Osteosarcoma and hemangiosarcoma: The ugly sarcomas (Proceedings)


Osteosarcoma (OSA) and hemangiosarcoma (HSA) are highly malignant tumors that have both a locally invasive and highly metastatic biologic behavior. Multimodality therapy must be employed to provide patients with these cancers the best chance for improved quality and quantity of life.

Osteosarcoma (OSA) and hemangiosarcoma (HSA) are highly malignant tumors that have both a locally invasive and highly metastatic biologic behavior. Multimodality therapy must be employed to provide patients with these cancers the best chance for improved quality and quantity of life. Despite these therapies, however, dogs with OSA or HSA generally succumb to their disease within 6-12 months. This presentation will highlight new information regarding diagnostics and therapeutics for these aggressive tumors.


Appendicular OSA is the most common primary bone tumor in dogs, accounting for up to 85% of all primary bone tumors. There is a bimodal peak in age of occurrence, at 2 years of age and middle-aged to older dogs. Males are reported to be over-represented with a male:female ratio of 1.5:1. Large to giant breed dogs are at greater risk with Irish setters, St Bernards, Rottweilers, etc. likely to be affected. One study found a mutation in a proto-oncogene (MET) in 70% of Rottweilers; this mutation was found in < 5% of all other breeds examined. This breed-specific finding suggests a heritable mutation with the Rottweiler breed.


 • Radiographs of the primary lesion will show a mix of bone proliferation and lysis; the degree of each finding can vary widely between tumors. The long bone locations for the tumors are at the metaphyseal region: away from the elbow, but towards or away from the knee (i.e. in the rear leg the tumor may arise at either end of the femur or tibia). Thoracic radiographs will show metastasis in only ~8% of dogs at diagnosis, but microscopic metastasis is present in another 80%. Finding visible metastatic lesions greatly changes the prognosis, as visible nodules do NOT respond to chemotherapy. Thus, a complete metastasis check with 3 radiographic views (right and left lateral and a DV) is critical. CT scan is a more sensitive way to detect nodules; the response to chemotherapy of CT visible yet non-radiographically visible nodules is unknown.

 • Fine needle aspirates may be diagnostic in the majority of cases. OSA is easily exfoliative and reveals large, immature mesenchymal cells that may be producing osteoid (pink matrix). OSA may have a plasmacytoid appearance, with round-ish margins and deep blue cytoplasm. Alkaline phosphatase staining is a highly sensitive and specific way to determine if a malignant cell is of bone origin. If cellularity is poor, a biopsy may be needed. The least invasive technique is with a Jam Shidi bone core biopsy instrument; most dogs are not significantly more lame after this type of biopsy.

 • Serum alkaline phosphatase is prognostic in that dogs with elevated levels have shorter survival times by 50% even when treated aggressively with surgery and chemotherapy.

Treatment and Prognosis

The treatment option associated with the longest disease control and overall survival times is surgical amputation followed by 4 to 6 doses of a platinum based chemotherapy drug (cisplatin or carboplatin). Doxorubicin may also be used; it may be slightly less effective. With amputation alone, median survival times range from 3-5 months; chemotherapy post-surgery will extend that to a median survival of 10-12 months. While combination chemotherapy would seem to offer benefits over single-agent treatments, no studies to date have shown strong support for the use of combination chemotherapy. Survival times with such combinations remain similar to single agent treatments, and toxicities are higher. As an alternative to amputation for local tumor control, limb salvage procedures, which remove the tumor but spare the limb, can be effective for tumors at the distal radius.

Palliative therapy provides median survival times from 3-6 months. One of the most affective palliative treatments is coarse (large) doses of radiation (RT) given weekly for 3-4 total doses. A recent paper showed good benefits in dogs receiving 2 doses of RT given daily on sequential days. Following RT, improved function can occur in up to 75% of patients. Interestingly, while dogs treated with amputation alone succumb to pulmonary metastasis at a median of 3-5 months, dogs treated only with palliative RT are typically free of pulmonary metastasis at the point at which they succumb to local pain, also at roughly 4 months. Studies have shown that the primary OSA tumor secretes anti-angiogenic factors that are likely involved with the suppression of growth of metastatic lesions. Thus, chemotherapy may not have a role in disease control in dogs treated palliatively with the tumor not removed.

