Joining me today is GeneDx. I'm the tools and diagnostics analyst, David Westenberg. Katherine Stueland, CEO, and Kevin Feeley, CFO. Sorry about that, so we'll just start off today by bringing you guys up a little bit. I'm pretty sure you are the best stock to date in diagnostics. Before I make it, why were investors so concerned about you that made your stock so well? Then what steps did you take to improve it?
Oh, thank you for having us. We're thrilled to be able to share our story, which today is a story of market leadership in terms of being able to diagnose anyone with a genetic disease as early as possible. Our focus today, of course, is in the pediatric segment. But the journey that we've been on to get here certainly has been one for us as a company and for investors as well. The company itself, GeneDx, has been around for 24 years. It was founded out of the National Institutes of Health. And the team that preceded me was really prescient in their thinking of how do we ensure that we can diagnose any genetic disease off of a genome backbone. That was really the original goal. And started developing that technology well in advance of the market being there.
But I joined the company three and a half years ago, really to establish a strategy for growth, to be able to deploy our industry-leading exome and genome to as many patients as possible, and to really establish a strategy for scaling the business in order to support diagnosing more patients cost-effectively. We were acquired in 2022 by Sema4, which had many different business lines that were established. So they had a cancer business line, a reproductive health business line. And with the acquisition of GeneDx, it was intended to provide that genome backbone. And I became CEO of the combined companies. And during that time, I had to wind down the Sema4 business. Capital markets changed, and investors really started to demand profitability versus growth at all costs. And so it took us some time to really restructure that business, wind it down.
And in Q4 of last year, we were able to deliver our first quarter that I think was emblematic of GeneDx and the strength of that underlying business unencumbered by the restructuring costs. So I think what we've seen by way of the stock reaction has been investors who have supported us through that restructuring are now being rewarded with a really strong foundation for growth. We think that there is sustainable growth in terms of being able to drive utilization of our testing. That's driving better clinical outcomes for children with rare diseases. And it's also driving a stronger P&L for us as a business.
It's interesting because coming from there, I mean, still when you cannot cut your way into growth, I mean, that's been like one of those principles that I've been told. It really is reality that cutting costs and still seeing no growth impact is almost impossible. It seems to have worked in your business. I still will doubt anyone who says to me, "We're going to cut and grow the same amount." What do you think actually went right to break that rule?
So, I think that we were very mindful of where we cut, and GeneDx had about 1,000 tests on our test menu, and the goal really was to get to and is to get to a place where we can run any number of genetic tests off of an exome and genome. We've been able to reduce that test menu by about 70% in the past year, and we were mindful not to cut any costs that would denigrate the quality of that backbone exome and genome, and so we were very mindful that those were the biggest levers that we had in terms of being able to, again, provide the very best clinical care through our exome and genome testing without in any way denigrating quality.
So last year, we organized all efforts in the company around growing exome and genome, around getting paid more and more regularly for our testing, so improving our average reimbursement rate, and really cutting back on key efforts. We did cut back in marketing. We did cut back on some R&D. But we wanted to make sure that the strongest levers for the business were well capitalized to ensure that we could arrive in this position today.
Gotcha. OK. And then you have a WGS ticker. So how do you use sequencing in terms of lower cost and faster output? And can you remind us what percentage of the tests are microarrays, targeted exome, and genome, and kind of what the areas of growth are of each of those?
Certainly. So the vast majority of our testing is whole exome sequencing, WES. But we do have an increasing focus on WGS, whole genome sequencing. So our entire team is focused on growing utilization of whole exome and whole genome sequencing. But there are important growth factors associated with chromosomal microarrays. That remains in guidelines for general pediatricians to be what guidelines would point clinicians to utilize. We see over time that being a really important catalyst for us. Once AAP guidelines flip to say that you can use an exome or a genome upfront, we'll be able to take that CMA business that shows up in our other testing line and flip that to exome and genome sequencing. And the gross margins on exome and genome combined are north of 70%, so a highly leveraged technology that we have. Do you want to add?
Yeah, the only thing I'd add, so in third quarter, 34% of all tests were exome and genome. Underlying that within the non-exome portfolios, Katherine said, we moved from 1,000 tests on our menu down to about 400, really going through a critical lens to say, are the tests gross margin positive and accretive to gross profit? And then also, what remains? Are they strategically important to growing the exome and genome business over time? Over time, it is our intention to cannibalize the rest of that test menu as well as the industry's test menu into whole exome and whole genome. 34% of all volume was exome and genome, but it represented north of 70% of all revenue and north of 90% of our gross profit. And so you should expect to see from us continued mix share gains into exome and genome over time.
Great. And you recently had the GUARDIAN study. What did you learn from that? And can this help open up a new market of all newborns? And maybe that's the dream, but what's the time frame to that happen?
So that is absolutely our goal, is to drive the entire market to get to a place where we can sequence every newborn at birth and get ahead of disease progressing. So the GUARDIAN study was designed to take a look at healthy newborns here in New York State. We're doing it with Columbia. We're doing it with the New York State Department of Health. And what really sets Guardian apart are a number of things. One, we were able to show that there is a viable market for this. So more than 70% of parents enrolled in the study. So that tells us that it's not just a dream. It's something that we know is something that will be in demand as we think about the future. Second, we were able to find a 4% positive rate in babies who were otherwise considered to be healthy.
