS as well as FENtrepid and PPMS. I will take you through the details of the RMS results in a moment, but let me make one thing really clear up fine. We are confident that fenebrutinib has the potential to become the first and only high-efficacy oral treatment for both RMS and PPMS. And we also reaffirm our belief in the potential peak sales of more than CHF 3 billion.

Now let's provide some background on where this conviction comes from, and we'll start with RMS results. So fenebrutinib met the primary endpoint, showing significant reduction of relapses, 51% in FENhance 1, 59% in FENhance 2 compared to teriflunomide. Together, these results translate into one relapse approximately every 17 years. Additionally, when looking at key secondary endpoints for disability progression, there is a constant trend favoring fenebrutinib. This is true for cCDP12 as well as the modified cCDP12, which is focused on EDS to the 9-Hole Peg Test, these findings further support fenebrutinib positive impact on progressive biology as seen in the fenebrutinib trial for PPMS.

Let me also cover the safety profile of fenebrutinib. The percentage of patients with liver enzymes in FENhance 1and 2 is lower than what we saw in fenebrutinib trial and is broadly similar to what we saw in teriflunomide. There was one Hy's law case in each of the fenebrutinib and teriflunomide arm of FENhance 1. Both cases were asymptomatic and resolved after study drug discontinuation. There were no additional Hy's law cases across all the other fenebrutinib trials in MS or any of the autoimmune indications where it has been studied after implementation of liver monitoring every 2 weeks during the first 20 weeks of treatment.

Lastly, to mention, we did see an imbalance in fatal AEs across the fenebrutinib arms of 1 and 2 as well as in FENtrepid. Various causes and time points were observed for those fatal AEs as were presented and discussed at AAN in more detail.

Again, let me be very clear. Patient safety is our #1 priority throughout the study. There was a continuous exchange with the FDA and the IDMC, including on AEs. And we are confident in the favorable risk-benefit profile of fenebrutinib in RMS and PPMS, and the totality of data from those Phase III trials will be submitted to regular authorities in the coming months.

So now let's move on to Enspryng. As previously mentioned, we're quite excited about the positive Phase III results for Enspryng and MOGAD. I'm guessing that not everyone is familiar with MOGAD or Myelin Oligodendrocyte Glycoprotein Antibody-Associated Disease. There will be a test on that later. So let me share some background. MOGAD is a demyelinating rare autoimmune disease that can cause severe neurological disability in children and adults. This includes loss of vision, cognitive dysfunction and loss of ambulation. There are currently no approved therapies and disease management frequently relies on acute treatment with high-dose corticosteroids, primarily aimed at reducing relapses. IL-6 signaling is implicated in pathophysiology of MOGAD and so therefore, Enspryng, our anti-IL-6 antibody sort of makes perfect sense in this disease. As I mentioned, full year, we believe MOGAD represents at least CHF 500 million peak sales opportunity for Enspryng, and now let's look at the data and what has us so excited. So here it is, what the METEOROID study just presented a few days ago for the primary endpoint, the portion of patients that are relapse-free, Enspryng achieved a strong risk reduction of 68% versus placebo. Patients demonstrated a response as early as 8 weeks and Enspryng achieved a significant reduction in annualized relapse rates, active MRI lesions and rescue therapy. Safety was comparable to placebo nicely rounding out these results, which have been extremely well received at AAN.

I think you can see why we and the KOL community are so excited for Enspryng and MOGAD and the opportunity that we have to bring a very meaningful benefit to patients. We expect to complete filing in the U.S. and EU for this indication later this year. And so with that, let's switch over to immunology.

Our immunology franchise grew at 8% at Constant exchange rates, it's CHF 1.5 billion in sales. The key growth driver here is Xolair, which continues to show exceptionally strong uptake in food allergy as well as continued growth in CSU. In terms of 2026 outlook, we continue to expect around 20% growth for Xolair this year, and this includes the expected impact of the first biosimilar entering the market in the second half of 2026. October sales declined by 4% in Q1, driven by biosimilar impact in the U.S. and EU.

Now again, there is a lot of positive news flow around Gazyva, so we're going to take this step by step. I'm going to go into the details of SLE, MN and INS on the next slide. whoever here, again, I want to briefly comment on the ongoing launch of lupus nephritis. I mentioned earlier, we are looking at splitting out the data between immunology and hematology sales. Once that happens, you will see because I have a begin to appear on this slide. In the meantime, let me share with you some qualitative insights on how the launch is going. Across early launch countries in the U.S. and EU, we see positive feedback from doctors and patients, including awareness and treatment satisfaction. And what we're hearing in the field also gives us good confidence that Gazyva is off to a strong start in immunology. For example, a doctor in the U.K. as remarked that there's no reason to just use MMF and steroids anymore, Gazyva is clearly a better option. So safe to say, I continue to be excited about the potential for Gazyva in immunology, and we remain confident that this is up to a CHF 2 billion opportunity across all of the immunology indications.

And finally, I'd like to mention the upcoming Phase III readout for sefaxersen and IgAN which is now expected for later in the year. So let's take a look at Gazyva in a little bit more detail. Gazyva is on a tremendous round achieving 4 out of 4 positive Phase III trials in immunology since we reported the first results back in 2024 in MN. The positive Phase III MAJESTY and membranous nephropathy -- but we're going to talk about more of that in a minute. First, I want to highlight the strong Phase III results from ALLEGORY and SLE, which were just recently presented at SLA [ Euro ]. You can see on the left, the primary endpoint and key secondary endpoints all show a clear benefit for GAZYVA treatment relative to placebo. And on the back of these results, we believe Gazyva has the potential to be the new standard of care in SLE. Again, when you look across these four indications, you can clearly see why we believe in the potential for that CHF 2 billion CSF upside opportunity comes from. You can also see that we are progressing our filing activities at pace. LN is already approved and the U.S. and EU launches are ongoing. SLE has been filed in the U.S. and the EU and the U.S. PDUFA has for the fourth of December. MN will be filed in the U.S. and EU later this year, and INS has been filed with the U.S. and EU filing is expected later this year. So a ton of momentum for Gazyva in immunology, and we are excited to update you.

So let's move on to ophthalmology. Ophthalmology grew by 10%, achieving CHF 1.1 billion in sales. The Vabysmo started in -- Vabysmo this year with a strong quarter with 13% growth at Constant exchange rates. While we continue to see global market share gains across countries, I did want to highlight the U.S. performance here in particular. Q1 marks a return to growth in the U.S. In fact, we see 4% sales growth with low double-digit volume growth and a steady market share expansion. Vabysmo continues to establish itself as a standard of care across AMD, DME and RVO as underscored by the fact that roughly 60% of Vabysmo starts are treatment naive. Regarding the branded market contraction we saw in the U.S. over the last quarters. We are now beginning to see the first signs of recovery, but for now, they are only the first signs. We continue to expect to see a gradual recovery of the branded market going forward. And last but not least, we are excited to have received an updated FDA label for the RVO indication, and that update adds the flexibility for treatment beyond 6 months based on long-term data for BALATON and COMINOstudies.

Let me also say a few quick words on the EU performance in Q1. In Q1, we had to digest a negative impact due to price impacts, but we saw quite good volume growth in the EU. So therefore, we do expect a return to growth in Europe in the coming quarters.

In 2026, the outlook for Vabysmo is a growth acceleration compared to the 2025 growth rate of 12%, which is what we have messaged earlier. Let me highlight that we expect to file 2 new potential medicines in our ophthalmology franchise later this year, the vamikibart in UME and Enspryng in thyroid eye disease.

So up next, let's cover the CVRM portfolio. We continue to progress our CVRM assets at pace, and there are two highlights for Q1 that I want to discuss in more detail. Let's start with petrelintide in Q1, together with our partner Zealand, we shared the positive results of the Phase II ZUPREME-1 study of petrelintide in obese patients without type 2. These results showed that petrelintide delivered meaningful double-digit weight loss. Importantly, this weight loss was accompanied by a placebo-like tolerability profile. So therefore, we believe in the potential of petrelintide to address key unmet needs for people living with overweight or obesity, namely to meet their expectations for weight loss and significantly limited -- life-limiting adverse events, thereby improving treatment persistence.

Moving on to the newly named enicepatide, formerly known as CT-388. At the full year event, we showed the positive Phase II top line results for week 48 for the once-weekly enicepatide in people with obesity. That was study 103. Let me just remind you the highlights from that data using the efficacy as demand. Enicepatide demand achieved a placebo-adjusted weight loss of 22.5%. We saw a clear dose response relationship with weight loss. And importantly, we are pleased by the absence of the physical efficacy plateau at 48 weeks for the 24 mg dose, which is the highest dose tested.

On the back of this strong Phase II data, we have successfully initiated 2 Phase III trials in obesity, it 1 and 2 in Q1. For both petrelintide and enicepatide, we are looking forward to presenting the results at ADA in June, including at a separate IR event.

In regards to what's to come in 2026, you will see we have quite a few milestones. So we expect additional Phase II readouts for enicepatide in obesity with type 2 and CT-996, our oral GLP-1 and obesity and finally, as data, GYMINDA for emugrobart enicepatide in obesity. We also expect multiple trial starts. We plan to initiate a Phase II combination study for enicepatide and petrelintide with the FPI expected towards midyear and the first Phase III studies for petrelintide plus CT-996 or it's petre and CT-996. As you can see, there continues to be a lot happening in our CVRM portfolio, and we will be sharing additional updates throughout the year.

So moving on to the news flow slide. So here as we have what is happening this year. I believe I've covered all of the changes shown previously with one exception. Earlier this quarter, we shared that we made the decision to discontinue development of emugrobart in SMA and FSHD. This follows a rigorous assessment of data from the Phase II studies, MANATEE and MANOEUVRE. We plan to present these data at upcoming conferences. And just to be clear, there is no impact from this decision on the ongoing Phase II obesity trial reading out towards year-end.

So let's go to my final slide. To close my section today, I wanted to walk you through an update of our well-known consensus GAAP analysis slide. As we've recently had a series of positive readouts representing significant commercial opportunity. To briefly orient ourselves, the value shown here are now based on the post full year 2025 consensus numbers from sell-side analyst models. And we've pushed out the time frame by 1 year to look at 2025 to 2030. According to consensus, Roche has a biosimilar gap of CHF 6.7 billion by 2030. However, if you look at the middle section of the slide, you can see where consensus sees sales growth coming from during the same time period. Taking the on-market portfolio and the Phase III pipeline together, consensus expects a total of CHF 5.9 billion -- sorry, CHF 15.9 billion and additional sales by 2030, which compares quite favorably to that projected biosimilar gap. If you take a closer look, these additional sales come from the on-market portfolio where future growth is expected to deliver an additional CHF 7.5 billion in sales with Vabysmo and Itovebi being at the top of a long list of growth contributors. And for the late-stage Phase III pipeline, the consensus forecast a total of CHF 8.4 billion in additional sales with giredestrant, fenebrutinib and enicepatide all projected to be post growth drivers, where most of our other Phase III assets all have rather modest sales projections at this point.

In addition, there are still a number of other assets in our pipeline that are not really covered yet by the consensus, presenting potential upside. These are on the right side of the slide, including Enspryng in MOGAD with the compelling data we just talked about today. Enspryng in TED to be filed later this year, or Gazyva in SLE again with very compelling data that has just been released. So in summary, if you remember Thomas' slide from earlier, we have the potential to launch 19 NMEs by 2030. There's, of course, also the possibility that BD might contribute more over time. And I hope this makes you at least as excited as I am for the future of Roche Pharma. We have a lot to look forward to. And with that, I will turn it over to Matt.

All right. Congratulations, Teresa. With that, good morning, good afternoon, everyone. It's my pleasure to present the Diagnostics financial results for the first quarter of 2026. So with sales of CHF 3.3 billion, the Diagnostics division grew sales by 3% or plus CHF 88 million compared with 2025 at Constant exchange. So this growth was achieved despite the ongoing impact of health care price reform in China, which continued their impact in 2026.

Now as you heard earlier from Alan and as well from Thomas, excluding China, the growth of the business was plus 5% and was additionally impacted by approximately CHF 40 million lower sales due to the weak Northern Hemisphere respiratory season.

Now with that, let me walk you through sales by customer area. Sales in our Core Lab increased at 4%. Now this was impacted by the previously mentioned policy impact from China. Excluding this, Core Lab grew at plus 8%. Now sales in Molecular Lab were flat, with strong growth in transplant at plus 15%, but this was offset by the previously mentioned mild respiratory season which lowered testing volumes. Now while blood screening decreased at minus 9%, this is due to the ongoing conflict in the Middle East, and this impacted some customer deliveries.

Sales in near patient care decreased at minus 5%. This was driven by lower [indiscernible] sales, again due to the mild respiratory season, but this was partially offset by growth in our blood glucose monitoring at plus 8% following some recent competitive wins.

Sales in pathology lab grew at plus 12%, mainly driven by advanced staining growth of plus 9% and companion diagnostics growth of plus 23%.

So now I'd like to shift and take you through the regional view. I'll take you through the performance by geography. In North America, the business grew at plus 6%. In EMEA, the business grew at plus 3%. As previously mentioned, this growth was impacted by the decrease in the blood screening business due to the Middle East conflict, in Lat Am, the business grew at plus 10%. In APAC, the business declined at minus 5%. Now as previously mentioned, sales growth was impacted by the health care pricing reform in China. As a result of this, China sales declined at minus 14%. Now as stated at the full year, we expect a lessened impact of the China pricing reform in 2026, and our ambition is to grow sales in the Diagnostics division at mid-single digits this year.

So now I'd like to continue with some updates to our assay pipeline, starting with the CE mark for our cobas MPX-E test which we announced in March. The blood screening nucleic acid testing market represents approximately CHF 800 million market globally and screening for hepatitis E or HEV is growing in APAC as well as EMEA due to increasing incidents. With the cobas MPX-E, we will enable labs to efficiently integrate HEV screening into their existing workflow. This will address a growing demand for comprehensive blood donation safety worldwide. Our test will improve lab efficiency by simultaneously detecting HIV, HCV, HPV and HEV from a single sample, and this will eliminate the need for retesting and reduce turnaround times, again, helping patients around the world have access to blood supply. This test will expand Roche's fully automated blood safety solution portfolio and strengthen our leading position being the only company with a full nucleic acid and serology testing solution in our portfolio.

So now I'd like to move to our leading neurology portfolio, starting with the CE mark of our blood-based neurofilament light chain assay. So as you heard quite a bit from Teresa, MS is a progressive disease, affecting around 2.8 million people worldwide. And about 45% of patients are currently managed with low efficacy treatments, and disease monitoring relies heavily on expensive and resource-intensive MRI scans. Our NFL test will enable detection of neuroinflammation in patients with relapsing remitting MS, which will support more informed treatment decisions. And what specific and differentiated about our test is we have established age-specific NFL percentile from a comprehensive reference data set. And this is critical for clinical interpretation because NFL values increase with age. And this provides a more robust approach as opposed to other approved tests that rely on a single cut off. This test will broaden Roche Diagnostics neurology portfolio with a simple, reliable and accessible test.

So continuing with our neurology portfolio, I'd like to talk a little bit about the data from our pTau217 blood-based biomarker study readout. So this is data that was presented at the ADPD Conference in March. And here, we assessed our blood-based test across primary, secondary and tertiary care settings. Evaluating its performance throughout the full Alzheimer's disease continuum in 2 cohorts with a total of 675 participants. Our test demonstrated high diagnostics accuracy consistently across all clinical stages and health care settings. And it will be the first pTau217 test with the potential to be approved across primary and secondary settings. Again, differentiating our neurology portfolio through the clinical evidence. This test will enable rapid, minimally invasive diagnosis across care settings, allowing clinicians to identify Alzheimer's disease earlier.

So now you heard a bit about this from Thomas, and now I would like to talk a little bit our forthcoming AXELIOS sequencing solution. Talk a little bit about some of the data we presented at AGBT, which is the leading genomics conference in February of this year.

On the left-hand side of the page, you can see the SBX-Duplex performance data from a large internal study of over 1,000 human genomes that were processed over 65 runs. Here, we demonstrated unprecedented throughput. As you can see from the chart, the systems baseline is very robust, reliably generated an average of 2.3 trillion bases of concordant duplex basis in 4 hours. So as you heard earlier, one of the key differentiators of our system will be the unprecedated throughput, unprecedented throughput. Now on the right side, you can see data presented by one of our early evaluators the Hartwig Medical Foundation. They compared performance of our SBX technology with a leading on-market technology for 118 tumor-normal pairs. The results demonstrated high throughput and improved batch flexibility per run and 99% concordance with the leading technology. These results reinforce the potential of Roche's SBX technology set a new standard in next-generation sequencing, unprecedented throughput and accuracy on par with leading technologies.

And I would also like to highlight a little bit of our recent acquisition or excuse me, our merger agreement with SAGA Diagnostics. So SAGA Diagnostics Pathlight tumor-informed MRD or minimal residual disease testing platform is currently available for early breast cancer and colorectal cancer. It relies on DNA sequencing followed by digital PCR assessment of structural variants for monitoring disease relapse. Pathlight's tumor-informed ultrasensitive MRD platform will be fully integrated into our Foundation Medicine organization. Foundation Medicine, we -- using our colleagues at Foundation Medicine team, we also plan to leverage the AXELIOS and our Roche Digital LightCycler PCR platform to develop a decentralized MRD solution, enabling international expansion and improved patient access. This minimal residual disease market is projected to grow at 31% compound annual growth rate, making it one of the fastest-growing segments in diagnostics overall. This acquisition will strengthen our leading oncology portfolio and support our ambition to build an end-to-end offering, spanning early detection, diagnosis, therapy selection and now disease monitoring, powered by our core technologies such as AXELIOS.

So to conclude, I'd like to report on progress of our key launch list for the Diagnostics division. For the 11 launches shown here, which you heard a little bit about earlier, we achieved 3 by Q1, and we're making good progress towards the other launches, which are on track. I look forward to providing future updates at our next meeting.

And now I'd like to hand it over to Bruno. Thank you.

Thank you, Matt. And just quickly to point out upcoming IR events. Next events, which are scheduled now will be the Diagnostics Day on May 12. This will be, again, a live event in London. We will have Thomas opening the day with a group and the strategy update. And then we, again, will take you through the entire portfolio. And as Matt already mentioned, sequencing will be, I think, in the center.

We have then two other events scheduled beginning of June. We will have an ASCO event, which will be live out of Chicago. So if you want to join us, you can come along. On the second of June, we will here primarily be focused on giredestrant, the key data again and also a bit the clinical future development program. And then on June 8, we will have an event around ADA, which will focus on the obesity franchise. Key data to presented will be our 388 Phase II data in petrelintide. And then also here an update on the next steps in terms of our combination development where we expect the Phase II to start around midyear.

And then I think, finally, in the second half, Pharma Day is now scheduled for the 28th of September. Again, a live event in London. We will have a plan for a similar setup like last year, taking you through pharma strategy then also the primarily late-stage pipeline and the portfolio. We also have included an update on our AI investments, especially focusing on early drug development here as well.

With that, I think we are closed with the presentations and would open the Q&A. First questions here would go to James Gordon from Barclays. James, please.

James Gordon from Barclays. First question was about giredestrant. So had failure of the persevERA trial. Is that at all surprising in light of lidERA? and how do you reconcile the differences in the outcomes? Do you think it could be anything to do with whether there's background CDK4/6 therapies? Or do you think it might be more about prior patient experience to hormonal therapies in the persevERA trial but not the lidERA trial. That will be the first question, please.

And then the second one would just be about obesity. So there's lots of different things going on in your pipeline. But we've seen some very low-cost GLP-1 generics launched in some places already such as in India. How do you think about obesity price in the next decade and what that might mean for your next-generation therapies is by the time you launch, there could be some or shortly after some much lower priced therapies. Where do you think a drug like petrelintide will be used and could pricing going to be tougher?

Great. Thanks, James. So let's start with persevERA versus lidERA. So I mean I think one of the most important things to always remember in oncology trials is that the context and the biology of the line of therapy really does matter a lot. Tumor biology differs across lines of treatments, especially in early breast cancer versus later-line setting. In EBC, tumor burden is low versus the metastatic setting, the tumor burden increases, the environment becomes more genetically complex, resistance mechanisms are more prevalent sort of the later line in therapy that you go. And so -- it's really important that we evaluated giredestrant or giredestrant in all of these different lines and settings in multiple different ways because this is really -- there's differences in the tumor biology. Although the study didn't reach its statistical significance in PFS, a numerical improvement was observed. So we do believe that giredestrant is active in the first-line metastatic setting. And we are confident in its potential certainly in early breast cancer, and I think we're very curious what we'll see in pionERA, which is the -- persevERA covered 60% of the first-line setting pionERA covers the 40% remaining. And I think we're very curious what we'll see in the pionERA setting. So there isn't anything about persevERA that changes our confidence and our belief in the importance of giredestrant in early breast cancer, again, where 70% of the opportunity sort of sits. So very yes, just to be super clear about that.

In terms of obesity pricing, this was very foreseeable. I mean, I think as we -- when we made the decision to enter this market, we assumed that by the time we got there, there would be significant price erosion. It's not unexpected that in certain parts of the world, you would have very low-priced options. But regardless, I think given the portfolio that we've put together, the clinical benefit think our products will provide, the clinical differentiation that we believe that they'll represent. We do believe that there continues to be a very robust opportunity for us in obesity.

James, did this answer all your questions?

Yes.

Yes. And we move on. Next one is Sachin Jain from Bank of America.

So SERD and BTK. So Teresa, thanks for the introductory comments on. I just had two around lidERA and then one on BTK. So lidERA, how should we think about the cadence of launch as you prepare for sort of commercial towards the end of this year early next year. Do you think this is a market which will require a lot of community oncologists, education versus CDK4/6 AI? Or that the data will be easily understood and we should think about this as a strong launch.

And then secondly, you and Thomas have commented, this could be one of Roche's top drug, which I guess alludes somewhere in the CHF 5 billion to CHF 10 billion range. So just interested in your thoughts as to what factors you see that's going to dictate where in that range you end up. Some of the factors we hear back, would love your perspective on is adoption of low-risk patients, whether you get any combination use of CDK4/6 and then impact from other adjuvant studies, notably the Switch studies that are due next year [indiscernible]. That's lidERA.

And then just a very quick one on BTK. You say you've been in continuous exchange with the FDA, IDMC. Any color you can give on the tone of those discussions. Clearly, part of your [ insurance ] comment, I guess, is referencing safety where investors just seen the death imbalance and I guess, lack of color on the suicides, driving concerns on approvability.

Yes. Great. So in terms of cadence of launch, I mean, we would expect that giredestrant will have a strong launch. I mean I think there's quite a bit of excitement and interest and this is a therapy for patients living with ER-positive HER2-negative breast cancer. I mean, where we expect to have the quickest uptake is in that sort of mid-risk population. And from there, there's quite a reasonable belief that we would then move into the high-risk population, given the talk that you see with the CDK4/6, over time, patients have a difficult time tolerating it. They drop off. We'll have additional data that we will generate in these areas as well in terms of combination therapy. But it is very tough to remain on these regimens over time. And given the safety profile that we saw with giredestrant, we would believe that there will be a number of patients even in that high-risk population that we'll want to make the switch.

Low risk. I mean, again, we hear KOL intent to use in this low-risk population. I mean with -- as with any kind of oncolytic indication, you really want to come out of the gate swing. What is your strongest weapon that you can use against the cancer. And I think the lidERA data very clearly show that we are highly efficacious and very tolerable in this setting. So I think we intend to resource and go after this indication, assuming that it is going to be strong right out of the gates.

In terms of the Switch studies, I mean, I think this is -- it's an interesting question. I mean, the Switch study is assumed that after 3 to 5 years of therapy, you're going to make a switch. Why would you save your biggest gun to last? I mean it doesn't kind of totally compute that that's what people would necessarily want to do. I mean I think at the end of the day, with lidERA, we asked a very clear scientific question. We got a very clear answer. And I think that very clear answer is one that's been quite well received by the KOL community.

I mean in terms -- so switching over to BTK in terms of the tone of discussion, obviously, we don't discuss ongoing communications with regulators. But I would say, as always, we've been very transparent with our data. They've been very productive discussions. And again, we believe that we've shown a very good benefit risk profile. There is no medicine out there that doesn't carry some amount of risk, but we believe that given what we've shown from an efficacy standpoint, we've got a really good treatment for MS on our hands. I mean, let's talk for a minute about suicides and suicidality. I mean, first, I think it's just important to put in context that the MS population, unfortunately, does have a higher risk of suicide and suicidality. And so sort of the backdrop under which you have to sort of consider some of this data, and I think you've heard that from the KOLs who've presented on the podium. These are -- there's not a huge number of these events. We clearly are trying very hard to understand them, but it's likely going to require more data and more time. And I think what's also important to remember is that we didn't see any suicide or suicidality in our Phase II trial, including long-term extension. And we haven't seen any in any of the autoimmune indications that we have studied as well.