Pain medications alone are generally not very effective at controlling bone pain. Combinations of a nonsteroidal antiinflammatory drug, tramadol, and gabapentin may provide increased comfort for some amount of time. These drugs are often used in conjunction with RT.

Bisphosphonates are widely used in humans with lytic bone disorders. These drugs inhibit osteoclast activity and thus may help decrease boney lysis at the tumor site. Other actions include induction of apoptosis of osteoclasts and potentially malignant osteoblasts; they may also alter the microenvironment within the bone. Interestingly, 3 groups of investigators have shown that bisphosphonates also have direct anti-cancer effects on OSA cell lines in tissue culture. The anticancer effect of bisphosphonates in OSA patients is under investigation. Pamidronate (1-2 mg/kg IV over 2 hrs in 250 mls 0.9% NaCl once every 4 weeks) alone increases comfort in about 1/4 of dogs with OSA. When pamidronate was given with palliative radiation and chemotherapy, an increase in pain control was not seen over radiation/chemotherapy alone. Zoledronate (0.1 mg/kg IV over 15 minutes in 60 mls 0.9% NaCl), a next generation bisphosphonate, showed a 50% response in dogs with OSA as a single agent. Preliminary findings of an ongoing study with zoledronate combined with palliative radiation are very encouraging. Oral bisphosphonates are very poorly absorbed in the dog and should not be used.

Immunotherapy using L-MTP-PE, a liposome-encapsulated immunostimulant derived from a bacterial cell wall, increased survival time to a median of 14.5 months when used after amputation and 4 doses of cisplatin. This drug, known as mifamurtide, has recently received approval in Europe and thus may be available for purchase for patients in the United States in the future.


About 50% of HSA are splenic, about 25% are right atrial, and 25% are cutaneous. Other primary sites include liver, kidney, bone, bladder, and brain. Middle-aged to older (8-10 years) dogs are at risk, with no sex predilection. Overrepresented breeds include German shepherds, boxers, Labrador retrievers, and golden retrievers. A recent study using gene expression profiling of HSA showed significantly different clustering of genes in golden retriever tumors vs. non-golden retriever tumors for a variety of transcription factors, survival factors, and pro-inflammatory and angiogenic genes including vascular endothelial growth factor receptor 1 (VEGFR1). These results support the theory that heritable factors affect gene expression and thus biologic behavior in natural cancers.

Differentials for splenic mass

As hemangiosarcoma carries a very guarded prognosis, and it is the most common tumor affecting the spleen, it is important to remember that benign splenic masses occur not infrequently in dogs. As a general rule, 1/3 of splenic masses in dogs are benign (hematomas), 2/3 are malignant. 2/3 of the malignant ones are HSA, with the other 1/3 comprised of an assortment of other tumors including but not restricted to leiomyosarcoma, lymphoma, plasma cell tumor, and histiocytic tumors. The odds of malignancy increased to 3/4 in one study if the mass had ruptured and the patient was anemic. A more recent study of 71 dogs with a splenic mass and hemoabdomen needing a blood transfusion confirmed the findings of 3/4 of the masses being malignant, with 1/4 being benign. 93% of the dogs with malignancy had HSA; these dogs had significantly lower total solids and platelet concentrations than dogs with benign disease. Additionally, hemoabdomen can also be caused by a number of diseases; a recent review of the final diagnosis in 60 dogs with hemoabdomen showed 63% HSA (splenic and/or hepatic, less frequently omentum or kidney), 27% splenic hematoma, 5% splenic torsion, 3 % hepatocellular carcinoma, and 1.6% carcinomatosis. Thus, in this patient population, benign disease accounted for 33% of the cases of hemoabdomen; a finding generally considered very suspicious for HSA.

