Outcome of inferior vena cava and noncaval venous leiomyosarcomas

ARTICLE IN PRESS

Outcome of inferior vena cava and noncaval venous leiomyosarcomas Giulio Illuminati, MD,a Giulia Pizzardi, MD,a Francesco Calio’, MD,b Maria Antonietta Pacil e, MD,a Federica Masci, MD,a and Francesco Vietri, MD,a Rome and Catanzaro Italy

Background. Leiomyosarcoma (LMS) is a rare tumor arising from the smooth muscle cells of arteries and veins. LMS may affect both the inferior vena cava (IVC) and non-IVC veins. Because of its rarity, the experience with the outcome of the disease originating from the IVC compared with that with non-IVC offspring is overall limited. In this study, we compared the clinical features and outcomes after operative resection of IVC and non-IVC LMS to detect possible significant differences that could affect treatment and prognosis. Methods. Twenty-seven patients undergoing operative resection of a venous LMS at a single tertiary care center and one secondary care hospital were reviewed retrospectively and divided into 2 groups: IVC-LMS (Group A, n = 18) and non-IVC LMS (Group B, n = 9). As primary end points, postoperative mortality and morbidity, disease-specific survival and, if applicable, patency of venous reconstruction were considered. Bivariate differences were compared with the v2 test. Disease-specific survival was expressed by a lifetable analysis and compared using the log-rank test. Results. No postoperative mortality was observed in either group. Postoperative morbidity was 28% in group A and 11% in group B (P = .33). The mean duration of follow-up was 60 months (range, 13– 140). Disease-specific survival was 60% in group A and 75% in group B at 3 years (P = .48), and it was 54% in group A and 62% in group B at 5 years (P = .63). Seven grafts were occluded in group A (39%) and 1of 3 were occluded in group B (33%) (P = .85). Conclusion. IVC and non-IVC LMS exhibit similar outcomes in terms of postoperative course and survival. Operative resection associated with vascular reconstruction, if applicable, eventually followed by radiation and chemotherapy may be curative and is associated with good functional results. (Surgery 2015;j:j-j.) From the ‘‘F. Durante’’ Department of Surgical Sciences,a the University of Rome ‘‘La Sapienza,’’ Rome; and the Unit of Vascular Surgery,b Sant’Anna Hospital, Catanzaro, Italy

LEIOMYOSARCOMA (LMS) IS A RARE TUMOR OF MESENCHYMAL ORIGIN, representing 5–7% of all soft-tissue sarcomas with an origin from large vessels in 2% of the cases.1 No predisposing risk factors of the development of LMS of vascular smooth muscle have been clearly identified. Venous LMS, derived from the smooth muscle cells of the vessel wall, affects the inferior vena cava (IVC) in 60% of the cases.2 LMS of the arterial smooth muscles is rarer than LMS of the venous system, is found more frequently in men than women, and usually exhibits a more aggressive clinical pattern.3 Overall, venous LMS is considered a rare and often lethal

Presented at the Annual Meeting of the American Association of Endocrine Surgeons on May 17–19, 2015, Nashville, TN. Accepted for publication August 26, 2015. Reprint requests: Giulio Illuminati, Via Vincenzo Bellini 14, 00198 Rome, Italy. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.surg.2015.08.026

disease2; nonetheless, curative operative resection, regardless of whether it is associated with venous reconstruction, may lead to prolonged survival and good functional outcome, both with IVC and non-IVC LMS.4-9 A single institution’s experience with IVC LMS is explained thoroughly in several studies,2,4-8,10,11 whereas experience with non-IVC LMS is lacking because of the greater rarity of this localization of the disease. Available information on non-IVC LMS mostly relies on single case reports or small case series.1,12 We reviewed our results with the treatment of both IVC and non-IVC LMS to define eventual differences in their clinical outcomes and to assess the best operative treatment, validity of venous reconstruction, and ultimate patient prognosis. PATIENTS AND METHODS From October 1, 1993, to December 31, 2014, 27 consecutive patients underwent operative resection of a primary venous LMS at the ‘‘Francesco Durante’’ Department of Surgery, ‘‘La Sapienza’’ University Hospital of Rome, Italy, and one SURGERY 1

