Results of limb-sparing surgery with vascular replacement for soft tissue sarcoma in the lower extremity

From the German, Austrian and Swiss Societies of Vascular and Endovascular Surgery

Results of limb-sparing surgery with vascular replacement for soft tissue sarcoma in the lower extremity Matthias H. M. Schwarzbach, MD,a,b Yura Hormann,b Ulf Hinz, MSc,c Ludger Bernd, MD,d Frank Willeke, MD,e Gunhild Mechtersheimer, MD,f Dittmar Böckler, MD, PhD,a Hardy Schumacher, MD, PhD,a Christian Herfarth, MD, FACS,b Markus W. Büchler, MD, FACS,b and Jens-R. Allenberg, MD, PhD,a Heidelberg, Germany Objective: To evaluate limb-salvage surgery with vascular resection for lower extremity soft tissue sarcomas (STS) in adult patients and to classify blood vessel involvement. Methods: Subjects were consecutive patients (median age, 56 years) who underwent vascular replacement during surgery of STS in the lower limb between January 1988 and December 2003. Blood vessel involvement by STS was classified as follows: type I, artery and vein; type II, artery only; type III, vein only; and type IV, neither artery nor vein (excluded from the analysis). Patient data were prospectively gathered in a computerized database. Results: Twenty-one (9.9%) of 213 patients underwent vascular resections for lower limb STS. Besides 17 type I tumors (81.0%), 3 (14.3%) type II and 1 (4.7%) type III STS were diagnosed. Arterial reconstruction was performed for all type I and II tumors. Venous replacement in type I and III tumors was performed in 66.7% of patients. Autologous vein (n ⴝ 8) and synthetic (Dacron and expanded polytetrafluoroethylene; n ⴝ 12) bypasses were used with comparable frequency for arterial repair, whereas expanded polytetrafluoroethylene prostheses were implanted in veins. Morbidity was 57.2% (hematoma, thrombosis, and infection), and mortality was 5% (embolism). At a median follow-up of 34 months, the primary and secondary patency rates of arterial (venous) reconstructions were 58.3% (54.9%) and 78.3% (54.9%). Limb salvage was achieved in 94.1% of all cases. The 5-year local control rate and survival rate were 80.4% and 52%, respectively. We observed a 5-year metastasis-free survival rate of 37.7% and found vessel infiltration and higher tumor grade (low-grade vs intermediate grade and high grade tumors) to be negative prognostic factors at univariate and multivariate analysis. Conclusions: Long-term bypass patency rates, the high percentage of limb salvage, and the oncologic outcome underline the efficacy of en bloc resection of STS involving major vessels in the lower limb. Disease-specific morbidity must be anticipated. The classification of vascular involvement (type I to IV) is useful for surgical management. ( J Vasc Surg 2005;42:88-97.)

Function-sparing resection has become the standard recommended surgical treatment for lower limb soft tissue sarcomas (STS).1-4 Local tumor control is related to the type of resection and whether patients received adjuvant radiotherapy.4-6 Complete resection improves both local tumor control and survival.6-8 However, STS that involve vascular structures remain a crucial issue. If STS arise from arterial or venous blood vessels, then vascular resection is inevitable.9-16 Furthermore, if STS infiltrate or surround vascular structures, then the vessels must also be resected to meet established procedural stanFrom the Department of Vascular and Endovascular Surgery,a Department of Surgeryb with the Unit for Documentation and Statistics,c Department of Orthopedic Surgery,d Institute of Pathology,f Heidelberg, Department of Surgery and Vascular Surgery, Mannheim,e University of Heidelberg. Competition of interest: none. Presented at the Joint Annual Meeting of the German, Austrian, and Swiss Societies of Vascular and Endovascular Surgery (Dreilaendertagung), Innsbruck, Austria, September 16, 2004. Reprint requests (current address): M. H. M. Schwarzbach, MD, Department of Surgery and Vascular Surgery, Mannheim University Clinic, University of Heidelberg, Theodor-Kutzer-Ufer 1-3, 68135 Mannheim, Germany (e-mail: [email protected]). 0741-5214/$30.00 Copyright © 2005 by The Society for Vascular Surgery. doi:10.1016/j.jvs.2005.03.017

88

dards.1,2,17-27 STS surrounded by a plane of normal tissue can be dissected from major blood vessels. By longitudinally splitting the adventitia opposite the tumor, a rim of normal tissue is preserved in the vessel-tumor interface.27,28 Vascular resection, however, increases the width of the resection margin.26,27 Therefore, it was proposed that whenever it is impossible to achieve a wide resection margin without vascular resection, vascular resection is indicated.26,27 Only few published studies have reported the outcome of limb-sparing of STS surgery with blood vessel replacement, and to our knowledge, only two series included more than ten patients.17-27 Furthermore, blood vessel involvement has not been systematically classified, nor has a treatment algorithm been suggested. The aim of this study was to classify vascular involvement by extremity STS, to suggest a treatment algorithm, and to analyze the results of limb-sparing surgery necessitating vascular resection. METHODS Selection of patients. Consecutive adult patients in whom STS involved major arterial or venous blood vessels in the lower extremity and who were treated by limbpreserving resection between January 1988 and December

JOURNAL OF VASCULAR SURGERY Volume 42, Number 1

Schwarzbach et al 89

Limb-preserving soft tissue sarcoma resection

Vascular embracement / attatchment / infiltration

Type I

Type II

Artery + Vein (n=17)

Artery (n=3)

Impaired collateral venous drainage Collateral venous (resection/ligation drainage (patent GSV) GSV)

Arterial + Venous Reconstruction (n=12)

No vascular involvement

Type III

Type IV

Vein (n=1)

None (n=192)*

Impaired collateral venous drainage Collateral venous (resection / ligation drainage (patent GSV) of GSV)

Arterial Reconstruction (n=8)

Venous Reconstruction (n=0)

No vascular Reconstruction (n=1)

Fig 1. Type of vascular involvement and algorithm for vascular reconstruction in patients with extremity soft tissue sarcoma. *Number of patients with type IV soft tissue sarcomas treated during the study period (excluded from further analysis). GSV, greater saphenous vein.