Differentials for pericardial effusion

Most (60 to 70%) clinically significant pericardial effusions are secondary to cancers, including right atrial hemangiosarcoma, heart base tumors (aortic body tumors, or chemodectomas), and mesothelioma. Idiopathic (benign) pericardial effusions make up about 30% of cases. Hemangiosarcoma is the most common cardiac tumor (10x greater than next tumor). They are most often solitary, affecting the right auricle or atrium. The tumor itself can bleed, or it can erode through the atrial free wall leading to hemorrhage into the pericardial sac and subsequent cardiac tamponade and right heart failure. Cardiac troponin I (cTnI) in the serum is a sensitive and specific marker for myocardial ischemia and necrosis and was found to be significantly elevated in dogs with pericardial effusion due to HSA (2.77 ng/dL; range: 0.09–47.118 ng/dL) compared to dogs with idiopathic effusion(0.05 ng/dL; range: 0.03–0.09 ng/dL). This serum test may be useful in differentiating dogs with HSA versus idiopathic effusion in cases where a mass cannot be detected.


Fluid that is removed can be evaluated via a hematocrit tube for PCV and total protein; if the fluid is hemorrhagic the sample should be watched to see if clotting occurs. If it does, it is a fresh blood sample and not from free abdominal fluid; another sample should be obtained. Hemorrhagic fluid (PCV > 15 or so) very rarely reveals anything noteworthy on cytology. In critical cases, looking at a slide of the fluid under a microscope is usually not worth the time. If the sample is not hemorrhagic, cytologic analysis can be very helpful.

Imaging for these tumors

Ultrasound is a highly sensitive, noninvasive technique to image splenic and hepatic lesions. With HSA, these masses usually have a mixed pattern with anechoic to hyperechoic areas. "Target lesions" in the liver are most consistent with metastatic neoplasia. However, there are many times when lesions in the liver have an appearance that may be benign or malignant. Contrast harmonic imaging (or "contrast ultrasound") involves administration of an IV ultrasound contrast medium (such as Definity or Sonovue) with subsequent imaging with contrast harmonic software on a conventional ultrasound machine. In one study of 32 dogs with hepatic nodules, a significant association was found at peak contrast enhancement between hypoechoic appearance and malignancy. Benign nodules were isoechoic to the surrounding normal liver at peak contrast enhancement. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were 100%, 94%, 94%, 100%, and 97%, respectively. Furthermore, a report of contrast ultrasound in 3 dogs with normal liver appearance on routine ultrasound revealed indistinct nodules in all 3 that were histologically confirmed to be HSA. Thus contrast ultrasound may not only be able to differentiate benign from malignant liver nodules after detection, but may lead to improvement in the detectability of such metastatic nodules. Subsequent studies evaluated the ability of contrast ultrasound to differentiate benign from malignant lesions in the spleen. In 60 dogs with splenic abnormalities (27 benign, 29 malignant, 4 normal), moderate to severe hypoechogenicity was seen primarily in malignant lesions, but both benign and malignant lesions had a variety of appearances, including mildly hypo-, iso-, or hyperechoic. Thus, the specificity for lesions with marked enhancement was extremely low. Another study in 19 dogs (11 HSA, 7 hematoma, 1 undifferentiated sarcoma) confirmed that benign versus malignant lesions in the spleen could not be distinguished with contrast harmonic imaging. For the liver, however, histologic assessment in 18 dogs showed the sensitivity and specificity of contrast ultrasound to characterize metastasis were both 100%.

Echocardiography is the best way to evaluate for right atrial lesions but still can provide a false negative unless pericardial effusion is surrounding the mass, making it easier to image. Additionally, a recent abstract presentation of 83 dogs with intra-abdominal (n=66) or SQ (n=17) HSA found that only 10 had atrial masses found with echocardiography. Of those, only 4 did not have evidence of metastatic disease elsewhere, either intra-thoracic (n=3) or intra-abdominal (n=3). Thus, in just 5% (4/83) of the cases was echo was the only test that detected metastasis or a second primary tumor. None of the dogs with SQ HSA had cardiac involvement identified.


Fine needle aspiration with cytology of HSA is typically very low yield. Masses in the liver or spleen that have a more solid appearance are more likely to provide cells and a diagnosis with cytology. Tumors that can be diagnosed in this manner include lymphoma, plasma cell tumors, and osteosarcoma. Tumor seeding or hemorrhage are possible concerns but are rarely clinically significant problems with splenic lesions.

