experience of angiosarcoma in the north of england bone and soft tissue tumour service

Journal of Plastic, Reconstructive & Aesthetic Surgery (2011) 64, 884e891

Experience of angiosarcoma in the North of EnglandBone and Soft Tissue Tumour Service*

C.J. Lewis a,*, C. Gerrand b, D.E. Barnes a, S. Murray b, R.H. Milner a,M. Ragbir a

aDepartment of Plastic Surgery, Royal Victoria Infirmary, Queen Victoria Road, Newcastle-Upon-Tyne, NE1 4LP, UKbDepartment of Orthopaedic Surgery, Freeman Hospital, High Heaton, Newcastle-Upon-Tyne, NE7 7DN, UK

Received 30 September 2010; accepted 16 November 2010

KEYWORDSSarcoma;Angiosarcoma;Radiotherapy;Management;Review

* This work was presented as a post* Corresponding author.E-mail address: drchristopherlewis

1748-6815/$-seefrontmatterª2010Bridoi:10.1016/j.bjps.2010.11.016

Summary Introduction: Angiosarcomas are rare aggressive sarcomas of vascular endothelialorigin. These tumours have the potential to be multicentric and are associated with high ratesof local recurrence, which makes treatment challenging. The gold-standard is that thesepatients are managed in specialist centres by a multidisciplinary team. We present our expe-rience of managing patients with angiosarcoma in the North of England Bone and Soft TissueTumour Service and a review of the literature.Methods: A prospectively collated electronic database was used to identify patients withangiosarcoma treated between 2000 and 2008, and an analysis performed of demographics,anatomical site, surgical excision and reconstruction, local disease recurrence and metastaticdisease.Results: Fifteen patients (ten female, five male, mean age 71 years) were identified. Eightpatients developed tumours in a previously irradiated area, after a mean of 11 years. Sixpatients had metastatic disease at presentation. Fourteen patients underwent wide surgicalexcision of the tumour, of which nine required defect reconstruction (five free latissimus dorsiflaps, two free anterolateral thigh flaps, two pedicled latissimus dorsi flaps). One patient wastreated with chemotherapy only. Five of 14 patients received adjuvant radiotherapy, and onereceived adjuvant chemotherapy. Two out of 14 patients developed local recurrence. Eightpatients developed metastases, the majority of which were pulmonary. Estimated five-yearsurvival was calculated as 33% in our patient cohort.

er at the BAPRAS Summer Meeting, 2009.

@googlemail.com (C.J. Lewis).

tishAssociationofPlastic,ReconstructiveandAestheticSurgeons.PublishedbyElsevierLtd.All rightsreserved.

http://dx.doi.org/10.1016/j.bjps.2010.11.016



Experience of angiosarcoma in the North of England Bone and Soft Tissue Tumour Service 885

Conclusions: Angiosarcomas are aggressive, difficult to treat tumours, which can occursecondary to a multitude of causes. Clinical suspicion, biopsy and early diagnosis are essentialto allow optimum treatment, which currently consists of radical surgery, together with adju-vant radiotherapy and chemotherapy.ª 2010 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published byElsevier Ltd. All rights reserved.

Figure 1 A patient treated in our unit with a radiation-induced angiosarcoma of the left breast following previous widelocal excision and adjuvant radiotherapy for breast carcinoma.

Introduction

The word sarkoma is derived from the Greek for ‘fleshygrowth’. Soft tissue sarcomas are a group of rare soft tissueneoplasms, with an incidence equivalent to less than onepercent of all cancers. Angiosarcomas are a rare aggressivetype of sarcoma of vascular endothelial origin, accountingfor 4.1% of all soft tissue sarcomas.1

Angiosarcoma has a peak incidence in the seventh decadeof life, though it can affect a wide age range of patients2 andboth sexes equally. Risk factors associated with the devel-opment of angiosarcoma include radiotherapy,3 polyvinylchloride,4 arsenic exposure5 and acquired lymphoedema.6

Angiosarcomas typically arise in the breast, skin, liver,skeleton, heart and extremities. Approximately 25% ofangiosarcomas occur in the head and neck, but they accountfor less than 0.1% of all head and neck malignancies.7 60% ofthese tumours are cutaneous, whilst 25% occur withinorgans, including the heart, liver and spleen.

