**Clinical and Molecular Biology of Angiosarcoma**

N.J. Andersen1, R.E. Froman1, B.E. Kitchell2 and N.S. Duesbery1 *1Laboratory of Cancer and Developmental Cell Biology, Van Andel Research Institute Grand Rapids, 2College of Veterinary Medicine, Michigan State University East Lansing, Michigan, USA* 

## **1. Introduction**

148 Soft Tissue Tumors

[14] Servant JM., Arnault E., Revol M., Damino A. Reconstruction of large thoraco-

Plast Reconstr. Aesthet. Surg. 2006; 59(4): 360-365.

abdominal defects using two-stage free tissue transfers and prosthetic materials. J

*'Skepticism is a healthy response to diagnosis of any tumor as angiosarcoma.* ' (Lane, 1952) Angiosarcoma (AS) is an aggressive malignancy of vascular tissue or vessel forming cells (Requena & Sangueza 1998). AS is rare in humans, making up 1-2 % of soft-tissue sarcomas (Young et al. 2010) and having an estimated incidence of 0.2/100,000 persons per year. Although AS can present anywhere in the body, in humans they typically arise in the skin or superficial soft tissues. It is most frequently noted on the face and scalp of elderly men where their persistent growth causes ulceration and infection, as well as on breasts, and extremities (Brennan et al. 2001; Fayette et al. 2007; Glazebrook et al. 2008). Less frequently AS arises in liver, heart, and spleen (Young et al. 2010). The literature is replete with retrospective analyses and case studies on AS but the rarity of patients diagnosed with this disease makes it difficult to perform more than a superficial investigation on the biology and clinical behavior of AS. However, the occurrence of AS is not restricted to humans and there are several alternative animal models with the potential to inform human studies. We have written this review to collect, compare, and contrast diverse reports on the biology and treatment of AS in humans and alternative animal models. Our objective is to establish a comparative framework to focus further discussion and scientific enquiry.

## **2. Types of AS**

While there is general agreement angiosarcomas arise from endothelial cells, its causes are uncertain. There appears to be no single cause of AS in humans; rather, a variety of factors have been reported to contribute to risk of disease. Broadly speaking the causes of this disease may be classified as environmental, familial, and iatrogenic.

#### **2.1 Environmental AS**

Chronic exposure to any of several environmental risk factors has been linked with AS. Each of these directly or indirectly induce genetic mutations and alter gene expression. Vinyl chloride was first identified as an AS inducing compound when it was noticed that industrial plastic and synthetic rubber manufacturers as well as beauty salon personnel had

Clinical and Molecular Biology of Angiosarcoma 151

However, a recent retrospective study of Finnish cancer patients showed different results (Virtanen et al. 2007). Surveying more than 300,000 cancer patient records accrued over a 50 year period, Virtanen et al. (2007) failed to identify a statistically significant association of AS with prior radiation treatment. These results agree with an earlier Swedish study of 122,991 women with breast cancer that showed no statistically significant correlation between radiation therapy and angiosarcoma, although they did observe a correlation with lymphedema (Karlsson et al. 1998). The Finnish research team did note a five-fold increase in AS among women with breast or gynecologic cancers when compared with the rest of the population. However, while the SEER-based studies focused on the incidence of subsequent sarcoma within the irradiated field of women with breast cancer, the Finnish study examined incidence irrespective of radiation field within a mixed population of men and women treated for several tumor diagnoses. Thus, the latter studies may have failed to demonstrate a statistically significant association between AS and radiotherapy because

Chronic lymphedema is another risk factor for AS. The most well known form is Stewart-Treves syndrome, in which lymphedema presents in the adjacent arm after radical mastectomy (Stewart & Treves 1948). Although rare, AS is also associated with other chronic lymphedema disorders such as Milroy disease and elephantiasis (Hallel-Halevy et al. 1999; Offori et al. 1993). It is not immediately apparent why failure of lymph node drainage should promote AS. One possibility is that edema restricts perfusion and induces an ischemic response involving endothelial proliferation. If this were the case one might expect to see increased incidence of AS in association with peripheral arterial disease. However, such an association has not been reported. Alternatively, edema-mediated changes may create a permissive microenvironment allowing phenotypic expression of an otherwise

Although patients with Von Hippel-Lindau disease show increased incidence of cranial hemangioblastoma (Kaelin 2007), no highly penetrant, causal mutation for AS in humans has been reported. This should not be interpreted as an indication that hereditary factors do not influence the risk of developing AS, but that the incidence of AS is so low that identifying such genetic factors may be exceedingly difficult. However, indirect evidence supports a role for hereditary factors in the onset or progression of AS. The incidence of all vascular tumors appears to be increased in offspring whose parents were diagnosed with kidney cancer, nervous system hemangioma, or hemangioblastoma, and AS of the trunk and extremities has been associated with maternal breast cancer (Ji & Hemminki 2007). Moreover, there is strong evidence for breed and strain-specific predisposition of AS in mice

The presentation of AS is tied to the site of occurrence. In cutaneous AS, tumor may begin as a bruise-like bluish coloration or a red nodular rash (Lane 1952; Morgan et al. 2004). It can also present as purplish multicentric lesions as seen in Stewart-Treves syndrome (Stewart & Treves 1948). Unfortunately due to the innocuous appearance of the initial dermal lesion, a large percentage of patients initially disregard the lesion and as a consequence have

they were underpowered.

**2.3 Familial AS** 

and dogs (see below).

**3. Clinical features** 

**3.1 Presentation** 

masked trait that promotes endothelial proliferation.

an elevated incidence of AS. Exposure to vinyl chloride was the only common element identified in both populations (Creech & Johnson 1974; Infante et al. 2009; Sahmel et al. 2009). Vinyl chloride was used for the production of industrial resins and as a hairspray propellant. Interestingly, vinyl chloride increases the rate of hepatic AS because vinyl chloride is a carcinogen that is predominantly directed towards hepatic endothelial cells. The metabolism of vinyl chloride produces reactive metabolites that form pro-mutagenic DNA adducts (reviewed by Bolt (2005)).

Chronic exposure to arsenic in pesticides or in a traditional medicinal tonic (Fowler's solution) has also been linked to hepatic AS (Centeno et al. 2002; Falk et al. 1993). The precise mechanism(s) of arsenic carcinogenesis is uncertain, although it appears to act in an epigenetic fashion rather than as a classical mutagen, causing pleiotropic cellular effects that promote tumorigenesis (reviewed by Flora (2011)).

No link between AS and UV exposure has been documented. However, there are published case reports of xeroderma pigmentosum patients developing cutaneous AS. Xeroderma pigmentosum is a rare autosomal recessive disease causing defects in nucleotide excision repair; the patients are sensitive to UV radiation and are at higher risk for developing skin cancers (de Boer & Hoeijmakers 2000; Kraemer et al. 1987). Whereas cutaneous AS normally develops in older patients with a mean onset of 63 years (Aust et al. 1997), cutaneous AS in xeroderma pigmentosum patients appears to have an earlier onset, ranging from 13 to 40 years with a mean of 19.2 years (Arlett et al. 2006; Arora et al. 2008; Ludolph-Hauser et al. 2000; Marcon et al. 2004; van Geel & den Bakker 2009). Further work is needed to establish whether there is a role for UV as a causative agent of cutaneous AS.

#### **2.2 Iatrogenic AS**

There are several examples of AS being induced as a consequence of prior medical treatment. Exposure to Thorotrast**®** (colloidal thorium-232 dioxide) has been linked to hepatic AS. Thorotrast**®** was used in the United States from the 1930's through the 1960's as a radiographic contrast compound to visualize blood vessels. Upon injection, Thorotrast is absorbed by bones, liver, spleen, and lymph nodes where it has a high bioavailability and releases alpha particles for imaging. While useful as an imaging agent, Thorotrast**®** was found to have unanticipated toxicities. Because it has a long biological half-life, patients were exposed to high levels of ionizing radiation for extended (lifelong) periods (Autenrieth & Lange 1979; Kiyosawa et al. 1989). Thorotrast**®** exposure has also been associated with increased risk of other cancers including hepatic ductal carcinoma and leukemia (Johnson et al. 1977; Looney 1960).

Radiotherapy is an infrequent but well-recognized risk factor for AS (Botros et al. 2009; Fury et al. 2005; Miura et al. 2003). A study of 274,572 breast cancer patients in the SEER Cancer Incidence Public-Use Database (Yap et al. 2002) found that incidence of a subsequent sarcoma while low, was modestly elevated in patients receiving breast irradiation compared to patients not receiving such therapy (87/82,296 [0.0011%] versus 176/192,276 [0.0009%] respectively). Further, whereas AS accounted for approximately 6% of subsequent sarcomas in unirradiated patients, it represented almost 60% of cases within a previously irradiated field (Yap et al. 2002). It seems radiation therapy not only increases the incidence of secondary malignancies but also changes the spectrum of tumors associated with treatment. These results were consistent with a prior study using the same SEER database that showed increased incidence of AS in irradiated breast cancer patients versus unirradiated patients (Huang & Mackillop 2001).

an elevated incidence of AS. Exposure to vinyl chloride was the only common element identified in both populations (Creech & Johnson 1974; Infante et al. 2009; Sahmel et al. 2009). Vinyl chloride was used for the production of industrial resins and as a hairspray propellant. Interestingly, vinyl chloride increases the rate of hepatic AS because vinyl chloride is a carcinogen that is predominantly directed towards hepatic endothelial cells. The metabolism of vinyl chloride produces reactive metabolites that form pro-mutagenic

Chronic exposure to arsenic in pesticides or in a traditional medicinal tonic (Fowler's solution) has also been linked to hepatic AS (Centeno et al. 2002; Falk et al. 1993). The precise mechanism(s) of arsenic carcinogenesis is uncertain, although it appears to act in an epigenetic fashion rather than as a classical mutagen, causing pleiotropic cellular effects that

No link between AS and UV exposure has been documented. However, there are published case reports of xeroderma pigmentosum patients developing cutaneous AS. Xeroderma pigmentosum is a rare autosomal recessive disease causing defects in nucleotide excision repair; the patients are sensitive to UV radiation and are at higher risk for developing skin cancers (de Boer & Hoeijmakers 2000; Kraemer et al. 1987). Whereas cutaneous AS normally develops in older patients with a mean onset of 63 years (Aust et al. 1997), cutaneous AS in xeroderma pigmentosum patients appears to have an earlier onset, ranging from 13 to 40 years with a mean of 19.2 years (Arlett et al. 2006; Arora et al. 2008; Ludolph-Hauser et al. 2000; Marcon et al. 2004; van Geel & den Bakker 2009). Further work is needed to establish

There are several examples of AS being induced as a consequence of prior medical treatment. Exposure to Thorotrast**®** (colloidal thorium-232 dioxide) has been linked to hepatic AS. Thorotrast**®** was used in the United States from the 1930's through the 1960's as a radiographic contrast compound to visualize blood vessels. Upon injection, Thorotrast is absorbed by bones, liver, spleen, and lymph nodes where it has a high bioavailability and releases alpha particles for imaging. While useful as an imaging agent, Thorotrast**®** was found to have unanticipated toxicities. Because it has a long biological half-life, patients were exposed to high levels of ionizing radiation for extended (lifelong) periods (Autenrieth & Lange 1979; Kiyosawa et al. 1989). Thorotrast**®** exposure has also been associated with increased risk of other cancers including hepatic ductal carcinoma and leukemia (Johnson et

Radiotherapy is an infrequent but well-recognized risk factor for AS (Botros et al. 2009; Fury et al. 2005; Miura et al. 2003). A study of 274,572 breast cancer patients in the SEER Cancer Incidence Public-Use Database (Yap et al. 2002) found that incidence of a subsequent sarcoma while low, was modestly elevated in patients receiving breast irradiation compared to patients not receiving such therapy (87/82,296 [0.0011%] versus 176/192,276 [0.0009%] respectively). Further, whereas AS accounted for approximately 6% of subsequent sarcomas in unirradiated patients, it represented almost 60% of cases within a previously irradiated field (Yap et al. 2002). It seems radiation therapy not only increases the incidence of secondary malignancies but also changes the spectrum of tumors associated with treatment. These results were consistent with a prior study using the same SEER database that showed increased incidence of AS in irradiated breast cancer patients versus unirradiated patients

DNA adducts (reviewed by Bolt (2005)).

**2.2 Iatrogenic AS** 

al. 1977; Looney 1960).

(Huang & Mackillop 2001).

promote tumorigenesis (reviewed by Flora (2011)).

whether there is a role for UV as a causative agent of cutaneous AS.

However, a recent retrospective study of Finnish cancer patients showed different results (Virtanen et al. 2007). Surveying more than 300,000 cancer patient records accrued over a 50 year period, Virtanen et al. (2007) failed to identify a statistically significant association of AS with prior radiation treatment. These results agree with an earlier Swedish study of 122,991 women with breast cancer that showed no statistically significant correlation between radiation therapy and angiosarcoma, although they did observe a correlation with lymphedema (Karlsson et al. 1998). The Finnish research team did note a five-fold increase in AS among women with breast or gynecologic cancers when compared with the rest of the population. However, while the SEER-based studies focused on the incidence of subsequent sarcoma within the irradiated field of women with breast cancer, the Finnish study examined incidence irrespective of radiation field within a mixed population of men and women treated for several tumor diagnoses. Thus, the latter studies may have failed to demonstrate a statistically significant association between AS and radiotherapy because they were underpowered.

Chronic lymphedema is another risk factor for AS. The most well known form is Stewart-Treves syndrome, in which lymphedema presents in the adjacent arm after radical mastectomy (Stewart & Treves 1948). Although rare, AS is also associated with other chronic lymphedema disorders such as Milroy disease and elephantiasis (Hallel-Halevy et al. 1999; Offori et al. 1993). It is not immediately apparent why failure of lymph node drainage should promote AS. One possibility is that edema restricts perfusion and induces an ischemic response involving endothelial proliferation. If this were the case one might expect to see increased incidence of AS in association with peripheral arterial disease. However, such an association has not been reported. Alternatively, edema-mediated changes may create a permissive microenvironment allowing phenotypic expression of an otherwise masked trait that promotes endothelial proliferation.

## **2.3 Familial AS**

Although patients with Von Hippel-Lindau disease show increased incidence of cranial hemangioblastoma (Kaelin 2007), no highly penetrant, causal mutation for AS in humans has been reported. This should not be interpreted as an indication that hereditary factors do not influence the risk of developing AS, but that the incidence of AS is so low that identifying such genetic factors may be exceedingly difficult. However, indirect evidence supports a role for hereditary factors in the onset or progression of AS. The incidence of all vascular tumors appears to be increased in offspring whose parents were diagnosed with kidney cancer, nervous system hemangioma, or hemangioblastoma, and AS of the trunk and extremities has been associated with maternal breast cancer (Ji & Hemminki 2007). Moreover, there is strong evidence for breed and strain-specific predisposition of AS in mice and dogs (see below).

## **3. Clinical features**

#### **3.1 Presentation**

The presentation of AS is tied to the site of occurrence. In cutaneous AS, tumor may begin as a bruise-like bluish coloration or a red nodular rash (Lane 1952; Morgan et al. 2004). It can also present as purplish multicentric lesions as seen in Stewart-Treves syndrome (Stewart & Treves 1948). Unfortunately due to the innocuous appearance of the initial dermal lesion, a large percentage of patients initially disregard the lesion and as a consequence have

Clinical and Molecular Biology of Angiosarcoma 153

Fig. 1. **Histology of human and canine angiosarcoma.** Stained with Hematoxylin and Eosin (a,b,e,f) or immunostained with antibodies against CD31 (c,d,g,h) both human (a, c) and canine (b, d) splenic angiosarcoma as well as canine kidney angiosarcoma (e, g) show extensive networks of endothelial (CD31 positive) cells with multiple irregular

anastomosing channels filled with blood. In contrast this canine cutaneous angiosarcoma (f, h) shows a highly vascular tumor with spindle-shaped CD31-positive endothelial cells. Bars

= 100 micrometers.

systemic disease at the time of ultimate diagnosis (Morgan et al. 2004). For visceral and cardiac disease, AS most often presents as organ failure. Cardiac AS patients commonly have atrial fibrillation, malaise, and resting tachycardia (Ge et al. 2011; Matzke et al.). Splenic AS presents with splenomegaly (enlarged spleen) or abdominal pain and fatigue from splenic rupture (Falk et al. 1993); these patients may also have cytopenia, leukocytosis, and thrombocytosis. Hepatic AS patients initially present with jaundice and may also experience abdominal pain and fatigue (Mahony et al. 1982; Valenzuela et al. 2009).

