Surgical management of primary retroperitoneal sarcoma

Transcription

1 Original article Surgical management of primary retroperitoneal sarcoma D. C. Strauss 1,A.J.Hayes 1,K.Thway 2,E.C.Moskovic 3,C.Fisher 2 and J. M. Thomas 1 1 Melanoma/Sarcoma Unit, Department of Surgery, and Departments of 2 Histopathology and 3 Radiology, Royal Marsden Hospital NHS Foundation Trust, London, UK Correspondence to: Mr D. C. Strauss, Academic Surgery, Royal Marsden Hospital NHS Foundation Trust, Fulham Road, London SW3 6JJ, UK ( Dirk.Strauss@rmh.nhs.uk) Background: Local recurrence after surgical resection is the main cause of disease-related mortality in patients with primary retroperitoneal sarcoma (RPS). This study analysed predictors of local recurrence and disease-specific survival. Methods: A prospective database was reviewed to identify patients who underwent surgery for primary RPS between 1990 and 09. Patient demographics, operative outcomes and tumour variables were correlated with local recurrence and disease-specific survival. Multivariable analysis was performed to evaluate predictors for local recurrence and disease-free survival. Results: Macroscopic clearance was achieved in 170 of 0 patients. The median weight of tumours was 4 0 kg and median maximum diameter 27 cm. Resection of adjacent organs was required in 126 patients. The postoperative mortality rate was 3 0 per cent. Seventy-five patients developed local recurrence during follow-up. At 5 years the local recurrence-free survival rate was 54 6 per cent and the diseasespecific survival rate 68 6 per cent. Inability to obtain macroscopic clearance at resection and high-grade tumours were significant predictors for local recurrence and disease-specific survival. Conclusion: Complete macroscopic excision should be the goal of surgical resection. Ability to resect a RPS completely and tumour grade are the most important predictors of local recurrence and overall survival. Paper accepted 10 December 09 Published online in Wiley InterScience ( DOI: 10.2/bjs.6994 Introduction Retroperitoneal sarcoma (RPS) is a rare tumour, accounting for approximately per cent of all soft tissue tumours 1. It constitutes a therapeutic challenge because of relatively late presentation and anatomical location, often close to vital structures in the retroperitoneal space. This close relationship to vital structures impacts on the ability to perform a radical wide resection. It is often not possible to obtain a margin of normal tissue around the tumour. Local recurrence is the main cause of failure, ranging from to per cent 2. Seventy-five per cent of sarcoma-related deaths involve uncontrolled local recurrence 2. Surgery plays a principal role in the management of RPS and provides the only opportunity for cure. No effective chemotherapy exists to influence survival in patients with RPS. Given that local failure remains the main cause of death after surgery in these patients, there is great interest in strategies that might improve local control 3,4. The role of radiotherapy in helping to achieve local control remains undefined with no prospective randomized controlled trials available to define indications, dose, route of administration or impact on overall survival. Initial attempts by the American College of Surgeons Oncology Group to investigate the role of preoperative radiotherapy in RPS in a randomized setting (trial ACOSOG Z9031) was closed early because of poor patient accrual. A European Organization for Research and Treatment of Cancer study of preoperative radiotherapy in RPS that will accrue patients only from centres seeing large numbers of patients is at the planning stage. Regarding surgery for RPS, two recent observational studies 5,6 investigated the role of liberal visceral en bloc resection in an attempt to include an envelope of normal tissue around the tumour in the hope of improving outcome. A number of criticisms, however, were raised in response to these two reports 7, based in part on Copyright 10 British Journal of Surgery Society Ltd British Journal of Surgery