The definitive diagnosis of HSA requires exploratory surgery with an excisional biopsy. Histopathology can be challenging as HSA will bleed, causing hematomas around the tumor. The ENTIRE mass must be evaluated before a diagnosis of a benign hematoma vs. HSA can be believed. Stage I HSA is tumor confined to the spleen, stage II is ruptured tumor or local lymph node involved, and stage III indicates distant metastasis.

Treatment and Prognosis

 • Surgery - For splenic hemangiosarcoma, surgery is an essential part of the standard of care, but splenectomy alone results in MSTs of only 19 to 86 days, with only ~6% of dogs living to one year. Rupture of the tumor did not correlate with a worse prognosis in cases treated with surgery alone. For right auricular masses, the mass may be able to be excised, with a MST of 4 months. Pericardiectomy alone can help relieve the signs of tamponade, and provides a similar MST.

 • Chemotherapy - Chemotherapy post-operatively for HSA (doxorubicin/cyclophosphamide) can extend the MST to 5-8 months. In 8 dogs with excised cardiac HSA treated with doxorubicin-based chemotherapy protocols, the MST was 6 months. Many variations of chemotherapy have been used in an attempt to improve survival times for this aggressive disease. Twenty dogs with HSA were given a dose-intensified doxorubicin protocol (30mg/m2 IV every 2 weeks instead of every 3) for 5 treatments. The protocol was well-tolerated, but did not show a significant improvement in survival times (stage I, II, and III living a median of 257, 210, and 107 days, respectively). In a retrospective study, epirubicin (30 mg/m2) every 3 weeks for up to 4 to 6 treatments was used in 18 dogs, while 41 dogs received splenectomy alone. The MST of dogs given epirubicin (144 days) was greater than the splenectomy alone dogs (86 days). For all dogs, MST with stage I disease (345 days) was significantly longer than for dogs with either stage II (93 days) or III disease (68 days). Dogs with stage I disease treated with splenectomy alone had a MST of 238 days, greater than the 86 days previously reported, and potentially signifying an improved survival for this stage of splenic HSA. Impressively, the MST for dogs with stage I HSA treated with epirubicin was 983 days, and 4 of 7 of these dogs survived > 1 year. Seven of 18 dogs were hospitalized for epirubicin-induced GI toxicity. Recently, low dose continuous chemotherapy, or metronomic chemotherapy, showed potential to be as effective, or possibly more effective, in dogs with post-operative HSA. Etoposide (50 mg/m2 PO daily for 3-weeks) alternated with cyclophosphamide (12.5 to 25 mg/m2 per day orally for 3 weeks). Piroxicam (0.3 mg/kg PO once daily) was administered throughout. Both the MST and the median DFI in 9 dogs treated with LDC were 178 days. By comparison, the MST and DFI in 24 historical control dogs treated with doxorubicin were 133 and 126 days, respectively.

 • Biologic Response Modifiers - In an older study in 32 dogs with splenic HSA, L-MTP-PE, as discussed above, increased survival time, when combined with splenectomy and chemotherapy, from 150 days to 277 days. Twenty eight dogs with a variety of stages of HSA were treated with an allogeneic HSA with liposome-DNA complex vaccine over 22 weeks. Most of the dogs also received chemotherapy. Antibody responses against a control antigen and against canine HSA cells developed in most dogs. Thirteen dogs with stage II splenic HSA that received the tumor vaccine plus doxorubicin chemotherapy had an overall MST of 182 days. The treatment was well-tolerated, and while survival times were not substantially better than those previously reported for chemotherapy alone, further studies with the vaccine in a multi-modality protocol are indicated.

 • Targeted Therapy - Cell culture studies performed with HSA cells and masitinib (Kinavet®), a c-Kit targeting tyrosine kinase inhibitor, showed greatly increased sensitivity of the cells to doxorubicin. Studies evaluating this combination of therapy in clinical patients are pending.


Despite multimodality therapy for these aggressive sarcomas, survival times are relatively short and few cases achieve prolonged survival. New targeted therapies may hold promise for improvements in patient management in the future.

Related Videos
© 2023 MJH Life Sciences

All rights reserved.