#### **3.2 Diagnostic pathology**

There are two subtypes of angiosarcoma defined by pathology. Angiosarcoma is classically defined by a growth of spindle-shaped endothelial cells (Figure 1). In 1982, Weiss and Enzinger described a new subtype, epithelioid AS, with large rounded or polygonal cells that have an epithelial-like morphology (Weiss & Enzinger 1982). In reality, most tumors have a mixture of both spindle and round endothelial cells (Morgan et al. 2004). Angiosarcoma tumors classically consist of sheets of endothelial cells with multiple irregular anastomosing channels. These channels can be perfused and blood-filled or void and can be lined with a single or multiple layers of atypical endothelial cells (Koch et al. 2008; Ohsawa et al. 1995; Yang et al. 2010). Areas of hemorrhage and necrosis are also common in AS (Armah et al. 2007; Ge et al. 2011; Gong et al. 2011; Neuhauser et al. 2000; Ohsawa et al. 1995; Yang et al. 2010). Low grade lesions display irregular vascular channels lined with atypical endothelial cells in single or multiple layers (Koch et al. 2008). High-grade lesions are comprised of sheets of undifferentiated, pleomorphic cells which can make them difficult to distinguish from carcinomas (Koch et al. 2008). High grade lesions also have large areas of hemorrhage. There is no clinical or survival advantage between low- and high-grade lesions. Definitive diagnosis is made by biopsy and immunohistochemistry with antibodies against the endothelial marker CD31 and Factor VIII-related antigen. Staining with antibodies against von Willebrand's factor II and CD34 may also be used to confirm a diagnosis of AS (Ohsawa et al. 1995).

#### **3.3 Differential diagnosis**

Cutaneous AS can initially be misdiagnosed as bruising, multiple different skin infections, and even insect bites. It is the duration and spreading of the lesion that prompts concern. Cutaneous AS is similar in appearance to another endothelial derived malignancy, Kaposi sarcoma. Kaposi sarcoma is a virus-induced cancer (Chang et al. 1994; Kemeny et al. 1996). Both Kaposi sarcoma and AS can be multicentric skin lesions with spindle cell pathology (Morgan et al. 2004; Nickoloff & Griffiths 1989). To further complicate the diagnosis, a percentage of Kaposi cells are CD31 positive and up to half of AS cells are positive for the endothelial lymph factor podoplanin by immunohistochemistry. However, only Kaposi sarcoma are positive for the presence of Kaposi's sarcoma causative virus (Human Herpes Virus 8) by immunohistochemistry using antibodies against Latency nuclear antigen-1 (LANA-1) or by amplifying viral DNA by PCR (reviewed in (Mesri et al. 2010; Schmid & Zietz 2005)). Cutaneous AS may also be confused with irregular growths of superficial vasculature or benign hemangiomas, mostly seen in infants (Lawley et al. 2005). These vascular malformations are commonly referred to as port wine stains or angel kisses. These may be distinguished from visceral AS by immunohistochemistry with antibodies against the Wilms tumor 1 transcription factor (Ge et al. 2011; Lawley et al. 2005).

systemic disease at the time of ultimate diagnosis (Morgan et al. 2004). For visceral and cardiac disease, AS most often presents as organ failure. Cardiac AS patients commonly have atrial fibrillation, malaise, and resting tachycardia (Ge et al. 2011; Matzke et al.). Splenic AS presents with splenomegaly (enlarged spleen) or abdominal pain and fatigue from splenic rupture (Falk et al. 1993); these patients may also have cytopenia, leukocytosis, and thrombocytosis. Hepatic AS patients initially present with jaundice and may also

There are two subtypes of angiosarcoma defined by pathology. Angiosarcoma is classically defined by a growth of spindle-shaped endothelial cells (Figure 1). In 1982, Weiss and Enzinger described a new subtype, epithelioid AS, with large rounded or polygonal cells that have an epithelial-like morphology (Weiss & Enzinger 1982). In reality, most tumors have a mixture of both spindle and round endothelial cells (Morgan et al. 2004). Angiosarcoma tumors classically consist of sheets of endothelial cells with multiple irregular anastomosing channels. These channels can be perfused and blood-filled or void and can be lined with a single or multiple layers of atypical endothelial cells (Koch et al. 2008; Ohsawa et al. 1995; Yang et al. 2010). Areas of hemorrhage and necrosis are also common in AS (Armah et al. 2007; Ge et al. 2011; Gong et al. 2011; Neuhauser et al. 2000; Ohsawa et al. 1995; Yang et al. 2010). Low grade lesions display irregular vascular channels lined with atypical endothelial cells in single or multiple layers (Koch et al. 2008). High-grade lesions are comprised of sheets of undifferentiated, pleomorphic cells which can make them difficult to distinguish from carcinomas (Koch et al. 2008). High grade lesions also have large areas of hemorrhage. There is no clinical or survival advantage between low- and high-grade lesions. Definitive diagnosis is made by biopsy and immunohistochemistry with antibodies against the endothelial marker CD31 and Factor VIII-related antigen. Staining with antibodies against von Willebrand's factor II and CD34 may also be used to confirm a

Cutaneous AS can initially be misdiagnosed as bruising, multiple different skin infections, and even insect bites. It is the duration and spreading of the lesion that prompts concern. Cutaneous AS is similar in appearance to another endothelial derived malignancy, Kaposi sarcoma. Kaposi sarcoma is a virus-induced cancer (Chang et al. 1994; Kemeny et al. 1996). Both Kaposi sarcoma and AS can be multicentric skin lesions with spindle cell pathology (Morgan et al. 2004; Nickoloff & Griffiths 1989). To further complicate the diagnosis, a percentage of Kaposi cells are CD31 positive and up to half of AS cells are positive for the endothelial lymph factor podoplanin by immunohistochemistry. However, only Kaposi sarcoma are positive for the presence of Kaposi's sarcoma causative virus (Human Herpes Virus 8) by immunohistochemistry using antibodies against Latency nuclear antigen-1 (LANA-1) or by amplifying viral DNA by PCR (reviewed in (Mesri et al. 2010; Schmid & Zietz 2005)). Cutaneous AS may also be confused with irregular growths of superficial vasculature or benign hemangiomas, mostly seen in infants (Lawley et al. 2005). These vascular malformations are commonly referred to as port wine stains or angel kisses. These may be distinguished from visceral AS by immunohistochemistry with antibodies against

the Wilms tumor 1 transcription factor (Ge et al. 2011; Lawley et al. 2005).

experience abdominal pain and fatigue (Mahony et al. 1982; Valenzuela et al. 2009).

**3.2 Diagnostic pathology** 

diagnosis of AS (Ohsawa et al. 1995).

**3.3 Differential diagnosis** 

Fig. 1. **Histology of human and canine angiosarcoma.** Stained with Hematoxylin and Eosin (a,b,e,f) or immunostained with antibodies against CD31 (c,d,g,h) both human (a, c) and canine (b, d) splenic angiosarcoma as well as canine kidney angiosarcoma (e, g) show extensive networks of endothelial (CD31 positive) cells with multiple irregular anastomosing channels filled with blood. In contrast this canine cutaneous angiosarcoma (f, h) shows a highly vascular tumor with spindle-shaped CD31-positive endothelial cells. Bars = 100 micrometers.

Clinical and Molecular Biology of Angiosarcoma 155

overall survival of 7.6 months (Penel et al. 2008). Paclitaxel did show a subtype specificity; the response rate in head and neck AS was 75%, while for other primary tumor locations the response rate was 58%. The median time to progression of head and neck disease was 9.5 months compared with 7 months for AS of other primary sites (Schlemmer et al. 2008). In another study, scalp AS had a progression-free survival of 6.8 months, but only 2.8 months of progression free survival was documented for AS arising below the clavicle (Fury et al. 2005). In comparison to spontaneous tumors that have a median survival expectation of nearly 9 months, radiation induced AS are more resistant to paclitaxel treatment and have a median overall survival of less than 6 months (Penel et al. 2008). Paclitaxel has also been used as a neoadjuvant therapy to decrease primary tumor size prior to surgery for splenic

The origin of AS tumor cells is still unknown. Angiosarcoma tumors are generally thought to arise from endothelial cells. The majority of cells stain positively for the endothelial cell markers: CD31, CD34, and Factor VIII-antigen. However, up to 70% of CD31 positive AS cells are also positive for the lymph endothelial marker podoplanin (Breiteneder-Geleff et al. 1999). This suggests contributions of both lymph and vascular endothelium in AS. Alternatively, the tumor microenvironment may differentiate a population of

Cancer stem cells or cancer progenitor cells have been documented in a variety of tumors, but the presence of angiosarcoma cancer stem cells is not known. Interestingly, Wnt1 was recently found to be expressed and activated in AS cells. Furthermore, Wnt1 increases endothelial progenitor cell (EPC) proliferation and EPC-dependent angiogenesis (Gherghe et al. 2011). These data suggest a possible role of EPCs in AS tumorigenesis and tumor angiogenesis through a Wnt1 mechanism. As of yet, no study has clearly determined the role of EPCs in tumorigenesis, and the importance of tumor stem cells in AS tumorigenesis

Angiosarcoma has no clearly defined causative mutation or chromosomal changes, but genomic abnormalities including chromosomal deletions, amplifications, and rearrangements, have been identified in a variety of these tumors. In general, the karyotypes of AS are complicated and display both clonal and non-clonal chromosomal changes. The breadth and variety of aberrations suggest they may represent chromosomal alterations from unbridled cell proliferation. The importance of these mutations or gene amplifications for tumorigenesis and tumor development is not known. The most common aberrations involve the loss of heterozygosity of chromosome 22 (Gil-Benso et al. 1994; Kindblom et al. 1991; Quezado et al. 1998; Schuborg et al. 1998; Wong et al. 2001; Zu et al. 2001). The genes of both platelet-derived endothelial cell growth factor (PD-ECGF) and the tumor suppressor NF2 mapped to chromosome 22 (Kindblom et al. 1991) but the significance of this observation has not been explored. Losses of the short arm of chromosomes 4 and 7 are also common in AS (Wong et al. 2001). MYC amplification has been found in radiation-induced

undifferentiated endothelial cells to lymph or vascular-like endothelium.

AS (Vakkalanka & Milhem 2010).

**4. Etiology** 

**4.1 Cellular origins** 

is still unknown.

**4.2 Molecular biology** 

AS tumors that have an 8q24 gain (Guo et al. 2011).

The main discernible clinical sign of splenic AS is splenomegaly (Neuhauser et al. 2000; Thompson et al. 2005). Any known cause of spleen enlargement is a differential diagnosis for splenic AS. This includes but is not limited to lymphoma or infection (reviewed in (Pozo et al. 2009; Suzuki et al. 2010)). In addition, visceral disease often presents with vague flulike symptoms such as malaise and abdominal pain (Falk et al. 1993).

Since the epithelioid AS subtype is defined by endothelial cells that have a large, round cell morphology similar to that of epithelial cells, AS may be confused with a carcinoma (Lin et al. 2010; Weiss & Enzinger 1982). However, epitheloid AS cells stain positive for CD31 and Factor VIII on immunohistochemistry (Fletcher et al. 1991; Lin et al. 2010).

#### **3.4 Prognosis and treatment**

Over the years AS treatment has changed very little. The usual treatment for primary AS is surgery followed by high-dose radiotherapy either alone or with doxorubicin-based chemotherapy. Even with such interventions survival rates are poor. In addition, there is no current standard of care based on disease subtype or location.

Surgery is the mainstay for the treatment of primary AS. Surgical intervention is limited to local control and is not recommended for widely disseminated disease. Current guidelines call for total surgical resection of the tumor with wide negative margins to reduce reoccurrence (Lahat et al. 2009; Mendenhall et al. 2006). However, the multicentric nature of cutaneous AS and its proximity to vital organs in visceral disease makes achieving complete tumor resection difficult. In a review of head and neck soft tissue sarcomas, positive margins were found in 50% of patients (Farhood et al. 1990).

Survival rates after surgery alone are poor (Holden et al. 1987). While the addition of postoperative radiotherapy improves survival, there are drawbacks to radiation. Effective radiation treatment requires wide radiation fields and often doses of more than 50 Gy to achieve tumor control (Mark et al. 1996; Pawlik et al. 2003). Furthermore, radiotherapy is not recommended for radiation-induced AS (Mark et al. 1996).

Doxorubicin is considered the most efficacious chemotherapy agent for local and metastatic AS but provides minimal tumor control or survival benefit. The response rate to doxorubicin has been reported to be 16 - 36%. In addition, doxorubicin causes long-term damage to the heart at cumulative total doses over 400 mg/m2 of body surface area (Chlebowski 1979). To reduce cardiotoxicity, doxorubicin can be encased in liposomes, reducing the amount of free doxorubicin (Cattel et al. 2003). Liposomal doxorubicin treatment of soft tissue sarcomas has equivalent activity but fewer adverse effects relative to free doxorubicin, and thus it is included in the current National Comprehensive Cancer Network guidelines for the treatment of soft tissue sarcomas (Judson et al. 2001). However, the lipid envelope increases doxorubicin absorption and retention by the skin. Side effects of liposomal doxorubicin may include dermal rash or sores on the palms and soles of the feet called palmar-plantar erythodysesthesia (Judson et al. 2001; Lorusso et al. 2007). The latter can be severe enough to require a dosage decrease, treatment interruption, or termination of treatment (Lorusso et al. 2007).

Other adjuvant treatments are reported to have had minimal success. The use of taxanes such as paclitaxel to treat AS has increased. In a small study of 9 patients, 8 responded to paclitaxel (Fata et al. 1999). A 32-patient retrospective paclitaxel study documented a 62% response rate with one complete response. A 30-patient prospective study demonstrated a 19% response rate after 6 months with 3 complete responses with surgery and a median overall survival of 7.6 months (Penel et al. 2008). Paclitaxel did show a subtype specificity; the response rate in head and neck AS was 75%, while for other primary tumor locations the response rate was 58%. The median time to progression of head and neck disease was 9.5 months compared with 7 months for AS of other primary sites (Schlemmer et al. 2008). In another study, scalp AS had a progression-free survival of 6.8 months, but only 2.8 months of progression free survival was documented for AS arising below the clavicle (Fury et al. 2005). In comparison to spontaneous tumors that have a median survival expectation of nearly 9 months, radiation induced AS are more resistant to paclitaxel treatment and have a median overall survival of less than 6 months (Penel et al. 2008). Paclitaxel has also been used as a neoadjuvant therapy to decrease primary tumor size prior to surgery for splenic AS (Vakkalanka & Milhem 2010).

#### **4. Etiology**

154 Soft Tissue Tumors

The main discernible clinical sign of splenic AS is splenomegaly (Neuhauser et al. 2000; Thompson et al. 2005). Any known cause of spleen enlargement is a differential diagnosis for splenic AS. This includes but is not limited to lymphoma or infection (reviewed in (Pozo et al. 2009; Suzuki et al. 2010)). In addition, visceral disease often presents with vague flu-

Since the epithelioid AS subtype is defined by endothelial cells that have a large, round cell morphology similar to that of epithelial cells, AS may be confused with a carcinoma (Lin et al. 2010; Weiss & Enzinger 1982). However, epitheloid AS cells stain positive for CD31 and

Over the years AS treatment has changed very little. The usual treatment for primary AS is surgery followed by high-dose radiotherapy either alone or with doxorubicin-based chemotherapy. Even with such interventions survival rates are poor. In addition, there is no

Surgery is the mainstay for the treatment of primary AS. Surgical intervention is limited to local control and is not recommended for widely disseminated disease. Current guidelines call for total surgical resection of the tumor with wide negative margins to reduce reoccurrence (Lahat et al. 2009; Mendenhall et al. 2006). However, the multicentric nature of cutaneous AS and its proximity to vital organs in visceral disease makes achieving complete tumor resection difficult. In a review of head and neck soft tissue sarcomas, positive

Survival rates after surgery alone are poor (Holden et al. 1987). While the addition of postoperative radiotherapy improves survival, there are drawbacks to radiation. Effective radiation treatment requires wide radiation fields and often doses of more than 50 Gy to achieve tumor control (Mark et al. 1996; Pawlik et al. 2003). Furthermore, radiotherapy is not

Doxorubicin is considered the most efficacious chemotherapy agent for local and metastatic AS but provides minimal tumor control or survival benefit. The response rate to doxorubicin has been reported to be 16 - 36%. In addition, doxorubicin causes long-term damage to the heart at cumulative total doses over 400 mg/m2 of body surface area (Chlebowski 1979). To reduce cardiotoxicity, doxorubicin can be encased in liposomes, reducing the amount of free doxorubicin (Cattel et al. 2003). Liposomal doxorubicin treatment of soft tissue sarcomas has equivalent activity but fewer adverse effects relative to free doxorubicin, and thus it is included in the current National Comprehensive Cancer Network guidelines for the treatment of soft tissue sarcomas (Judson et al. 2001). However, the lipid envelope increases doxorubicin absorption and retention by the skin. Side effects of liposomal doxorubicin may include dermal rash or sores on the palms and soles of the feet called palmar-plantar erythodysesthesia (Judson et al. 2001; Lorusso et al. 2007). The latter can be severe enough to require a dosage decrease, treatment interruption, or termination of

Other adjuvant treatments are reported to have had minimal success. The use of taxanes such as paclitaxel to treat AS has increased. In a small study of 9 patients, 8 responded to paclitaxel (Fata et al. 1999). A 32-patient retrospective paclitaxel study documented a 62% response rate with one complete response. A 30-patient prospective study demonstrated a 19% response rate after 6 months with 3 complete responses with surgery and a median

like symptoms such as malaise and abdominal pain (Falk et al. 1993).

current standard of care based on disease subtype or location.

margins were found in 50% of patients (Farhood et al. 1990).

recommended for radiation-induced AS (Mark et al. 1996).