So I think there are certainly questions that we have to ask and answer here. Patient safety, as I've said previously, is always our #1 priority. We are looking into this very, very closely. And as we have more information, we will certainly share it.

Yes. Maybe, Sachin, I can add two things here as well. One on giredestrant and on the Switching studies. Well, first of all, just on giredestrant, giredestrant has demonstrated the highest preclinical potency among the SERDs. Second, if you look at Switching studies and people have been pretreated for 3 to 5 years, you are just also in a different setting. So I don't think that the data that we have shown necessarily needs to translate to another SERDs, which is tested in a very different setting and it's a very different molecule. And if you then look at when the next study will come in terms of true adjuvant study, this is a number of years away.

On the BTK, one thing that I can also give as an indicator is that on the PPMS side, we have already achieved the approval of getting all of the patients that happen on the trial into the open-label extension. And not only that, also the people that were in a placebo arm have switched to fenebrutinib. So I think this is at least a good sign if you ask me because it shows that the benefit is seen and yes, these people are allowed to be on this medicine.

Did you mention that fenebrutinib was used in other indications as well?

Yes. It did.

Very good. Let's move on in the queue. The next question go to Peter Verdult from BNP Paribas.

Yes. BNP, just Teresa, a couple of questions on your obesity portfolio. If we look through the lens of your R&D bar strategy about prosecuting best or first-in-class assets. I get it why you're proceeding enicepatide forward. Interested to hear more about the Phase III development plans there. But I don't see how the amylin petre can be considered first or anywhere near best-in-class. So interested to hear more your go-forward strategy there on amylin. And just to check, is there any debate internally at Roche, whether to make that sort of CHF 2 billion commitment to petre for a full Phase III development program?

Sure. So I mean, I think, first, we should talk about the fact that the amylin class is fundamentally different than the GLP, GLP-1 class. It serves a very different role in what we think is going to be a very heterogeneous obesity market. We do see a very different tolerability profile for petre. And so when you think about the over -- when you think about the 50% of the population globally, that is going to be obese by 2030, 2035, there are certain portion of them that are not going to tolerate any kind of adverse event. And petre in combination or in other uses could potentially be really big benefit for those people don't need to loose 20% of their body weight but do need to llose in the low teens, which is exactly where we saw the data come out for petri along with that placebo-like efficacy profile. So I think for us, as we look across the entire portfolio and the kind of offering that we want to bring to the market in terms of having a differentiated holistic portfolio that addresses all of the different needs that patients will have. We think that there is a role for amylin play and a role for petre to play. And so more to come on exactly what our Phase III trials will look like in the coming months. But I think we still continue to believe that there is a place for petre.

And we have announced that we will do a Phase II combination of petrelintide and CT-388. And especially in the combination, I think it's important to have molecules that are tolerable in the combination. And especially in the conation, we do see the potential between the molecule.

So something very different unit.

Okay. Then let's go on. The next questions come from Sarita Kapila, Morgan Stanley.

Just one for me. Ahead of the Q4 data in DME, how should we think about competition to Vabysmo from Merck's MK-3000. So the early data suggested a similar visual acuity improvement at 12 weeks versus 52 weeks for the Vabysmo. And how should we think about potential read or risk to AMD?

So I mean I think the Vabysmo has done an exceptional job over the course of the last couple of years establishing itself as the standard of care in these in retinal disease. And we see that being consistently reinforced in all of the interviews and information that -- and exchanges that we have with our retinal specialists.

So while it's always interesting to see another mechanism of action kind of entered the space. I think Vabysmo has set a high bar and they will have to sort of demonstrate that they can cross over that over the long term with their safety and efficacy data. Again, we also have a win. So we're in this race as well. Ophthalmology is a core area for us, but we do believe that at the end of the day, people will need to see something that works better than Vabysmo. And as of yet, we don't have anything in a Phase III that's proven that.

Very good. Then let's move on. The next questions go to Luisa Hector from Berenberg.

So I have a question around the derisking really of the legacy pipeline and your levels of confidence in those drugs. Has that changed your ambitions in terms of business development, R&D investments and your thoughts around the longer-term sales and margin. Just any color you can give there.

When you refer to sort of legacy pipeline, what are you referring to?

Fene and giredestrant. So hearing your levels of excitement around those assets and the positive data that you have, does that lead you to behave differently and think differently towards future business development?

Yes. So I mean, both fenebrutinib and giredestrant in what we call end-to-end disease areas in our pharma strategy. So MS and breast cancer are two areas that we have said we have a long-term strategic interest being leaders in both today and tomorrow. So I think it doesn't change our interest. We remain deeply interested in both of these spaces. And are going to be constantly on the hunt for opportunities to really improve patient efficacy in both breast cancer and MS.

I mean from an M&A perspective, I can say that we'll continue to do M&A like you've seen in the last couple of years, but we're not dependent on it, right? I mean if you look at the last couple of years, we could actually finance our M&A that we did with the earnings that we had in that year. And I think there are other companies that are in a much more different situation where they actually do have to do M&A. So continue to see what we've done over the last couple of years, very disciplined spending and again, we are in a good position. We will have continued growth even in absence of M&A.

And we don't have a patent clip, correct?

We don't have a patent [ cliff ], correct?

Yes.

Luisa did we answer all your questions?

Yes.

And next one is Graham Parry from Citi.

Just going back to giredestrant at [indiscernible] and 70% of the market opportunity in adjuvant, over CHF 20 million to CHF 30 billion would imply CHF 14 billion to CHF 20 billion opportunity in adjuvant. Just could you clarify how much of that would assume CDK4/6 combo use or erosion? Or how much is just essentially what is currently a monotherapy aromatase inhibitor market, which is aligned with your lidERA study and, therefore, fairly easy to go after.

And then secondly, do you think that you could get a label which would allow aromatherapy switches on launch or just de novo patients per the lidERA study?

And then thirdly, one for Alan. Just the thoughts on the margin trajectory of Roche as you go forward and have this multibillion-dollar in-house developed asset, that's a small molecule at a high price, driving your revenue growth.

Yes. So you have asked 2 sneaky questions. So we will not comment on what we -- on the label negotiations that we're having with the FDA. So more to come on that. Right now, about 20% to 25% of early breast cancer patients get CDK4/6 combination therapy, but there are very high discontinuation rates due to adverse events. And so we are -- we don't provide peak estimates for our products at this stage, but I will say that we would expect that we will have a significant share of that market. I mean it's in the [indiscernible].

Yes. Well, look, I think from the margin point of view, as Thomas always said, I think we defend the margin or increase it. I think these are the combo, so at least, we defend the margin moving forward. I think we've said it also for this year already. I think without increasing the margin, honestly, I think very hard to do and to fulfill the guidance. I think we have a higher burden coming from the tech side. We have a higher burden coming from the financing side that we have to cover. So we're going to do that. So that's automatic. And I think really now distinguishing what's a small molecule or whatever mechanism it is, when you look at the longer term of Roche, I think we've been pretty, pretty solid when it comes to the gross margin, especially on the pharma side, roughly 80%. That's something that we always had somehow here, was it small molecules or even other mechanisms. So I think that looks fine. Certainly, I think that's a great opportunity. We'd love to [indiscernible]. I think that's a huge opportunity moving forward. And very clearly, I think that looks promising.

I think what's probably fair to mention that we are well entrenched on the commercial side in these areas. So we have a bit breast cancer sales force and we have a multiple sclerosis sales force, so there will be a benefit on...

Exactly, from an M&D perspective, these are launching right in our sweet spot.

Yes. Graham, any additional questions, follow-on questions -- what's the [indiscernible]?

Then let's move on. Next question come from James Quigley with Goldman Sachs.

So I got 2, please. And apologies, I missed the first one on the Vabysmo. So what is the current bottleneck for the recovery or acceleration of Vabysmo sales in the U.S.? We've sort of seen increased funding towards the back end of the last year, but it seems to be taking a lot of flow through to patients. So what is the -- what is going on here? Is that normal from what you've seen before? And have you started to see the bottlenecks lifting as we've gone through the quarter? That's the first one.

Second one, again, revisiting CT-388. So you started the Phase III trials now. So there's two of them that are seen on clinicaltrials.gov. And these look like traditional Phase III designs, can you confirm to what extent there is any flexible dosing here? Did you also consider adding semaglutide as a reference arm or even an active comparator to the study?

And then also related to 388, what was the progress? What's the next steps in developing combinations with petrelintide? Is that still the plan? And when will we start to see some of that data?

Yes. Great. So Vabysmo, just to reiterate, good quarter, 13% for the division back to growth in the U.S., 4% up. We continue to reiterate that we will see an acceleration from the 12% that -- growth that we saw overall last year. To reiterate, patient assistance foundations are not part of our commercial strategy, and we do not look at them when we think about sales. That having been said, what we have been hearing from the market is that based on what happened last year, physicians have just gotten savvier about where to use patient assistance dollars and they're using them for patients who truly need them. And they're just thinking differently about how they're utilizing those funds as they're made available. But that is a completely different and separate thing to anything that we do on the commercial side. And I just want to be crystal clear about that.

CT-388, you are right, the first two trials are relatively straightforward and standard against placebo. And that is because that is what is currently required for regulatory approval. So when you talk about bringing a new molecule in this space to the market. These are the trials that the FDA asks you run. And so therefore, that is what we are doing. We will be looking at other things going forward with CT-388 and more to come on our plans in that space. And then obviously, we're very excited about launching that fixed-dose combination you've now heard us talk about several types.

I think, James, regarding the -- you had a question on the combination development. I think this has to be kicked off of midyear. So midyear, we would expect the first patient in. And I think we also have the opportunity at the IR event coming up around ADAs then to drill a bit more and go into it more into the details on the trial design.

Okay. Then we move on, and the next one is Simon Baker from Redburn.

Thank you, Bruno. 2 if I may, please. Firstly, for Teresa. I was wondering if you could share some of the feedback from physicians following the AAN presentations of fenebrutinib. We've had a few drugs recently where there's been a should we say, a disconnect between a negative market impression and a positive physician impression on drug. So I just wanted to know if that was the case here.

And then a slightly different question, going back to the AI factory. I just wonder if you could give us an idea of how and when you will assess the impact? And how and when will we be able to assess the impact of your significant AI investments?

Yes. So I will say, I was at ACTRIMS when we presented the PPMS data, and I have had multiple conversations with the team on the ground at AAN, and the KOL feedback is extraordinarily positive. I think people are very encouraged by the data that they're seeing. They are they're reassured in some ways by what they're seeing on the safety side with the liver, with the relative understandability of the liver enzymes and the fact that we're not really seeing a pattern with some of the other adverse events. So there is a huge amount of excitement here. I mean when you think about the MS population, 45% of patients today are either on orals or low-efficacy therapies. This is extremely low-hanging fruit for something like a fenebrutinib that acts on both the progressive and relapsing portions of the disease. So I think we see a ton of positive KOL excitement and enthusiasm. And I think they're really looking forward to having this tool in their armamentarium. And I can pass it on to you, Thomas.

Yes. On the AI side, you saw on the slide from Bruno that you will have an update at Pharma day. And I know Aviv and some of our people will be there. So you see exactly how many molecules are ready in our pipeline coming through the use of AI tools in research. And how we're going to really track that in the future and how we can make it transparent for you so that you can see the benefit that AI has already today. But I can say our teams are now really utilizing AI constantly in research. And actually, a lot of the molecules have some sort of AI tools used in the development. Bruno, do you want to comment on that?

Yes, I can quickly comment. So really stay tuned for Pharma Day. We will have Avon on stage, and I think we will roll out a bit our strategy here also and how we will track the progress in this field. I think it's fair to say it's early days. We have now generated, I think, a couple of NMEs. Basically, all new NMEs have another AI component to them. And we have established a system where we will track this progress. So these molecules, I think, are to enter the clinic within the next couple of quarters. And then, of course, we have to wait 1 to 3 years to really collect enough data to get a feel on how this has impacted success rates, for example, or whether we were able to identify new drug targets and many more questions to be asked. But I think September, that's the point in time when we really will kick off, I think, this AI communication and provide also some mid- and longer-term perspectives.

Then we go on. Next question go over to Justin Smith from Bernstein.

Just a couple of. Firstly, on Alzheimer's, just wondered if anyone could share any thoughts about how the launch of the pTau217 test could actually expand the market for biologics.

And then just second one on Polivy given potential competition coming up? How will you think about whether the sequential share gains we've seen in the U.S. would be sustainable or not?

So yes, I'll take the first one on pTau217. So if you think about the data set, as you heard on the slide, we have data across primary, secondary and tertiary care. Now one of the biggest challenges people need to be diagnosed with Alzheimer disease is the availability of specialists. Oftentimes, it takes over 2 years. By having a primary care claim, potentially with the blood-based biomarker, you have the potential to greatly expand access. And for this to get into guidelines, it would need to have a 90% negative predictive value, 90% positive predictive value with a 90% sensitivity and 90% specificity. If that is achieved, then absolutely, that becomes possible. So that's how I'd characterize 217.

Yes. And I would say from the pharma standpoint, we're extremely excited to have these assays in the market because we do think this has been one of the limiting factors to uptake, is that just getting proper diagnosis and being able to get those eligible patients in. So I think we're super excited to have that partnership.

In terms of Polivy, yes, I do think those share gains are sustainable. And I think in addition, we will be releasing even additional data about combinations with Polivy. So we set the standard of care 20 years ago plus with Rituxan. We sent it again with Polivy and stay tuned.

Yes. And also to remind you, we have our own first-line development program ongoing with Tigo, so this is to come. So the other thing is, I think if you look at all the recent study starts, they all have basically been building on Polivy regimen. So I think this really tells you that this has been widely perceived as the standard of care in first line, which is roughly CHF 2 billion opportunity, and I think we are really nicely tracking towards the scale. Justin, any other questions?

No. That's very clear.

Yes. And we move on with Richard Vosser from JPMorgan, Richard.

A couple of questions, please. Maybe on Hemlibra. Teresa, you talked about very strong growth across the world, but also seeing switches back from ALTUVIIIO. So I just wanted to think about the sustainability of those switches and how much that's contributing now?

And just also, we're going to see a competitor to come later this year. So just thoughts on how you defend against that and what they could do to the growth, which is very good at the moment.

And then the second question on Gazyva you alluded to the impact from competition in CLL. Just could you give us a context of what sort of drag that could be for the rest of the year as we think about the rollout into lupus, just so that we can get an idea of the growth there.

Absolutely. So sort of exactly what we had thought was likely to happen with Hemlibra is what we're seeing happening, which is patients who were well controlled on HEMLIBRA, maybe were tempted to try ALTUVIIIO. We see actually a very large percentage of patients switching back. It's around 30% of patients who went to ALTUVIIIO have switched back. We do not see that trend have slowed down dramatically. The growth that ALTUVIIIO is getting is really coming from other Factor VIII at this point. And so I think what it boils down to is that patients who are on Hemlibra have an incredibly high satisfaction with their therapy. And it's not uncommon that they switch back after time. So that's -- this is a great therapy. It has changed the lives of many, many patients living with hemophilia. And I think we -- we expected to see a lot of people come back and that's exactly what's happened. We also know that Hemlibra has set a very high bar in the treatment of hemophilia. And we don't actually really see much clinical -- we don't really see any clinical differentiation to be frank with MMAE. And so the efficacy profile doesn't look differentiated in terms of ABR reduction patients with 0 bleeds. I mean we do expect increased competition, but we are also doing things to benefit the patient experience of Hemlibra with the new injection kit, with the auto-injector that's in development. And so we are -- we remain confident that Hemlibra will continue to be the standard of care. That having said, we do believe that there will be competitive impact to the back half of the year, and we took that into account when we gave the outlook for this year. But as we are looking forward to NXT007, I think it is worth pointing out that we are the only people who are -- have been kind of brave enough to go up against that standard of care, raising the bar on ourselves yet again. by going head-to-head with Hemlibra with NXT007. So I think we're yes, I think we are -- we remain confident that we've got a great product on the market and that we'll be bringing you the next generation here in the coming years.

In terms of Gazyva, I think that the drag on the oncology side, it's probably a little too soon to tell. I'll have that have a better look for you as we get into half year. it's sort of been just enough to kind of tamper down, I think, what we think the initial uptake is in the immunology side of things. So again, it's earlier days, so maybe give us to half year, and then we'll be able to give you a better sense.

Yes, I think it was launched in the fourth quarter or something...

Exactly. It's just a little soon to actually see the data in the market, but that is what we're hearing. They're just starting to get uptake.

One more comment on the Hemlibra ALTUVIIIO situation. I think what has also been driving switching back is that just remember that Hemlibra doesn't cause any inhibitors. And I think this has been an issue constantly. Richard, any other questions?

No, that's perfect.

Then next one is Matthew Weston from UBS.

Two questions, please, both on giredestrant. Teresa, you laid out the market at the very beginning in your introductory comments. I'd be intrigued to know what proportion of ESR1 mutant patients, you think already know their status. Because you talked about a strong launch, I can imagine it's even stronger in that late-line population but only if they know whether they're ESR1 mutants?

And then the second question on giredestrant is a much more overarching one. I don't know it's for you, Teresa or whether it's for Thomas. Based on the numbers you've laid out, you have 1 of the most exciting launches coming up in the next couple of years. And historically, breast cancer has been a setting where ex U.S. and U.S. have actually been quite balanced in terms of peak sales potential. So given MFN, how are discussions ongoing with European governments about the idea of accepting higher prices? And do you think we're going to see a rapid rollout of giredestrant ex U.S. as well as U.S.

So I would say that within breast cancer, testing in breast cancer is actually quite high. And so in the more mature markets, I would say that it is pretty common that people would know what their statuses are. And so therefore, I think that very much enables us with the strong launch. Thomas, do you want to take the [indiscernible] question or do you want me to take it?

Sure, I can do that. So on the U.S., we know that we'll be on the market towards the end of the year. So this is when we set the price in the U.S. But of course, we are already in discussions with politicians across Europe. And they do understand that there will be a change in terms of how these medicines will be priced in the future. And the good thing is it's not a one-step change because medicines that they already on the market are not going to be priced higher. It's just for new introductions. So it's not a very rapid and steep increase in terms of spending that they have on medicines.

But I do believe what will happen is you will have more differentiation between those medicines that have extremely strong data and maybe medicines that don't have a strong data. I think with your giredestrant being, I think, one of the first molecules we have, where we will go through this process, we're actually in a beneficially -- beneficial situation that we have very, very strong data. And especially the breast cancer space has also a lot of patient groups behind it that are very interested to get this to patients as quickly as possible. So we are hopeful that with all of that, that we will get it to a point where we can get the right kind of pricing also in Europe.

Very good. Matthew, did this answer your questions?

Perfect.

Yes. Then next question go to Stephen Scala from Cowen.

I have 3 brief questions. First, it sounds like Roche still has confidence in the pionERA trial post failure persevERA. And I'm just wondering what is the basis for that confidence?

Second, why isn't Roche doing a cardiovascular risk-reduction study of its IL-6 assets following in the footsteps of its competition.

And then lastly, Thomas, I don't believe you have opened the diagnostics meeting in the past. I'm curious why the change this year?

So just to reiterate, though persevERA didn't reach statistical significance at PFS, we did see a numerical improvement. We do believe that it's active in the first-line setting. And the reason that we continue to have confidence for persevERA is it's just being studied in a very different population than persevERA. And so again, in that first-line setting, context really does matter. And so we still believe that biologically, there's a very good rationale on why we work with with -- in persevERA, 20% to 25% of patients in that setting are mutated and pionERA has enriched CER mutation patients to 40%. So again, we're sort of designed to really uncover that signal.

In terms of the IL-6 assets, I don't assume that we haven't looked. So I'll say that. And I don't know who would like to take questions.

Yes. The last one, I can take. Yes, I just want to also reiterate the reason for our higher confidence in pionERA is simply because it's enriched with the ESR1 mutant population. And if you look at the hazard ratio in past trials in this population, then you would understand, of course, why we have a higher chance on this one. And regarding the diagnostics, I did open Diagnostics Day before because I was the CEO of Diagnostics. But it's the first time that I will actually open it when I'm not the CEO of Diagnostics. And the way Bruno and I discussed it is that I would go once to Pharma Day, once the Dia Day, once for Pharma Day so that I just sit back and forth. And every time I give you more of a group update and would be available for questions in that setting as well.

Yes. Very good. Steve, maybe one comment on cardiovascular diseases. You might be reminded that we have other modalities we are focusing on. For example, we highlighted NLRP3 as one of our key targets, and we can imagine that we have also looked into IL6 in that context. Any other questions?

May I assume since you've looked -- can we assume that you looked in the trials who were not successful with?

I would -- all I'll say is just don't assume we haven't looked.

Okay. Then we have next in the row, Michael Leuchten from Jefferies.

Two questions. One question, one clarification. The FDA's increased level of disclosure seems to be triggering some petitions for label changes. And I think Ocrevus is one of them. I just wondered if you could talk to the process here, time lines. Is this something that you think is going to happen more broadly? And how long does it take to get some clarity on that?

And then a clarification question for Teresa, please. Just going back to the comment on Vabysmo U.S. returning to growth. The revenue sequentially are and so there's 3% growth in Q1, it was 7% in Q4. So when you say returning to growth, what's the variable that you're looking at that gives you that lead indicator that we're seeing that?

Yes. So just speaking to Vabysmo first, I mean we saw a 4% increase in the U.S. But we saw a more significant volume increase. So again, this is a highly contracted market. So those things can sometimes look a little bit different. We're also seeing the unbranded use of Avastin start to trickle down. So that's another thing that kind of gives us confidence that the branded market may be coming back or is coming back.

In terms of the FDA, I assume that you're talking about the letter that was published by the British Medical Journal, is that what you're referring to?

Yes, that's correct.

On PPMS. Yes. So I just want to be really clear that Ocrevus and PPMS has been used in 450,000 patients it was approved. We have 1.4 million patient years of exposure. We are approved in 130 countries. This is an approval that KOLs and patients in the FDA are not questioning. We believe strongly in the benefit and the safety of Ocrevus in the PPMS population that was reinforced with the [ OHAN ] study, which was just released September of last year, which is not at all mentioned in the letter that you're referring to. So I think we have an incredible level of confidence in the benefit and the safety of Ocrevus and PPMS. And I will leave it there.

Maybe just, Michael, one add-on the Vabysmo side, how are we tracking our performance. I think if we just zoom in on the branded segment and exclude all biosimilars which are in the market, then I think we still keep gaining shares in all the indications, and that's a picture, I think you can see globally. And the run rate of share gains versus the other branded competitor in the market, I think, is very constant quarter-over-quarter. So I think, therefore, I think we are reasonably confident.

Exactly correct. And that is, by the way, we continue to hear every time we pull KOLs that hands down Vabysmo is seen as the most efficacious drug that they have in their armamentarium.

Any other questions, Michael?

No.

Yes. Then we come to the final questions, which go to Rajesh Kumar from HSBC.

A Couple of questions, if I may. First is on breast cancer opportunity, obviously, a very exciting opportunity for you on giradestrant. Your oncology portfolio looks a bit light on the pipeline side. Would you think -- or talk us through how you're thinking about capital allocation to augment your oncology portfolio going forward. Definitely interested in understanding if you need to add something more there to utilize your strong position in the market?

Second is on the obesity clinical trials. What we have seen more recently is the patient churn increases once patients work out, they are not on an active arm. And when you're thinking of your Phase III trials, would you consider running active control trials for registrational purposes to ensure patient churn is not very hot, and you have credible data? Or have you found a work around that would be very interested in hearing your thoughts

And as always, being an analyst who can't count, the third question is, I totally appreciate your excitement around NMEs. If you can help us understand where in the financials, will we start seeing the benefit of AI? Would it be in R&D intensity, SG&A intensity in the next 3 to 5 years? That would be much appreciated. Okay.

Do you want me to start?

Yes.

Yes, sure. So first, let me start on the capital allocation. We have the 5 therapeutic areas that we're active in, and we continue to look for opportunities in all of those therapeutic areas that fits in terms of from a scientific perspective, if we believe that it can be a medicine that could be best-in-class or first-in-class. At the same time, we also look at the financials. And the two things have to come together. We want to make sure that we are not in a situation that we overpay for some of these assets. And I think the combination is very important. And especially in oncology, what we have seen sometimes is that there has been quite large amounts of money being paid for very, very early stage assets. So we keep working on that. We brought in CDKs, we brought in a number of different ADCs. So you see we keep evolving also our oncology pipeline, but we are also disciplined, right? And the reason is there's just many more opportunities than money. So there -- if you don't get an opportunity where you pay a couple of billion, I can tell you, you will get another opportunity that may be as good or even better for less money. And so you need to keep looking. But what I can say is we look at everything. So everything that goes across the table, I can tell you we've looked at it, and we've made our own assessment on that.