2003 at the University Clinic of Heidelberg are described here. Patients were identified from a prospectively gathered database containing prospective data from all patients with STS treated during this period. Patients were diagnosed with STS (color duplex sonography, magnetic resonance [MR] imaging, computed tomography [CT], angiography, or a combination of these) arising in a major blood vessel or extending toward a major blood vessel in the lower extremity. Vascular involvement was diagnosed when MR or CT imaging did not show a rim of normal tissue in the tumorto-vessel interface. The indications for vascular resection were standardized. Patients underwent vascular resection when STS were of primary vascular origin (primary vascular involvement) or were infiltrating, encasing, or affecting major blood vessels (secondary vascular involvement). Classification of vascular involvement. Types of vascular involvement were assessed before surgery by highresolution CT or MR imaging. STS involving both major arteries and veins were classified as type I. STS affecting only arterial blood vessels were classified as type II. Sarcomas involving major veins without altering an artery were classified as type III. STS without involvement of arterial or venous blood vessels were classified as type IV (Fig 1). Surgery, radiation, and vascular replacement. STS were resected en bloc together with blood vessels according to the type of vascular involvement and the surgical standards described by Enneking et al.1,2 Neural structures were included in the en bloc specimen if they were encased or infiltrated by the tumor. One pathologist (G.M.) assessed the resected specimens for infiltration of the blood vessel wall, tumor grade, and microscopic margin. Grade was determined by the degree of cellularity, differentiation, pleomorphism, necrosis, and mitotic activity and was cate-

gorized into low grade, intermediate grade, and high grade.29,30 Radiotherapy was given in higher-grade (intermediate grade and high grade) tumors and recurrent STS by using an intraoperative and postoperative protocol (during surgery only in patients who had received radiation treatment previously or selectively as external beam radiotherapy).31,32 Radiosensitive structures (vessels or nerves) were spared from radiation by dissection, mobilization, and positioning of the collimation tube. Arterial and venous blood flow were restored with respect to the resection site, the extent of the vascular defect, and residual venous drainage. In type I and II situations, arteries were reconstructed either by ring-enforced synthetic prostheses (expanded polytetrafluoroethylene [ePTFE] or Dacron or by autologous vein (reversed greater saphenous vein [GSV]). In type I and III settings, deep iliac and/or femoral veins were reconstructed by ring-sustained ePTFE prostheses or autologous vein grafts (nonreversed GSV). If the GSV had not been removed previously and was patent, venous replacement was not routinely performed. Anatomic repair was the preferred reconstruction, and the contralateral GSV was the preferred autologous graft. Routinely, patients received prophylactic antibiotic treatment (cephalosporin), and suction drains were placed. Perioperatively, a low-dose regimen of heparin was administered as unfractionated heparin during surgery and low-molecular-weight heparin in the postoperative phase. After discharge, oral anticoagulants (phenprocoumon) were administered restrictively only in below-knee reconstructions. Study population. A total of 385 adult patients with STS (extremities, retroperitoneum, and trunk) underwent operation. In 45 (11.7%) of these patients, vascular resections were performed. For the purpose of this analysis,

JOURNAL OF VASCULAR SURGERY July 2005

90 Schwarzbach et al

Table I. Clinicopathologic factors in 21 patients with soft tissue sarcomas of the lower extremity involving major blood vessels* Characteristic

Table II. Treatment type in 21 patients with soft tissue sarcomas of the lower extremity involving major vascular structures

No. (N ⫽ 21)

%

10 11

48% 52%

15 6

71% 29%

11 4 3 2 1

52% 19% 14% 10% 5%

16 5

76% 24%

3 5 13

14% 24% 62%

1 4 4 12

5% 19% 19% 57%

Sex Male Female Presentation status Primary tumor Local recurrence Histologic diagnosis Liposarcoma Leiomyosarcoma Malignant fibrous histiocytoma Synovial sarcoma Angiosarcoma Tumor localization Thigh Groin Tumor grade Low Intermediate High Vessel infiltration Artery Vein Artery and vein No *Median tumor size was 10.5 cm (range, 4-24 cm).

retroperitoneal STS (n ⫽ 23) and upper extremity STS (n ⫽ 1) were excluded. Therefore, data from 21 patients were analyzed, giving a 9.9% rate of vascular resections in the group of lower extremity STS (n ⫽ 213). The median age of the patients was 56 years (interquartile range [IQR], 32-75 years). Swelling (n ⫽ 18), pain (n ⫽ 9), and neurologic disorders (n ⫽ 2) were observed. Most of the patients were referred for primary STS. Two patients were diagnosed with the first, two with the second, one with the third, and one with the fifth local recurrence. A total of 19 (90%) patients were treated for locally advanced STS without concomitant metastatic disease, and 2 (10%) patients were treated for locally advanced STS with distant disease. Liposarcoma was the predominant histopathologic diagnosis, and higher-grade (intermediate grade and high grade) tumors were found in 62% of the patients. In most STS (90%), secondary involvement of blood vessels was observed, whereas two STS originated in the wall of the blood vessel. Sixty-five percent of the patients received adjuvant radiation therapy. The clinicopathologic data are summarized in Table I. Arterial reconstruction. In four patients, iliofemoral repair was performed; this repair consisted of two ePTFE prostheses (8 mm), one Dacron bypass graft (8 mm), and one autologous vein graft. In one of these patients, the deep femoral artery was ligated. One iliocrural (anterior tibial artery) reversed venous bypass (extra-anatomic) was implanted after extensive compartmental resection. Femorofemoral replacement in three patients was performed by

Characteristics Resection type CR with negative margin CR with positive margin ICR Vascular replacement Artery Vein Artery and vein Arterial reconstruction Vein/ePTFE/Dacron Venous reconstruction Vein/ePTFE Adjuvant treatment Only surgery Surgery ⫹ IORT ⫹ ERT Surgery ⫹ IORT-ERT Surgery ⫹ ERT

No. (N ⫽ 21)

%

14 5 2

66% 24% 10%

8 0 12 20 8/10/2 12 2/10

40% 0 60% 100% 40%/50%/10% 60% 17%/83%

8 7 4 2

38% 33% 19% 10%

CR, Complete resection; ICR, incomplete resection; ePTFE, expanded polytetrafluoroethylene; IORT, intraoperative radiotherapy; ERT, external radiotherapy.

