Today, we're fortunate to be joined by Daniel S. Grosu, MD, MBA, Founder, President and CEO of PetDx, a liquid biopsy company that intends to revolutionize how we detect cancer in man’s best friend — canines. Daniel brings an impressive background in medicine, healthcare, and biotechnology, including an MD with Distinction in Research from Saint Louis University, an MBA from Oxford, and past roles as Chief Medical Officer (CMO) at Illumina and at Sequenom. Daniel founded PetDx in 2019, and the company has secured $72M in funding to date (including a $62M Series B round in late 2021) to support the development and wide-scale commercialization of OncoK9, a liquid biopsy test for dogs that can detect 30 types of canine cancer. PetDx released clinical validation results for OncoK9 on April 26th, 2022 from their CANDiD (CANcer Detection in Dogs) study — the largest clinical validation study in veterinary cancer diagnostics to date — as a peer-reviewed publication in PLOS ONE.
To start, it would be great to dive into the founding of PetDx. We understand that PetDx came about when you recognized the difference in cancer treatment between humans and animals when both your parents and your pet dog, Poppy, were diagnosed with cancer. Can you tell us more about how you leveraged your experience at Illumina to create PetDx?
Absolutely. It's a pleasure to speak with DeciBio. Thank you for allowing us to get the story out.
I spent my entire career in the human diagnostics space with companies like Siemens Medical, Bayer Pharmaceuticals, and Johnson & Johnson. About 11 years ago, I came to San Diego to be the first Chief Medical Officer of Illumina. After Illumina, I was the last Chief Medical Officer of Sequenom before it was acquired by LabCorp. At Sequenom, which was a pioneer in noninvasive prenatal testing (NIPT), we also did a lot of liquid biopsy work for cancer because many of the principles are the same. You work with cell-free DNA (cfDNA) from a blood sample.
After Sequenom was acquired, I took a break from the corporate world, took early retirement, and reinvented myself as an angel investor. I joined a local angel club here in San Diego, which is one of the largest angel investor groups actually in the country. It was called Tech Coast Angels. Now, it's called NuFund Venture Group as of a few weeks ago. I have made a lot of investments in biotech startups and didn't have any plans to work full time again, let alone start a company.
Then in 2018, as you described, both my parents were diagnosed with cancer in quick succession. As we were working through that as a family, I was able to get their cancers sequenced to get them the best care available today. About nine months into that, our little dog Poppy got diagnosed with advanced pancreatic cancer. She was just four years old, a tiny little thing, a rescue mixed-breed dog, but had a heart of gold, infinite love and loyalty, and it was really devastating for us to get that diagnosis, especially coming on the heels of the challenges we had in the family; eight weeks and a lot of money later, she was gone. The cancer was just diagnosed too late.
As I was looking into ways to help her, of course, I reached into the familiar bag of tricks, things I was doing for my parents. I have been privileged to be at the cutting, bleeding edge of genomic oncology from the early days with Illumina and Sequenom. I realized literally in the first few hours of searching that this world of genomic oncology had not yet intersected the world of veterinary medicine.
For one thing, there are no cancer-screening guidelines for dogs. The American Cancer Society has cancer-screening guidelines for five cancer types. The World Health Organization has cancer-screening guidelines. European countries have cancer-screening guidelines. No such thing in the veterinary space, not for cats and not for dogs. Also, the veterinary diagnostic space is completely unregulated. I spent a few days looking for who's going to certify our test, who's going to give us the gold star for doing a good job, and discovered that there is no regulation in this space. There is no FDA or USDA regulation and no CAP (College of American Pathologists) and no CLIA (Clinical Laboratory Improvement Amendments).
Unfortunately, dogs get diagnosed very late. Dogs don't have “symptoms”. Symptoms are something that you and I have when, for example, you have a tummy ache and you go to your doctor and describe your symptoms. Dogs only have clinical signs. That's when they're hunched over in pain because of the tummy ache and that's what you as a human perceive it as a clinical sign. So dogs are typically diagnosed late with cancer, and by that time, the disease is often disseminated.
Another thing that really shocked me as I started looking into this as both a way to fight back against the things that were happening to me personally but also as a way to give back and make sense of Poppy's really senseless passing, to give some sense to her life and to her legacy, was the fact that cancer is actually the biggest killer in dogs.