ARTICLE IN PRESS 2 Illuminati et al

affiliated center by 2 surgeons performing both general and vascular surgery. The study was approved by the institutional ethics committee, and informed consent was obtained from all the patients. Eighteen patients (Group A) presented with a primary LMS of the infrahepatic IVC. This sample represents 56% of 32 retroperitoneal tumors, including 2 angiosarcomas of the visceral aorta, treated at both centers in the same period. Nine patients (Group B) underwent resection of a primary, non-inferior vena cava (non-IVC) LMS. In this group, the tumor arose from the superficial/ common femoral vein in 3 patients, from the great saphenous vein in 3, from the renal vein in 2, and from the innominate vein in 1. Demographic characteristics were similar between the groups, with a prevalence of female patients (67% in both groups, P = 1) and a mean age of 51 years (range, 32–77) in group A and 57 years (range, 39–78) in group B (P = .41). In group A, the presenting symptoms and signs were abdominal pain in 15 patients (83%), a palpable abdominal mass in 12 (67%), edema of the lower limbs in 9 (50%), and abdominal distention in 5 (28%). In group B, the symptoms and signs were related to the primary location of the tumor, consisting of edema of the lower limb in 3 patients (superficial femoral vein), a growing mass at the thigh in 3 (great saphenous vein), abdominal discomfort and mass in 2 (renal vein), and left upper limb swelling in 1 (innominate vein). The patients’ preoperative workup included ultrasonography and computed tomography (CT) scan or magnetic resonance imaging in all the patients of both groups. In particular cases, positron emission tomography scanning was obtained to confirm the neoplastic origin of the venous thrombosis. No patient presented with evident metastatic disease at the time of surgery in each group. In group A, the tumor involved the infrarenal portion of the IVC in 14 patients (78%) and the interrenal or suprarenal portion in 4 (22%). In this group, the growth pattern of the sarcoma was mainly extraluminal in all the cases, with also a partial intraluminal extension and a partially patent IVC at the time of diagnosis. In group B, the pattern of growth was almost exclusively intraluminal in 1 case (innominate vein) and both intra- and extraluminal in all the other cases (superficial/common femoral, great saphenous and renal vein) (Fig 1). Because incomplete resection may be worse than no resection at all, the extent of resection was anticipated and planned with the preoperative CT scan or magnetic resonance imaging. The first technical step was to gain proximal and distal

Surgery j 2015

venous control in a zone clear from the tumor, followed by systemic heparinization (nonreversed at the end of the operation), en-bloc resection of the tumor and affected venous segment, and venous reconstruction if indicated. In doubtful cases, we did not hesitate to widely extend resection depending on the affected venous segment in the retroperitoneal fat and muscles (such as the psoas muscle), thigh muscles, such as the sartorius, other adductors, and quadriceps muscle. A perioperative frozen section examination of the resection margins was never performed in this study. In group A, all the patients underwent complete gross resection of the tumor en bloc with the involved segment of IVC, with curative intention. Reconstruction of the IVC was performed in all the patients by the use of a polytetrafluoroethylene (PTFE) graft (W.L. Gore and Associates, Flagstaff, AZ), ranging from 14 to 18 mm in diameter. The venous reconstruction was cavocaval in 16 patients and cavobiliac in 2. A right nephrectomy was associated in 2 patients. In 4 patients, the left renal vein was reimplanted in the graft. An adjunct arteriovenous fistula was not performed in any case. In group B, treatment consisted of resection of the mass en bloc with the involved venous segment. Resection included an associated nephrectomy in both the patients with the tumor arising from the renal vein. Venous reconstruction was performed only in the 3 patients in whom the tumor involved the superficial/common femoral vein and consisted of a femoro-femoral 8-mm PTFE graft without associated arteriovenous fistula. Histologic analysis revealed a LMS with negative tumor margins (>1 cm) in the resected specimens in all the patients of both groups, except in one patient of group B who had a LMS of the superficial femoral vein with microscopically involved resection margins. Postoperatively, all the patients undergoing venous reconstruction received low molecular weight heparin for 6 weeks and were subsequently put on 100 mg/day oral aspirin. Warfarin-based, oral anticoagulation was prescribed only in case of graft thrombosis associated with lower-limb edema. After surgery, all the patients were referred to the oncology department for eventual adjuvant radiation therapy and chemotherapy. In group A, no patient received adjuvant radiation therapy, whereas 7 patients with recurrent disease received chemotherapy consisting of doxorubicin at a dose ranging from 50 to 70 mg/m2 per cycle (400–420 mg/m2 total). In group B, 6 patients with an LMS of the thigh (superficial femoral vein and great saphenous vein),

