Angiosarcoma

AJCC 8th Edition (Soft Tissue Sarcoma)

Angiosarcoma is staged using the AJCC soft tissue sarcoma staging system. However, most angiosarcomas are high-grade and present at advanced stage due to multifocal disease and early metastases.

Most angiosarcomas present as Stage III or IV due to high grade and extent of disease.

Clinical Subtypes

Cutaneous (Head/Neck): Elderly patients, scalp most common, multifocal, worst prognosis of the subtypes — 10-20% 5-year survival in unselected series, 43-56% in non-metastatic disease treated with surgery plus radiotherapy

Radiation-Induced: Post-breast cancer radiation, 5-10 year latency, requires mastectomy and chest wall resection

Lymphedema-Associated (Stewart-Treves): Post-mastectomy arm lymphedema, amputation often required, 15-25% 5-year survival

Deep Soft Tissue: Large extremity or trunk mass, best chance of negative margins, 30-40% 5-year survival

Clinical subtype is the most important prognostic factor.

Planning the Biopsy

Image before you biopsy. Biopsy through an unimaged mass loses the information the radiologist needs — the pre-biopsy signal characteristics, the true extent, and the sites of viable rather than necrotic tumour — and the haematoma from a vascular tumour will distort every subsequent scan.

For a deep soft-tissue or bone lesion the biopsy is planned by, and ideally performed at, the unit that will carry out the definitive resection. This is not a courtesy referral. In the Musculoskeletal Tumor Society's multicentre review of 597 biopsies of primary musculoskeletal sarcomas, the diagnosis was wrong in 17.8%; a problem with the biopsy forced a different and often more complex operation, or the addition of radiotherapy or chemotherapy, in 19.3%; the outcome itself was changed — a more disabling resection, loss of function, local recurrence or death — in 10.1%; and 18 patients underwent an unnecessary amputation because of the biopsy. Errors, complications and changes in the course of treatment were two to twelve times more frequent when the biopsy was performed at the referring institution rather than at the treating centre (p less than 0.001).

The mechanics that follow from this are the examinable part:

• Place the tract where the resection will take it out. Longitudinal on a limb, in line with the planned incision, through one compartment only. A transverse or badly sited tract contaminates tissue that would otherwise have been spared and can convert a limb-sparing resection into an amputation.
• Mark and record the tract so the resecting surgeon can excise it en bloc, and use the shortest route to the lesion that does not cross an uninvolved compartment or a neurovascular bundle.
• Angiosarcoma bleeds. It is a vasoformative tumour, so achieve meticulous haemostasis and avoid a drain; where a drain is unavoidable it exits in line with the incision so that it too can be excised.
• Send fresh tissue as well as formalin-fixed, and request endothelial markers up front (CD31, ERG) with MYC FISH whenever the lesion sits in a previously irradiated field, because that single test is what separates angiosarcoma from a benign atypical vascular lesion.

The general principles are set out in biopsy principles and techniques.

View larger

Histopathology

View larger

Morphology:

• Vasoformative: Blood-filled vascular channels
• Solid areas: Sheets of atypical endothelial cells
• High mitotic rate, nuclear atypia
• Variable differentiation (well-differentiated can appear bland)

Immunohistochemistry is essential:

• ERG: nuclear; the most sensitive endothelial marker and retained in poorly differentiated and epithelioid tumours where CD31 becomes patchy. Not specific - also positive in prostate adenocarcinoma and Ewing sarcoma
• CD31: membranous; the workhorse marker, roughly 90% positive, but also expressed by macrophages, leukocytes and non-angiosarcoma vascular tumours
• CD34: roughly 80% positive; widely expressed in fibroblastic tumours and GIST, so of limited discriminant value
• FLI-1: nuclear, sensitive, shares ERG's specificity problem
• Factor VIII (vWF): roughly 60% positive - low sensitivity and prone to serum contamination artefact; largely superseded

No single marker is both sensitive and specific for angiosarcoma, which is why lineage is established on a panel rather than one stain, and why MYC testing is added whenever the lesion arises in an irradiated field (PMID 40666093).

Differential Diagnosis

The most clinically dangerous error is mistaking a benign atypical vascular lesion (AVL) in an irradiated field for angiosarcoma, or vice versa. MYC amplification is the decisive discriminator.

View larger

View larger

Surgical Resection Principles

Goals:

• Wide excision with 2-3cm margins (if feasible)
• En bloc resection of all involved areas
• Reconstruction as needed

Site-Specific Surgery:

Amputation indications:

• Stewart-Treves syndrome with extensive involvement
• Neurovascular encasement precluding limb salvage
• Recurrent disease after prior limb-sparing surgery
• Patient choice for local control

Amputation is often necessary for Stewart-Treves syndrome to achieve local control.

Radiation Therapy

Indications:

• Adjuvant after surgery (all cases with high-grade disease)
• Positive or close margins (under 1cm)
• Unresectable disease (palliative)
• Head/neck angiosarcoma (often unresectable)

Dosing:

• Adjuvant: 60-66 Gy in 1.8-2 Gy fractions
• Preoperative: 50 Gy (may downsize the tumour)
• Palliative: 30-40 Gy

Efficacy: Angiosarcoma has variable radiosensitivity. Radiation improves local control but does not eliminate risk of recurrence. Essential component of multimodal therapy.

Radiation therapy is standard adjuvant treatment for angiosarcoma despite variable response.

Systemic Chemotherapy

Indications:

• Metastatic or unresectable disease — this is where the evidence sits, and both trial and cohort data come from this population
• Neoadjuvant for large or locally advanced tumours, to make an unresectable lesion resectable
• Adjuvant after complete resection of localised disease — widely given for high-grade angiosarcoma, but be clear in a viva that this is extrapolated from high-grade soft tissue sarcoma generally. No trial has shown that adjuvant chemotherapy improves survival in resected localised angiosarcoma; the survival gains demonstrated in the face and scalp cohorts came from adding radiotherapy to surgery, not chemotherapy. Decide it at the multidisciplinary meeting rather than reflexively

First-line regimens:

• Weekly paclitaxel (80 mg/m² on days 1, 8, 15 of a 4-week cycle) - the ANGIOTAX phase II non-progression rate was 74% at 2 months and 45% at 4 months
• Doxorubicin-based (doxorubicin +/- ifosfamide) - traditional high-grade sarcoma regimen; French Sarcoma Group data show doxorubicin and weekly paclitaxel give comparable first-line survival
• Paclitaxel + bevacizumab (anti-VEGF) - studied but adds toxicity without a clear survival gain

Second-line and emerging options:

• Gemcitabine +/- docetaxel
• Pazopanib (tyrosine kinase inhibitor, licensed for non-adipocytic soft tissue sarcoma)
• Propranolol - a non-selective beta-blocker repurposed as an adjunct; beta-adrenergic signalling drives angiosarcoma proliferation/angiogenesis, and propranolol combined with chemotherapy (e.g. metronomic vinblastine/methotrexate or paclitaxel) has produced responses in case series, particularly cutaneous angiosarcoma
• Clinical trials (immune checkpoint inhibitors under investigation, particularly in cutaneous/UV-related high-tumour-mutational-burden disease)

Efficacy: Angiosarcoma has moderate chemosensitivity relative to other sarcomas. Taxanes are particularly favoured for cutaneous and scalp disease; doxorubicin remains a reasonable first-line alternative. Chemotherapy delays progression and can downstage locally advanced disease, but cure is rare once metastatic.

Weekly paclitaxel is a preferred first-line systemic option, with doxorubicin-based therapy an equivalent alternative.