NEW YORK – Veterinary startup PetDx is applying lessons learned from the liquid biopsy space in humans to the world of dogs by launching its OncoK9 assay to screen for multiple early-stage cancers from a dog's blood sample.
The San Diego-based firm believes that its newly-launched panel — which spots genetic biomarkers in a canine's cell-free and genomic DNA and will cost $400 per test at select Petco locations — can be used as a screening test in elderly dogs and as an aid for diagnosing dogs suspected of cancer based on clinical signs.
Luis Diaz, head of the division of solid tumor oncology at Memorial Sloan Kettering Cancer Center and an adviser to PetDx, said he also envisions the test's use for additional cancer-based applications, including monitoring for minimal residual disease and as a companion diagnostic.
According to Bruce Smith, a pathobiology professor at the Auburn University College of Veterinary Medicine, cancers in dogs are most often detected after identifying "lumps, bumps," and odd behaviors, such as a limping back leg.
While the most common human cancers include breast, lung, and prostate cancer, dogs are most often diagnosed with lymphoma, mast cell tumors, sarcomas, mammary gland carcinoma, and malignant melanoma.
Following a suspicious finding, vets typically refer the owner to a veterinary oncologist, who confirms the diagnosis with a tissue biopsy and staining. By that point, however, the dog's tumor may have advanced too far to benefit from surgery or other treatment.
PetDx CEO Daniel Grosu said he had the initial idea to launch PetDx after his dog was diagnosed with advanced pancreatic cancer and passed away in 2019. Grosu, who has served as chief medical officer at Sequenom and Illumina, began speaking with industry experts about the landscape of early-stage, noninvasive liquid biopsy-based methods for cancer detection in veterinary medicine.
"When I looked at the options, there was nothing for dogs," Grosu said. "Everything that had happened in the human precision oncology space hadn’t intersected in the space of veterinary medicine."
Grosu therefore decided to develop a next-generation sequencing, or NGS, platform based off work done by liquid biopsy firms. However, he pointed out that the biology and genomes of humans and dogs differ widely, forcing his team to build a bioinformatic algorithm to incorporate the differences, followed by clinical validation.
PetDx's OncoK9 assay begins by separating plasma and buffy coat from a blood sample and extracting DNA from both, followed by proprietary library prep methods and sequencing on an Illumina NovaSeq instrument.
"We interrogate the cfDNA and genomic DNA in [white blood cells] for multiple classes of genomic alterations and use pretty sophisticated proprietary bioinformatic algorithms to generate cancer signals to produce a result," Grosu said. "[This is] followed by a manual review process to ensure all the metrics are met, and we can issue a result."
A dog owner interested in PetDx's OncoK9 test can visit select Petco full-service veterinary hospitals across the US. The vet extracts two 8.5 ml tubes of the dog's blood and ships it overnight to PetDx's lab at the Center for Novel Therapeutics — located in the University of California, San Diego's Science Research Park — with an expedited shipping label provided by the firm.
When the sample arrives at PetDx, the firm performs testing on the sample and generates a report within 12 to 18 days. Grosu said the company aims to lower the turnaround time to 10 days or less within the next year or two.
If PetDx's report states that the assay has detected a cancer signal, the firm will recommend potential next steps to the referring vet and pet owner. Grosu said the firm's clinical and genomic experts are open to speaking with the clinician over the phone to further discuss the test's results.
While PetDx has initially partnered with Petco to offer OncoK9 at "more than 30" locations by the end of May, Grosu said Petco plans to roll out the test to its growing network of partner clinics throughout 2021 and beyond. He added that PetDx expects to offer the test at veterinary clinics across the US this summer.
Validating the first canine liquid biopsy assay
To validate OncoK9's detection abilities, Grosu's team launched the "CANCer Detection in Dogs" (CANDiD) trial in November 2019, collecting blood samples from about 1,600 dogs with and without cancer at more than 50 clinical sites in North America, South America, Europe, and Asia.
As part of the internal study, results of which PetDx recently published as a white paper on its website, Grosu and his colleagues analyzed 332 control samples and 245 samples from dogs that were either diagnosed with cancer or were scheduled for tissue biopsies because they were suspected to have cancer (and later diagnosed as positive).
"We designed [OncoK9] not to be specific to any type of cancer, [but instead] looked across the genome for many genomic alterations that we knew were linked to cancers in general," Grosu said. "We then applied it to a large population of dogs with cancer."
The firm found that the OncoK9 assay had about 48 percent sensitivity and 97 percent specificity, which Grosu said is comparable to current human liquid biopsy assays. Among the seven most common canine cancers, he noted that the assay had a sensitivity of 60 percent at 97 percent specificity.
"The launch of the [OncoK9] test was based on baseline samples, and we used about 600 from this 1600 [cohort], because our results began to converge very quickly in terms of validation performance," Grosu said. "In our final cohort of 245 dogs with cancer, we showed the assay had a detection power in at least one subject for each of the 24 cancer types."