On AI. So we believe that AI will really impact us across actually all of our cost lines. So we already utilize AI in manufacturing. We use AI to really reshape processes in our organization, where we take processes that in the past will take maybe months, we try to limit that to weeks or even days. And we've seen that also across the entire R&D spectrum. So this is more, I would say, manual work that people have done in the past where we can really automate that in a very -- in a much better way. So you can really start to see that happening across the organization. And that's why I believe it's going to happen everywhere. It's going to happen or it is happening also on the commercial side. Where the big disruptive opportunity is, I believe, in AI is to really change on how fast we do drug development and also how many molecules we can screen. With these virtual models, you can screen through many more molecules, then you look deep into some of those, you test them in the wet lab, you feed back the data. And this is what we call lab in the loop. And so you can really continuously build up these models and these are proprietary models. It's not something that other companies have and where we believe we're quite far ahead compared to other companies. And as we continue to build these models, you will see increasing benefit over time. But from a process perspective, you already see some of the benefit in the organization and we'll continue to see more benefit.

In fact, we've trained 100,000 people on AI. It's something we call every day AI. We've trained everyone using AI tools. And feedback has been really amazing. People can really see the productivity benefits in the daily work.

Yes. Let me quickly add here. We also -- how should I say it, we are investing heavily into our data infrastructure anyway. I think really, we have a huge ERP program running. And why do I say this? Because that will facilitate the use of AI moving forward. As we will have the whole data in the cloud that we can really tap into this much more easily than we could in the past.

And to answer your final question, as I indicated earlier, placebo-controlled trials are still mandated by the FDA. That is why we run them. Obviously, for anybody running large placebo-controlled trials, patient retention is quite key. We have a number of -- we have a broad suite of patient retention, tools and strategies to design to help keep patients in the CT-388 studies and for obvious competitive reasons, I won't say more.

I think we had a third question, which was, is our next M&A deal an oncology deal? I think I just translated this -- the question. No problem at...

I don't know what our next deal is. We will find out, yes. But are we looking for opportunities, Thomas said it, happy to do that. But you know how that is, it takes two to tango. And the question is, who will be our next partner do a great deal.

So we're looking at all therapeutic areas. Yes. With that, I think we have closed the Q&A session. I hand back to Thomas for the final words.

Yes. Thank you very much for attending today's call. As you can see, we have continued strong sales momentum as we had over the last couple of years. And based on some recent readouts that we have seen, this gives us much more confidence also in the short to midterm that we can continue good growth outlook there as well. And you have also seen that we have between '27 and '30, the opportunity to launch up to 19 MMEs. That's status today. We continue to work on M&A so that we can increase the number in that time period, but this is really substantial, and many of those have a lot of value attached to them. So that also gives us confidence in the longer-term growth outlook of the company. All I can say is you can count on us, and we will deliver.

My name is Henrik, and I'm the technical operator for today's call. Kindly note that the webinar is being recorded. [Operator Instructions] One last remark, if you would like to follow the presented slides on your end as well, please feel free to go to roche.com/investors to download the presentation.

At this time, it's our pleasure to introduce you to Bruno Eschli, Head of Investor Relations. Bruno, the stage is yours.

Thanks, Henrik. And could I have the agenda slide, please. So welcome to our third IR event in 2026, focusing on the Phase III FENhance 1 and 2 results for fenebrutinib in RMS, which just got presented yesterday at AAN in Chicago. Today's call is scheduled for 60 minutes, and we will have our Roche team dialing in from Chicago. Let me quickly take you through today's agenda. We have 2 speakers with us. Following my opening, the first speaker for today will be Alexandra Goodyear. She is an MD and our life cycle leader for fenebrutinib.

Alexandra will provide a quick introduction to fenebrutinib and how this molecule truly differentiates from other BTKs who have been in late-stage development for MS. She will also highlight again the positive Phase III FENtrepid results for fenebrutinib in PPMS, which have been presented earlier this year at ACTRIMS. Our second speaker for today will be then Jiwon Oh. Jiwon is an MD and PhD FRCPC, a Medical Director of the Barlo Multiple Sclerosis Program and Associate Professor and Staff Neurologist at the Division of Neurology at the St. Michael's Hospital at the University of Toronto and the scientist at the Keenan Research Center of the Li Ka Shing Knowledge Institute. She is the lead author for FENhance 1 and 2 and has been sitting on the Study Steering Committee.

Jiwon will take us again through the positive Phase III FENhance 1 and 2 results for fenebrutinib in RMS, which we presented yesterday at AAN. Afterwards, we will have a Q&A session. In addition to our 2 speakers, we will be joined for the Q&A session by Hideki Garren, Senior Vice President and our Global Head of Product Development for Neurology; and by Levi Garraway, our Executive Vice President and Head of Product Development and our Chief Medical Officer.

And with that, over to you, Alexandra.

Thank you very much. I will take us through an overview of fenebrutinib's development program. Next slide. With the positive Phase III study data in RMS and primary progressive MS, fenebrutinib has the potential for both first and best-in-class. And fenebrutinib has the potential to be acting on B cells and macrophages in the periphery as well as B cells and microglia within the central nervous system. And with this ability, we're hoping that fenebrutinib as seen in the Phase III that Dr. Oh will take us through, is decreasing relapsing biology as well as progressive biology. And these 2 together are leading to overall decrease in disability accumulation.

And with this, fenebrutinib has the potential to be the first and only high-efficacy oral treatment for both RMS and PPMS. And with this has the opportunity to increase high-efficacy treatments for patients as well as expand the Roche MS franchise footprint. On the right-hand side, you see the fenebrutinib development program, where we have now had 4 positive trials in MS, including the Phase II RMS trial, the FENtrepid primary progressive MS trial and the 2 nearly identical relapsing MS trials. Fenebrutinib has been studied in a large number of patients across the entire development program and data will be submitted for filing midyear.

Next slide. Here is the primary progressive MS data that Bruno alluded to that was presented earlier this year. On the left-hand side, we're looking at the Kaplan-Meier curve for disability progression that was analyzed by the composite endpoint of cCDP. Here, you can see that the treatment curves separate already at week 24, fenebrutinib being the blue line on the bottom. And that separation in curves is maintained throughout the study duration. And with this, fenebrutinib achieved noninferiority versus ocrelizumab in the only head-to-head trial of any MS therapy versus ocrelizumab.

And on the right-hand side, you can see the breakdown in the composite disability progression endpoint. So the makeup was EDSS, Timed 25-Foot Walk and 9-Hole Peg Test. You can see that the most profound treatment effect was seen with 9-Hole Peg Test, which is an upper limb mobility test, very important for patients for them to be able to maintain independence throughout their lives. Second, the -- followed by the effect on EDSS and then Timed 25-Foot Walk. Notably, in a post-hoc analysis that looked at the same composite endpoint that was used in ORATORIO-HAND, which was a contemporaneous Phase III study of ocrelizumab versus placebo, which used the endpoint of EDSS and 9-Hole Peg Test. When we look at that endpoint in the FENtrepid study, we saw a 22% reduction versus ocrelizumab.

Next slide. This differentiated data that we have been showing, we think it really stems from the unique structure and binding mechanism of fenebrutinib. On the left-hand side, we show the chemical structures of fenebrutinib as well as some of the other BTK inhibitors being explored in late-stage development for MS. And you can see that fenebrutinib structure is very unique with highlighted portions that get into both the H2 and H3 pockets of the Bruton's tyrosine kinase and allow it to bind noncovalently and very selectively.

If we -- on the right-hand side, we see that we were able to achieve the selectivity that we were aiming for when designing this molecule. What you're looking at on the right-hand side is a panel of very similar kinases to BTK. BTK is the first row right at the top. We can see that fenebrutinib binds to BTK at a low concentration and then you -- and then really no other off-target kinases until you get to very high concentrations. This compares very favorably to -- compared to some of the other covalent inhibitors, which you can see are hitting multiple kinases. So we're hoping with this, that this helps improve long-term safety outcomes with fenebrutinib.

Next slide. On this, fenebrutinib has high systemic exposure as well as CNS penetration. On the left-hand side, we're looking at free plasma concentration curves and fenebrutinib being the red line at the top, you can see achieves much higher concentrations over the whole 24-hour dosing period. And this really reflects the different strategies of having a noncovalent drug versus covalent drugs and then optimizing the noncovalent drug to have high exposure. And this is important because the free plasma fraction is the part that is able to cross the blood-brain barrier in order to bind with microglia and B cells within the central nervous system.

And on the right-hand side, we are showing the data that fenebrutinib is able to achieve that. What we're looking at on that side is a human CSF concentrations of these drugs, fenebrutinib being the blue bar all the way on the left. Each drug has its own IC90, the concentration needed to inhibit 90% of the target. And we can see that fenebrutinib is able to cross the blood-brain barrier and being concentrations within the CSF well above the IC90, which is really differentiated compared to the other BTK inhibitors as shown here.

Next slide. I think that brings us to Dr. Oh's presentation of our Phase III data in RMS. Dr. Oh.

Thank you, Alexandra. So I will now, with that background, take you through the very recently reported results from FENhance 1 and 2, which were 2 nearly identical Phase III clinical trials evaluating the efficacy and safety of fenebrutinib in comparison to an active comparator, teriflunomide in people with relapsing MS.

Next slide. So Alexandra has already very nicely outlined this, but BTK inhibitors are of very high interest in the MS field because we know that BTK signaling is crucial to activation of both B cells and myeloid cells in the periphery and the central nervous system. And so we know that inhibiting BTK has the potential to address both relapsing and progressive disease biologies across the entire MS disease spectrum. As you've already heard, fenebrutinib is a BTK inhibitor that was uniquely purposefully designed. It is CNS penetrant, which is very important because when we're talking about modulating both relapsing and progressive disease biologies, we want to ensure that a drug that is active in the periphery can easily get into the central nervous system, which fenebrutinib can clearly do. It's noncovalent. It's highly selective.

And importantly, it has an optimized PK/PD profile, which is likely in large part of the reason we see a clear effect of fenebrutinib on relapsing disease biology that I'll tell you about in a second. Already in a Phase II clinical trial, the FENopta study, fenebrutinib demonstrated near complete suppression of disease activity in people with relapsing MS. And as you heard, in the Phase III FENtrepid study that evaluated fenebrutinib versus ocrelizumab in people with primary progressive MS, fenebrutinib demonstrated efficacy and that it was noninferior in reducing the risk of composite confirmed disability progression.

And in that post-hoc analysis that Alexandra presented, where there was a modified composite endpoint, fenebrutinib demonstrated superiority in this post-hoc analysis in comparison to ocrelizumab. Today, right now, we'll be focusing on the FENhance 1 and 2 trial results and the objective of these 2 trials was to evaluate the efficacy and safety of fenebrutinib in comparison to teriflunomide in people with relapsing MS.

Next slide. So this was the study design of FENhance 1 and 2, which were 2 largely identical Phase III multicenter, randomized trials. As you can see, the inclusion criteria were quite typical for what we usually see in contemporary relapsing MS clinical trials. People with relapsing MS were randomized in a 1:1 ratio to either fenebrutinib at 200 milligrams twice daily or the active comparator of teriflunomide at 14 milligrams once daily, which is a very commonly used first-line oral disease-modifying treatment in relapsing MS.

And at the end of the study, they were given the option to continue in the open-label extension period. The trial went for a minimum of 96 weeks for each participant because it was an event-driven trial. And as you can see, study visits were done periodically as is expected with both contemporary relapsing MS trials, and there were MRIs performed as well and of course, safety was evaluated for the duration of the study.

Next slide. So the primary endpoint as expected because these were trials looking at relapsing MS was the annualized relapse rate. For key secondary endpoints, the key clinical disability progression endpoint, again, a composite confirmed disability progression was used. And so this is an endpoint that consists of either EDSS progression, which is the typical global neurological disability scale that we use in really every MS clinical trial. However, a participant was deemed to have progressed if they showed progression in either the EDSS or the 9-Hole Peg Test, which is a test of manual dexterity or the 25-Foot Timed Walk, which is a semi-quantitative measure that measures ambulatory speed.

In addition, key MRI endpoints were evaluated as well, which included T1 gadolinium-enhancing lesions and the development of new or enlarging T2 lesions. I wanted to note as well that as with any clinical trial, there was a statistical hierarchy. And each of the endpoints that we discussed, it was prespecified that in the statistical hierarchy, they would be evaluated in each individual FENhance clinical trial separately. However, there was a planned prespecified pooled analysis of the disability -- the composite confirmed disability progression endpoint. And this is relevant when we discuss what was observed with respect to these secondary clinical endpoints.

Next slide. So when we look at baseline characteristics across FENhance 1 and 2, as you can see, clinical characteristics were quite well matched between treatment arms across the 2 FENhance studies. And also just wanted to know when you look at the time since MS diagnosis, and I'm just going to draw your attention to the fourth row down, and I know it's a very busy table, but it's very clear that this was a really early group of people with relapsing MS because the median time since diagnosis was 1 year.

So really, a lot of these participants were newly diagnosed. And similarly, when you look at the proportion of people who have previously been on disease-modifying treatment, which is [indiscernible], as you can see, only about 15% to 20% of participants were previously treated. So the majority of these participants were within a year of diagnosis as well as treatment naive. And so these points just illustrate that this was a very early population of people with relapsing MS.

Next slide. So when the primary endpoint was evaluated, as you can see, fenebrutinib very clearly substantially reduced the annualized relapse rate versus teriflunomide in both FENhance 1 and 2, and there was a substantial reduction. And as you can see, the reduction -- the relative reduction was 51% and 58% in FENhance 1 and 2, respectively.

Next slide. When you look at different subgroups, as you can see, generally, the treatment effect was consistent across different subgroups. However, in people who were younger and had a lower time since symptom onset as well as in those who had gadolinium-enhancing lesions and those who had less disability, it seemed that the treatment effect was accentuated. So overall, it seems that the treatment effect is consistent, but not really surprisingly and those who are earlier on in the disease course, the treatment effect seems to be accentuated.

Next slide. Moving on, when key secondary MRI endpoints were evaluated, as you can see with T1 gadolinium-enhancing lesions, fenebrutinib very clearly decreased T1 Gd lesions versus teriflunomide at 70% as well as 77% in FENhance 1 and 2, respectively. Not surprisingly, when new or enlarging T2 lesions were evaluated as well, as you can see, fenebrutinib very clearly reduced the formation of new or enlarging T2 lesions in comparison to teriflunomide by 76% and 82% in FENhance 1 and 2, respectively.

Next slide. Now when the composite confirmed disability endpoint was evaluated in each individual trial, as you can see, although there was a consistent trend towards favoring fenebrutinib and FENhance 1 and FENhance 2, the confidence interval is still slightly over left one. However, when the pooled cCDP endpoint was evaluated across both FENhance 1 and 2, as you can see, the confidence interval did not over one and this favored fenebrutinib. Similarly, in that post-hoc modified cCDP analysis, which included only EDSS and 9-Hole Peg Test. And just as a reminder, as Alexandra said, this was the endpoint that was actually recently used in the ORATORIO-HAND trial. As you can see, you see similar trends. Consistently, there is a trend favoring fenebrutinib with this modified cCDP endpoint as well. And then when you look at the pooled analysis, you can see that there is a clear treatment effect with the confidence interval not overlapping.

Next slide. Moving on, when you look at the treatment effect of the individual components of the cCDP measure, so this composite measure, as you can see first, the greatest number of events was related to changes in the 25-Foot Timed Walk. However, when you look at the treatment effect of individual components of the cCDP score, it really was the 9-Hole Peg Test that seemed to have the strongest treatment effect of fenebrutinib versus teriflunomide. So this is interesting and probably informative about the cCDP measure because it seems that the treatment effect is greatest with the 9-Hole Peg Test, but the most frequent events that were observed were with the 25-Foot Timed Walk.

Next slide. Moving on to safety. When you look at the overall safety profiles, these were generally comparable between fenebrutinib and teriflunomide with similar rates of any adverse events as well as serious adverse events and adverse events leading to withdrawal from treatment. However, there was an imbalance in fatal adverse events with 7 fatalities observed in fenebrutinib-treated participants and a single fatality observed in teriflunomide-treated participants.

However, when you break down and take a look at each of these individual fatalities, and we'll do that on the next slide, it's reassuring because there was no clustering of causes that seem to suggest that there is some signal related to fatalities and fenebrutinib, and we'll look at that in detail on the next slide. Also wanted to highlight, obviously, with any immunomodulatory treatment, there is a concern about infections and particularly serious infections. And it was reassuring because in the FENhance 1 and 2 clinical trials, the proportion of people who had any infection was similar between treatment arms.

And importantly, the proportion of people who had any serious infections was similar across treatment arms between fenebrutinib and teriflunomide, which was reassuring to see. Getting back to the fatalities. This is a table that summarizes all of the fatalities seen across the FENhance 1 and 2 clinical trials. So the first 7 cases are fatalities of participants who are on fenebrutinib and that final case listed here is the fatality that was observed in a participant who was on teriflunomide. So a couple of things. First of all, it's reassuring that there is no clear clustering of events that's related to something that potentially could be related to a mechanism of action that we know of fenebrutinib, which is a BTK inhibitor.

And of the 7 fatalities that were observed in fenebrutinib-treated participants, most of them were considered to be not relevant to fenebrutinib, but 2 of the cases were deemed to be likely relevant to fenebrutinib. And these were the 2 infectious cases, which included a participant who had pneumonia as well as a participant who had neuro-cryptococcosis. However, with both of these infectious cases, it's important to note that there were complicated other contributing factors. With the case of pneumonia, the family actually declined hospitalization despite the physician recommending this.

And with the neuro-cryptococcosis case, probably some people may not be familiar with this bug, but this is a bug that is endemic to some regions in Latin America, often related to occupational exposure, which was the case in this particular participant. And this is a bug that can actually cause even in immunocompetent individuals serious infections. As you can see with the other fatalities that were observed in the fenebrutinib treatment arm, there was a case of intestinal ischemia, a case of hemorrhagic stroke, a case of death, a case of suicide, multiple injuries and a case -- another case of suicide. So overall, these other deaths were not considered to be relevant by the study site PI to fenebrutinib. And generally, it seems that when you look at these individual cases, it's reassuring that there is no clustering of events that raise any red flags that this might be related directly to fenebrutinib.

Next slide. Liver safety, as you know, is a signal that we need to pay attention to with BTK inhibitors because a number of BTK inhibitors have shown that they can cause liver enzyme elevation. So when we look at this in FENhance 1 and 2, as you can see, with any liver enzyme elevations, 3x or more in the upper limit of normal, the proportion of individuals that had this looked similar across treatment arms and was in keeping with what has been seen with other BTK inhibitors that have been evaluated in Phase III clinical trials. Of note, there was only a single Hy's Law case in a fenebrutinib-treated participant as confirmed by the hepatic adjudication committee. And there was also a single Hy's Law case in the teriflunomide treatment arm that was confirmed by the hepatic adjudication committee. And it's important to note that with all of these liver enzyme elevations, these were all reversed.

Next slide. So to conclude, the results of FENhance 1 and 2 demonstrate that fenebrutinib is clearly superior to teriflunomide in significantly reducing relapses as well as disease activity on MRI, demonstrating that there is likely a substantial impact of fenebrutinib on relapsing disease biology in MS. With respect to the disability measures, there was a consistent trend favoring fenebrutinib over teriflunomide, which together, when you keep in mind the results of FENtrepid suggests that fenebrutinib has also an impact on progressive disease biology in MS. For safety, the rates of adverse events were comparable, including hepatic transaminase level elevations as well as infections. There was an imbalance observed in fatal AEs, but we've gone over each of these in detail.

And overall, it seems to be reassuring that, again, there is no clustering of causes of fatalities that seem to be relevant to the known mechanism of action of fenebrutinib. So taken together, these results demonstrate that fenebrutinib is the first BTK inhibitor to show clear efficacy on relapsing disease biology versus an active comparator. And again, when we bring into this, the results that we've seen with FENtrepid, fenebrutinib is the first and only oral treatment option to demonstrate efficacy across the entire MS disease continuum. Thank you very much. And I will just hand back over to Bruno.

Thanks, Dr. Oh, for the presentation. And with that, I think we open the Q&A. The first questions go to Luisa Hector from Berenberg.

Great. I would love to hear from Dr. Oh, just in terms of now that we've got all of the data in for fenebrutinib, who are the patients that you feel should be on fenebrutinib, where you're positioning it both in the relapsing and the progressive group?

Sure. So I think FENhance 1 and 2 very clearly demonstrate that even in very, very early active relapsing MS patients, this is a drug that has a very profound effect on clinical and MRI measures of relapsing disease activity. In fact, the effect that we see is akin to our currently available highest efficacy therapies. So definitely an agent that is suitable in early relapsing MS, which was the trial population of FENhance 1 and 2.

However, based on what we see in FENtrepid, it's also clear that it is noninferior, and there's good suggestion that based on what we know about the drug and the mechanism of action and the fact that it gets into the central nervous system and the magnitude of benefit that was seen across FENtrepid that this is a drug that is at least noninferior, maybe even better than our only available agent in PPMS. So taking together all of this and the fact that in the field as clinicians, when we think about relapsing MS, we know that both relapsing and progressive disease biologies are at play across the spectrum of the disease, which includes PPMS, but also many different stages, if you will, of relapsing MS.

So this is to say that based on what we see across the entire fenebrutinib clinical development program as well as what we as clinicians and scientists are thinking in the field right now about MS, it seems that this is a drug that is likely will be useful across the entire spectrum of MS, even people with relapsing MS who are clearly having progressive components of disease because of what we see in FENtrepid.

Luisa, did this answer your question? Or do you have a follow-on question?

Yes. Can I then ask, do you see yourself choosing whether to start a patient on fenebrutinib or Ocrevus? Or is there a sort of sequencing that seems obvious to you now the data is in hand?

So treatment decisions are always pretty individual. But in certain people -- so there's always kind of a preference often that patients have for mode of administration and frequency of therapy. However, in the end, as with most cases, it all depends on the individual patient. But the benefit that fenebrutinib may have over a drug like ocrelizumab, which is a very commonly used drug, as you know, in the MS field is that it may have a clearer effect on progressive disease biology.

So although with the pure relapsing component of disease, I think we would use both drugs in a similar fashion. If there is a relapsing MS patient that clearly has evidence of progressive disease, which we see in the majority of people with relapsing MS, even in the earlier stages of disease, it may make sense to use a drug like fenebrutinib because we know that it has an effect on relapsing as well as progressive disease biologies.

Next questions go to James Gordon from Barclays.

James Gordon, Barclays. A couple of questions, please. The first question was assuming Fene does get approved for RMS and PPMS, how do you think the label might capture any potential liver infection or suicide profile of the drug? And how would that impact uptake? Are there also other precedents for other drugs, which you could think about, which had similar profiles in any of these attributes? And how did that impact uptake? And maybe just squeezing in one other one. I've been asked by some people like in terms of suicides, could there be a mechanistic explanation? Or do you think it would be noise? How to think about that, please?

Maybe I'll start at a high level on the label. you might expect us, but it's too early to comment. In fact, we're -- as you know, we'll be submitting the entire data package as a composite package to the FDA by midyear, and we'll go through the process thereafter. So it's too early to speculate on what the label might or might not say. But first for the other aspects, I'll turn to Jiwon or others.

Bruno, would you mind repeating some of the key aspects there? It was a bit hard to hear here on what aspects he might want Dr. Oh or myself to elaborate on.

James, can you? Sorry.

Assuming you can still hear me, I was just...

We can hear you, yes.

I've been asked about a couple of attributes. So one was about the liver profile. One was about a potential increased risk of infection and then whether it's connected or not, but in terms of suicide profile. So there are other MS drugs that also have some of these things. So if some of those were -- we'll see if they are, but if some of those were effects on the label, how would that impact how a doctor might go about using a drug which had a profile in terms of liver infection or suicide?

So perhaps, Dr. Oh, you'd like to start on how we would use the drug and how would that impact using [indiscernible].

So liver enzyme monitoring is something that most MS neurologists are very familiar with. And with probably almost all of our existing drugs, there's different schedules that we use for liver enzyme monitoring. And even going back to the days of the old interferon beta agents, this is something that we're quite accustomed to doing. The liver enzyme monitoring protocol in FENhance 1 and 2 was every 2 weeks after initiation of the study drug. And it seems like with all of the cases that were observed, it's really within the first 20 weeks that we may see liver enzyme elevation.

