Hey, thanks so much for joining us. My name's Matt O'Brien. I'm one of the med tech analysts here at Piper. We're very lucky and excited to have the HeartFlow team with us. You've got John, who's the CEO. Vikram, who is the CFO. Nick's out in the audience. We're gonna reserve all the questions for you a little later. But yes, really do appreciate you guys coming all the way across country.
Yeah.
And yes, Vikram, that was a terrible joke, so I'll stick with my day job. But maybe start a little bit, you know, for those in the room or on the webcast that are a little bit newer to the HeartFlow story. You're just recently went public. What are you doing? What are you, you know, trying to solve for, and where are you going?
Yeah. So, first off, thanks for having us. Great to be here, and I appreciate everybody's time. So HeartFlow, we're a software company. We use AI in everything we do, and we diagnose coronary artery disease. And that's. We have an AI platform across the full continuum of care for coronary artery disease. And we're driving adoption of that, and it's our goal to create a new standard of care in that regard. Obviously, Healthcare Conference, I don't need to belabor the issue that heart disease is, but number one killer. It kills unfortunately more people than all cancers combined. Every 40 seconds, somebody in this country's having a heart attack. So there's a huge, huge problem to solve, and we think our technology is really the next wave of, you know, what can be better medicine to help address this.
We find disease, and the current standard of care does not do this, but we find disease down to the cubic millimeter. So we can diagnose if somebody in here gets a HeartFlow scan, we can tell you how many cubic millimeters of atherosclerosis or plaque that you have. We can tell you what type: calcified, non-calcified, low attenuation. And we can model blood flow across that disease or that stenosis to tell you, a very separate insight, which is how significant is it, and do you need to be, intervened on. This is all part of our software platform. We have workflow tools that improve efficiency. We have a Cath Lab tool that we're excited to launch in 2026. But at the end of the day, we're here to create a new standard of care. A little bit of history.
HeartFlow was founded out of Stanford University in 2010. Our first FDA clearance came in 2014. We've invested, since the get-go, and we continue to invest in clinical data. We believe in prospectively validated, published clinical data. We think when you're working with cardiologists, that's really the currency of the realm, and you need that in order to really change practice patterns. We have over 600 peer-reviewed journals to date, and we're continuing to invest. We've treated over 500,000 patients commercially since inception. If you include clinical, the number's quite a bit more. At our core, we have a very large data set. So we get every coronary CTA that a customer does, and we now have over 110 million annotated CTA images, okay? We use that database to expand gross margins, and I'm sure you'll, you might have a question on that.
We can explain how that's working. We use that database to introduce new technologies. We're super excited about, and like I said, like you know we did the IPO in August, so our balance sheet is stronger than ever, and we're gonna continue to invest to create a new standard of care.
Got it. Got it. Very helpful there. So, and I really hate to talk about this topic first. Let's just get it out of the way. You know, the DOJ investigation was announced, you know, in your Q on the Q3 update. You had a great quarter, but that got a lot of attention. Can you maybe just talk a little bit about what you saw there? The language in the 10-K was pretty vague. What are they looking for, and have you started giving them documentation?
Yeah. So, yeah, that was a surprise, obviously, and I think it's a shame that it got a little bit of a, you know, a headline relative to what I thought was a really great quarter and really, really great work by the team executing. Listen, we're super early in this process. This is a request for information. Right now, we've started providing the information request. It's a very broad information request, so for me to speculate on exactly what they might or might not be looking for, I don't think would be wise. I personally don't know. I don't wanna speculate. What I can say is, as a company, and certainly I hold myself to a high regard on this, I believe HeartFlow has a culture of integrity and compliance. And I personally believe that I do. I know our leadership team does.
I don't see this as we fully cooperate, and we are, and we will fully cooperate in any way slowing us down, okay? We have rigorous compliance procedures, systems, processes in place. Because of that, I have really high confidence in how we go to market, and I do not see this slowing us down in how we go to market.
Got it. Can you just talk? Just last question on this?
Yeah.
On the compliance side, when you talk about this rigorous compliance process, I work at an investment bank. I'm very, very aware of the compliance side of things. What does yours look like, and why the confidence in what you have in place there?
