An aggressive type of cancer that originates from blood vessels is called hemangiosarcoma. Though it can affect any part of the body, dogs’ spleens, hearts, livers, or skin are most frequently affected.
Symptoms of hemangiosarcoma vary depending on the body systems involved. The abnormal blood and blood vessels that make up hemangiosarcoma tumors are delicate, invasive, and prone to rupture. Many of the clinical signs associated with hemangiosarcoma are caused when a tumor ruptures and bleeds into a body cavity like the chest or abdomen.
Although it has been reported in puppies as young as a few months old, hemangiosarcoma typically affects middle-aged to older dogs. Some breeds may be predisposed to hemangiosarcoma, including:
Hemangiosarcoma has a vascular (many blood vessels) nature, which allows it to spread throughout the body and cause cancer. The most common types of hemangiosarcoma are:
Clinical signs of hemangiosarcoma vary with the organs involved. Hemangiosarcoma is characterized by its propensity to bleed heavily, primarily because the tumors are composed of blood vessels and cells. The tumors are distantly aggressive, rapidly growing, and spread to other parts of the body in addition to being invasive at the primary tissue site.
Most causes of hemangiosarcoma are not known. Given the predisposed breeds, it most likely has a genetic component. UV exposure has been linked to skin hemangiosarcoma, particularly in thin, light-coated dogs.
Some chemicals, insecticides, toxins, and radiation have been associated with the emergence of hemangiosarcomas in people. Although there are no direct studies on dogs, it is also thought that these factors contribute to hemangiosarcoma in our animals.
Gene Expression Analysis Segregates Canine Hemangiosarcoma According to Breed
Although numerous human cancer cells have been found to have distinct gene expression profiles from their healthy counterpart cells (e g. Little work has been done to identify the gene expression patterns in canine tumors ([17]), for example. Furthermore, no research has been done to describe how heritable factors in any species affect these phenotypes. We previously demonstrated that based on gene expression profiles, hemangiosarcoma cells can be distinguished from non-malignant splenic hematoma cells (Tamburini et al., manuscript in preparation). Unsupervised clustering was used in this analysis to distinguish between tumors that originated from Golden Retrievers and those that did not (“type”:”entrez-geo”,”attrs”:”text”:”GSE15086″,”term_id”:”15086″GSE15086) To make sure there were no unintentional biases in the sample population, however, before we addressed potential differences between these two groups. 76 dogs with pathologically confirmed hemangiosarcoma, including 48 Golden Retrievers and 28 non-Golden Retrievers, provided blood samples for our study. Comparing the age at diagnosis (meanS) between the dogs in these two groups, there were no differences. D. = 9. 3±2. 6 and 8. 6±2. 6 years, respectively), gender (male vs. female, intact, or neutered), the primary tumor’s location, the quantity of dogs treated, or the outcome The population’s characteristics mirrored those previously discussed for both Golden Retrievers [15] and all dogs, regardless of breed [14], [19].
Every sample for which viable tumor tissue could be used to create at least short-term cell cultures was used in our gene profiling experiments (N = 10). In this subgroup, the mean ages of the Golden Retrievers (N = 6) and non-Golden Retrievers (N = 3) were 10 and 8, respectively. 3 years, respectively, while only male dogs made up the latter group. The final sample came from a male Golden Retriever/Great Pyrenees F1 dog (F1) who was 9 years old. When we divided the 10 tumor samples into groups where affected dogs were younger than 7 years vs. older dogs, age and gender as variables did not explain the observed clustering of the samples. older than 7 years or into male vs. Gene expression profiles in female dogs did not differ significantly from one another. However, a pattern persisted after the nine purebred dog tumor samples (aside from the F1 sample) were divided by breed. Golden Retriever and non-Golden Retriever samples were divided into separate groups by False Discovery Rate (FDR) analysis using a five-gene signature consisting of MHC DLA88, Forkhead box protein F1, thrombospondin-3 precursor, zinc finger protein 322A, and NAD(P) dependent steroid dehydrogenase. Given the relatively low expected discovery rate for this sample size, the relatively small number of genes was not unexpected because we expected that differences between hemangiosarcomas from dogs of different breeds would be subtle. The predicted true positive rate allowed us to take advantage of it and find more genes that were significantly different between the two groups at p 0 level. 001. is a heat map that shows the hierarchical clusters created by 12 known genes, 4 unknown genes, and 1 repeated gene (acid ceramidase) that was found by two different probes. An additional MHC gene, genes responsible for maintaining and replicating DNA, and genes that control cellular metabolism are all included in the list of known genes () Few significant changes were observed when gene differences were plotted according to their cytogenetic location. Samples from Golden Retrievers demonstrated a net increase in the sum of expression of genes located in CFA 3, CFA 25, and CFA 30, and a net decrease in the sum of expression of genes located in CFA 12, CFA 14, CFA 29, CFA 32, CFA 33, and CFA 34. This contrasted with the significant global underexpression observed when tumors were compared to non-malignant cells. Three females were included in the samples from Golden Retrievers, so it was expected that the group would exhibit a net rise in the expression of genes on the X chromosome. shows the location of specific genes that were consistently and noticeably over- or underexpressed in the Golden Retriever samples.
