Hemangiosarcoma (HSA) is a highly malignant cancer originating from vascular endothelial cells. More frequent in dogs than in any other domestic species, with a reported prevalence of 2 % of all canine tumors, it is associated with a high fatality rate. Hemangiosarcoma typically affects older dogs, averaging 10 years of age at diagnosis, and a strong sex overrepresentation has not been identified.
Hemangiosarcoma (HSA) is a highly malignant cancer originating from vascular endothelial cells. More frequent in dogs than in any other domestic species, with a reported prevalence of 2 % of all canine tumors, it is associated with a high fatality rate. Hemangiosarcoma typically affects older dogs, averaging 10 years of age at diagnosis, and a strong sex overrepresentation has not been identified. While any canine breed can be affected with HSA, large and giant breeds are overrepresented, and Golden retrievers and German shepherds seem predisposed. Though not as frequent in cats, HSA is nevertheless occasionally diagnosed and recent studies helped describe the disease better.
While the strong breed association suggests a genetic predisposition in dogs, the definitive etiology of HSA remains elusive. Chronic ultraviolet (UV) irradiation is a recognized risk factor for dermal HSA (superficial, stage I) in lightly pigmented short-haired breeds of dogs (Whippet, Am Staff, etc.). Similarly, in recent studies, chronic UV light exposure was suggested to predispose to HSA of the conjunctiva (nictitating membrane) of both dogs and cats.
A study demonstrated overexpression of the STAT3 protein, a transcription factor playing a role in many human cancers (oncoprotein), to be common in canine vascular tumors, and a higher percentage of positive cells was observed in HSA than hemangiomas. In contrast, a study on 19 canine HSA samples failed to reveal Cox-2 positive immunostaining. Mutations in p53 and PTEN, two important tumor suppressor genes, were recently demonstrated in canine HSA, and may play a role in the progression of the disease.
Studies showed vascular endothelial growth factor (VEGF), one of the most potent angiogenic factors, to be elevated in the plasma and effusions of dogs with HSA when compared to healthy dogs. Similarly splenic HSA samples showed overexpression of angiopoietins, proteins playing a key role in the regulation of angiogenesis, when compared to normal canine spleen. Interestingly endostatin, a negative modulator of angiogenesis, was found to be elevated in the serum of dogs with HSA. In vitro research outlined the capacity of HSA cells to produce many angiogenic factors, including VEGF and the potent basic fibroblastic growth factor (bFGF). Furthermore, in vivo exposure to interleukin-12 (IL-12), a cytokine known to possess antiangiogenic properties, inhibited the growth of canine HSA in immunodeficient mice.
Arising from endothelial cells, HSA may involve any tissue in the body. The four most frequent primary sites in dogs are the spleen, heart (right atrium or auricle), skin or SQ tissues, and liver. In cats, the most common site is skin and SQ tissues, followed by visceral organs. Many other primary sites are known to occur. Metastatic dissemination and local infiltration occur early in the disease progression via local seeding following tumor rupture, or hematogenously. The most frequent sites of metastasis are the omentum, liver, and lungs. Being one of sarcomas most commonly spreading to the CNS, HSA has been found to cause brain metastasis detected in up to 15% of dogs on post-mortem examinations. Dermal and conjunctival HSA, possibly UV-induced, are associated with a better prognosis and lower metastatic rate than other locations. In general, a more advanced clinical stage predicts shorter survival times. When multiple sites are involved, it is often impossible to identify the primary site, and the prognosis is guarded, although occasional responses to therapy are observed. In cats, visceral HSA appears to be more frequently associated with metastasis than cutaneous and SQ disease, with 77% of cats having multifocal disease in one study.