Well, anything we do is vetted, first and foremost, and there's trackability towards that vetting processes. Obviously, we've been through an extremely thorough vetting process through the S-1 process. We brought on a new general counsel close to two years ago, independent of anything IPO-related. She doesn't take anything at face value. She did her own, you know, audits, so to speak. And then with that, we have trackability, systems, training. So that's kind of like the wiring of the organization. But more importantly, I think we have a culture, okay? If you work at HeartFlow, you're here to help patients. Right now, there's better patient care available with HeartFlow technology, and I think that motivates the people that join this company. I think it motivates certainly the customers that we partner with.
And so from a culture standpoint, this is not a space where you're even close to these risks, you know? And I know other parts of med tech, sometimes you are based on misaligned incentives or whatever. We're not there. I mean, we're in a different playing field.
Yeah. That's what makes the investigation so strange to me, but who knows?
Well, I'll tell you what I've learned. I mean, obviously, I'm a first-time, you know, public company CEO. I've learned information requests, you know, are fairly common.
Yep.
And that's the phase we're in now, and we'll let the process play out. And, you know, like I said, we're fully cooperating.
Got it. Okay. So let's move on from that and get to the business itself. You know, the market opportunity for HeartFlow, about 900,000 CCTA images done today. You're, you know, just over 13% penetrated of those cases. So, you know, how many of those are you really indicated for? And then where do you think that 13% number can go over the next several years?
Yeah. So, first thing, I think the first number you quoted was like 900,000. Is that right?
That's right.
As total CTAs and just so everybody knows, our software, and I should have said this upfront, our software sits on top of a coronary CTA, okay? Coronary CTA, I think that's a 2024 number, right? 2023 number. Okay.
That's a 2024 number.
23 number, yeah.
Yeah.
So that number in and of itself is growing, okay? As coronary CTAs are in the guidelines, a level 1A test ahead of the standard of care, historically, it's grown about 22%. I believe that's a durable trend. So the pie's gonna continue to get bigger, and we like to say at HeartFlow, we're on the right side of history here. Okay. Now, when it comes to our reimbursed technology, 'cause we've got other technologies that aren't reimbursed, but reimbursed technology, our core, FFRCT technology applies to 33%, okay? And the reason why is it's paid for and covered if the stenosis is between 40% and 90%, okay? So that's a 33% population. Our Plaque Analysis, which we're in the very early innings of launching here, applies to a much bigger, okay? That's any visible plaque up until 70% stenosis, that's 60%.
So you put them together, you got 93, okay? And hopefully, over time, we can drive full utilization across all that.
Got it. Okay. So the CCTA market's growing 22%. You're growing a little bit faster than that because you're getting more adoption among all these centers.
Yeah. Yep. Yep.
Okay. What's the gating factor for a lot of these centers to go ahead and use you guys? I mean, what, why wouldn't they want to use HeartFlow?
Yeah. So our commercial model, we go to centers, hospitals, and clinics that are already up and running with active coronary CTA programs. We've kinda got lines of sight to other centers that might be starting one, but we're not selling into them right now. Once we're there, we get pretty good clinical buy-in based on the data we have. The issue we have is this is a platform that works across a health system, and there's multiple stakeholders. I think everybody knows, healthcare is slow to move and slow to evolve. So aligning all these different stakeholders to a new diagnostic pathway, which another way of saying that, that's a new diagnostic workflow, takes time.
Yep.
Okay? And what we see is once we're in, the growth happens, but it is not as simple as we have a new widget, and we're swapping out the inventory, and then you see the adoption the next quarter. This is a lot more change, and dominoes need to line up.
Got it. Okay. Okay. Appreciate that. What about the Salesforce expansion? You executed that earlier this year. How many reps did you add, and where are you really focusing these reps?
Yeah. Good. So we've got two arms of the sales force, and this is really kind of a land-and-expand type model. One arm lands the new accounts, okay? So there's a hunting ground of about 2,900 accounts out there with active coronary CTA programs. This arm of the sales force, we call them territory sales managers. We've got about 46, 47 of them. And they're in charge of hunting and getting the contracts signed, okay? We have a second arm, and this is the arm that we've expanded. We call them territory account managers, and they work out in the community really to raise awareness of a CT plus HeartFlow pathway. They drive a lot of medical education. They're targeting general cardiologists, okay? We're not in a world where we're going after primary care or whatever.
And we started the year with about 45, and we're gonna end the year with 95.