What that means is we were able to actually diagnose disease before symptoms started. We screened off of a whole genome sequence. We screened for, at first, 200 clinically actionable conditions and increased that to 450 clinically actionable conditions. I think importantly, part of what we're proving through the Guardian study is that there is an ethical and responsible way to be able to deploy this sort of technology. Other studies that had been done previously, frankly, just put in the hands of parents some really scary risk-related information. No parent needs to know that their child may be at a higher risk of prostate cancer. We really took a careful lens to how do you do this in an ethical and responsible way. 100% of the conditions that we screened for were clinically actionable.
Finally, I think the other really important element of this is we took a look in GeneDx's database retrospectively to say, what was the average age of diagnosis for these several hundred conditions? And we were able to determine that it was eight years old. So all of these conditions would have progressed over the course of, on average, eight years. You would have had the clinical progression of disease, and you would have had enormous costs to the health care system and to these families that contributed over the course of eight years. Now what we're proving through Guardian is you can do that at birth, get ahead of disease progression, potentially stop disease progression, and save the health care system valuable dollars.
I think as we move forward, while we don't think that this is going to be a market that becomes a reality in the next five years necessarily, we have the framework that we need to bring it to policymakers. Ultimately, being able to show that there is a cost-effective way to sequence babies at birth and get ahead of disease progression, that plays very favorably in terms of the environment moving forward.
Gotcha. So I want to talk about maybe some of the ways to increase diagnostic yield through data. And it was interesting. I was at Stanford this summer. And just a couple of different things that we noticed. When we walked into the lab, there was a box that said NVIDIA that took the entire length of the overall lab. And then a second thing we learned was that just the acceleration in terms of data processing was reducing the bottleneck from days to hours. So as we're thinking about data and AI learning in rare disease, do you think there actually could be an acceleration of data and knowledge of new variants where we could see a diagnostic yield increase quite significantly?
Can you just remind us, because you're going to be, if there is this really rapid progress, patient in time, 2024, December 3rd, maybe does not have a disease? But if that is discovered nine months from now, what happens?
So one, AI is going to be beneficial to our business and to the entire ecosystem in terms of, yes, being able to more efficiently and rapidly help us upgrade and downgrade variants of unknown significance. It's going to help us with our overall workflow to make it as efficient as possible. But then when we think about our data assets, so we have a data asset that is comprised of more than 700,000 clinical exomes. It's the largest in the United States. How do we take that and put it to work for more biopharma companies? And I think this is where AI can help really shift from a failure rate of 90% in R&D in the pharma industry to be able to have more accurate target discovery, faster clinical trial recruitment and development, faster time in front of FDA, faster time to get to more and more patients.
And so we think that there's a tremendous opportunity. In terms of the ability to retrospectively go back once we have identified or upgraded or downgraded a variant of unknown significance, we do that today. We do it in a way that I think AI can enable. So it is more automated, and it's happening more rapidly. But for every patient that we are testing, we're actually generating more genotypic information. We're generating more phenotypic information that is helping us today upgrade and downgrade variants of unknown significance, which then ensures that for the next patient that we're testing, there's an automatic product improvement, and it becomes this virtuous cycle. So AI can certainly help automate and accelerate that.
But absolutely, for patients today and patients in the future, that retrospective ability to be able to provide information on demand as we're making new gene disease discoveries and developments, that will become an automated part of our service.
Gotcha. Thank you. So can you just give us the size of the rare disease market? What's it growing at? And what are the barriers to entry?
Absolutely. I think one of the misperceptions of rare diseases is that they're rare. One in 10 Americans has a rare disease. Half of them are children, and so we remain in the very early stages of being able to diagnose the total number of patients, and what we're finding is, as we open up access to our testing, that these conditions are actually far more prevalent than what people had previously estimated. That's true of all hereditary disease, as we've been able to open up access to testing. It most certainly is true as it pertains to rare diseases as well, so as we continue to make really important discoveries along the way, for example, if you look at epilepsy, there's 768 genes associated with epilepsy. The importance of using a whole exome or a whole genome upfront is the ability to interrogate the totality of the genome.
As we continue to find and upgrade more variants of unknown significance and learn that there are more than 768 genes associated with epilepsy, we can then go back and be able to provide updated information to patients, so our technology makes multi-gene panels obsolete. It ensures that it's the most efficient and cost-effective way of being able to diagnose disease, and it ensures that we are continuing to benefit future patients from current patients that we're testing and analyzing today.
Gotcha. So you have high growth, reimbursement's good, and there's not really a lot of competitors when I think about rare disease focus. So obviously, we're going to have some entrants. What do you think these entrants are going to look like? Do you think it's going to be more a national reference lab? Or do you see more esoteric companies with a specific focus on rare disease?
