Researchers launch largest open-source pet tumor database

A research team from the University of Liverpool and the University of Las Palmas de Gran Canaria has launched what it describes as the world’s largest open-source database of canine and feline tumors, bringing together more than 1 million records in a bid to sharpen cancer surveillance in companion animals. The registry aims to solve a longstanding problem in veterinary oncology: most pathology data are generated routinely, but much of it sits in fragmented laboratory systems that are difficult to aggregate and analyze at population scale. (liverpool.ac.uk)

The project didn’t appear overnight. Liverpool researchers have been building toward this for years through the Small Animal Veterinary Surveillance Network, or SAVSNET, which already collects large volumes of diagnostic laboratory data from across the UK. In a 2021 Scientific Data paper, the group described a text-mining pipeline that extracted and normalized tumor information from 180,232 canine and feline electronic pathology records submitted by three UK laboratories, producing a pathology tumor registry of 109,895 tumors. At that point, the authors said the Swiss registries, with more than 85,000 tumor cases, were the largest animal pathology tumor registries reported. (nature.com)

The new announcement suggests that effort has now expanded dramatically. According to the University of Liverpool, the database contains more than 1 million records and is intended to let researchers study rare cancers and uncommon breeds in ways that weren’t previously possible. The university said the work combines veterinary pathology, epidemiology, data science, and clinical expertise, with diagnostic laboratory data standardized into a unified research resource. A related paper published in 2026, “Epidemiology of Four Major Canine Tumours in the UK: Insights From a National Pathology Registry With Comparative Oncology Perspectives,” is one of the first examples of how the registry can be used. (liverpool.ac.uk)

The comparative oncology angle is important. The earlier Liverpool registry paper explicitly framed companion animals as both sentinels and models of human health, and that case has only strengthened as newer genomic studies come online. In February 2026, an international team published what was described as the first large-scale genetic map of feline cancer in Science, sequencing 493 tumor-normal pairs across 13 feline cancer types collected through routine veterinary diagnostics in five countries. The study found recurrent mutations in genes including TP53, FBXW7, CTNNB1, PTEN, and TRAF3, with TP53 mutated in about one-third of tumors overall, and identified strong parallels between feline and human cancers. In mammary carcinoma, one of the most common and aggressive feline tumors, FBXW7 mutations appeared in more than half of cases and PIK3CA alterations were also common, findings that may eventually matter for targeted therapy development in both species. (nature.com) (science.org)

That doesn’t mean this new database is a clinical tool yet, but it does suggest a growing infrastructure for studying how tumor biology, breed predisposition, environment, and demographics intersect across species. The feline genomics work also reinforces why large registries matter: cats share household environments and at least some carcinogenic exposures with people, and naturally occurring tumors in pets may offer a more realistic comparative model than induced disease in laboratory animals. A recent comparative oncology review made a similar point, noting that dogs and cats can mirror human malignancies in histopathology, molecular alterations, tumor microenvironment, and even some treatment-response patterns, while still requiring species-specific dosing and safety considerations. (life-basel.com)

Public expert reaction has so far come mainly from the institutions involved. In Liverpool’s announcement, David Killick said the goal was to turn data that are usually “locked away in private laboratories” into a research-ready resource, while José Rodríguez Torres said veterinary cancer analysis has lagged behind human medicine because of fragmented data. Their comments align with a broader trend in veterinary oncology toward larger registries and shared datasets, including efforts such as Portugal’s Vet-OncoNet and other national registry initiatives. They also line up with calls from other epidemiology groups for broader registry infrastructure: a 20-year UC Davis study of 150,063 dogs and cats seen at a tertiary hospital found 26,883 cancer diagnoses and concluded that a California-wide companion animal cancer registry would provide a more complete picture than hospital data alone. In that study, cancer incidence declined over time in cats but remained stable in dogs, age was the strongest predictor in both species, and age-by-sex-neuter interactions affected risks for sarcoma and carcinoma. (liverpool.ac.uk) (sciencedirect.com)

Why it matters: For veterinary professionals, the immediate significance is less about a new test or treatment and more about better evidence. Cancer risk discussions with pet parents are often shaped by relatively small institutional datasets, referral bias, or breed anecdotes. A registry with this scale could help validate which associations are real, clarify how age, breed, sex, and neuter status interact across tumor types, and surface patterns in uncommon cancers that primary care teams and specialists rarely see often enough to characterize alone. That matters because existing studies already show the field is dealing with nuanced, cancer-type-specific patterns rather than simple one-variable risks. The California hospital analysis, for example, found that older age strongly increased risk in both species, but odds of sarcoma, lymphoid neoplasia, and mast cell tumor declined in senior dogs, and neuter-status effects differed by sex and tumor type. Over time, better population-scale data could influence everything from preventive counseling and earlier workups to referral pathways and research enrollment. (nature.com) (sciencedirect.com)

It may also help connect clinical practice to precision medicine and One Health research. Large, standardized registries can support downstream work in genomics, pathology, and treatment response, especially when paired with biobanks or molecular datasets. SAVSNET’s tumor registry page indicates the platform is meant to enable longitudinal epidemiology and identify novel risk factors, while also extending toward genomic analyses of cancer biology. That makes this database potentially useful not just for counting tumors, but for building the evidence base behind future oncology tools. The feline Science study offers a glimpse of what that future could look like: researchers reported potentially actionable alterations in a subset of tumors, highlighted mammary-cancer parallels with human breast cancer, and suggested that mutations such as FBXW7 and PIK3CA could help prioritize translational drug research. On the canine side, other groups are already testing less invasive precision-oncology approaches, including blood-based immune-signature analysis in pet dogs with diffuse large B-cell lymphoma, where expression of genes such as CD1E and CCL14 was associated with longer survival in an experimental chemo-immunotherapy setting. (liverpool.ac.uk) (science.org) (newswise.com)

Even relatively niche pathology studies underline why broad registries are useful. A long-term institutional cohort from Italy reviewing mammary tumors in male dogs and cats found that these rare cases were predominantly malignant in both species, with a particularly aggressive pattern in male cats, where intermediate- and high-grade tumors were more common. Findings like that are clinically relevant, but hard to interpret in isolation without larger background datasets that show how uncommon presentations fit into the wider epidemiologic picture. (wiley.com)

What to watch: The next milestones will be peer-reviewed outputs showing what the 1 million-record database can actually reveal, whether additional labs and countries contribute data, and whether the findings can be translated into practical risk stratification or comparative oncology studies with clinical relevance. Just as important will be whether registry-scale surveillance can be linked to genomics, biobanking, and treatment-response datasets in ways that support species-aware precision medicine rather than one-size-fits-all extrapolation from human oncology. (livrepository.liverpool.ac.uk) (life-basel.com)