| Sample ID | Diagnosis | Breed | Sex | Age |
| CHAD G4 | Hemangiosarcoma | Golden Retriever | Male | 10 |
| CHAD G6 | Hemangiosarcoma | Golden Retriever | Female | 12 |
| CHAD G8 | Hemangiosarcoma | Golden Retriever | Male | 12 |
| FROG | Hemangiosarcoma | Golden Retriever | Female | 10 |
| JOURNEY | Hemangiosarcoma | Golden Retriever | Female | 11 |
| TUCKER | Hemangiosarcoma | Golden Retriever | Male | 6 |
| JOEY | Hemangiosarcoma | Rottweiler | Male | 9 |
| DD-1 | Hemangiosarcoma | Golden Retriever×Great Pyrenees | Male | 9 |
| CHAD P9 | Hemangiosarcoma | Portuguese Water Dog | Male | 9 |
| DAL-4 | Hemangiosarcoma | Dalmatian | Male | 7 |
| FOREST | Unaffected | Golden Retriever | Male | 10 |
| HANK | Unaffected | Golden Retriever | Male | 10 |
| TUX | Unaffected | Golden Retriever | Male | 10 |
| JASPER | Unaffected | Boxer | Male | 8 |
| T | Unaffected | German Shorthair Pointer | Male | 11 |
| INGO | Unaffected | Rottweiler | Male | 10 |
| QUANTUM | Melanoma | Golden Retriever | Male | 13 |
| CHESTER | Melanoma | Golden Retriever | Male | 14 |
| REP | Melanoma | Miniature Schnauzer | Male | 11 |
| BAXTER L | Osteosarcoma | Golden Retriever | Male | 8 |
| JAZZ | Osteosarcoma | Golden Retriever | Male | 7 |
| KODIAK | Osteosarcoma | Great Pyrenees | Male | 9 |
| STRETCH | Osteosarcoma | Greyhound | Male | 8.5 |
| NELLIE | T-cell lymphoma | Golden Retriever | Female | 6 |
| PUEBLO | T-cell lymphoma | Golden Retriever | Male | 11 |
| MURPHY | T-cell lymphoma | Boxer | Male | 9 |
| RUFFIAN | B-cell lymphoma | Boykin Spaniel | Male | 11 |
| Gene title | Chrom. | Function | Fold Change | p-value |
| similar to N-acylsphingosine amidohydrolase (acid ceramidase) 1 | 16 | Metabolic processing | 1.79 | 3.6E-04 |
| MHC class 1 DLA-88 | 12 | Cell-cell interaction | −603.8 | 1.7E-08 |
| similar to Thrombospondin-3 precursor | 7 | Cell-cell interaction | −2.08 | 1.1E-04 |
| similar to NAD(P) dependent steroid dehydrogenase-like | 5 | Cell-cell interaction | −1.91 | 1.7E-04 |
| MHC class 1 DLA-64 | 12 | Cell-cell interaction | −2.81 | 6.9E-04 |
| similar to Wiskott-Aldrich syndrome gene-like protein | 14 | Cell-cell interaction | −1.60 | 7.4E-04 |
| similar to staufen, RNA binding protein, homolog2 isoform LL (A) | 29 | Cell-cell interaction | −2.02 | 8.2E-04 |
| MHC class 1 DLA-88 | 12 | Survival/apoptosis | −603.8 | 1.7E-08 |
| similar to interferon stimulated exonuclease gene 29 kDa-like 1 | 3 | Survival/apoptosis | −2.23 | 5.1E-04 |
| MHC class 1 DLA-64 | 12 | Surivival/apoptosis | −2.81 | 6.9E-04 |
| similar to SWI/SNF-related matrix associated actin-dependent regulator of chromatin remodeling | X | Signaling/cell cycle | 3.42 | 5.7E-04 |
| similar to Structural maintenance of chromosomes4-like 1 protein | 34 | Signaling/cell cycle | 1.72 | 6.8E-04 |
| similar to Forkhead box protein F1 | 5 | Transcription | −1.65 | 9.3E-05 |
| similar to zinc finger protein 322A | 35 | Transcription | −1.68 | 1.5E-04 |
| similar to SWI/SNF-related matrix associated actin-dependent regulator of chromatin remodeling | X | Transcription | 3.42 | 5.7E-04 |
| MHC class 1 DLA-88 | 12 | Immune response | −603.8 | 1.7E-08 |
| similar to interferon stimulated exonuclease gene 29 kDa-like 1 | 3 | Immune response | −2.23 | 5.1E-04 |
| MHC class 1 DLA-64 | 12 | Immune response | −2.81 | 6.9E-04 |
| Transcribed locus [Cfa.6637.1.A1_at] | 9 | Unknown | −6.10 | 5.3E-04 |
| Transcribed locus [Cfa.14890.1.A1_at] | 28 | Unknown | 2.82 | 9.8E-04 |
| — [CfaAffx.1401.1.S1_at] | 1 | Unknown | −1.59 | 8.4E-04 |
| — [Cfa.11358.1.A1_at] | 16 | Unknown | 2.02 | 1.0E-03 |