ARTICLE IN PRESS Surgery Volume j, Number j

Illuminati et al 3

Fig 1. CT scan of the thigh showing a LMS of the left common-superficial femoral vein, with intra and extraluminal growth (A, arrow), displacing the superficial and profunda femoris arteries (B, arrow).

in which one had residual disease, received a total of 60 Gy adjuvant radiation therapy over 6 weeks, whereas 3 patients with subsequent metastatic disease also received chemotherapy (doxorubicin, 50–75 mg/m2 per cycle, 400–450 mg/m total). A possible treatment algorithm for both IVC and non-IVC LMS is proposed in Fig 2. All the patients were reviewed on an outpatient basis every 6 months with a CT scan of the chest, abdomen, and pelvis and an ultrasonographic scan of the venous grafts during the first 2 postoperative years and then on a yearly basis or earlier if necessary. As primary end points of the study, postoperative mortality and morbidity, diseasespecific survival and, if applicable, graft patency, were considered. Disease-specific survival was expressed by a life-table analysis13 and was defined as patients’ survival minus any deaths attributable to recurrent disease. Graft patency was defined as the absence of graft thrombosis on CT scan or duplex ultrasonography. Differences between the 2 groups were compared with the v2 test for bivariate analysis. Life-table curves were compared with the log-rank test.

RESULTS No patient died during the postoperative period in either of the 2 groups. In group A, the incidence of nonfatal complications was 28% and included 2 dehiscences of the abdominal wound, one myocardial ischemia, 1 acute renal insufficiency, and 1 respiratory distress syndrome. In group B, 1 postoperative complication was observed (11%), consisting of a dehiscence of the operative incision at the groin with delayed healing in a patient undergoing resection/PTFE grafting of a LMS of the superficial femoral artery. Differences in postoperative outcomes were not statistically significant (P = .63). The mean length of follow-up was 60 months (range, 13–140). In group A, 7 patients (39%) died of recurrent disease at a mean interval of 21 months after operation (range, 13–29). In group B, 2 patients (22%) died of recurrent disease at19 and 26 months after operation. The mean disease-free interval, which was not significantly different between the 2 groups, was 11 months (range, 8–17). Overall, cumulative survival in both groups was 46% in group A and 62% in group B (P = .49). Disease-specific survival was 60% in group

ARTICLE IN PRESS 4 Illuminati et al

Surgery j 2015

Fig 2. Possible treatment algorithm for both (IVC) (A) and (non-IVC) (B) venous LMS.

A and 75% in group B at 3 years (P = .48), and it was 54% in group A and 62% in group B at 5 years (P = .63) (Fig 3). One patient in group B underwent re-resection of residual disease 4 months after resection/PTFE grafting of an LMS of the superficial/common femoral vein. He was free from disease with a patent graft at the 18-month follow-up. No correlation could be found in either group between size and histological grade of LMS and likelihood of metastatic disease.

During the follow-up, 7 graft occlusions were observed in group A (39%), and 1 of 3 occlusions was observed in group B (33%) (P = .85). They were all associated with recurrent edema of the lower limbs. In group A, the diameter of the thrombosed grafts was 16 mm in 4 cases, 18 mm in 2, and 14 mm in 1. Of the 7 patients with graft thrombosis, 4 had metastatic disease, and 2 had a cavobiliac graft configuration. The patient whose graft thrombosed in group B was free from disease

ARTICLE IN PRESS Surgery Volume j, Number j

Illuminati et al 5

Fig 3. Life-table analysis of disease-specific survival after resection of IVC and non-IVC LMS showing no significant difference at both 3 (P = .48) and 5 years (P = .63). The numbers above the dotted line indicate the number of patients at risk in each interval.

at 65 months of follow-up. Essential clinical data and outcomes of patients in both groups are listed in Tables I and II. DISCUSSION This study shows that the results of resection of the IVC and non-IVC LMS in terms of postoperative outcome, cumulative survival, and, if applicable, patency of venous reconstruction are similar. This pattern can be related to the similar biologic behavior of the LMS, irrespective to its primary site of origin.14 Overall, the cumulative survival of 54% in group A and 62% in group B at 5 years is similar to other reports on caval and non-caval venous LMS.4-8,10,15-19 Limited to non-IVC tumors, our results also are consistent with those of other larger studies reporting a survival of 25% at 5 years and 44% of patients harboring metastatic disease at the time of first presentation.12 Overall, in our study, the average survival of both IVC and non-IVC LMS was better than expected and also a few, long-term, disease-free survivors were observed in both groups, as outlined in previous reports.4,9,20 A possible explanation for this trend may be because, especially for non-IVC LMS, cases or short studies reported in the literature may have included other sarcomas arising from the venous wall under the term LMS, such as angiosarcoma, fibromyosarcoma, and hemangiopericytoma, which may have more aggressive behavior compared with that of typical LMS. Currently, the best predictors of long-term survival are curative operative resection with R0 margins and the absence of metastatic disease at the time of surgery. Aggressive operative resection with venous reconstruction, if feasible, is therefore