2 D. C. Strauss, A. J. Hayes, K. Thway, E. C. Moskovic, C. Fisher and J. M. Thomas the difficulty in drawing conclusions from multicentre retrospective series and also questioning whether a selective approach to resection of structures adjacent to the tumour could ever succeed in improving the margin status unless all surrounding structures were resected. The present study describes a large series of patients undergoing surgery for primary RPS. The series is based on the experience of only two surgeons within one centre, with a consistent operative approach. In contrast to the reports cited above, the surgical approach involved obtaining complete clearance of all macroscopic disease, resection of contiguous organs or vascular structures that were macroscopically involved with the tumour, but preservation of macroscopically uninvolved organs even when they were closely apposed to the surface of the tumour. The aim of the study was to investigate predictors of local recurrence and disease-specific survival within the context of current surgical treatment in order to indicate baseline outcomes of current treatment in any future multimodality trials. Methods Information was obtained from a prospectively maintained database for all patients with soft tissue tumours treated at the Royal Marsden Hospital from Patients undergoing resection for primary RPS between March 1990 and March 09 were the subject of this study. Patient characteristics, operative outcomes and tumour variables were correlated with the local recurrence and disease-specific survival. The series included only patients presenting with primary RPS who underwent initial surgical resection at this hospital. Histology was reported and reviewed if necessary by a dedicated specialist sarcoma pathologist. Pathological variables included maximum dimension, weight, microscopic margins, histological subtype and tumour grade. Grade was determined using the French Federation of Cancer Centres Sarcoma Group Grading System (FNCLCC) 8. Microscopically clear margins (R0) were defined as no microscopic residual disease and tumour more than 0 1 mmfrommargin; microscopically involved margins (R1) were defined as macroscopically complete resection but with microscopic residual disease extending to or within 0 1 mmofthemargin. Macroscopically incomplete resection was defined as gross residual disease. Patients with recurrent RPS, gastrointestinal stromal tumours, uterine leiomyosarcoma, carcinosarcoma and metastatic retroperitoneal sarcoma were excluded from the study. Tumours with extensive encasement of vital vascular structures demonstrated on preoperative imaging where vascular resection and reconstruction was not an option (for example mesenteric vessels, portal vein) were deemed unresectable. Vascular encasement of major vessels that could be resected and reconstructed (inferior vena cava (IVC), iliac arteries) was not a contraindication to resection. Patients with major medical risks or poor general health that prohibited a major operation were not considered for surgical intervention. Organs were resected to obtain macroscopic clear margins when the contiguous organ appeared clinically involved or to facilitate safe dissection around the tumour. Clinically uninvolved but adjacent organs were not an indication for resection. Endpoints of the study were local recurrence and disease-specific survival. The interval to local recurrence was defined as the time from initial surgery to radiological confirmation of local recurrence. Disease-specific survival was measured from the time of the first operation for RPS to either sarcoma-related death or last followup. After surgical resection, a follow-up programme consisted of review at 3-month intervals for 1 year, at 6-month intervals for the next 4 years and then annually. Adjuvant or neoadjuvant chemotherapy or radiotherapy was not used routinely, but selectively in a small group of patients as deemed appropriate after discussion in a multidisciplinary meeting. This has been omitted from the outcome analysis. Patients were further stratified into three groups based on tumour subtype well differentiated liposarcoma/atypical lipomatous tumour (ALT), non- ALT liposarcoma (dedifferentiated and pleomorphic liposarcoma) and other histologies in order to assess whether a pathological subtype-based paradigm might predict outcome 9. Statistical analysis Univariable Cox regression was used to evaluate the following potential predictors of local recurrence-free survival: grade, macroscopic clearance, age, tumour size and tumour weight. Significant variables (P < 0 050) were combined using forward stepwise selection methods in a multivariable analysis. Univariable Cox regression was used to evaluate the following potential predictors of disease-specific survival: grade, macroscopic clearance, pathology, age, tumour size and tumour weight. Significant variables (P < 0 050) were combined using forward stepwise selection methods in a multivariable analysis. Five-year disease-specific survival and local recurrence rates were calculated and plotted using the Kaplan Meier method 10. Death from other causes was treated as a censored event for analysis of disease-specific survival. The statistical program SPSS version 17.0 (SPSS, Chicago, Illinois, USA) was used for analyses.