About 6 million dogs are newly diagnosed with cancer each year in the U.S. even though the population of dogs is only about 90 million, while in humans, there are just under 2 million new cancer diagnoses each year in a population of over 330 million. If you do the math, you figure that dogs get cancer at an annual incidence rate of about 10 times higher than humans: One third the population with three times the number of cases. That really comes down to the fact that dogs and humans have the same lifetime risk of cancer and about one-in-three dogs and one-in-three humans will develop cancer during their lifetime, so from that point of view, it's not more common; but dogs live much shorter lives, so they need to compress that lifetime risk into a few short years of incidence, and that's what gives rise to these highly discrepant numbers.
We started PetDx with a number of colleagues from organizations including Illumina, Memorial Sloan Kettering Cancer Center, and Sequenom to address this biggest unmet need in dogs. Cancer is in fact by far the biggest cause of death in dogs. You need to add up the next five leading causes of death just to add up to cancer as the number one killer in dogs, yet there are no screening guidelines. There is no early-detection paradigm, dogs are just diagnosed late and then they die. Veterinarians are typically taught, when you hear cancer, start talking palliation, pain control, get the family ready to accept the inevitable.
Of course, in the human space, that's not what we do. In the human space, we know that early detection saves lives. That's the motto of GRAIL. That’s why the American Cancer Society has formal screening guidelines, and that's why insurance, both private and public, pays for those screening tests. Insurance is not in the business of paying for stuff that doesn’t work.
Those are fascinating points you've shared with us. The disproportionate rate of cancer in dogs compared to humans immediately stuck out during our initial research as well as the lack of a clear veterinary regulatory pathway and the need for that. To make sure we go over your recent exciting news, let’s talk about the CANDiD study.
PetDx recently published the results of your CANDiD international multi-center clinical validation study for the OncoK9 multi-cancer early detection (MCED) liquid biopsy test for canines. The study showed OncoK9’s impressive capability to identify 30 total canine cancers with a focus on 3 main aggressive cancers, namely lymphoma, osteosarcoma, and hemangiosarcoma, as well as 8 common cancers. OncoK9 displayed strong specificity and comprehensive all-around performance data (e.g., sensitivity, PPV/NPV, site of origin) across pure-breeds and mixed-breeds. What was your goal for the CANDiD study and what do you believe are the major takeaways from the findings in regard to the OncoK9 liquid biopsy test?
When we started planning the clinical development and validation program for OncoK9 back in 2019, we wanted to model it after best-in-class human liquid biopsy studies, which automatically meant a large sample set. We aimed from the beginning for over one thousand dogs. When we were telling investors this in the Summer of 2019, for many specialists from the veterinary space, it gave them pause because they said this has never been done in the veterinary space. They asked: “How will you do this as a totally new startup when even established companies don't do thousand-dog studies?” We said: “Well, we'll do it with your money and with the expertise that we have from doing these kinds of studies in the human space.” Our goal was to do a clinical study that was very compelling to veterinarians exactly because there is no regulation in this space.
You see, in the human space, as an MD, you learn that the FDA has approved a drug or a diagnostic, and even though you haven't read the details, you know that the FDA has pretty high standards, so nobody's going to question you for using that for the indicated use when the FDA has approved it. The veterinary space has no regulatory pathway, and veterinarians understandably are a skeptical bunch because you can literally set up a diagnostic company in your garage and flip a coin every time a sample arrives and you show results, and there's no regulator to come after you and send you a cease-and-desist letter and shut you down. A claim as bold as — “we can detect cancer with a blood draw” — is a bold claim, and companies have come and gone in the past in this space saying they can do that, but in fact, they could not do that. So the only way to overcome this very legitimate point of resistance and skepticism was to set up a very large, very compelling study to show the performance of the test.
We also wanted this to be representative of the typical patient population that the veterinarian sees. That's why we enrolled at about 40 clinical sites on four continents, a mix of private and public institutions, universities, smaller clinics, and larger clinics. This was truly an “all-comers” study. We deliberately did not want to specify and only enroll dogs with a certain cancer type or only enroll dogs of a certain breed. Different breeds may have different predispositions for getting cancer, but once you do have the cancer, the genomic features of these cancers tend to be very similar regardless of breed. So we deliberately set it up as a very large study that includes all comers, all different cancer types, all different breeds, et cetera, to mirror the real-world experience of our veterinary customers, who see a mixed population of dogs coming in.