Although the exact pathogenesis of angiosarcoma remainsunclear, it is hypothesised that local cellular immunode-pression secondary to destruction of regional lymph nodesand their respective networks results in a complex disruptionin local immunity.8 Lymphoedema has also been implicatedto be ‘tumour-inducing’ because oedematous regions areimmunodeficient, being ‘immune privileged sites’ in whichtumours can develop.9 This may also include delayed recog-nition of abnormal cells or antigens presented to the immunesystem.The fact that there havebeennopublished reports ofangiosarcoma related to oedema from cardiovascular, renalor hepatic disease suggests a possible initial inflammatorycause in the pathogenesis of this soft tissue tumour.

Cutaneous angiosarcomas usually present as a slow-growing subcutaneous mass or nodule, with discolourationand/or bruising of the skin1,10 (Figure 1). Associated purplishlesions are often multiple and of variable size and shape,1,10

and can be confused with benign haemangiomas. Occasion-ally, vesicles or pustulesmay appear on the lesion, which canulcerate (Figure 1). Ulcerated lesions are usually associatedwith local and distant invasion,which can occur over a periodof months rather than years. Angiosarcomas may present asmulticentric disease, presenting as multiple primarytumours; this phenomenon has been described in bone11 andthe gastrointestinal tract.12 Deep tumours present witha myriad of symptoms dependent on their location. Thesemay include hepatic dysfunction, anaemia, pathologicalfracture13 or symptoms of malignant vascular proliferationsuch as thrombocytopenia or bleeding. These complex andvaried presentations make diagnosis difficult.

Histologically, angiosarcomas have a typical papillaryarchitecture (Figure 5), with or without associated vascularand lymphatic invasion. Sarcomas are graded with theTrojani score,14 which involves the staining of reticulin and

immunohistochemistry. First described in 1984, it is basedon tumour necrosis, differentiation and mitotic count, andcorrelates well with tumour prognosis. The American JointCommittee on Cancer (AJCC) is used for sarcoma staging,and takes into consideration tumour size, regional lymphnode status, metastases and histological grade (Table 1).

The North of England Bone and Soft Tissue Tumour Service(NEBSTS) is a multidisciplinary team of plastic, orthopaedic,thoracic, head and neck and general surgeons, medical andclinical oncologists, specialist nurses, radiologists andpathologists. The service is based in Newcastle-upon-Tyne aspart of the Northern Cancer Network. It supports a pop-ulation of approximately 3 million people, and covers thegeographical area of Northumberland, Tyne and Wear,County Durham, Cleveland, Cumbria and North Yorkshire.The aim of this paper is to review the experience of theNEBSTS in the treatment of angiosarcoma, with particularregard to local and systemic failure rates.

Methods

A prospectively collected computerised database was usedto identify patients with an angiosarcoma treated by theNEBSTS between 2000 and 2008. Electronic and paperrecords were reviewed and the following data collected;

- Patient age and gender- Anatomical site- Previous radiotherapy exposure and time to subsequentangiosarcoma development

- Surgical management- Resection margin status- Histological grade

Figure 2 The defect left following surgical resection of thebreast angiosarcoma.

Figure 4 Post-operative result seen 6 months followingsurgery.

886 C.J. Lewis et al.

- Surgical reconstruction performed- Peri-operative complications- Tumour recurrence and metastases- Management of tumour recurrence

Descriptive statistical analysis and estimated survivalprobability (Kaplan-Meier equation) was calculated usingMicrosoft Excel.