using autologous vein (n ⫽ 2) or ePTFE (6 mm; n ⫽ 1) prostheses. In the latter reconstructions, the deep femoral artery was reconnected to the axial bypass with a segment of reversed GSV (n ⫽ 2) or was ligated (n ⫽ 1). Femoropopliteal bypasses above the knee were implanted in ten patients. Six of these patients received ePTFE prostheses (6 mm), three patients were treated by reversed GSV, and in one case a Dacron bypass graft (6 mm) was used. One autologous femoropopliteal graft ended at the level of the knee joint, and one femoropopliteal ePTFE prosthesis (6 mm) ended below the knee joint. In 19 patients, anatomic reconstructions were performed. Autologous grafts consisted of contralateral GSV (n ⫽ 6) and ipsilateral GSV (n ⫽ 2). GSV was implanted in the reverse direction (Table II). Venous reconstruction. Venous reconstructions consisted of three iliofemoral and nine femorofemoral/ popliteal reconstructions. For proximal iliofemoral reconstruction, ePTFE prostheses (8 or 12 mm, n ⫽ 2) or autologous venous grafts (GSV, n ⫽ 1) were used. In the latter cases, the deep femoral vein was ligated. Four femorofemoral bypasses were implanted by using either ePTFE prostheses (8 mm, n ⫽ 2 and 10 mm, n ⫽ 1) or ipsilateral autologous GSV. In one of these patients, the deep femoral vein was ligated. Another five venous reconstructions were performed in the femoropopliteal region above the knee. For these procedures, ePTFE prostheses were used (8 mm, n ⫽ 2; 10 mm, n ⫽ 1; or 12 mm, n ⫽ 2). A resected femoral vein was ligated in six patients in whom it was possible to preserve a functional GSV during sarcoma resection (Table II). Follow-up. Patients were seen regularly during the observation period in our outpatient clinic. The standard follow-up was at 6 weeks and 6 months after surgery and yearly thereafter. Patients were questioned about symp-

JOURNAL OF VASCULAR SURGERY Volume 42, Number 1

toms suggestive of graft thrombosis, including pain and discomfort. Clinical examinations were usually combined with duplex sonography. In addition, patients were seen according to the routine oncologic follow-up schedule. To evaluate the outcome and patient acceptance of the limbpreserving operation, the following data were gathered at the most recent assessment. (1) Patients were asked their opinion about the surgical result (categories: excellent, good, and poor); (2) patients were asked whether they would have preferred amputation (categories: amputation or limb salvage); and (3) limb function was assessed by using the Enneking system for reconstructive procedures after surgical treatment of tumors of the musculoskeletal system.33 Statistical analysis. SAS software (release 9.1; SAS Institute, Inc, Cary, NC) was used for statistical analysis. The distribution of age at operation, tumor size, and follow-up time were described as median with IQR. KaplanMeier estimations were used to analyze the rates of patency, limb salvage, and local tumor control; the metastasis-free period; and overall survival from the date of operation.34 The 2- and 5-year rates with a corresponding 95% confidence interval (CI) are presented for each end point. Patients in whom no event was observed for one of the end points were censored at the last follow-up and are indicated in the figures as a vertical line. The end of the follow-up period for all patients alive was in February 2004. One patient was lost to follow-up 23 months after operation and was included as a censored observation. The log-rank test was performed to compare survival time distributions between curves regarding tumor grade, margin of resection, and vessel infiltration. Cox proportional hazards regression analysis35 was performed to analyze the multivariate effect on overall survival of the factors tumor grade, margin of resection, and vessel infiltration. The estimated hazard ratios were given. Two-sided P values were always computed, and an effect was considered statistically significant at P ⱕ .05. RESULTS Resectability and vascular infiltration. All tumors could be resected without amputation. In all, 19 (90%) of 20 patients were treated by complete en bloc resection for locally advanced STS without concomitant metastatic disease (Table II). Vascular resection was performed for 17 (81%) type I, 3 (14.3%) type II, and 1 (4.7%) type III STS (Fig 1). Histopathologic examination of the specimens showed that blood vessels had been infiltrated in 43% of the STS. Four (19%) specimens showed both arterial and venous infiltration, four (19%) specimens showed venous infiltration, and one (5%) specimen showed arterial infiltration (Table I). Involvement of major nerves was observed in nine (42.8%) patients. Of these, the femoral nerve was partially resected in six patients, and the sciatic nerve was partially resected in three patients. Involvement of nerves was diagnosed only in patients with type I STS (53%). Morbidity and mortality. Surgical morbidity during the hospital stay developed in 12 (57.2%) of 21 patients.

Schwarzbach et al 91

Table III. Morbidity, reoperation rate, and mortality in 21 patients with lower extremity soft tissue sarcoma involving major vascular structures Variable Morbidity (hospital) Surgical morbidity Soft tissue/skin necrosis Hematoma Wound infection Graft thrombosis Lymphatic fistula Nerve damage (peroneal paresis) Reoperation Bypass revision Wound revision Hematoma revision Medical morbidity Pulmonary embolism Mortality (hospital)*

Total (N ⫽ 21)

%

12 5 4 2 1 3 1 7 2 3 2 1 1 1

57.2%

33%

4.8% 4.8%

*Fatal pulmonary embolism.

The most common complications observed were soft tissue and skin necrosis, hematoma, and lymphatic fistula (Table III). Seven (33%) patients required operative reintervention. Most patients with wound complications were treated conservatively (57%), as were half of the patients with postoperative hematoma. At the 17th postoperative day, one above-knee ePTFE bypass had become occluded because of a technical failure and had to be replaced. One massively obese woman died 12 days after surgery of a fatal pulmonary embolism (mortality 5%). After discharge from the hospital, bypass infections were seen in four patients (19%). These infections were diagnosed 1.5, 2, 4, and 5 months after surgery. In these patients, vascular replacement was performed for three type I and one type II STS. The infected material included ePTFE (n ⫽ 2), Dacron (n ⫽ 2), and vein (n ⫽ 1). All patients were successfully treated by limb-preserving reoperation. One patient had an infected venous ePTFE prosthesis and an uninfected arterial GSV graft. The treatment consisted of explantation of the infected ePTFE graft. Another patient presented with an infected arterial and venous ePTFE prosthesis. Subsequently, the graft was explanted, and the arterial bypass was replaced. Another patient had an infected venous Dacron graft and an infected arterial GSV bypass that necessitated explantation. Another patient was successfully treated by operation for a type II STS in which an infected Dacron graft was replaced by an autologous graft. Arterial resection and graft function. Twenty (95.2%) of 21 patients showed arterial involvement. Arterial reconstructions were performed in all patients with resections for type I, or II, or III disease. Synthetic grafts were used in 12 reconstructions (60%). Median follow-up was 34 months for surviving patients (IQR, 20-62 months). Seven (35%) arterial bypass occlusions were observed in 20 reconstructions. The primary 2- and 5-year patency rates of arterial reconstructions were 58.3% and 58.3%, respectively (2- and 5-year 95% CI: 31.5%-77.8%