**3.4 Prognosis and treatment** 

treatment (Lorusso et al. 2007).

Factor VIII on immunohistochemistry (Fletcher et al. 1991; Lin et al. 2010).

#### **4.1 Cellular origins**

The origin of AS tumor cells is still unknown. Angiosarcoma tumors are generally thought to arise from endothelial cells. The majority of cells stain positively for the endothelial cell markers: CD31, CD34, and Factor VIII-antigen. However, up to 70% of CD31 positive AS cells are also positive for the lymph endothelial marker podoplanin (Breiteneder-Geleff et al. 1999). This suggests contributions of both lymph and vascular endothelium in AS. Alternatively, the tumor microenvironment may differentiate a population of undifferentiated endothelial cells to lymph or vascular-like endothelium.

Cancer stem cells or cancer progenitor cells have been documented in a variety of tumors, but the presence of angiosarcoma cancer stem cells is not known. Interestingly, Wnt1 was recently found to be expressed and activated in AS cells. Furthermore, Wnt1 increases endothelial progenitor cell (EPC) proliferation and EPC-dependent angiogenesis (Gherghe et al. 2011). These data suggest a possible role of EPCs in AS tumorigenesis and tumor angiogenesis through a Wnt1 mechanism. As of yet, no study has clearly determined the role of EPCs in tumorigenesis, and the importance of tumor stem cells in AS tumorigenesis is still unknown.

#### **4.2 Molecular biology**

Angiosarcoma has no clearly defined causative mutation or chromosomal changes, but genomic abnormalities including chromosomal deletions, amplifications, and rearrangements, have been identified in a variety of these tumors. In general, the karyotypes of AS are complicated and display both clonal and non-clonal chromosomal changes. The breadth and variety of aberrations suggest they may represent chromosomal alterations from unbridled cell proliferation. The importance of these mutations or gene amplifications for tumorigenesis and tumor development is not known. The most common aberrations involve the loss of heterozygosity of chromosome 22 (Gil-Benso et al. 1994; Kindblom et al. 1991; Quezado et al. 1998; Schuborg et al. 1998; Wong et al. 2001; Zu et al. 2001). The genes of both platelet-derived endothelial cell growth factor (PD-ECGF) and the tumor suppressor NF2 mapped to chromosome 22 (Kindblom et al. 1991) but the significance of this observation has not been explored. Losses of the short arm of chromosomes 4 and 7 are also common in AS (Wong et al. 2001). MYC amplification has been found in radiation-induced AS tumors that have an 8q24 gain (Guo et al. 2011).

Clinical and Molecular Biology of Angiosarcoma 157

including VEGF, which are necessary to promote KS lesions. Kaposi sarcoma is caused by human herpes virus 8 (HHV8 or KSHV; (Ablashi et al. 2002; Chang et al. 1994)). The expression of a single viral gene encoding a constitutively active G-protein coupled receptor (vGPCR) in endothelial cells is sufficient to induce angioproliferative tumors in mice that closely resemble human KS (Montaner et al. 2003). Among other functions, Kaposi sarcoma vGPCR has been demonstrated to activate the mitogen- and stress-activated protein kinases MAPK1/2 and p38 MAPK in a variety of cell lines (Arvanitakis et al. 1997; Bais et al. 1998; Burger et al. 1999; Cannon et al. 2003; Cannon & Cesarman 2004; Sodhi et al. 2000). Significantly, KS vGPCR-induced expression of VEGF depends upon increased activity of both MAPK1/2 and p38 MAPK (Bais et al. 1998; Sodhi et al. 2000). Since these are direct downstream targets of mitogen activated protein kinase kinases (MKKs) (Lewis et al. 1998), MKKs present an attractive potential therapeutic target for treatment of AS. In support of this we have observed that when mouse endothelial cell tumor xenografts expressing vGPCR are treated with a proteolytic inhibitor of MKK tumor growth is blocked and vascularization is decreased (Depeille et al. 2007). MEK inhibitors tested in clinical trials for a variety of tumors, most notably melanoma, have not shown marked efficacy (Lee & Duesbery 2010). However, these drugs target only MEK1 and 2, leaving other MKK

The similarity between Kaposi sarcoma and AS raises interesting questions regarding the etiology and molecular biology of AS. Are there spindle cell equivalents in AS that release cytokines and drive tumor growth? Could AS also have a viral origin? Indeed, several other human cutaneous cancers have viral origins (e.g. Merkel cell carcinoma and polyomavirus, squamous cell carcinoma and HPV) and viruses can cause AS in mouse models (see below). Moreover, similar to iatrogenic KS, cases of AS have been reported to develop in kidney transplant patients taking immune suppressants (e.g. Farag et al. (2005) and references therein). Evidence for a viral origin of AS has been sought through detection of HHV8 viral expression sequences. However, an initial report showing evidence of an HHV8 sequence in a subset of AS (McDonagh et al. 1996) was contradicted by subsequent studies (Lasota & Miettinen 1999). Though it appears HHV8 does not play a role in the

Hemangiomas are benign growths arising from vascular endothelial cells. In hemangiomas, endothelial cells form a tortuous network of blood vessels. These lesions are common in newborns as strawberry patches or port wine stains. The causative mechanism of growth is unknown, but serum levels of VEGFA and HIF-1alpha are elevated in children with hemangiomas (Kleinman et al. 2007) suggesting hypoxic signals may be driving AS growth. Hemangiomas have been successfully treated with the beta-blocking agent propranolol (Leaute-Labreze et al. 2008; Sans et al. 2009). While the mechanism of action is not known, Sans et al. (2009) speculate that propranolol inhibits beta-adrenergic receptors expressed in endothelial cells, and in this manner down-regulates MAPK signaling and VEGF release.

The incidence of AS has been documented in a variety of species. For discussion purposes we have separated these into 1) genetically engineered or chemically induced AS models

origin of AS, we can not yet rule out a viral contribution to AS tumorigenesis.

signaling pathways intact.

**5.1.2 Hemangioma** 

**5.2 Similar tumors in other species** 

and 2) spontaneous or naturally occurring AS.

The loss of tumor suppressor function likely plays an important role in AS development. The tumor suppressor p53 is mutated in a subset of AS tumors. In cardiac AS, mutant p53 is overexpressed and localizes to the nucleus (Hollstein et al. 1994; Soini et al. 1995; Zu et al. 2001). Mutations in p53 codons 141 and 136 were identified in a small cohort of hepatic AS (Soini et al. 1995) and mutations in p53 codons 249 and 255 have been associated with vinyl chloride exposure (Hollstein et al. 1994). Other studies indicate the INK4a-ARF locus is frequently inactivated in liver AS independent of p53 mutations (Weihrauch et al. 2002). Mutations in the PTEN tumor suppressor have also been identified in hepatic AS; an exon 7 mutation codes for a premature stop codon resulting in a nonfunctional protein (Tate et al. 2007).

Activating mutations in key signaling molecules promote AS tumorigenesis. K-Ras2 mutations have been associated with AS (Marion 2005; Przygodzki et al. 1997) and have been detected in hepatic AS after exposure to vinyl chloride (Weihrauch et al. 2002). Ten percent of AS tumors showed activating KDR (VEGFR2) mutations (Antonescu et al. 2009). Also, 25% of a 20 patient study of radiation-induced AS showed a co-amplification of FLT4 (VEGFR3) and MYC (Guo et al. 2011).

Interestingly, vascular endothelial specific cadherin (VE-cadherin) is absent or decreased in AS (Martin-Padura et al. 1995; Tanioka et al. 2001; Zanetta et al. 2005). Loss of VE-cadherin may promote AS growth in several ways. Like other cadherins, VE-cadherin interacts with and retains β-catenin at adherens junctions. Through this mechanism, cadherins indirectly control the free (cytoplasmic) pool of β-catenin. Free β-catenin can be transported to the nucleus to regulate transcription. Thus, loss of VE-cadherin may represent a pathway to induce cancer gene dysregulation. Retention of β-catenin at adherens junctions by VEcadherin has been inversely correlated with vascular tumor growth and hemorrhage (Zanetta et al. 2005). During angiogenesis, endothelial cells detach and invade surrounding tissue, where they replicate and branch into a vascular network. The loss of VE-cadherin may facilitate angiogenesis. In addition, change in the cadherin profile of cells, or cadherin switching, has been implicated in tumor cell invasion and metastasis (reviewed in Maeda et al. (2005)). Whether cadherin misregulation in AS cells explains the invasive and metastatic nature of AS is not known.

#### **5. Insight from other sources**

The rarity of angiosarcoma has restricted the scope of basic and clinical research on this disease. In this section we will discuss experimental models for generating further insight into the biology and treatment of AS, including the study of more common (but related) human tumors, genetically engineered or chemically induced mouse models, and the emerging relevance of studies of naturally occurring AS in companion dogs.

#### **5.1 Related tumors in humans**

#### **5.1.1 Kaposi sarcoma**

The similarity between Kaposi sarcoma (KS) and AS may provide a foundation for understanding pathways vital for AS tumorigenesis and development. Like AS, KS is an endothelial cell-derived tumor, but Kaposi is derived from lymphatic endothelial cells. Kaposi sarcoma is a vascular tumor characterized histologically by a high degree of angiogenesis and the presence of proliferating spindle cells (Brash & Bale 2001). The spindle cells act in a paracrine fashion, releasing pro-inflammatory and angiogenic factors,

The loss of tumor suppressor function likely plays an important role in AS development. The tumor suppressor p53 is mutated in a subset of AS tumors. In cardiac AS, mutant p53 is overexpressed and localizes to the nucleus (Hollstein et al. 1994; Soini et al. 1995; Zu et al. 2001). Mutations in p53 codons 141 and 136 were identified in a small cohort of hepatic AS (Soini et al. 1995) and mutations in p53 codons 249 and 255 have been associated with vinyl chloride exposure (Hollstein et al. 1994). Other studies indicate the INK4a-ARF locus is frequently inactivated in liver AS independent of p53 mutations (Weihrauch et al. 2002). Mutations in the PTEN tumor suppressor have also been identified in hepatic AS; an exon 7 mutation codes for a premature stop codon resulting in a nonfunctional protein (Tate et al.

Activating mutations in key signaling molecules promote AS tumorigenesis. K-Ras2 mutations have been associated with AS (Marion 2005; Przygodzki et al. 1997) and have been detected in hepatic AS after exposure to vinyl chloride (Weihrauch et al. 2002). Ten percent of AS tumors showed activating KDR (VEGFR2) mutations (Antonescu et al. 2009). Also, 25% of a 20 patient study of radiation-induced AS showed a co-amplification of FLT4

Interestingly, vascular endothelial specific cadherin (VE-cadherin) is absent or decreased in AS (Martin-Padura et al. 1995; Tanioka et al. 2001; Zanetta et al. 2005). Loss of VE-cadherin may promote AS growth in several ways. Like other cadherins, VE-cadherin interacts with and retains β-catenin at adherens junctions. Through this mechanism, cadherins indirectly control the free (cytoplasmic) pool of β-catenin. Free β-catenin can be transported to the nucleus to regulate transcription. Thus, loss of VE-cadherin may represent a pathway to induce cancer gene dysregulation. Retention of β-catenin at adherens junctions by VEcadherin has been inversely correlated with vascular tumor growth and hemorrhage (Zanetta et al. 2005). During angiogenesis, endothelial cells detach and invade surrounding tissue, where they replicate and branch into a vascular network. The loss of VE-cadherin may facilitate angiogenesis. In addition, change in the cadherin profile of cells, or cadherin switching, has been implicated in tumor cell invasion and metastasis (reviewed in Maeda et al. (2005)). Whether cadherin misregulation in AS cells explains the invasive and metastatic

The rarity of angiosarcoma has restricted the scope of basic and clinical research on this disease. In this section we will discuss experimental models for generating further insight into the biology and treatment of AS, including the study of more common (but related) human tumors, genetically engineered or chemically induced mouse models, and the

The similarity between Kaposi sarcoma (KS) and AS may provide a foundation for understanding pathways vital for AS tumorigenesis and development. Like AS, KS is an endothelial cell-derived tumor, but Kaposi is derived from lymphatic endothelial cells. Kaposi sarcoma is a vascular tumor characterized histologically by a high degree of angiogenesis and the presence of proliferating spindle cells (Brash & Bale 2001). The spindle cells act in a paracrine fashion, releasing pro-inflammatory and angiogenic factors,

emerging relevance of studies of naturally occurring AS in companion dogs.

2007).

(VEGFR3) and MYC (Guo et al. 2011).

nature of AS is not known.

**5. Insight from other sources** 

**5.1 Related tumors in humans** 

**5.1.1 Kaposi sarcoma** 

including VEGF, which are necessary to promote KS lesions. Kaposi sarcoma is caused by human herpes virus 8 (HHV8 or KSHV; (Ablashi et al. 2002; Chang et al. 1994)). The expression of a single viral gene encoding a constitutively active G-protein coupled receptor (vGPCR) in endothelial cells is sufficient to induce angioproliferative tumors in mice that closely resemble human KS (Montaner et al. 2003). Among other functions, Kaposi sarcoma vGPCR has been demonstrated to activate the mitogen- and stress-activated protein kinases MAPK1/2 and p38 MAPK in a variety of cell lines (Arvanitakis et al. 1997; Bais et al. 1998; Burger et al. 1999; Cannon et al. 2003; Cannon & Cesarman 2004; Sodhi et al. 2000). Significantly, KS vGPCR-induced expression of VEGF depends upon increased activity of both MAPK1/2 and p38 MAPK (Bais et al. 1998; Sodhi et al. 2000). Since these are direct downstream targets of mitogen activated protein kinase kinases (MKKs) (Lewis et al. 1998), MKKs present an attractive potential therapeutic target for treatment of AS. In support of this we have observed that when mouse endothelial cell tumor xenografts expressing vGPCR are treated with a proteolytic inhibitor of MKK tumor growth is blocked and

vascularization is decreased (Depeille et al. 2007). MEK inhibitors tested in clinical trials for a variety of tumors, most notably melanoma, have not shown marked efficacy (Lee & Duesbery 2010). However, these drugs target only MEK1 and 2, leaving other MKK signaling pathways intact.

The similarity between Kaposi sarcoma and AS raises interesting questions regarding the etiology and molecular biology of AS. Are there spindle cell equivalents in AS that release cytokines and drive tumor growth? Could AS also have a viral origin? Indeed, several other human cutaneous cancers have viral origins (e.g. Merkel cell carcinoma and polyomavirus, squamous cell carcinoma and HPV) and viruses can cause AS in mouse models (see below). Moreover, similar to iatrogenic KS, cases of AS have been reported to develop in kidney transplant patients taking immune suppressants (e.g. Farag et al. (2005) and references therein). Evidence for a viral origin of AS has been sought through detection of HHV8 viral expression sequences. However, an initial report showing evidence of an HHV8 sequence in a subset of AS (McDonagh et al. 1996) was contradicted by subsequent studies (Lasota & Miettinen 1999). Though it appears HHV8 does not play a role in the origin of AS, we can not yet rule out a viral contribution to AS tumorigenesis.

## **5.1.2 Hemangioma**

Hemangiomas are benign growths arising from vascular endothelial cells. In hemangiomas, endothelial cells form a tortuous network of blood vessels. These lesions are common in newborns as strawberry patches or port wine stains. The causative mechanism of growth is unknown, but serum levels of VEGFA and HIF-1alpha are elevated in children with hemangiomas (Kleinman et al. 2007) suggesting hypoxic signals may be driving AS growth. Hemangiomas have been successfully treated with the beta-blocking agent propranolol (Leaute-Labreze et al. 2008; Sans et al. 2009). While the mechanism of action is not known, Sans et al. (2009) speculate that propranolol inhibits beta-adrenergic receptors expressed in endothelial cells, and in this manner down-regulates MAPK signaling and VEGF release.

#### **5.2 Similar tumors in other species**

The incidence of AS has been documented in a variety of species. For discussion purposes we have separated these into 1) genetically engineered or chemically induced AS models and 2) spontaneous or naturally occurring AS.