Results

Fifteen patients (ten female, five male; mean age 71 years(range 39e90 years)) met the inclusion criteria (Table 2).

Previous radiotherapy exposure

Eight patients had previously been treated with radio-therapy. Five patients had received radiotherapy for breastcarcinoma, one for prostatic carcinoma, one for bladdertransitional cell carcinoma and one for a melanoma on thecalf. The mean time from radiotherapy exposure to angio-sarcoma diagnosis was 11 years (range 5e33 years).

Figure 3 Reconstruction of the defect using a pedicled lat-issimus dorsi flap and split skin grafting.

Anatomical site

Four angiosarcomas developed in the breast, three in thethigh, three in the scalp and forehead, two in the upperlimb, one in the calf, one in the myocardium and one ina mastectomy scar. At presentation, nine were larger than5 cm in their greatest dimension; six were less than 5 cm insize. No tumours were multicentric in their presentation.

Histological grade and AJCC staging

Nine angiosarcoma tumours were high-grade (Trojani grade3), three were moderate-grade (Trojani grade 2) anda further three were low-grade (Trojani grade 1) tumours.Six patients had metastatic disease at presentation, four ofwhich were to the lung, one to the parotid and one to theliver and skeleton.

Surgical resection and margin status

Fourteen patients underwent surgical resection as a primarytreatment. One individual with metastatic lung disease andmalignant pleural effusions received doxorubicin chemo-therapy alone. Nine patients had negative surgical resection

Figure 5 Haematoxylin and eosin staining of an angio-sarcoma illustrating characteristic papillary architecture.

Table 1 The AJCC sarcoma scoring system.

Tumour (T) Regional lymph nodes (N) Distant metastasis (M) Histological grade (G)

TX: Primary tumour cannotbe assessed

NX: Regional lymph nodescannot be assessed

MX: Presence of distantmetastasis cannot beaccessed

GX: Grade cannot beassessed

T0: No evidence of primarytumour

N0: No regional lymph nodemetastasis

M0: No distant metastasis G1: Well differentiated

T1: Tumour 5.0 cm or lessin greatest dimensionT1a: Superficial tumourT1b: Deep tumour

N1: Regional lymph nodemetastasis

M1: Distant metastasis G2: Moderatelydifferentiated

T2: Tumour more than 5.0 cmin greatest dimensionT2a: Superficial tumourT2b: Deep tumour

G3: Poorly differentiated

G4: Undifferentiated


margins following initial surgery. Two patients had plannedpositive margins; one due to proximity of a forearm angio-sarcoma to the median nerve, and the second for a myocar-dial angiosarcoma.

Three patients were found to have unexpectedly positivesurgicalmargins followingprimaryresection.Of these,onewasin a high-grade breast tumour presenting as a subcutaneousmass (Trojani grade3), onewas inamoderate-gradecutaneousscalp tumour (Trojani grade 2) and one was low-grade cuta-neous tumour on the dorsum of the hand (Trojani grade 1).Deep margins were involved in all three cases, despiteattempted wide excision. These patients underwent furtherwider resection, of which the re-excision margins were clear.

Table 2 Summary of patient data.

Age/sex

PreviousDXT

Time fromDXT to AS(years)

Previousurgery

1 79/F e e e

2 77/M Prostate Ca 9 e

3 65/M e e e

4 53/F e e e

5 62/F Breast Ca 10 Right bWLE

6 50/F Breast Ca 11 Rightmastec

7 80/F Breast Ca 5 Leftmastec

8 73/F e e e

9 39/M e e e

10 75/M Calfmelanoma

33 Excisionmelanogroin d

11 77/F Breast Ca 8 Breast12 82/F e e e

13 83/F Breast Ca 6 Breast14 85/F e e e

15 90/M Bladder Ca 6 e

DXT: Radiotherapy; Ca: Carcinoma; WLE: Wide local excision; ASthigh flap.

