Abiomed, Inc.
Q3 2020 Earnings Call Transcript

Published: 6 Feb 2020

[Operator:]

Ladies and gentlemen, thank you for standing by. and welcome to the Third Quarter 2020 Abiomed Earnings Conference Call. At this time, all participants lines are in a listen-only mode. After the speakers' presentation there will be a question-and-answer session. Please be advised that today's conference is being recorded. I would now like to hand the conference over to your speaker today, Mr. Todd Trapp. Sir, you may begin.

[Todd Trapp:]

Thank you, Chrystal. Good morning and welcome to Abiomed's third quarter of fiscal 2020 earnings conference call. This is Todd Trapp, Vice President and Chief Financial officer and I'm here with Mike Minogue, Abiomed's Chairman, President, and Chief Executive Officer.

[Michael Minogue:]

Thanks, Todd. Good morning, everyone. In the third quarter, Abiomed delivered $222 million of revenue, up 10% year-over-year with an operating margin of 31.7%. As previously disclosed, we had a strong start to the quarter highlighted by 24% global revenue growth in October. However, within the quarter, the company was negatively impacted by two misleading presentations at the American Heart Association, American Heart Conference in mid-November. Abiomed remains steadfast on our goal of creating the new field of heart recovery and becoming the standard of care for circulatory support which requires Class I clinical guideline recommendations for high-risk PCI and cardiogenic shock. Today, we are announcing the start of the next wave of clinical studies designed by the true experts in the field, with best practice protocols, derived from 15 years of clinical experience and studies. In the history of the company I feel most confident now about the strength of our innovation, clinical outcomes with protocols, and field team. We also have the infrastructure and expertise to execute multiple clinical trials. Our customers are energized and motivated to attempt again to randomize high risk patients and further validate the clinical benefits of Impella’s support. Our formula for success has been training, data, and time, while we continue to innovate and improve patient outcomes. On today's call, I will provide updates on physician education, current and expanding clinical data and studies, and I will also highlight our breakthrough innovation. Before I get into these initiatives, I would like to comment on the AHA presentation published in its journal called Circulation. To be clear, Abiomed believes that the medical field is enriched and advanced by free debate about the merits of technology and its clinical benefits.

[Todd Trapp:]

Thank, Mike and good morning everyone. As Mike mentioned, we delivered revenue of $222 million in the quarter, an increase of 10% on a reported basis versus a tough comparison of 30% growth in Q3 of last year. By region, U.S. revenue grew 8% to $186 million driven by 5% increase in patient utilization. For modeling purposes, U.S. product revenue was $177 million in the quarter. In October we had a record month with 24% global revenue growth and 16% U.S. patient utilization which we believe was the result of the progress on our key initiatives. Unfortunately, the presentations in November impacted our growth. Our elective business saw more of an impact with high-risk PCI declining 4% in the quarter, while cardiogenic shock grew 13%. Outside the U.S. revenue totaled $36 million, up 30% on constant currency driven by strength in Europe and Japan. In the U.S. at the end of our fiscal Q3 the Impella 2.5 and CP have reached 1425 sites. The Impella 5.0 has been placed in 640 sites and the RP is currently in 521 sites.

[Operator:]

Thank you. And our first question comes from Raj Denhoy from Jefferies. Your line is open.

[Raj Denhoy:]

Good morning. I wonder maybe I could start, Todd, you updated the guidance toward the very end. You know, I guess I'm curious to know how things are trending in the field. Right? You noted January continues to be somewhat weak similar to December. Sow has your response to the AHA publication has been received in the marketplace? Are you still hearing some pushback? And really the question is, when do you expect that the pressure will start to ease such that growth might start to accelerate again?

