Hemangiosarcoma: Look for endothelial precursors as sign

Q. Please review new concepts in diagnosing and managing hemangiosarcoma in dogs.

A. Drs. Angela R. Lamerato-Kozicki and Jaime F. Modiano at the 2005 American College of Veterinary Internal Medicine Forum in Baltimore gave a lecture on early detection of hemangiosarcoma. Some relevant points in this lecture are provided below.

Canine hemangiosarcoma (HSA) is an incurable tumor of vascular endothelial cells. HSA accounts for 7 percent of all cancers; thus, of the 65 million pet dogs living in the United States today, between 1.5 million and 2.5 million could get HSA and succumb to it. Although dogs of any age and breed are susceptible to HSA, it occurs more commonly in dogs beyond middle age and in breeds such as Golden Retrievers, German Shepherds, Portuguese Water Dogs and Skye Terriers. The estimated life-time risk of HSA in Golden Retrievers is one in five, illustrating the magnitude of this problem.

The common primary sites for HSA are spleen, right atrium and subcutis. Local infiltration and systemic metastases are the common growth patterns; metastatic sites are widespread with lung and liver being the most frequently affected organs.

The disease is extremely indolent, so clinical signs are usually not evident until the advanced stage when metastasis has occurred. The tumors at this stage are largely resistant to chemotherapy, and standard-of-care (surgery and intensive chemotherapy) provides a median survival of little more than six months. Morbidity and mortality are usually due to acute internal hemorrhage secondary to tumor rupture. Many dogs die from severe abdominal or thoracic hemorrhage before any treatment can be instituted.

There is presently no effective technology for early diagnosis of HSA. Careful examination of blood smears may hint at the presence of chronic hemorrhage (anemia and thrombocytopenia) and vascular abnormalities (red blood cell fragmentation) that are consistent with HSA; however, this method is neither sensitive nor specific to confirm the diagnosis. Non-invasive imaging methods are useful aids to diagnose the disease (for cavitary tumors such as those that occur in the spleen or heart). Ultrasound is moderately specific, but it is not sensitive.

Moreover, biopsies are required for confirmation of imaging results, and even then, distinction between hemangiosarcoma and benign lesions (hemangioma, hematoma) can be difficult. Skin biopsies where there is no lesion would be of little use to provide early diagnosis for cutaneous HSA. The same is true for splenic, hepatic or cardiac tumors with the added issue that the risk of these procedures in the absence of a visible tumor (on radiographs or ultrasound) is unacceptable. Given the severity of canine HSA and the lack of effective treatment options for advanced disease, it would be useful to have a method for early detection, which might improve the outcome of dogs treated with standard-of-care as well as the design of novel treatment options that may have a better chance of eradicating the tumor.

The endothelial origin of HSA characterizes this tumor in the broad group of hematopoietic or bone marrow-derived malignancies. Endothelial cells arise from bone marrow stromal cells that are triggered to differentiate along the hemangioblastic lineage. Hemangioblasts are pluripotent hematopoietic stem cells. Under conditions of enhanced angiogenesis, these cells give rise to angioblasts, which further differentiate into endothelial precursor cells (EPC). EPC have high proliferative capacity and are largely responsible for physiologic and pathologic angiogenesis. Limited numbers of EPC (and perhaps more primitive stem cells) are found among differentiated blood vessel lining cells in adult animals (humans and mice) and in the peripheral circulation. These cells (in humans and mice) are characterized by the expression of CD34, CD133 and the vascular endothelial growth factor (VEGF) receptor 2; however, the number of EPC in the peripheral circulation of normal individuals is virtually undetectable, except by the most sensitive methods or by prolonged expansion in vitro using specific growth conditions.