JOURNAL OF VASCULAR SURGERY July 2005

%

%

100

100

80

80

60

40

Primary 2- and 5-year patency rates (n=20): 58.3% (95% CI 31.5-77.8) and 58.3% (95% CI 31.5-77.8) Secondary 2- and 5-year patency rates (n=20): 78.3% (95% CI 51.9-91.3) and 78.3% (95% CI 51.9-91.3)

Patency Distribution Function

Patency Distribution Function

92 Schwarzbach et al

40

Primary/secondary 2- and 5-year patency rates (n=12): 54.9% (95% CI 18.7-80.6) and 54.9% (95% CI 18.7-80.6)

20

20

0

0

20 20

12 10 15

24 7 10

36 4 7

48 60 Months after Resection 4 7

4 5

Fig 2. Primary and secondary arterial patency in 20 patients with resection of soft tissue sarcoma and major arteries in the lower limb. The median follow-up was 34 months (interquartile range, 20-62 months). Censored patients are marked (as a vertical line); patients at risk and the 95% confidence interval (CI) are shown.

and 31.5%-77.8%, respectively). Graft occlusion was observed in autologous and prosthetic bypasses. Specifically, thrombosis developed in 3 (37.5%) of 8 venous bypasses (femoropopliteal GSV grafts) and 4 (33.3%) of 12 prosthetic reconstructions (femoropopliteal ePTFE grafts). Bypass revision was performed in two patients, bypass replacement in three patients, and graft explantation in one patient. One patient with an asymptomatic femoropopliteal bypass occlusion 16 months after surgery was treated conservatively. The secondary 2- and 5-year patency rates of arterial reconstructions were 78.3% and 78.3%, respectively (5 events, 25%; 2- and 5-year 95% CI: 51.9%-91.3% and 51.9%-91.3%, respectively; Fig 2). Venous resection and graft function. Among all 21 patients, 18 (85.7%) were treated by venous resection: 17 (81%) for type I involvement and 1 (4.8%) for type III involvement. Of type I resections, 12 (70.5%) were treated by venous replacement. The vein was not reconstructed in one patient with a type III resection. Thus, of 18 patients in whom venous resections were performed, veins were reconstructed only in 12 (66.6%). Most of the patients received prosthetic grafts (ePTFE prostheses, n ⫽ 10; GSV, n ⫽ 2) to replace the resected veins. A patent venous reconstruction was found in seven (58.3%) patients, whereas venous occlusions occurred in five (41.7%) patients. The primary and secondary venous patency after 2 and 5 years was 54.9% and 54.9%, respectively (2- and 5-year 95% CI: 18.7%80.6%; Fig 3). The median interval until venous thrombosis was observed was 12 months (range, 1.7-64 months). None of these patients required operative or interventional treatment. Venous thrombosis was clinically asymptomatic or presented with well-tolerated edema. Limb salvage and limb function. The 2- and 5-year limb-salvage rate was 94.1% (one event, 5%; 2- and 5-year 95% CI: 65%-99.2%; Fig 4). One patient required hemipelvectomy because of local recurrence in the proximal thigh 12 months after surgery. In all patients in whom arterial

Patients at risk:

0

12

24

36

12

4

4

3

48 60 Months after Resection 3

1

Fig 3. Primary and secondary venous bypass patency in 12 patients who underwent resection of soft tissue sarcoma and major deep veins in the lower limb. The median follow-up was 34 months (interquartile range, 20-62 months). Censored patients are marked (as a vertical line); patients at risk and the 95% confidence interval (CI) are shown.

% 100 Limb Salvage Distribution Function

0 Patients at risk:

60

80

2- and 5-year limb salvage rates (n=20): 94.1% (95% CI 65-99.2) and 94.1% (95% CI 65-99.2)

60

40

20

0

Patients at risk:

0

12

24

36

20

17

13

9

48 60 Months after Resection 9

7

Fig 4. Limb-salvage rates in 20 patients who underwent vascular reconstruction during soft tissue sarcoma surgery in the lower limb. The median follow-up was 34 months (interquartile range, 20-62 months). Censored patients are marked (as a vertical line); patients at risk and the 95% confidence interval are shown.

occlusion occurred during the observation period, ischemic limb loss was prevented by reoperation (except for one patient, who was treated conservatively). Nine (42.8%) of 21 patients were interviewed and evaluated regarding limb function. Concerning the acceptance of the surgical result, five patients thought that the result was excellent, three patients thought that the result was good, and one patient thought that the result was poor (because of a muscular contracture that impaired walking). None of the patients interviewed would have preferred primary amputation instead of the limb-sparing operation with vascular replacement. With the Enneking score (with a maximum of 5 points), pain was assessed with an average of 4.2 points, function with an average of 2.9 points, emotional acceptance with an average of 3.3 points, support (brace, cane, or

JOURNAL OF VASCULAR SURGERY Volume 42, Number 1

DISCUSSION Limb-preserving resection is recommended for patients with STS of the lower extremity.1-8 Locally advanced STS involving major vascular or neural structures are also considered as an indication for limb salvage.23-28,36,37 With modern surgical techniques, STS can be resected en bloc together with affected neurovascular structures.26 The first description of sarcoma resection in the extremities with subsequent arterial replacement was reported by Fortner et al17 in 1977. In addition, Fortner et al17 and Imparato et al18 described venous replacement during limb-sparing surgery. The importance of vascular involvement by extremity STS has been underestimated. In this analysis, blood vessels were affected in almost 10% of all adult patients who were treated for lower limb STS at our institution during the study period. In a review of the literature, blood vessel involvement was reported in only 5% of the cases.23,28 Most of the clinical series, however, are not representative because not all patients undergoing operation for extremity STS during the periods studied were analyzed.17-27 Furthermore, several investigators did not specifically focus on the histopathologically well defined tumor entity of adult STS but included other malignancies, such as carcinoma, sarcoma of the bone (osteosarcoma and Ewing’s sarcoma), and benign lesions (Table IV).17,18,21,22,28 From an oncologic point of view, there is little use in reporting tumor control and survival in such an eclectic groups of patients. These findings hamper the interpretation of the clinical