[Michael Minogue:]

Well, I'll cover the response in the field, Raj. The physicians that are the published experts are the ones that are currently enrolling. They are frustrated because the paper itself got more media attention and more press than some of the publications that show improvements in survival with nearly 95% native heart recovery. So, publications on best practices and shock like Inova or the Detroit Cardiogenic Shock Initiative, or the National Cardiogenic Shock Initiative, or the Hanover publication, or even the Shock Working Group. So, they are frustrated. The second component, the Circulation paper declared that the sites that did the most or that the most experience had the worst outcomes, implying that they have the worst outcomes not that they're treating this -- a sicker population, and the challenge with the Circulation is that they're mixing high-risk PCI patients with shock patients, and we had sicker patients. We also know that they eliminated patients that are escalated on a balloon pump, so they biased the data and took out the sickest balloon pump patients. So, they are frustrated and they are shocked it got published in Circulation, and they are also shocked that their conclusions were allowed to be so strong on an observational database. They have asked us to buy the data and get access to it, so that we can evaluate it and counter with the specific details, which we're doing for both the NCDR and this premier database, and they are galvanized. So there, we - the positive of this is they are totally engaged. We know there has been more than five letters written to Circulation, none of which have been published that are very strong in favor of the clinical science. And again, the slides that we've posted, many of them and a couple more are being utilized and shared with our key opinion leaders to show that no other company has the totality of the data. And in fact what some of these centers that are authors, their standard of care and their usage is really inotropes, balloon pump, and ECMO. So, I think the positive of this is it's going to allow us to get to that next level to push forward on the new studies, and I'll let Todd comment on the guidance question.

[Todd Trapp:]

Yes. So as you remember back at JPMorgan, we talked about the lower end of the range really assumed that December run rate was - for most of the quarter kind of continued and the higher end assumed that what we saw in October, we recovered earlier in the quarter and unfortunately January has played out more like December from a U.S. patient utilization perspective. So Raj, when I look at the patient utilization in January, it was up 3% sequentially from December, but from a year-over-year perspective, still relatively flat. So again, it's something that we watch every day. We did kick off. We just launched our multi-city tour. We're in the Southwest this week, and so we should see how that progresses. And then again in the fourth quarter, we're still launching some of our new products like CP with SmartAssist as well as the 5.5, and then some of the investments we've made in the U.S. distribution structure around the field leadership should start yielding some benefits. So, it's something that we're watching pretty closely, but unfortunately we haven't seen that rebound yet in January.

[Raj Denhoy:]

Right, and I guess, the math suggests that then the fourth quarter is going to come in something in the 2% range. And I guess the question is, really as you start thinking about fiscal 2021, right, and what kind of revenue growth one should expect out of the business, what we're all trying to figure out is kind of the pace of that recovery and when we might start to get back to more positive growth. And so, I don't know if there's anything you can offer in terms of when that might start to lift off at very low single-digit growth rate.

[Michael Minogue:]

So Raj, we can't comment on next year's forecast yet, but the - we know that our data is better, our clinical outcomes are better. The new innovation that's rolling out for SmartAssist also helps improve outcomes and ease of use. This is a bump, and we stand by the data we have, and we are working with the leaders in this space to correct for this misinformation. In the end it's up to the clinical community to digest the data, but we feel very confident that we will in the end become the standard of care, and we will utilize this time to leverage more studies.

[Raj Denhoy:]

Okay, fair. Maybe just one clarification, you mentioned the expandable sheath, you are saying the end of next fiscal year. So, you get yourself the better part of 12 or 14 months. Given that's a 510 (k), is there still some design work or something that has to happen on that product before you begin the regulatory process in the United States? It just seems like it's 12 or 14 months, it seems like a long time to get through the 510 (k).

[Michael Minogue:]

Well, we want to be conservative to make sure that we're comfortable with the timeline. We've already done 16 patients, and in April we're going to do another 8 to 10. We want to make sure as we enter this space that we don't create any new complications, and our process is always to get the best outcomes, and sometimes we go a little slower for that, but we think that's most prudent.

[Raj Denhoy:]

Okay, thank you.

[Operator:]

Thank you. Our next question comes from Matthew O'Brien from Piper Sandler.