Distinguishing from benign cells

Phenotypic properties of HSA cells can be used to distinguish them from normal (or benign) endothelial cells in tissue biopsies. As expected from their malignant nature, HSA cells are incompletely differentiated hemangioblastic cells that express markers associated with immature hematopoietic precursors, such as CD117 (c-Kit). Hence, a unique set of markers to distinguish HSA cells from other committed hematopoietic cell lineages and from differentiated bone-marrow derived cells can be developed. Specifically, these cells should express proteins that belie their hematopoietic origin with incomplete differentiation (e.g., CD117, CD34, CD133, CD45) along with proteins that document their commitment to the endothelial lineage (e.g., CD51/CD61 or v beta 3 integrin, CD105, CD146). Similarly, these cells would lack expression of proteins normally found in hematopoietic cells that are committed to leukocyte lineages, such as CD18, CD11b, CD3 and CD21. Moreover, the intimal contact of HSA cells with the circulation provides malignant cells easy access to the circulation. It is now widely accepted that most tumors shed cells into the microenvironment, and the process of metastasis may be more closely associated with an anti-apoptotic phenotype than with an invasive phenotype. Thus, the frequency of primitive endothelial cells (or EPC-like cells) should be increased naturally in dogs with HSA.

The number of EPC in the circulation has been reported to increase in a variety of conditions in humans and mice. These conditions include stem-cell mobilization using cytokines, physical training and vascular injury (including myocardial infarcts and early congestive heart failure). Other conditions, such as exposure to nicotine, are reported to both increase and decrease circulating EPC suggesting that even under some pathological conditions the frequency of these cells in the circulation is fickle and generally remains well below 1 percent. HSA lesions should shed EPC-like cells into the circulation continuously, meaning that a diagnostic procedure to detect EPC-like cells in excess of a normal reference range would be a sensitive means to diagnose canine HSA, even in early stages. In addition, these EPC-like cells would possess clonal abnormalities that would allow both their characterization as malignant cells and their exploitation to develop new strategies for targeted therapies.

The development of such a method to provide early detection of canine HSA using multi-parameter flow cytometry has currently been pursued. Flow cytometry has proven to be a robust technology to detect circulating cells present at extremely low frequencies. In addition, it also provides the opportunity to isolate and purify HSA cells (by sorting) for confirmatory tests and therapeutic development. Preliminary tests in dogs with biopsy-confirmed visceral hemangiosarcoma (N=7) show that these dogs harbor greater than 1 percent EPC-like cells in the peripheral circulation. This assay requires less than 3 ml of anticoagulated blood, making it a practical, non-invasive means to confirm a suspected diagnosis of HSA. The ultimate goal, however, is to refine the sensitivity and specificity of this assay to detect the presence of HSA lesions in dogs at risk before they pose a clinical problem. Early detection is likely to offer the highest probability of successful treatment outcomes.

The treatment options presently for HSA, even if detected early, are limited. The possible success of chemotherapy in asymptomatic dogs would have to be balanced against the potential toxicity associated with treatment. However, customized experimental treatments for various tumor types have reached (or are close to) clinical trials, suggesting that tailored cancer therapy treatments are on the horizon. Until recently, such tailored therapies were not practical. In particular, immunologic approaches (the gold standard for customized therapy) using tumor vaccines specific for an individual patient's tumor were laborious and slow, and small molecules are only effective when the correct molecular target is present.

In the case of dogs where survival with standard-of-care is short-lived, the time required to generate tumor-specific vaccines and the absence of defined molecular targets meant that customized treatments would not be available in a timely manner to benefit the dog. Yet, canine HSA is clearly amenable to treatment with immunologic approaches. Rapid isolation of malignant cells could be used to confirm the presence of targets that improve delivery of immunomodulators like interleukin (IL)-12. Moreover, a recent technology that generates tailored vaccines in a matter of days has been shown to be practical for clinical applications. In the case of HSA, this technology could be used to construct vaccines against mutant targets that regulate production of vascular endothelial growth factor (VEGF), which is constitutively elevated in the tumors and found at increased levels in blood samples from affected dogs.

Among other abnormalities that occur frequently in tumors, inactivation of the tumor suppressor genes PTEN and VHL can lead to increased VEGF production. Inactivation of PTEN occurs in canine HSA providing cells a growth advantage within their microenvironment. VHL regulates various distinct pathways that converge on endothelial cell proliferation under hypoxic conditions. Currently investigations are underway on how mutations of this gene influence the pathogenesis of HSA in dogs. These are but two potential targets that could be used to develop highly specific approaches with no or acceptable toxicity to treat HSA in its early stages, offering new hope for dogs that develop this disease.

Dr. Hoskins is owner of DocuTech Services. He is a diplomate of the American College of Veterinary Internal Medicine with specialities in small animal pediatrics. He can be reached at (225) 955-3252, fax: (214) 242-2200, or e-mail: jdhoskins@mindspring.com.