Got it. Got it. Okay. What kind of productivity can you expect out of this group, you know, and how quickly can they ramp?
You know, over time, I think, you know, a year, year and a half, we see pretty good productivity. Obviously, there's a learning curve.
Yep.
For the hunters, this is a longer sales cycle. Converting a new account can take upwards of 12 months by the time you work across all the various stakeholders in a health system. For the folks out in the community, that's much more like a pharma model, so it's a lot higher touch, a lot more reach and frequency. And we do our best to train them and get them as productive as fast as we can.
Got it. Okay. All right. Appreciate that. Vikram, I swear I'll get to you in a few minutes. You guys j ust bear with me. But the other thing that was, you know, was great last year, John, was you guys added 300 centers. That was up about 20% year over year. I think, you know, you're on pace to do a similar number this year. You know, how do we think about, you know, that's. I'm not quite sure if that's about a doubling of the number of centers that you had, you know, before.
Not quite.
Okay. But how quickly can they ramp up?
So I think there's probably two parts to that question. So one, we started this year, 2025, with a little over 1,100 centers in our installed base, okay? And the 300 that we added was to get us to that 1,100 in Q1 of this year. So coming up, I'll give guidance on how this year was, but it's safe to say, you know, publicly, it's been a great year.
Yep.
So this model that we have of hunting new accounts and bringing them into the fold, so to speak, I think is as strong as ever, you know? And with plaque coming on board, demand is higher than ever. The CT guidelines took effect in 2022, so now we're kind of, what, 2023, 2024, 2025. You know, we got some momentum on guidelines.
Yep.
We're launching our PCI Navigator technology for to round out our platform. There's more demand there. So I'm super confident that this model of landing new accounts and bringing them in is clicking on all cylinders. Now, the hunting ground starting in 2025 will be you gotta take that 2,900 number.
Yep.
And that gets you to, what, 3,200. And then you get about 300 new accounts that are gonna come on board during the course of the year. So there's plenty of.
Yep.
Hunting ground, still to have.
'Cause you're not even halfway penetrated.
Yeah. Not even halfway.
Okay. And do you get to the point where, I mean, it's kinda interesting that you added all these account managers to really drive utilization at a time where you've seen a big increase in the number of hospitals that you're working with?
Mm-hmm.
Is that how we think about the next big? I mean, it seems like the next big leg of growth for you guys is not necessarily just adding centers, but it's really going much deeper in existing ones.
Yeah. I mean, the thing to know is when we add a center, we set up bidirectional data pipes. So we take all the data coming out of a Coronary CTA, we take it into our cloud, and that's where all the analysis comes in. So as we launch new products like plaque, we're leveraging that same infrastructure. And it's literally just a new button on the existing user interface. It's to the same customer. It's the same sales rep. We're dropping it in the bag. So there's lots of efficiencies there. So as plaque comes online, and we're not forecasting anything material on plaque until tail end of 2026, it's gonna come online in a really efficient way.
Mm-hmm.
The utilization of HeartFlow technology in any of our existing accounts is gonna be, you know, quite attractive.
Okay. So I was gonna go into reimbursement, but let's hit plaque, for now. Maybe just talk about the utility of plaque versus FFRCT.
Yeah. Yeah. So again, we wanna provide clinically validated AI insights to our physician across the full continuum of coronary artery disease management, okay? Plaque answers a fundamentally different question than FFRCT. FFRCT is, does this specific lesion that's identified need to be reintervened, okay? And if you know that, you know, precisely, if you're an interventional cardiologist, that any patient that comes into my cath lab is gonna need an intervention, not a diagnosis, and you know exactly where you need to treat, okay? So that's FFRCT. And like I said, that's a separate revenue stream for us and a separate reimbursement revenue stream for our customers. Plaque answers a fundamentally different question. This is, irrespective of any disease that needs to be treated with a stent, how do I medically manage a patient, okay?
We really believe over time, coronary artery disease is gonna be managed very analogous to how cancer is getting managed. We have a staging system where we can tell patients what stage they're sitting in, you know, mild, moderate, severe, and extensive. Based on that, we know the treatment algorithm for all these great new therapeutics that are being put into the market that best apply for that for that risk level, okay? So it's a medical management decision that's gonna allow cardiologists to hopefully find disease earlier and manage it from getting worse, manage it so it doesn't progress.