Another thing that we did that’s really unique — to our knowledge — in the veterinary diagnostics validation space is, we used an independent training set and testing set. It's easy to take a new assay, apply it to a population of dogs, really tweak all the aspects of the test and overfit the data, and say: “Well, this is the performance of our test.” Then, you go to the market with that and that's in fact what happens... you just set the cutoff in one set of subjects and that's your test.
The gold standard, of course, for doing this — and GRAIL has shown that in their paper last year in Annals of Oncology with over 4,000 patients — is that you validate your test through proper cross-validation and controls for overfitting using a training set. Then, you establish your cutoffs, your algorithm, your pipeline based on that training set. Then you say, “okay, let's apply that without any further tweaks to an entirely different population,” a testing set, or sometimes it's also called a validation set. We chose to call it a testing set because the entire study is a validation study that includes a training set and a testing set. That's why even in the title, we wanted to emphasize that we used an independent testing set for validating this assay, to further convey the point that this is done at very high standards of clinical study design to really avoid the overfitting of the data.
In terms of the significance of the findings, we showed that our test has a detection rate of over 85% in these three very aggressive cancers, lymphoma, hemangiosarcoma, and osteosarcoma, which together account for 25-plus percent of all cancers in dogs. In eight of the most common cancers, which include these three but also include things like mammary gland carcinoma, soft tissue sarcoma, and mast cell tumors, the overall detection rate was about 62%, and across all the 42 cancer types that we evaluated in the study, the detection rate was 55%. Out of those 42 cancer types, we were able to detect cancer in at least one patient per cancer type in 30 cancer types. Overall, we were able to do this at a specificity of 98.5%, which corresponds to a very low false-positive rate of just 1.5%.
We actually believe that the performance in the real world for specificity is probably a bit better even than 98.5%, which itself is quite impressive. This is because we collected samples from November 2019 until the Summer of 2021, but the test was not ready to be run until late spring of 2021. So all those samples had been frozen at minus 80 (degrees Celsius) ready to be tested.
When we tested the samples, when we looked at this presumably cancer-free cohort, we found a number of subjects that had a cancer signal detected on our test. We went back to their veterinarian and asked: “Well, what can you tell us about what has happened to the patient between maybe a year ago to a year and a half ago, when we collected the blood sample, and today, when we have a result?”
In some of those cases, we were told: “You know what? The dog was diagnosed with cancer eight months after the blood sample was collected.” Okay, so that obviously was not a false positive. We just took it out of the specificity analysis. In other cases, the dog was still alive and we paid for a cancer workup and we could not find the cancer at that time, so those were very legitimate false positives, and there were a few of those.
Then, there were a few patients who had died in the intervening period, but there was no documented cause of death: “The patient died at 14 in her sleep in late 2020.” “Well, what did they die of?” “We don’t know. She died in her sleep.” So, we couldn't prove that the dog had cancer at the time of death, but we were looking at the results of the test, and it was pretty obvious that there was a cancer signal in the body. However, we don't know if the dog died of cancer. We're fairly certain that they died with cancer, but because we did not have proof that the dog had a definitive diagnosis of cancer at the time of death, we had to accept them as false positives. So even when including those, we had basically eight out of 519 false positives that included three or four of such cases, and that's what gave us a false positive rate of 1.5%.
Now, let’s move more to the technology side of things regarding your OncoK9 assay. What are some of the specifics in technology used for ctDNA analysis in pets, for example, biomarkers and analysis approaches? What are some of the major technical and clinical differences between liquid biopsy applications or biomarkers in dogs and humans? Can we learn from data and results in dogs to inform human health, or vice versa, and if so, how?
Absolutely. We are a little bit protective, for lack of a better term, of our data in the sense that although much of this research has been done in the human space, it hasn't really been done in the canine space, and we have discovered certain features of the canine cancer genome or canine cell-free DNA in plasma that haven't been seen in humans, so we filed IP on that. The IP is not public yet, so given the very fluid nature of the central-lab space, we just would rather keep these as trade secrets.
However, if you look, for example, at this case study that we feature in our publication, which is one of many case studies we have on file, you can see the types of alterations we look for. This shouldn't be a surprise to you guys, you are familiar with this space. You can see, for example, this dog has very clear copy number alterations (CNAs) on multiple chromosomes. I can walk you a bit through this case (Figure 4 in the publication).