Diagnosis and Staging
With stable patients, complete staging is mandatory at diagnosis because HSA is a highly metastatic disease and advanced clinical stage negatively impacts prognosis. When emergency surgery should be performed in a critical patient with suspected HSA, the clinical staging may be temporarily postponed. Close to 25% of dogs with HSA in one primary location may concomitantly have another primary HSA in a distant site (ex: spleen and right atrium), a process known as synchronous disease and possibly distinct from metastatic dissemination.
The recommended complete clinical staging includes three-view thoracic radiographs, abdominal ultrasonographic study, +/- echocardiography, in addition to a complete blood cell count (CBC), serum biochemistry, and urinalysis. The typical radiographic appearance of pulmonary metastatic disease with HSA is a coalescing military pattern. However, nodular or generalized military interstitial patterns can also be occasionally encountered. Abdominal ultrasonography may identify free abdominal fluid, visceral nodular lesions, and omental nodules when they are large enough. It is nonetheless crucial to remember that the appearance of any lesion is not diagnostic for HSA, and that benign lesions in the same locations (hematoma, hemangioma, ectopic spleens) may have an identical appearance. The preferred imaging technique to identify cardiac HSA remains echocardiography, and cardiac masses are best viewed when pericardial fluid is present. The CBC typically will shows changes suggesting microangiopathy, including regenerative anemia possibly with fragmented red blood cells (schistocytes), thrombocytopenia, and neutrophilic leukocytosis. When the spleen is affected, increased numbers of nucleated red blood cells (metarubricytes) are frequently observed, secondary to the decreased ability of the reticuloendothelial system lining the splenic sinusoids to promptly eliminate them from the circulating pool.
Additional imaging techniques may be used to detect primary or metastatic lesions based on clinical findings or suspicion. Plain bone radiographs or nuclear scintigraphy can help identifying skeletal lesions, while computed tomography (CT) or magnetic resonance imaging (MRI) will be useful for invasive subcutaneous or intramuscular HSA, as well as suspect CNS lesions. These advanced imaging modalities provide information on the invasiveness of the lesions in surrounding tissues, and help planning surgical excision and radiation therapy, as indicated. In addition, studies with CT and MRI have shown these techniques to be highly sensitive and specific in differentiating benign from malignant hepatosplenic processes. Other investigations are confirming the increased sensitivity of thoracic CT for early detection of pulmonary metastasis. These diagnostic techniques may soon become a standard component of the thorough clinical staging procedures in dogs and cats with HSA.
Biopsy and histopathology remain the only acceptable way to definitively diagnose HSA. Canine patients with hemoabdomen should not be euthanized following the observation of a splenic mass "typical of HSA", because nonmalignant splenic lesions, such as hematoma or hemangioma, appear identical in many cases, bear a good prognosis with splenectomy alone, and are often observed in the same breeds as HSA. Certain splenic malignancies, including indolent lymphomas and leiomyosarcoma, can also present with lesions mimicking HSA on ultrasound, but have a better prognosis with therapy. As a general rule with splenic nodules, approximately a third may be benign and curable with splenectomy alone. However, two studies of dogs presenting with nontraumatic hemoabdomen concluded that 60-70% were caused by HSA. A similar study on 65 cats showed that 28% of spontaneous hemoabdomen were caused by visceral HSA in that species, which made it the most common neoplasm to do so (60% of neoplastic causes). While cytology is a noninvasive and clinically valuable technique for the diagnosis of cancer in general, the diagnostic yield for HSA is reported to be low (25%), and the findings should not alter the therapeutic plan when a solitary lesion is deemed resectable.
Early detection of canine HSA with sensitive methodologies could potentially lead to the institution of therapy on a smaller tumor burden (lower stage), and an improved overall prognosis for survival. Novel detection methods may eventually lead to early diagnosis for high risk dogs, representing an improvement over the current standard diagnostic tests for diagnosing HSA (histopathology), usually after regional and distant metastasis has already occurred. Currently the use of such diagnostic techniques remains investigational.
The standard therapy for HSA involves the traditional therapeutic modalities used for most cancers: surgery, radiation therapy, and systemic chemotherapy. These can be used alone or in a multimodality setting.
