Simple Blood Test Shows Promise for Detecting Multiple Cancers in Horses

Diagnosing cancer in a horse has never been easy. Skin tumors such as sarcoids, squamous cell carcinomas and melanomas can often be recognized from their appearance and confirmed with a biopsy, but internal malignancies like lymphoma, adenocarcinoma and hemangiosarcoma hide deep in the chest and abdomen, where standard imaging frequently fails. Horses are simply too large to fit into the computed tomography and magnetic resonance scanners used routinely in small-animal practice, and radiographs and ultrasound rarely provide a definitive answer. Exploratory surgery and organ biopsies carry anatomical, financial and ethical obstacles, and in some cases an internal tumor is only confirmed after death. Against this backdrop, an international team of researchers has now reported that a simple blood test measuring a cell proliferation enzyme could transform how veterinarians approach equine cancer diagnosis, offering a non-invasive window into disease that previously required invasive procedures.

The new study, published in the Springer journal Veterinary Oncology, evaluated two candidate blood biomarkers across an unusually broad spectrum of equine tumors. Serum thymidine kinase 1, or TK1, is an enzyme central to DNA synthesis that becomes upregulated in rapidly dividing cancer cells and leaks into the circulation, where its activity can be measured. It has already attracted attention in human oncology and in canine medicine, particularly for hematopoietic cancers. Serum amyloid A, or SAA, is the major acute-phase protein of the horse, produced by the liver in response to inflammatory cytokines and widely used in equine practice to monitor infection and inflammation, yet it had never been systematically assessed as a tumor marker. The researchers, led by Lucia Unger of the Swiss Institute of Equine Medicine in Bern together with colleagues at the Swedish University of Agricultural Sciences, set out to determine whether either molecule could reliably distinguish tumor-bearing equids from healthy and sick but cancer-free animals.

The team drew on the tumor biobank of the Institut Suisse de Médecine Equine, assembling serum samples from 91 equids with histologically or clinically confirmed cancers and 149 control horses. The tumor cohort spanned eight groups: 14 cases of lymphoma, 8 of (adeno)carcinoma, 4 of hemangiosarcoma, 8 of other solid internal tumors, 11 of ovarian granulosa cell tumor, 14 of sarcoid, 17 of squamous cell carcinoma and 15 of melanoma. Crucially, the control group was split into 42 healthy horses and 107 horses suffering from non-neoplastic diseases, a design choice that directly tested whether the biomarkers rise because of cancer itself or merely because of general illness and inflammation. All 240 samples were analyzed for TK1 activity using a radiochemical assay with tritium-labeled thymidine as the natural substrate, while SAA concentrations were measured by enzyme-linked immunosorbent assay in a subset of 112 samples.

The technical details of the TK1 assay matter, because the field has been burned before. A commercially available chemiluminescence immunoassay that uses the synthetic substrate azidothymidine instead of thymidine has previously failed to diagnose equine lymphoma reliably, an outcome now explained by the finding that equine TK1 reacts with azidothymidine at only about 55 percent of the rate measured with the natural substrate. The tritium-labeled thymidine approach avoids this pitfall. In the new work, tumor samples were run in triplicate across two independent assay runs, producing six measurements per sample, and the investigators applied rigorous statistical testing, including Mann-Whitney U comparisons with Bonferroni correction for multiple comparisons and effect size calculations to quantify the magnitude of differences between groups.

The headline result was striking. Serum TK1 activity was significantly elevated in equids with lymphoma, other solid internal tumors, sarcoid, squamous cell carcinoma and melanoma compared with both healthy controls and controls with non-neoplastic comorbidities, with effect sizes ranging from moderate to large. In contrast, horses with (adeno)carcinoma, hemangiosarcoma or ovarian granulosa cell tumor generally did not show significant elevations, with the exception of hemangiosarcoma cases, which had modestly higher TK1 than the diseased control group. Receiver operating characteristic analysis reinforced the pattern: the area under the curve, a standard measure of diagnostic accuracy, was rated excellent for lymphoma, hemangiosarcoma, other solid internal tumors, sarcoid, squamous cell carcinoma and melanoma, with values between 0.90 and 0.98 in comparisons against both control groups, and good for ovarian granulosa cell tumor at around 0.75. Only carcinoma fell into the poor-to-fair range and did not reach statistical significance, indicating that TK1 is not currently a reliable marker for this heterogeneous group.