%

Survival Distribution Function

100

80

60

40 2- and 5-year local tumor control rates (n=19): 93.8% (95% CI 63.2-99.1) and 80.4% (95% CI 37.6-95.2)

20

2- and 5-year overall survival rates (n=21): 70.4% (95% CI 45.5-85.5) and 52.0% (95% CI 27.3-71.9) 2- and 5-year metastasis-free survival rates (n=19) : 63.6% (95% CI 36.0-81.8) and 37.7% (95% CI 13.8-62.5)

0

Patients at risk:

0

12

24

36

19 21 19

14 17 13

13 14 10

9 9 7

48 60 Months after Resection 9 9 7

6 7 4

Fig 5. Local tumor control, survival, and metastasis-free survival after complete resection with vascular reconstructions in 19 patients for lower-limb extremity soft tissue sarcoma. The median follow-up was 34 months (interquartile range, 20-62 months). Censored patients are marked (as a vertical line); patients at risk and the 95% confidence interval are shown. % 100

Survival Distribution Function

crutches) with an average of 3.7 points, walking ability with an average of 3.7 points, and gait with an average of 3.1 points. Taken together, all patients achieved an average of 20.9 points, which is an average of 69.6%. Tumor control and survival. During follow-up, 3 (15.8%) of 19 patients who were treated by complete tumor resection developed local recurrence. The 2- and 5-year local tumor control rates of these patients were 93.8% and 80.4% (2- and 5-year 95% CI: 63.2%-99.1% and 37.6%-95.2%, respectively; Fig. 5). At last follow-up, 11 (52.4%) patients had died. The 2- and 5-year overall survival rates were 70.4% and 52.0%, respectively (2- and 5-year 95% CI: 45.5%-85.5% and 27.3%-71.9%, respectively; Fig 5). Nine patients (47.4%) died because of hematogenous metastatic disease after complete resection. The 2- and 5-year metastasis-free survival rates were 63.6% and 37.7% (2- and 5-year 95% CI: 36.0%-81.8% and 13.8%62.5%, respectively; Fig 5). Univariate analysis determined vessel infiltration at histopathologic examination (positive vs negative, P ⫽ .0027), tumor grade (high grade vs intermediate grade and low grade, P ⫽ .0266), and margin of resection (complete resection with clear margin vs complete resection with positive margin, P ⫽ .0215) to be significant prognostic factors for survival. By multivariate analysis, high tumor grade (P ⫽ .0439) and blood vessel infiltration (P ⫽ .0124) were confirmed as significant risk factors associated with an increased risk of death (hazard ratios of 5.914 and 13.002, respectively; Fig 6).

Schwarzbach et al 93

80

60

No vessel infiltration (n=11) Vessel infiltration (n=8)

40 P-value = .0027

20

0

Patients at risk:

0

12

24

36

11 8

10 5

10 3

8 1

48 60 Months after Resection 8 0

7 0

Fig 6. Survival according to vessel infiltration after complete resection with vascular reconstructions in 19 patients for lower-limb extremity soft tissue sarcoma. The median follow-up was 34 months (interquartile range, 20-62 months). Censored patients are marked (as a vertical line); patients at risk and the 95% confidence interval are shown.

results and comparability of different studies. Furthermore, only two series reported on more than ten patients with extremity STS, and only one of these analyses was restricted to the lower extremity.23,26 This series is the largest series of a consecutive group of adult patients with lower limb STS analyzed to date for vascular involvement. Another aspect underlining the importance of oncologic vascular surgery in patients with extremity STS is the considerable number of specimens in this analysis with histopathologically proven sarcoma infiltration of the blood vessel wall. Infiltrative growth was found to affect not only arteries and veins together, but also arteries or veins alone. Until now, this important clinical feature of extremity STS has been evaluated by only a few investigators.26,28 Not only the necessity of vascular en bloc resection and repair, but also

JOURNAL OF VASCULAR SURGERY July 2005

94 Schwarzbach et al

Table IV. Review of the literature: Clinicopathological data and clinical results of surgery in lower extremity soft tissue sarcoma involving major blood vessels Years studied

No. Patients

Tumor type

Fortner 197717

1955-1976

7

Imparato 197818

1967-1976

13

Nambisan 198619

1977-1985

10

5 STS 1 CA 1 benign 9 STS 4 OS 10 STS

Reference

Ohara 199120 Wuisman 199421

2 12

Highergrade tumor

Tumor size (cm)

Type of STS (resected vessels)

PT/LR

#

7

#

6 type I 1 type II

2 iliac 5 femoral

#

5-15

#

13 type I

80%

#

#

7 type I 3 type II

# #

# #

2/0 #

2 type III 12 type I

11 femoropopliteal 2 subclavian 3 iliac 4 iliofemoral 2 femoropopliteal 1 crural 2 femoral 12 popliteal

6 type I 1 type II 7 type III 8 type I 13 type II

Koperna 199622

1984-1992

14

Karakousis 199623

1979-1994

21

2 STS 11 OS 1 ES 5 STS 8 OS 1 ES 21 STS

Kawai 199624

1982-1994

8

8 STS

#

#

8/8

8 type I

Reix 199825 Hohenberger 199926

1985-1995 1984-1992

7 20

7 STS 20 STS

# 90%

5 #

7/0 15/5

7 type III 11 type I 8 type II 1 type III

Bonardelli 200027

1995-1999

7

7 STS

85%

#

5/2

Present series 2004

1988-2003

21

21 STS

86%

10

15/6

4 type I 2 type II 1 type III 17 type I 3 type II 1 type III

#

#

#

86%

9.4

17/4

Graft location

2 iliofemoral 10 femoropopliteal 2 popliteal 7 iliofemoral 13 femoropopliteal 1 subclavian 6 femoral 2 popliteal 7 femoral 6 iliofemoral 1 femorofemoral 9 femoropopliteal 3 popliteopopliteal 4 femoropopliteal 4 iliofemoral 1 iliocrural 3 femorofemoral 12 femoropopliteal