So overall, I mean, I'm not sure exactly what the label will say. And obviously, these discussions what we had in the next little while. But it is probably expected that there will be a requirement for liver enzyme monitoring somewhat frequently, but this will be time limited. So as a clinician, it's -- you don't ever want to be doing some sort of intense monitoring forever. And every 2 weeks, it is relatively frequent. However, I think this will be very manageable in clinical practice because it's time limited. And again, this issue of liver enzyme monitoring is something that we do with almost all of our current MS therapies. So something that is pretty familiar and that many MS neurologists are quite accustomed to.

With respect to the question about infection, I think the results from the fenebrutinib clinical development program are actually quite reassuring. In FENhance 1 and 2, when you look at any infections and particularly any serious infections, there really wasn't a difference between fenebrutinib and teriflunomide, which is reassuring. But obviously, this will continue to be monitored in the open-label extension.

Finally, with respect to the cases of suicide that were observed, many of these cases did have some confounding factors. But obviously, this is something that we will need to pay attention to. There have been precedents for this with respect to MS disease-modifying treatments. Again, back in the days of the interferon beta agents, Betaseron actually did have a signal for this. And so this is something that obviously clinicians will have to keep in mind and probably we'll have to have some counseling with patients about this beforehand. I think we also need to keep in mind that with these cases that were observed, there isn't really a biologically plausible mechanism by which fenebrutinib or BTK inhibitors in general might do this.

And so really need to delve deeper into the cases and to continue to monitor to understand this, but also need to keep in mind that in MS, depression is very, very common and seen in kind of 50% to 80% of people even at the very beginning of disease. So there are a number of complicating factors here that make it unclear really if this is related to fenebrutinib. And then there's other reassuring pieces of data, including the fact that fenebrutinib has been studied in other autoimmune conditions. And there really wasn't any sort of suicide signal that was observed. And in the Phase II FENopta trial as well, which was obviously a smaller study, there really weren't any cases observed here. So this is something that we'll need to continue to monitor and not take lightly. However, it's really not clear whether these cases were really related to [indiscernible].

If I could just jump in, just to say that for Roche and Genentech, patient safety is extremely important. It's the #1 concern we have. And of course, we want to work with regulators to make sure that the patients are safe while getting the [indiscernible] to patients [ as soon as ] possible.

We do want to add that in these 2 studies and in the PPMS study as well, patients with suicides were not excluded as part of the study. And as Dr. Oh mentioned, in other studies, outside of MS with fenebrutinib, we have over [ 60 fenebrutinib ] patients where suicide risk was not seen.

James, did this answer your questions on the potential safety profile and how you would manage it in practice?

That's great.

Next questions go to Alek Ebbeling from UBS.

Two, please. First on liver enzyme elevations. So here there were lower in the FENhance 1 and 2 RMS studies relative to the FENtrepid PPMS study. I'm wondering what you believe is the reason behind this. Do you think it's related to the age of the patients enrolled or anything else?

And second question on one of the deaths. I was just wondering if you could provide any more color on the fatality from neurocysticercosis gattii. I mean the slides state that the pathogen and you said that it can cause serious infection even in immunocompetent patient, but the death was related to fenebrutinib. So what gives you confidence it's not a signal of excessive immunosuppression?

Sure. So I'll address the neurocryptococcosis issue. So I think it's always difficult in study settings to know definitively what is related to a study drug or not. And because of what we know about the mechanism of action of BTK inhibitors, they are immunomodulatory agents. And this is the case for all MS disease-modifying treatments because you need to modulate the immune system to show an effect in the disease.

If there is an adverse event that plausibly is related to the mechanism of action of a drug, this is when it's deemed by the study site investigator usually that it is likely relevant. And so it's not surprising that the fatalities that were related to infections because of exactly what we know about the mechanism of action of the drug was deemed to be likely related to fenebrutinib. And of course, there's a lot of rationale behind why this would be the case.

With neurocryptococcosis, we talked about how it's endemic and it can even in immunocompetent post cause serious infections, I think that's one piece of information that's reassuring that fenebrutinib is not a drug that causes substantial immunosuppression so that you'll see these somewhat [indiscernible] infections. However, I think it's reassuring that there aren't other opportunistic infections that have popped up in this pretty large clinical development program.

In addition, I think we need to keep in mind, also when you look at the totality of infections and particularly serious infections across the fenebrutinib clinical development program, which also included older patients with PPMS, there really wasn't a different scene with respect to serious infection.

So although it's a rare infection and potentially can be related, when you look at the totality of what was observed with respect to infections, which is obviously the adverse event that we tend to pay most attention to, it doesn't seem like there is some signal for severe immunocompromisation as well as opportunistic infections.

Thanks, Dr. Oh. I'm happy to take the question on PPMS liver enzyme data versus RMS. And you are correct that there is a big difference in age of the populations that we've studied in the RMS study versus the PPMS study. In the RMS population, the average age was approximately 35, 36 versus in the primary progressive MS trial in which we enroll patients up to age 65, and the average age was 49. So you're correct that age does play a factor.

And also, it is a signifier of also the increased comorbidities that you see in the PPMS population and polypharmacy that you see in the PPMS population and older population. So it's probably those three things together: age, comorbidities and [indiscernible] that are really contributing to that. So we found that unsurprising to see that difference between the RMS and PPMS populations.

Alex, did we answer all your questions?

Yes.

Then let's move on. Next question go to Sachin Jain from Bank of America.

A couple more on safety from the Roche of side and then one for Dr. Oh. So maybe if you could just provide any color on your FDA interactions around safety. As you can tell from the call, investors clearly worried by the safety profile and approvability. Just wonder how you would frame your dialogue with the regulator, I'm assuming, ongoing around the safety issues that have come up?

And then the second question is, as we think about what questions the FDA may ask on a potential outcome, just wondering what additional work you can do prior to submission to address what questions you think they may ask on suicide and infection? Would you present that data? Because obviously, we've only got limited information on the infection and suicidality so far? And then how you're thinking about post-marketing commitments as you put the package together. So those are the questions for Roche folks.

And just one simple one for Dr. Oh. The PPMS tool will be unique to fene within the BTKs, to what extent does that data influence your potential adoption in RMS?

Would you like to start?

Sure. So in the MS field as clinicians, we're recognizing more and more that the current classifications that we have of MS are not adequate nowadays. Just because when we designate someone as relapsing MS, we often simultaneously very early on in disease see clinical and when you have the tools imaging evidence of progressive disease biology. And then in progressive MS and PPMS is considered to be probably the most purely progressive form of MS, we often see evidence of relapsing disease biology as well.

So because of this, we know that MS is a disease continuum. And having the totality of the data available from fenebrutinib where it really has gone from really early relapsing MS all the way to the most purely progressive form of MS, which is PPMS and to demonstrate a clear effect across that spectrum, that is what gives us reassurance that this is a drug that probably will be beneficial from the beginning to kind of the very later progressive stages of disease. So from kind of a clinical standpoint, it gives us reassurance that individuals will likely benefit from fenebrutinib probably for different reasons across the entire spectrum of disease.

The issue though, of course, is that, in the real world, usually medications are approved for specific indications. And so when you have an indication both for PPMS as well as relapsing forms of MS and within relapsing forms of MS, we, as clinicians tend to be pretty liberal with what we designate as relapsing forms of MS because, again, we know that relapsing disease biology happens across the spectrum as is progressive disease biology, this just gives us flexibility to use the drug, again, pretty liberally across the spectrum of MS.

Thanks, Dr. Oh. And we appreciate the interest in how things are going with the FDA. Of course, we were following the typical processes for interactions with the FDA and health authorities. We are very early in that process. As we've shared previously, we did share the top line data with the FDA right away following our data readouts. And currently, we are in the process of putting together our filing dossier. So we're in contact, but it is early to comment on that before we are in -- formally in the final process.

Levi, is there anything you would like to add to that?

No.

Sachin, did this help?

I guess the second question was just what additional work do you plan on doing around the suicidality infection? And would you present that at some point?

Thank you. We, of course, have an ongoing open-label extension study in which we are following the patients from the original Phase II FENopta study as well as the relapsing studies and the primary progressive study. We are following all of those patients in an open-label extension study, including patients that were originally on the active comparator and have now switched to fenebrutinib.

So we think that will be very enlightening on confirming or refuting the signals that we're seeing in the double-line treatment period. Whether or not we need to do post-approval commitment studies, there will be something determined in filing, that would, of course, be very typical for MS drugs. Most, if not all, of whom have a post-approval commitment studies. What those will be will come out in -- during the final process.

Then let's move on. And the next question is going to Peter Verdult from BNP Paribas.

Peter Verdult here from BNP. Two questions. So Alexandra or Levi, I know your focus right now on the filings, but if you allow me just to look ahead, is there any future clinical development that you want to do with fene in SPMS, within MS or any indications beyond MS? That's question #1.

And then Dr. Oh, forgive me, for returning to the first question, but pushing you forward maybe a little bit of numbers. But assuming fenebrutinib is approved end of the year, can you give us some ballpark numbers in regards from [indiscernible] MS and PPMS in terms of what proportion of new or current patients would you consider starting on fene or switching? If I could push you with some numbers, that would be appreciated.

I'm happy to start talking about the future of fenebrutinib development program. We are exploring different opportunities because of Bruton's tyrosine kinase is an excellent target for many autoimmune diseases. So we are exploring where that could fit well with the Roche portfolio.

Within MS specifically, we, at Roche feel very much like Dr. Oh. that MS is one disease, and there is relapsing and progressive biology in all patients starting from very beginning. So medically and clinically, we think we have the data necessary already to really guide physicians on use of fenebrutinib in clinic.

So at the moment, there is no plans for a SPMS study.

I can add with that. I think Dr. Oh and Alex said this very well. Certainly, there may be additional studies that we would do. For example, here at AAN, there have been conversations about can one refine understanding of the difference between the inflammatory component and the progressive components. So there may be other things we can do to fully elaborate that and the mechanisms that are happening and that may help inform in practice the conversation about which -- whether to use fenebrutinib, for example. But from the same point of additional states in subtypes, fully agree that we don't think that will be indicated.

And I can answer the kind of ballpark numbers fenebrutinib cases. So these are really, really ballpark numbers. But when it comes to choice of treatment when a patient is treatment naive, so newly diagnosed, I think very clearly, the field is evolving to start with high-efficacy treatment in almost everybody with a new diagnosis of relapsing MS from the beginning.

And what we currently have now that is probably the most widely used treatment in general are the anti-CD20 agents, and there's a number of them. But fenebrutinib would very early be a choice that patients would make from amongst these high-efficacy treatments against relapsing disease biology.

And that choice is often made by patients, to be honest, just because people have very like strong preferences about mode of administration and how frequently they need to take therapy, some people just really want to be on oral therapies and don't want to think about having to go into an infusion center and needles and all of this. And then some people like the freedom of just having to go somewhere once every 6 months and not having to take it off.

So probably in a decent proportion of newly diagnosed people with MS, fenebrutinib would be a very good drug because it has high efficacy against relapsing disease, and we have excellent evidence in the field, demonstrating how important it is for most people from the very beginning of the disease to start treatment with high efficacy against relapsing disease activity, which tends to be the dominant driver of disease in very, very early stages.

I wanted to add to that, though, in people who, even from the beginning of disease, clearly have evidence of some progressive components of disease, fenebrutinib will then be the treatment of choice because you have the benefit of having the agent that has very high efficacy against relapsing disease but also has an effect on progressive disease mechanisms, particularly when we take into account what we know about the drug, which is that it easily gets into the central nervous system. And what we know from the FENtrepid trial, where it was non-inferior to oprolizumab, but there is a clear suggestion that there was a benefit on these progressive disease processes in the FENtrepid trial.

With respect to agents who are switching, reasons were switching will vary quite a bit. But again, given what we are understanding in the field that currently, we have any people on high efficacy treatments, and they demonstrate clear evidence of slow progression over time. Fenebrutinib would probably be the first treatment option in that setting. So for treatment sequencing because we're reassured that it is very potent against relapsing disease activity, but again, has that benefit of targeting progressive disease mechanisms that many of our current therapies cannot really do in a meaningful way.

So ballpark figures, I'd say in treatment naive, probably 30% to 50% of people would consider starting on a drug like fenebrutinib so long as the safety profile is similar to what we used with the current high-efficacy treatment options.

And then in many switches, obviously, not necessarily switches related to family planning or tolerability, things like that, but many switches can be related to efficacy. In fact, I'd say probably 50% to 60% of switches tend to be related to efficacy. This would be a very natural probably one of the top choices. So I would say over 50% of people in a first switch setting would probably consider a drug like fenebrutinib given what we've seen with the clinical trial data as well as what we know about [ the medicine ].

Sorry, I mentioned you got a lot of ballpark numbers here for your model, Peter. Any additional questions or...

Well, I could try my luck and say do it again through PPMS, but I will be cheeky. So I'll refrain and let the next person ask that question or do something else.

Okay, very good. Let's move on. Next question will go to James Quigley from Goldman Sachs.

I've got a couple of follow-ups as well here. So on the suicides, I mean, you've been pretty clear, you don't think there is a biologic rationale for BTK and driving potential suicide ideation. But did you take any measurements of mood changes or changes in depression levels as part of the adverse events monitoring for FENhance of FENtrepid? Did that data give you any more confidence that there is no or very low suicidality risk?

And the second question for Dr. Oh, how do you compare the disability progression data across trials? Obviously, taking into account the limitation of cross-trial comparison, but in the Kesimpta trials versus Aubagio, there was a 34% benefit in CDP EDSS score while fenebrutinib, but when we look at the CDP EDSS, the benefit was around 19% to 23% across both the trials and then also in PPMS, numerically higher reduction versus [indiscernible] as you said. So how do you extrapolate these data through to the MS population -- the RMS population when you're thinking about disease and disability progression?

I will start with the question about mood in the trials. As was stated earlier, we did not have this previewed as there had not been any signal of depression or suicidality in our early phases of studies. So the Phase III study was not set up in order to evaluate mood. There was -- we did not have a mood scale there.

Going forward, there was a question earlier about whether or not we would be doing future studies to evaluate this. If we were to do so, perhaps including a mood scale to be able to evaluate depression and evaluate whether it gets worse would be something that we would consider. That is not being planned now, but throughout our filing processes, if such a study came forward, including an mood scale would be very appropriate.

We did have the C-SSRS scale as is recommended by the FDA for any neuro treatment that's in development. And there, we saw -- it's not a depression scale. There is suicidal ideation to evaluate that. And in the C-SSRS scale, we did not see any clear trend in suicidal ideation. So you would kind of think mechanistically that if there was a biologic reason for this, that you would see patients would worsen in ideation, patients would worsen in behavior. And this we didn't see. We have these rare events, but we don't have a clear signal on suicidal ideation.

So to us, this shows that there are gaps. And so perhaps we would include, as you suggest, we have some indicators in major studies, but we don't have that data today.

And I can address the -- comparing disability progression across the clinical trials. Well, there were always talks in school to never compare across clinical trials. And particularly with disability progression, it's tough because there really is a period effect related to clinical trials these days.

When you looked simply at relapse rates in the placebo arms of relapsing MS clinical trials in the past 2 decades, you see a dramatic decline over time. And this is directly related to the fact that our diagnostic criteria continue to evolve over time, more widespread availability of MRI and kind of loosening of MRI criteria required to come to a diagnosis of MS. And so overall, this has really resulted in the face of MS being completely different from what it was actually 2 decades ago.

And so what we know is that we're in clinical trials, we're proving mild -- milder and milder population of people with MS. And this is reflected in the baseline characteristics of the FENhance clinical trials, where generally, the average age is similar. But clearly, people have less disability and are really within a year of MS diagnosis in these clinical trials. So already, you have a milder MS population.

And so with that, it becomes harder to detect disability progression when people are still early in their disease course. And so this is why it's challenging to compare disability progression, which is the harder measure by far in comparison to the clinical measure relapses. It's very challenging to compare that across clinical trials.

I also wanted to note that in contemporary clinical trials, we do see this issue of problems detecting differences in disability progression, even though we know clearly, these are drugs that have a profound effect on relapses and therefore, should also have an effect on disability progression.

Take, for example, the ublituximab Phase III clinical trials, which were the most recent anti-CD20 clinical trials, where very clearly, there was an effect, a profound effect on relapse in disease activity clinically, but you don't see that corresponding effect on disability progression. And this is similar to this temporary to the study population of fenebrutinib.

So all this to say, I think it's really challenging to look at numbers and try to compare across clinical trials because of this evolution of what MS look like, the milder disease population. And so it would be a disservice to draw conclusions based on just that magnitude of that on disability progression, which is already a very challenging measure.

And I just want to add, I guess, probably agree with Dr. Oh that crosstalk compares have had lots of biases. It's important to note FENhance study when compared. So we've been the relapsing MS [indiscernible] compared to [indiscernible], which has already shown a [indiscernible] benefit of 30% of placebo in the past and our [indiscernible] MS [ compares ] where we have shown 24% visibility improvement better placebo. It's already in our studies, we have raised the bar in terms of measuring the slow progression.

Very good. James, does that answer all your questions?

Yes, it did.

Next questions go to Stephen Scala from Cowen.

Congratulations on the data. I do have two questions, but they are both on safety. So first, Dr. Oh, I'm just curious, have you ever seen a trial in neurology with such a significant adverse SKU in deaths and have it just be a random finding and that, that drug was ultimately approved despite that finding?

And then the second question, perhaps to the Roche folks, was the liver monitoring instituted while the trial was already underway? And if yes, how many liver enzyme elevations were observed before that implementation and how many were after that liver function monitoring implementation?

I can start with that right away. You're very astute in remembering that this was implemented during the clinical trial process. We started all 3 of the Phase III studies with monthly monitoring. And at the time that we implemented the change to every 2-week monitoring, we had well recruited in the studies that were over 50% recruited. The PPMS trial was probably over 70% recruited at that time so that the data that we're looking at in terms of liver enzymes is predominantly collected with the monthly monitoring.

Once we implemented the every 2-week monitoring, we did see a shift in detecting the liver elevations earlier, so we -- which also gives us the reassurance that our every 2-week monitoring is working. After we implemented the every 2-week monitoring, we did not see any further high elevations. And of course, the single highest block case occurred while monthly monitoring is in place.

I can address the question about death. So I don't know all the numbers by heart. But in recent memory, I don't think I have seen an MS Phase III clinical ride with this significant imbalance, this magnitude of an imbalance in deaths. However, I think these rare, but serious events, by chance, can result in these imbalances that we see. And so this why it is so important to tease apart each and every single case to see if there is some plausible mechanism that may be relevant to fenebrutinib.

And as you saw from the FENhance data as well as the FENtrepid data, there really isn't any sort of clustering of events. It is a little bit odd that there is this -- just based on sheer numbers alone and imbalance. But I think the most responsible thing to do is to tease apart every single case. And overall, there didn't seem to be that is relevant to what we know about the mechanism of BTK inhibitor. So obviously, this will need to continue to be evaluated, but the short answer is no, I have not seen and any recent Phase III trial with the similar magnitude of deaths.

Then we go to the final questions. Final questions go to Urban Fritsche from ZKB.

A question on the subgroup analysis. Do you have any idea of why prior DMT use could be beneficial for fenebrutinib treatment? And could this data be informative for the real-world sequencing of therapies, which indeed will be a bit contradicting to your prior comments that fenebrutinib would be ideal for naive patients?

Perhaps I can just comment before handing this to Dr. Oh, that, of course, the -- as Dr. Oh described in the baseline characteristics, patients previously treated with DMTs was a very small proportion. And because of that, we have to be cautious when making inferences there that their confidence intervals around that group are large. So just reminding us that that's actually a small data set, which we were not powered to make conclusions with. Dr. Oh, any further comments there?

No. I mean I think it was really only 15% of the entire setting population that contributed to that subgroup. So I completely agree. I think we just need to be cautious about which subgroup analyses are more meaningful than others just based on first.

Excellent. I think with that, we are at the end of our call. I would like to thank again all of our speakers for their time and their efforts exploring new treatment options for MS patients. Let me also thank the IR team members who worked on the slides and prepared the event to call out here, [ Alex Mora ] and Loren Kalm and also [ Les Lloyd ] for event organization. I hope event was helpful, providing a timely update of our MS franchise.

If there are any remaining questions, please reach out to the Roche IR team. We will be happy to help. And otherwise, we're about to meet again soon at our Q1 results call tomorrow or on one of the roadshows on Friday.

With that, we'll close the call. Have a good evening. Have a good day. Bye-bye.

Awesome. So let's get started. Thank you, everyone, for joining us this morning. Kicking off day 3 of the Morgan Stanley TMT Conference for me. Very pleased to be hosting from Roper Technologies, Neil Hunn, President and CEO; and Jason Conley, EVP and CFO. So gentlemen, thank you so much for joining.

So maybe just as sort of to open up the conversation for maybe some people in the room maybe not as familiar with Roper Technologies, a little bit of a kind of one-on-one. This is a little bit of a different story from a lot of the [indiscernible]. What is the Roper story all about?

Yes. Thanks for having us. It's always great to be here, Keith. So yes, those that are newer to our story, Roper is this year, roughly $8.5 billion business, 40% EBITDA margins, 30-plus percent free cash flow margins. And we're a durable, steady, vertical market software compounder. I think that's what's unique about sort of our model versus others. So really a dual thread offense sort of steady organic cash flow generation and then we take all the cash flow generation and deploy it offensively.

Most of the time, historically through M&A, more recently through buyback. When you put it together, sort of we think about how durably and consistently can we compound our free cash flow per share. That's been in the mid-teens area, and we have the strategies and operational rigor in place to try to push that to be in the high teens area. In terms of the portfolio construct itself, it's vertical markets, right? So it's -- all of our businesses are leaders in small markets. We're super vertical-oriented, application-specific, high retention, high gross margins, high cash flow dynamics attached. And we'll get into all that, but that's a flyby of Roper, durable compounding vertical market business.

Got it. And you guys often talk about edge markets as kind of where you're looking for those either market leaders or emerging market leaders. How do you define [indiscernible] market? And within that, what are the key kind of KPIs? What are the key attributes you're looking for in terms of an emerging leader or a leader in that market?

Sure. So just to double-click on the high level. So we love owning businesses that are leaders in small markets. So small market, we mean small, like the largest TAM that we have in the portfolio of 29 businesses is about $4 billion, plus or minus. Like a lot of our TAMs are $1 billion plus or minus. So that is the first thing as it's a very small TAM. So these are TAM like software for early childhood education centers, software for autism therapy clinics, software for large law firms, software for tax accounting for publicly owned utilities. I mean these are niches within niches and very, very sort of specialized. And we like these small markets for 2 reasons. One is they are protective.

So very rarely do we have a new entrant because the size of the prize is quite small, and they're well served by us. But also, I think more importantly, it means we compete on this notion of customer intimacy. Like we are just ingrained in the operations of what our customer does across the totality of the portfolio, our customers can't do what they do without us, right? So we are sort of the backbone of their business, system of record or whatever network that they operate within. So that's what we mean niche and sort of edge market. That's what it is. small markets where we have a clear leadership position.

Got it. I'm going to skip around the question list because I think that brings up a really interesting point. And let's just get straight into what has been sort of the major debate point of the TMT confidence really within the software industry for the last 2 years is the -- how insulated certain markets are or how well positioned certain markets are for AI. So when you're talking about these smaller markets feels more protected, do you guys see that as a more insulated kind of market from an emerging competition from a large language model lab? Yes, that's basically the question.

So our empirical evidence to date is absolutely, right? So again, we have 29 business in the portfolio, 21 are vertical market software businesses. The balance are sort of application-specific product businesses. So relative to the 21 -- it is our -- we're not aware of a model company or an AI native start-up that's attacking the core of what we do. I think there's logic behind that, by the way, which is if you're an investor and you're starting company, do you want to really sort of attack, if you will, a system of record and a $1 billion TAM that's well served, your rate of growth is going to be sort of rate limited by dislodging of a strong incumbent that has all the incumbent advantages. And so we haven't seen any of that across the portfolio in any capacity. So it proves to be quite insular in that regard.

I think more importantly, though, I'm sure in this reading your recap notes for the past couple of days, I don't think what we're about to say is like unique from what other companies are seeing, incumbency has like real tangible advantages as we're trying to develop AI tools at whatever term you want to use, the agentic layer, the task augmentation layer, right? So at the highest level, I've said this for a couple of years, we've said this for a couple of years, it's a combination of 2 things, and everybody focuses on the first one, which is like do you have the data, the knowledge graph, the context to sort of take a sort of a predictive model, make it deterministic because that's what has to happen when you're augmenting and replacing work.

