Diagnosis is performed clinically, but it is recommended to obtain confirmation by biopsy. Histological study reveals spindle cells, proliferation of irregular vessels with slit-like forms, red blood cell extravasation, and leukocytic infiltrate with plasma cells and intra- and extracellular hyalin globules throughout the thickness of the dermis, as well as the so-called promontory sign (Fig. 3). Polymerase chain reaction and immunohistochemical staining for latency-associated nuclear antigen (LANA-1) of the HHV-8 virus are positive.1

Figure 3.

Histology of Kaposi sarcoma.

A, Low-magnification view of Kaposi sarcoma in nodular phase. Well-defined dermal nodule.

B, Highly cellular lesion with some gaps in the form of cracks.

C, Detail of spindle cells and red blood cells within the small vessels.

D, Immunohistochemical nuclear positivity for HHV-8 specific to Kaposi sarcoma.

(0.56MB).

Classic Kaposi sarcoma. In view of the clinical presentation (age, local involvement, infrequent involvement of internal organs, and indolent course), skin and lymph node examination is sufficient. Complementary tests are performed if the patient presents symptoms of visceral compromise.5

Immunosuppression-associated Kaposi sarcoma. As for HIV-associated KS discussed below, there is no consensus staging for immunosuppression-associated KS. Given the lack of consensus, the tests and recommended applicable criteria are usually the same as for HIV-associated KS.

HIV-Associated Kaposi Sarcoma There is no accepted staging system for this variant. Chest X-ray is recommended and, in the event of abnormalities suggestive of respiratory compromise, bronchoscopy or computed chest tomography should be performed. It is also recommended to rule out fecal occult blood; if this test is positive, digestive endoscopy should be performed.

In 1989, the AIDS Clinical Trials Group Oncology Committee proposed staging (Table 2) based on disease extension, CD4T-cell count, and systemic compromise (opportunistic infections, B symptoms such as fever, weight loss, or persistent diarrhea, and Karnofsky performance status below 70 points). A prospective analysis showed that these variables were related to patient survival, with factors associated with good prognosis being limited disease, CD4 lymphocyte count greater than 150cells/mm3, and absence of systemic compromise.9

These proposals for staging were put forward before HAART was available and do not include viral load, and so application of this staging system has been limited to clinical trials.

KS is an angioproliferative tumor with different subtypes associated with advanced age, certain African populations, iatrogenic immunosuppression, or HIV infection. Although HAART has led to a dramatic decrease in the incidence and severity of KS in individuals with HIV infection, it is important to be aware of the different therapeutic options according to the KS variant and its clinical presentation.

The incidence of cutaneous angiosarcoma is difficult to calculate, but angiosarcomas together are reported at a rate of 0.4 cases per million inhabitants in the United States.31 Cutaneous angiosarcomas account for between 35% and 60% of all angiosarcomas, with an approximate incidence of 0.2 cases per million inhabitants. They predominate in elderly patients with a mean age of 73 years. These lesions are very rare in children or young patients. The classic Wilson-Jones angiosarcoma predominates in men and after radiotherapy in women.29 It also predominantly affects whites.25,29 The majority of idiopathic angiosarcomas are located on the head and neck (62%), postradiation lesions predominate on the trunk (24%), particularly in the pectoral area (postradiation of the breast) and after lymphedema on the limbs (11%). Most postradiation angiosarcomas appear after radiotherapy due to breast cancer,32 but there are cases in other irradiated areas and not just because of malignant processes. The latency time between radiotherapy and the development of angiosarcoma varies, with a mean of 5 years for breast sites and 10 years for other sites. Angiosarcomas due to lymphedema predominately occur in areas of chronic lymphedema on the arms of women who undergo radical mastectomy (Stewart-Treves syndrome),33 but cases have been reported of lesions over lymphedema of any etiology. The time between presentation of lymphedema and development of angiosarcomas varies greatly between 1 year and 30 years.