To translate laboratory statistics into clinical meaning, the researchers calculated post-test probabilities for the three tumor types with known prevalence in comparable equine populations: lymphoma at 0.002 to 0.5 percent, sarcoid at 19 percent, and melanoma at 40 to 80 percent in grey horses. For lymphoma, where the disease is extremely rare, a positive TK1 result only modestly increased the probability of cancer and a negative result barely lowered it, meaning the test cannot serve as a population-wide screening tool. For sarcoid and melanoma, however, where pre-test probability is moderate to high, a positive result substantially increased the likelihood of disease and a negative result reduced it essentially to zero, regardless of whether controls had other illnesses. In practical terms, for a grey horse with a suspicious lump or a horse whose biopsy is impractical, a TK1 measurement could meaningfully sharpen the diagnostic picture.

SAA told a very different story. Despite its central role as the horse’s principal acute-phase protein, SAA concentrations did not differ significantly between any of the eight tumor groups and the diseased control horses. The finding underscores a fundamental limitation of inflammation-based markers: tumors do not always provoke the systemic inflammatory response that drives hepatic acute-phase protein production, and when inflammation is present, non-neoplastic conditions produce it just as vigorously. The result parallels mixed evidence from other species, where C-reactive protein shows promise in canine cancer but SAA has not been validated as a relapse marker in dogs with multicentric lymphoma, even though cats with lymphoma do show elevated SAA that declines with successful treatment. For equine practitioners, the message is that SAA remains an excellent inflammation marker but has no place as a cancer diagnostic.

The study also surfaced intriguing biological nuances. TK1 activity correlated weakly but significantly and negatively with age among control horses, with younger animals showing higher levels, possibly reflecting greater cellular turnover in growing tissue. Because horses affected by sarcoids were younger than the healthy control group, the authors examined whether age could explain their elevated TK1, but the effect persisted when sarcoid cases were compared with the more age-matched diseased control group, supporting a genuine association with the tumor. One horse with multicentric lymphoma and extensive metastases displayed exceptionally high TK1 activity of 17.72 pmol/min/mL, echoing human data in which serum TK1 rises with metastatic burden in melanoma, lung, gastric and colorectal cancers. In dogs with lymphoma, serial TK1 monitoring tracks disease progression and treatment response, raising the prospect that repeated measurement could one day serve a similar prognostic role in horses. Meanwhile, the three highest values among diseased controls came from horses with severe inflammatory conditions, including extensive necrotic lipoma, cellulitis and endotoxemia with laminitis, a cautionary observation that inflammation may occasionally confound the test.

The authors are careful about the limits of their work. Several tumor groups, particularly carcinoma, hemangiosarcoma and other solid internal tumors, were small and heterogeneous, making those results exploratory, and no power calculation was possible for SAA. Yet for lymphoma the minimal required sample size was met, and the findings for sarcoid, squamous cell carcinoma and melanoma rest on the largest cohorts. The bigger obstacle is practical: the tritium-labeled thymidine assay requires radiochemical laboratory infrastructure that few veterinary diagnostic services possess, while the commercial chemiluminescence alternative is unreliable in horses. The researchers call for the development of a horse-specific TK1 ELISA or point-of-care test that could bring the assay into routine practice, and for multicenter studies to validate the cut-offs in larger, prospectively recruited cohorts.

If those steps succeed, the implications for animal welfare are considerable. A validated blood test would spare horses invasive biopsies and exploratory surgeries, help clinicians decide when treatment is justified, and potentially enable monitoring of recurrence in tumors like sarcoids, which recur unpredictably after treatment. The study also clarifies that TK1 is not a lymphoma-specific marker but a broad proliferation signal shared across equine neoplasia, mirroring its behavior in human oncology where it performs best in combination with other tests rather than alone. For a field that has long lagged behind small-animal and human cancer diagnostics, this work marks a credible first step toward non-invasive, blood-based tumor detection in horses, built on a decade of careful enzyme biochemistry and one of the largest equine tumor serum collections ever assembled.

Subject of Research: Evaluation of serum thymidine kinase 1 and serum amyloid A as blood-based tumor biomarkers in equids

Article Title: Evaluation of serum thymidine kinase 1 and serum amyloid A as tumor biomarkers in equids: a comparative cross-sectional study

Article References: Unger, L., Rayhan, A. B. H., Sharif, H., Gerber, V., Rönnberg, H., Eriksson, S., & Wang, L. (2026). Evaluation of serum thymidine kinase 1 and serum amyloid A as tumor biomarkers in equids: a comparative cross-sectional study. Veterinary Oncology, 3(1), Article 6. https://doi.org/10.1186/s44356-026-00057-4

Image Credits: AI Generated

DOI: 10.1186/s44356-026-00057-4

Keywords: thymidine kinase 1, serum amyloid A, equine cancer, biomarkers, lymphoma, sarcoid, melanoma, squamous cell carcinoma, veterinary oncology, liquid biopsy, diagnostics, horses