[Matt O'Brien:]

Good morning, thanks for taking my questions. Just, I guess, Todd for starters on a finer point on guidance for Q4 here, I mean it assumes U.S. Impella revenue is down year-over-year, is that right?

[Todd Trapp:]

I don't think it assumes U.S. Impella revenue is down, but it's probably more in the flattish range, flat to up low-single digits.

[Matt O'Brien:]

Okay, so then maybe O.U.S. slows a little bit more?

[Todd Trapp:]

Yes.

[Matt O'Brien:]

Okay, got it. And then as you guys and I think Raj kind of touched on it a little bit, but as you think about the high-risk PCI group in the U.S., as you are - there is clearly a contraction going on right now at that group, not using the product at all. Have you been able to isolate or identify a bigger chunk of that group has decided to start to reduce the number of Impellas that they're using and how do you reach out to those guys directly, and then how do you get them to come around to the totality of the data that you have to get them back using more aggressively and how long does that take, is it 6, 9, 12 months?

[Michael Minogue:]

So, Matt as we stated, we're seeing more of an impact on high-risk PCI patients, because they are elective and they're more dependent on the referring physicians in the community. Unfortunately, the referring physicians in the community saw some negative headlines, severe headlines on safety that were unjustified and they sometimes only read the abstract. So we're working with them. The other impact we've had is for high-risk PCI with the dabblers, the people who are starting to use it. They may decide to put off treatment or get in and out or stage of the patients. So we're working through that. We also saw some of these headlines have impacted a little bit of what the hospitals are doing from a media perspective. So we're going to try to transition that phase from people that are being aggressively marketing against us to the sites that are now going to start marketing these examples where they can do high-risk PCI patients and they can see improvements in EF. And to remind all the investors, PCI in general, historically does not show an improvement in ejection fraction or heart function. It usually relieves symptoms, chest pain. CABG or open heart surgery has increase to EF because you have complete revascularization with the vein and every publication on PCI shows that complete revascularization reduces major adverse events and that's the ultimate goal. So getting in and out, and not getting complete revascularization limits the performance of PCI and that's the bigger goal and that's what people are focused on and that's what PROTECT IV will highlight.

[Matt O'Brien:]

Okay. And then last one from me Mike, you kind of touched on it, PROTECT IV, but I think what PROTECT II it took you, I think that study was about three or four years to be completed. How do we think about how long this one will take? And then PROTECT II, I'm pretty sure was randomized versus balloon pumps and in this case it's not, is that going to be an issue with the clinical community given that you're not randomizing versus some other type of hemodynamic support?

[Michael Minogue:]

Well, that's a great question. So the PROTECT II does give us the data against the balloon pump. What we see now as far as the critics or the skeptics, is they are not necessarily promoting the balloon pump because the balloon pump has complications, it does have an increase of stroke, it does have vascular complications and has major bleeding as published in the literature. However, the folks that talked about getting in and out. So they will do a radial, they will get in and out, they won't inflate the balloon for the mandated time by the manufacturer for the stent. They don't necessarily do atherectomy. They limit how much contrast they have for visualization. And that's really the broader bigger market, that's also gives us the ability to look at reducing readmissions or improving complete revasc or reducing acute kidney injury or even reducing readmission. So that's the bigger, broader market. That's when you're in and out to hundreds of thousands of patients every year with PCI and that's the focus. I think from a speed perspective, we'll give that data out as we get closer, but we're also doing more patients, a factor of 10 times more patients per week from when we were starting with PROTECT II. And when we started with PROTECT II, we didn't have 510 (k) and then we had 510 (k). So we have the infrastructure, the band work. We have the field team, but we're going to do it right, because we want to have a Class I recommendation. So patients that are termed high-risk PCI are turned down for surgery, can get the benefit of Impella support, so the physician can do complete revascularization.

[Matt O'Brien:]

Very helpful, thank you.

[Operator:]

Thank you. Our next question comes from Danielle Antalffy from SVB Leerink. Your line is open.