But I think what also is often forgotten about is you have to know where to point that capability to and what problem to solve. And these are like very esoteric "boring tasks" that are repetitive, but it's truly like magic to our customers. Like how do you automate account reconciliations that take 17 minutes into a minute? Or how do you auto verify order entry for pharmaceuticals, which is done by a pharmacist today, a high-dollar pharmacist? Or how do you take 10 phone calls out of matching a freight load in the spot market and do it with no calls? And so these are like it's truly magic, but you have the capability, the knowledge, the data, the context, but the problem to point it to. And in these little niches that we sort of tend to own and that we're the #1 player again in each of them, it's -- that's the magic of what's happening, right? So lots of opportunity.

And back to deterministic, I mean, there's a regulatory compliance overlay in almost all of our vertical market businesses, and that's very hard to replicate. That's your decades and decades of getting compliance. So our customers don't want to sort of violate any sort of regulatory hurdles, and it's harder to get in. It just increases our moat.

Yes. And that was exactly what kind of triggered me to jump down into AI. When you're talking about intimately knowing your customers, software is all about solving your customers' problems and solving the sort of the business problems and making those businesses more efficiently. And you're right, everybody just focuses on the data. And there's a little bit of an inconsistency in that argument because ultimately, it's the customer's data, right? It's not -- you're managing it for them, you have visibility into it.

But what you have is that domain expertise, you understand these domains. And from my perspective, generative AI, the real opportunity is not to displace existing transactional systems is to automate workflows around those systems, and you need to know the problems to be automated. You need to know the business process to be automated. So the question in all of that is, I like it that you think about the world in the same way I do. But how do you push that down into this distributed portfolio of companies?

Yes. So very important context setting. So let me just take a minute to sort of talk about our org structure and why we operate the way we do. And then in this regard, it's -- I'll say it's better to be sort of lucky than good in the context of what we have to do with AI in the org structure. So because of the portfolio we have, like I mentioned, we're the largest player across all 29 of our markets, which means this is over a long arc of time. This has been since I've been at the company 15 years, Jason, 20. We've been in the modern day Roper for going on 25 years. And so all of our competitors, by definition, are smaller, which means probably they're generally a little more nimble.

So our entire organizational structure from the dawn of the modern day Roper been super decentralized, right? To optimize on 2 things: speed coefficient and proximity to customer, right? So 20,000 people in the company, 130 in the corporate center. So every other resource in the organization is closest to the customer, organized around speed. Like that's been the design principle for 25 years. So that's why I say it's better to be lucky than good. If we happen to have been organized for like cost of development or more centralized resources and not on customer centricity and speed based on our heritage, I think we have a real problem in the AI era, but we're not. And so we're -- as soon as the sort of light bulb goes on inside of our organization -- organizations, then the speed coefficient sort of builds on itself.

And so going on now 2 years ago, we said to our organization, all 29 of our leaders, like this AI thing is not a trendy paradigm. It's the next tech wave. We have to lean into it. And we've had each business now 1.5 years ago reunderwrite their entire business model, be AI native to get the thinking going. We've had them sort of change the way they do development. So smaller teams closer to the customer, collapsing product and development, sort of minimizing design, making that part of product, all the things that you do across all 21 of the software businesses. And so we really sort of have this increasing sort of flywheel speed happening at the point of impact. Now what we also identified is this is in the last maybe year, and we just started hiring the team about 6 months ago, where we were hiring sort of an AI accelerator tech team at the center. We'll be small to start, 20 or so people.

We're fortunate enough to have 2 exceptional leaders come from another large application software business to sort of lead the team for us. And what we found is as our businesses were experimenting and learning how to develop an AI, it's a very different development pattern pathway than traditional development, where you can have a -- traditionally, you can have a sort of an architecture and sort of deterministically work your way through. There's a little bit of art in AI development. And so we very much wanted to sort of have a team that has sort of grew up in machine learning, trained in machine learning and trained in AI, had done this at scale in the application layer. And so their mandate in the organization is to accelerate the technical teachings to speed things along even further, but also have a build team like they're going to be a small sort of, if you will, "10x" build team that partners with our companies, and then finally, there's going to be a shared Roper repo.

When you have common functions like whether it is document retrieval or document sort of parsing, that's a common AI sort of run time routine. There will be a central repo of that, so we don't have to multiply that times 21 -- times across the organization. So sort of this sort of perfect storm of our legacy close to customer, have speed built into the organization and accelerating it with the center team.

It almost reminds me a lot of companies that we're talking to have really brought forward the concept of forward deployed engineers, right? You need people who are really skilled in the AI to be able to help the businesses, right, and usually in the customer concept to understand how to utilize these technologies that really don't work very much in the same way the traditional technologies work. You guys do within your portfolio companies. You guys have talked about approximately 25 Gen AI initiatives underway within the broader organization. Any like early wins, any initiatives that are seeing significant or early traction, and on the other side of the equation, where do you guys see the white space? Like what looks most interesting in terms of new areas to go out and automate?

Yes, sure. So when I talked about that and we talked about this 25, that was probably 6 months ago. It was the culmination of the early product development work we've done in the organization. So this is product-oriented, playing offense sort of monetizing in the tech stack. That number is basically not discernible now. It explodes. So we're not going to talk to the 25, then 40 to 50 to 100, but it's much more than that now. Going back to a couple of years ago, we made the early decision that between using AI to drive play offense and put in the product stack and grow and expand versus use AI for productivity, we're tilting demonstrably more towards playing offense. And so everything that we've talked about today has been about that. So that's where the principal amount of our focus is.

Obviously, we're working like every company on coding sort of efficiency and customer service and sales. But that's -- we make -- we have 40% margins. Like we're not about -- we will improve margins over time, take that productivity, put it into the product road map, but the principal of our focus is playing offense. So a couple of examples where we've seen real success there. I'll give you maybe 2 or 3, if it's okay, to make it sort of tangible. I'd start with the best story, which is our autism therapy business is called CentralReach. This is the backbone that autism therapists sort of run their practice on.

To set the context here, there's something like 800 million therapy hours demanded a year in the U.S. and 300 million therapy hours supplied. So a huge supply-demand imbalance, waitlist sort of people -- these families waiting for access to care. And also inside of the care, the people that deliver the care, they turn over 85% a year. So it's a very difficult problem. And so the tools, the AI tools that Central Reach has developed and deployed into the market is a scheduling tool. It's a very complicated schedule. When you're -- you have a demand and your people are turning over all the time, you can't have a blank and schedule. It's -- and by the way, these learners are in between 2 and 8 hours a day for 6 months.

So if you have one person leaving a huge block in the schedule, it's a problem. Also all the note taking and sort of medical records and then all the revenue cycle back in the billing and collections. The punchline of this is the business has increased its growth rate from the low 20s to the high 20s. Its win rate in the second half of last year from the primary competitor went from 75% to 100%. And fortunately, for our customers, the attrition rate for the therapist has gone from 85% down to 40% because it takes a lot of the bad part of the job out. So it's like this win-win-win sort of in terms of customer, the market, the learners, the families and sort of our business.

Another example I would share is earlier days, but we're very excited about it. We have a business that is -- it is the technology that -- and the network that organizes the spot freight market in the U.S. So if you're a broker or a trucker, you subscribe to our network to find out where your load is. And today, traditionally, it takes about 10 phone calls for a load to be brokered. Are you available? Can you pick it up at 3? Will you do it at this rate? Oh, I need you there at 4. Can you deliver it by Tuesday at 3? They take those phone calls. And on average, that's about $100 or $200 of labor to broker a load at the broker level. We now have in place the technology to do that without a human in the loop. We charge a fraction of that cost and of the labor cost, and that's happening in the real world.

It's a magic button that sort of takes that inefficient sort of human connectivity out of loop to sort of accelerate the brokerage, the brokering of loads, which we think, by the way, it's great for a broker, it's great for a trucker, but actually makes the whole spot market a more fluid market. So we actually think there could be more volume that actually comes into the spot market. The spot market won't be the market of last resort. It could be the market of sort of middle resort because it's a more fluid, transactionally sort of dynamic market because -- now this, by the way, is very expensive technology. It's a lot of machine learning. It has to be absolutely deterministic.

You have to have the volume of rate data because part of this is how do you price a lane. We have happened to have in our network because of the scale of it, like 7x relative market share to our largest competitor. We know what the rates are per lane, per day by truck type, and if you don't have that, you can't do the match, right? So it's this combination of technology and data. I can give you other examples, but those are 2 that we're quite excited about.

I think what's exciting is the companies that are farthest along in AI are growing the fastest in our portfolio. So just to give you a sense of like we have this incumbency right to win. And so it gets our other businesses really excited when they see that and it fostered a lot of learning across the portfolio.

Got it. Have you guys had to adjust the pricing model? There's another big debate in software is kind of what happens with pricing models. But have you guys come up with new pricing models, more outcome-based, more consumptive-based to align to the newer technologies?

I mean we're certainly doing a lot of voice of the customer work on this right now in some of our businesses and just looking at where the puck is going and how we should be thinking, especially in the businesses where we have the highest right to sort of win that game and increase the productivity of our customers where you could see some seat degradation over time. And so I think what we're seeing is that there's just pointing it out another direction, like maybe it's a number of customers, maybe it's a number of whatever the driver of that business, maybe a number of policies, that sort of thing.

And our voice of the customer work so far has said that as long as we are capturing value for them, we get our fair share of that value. Our position on this, by the way, is to get the customers really excited about the product and really using the product before we sort of capture our share of value, again, just to continue to protect our moat and then play the long game in terms of AI monetization.

The only thing I'd add on top of that, certainly on the system of work agentic layer, our clients are telling us very clearly like, yes, we'll pay you for the value, but we don't -- we're not that particularly excited about an open-ended pay as you go. So it's probably going to be some sort of subscription with overage or something. So they have some certainty of what they're going to spend.

Tiers.

Some tiers, exactly.

I think it's a part of the equation a lot of investors have over-rotated on the idea that everything needs to move to a consumption model because when you talk to customers, customers don't love...

They don't want to...

Because they look at it as a variable -- a variable cost model, they'd much rather have certainty in terms [indiscernible] and enables them to get leverage on the technology.

And our current cost as a percent of their revenue on average for enterprise software businesses is about 50 to 100 basis points. We're not -- we're not a huge part of their cost part today. So we have latitude if we're demonstrating value to capture that value over in a fixed way if it turns out that way.

And just to double-click or just to round out the answer on the subscription plus. We are fully confident we'll be able to capture like the power users that are just driving token consumption. We're very comfortable with the margin profile of the subscription and then sort of like an old school cell phone plan. You get this many units of work. And if you go over, then you're going to pay, but there's some predictability for the customer.

Got it. Got it. I want to switch gears and talk about sort of recent results and how the company has been operating. We've been having this conversation for a couple of years running now. And I know the center of the Roper story is consistency in results, consistency in sort of the yields that you guys are able to bring to the marketplace. So let's talk about 2025. Revenue growth came in at 12%, $7.9 billion in revenues. Adjusted EBITDA grew 11% to $3.14 billion. Can you talk us through what worked well in 2025? And maybe what are some of the areas for improvement into 2026?

Sure. Yes, I can cover that. So I think as we obviously had some great business building in the year, but our growth wasn't really where we thought it was going to be at the beginning of the year. And there's a few reasons behind that. One is Deltek. It's our biggest business. It's exposed to government contractors. And that, as you know, when Doge hit and then there was nobody to deal with at the agencies with our customers not being able to talk to anybody at the agencies. And then the shutdown, that business was severely impacted in 2025. So you look at a business that has been a steady mid-single-digit plus grower and it was in the low single digits in 2025.

So a big part of our business and a big segment. And then -- so if you look at our Application Software segment, it actually organic revenue improved 70 basis points year-over-year if you sort of exclude that. Of course, you got to own all of Roper, but just to give you some context of the impact of that. Procare was a little off to a lower start. We'll talk about that a little bit later. I think Neil will touch on that. That's a business we acquired a couple of years ago, but we feel confident in the ability to inflect the growth rate there.

And then Neptune, which is in our TEP segment, our products segment, second half just impacted by tariffs and just skyrocketing input costs. And so we had to sort of process that with our customers in the second half of last year. And so again, we own it. We know we're going to do better going forward, but just a lot of things happened last year. As we look forward to 2026, we're sort of -- we're guiding organic growth 5% to 6%, taking a view that at least coming out of the gate, things aren't going to get demonstrably better. But we will have, like in application software, we've got a mid-single-digit plus. We're going to have a better comp just on Neptune alone -- or not on Neptune, on Deltek alone. So we're not expecting a lot of improvement, just a little bit better comp there.

And importantly, our CentralReach business becomes organic in the second half. So that will add 80 basis points or so to that segment. Network Software will be mid-single-digit plus. We feel very good about that. That's just our Foundry business, which was dealing with writers and actors strikes to media and entertainment. That will get better this year. And then we'll have our Subsplash business will also become organic in the fourth quarter. And then our TEP segment, that's sort of a little bit more wider range of outcomes as we sit today. We have guided low to mid-single digits with the first half being a little bit lower. We have Neptune down for the year, but we'll see how that plays out. Neptune, their water meter business, and we're waiting to see when the COVID super cycle sort of normalizes out. So that's sort of the one watch out that could provide some upside. But as we sit today, we're sort of -- that's the approach we're taking.

Got it. That the same conservative. Double-clicking on the Deltek side of the equation. A lot of the executives that I've talked to about sort of the federal disruptions in 2025 see the opposite side of the coin of that. On the other side of the disruptions of DOGE is also the idea that we got to run the government more efficiently and software is going to be a big part of the solution. Do you see an upside scenario in Deltek of that you guys become a bigger part of the solution in terms of how to garner efficiencies in the government?

I mean I think it's -- the value prop for Deltek has been compliance, right? If you're doing highly regulated projects, but also efficiency, how do you run a really tight project and increase margins. So I think we're always on the side of the angels, no matter which way you look at that. For us, though, I think -- and this is the big sort of what's going to happen is the O-Triple-B spending and when that ultimately cascades down to the contractors. We're starting to see some maybe signs of life, not at Deltek, but just in the marketplace in general. And that -- I mean that should accrue very nicely to that business.

And just so there's -- we're all clear, like our Deltek customers are the contractors who serve the federal government, not the government itself. And so O-Triple-B is an unmitigated tailwind for the industry. It's just when is the government going to be functional enough to basically award the budgeted dollars or allocated dollars and then that will trickle into our customers, which will sort of benefit us. But the timing is what the question mark is there.

Got it. Got it. And then on the capital deployment side of the equation, $3.3 billion in M&A in 2025. Could we start to get into the high-level conversation about how you guys are deploying that capital, the shifting mix of acquired assets. You guys have been talking about acquiring assets a little earlier in the life cycle. So why the shift? Why shift towards earlier stage? And should investors be concerned about having to pay a higher multiple for earlier-stage companies for that slightly higher growth?

So hold Jason and I accountable to all parts of that question. We're not avoiding if we forget to round it out. So to focus everybody, so I started by saying our cash flow compounding algorithm is sort of mid-teens, and we aspire to be high teens. That's our mantra, that's our rally cry. That's where we're focused on building a sustainable business. There's 2 levers to make that happen. The first is to improve the organic growth rate of the business by a little 100, 150 basis points, which you go to work every day to do on a sustainable basis. The second is to, if you will, recapture more value or higher returns from the capital we have to deploy, we do those 2 things, we can get to the high teens. So on -- when we look at how to capture more value, have higher returns on the amount of capital we have to deploy, it comes in 2 flavors. One is more tuck-ins. So that is a -- historically, a tuck-in was kind of like a 4-letter word inside of Roper going back to my predecessor.

They're actually the best deals we've been able to do, as you can imagine. We get back in sort of back-office G&A synergies and our adjudicating factor for a bolt-on is we only want to do a bolt-on, it's going to improve the organic growth rate of the company we're tucking into. So they become growthy and they're highly valued because of generally the G&A synergies that happen. But that's going to be -- we think when you do the time and motion study, the size of the portfolio, the digestibility of our companies, we think that could be about 1/3 of our capital deployed over a long arc of time.

The balance of the 2/3 will be platforms. And as we look at the range -- array of platform opportunities, whether it's, call it, a business as usual pre-2022 Roper-style deal, call it, 5% to 7% organic growth business, optimized margins that you pay a certain price for or you buy that same business just one owner earlier. where you have -- you're catching with a higher growth rate and you still have margin opportunity. Yes, on tZERO, it's a higher multiple. But on T5, it's substantially lower, like 30% to 50% lower because you have the value creation of both the growth and the margin improvement that happens. And so that's how we get the compounding flywheel to happen and sort of accelerate the compounding on the capital deployment front.

Got it. Got it. So not necessarily -- maybe a little bit of near-term pain, but long-term gain because you guys get to apply more of your expertise in terms of improving these businesses. And as long as you have patience in that capital deployment, the yield should be higher. So maybe let's apply that framework to some of the recent M&A, start with CentralReach. You talked about that as one of the kind of AI use cases of where you guys could add a lot of value by optimizing what is a really tough supply and demand imbalance. Can you talk to us about sort of those 2 vectors in terms of how do you guys improve kind of the revenue growth profile there? And sort of what are the main levers for margin improvement?

This is a business that is a mid-20% growth business for us. Its margins are -- can -- will be optimized, but this is more about how they scale their margins versus they're being inefficiently run sort of out of the gate. For us, it's about -- when you own a business forever, right, every business we bought has come from private equity and they own a business for 3 to 5 years, and they have to invest in a sine wave, invest for 2 or 3 years harvest, invest 2 or 3 years of harvest. And CentralReach and the others are no different. In our case, we get to have sort of Horizon 1, 2 and 3 sort of investments, and in this particular case, we have great near term, like next 5- to 7-year dynamics in supply and demand of the learners and the caregivers of the market position that we talked about before.

So the next intermediate period of time feels pretty solid. But as the market dynamics in this business shift and supply and demand become more balanced, then all of a sudden, what will happen like it happens in every class of trade in health care is people start discerning on quality of care, not just can you get access to care. So in the first year of ownership, we bought 2 assets, 2 small assets that are leading the industry and how you discern quality of care. And so not only will -- so the vision there is the CentralReach's customers won't just be sort of access to care, but they'll be the highest quality sort of caregivers, and that will sort of drive sort of the extended revenue growth there, for instance, is a way to do that. Again, that's a Horizon 3 investment, but that's where you have a long-term ownership will pay dividends over time.

Got it. Then the next big deal with Subsplash. It sounds like that's more in the vein of the maturing leaders. That's just the traditional Roper strategy?

So no, this is the new strategy. So we consider CentralReach and Subsplash very much the sort of one owner earlier. Subsplash is the software that churches run on. So it's the church management system. It's the engagement, both online and offline. It's the giving platform. It's all that integrated. It's beautifully designed software across all of this. This is a business that has 20,000 or 25,000 churches as customers, there's several million churches in the world, and it's a wildly underpenetrated category from a technology point of view.

So the top line growth is all about the go-to-market motion and just and dialing in the unit economics of that then pouring gas on it. This is also one where we underwrote pretty substantial improvement in margins. So this was run by the prior owner to not optimize margins or even sort of I would say modestly optimize margins. It was like a high 20% margin business. We think in the duration of this business, it can be high 30s to low 40s. And some of the very first value creation levers we pulled were to move margins higher, especially in the payments apparatus. And so -- and we've had some great success in that regard.

Outstanding. All right. And Jason, you previewed this before, Procare. It sounds like that got off to a little bit of a slower start. What are some of the lessons learned there? And what's the plan for turning that one around?

Yes. Well, Procare was our first maturing leader. And I think what we've learned and we've taken that to the next 2 businesses is really just around like if the team sort of got you from A to B, can they get you from B to C. And so we didn't move fast enough, I think. And so we've made those changes when we need to. And that just becomes from our history where we were a little bit more differential to management. And so kind of turning that crank a little bit more.

Our governance and our telemetry and our just cadence around the value creation plan is much tighter now. So we just learned lessons on how to stand that up and how to operationalize that. And so those are the probably the 2 biggest lessons. And then I think the other part is just working with management on the value creation thesis to the value creation plan and having that be really tight. And we did -- I think we did a really good job with that CentralReach and Subsplash such that in the second week of ownership, we know exactly where we need to go, and we know exactly if things are off the track, how do we do countermeasures and how do we partner together to work on those solutions rapidly.

Outstanding. So if you look into 2026, how does the M&A pipeline look? And what's going on with multiples? I mean we definitely see what's going on in the public market with multiples. Is that getting reflected at all into the private markets?

Yes. What has changed since the -- since how long ago we did our earnings call 5 or 6 weeks ago, it was on the call, we're like, hey, this is as busy as we've been in a long time in terms of just the sector of -- there's not a whole lot that happened in the last 4 months of last year, but there's a lot of organizing and a lot of pipeline building and a lot of meetings and a lot of processes that we're launching. And then we did earnings call, that was absolutely the case. And then what's happened in the public markets, just the private sellers have just tapped the brakes on that as you expect. I mean that is -- so right now, there's not a lot of activity and everybody is just trying to understand where valuations are.

If you're a private equity seller and the types of assets that we're owners of, you're under no stress, right? You're not a venture, you're not running out of money. There is some option value to wait to sort of see where the world sort of settles. That said, there is -- this is now we're in the -- solidly into the third year of very serious LP pressure on the asset class. And so that's -- so they will have to -- this will settle itself. It's hard to predict what timing. I don't know if it's second half of this year, second quarter this year or the next year, which is okay. We're patient. We've always been patient.

We've never viewed our capital allocation capacity as a budget because we're owned businesses forever. You have to sort of be patient, find the right asset, the right quality asset at the right price. In the meantime, the markets presented an opportunity for us and a lot of other software companies to be pretty aggressive in the buyback, which we've been. So I think between Q4 and what was then we put in our K, it's been $1.8 billion-ish buyback, is what we've done in the last 5 months or so, and that's what, 5-ish percent of our shares outstanding. So it's sort of a once-in-a-generation opportunity for us.

Got it. I have a couple more questions. I want to take the opportunity to see if anyone in the audience has questions.

Yes. You guys talked [indiscernible] what have you guys learned over time...

Do you want to take that to...

You can start. I have got it up.

Yes. So I think it's -- the process is -- I think if it starts -- a lot of that starts with really talking to the customers, right? So you look at sort of the competitive positioning, you look at the Net Promoter Score, you understand sort of the key purchasing criteria and how each customer -- competitor sort of ranks on that. So generally, we're buying again the largest player that has -- that wins on the key purchasing criteria, whether it's features or integration, whatever it is. And I think that gives you a sense they're making the right investments into the tech stack or into whatever it is or if customer support and Net Promoter Score is horrible, then they may not have enough in customer support will be sort of one.

Number two, when you run a portfolio of very like-oriented businesses, we have internal benchmarks about kind of what good generally looks like from a cost structure in a business. If you're running a $300 million vertical market business, is 3x market share of the next closest competitor. and you have 40% sales and marketing as a percent of revenue, something that's not right about that business. Conversely, if you're like 5% of revenue is R&D, something is not right about that business. And so we've got the internal benchmarks that we throw things against. Those are 2 things that come to mind.

Yes. I mean the other thing is it's really about the focus of the organization. And so the -- it's about reallocating resources where the best returns are. What we found -- pattern is that typically private equity, they do a lot of bolt-ons. Sometimes they don't have -- like we wouldn't do that bolt-on because we get to own the business forever.

They don't have a high right to win.

So it's like a new TAM. So it distracts, I think, the organization around some of these edge businesses. So our pattern is to focus and then get the resources focused on the core. And typically, you don't have to do a lot of -- you don't have to sort of surge invest for that because it's just getting focused on the things that are going to drive the most growth.

Any additional questions?

[indiscernible].

Yes, fair question. So just for those who might not have heard it, like are we over monetizing and sacrificing growth, essentially, is that a fair representation of the question? So the examples we gave on CentralReach and Subsplash, first of all, let me back up. At the enterprise layer, when you look at R&D as a percent of revenue over the last 5 years, it's gone up like 200 or 250 basis points across the entire organization, almost like 11% of revenue today, and that includes the product businesses that are like sub 5, for Apple. So from an R&D perspective, that's been going up over time steadily. The 2 assets that were -- the 2 businesses that we just described, in the case of Subsplash, this is like just wildly under monetized on how they deal with the payments infrastructure.

This is just literally like how they have the relationship with the processors and how you monetize with the processor. It's not taken away from R&D. It's just an under -- under sub-optimized sort of part of that business model that we had a pattern recognition on with 2 or 3 of our other businesses that we monetize through payments. So we're not sort of taking dollars, if you will, out of the company. We're sort of monetizing the environmental sort of factor a little bit better. And on CentralReach, like I said, that is -- that's kind of margin, is going to scale over time. There's not like a cost action that's happening inside of that business.