You can see here that this dog actually was enrolled in December 2020. We ran the test in April 2021. A cancer signal was detected and the result was reported. By May, the patient came back in, and we asked for another blood sample. Again, it was a cancer signal detected. The patient was also examined. The patient was completely fine at the time, so even five months after this blood collection, an initial examination did not reveal anything.
An x-ray started getting a bit suspicious, and to make a long story short, the CT scan showed a left ventricle mass and clear pulmonary infiltrates suggestive of metastatic hemangiosarcoma. A third blood sample in July 2020 also had a cancer signal detected. This is where the dog started showing mild clinical signs, and eventually the disease progressed, and the dog was euthanized in August 2021.
To give you a sense of how these hemangiosarcoma cases almost universally present, they tend to be asymptomatic, and then that heart mass ruptures, and then, there's bleeding into the pericardium. The dog presents in the ER at two o'clock in the morning, is diaphoretic, cannot breathe, has pale gums, and lost a lot of blood, and the owner suddenly has to make life and death decisions under very stressful and time-constrained circumstances.
In this case, the owner is actually a veterinary tech, and she was very grateful because she knows how these dogs present, and she was very grateful to receive this kind of advance notice so she could do palliation. She could say goodbye to their dog and do their bucket-list things.
You can see here, in Fig 4 of the publication, that from blood, we sequence both the gDNA (genomic DNA) from WBCs (white blood cells) as well as the cell-free DNA from plasma. Each dog acts as their own intrasubject control because we compare the gDNA and the cell-free DNA.
Just as an interesting tidbit, when we see penetration of a somatic signal into the gDNA, that is sometimes, not always but sometimes, a sign of a hematological malignancy. Because in a lymphoma and a leukemia situation, it's the white blood cells that are the tumor and then you see mutations or genomic alterations in both plasma and gDNA.
Here, you can see in this patient’s case, the gDNA in December 2020 was normal. However, in the plasma from the same blood sample, you can see there’s multiple copy number alterations partially on chromosome 2, chromosome 5, and so on, loss on chromosome 16, et cetera. Then, on the second test, in May 2021, you see the same copy number variations, except more pronounced. If you look at how far these alterations are from the baseline, you can see that it's more pronounced. Then, the patient was put on palliation, anti-inflammatory drugs and so on, which are not curative for cancer, but they do dampen the signal because they slow down the inflammation and therefore slow down the necrosis of the tumor.
Again in July 2021, you see the same signal, but it's not necessarily more pronounced compared to May, because of this palliation, meaning that at least the patient's subjective experience of the disease is palliated. You can also see here that the patient, in addition to these copy number variations, had mutations in two genes including the p53 gene, which is of course the guardian of the genome — very commonly mutated in human cancer and, not surprisingly, very commonly mutated in canine cancers. There’s again an increase in VAFs (variant allele frequencies) from the December blood sample to the five-months-later time sample.
On the right-hand side, the owner was very generous again in allowing us to collect tumor tissue samples at necropsy. Unfortunately, the patient did have metastatic disease, with tumors in the spleen, the heart, and also in the ribs and in the lungs. You can see how there is even a bit of a tumor heterogeneity here where the mass in the spleen has no alterations on chromosome 1 that are visible at least, whereas the masses in the rib and in the lung have very clear alterations on chromosome 1. You can pick up on some of the heterogeneity in other places as well, but by and large, you can see that many of these are truncal alterations that are recurrent throughout the various tissue types and, very importantly, are captured in the plasma at the various time points.
So that's one example. We look at other things as well that we plan to publish on in the future. Again, things that haven't really been documented in the canine genome and in the canine cell-free-DNA compartment, but I think this gives you a good example of our approach.
I do want to also pick up on the point about what we can learn from dogs and apply to humans. It turns out it's a lot. The two genomes are about 85-86% similar, but when you look at just the tumor suppressor genes and oncogenes, the level of homology is well over 90%. In fact, if you look at the COSMIC database and you look at the top 100 variants in the COSMIC database for humans, 95 out of 100 have a perfect orthologue in the canine genome. If you go to petdx.com/resources, you will see we have a number of posters, including one we just presented at AACR about this comparative oncology angle. We can learn a lot about the genomics of cancer in dogs, and apply those lessons to humans.
That's particularly valuable, we think, for cancers like sarcomas, which are very rare in humans but very common in dogs. Sarcomas account for about 1% of adult cancers in humans. They're very painful cancers, very devastating cancers, and they typically affect younger people such as adolescents and young adults.