[Danielle Antalffy:]

Hey, good morning everyone. Thanks so much for taking the question. Just wanted to ask about Japan here for a second. I know you said in your prepared remarks it was in line with expectations. It was down sequentially, though, and this is the first time it was down since sequentially since you've launched. So just wanted to get more color on why, what the drivers there were? And then I have one follow-up?

[Todd Trapp:]

Sure. Thanks for the question Danielle. So, as we mentioned in the last quarter, we opened up far fewer sites in the second half of the year as we did in the first half. So just to remind you, we opened up about 49 sites in Japan in the first half of this year and in the second half we opened up 10 in this quarter and we're going to open up a little bit less next quarter. And it's really as we close out the post-approval study, and really the team is focused on rolling out CP in Japan. And so really that's the difference between the $9.5 million that we had in Q2 to the $8.6 million, so it's literally lower site openings in console shipments.

[Danielle Antalffy:]

Okay, so nothing about the AHA data, that's not having an impact ex-U.S., Japan or Germany?

[Todd Trapp:]

No, I would say, it's, we haven't seen an impact in Japan. In Germany, I would say, we did see a little of that AHA noise start to spill over in the month of January. So we're seeing a little bit of softness in Germany as a result of AHA.

[Danielle Antalffy:]

Got it, and then my one follow-up is, maybe this is for you, Mike. One of the questions we're getting is how we should think about the government agencies looking at these AHA data presentations and Circulation publication. I mean, obviously this data set is flawed, but CMS over the last two years has proposed rate cuts and every year they've reversed those. But how does this, this was – the one AHA data presentation was a cost-effective analysis? I mean does this give them leverage to potentially cut reimbursement? I don't know what you can say about that, but would love your views?

[Michael Minogue:]

Sure, two things. So one, as I said in my comments, we want the FDA involved. The FDA has, we've provided the data that they - in the publication. We are getting access to that data. And we also are getting access to the NCDR data. But remember, from 2015 to 2017 we have all those patients in the IQ Database and we have some of those patients in multiple databases. So we can break it out, the high-risk PCI, shock, right heart failure, 5.0, by name, by hospital, by indication. So we have more data than exists in those databases. The NCDR database did not actually have the ability to track some of the hemodynamic information until the end of 2017. And so, we have access to the data and we'll continue to push that. For CMS it's a great question and we are in with CMS on a routine basis showing them our new data, talking about heart recovery, talking about our ability to enable improvement in EF and reduce readmissions, and acute kidney injury. And we want CMS to look at all this information. We want them to look at these other alternative therapies, inotropes, intra-aortic balloon pumps, ECMO. We want the FDA now to look at all those adverse events that they have for vascular complications and bleeding and stroke. But again, the reason that both presentations are interesting and somewhat flawed is they eliminated all the patients that didn't do well on the balloon pump. I don't think any paper that we would submit or any study we would try would ever get completed if we said we're going to published a paper on patients that got Impella only in the cath lab. The intent to treat, if it's balloon pumps, then the intent to treat is the balloon pump and if they have to go onto other technology, then – that should be counted, especially from a cost effectiveness perspective. If a patient gets ECMO and as a result has to go on to and LVAD then do a transplant that's $1 million patient, that's $1 million patient to CMS and hospital charges, that's $1 million patient to an insurance company and all those things should be looked at. So in the end CMS looks at the data. It's a hospital charge equation and remember we have high-risk PCI and shock patients and what you're expected to do is try to find ways to improve productivity, which is what we do at Abiomed. We expect that to come down every year. But what CMS has instituted is a network of ways to get paid for Impella whether it's hub-and-spoke where you put it in and transfer or if it's biventricular or if it's just left side or right side. So we'll work with both agencies. We stand by our data and our data is more comprehensive and extensive and we have more cost effective as data than any database out there, and we have the largest database with high-risk PCI and shock patients that exist in the world and will continue to utilize that.

[Danielle Antalffy:]

Thank you so much.