And we did some tuck-ins for CentralReach. So I mean if you wanted to penalize the P&L for that, but that just got this market probably 3 or 4 years ahead of time with a solution that's all about outcomes-based care. So sometimes we will do the tuck-ins to sort of get us where we need to get to with that platform. And so you could argue that's part of the investment of the business. But these businesses just typically run -- if you think about the go-to-market, a lot of what we're doing is cross-sell or we got expansions, and there's just not a lot of go-to-market costs around that because you already have the relationship with the customer.

Outstanding. Unfortunately, that takes us to the end of our time. But Jason, thank you so much for joining.

Thank you. Appreciate it.

Good morning. The Roper Technologies conference call will now begin. Today's call is being recorded. [Operator Instructions]

I would now like to turn the call over to Zack Moxcey, Vice President of Investor Relations. Please go ahead.

Good morning and thank you all for joining us as we discuss the fourth quarter and full year 2025 financial results for Roper Technologies. Joining me on the call this morning are Neil Hunn, President and Chief Executive Officer; Jason Conley, Executive Vice President and Chief Financial Officer; Brandon Cross, Vice President and Principal Accounting Officer; and Shannon O'Callaghan, Senior Vice President of Finance.

Earlier this morning, we issued a press release announcing our financial results. The press release also includes replay information for today's call. We have prepared slides to accompany today's call, which are available through the webcast and are also available on our website.

And now if you'll please turn to Page 2. We begin with our safe harbor statement. During the course of today's call, we will make forward-looking statements, which are subject to risks and uncertainties as described on this page, in our press release and in our SEC filings. You should listen to today's call in the context of that information.

And now please turn to Page 3. Today, we will discuss our results primarily on an adjusted non-GAAP and continuing operations basis. For the fourth quarter, the difference between our GAAP results and adjusted results consists of the following items: amortization of acquisition-related intangible assets and financial impacts associated with our minority investment in Indicor. Reconciliations can be found in our press release and in the appendix of this presentation on our website.

And now if you please turn to Page 4, I'll hand the call over to Neil. After our prepared remarks, we will take questions from our telephone participants. Neil?

Thank you, Zack, and thanks to everyone for joining our call. As we turn to Page 4, you'll see the topics we plan to cover today. We'll start by highlighting our Q4 and full year performance, then Jason will walk through our enterprise financials, our Q4 segment performance, our balance sheet and capital deployment capacity. Next, we'll discuss our segment highlights and introduce our 2026 guidance, and then we'll close with a few summary thoughts before opening the call for questions. So let's go ahead and get started. Next slide, please.

As we turn to Page 5, I want to highlight 3 takeaways for today's call. First, we delivered solid execution in 2025. Revenue was up 12%, EBITDA was up 11% and free cash flow was up 8%. Importantly, enterprise software bookings grew in the low double-digit range for the year, providing strength as we head into 2026. Second, we continue to invest for a long-term and sustainable growth. That said, organic growth this past year was below our expectations in 2025 and we own that. Our organizational focus and resolve are even stronger coming into this year.

We've upscaled talent, sharpened strategy and improved execution across the portfolio and that work is showing up for the enterprise. To this end, our application software businesses, save for Deltek, improved organic growth in the 70 basis point area, demonstrating broad-based growth improvements occurring within the segment. Importantly, we're not starting the year assuming organic growth will inflect in 2026 despite the traction we believe we're starting to achieve. We're going to execute and will reflect any improvement in organic growth in our guidance as it materializes throughout the year. We'll have much more to say on this later in the call.

On AI specifically, we continue to be excited about the AI product opportunity because our businesses sit directly inside mission-critical, high-frequency workflows where we already have deep domain knowledge, proprietary data and trusted distribution. So AI can move from productivity to on-stack embedded automation that improves outcomes for our customers and is highly monetizable. Importantly, our decentralized model lets each business deploy AI with the appropriate domain specificity across their various end markets.

To further accelerate our pace of AI product development, we hired Shane Luke and Eddie Raffaele to lead the Roper AI accelerator team. They will coach and partner directly with our businesses, build a small AI development strike team and leverage reasonable elements and best practices across the portfolio so we can deploy AI with increasing speed and market-specific precision while scaling what works, exciting stuff for sure.

And our third key takeaway centers on capital allocation. During 2025, we materially advanced our portfolio and foundation through capital deployment, deploying $3.3 billion towards high-quality vertical software acquisitions during the year, highlighted by CentralReach, Subsplash and several tuck-in acquisitions. Also and importantly, we leaned into opportunistic repurchases buying back 1.1 million shares for $500 million in Q4.

As we look to 2026, we have north of $6 billion of capacity for potential M&A and share repurchases. We're very encouraged by the size and quality of our acquisition pipeline, and we expect to remain active while staying highly disciplined on price and business quality. And in parallel, we'll continue to use buybacks opportunistically when they represent the most attractive risk-adjusted path to durable cash flow per share compounding.

So with that, Jason, let me turn the call over to you so you can walk through our quarterly and full year results. Jason?

Thanks, Neil, and good morning, everyone. We'll start off here with the fourth quarter results. To summarize, we finished ahead of expectations on DEPS driven by very strong margin performance. Revenue of $2.06 billion was up 10% over prior year with acquisitions contributing 5% and organic growth of 4% which was below our expectations. I'll expand on this shortly. EBITDA of $818 million was also up 10% over prior year. Notably, our core EBITDA margin expanded 60 basis points in the quarter, representing 54% incremental margin. DEPS of $5.21 was above our guidance range of $5.11 to $5.16 and up $0.40 over the prior year. Shares were reduced by $1.1 million in the quarter for repurchases, which you see partially showing up here in our diluted share count on a year-over-year basis. However, the repurchase did not impact DEPS in the quarter versus our guidance given the partial quarter share count benefit and higher interest expense.

Now if you turn with me to Slide 7, I'll walk through the Q4 segment performance. Application Software revenue grew 10% with organic growth of 4% and margins were solid, expanding 70 basis points to 42.2%. It's important to outline some details on organic revenue. Recurring revenue grew 6% in the quarter. However, nonrecurring revenue was down 8% in the quarter and was the primary driver to the lower end of our mid-single-digit outlook.

In our last call, we talked about Deltek being the big swing factor in the quarter. With the prolonged government shutdown, large GovCon commercial activity and perpetual license revenue was meaningfully impacted, leading to Deltek being up at the lower end of mid-single digits for the year as compared to the solid mid-single digit plus grower it's been over the decade that we've owned the business. That said, we are cautiously optimistic about a 2026 improvement for Deltek given both the 2025 disruptions caused by DOGE and the shutdown and the forward benefit of the OBBB appropriations coming into the market. As improvements occur, we will reflect this in our outlook.

For Network Software, revenue grew 14% with organic growth of 5%. Margins were lower at 52.8% due to the recent bolt-ons for DAT that are currently scaling into profitability. On organic revenue, recurring growth here was also 6%. The recurring performance was consistent with patterns over the last 2 quarters with mid-single-digit growth at DAT despite a muted market backdrop and steady improvement at Foundry. However, nonrecurring revenue was down 3% on lower services revenue and some customers electing to move from perpetual to SaaS which negatively impacts the quarter but benefits long-term growth and customer lifetime value.

For our Tech segment, revenue grew 6% or 5% organically while margins held flat to prior year at 34.8%. NDI outperformed in the quarter given strong demand for solutions in the cardiac ablation space, while Neptune was down slightly as expected as we comped against a stronger prior fourth quarter and worked through the final surcharge negotiations.

Now let's turn to Slide 8, where I'll summarize our 2025 full year results. 2025 was a solid year in terms of cash flow and DEPS performance, despite lower-than-expected organic revenue. Revenue posted at $7.9 billion or up 12% over prior year. Acquisitions contributed nearly 7% growth. Of note, we acquired 2 great platform businesses in CentralReach and Subsplash that will be accretive to 2026 second half organic growth. We also made 3 strategic bolt-ons for DAT that significantly automate workflow in the spot freight market and will gain adoption in the years to come which will ultimately inflect the growth rate for DAT. Organic growth was nearly 5.5%, which Neil will discuss in the segment detail.

EBITDA reached $3.1 billion or 39.8% margin and was up 11% over prior year. Of note, core margin improved 30 basis points and represented 47% incremental margin, which is in line with our long-term growth algorithm. DEPS of $20 was up 9% over prior year and reflects the top end of our 2025 guidance range provided in January despite lower organic revenue and in-year dilution from recent acquisitions. Free cash flow of nearly $2.5 billion was up 8% and represented 31% of revenue, which is in line with our initial free cash flow margin framing for the year.

This represents an 18% CAGR since 2022 or excluding the impact from Section 174. In both periods, it was at 14%. As we look forward to 2026, we expect higher growth than in 2025 through benefits from working capital and cash tax improvements. This will put us safely over 30% of revenue next year. However, Q1 will be a bit lower given timing of coupon payments for new bonds issued in the third quarter of 2025.

This, of course, does not contemplate future capital deployment towards either M&A or share repurchases, which brings us to our balance sheet discussion on Slide 9. We're entering 2026 in a strong financial position with net leverage ratio of 2.9x and ample near-term liquidity with about $300 million of cash, nearly $2.7 billion available on our revolver. With this position and strong forward cash generation, we have over $6 billion in capacity for capital deployment this year.

Regarding M&A opportunities, we've been proactive and successful in executing high-quality acquisitions for the last couple of years despite a weak M&A market. Most anticipate the market to pick up in 2026, which we view as a net positive given Roper is a home of choice for many acquisition target CEOs. Additionally, we have the attractive optionality of a share repurchase program which was authorized and commenced in the fourth quarter.

As Neil mentioned, we deployed $500 million to acquire 1.1 million shares in the quarter at an average price of just under $446. This leaves us $2.5 billion remaining on our current $3 billion authorization. We will remain agile in deploying capital to the best return for shareholders. Given the current valuation dislocation, we are now very pleased to have the buyback option available.

With that, I'll turn it back over to Neil to discuss the segment performance and outlook.

Thanks, Jason. As we turn to Page 11, let's review our Application Software segment. Revenue for the year grew by 16% in total and organic revenue grew by 5%. EBITDA margins were 42.5% and core margins improved 80 basis points in the year. For the segment, we saw recurring and reoccurring revenue grow on an organic basis 7% for the year and total organic revenue improve about 70 basis points, save for the Deltek-related market weakness, both of which provide evidence of underlying strength for this business -- for the businesses in this group.

Aderant continues to execute from a position of strength. FY '25 revenue grew in the mid-teens area with strong bookings throughout the year. Importantly, they're leaning into the right long-term work, accelerating SaaS and AI-led innovation while modernizing their tech platform and data lake.

Deltek was the primary weaker part of the story for this segment and has been straightforward all year with GovCon remaining a challenging market throughout most of 2025. That said, we view the passage of the OBBB as a positive development for the market. It should drive upside over time, but we've not included any benefit in our 2026 guidance and will monitor customer activity as the year progresses.

Vertafore had another solid year with growth driven by strong recurring revenue performance and continued execution on product and customer outcomes. Looking ahead, the team is leaning into a focused set of priorities, scaling automation, particularly AI-enabled workflow improvements while continuing to deliver steady innovation to the agency and carrier ecosystem.

PowerPlan delivered another strong year with healthy recurring growth and steady progress on product modernization and cloud migration. They continue to invest in product innovation, customer experience and internal operating capabilities, improving their long-term organic growth profile. Shout out to Rafi for carrying the leadership mantle forward at PowerPlan and great job managing the transition from Joe.

Illumia, formerly known as CBORD and Transact, continues to execute well and is progressing in its integration and platform road map while maintaining solid commercial momentum. And we're excited to welcome Greg Brown, our new CEO at Illumia, who brings a long and successful history of leading scaled software businesses. Congrats and thanks to Laura, Rachael, Taran and Rob for executing the VCP, driving the business combination and achieving the year 1 target.

We look at the broader portfolio of businesses we've acquired over the last couple of years, Syntellis, Transact, Subsplash, CentralReach and Procare. We feel very good about the quality and long-term growth potential of this group. However, Procare did not perform to our expectations in 2025, although we do feel good about the business building that occurred last year.

Specifically, we improved payments execution, upgraded the entire leadership team and continued to win competitively in the market where Procare remains the category leader. The biggest constraint was implementation timing across both software and payments which delayed customer time to value and weighed on payments volumes. Improving implementation speed and delighting the customer base are the top priorities and Procare's leader, Joe Gomes, has executed this playbook before at PowerPlan.

CentralReach is off to an outstanding start and ahead of our deal model. The business is scaling well with strong recurring software momentum and expanding profitability, and they're building a broader growth engine through cross-sell and a steady cadence of new product releases, including AI-enabled offerings.

Now turning to our outlook for 2026. We expect organic growth to be in the higher end of the mid-singles range. We also expect a modest back half weighting as CentralReach turns organic and nonrecurring comparables ease in the second half. As mentioned previously, we're maintaining a conservative posture in GovCon at Deltek until we see sustained improvement in commercial activity. So overall, application software remains a durable growth engine supported by recurring revenue momentum and continued product execution across this portfolio.

Please turn with us to Page 12. Total revenue growth in our Network segment was 8% and organic revenue grew 4% for 2025. EBITDA margins came in at 54.1%. DAT continues to execute well on what they can control, broker integrations, value capture and trust in the network, all leading to ARPU expansion.

Although the freight recession persisted throughout 2025, DAT is continuing its evolution from a traditional load board into a more automated marketplace where brokers and carriers can match loads with greater trust and efficiency and increasingly transact within the platform. And as this happens, DAT's TAM and monetization opportunities grow. To this end, DAT is advancing its AI-first operating model with concrete use cases across carrier onboarding, fraud detection and freight matching automation. This is a pattern we like. AI that improves customer outcomes, lowers transaction friction and expands our TAM where we have a very high rate to win.

ConstructConnect had another strong year of recurring revenue growth, and the team made material technical advances with their AI-based takeoff solution, Boost. Foundry is making steady progress with year-over-year growth in ARR as the market continues to recover. We continue to be excited about the AI product development at Foundry because it fits naturally in the creative workflows where small improvements can materially improve artist throughput. Importantly, these are high-frequency, high-value tasks that Foundry already sits inside, so AI is being delivered as embedded features that customers should adopt quickly given the clear and integrated efficiency gains offered.

MHA, SoftWriters, SHP continue to execute well, supported by stable end market demand and strong reoccurring revenue models. Each team is advancing its road map with targeted investments in functionality, workflow efficiency and service levels to deepen customer value and retention. Subsplash is off to a great start in the portfolio with strong execution and solid momentum across the business. We're encouraged by the durability of the revenue model and the opportunity to continue expanding value delivered to customers over time.

As we turn to the outlook for the year, we expect network software organic growth to be in the higher end of the mid-singles range representing a modest improvement versus 2025. We expect a stronger Q4 driven by Subsplash turning organic in the quarter. Of note, we remain conservative on DAT by assuming no meaningful improvement in the freight market.

Now please turn to Page 13 and let's review our TEP segment's full year results. Revenue here grew 7% on a total and 6% on an organic basis. EBITDA margins remain strong at 35.7%. We'll start with NDI whose growth is being driven primarily by sustained momentum in its electromagnetic tracking solutions supported by strong OEM demand and program ramps. Importantly, OEM order activity has remained strong, and the business is converting that demand into higher revenue scale and operating leverage. Great job by Dave and the entire team at NDI.

Verathon continues to perform very well with solid growth across its GlideScope and BFlex franchises. Importantly, Verathon is the U.S. market share leader in single-use bronchoscopes, which reflects several years of consistent execution and reinforces the durability of a model as the business continues to take share in an attractive procedural workflow area. Looking to 2026, we're optimistic about several new product launches planned throughout the year.

For the full year, Neptune grew modestly notwithstanding the year-long backlog normalization supported by demand for its static ultrasonic meters and its cloud-based software solutions. Although the second half commercial challenges tied to our tariff surcharging program eased late in the year, we remain cautious and are not underwriting a recovery in our 2026 guidance.

Finally, the balance of the businesses in this segment, CIVCO, FMI, Inovonics, IPA and rf IDEAS were really strong throughout 2025 and were meaningful contributors to the segment's results. For the full year, we expect segment organic growth in the mid-single-digit range, with first half being more in the low singles area as Neptune's backlog continues to normalize. Given the more limited visibility of Neptune, we're taking a cautious approach as we monitor underlying demand over the next couple of quarters.

With that, please turn with us to Page 15. So now let's turn to our Q1 and full year 2026 guidance. Based on what we previously discussed in our segment overviews, we're initiating our 2026 financial guidance to grow full year revenue in the 8% area, organic revenue growth between 5% and 6% and adjusted DEPS of $21.30 to $21.55. Our guidance assumes a full year effective tax rate in the 21% area and more in the 22% area for Q1.

To reiterate from earlier, our full year guidance does not bake in improvement at Deltek's GovCon business or in DAT's freight market and assumes modest top line weakness at Neptune versus 2025. As discussed, we expect stronger second half organic growth driven largely by CentralReach and Subsplash turning organic and easing nonrecurring comparables. Our guidance does not assume a meaningful revenue uplift from our AI development work either. We view AI as incremental upside as we scale commercialization across the portfolio.

Finally, we remain positioned to be active and opportunistic on capital deployment. We continue to have a robust M&A funnel, a meaningful remaining share repurchase authorization and substantial financial flexibility and will remain disciplined and unbiased between acquisitions and buybacks based on what drives the highest and most durable cash flow per share compounding. For the first quarter, we expect adjusted DEPS to be in the range of $4.95 to $5, reflecting the dynamics previously discussed.

Now please turn us to Page 16, and we'll open up for your questions. We'll conclude with the same 3 takeaways with which we started. First, in 2025, we delivered both double-digit revenue and EBITDA growth and solid free cash flow. Enterprise software bookings grew in the low double-digit range which positioned us well entering 2026. Second, we're investing for long-term sustainable growth improvements while staying disciplined in our expectations. Throughout 2025, we upskilled talent, sharpened strategy and improved execution across the portfolio, and we are accelerating AI product development. We're not baking in an organic inflection in 2026, and our guidance will reflect improvement as it materializes.

Third, we materially advance our portfolio through capital deployment. We deployed $3.3 billion into high-quality vertical software acquisitions, executed opportunistic repurchases and maintained more than $6 billion of forward capacity. As we look ahead, Roper remains an advantaged and preferred buyer for both management teams and private equity sellers, and we believe the M&A backdrop remains constructive as private equity firms face increasing pressure to generate liquidity for limited partners.

Our pipeline is robust, and our team is deeply engaged, and we will remain disciplined and unbiased on valuation and business quality. In parallel, we'll continue to balance acquisitions with opportunistic buybacks, allocating capital to whichever path drives the best risk-adjusted and long-term cash flow per share compounding.

As we turn to your questions, please flip to the final slide, our strategic compounding flywheel. What we do at Roper is simple, we compound cash flow over a long arc of time through a disciplined strategy anchored on 3 things. First, we are market-leading vertical focused businesses, application-specific, deeply embedded and mission-critical. These are durable franchises with highly recurring revenue and organic cash flow growth that can improve over time.

Second, we run a decentralized operating model so our teams stay exceptionally close to customers and their workflows, so we can consistently compete and win. That customer intimacy is a core competitive advantage. That's also how we win in AI. In our markets, AI isn't a generic overlay. It has to be grounded in a real workflow context, tuned to domain-specific edge cases and deployed through trusted embedded relationships. So our AI delivers measurable value and better customer results.

Third, we pair that with disciplined, central-led capital deployment, focused on high-quality M&A and opportunistic share repurchases, allocating capital objectively to maximize durable cash flow per share compounding, niche-leading businesses, decentralized operations close to customers and disciplined capital deployment. That's our long-term compounding flywheel. We're excited to compete and win and continue delivering long-term and improving cash flow compounding per share.

So with that, thank you for your continued interest and support. Let's open it up to your questions.

[Operator Instructions] Your first question comes from Brent Thill with Jefferies.

Neil, regarding Deltek, I'm curious if you could just give everyone a sense of what you're baking into the '26 guide and how you're protecting against another potential government shutdown.

Yes, go ahead, Jason.

Yes. So I think we are not assuming an improvement this year. The fourth quarter was depressed by the perpetual license revenue. As I talked about, most of GovCon Enterprise still buys perpetual licenses. So that's what drove our lower organic in AS in the fourth quarter. So we don't think that's going to repeat next year. So we do have a comp benefit but otherwise we're not really assuming improvement in that market until we see it.

Okay. And on Procare, Neil, what do you think needs to happen to get that back meeting expectations?

Yes. I think it's -- just to go through a little bit of what we talked about in the prepared remarks and then add a little bit more to it. So hey, the business is the leader in the marketplace. It is the clear leader. We've done a lot of good things there. We sort of cleaned up and fixed the payments, cost infrastructure and processing capability. We fixed and improved the go-to-market. So we're competing and winning the marketplace. We're winning a majority of the jump balls versus the primary competitor.

And so the problem now is just push to the right. So now we're winning these opportunities, and we're slow to implement the software, which means we're slow to implement the payments, and that's the next sort of objective in front of the team there. So once we get that done, we feel much better about that. It's a completely fixable problem. It's one of the problems that you don't like to have problems generally, but when you do have new ones that are eminently fixable, which this one is the larger problem would be if we had a competitive situation or something like that, which we do not have.

Your next question comes from Clarke Jeffries with Piper Sandler.

I wondered if specifically on the GovCon business, you could maybe give a rank order of the kind of appropriation bills, what would have the most impact? Or within Deltek's exposure, what segments of the government getting those appropriation bills passed would be most significant?

Yes. So I'll just draw back to the OBBB. And this is, as you know, we're not -- Deltek doesn't have direct exposure to the government. It's our customers that are the federal contact -- that have the direct exposure to the government, just to remind everybody.

The OBBB is heavy on defense -- Department of War, Department of Defense and DHS funding and spending, that tends -- those categories tend to have larger percentage of contractor spend. It could be north of 50% of the whole category. It can be contractor spend, so that's definitely a tailwind. The civilian programs tend to have lower percentage. So it's not necessarily a bad thing for Deltek's customers, but it's certainly better on the current appropriations, the OBBB.

Perfect. And then the last 2 years hovered around $3 billion deployed towards acquisitions. Wondering if you could talk about expectations for how much you might deploy in '26. What scenario might push you towards a number closer to $4 billion or a number closer to the $2 billion? So what are you factoring into the deployment outlook for '26?

As we mentioned there's about $6 billion sort of is what the forward capacity is over the next 12 months. We have the 2 levers available to us on the capital, the M&A and the buyback. On the M&A side, the thing for us -- I've been here 15 years. Jason has been here 20. When you're building a business that has an M&A lever, we never view the amount of capital next 12 months in the budget or we got to spend it because we're building a business that's going to own businesses in the perpetuity. So you have to buy very high-quality businesses at an appropriate price.

And so that discipline guides us. So it's hard to set an expectation that says we'd be -- we're going to get X dollars deployed against the $3 billion -- $6 billion, excuse me, in M&A or buyback, but we like having both levers available to us, and we're just going to do what's best objectively to compound cash flow per share at the best rate we can.

I will say that we think this -- coming into this year, I think the market is ripe for more assets to become available. I mean we've been very proactive the last couple of years with -- in a very muted market. So as I mentioned, I think it's a net positive for us, but we'll just stay disciplined and focused.

Don't mistake anything I'm saying, like I was just saying, as Jason said, the opportunity -- the number of deals, the number -- the amount of LP pressure on the GPs and private equity just continues to mount. There's going to be -- there's an aging portfolio of very high-quality assets and number of those assets that we have relationship with and are meeting with management teams and becoming the preferred owner, all that is very, very ripe for opportunity, but we're going to remain, as we always do, disciplined.

Finally, on that, over the last really 3 years -- 2.5 or 3 years, we've really leaned in and built capacity for tuck-ins and bolt-ons. That's a more predictable pace. I think we did 7 or maybe 8 small tuck-in acquisitions last year, that'll be more sort of predictable because it's a lower dollar per transaction but more of them.

Your next question comes from Joseph Vruwink with Baird.

Great. On AI, when would you expect to get to the point of quantifying what AI means for Roper at maybe a more precise level? I think it's evident in certain areas already, the Aderant call out their mid-teens growth, they're 10 points above the segment. I would imagine customers want the cloud as part of their AI initiatives and so there's an inherent uplift and positive correlation between Roper and AI for the legal space. Can you do that more holistically and attribute some of the organic improvement ex Deltek you're already seeing and say that's directly or indirectly related to AI investments?

Yes, so appreciate the question. We spend a fair amount of time talking about that internally. Just a couple guiding principles that we have internally is, one, we're not going to AI wash or allocate revenue like other companies have done or doing. We're not going to say X dollars R&D, so therefore Y dollars of revenue is AI related. So we're not going to AI wash our revenue stream.