Got it. And that opens up that much bigger patient population you're working with.
Yeah, and that's, you know, that's 60% of the symptomatic patients. We're only talking symptomatic at this point. When we talk asymptomatic, we're orders of magnitude bigger.
Sure.
Yeah.
Sure. Got it. Okay. I think you said on the call you're about 57% covered lives at this point for plaque.
Yep.
It seems like you need one more big payer. What do their cycles look like in terms of when they're determining coverage on this, and could we get an update, you know, or first half of next year?
Yeah. Good, good question. I was gonna ask you the same one. So yeah, we have 57% coverage right now. We got United. This is for plaque. We got United, and we got Cigna effective October 1st. So as we thought about kinda where we were gonna land the year, we're pretty close. We're a tick or two above that than our initial assessments. The next two big ones to drop are Aetna and Anthem, okay? We've engaged with them. We're a part of their process. You know, they ask for clinical dossiers, etc. We have physicians speak on our behalf, so there's a whole process that goes into it. Unfortunately, it's a little bit of a one-sided process. You know, payers don't exactly tell you what they're thinking, or they don't really tell you when to expect it. So it, I can't call it.
I am confident we're gonna get coverage on plaque. Hopefully it happens as soon as we can, but I can't call exactly when it's gonna hit. Now, if I could, one other element of plaque coverage is really important.
Yep.
Okay? And we based on our FFR journey, we learned you kinda gotta get to that 70% neighborhood before you see real adoption. The equally important piece is medical education, okay? And with plaque, when I talk to physicians, the interest level is off the charts, okay? But the very first question I get is, what do I do with this information, okay? That's a different medical education effort than we undertook with FFRCT. FFR, their FFR value existed in the marketplace prior to us. There was just an invasive FFR. So once we proved accuracy with our algorithm, everybody knew what to do with a 0.64 FFR value. You didn't have to educate them. Plaque's a different game. Nobody knows what to do with 278 cubic millimeters of plaque if you're 44 years old.
We need to educate them, which is why we've invested in the DECIDE Registry, which is the largest prospective registry of its kind, to educate physicians on how to use this tool. We risk stratify, as I mentioned. We've got the largest risk cohort, database out there, and then we've paired it with this medical management framework, so based on whatever stage you're in, you know what, corresponding FFR cocktail, so to speak, to be treated with. All of that, as you can tell by the length of that answer, is a lot more nuanced than, you know, 0.64 bad, 0.82 good.
Yeah.
Okay.
The only thing I would add to that is, even though we're not able to pinpoint when exactly, we're confident we'll close out 2026 with 70%.
Yeah. Absolutely.
coverage.
Okay.
On the commercial side.
Okay.
That we've highlighted John's points as the trigger for more widespread adoption in conjunction with our medical education efforts.
Okay. I mean, I'm just what I'm trying to make sure I understand, Vikram or John, is that you're still super bullish about everything. I'm just making sure there's not this extra layer we hadn't kinda contemplated of, oh, hey, don't expect it to go quick once we get to 70 because we have to educate. You still think it'll go fairly quickly even as you're educating and as. Yeah. What I'm saying is, well, first off, we're not waiting to get to 70 to start the education journey. Okay.
What I'm saying is, our guide is not to expect any material plaque contribution until the tail end of 2026. That's, that's one message that needs to be delivered. Now, the drivers of that are getting to 70% and achieving medical education. Now, the medical education is not a quantitative number that's, you know, everybody's gonna be on their own journey there. And so the reason why I raise it is it's a fundamentally different journey than the FFRCT journey.
Got it. Okay. Makes sense. All right. So Vikram, I told you.
Yeah.
Thanks for being patient. You know, few questions for you. The plaque reimbursements.
Yep.
Came in at $950 was really good. We're all modeling, you know, ASP of, you know, $350 for you guys on the plaque side. You get a lot more of the economics on the FFRCT side. Why wouldn't you get more of the economics on the plaque side?
Yeah. It's a good question. The ASP is really a reflection of reimbursement dynamics. The way the customers look at this is not on a per-patient basis. They look at it as a diagnostic pathway. What is the weighted average of reimbursement across the pathway? And so that's an important nuance. You know, and we think, you know, until we get broad coverage, commercial coverage, reimbursement is gonna be relatively low. And we've priced plaque appropriately to really, you know, endorse the adoption of the product, the service lines. Longer term, however, as coverage ramps, we'll have the flexibility to adjust pricing upwards.