[Operator:]

Thank you. Our next question comes from Chris Pasquale from Guggenheim. Your line is open.

[Chris Pasquale:]

Thanks, and Mike I have to say first kudos to you and the team for being willing to do the kind of clinical studies necessary to generate some of this level 1 evidence. I wanted to touch on one point with each of the studies. So on RECOVER IV I just want to make sure I understand your comments there, do you plan to hold off on initiating that study until STEMI DTU is enrolled?

[Michael Minogue:]

So the answer is not until it's enrolled. We want to get the RV sites up and running, feel comfortable that we can now add it because we have the infrastructure. We've made the investment. These are some of the top users. And remember, they are putting the device in at all hours. The device can be in for six plus hours, the patients will be in the ICU in some cases, in most cases and they are already doing something that has some ethical logistical issues by waiting 30 minutes before they open up the blocked artery. The flip side of that is since we are screening now for all STEMI non-STEMI patients, we're going to get 10% of that population is going to be in shock. And therefore at many of these centers, we can just go to the next protocol, which would be shock and then we're going to utilize the protocol of NCSI or Inova for best practices, compare it to the other arm, which is going to be inotropes, balloon pump, ECMO, all the things that they do today at sites that don't have Impella. And to point out and I appreciate the comments to continue to do studies, we've continued to do studies since 2006. It tells you something about the technology that it's such an ethical challenge and logistical challenge to withhold hemodynamic support. LVADs for acute shock have never been randomized in the history of the FDA. Impella - this is the first time we've done it. For DT and bridge and transplant, the only randomization to optimal medical management was in HeartMate I and it wasn't necessarily an ethical challenge because you were actually giving a pump to a patient that was dying of heart failure. So these are very difficult and challenging studies to do for ethical reasons, but we're going to push through that and have certain ways to crossover, but they're going to be really hard endpoints to crossovers, so there is no confusion over how sick these patients are.

[Chris Pasquale:]

Thanks and then on PROTECT IV, I appreciate the importance of comparing supported PCI to unsupported PCI and the idea that perhaps it's that unsupported group that's the bigger opportunity. But the recent papers do muddy the water in terms of the incremental benefit of Impella versus balloon pumps? So why not have a third arm in this study where patients get balloon pumps and where you just count escalation of therapy perhaps as a failure on a composite endpoint to show the difference between that sort of support and the support that Impella provides?

[Michael Minogue:]

That's a good point. And as I said, we've spent 8 months working on this study with the experts. The balloon pump itself has adverse events and so going again nothing is actually a tougher competitor. Most people that do high-risk PCI, again the people on these publications are not published and are not high-risk PCI practitioners. They don't use a balloon pump because the hemodynamic support proven and measured is about 0.2 to 0.4 liters per minute if it's timed perfectly, which it's not in most cases. So it's a harder bar and if they're going to do and get in and out that's really what they want to compare to. That's the bulk of the patients that are out there. We don't necessarily see high-risk PCI. We don't see the balloon pump as a competitor anymore. And even from PROTECT II and PROTECT III we are now going to have 1,500 patients to compare to PROTECT II and in PROTECT II remember, if you had atherectomy with the balloon pump, you had a 30% repeat revasc rate in 90 days. So, one out of three of all those patients had to come back, because the balloon pump didn't provide hemodynamic support. So we're going to go for the larger opportunity. That's what our advisors believe and most of these patients in the future, it's really going to be almost similar to on pump surgery versus off pump surgery, which about 85% of the patients in the U.S. have surgery on pump. And so to some extent, we're looking at on pump PCI versus off pump and that will clarify and speed up the study.

[Chris Pasquale:]

Thanks.

[Operator:]

Thank you. Our next question comes from Jayson Bedford from Raymond James. Your line is open.

[Jayson Bedford:]

Good morning and thanks for taking the questions. Just a couple, Todd, you mentioned trends in January earlier and I think you said 3%, up 3% month-over-month, flat year-over-year in the month of January. Just curious, was that in reference to worldwide or U.S. growth?