That said, we do aspire to be able to report a number, like AI revenue SKU related is X or Y. Unfortunately, if we do that, I mean we're going to monetize AI in more ways than just AI SKUs. It's going to be cloud uplift. It's going to be in packaging. There's going to be lots of ways that we monetize this. So this is actually -- it says simple, it does pretty hard from how we're going to be able to sort of report this where it's credible.

At the end of the day, Joe, you highlight the most important thing, which is we believe this is a TAM -- meaningful TAM expander for us, which means it should be a growth driver for us, and you'll see it show up initially in bookings and eventually into the recurring or reoccurring base, so we see that at Aderant, see at CentralReach. We'd expect to see it across several of our businesses starting this year. More broadly, as we sort of write the chapters on Roper, 2025, the chapter in AI would be how we learned to develop the initial set of products across our software business. Essentially, every one of our software businesses either has or is right on the precipice of having AI-related product to deliver to our customers.

'26, I think the chapter is going to be how we commercialize, how do you sell, deploy, drive implementation and ultimately sort of monetize all of the product. And so that's going to be the journey of learning for us across the portfolio organization. We look forward to providing updates on that as we get through the year.

Great. That's helpful. On your approach to guidance this year, I think it's very clear that you're going to let the upside come to you and future changes are going to happen. As you see it, can you maybe put some guardrails in magnitude of what that could ultimately mean? And I'm thinking in the past you've talked about how your current portfolio could be capable of 7% and in a best case scenario, 8% to 9%, that's more of a long-term framework. Within FY '26, if things go right and you get some redirection and where the pressures within the portfolio have been, what sort of upside possibility could there be?

Yes. So I would say the long term -- to first point, the destination, if you will, the longer term, certainly not a '26 comment. The longer-term entitled growth in the portfolio, we still have conviction is north of 8%. We go company by company about what their entitled realistically achievable growth can be and so that number sort of the target destination has not changed.

We are definitely, as you've heard from our commentary, taking a much more appropriate and balanced view for the initial guide here. No improvement at Deltek on the government contracting side, no DAT market recovery actually underwriting a slight decline in Neptune. And so if you sort of thumb each one of those, I don't want to get into order of magnitude of what it could look like, but I would say it definitely tilts more conservative than this past year, for instance.

Your next question comes from Dylan Becker with William Blair.

Maybe kind of following up on one of Joe's questions, too. I think the Deltek perpetual piece makes sense, but you also called out some softness on nonrecurring due to some of those cloud migrations and maybe that's just kind of rev rec of upfront for ratable. I guess as you think about kind of AI's opportunity to accelerate this modernization and cloud journey given kind of the heavy maintenance space you still have there, how do you think about kind of that trade-off in those long-term economics? It seems favorable. I think it's kind of evident in the subscription bookings in that low double-digit framework, but maybe kind of walk through some of the nuance between those as well, too, if you can.

Yes, certainly. Appreciate the question. I think we've seen some of this just in our AS segment over the last year's nonrecurring revenue has been sort of flattish up a little bit here and there because some of our businesses have moved more to the cloud, be it Aderant or in recent years, PowerPlan. We see that at Deltek is going to be a significant opportunity. So that's a part of our thinking in '26. Deltek's really put a lot of AI functionality into their cloud product and so some of these large government contract customers are contemplating going to the cloud, and they have a big push for that.

So you're right that will obviously increase customer lifetime value, but it'll have a more muted impact in year. So we thought through a little bit of those dynamics this year and I think that's probably the biggest area where we will see that because a lot of our -- the rest of our businesses are sort of on their cloud journey, but they will continue, to your point, to include only AI features in the cloud and so that will just increase adoption as we go forward.

Yes. And just to add to what Jason said, we don't expect a pronounced J curve because we have a very large installed base that's on-premise. That's going to lift -- that is lifting and shifting at 2 to 3x recurring. So it's -- the J curve is less pronounced because you're converting an existing recurring base at a higher level. And as you sort of, if you will, attrib or convert net new perpetual to recurring. So they offset one another, and we think we have that harnessed in our guidance.

Okay. Great. And then maybe, Neil, for you, too, on the topic between platform and bolt-on M&A, I guess, could you kind of give us a sense if you see any opportunity maybe as a part of AI, maybe not, but given kind of the current backdrop to accelerate some of the initiatives in one effort. I know we've built out kind of the bolt-on team. There's a little bit more visibility into those platform valuations maybe have come down to a particular level as well. But just kind of think about kind of the mix between capital deployments between bolt-on and platform, if you can.

Always hard to predict mix. I can tell you that the -- if there is a -- generally speaking, if there's a rank order, bolt-ons or tuck-ins are going to be first order because they are advancing sort of the organic growth and strategic direction of one of our platform businesses. At the same time, you always have a little bit of back-office G&A synergy which enables to sort of buy down the initial purchase price pretty quickly and then you get to the growth orientation. So that continues and will always be a focus of ours.

What we see, by the way, on that front is it takes a little bit of time as we added the resources, they get to know the company, they build a relationship with our companies, they build a relationship with sponsors and targets and founders. I think something like 60% of our pipeline for bolt-ons is either proprietary or founder-driven. That's a completely new motion for us. That would not have been the case 3 years ago. I think that bodes well, but these things do take time to matriculate through the system and mature to where they can become actionable.

On the platform side, it's a -- the number of opportunities and the quality of assets is very interesting. The question on the table is going to be what happens relative to valuation. We've never been a short-term next 12-month multiple arbitrager. So we're not going to do that. But we definitely have to sort of -- we have to look at buybacks versus bolt-ons versus platforms on what is the best long-term compounding in terms of value-creation opportunities for us.

Your next question comes from Brad Reback with Stifel.

I know you guys gave the software bookings for the entire year. Can you give us what it was in 4Q?

Yes, it was up high single digits and that is with Deltek being down in the low double-digit area. Actually, Deltek's SaaS was strong but like I mentioned, perpetual was down meaningfully. Yes, the rest of the portfolio performed pretty well. Vertafore had a strong quarter off of a really tough comp last year. So they're continuing to just have success in 2025 and that should be good for them for '26. And then as I mentioned last quarter, health care has been really strong for us and that was the same in the fourth quarter.

Great. Neil, this is the second quarter in a row you've missed expectations, and you're guiding to a back half acceleration in '26. So maybe take a moment and help us understand where the incremental conservatism is in the '26 guide versus the last couple of quarters?

Sure. So we did say -- and we're certainly disappointed over the last couple quarters, but we did say this time last quarter we had a wider range of outcome, fan of outcome, especially because of the uncertainty at Deltek. And so while disappointing, we tried to be sort of very straightforward in that regard.

I think for this year, I said it before, I'll say it again, when you look at where we're exiting this year, look at '25 compared to '26 and you just go -- let me back up. When you look at '25, the initial guide versus where we ended up, it really reconciles to 3 things we talked about. It's Deltek because of GovCon, it's Neptune because of the dynamics we talked about and it's Procare. That almost fully reconciles the difference between the initial guide and where we ended up.

You then have that in mind, you take it, you carry it forward. We're assuming in 2026, there's no improvement in Deltek. There is no acceleration at DAT. There's actually -- we're underwriting a modest decline at Neptune versus '25. And the only thing that sort of -- then you get the accretion to organic growth from Subsplash and CentralReach that turn organic this year. And then we have this easing second half sort of nonrecurring. So optically, it looks like an acceleration through the year, but when you look at the pieces, it's actually pretty steady, say, for the things that we just said.

Yes. And just to remind you that, all the CentralReach and Subsplash becomes organic second half. So it's got to create some of that ramp. And I would just call out, too, just again, just a small comp -- a couple comp issues. Foundry gets a little bit better this year, and I know it's small dollars, but in network, it matters. And then we had the first quarter of '25, a pretty depressed network number because we were comping against a bigger number in '24. So that comp goes away. So there's some math too, just going from '25 to '26.

Your next question comes from Terry Tillman with Truist Securities.

The first one's going to be on Deltek. The second one, the follow-up is going to be on DAT. But on Deltek, and I know these months or these part of the quarters are probably less seasonally strong, but did you actually see any improvement in order volumes for perpetual in December or January? And also with Deltek, are the effects of DOGE kind of lessening? Or is that still impactful? And then I had a follow-up.

Yes. So Terry, this is Jason. I think the -- December is always stronger than the other months and that's just the natural kind of inertia of how orders flow in Deltek. I will say we had 2 large contractor -- government contractor deals that slipped. So it was like right at the end. And so we think they'll both land in the first half of next year, but we've also sort of hedged that just in case. But -- so usually, we get some big deals, and they were right at the finish line and they didn't close, but they're still in the queue, and we still think we're going to...

Close the rest of this year.

Yes, exactly this year.

And just to pick up on that, Terry, as well, just to add, the -- while the signatures on paper are slower because of the shutdown, the commercial activity, the pipeline build has actually been encouraging. It's been encouraging throughout. It's because all the -- this is an environment, unfortunately, that our customers live in and they sort of -- they're subject to the vagaries of what's happening in the government, but they have a business to run and they have contracts that are likely going to get awarded, and they have to sort of manage sort of our software in that regard. And so there's no competitive issue here at all. Let's take it hasn't been asked, but 0 competitive issue here. It's just deals that are building that are pushed to the right a little bit given the uncertainty. DOGE, I would say, is, to your question, is lingering impact, but it's not the topic that anybody's talking about the way it was in the first, third -- the first half of the year of last year.

Got it. I appreciate that. And just a follow-up is on DAT. Do you see ARPU lift continuing to play out through the year and are you on track for that autonomous kind of load matching technology innovation to play out in '27?

Yes. So we do expect ARPU to continue improving, growing in 2026. There's a couple reasons for that. One is you just have like-for-like pricing opportunity that'll sort of get cascaded in during the year as it normally does. But for -- the second reason is we have more value to sort of sell to both sides of the network. And so you're going to get -- in the past, it was just load board, so it's load board and pricing. Now it's load board and automation. It's load board and data. It's load board and a number of things on both the carrier and the broker side.

In terms of the automated matching, we're -- it's early days but we're encouraged by the progress. The tech -- unambiguously, the technology does the job. Let's just be clear. It is the ability for a broker to tender to the DAT One platform and automatically match a load and have a carrier pick it up and complete the commerce with payment sort of overlay across all that. It works and it's working every day in the marketplace. The #1 focus of that business is to build both sides of the network that starts on the broker side by getting native integrations with their TMS systems. And you have seen and will continue to see during 2026 a cascade of announcements about the various TMSs that we're integrating with that allows the brokering or tendering to our platform to be native and the work for the brokers.

And then we continue to build the carrier side of the network. It's got to be a high trust, no fraud environment. That's part of the core technology that we have and we're integrating. So early days, but we like the tendering percentage, we like the completion percentages. We like the factoring percentages, and we'll want to see that business scale as Jason and Shannon and Satish and the team at DAT look at this on a monthly basis.

Your next question comes from Ken Wong with Oppenheimer.

You guys are guiding the 5% to 6% organic for '26. As we think about 1Q first half with a lot of faster growth businesses coming in second half 4Q and no Deltek tailwind, like is there the possibility that you could be below that 5% low end? Any context there so we could properly level set our numbers?

Yes. No, I don't think so. We're kind of thinking for AS we'll be sort of in the mid-single-digit range with nonrecurring being flattish and the recurring reoccurring being like mid-single digit plus which is consistent with Q4 levels. I mean you're not -- we're not going to have the same nonrecurring decline like we did in the fourth quarter. And then as you mentioned, the second half gets better on the CentralReach turning organic. And we've got this, the nonrecurring, as I just mentioned, we get a better comp in the fourth quarter. So not a lot of -- I would just say not a lot of go-get in that second half number.

And then on NS, I think sort of the recurring revenue will be just sort of continue to be mid-single digit plus out of the gate. And then as we go throughout the year, Subsplash actually comes in, in the fourth quarter, so that'll be helpful. And so I think that's sort of how it sets up, nothing outside of that to call out.

Got it. Really appreciate the color there. And then perhaps just any additional context you could provide in terms of what the new business activity pipeline conversion look like versus maybe the renewal business, term expansion, contraction? And any details would be helpful.

Ken, is that a broad portfolio question or specific to a business, broad, broad question?

Yes, just like -- yes, correct. It's more of a broad kind of software selling kind of trends that you guys are noticing across the group.

I understand the question. Yes, I would say we're broadly encouraged by what we saw in the finishing the year. The bookings and retention statistics were quite good. As you know, our enterprise -- our gross retention is in the mid-90s for our enterprise businesses. That's steady to tick up a little bit during 2025. And then in terms of the bookings activity, hey, while there's a little -- we expect there to be volatility quarter-to-quarter, low double-digit bookings growth in the year and being pretty broad-based with sort of weakness at Deltek, I think, is all you need to see. We -- and so it's been pretty good.

Your next question comes from George Kurosawa with Citi.

You guys brought in some new AI leadership, Shane and Edward. Would love to hear a little bit about the team they're building out, what you have them focused on and if there's any kind of low-hanging fruit that learnings they can apply across the portfolio.

Yes. So we're excited to have Shane and Eddie join and the team they're starting to build out. 3 things they're generally focused on. First is it's all in pursuit of accelerating the top line goals, accelerating sort of our pace of AI product development and ultimately, shipping, selling monetization. That's where the focus is. So 3 subcomponents of that, it's coaching and teaching, right. Our businesses did a meaningfully above average, above expectation job in 2025, late '24 or '25, learning, making mistakes, learning, making mistakes, learning around AI, AI development, what works, what doesn't and getting products into the hands of customers. That was great to see.

But some of these -- a lot of these AI tasks are quite complicated, complex from a technical point of view. And also, unlike regular way software development, there's some art in this AI development. And so Shane and Eddie and the team they're bringing -- really bring just a history. These are people that studied machine learning and AI in university quite a while ago and spent their entire careers. They've seen a lot of pattern recognition. They're going to coach and teach our leadership teams, our technical leaders, our product teams on all things AI, ML related. That's one. Number two, they are going to build an AI sort of development strike team or accelerator team to where when there are -- a company might have more than it can do from its internal resources, and we'll supplement those teams to accelerate in some pockets.

And then third, there is in their first quarter with the business and they met with most of our software businesses there is clear opportunity for some reuse inside the portfolio, certain AI-based sort of capabilities we can sort of produce and sort of have, if you will, a Roper open source model where we can reuse some components and componentry. And so we're going to focus there. And early days, it's been just really great. They understand our culture, our teams have really engaged them, and they're just looking forward to scaling the team and getting to the work.

Okay. Great. I did also want to touch on the margin side. Core gross margins were up over 1 point in the quarter. Maybe just talk through the tailwinds there. How should we think about any sustainability to improvements?

Yes. I mean I think we've always said in our long-term incremental margins at the EBITDA lines around 45%, did a little bit better this year. I think as we go into next year, I think AS might be up a little bit. Network's going to be down just because we've got the full year of Convoy and our algo rolling through. And so that'll accrete up over time, but a little bit of a drag in '26. And then I think our TEP segment will be for the full year sort of flattish. We've got more consumables rolling through next year than normal, which has a little bit lower margin. And that's really more pronounced in the first half. So maybe down a little bit in TEP in the first half and then it'll improve throughout the year.

Your next question comes from Josh Tilton with Wolfe Research.

Maybe just first kind of a simple high-level one on the guide for next year. I appreciate all the color that you gave on Deltek and DAT and Neptune, but if you were to take those 3 businesses aside and treat the rest of the organic business as one, like what would be the one line color on the rest of the organic business? Does the guidance assume that everything ex DAT, Neptune and Deltek gets better, stays the same, gets worse? Like how would you characterize what the rest of the business has to do that's baked into the guidance? Does that make sense?

It does. Yes. So I mean, I would say broadly, it gets a little bit better, but not a lot, not meaningful enough to draw inflection. So that's baked in. I mean when you think about -- we finished around 5.4%, the deals are a tailwind. This nonrecurring in AS is going to be somewhat of a tailwind. Foundry does get a little bit better. And so maybe 10 basis points or so to the enterprise. And then really the swing factors, I talked about the comp in NS in Q1 of '25 that didn't repeat. Then you just talk about like the swing factors, it's all within Neptune, and that's what our low single-digit to mid-single-digit guidance has for TEP, and that's what kind of bridges you to the -- from the low to the high end.

And then maybe just a quick follow-up. I understand that some businesses go organic in the second half. Is there any -- I don't know if conservatism is the right word, but is there any conservatism? Is there any learning that you saw following like kind of the little hiccups that you saw in Procare that you're kind of applying or embedding or assuming will happen as some of these inorganic businesses convert to organic in the second half of next year?

Yes, Josh, it's Neil. I'll take that one. So short answer is heck, yes, there's a lot of learning from our Procare governance, what worked, what didn't work and how we're governing both Subsplash and CentralReach. And we can spend more time talking about offline. But in essence, when we see a small variance in a monthly reporting package relative to one of the key levers in value creation plan, in Procare, we observe that variance for a longer time before we decided to take action to correct it. Now we immediately jump to a corrective action, a countermeasure. And we don't let small variances turn into large variances. And as a result, you have CentralReach that's ahead of the underwrite model and Subsplash is on the underwrite model for the outlooks for those businesses. Just -- we can get in much more detail when we have more time offline, but that's the essence of it.

And I would just say that for CentralReach and Subsplash, feel good about the contribution in the second half, the path that we're on, bookings momentum, the recurring, the gross retention, just the path to get to that accretion in the second half, we feel very good about that.

Your next question comes from Deane Dray with RBC Capital Markets.

This has come up several times today about the M&A bias and looking for durable cash flow compounding. I'd be interested in hearing your thoughts about how do you rank looking at absolute dislocations in some asset prices today versus what you perceive as where there might be a wider moat against AI in these assets. So how are you weighing those?

I want to reframe, Deane, the question to make sure that we answer the right question. If not, you can correct us. And so the question is looking at both private and public companies that have a valuation dislocation, are we looking at that? And then how does the AI moat influence our thinking? Can you just reframe that? I want to make sure we answer the right question.

Yes. So yes, I wasn't specifically talking about public valuations, but that would be great to hear that as well because you've done those in the past versus thinking more strategically about where there might be wider moats.

Yes. So I would -- so -- and feel free, we won't ding you on one of your questions if again, I answered the wrong question. We're always -- for the long history of Roper, we're always investing in these vertical market application-specific businesses with deep moats, right. And so that does not change. We believe in the AI world, these -- the moats where you're intimate with the customer, you have unique and proprietary data, you're embedded in high-frequency workflows where on-stack AI is easier to implement, easier to monetize and ultimately translates to the automation of tasks, which is this TAM expansion. We really like and are leaning in it. We're seeing it playing across our 21 software business days, so we'll continue to lean in that thesis from a capital deployment point of view.

That's really helpful. That's what I was looking for there. And just a quick one on Neptune, we've talked about the order delays. Is there -- how much of an impact is the spike in copper played? Is there a sticker shock? Is that -- does that need to be kind of rippled through the market to reprice? Just what's the impact there?

I would say that -- what we talked about last quarter, largely in the bucket of tariffs, but it's tariffs, it's copper pricing, generally the shock to the cost structure of a water meter when we started in really July of last year pushing a surcharge to accommodate for that increase in cost of goods. It was definitely a shock in the system in Q3, and it really abated during Q4. So I think our base case assumption is that is really in the rearview mirror and set to the side. And moving forward, it's just about the normalization of volumes in the market sort of on the very tail end of the COVID spike in volumes, and now we're on the backside of that spike into more normalizing range of volumes in the market.

Your next question comes from Joe Giordano with TD Cowen.

I'm just curious how you're now weighing like in terms of capital deployment, like different, like timing horizons here, right. Like you have -- stock today is, I don't know, 15% below the average price of the buyback in the fourth quarter. You're trading at like almost a high single-digit free cash flow yield now. And it's a portfolio that you're intimately like close to relative to something that you might buy that drives top line that is something that inherently has more risk because you don't know it as well. Like how are you weighing something that -- like the certainty of what you know versus like the risk reward of something you don't know at the price that you're paying?

So I'll take the first pass of that. I'm sure Jason will have some color he may want to add. So again, just to -- we've said it, it's on repeat, we'll say it again. The objective of M&A versus buybacks, sort of the levers available to us is what's the best risk-adjusted path to long-term cash flow per share compounding, period, full stop. We're totally objective and just passionate about the allocation of the 2. There's $6 billion available so a big sort of -- a large amount of capacity. On the buyback, we just -- the valuation dislocation is just -- is silly. And so we leaned into it in Q4, and we find it obviously more attractive today. And it's a great opportunity to drive long-term cash flow compound that way.

On top of that, we're very excited and confident about our future, right. I mean the growth, the AI, the leadership, the strategy, execution prowess, I mean all of it feels very, very good to us and what we see internally. At the same time, M&A is a real lever. I mean there's not -- to somewhat to my surprise, we introduced the buyback last quarter, there is commentary about, oh my gosh, is the M&A thesis not intact. That is one of the most absurd things I've heard in my 15 years at Roper. We are a preferred buyer of vertical market software leaders. We're absolutely preferred from a management point of view. We're preferred from a seller point of view. The pipeline is enormous.

The LP pressure is legit. The number of assets in private equity portfolio have to get liquidity are at levels we've not seen. So that thesis just needs to be eliminated from the talk track because it's not real. And so for us, it's balancing those 2. Buybacks are great in the short run. M&A generally is going to beat in the long run, and we like having the balance between the 2 options in front of us.

Yes. I would just add that around the confidence, and we've had just these unusual things happen with 3 of our businesses, but like the underlying quality is getting better, so there's that. And I would just say the AI, we have 21 different businesses working through AI right now. That are annual operating plan reviews and came out with an increased level of conviction that we're going to win relative to AI. We're just -- we're so -- our customer intimacy is really proven to be a competitive advantage, and we've got the tools and resources to get after the AI just as fast as anyone else. So we feel really good about that. And so buying ourselves in that scenario where there's this dislocation makes all the sense in the world, but it's also going to be an incredibly active year on M&A. So we're just really -- just got an abundance of opportunity in front of us this year.

Now that you brought in this AI talent on the accelerator team, were there any instances where like negative instances where these guys coming in as experts kind of identified that maybe parts of your business where you thought you had more of an opportunity is going to be harder to drive? Or I mean, I'm sure they're identifying places that you have opportunities, but was there any unlike the negative side where something was like, well, maybe this isn't as attractive as I thought in a particular part of the company.

Yes. So I would say, on balance, their reviews and early takes are quite positive about the opportunity market-wise, technical-wise, the prowess of the teams that we have in place. But we also -- we had them sort of do a short readout to our Board last week. And then there was a few bullet points of things that were on the constructive ledger. None of it was market opportunity, lack of market opportunity or lack of opportunity to win.

Again, these are more technical resources, so they might not have like the best acumen in like these vertical market spaces to judge that anyway. But it was like, hey, as you'd expect, maybe there's -- we definitely need to improve the quantity of AI talent in the businesses. I mean that's a little bit of why we're adding the central team to sort of spark some acceleration. And it's just going to take some time because we've got to build these people. It's not something that we're going to be able to hire en masse. We got to build these people, and we did a good job last year. We'll continue to scale that and compound the learning on that this year.

Yes. I think the ideas have been well received by Shane and Eddie. I mean they understand the specificity of what we're trying to solve at the individual sort of vertical level, and that's they view that as very unique, right, coming from a horizontal player. So I think they see the opportunity just like our businesses do.

This concludes our question-and-answer session. We will now return back to Zack Moxcey for any closing remarks.

Thanks, everyone, for joining us today. We look forward to speaking with you during our next earnings call.

The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.

Good morning. The Roper Technologies conference call will now begin. Today's call is being recorded. [Operator Instructions] I would now like to turn the call over to Zack Moxcey, Vice President of Investor Relations. Please go ahead.

Good morning, and thank you all for joining us as we discuss the third quarter 2025 financial results for Roper Technologies. Joining me on the call this morning are Neil Hunn, President and Chief Executive Officer; Jason Conley, Executive Vice President and Chief Financial Officer; Brandon Cross, Vice President and Principal Accounting Officer; and Shannon O'Callaghan, Senior Vice President of Finance. Earlier this morning, we issued a press release announcing our financial results. The press release also includes replay information for today's call. We have prepared slides to accompany today's call, which are available through the webcast and are also available on our website.

Now if you please turn to Page 2. We begin with our safe harbor statement. During the course of today's call, we will make forward-looking statements, which are subject to risks and uncertainties as described on this page, in our press release and in our SEC filings. You should listen to today's call in the context of that information.