Okay. Will you have the room to take it up if you set it at a certain level? And, you know, have you done that in the past with FFRCT?
Over time, I think there's absolutely the flexibility to adjust pricing upwards.
Got it. Okay. Then, Vikram, I also wanted to talk about your Q4 guide. It's viewed by everybody as super conservative, but you know, given the big Q3 beat, that sequential step-up was a lot higher than normal, so historically, you know, Q4 sequentially is up kinda the mid-single-digit range, but that's gonna be really difficult to do, right, because of that bigger sequential step-up in Q3 that you saw, so should we not expect, you know, an acceleration, you know, in Q4? I mean, still good performance, but you know, how do we kinda think about the, you know, the cadence that we're on the top line?
Yeah. Good question. So I'll start by saying Q3 was a really strong quarter for us. We saw broad-based growth, really meaningful expansion of the install base as well as sequential, you know, growth in, in utilization in both new and existing sites. We'd also called out particular strengths in the clinic side of service. So the sales team's been, you know, firing on all cylinders, and we're very pleased with the growth we saw. Relative to Q4, you know, it's, you know, it starts with our guidance philosophy. It's anchored on a high conviction forecast, you know, that sets a reasonable baseline upon which we can overdeliver. The strength in Q3 obviously makes the sequential comps a bit tougher in Q4.
So we felt it was prudent to, you know, embed a measure of conservatism in our guide, especially since it's our, you know, as a newly public company, right?
Yeah.
You know, and then, moreover, relative to, you know, what your comments about sequential growth in the second half versus the first half, that's a dynamic we've seen historically. So those are, you know, directionally make sense, and that was what was factored into our guide.
Got it. Makes sense. Okay. So still very, you know, very, confident in the outlook for the business. Still really good growth, but just, you know, don't, don't forget about the cadence.
Absolutely.
Okay. And then, on the gross margin side, you know, it was kinda eye-popping, the Q3 gross margin performance. What drove that? And then I think the street's modeling things flat in Q4. Why would that be the case in a, you know, as you bump up a little?
Maybe, I'll let Vikram speak to the specifics, but I probably should've mentioned this earlier on. We are a case study of using AI to take costs out of gross margin. And we have a long history of doing that. The reason we can do that is we've got this database of 110 million annotated CT images, and we train new algorithms. And by virtue of doing that, the humans that we have in the loop, we have quality control humans, spend less time, over time. So that's one really important long-term driver of gross margin expansion. The other two, of course, are volume, right? As we grow, volume helps. And then the third is plaque.
Yep.
As plaque comes on, that's incremental, sitting on the same algorithm we're already doing. So I just wanna set the stage on that. From a bullishness perspective, I really like our long-term gross margin opportunity, but I'll let Vikram speak to this.
Yeah. So long-term, you know, we're confident that gross margin's gonna be the hallmark of our financial story. In Q3 specifically, you know, plaque revenues were immaterial. So it was really driven by volumes as well as the automation of the algorithm. To your question on Q4, again, you know, we've based on the midpoint of the revenue guide, you know, we've assumed sort of similar gross margins to Q3, similar revenue, similar cost of revenue, that gets you to flat gross margins in Q4.
Got it. Okay. And I'll just sneak in one more real quick. If plaque outperforms over the next several years, how much of that are you gonna let fall to the bottom line versus reinvest in asymptomatic or other areas?
Yeah. Great, great question. You know, I think we could thread that needle, and do both, you know, provide meaningful operating leverage over time as well as reinvest in some of these growth vectors we talked about. You know, longer term, I'd say we're highly confident this is a business with strong earnings potential, the durable revenue growth, consistent margin expansion, a commercial model that's highly scalable that lends itself to strong cash flow generation over time. In the near term, I think we'd have a bias to reinvesting some of the plaque upside back into those growth-oriented areas, automation, platform expansion, or selective investments in sales and marketing, but longer term, you know, we're confident we can reinvest into the business, but we're also reinforced in our conviction that we are on this glide path to profitability.
We've talked about this three-year timeframe from IPO, and the most recent quarters give us a conviction around that forecast.
Understood. Okay. I think we're out of time. So I'll have to end it there. John, Vikram, thank you so much. Appreciate it.