STS, Soft tissue sarcoma; BP, bypass; ePTFE, expanded polytetrafluoroethylene; #, not specified; OS, osteosarcoma; ES, Ewing’s sarcoma; CA, carcinoma; LR, local recurrence; DR, distant recurrence; PT, primary tumor. *Five-year-survival rate. **Absolute number of survivors.

the necessity of classifying the different forms and providing a treatment algorithm, is indicated by this finding. The decision of whether or not to resect blood vessels and to what extent depends on preoperative MR/CT imaging and intraoperative findings.17,18,26,38 MR angiography might be augmented by conventional angiography or duplex sonography. The value of positron emission tomography is unclear.39 Using standard radiologic diagnostic procedures such as MR or CT imaging, we have suggested a four-stage classification that describes the growth pattern of extremity STS with respect to the major vascular structures of the limbs. Patients with a type I STS (blood vessel involvement of artery and vein) require arterial reconstruction after en bloc resection, whereas venous repair depends on the collateral venous drainage. If sufficient collateral venous drainage is preserved, then venous reconstruction is not necessary. Patients in whom only arteries are affected by the STS (type II involvement) regularly require blood vessel replacement after tumor clearance. Those with a type III STS (involvement of a vein) will benefit from venous

reconstruction if sufficient collateral venous drainage has not been maintained. Type IV, by far the most common type of STS in the extremities, does not require blood vessel replacement because an adequate margin can be obtained without vascular resection. We present our algorithm in Fig 1. In the iliofemoral and femoropopliteal region above the knee, we used predominantly synthetic grafts for arterial reconstruction. The rationale for this was to reduce the operative time and to preserve the ipsilateral GSV. Alternatively, autologous vein grafts, such as the GSV, from the contralateral leg can be used.17,23,26 If the superficial and deep femoral arteries were affected by the STS, we reconstructed both. In patients with distal arterial reconstructions, autologous bypasses were preferred. Venous repair has to be performed according to the functional integrity of the ipsilateral GSV. If the ipsilateral GSV was preserved and was patent, then the remaining femoral vein stumps were ligated.26 However, if the GSV was not functioning (sclerosis) or was absent or resected together with the femoral vein, then reconstruction was performed. Venous repair in

JOURNAL OF VASCULAR SURGERY Volume 42, Number 1

Schwarzbach et al 95

Table IV. Continued.

Arterial repair

Venous repair

Follow-up (mo)

2 venous BP 2 Dacron BP 3 ligation 13 venous BP

3 venous BP 3 ligation

#

3 venous BP

46

1 venous BP 6 ePTFE BP 3 Dacron BP

6 ePTFE BP

24.8

None

2 ePTFE BP

#

12 venous BP

12 venous BP

#

7 venous BP

8 venous BP 1 Dacron BP 4 ePTFE BP 1 venous BP 2 ePTFE BP 5 ligation 2 venous BP 5 ePTFE BP 1 ligation 2 venous patch 5 ligation 6 venous BP 4 ePTFE BP 1 anastomosis 3 venous BP 2 transposition

4 venous BP 17 ePTFE BP 4 venous BP 4 Dacron BP subad. dissection 8 venous BP 11 ePTFE BP 2 venous BP 2 ePTFE BP 3 subad. dissection 8 venous BP 10 ePTFE BP 2 Dacron BP

2 venous BP 10 ePTFE BP 6 ligation

Arterial/venous patency

Limb-salvage (primary/ secondary)

Morbidity/ mortality

LR/DR

Survival

4/4 (100%)/ 3/3 (100%)

7/7 (100%)/ 5/7 (71%)

4/7 (57%)/ 0/7 (0%)

2/7 (29%)/ 2/7 (29%)

86%*

12/13 (92%)/ 1/3 (33%) 9/10 (90%)/ 2/6 (33%)

11/13 (85%)/ 10/13 (77%) 9/10 (90%)/ 8/10 (80%)

1/13 (8%)/ # 6/10 (60%)/ #

2/13 (15%)/ 5/13 (38%) 0 6/10 (60%)

8/13(62%)**

# 1/2 (50%) 11/12 (92%)/ 10/12 (83%) 6/7 (86%)/ 10/13 (77%)

2/2 (100%)/ 2/2 (100%) 10/12 (83%)/ 10/12 (83%) #/#

2/2 (100 %)/ # 6/12 (50%)/ 0/12 (0%) 7/14 (50%)/#

0 # 0 0 0 6/14 (43%)

2/2 (100%)**

20/21 (95%)/ 20/21 (95%)

7/21(33%)/ 0/21 (0%)

3/21 (14%)/ 8/21 (38%)

63%*

6/8 (75%)/ 6/8 (75%)

2/8 (25%)/#

0 2/8 (25%)

8/8 (100%)**

0/7 (0%)/ 0/7 (0%) 7/20 (35%)/ 1/20 (5%)

1/7 (14%)/ 4/7 (57%) 4/20 (20%)/ 11/20 (55%)

3/7 (43%)**

19/19 (100%)/ 9/10 (90%)

7/7 (100%)/ 7/7 (100%) 19/20 (95%)/ 19/20 (95%)

25

4/4(100%)/ 5/5 (100%)

7/7 (100%)/ 7/7 (100%)

1/7 (14%)/ 0/7 (0%)

0 2/7 (29%)

5/7 (71%)**

34

15/20 (75%)/ 7/12 (58%)

21/21 (100%)/ 20/21 (95%)

12/21 (57%)/ 1/21 (4.8%)

3/21 (14%)/ 10/21 (48%)

52%*

55 31 42.5 36.4

#/# 5/8 (63%)/ 2/7 (29%) #/#

such situations prevented limb edema and procedure-associated clinical sequelae (complicated wound healing, pain, swelling, tension, and skin alterations).17,26 In our series, a long-term venous patency of 54.9% was observed with predominantly ePTFE grafts. Because of rapid formation of collateral drainage, an occlusion of a venous bypass occurring some weeks after surgery usually did not require operative revision.26 The use of the autologous vein has not proven to be superior to synthetic grafts with respect to the long-term patency rate after limb-salvage surgery.17,18,21 The results with respect to limb salvage were excellent in this study. Comparably high salvage rates have been reported before, ranging from 80% to 100% (Table IV).17,23,26 Surviving patients at the latest follow-up in our study gave a good overall assessment of limb function and showed good results after being classified by the Enneking score.33 None of these patients would have preferred primary amputation instead of limb-preserving resection with vascular replacement. Furthermore, all of these patients