Take osteosarcoma. There are under a thousand cases per year in the U.S., so even a large cancer center would have difficulty compiling a large cohort to run a large genomic profiling study. Osteosarcoma is very common in dogs, about 20-30 times more common in dogs than it is in humans, which is obviously unfortunate for dogs, but from a learning perspective, we can learn very quickly from dogs with osteosarcoma and apply those lessons to humans.
Finally, think about how cancer drug development has happened in the past two decades: you don't want to put a new cancer drug directly into a human being, so all these pharma companies have developed ways of preclinical research. Before you do a first-in-human study, you try to test it in preclinical models. So what are those models? Petri dishes with cell cultures, and rodent xenografts: a little bit of human tissue is transplanted into a mouse or a rat either under the skin or in an orthotopic way (in the actual organ where the cancer came from in the human) and allowed to grow for a while. Then, the Petri dish or the rodent is treated with a drug.
As you well know, many Petri dishes and many xenografted rodents have been “cured” of cancer only for the drug to then move into human clinical trials and then fail for efficacy reasons, not safety reasons. Everybody expects these drugs to be very toxic. They fail for efficacy reasons. You just do not see that overall survival improvement. You don’t see that PFS (progression-free survival) improvement.
So now, think of dogs. Dogs have naturally- occurring cancers. They don't have to be genetically modified to accept a xenograft or to develop cancer. They unfortunately develop these cancers naturally. It’s much of the same types of cancers that humans develop. They're exposed to the same carcinogens in the environment that humans are because they live with us.
We do see a future where we will be able to work with CROs (contract research organizations) and pharmaceutical companies where we provide them access to pet owners who say: “Okay, you found cancer in my dog. I cannot treat the cancer. I understand there is a drug developed for humans that targets this exact genomic alteration that is found in my dog's cancer. I would be willing to enroll my dog in a clinical study, just like a human being would enroll in a clinical study, in the hope of both helping my dog as well as furthering research.” This wouldn't be experimenting on these pets any more than we experiment on humans who volunteer for these human clinical trials.
Unfortunately, my father did pass away in the end from cancer, and in the last few months of his life, he did go for some phase I studies because that's the only thing that was left to try. So there’s a similar mindset here, providing access to experimental treatments for these genomically orthologous cancers that are naturally occurring in dogs, with an intact immune system and very similar genomically to human cancers.
We hope that in the future, this could become an alternative model for preclinical efficacy proof points before first-in-human trials; not in an experimental sense of experimenting on dogs, but rather in the sense that these pet owners would want their dogs to be involved in these studies because there would be a high probability for the drug being efficacious in the tumors based on the genomic profiles.
Thank you so much for sharing that. Great, and from here, I just wanted to switch gears and discuss more about how you've been navigating this space. In the veterinary diagnostics field, there is a lack of clear regulatory approval processes in the United States. How is PetDx working to establish a formal regulatory approval process for veterinary diagnostics? The United States certainly has different standards of treatment for humans as compared to animals however, PetDx is working to bring veterinary medicine up to higher standards. What sort of veterinary diagnostic regulatory processes does PetDx currently look to for guidance?
We built this test with expertise from the human side of things. My entire career in industry has been in medical affairs and clinical development.
I built and led the team that did the clinical trials for the MiSeqDx FDA clearance back in 2013. That was the very first next-generation-sequencing platform to get FDA clearance and also CE marking in the European Union. Also, the very first IVD (In-Vitro Diagnostic) assay, the Cystic Fibrosis 139-Variant Assay, was developed and launched together with the platform back in 2013. I was also part of the team that did the first PMA (Pre-Market Approval) for a companion diagnostic (CDx), using tissue, for a targeted cancer drug during my time at Illumina.
Many of the leaders on our team come from a similar environment. Our VP of Laboratory Operations, for example, was previously the director of the entire clinical lab at Sequenom, bringing a lot of experience from the CAP and CLIA regulated space. So between the FDA, CAP, and CLIA, literally all of our R&D team is coming from the human space with that sort of regulation in the back of our minds. That's the kind of SOPs (standard operating procedures) we have built. That's the kind of work we have done. You can see it’s reflected in the clinical-study design and it's also reflected in the day-to-day operations of our lab. We do everything in-house. We don't outsource anything. We have two Illumina NovaSeq 6000’s. Everything from sample accessioning to final reporting is done in-house to those kinds of standards.