Defect reconstruction

Of the 14 patients who underwent surgical resection, ninerequired subsequent defect reconstruction. Five patientsreceived a free latissimus dorsi flap and split skin graftcoverage (Figures 1e4). Two patients underwent a freeanterolateral thigh flap and split skin grafting. Two patientsunderwent a pedicled latissimus dorsi flap.

Peri-operative complications

One pedicled latissimus dorsi flap failed and was subse-quently removed and revised. One patient developed

s Anatomicalsite

Trojanigrade

Greatestdimension(cm)

Left calf 3 7Right thigh 3 10.5Right thigh 3 4Left breast 1 6

reast Right breast 2 4

tomyRight forearm 3 7

tomyLeftmastectomyscar

2 11.9

Forehead 3 2.3Myocardium 3 1.5

calfma &issection

Right thigh 3 6.5

WLE Right breast 3 3.5Scalp 2 11

WLE Left breast 1 18Right hand 1 1.5Scalp 3 12

: Angiosarcoma; LD: Latissimus dorsi flap; ALT: Anterolateral


a wound infection, which resolved with antibiotics. Onepatient developed a chest infection, which responded toantibiotics. Surgical resection of a forehead angiosarcomaresulted in the division of the frontal division of the facialnerve, which was involved in the tumour. One patientdeveloped cardiac tamponade following progression ofa myocardial angiosarcoma. The remaining nine patientssuffered no peri-operative complications.

Adjuvant therapy

Eight patients received no adjuvant therapy followingsurgical resection. Five patients received adjuvant radio-therapy alone. Of these five patients, two had previouslyreceived radiotherapy for previous carcinoma; one forbreast cancer and the second for a calf melanoma (Table 2).One patient with myocardial angiosarcoma received adju-vant chemotherapy alone. One individual was deemedunsuitable for surgical resection due to extensivemetastaticdisease at presentation, and was treated palliatively withdoxorubicin chemotherapy.

Local control

Two patients developed local recurrence of the primarytumour, one within 9 months and the second within 11months from initial tumour excision. There was no rela-tionship between initial tumour margin status and subse-quent local tumour recurrence. The first case of recurrentdisease occurred in a patient with an angiosarcoma of thebreast, excised with negative margins. There was localrecurrence at 9 months, which was treated with wider

Histologicalmarginstatus

Surgicalreconstruction

Peri-operativecomplications

Adjuvanttherapy

Negative e e e

Negative Free LD & SSG e e

e e e Doxorubicchemothe

Negative Free LD & SSG e RadiotheNegative Pedicled LD e RadiothePlannedpositive

Free ALT & SSG Wound infection e

Negative Pedicled LD Flap failure e

Negative Free ALT &SSG

Sacrifice frontal branchof facial nerve

Radiothe

Plannedpositive

e Cardiac tamponade Chemoth

Negative e e RadiotheUnplannedpositive

e e e

Unplannedpositive

Free LD & SSG Chestinfection

Radiothe


Unplannedpositive

e e e


DXT: Radiotherapy; Ca: Carcinoma; WLE: Wide local excision; AS: Angi

excision. The second case of recurrence occurred ina patient with a lower limb angiosarcoma excised withnegative margins and treated with adjuvant radiotherapy.Local recurrence developed after 11months, and the patientdeclined further treatment.

Metastasis-free survival

Eight patients hadmetastases, which were pulmonary in fivecases, spinal andhepatic in onecase, to theaxilla and ilium inone case, and to the parotid gland in another. Six of theseeight patients had metastatic disease on presentation to ourunit. Two further patients developed metastatic diseaseafter treatment, one at 24 months and the other at 13months. Metastatic disease was associated with primarytumour size but there was no association with tumour grade.Six patients with high-grade tumours and twowith low-gradetumours developed metastatic disease. Systemic spread wasassociated in five out of eight cases with tumours greaterthan 5 cm in size, and in three out of eight caseswith tumourssmaller than 5 cm in size. There was no relationship betweenanatomical site and the development of metastatic disease.