80% of clinical exomes run through GeneDx today. Other companies have tried to enter the exome and the genome space. We welcome the competition. We think the competition is really healthy and that it's important in terms of influencing policy, both state-by-state policy, as we bring on more states through Medicaid, medical policy through private payers. Competition overall is a really good thing. What it takes to actually reach our standard of quality and accuracy, that's a totally different ball game in terms of somebody being able to displace us. The 700,000 clinical exomes that we've run, they are all enriched for rare disease. They are complemented by more than 5 million phenotypic data points.
So if you think about all of the clinical notes that we are getting and annotating, that plays a really important role in complementing the genotypic information that we're getting. We are running 60% of the time comparator samples with mom and dad. And because we have invested in Medicaid populations, we have the most diverse data set that exists in rare disease. In order to recreate that, to reach our standard, and our standard is the one that clinicians have relied on for decades now, it would require a heavy investment. And it's not something you can turn on overnight. It will require actually recreating the quality and specificity of that data asset. So it takes a really long time, in addition to the capital, to do that.
That, on top of the fact that we have a really fantastic commercial engine that we have built that is ensuring that more and more clinicians understand exactly why GeneDx has been the one that the most discerning genetics experts have relied on over time, that will ensure that we will continue to be the leader, both in exome and as we enter a genome-forward market. But again, we have, of course, welcome competition to ensure that we can expand utilization for the growing number of families who would benefit from this information.
I haven't given enough to Kevin here. So just in terms of direction of your reimbursement, has it been up or down? Could there be cuts to CMS? Kind of, what's your covered lives? And then can you give us a breakdown on how much is Medicaid versus, because I'm assuming it'd be more Medicaid than Medicare. I hope no newborns on Medicare. And then versus private insurance.
Yeah. About 10% of the business today is institutional pay, so a contractor rate direct from a hospital system. What that leaves is 90% of all volume today being submitted through insurance channels, with a roughly 50/50 split there between commercial insurers and Medicaid. So north of 40% of all volume submitted to Medicaid programs. We're now up to 30 states that cover exome in the outpatient setting, 14 covering our use of a rapid genome. Inpatients, if you were to go back just five years, it was zero across those settings. And so great progress in more and more states recognizing the clinical and economic benefit of exome and genome and picking up coverage. And we therefore have seen the denial rates in those populations come down over time. On the commercial side, we're in network and contracted with over 80% of commercial lives, all large national payers.
Most have some level of policy coverage for exome and genome. If you take the totality of all volume submitted to insurance across both commercial and Medicaid, we now see a roughly 55% payment rate, so being denied 45% of the time. We obviously believe that there is room to further improve that by reducing denials over time. In fact, in the last 18 months, we've radically reduced that denial rate. We were being paid only 35% of the time across insurance channels in the second quarter of 2023, really spending a lot of time to design bespoke, customized order capture claim submission processes for commercial payers to avoid denials. And on the Medicaid side, benefiting as we've ramped up to that 30 states and 14 that have policy coverage. And we expect that to continue for some time to come.
Thinking of that on the ASP front, from a COGS standpoint, how do you improve it? Is this just purely same volume going through or more volume going through kind of the same fixed cost structure? This new sequencers, lab automation, simplified library prep, and then essentially, where do you see peak gross margins?
Yeah. So exome and genome today operating mid-70% gross margins. In the third quarter, we posted an average reimbursement rate after all of those denials of $3,100. Believe there's room to reduce denials to improve that rate. At the same time, really proud of the overall COGS stack. About a third of all costs are input cost reagents. And so we've benefited from some price rationalization as manufacturers are starting to see more competition in their space. We think that'll continue. A lot of automation robotics in the laboratory. The labs went through a full retrofit during the COVID years. Feel fairly optimized from a process and a cost perspective across the wet lab. About a third of overall COGS today still reside in those dry side steps after the sequencer, but before report writing.
And we believe those steps, which are fairly manual today, will benefit over the next couple of years as we deploy further automation, AI across the spectrum. Certainly believe there's room over the next several quarters to reduce COGS and therefore see expansion of gross margin.
Gotcha. I want to go back to the ASP side here. What was that? 45% denial rate? Wait, that was what? 45? Sorry. What was that again?
Today, being paid 55% of the time.
55%.
If we go back 18 months, we were being paid 35%.
Yeah. So 45% denial rate. Is there a committee, a guideline committee that could help open up the spigot of payer coverage to maybe accelerate that? I mean, 1,000 basis points of improvement is pretty good, but yeah.
What have you done for me lately?
Yeah.
The largest, which would radically grow the addressable market today, would be the American Academy of Pediatrics, AAP. Last issued their genetic testing guidelines for pediatricians in 2014. So they're 10 years old and would be due for a refresh. And in fact, those guidelines 10 years ago called for the use of exome and genome as a test of last resort because of realities at the time a decade ago that these tests would take months, that they would cost tens of thousands of dollars, and be too complex to understand for a day-to-day pediatrician. That was true 10 years ago, and it's precisely what GeneDx has solved for over the last decade. And so certainly, we believe those guidelines at some point should be changed in favor of an exome-genome-first approach.
And that would be a major catalyst for us if and when it comes, and we look forward to that day.
All right. One second. Not enough time to get anything in there. Thank you so much for coming.
Thank you.
Thank you.
We appreciate it.