6/9 (67%)**

12/12 (100%)** 57%*

40%*

were satisfied with the functional result. Compared with the studies available, we report the important criterion of functionality after limb-preserving resection with vascular replacement.17-27 However, we do acknowledge that limbfunction assessment was based on a small number of patients. During observation, the overall local tumor control was high in this series (86%). Comparably high rates of local tumor control, ranging from 80% to 100%, were reported in the few studies that consisted of more than five patients.18,19,23-26 The use of adjuvant radiation therapy in patients with limbsparing surgery for extremity STS has been widely accepted.4,5 We selectively apply combined intraoperative radiotherapy with postoperative boost radiotherapy for extremity STS in high-risk patients or deliver postoperative radiotherapy.31,32 Data from our observational analysis and data from the literature show that intraoperative radiotherapy improves local tumor control.31,32 The high local control after limb-sparing surgery combined with adjuvant local treatment and the re-

JOURNAL OF VASCULAR SURGERY July 2005

96 Schwarzbach et al

constructive efficacy of vascular surgery explain the excellent limb-salvage rates in extremity STS involving major blood vessels (Table IV). The goals of limb salvage and reasonable functional results must be evaluated in the light of the high rate of perioperative morbidity, which seems to be related specifically to these procedures.17,19,20,22 Hematoma, wound infection, skin and soft tissue necrosis, and lymphatic fistulas are the major causes of a complicated postoperative course.17,19,20,22 Cautious interpretation, however, is mandatory because en bloc resection alone without vascular repair or intraoperative adjuvant treatment modalities has shown a wound morbidity rate of 34.7%.40,41 Here, a long procedural exposure time, the use of prosthetic material, and manipulation as a result of adjuvant intraoperative treatment (radiotherapy, brachytherapy, or regional chemotherapy) are possible risk factors for the development of complications. The question of whether synthetic material influences morbidity cannot be answered because of the small number of patients studied so far. Nevertheless, aseptic procedural standards, the correct placement of suction drains, and prophylactic administration of antibiotics (cephalosporin) are recommended. Careful hemostasis and cautious administration of heparin are mandatory to reduce bleeding complications. Because patients do not experience peripheral arterial disease and usually have an excellent peripheral runoff, low-dose heparinization is sufficient for vessel replacement at the iliofemoral level.28 Concerning patient survival, this study shows that a stage-dependent long-term survival can be achieved. The 2- and 5-year survival rates of 74.1% and 54.7% in this selective high-risk group of STS patients in this series are respectable and comparable to the results of Hohenberger et al26 and Karakousis et al.23 With respect to the survival rates, most patients in these studies had at least three adverse prognostic factors as diagnosed from the local tumor growth (higher tumor grade [intermediate grade and high grade tumors], subfascial location, and expansive growth).2,6-8,23,26,29 It has now been shown that histopathologically proven infiltrative growth is a negative predictor for survival at univariate and multivariate analysis. This pivotal issue calls for further clinical investigation considering the high risk for distant recurrence and subsequent tumor-related death. Nevertheless, even in the light of the high risk for distant recurrence, whether adjuvant chemotherapy should be given remains an individual and interdisciplinary decision.42 In this study, we demonstrated the clinical significance of surgical resection in patients with extremity STS involving major blood vessels. A novel classification system and a treatment algorithm have been described. Vascular replacement enables a limb-sparing resection with excellent local tumor control, acceptable limb function, and stage-dependent oncologic long-term results. We thank Sherryl Sundell for her help preparing this manuscript.

REFERENCES 1. Enneking WF. Musculoskeletal tumor surgery. New York: Churchill Livingstone; 1983. 2. Enneking WF, Spanier SS, Goodman MA. A system for surgical staging of musculoskeletal sarcoma. 1980. Clin Orthop Relat Res 2003;415:4-18. 3. Williard WC, Hajdu SI, Casper ES, Brennan MF. Comparison of amputation with limb-sparing operations for adult soft tissue sarcoma of the extremity. Ann Surg 1992;215:269-75. 4. Suit HD, Mankin HJ, Wood WC, Proppe KH. Preoperative, intraoperative, and postoperative radiation in the treatment of primary soft tissue sarcoma. Cancer 1985;55:2659-67. 5. Karakousis CP, Emrich LJ, Rao U, Khalil M. Limb salvage in soft tissue sarcomas with selective combination of modalities. Eur J Surg Oncol 1991;17:71-80. 6. Herbert SH, Corn BW, Solin LJ, Lanciano RM, Schultz DJ, McKenna WG, Coia LR. Limb-preserving treatment for soft tissue sarcomas of the extremities. The significance of surgical margins. Cancer 1993;72: 1230-8. 7. Stojadinovic A, Leung D, Hoos A, Jaques DP, Lewis JJ, Brennan MF. Analysis of the prognostic significance of microscopic margins in 2,084 localized primary adult soft tissue sarcomas. Ann Surg 2002;235:42434. 8. Eilber FC, Rosen G, Nelson SD, Selch M, Dorey F, Eckardt J, Eilber FR. High-grade extremity soft tissue sarcomas. Factors predictive of local recurrence and its effect on morbidity and mortality. Ann Surg 2003;237:218-26. 9. Caldarelli GF, Barellini L, Faviana P, Guerra M. Leiomyosarcoma of the popliteal artery: a case report and a review of the literature. J Vasc Surg 2003;37:206-9. 10. Gross E, Horton MA. Leiomyosarcoma of the saphenous vein. J Pathol 1975;116:37-41. 11. Charette S, Nehler MR, Whitehill TA, Gibbs P, Foulk D, Krupski WC. Epithelioid hemangioendothelioma of the common femoral vein: case report and review of the literature. J Vasc Surg 2001;33:1100-3. 12. Haug W, Losli EJ. Primary leiomyosarcoma within the femoral vein. Cancer 1954;7:159-62. 13. Gonzalez EB. Leiomyosarcoma of iliac and femoral veins. Arch Surg 1965;91:923-4. 14. Kim GE, Kim do K, Park IJ, Ro JY. Mesenchymal chondrosarcoma originating from the femoral vein. J Vasc Surg 2003;37:202-5. 15. Dzsinich C, Gloviczki P, van Heerden JA, Nagorney DM, Pairolero PC, Johnson CM, et al. Primary venous leiomyosarcoma: a rare but lethal disease. J Vasc Surg 1993;17:447-8. 16. Kogon B, Kabeer M, Sawchuk AP, Dalsing M, Billings S. Angiosarcoma presenting as an occluded popliteal artery pseudoaneurysm. J Vasc Surg 1998;27:970-3. 17. Fortner JG, Kim DK, Shiu MH. Limb-preserving vascular surgery for malignant tumors of the extremities. Arch Surg 1977;112:391-4. 18. Imparato A, Roses D, Francis KC, Lewis MM. Major vascular reconstructions for limb salvage in patients with soft tissue and skeletal sarcomas of the extremities. Surg Gynecol Obstet 1978;147:891-6. 19. Nambisan RN, Karakousis CP. Vascular reconstruction for limb salvage in soft tissue sarcomas. Surgery 1987;101:668-77. 20. Ohara T, Fukushima K, Hasegawa T, Ohki I. Reconstruction of femoral vein using ePTFE grafts. Nippon Geka Gakkai Zasshi 1991;92:753-6. 21. Wuisman P, Grunert J. Blood vessel transfer allowing avoidance of surgical rotation or amputation in the management of primary malignant tumors of the knee. Rev Chir Orthop Reparatrice Appar Mot 1994;80:720-7. 22. Koperna T, Teleky B, Vogl S, Windhager R, Kainberger F, Schatz K-D, Kotz R, Polterauer P. Vascular reconstruction for limb salvage in sarcoma of the lower extremity. Arch Surg 1996;131:1103-7. 23. Karakousis CP, Karmpaliotis C, Driscoll DL. Major vessels resection during limb-preserving surgery for soft tissue sarcoma. World J Surg 1996;20:345-50. 24. Kawai A, Hashizume HV, Inoue H, Uchida H, Sano S. Vascular reconstruction in limb salvage operation for soft tissue tumors of the extremities. Clin Orthop 1996;332:215-22.