Overall survival

Five patients died at a mean of 16 months (range 3e62months) from diagnosis. The mean follow-up of theremainder was 23 months (range 9e66 months). Onepatient was lost to follow-up. Estimated survival probabilitywas calculated using the Kaplan-Meier equation, with anestimated five-year survival of 33% (CI 1.26 to �0.59) in ourpatient cohort.

Localrecurrence

Metastaticdisease

Survival(months)

e e 5 (death)e Lung (at presentation) 10

inrapy

e Lung (at presentation) 6 (death)

rapy e Chest 9rapy e e 66

e Axilla and ilium 44

9 months e 26rapy e e 14

erapy e Liver, spine (at presentation) 3 (death)

rapy 11 months e 62 (death)e Lung (at presentation) 5 (death)

rapy e e 18

e e 20e Lung (at presentation) Unknown

e Parotid (at presentation) 23

osarcoma; LD: Latissimus dorsi flap; ALT: Anterolateral thigh flap.


Discussion

We have reported our experience of angiosarcoma in theNorth of England Bone and Soft Tissue Tumour Service. It canbe appreciated that even within a regional service coveringa large geographical area, angiosarcomas remain rare andpose significant diagnostic and management challenges, themost difficult being local recurrence and distant metastaticspread. Many patients are found to have metastatic diseaseat presentation, so a method of earlier identification ofcases would be beneficial. It is estimated that generalpractitioners will encounter one or two cases of sarcomaduring their medical career, with a cumulative incidence inour region of <0.001% cases per population over the studyperiod. Combined with their varied presentation and speedof growth, this may help to explain the number of patientswho present with extensive and metastatic disease.

There is a paucity of clear information or meta-analyseson the management of these tumours in the literature.Much of the published literature is based on case seriesresults, with no prospective phase 3 and few phase 2treatment trials,15 making the interpretation of andcomparison of results difficult.

The interval between radiotherapy and angiosarcomadevelopment is highly variable, and is reported at between1 and 54 years, with a mean interval of 15 years.16 Wereport a mean latency period of 11 years in our patientgroup, the majority of which were associated with adjuvantbreast carcinoma therapy. The Surveillance of Epidemi-ology and End Results (SEER) database assessed 195,000breast cancer patients, which showed an increase ofangiosarcomas after adjuvant radiotherapy, with a peak5e10 years after treatment.17 Indeed, patients have beenfound to have a sixteen-fold increased risk of developingangiosarcoma following radiotherapy treatment.17

The majority of angiosarcomas in our patient group werefound in the extremities, followed by the breast, head andneck, and myocardium. The distribution of our caseload issomewhat different to that found in the literature. Young15

et al. pooled data of 534 angiosarcoma cases, and found that27%occur in theheadandneck,19.7% in thebreastand15.3% inthe extremities, with smaller numbers elsewhere in the body.

These rare and highly aggressive tumours should bemanaged in tertiary centres, with available oncology,plastic surgery and orthopaedic expertise. Due to theiraggressive nature, we choose early surgical resectioncombined with immediate/delayed defect reconstructionand adjuvant radiotherapy and chemotherapy18,19 as thecurrent gold-standard treatment.20 The Vortex trial, pub-lished in 2010, set out to establish the role of post-opera-tive radiotherapy in extremity soft tissue sarcoma and hasissued radiotherapy guidelines. These categorise angio-sarcomas according to the AJCC staging criteria; large,high-grade, deeply located tumours are treated withcombined radiotherapy and surgery. Smaller, low-gradetumours are treated with surgical excision, combined withadjuvant radiotherapy if only narrow margins are taken.Only five of our patients met the criteria for adjuvantradiotherapy. It is important to note that these guidelinesare for limb sarcoma: no specific guidelines for trunkangiosarcomas exist. Radiotherapy has no role when

performing radical surgery, for example, following a limbamputation, and its overall use remains judicious.