[Todd Trapp:]

That was U.S. patient utilization.

[Jayson Bedford:]

Okay, perfect. Mike, you've mentioned a couple of times that a Class I guideline recommendation for percutaneous unloading is a goal. Just in terms of timing, when do you expect to get a Class I recommendation?

[Michael Minogue:]

So the PROTECT IV is our priority study right now and it's already in the works. We already have a committee, and we are now going through the process of starting to screen the centers that will accept and we hope to be enrolling patients within the next year.

[Jayson Bedford:]

Okay. And I apologize if I missed this, if you talked about the size and the follow-up of that trial?

[Michael Minogue:]

What we said is we're going to give details, more details at ACC and we will give a lot of details at the Investor Day in May. Just for the record, currently today for our guideline high-risk PCI is Class II B. The intra-aortic balloon for shock in Europe and Japan is Class III, which means it's harmful for cardiogenic shock, which is why the paper that's currently out is so questionable because some of the physicians have questioned, what is the purpose of that paper, is it to try to encourage a Class III recommended device or is it too softly encourage ECMO for shock patients.

[Jayson Bedford:]

Okay, thanks, that's it from me.

[Michael Minogue:]

Thanks .

[Operator:]

Thank you. Our next question comes from Marie Thibault from BTIG. Your line is open.

[Marie Thibault:]

Hi, good morning. Thanks for taking the questions. I wanted to ask one on the date here and then one away from it. With the early efforts we've been making to make calls of KOLs you're starting the road show to get the data out to centers. Just on an early feedback basis, and kind of a one-on-one basis, are you seeing a positive impact person by person at this point?

[Michael Minogue:]

Yes.

[Marie Thibault:]

Great, that's what I wanted to hear. One asking this on away from all the confusion here, I would love to hear more about how the Impella 5.5 launch is going. Is it on track with your expectations and what can we look for over the coming quarters?

[Michael Minogue:]

So Marie the Impella 5.5 is what I believe heart surgeons have wanted for the last 15 years. When I started at Abiomed we were a surgical company and people are always talking about a minimally invasive, longer term, wean-able pump, where you didn't have to core out the ventricle or you didn't have to crack open the chest. And so the Impella 5.5 is our first step into that equation, the Impella BTR pump will follow it. That will be a longer-term VAD implanted a similar way, but it will run for more than a year and allow for patient discharge. On both the Impella 5.5 and the BTR pump, they will have SmartAssist. So we're able to actually monitor and see how the patient is doing. We're able to look for signs and predict recovery. And of course with Impella Connect we track that patient in the cloud. And so – it’s an absolute breakthrough. Fortunately the surgeons are not as susceptible to the noise from AHA. We have huge demand. The feedback has been incredibly positive, especially since we've been able to duplicate that cardiomyopathy shock patients can actually return to baseline come off their inotropes and have their kidneys either recover or protected, which allows them to be explanted and go home without any further technology. Such a big win, especially for a large population out there that’s 72 that's acutely decompensating and when they come in, you don't have a lot of options to get them back to baseline unless it's more invasive or an escalation to inotropic therapy, which has been proven to increase mortality. So we're super excited. The top centers, the top heart surgeons in the country have embraced it. We've done over 100 patients and the publication is pending. We also believe it's the most blood compatible pump and VAD ever created and we will be measuring and publishing that data as well. And there's just a huge opportunity here as a forward flow pump because it does rest and recover the device with the heart, but as you wean it down, we can measure via the pressure volume loops, we can look at pressures in the LV and really get a good or have good timing on weaning somebody back almost the way you'd wean a hamstring.

[Marie Thibault:]

Great, thank you so much.

[Operator:]

Thank you. Our next question comes from Chris Cooley from Stephens. Your line is open.