indicated in any case without synchronous metastases at the time of referral for the following 2 reasons: (1) it remains the only chance of cure and (2) it allows excellent local control of the disease, provided that R0 margins are anticipated at preoperative CT scan and can actually be obtained at operation. Cytoreductive operations with predictable R1 margins are strongly contraindicated, as for all retroperitoneal sarcomas, because they may expose the patients to faster local growth of the residual tumor with a very poor quality of residual life. LMS are tumors deemed to be poor responders to both radiation and chemotherapy.1,2,6,21-23 In this study, chemotherapy was offered to patients of both groups limited to the metastatic stage of the disease. Adjuvant radiation therapy of the primary site of the disease was offered only to patients with LMS of the lower limbs, both after apparently curative surgical resection or in cases of residual disease. The issue of whether radiation therapy and chemotherapy were effective in prolonging survival cannot be addressed based on our data. In one patient with residual LMS of the superficial femoral vein, radiation therapy probably contributed to allow re-resection, limb preservation, and good quality of life for over a year. On the basis of our empirical experience, we would suggest adjuvant radiation therapy for all LMS of the limbs and chemotherapy for metastatic disease of any primary origin. Radiation therapy for LMS of the IVC should be avoided due to the limited responsiveness of the tumor and its potential harmful side effects, including radiation enteritis. Other similar features of the disease, in both the study groups, were its prevalence in the female

ARTICLE IN PRESS 6 Illuminati et al

Surgery j 2015

Table I. Clinical data of patients with leiomyosarcoma of the inferior vena cava (Group A) n

Sex

Age

1

F

40

2

F

42

3

F

58

4

F

67

5

F

6

Caval reconstruction

Radiation therapy

Chemotherapy

Follow-up, mo

Cavocaval

No

No

122

Cavocaval, reimplantation of left renal vein Cavocaval

No

No

112

No

No

111

No

No

101

67

Cavocaval, reimplantation of left renal vein Cavocaval

No

No

92

M

55

Cavobiliac

No

No

92

7

F

40

No

No

67

8

M

50

Cavocaval, right nephrectomy, reimplantation of left renal vein Cavocaval

No

No

48

9

M

44

Cavocaval

No

No

43

10

M

74

Cavocaval

No

No

40

11

M

43

Cavocaval

No

Yes

29

12

M

49

No

Yes

27

13

F

56

Cavocaval, right nephrectomy, reimplantation of left renal vein Cavocaval

No

Yes

26

14

F

32

Cavocaval

No

No

25

15

F

46

Cavocaval

No

Yes

20

16

F

77

Cavocaval

No

Yes

19

17

F

56

Cavobiliac

No

Yes

19

18

F

38

Cavocaval

No

Yes

16

Outcome Lost at follow-up, free from disease, graft thrombosed Alive, free from disease, patent graft Lost at follow-up, free from disease, patent graft Lost at follow-up, free from disease, patent graft Alive, free from disease, graft thrombosed Lost at follow-up, free from disease, graft thrombosed Died of stroke, free from disease, patent graft Lost at follow-up, free from disease, patent graft Alive, free from disease, patent graft Died, myocardial infarction, free from disease, patent graft Died, pulmonary metastases, patent graft Died, nonresected, local recurrence, graft thrombosed Died, pulmonary and hepatic metastases, graft thrombosed Alive, free from disease, patent graft Died, pulmonary metastases, patent graft Died, pulmonary metastases, graft thrombosed Died, hepatic metastases, graft thrombosed Died pulmonary metastases, patent graft

F, female; M, male.

gender in the 5th decade of age and the overall low incidence of postoperative morbidity, as observed in other reports.4,5,7,24 The results of this study also support the value of venous reconstruction after resection, if applicable. This evidence is apparent in group A, as a 60% patency of caval reconstructions ensured a low rate lower limb edema. The appearance of edema in the 8 patients experiencing graft thrombosis and the need for ligation and disruption of venous collaterals during resection of large IVC tumors suggests

that caval reconstruction should be considered, if feasible.4,8,25 The rationale for using PTFE as a caval substitute in the setting of resection for LMS is because operations are conducted in a noninfected field, thus not requiring autogenous material; PTFE is readily available in various sizes, thus significantly reducing the possibility of a mismatch in anastomotic sites, compared with the spiral saphenous vein and superficial femoral vein; it also has the advantage of being more resistant than autogenous venous conduits to