Wide excision alone has been performed with littlecurative success and a resulting high level of local recurrence(32.5% local relapse rate after a median-free interval of 5months).21 Fourteen of our 15 patients were treated withwide surgical resection; of these, five received adjuvantradiotherapy and one patient with myocardial angiosarcomareceived adjuvant chemotherapy. High dose adjuvantradiotherapy (>50 Gy) has been retrospectively shown toimprove overall survival,15 but there are no formal trialresults. Adjuvant chemotherapy with or without concurrentradiotherapy has also been shown to be associated withimproved survival.22 Stand-alone treatment solely withradiotherapy or chemotherapy has a poor success rate.23

Diffuse tissue infiltration and contamination can result incontaminated surgical margins despite radical surgery,15

which may account for our three cases of unexpected posi-tive margins, despite wide excision. The anatomical locationof certain tumours can make wide excision difficult orimpossible, such as myocardial and head angiosarcomas.Histological Trojani grade did not correlate with positivemargin status. Despite our cases of unexpectedly positivemargins, we have demonstrated that these cases aresalvageablewith furtherwide excision,which is encouraging.

We report eight cases of metastatic disease, principallyto the chest. Local and distant recurrence is common dueto the aggressive nature of angiosarcoma even after radicalsurgery with the lungs noted to be a prime site for metas-tasis in 50% of cases.23e27 Of the six patients presentingwith metastatic disease, five had high-grade tumours.Whether this suggests that Trojani 3 grade angiosarcomasare more aggressive than the lower grade tumours, or thatthis patient group were delayed in their presentation is notfully clear. Five-year survival is reported at a rate ofbetween 20e25%.18,28 Due to the short time-scale of ourstudy over an eight-year period, we are unable to fullyreport five-year survival data but have a mean follow-uptime of 23 months for our patient group. However, of theeight patients who have available five-year data, five havedied from metastatic disease and three are alive. Thistherefore is a limitation of our study. Our estimated five-year overall survival probability is 33%, which is in keepingwith that reported in the literature.18,28

Our study is limited by what appears to be a relativelysmall caseload. Despite this, we have obtained good resultswith few cases of local recurrence and peri-operativecomplications. Our one case of flap failure occurred ina pedicled latissimus dorsi reconstruction following exci-sion of an angiosarcoma in a mastectomy scar. This patienthad been treated with adjuvant radiotherapy following hermastectomy, and one possible theory for flap failure is thatthe pedicle had suffered damage secondary to radiotherapytreatment. Both our local recurrence and metastasis ratesare in keeping with that described in other units, and ourfive-year survival rate of 37% is in keeping with that previ-ously reported.

What is the future direction of angiosarcoma diagnosisand management? Positron emission tomography (PET)scanning using 18F-fluorodeoxyglucose is able to accuratelydetect primary tumours as well as lymph node and bonemetastases in sarcoma patients.29 Both doppler ultrasound30


and magnetic resonance angiography (MRA)31 allow theassessment of vascular architecture, which can be useful inthe differentiating benign and malignant tumours.

Hyperthermic isolated perfusion of the affected limb withmelphalan and tumour necrosis factor alpha (TNF-a) repre-sents an alternative to extensive surgical resection andradiotherapy, which in itself may cause severe disability ofthe limb.32 Combined with mild hyperthermic conditions,TNF-a selectively targets tumour vasculature, leading tomicrovascular damage and loss of endothelial cohesion.33 Aswell as the vasculotoxic and cytotoxic effects of TNF-a to thetumour, the microvascular damage it causes permits entry ofmelphalan to cells to exert their effect, withminimal damageto surrounding normal tissue.32 Response rates of up to 75%have been demonstrated, allowing limb-salvaging surgery tobe performed.33,34 Its success in the treatment of lymphoe-dema-induced lymphangiosarcoma has also been high-lighted,35 with a 56% complete and 31% partial response rate.