To validate the microarray data, we performed reverse transcriptase PCR followed by quantitative real-time PCR analysis on the expression of DLA-88 (MHC), TSP-3, and SMARCA-1 (SWI/SNF). For this analysis, we took into consideration all four dogs: Dal-4, Joey, CHAD-P9, and three Golden Retrievers: CHAD G6, CHAD G4, and Frog. The genes were chosen due to their potential role in the definition of MHC haplotypes or due to their relevance to tumor biology; i e. , SMARCA-1 is an ATP-dependent chromatin remodeler important for the regulation of transcription, DNA replication, and DNA repair that is abnormally expressed in some tumors [22], TSP-3 homologues are known to regulate angiogenesis [21], and DLA-88 is an MHC class I gene. and demonstrate that SMARCA-1 was consistently overexpressed while TSP-3 and DLA-88 were consistently underexpressed in Golden Retriever hemangiosarcomas. SMARCA1 expression was actually highest in cells from CHAD G4, a male dog (the middle dog in the Golden Retriever group), even though this latter gene is encoded on the X chromosome, suggesting that this is not just a female bias. P-values, a measure of statistical significance, were calculated using the Students T-test for equal variance () We were able to ask intriguing, if anecdotal, questions thanks to the availability of a sample coming from an F1 mix-breed dog with luckily known parentage (Golden RetrieverGreat Pyrenees). When we included this sample in the hierarchical clustering, the features that distinguished the two groups were less distinct: was this dog more similar to Golden Retrievers, to non-Golden Retrievers, or would it reflect a “mixture” of both? Among the 35 genes with the lowest p-values (p0), only 7 of the 17 signals on the hit list were still present. 0123), and 11 of the 17 were discovered among the top 200 (p 0 04). This implied that “Golden Retriever” affected this F1 dog tumor’s gene expression signature, but not entirely. demonstrates that the expression of TSP-3 and MHC DLA-88 in the F1 (Golden Retriever mix) were similar to the Golden Retriever group and the non-Golden Retriever group, respectively. As a result, the dogs’ Golden Retriever and non-Golden Retriever backgrounds modulated the expression of genes in the tumor in a predictable manner.
That hierarchical clustering by breed reflected distinct traits of genetic variants within the breed, rather than a specific influence of breed on tumor phenotypes, is one explanation for why Golden Retrievers separate from non-Golden Retrievers in this analysis. Since breed and MHC haplotypes have not been reported to be correlated, this was unlikely. However, we investigated whether the association between breed (Golden Retriever) and TSP-3, DLA-88, or SMARCA-1 expression would persist in non-hemangiosarcoma samples. Blood leukocytes from healthy Golden Retrievers and non-Golden Retrievers, from Golden Retrievers and non-Golden Retrievers who were diagnosed with another cancer (melanoma, non-Hodgkin lymphoma, or osteosarcoma), and from each affected dog’s hemangiosarcoma cells were among the samples examined. TSP-3, MHC DLA-88, or SWI/SNF (SMARCA1) expression in blood samples from cancer-free dogs and dogs with other types of cancer did not differ significantly between the two groups. TSP-3 and DLA-88 expression, however, was consistently lower and SMARCA1 expression, was consistently higher in hemangiosarcoma samples from Golden Retrievers as compared to non-Golden Retrievers (p0. 03). One intriguing finding is that although these genes’ ranges of expression were narrow in hemangiosarcoma samples and blood samples from healthy dogs, they were relatively wide in blood samples from dogs with tumors other than hemangiosarcoma. However, the TSP-3 and DLA-88 expression trends are reversing in these samples. This suggests that the differences were not caused by breed variations but rather by the breed’s genetic background and its effects on the phenotypes of hemangiosarcoma. We tested whether these genes had significantly different calls by comparing our hemangiosarcoma Golden Retriever expression arrays to expression arrays from lymphoma and leukemia (30 Golden Retrievers) and from osteosarcoma (9 Golden Retrievers), as another possibility was that this difference would only be reflected on tumor samples. These analyses confirmed the link between hemangiosarcoma and acid ceramidase overexpression, but neither TSP-3, DLA-88, nor SMARCA1 showed breed-dependent differential expression in lymphoma, leukemia, or osteosarcoma, even though those samples also appear to have distinct and unique sets of genes whose expression varies as a function of breed (TSP-3, DLA-88, SMARCA1). Phang, K. Gavin, A. Sarver, and J. Modiano, unpublished data).