While Grosu acknowledged the OncoK9 missed some tumors, he pointed out the test's high specificity, which helps avoid false positives that can lead to anxiety and unnecessary invasive procedures. He said that his team will follow up on all the dogs at three-month intervals — for up to a year — to track disease progression, establish what he calls the "cumulative detection rate" over time, and identify if the dogs in the cancer-free cohort eventually develop tumors.
PetDx aims to submit the results of the CANDiD study to a "leading peer-reviewed journal" later this year, but Grosu declined to provide additional details.
Ahead of the anticipated submission, Grosu and his colleagues demonstrated the assay's use in a cohort of 11 dogs with a variety of cancer diagnoses, as well as in five cancer-free dogs, in a preprint posted on BioRxiv earlier this month.
In five of the nine dogs with a matched tumor and plasma sample, the assay identified genomic alterations in plasma, including single nucleotide variants and copy number variants, that matched alterations independently spotted in corresponding tumor tissue samples. The study authors emphasized that OncoK9 identified alterations found in "spatially separated tissues from the same subject," indicating the test's potential "for comprehensive genomic profiling of heterogenous tumors."
Smith, who is not affiliated with PetDx, believes the white paper and preprint studies have certain issues that the firm should address in future trials.
"Both the studies … are focused on dogs that already have cancer and have been diagnosed by the normal technology, [which] does not test the idea that this test can be used to identify animals with cancer prior to it becoming obvious by other means," Smith said. "That's not to say the test can't do that, but neither study is designed to show that."
Smith also noted that the discrepancy in median age of the two cohorts in the white paper could potentially skew diagnostic results. While the cancer-free group's median age was 6 years old, the median age of the dogs with cancer was 9 years old. Smith believes that the difference in age makes it difficult to decide whether "your cancer-free group is really cancer free … as opposed to being undiagnosed cancer patients."
Smith acknowledged that OncoK9 may offer advantages over standard cancer testing methods, with a few caveats. He believes the test may help inform surgeons whether they have removed all of a dog's tumor, but it is still unclear "how much tumor needs to remain for the assay to be positive."
Smith also wondered if OncoK9 can potentially compensate for tumor sampling bias because liquid biopsy tests report "all or nearly all of the alterations, giving a more complete picture" of the genetic alterations in the tumor. However, he noted that liquid biopsy assays can't provide information about what proportion of a tumor has any specific alteration.
In addition, Smith said the test can offer some indication of specific genetic alterations that may be targeted by a drug and therefore guide treatment. But he emphasized that the studies' results are not clear enough to identify if the treatable mutation "was in every cell or only a few cells."
Since launching in 2019, PetDx has raised over $18 million in funding, including an investment from Petco and two undisclosed investors, which Grosu described as "a large player in the human diagnostic space and a large player in the veterinary diagnostic space," as well as from leaders in the genomics space.
While PetDx has filed a number of patents related to its workflow, from sample extraction to bioinformatics analysis, Grosu noted that the firm is also holding trade secrets.
With an introductory price of $400 per test through Petco, Grosu argued that OncoK9 costs less than "any other invasive test" that a vet might use for a dog suspected of having a tumor. He recommended OncoK9 as an annual screening test for dogs that are at least 8 years old or belong to breeds that are highly predisposed to cancer, such as golden retrievers.
However, PetDx has not decided the price of the test for other clients beyond Petco.
"If you consider the vast costs of treating cancer in a dog, a cancer screening test below the $1,000 range is a very good value proposition," said Grosu.
Smith, on the other hand, said he believes the test's $400 price tag is "definitely on the top end of the price spectrum," as he noted a pet owner on average spends $300 to $400 for vet care per year. He acknowledged that owners of older dogs could consider the test valuable since cancer is largely a geriatric disease in dogs.
"It might be useful for the health of a purebred dog that you've invested in, that has cost you a lot to begin with, and you're willing to pay for it," Smith said. "Honestly, I don't see it in every dog's future, [as] there are some dogs that aren't going to get this test, and that's just fine."
While no official regulation exists for veterinary diagnostic testing, Smith said that veterinary diagnostic labs can voluntarily apply to a group called the "American Association of Veterinary Laboratory Diagnosticians," or AAVLD, and undergo an accreditation and certification process.
However, Grosu said that while the AAVLD offers an accreditation program, it is not available to commercial laboratories. Instead, he noted that the program is limited to government-funded diagnostic laboratories and focuses on infectious disease surveillance in animals.
Diaz noted that there are cancer research opportunities for PetDx's NGS' platform, where researchers could test therapeutics that focus on similar targets in human and canines.
Following the large-scale launch of OncoK9 this summer, Grosu said PetDx may indeed explore performing long-term follow-up of "certain categories of canine patients who initially undergo commercial testing" at the firm. He highlighted that PetDx "is already engaged" with several veterinary pharmaceutical companies and is developing a program to use its platform and data to develop therapeutics for both dogs and humans.
"There is a high degree of homology between humans and dogs across actionable target regions in oncogenes and tumor suppressor genes, given the high degree of evolutionary conservation in those areas of the genome," Grosu said. "We see a massive long-term opportunity in comparative oncology with 'big data' mining and enabling large-scale pre-clinical efficacy studies in dogs with naturally occurring cancers featuring canine orthologs of targetable human variants."