ARTICLE IN PRESS Illuminati et al 7

Surgery Volume j, Number j

Table II. Clinical data of patients with noncaval venous leiomyosarcoma (Group B) Radiation therapy

Chemotherapy

Follow-up, mo

Resection

Es

No

140

Great sephenous

Resection

Yes

No

132

50

Innominate

Resection

No

No

108

F

71

Great saphenous

Resection

Yes

No

68

5

F

42

Superficial femoral

Resection, PTFE grafting

Yes

No

65

6

M

57

Superficial femoral

Resection, PTFE grafting

Yes

Yes

41

7

F

64

Renal

Nephrectomy

No

Yes

26

8

M

68

Renal

Nephrectomy

No

Yes

19

9

M

78

Superficial femoral

Resection, PTFE grafting

Yes

No

18

n

Sex

Age

Venous location

1

F

39

Great saphenous

2

F

42

3

F

4

Surgery

Outcome Alive, free from disease Alive, free from disease Alive, free from disease Alive, free from disease Alive, free from disease, graft thrombosed Died, myocardial infarction, patent graft Died, pulmonary metastases Died, pulmonary metastases Alive, residual disease/reresection, free from disease, patent graft

F, female; M, male; PTFE, polytetrafluoroethylene.

compression by overlying abdominal viscera.26 The greater saphenous vein for replacement of the superficial/common femoral vein is not recommended in this setting, as it would eliminate a source of collaterals in case of thrombosis. Graft diameter was not correlated with thrombosis; however, the sample size in this study does not allow significant conclusions. Considering that 4 of 7 patients with IVC graft thrombosis also presented with metastatic disease, it can be speculated that metastatic disease may contribute to graft thrombosis by inducing a para-neoplastic hypercoagulable state. This speculation does not apply to group B, in which the single graft thrombosis occurred in a patient free from disease. The benefit of venous reconstruction is less evident for non-IVC tumors, for which this option is applicable. In our study, 2 of 3 patients undergoing reconstruction of the superficial/common femoral vein in group B were free of edema with a patent graft at approximately one year after surgery despite the absence of an arteriovenous fistula. Although encouraging, these results cannot be considered conclusive. Our study was limited by the small sample size, especially in group B; however, the rarity of venous LMS does not allow single institutions to gather

large patient cohorts in a convenient time period to perform statistically valid comparisons. The same consideration applies to meta-analyses because most of the cases of non-IVC LMS in the literature consist of single case reports, which may not be an accurate representation of all actual cases of non-IVC LMS in existence.1 In addition to this weakness, this study’s strength is that it addresses a homogeneous series of patients treated and followed by the same surgeons, which allows us to objectively assess the results over time. The results of this study indicate that venous LMS tend to exhibit similar oncologic behavior irrespective to their primary site of origin. Although rare, long-term survivors with good functional outcome are observed in both IVC and non-IVC disease, thus supporting aggressive resection with venous reconstruction, if applicable, as the best treatment option currently available. More extensive experience is needed to confirm these conclusions and to better define the role of radiation and chemotherapy. REFERENCES 1. Gage MJ, Patel AV, Koenig KL, Newman E. Non-vena cava venous leiomyosarcomas: a review of the literature. Ann Surg Oncol 2012;19:3368-74.

