Insulet Corporation (PODD) Earnings Call Transcript & Summary

All right. Good afternoon from Barcelona. Good morning, I think, in the states where most of you are. Welcome to what I think I'm going to call our third annual podcast webcast here from the floor of the ATTD conference. I think 2 years ago, we started this in Florence, last year in Amsterdam. And this year, where are we? We're in Barcelona. And God willing, maybe we'll do one from Paris next year. It's not a bad thing to come to this meeting in such fantastic location. With us, we've got Insulet's Chief Medical Officer, Dr. Trang Ly. Trang, good to see you again. Thanks for taking the time.

Nice to see you. Of course.

All right. I do have to read one disclosure here for everyone. Please refer to the event calendar published research or Baird's website for important disclosures regarding the companies discussed during this event. With that, why don't we jump into it? I think we've got about 35, 40 minutes or so we'll spend here.

Let's start with the elephant in the room. A little bit of news last night, a medical device recall, urgent letter went out. There has been some manufacturing issues with some lots of Omnipod 5. I think there are 18 adverse events or serious adverse events. So Dr. Ly, I mean, we'd love to hear from you what happened? And then I've got maybe a couple of follow-up questions on how you're hearing doctors responding to this issue in the field.

Yes. Thanks, Jeff, and really great to be here. As part of our routine post-market surveillance and monitoring, we look thoroughly at the complaints coming back from customers and customer feedback. And through that process, we uncovered a manufacturing defect. And the issue is that there was a small tear detected inside the pod in the cannula, which is the tubing that carries the insulin. And as a consequence of that, insulin that is supposed to be infused inside the body is actually leaking in the pod, which leads to under delivery of insulin for the patient. And in the worst-case scenario, prolonged under infusion can cause high blood glucose levels and in the case of patients with diabetes can lead to diabetic ketoacidosis and hospitalization. So that's the issue that can potentially cause harm for patients. And the reason why we notified patients is that if they encounter this issue in multiple pods that is the -- where a serious harm can occur. So we wanted to let patients know the affected lot numbers. So there is a lookup tool on our website where they can come and enter in their lot numbers and then we tell them whether their lot is affected or not. And then we -- of course, we replace those pods for our patients. And then on top of that, we're making improvements to our manufacturing and our quality release of lots so that our patients can be assured of the safety and effectiveness of our product. And that's the most important thing is really the safety of our patients.

And is it safe to assume of the 18 adverse events that were called out, those were all hospitalizations for DKA or most of them?

Yes. So we would count an adverse event if patients had high blood glucose levels and needed to go to hospital. So whether they were in hospital for a couple of hours or a few hours or they were hospitalized for diabetic ketoacidosis and needed days of infusion, it all would count. So yes, there were cases of hospitalization and diabetic ketoacidosis and then those patients recovered.

Yes. And we know -- I mean, I've covered medical devices for 23 years. You've lived in the world of medical devices, most of your career, I believe. So medical devices do cause problems as much as they solve and can be life-saving devices, they can also cause issues, so not unusual. We've seen CGM recalls here in the last year or 2. We've seen pump issues in the past. So I guess -- and I know you're not a manufacturing person, but was it something contained to one line you could find that something was out of spec or something and can be corrected easily. This isn't something that investors, physicians, patients will have to worry about moving forward. Just generally, and again, I'm sure you're limited in what you could say, but...

Yes. What we did let our patients know is that this affected only a very small number of pods. We said 1.5% of pods that were produced in the last year. So this is a very small issue. It's -- people can be assured that the pods that are not recalled are very safe to use.

Okay. All right. And then I'm assuming -- I know when you and I spoke the last couple of days, you obviously couldn't say anything about this until the news broke last night. But I'm assuming as a company, you've probably been investigating this for at least a few days, if not longer. So that's potentially given you some interactions with physicians, I'm sorry, some of your customers. Just has there been any feedback from customers, a loss of confidence, a concern about prescribing Omnipod 5 in the future, anything like that?

I've really only spoken to doctors last night externally since this news came out. And I think people who have been in this space for a long time, physicians do understand that these devices are not perfect. As you mentioned, in the space, there have been several issues from the big companies. And it just goes to show that this space, it's really hard to produce products at scale of high quality for our patients, whether that's CGM or pumps. And I think what people really appreciate is the transparency of the issue. I think people are understanding, and I don't think that they would lose confidence in our product and would not stop prescribing.

Yes. Did that safety mechanism fail in this 1.5% of pods, I just...