3 Surgical management of primary retroperitoneal sarcoma Results From March 1990 to March 09, 382 patients had resection of retroperitoneal tumours, of whom 0 had primary RPS. Of the remainder, 123 underwent resection for recurrent RPS, seven for other malignancies and 52 for benign tumours. In the RPS group, median follow-up was 29 (range 0 198) months. Patient characteristics, operative outcomes and tumour pathology variables are shown in Table 1. There were 114 men (57 0 per cent) and 86 women (43 0 per cent). Their median age was 56 (range 18 89) years. Macroscopic clearance was achieved in 170 (85 0 per cent) of the 0 patients. In half of the patients ( of 0) a preoperative biopsy was performed and a histological diagnosis was available. Seventy-six of these biopsies were performed before referral to the authors sarcoma unit. When the radiological characteristics of retroperitoneal liposarcoma were not in doubt, a preoperative biopsy was not undertaken. The median weight of primary tumours was 4 0 (range ) kg and the median maximum diameter 27 (range 6 ) cm. Table 1 Demographics, operative outcomes and pathology characteristics of 0 patients with primary retroperitoneal sarcoma Median (range) tumour diameter (cm) 27 (6 ) Tumour size distribution (cm) (2 0) (25 5) (54 0) 37 (18 5) Median (range) tumour weight (kg) 4 0 ( ) Tumour grade 1 93 (46 5) 2 81 ( 5) 3 26 (13 0) Tumour subtype Liposarcoma 152 (76 0) Well differentiated 84 (42 0) Dedifferentiated 57 (28 5) Myxoid 8 (4 0) Pleomorphic 3 (1 5) Leiomyosarcoma 27 (13 5) Solitary fibrous tumour 11 (5 5) Synovial sarcoma 3 (1 5) Fibrosarcoma 2 (1 0) Pleomorphic rhabdomyosarcoma 1 (0 5) Undifferentiated round cell sarcoma 1 (0 5) Malignant PNST 1 (0 5) Epithelioid smooth muscle tumour 1 (0 5) Desmoplastic small round-cell tumour 1 (0 5) Value in parentheses are percentages if not indicated otherwise. PNST, peripheral nerve sheath tumour. Resection of adjacent organs was required in 126 patients (63 0 per cent), with the ipsilateral kidney and colon the most common organs resected (Table 2). Sixty patients required resection of more than one organ. In 26 patients (13 0 per cent) the resected organ showed infiltration by the tumour. The kidney (nine patients), colon (seven), pancreas (three) and IVC (two) were the most common structures infiltrated, and the abdominal wall, diaphragm, rectum, uterus and stomach wall were each reported in one patient to be infiltrated by tumour. Median blood loss was 1 5 (range ) litres, and median length of intensive care unit stay was 1 (range 0 67) days. The postoperative mortality rate was 3 0 per cent (six patients). Three patients died from multisystem organ failure secondary to sepsis (two with proven anastomotic leaks), one patient had cardiac dysrhythmia and haematemesis, one a myocardial infarction, and one died from heart failure secondary to dysrhythmia. Postoperative complications that required reoperation occurred in 24 patients (12 0 per cent). The major surgical morbidity was bleeding, which required intraoperative packing in eight patients, and postoperative bleeding required relaparotomy in six patients. Seventy-five patients (37 5 per cent) developed local recurrence during follow-up. Median local recurrencefree survival was 3 8 years for the whole group of 0 patients and 6 8 years for the 170 patients who had a macroscopically complete resection. For the cohort of patients who had a macroscopically complete resection, the 2-year local recurrence-free survival rate was 74 6 per cent and the 5-year local recurrence-free survival rate was 54 6 per cent (Fig. 1a). High tumour grade and inability to obtain macroscopically complete clearance were the Table 2 Involved/adjacent organs resected and associated procedures No. of patients (n = 0) Kidney 71 (35 5) Colon 44 (22 0) Spleen 28 (14 0) Distal pancreas 17 (8 5) Small bowel 5 (2 5) Partial stomach 4 (2 0) Inferior vena cava 4 (2 0) Partial bladder 3 (1 5) Other* 17 (8 5) Abdominal wall plus mesh reconstruction 1 (0 5) Abdominal aortic aneurysm repair 1 (0 5) Values in parentheses are percentages. *Iliac artery (one), iliac vein (two), spermatic cord and testis (three), diaphragm (three), duodenum (two), ovary (two), uterus (two), gallbladder (two).