ARTICLE IN PRESS 8 Illuminati et al

2. 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 1992;15: 595-603. 3. Plissonnier D, Kieffer E. Tumeurs primitives de l’aorte et des grosses arteres. In: Kieffer E, Godeau P, editors. Maladies art erielles non atherosclereuses de l’adulte. Paris: AERCV; 1994. p. 415-36. 4. Illuminati G, Calio’ FG, D’Urso A, Giacobbi D, Papaspyropoulos V, Ceccanei G. Prosthetic replacement of the infrahepatic inferior vena cava for leiomyosarcoma. Arch Surg 2006;141:919-24; discussion 924. 5. Kieffer E, Alaoui M, Piette JC, Cacoub P, Chiche L. Leiomyosarcoma of the inferior vena cava: experience in 22 cases. Ann Surg 2006;244:289-95. 6. Wachtel H, Jackson BM, Bartlett EK, Karakousis GK, Roses RE, Bavaria JE, et al. Resection of primary leiomyosarcoma of the inferior vena cava (IVC) with reconstruction: a case series and review of the literature. J Surg Oncol 2015; 111:328-33. 7. Hollenbeck ST, Grobmmyer SR, Kent KC, Brennan MF. Surgical treatment and outcomes of patients with primary inferior vena cava leiomyosarcoma. J Am Coll Surg 2003;197:575-9. 8. Hardwigsen J, Baque P, Crespy B, Moutardier V, Delpero JR, Le Treut YP. Resection of the inferior vena cava for neoplasms with or without prosthetic replacement: a 14patient series. Ann Surg 2001;233:242-9. 9. Illuminati G, Miraldi F, Mazzesi G, D’Urso A, Ceccanei G, Bezzi M. Primary leiomyosarcoma of the innominate vein. Ann Vasc Surg 2007;21:75-8. 10. Huguet C, Ferri M, Gavelli A. Resection of the suprarenal inferior vena cava: the role of prosthetic replacement. Arch Surg 1995;130:793-7. 11. Quinones-Baldrich W, Alktaifi A, Eilber F, Eilber F. Inferior vena cava resection and reconstruction for retroperitoneal tumor excision. J Vasc Surg 2012;55:1386-93; discussion 1393. 12. Abed R, Abudu A, Grimer RJ, Tillman RM, Carter SR, Jeys L. Leiomyosarcomas of vascular origin in the extremity. Sarcoma 2009;2009:385164. 13. Peto R, Pike MC, Armitage P, Breslow NE, Cox DR, Howard SV, et al. Design and analysis of randomized clinical trials requiring prolonged observation of each patient. II. Analysis and examples. Br J Cancer 1977;35:1-39.

Surgery j 2015

14. Worhunsky DJ, Gupta M, Gholami S, Tran TB, Ganjoo KN, Van den Rijn M, et al. Leiomyosarcoma: one disease or distinct biologic entities based on site of origin? J Surg Oncol 2015;111:808-12. 15. Gloviczki P, Pairolero PC, Cherry KJ, Hallett JW. Reconstruction of the vena cava and of its primary tributaries: a primary report. J Vasc Surg 1990;11:373-81. 16. Daylami R, Amiri A, Goldsmith B, Troppmann C, Schneider PD, Khatri VP. Inferior vena cava leiomyosarcoma: is reconstruction necessary after resection? J Am Coll Surg 2010;210:185-90. 17. Mann GN, Mann LV, Levine EA, Shen P. Primary leiomyosarcoma of the inferior vena cava: a 2-institution analysis of outcomes. Surgery 2012;15:261-7. 18. Fu TY, Hsieh PP, Chen LW, Tseng HH, Wang JS. Leiomyosarcoma of the cephalic vein: case report and review of the literature. Ann Vasc Surg 2007;21:508-11. 19. Jefferson KP, Dixon JH. Leiomyosarcoma of the cephalic vein. Sarcoma 2001;5:27-30. 20. Illuminati G, Miraldi F, A Pacile M, Palumbo P, Vietri F. Long-term survival after resection of a primary leiomyosarcoma of the innominate vein: report of a case. Ann Ital Chir 2012;2012. 21. Miller ED, Xu-Welliver M, Haglund KE. The role of modern radiation therapy in the management of extremity sarcomas. J Surg Oncol 2015;111:599-603. 22. De Amorim Bernstein K, Delaney TF. Role of radiation therapy for non-extremity soft tissue sarcomas. J Surg Oncol 2015;111:604-14. 23. Liebner DA. The indications and efficacy of conventional chemotherapy in primary and recurrent sarcoma. J Surg Oncol 2015;111:622-33. 24. Hines OJ, Nelson S, Quinones-Baldrich WJ, Eilber FR. Leiomyosarcoma of the inferior vena cava: prognosis and comparison with leiomyosarcomas of other anatomic sites. Cancer 1999;85:1077-83. 25. Sarkar R, Eilber FR, Gelabert HA, Quinones-Baldrich WJ. Prosthetic replacement of the inferior vena cava for malignancy. J Vasc Surg 1998;28:75-81; discussion 82-83. 26. Bower TC, Nagorney DM, Cherry KJ, Toomey BJ, Hallett JW, Panneton JM, et al. Replacement of the inferior vena cava for malignancy: an update. J Vasc Surg 2000;31:270-81.