No, no, no.

Okay. I guess one other question or not even a question, I guess, a point, it came to mind to me just as you were giving your answer, I think back to your Analyst Day and investors are going to think I'm throwing you a softball here. I'm just being too nice here. But I think back to your Analyst Day, you guys build up such goodwill with your pediatric patients and just the whole Insulet experience of caring for your patients that I got to think that reputation can help carry you through these kind of issues.

Yes. Well, I think -- exactly, but that doesn't even need to be said, right? Like we...

No, no, but my point being, patients love you, love your company. They know you're looking out for their best interest that when something goes wrong, it's not like, oh, it's these guys again. I mean I think you built up so much goodwill. My point is I think that helps you and protect you in a time like this.

Yes, for sure, for sure. But I think in these cases, companies have to be transparent and have communicate exactly what the risk is and how to protect patients from that. I think people -- it's a very serious condition, whether it's underdosing of insulin, overdose, it is a community where the delivery of insulin has such a narrow therapeutic window that, that precision of insulin delivery is so important. And that's something we take great pride in. And I think it's to the credit of the teams that when we detected this signal, we went and investigated clearly and we found the issue and we notified our patients as soon as we could.

Okay. Well, let's maybe move on to happier times or what we really came here to talk about or what I thought we came here to talk about initially. It all starts with the EVOLUTION 2 trial. That is the kind of latest and greatest from Insulet. You published those trial results just a couple of days ago in your symposium. And just to remind investors, that trial is really the second feasibility trial on the path towards fully closed loop in type 2 diabetes. So the EVOLUTION trial, 24 patients. Maybe why don't you just hit us with kind of your view of what the highlights were, and then I have a few questions on it.

Yes. So just on Wednesday at our symposium, we shared results from 24 patients, adults living with type 2 diabetes. And these were adults who were in New Zealand who've been using an investigational product of our latest Omnipod, fully closed loop for type 2 diabetes product. And results were excellent, 68% time in range and very low hypoglycemia. And this will be the basis of the algorithm that is going to be used in the EVOLVE pivotal study. That's our pivotal trial for our fully closed loop type 2 product that's going to start this year. And as we mentioned at Investor Day, that will be the product that is launched in 2028. And what I'll share is that it was one study in a series of studies that we've done. We -- actually, 2 years ago in Florence, we shared results from EVOLUTION 1, which was fully closed loop in both type 1 and type 2 diabetes, small numbers, 12 type 1, 8 type 2, and we got pretty good results, and we've spent the last couple of years refining that algorithm and getting it to the best place it can be so that we could enter studies last year. And we -- what Dr. de Bock talked about on Wednesday was really testing several iterations of the fully closed loop so that we could push efficacy as much as possible. So you could see timing range increased, and we got 66 with the first version, 68 with the second and the third one, we were back down to 66. And the combination that we picked had the highest efficacy, highest timing range, but also very safe with low hypoglycemia. So that will be the basis going into all future studies.

Okay. Fair enough. And yes, you brought up the kind of open concept of the study where you were able to toggle that algorithm. I'm sure there's more scientific term than that, but you were able to make some changes to that algorithm and see in the field during the study how that impacted. Is that different than past studies? I don't remember seeing you kind of change things midstream. And I think is that -- I would assume that's a positive iteration of how comfortable you are with the algorithm or how comfortable you are making changes like that on a short-term basis?

Yes. I think we're just getting better and better with our capabilities. Our computer modeling and computer simulation has gotten better, so we can have better predictability in terms of what certain settings changes can affect outcomes. And then on a research and development side, we were also able to really set parameters that were broad enough where we could really test that full spectrum of how much insulin to deliver. Sometimes when you deliver a lot of insulin, if you deliver a lot, but it's no longer effective, you can cause issues such as unnecessary weight gain for someone with type 2 and actually put them at risk for more hypoglycemia. So getting that recipe just right in terms of safety and efficacy is really what feasibility studies should be about. And they need to be in real-world situations and have a broad range of patients, which not all companies do. But we're really committed to testing things in very heterogeneous populations so that we can really understand how these products will work in the real world because that's where it matters for our patients.

Yes. Fair enough. Can I just ask one question and a couple of investors who are probably on this webcast pushed back a little bit on me when I published that data on Wednesday. The hypo readings, 0.04% at baseline. I think in the algo of the middle of the 3, the one that you've chosen to stick with, that percentage hypoglycemia went up to 0.14%, so up 0.1%. How is that in the context of like what you worry about or don't worry about? Is 0.14%, 0.1% increase clinically almost nonexistent? Is that...