4 D. C. Strauss, A. J. Hayes, K. Thway, E. C. Moskovic, C. Fisher and J. M. Thomas Local recurrence-free survival (%) Local recurrence-free survival (%) Grade 1 Grade 2 Grade a Overall local recurrence-free survival Grade Grade Grade b Local recurrence-free survival according to grade Macroscopic clearance Incomplete resection Local recurrence-free survival (%) Macroscopic clearance Incomplete resection c Local recurrence-free survival according to completeness of resection Fig. 1 Local recurrence-free survival a in patients who had a macroscopically complete resection, b relative to tumour grade and c comparing macroscopically complete clearance with incomplete resection. b,c P < (log rank test)

5 Surgical management of primary retroperitoneal sarcoma Table 3 Univariable and multivariable analyses of risk factors for local recurrence and disease-specific survival after surgery Univariable analysis Multivariable analysis Hazard ratio P Hazard ratio P Local recurrence-free survival Tumour grade (2 2, 5 8) < (2 1, 5 5) < (2 7, 9 8) < (2 4, 9 0) < Macroscopic clearance Yes 1 1 No 2 0 (1 6, 2 5) < (2 3, 5 9) < Age (continuous; 1-year increase) 1 0 (1 0, 1 0) Age (categorical) Median (57 years) or less 1 Above median 1 2 (0 8, 1 8) Tumour size (cm) 30 1 > (0 5, 1 2) Tumour weight (kg) > (0 6, 1 4) Disease-specific survival Tumour grade (3 3, 16 6) < (3 3, 17 5) < (6 1, 8) < (6 5, 46 3) < Macroscopic clearance Yes 1 1 No 1 9 (1 4, 2 7) < (1 5, 5 8) Pathology ALT 1 Non-ALT liposarcoma 6 0 (2 7, 13 3) < Other 4 4 (1 9, 10 1) < Age (continuous; 1-year increase) 1 0 (0 99, 1 04) Age (categorical) Median (57 years) or less 1 Above median 1 2 (0 6, 2 1) Tumour size (cm) > (0 3, 1 0) (0 2, 1 0) Tumour weight (kg) 4 1 > (0 6, 2 2) Values in parentheses are 95 per cent confidence intervals. ALT, atypical lipomatous tumour. only factors found to be statistically significant predictors of local recurrence (Table 3 and Fig. 1b,c). Median disease-specific survival was 115 months. The disease-specific survival rate at 2 and 5 years was 86 2 and 68 6 per cent respectively (Fig. 2a). Tumour grade and the ability to obtain macroscopic clearance emerged in multivariable analysis as the most important factors influencing disease-specific survival (Table 3 and Fig. 2b,c). Microscopic margin status was available for 1 patients: margins were clear (R0) in 55 patients and involved (R1) in 85. Microscopic margin status did not influence local recurrence or survival (P = 0 109). Regarding a pathological subtype-based paradigm, on univariable analysis non-alt liposarcoma and other histologies were found to be predictors of survival (Table 3). After correcting for grade, macroscopic clearance and size, the pathological subtype did not reach statistical significance in multivariable analysis (P = 0 096). Seventeen patients were diagnosed with systemic metastasis during follow-up. Metastatic disease occurred exclusively in patients with grade 2 or 3 tumours. All systemic metastases presented within 33 6 months of surgery. Eighteen patients (9 0 per cent) received radiotherapy and 25 (12 5 per cent) had chemotherapy. At

6 D. C. Strauss, A. J. Hayes, K. Thway, E. C. Moskovic, C. Fisher and J. M. Thomas Disease-specific survival (%) Disease-specific survival (%) Grade 1 Grade 2 Grade a Overall disease-specific survival Grade Grade Grade b Disease-specific survival according to tumour grade Disease-specific survival (%) Macroscopic clearance Incomplete resection Disease-specific survival (%) ALT Non-ALT Other histology Macroscopic clearance Incomplete resection c Disease-specific survival according to completeness of resection 0 ALT 84 Non-ALT Other d Disease-specific survival according to pathology subtype Fig. 2 Disease-specific survival a in 0 patients who underwent resection of a primary retroperitoneal sarcoma, b relative to tumour grade, c relative to macroscopically complete resection and d according to pathology group. ALT, atypical lipomatous tumour; non-alt, dedifferentiated and pleomorphic liposarcomas. b d P < (log rank test)

