A large VetCompass study has estimated the frequency of haemangiosarcoma diagnosis in dogs under UK primary veterinary care and identified several breed, age and bodyweight associations that may be relevant to first-opinion practice.
The paper, published in Veterinary Oncology, analysed electronic health records from more than 2.25 million dogs under primary veterinary care in 2019. The authors reported an annual incidence risk of 0.036% across dogs of all ages, rising to 0.074% in dogs aged five years and older.
Haemangiosarcoma can be difficult to recognise early, especially in visceral cases where presenting signs may be non-specific. Dogs may present with lethargy, intermittent weakness, collapse, abdominal signs or other findings that can initially overlap with more common conditions.
Knowing which dogs are at higher odds of diagnosis may help vets keep haemangiosarcoma in mind when assessing older patients, particularly larger dogs or breeds with identified associations. It may also help frame discussions with owners about why further investigations, such as imaging, blood tests, cytology, biopsy or referral, may be appropriate when the clinical picture is concerning.
The authors note that identifying breeds at increased odds can support earlier clinical suspicion and more targeted diagnostic work-up, especially for visceral haemangiosarcoma where early signs may be vague and easily attributed to other conditions.
The study used VetCompass records from UK primary care practices. The researchers identified dogs with a recorded final diagnosis of haemangiosarcoma during 2019, including cases with and without histopathological confirmation. They then ran two nested case-control studies: one including all clinically diagnosed cases and one limited to histopathologically confirmed cases. Risk factor modelling focused on dogs aged five years or older.
The risk factor analysis compared 788 clinically diagnosed cases with 1,064,187 controls. The histopathological analysis included a smaller subset of 273 incident cases confirmed by histopathology.
Of 2,250,741 dogs of all ages, 801 were diagnosed with haemangiosarcoma in 2019, giving an annual incidence risk of 0.036%. In dogs aged at least five years, there were 790 incident cases among 1,066,381 dogs, giving an annual incidence risk of 0.074%.
In the clinical diagnosis model, ten breeds had increased odds of haemangiosarcoma diagnosis compared with crossbreed dogs. The breeds with the highest odds were Dogue de Bordeaux, Flat Coated Retriever, German Shepherd Dog and Hungarian Vizsla.
In the histopathologically confirmed subset, five breeds had increased odds compared with crossbreeds: Flat Coated Retriever, Hungarian Vizsla, German Shepherd Dog, Beagle and Boxer. The Flat Coated Retriever had the highest odds in this subset.
Age was also associated with diagnosis. Compared with dogs aged 7 to under 9 years, dogs aged 11 to under 13 years had the highest odds of clinical haemangiosarcoma diagnosis. In the histopathological model, dogs aged 9 to under 11 years and 11 to under 13 years had increased odds compared with the 7 to under 9 year reference group.
Bodyweight also mattered. When bodyweight was assessed in place of breed, the odds of clinical haemangiosarcoma diagnosis increased across bodyweight categories above 15kg, with the highest odds in dogs weighing 37.5kg or more. A similar pattern was seen in the histopathologically confirmed analysis.
This was an electronic health record study, so the data were not originally collected for research purposes. Cases may have been missed if haemangiosarcoma was not recorded in the clinical notes, or if it appeared only as a differential diagnosis rather than a final diagnosis.
Only around 35% of incident clinical cases had histological confirmation. That reflects the realities of primary care, where some owners may not pursue invasive or costly diagnostics, particularly if disease is advanced or prognosis is poor. However, it also means that some clinically diagnosed cases could have been misclassified.
Some breed findings need particular caution. The authors note that none of the haemangiosarcoma diagnoses in Dogue de Bordeaux or Rottweiler dogs were supported by histopathological confirmation. They suggest this could reflect reduced surgical intervention in some larger breeds, potentially because of cost, or a diagnostic bias where very large dogs are more likely to be diagnosed clinically based on suspicion alone.
The histopathological analysis was smaller and may have been underpowered for less common breeds. The risk factor analysis was also restricted to dogs aged five years or older, making the findings most relevant to mature and senior dogs rather than younger patients.
This study does not mean that every older, large-breed or predisposed dog with vague signs has haemangiosarcoma. It also does not show that breed or bodyweight should override clinical judgement.
What it does provide is useful UK primary-care evidence to support risk-aware decision-making. In dogs with compatible signs, especially older or larger dogs and breeds identified in this paper, haemangiosarcoma may deserve a place on the differential list earlier in the work-up.
For GP vets, the paper may be particularly useful when explaining diagnostic recommendations to owners. Breed, age and bodyweight cannot diagnose haemangiosarcoma, but they can help explain why further investigation may be warranted when clinical signs are concerning.
The study also reinforces the importance of being careful with apparently simple risk messages. Some findings were stronger in clinically diagnosed cases than in histopathologically confirmed cases, and some associations may be influenced by owner decisions, insurance status, cost, diagnostic access or clinician expectations. The practical message is therefore not “these dogs will get haemangiosarcoma”, but rather “these dogs may justify a lower threshold for considering it”.
Source: Barry GJ, O’Neill DG, Guillén A, et al. Epidemiology of haemangiosarcoma diagnosed in dogs under primary veterinary care in the United Kingdom: Frequency and risk factors. Veterinary Oncology, 2, 26 (2025).