There are no regulations though in this space and I want to clarify that doesn't mean that there is no requirement for regulations. It literally means, there are no regulations. There is nobody that we can take our work to and say: “Here is our work. Please give us that certificate of achievement, that gold star, that clearance, that approval, that regulatory imprimatur.”
I don't know when that will change. I don't know what would bring about that change. It's not our place to speak to those things, but it is our place to recognize the fact that this creates a bit of a Wild West environment where then the only thing standing between good science and the very justified skepticism of the veterinarian is doing work of the level that we have done.
That's why we've invested so much money and time. We could’ve turned this around with just a few hundred dogs in nine months, but that's not what we chose to do. We chose to do a very large study, build a very robust product and a very robust pipeline and workflow because this lack of regulation creates an environment where you have to work that much harder at demonstrating good science so people can believe your claims.
Got it. Thank you so much for clarifying that. It's really amazing to hear that even if there are no standards, you hold yourself to such high standards. Just to wrap up here we wanted to ask a little bit more about the future outlook of what you have for PetDx. What are some future trends or goals that you see for PetDx in the next 5 years?
Absolutely, from day one, we mapped out the universe of possibilities for liquid biopsy. This is no surprise to you guys, who are experts in liquid biopsy, but this is, basically, the universe of liquid biopsy, and based on our market research, we assessed that the biggest opportunity, both from a commercial perspective as well as from an impact on populations and health of many dogs perspective, we should launch the test in these two use cases prior to diagnosis, which means screening and aid-in-diagnosis.
Aid-in-diagnosis is not such a big thing in the human space yet, although, for example, if you read the GRAIL SEC filings, they do talk quite a bit about aid-in-diagnosis. But in humans, once you suspect cancer, you don't typically go to liquid biopsy as the next stage, instead you often go to very advanced imaging; you do biopsies and so on.
It is not so in the veterinary space. In the veterinary space, very few people have insurance, with less than 5% of dog owners having insurance. It's an all-cash-based space and people are often squeezed between, do I spend $2,000-3,000 to pursue a definitive diagnosis, or do I save that money for treatment in case my dog actually turns out to have cancer?
In many cases, you have a dog that's coming in, the “technical” term is ADR, “ain't doing right,” they lost weight, they don’t want to go on walks anymore, they're lethargic. We can't quite pinpoint a location for the tumor, but the clinical intuition of the veterinarian is that cancer is one of the top three items on their differential diagnosis. It will often be a very expensive diagnostic odyssey to really chase it down and make that definitive diagnosis of cancer.
In other cases, you do a physical exam, you do an imaging exam, and you find out, for example, there's a heart-base mass. Can you be a hundred percent sure that that's malignant? Not a hundred percent, but you correlate the clinical presentation with the findings on imaging, and now, you get the results from OncoK9, which has the cancer signal detected where you see these somatic alterations, which you know very well don't happen from any other disease. Diabetes doesn't change your DNA at a somatic level and neither does arthritis and neither does a worm infection. Only cancer changes your DNA at the somatic level. That's why our test is so specific for cancer. So now, you're triangulating between a clinical picture, some imaging finding, and an orthogonal method of saying: “Hey, we found somatic alterations in the DNA from your dog.”
Aid-in-diagnosis is actually a very important use-case for us and a significant proportion, about 30% of our test volume today, comes for aid-in-diagnosis because the moment a veterinarian suspects cancer in the clinic the first time, they can act right away. It can take weeks to schedule an ultrasound or a CT scan. You don't really want to tell the owner: “Well, come back in six weeks. If it's gotten worse, then we will know that back then it was cancer.” This gives the vet and the pet owner an opportunity to do something right away: “Let's draw that blood today. We'll get the result back in about 10 business days or less, and we'll already make some headway towards a diagnosis.”
Then, there are these four post-diagnosis use cases. In the next few months, literally, we plan to put out a white paper about minimal-residual-disease (MRD) detection, recurrence monitoring, and monitoring for treatment response. We can do that because when we collected samples from cancer-diagnosed subjects in our study, we didn't just collect one blood sample, we collected samples at month 3, 6, 9, all the way to 12.