Yes, it's clinically negligible. Yes, it's a tiny amount. So it's not...

Yes. I mean just when I look at some of the other trials out there, I see 1%, 1.3% in hypo and things like that. I was surprised I got pushback on, is that a risk? Is that a concern? But you would see that as a very -- let me ask you it this way, I phrased it to investors. To me, that's going to be seen as a good number by doctors, not as a bad number. Do you agree with that?

That's right. Yes, that's a low number.

Okay. Okay. Fair enough. You also mentioned type 1. Obviously, we know after the first EVOLUTION study, you decided to focus first on type 2. Maybe did get a signal on a little bit higher hypo in the type 1 population, so you wanted to refine some things. I think one of the nuggets you put in your Wednesday slide was that there are still some type 1 fully closed loop efforts continuing this year, maybe even some feasibility trials, things like that. As you have that open concept in type 2, and we're able to iterate on the fly as you did and now that you've started type 1 fully closed loop again, it seems to me like the path -- you can run 2 paths at once here, the type 2 path and the type 1. And even in your Analyst Day, while you kind of laid out the '25 through '27 or '28...

'28.

Through '28, but there could be things beyond '28. That type 1 fully closed loop may not be dramatically beyond that '28 window that you kind of went up to in your Analyst Day. So I'm sure you won't confirm a type 1 fully closed loop time for us. But just help us understand what your efforts in type 1 are going to look like over the next few years.

Yes. I think when we think about fully closed loop, I'll just share that our vision of fully closed loop is different to what other companies are saying. And for fully closed loop for type 2, our product has no meal announcement, no boluses, no settings. And what that means is easier onboarding for clinicians and ease of use and simplicity for the patient. All they have to do is take the adhesive off, put the pod on, sensor on and then off they go. Omnipod takes care of the rest. And that is visionary. And like we had doctors come up saying, can you really do that, Trang?

So you're essentially raising the bar for what the definition of fully closed loop is.

Yes. Exactly. And we think that's possible in type 2 because of the safety profile and the risk of hypoglycemia is less and the risk of DKA, diabetic ketoacidosis is less in type 2. So we think it's a doable process. And that's why we're investing in a large clinical trial to get this product to market. And we really feel that it will unlock a huge market that is out there that's not accessing automated insulin delivery today. So I would say in type 1, I think we're going to definitely continue advancing hybrid closed loop, and we've talked about the enhancements with Omnipod 5 today and also next-generation hybrid closed loop with Omnipod 6 and then continuing on the hybrid closed loop side in terms of making things more and more bolus optional. How I see it is that every time we iterate for our hybrid closed loop side, we're going to be doing more of the automated insulin delivery and then there is going to release that burden for the patients. I don't think any company is going to be able to go from like where we are today all the way to fully closed loop for type 1 for all patients. But I do see that there will be a spectrum. There's going to be patients who want to engage and want bolus in their system and products like Omnipod 6 is going to allow them to do it. And then there's going to be patients who want to set it and forget it and not engage. And they might have a different set of outcomes with that product. And so getting that recipe just right for our type 1s is something that we're working on, and we are conducting studies this year on that. But as you know, the risk of hypoglycemia in our type 1s is greater. And what we care about is our children and our little ones. And I'm always thinking about, okay, if we give that little [ cramming ] bolus when we detect a meal, how is that going to work in my little 2-year-old. That's what I'm thinking about. So that safety of that algorithm is really important. So we've got to make sure that it's safe enough before we go into a pivotal study.

Yes. I think one thing I've noticed at this meeting, especially as there's been a big focus from some of your competitors and you on fully closed loop. One thing I've noticed is people are starting to choose their words very carefully. They're calling some things fully closed loop. And I think some are calling them fully closed loop that would not meet your definition. But I also think some are starting to use terms like meal announcement optional or optional meal announcements, things like that.

Like bolus optional.

Fat bolus optional, exactly. I'm starting to hear those terms. But I think that can still be a fantastic improvement for a lot of these type 1s, right? I mean they don't have bolus if they don't want to. They don't have to put in 12 different things at the start of the pump. You don't have to tweak the settings as frequently. I mean there's still a lot of simplification that can happen between type 1 AID use today and fully closed loop never have to touch it. There's still a lot of benefit that can come.

Absolutely. Yes, that's right. And we're committed to our patients all the way along that path.