Several recent studies have commented on the use ofanti-angiogenic monoclonal antibodies in the treatment ofangiosarcoma. Bevacizumab,15 a vascular endothelialgrowth factor-A (VEGF-A) monoclonal antibody has ach-ieved complete response when used to treat nasal angio-sarcoma in combination with radiotherapy. Sorafenib isa tyrosine-kinase inhibitor, acting on vascular endothelialgrowth factor receptors. Overall tumour response rate hasbeen reported as 14%15 with some partial and completeresponses seen. These anti-angiogenic molecules may offerangiosarcoma-specific treatments in the future, thoughtheir use is explicitly limited to trials at the moment.

Conclusion

Angiosarcomas are aggressive, difficult to treat tumours,which can occur secondary to a multitude of causes. Clin-ical suspicion, biopsy and early diagnosis are essential toallow optimum treatment, which currently consists ofradical surgery, together with adjuvant radiotherapy andchemotherapy. Biological therapy offers some hope ofa more specific treatment in the future.

Conflicts of interest

None.

Funding

Nil.

References

1. Enzinger FM, Weiss SW, editors. Soft Tissue Tumors. 3rd ed.St. Louis, Mo: Mosby; 1995. p. 648e77.

2. Meis-Kindblom JM, Kindblom LG. Angiosarcoma of soft tissue:a study of 80 cases. Am J Surg Pathol 1998;22(6):683e97.

3. Simonart T, Heenen M. Radiation-induced angiosarcomas.Dermatology 2004;209(3):175e6.

4. Khalil MF, Thomas A, Aassad A, et al. Epithelioid angiosarcomaof the small intestine after occupational exposure to radiationand polyvinyl chloride: a case report and review of literature.Sarcoma 2005;9(3e4):161e4.

5. Liu J, Waalkes MP. Liver is a target of arsenic carcinogenesis.Toxicol Sci. 2008;105(1):24e32.

6. Ruocco V, Schwartz RA, Ruocco E. Lymphoedema: an immu-nologically vulnerable site for the development of neoplasms.J Am Acad Dermatol 2003;48(6):983.

7. Sturgis EM, Potter BO. Sarcomas of the head and neck region.Curr Opin Oncol 2003;15(3):239e52.

8. ShreiberH,Barry FM,RussellWC,etal. StewartTreves syndrome:a lethal complication of postmastectomy lymphedema andregional immune deficiency. Arch Surg 1979;114:82e5.

9. Janse AJ, Coevorden F, Peterse H, et al. Lymphedema-inducedlymphangiosarcoma. Eur J Surg Oncol 1995;21:155e8.

10. Morgan MB, Swann M, Somach S, et al. Cutaneous angio-sarcoma: a case series with prognostic correlation. J Am AcadDermatol 2004;50(6):867e74.

11. Kakouri E, Whelan JS, Coltart S, et al. Multi-focal, multi-centric angiosarcoma of bone. Sarcoma 1997;1(3e4):183e7.

12. Usuda H, Naito M. Multicentric angiosarcoma of the gastroin-testinal tract. Pathol Int 1997 Aug;47(8):553e6.

13. Wenger DE, Wold LE. Malignant vascular lesions of bone:radiologic and pathologic features. Skeletal Radiol 2000;29(11):619e31.

14. Trojani M, Contesso G, Coindre JM, et al. Soft tissue sarcomasof adults; a study of pathological prognostic variable anddefinition of a histopathological grading system. Int J Cancer1984;33(1):37e42. 15.

15. Young RJ, Brown NJ, Reed MW, et al. Angiosarcoma. LancetOnco.; 26th May 2010 [Epub ahead of print].

16. De Smet S, Vandermeeren L, Christiaens MR, et al. Radiation-induced sarcoma: analysis of 46 cases. Acta Chir Belg 2008;108(5):574e9.