Now, we have tested all of these longitudinal samples. We collected the case report form in each one of those clinic visits: What is the status of their disease by RECIST (Response Evaluation Criteria in Solid Tumors) criteria? Has the dog died, perhaps, in the interim? And we collect that as a closeout for that patient. What kind of treatment have they been receiving? So this allowed us to then go back and determine the performance of the test in these three use cases: MRD, recurrence monitoring, and monitoring for treatment response.
All I can say is, the early results are very encouraging. Honestly, they're not that much of a surprise. It’s what you see in the human space. It just hasn't been shown in the veterinary space, but we know in the human space, for example, that detection of circulating tumor DNA (ctDNA) post-surgery gives you a far higher chance of clinical disease recurrence than absence of ctDNA after surgery. So we basically see the recurrence of molecular signal weeks to many months prior to the onset of clinical recurrence. That's the bottom line.
In the future, we will also continue research into targeted treatment selection. This is, of course, where liquid biopsy was born in the human space, things like early PCR (polymerase chain reaction) tests from Roche, which was like: “Hey, let's take the testing away from tissue and do it in plasma, because you don't want to do tissue biopsies on lung-cancer patients every time you are worrying about recurrence and emergence of resistance mutations.” There are obviously a number of products and panels in the human space that target specific markers for which, instead of doing it from tissue, you want to do it from plasma for cost reasons, for risk reasons.
In the veterinary space, this is actually the smallest market for a number of reasons. One is, only specialists can give these types of highly targeted treatments and there are less than 500 boarded veterinary oncologists among more than 50,000 veterinarians in the U.S. Two, these drugs have not been tested for efficacy in dogs. Just because they work in humans and the dog has an orthologous KRAS, NRAS or EGFR mutation doesn’t mean that it is actually going to have efficacy in that dog for that tumor type.
You guys know — now it's a classic story — about how everybody got excited about vemurafenib working in melanoma patients with a V600E mutation and then people got excited when they found V600E mutations in colon cancer in humans. They said: “Surely vemurafenib is going to work because the mutation is the same gene, same exact locus,” and it didn't work. So even within the same species, just jumping across one tumor type, a drug that worked very well against a certain target didn't work in another cancer type that featured the same exact target.
So there is a lot of reason to be optimistic about the future of precision oncology treatment in dogs, but we feel that there's a lot of research that still needs to be done on the efficacy side and demonstrate that these drugs actually work in dogs before we roll that out clinically. So this is something that we have ongoing research on, but we do not plan to roll it out as a product in the near future.
Thank you so much for shedding light on all of your future plans. It's all very exciting, and I look forward to hearing about it in future years. So with that, that sums up all of our questions today, but if there's any last-minute comments that you want to leave us with, we're happy to listen to those.
Thank you. I really appreciate your time and very thoughtful questions. It's always a pleasure to speak to individuals who are really steeped in the science and know the space well on the human side and can appreciate what we've built here for humanity's best friend.
I just want to close on the thought that cancer is a huge problem in dogs. We really have developed a pioneering novel solution for early detection. We're very proud of our science. We are keen to see it adopted as part of the standard of care.
The test is recommended for dogs seven years of age and older just like human cancer screening tests are typically recommended in humans 45 to 50 and older. That's where the risk of cancer is highest. Dogs above the age of seven have a nine-fold higher risk of cancer than dogs below the age of seven. It’s also recommended at somewhat earlier ages in dogs from predisposed breeds. For example, some giant breeds tend to have cancer diagnosed when they're six or seven and you obviously want to start screening about two years before the median age of diagnosis, to give yourself a chance of catching that cancer preclinically. So, in some breeds, as early as age four. It's also recommended as an aid-in-diagnosis for dogs in which cancer is suspected based on clinical findings.
So we're really excited that, like you said in the beginning, this is the first application of liquid biopsy outside of the human space. We also have plans to develop feline products; we trademarked OncoFeline, for example. But dogs get cancer a lot more than cats, and that's why we started here. It was really shocking to see that there is so little done for dogs.
Historically, we have, for example, used dogs in pharmaceutical preclinical studies for safety. We always basically use them and then move on and develop the drug for humans and we never look back and say: “Let's develop this for dogs.” There are only three drugs available specifically for dogs with cancer that have regulatory approval; there are hundreds available for humans. There are obviously dozens of liquid biopsy companies for humans; we are the only one for dogs. We just haven't given back enough. Many of us have a sense of mission here at PetDx. We've lost dogs to cancer. We want to give back as a human species and help address this single biggest problem in dogs, cancer, which is currently really unaddressed.