| Tissue Type | Average fold change of TSP-3 GR vs nGR (Mean [Range]) | p-value | Average fold change of MHC GR vs nGR (Mean [Range]) | p-value | Average fold change SWI/SNF (SMARCA1) GR vs nGR (Mean [Range]) | p-value |
| Hemangio-sarcoma (tumor) | 0.47 [0.39–0.56] | 0.025 | 0.16 [0.07–0.26] | 0.029 | 3.18 [2.04–4.32] | 0.028 |
| Healthy (blood) | 0.84 [0.52–1.16] | 0.623 | 1.14 [0.98–1.30] | 0.783 | 0.78 [0.40–1.15] | 0.571 |
| Tumors (blood) | 3.95 [0.82–7.08] | 0.558 | 6.93 [2.49–11.37] | 0.279 | 22.60 [1.98–43.22] | 0.174 |
Although the exact cause of this cancer is unknown, a combination of genetic and environmental factors is thought to be responsible. Exposure to sunlight is regarded as a significant risk factor in animals who experience the skin (cutaneous) version. Unfortunately, the prognosis for most hemangiosarcoma patients is poor. Even with aggressive treatment, median survival times typically only last around six months.
The cost of this illness can be quite high because, like many cancers, it requires surgery for both diagnosis and treatment and chemotherapy for the best results. Dogs with hemangiosarcoma may even incur diagnosis and treatment costs that reach the tens of thousands of dollars, despite the fact that it is rare.
50% of dogs with hemangiosarcoma, a common cancer of middle-aged to older dogs, develop the disease in the spleen. In fact, its the most common tumor of this organ. Less frequently, this cancer can cause tumors in the heart, the pericardium, the skin, and the tissues just beneath the skin. The lungs, kidneys, mouth, muscle, bone, bladder, and uterus are additional locations.
Unless the tumor only appears to affect the lining of the heart, dogs with heart tumors are typically not surgically treated. To help keep the heart pumping normally, these are occasionally treated by a procedure known as a pericardectomy (removal of the heart’s lining). Chemotherapy is often recommended for these internal hemangiosarcoma patients, too.
IntroductionDiligent clinical research over the past 50 years has allowed veterinarians to adapt or develop protocols to treat companion animals with cancer, providing pet owners reasonable options ranging from palliative care to therapies with curative intent. In addition to adapting protocols using conventional modalities (surgery, radiation, and cytotoxic chemotherapy), targeted drugs and immunotherapy approaches have been developed specifically for the veterinary market [
Any tissue or organ with vascular structures can develop HSA. The right atrium of the heart, the spleen, the subcutis, and the dermis are the most typical primary sites for HSAs. Visceral HSAs occur more frequently than cutaneous HSAs and have a worse prognosis.
FAQ
Why are so many dogs getting hemangiosarcoma?
In most cases, the cause of hemangiosarcoma is unknown. Some dogs are susceptible to developing skin tumors from sunlight exposure, especially in areas with thin coats like the belly, inner thighs, and eyelids.
How does dog hemangiosarcoma start?
There is mounting proof that the cancer cells responsible for hemangiosarcoma begin in the bone marrow before quickly migrating to other parts of the body. The heart and spleen, the two most typical locations for this type of tumor to be found, are frequently where hemangiosarcoma is first discovered.
Are tumors in dogs hereditary?
Unfortunately, many cancers in dogs have genetic causes, so owners can do little to prevent it.
Can hemangiosarcoma in dogs be prevented?
Internal hemangiosarcoma cannot currently be prevented by any known means. However, cutaneous (skin) hemangiosarcoma can be prevented by avoiding ultraviolet light or by covering areas of the body with fine hair with a sunscreen designed for pets.