Yes. All right. Fair enough. Anything else on EVOLUTION, on fully closed loop that we should talk to? I do want to move over to some of the 2026 Omnipod improvements that you're launching and maybe even Omnipod 6.

Yes. I would just say I'm just really proud of the team and really proud of the fact that it was a very heterogeneous population, total insulin requirements going all the way up to 200 units a day. And I think that -- our approach to clinical studies is different from the other companies, and it's a really robust data set that I'm proud of.

All right. One thing I did think of -- I'm sorry, I said last question, but as you get type 2 approved, let's say, you get the type 2 fully closed loop approved in 2028 or launched in 2028, that would be a separate prescription, right? The physician would have to write a prescription for that pump as opposed to a standard O5.

Yes.

In that case, and I know you would never advocate for off-label use. But if a physician had a patient, let's say, was 26 years old, didn't want a bolus or maybe was very in tune with his or her diabetes and could take some breaks at times. The physician could write that prescription for that type 1 patient, I would assume, and you could have like a fully closed-loop option for type 1 in that year. Again, not that you would advocate that, but it could happen that way.

It's off-label, and we need to better understand the safety profile of the product in type 1s, and we'll need to include that in labeling. So more to come there. But right now, it's intended to only be for type 2 diabetes.

Okay. All right. Let's move on to some of the launches that are happening this year. You've got some algorithm improvements. You've got the 100-milligram per deciliter set point. Is that out in the U.S.? Now I can't remember.

It's a limited market release.

Limited market release now, yes. Is that going to launch anywhere outside the U.S. anytime soon, number one? But number two, just the 100-milligram setting, what does that bring for patients?

So that will be U.S. only this year. And what it brings for patients is another option for them to have tighter glucose control. So I think the way we explained it on Wednesday was it's not just another option because if you -- you can imagine that when you design an algorithm, you could set certain constraints so maybe there's not much difference in insulin delivery. But what we actually showed people was a simulation where we conducted if someone had chosen 120 versus 110 versus 100. And actually, today, looking at all of our users across the world, only about half of our users use our lowest target of 110. So there's a lot of people using 120 and higher because they're quite rightly afraid of hypoglycemia. So they might want to set their targets a little bit higher. And so what we showed is if you move from 120 to 100, just a simple toggle that you can actually gain up to nearly 5% time in range improvement. And actually, the algorithm can deliver up to 12% more automated insulin, which is really quite remarkable from just one toggle change where the patient doesn't have to do anything. I think that's going to be a really welcome change for patients. And then -- and also competitively, I think it's also going to be important to make it a much more competitive system in the marketplace. And then we also announced another change, which is the maximum insulin delivery alert. So one of the -- so we care a lot about safety and the alert is exactly, as I described, like when we've delivered what we think is a lot of insulin for this patient based on their history, we tell the patient, we're like, hey, it's -- we've been delivering a lot of insulin in the last few hours. You might want to check your blood glucose level because maybe the pod is not on the body or maybe your CGM is overreading or this is an unusual situation for you. But what we found in the real world is that in the majority -- in the vast majority of cases, it was because people had missed boluses, multiple boluses where they should have bolus. And so in that scenario, people tend to reach what we call limited mode. And in the -- with the change, what it means is that people who tend to not bolus frequently, we're going to keep them in automated mode for longer and they're going to have fewer interruptions. And I think that will be a welcome change for both patients and clinicians. It's one of the objections that we often hear from clinicians about adolescents and teens.

Yes. If they don't bolus, then they kind of get kicked out of automated mode?

That's right, yes. And then they forget to go back.

Okay. Let's say they forget 1 or 2 meals in a row and they get kicked out. What do they have to do to get back into kind of closed...

It's actually just a simple toggle of going back into automated mode, but life gets busy and people forget. But we can tell all of this from our data actually. So that is what gives us a lot of confidence.

And wasn't there in the algo improvements this year? I know there's going to be some changes in Omnipod 6 as well. But for this year, some increased algorithm responsiveness, too? Or is that the high?

It is more responsive to a missed meal bolus. So when you set a 100 target, it means that the algorithm is going to kick in and give you more insulin on that upswing, so as...

We saw the downswing kind of in that...

That's right. So it just flattens everything a little bit. So there is that improvement, but it really is just those 2 changes. And then with Omnipod 6, there'll be even more enhanced algorithm changes for a missed meal boluses.

Would you think as this algorithm gets out in the user base and then as we'll go talk in a few minutes about Omnipod 6 next year?

Yes.