7 Surgical management of primary retroperitoneal sarcoma Discussion a b Fig. 3 Computed tomograms showing a a retroperitoneal dedifferentiated liposarcoma in contact with the aorta, superior mesenteric artery, colon, pancreas and all surrounding musculature of the abdominal wall, and b the same liposarcoma in contact with the aorta, inferior vena cava, colon, stomach and all surrounding musculature of the abdominal wall. Macroscopically complete clearance necessitated nephrectomy, left hemicolectomy, splenectomy and distal pancreatectomy as these structures or their vascular supply were enveloped by tumour the end of the study 141 of the 0 patients were alive at last follow-up. Of late deaths, 45 were from sarcoma-related causes and eight from unrelated causes. A surgical approach with the aim of obtaining macroscopically complete clearance should form the basis of management in patients with RPS 2, The present study has confirmed that many retroperitoneal sarcomas are large at presentation. The median maximum diameter was 27 cm and median weight 4 0 kg. Surgical resection is often difficult owing to tumour bulk with displacement of surrounding structures. The sheer size and bulk of these tumours creates difficulty gaining access behind the tumour to identify retroperitoneal structures and to obtain vascular control of major vessels. The postoperative mortality rate in the present series was 3 0 per cent. This compares favourably with reported mortality rates of 2 7 per cent in other large series 2,5,9,13. In this study, low tumour grade and the ability to obtain macroscopic clearance were the only variables that were significantly associated with a decrease in local recurrence. Complete macroscopic clearance was obtained in 170 patients (85 0 per cent). In 126 patients (63 0 per cent), resection of contiguous viscera was required. Much has been written recently 5 7 advocating a policy of more liberal visceral en bloc resections to include an envelope of normal tissue around the tumour. Two recent European studies 5,6 claim improved local control through resection of uninvolved adjacent organs. However, Pisters 7 has raised criticism and limitations related to the methodology of both of these observational studies. Bonvalot and colleagues 5 reported on data from an unspecified number of institutions across France over a -year period. Gronchi and co-workers 6 compared a patient group in which liberal visceral en bloc resections were performed with a historical group that underwent surgery before the introduction of this more aggressive approach. Although the local recurrence rate was lower in the recent patient group that had more aggressive surgery, this outcome was confounded by a significantly shorter follow-up period for these patients and no overall survival benefit was achieved. Neither study standardized operative techniques or the processing and reporting of pathology specimens. This was further confounded by the use of preoperative or postoperative radiotherapy in 31 per cent 6 and 32 per cent 5 of patients. The concept of compartmentalization that applies to limb sarcomas is difficult to extrapolate directly to RPS. The proponents of a liberal approach to resection of abdominal viscera have argued that resection of uninvolved adjacent organs could impact upon outcome, as this will affect local relapse. The criticism with this argument is that, although it is clear that certain structures were liberally resected (the colon and psoas), other structures

8 D. C. Strauss, A. J. Hayes, K. Thway, E. C. Moskovic, C. Fisher and J. M. Thomas intimately associated with the tumour (for example the aorta or IVC) were not resected. Although this approach might influence the margin of normal tissue separating the tumour from the resection margin at certain points, it is not possible to perform a radical resection that could provide a wide margin in all directions. The close relationship between the tumour and major vessels (aorta, vena cava, portal vein, superior mesenteric artery), visceral organs (duodenum, pancreas, liver), the axial skeleton (vertebral bodies), pelvic bones, muscles (abdominal wall, diaphragm) and vitalneuralstructures (femoral nerve roots) impacts on the ability to perform radical wide resection with normal tissue surrounding all aspects of the tumour. In the present study, the surgical approach involved a low threshold for organ resection of contiguous structures that were clearly involved or enveloped by the tumour, in order to gain complete macroscopic clearance, but with no attempt to resect organs that merely lay adjacent to the tumour but were not involved. The distinction between these two approaches is shown in Fig. 3, which shows computed tomography slices of a retroperitoneal sarcoma. In order to gain macroscopically complete clearance, the resection mandated a nephrectomy, left hemicolectomy, splenectomy