Surgery – Whenever the measurable tumor burden can be safely resected, decreasing further risk of severe acute hemorrhage, disseminated intravascular coagulation or death, surgery is indicated. Splenectomy, liver lobectomy, excision of a dermal or resectable subcutaneous nodule, and right auriculectomy have all been described. In general surgery alone for visceral HSA is considered purely palliative, leading to median survival times (MST) averaging 1-3 months in dogs. A study of 23 dogs with cardiac HSA reported a MST of 43 days (n=8) and 118 days (n=15) following resection of right atrial or auricular HSA, respectively. In 8 dogs treated with adjuvant chemotherapy the MST was significantly longer (175 days) when compared to the 15 dogs treated with surgery alone (42 days). Partial or subtotal pericardectomy, via thoracotomy or thoracoscopy, can prevent life threatening cardiac dysfunction, but is unlikely to markedly prolong survival without the addition of systemic chemotherapy. When resection of superficial dermal or conjunctival HSA leads to complete excision, prolonged survival times will generally follow, in both dogs and cats, owing to the lower metastatic rate observed with these primary locations. Surgical removal of subcutaneous HSA without intramuscular involvement is associated with a much longer survival time (> 1000 days) than when muscle is involved (~270 days). In cats, visceral HSA has a guarded prognosis (< 3 months) when compared to cutaneous and SQ disease (> 1 year).
Conventional chemotherapy – Adjuvant doxorubicin-based chemotherapy has provided the best survival times so far. Typical protocols in dogs include single agent doxorubicin, doxorubicin with cyclophosphamide (AC protocol), and doxorubicin with cyclophosphamide and vincristine (VAC protocol). Reported survival times for canine splenic HSA range from 140 to 202 days for the various doxorubicin-based protocols, and none clearly stands out as superior. Attempts to prolong the MST with a dose intense protocol (doxorubicin every two weeks) or with intraperitoneal administration of the liposome-encapsulated form of doxorubicin (Doxil) did not result in significant survival improvements. Ifosfamide, an alkylating agent with efficacy on human sarcomas, was administered to 6 dogs with splenic HSA, resulting in a MST of 147 days. When ifosfamide was combined with doxorubicin post splenectomy to treat 27 dogs with HSA, a disappointing MST of 149 days was obtained. A response rate approaching 40% was recently described in dogs with inoperable subcutaneous HSA, but the median response duration was short (53 days). Dogs having a complete surgical removal after neoadjuvant doxorubicin had a median survival time of 207 days. Too few cats with HSA received chemotherapy, in the various recent studies, to draw any conclusions regarding its efficacy in that species. Additional or alternative therapeutic modalities are needed for longer survival times to be achieved.
Radiation therapy – There is little published information discussing radiation therapy for canine or feline HSA patients. Clinical responses (14 of 20 dogs in a study – 4 complete responses) and quality of life improvements have been observed with large fractions of megavoltage radiation therapy delivered to invasive and non-resectable subcutaneous or intramuscular HSA lesions. Radiation therapy is considered palliative in this setting, and will typically be combined with systemic doxorubicin-based chemotherapy.
Because traditional therapeutic modalities appear to have reached a plateau, with MST averaging 6-7 months with splenic HSA in dogs, it seems logical to investigate novel therapies that may lead to superior results. A better understanding of the genetic and molecular alterations in HSA cells may hopefully lead to identification of specific novel therapeutic targets.