And now if you please turn to Page 3. Today, we will discuss our results primarily on an adjusted non-GAAP and continuing operations basis. For the third quarter, the difference between our GAAP results and adjusted results consists of the following items: amortization of acquisition-related intangible assets, transaction-related expenses associated with completed acquisitions; and lastly, financial impacts associated with our minority investment in Indicor. Reconciliations can be found in our press release and in the appendix of this presentation on our website. And now if you please turn to Page 4, I'll hand the call over to Neil. After our prepared remarks, we will take questions from our telephone participants. Neil?

Thank you, Zack, and thanks to everyone for joining us and excited to be with you this morning. As we turn to Page 4, you'll see the topics we plan to cover today. We'll start with our third quarter highlights and financial results. Next, we'll review our segment performance, our AI progress and momentum and our most recent set of bolt-on acquisitions. Then I'll get into our guidance details and of course, wrap up with your questions. So with that, let's go ahead and get started. Next slide, please.

Turning to Page 5. Let me run through the 4 key takeaways for today's call. First, we had a strong third quarter. Total revenue grew 14%, organic revenue grew 6%, software bookings grew in the high singles area, and we continue to deliver impressive free cash flow with free cash flow growing 17%. And of note, free cash flow margins posted at 32% for the TTM period, really impressive financial results. Second, we're super encouraged by the progress and momentum we're seeing across all of our businesses as it relates to our AI enablement and our product stacks and our internal operations and more on this in a moment. Third, we're announcing today our first share repurchase authorization, $3 billion in total. And lastly, we continue to execute on our M&A strategy of acquiring faster growth platforms and bolt-on or tuck-in acquisitions at a high fidelity rate.

In the quarter, we deployed $1.3 billion, $800 million for Subsplash, which we detailed this time last quarter, and $500 million on a series of tuck-in acquisitions. Also more on this later, but worth highlighting here, we are very encouraged by this recent capital deployment execution and the future growth potential that's being layered into our enterprise. Importantly, we remain very well positioned for the continued execution of our M&A strategy and continue to have north of $5 billion of capital deployment capacity available over the next 12 months or so.

As I turn the call over to Jason, reflecting on the quarter, I'm quite bullish on most of what we're seeing, a very strong 3Q, real demonstrable AI progress, which is a long-term growth driver for us, excellent execution of our higher growth, higher returning capital deployment strategy and the announcement of our first-ever buyback authorization. This all bodes very well for the future. That said, we'd like to see some of our markets start to cooperate a bit better, namely the government contracting and freight markets, and we have some delays in Neptune. Much more on this as we walk through today's call. So with that, let me turn the call over to Jason to talk through our P&L and our balance sheet.

Thanks, Neil. Good morning, everyone, and thanks for joining us today. I'm pleased to take you through our third quarter results and strong financial position. Turning to Page 6. Q3 and TTM results reflect the long-term financial profile of Roper, which is to compound cash flow in the mid-teens area. We'll start with revenue, which was 14% over prior year and surpassed the $2 billion mark. Acquisitions contributed 8%, led by the final quarter of Transact before it turns organic and CentralReach, which we acquired in April this year. Of note, these businesses are tracking very well against our acquisition expectations. We printed 6% organic growth, both for the consolidated enterprise and across each of our 3 segments.

Our Application and Network Software segments were in line with expectations, while TEP was a bit below given near-term timing at Neptune, which Neil will discuss further. EBITDA of $810 million was 13% over prior year with EBITDA margin of 40.2%. Core margins expanded 10 basis points and segment core margins expanded 30 basis points, led by our software segments. DEPS of $5.14 was 11% over prior year and $0.02 above the high end of our guidance range despite absorbing $0.05 of dilution from Q3 acquisitions that were not reflected in previous guidance. Free cash flow was outstanding at $842 million, up 17% over prior year and representing 32% of revenue on a TTM basis.

Our software businesses captured strong renewals, and we drove great working capital performance across the board. Broadening out a bit, TTM cash flow of over $2.4 billion is a 17% CAGR over a 3-year period. And for those looking at per share metrics, you'll note that our share count has compounded at about 0.5% over that same time period. At Roper, we have been and will continue to be relentlessly focused on cash flow and shareholder value creation.

Now let's turn to Slide 7 and discuss our very strong financial position. Our net debt-to-EBITDA stands at 3x, which is up only modestly from Q2 at 2.9x despite deploying $1.3 billion towards acquisitions. This places us in a great position with over $5 billion in next 12-month capacity for capital deployment. Regarding M&A, you can see that we've been quite active this year in acquiring high-quality growth businesses and several strategic bolt-ons. This is against the backdrop of a muted PE deal environment. The pipeline of high-quality acquisitions continues to build as assets mature in PE portfolios and a return of capital to LPs becomes paramount.

Additionally, as Neil mentioned, we're pleased to announce another capital deployment lever that was previously unavailable. Our Board has authorized a $3 billion share repurchase program with an open-ended time period to execute. While M&A will continue to be the majority of our capital deployment allocation, our share repurchase program will allow us to opportunistically complement our M&A program. Over the last year or 2, we have talked about the great business building taking place across the Roper portfolio from strategy to talent to execution, all now greatly turbocharged by AI.

Our repurchase program reflects both confidence in our strategy and our commitment to delivering long-term shareholder value. So with that, I'll turn it back over to Neil to talk about our segment performance. Neil?

Thanks, Jason. Turning to Page 8. And before we get into our segment details, we want to discuss why AI is a powerful and durable growth driver for Roper. To start, AI represents a meaningful expansion of our TAM across the portfolio. We can now deliver transformational software solutions that automate labor-intensive work adjacent to our existing platforms. This creates substantial new value streams for our customers and correspondingly facilitates long-term growth for Roper and our businesses. Importantly, our businesses are uniquely positioned to win in AI, in fact, having a very high right to win in the AI world.

Our software solutions are deeply embedded system of record applications with workflow-oriented domain-specific architectures. The decades of cumulative workflow knowledge built into our platforms, combined with the proprietary vertical market data, provide the precise context needed to develop Agentic AI solutions. Because of this, our businesses have an exceptionally high right to win as we deploy these capabilities across our VMS end markets. Internally, we're becoming AI native across all functions to drive productivity gains. We're excited to reinvest these gains to further accelerate our product development and go-to-market initiatives. It's important to note, we've always had more great ideas than resources needed to execute, and AI has the potential to attack this challenge.

Finally, we have tangible proof points, though it's still early. Aderant has claimed a technology leadership position in legal tech, accelerating their bookings growth. CentralReach now has roughly 75% of their bookings attributed to AI-enabled products, which have automated 100 million reimbursement rule evaluations, over 3.5 million learner appointments and over 1 million clinical summaries being generated, great real-world examples of the power of AI. Deltek has released over 40 AI features into their cloud offerings, driving increased cloud conversion activity. And DAT has industry-leading AI/ML-enabled freight matching capabilities, which I'll detail shortly. These are but a few examples from across the portfolio. Very exciting times for sure.

With that, let's now turn to our segment review, starting with Page 10 and our Application Software segment. Revenue for the quarter grew by 18% in total and organic revenue grew by 6%. EBITDA margins were 43.4% and core margins improved 40 basis points in the quarter. Starting with Deltek. Deltek delivered solid performance in the quarter with particularly strong results in their private sector end markets. Construction, architecture and engineering remained robust throughout. The GovCon business experienced softness in September as agencies paused activity ahead of the pending government shutdown. This timing is unfortunate.

Pipeline activity and commercial momentum had been building nicely following the passage of the one big beautiful bill in July, and we are seeing increased engagement across our customer base heading into the new fiscal year. The fundamentals remain strong. The OB3 authorized significant increases in defense and infrastructure spending that will flow through to our customers once appropriations are finalized. This is simply the timing issue, not a demand issue. Finally, retention levels across the entire Deltek franchise remain very high.

Aderant continues to be incredibly strong and continues to post impressive bookings and recurring revenue growth. The booking strength is broad-based, fueled by their AI-enabled solutions, especially as it relates to AI-enabled compliant time capture and billing and is a combination of market share gains, cloud migration and SaaS growth. Vertafore continues once again to be steady and solid for us. We continue to see consistent ARR growth and strong customer retention and strength across their agency, MGA and carrier solutions. This growth is enabled by their strong go-to-market capabilities and their long-term commitment to product strength.

PowerPlan's performance has been terrific. Their success is a result of several years of business building in the product stack, the go-to-market capabilities, their service delivery really across all functions. In addition, to remind everyone, they serve power generation customers, which are adding capacity as quickly as possible to handle the AI workloads. The setup here should be quite good for a long time. Also in the quarter, we completed the acquisition of Orchard, a tuck-in acquisition for our CliniSys business. Orchard brings additional clinical laboratory capability to CliniSys with particular strength in reference, physician office and public health labs.

Finally, the balance of our application software portfolio continues to execute very well. CentralReach was awesome again in the quarter, driving accelerating adoption of their AI tools and capturing ABA therapy capacity additions. Procare made a great installment of progress with new bookings continuing to be strong, posting low double-digit growth in payments with improved gross margins, though still work to do, in particular, with faster implementation time frames and share of wallet expansion, but meaningful progress for sure. Finally, Strata and Transact were steady and solid in the quarter.

As we look to the final quarter of the year, we expect to deliver mid-single-digit organic revenue growth. This outlook reflects high single-digit growth in our recurring revenue base, offset by declines in nonrecurring revenue, primarily due to anticipated softness in our Deltek business stemming from the ongoing government shutdown. Given the uncertainty surrounding the duration and impact of the shutdown, we see potential outcomes across the full range of our MSD outlook from the lower to the higher end. That said, our businesses in this segment continue to compete and execute exceptionally well. The primary variable remains a higher level of market uncertainty than we typically experience for our Deltek business.

Please turn with us to Page 11. Total revenue in our Network segment grew 13% and organic revenue 6% in the quarter. EBITDA margins remained strong at 53.7% with core margins improving 60 basis points. As we dig into the individual businesses, we'll start with DAT. DAT was solid in the quarter and had strong ARPU improvements. DAT continues to execute exceptionally well on their core strategy of driving enhanced network value for both brokers and carriers. This dual-sided approach positions DAT to better monetize their entire network ecosystem and more on this when we turn to the next page.

ConstructConnect was solid again for us in the quarter. The growth was fueled by strong customer bookings activity and improved customer net retention. Of note, this business continues to make good progress with our emerging AI-enabled takeoff and estimating solution. Foundry is turning the corner on growth, posting continued sequential improvements in ARR, and we expect our Q4 exit ARR to grow year-over-year in the HSD area. Really happy for the team there as they've had to work through some tough market conditions. Next, our network health care businesses, MHA, SHP and SoftWriters were very good in the quarter. Of particular note, SoftWriters is executing at an exceptional level, winning a few very large pharmacy customers and making substantial progress on a high-impact AI solution, which is being beta tested in the market currently. Congrats to Scott and his entire team for their success.

Finally, Subsplash, our most recent acquisition that closed on July 25, is off to a great start, delivering financial results in line with our deal model expectations. Of note, they saw very good market traction with their AI-driven sermon content offering, Pulpit AI, and they deepened its integration with their core engagement platform, driving strong product-led growth, exciting stuff. As we turn to the outlook for the final quarter of the year, we expect to see organic revenue growth at the higher end of the mid-singles area.

As we turn to Page 12, we'd like to spend a few minutes describing the strategic evolution of our DAT business and why we're so excited about its future growth prospects. To start, our legacy DAT platform is the largest freight matching network across the U.S. and Canada. The scale is remarkable, over 1.2 million loads posted and 15 million rate views every single day. DAT is the clear market leader, delivering tremendous value to both freight brokers and carriers, both of whom pay to participate in this powerful network. As strong as the legacy business is, we're even more bullish about where DAT is headed. To bring this vision to life, DAT is building capabilities across the entire freight automation workflow from carrier vetting to broker carrier matching to AI-driven rate negotiation, load management tracking and finally, payment and settlement.

Through deep customer partnering with the brokerage community, DAT is working to fully automate the freight matching process. As this happens, DAT will generate $100 to $200 per load in savings for brokers while giving carriers greater predictability and faster payments on their invoices. What sets DAT apart is this end-to-end product capability and its role as a neutral trusted partner, a Switzerland-like player that equally serves the entire freight brokerage market. This is a truly unique position in the market. This evolution also highlights the Roper DAT partnership at its best. We work closely with the DA team to craft this strategy, then we executed a focused M&A program to strengthen it through 3 strategic tuck-ins: Trucker Tools, Outgo and Convoy.

With the deals complete, DAT is now fully focused on delivering against this strategic opportunity. Important to note, Convoy is an unusual transaction for us as it currently is not profitable, but we expect the financial returns over the next several years to be extremely attractive. The key to success is scaling efficiently, leveraging DAT's advantaged customer unit economics for both brokers and carriers to drive sustained growth and profitability. We are confident in this strategy, market position and DAT's ability to execute. I know this was a bit of a deep dive, but we wanted to share with you why we're so excited about the growth opportunity that sits in front of DAT, true AI-based freight automation.

Now let's turn to Page 13 and review our TEP segment's quarterly results. Total revenue here grew 7% and organic revenue grew 6%. EBITDA margins came in at 35.2%. Let's start with Neptune. As we've said before, Neptune continues to execute really well, particularly around its ultrasonic meter strategy, and we're seeing strong traction in its data and software billing solutions. The new copper tariff that took effect on August 1 caused some short-term disruption. Neptune responded by implementing surcharges to offset the tariffs impact, which temporarily slowed order timing. These actions reflect the benefit of being part of Roper, doing the right long-term thing for customers and the business even when it creates near-term headwinds.

Verathon continues to perform well. In particular, during the quarter, Verathon saw continued strength in its single-use recurring product lines, both BFlex and GlideScope, which remain key growth drivers. NDI also delivered an excellent quarter. As we discussed previously, NDI provides proprietary world-class precision measurement technologies to a range of health care OEMs. These technologies in turn enable guidance-enabled solutions across multiple clinical markets, including orthopedic surgery, interventional radiology and cardiac ablation. Finally, we saw strong execution and growth across CIVCO, FMI, Inovonics, IPA and rf IDEAS, rounding out a solid overall performance for this group of companies. Looking ahead to the fourth quarter, we expect organic growth in the low single-digit area given the very difficult prior year comp and the timing we discussed at Neptune.

Now let's turn to Page 15 and review our Q4 and updated full year 2025 guidance. Starting with the full year outlook, we continue to expect total revenue to remain in the 13% area. Also, given the delays at Neptune and the temporary impact of the government shutdown, which is slowing year-end commercial activity at Deltek, we now expect organic revenue to land in the 6% area versus our previous 6% to 7% range. Relative to our full year DEPS outlook, we're tightening guidance to the high end of our prior range after adjusting for $0.10 of dilution from the $500 million of tuck-in acquisitions completed during the quarter. Specifically, we now expect adjusted DEPS to be in the range of $19.90 and $19.95. We expect to see our tax rate at the lower end of our 21% to 22% area for the full year. For the fourth quarter, we're establishing adjusted DEPS guidance to be between $5.11 and $5.16, which includes $0.05 of dilution for last quarter's tuck-in deals.

Now please turn with us to Page 16, and we'll open it up to your questions. We'll conclude with the same key takeaways with which we started. First, we had a very good third quarter with exceptional free cash flow. Second, we're super excited about the pace of AI innovation and the growth potential in front of our enterprise. Third, we're announcing a $3 billion authorization for a share repurchase. And finally, we remain super well positioned for further M&A activity. Relative to our financial results, we grew total revenue 14% and organic revenue 6%, grew EBITDA 13% and delivered 17% free cash flow growth in the quarter. AI is a significant growth driver for Roper, expanding our TAMs by automating tasks and work across our vertical market offerings. With deep workflow integration, proprietary data and vertical market-specific architectures, our businesses are well positioned to succeed in AI, in fact, have a very high right to win and are already seeing measurable yet early product and commercial results.

DAT exemplifies this strategy in action, evolving from a traditional freight matching network to a fully automated freight marketplace powered by AI. Through this transformation, DAT is unlocking significant efficiency and economic value for brokers and carriers alike, positioning itself for improved high-quality growth. As Jason mentioned earlier, we're excited to announce a $3 billion share repurchase authorization, which we will deploy opportunistically, enabling us to take advantage of dislocations in the market. We're super confident with our talent advantage, our strategy and our execution capabilities, and this first-ever buyback is evidence of such.

Importantly, we remain exceptionally well positioned to execute our M&A strategy. We have north of $5 billion of available firepower over the next 12 months and a very active, large and attractive pipeline of opportunities. Importantly, Roper continues to strengthen its position as an acquirer of choice for both target CEOs and their private equity owners. As always, we'll pursue these opportunities with our consistent, unbiased, patient and disciplined approach.

Prior to turning to your questions and if you flip to the final slide, our strategic compounding flywheel, we'd like to remind everyone that what we do as Roper is simple. We compound cash flow over a long arc of time by executing a low-risk strategy and running our dual threat offense. First, we have a proven powerful business model that begins with operating a portfolio of market-leading, application-specific and vertically oriented businesses. Once the company is part of Roper, we operate a decentralized environment so our businesses can compete and win based on customer intimacy. We coach our businesses on how to structurally improve their long-term and sustainable organic growth rates and underlying business quality.

Second, we run a centralized process-driven capital deployment strategy that focuses in a deliberate and disciplined manner on cultivating, curating and acquiring the next great vertical market-leading business or tuck-in acquisition to add to our cash flow compounding flywheel. Taken together, we compound our cash flow over a long arc of time in the mid-teens area, meaning we double our cash flow every 5 years or so. So with that, we'd like to thank you for your continued interest and support and open the call to your questions.

[Operator Instructions] Your first question comes from the line of George Kurosawa with Citi.

Great to be on the call here. Wanted to first touch on kind of the high-level organic growth picture. I think you can -- took a step back this quarter, but I think you can certainly argue there are some onetime or short-term dynamics at play here. Maybe just if you could frame your confidence in a reacceleration from here, particularly as we start to sharpen our pencils for '26.

Yes. Appreciate it. Thanks for being on the call this morning. So the -- yes, I think you're right. I mean the reason that was a little rough this quarter were the 2 reasons we talked about, the commercial activity at Deltek with the government shutdown and then this tariff-related impact at Neptune. As we think about '26, I mean, it's a little early for us to get super detailed about '26. But if you sort of roll sort of segment by segment, it's been pretty -- in application, it's been pretty consistent trends there throughout '25. Deltek and government contracting should improve next year given the passage of OB3. I think the timing of when that improves is still up in the air a little bit. We'll see that as we get through our planning and roll into next year. But there's definitely sort of improvement happening in that market given the spending attached to OB3.

In the Networks segment, it's been pretty consistent over the last 3 quarters. There is sort of a comp thing in the first quarter. So pretty consistent over the last 3 quarters despite the sort of the headwinds in the freight market. We'll have to see how the freight market evolves next year, but really like the business building we're doing at DAT, as I talked about. As I also mentioned, foundry is going to be better next year. And then on TEP, the Neptune order patterns likely continue normalizing the pre-COVID sort of lead time levels. Orders there have been pretty good. It's just the lead -- and the lead times are going to continue to shorten.

NDI is poised for a couple of strong years, but we really need to get through our planning process to have more clarity on how TEP is going to play out next year. But all in all, we feel pretty good about the trends in GovCon, Foundry, CentralReach and Subsplash turning organic in the second half of next year and the general business building. But as usual, we go through a pretty exhaustive Q4 planning process, which we kick off in a couple of weeks.

Okay. That's super helpful color. And then maybe just one quick follow-up here on the AI strategy. I think you disclosed 25 products last quarter. I'm curious if you have an updated number just to give us a sense for the pace of innovation and just more generally, how you feel businesses are coming up the AI curve here.

Yes. We feel very, very good. I won't rehash all the prepared comments about why we feel that way. But we're going from having a large number of products and key features. We talked about the 40 AI features in the Deltek core that's driving sort of the cloud migrations and SaaS. But increasingly, we're seeing sort of AI SKUs. So we feel real good about that. Now we've got to get through the commercial activities as we release these SKUs across essentially every one of our software businesses now and in the first half of next year that we got to go sell them and commercialize them and then the momentum will sort of pick up from there.

But feel very good, very high right to win, a lot of compounding of knowledge about how to do all this stuff internally. This is -- you can hire some talent, you got to build it. So we feel real good about that. A lot of internal sharing that's going on, which is great to see, a lot of momentum. So we can certainly talk more about that, but feel great about where we are on the AI front.

And the next question comes from the line of Brent Thill with Jefferies.

Just on the buyback, I guess, maybe walk through the strategy, why not leaning harder into M&A versus -- I believe this is your first buyback ever. What drove that decision?

Yes, I appreciate it. The -- so just to be clear on what it is. So it's $3 billion. It's open-ended timing. It's opportunistic and in no way, shape or form, a change in our strategy. Set this in the context of the amount of capital we have to deploy over the next 3 years is somewhere in the $15 billion to $20 billion range. So it's not a change in any way, shape or form. The rationale for it is pretty straightforward. We just have a ton of conviction in what we're doing. And in terms of the talent we have on the team and that lead our companies, the strategies, the AI execution, the general continuous improvement execution, the business building we're doing. And we think this buyback is just clear evidence and support for our conviction there. But we're going to maintain a strong bias towards M&A.

The compounding nature of the numerator is better than the denominator. It's just straight math. We're super active on the M&A front. We cultivate every day. In fact, our -- Janet Glazer leads our capital deployment efforts, had a fantastic meeting 3 or 4 weeks ago, I think, with 18 CEOs of companies that are in the pipeline, so a marketing event, and it was met with great reviews, and we're really becoming sort of a buyer of choice, both for the CEOs of companies and also the private equity sellers. So we feel real good about the execution of our M&A strategy, and this buyback is just a small complement to the overall strategy of Roper.

Okay. Neil, I know the last couple of years, we've had a couple of things that maybe haven't gone the way you wanted to. The question is just how do you derisk this out of the guide? And I think investors have looked at the portfolio and said that you get the diversification aspect, but why do we keep having kind of the setbacks if we're that diverse. So that's the question I'm getting.

Yes. So you're right. I mean we've built this portfolio to essentially take as much cyclicality and cycle risk as you can take out of an enterprise. If you look back over our long history, before we sold and divested all the industrial businesses, we'd cycle up or down 5 to 10 points. Now we're cycling like a point here or there. So we've essentially beaten out ostensively all the cycle risk you can in an enterprise. In this case, it's just -- it's frustratingly bespoke situations. Government contracting, normally, you're in GovTech because of the stability of the end market here. It's been anything but that the last couple of years. Transportation, who would have predicted like a 3-year freight recession. And so it is frustrating these things are stacking on top of each other, but they're bespoke, and we like the construct of the portfolio for sure.

I think the cash flow generation continues to be strong and consistent with what we thought. Obviously, we've had some new deals come in that have been dilutive, but we have been able to sort of push through that. Our guidance is -- adjusted for dilution is pretty close to where we were before. So despite some of the softness we've seen, we've been able to sort of maintain the bottom.

The next question comes from Brad Reback with Stifel.

Software bookings decelerated a little bit sequentially, I think, from the mid-teens to the high singles. Was that predominantly Deltek? Or were there other drivers there?

Yes. It was mainly Deltek, a little bit of frontline. We've talked about the -- some of the funding from the DOE. We don't get a ton of funding down to the states, but at the margin, it does slow down a little bit in K-12. But yes, it's Deltek Frontline. Outside of that, it's very strong. So if you look on a TTM, also a very lumpy dynamic, right? Software bookings are -- can be lumpy quarter-to-quarter. TTM is up low single digits -- sorry, low double digits. So I feel good about that trend. And I would also just call out that health care has been particularly strong this quarter.

Our Strata business, we combined Strata and Syntellis a couple of years ago, and that's really starting to take hold in the market. So bookings are really strong there. And actually, our CliniSys business is doing quite well, too, in Europe and even in the U.S. with some of the bolt-ons we've done for them to get outside of the hospital, that's starting to gain traction as well. So just some color behind the bookings this quarter.

Great. And then, Neil, I think 2 questions ago, you talked about the rollout of the AI SKUs happening now through the first half of '26 and then needing to sell it. That all seems like we should be thinking about this more of a '27 and beyond organic driver as opposed to '26?

I think that's a fair -- we're certainly viewing it that way. I think it's a fair assumption. We'll certainly see progress in bookings throughout next year because again, this is across 20-plus software companies and multiple products across 20 software companies. And so we'll see building momentum. But before it has a meaningful impact, I think it's '27 because of the commercial activity that has to go along with the innovation.

The next question comes from Ken Wong with Oppenheimer.

Fantastic. I wanted to maybe drill in a little bit on just the organic growth. Any way for you guys to help kind of slice what you might have seen from maybe the same-store sales versus maybe the net new organic that's coming on to the P&L. Hopefully, that question makes sense.