[Chris Cooley:]

Good morning and thanks for taking my questions. Just two quick ones from me at this point. Todd, maybe this is something we have to wait for the upcoming fiscal year guide, but could you maybe help us think a little bit about the related clinical trial expense that would still fall into fiscal 4Q? Just want to make sure we can triangulate to more correctly there. I think it’s great effort to be undertaking, but just - from a modeling perspective want to see if how we should maybe contemplate that? And then maybe just as my follow-up Mike, I think I know the answer here too, but I've got to ask and one of the prior questions, you mentioned that you were in contact with the FDA on the data. I guess, I just - can you give us some color regarding the characterization of those discussions and what we should interpret from that or is this just the acknowledge that the agency is aware of the discrepancies in the Circulation paper? Thanks so much.

[Michael Minogue:]

Chris, let me answer the second question first. I don't think we heard the first question. There was a little bit of static. So we worked all the time with the FDA. We collaborate with them. We have an ongoing FDA study now, we're doing a review of some of the future studies. We have the five post-market approvals. So we're always talking with them. We're always communicating our database. And the reason that we want to communicate with them is we have an obligation as a company for patient safety. We track nearly all our patients in the U.S. I don't believe there is another med tech company that does that. And certainly, no one in this space does that. And so, if things progress and we were not able to get access to the data, we have Circulation to have the authors validated and provided at least to the FDA, if they want so we can fulfill our – regulatory obligation. That request was denied and so the FDA is its own separate agency. They have seen the publication, but they also have seen all our data. And if they were concerned, I think you'd know that because we currently have a STEMI DTU study going on for patients that aren't in shock where we're moving now for patients that just for heart attack. So they see the real-time data. The frustrating part is, this – I mean paper, it basically omitted every single FDA study. It even omitted the STEMI DTU publication that was in Circulation and presented at AHA as well as the PROTECT II study that was published in Circulation. So the data for vascular complications, major bleeding and stroke, the information for high-risk PCI comes from an FDA randomized study PROTECT II. The STEMI numbers we have come from our pilot randomized study and now in process. And for shock it comes from our registry data and other studies as well, so we stand by our data. We continue to work with the FDA to improve outcomes and innovate. Now what was the first question?

[Chris Cooley:]

Just, I appreciate the additional color there. I just wanted to clarify from Todd, the PROTECT IV and the additional clinical trial ramp up here. How should we think about that expense? I realize it's probably more of the fiscal 2021, but just helping us kind of dial into through remainder of this fiscal year, just in terms of the 4Q if there is a material spend here in 4Q?

[Todd Trapp:]

Yes, I would say the PROTECT IV is more of a fiscal probably towards the end of the fiscal 2021 cost element. As you think about this year in Q4, obviously from a margin standpoint, it's going to - I imagine will be down year-over-year and mostly driven by just the investments we're making in the business. And so, if you think about from an investment standpoint, we have to hire heads that are hitting us in Q4 for what we hired this year. We had the Japan post-approval study that's more back-end loaded, that will impact Q4. STEMI is ramping up over the course of the year and again more back-end loaded this year. And then it’s also that we have the start of the next calendar year, we typically see higher payroll taxes in fringe benefits in Q4. So from a year-over-year perspective, margins in Q4 I do expect them to be down versus what we saw in Q3 as well as year-over-year.

[Chris Cooley:]

Thank you for clarifying that.

[Operator:]

Thank you. Our next question comes from Margaret Kaczor from William Blair. Your line is open.

[Margaret Kaczor:]

Hey, good morning guys. Thanks for taking the questions.

[Michael Minogue:]

Hi, Margaret.

[Margaret Kaczor:]

So what I wanted to follow-up on was the referral channel and this is maybe something you guys don't have quite as much control over is going out and talking to your top accounts and administrators. So whether it's a 60 site road show have you been able to reach out to that referral channel? How are the - top users - the top accounts for you guys thought about how to effectively address that kind of decrease in referral network and just kind of general thoughts on that over the next 12 to 18 months?