The initial most characteristic form is a bruise-like lesion, at times edematous, with a poor defined frequency, that tends to go unnoticed initially, particularly on the scalp of patients with hair.34 In the case of angiosarcoma of the head and neck, the true extent of the lesion can be better appreciated if the patient holds his or her head below the level of the heart for a few seconds. This maneuver will make the subclinical part become more visible as it takes on a violaceous tone and edematous appearance.35 Later in the clinical course, papules and nodules appear, occasionally with ulceration and bleeding in advanced phases. Some cases present directly with papules and nodules with hardly any prior bruise-like phase. The mean size on diagnosis is 3-5 cm.28,29 Angiosarcomas can appear as a solitary or multifocal tumor and the lesion will often extend beyond the clinically appreciable limits. Clinical suspicion of cutaneous angiosarcoma should be confirmed by biopsy.

The 3 types of cutaneous angiosarcoma have overlapping histopathologic characteristics. Well-differentiated angiosarcomas show vascular lumina lined by flattened endothelial cells, dissecting collagen bundles, with limited cell atypia. Histologic diagnosis is complex in these phases and it is useful to recognize some endothelia with prominent, pleomorphic cells with hyperchromatic nuclei that tend to protrude and create some papillae, with several endothelial cell layers in the vascular lumina.36,37 The vascular conduits are irregular and tend to form anastomosing channels (Fig. 6). In the worst differentiated cases, tumor cells are epithelioid or fusiform, with marked atypia and abundant mitosis and a more solid growth pattern with few vascular lumina, such that they can be confused with carcinoma or even melanomas or fibrosarcomas. The presence of intracytoplasmic vacuoles as expression of primitive vascular differentiation can be very useful for suspicion of correct diagnosis in these cases. Infiltration in the case of undifferentiated angiosarcomas is very destructive and normal components of the dermis and skin appendages disappear. A patchy inflammatory infiltrate often accompanies the process and this infiltrate may at times be so dense that it resembles lymphoma.38 The extent of accompanying red blood cells and hemosiderin is very variable. Different degrees of differentiation may often be found in the same angiosarcoma. The epidermis may be normal, atrophic, or ulcerated. The extent of differentiation of cutaneous angiosarcomas is not traditionally thought to have an impact on prognosis, and so, unlike with other sarcomas, the histologic grade is not taken into account for staging.39 This approach may change in the future, as a recent study with the largest series of cutaneous angiosarcomas and soft tissue angiosarcomas published to date, with 821 patients, developed a prognostic model for survial which included histopathologic grade.25

Figure 6.

A, Immunohistochemistry of cutaneous angiosarcoma that is positive for ERG (typically with a nuclear pattern). B, Angiosarcoma with predominance of areas with vasoformative pattern. C, D, Detailed images of neoplastic endothelial cells, which in this case are prominent but without noteworthy atypia.

(0.71MB).

Study of angiosarcoma should be completed with an immunohistochemical panel that includes a basic panel for spindle-cell tumors (CD31, pancytokeratins, S110, and actin) and additional vascular markers (CD34, erythroblast transformation-specific–related gene [ERG], podoplanin). Some cases of angiosarcoma with predominance of epithelioid cells may be positive for cytokeratins; however, positivity for vascular markers such as CD31, ERG, and/or podoplanin can rule out undifferentiated carcinoma. In recent years, it has been shown that many angiosarcomas have MYC amplification/overexpression. In most studies, MYC amplification is found in between 50% and 100% of secondary angiosarcomas but not generally in idiopathic ones,40–42 but MYC amplification or overexpression has also been detected in some idiopathic angiosarcomas.43,44 However, almost all studies have shown the absence of MYC amplification or overexpression in atypical postradiation vascular proliferations, and so positivity in a doubtful case of vascular proliferation in the irradiated area almost rules out angiosarcoma.

The origin of angiosarcoma in blood vessels or lymphatic vessels is subject to debate. Expression of CD31 or CD34 is greater in the most differentiated areas, but in neither case is it constant.45 Immunohistochemical markers relatively specific for lymphatic vessels, such as podoplanin, D2-40, LYVE-1, and PROX-1, are usually positive in cutaneous angiosarcomas, often expressed with a mixed immunohistochemical pattern of endothelial blood vessels and endothelial lymphatic vessels.38,46,47 Although exceptional, cases have been reported of cutaneous angiosarcomas that express S-100 protein,48 or neuroendocrine markers.49