Clinical and Molecular Biology of Angiosarcoma 159

et al. (1998) noted that 2.5-2.7% spontaneously developed AS. Although AS arose at many sites, the predominant locations were liver (male/female; 0.7/1.1%), spleen (0.7/1.0%), and uterus (0.4%). Similarly, a second study of B6C3F1 mice (Chandra & Frith 1992) noted 4.5/1.0% (male/female) of mice spontaneously developed AS. In contrast CD-1 and Icr:Ha mice have spontaneous rates of AS around 0.7% and 0.4%, respectively (Chandra & Frith 1992; Eaton et al. 1980). The increased incidence in B6C3F1 mice indicates that certain strains

Naturally occurring (spontaneous) AS are not restricted to humans and mice. They have been detected in a variety of animals including rhesus macaques (Mejia et al. 2009; Myers et al. 2001), horses (Schultheiss 2004), cows (Sutton & McLennan 1982), ferrets (Schultheiss 2004), dogs (Hargis et al. 1992), cats (Miller et al. 1992), rats (Schultheiss 2004), and snakes (Tuttle 2006). Any of these animals could potentially serve as a surrogate model system for human disease in an experimental context but dogs seem best suited for this purpose. While sharing many features in common with humans as patients, dogs have a much higher incidence of AS and a more rapid time course of disease progression. The canine model also offers some unique, advantageous features distinguishing it from other animal models and opening novel experimental opportunities. First, because of selective breeding, genetic variation within canine breeds is very low. Second, since each breed is derived from a small group of founders, most tracing back approximately 150 years, many of the genes associated with polygenic traits are fixed, so only a few variable genes determine phenotype. This means that it will be much easier to identify genetic disease determinants in dogs than in humans. Finally, companion dogs share the same environmental exposures as humans and thus may more accurately reflect the human condition. The ability to identify, recruit, and study cancers within a breed of dog offers new avenues of hope for research into clinical oncology and the underlying causes of AS (Gordon et al. 2009; Paoloni & Khanna 2008).

In contrast to AS in humans, AS is relatively common in dogs, with an overall incidence of 24/100,000 (Dobson et al. 2002). There is substantial variability in the incidence of canine angiosarcoma (cAS, more commonly referred to as hemangiosarcoma) among breeds, with large breed dogs such as Golden retrievers and German shepherds having a much higher incidence. A health survey of Golden retrievers attributed 62% of all deaths to cancer, with cAS representing 16% of reported cancers (Glickman et al. 1999). Why this tumor is more common in dogs than in humans is uncertain. It is possible that disease-causing genes were over-represented in the founding populations for these dogs. Alternatively, genes promoting susceptibility to cAS may have been co-selected along with artificially selected traits in the directed evolution of these breeds. Regardless, the genetic uniformity within

cAS may present in virtually any tissue of the body, including bone, kidney, skeletal muscle, liver, lung, aorta, urinary bladder, intestine, oral mucosa, tongue, prostate, vulva/vagina, perineum and the cornea, nictitans or conjunctival tissue of the eye (Bergman 2010; Bulakowski et al. 2008; Hargis et al. 1992; Schultheiss 2004; Withrow & MacEwen 1996). Metastasis can involve any portion of the body, including the central nervous system. While most dogs present with splenic or liver lesions, some breeds show elevated incidence of

breeds offers a unique and unbiased opportunity to identify genes that cause AS.

are genetically predisposed to develop AS.

**6. Canine AS** 

**6.1 Presentation and diagnosis** 

#### **5.2.1 Induced angiosarcoma**

Mice develop AS in response to a viral infection, toxicologic treatment, or genetic manipulation. Study of the common properties or convergent signaling of these diverse agents may reveal a unified mechanism for AS induction.

**Viral Infection** – Injection of Moloney mouse sarcoma virus (Stoica et al. 1990; Yuen et al. 1991), Harvey Sarcoma virus (Chesterman et al. 1966; Harvey 1964), or Kirsten Sarcoma virus (Pitts et al. 1983) into mice or rats induces the formation of angiomatous lesions resembling angiosarcoma or Kaposi's sarcoma. These viruses encode mutant, oncogenic forms of cellular oncogenes encoding c-Mos, H-Ras, and K-Ras, respectively. Each of these genes stimulates cellular proliferation and oncogenic transformation by activation of the mitogen-activated protein kinase (MAPK) signaling pathway.

**Toxicologic treatment** – Rodent bioassays have been used to identify several carcinogenic compounds capable of inducing angiosarcoma (reviewed by Cohen et al. (2009)). Some of these compounds react with DNA, acting in a manner similar to vinyl chloride or Thorotrast. Others, such as 2-butoxyethanol, are non-genotoxic and act indirectly through production of reactive oxygen species and cytokines to stimulate endothelial cell proliferation (Corthals et al. 2006; Kamendulis et al. 2010; Klaunig & Kamendulis 2005).

**Genetic manipulations** – Mice engineered to express knock-in mutations (D1226N or Y1228C) in the activation loop of Met develop a high incidence of angiosarcoma with moderately pleiomorphic endothelial cells, cavernous blood vessels, and palisading epithelioid-like cells (Graveel et al. 2004). Met is a tyrosine kinase receptor for the hepatocyte growth factor/scatter factor and is a potent activator of the MAPK signaling pathway, regulating among other things the epithelial-to-mesenchymal transition and metastatic behavior (Birchmeier et al. 2003). There are several Met antagonists in clinical trials for a variety of tumors (Eder et al. 2009).

As noted above, the loss of tumor suppressor function likely plays an important role in AS development. Not surprisingly then, loss of tumor suppressor function in mice induces AS formation. Heterozygous deletion of p53 causes a broad-based cancer predisposition in which 57% of mice develop a variety of sarcomas including AS (Jacks et al. 1994). The incidence of AS in p53-null mice is dramatically increased from 15% to 53% when mice are crossed to Wrnhel mice harboring homozygous deletions in the helicase domain of the *Wrn* gene (Lebel et al. 2001). Similarly, p53-null mice crossed with mice lacking alleles of Ink4d and/or Ink4c show as high as 75-85% incidence of angiosarcoma (Zindy et al. 2003). The Wrn protein interacts with p53 to promote genomic stability. The Ink4 proteins bind and inactivate cyclin dependent kinases, preventing cell cycle progression and maintaining cells in a quiescent state.

These data show that genomic manipulation resulting in deregulation of cell cycle control and genomic instability in mice readily promotes endothelial cells to switch to a malignant phenotype. Why then do mice develop AS at such high rates in response to a viral infection, toxicologic treatment, or genetic manipulation? The answer may be linked to unidentified hereditary factors since studies of spontaneous tumorigenesis in mice indicate mice are predisposed to develop AS (see below).

#### **5.2.2 Spontaneous tumors**

Mice have a substantially higher incidence of spontaneous AS than do humans, but the incidence varies by strain. In a two-year study of approximately 2000 B6C3F1 mice, Haseman

Mice develop AS in response to a viral infection, toxicologic treatment, or genetic manipulation. Study of the common properties or convergent signaling of these diverse

**Viral Infection** – Injection of Moloney mouse sarcoma virus (Stoica et al. 1990; Yuen et al. 1991), Harvey Sarcoma virus (Chesterman et al. 1966; Harvey 1964), or Kirsten Sarcoma virus (Pitts et al. 1983) into mice or rats induces the formation of angiomatous lesions resembling angiosarcoma or Kaposi's sarcoma. These viruses encode mutant, oncogenic forms of cellular oncogenes encoding c-Mos, H-Ras, and K-Ras, respectively. Each of these genes stimulates cellular proliferation and oncogenic transformation by activation of the

**Toxicologic treatment** – Rodent bioassays have been used to identify several carcinogenic compounds capable of inducing angiosarcoma (reviewed by Cohen et al. (2009)). Some of these compounds react with DNA, acting in a manner similar to vinyl chloride or Thorotrast. Others, such as 2-butoxyethanol, are non-genotoxic and act indirectly through production of reactive oxygen species and cytokines to stimulate endothelial cell proliferation (Corthals et al. 2006; Kamendulis et al. 2010; Klaunig & Kamendulis 2005). **Genetic manipulations** – Mice engineered to express knock-in mutations (D1226N or Y1228C) in the activation loop of Met develop a high incidence of angiosarcoma with moderately pleiomorphic endothelial cells, cavernous blood vessels, and palisading epithelioid-like cells (Graveel et al. 2004). Met is a tyrosine kinase receptor for the hepatocyte growth factor/scatter factor and is a potent activator of the MAPK signaling pathway, regulating among other things the epithelial-to-mesenchymal transition and metastatic behavior (Birchmeier et al. 2003). There are several Met antagonists in clinical

As noted above, the loss of tumor suppressor function likely plays an important role in AS development. Not surprisingly then, loss of tumor suppressor function in mice induces AS formation. Heterozygous deletion of p53 causes a broad-based cancer predisposition in which 57% of mice develop a variety of sarcomas including AS (Jacks et al. 1994). The incidence of AS in p53-null mice is dramatically increased from 15% to 53% when mice are crossed to Wrnhel mice harboring homozygous deletions in the helicase domain of the *Wrn* gene (Lebel et al. 2001). Similarly, p53-null mice crossed with mice lacking alleles of Ink4d and/or Ink4c show as high as 75-85% incidence of angiosarcoma (Zindy et al. 2003). The Wrn protein interacts with p53 to promote genomic stability. The Ink4 proteins bind and inactivate cyclin dependent kinases, preventing cell cycle progression and maintaining cells

These data show that genomic manipulation resulting in deregulation of cell cycle control and genomic instability in mice readily promotes endothelial cells to switch to a malignant phenotype. Why then do mice develop AS at such high rates in response to a viral infection, toxicologic treatment, or genetic manipulation? The answer may be linked to unidentified hereditary factors since studies of spontaneous tumorigenesis in mice indicate mice are

Mice have a substantially higher incidence of spontaneous AS than do humans, but the incidence varies by strain. In a two-year study of approximately 2000 B6C3F1 mice, Haseman

**5.2.1 Induced angiosarcoma** 

agents may reveal a unified mechanism for AS induction.

mitogen-activated protein kinase (MAPK) signaling pathway.

trials for a variety of tumors (Eder et al. 2009).

predisposed to develop AS (see below).

**5.2.2 Spontaneous tumors** 

in a quiescent state.

et al. (1998) noted that 2.5-2.7% spontaneously developed AS. Although AS arose at many sites, the predominant locations were liver (male/female; 0.7/1.1%), spleen (0.7/1.0%), and uterus (0.4%). Similarly, a second study of B6C3F1 mice (Chandra & Frith 1992) noted 4.5/1.0% (male/female) of mice spontaneously developed AS. In contrast CD-1 and Icr:Ha mice have spontaneous rates of AS around 0.7% and 0.4%, respectively (Chandra & Frith 1992; Eaton et al. 1980). The increased incidence in B6C3F1 mice indicates that certain strains are genetically predisposed to develop AS.

Naturally occurring (spontaneous) AS are not restricted to humans and mice. They have been detected in a variety of animals including rhesus macaques (Mejia et al. 2009; Myers et al. 2001), horses (Schultheiss 2004), cows (Sutton & McLennan 1982), ferrets (Schultheiss 2004), dogs (Hargis et al. 1992), cats (Miller et al. 1992), rats (Schultheiss 2004), and snakes (Tuttle 2006). Any of these animals could potentially serve as a surrogate model system for human disease in an experimental context but dogs seem best suited for this purpose. While sharing many features in common with humans as patients, dogs have a much higher incidence of AS and a more rapid time course of disease progression. The canine model also offers some unique, advantageous features distinguishing it from other animal models and opening novel experimental opportunities. First, because of selective breeding, genetic variation within canine breeds is very low. Second, since each breed is derived from a small group of founders, most tracing back approximately 150 years, many of the genes associated with polygenic traits are fixed, so only a few variable genes determine phenotype. This means that it will be much easier to identify genetic disease determinants in dogs than in humans. Finally, companion dogs share the same environmental exposures as humans and thus may more accurately reflect the human condition. The ability to identify, recruit, and study cancers within a breed of dog offers new avenues of hope for research into clinical oncology and the underlying causes of AS (Gordon et al. 2009; Paoloni & Khanna 2008).

## **6. Canine AS**

In contrast to AS in humans, AS is relatively common in dogs, with an overall incidence of 24/100,000 (Dobson et al. 2002). There is substantial variability in the incidence of canine angiosarcoma (cAS, more commonly referred to as hemangiosarcoma) among breeds, with large breed dogs such as Golden retrievers and German shepherds having a much higher incidence. A health survey of Golden retrievers attributed 62% of all deaths to cancer, with cAS representing 16% of reported cancers (Glickman et al. 1999). Why this tumor is more common in dogs than in humans is uncertain. It is possible that disease-causing genes were over-represented in the founding populations for these dogs. Alternatively, genes promoting susceptibility to cAS may have been co-selected along with artificially selected traits in the directed evolution of these breeds. Regardless, the genetic uniformity within breeds offers a unique and unbiased opportunity to identify genes that cause AS.

#### **6.1 Presentation and diagnosis**

cAS may present in virtually any tissue of the body, including bone, kidney, skeletal muscle, liver, lung, aorta, urinary bladder, intestine, oral mucosa, tongue, prostate, vulva/vagina, perineum and the cornea, nictitans or conjunctival tissue of the eye (Bergman 2010; Bulakowski et al. 2008; Hargis et al. 1992; Schultheiss 2004; Withrow & MacEwen 1996). Metastasis can involve any portion of the body, including the central nervous system. While most dogs present with splenic or liver lesions, some breeds show elevated incidence of

studies.

2004; Withrow & MacEwen 1996).

**6.2 Differential diagnosis** 

collection bias was a concern.

**6.3 Prognosis and treatment** 

examination and confirmation of clear surgical margins.

Clinical and Molecular Biology of Angiosarcoma 161

cAS may also occur on haired skin of darker colored dogs. Solar elastoses has been associated with cutaneous cAS in a large retrospective study involving many dogs from California practices (Hargis et al. 1992). A more recent retrospective study conducted at the University of Colorado did not confirm this finding (Schultheiss 2004), but the patient population for this study may have been comprised of more dogs from northern latitudes. The geographic location of the patients themselves was not annotated in either of these

The small, solitary or multiple dermal lesions of cutaneous cAS are often less than 1 cm in diameter, but may exceed 3 cm (Hargis et al. 1992). Mitotic figures are usually seen in malignant cAS. Conclusive diagnosis is always via histopathology, regardless of tumor location. Several markers can be used (e.g. CD31, factor VIII, Claudin-5, and cardiac troponin) to aid in confirmation of diagnosis (Chun et al. 2010; Jakab et al. 2009; Shaw et al.

As in human AS, the differential diagnosis of canine AS varies with the location of the disease. Visceral tumors must be distinguished from other neoplasms such as hemangiomas, other sarcomas, lymphoma, leiomyomas, and hematomas (Jakab et al. 2009). Signs of acute anemia, extreme lethargy, or collapse, requires exclusion of immune mediated hemolytic anemia (IMHA), trauma, coagulopathy, toxin exposure and allergic reaction. One retrospective study reported that among anemic dogs presenting with a splenic mass and abdominal bleeding requiring transfusion, 70% had splenic cAS (Hammond & Pesillo-Crosby 2008). Conclusive diagnosis of cAS requires histopathology of the excised tissue. Cardiac cAS must also be differentiated from other tumors and from idiopathic pericarditis (IP), both of which can cause pericardial effusion. In one retrospective study conducted in the UK, dogs with cardiac masses detected by ultrasound (echo-positive) had significantly shorter survival times than echo-negative dogs (Johnson 2004). Patients with a discernible cardiac mass, ascites, or collapse had a much less favorable prognosis than patients with IP. Another retrospective study reported that dogs with right atrial cAS treated surgically by pericardectomy and chemotherapy had a mean survival time of 164 days, compared to 46 days for dogs treated by pericardectomy alone (Weisse et al. 2005). However, this report also noted that dogs receiving chemotherapy were significantly younger and had significantly lower white blood cell counts than the dogs that did not receive chemotherapy, so case

Cutaneous cAS must be differentiated from other cutaneous tumors, nevi, and cysts. Histopathology is the only conclusive method of diagnosis. Because dogs may suffer from hemangiomas as well as hemangiosarcomas, and because depth of invasion can be a prognostic indicator, it is recommended that all tumors be submitted for histopathologic

The behavior of cAS is also linked to the site of origin, with visceral lesions being more aggressive than cutaneous lesions (Schultheiss 2004). For visceral disease, the high rate of early metastasis and the lack of early clinical signs or hematologic changes lead to late diagnosis and extremely poor prognosis. The average life span upon diagnosis of splenic cAS is less than three months for dogs treated with splenectomy alone (Spangler &

cutaneous (Whippets, Italian greyhounds) or cardiac (Saluki) cAS (Hargis et al. 1992; Prymak et al. 1988; Schultheiss 2004). This demonstrates that not only do genetic factors influence susceptibility to cAS, but also they determine the location at which the disease will present. From a research perspective, it opens new avenues for discovery of the basic mechanisms of the origin and progression of cancer that could be translated to further our understanding of the etiology, pathophysiology, and treatment of AS in human patients.

Canine AS is the most common malignancy of the spleen in dogs (Spangler & Culbertson 1992). Middle-aged, large (> 20kg) dogs are most prone to splenic cAS. Golden retrievers, German shepherds, and Labrador retrievers are overrepresented, as are other large breeds such as Clumber spaniels. A health survey conducted by the Golden Retriever Club of America/Golden Retriever Foundation reported that 16% of Golden retrievers with cancer die from cAS at an average age of 10 years (Glickman et al. 1999). Mixed breeds are also commonly affected. Dogs with visceral cAS typically present acutely, with profound lethargy, anemia, or sudden collapse. There may be intermittent, nonspecific clinical signs such as inappetance, weight loss and occasional episodes of weakness. Hemoperitoneum most often leads to diagnosis, often on an emergency basis.

Splenic tumor size varies, but the lesions are frequently very large and multinodular, with one to many saccular, blood filled lesions. The tissue is extremely friable, which often leads to spontaneous rupture. It is the acute loss of blood which causes the presenting signs. Dogs may recover from intra-abdominal hemorrhage through resorption of the blood (autotransfusion) and erythrocyte regeneration. However, there will frequently be further hemorrhaging. Anecdotally, more dogs are lost to euthanasia than to spontaneous hemorrhage, as the owners do not wish their dogs to suffer further such episodes.

In contrast to humans, in whom left atrial AS is more frequent (Casha et al. 2002), cardiac cAS most commonly occurs in the right atrium and dogs of the Saluki breed are particularly prone to development of such tumors. Right atrial cAS is also seen in Golden retrievers and German shepherd dogs. Cardiac cAS can be devastating from the outset, with many dogs presenting for cardiac tamponade and signs of right heart failure. The pressure of blood from a ruptured cAS in the right atrium (RA) fills the pericardial sac and leads to extreme respiratory distress, anxiety, and collapse. Alternatively, dogs may simply die acutely from cardiac arrest, and the disease is discovered upon necropsy of what seemed to be an otherwise healthy dog. Many cases are likely never conclusively diagnosed, as necropsy is infrequent in veterinary medicine.

Cardiac cAS are typically red to dark red or purple in color, and ranging from millimeters to centimeters in length. Such tumors are prone to spontaneous hemorrhage, which may lead to pericardial effusion and cardiac tamponade. While no definitive studies have delineated time course for these tumors, anecdotal reports suggest progression may be very rapid. Cardiac cAS can occur with or without splenic involvement and is more likely to present as a solitary lesion in some breeds such as the Saluki. It has not been determined if right atrial involvement is a primary tumor site or reflects a site of metastasis when concurrent visceral disease is present.

As in human AS, cutaneous cAS most often presents as solitary or multiple small dermal masses, which can extend into the subcutis or even into underlying musculature. They are frequently seen on non-haired skin on the ventral abdomen of light-colored dogs (Italian greyhounds, whippets, Dalmatians, pit bulls), especially when those dogs have excess exposure to solar radiation (southern latitudes, tendency towards sunbathing). Cutaneous

cutaneous (Whippets, Italian greyhounds) or cardiac (Saluki) cAS (Hargis et al. 1992; Prymak et al. 1988; Schultheiss 2004). This demonstrates that not only do genetic factors influence susceptibility to cAS, but also they determine the location at which the disease will present. From a research perspective, it opens new avenues for discovery of the basic mechanisms of the origin and progression of cancer that could be translated to further our understanding of the etiology, pathophysiology, and treatment of AS in human patients. Canine AS is the most common malignancy of the spleen in dogs (Spangler & Culbertson 1992). Middle-aged, large (> 20kg) dogs are most prone to splenic cAS. Golden retrievers, German shepherds, and Labrador retrievers are overrepresented, as are other large breeds such as Clumber spaniels. A health survey conducted by the Golden Retriever Club of America/Golden Retriever Foundation reported that 16% of Golden retrievers with cancer die from cAS at an average age of 10 years (Glickman et al. 1999). Mixed breeds are also commonly affected. Dogs with visceral cAS typically present acutely, with profound lethargy, anemia, or sudden collapse. There may be intermittent, nonspecific clinical signs such as inappetance, weight loss and occasional episodes of weakness. Hemoperitoneum

Splenic tumor size varies, but the lesions are frequently very large and multinodular, with one to many saccular, blood filled lesions. The tissue is extremely friable, which often leads to spontaneous rupture. It is the acute loss of blood which causes the presenting signs. Dogs may recover from intra-abdominal hemorrhage through resorption of the blood (autotransfusion) and erythrocyte regeneration. However, there will frequently be further hemorrhaging. Anecdotally, more dogs are lost to euthanasia than to spontaneous

In contrast to humans, in whom left atrial AS is more frequent (Casha et al. 2002), cardiac cAS most commonly occurs in the right atrium and dogs of the Saluki breed are particularly prone to development of such tumors. Right atrial cAS is also seen in Golden retrievers and German shepherd dogs. Cardiac cAS can be devastating from the outset, with many dogs presenting for cardiac tamponade and signs of right heart failure. The pressure of blood from a ruptured cAS in the right atrium (RA) fills the pericardial sac and leads to extreme respiratory distress, anxiety, and collapse. Alternatively, dogs may simply die acutely from cardiac arrest, and the disease is discovered upon necropsy of what seemed to be an otherwise healthy dog. Many cases are likely never conclusively diagnosed, as necropsy is

Cardiac cAS are typically red to dark red or purple in color, and ranging from millimeters to centimeters in length. Such tumors are prone to spontaneous hemorrhage, which may lead to pericardial effusion and cardiac tamponade. While no definitive studies have delineated time course for these tumors, anecdotal reports suggest progression may be very rapid. Cardiac cAS can occur with or without splenic involvement and is more likely to present as a solitary lesion in some breeds such as the Saluki. It has not been determined if right atrial involvement is a primary tumor site or reflects a site of metastasis when concurrent visceral

As in human AS, cutaneous cAS most often presents as solitary or multiple small dermal masses, which can extend into the subcutis or even into underlying musculature. They are frequently seen on non-haired skin on the ventral abdomen of light-colored dogs (Italian greyhounds, whippets, Dalmatians, pit bulls), especially when those dogs have excess exposure to solar radiation (southern latitudes, tendency towards sunbathing). Cutaneous

hemorrhage, as the owners do not wish their dogs to suffer further such episodes.

most often leads to diagnosis, often on an emergency basis.

infrequent in veterinary medicine.

disease is present.

cAS may also occur on haired skin of darker colored dogs. Solar elastoses has been associated with cutaneous cAS in a large retrospective study involving many dogs from California practices (Hargis et al. 1992). A more recent retrospective study conducted at the University of Colorado did not confirm this finding (Schultheiss 2004), but the patient population for this study may have been comprised of more dogs from northern latitudes. The geographic location of the patients themselves was not annotated in either of these studies.

The small, solitary or multiple dermal lesions of cutaneous cAS are often less than 1 cm in diameter, but may exceed 3 cm (Hargis et al. 1992). Mitotic figures are usually seen in malignant cAS. Conclusive diagnosis is always via histopathology, regardless of tumor location. Several markers can be used (e.g. CD31, factor VIII, Claudin-5, and cardiac troponin) to aid in confirmation of diagnosis (Chun et al. 2010; Jakab et al. 2009; Shaw et al. 2004; Withrow & MacEwen 1996).

## **6.2 Differential diagnosis**

As in human AS, the differential diagnosis of canine AS varies with the location of the disease. Visceral tumors must be distinguished from other neoplasms such as hemangiomas, other sarcomas, lymphoma, leiomyomas, and hematomas (Jakab et al. 2009). Signs of acute anemia, extreme lethargy, or collapse, requires exclusion of immune mediated hemolytic anemia (IMHA), trauma, coagulopathy, toxin exposure and allergic reaction. One retrospective study reported that among anemic dogs presenting with a splenic mass and abdominal bleeding requiring transfusion, 70% had splenic cAS (Hammond & Pesillo-Crosby 2008). Conclusive diagnosis of cAS requires histopathology of the excised tissue.

Cardiac cAS must also be differentiated from other tumors and from idiopathic pericarditis (IP), both of which can cause pericardial effusion. In one retrospective study conducted in the UK, dogs with cardiac masses detected by ultrasound (echo-positive) had significantly shorter survival times than echo-negative dogs (Johnson 2004). Patients with a discernible cardiac mass, ascites, or collapse had a much less favorable prognosis than patients with IP. Another retrospective study reported that dogs with right atrial cAS treated surgically by pericardectomy and chemotherapy had a mean survival time of 164 days, compared to 46 days for dogs treated by pericardectomy alone (Weisse et al. 2005). However, this report also noted that dogs receiving chemotherapy were significantly younger and had significantly lower white blood cell counts than the dogs that did not receive chemotherapy, so case collection bias was a concern.

Cutaneous cAS must be differentiated from other cutaneous tumors, nevi, and cysts. Histopathology is the only conclusive method of diagnosis. Because dogs may suffer from hemangiomas as well as hemangiosarcomas, and because depth of invasion can be a prognostic indicator, it is recommended that all tumors be submitted for histopathologic examination and confirmation of clear surgical margins.

#### **6.3 Prognosis and treatment**

The behavior of cAS is also linked to the site of origin, with visceral lesions being more aggressive than cutaneous lesions (Schultheiss 2004). For visceral disease, the high rate of early metastasis and the lack of early clinical signs or hematologic changes lead to late diagnosis and extremely poor prognosis. The average life span upon diagnosis of splenic cAS is less than three months for dogs treated with splenectomy alone (Spangler &

Clinical and Molecular Biology of Angiosarcoma 163

Molecular insight into cAS is limited and focuses primarily on visceral disease, but what is known is consistent with the biology of human AS. Tumor-derived cell lines and cell isolates express endothelial and hematopoietic stem cell markers, suggesting they may arise from pluripotent bone marrow-derived stem cells and/or cells committed to the endothelial lineage (Fosmire et al. 2004; Lamerato-Kozicki et al. 2006; Tamburini BA 2010; Thamm et al. 2006). Cells form branching vascular structures when grown on Matrigel (Thamm et al. 2006), secrete pro-angiogenic factors including VEGF and bFGF , and express elevated levels of Ang-1 and Ang-2 mRNA (Kato et al. 2006). Upon injection into immune-compromised mice, such cells readily form AS-like tumors expressing endothelial markers including CD31, vWF, VEGF-A, bFGF, flt-1, flk-1, FGFR-1, HoxA9, HoxB3, HoxB7, HoxD3, Pbx1, and Meis1 (Kodama et al. 2009). Although mutations in p53 have not been detected, cAS stain positively for nuclear p53 (Yonemaru et al. 2007). As in the human disease, cAS lesions have been found to harbor point mutations or deletions in the C-terminal domain of PTEN

The rarity of angiosarcoma has had a profound impact on its clinical management and research. Most clinicians are unfamiliar with this disease and thus do not have experience to guide management decisions. Much of the literature regarding these rare tumors is comprised of anecdotal case reports or retrospective analyses which are difficult to assimilate and interpret. Thus, we must be especially wary not to assume all AS are equivalent. The varied morphology and presentation of these lesions indicates we may be dealing with not one, but several diseases with unique physiologic or clinical properties and pharmacologic responsiveness. For instance, it has been reported that tumor behavior might depend on the site of origin, with superficial tumors associating with a longer progressionfree survival after initial treatment (Fury et al. 2005; Schlemmer et al. 2008). However, it is unclear whether this is caused by intrinsic biologic differences or differences in clinical presentation and prior treatment. Unfortunately, when AS is examined prospectively in clinical trials it is usually included as a limited number of cases within a larger study of various soft-tissue sarcomas. Further progress in our understanding of the biology and treatment of AS depends on the willingness of the medical research and clinical oncology communities to share resources through multi-center collaborations. In the absence of such efforts, alternative strategies are needed to generate meaningful insight into this disease. Significant insight into the biology of AS has been generated through translational studies of related human tumors and mouse models. Collectively, studies of Kaposi sarcoma and hemangiomas, plus studies conducted in mouse models provide strong support for the hypothesis that MAPK signaling is a key event in the induction of endothelial cell proliferation and AS in mice and humans. This pathway has been invoked to explain the origins of many other tumor types. MAPK signaling pathways are frequently activated in endothelial cells during developmental and pathologic angiogenesis (reviewed in (Depeille et al. 2007)). Further, MAPK signaling is physiologically activated in response to shear stress (Azuma et al. 2000) and in pre-atherosclerotic plaques (Muslin 2008). A key question that needs to be addressed may therefore be not how these events trigger AS, but why they do not trigger AS more frequently. An examination of the role of MAPK phosphatases in

endothelial cells and in AS tumorigenesis may be instructive.

**6.4 Cell and molecular biology** 

(Dickerson et al. 2005).

**7. Conclusions** 

Culbertson 1992). Dogs treated with adjunctive chemotherapy (usually doxorubicin-based protocols) average less than six months survival (Kim et al. 2007; Sorenmo et al. 2004). Owners frequently elect to euthanize the dog at the time of diagnosis or shortly thereafter.

Cardiac cAS is equally grim, with many dogs dying of acute cardiac arrest prior to any overt clinical signs. This sudden death syndrome is seen particularly among Saluki. Dogs may appear clinically normal and sometimes die in their sleep. Others may develop exercise intolerance, weakness, lethargy, and inappetance prior to developing ascites secondary to cardiac tamponade. One retrospective study of 23 cases of right atrial cAS treated surgically, with or without chemotherapy, reported a mean survival time of 175 days or 42 days, respectively (Weisse et al. 2005). Another retrospective study from the UK which evaluated dogs with pericardial effusion reported a mean survival time of 26 days for dogs with echocardiograms revealing a cardiac mass versus 1068 days for those with effusion but no detectable mass. However, in this study, no dogs with detectable masses had treatment beyond pericardiocentesis, and many were euthanized at diagnosis (Johnson 2004).

In stark contrast is the positive prognosis for cutaneous cAS. A majority of dogs with strictly cutaneous cAS survive many years with local recurrent disease (Schultheiss 2004). Schultheiss reported that complete surgical resection of cutaneous cAS was the most important prognostic factor for survival (Schultheiss 2004). A retrospective study of 25 dogs reported that those with cutaneous cAS confined to the dermis had a mean survival time of 780 days, compared with 172 and 307 days for dogs with a primary tumor involving the hypodermis or underlying muscle, respectively (Ward et al. 1994). Confinement to the dermis also was correlated with better outcome, most likely due to complete surgical resection (Schultheiss 2004).

The response of cAS tumors to treatment varies greatly depending on the site of the tumor. Dogs with splenic tumors survive a few months beyond splenectomy but live longer than those that do not have surgery. Survival time is measured in months regardless of treatment, but owners frequently elect splenectomy for palliative care, extension of quality time, and reduction of the worry regarding continued intra-abdominal hemorrhage. The addition of doxorubicin-based chemotherapy protocols may add a few more months of survival. Single agent doxorubicin, administered intravenously every three weeks for up to five cycles is commonly used if chemotherapy is elected. Alternate protocols include doxorubicin plus cyclophosphamide, or doxorubicin, vincristine and cyclosphosphamide. No protocol is superior. Doxorubicin can induce dose-dependent cardiotoxicities in dogs; the total cumulative dose should not exceed 180-240 mg/m2. A study evaluating the potential use of epirubicin in splenectomized cAS patients reported no cardiac toxicity *(Kim et al. 2007)*. However, gastrointestinal side effects occurred in a majority of the patients, and neutropenia was also common. The expense of this treatment may also preclude this drug from becoming more commonly used, although it may provide an option for dogs with preexisting cardiac disease *(Kim et al. 2007)*.

Cutaneous cAS is generally treated with complete surgical excision. There are minimal longterm consequences apart from the necessity of multiple surgeries often being required. This is not generally due to recurrence of excised tumors; rather, patients tend to develop tumors spontaneously in multiple sites. It is unclear whether these represent new primary tumors or metastatic disease (Hargis et al. 1992; Ward et al. 1994). Multiple surgeries can create a financial hardship for the owner, and may be the cause for dogs to be relinquished to rescue organizations or euthanized.

#### **6.4 Cell and molecular biology**

162 Soft Tissue Tumors

Culbertson 1992). Dogs treated with adjunctive chemotherapy (usually doxorubicin-based protocols) average less than six months survival (Kim et al. 2007; Sorenmo et al. 2004). Owners frequently elect to euthanize the dog at the time of diagnosis or shortly thereafter. Cardiac cAS is equally grim, with many dogs dying of acute cardiac arrest prior to any overt clinical signs. This sudden death syndrome is seen particularly among Saluki. Dogs may appear clinically normal and sometimes die in their sleep. Others may develop exercise intolerance, weakness, lethargy, and inappetance prior to developing ascites secondary to cardiac tamponade. One retrospective study of 23 cases of right atrial cAS treated surgically, with or without chemotherapy, reported a mean survival time of 175 days or 42 days, respectively (Weisse et al. 2005). Another retrospective study from the UK which evaluated dogs with pericardial effusion reported a mean survival time of 26 days for dogs with echocardiograms revealing a cardiac mass versus 1068 days for those with effusion but no detectable mass. However, in this study, no dogs with detectable masses had treatment

beyond pericardiocentesis, and many were euthanized at diagnosis (Johnson 2004).

resection (Schultheiss 2004).

preexisting cardiac disease *(Kim et al. 2007)*.

organizations or euthanized.

In stark contrast is the positive prognosis for cutaneous cAS. A majority of dogs with strictly cutaneous cAS survive many years with local recurrent disease (Schultheiss 2004). Schultheiss reported that complete surgical resection of cutaneous cAS was the most important prognostic factor for survival (Schultheiss 2004). A retrospective study of 25 dogs reported that those with cutaneous cAS confined to the dermis had a mean survival time of 780 days, compared with 172 and 307 days for dogs with a primary tumor involving the hypodermis or underlying muscle, respectively (Ward et al. 1994). Confinement to the dermis also was correlated with better outcome, most likely due to complete surgical

The response of cAS tumors to treatment varies greatly depending on the site of the tumor. Dogs with splenic tumors survive a few months beyond splenectomy but live longer than those that do not have surgery. Survival time is measured in months regardless of treatment, but owners frequently elect splenectomy for palliative care, extension of quality time, and reduction of the worry regarding continued intra-abdominal hemorrhage. The addition of doxorubicin-based chemotherapy protocols may add a few more months of survival. Single agent doxorubicin, administered intravenously every three weeks for up to five cycles is commonly used if chemotherapy is elected. Alternate protocols include doxorubicin plus cyclophosphamide, or doxorubicin, vincristine and cyclosphosphamide. No protocol is superior. Doxorubicin can induce dose-dependent cardiotoxicities in dogs; the total cumulative dose should not exceed 180-240 mg/m2. A study evaluating the potential use of epirubicin in splenectomized cAS patients reported no cardiac toxicity *(Kim et al. 2007)*. However, gastrointestinal side effects occurred in a majority of the patients, and neutropenia was also common. The expense of this treatment may also preclude this drug from becoming more commonly used, although it may provide an option for dogs with

Cutaneous cAS is generally treated with complete surgical excision. There are minimal longterm consequences apart from the necessity of multiple surgeries often being required. This is not generally due to recurrence of excised tumors; rather, patients tend to develop tumors spontaneously in multiple sites. It is unclear whether these represent new primary tumors or metastatic disease (Hargis et al. 1992; Ward et al. 1994). Multiple surgeries can create a financial hardship for the owner, and may be the cause for dogs to be relinquished to rescue Molecular insight into cAS is limited and focuses primarily on visceral disease, but what is known is consistent with the biology of human AS. Tumor-derived cell lines and cell isolates express endothelial and hematopoietic stem cell markers, suggesting they may arise from pluripotent bone marrow-derived stem cells and/or cells committed to the endothelial lineage (Fosmire et al. 2004; Lamerato-Kozicki et al. 2006; Tamburini BA 2010; Thamm et al. 2006). Cells form branching vascular structures when grown on Matrigel (Thamm et al. 2006), secrete pro-angiogenic factors including VEGF and bFGF , and express elevated levels of Ang-1 and Ang-2 mRNA (Kato et al. 2006). Upon injection into immune-compromised mice, such cells readily form AS-like tumors expressing endothelial markers including CD31, vWF, VEGF-A, bFGF, flt-1, flk-1, FGFR-1, HoxA9, HoxB3, HoxB7, HoxD3, Pbx1, and Meis1 (Kodama et al. 2009). Although mutations in p53 have not been detected, cAS stain positively for nuclear p53 (Yonemaru et al. 2007). As in the human disease, cAS lesions have been found to harbor point mutations or deletions in the C-terminal domain of PTEN (Dickerson et al. 2005).

## **7. Conclusions**

The rarity of angiosarcoma has had a profound impact on its clinical management and research. Most clinicians are unfamiliar with this disease and thus do not have experience to guide management decisions. Much of the literature regarding these rare tumors is comprised of anecdotal case reports or retrospective analyses which are difficult to assimilate and interpret. Thus, we must be especially wary not to assume all AS are equivalent. The varied morphology and presentation of these lesions indicates we may be dealing with not one, but several diseases with unique physiologic or clinical properties and pharmacologic responsiveness. For instance, it has been reported that tumor behavior might depend on the site of origin, with superficial tumors associating with a longer progressionfree survival after initial treatment (Fury et al. 2005; Schlemmer et al. 2008). However, it is unclear whether this is caused by intrinsic biologic differences or differences in clinical presentation and prior treatment. Unfortunately, when AS is examined prospectively in clinical trials it is usually included as a limited number of cases within a larger study of various soft-tissue sarcomas. Further progress in our understanding of the biology and treatment of AS depends on the willingness of the medical research and clinical oncology communities to share resources through multi-center collaborations. In the absence of such efforts, alternative strategies are needed to generate meaningful insight into this disease.

Significant insight into the biology of AS has been generated through translational studies of related human tumors and mouse models. Collectively, studies of Kaposi sarcoma and hemangiomas, plus studies conducted in mouse models provide strong support for the hypothesis that MAPK signaling is a key event in the induction of endothelial cell proliferation and AS in mice and humans. This pathway has been invoked to explain the origins of many other tumor types. MAPK signaling pathways are frequently activated in endothelial cells during developmental and pathologic angiogenesis (reviewed in (Depeille et al. 2007)). Further, MAPK signaling is physiologically activated in response to shear stress (Azuma et al. 2000) and in pre-atherosclerotic plaques (Muslin 2008). A key question that needs to be addressed may therefore be not how these events trigger AS, but why they do not trigger AS more frequently. An examination of the role of MAPK phosphatases in endothelial cells and in AS tumorigenesis may be instructive.

Clinical and Molecular Biology of Angiosarcoma 165

Aust, M. R., K. D. Olsen, J. E. Lewis, A. G. Nascimento, N. B. Meland, R. L. Foote & V. J.

Autenrieth, J. & S. Lange (1979). [Thorotrastoses--viewed in retrospect (author's transl)].

Azuma, N., S. A. Duzgun, M. Ikeda, H. Kito, N. Akasaka, T. Sasajima & B. E. Sumpio (2000). Endothelial cell response to different mechanical forces. *J Vasc Surg* 32(4): 789-94. Bais, C., B. Santomasso, O. Coso, L. Arvanitakis, E. G. Raaka, J. S. Gutkind, A. S. Asch, E.

Bergman, P. (2010). Hemangiosarcoma. In: *Textbook of Veterinary Internal Medicine*. S. J. a. F.

Birchmeier, C., W. Birchmeier, E. Gherardi & G. F. Vande Woude (2003). Met, metastasis,

Bolt, H. M. (2005). Vinyl chloride-a classical industrial toxicant of new interest. *Crit Rev* 

Botros, M., J. F. Quevedo & R. C. Miler (2009). Angiosarcoma of the liver after multimodality

Brash, D. & A. Bale (2001). Cancer of the Skin. In: *Cancer. Principles and practice of oncology*.

Breiteneder-Geleff, S., A. Soleiman, H. Kowalski, R. Horvat, G. Amann, E. Kriehuber, K.

Brennan, M. F., K. M. Alektiar & R. G. Maki (2001). Sarcomas of the soft tissue and bone. In:

Bulakowski, E. J., J. C. Philibert, S. Siegel, C. A. Clifford, R. Risbon, K. Zivin & K. L. Cronin

Burger, M., J. A. Burger, R. C. Hoch, Z. Oades, H. Takamori & I. U. Schraufstatter (1999).

Cannon, M., N. J. Philpott & E. Cesarman (2003). The Kaposi's sarcoma-associated

Cannon, M. L. & E. Cesarman (2004). The KSHV G protein-coupled receptor signals via

Casha, A. R., L. A. Davidson, P. Roberts & R. U. Nair (2002). Familial angiosarcoma of the

V. DeVita Jr., S. Hellman&S. Rosenberg. Philadelpia, Lippincott Williams &

Diem, W. Weninger, E. Tschachler, K. Alitalo & D. Kerjaschki (1999). Angiosarcomas express mixed endothelial phenotypes of blood and lymphatic capillaries: podoplanin as a specific marker for lymphatic endothelium. *Am J Pathol*

*Cancer. Principles & Practice of Oncology*. V. DeVita Jr., S. Hellman&S. Rosenberg.

(2008). Evaluation of outcome associated with subcutaneous and intramuscular hemangiosarcoma treated with adjuvant doxorubicin in dogs: 21 cases (2001-2006).

Point mutation causing constitutive signaling of CXCR2 leads to transforming activity similar to Kaposi's sarcoma herpesvirus-G protein-coupled receptor. *J* 

herpesvirus G protein-coupled receptor has broad signaling effects in primary

multiple pathways to induce transcription factor activation in primary effusion

therapy for gallbladder carcinoma. *Radiol Oncol* 43(2): 126-131.

Philadelphia, Lippincott Williams & Wilkins: 1841-1891.

characteristics. *Ann Otol Rhinol Laryngol* 106(11): 943-51.

and angiogenesis activator. *Nature* 391(6662): 86-9.

Ettinger, E.C. St. Louis, MO, Saunders. 2: 2175-2180.

motility and more. *Mol Biol Cell* 4: 915-925.

*Rontgenblatter* 32(2): 71-4.

*Toxicol* 35(4): 307-23.

Wilkins: 1971-2002.

154(2): 385-94.

*J Am Vet Med Assoc* 233(1): 122-8.

effusion lymphoma cells. *J Virol* 77(1): 57-67.

lymphoma cells. *Oncogene* 23(2): 514-23.

heart. *J Thorac Cardiovasc Surg* 124(2): 392-4.

*Immunol* 163(4): 2017-22.

Suman (1997). Angiosarcomas of the head and neck: clinical and pathologic

Cesarman, M. C. Gershengorn, E. A. Mesri & M. C. Gerhengorn (1998). G-proteincoupled receptor of Kaposi's sarcoma-associated herpesvirus is a viral oncogene

However, important questions remain to be asked and answered. What are the key biochemical pathways involved in this disease? Does misregulation of MAPK signaling by phosphatases allow constitutive signaling to promote tumorigenesis? Do genetic mutations contribute to or establish a microenvironment that predisposes scalp, spleen, liver, or heart to develop AS? Why is angiosarcoma infrequent in humans, while it is more prevalent in dogs? Can we identify molecular targets for drug therapies? Can we devise a fundable and meaningful clinical trial to test novel therapies for such a rare disease? Will the use of spontaneous animal tumor models prove useful in informing human clinical trials?

The canine model is of growing interest in this regard since it closely resembles human disease. Studies of spontaneous cAS offer a clear experimental advantage through genomewide association studies of single nucleotide polymorphisms that provide an unbiased approach to identify the genes contributing to the onset and progression of disease in a compressed time frame. This, combined with the availability of tumor material and the presence of multi-institution consortia such as the Canine Comparative Oncology and Genomics Consortium, the LUPA consortium, the Canine Hereditary Cancer Consortium, and the Canine Oncology Trials Consortium, may facilitate rapid identification and translation of novel therapies through clinical trials that may benefit dogs and humans alike.

## **8. Acknowledgements**

We thank Drs. Jennifer Bromberg-White and Chih-Shia Lee for their critical comments on the manuscript, as well as Ms. Diana Lewis for administrative support and Mr. David Nadziejka for scientific editing of the manuscript. We also acknowledge financial support from the National Institutes of Health/National Cancer Institute (RC2CA148149), the Animal Cancer Foundation, and the Dwight Reed Memorial Foundation.

#### **9. References**


However, important questions remain to be asked and answered. What are the key biochemical pathways involved in this disease? Does misregulation of MAPK signaling by phosphatases allow constitutive signaling to promote tumorigenesis? Do genetic mutations contribute to or establish a microenvironment that predisposes scalp, spleen, liver, or heart to develop AS? Why is angiosarcoma infrequent in humans, while it is more prevalent in dogs? Can we identify molecular targets for drug therapies? Can we devise a fundable and meaningful clinical trial to test novel therapies for such a rare disease? Will the use of

The canine model is of growing interest in this regard since it closely resembles human disease. Studies of spontaneous cAS offer a clear experimental advantage through genomewide association studies of single nucleotide polymorphisms that provide an unbiased approach to identify the genes contributing to the onset and progression of disease in a compressed time frame. This, combined with the availability of tumor material and the presence of multi-institution consortia such as the Canine Comparative Oncology and Genomics Consortium, the LUPA consortium, the Canine Hereditary Cancer Consortium, and the Canine Oncology Trials Consortium, may facilitate rapid identification and translation of novel therapies through clinical trials that may benefit dogs and humans alike.

We thank Drs. Jennifer Bromberg-White and Chih-Shia Lee for their critical comments on the manuscript, as well as Ms. Diana Lewis for administrative support and Mr. David Nadziejka for scientific editing of the manuscript. We also acknowledge financial support from the National Institutes of Health/National Cancer Institute (RC2CA148149), the

Ablashi, D. V., L. G. Chatlynne, J. E. Whitman, Jr. & E. Cesarman (2002). Spectrum of

Antonescu, C. R., A. Yoshida, T. Guo, N. E. Chang, L. Zhang, N. P. Agaram, L. X. Qin, M. F.

Armah, H. B., U. N. Rao & A. V. Parwani (2007). Primary angiosarcoma of the testis: report

Arora, R., A. Sharma, R. Gupta & M. Vijayaraghavan (2008). Cutaneous angiosarcoma in a patient with xeroderma pigmentosum. *Indian J Pathol Microbiol* 51(4): 504-6. Arvanitakis, L., E. Geras-Raaka, A. Varma, M. C. Gershengorn & E. Cesarman (1997).

of a rare entity and review of the literature. *Diagn Pathol* 2: 23.

receptor linked to cell proliferation. *Nature* 385(6614): 347-50.

Kaposi's sarcoma-associated herpesvirus, or human herpesvirus 8, diseases. *Clin* 

Brennan, S. Singer & R. G. Maki (2009). KDR activating mutations in human angiosarcomas are sensitive to specific kinase inhibitors. *Cancer Res* 69(18): 7175-9. Arlett, C. F., P. N. Plowman, P. B. Rogers, C. N. Parris, F. Abbaszadeh, M. H. Green, T. J.

McMillan, C. Bush, N. Foray & A. R. Lehmann (2006). Clinical and cellular ionizing radiation sensitivity in a patient with xeroderma pigmentosum. *Br J Radiol* 79(942):

Human herpesvirus KSHV encodes a constitutively active G-protein-coupled

Animal Cancer Foundation, and the Dwight Reed Memorial Foundation.

spontaneous animal tumor models prove useful in informing human clinical trials?

**8. Acknowledgements** 

*Microbiol Rev* 15(3): 439-64.

**9. References** 

510-7.


Clinical and Molecular Biology of Angiosarcoma 167

Eaton, G. J., F. N. Johnson, R. P. Custer & A. R. Crane (1980). The Icr:Ha(ICR) mouse: a

Eder, J. P., G. F. Vande Woude, S. A. Boerner & P. M. LoRusso (2009). Novel therapeutic inhibitors of the c-Met signaling pathway in cancer. *Clin Cancer Res* 15(7): 2207-14. Falk, S., J. Krishnan & J. M. Meis (1993). Primary angiosarcoma of the spleen. A

Farag, R., J. A. Schulak, F. W. Abdul-Karim & J. K. Wasman (2005). Angiosarcoma arising in

Farhood, A. I., S. I. Hajdu, M. H. Shiu & E. W. Strong (1990). Soft tissue sarcomas of the head

Fata, F., E. O'Reilly, D. Ilson, D. Pfister, D. Leffel, D. P. Kelsen, G. K. Schwartz & E. S. Casper

Fayette, J., E. Martin, S. Piperno-Neumann, A. Le Cesne, C. Robert, S. Bonvalot, D. Ranchere,

Fletcher, C. D., A. Beham, S. Bekir, A. M. Clarke & N. J. Marley (1991). Epithelioid

Flora, S. J. (2011). Arsenic-induced oxidative stress and its reversibility. *Free Radic Biol Med*

Fosmire, S. P., E. B. Dickerson, A. M. Scott, S. R. Bianco, M. J. Pettengill, H. Meylemans, M.

Fury, M. G., C. R. Antonescu, K. J. Van Zee, M. F. Brennan & R. G. Maki (2005). A 14-year

Gherghe, C. M., J. Duan, J. Gong, M. Rojas, N. Klauber-Demore, M. Majesky & A. Deb

Gil-Benso, R., C. Lopez-Gines, P. Soriano, S. Almenar, C. Vazquez & A. Llombart-Bosch

Glazebrook, K. N., M. J. Magut & C. Reynolds (2008). Angiosarcoma of the breast. *AJR Am J* 

Glickman, L. T., N. Glickman & R. Thorpe (1999). The Golden Retriever Club of America

retrospective study of 161 cases. *Ann Oncol* 18(12): 2030-6.

primitive angiogenic endothelium. *Lab Invest* 84(5): 562-72.

epithelial neoplasm. *Am J Surg Pathol* 15(10): 915-24.

clinicopathologic study of 40 cases. *Am J Surg Pathol* 17(10): 959-70.

literature review. *Clin Nephrol* 63(5): 408-12.

and neck in adults. *Am J Surg* 160(4): 365-9.

face. *Cancer* 86(10): 2034-7.

51(2): 257-81.

*Diagn Pathol*.

10(3): 210-2.

*Faseb J* 25(6): 1836-43.

*Roentgenol* 190(2): 533-8.

National Health Survey: 182.

24.

current account of breeding, mutations, diseases and mortality. *Lab Anim* 14(1): 17-

an arteriovenous fistula site in a renal transplant patient: a case report and

(1999). Paclitaxel in the treatment of patients with angiosarcoma of the scalp or

P. Pouillart, J. M. Coindre & J. Y. Blay (2007). Angiosarcomas, a heterogeneous group of sarcomas with specific behavior depending on primary site: a

angiosarcoma of deep soft tissue: a distinctive tumor readily mistaken for an

Padilla, A. A. Frazer-Abel, N. Akhtar, D. M. Getzy, J. Wojcieszyn, M. Breen, S. C. Helfand & J. F. Modiano (2004). Canine malignant hemangiosarcoma as a model of

retrospective review of angiosarcoma: clinical characteristics, prognostic factors, and treatment outcomes with surgery and chemotherapy. *Cancer J* 11(3): 241-7. Ge, Y., J. Y. Ro, D. Kim, C. H. Kim, M. J. Reardon, S. Blackmon, J. Zhai, D. Coffey, R. S.

Benjamin & A. G. Ayala (2011). Clinicopathologic and immunohistochemical characteristics of adult primary cardiac angiosarcomas: analysis of 10 cases. *Ann* 

(2011). Wnt1 is a proangiogenic molecule, enhances human endothelial progenitor function, and increases blood flow to ischemic limbs in a HGF-dependent manner.

(1994). Cytogenetic study of angiosarcoma of the breast. *Genes Chromosomes Cancer*


Cattel, L., M. Ceruti & F. Dosio (2003). From conventional to stealth liposomes: a new

Centeno, J. A., F. G. Mullick, L. Martinez, N. P. Page, H. Gibb, D. Longfellow, C. Thompson

Chandra, M. & C. H. Frith (1992). Spontaneous neoplasms in aged CD-1 mice. *Toxicol Lett*

Chandra, M. & C. H. Frith (1992). Spontaneous neoplasms in B6C3F1 mice. *Toxicol Lett* 60(1):

Chang, Y., E. Cesarman, M. S. Pessin, F. Lee, J. Culpepper, D. M. Knowles & P. S. Moore

Chesterman, F. C., J. J. Harvey, R. R. Dourmashkin & M. H. Salaman (1966). The pathology

Chlebowski, R. T. (1979). Adriamycin (doxorubicin) cardiotoxicity: a review. *West J Med*

Chun, R., H. B. Kellihan, R. A. Henik & R. L. Stepien (2010). Comparison of plasma cardiac

Cohen, S. M., R. D. Storer, K. A. Criswell, N. G. Doerrer, V. L. Dellarco, D. G. Pegg, Z. W.

Corthals, S. M., L. M. Kamendulis & J. E. Klaunig (2006). Mechanisms of 2-butoxyethanol-

Creech, J. L., Jr. & M. N. Johnson (1974). Angiosarcoma of liver in the manufacture of

de Boer, J. & J. H. Hoeijmakers (2000). Nucleotide excision repair and human syndromes.

Depeille, P., J. J. Young, E. A. Boguslawski, B. D. Berghuis, E. J. Kort, J. H. Resau, A. E.

Depeille, P. E., Y. Ding, J. L. Bromberg-White & N. S. Duesbery (2007). MKK signaling and

Dickerson, E. B., R. Thomas, S. P. Fosmire, A. R. Lamerato-Kozicki, S. R. Bianco, J. W.

Dobson, J. M., S. Samuel, H. Milstein, K. Rogers & J. L. Wood (2002). Canine neoplasia in the

Frankel & N. S. Duesbery (2007). Anthrax lethal toxin inhibits growth of and vascular endothelial growth factor release from endothelial cells expressing the human herpes virus 8 viral G protein coupled receptor. *Clin Cancer Res* 13(19): 5926-

Wojcieszyn, M. Breen, S. C. Helfand & J. F. Modiano (2005). Mutations of phosphatase and tensin homolog deleted from chromosome 10 in canine

UK: estimates of incidence rates from a population of insured dogs. *J Small Anim* 

nonhemangiosarcoma origin. *J Am Vet Med Assoc* 237(7): 806-11.

induced hemangiosarcomas. *Toxicol Sci* 92(2): 378-86.

polyvinyl chloride. *J Occup Med* 16(3): 150-1.

vascularization. *Oncogene* 26(9): 1290-6.

hemangiosarcoma. *Vet Pathol* 42(5): 618-32.

& E. R. Ladich (2002). Pathology related to chronic arsenic exposure. *Environ Health* 

(1994). Identification of herpesvirus-like DNA sequences in AIDS-associated

of tumors and other lesions induced in rodents by virus derived from a rat with

troponin I concentrations among dogs with cardiac hemangiosarcoma, noncardiac hemangiosarcoma, other neoplasms, and pericardial effusion of

Wojcinski, D. E. Malarkey, A. C. Jacobs, J. E. Klaunig, J. A. Swenberg & J. C. Cook (2009). Hemangiosarcoma in rodents: mode-of-action evaluation and human

frontier in cancer chemotherapy. *Tumori* 89(3): 237-49.

Kaposi's sarcoma. *Science* 266(5192): 1865-9.

Moloney leukemia. *Cancer Res* 26(8): 1759-68.

relevance. *Toxicol Sci* 111(1): 4-18.

*Carcinogenesis* 21(3): 453-60.

34.

*Pract* 43(6): 240-6.

*Perspect* 110 Suppl 5: 883-6.

61(1): 67-74.

131(5): 364-8.

91-8.


Clinical and Molecular Biology of Angiosarcoma 169

Johnson, S. A., C. J. Bateman, M. E. Beard, J. M. Whitehouse & A. H. Waters (1977). Longterm haematological complications of thorotrast. *Q J Med* 46(182): 259-71. Judson, I., J. A. Radford, M. Harris, J. Y. Blay, Q. van Hoesel, A. le Cesne, A. T. van

Kaelin, W. G., Jr. (2007). The von hippel-lindau tumor suppressor protein: an update.

Kamendulis, L., S. Corthals & J. Klaunig (2010). Kupffer cells participate in 2-butoxyethanol-

Karlsson, P., E. Holmberg, A. Samuelsson, K. A. Johansson & A. Wallgren (1998). Soft tissue

Kato, Y., K. Asano, I. Mizutani, T. Konno, Y. Sasaki, K. Kutara, K. Teshima, K. Edamura, R.

Kemeny, L., M. Kiss, R. Gyulai, A. S. Kenderessy, E. Adam, F. Nagy & A. Dobozy (1996).

Kim, S. E., J. M. Liptak, T. T. Gall, G. J. Monteith & J. P. Woods (2007). Epirubicin in the

Kindblom, L. G., G. Stenman & L. Angervall (1991). Morphological and cytogenetic studies

Kiyosawa, K., H. Imai, T. Sodeyama, S. T. Franca, M. Yousuf, S. Furuta, K. Fujisawa & C.

Koch, M., G. P. Nielsen & S. S. Yoon (2008). Malignant tumors of blood vessels:

Kodama, A., H. Sakai, S. Matsuura, M. Murakami, A. Murai, T. Mori, K. Maruo, T. Kimura,

Kraemer, K. H., M. M. Lee & J. Scotto (1987). Xeroderma pigmentosum. Cutaneous, ocular, and neurologic abnormalities in 830 published cases. *Arch Dermatol* 123(2): 241-50.

associated homeobox genes. *BMC Cancer* 9: 363.

sarcoma after treatment for breast cancer--a Swedish population-based study. *Eur J* 

Kano, K. Suzuki, H. Shibuya, T. Sato, A. Hasegawa & S. Tanaka (2006). Gene expressions of canine angiopoietin-1 and -2 in normal tissues and spontaneous

Human herpesvirus 8 in classic Kaposi sarcoma. *Acta Microbiol Immunol Hung* 43(4):

adjuvant treatment of splenic hemangiosarcoma in dogs: 59 cases (1997-2004). *J Am* 

of angiosarcoma in Stewart-Treves syndrome. *Virchows Arch A Pathol Anat* 

Kido (1989). Comparison of anamnestic history, alcohol intake and smoking, nutritional status, and liver dysfunction between thorotrast patients who developed primary liver cancer and those who did not. *Environ Res* 49(2): 166-72. Klaunig, J. E. & L. M. Kamendulis (2005). Mode of action of butoxyethanol-induced mouse liver hemangiosarcomas and hepatocellular carcinomas. *Toxicol Lett* 156(1): 107-15. Kleinman, M. E., M. R. Greives, S. S. Churgin, K. M. Blechman, E. I. Chang, D. J. Ceradini, O.

M. Tepper & G. C. Gurtner (2007). Hypoxia-induced mediators of stem/progenitor cell trafficking are increased in children with hemangioma. *Arterioscler Thromb Vasc* 

angiosarcomas, hemangioendotheliomas, and hemangioperictyomas. *J Surg Oncol*

T. Masegi & T. Yanai (2009). Establishment of canine hemangiosarcoma xenograft models expressing endothelial growth factors, their receptors, and angiogenesis-

induced liver hemangiosarcomas. *Toxicology* 270(2-3): 131-136.

Sarcoma Group. *Eur J Cancer* 37(7): 870-7.

*Methods Enzymol* 435: 371-83.

tumours. *Res Vet Sci* 81(2): 280-6.

*Vet Med Assoc* 231(10): 1550-7.

*Histopathol* 419(5): 439-45.

*Biol* 27(12): 2664-70.

97(4): 321-9.

*Cancer* 34(13): 2068-75.

391-5.

Oosterom, M. J. Clemons, C. Kamby, C. Hermans, J. Whittaker, E. Donato di Paola, J. Verweij & S. Nielsen (2001). Randomised phase II trial of pegylated liposomal doxorubicin (DOXIL/CAELYX) versus doxorubicin in the treatment of advanced or metastatic soft tissue sarcoma: a study by the EORTC Soft Tissue and Bone


Gong, Y., T. Hong, M. Chen & Y. Huo (2011). A right heart angiosarcoma with rapidly progressing hemorrhagic pericardial effusion. *Intern Med* 50(5): 455-8. Gordon, I., M. Paoloni, C. Mazcko & C. Khanna (2009). The Comparative Oncology Trials

Graveel, C., Y. Su, J. Koeman, L. M. Wang, L. Tessarollo, M. Fiscella, C. Birchmeier, P.

Guo, T., L. Zhang, N. E. Chang, S. Singer, R. G. Maki & C. R. Antonescu (2011). Consistent

Hallel-Halevy, D., J. Yerushalmi, M. H. Grunwald, I. Avinoach & S. Halevy (1999). Stewart-

Hammond, T. N. & S. A. Pesillo-Crosby (2008). Prevalence of hemangiosarcoma in anemic

Hargis, A. M., P. J. Ihrke, W. L. Spangler & A. A. Stannard (1992). A retrospective

Harvey, J. J. (1964). An Unidentified Virus Which Causes the Rapid Production of Tumours

Haseman, J. K., J. R. Hailey & R. W. Morris (1998). Spontaneous neoplasm incidences in

Holden, C. A., M. F. Spittle & E. W. Jones (1987). Angiosarcoma of the face and scalp,

Hollstein, M., M. J. Marion, T. Lehman, J. Welsh, C. C. Harris, G. Martel-Planche, I. Kusters

Huang, J. & W. J. Mackillop (2001). Increased risk of soft tissue sarcoma after radiotherapy

Infante, P. F., S. E. Petty, D. H. Groth, G. Markowitz & D. Rosner (2009). Vinyl chloride

Jacks, T., L. Remington, B. O. Williams, E. M. Schmitt, S. Halachmi, R. T. Bronson & R. A. Weinberg (1994). Tumor spectrum analysis in p53-mutant mice. *Curr Biol* 4(1): 1-7. Jakab, C., J. Halasz, A. Kiss, Z. Schaff, M. Rusvai, P. Galfi, T. Z. Abonyi & J. Kulka (2009).

Johnson, M. S., Martin, M., Binns, S., Day, M.J. (2004). A Retrospective Study of Clinical

diagnosis of canine hemangiosarcoma. *Histol Histopathol* 24(7): 801-13. Ji, J. & K. Hemminki (2007). Familial blood vessel tumors and subsequent cancers. *Ann* 

drug development pathway. *PLoS Med* 6(10): e1000161.

*Natl Acad Sci U S A* 101(49): 17198-203.

(2003-2005). *J Am Vet Med Assoc* 232(4): 553-8.

hemangiosarcomas. *Vet Pathol* 29(4): 316-28.

prognosis and treatment. *Cancer* 59(5): 1046-57.

Toxicology Program update. *Toxicol Pathol* 26(3): 428-41.

chloride-exposed factory workers. *Carcinogenesis* 15(1): 1-3.

barbers: case reports. *Int J Occup Environ Health* 15(1): 36-42.

in women with breast carcinoma. *Cancer* 92(1): 172-80.

in Mice. *Nature* 204: 1104-5.

*Oncol* 18(7): 1260-7.

*Small Animal Practice* 45: 546-552.

50(1): 25-33.

50.

Consortium: using spontaneously occurring cancers in dogs to inform the cancer

Swiatek, R. Bronson & G. Vande Woude (2004). Activating Met mutations produce unique tumor profiles in mice with selective duplication of the mutant allele. *Proc* 

MYC and FLT4 gene amplification in radiation-induced angiosarcoma but not in other radiation-associated atypical vascular lesions. *Genes Chromosomes Cancer*

Treves syndrome in a patient with elephantiasis. *J Am Acad Dermatol* 41(2 Pt 2): 349-

dogs with a splenic mass and hemoperitoneum requiring a transfusion: 71 cases

clinicopathologic study of 212 dogs with cutaneous hemangiomas and

Fischer 344 rats and B6C3F1 mice in two-year carcinogenicity studies: a National

& R. Montesano (1994). p53 mutations at A:T base pairs in angiosarcomas of vinyl

propellant in hair spray and angiosarcoma of the liver among hairdressers and

Claudin-5 protein is a new differential marker for histopathological differential

Findings, Treatment and Outcome in 143 Dogs with Pericardial Effusion. *Journal of* 


Clinical and Molecular Biology of Angiosarcoma 171

Matzke, L. A., M. A. Knowling, D. Grant, J. B. Cupples, J. Leipsic, A. Ignaszewski & M. F.

McDonagh, D. P., J. Liu, M. J. Gaffey, L. J. Layfield, N. Azumi & S. T. Traweek (1996).

Mejia, A. F., L. Gierbolini, B. Jacob & S. V. Westmoreland (2009). Pediatric hepatic

Mendenhall, W. M., C. M. Mendenhall, J. W. Werning, J. D. Reith & N. P. Mendenhall (2006).

Mesri, E. A., E. Cesarman & C. Boshoff (2010). Kaposi's sarcoma and its associated

Miller, M. A., J. A. Ramos & J. M. Kreeger (1992). Cutaneous vascular neoplasia in 15 cats: clinical, morphologic, and immunohistochemical studies. *Vet Pathol* 29(4): 329-36. Miura, K., Y. Kum, G. Han & Y. Tsutsui (2003). Radiation-induced laryngeal angiosarcoma

Montaner, S., A. Sodhi, A. Molinolo, T. H. Bugge, E. T. Sawai, Y. He, Y. Li, P. E. Ray & J. S.

Morgan, M. B., M. Swann, S. Somach, W. Eng & B. Smoller (2004). Cutaneous angiosarcoma: a case series with prognostic correlation. *J Am Acad Dermatol* 50(6): 867-74. Muslin, A. J. (2008). MAPK signalling in cardiovascular health and disease: molecular

Myers, D. D., Jr., R. C. Dysko, C. E. Chrisp & J. L. Decoster (2001). Subcutaneous

Neuhauser, T. S., G. A. Derringer, L. D. Thompson, J. C. Fanburg-Smith, M. Miettinen, A.

Nickoloff, B. J. & C. E. Griffiths (1989). The spindle-shaped cells in cutaneous Kaposi's

Offori, T. W., C. C. Platt, M. Stephens & G. B. Hopkinson (1993). Angiosarcoma in congenital

Ohsawa, M., N. Naka, Y. Tomita, D. Kawamori, H. Kanno & K. Aozasa (1995). Use of

Paoloni, M. & C. Khanna (2008). Translation of new cancer treatments from pet dogs to

mechanisms and therapeutic targets. *Clin Sci (Lond)* 115(7): 203-18.

immunophenotypic study of 28 cases. *Mod Pathol* 13(9): 978-87.

*Pathol* 175(1): 51-7.

*Cardiovasc Pathol*.

4.

127-30.

135(5): 793-800.

angiosarcoma. *Am J Pathol* 149(4): 1363-8.

herpesvirus. *Nat Rev Cancer* 10(10): 707-19.

the literature. *Pathol Int* 53(10): 710-5.

viral latent genes. *Cancer Cell* 3(1): 23-36.

the literature. *Clin Exp Dermatol* 18(2): 174-7.

cases. *Cancer* 75(12): 2867-74.

humans. *Nat Rev Cancer* 8(2): 147-56.

Cutaneous angiosarcoma. *Am J Clin Oncol* 29(5): 524-8.

endothelial)-cadherin and other endothelial-specific markers in haemangiomas. *J* 

Allard A rare cardiac neoplasm: case report of cardiac epithelioid angiosarcoma.

Detection of Kaposi's sarcoma-associated herpesvirus-like DNA sequence in

hemangiosarcoma in a rhesus macaque (Macaca mulatta). *J Med Primatol* 38(2): 121-

after cervical tuberculosis and squamous cell carcinoma: case report and review of

Gutkind (2003). Endothelial infection with KSHV genes in vivo reveals that vGPCR initiates Kaposi's sarcomagenesis and can promote the tumorigenic potential of

hemangiosarcomas in a rhesus macaque (Macaca mulatta). *J Med Primatol* 30(2):

Saaristo & S. L. Abbondanzo (2000). Splenic angiosarcoma: a clinicopathologic and

sarcoma. Histologic simulators include factor XIIIa dermal dendrocytes. *Am J Pathol*

hereditary lymphoedema (Milroy's disease)--diagnostic beacons and a review of

immunohistochemical procedures in diagnosing angiosarcoma. Evaluation of 98


Lahat, G., A. R. Dhuka, S. Lahat, K. D. Smith, R. E. Pollock, K. K. Hunt, V. Ravi, A. J. Lazar &

Lamerato-Kozicki, A. R., K. M. Helm, C. M. Jubala, G. C. Cutter & J. F. Modiano (2006).

Lasota, J. & M. Miettinen (1999). Absence of Kaposi's sarcoma-associated virus (human herpesvirus-8) sequences in angiosarcoma. *Virchows Arch* 434(1): 51-6. Lawley, L. P., F. Cerimele, S. W. Weiss, P. North, C. Cohen, H. P. Kozakewich, J. B. Mulliken

Leaute-Labreze, C., E. Dumas de la Roque, T. Hubiche, F. Boralevi, J. B. Thambo & A. Taieb

Lebel, M., R. D. Cardiff & P. Leder (2001). Tumorigenic effect of nonfunctional p53 or p21 in mice mutant in the Werner syndrome helicase. *Cancer Res* 61(5): 1816-9. Lee, C.-S. & N. Duesbery (2010). Highly Selective MEK Inhibitors. *Current Enzyme Inhibition*

Lewis, T. S., P. S. Shapiro & N. G. Ahn (1998). Signal transduction through MAP kinase

Lin, C. F., D. DeFrias & X. Lin (2010). Epithelioid angiosarcoma: a neoplasm with potential

Looney, W. B. (1960). An investigation of the late clinical findings following thorotrast

Lorusso, D., A. Di Stefano, V. Carone, A. Fagotti, S. Pisconti & G. Scambia (2007). Pegylated

Ludolph-Hauser, D., E. Thoma-Greber, C. Sander, C. P. Sommerhoff & M. Rocken (2000).

Maeda, M., K. R. Johnson & M. J. Wheelock (2005). Cadherin switching: essential for

Mahony, B., R. B. Jeffrey & M. P. Federle (1982). Spontaneous rupture of hepatic and splenic angiosarcoma demonstrated by CT. *AJR Am J Roentgenol* 138(5): 965-6. Marcon, I., P. Collini, M. Casanova, C. Meazza & A. Ferrari (2004). Cutaneous angiosarcoma in a patient with xeroderma pigmentosum. *Pediatr Hematol Oncol* 21(1): 23-6. Marion, M. J., Froment, O., Trepo, C. (2005). Activation of Ki-ras gene by point mutation in

Mark, R. J., J. C. Poen, L. M. Tran, Y. S. Fu & G. F. Juillard (1996). Angiosarcoma. A report of

Martin-Padura, I., C. De Castellarnau, S. Uccini, E. Pilozzi, P. G. Natali, M. R. Nicotra, F.

67 patients and a review of the literature. *Cancer* 77(11): 2400-6.

diagnostic challenges. *Diagn Cytopathol* 38(2): 154-8.

for endothelial differentiation. *Exp Hematol* 34(7): 870-8.

Lane, O. G. (1952). Cutaneous angiosarcoma with metastases. *Br J Cancer* 6(3): 230-5.

*Oncol* 16(9): 2502-9.

300.

51.

6(3): 146-157.

Academic Press. 74: 49-139.

syndrome). *Ann Oncol* 18(7): 1159-64.

transition. *J Cell Sci* 118(Pt 5): 873-87.

*Carcinogenesis* 4(6): 450-454.

old girl. *J Am Acad Dermatol* 43(5 Pt 2): 900-2.

D. Lev (2009). Outcome of locally recurrent and metastatic angiosarcoma. *Ann Surg* 

Canine hemangiosarcoma originates from hematopoietic precursors with potential

& J. L. Arbiser (2005). Expression of Wilms tumor 1 gene distinguishes vascular malformations from proliferative endothelial lesions. *Arch Dermatol* 141(10): 1297-

(2008). Propranolol for severe hemangiomas of infancy. *N Engl J Med* 358(24): 2649-

cascades. In: *Advances in Cancer Research*. G. F. V. Woude&G. Klein. San Diego, CA,

(thorium dioxide) administration. *Am J Roentgenol Radium Ther Nucl Med* 83: 163-85.

liposomal doxorubicin-related palmar-plantar erythrodysesthesia ('hand-foot'

Mast cells in an angiosarcoma complicating xeroderma pigmentosum in a 13-year-

behavioral but not morphological changes during an epithelium-to-mesenchyme

human liver angiosarcoma associated with vinyl-chloride exposure. *Molecular* 

Ughi, C. Azzolini, E. Dejana & L. Ruco (1995). Expression of VE (vascular

endothelial)-cadherin and other endothelial-specific markers in haemangiomas. *J Pathol* 175(1): 51-7.


Clinical and Molecular Biology of Angiosarcoma 173

Soini, Y., J. A. Welsh, K. G. Ishak & W. P. Bennett (1995). p53 mutations in primary hepatic

Sorenmo, K. U., J. L. Baez, C. A. Clifford, E. Mauldin, B. Overley, K. Skorupski, R. Bachman,

Stoica, G., J. Hoffman & P. H. Yuen (1990). Moloney murine sarcoma virus 349 induces Kaposi's sarcomalike lesions in Balb/c mice. *Am J Pathol* 136(4): 933-47. Sutton, R. H. & M. W. McLennan (1982). Hemangiosarcoma in a cow. *Vet Pathol* 19(4): 456-8. Suzuki, H., A. Komatsu, Y. Fujioka, K. Yamashiro, H. Takeda & T. Hamada (2010). Angiosarcoma-like metastatic carcinoma of the liver. *Pathol Res Pract* 206(7): 484-8. Tamburini BA, P. T., Fosmire SP, Scott MC, Trapp SC, Duckett MM, Robinson SR, Slansky

Tanioka, M., A. Ikoma, K. Morita, H. Fujii, K. I. Toda, K. Takahashi, T. Tanaka, C. Nishigori,

Tate, G., T. Suzuki & T. Mitsuya (2007). Mutation of the PTEN gene in a human hepatic

Thamm, D. H., E. B. Dickerson, N. Akhtar, R. Lewis, R. Auerbach, S. C. Helfand & E. G.

Thompson, W. M., A. D. Levy, N. S. Aguilera, L. Gorospe & R. M. Abbott (2005).

Tuttle, A. D., Harms, C.A., Van Wettere, A., Grafinger, M.S., Lewbart, G.A. (2006). Splenic

Vakkalanka, B. & M. Milhem (2010). Paclitaxel as neoadjuvant therapy for high grade

Valenzuela, E. J., L. M. J. Lopez Poveda, P. F. J. Fuenzalida, G. C. Garre Sanchez, M. E. Barba

van Geel, A. N. & M. A. den Bakker (2009). Bilateral angiosarcoma of the breast in a

angiosarcoma. *Cancer Genet Cytogenet* 178(2): 160-2.

hemangiosarcoma-derived cell line. *Res Vet Sci* 81(1): 76-86.

report of six cases in elephantiasis chirurgica. *Cancer* 1(1): 64-81.

1alpha. *Cancer Res* 60(17): 4873-80.

2879-81.

10: 619.

106-15.

*Dermatol* 144(2): 380-3.

*and Surgery* 16(4): 140-143.

*Enferm Dig* 101(6): 430-4, 434-7.

fourteen-year-old child. *Rare Tumors* 1(2): e38.

*Oncol* 4: 107-110.

vascular endothelial growth factor expression and secretion through mitogenactivated protein kinase and p38 pathways acting on hypoxia-inducible factor

angiosarcomas not associated with vinyl chloride exposure. *Carcinogenesis* 16(11):

M. Samluk & F. Shofer (2004). Efficacy and toxicity of a dose-intensified doxorubicin protocol in canine hemangiosarcoma. *J Vet Intern Med* 18(2): 209-13. Spangler, W. L. & M. R. Culbertson (1992). Prevalence, type, and importance of splenic diseases in dogs: 1,480 cases (1985-1989). *J Am Vet Med Assoc* 200(6): 829-34. Stewart, F. W. & N. Treves (1948). Lymphangiosarcoma in postmastectomy lymphedema; a

JE, Sharkey LC, Cutter GR, Wojcieszyn JW, Bellgrau D, Gemmill RM, Hunter LE, Modiano JF. (2010). Gene expression profiling identifies inflammation and angiogenesis as distinguishing features of canine hemangiosarcoma. *BMC Cancer*

G. Jin, S. Higashi, S. Toyokuni & Y. Miyachi (2001). Angiosarcoma of the scalp: absence of vascular endothelial cadherin in primary and metastatic lesions. *Br J* 

MacEwen (2006). Biological and molecular characterization of a canine

Angiosarcoma of the spleen: imaging characteristics in 12 patients. *Radiology* 235(1):

Hemangiosarcoma in a Corn Snake, *Elaphe guttata*. *Journal of Herpetological Medicine* 

angiosarcoma of the spleen: a brief report and literature review. *Clin Med Insights* 

& C. F. Alvarez (2009). Hepatic angiosarcoma. Presentation of two cases. *Rev Esp* 


Pawlik, T. M., A. F. Paulino, C. J. McGinn, L. H. Baker, D. S. Cohen, J. S. Morris, R. Rees & V.

Penel, N., B. N. Bui, J. O. Bay, D. Cupissol, I. Ray-Coquard, S. Piperno-Neumann, P. Kerbrat,

Pitts, O. M., J. M. Powers & P. M. Hoffman (1983). Vascular neoplasms induced in rodent central nervous system by murine sarcoma viruses. *Lab Invest* 49(2): 171-82. Pozo, A. L., E. M. Godfrey & K. M. Bowles (2009). Splenomegaly: investigation, diagnosis

Prymak, C., L. J. McKee, M. H. Goldschmidt & L. T. Glickman (1988). Epidemiologic,

Quezado, M. M., L. P. Middleton, B. Bryant, K. Lane, S. W. Weiss & M. J. Merino (1998). Allelic loss on chromosome 22q in epithelioid sarcomas. *Hum Pathol* 29(6): 604-8. Requena, L. & O. P. Sangueza (1998). Cutaneous vascular proliferations. Part III. Malignant

Sahmel, J., K. Unice, P. Scott, D. Cowan & D. Paustenbach (2009). The use of multizone

Sans, V., E. D. de la Roque, J. Berge, N. Grenier, F. Boralevi, J. Mazereeuw-Hautier, D.

Schlemmer, M., P. Reichardt, J. Verweij, J. T. Hartmann, I. Judson, A. Thyss, P. C.

Schuborg, C., F. Mertens, A. Rydholm, O. Brosjo, M. Dictor, F. Mitelman & N. Mandahl

Schultheiss, P. C. (2004). A retrospective study of visceral and nonvisceral hemangiosarcoma and hemangiomas in domestic animals. *J Vet Diagn Invest* 16(6): 522-6. Shaw, S. P., E. A. Rozanski & J. E. Rush (2004). Cardiac troponins I and T in dogs with

Sodhi, A., S. Montaner, V. Patel, M. Zohar, C. Bais, E. A. Mesri & J. S. Gutkind (2000). The

EORTC soft tissue and bone sarcoma group. *Eur J Cancer* 44(16): 2433-6. Schmid, H. & C. Zietz (2005). Human herpesvirus 8 and angiosarcoma: analysis of 40 cases

angiosarcoma: the ANGIOTAX Study. *J Clin Oncol* 26(32): 5269-74.

approach. *Cancer* 98(8): 1716-26.

ras-2. *Lab Invest* 76(1): 153-9.

Pt 1): 143-75; quiz 176-8.

e423-31.

and management. *Blood Rev* 23(3): 105-11.

1960s and 1970s. *Risk Anal* 29(12): 1699-725.

and review of the literature. *Pathology* 37(4): 284-7.

pericardial effusion. *J Vet Intern Med* 18(3): 322-4.

tissue. *Cancer Genet Cytogenet* 100(1): 52-6.

K. Sondak (2003). Cutaneous angiosarcoma of the scalp: a multidisciplinary

C. Fournier, S. Taieb, M. Jimenez, N. Isambert, F. Peyrade, C. Chevreau, E. Bompas, E. G. Brain & J. Y. Blay (2008). Phase II trial of weekly paclitaxel for unresectable

clinical, pathologic, and prognostic characteristics of splenic hemangiosarcoma and splenic hematoma in dogs: 217 cases (1985). *J Am Vet Med Assoc* 193(6): 706-12. Przygodzki, R. M., S. D. Finkelstein, P. Keohavong, D. Zhu, A. Bakker, P. A. Swalsky, Y.

Soini, K. G. Ishak & W. P. Bennett (1997). Sporadic and Thorotrast-induced angiosarcomas of the liver manifest frequent and multiple point mutations in K-

neoplasms, other cutaneous neoplasms with significant vascular component, and disorders erroneously considered as vascular neoplasms. *J Am Acad Dermatol* 38(2

models to estimate an airborne chemical contaminant generation and decay profile: occupational exposures of hairdressers to vinyl chloride in hairspray during the

Lipsker, E. Dupuis, K. Ezzedine, P. Vergnes, A. Taieb & C. Leaute-Labreze (2009). Propranolol for severe infantile hemangiomas: follow-up report. *Pediatrics* 124(3):

Hogendoorn, S. Marreaud, M. Van Glabbeke & J. Y. Blay (2008). Paclitaxel in patients with advanced angiosarcomas of soft tissue: a retrospective study of the

(1998). Cytogenetic analysis of four angiosarcomas from deep and superficial soft

Kaposi's sarcoma-associated herpes virus G protein-coupled receptor up-regulates

vascular endothelial growth factor expression and secretion through mitogenactivated protein kinase and p38 pathways acting on hypoxia-inducible factor 1alpha. *Cancer Res* 60(17): 4873-80.


**10** 

Muna Sabah

*Ireland* 

*Connolly Hospital, Dublin,* 

**Gastrointestinal Stromal Tumours:** 

**Morphology and Prognosis** 

**A Contemporary Review on Pathogenesis,** 

Gastrointestinal stromal tumours (GISTs) comprise the largest subset of mesenchymal tumours of the digestive tract. Over the past 10 years, this group of tumours has emerged

The first accurate description of mesenchymal neoplasms of the gastrointestinal tract was in 1941 (Golden T & Stout AP, 1941). Traditionally, these tumours were thought to be derived from smooth muscle cells, based on their resemblance to smooth muscle tumours. They were referred to as leiomyomas, bizarre leiomyomas (Stout AP, 1976), cellular leiomyomas (Appelman, 1977) and leiomyosarcomas. However, the advent of electron microscopy revealed that only a few of them have convincing ultrastructural evidence of smooth muscle differentiation. In addition, the application of immunohistochemistry clearly demonstrated that many of these tumours lack the features of smooth muscle differentiation. This led Mazur and Clark in 1983 to introduce the generic designation "stromal tumour" (Mazur &

A little later, in 1984, Herrera et al (Herrera et al., 1984) introduced the concept of "plexosarcoma" to acknowledge the existence of a small subset of stromal tumours with autonomic neuronal differentiation which became better known as gastrointestinal

A considerable controversy arose as to the line of differentiation these tumours take. Some tumours exhibit a myogenic phenotype, while others may show a neural differentiation, mixed or no differentiation at all, the so called "null phenotype". Some enthusiasm was generated by the recognition that a significant proportion of stromal tumours express CD34 (Miettinen et al., 1995). However the diagnostic utility of this marker was hampered by its poor specificity. As a consequence of the lack of reproducible diagnostic criteria, GIST represented a generic term that indicates any mesenchymal tumour arising in the

The recognition of the central role of *c-kit* mutations in the pathogenesis of GISTs (Hirota et al., 1998; Rubin et al., 2000) and in most cases the associated expression of KIT protein in these tumours has provided a reproducible genotypic and phenotypic marker (Kindblom et al., 1998). Therefore KIT (CD117 in the standardised terminology of leucocyte antigens) expression has emerged as a marker for discriminating GISTs from other mesenchymal

from a poorly understood neoplasm to a well defined tumour entity.

autonomic nerve tumours (GANTs) (Lauwers et al., 1993).

**1. Introduction** 

Clark, 1983).

gastrointestinal tract.