[Michael Minogue:]

Thanks Margaret for the question. So it's – the road shows is kind of a three-pronged approach to clarify this data. Number one is, we've already done calls with the key opinion leaders, where we're providing the slides you see on the Internet and more in detail to make sure that they have the information summarized on what are the – what's the safety data available for balloon pump, Impella and ECMO and how does that stack up for vascular - master complications, major bleeding and stroke and Impella does very well there. The second with that information is asking them now either with or without us to work with their referring physicians. So one physician told me his strategy is when he has gotten these calls from referring physicians saying they are concerned about the safety of Impella and they don't want to have a Protected PCI, his standard line is, have you read the paper in whole, and usually the answer is no. And he comments to them well, why don't you read the paper, look at the limitations of it and then call me back if you really are still concerned. And so, those local meetings are important, but what we're also doing is we're going out with some of them and having them present to their community and then as part of that road show we're talking about the clinical data, but we're also talking about the innovation, some of the new things coming, whether it's expandable sheath or single access or SmartAssist. And then last is we're creating this CAMP user base network so that our users can communicate with each other, can share information, can post slides, can post case studies, and we can share the videos with them, both internally and then externally as appropriate. So we are very fortunate to have such a dedicated group of physicians. They are the experts in this space. They are the published leaders and they're the ones treating these patients. So I would say in summary that's the most important thing and investors should take notice of is that the people in this space that are leading the field are using Impella, are publishing their actual data, not taking a billing database and pulling it to take other people's data. So we'll continue to lead on that front and we feel like our physicians have never been so supportive and energized as they are today.

[Margaret Kaczor:]

Okay, that's great. And then just to follow-up and you kind of referenced this a little bit, but I think earlier you said you've been asked and you did may be purchase some of the data from some of the other networks, and so are you going to take that data, review it, and then potentially publish your response or leave it in the clinicians hands and just wait for kind of the single site and RCT data to be published? Thanks.

[Michael Minogue:]

I'm going to do all of the above. So the first thing is, we have data that's already like that. So we have the NIH database that was published in JACC in 2015, the author was stretched. That article showed that when the advent of PVADs hit the market and that's 99% Impella in 2008, it lowered mortality. The paper also showed that when you utilize the balloon pump costs went up 25% absolute costs because many of these patients get escalated and when patients were escalated or used before the PVAD, it was a statistical predictor of death. This is why this paper by eliminating those escalated on a balloon pump took out the sickest and most costly patients and also eliminated all ECMO patients, which many of those survivors because the heart has been loaded require more invasive and more expensive procedures like LVADs and then transplant. So that's a very important component, but then with the premier database itself we will go through it. We will put everything back into it with the NCDR data, we're going to run the numbers from 2015 and 2019. Some of those patients will be in IQ. Some of them will be in IQ and in our post-approval studies. We also are pulling the data from these six centers that have very limited use of Impella. So in 2018 the authors, the hospitals, they did 1,400 ECMO patients. They did 1000 plus balloon pump patients and they did 221 Impella patients. In those patients at those centers, about half of the time they do not follow the protocol. So Impella patients that get Impella come usually with a balloon pump before or many of our patients transferred in on Impella to their hospitals, Impella is immediately removed and switched to ECMO. So we'll be continuing to publish data from our real world evidence and will complement it with the other publications that are out there.

[Margaret Kaczor:]

Great, thank you, guys.

[Todd Trapp:]

Thanks Margaret.

[Operator:]

Thank you. And that does conclude our question-and-answer session for today's conference and I'd like to turn the conference back over to Mike Minogue for any closing remarks.

[Michael Minogue:]

Well, I want to end with thanking our investors for their support. This is part of our journey to become the standard of care, and it's probably the best opportunity we've had in a long time to push through to get the guidelines required so that Impella can be worldwide a Class 1 guideline. If you have any follow-up questions, please reach out to us, and we thank you for your time today.

[Operator:]

Ladies and gentlemen, thank you for participating in today's conference. This does conclude the program and you may all disconnect. Everyone, have a wonderful day.