and distal pancreatectomy, because these structures or their vascular supply were enveloped by the tumour. It can be seen, however, that this tumour abuts many other structures (aorta, IVC, abdominal wall). The achievement of surgical margins that were clear circumferentially would require resection of the vena cava, diaphragm, stomach and surrounding musculature of the abdominal wall (quadratus lumborum and lateral abdominal wall). The authors unit s policy remains principally to obtain macroscopically complete clearance, with en bloc resection of organs only when directly involved by the tumour or organ resection to facilitate exposure, but with preservation of adjacent uninvolved organs. The overall disease-specific survival rate at 2 and 5 years was 86 2 and 68 6 per cent respectively, similar to the results of other series 2,9, Multivariable analysis of risk factors indicated that tumour grade and macroscopically complete clearance were the only significant predictors of disease-free survival. Size was not a predictor of survival, as in other large series 9,11,12. A histology-based paradigm has been proposed to stratify patients 9. The 0 patients in the present study were divided into three groups on this basis: well differentiated liposarcomas (also known as ALT), non-alt liposarcoma (dedifferentiated and pleomorphic liposarcomas) and other histologies. Multivariable analysis did not, however, show this to be a predictor of disease-specific survival. One possible reason for this was that pathology and tumour size were related based on cross-tabulation and the associated χ 2 test. The importance of microscopically clear margins (R0 resection) in RPS and how this relates to local recurrence has been investigated Microscopically clear margins do not reliably predict local control, but local control is closely determined by the ability to obtain macroscopic clearance. Whether macroscopic clearance has been obtained is ascertained largely by the operating surgeon at the end of the operation based on whether the tumour has been removed completely. The evaluation of microscopic margins for RPS is problematic as these large tumours have an extensive surface area, making it hard to assess all microscopic margins reliably. The ability to determine microscopically clear margins accurately depends on the methodology of pathological sampling 16. Microscopic margins can also be disturbed by tumour handling and retraction artefacts after resection, and do not necessarily indicate microscopic residual disease. A comparison between patients who have negative microscopic margins (R0) and those with involved microscopic margins (R1) showed no difference in local recurrence rates in the present study. The high incidence of local recurrence and its close relationship with sarcoma-related death has prompted investigation of combined-modality treatment in an attempt to lower the rate of local failure. Evidence from prospective randomized trials has demonstrated that radiotherapy improves local control in extremity sarcomas 18.Thishas become standard practice in extremity sarcoma management and prompted investigations into combined-modality treatment for RPS 19,. Retroperitoneal sarcomas, however, present several radiotherapeutic challenges. Tumours are large and often adjacent to structures with low radiation tolerance. RPS comprise a heterogeneous group of pathologies with variable radiosensitivities. Several observational studies have been published to evaluate the outcome and feasibility of radiotherapy in the management of RPS 22. The advantages of preoperative radiotherapy include the tumour displacing some of the adjacent radiosensitive organs, clear demarcation of the tumour for radiotherapy planning, and a lower biological radiotherapy dose in the preoperative setting 22. Postoperative radiotherapy makes it possible to select patients at highest risk for recurrence based on the grade and margin status. However, in the postoperative setting, the normal adjacent organs will move into and become adherent to the tumour bed, leading to a higher risk of radiation-associated toxicity. In an attempt to reduce such toxicities, studies have evaluated the treatment planning with conformal therapies such as intensity-modulated radiation therapy 23 or the