Although there are no guidelines for the management of cutaneous angiosarcoma, given that the most frequent site of metastasis is the lung50 followed by the lymph nodes,29,51 thoracoabdominal computed tomography is usually recommended after pathological diagnosis. This imaging study should include the cervical region in the case of an angiosarcoma of the head and neck and the pelvis in in the case of abdominopelvic postradiation angiosarcomas. The presence of regional or distant dissemination is not uncommon with angiosarcomas, and is reported in 30% and 10% of cases, respectively.29

There is no TNM staging specific to angiosarcomas, and so the TNM staging for soft tissue sarcomas of the American Joint Committee on Cancer classification, adapted to angiosarcomas, is used. In the case of angiosarcomas, as histologic grade is not currently considered to impact prognosis, stage IA andIIA is grouped in some studies with stages IB andIIB, which in other sarcomas are differentiated only by histologic grade.

The only treatment shown to be potentially curative in cutaneous angiosarcoma is surgery, although cure only occurs sporadically. However, in inoperable or metastatic cases, radiotherapy and/or chemotherapy have a recognized palliative role. Moreover, some recent studies have also included radiotherapy as an adjuvant to surgery in localized cases of angiosarcoma,50,51 and some authors have even recommended to irradiate regional lymph nodes,52 but this is not common practice. The problem with surgery in cutaneous angiosarcoma is at times the multicenter character and the poor clinical delimitation in others, as well as the fact that they are often diagnosed when lesions exceed 5 cm in size. In addition to these factors, patients are often elderly which makes it more difficult to obtain suitable surgical margins. In general, if the tumor characteristics and general state of the patient permit, treatment of cutaneous angiosarcoma is surgical excision with sufficient margins. The most accepted is surgery with 3 cm margin with respect to the clinically appreciable limits.53 The depth of the margin is not well established, given that this is a dermal sarcoma, though it would appear reasonable to reach the fascia without excising them. In the most infiltrative cases, however, the muscle should be included to achieve clear margins. Margins with angiosarcoma involvement have been shown to be a factor for poor prognosis in several studies.28,30,51 In the case of breast involvement, most studies suggest total mastectomy or more or less extensive excisions of irradiated skin. In complex cases, prior biopsy mapping can help assist with presurgical planning. Whenever possible, in cutaneous oncology, direct closure, grafts, or second intent closures are preferred to facilitate follow-up and not mask possible local recurrence with surgical reconstruction, but this may be difficult or impossible after more radical excisions of mammary angiosarcomas that require total mastectomy with irradiation of the entire skin. In the case of angiosarcomas caused by lymphedema, a study reviewed reports of 160 patients with Stewart-Treves syndrome and found absence of benefit of amputation compared with radical surgery (with 2 or 3 cm margin) in these cases, and so limb amputation does not appear to be justified in such angiosarcomas.54

In cases in which surgery is impossible, that is multicentric or extensive lesions, or those that affect areas that complicate surgery, radiotherapy is the treatment of choice.55 The dose of radiotherapy for cutaneous angiosarcoma is usually 60 Gy distributed in 20 sessions of 3 Gy each. When used as an adjuvant to surgery, the doses are similar, except when indicated for postradiation angiosarcomas, in which the dose will usually be lower (45-50 Gy).

The only role of chemotherapy in angiosarcoma is as a palliative treatment, reserved for relapsed or metastatic lesions not amenable to surgery. A neoadjuvant role has also recently been proposed for chemotherapy prior to surgery in periorbital sites.56 The most widely used chemotherapies in angiosarcoma are docetaxel,57,58 paclitaxel,59 and liposomal doxorrubicin,60 but the current NCCN guidelines also include vinorelbine, sorafenib, sunitinib, and bevacizumab, although the results with these latter 3 antiangiogenic agents have been disappointing. Combination with beta blockers in this phase of palliative treatment may be of some benefit for the patient.61,62

There are no standard guidelines for follow-up of cutaneous angiosarcomas. In our group, we have close clinical follow-up, with check-ups every 3-6 months for the first few years, and then yearly check-ups for 10 years. In these visits, we examine the entire skin and palpate the corresponding territorial lymph nodes. At least once a year, we perform a laboratory analysis and thoracoabdominopelvic computed tomography study.