Immunotherapy has long been an active field of cancer research, and HSA provides a great opportunity for even small improvements to be detected. Initial studies evaluating mixed bacterial vaccines failed to provide positive outcomes. Fifteen years ago, a prospective evaluation on canine HSA patients combined the AC protocol with liposome encapsulated muramyl tripeptide phosphatidylethanolamide (L-MTP-PE), a synthetic macrophage activator, and resulted in the only published significant improvement in MST of canine HSA over doxorubicin-based chemotherapy alone, at 273 days. However, while this result established a role for immunotherapy in treating canine HSA, the clinical use of L-MTP-PE has been hindered as the drug has yet to be commercially available. Combined administration of doxorubicin and an intraperitoneally delivered allogeneic vaccine derived from canine HSA cell lines resulted in a median survival time of 182 days in 13 dogs with stage II HSA. Further clinical trials with this vaccine are awaited.
Since angiogenesis plays a key role in tumors demonstrating an aggressive biological behavior, antiangiogenic therapy is an important field of investigational anticancer therapy in humans, and is the subject of ongoing clinical studies on canine HSA. Studies evaluating interleukin-12 (IL-12), interferon-α-2a (IFNα2a), thrombospondin-1 (TSP-1), HIV protease inhibitors, and thalidomide may eventually demonstrate benefits with improved MST. Targeted therapy inhibiting dysregulated signaling pathways is rapidly becoming a major class of anticancer therapy agents in human oncology. Humanized monocloncal antibodies targeting specific angiogenic proteins (ex: bevacizumab for VEGF) are unlikely to be helpful to dogs for obvious reasons, including the generation of neutralizing canine anti-human antibodies. Nevertheless, the active ongoing research on small molecule inhibitors against the same angiogenic proteins (VEGF, bFGF, integrins) raises cautious optimism for future use in veterinary oncology, without the abovementioned shortcomings.
As previously stated, mutated or deleted tumor suppressor genes may contribute to malignant transformation. Such genes may also be transiently silenced via epigenetic modifications, raising the possibility to use certain therapies aiming to reverse such alterations. A case report described a dog living more than 30 months following splenectomy for HSA, after receiving a histone deacetylase (HDAC) inhibitor as the sole adjuvant therapy. Larger prospective in vivo studies are required to determine if HDAC inhibitors will provide a significant survival advantage to dogs with HSA.
Canine HSA typically does not appear to overexpress Cox-2, contrary to most canine carcinomas surveyed so far and a few tumors of mesenchymal origin. However, because NSAIDs have also resulted in clinical responses in certain non-Cox-2 overexpressing tumors, and are inexpensive and relatively safe, it seems reasonable to combine their use to standard chemotherapy as an attempt to prolong the MST of dogs with HSA. A recent unpublished prospective pilot study failed to demonstrate an improvement in the overall MST of dogs with splenic HSA when a selective Cox-2 inhibitor was coadministered with adjuvant doxorubicin therapy.
Lastly, the administration of traditional cytotoxic chemotherapy agents can be revisited via novel methods or schedules of administration. A pilot study evaluated inhalational therapy with paclitaxel and doxorubicin, yielding encouraging results on dogs with pulmonary metastatic disease from HSA. Another novel approach, known as metronomic therapy, uses frequent low dosing of traditional cytotoxic chemotherapy agents. By so doing, the therapeutic target is shifted from the cancer cells to the endothelial cells, and the typical issues of toxicities and drug resistance can be largely eluded. A pilot study with 9 dogs evaluated low-dose oral chemotherapy combining piroxicam and cyclophosphamide and demonstrated an encouraging median survival time of 178 days in the adjuvant use, post-splenectomy.
Further prospective controlled studies are required, with larger numbers of patients, and some are underway combining conventional chemotherapy, metronomic therapy, and novel therapeutic agents such as protein tyrosine kinase inhibitors inhibiting VEGF.
Canine HSA remains a common, aggressive and highly metastatic cancer, and no marked improvement in survival times have been obtained in recent years. Hope remains for improved prognosis in the near future, and will rely on current research promising earlier diagnosis and successful novel targeted therapies. Recent studies have provided important information on the behavior of feline HSA, but further research and prospective studies are required in that species as well to improve diagnostic accuracy and therapeutic options.
Available upon request.