9 Surgical management of primary retroperitoneal sarcoma use of intraoperative radiotherapy 24. The rarity and heterogeneous biological behaviour of RPS will necessitate multi-institutional combined efforts to define the precise role of combined-modality treatment. Acknowledgements The authors acknowledge the assistance of Ms Karen Thomas, statistician at the Royal Marsden NHS Foundation Hospital, for statistical analysis of the data. We acknowledge NHS support to the NIHR Biomedical Research Centre at the Royal Marsden NHS Foundation Trust. The authors declare no conflict of interest. References 1 Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ. Cancer statistics, 07. CA Cancer J Clin 07; 57: Lewis JJ, Leung D, Woodruff JM, Brennan MF. Retroperitoneal soft-tissue sarcoma: analysis of 500 patients treated and followed at a single institution. Ann Surg 1998; 228: Katz MH, Choi EA, Pollock RE. Current concepts in multimodality therapy for retroperitoneal sarcoma. Expert Rev Anticancer Ther 07; 7: Tzeng CW, Fiveash JB, Heslin MJ. Radiation therapy for retroperitoneal sarcoma. Expert Rev Anticancer Ther 06; 6: Bonvalot S, Rivoire M, Castaing M, Stoeckle E, Le Cesne A, Blay JY et al. Primary retroperitoneal sarcomas: a multivariate analysis of surgical factors associated with local control. JClinOncol09; 27: Gronchi A, Lo Vullo S, Fiore M, Mussi C, Stacchiotti S, Collini P et al. Aggressive surgical policies in a retrospectively reviewed single-institution case series of retroperitoneal soft tissue sarcoma patients. JClinOncol09; 27: Pisters PW. Resection of some but not all clinically uninvolved adjacent viscera as part of surgery for retroperitoneal soft tissue sarcomas. JClinOncol09; 27: Trojani M, Contesso G, Coindre JM, Rouesse J, Bui NB, de Mascarel A et al. Soft-tissue sarcomas of adults; study of pathological prognostic variables and definition of histopathological grading system. Int J Cancer 1984; 33: Anaya DA, Lahat G, Wang X, Xiao L, Tuvin D, Pisters PW et al. Establishing prognosis in retroperitoneal sarcoma: a new histology-based paradigm. Ann Surg Oncol 09; 16: Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: Neuhaus SJ, Barry P, Clark MA, Hayes AJ, Fisher C, Thomas JM. Surgical management of primary and recurrent retroperitoneal liposarcoma. Br J Surg 05; 92: Singer S, Antonescu CR, Riedel E, Brennan MF. Histologic subtype and margin of resection predict pattern of recurrence and survival for retroperitoneal liposarcoma. Ann Surg 03; 238: Lehnert T, Cardona S, Hinz U, Willeke F, Mechtersheimer G, Treiber M et al. Primary and locally recurrent retroperitoneal soft-tissue sarcoma: local control and survival. Eur J Surg Oncol 09; 35: Hassan I, Park SZ, Donohue JH, Nagorney DM, Kay PA, Nasciemento AG et al. Operative management of primary retroperitoneal sarcomas: a reappraisal of an institutional experience. Ann Surg 04; 239: Brennan MF. Local recurrence in soft tissue sarcoma: more about the tumor, less about the surgeon. Ann Surg Oncol 07; 14: Anaya DA, Lev DC, Pollock RE. The role of surgical margin status in retroperitoneal sarcoma. JSurgOncol08; 98: Stojadinovic A, Leung DH, Hoos A, Jaques DP, Lewis JJ, Brennan MF. Analysis of the prognostic significance of microscopic margins in 84 localized primary adult soft tissue sarcomas. Ann Surg 02; 235: Yang JC, Chang AE, Baker AR, Sindelar WF, Danforth DN, Topalian SL et al. Randomized prospective study of the benefit of adjuvant radiation therapy in the treatment of soft tissue sarcomas of the extremity. JClinOncol1998; 16: Zlotecki RA, Katz TS, Morris CG, Lind DS, Hochwald SN. Adjuvant radiation therapy for resectable retroperitoneal soft tissue sarcoma: the University of Florida experience. Am J Clin Oncol 05; 28: Feng M, Murphy J, Griffith KA, Sondak VK, Lucas DR, McGinn CJ et al. Long-term outcomes after radiotherapy for retroperitoneal and deep truncal sarcoma. Int J Radiat Oncol Biol Phys 07; 69: Youssef E, Fontanesi J, Mott M, Kraut M, Lucas D, Mekhael H et al. Long-term outcome of combined modality therapy in retroperitoneal and deep-trunk soft-tissue sarcoma: analysis of prognostic factors. Int J Radiat Oncol Biol Phys 02; 54: Raut CP, Pisters PW. Retroperitoneal sarcomas: combined-modality treatment approaches. JSurgOncol06; 94: Bossi A, De Wever I, Van Limbergen E, Vanstraelen B. Intensity modulated radiation-therapy for preoperative posterior abdominal wall irradiation of retroperitoneal liposarcomas. Int J Radiat Oncol Biol Phys 07; 67: AlektiarKM,HuK,AndersonL,BrennanMF,HarrisonLB. High-dose-rate intraoperative radiation therapy (HDR-IORT) for retroperitoneal sarcomas. Int J Radiat Oncol Biol Phys 00; 47: