LivaNova PLC (LIVN) Earnings Call Transcript & Summary

Good morning, Matt Miksic.And I cover U.S. medical devices here at Barclays. We're very pleased to have with us LivaNova. From the team, we've got Matt Dodds, Senior Vice President, Corporate Development; and Briana Gotlin, Director of IR. So thanks again, both of you for coming and taking the time to chat with us.

No, Matt. Thanks for having us. And Damien sends his apologies, his knee gave out yesterday when we were in Houston. And I told them Matt's got 20-plus years of orthopedic experience, but it didn't get them over. But he apologizes for not being able to be here today with us.

No worries at all. So I thought maybe we start with sort of the clinical pipeline. And I know recover and difficult-to-treat depression has been kind of a -- to say it's been a big part of the narrative over the past year would be bit of an understatement. So we can't not start with that. Why don't we just maybe chat about some of the sort of the -- where you're at now and sort of like enrolling the study on time, submitting the study take us through the time line, which you could probably do in your sleep by now, Matt.

Sure. And we're getting close to all the different permutations. So this trial, the RECOVER trial, very unique in the fact that we have an independent statistician that can look every 25 patients enrolled are randomized. And we're now -- the next one will be 475 patients. We expect to hear back and release that by the end of the month. If, in fact, we hit the upper bound of predictive probability, we'll be able to stop at 475 and just follow them for 12 months. If not, we only have 1 more rev left. It's 500. The trial is designed for each arm, unipolar and bipolar and this is unipolar up to 500 patients. So either way, we have to follow them for 12 months. So we'll get the data sometime in mid-'24 either way. And then a little bit of time to analyze the data, submit it to CMS. So we still expect CMS to make a decision on coverage by the end of 2024.

Okay. And still aiming for a National Coverage Determination on -- the NCD, is that...

Correct. Right. So ultimately, there's 3 paths that CMS can choose. Again, we think, obviously, depending on the quality of the overall data, how significant it is. They can either decide to give us a national coverage decision, which was the whole premise for this trial, they could decide to give us back local coverage, which you know less or most medical devices have we still say would be a big win or they could say no coverage, which if, again, we think we have really positive data that would be very surprising and we think unlikely. So those are their options when we get them the data.

Okay. And I thought it would be helpful maybe just to point out, if anyone's wondering why it is that we've gotten into so many conversations about their statistician and the different enrollment checks. And the way the trial was set up, there is a rationale behind all that and why it's different from other studies? Do you want to talk a little bit about that?

Sure. So the trial design is very unique. It's not medical devices, I think, in health care in general, in that this is a 1-year randomized sham-control arm, 50-50. So you've got half the patients that have a device implanted and turned off for a year. So our argument to CMS was if we hit a very strong upper bound of predictive probability. So the statisticians when they look, if they say with very high confidence, 90% plus, you will hit the P-value at the end of the trial.You can stop enrolling early and then just follow those patients 12 months. And we put a paper out on this. So it was very public and a lot of people were able to really follow along. And the concept is as we get closer to the end, the predictive probability does decline a bit and it's still pretty high. But also, there's a futility curve in the bottom. We have to stay above a certain percentage. And what's encouraging while we haven't hit the upper bound is that the futility curve at 450 patients, we had to show at least a 35% spread between the therapy and control arm. So there is an effect. It's just that it hasn't gotten to the point where we have that we hit the upper end. So we're somewhere in the middle. So we still feel that there's a very high probability of this trial is successful. And either way, we won't know the final P-value in the data no matter how this turned out until the 12-month data comes in. And then the one other thing to point out is there's actually 13 different endpoints per se. It's not an FDA trial, so CMS used questions. But the time and response is an important one, but it's just 1 of 13.

Okay. And yes, it's good to point out. This is an FDA-approved device. So it has been shown to be safe and efficacious enough to be approved. This is about reimbursement. And also that is like why have all these checks why just run the study and spares all the [indiscernible] have gone back-and-forth is there's an ethical -- like in other words, if you were to hit the treatment metrics, you wouldn't want those patients to be sitting there with a device that could be effective but isn't switched on for another year.

Correct.

Right? So does it make sense when you step back and look at it a little bit. Can you talk a little bit about the debate that came up as we get closer and closer to the sort of end of enrollment around the sort of separation of the curves and how the clinical effect of the therapy has some effect on that?

Sure. So if you look at the premise of the RECOVER trial, it was based on 6 separate studies, which in total on 1,400 patients. The largest one was a paper by Dr. Aaronson. And that followed patients for 5 years. And what it showed was that the VNS therapy because this was the longest trial, it really kind of gotten going and you saw the separation between the 2 arms at 6 months. And it continued from there. Aaronson ran out 5 years. RECOVER does go out 5 years, but the 1 difference is Aaron's in the treatment as usual arm just stay treatment as usual, whereas in RECOVER at 12 months. Everybody switches to therapy on. So you become a bit of a self-control. But our expectation is a lot -- you'll see a lot of benefit between months 6 and 12. And then hopefully, we would continue to see a benefit as we follow the patients out even longer. So that is why it's still important that for the trial where we are today, about 65% of the total scores are in. And a lot of the scores, if you think about where we are in the trial, our numbers between months 6 and 12 for the patients that haven't gotten there yet. So hopefully, that will continue to drive sort of higher response rates based on all the prior data.

Right. As they inflect as you get past 6 and start to improve through 12.

Exactly.

Yes. Because obviously, one of the questions that comes up is, is if it hasn't stopped if the hope was that it's going to stop really and it hasn't stopped, maybe we're not getting the results that we want to get. And you'd say that this separation accrues effect would maybe back-end load some of those benefits.

I mean that was -- that's part of it. The trial was always designed to go up to 500 patients. And there's a P-value at the end with a certain level of separation. The early stoppage was on the expectation based on the earlier data that we might just exceed it at such a level early on. There's a possibility. And then we would have high confidence that we would hit the P-value. So we don't have that yet. Still a chance of 475 we get there. But I think what was underappreciated is on the lower end of futility, the curve does bend quite a bit. So we know we've got at least a 35% spread. It will go up in the next look. And so we're either up -- we're at a higher level or certainly trending to hopefully get us there at the end. So the other thing I would highlight is we finally did put out some patient demographic data. It was a poster by Dr. Conway, who is one of the lead investigators in the study. And it just showed this patient group is very sick. There's no prior study in device land, but even in pharma that treated these patients. So we're still very adamant that even an improvement of 65% or more between the arms, very clinically meaningful for patients that really have no option.

Right. Okay. So we'll come back to maybe the sort of the structure in the market and the way it's segmented and where the sort of obvious early opportunity is for adoption in a minute. But maybe if we just want to touch on the OSA, obstructive sleep apnea opportunity and then just quickly on ANTHEM and the reason for shutting it down in heart failure.

Sure. So OSPREY is, that's the trial, the confirmatory trial for sleep apnea. That's moving along. We're getting a lot of consents. We're getting our implants it's still on track for a 2024 FDA submission. The trial 6-month follow-up. So a little shorter than the prior study. And then we do believe that, again, knock on wood and somewhere at this, maybe not wood, that if it's successful, we wouldn't be able to launch relatively quickly after PMA approval. So that one remains on track.

Okay. And then the sort of -- so I have prep study that was stopped was just normal reasons for stopping a trial, weren't seeing the efficacy and took it out of the program. I'd say, sort of zooming out for a minute there was some kind of reaction to that, I think, because there was this anticipation around RECOVER, it was sort of like well maybe this is a little bit treatment or a mechanism of action effect across the rest of your portfolio. But going back to the point about this is an FDA-approved therapy and difficult-to-treat depression, whether or not VNS works and has almost nothing to do with whether it works in DTD. So I think those connection of those dots seemed a little bit major to me around when the announcement came. But it's worth mentioning that there was that sort of confusion, I think.

Yes. No, no, I agree. And a couple of key differences. One, for VNS and for Depression and epilepsy, implants on the left side of the heart and the left side of the clavicle and its stimulates the left vagus nerve because 80% of the signal goes up. Heart failure is the opposite. It's on the right side because 80% of the vagus nerve goes down on the right side. As I said earlier, in Depression, we have data on 1,400 patients. And a lot of it is very powerful data. In heart failure, we only had data on 60-plus patients. But again, it was encouraging data. And we there -- in that trial, the ANTHEM-HFrEF, as you mentioned, we're allowed to look every 100 patients, so -- or the statisticians are, so a much slower cadence. And there, we have to hit 5 different sort of endpoints in order to unblind some of the data and potentially file a PMA on functional endpoints, not mortality morbidity but functional endpoints. And we didn't hit all 5 in the second look. We didn't it in the first. We did them in the second and when we did a deeper review, there was some, I think, some positive information there. But we didn't think it was clinically significant and enough to continue the study potentially all the way out to 1,000 patients. So that's why we decided to stop enrollment, ultimately stop the trial and eventually shut down the program.

Got it. So got about 9 minutes left. I think one of the subjects that came out of the fourth quarter and that sort of caught people's attention, I think, was sort of the direction of travel for new patient implants in epilepsy. So came in a little slower than you were hoping. I think your guide makes some assumptions around that starting to kind of turn the corner here in '23. If you could talk a little bit about sort of the replacement and NPI trends, kind of what gives you confidence that in the next quarter or 2, we are going to see that turn?

Sure. Great question. So for U.S. epilepsy, last year, the replacements did a little better than we thought overall close for the tiny bit better. And then as you said, the new patients were below our expectations. Just to kind of go back a little bit in history to explain what's changed there. Before COVID, we were doing 4,000 plus new patients a year in the U.S. pretty consistently give or take. And then in 2019, we got caught off guard a bit by Epidiolex and the initial enthusiasm around that drug. And we took a really hard look overall at the commercial model. And our view is that we were a little too focused on kind of short-term revenue achievement, and not focused enough on driving physician awareness adoption referrals. So we decided in 2019 to really change the model of this organization. And the problem was we went right into COVID. And so I would say, in 2020 and even 2021, a, change always has some effect. But b, it was hard to really track how they were doing because there's not a whole lot of touches in high-touch model, low-touch environment. So I would say 2022 was the first year again, COVID not completely out of the equation, but the first year that we could really assess what was going on. And we thought the results were a bit mixed. So -- and then also tied into the fourth quarter, we had hoped for a little more on the new patients. So if you look at what we did we essentially removed the bottom 10% of performers. We have much more, I think, aggressive oversight by the team on how things are trending, what are the leading indicators, how are we doing? And then the really encouraging thing is of the replacements for the turnover, again, mostly involuntary but some voluntary we've really upped the game on who we've hired, 2/3 have been people with neuromod experience this time around and on average 8 years. So we're getting, I think, there's a lot more traction in a group of salespeople who can hit the ground running much faster and traditionally, we've had a lot of -- I think, a lot of success in that type of selling model. So that's why we feel really good about how we're going to turn the corner in '23. So the long answer to your question is, we do think replacements are going to be down a little bit this year because our backlog has gone. But we do think new patients will cover it and grow at a pretty good clip sequentially.

Okay. And how long will you bring someone in from sort of a neuromod, I guess, SVS kind of experience how long does it take for that person to be delivering [indiscernible].

So normally, it's about a year before they can get up to close to a level of a multiyear rep. In this case, we think it could be between 6 and 12 months. So -- and a lot were higher earlier this year. So again, I think second quarter, third quarter, we'll see the benefit of this kind of new group.

Right. So maybe side of the gallows for the other 90%. We see improvement in the second quarter. And then back half, we start to see some of the new folks...

Correctly. And there are definitely teams out there within our U.S. epilepsy that are succeeding in the new model. It's just not as many as we had hoped by this time.

Okay. I'm a little curious where some of the reps are coming with all of a sudden, the opportunity set is presenting itself up.

I've got the list of -- I'll tell you offline.

Okay.

But, on spinal cord stem?

Yes. [indiscernible]. And I'm sure you assume that.

Not at all, but -- on spinal cord stem.

Okay. Fair enough. All right. So that, I think, covers some of the dynamics, much of the dynamics in epilepsy. Maybe talking about cardiopulmonary. I guess, what -- you got a strong finish to the year. Do you start to run into comps in the back half of this year? How are you thinking about that, the trends in that business sort of first half, second half?

Sure. So that business definitely has a tough comp. So we guided, we hope conservatively 3% to 5%. There's 2 big swing factors there. The first is the Essenz, next-generation heart-lung machine we've been talking about for a while, this is the year. We just got FDA approval, right? That's a big deal. We've done over 200 cases in Europe. They've been very successful. So we are fully rolling out Essenz in the U.S. and Europe. The other markets will follow later this year, '24, '25, but that's a big deal. We're #1 in heart-lung machines. The second piece is in oxygenators. We assume we're going back to market growth, which is 1% to 3%. But part of the reason we believe we grew a little faster last year was some market share gains. And so there's the potential. We never can assume that's going to be something that maintains because I think it might have been a supply issue. There's upside there if we either hold that share or some of our competitors still have their supply issues. So those are the 2 I would look for, for the areas of upside.

Okay. And then ACS smaller part, but just a quick comment on trends or expectations there.

Can you handle that one?

I got it. So our full year forecast calls for 4% to 6% growth, which implies an increase in non-COVID cases and it assumes a return to growth once the anniversary out of that COVID impact in the second quarter. And then looking towards the back half of the year, we assume a return to growth in the double-digit range.

Excellent. Okay. And again, a smaller part, but still important part, particularly is growing double digits in the back half. So maybe on margins, I don't know that it's got a ton of attention coming out of the quarter, but there's a little bit more sort of flatness and margin expectation, a little less leverage to the bottom line. If you could maybe talk about where those effects are happening in the P&L.

Sure. It's unfortunately, it's macro. So what you don't see -- and Alex has been highlighting quite often is we're getting some nice manufacturing efficiencies. Even in SG&A, I mean, we are definitely driving down cost. And selectively, we're getting price. It is all getting chewed up by labor, freight and higher component prices. And some of that may abate a little bit in '23, I think freight probably the most likely. But a lot of this is it's going to just be whatever we negotiated last year, it carries into this year. So it's still going to impact us in '23, and it will hold the gross margin largely flattish year-over-year. The other area is, obviously, in R&D, if you want to comment a little bit about the portfolio initiatives.

Yes. So our SPI is cost approximately $80 million per year. And this spend is actually forecasted into 2023. Winding on the heart failure program could provide upside to the bottom line. However, 2023 guidance includes the full spend on the program.

Okay. So maybe by the back end, there's some relief if there's not reallocation or timing that you decide to sort of pull a lever on.

We're still working through all the ramifications, but we hope pretty soon we'll have an update with a range of -- if there's savings in '23, how much they can be.

Okay. And then, of course, by the mid-year, we'll have an idea as to whether which direction all these headwinds are going in. Hopefully, the right direction, but I think everyone's kind of leaving the door open for other exogenous that may be or -- who knows. But -- and everybody has sort of a different impact, a different view on headwinds. But I guess the good news is you're not being supply constrained. It sounds like for chips or anything like that. But...

It's always something we're watching. And I wouldn't say over the last 2 years, we haven't had any issues with that. But it hasn't had a big impact on us. Yes.

Okay. Great. Well, we that, we're at time. So we'll stop there. But thanks again for coming and spending the time.

No. Matt, thanks for having us.

You bet.

Thank you.