Canine tumors closely model many adult human cancers, study finds

A comparative analysis of 5,875 tumor samples found that canine cancers closely resemble many human adult cancers at the transcriptomic level, suggesting that dogs can be useful models for studying human malignancies. Researchers analyzed 913 canine and 4,962 human samples spanning 11 tumor types, including 806 pediatric human samples. Overall, the study found stronger similarity between dog tumors and adult human cancers than between dog tumors and pediatric human cancers, with gliomas and sarcomas standing out as notable exceptions.

The work, published in BMC Cancer , was designed to address a gap in cross-cancer research. Individual studies have described similarities and differences between particular human and canine tumors, but the broader degree of molecular similarity across tumor types remained unclear. Dogs offer several practical advantages for comparative oncology: they share aspects of human diet and environment, are closely related to humans from a genomic perspective, develop spontaneous tumors that closely resemble human tumors, and have a high tumor incidence rate.

Their status as popular pets also makes a wide range of samples available for scientific study. To make this pancancer comparison, the researchers developed a formula for quantifying transcriptomic similarity using several metrics. The method was applied to samples drawn from 11 tumor types, with both adult and pediatric human cancers included.

This design allowed the investigators to distinguish general patterns across the cancer landscape from cases in which the relationship between canine and human disease differed by tumor type or age group. The central result was that dogs were generally excellent models for many human cancers, but the degree of correspondence varied. The authors reported that canine tumors tended to be better models of adult human cancers than pediatric human cancers.

This pattern was not universal: gliomas and sarcomas were identified as notable exceptions to the broader tendency. The study therefore shows why similarity should be evaluated at the level of particular tumor types rather than inferred for all cancers from a single species-wide conclusion. Its findings support using canine tumors as relevant comparators for many adult human cancers, while indicating that the pediatric comparison requires more careful evaluation.

Although the analysis identifies promising model systems, it does not demonstrate that canine tumors are interchangeable with human disease or that findings in dogs automatically translate into clinical outcomes in people. The study assesses transcriptomic similarity, one dimension of molecular correspondence, across a defined collection of tumor types and samples; the supplied evidence does not establish equivalence in every biological feature or treatment response. The authors describe their approach as scalable and capable of rapidly identifying model systems from transcriptomic data, offering a way to compare cancer biology across species.

Its practical value lies in mapping where canine and human tumors align, where differences persist, and which comparisons are most appropriate for subsequent comparative oncology research.

Subject of Research: Comparative transcriptomic analysis of canine and human cancers across 11 tumor types.

Article Title: Leveraging dog models to uncover human cancer insights

Article References: Daluwatumulle, G., Smith, L. A., Glen, N., Lee, J.-H., Seligson, N. D., Cahill, J. A., & Graim, K. (2026). Leveraging dog models to uncover human cancer insights. BMC Cancer. https://doi.org/10.1186/s12885-026-16813-6

Image Credits: AI Generated

DOI: 10.1186/s12885-026-16813-6

Keywords: comparative oncology, canine cancer, human cancer, transcriptomics, comparative genomics, pancancer analysis, pediatric cancer, glioma, sarcoma, cancer models, Leveraging, models