17. Huang J, Mackillop WJ. Increased risk of soft tissue sarcomaafter radiotherapy in women with breast carcinoma. Cancer2001;92(1):172e80.

18. Mark RJ, Poen JC, Tran LM, et al. Angiosarcoma. A report of 67patients and a review of the literature. Cancer 1996;77(11):2400e6.

19. DeVita Jr VT, Hellman S, Rosenberg SA, editors. Cancer: prin-ciples & practice of oncology. 8th ed. Philadelphia, Pa: Lip-pincott Williams & Wilkins; 2008.

20. Koch M, Nielsen GP, Yoon SS. Malignant tumours of bloodvessels; angiosarcomas, haemangioendotheliomas and hae-mangiopericytomas. J Surg Oncol 2008;97(4):321e9. 15.

21. Sordillo EM, Sordillo PP, Hadju SI, et al. Lymphangiosarcomaafter filarial infection. Cancer 1981;48:1671e9.

22. Naka N, Ohsaawa M, Tornita Y, et al. Prognostic factors inangiosarcoma: a multivariate analysis of 55 cases. J Surg Oncol1996;61:170e6.

23. Gebhart M, Chasse E, Petein M. Lymphangiosarcoma. Report of 3cases and review of the literature. J Surg Oncol 1995;21:211e4.

24. Neuhaus SJ, Pinnock N, Giblin V, et al. Treatment and outcomeof radiation-induced soft tissue sarcomas at a specialist insti-tution. Eur J Surg Oncol 2009;35(6):654e9.

25. Billingsley KG, Burt ME, Jara E, et al. Pulmonary metastasesfrom soft tissue sarcoma. Ann Surg 1999;229:602e12.

26. Grobmyer SR, Daly JM, Glotzbach RE, et al. Role of surgery in themanagement of postmastectomy extremity angiosarcoma(Stewart-Treves syndrome). Eur J Surg Oncol 2000;73:182e8.

27. Stewart NJ, Pritchard DJ, Nascimento AG, et al. Lymphangio-sarcoma following mastectomy. Clin Orthop Relat Res 1995;320:135e41.

28. Lezama-del Valle P, Gerald WL, Tsai J, et al. Malignant vasculartumors in young patients. Cancer 1998;83(8):1634e9.

29. Benz MR, Tchekmedyian N, Eilber FC, et al. Utilization ofpositron emission tomography in the management of patientswith sarcoma. Curr Opin Oncol 2009;2(4):345e51.

30. Varma DG. Optimal radiologic imaging of soft tissue sarcomas.Semin Surg Oncol 1999;17(1):2e10.


31. Knapp EL, Kransdorf MJ, Letson GD. Diagnostic imaging update:soft tissue sarcomas. Cancer Control 2005;12(1):22e6.

32. Verhoef C, deWilt J, Grunhagen D, et al. Isolated limb perfusionwith melphalan and TNF-a in the treatment of extremitysarcoma. Curr Treat Options Oncol 2007;8(6):417e27.

33. Grunhagen DJ, de Wilt JH, van Geel AN, et al. Isolated limbperfusion with TNF-alpha and melphalan in locally advancedsoft tissue sarcomas of the extremities. Recent Results CancerRes 2009;179:257e70.

34. Grunhagen DJ, de Wilt JH, Graveland WJ, et al. Outcome andprognostic factor analysis of 217 consecutive isolated limbperfusions with tumor necrosis factor-alpha and melphalan forlimb-threatening soft tissue sarcoma. Cancer 2006;106(8):1776e84.

35. Lans TE, de Wilt JH, van Geel AN, et al. Isolated limb perfusionwith tumor necrosis factor and melphalan for nonresectableStewart-Treves lymphangiosarcoma. Ann Surg Oncol 2002;9(10):1004e9.

experience of angiosarcoma in the north of england bone and soft tissue tumour service

Documents