JOURNAL OF VASCULAR SURGERY Volume 42, Number 1

25. Reix T, Sevestri-Pietri MA, Szychta P, Pietri J. Primary malignant tumors of the venous system in the lower extremities. Ann Vasc Surg 1998;12:589-96. 26. Hohenberger P, Allenberg JR, Schlag PM, Reichardt P. Results of surgery and multimodal therapy for patients with soft tissue sarcoma invading to vascular structures. Cancer 1999;85:396-408. 27. Bonardelli S, Nadari F, Maffeis R, Ippolito V, Saccalani M, Lussardi L, et al. Limb salvage in lower-extremity sarcomas and technical details about vascular reconstruction. J Orthop Sci 2000;5:555-60. 28. Ceraldi CM, Wang T-N, O’Donnell RJ, McDonald PT, Granelli SG. Vascular reconstruction in the resection of soft tissue sarcoma. Perspect Vasc Surg 2000;12:67-83. 29. Sobin LH, Wittekind CH, editors. TNM: classification of malignant tumours. 6th ed. New York: Wiley-Liss; 2002. p. 115-8. 30. Trojani M, Contesso G, Coindre JM, Rousse J, Bui NB, de Mascarel A, Goussot JF, David M, Bonichon F, Lagarde C. Soft tissue sarcomas of adults. Study of pathological prognostic variables and definition of a histological grading system. Int J Cancer 1984;33:37-42. 31. Eble MJ, Quentmeier A, Ewerbeck V, Herfarth C, Wannenmacher M. Methodology, technical prerequisites and postoperative morbidity of intraoperative radiotherapy (IORT) of soft tissue sarcomas. Heidelberg Krankengut 6/91-9/92. Radiologe 1993;33:513-9. 32. Van Kampen M, Eble MJ, Lehnert T, Bernd L, Jensen K, Hensley F, et al. Correlation of intraoperatively irradiated volume and fibrosis in patients with soft-tissue sarcoma of the extremities. Int J Radiat Oncol Biol Phys 2001;51:94-7. 33. Enneking WF, Dunham W, Gebhardt MC, Malawar M, Pritchard DJ. A system for the functional evaluation of reconstructive procedures after

Schwarzbach et al 97

34. 35. 36. 37.

38.

39.

40.

41.

42.

surgical treatment of tumors of the musculoskeletal system. Clin Orthop Relat Res 1993;286:241-6. Kaplan EL, Meier P. Non-parametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457-81. Cox DR. Regression models and life tables (with discussion). J R Stat Soc B 1972;34:187-220. Lawrence W Jr. Concepts in limb-sparing treatment of adult soft tissue sarcomas. Semin Surg Oncol 1988;4:73-7. Potter DA, Kinsella T, Glatstein E, Wesley R, White DE, Seipp CA, Chang AE, Lack EE, Costa J, Rosenberg SA. High-grade soft tissue sarcoma in the extremities. Cancer 1986;58:190-205. Panicek D, Go SD, Healey JH. Soft-tissue sarcoma invading bone or neurovascular structures: MR imaging prognostic factors. Radiology 1997;205:871-5. Schwarzbach MH, Dimitrakopoulou-Strauss A, Willeke F, Hinz U, Strauss LG, Zhang YM, et al. Clinical value of [18-F] fluorodeoxyglucose positron emission tomography imaging in soft tissue sarcomas. Ann Surg 2000;231:380-6. Arbeit JM, Hilaris BS, Brennan MF. Wound complication in the multimodality treatment of extremity and superficial truncal sarcomas. J Clin Oncol 1987;5:480-8. Skibber JM, Lotze MT, Seipp CA, Salcedo R, Rosenberg SA. Limbsparing surgery for soft tissue sarcomas: wound related morbidity in patients undergoing wide local excision. Surgery 1987;102:447-52 Adjuvant chemotherapy for localised resectable soft-tissue sarcoma of adults: meta-analysis of individual data. Sarcoma Meta-analysis Collaboration. Lancet 1997;350:1647-54.

Submitted Dec 2, 2004; accepted Mar 7, 2005.