Retroperitoneal Soft Tissue Sarcomas: Prognosis and Treatment of Primary and Recurrent Disease in 117 Patients

Retroperitoneal Soft Tissue Sarcomas: Prognosis and Treatment of Primary and Recurrent Disease in 117 Patients INGO ALLDINGER 1,2, QIN YANG 3, CHRISTIAN PILARSKY 1, HANS-DETLEV SAEGER 1, WOLFRAM T. KNOEFEL 2,4 and MATTHIAS PEIPER 2,4 1 Department of Visceral-, Thoracic- and Vascular Surgery, University Hospital Dresden, Fetscherstr. 74, D-01307 Dresden; 2 Department of General and Visceral Surgery and 3 Center of Clinical Studies, University Hospital Düsseldorf, Moorenstr. 5, D-40225 Düsseldorf; 4 Department of General-, Visceral- and Thoracic Surgery, University Hospital Eppendorf, Martinistraße 52, D-20246 Hamburg, Germany Abstract. Background: The objective of this study was to define prognostic factors for patients with primary soft tissue sarcomas (STS) arising from the retroperitoneum. Patients and Methods: One hundred and seventeen consecutive patients, resected in our institutions between July 1972 and November 2002, were reviewed. Results: The prognostic factors predicting survival were incomplete resection, a tumor of high grade (G3), metastases to lymph nodes and distant metastasis. Patients with a malignant fibrous histiocytoma (MFH) or a malignant peripheral nerve sheath tumor (MPNST) had a worse prognosis than those patients with other tumors. The prognostic factors predicting local recurrence were incomplete resection and high grade (G3). The prognostic factors predicting metastasis were incomplete resection, lymph node metastasis at the time of the resection of the primary tumor and tumor histology. Conclusion: Since only complete tumor resection offers a chance for cure, it is mandatory, and local control remains the most significant challenge in the management of retroperitoneal sarcomas. Other therapies can support surgical treatment, depending on the tumor localization and histological entity. The management of patients with a STS should be provided by a specialized team of surgeons, oncologists and radiotherapists, and patients should be enrolled in a treatment study whenever possible. Soft tissue sarcomas (STS) constitute a heterogeneous entity of neoplasms of different histological subtypes. Ten 15% of all STS arise from the retroperitoneum (1, 2). Other than radical surgery, no other effective treatment is known to Correspondence to: Dr. Ingo Alldinger, Dept. of General and Visceral Surgery, University Hospital Düsseldorf, Moorenstr. 5, 40225 Düsseldorf, Germany. Tel: +49 211 7350, Fax: +49 211 7359, e-mail: Ingo.Alldinger@med.uni-duesseldorf.de Key Words: Soft tissue sarcoma, retroperitoneal, surgery, survival. achieve local control and improved survival. Even if radiation is often applied after complete or incomplete resection, its role remains controversial. Patients (n=117) with primary retroperitoneal soft tissue sarcomas (RSTS), resected in the Hamburg and Dresden University Hospitals, Germany, were reviewed. Patients and Methods Between July 1972 and November 2002, 117 patients with RSTS were admitted to our hospitals and operated on. Fifty-nine of the patients were male (50.4%) and 58 (49.6%) were female. The quality of resection was determined according to the UICC R-classification (3). The clinical course of all patients was monitored prospectively and reviewed retrospectively. The data were collected, in part, from the patient, in part by interview of the patient s General Practitioner and, in part, extracted from the patient s chart. The patients were followed until death or the doctor in charge of the follow-up was contacted for the latest status. The neoplasms were characterized by histology and tumor grade. Resection was rated R0 when both the surgeon and the pathologist had no doubt that the tumor had been resected in total without opening of the tumor. In this case the resection was rated R1, which, by consensus, accounts for the large group of R1 resections. Statistical analysis. Descriptive statistics are reported as proportions. Survival and local recurrence curves were estimated using the Kaplan-Meier method (4). Step-wise logistic analysis was performed to identify prognostic risk factors (5) and the likelihood ratio test was used to assess statistical significance. Odds ratios and 95% confidence intervals were calculated as a measure of risk. All comparisons were made by means of a two-sided log-rank test with a 0.05 significance level (6). The Cox proportional hazards model was used to derive hazard ratios (7). Two-tailed values of p<0:05 were considered significant. All variables having a p value in the univariate analysis of 5% or less were included in the Cox regression model. Data analysis was conducted using the GENMOD, LIFETEST and PHREG procedures in SAS for Windows (version 6.11, SAS Institute Inc., Cary, NC, USA). 0250-7005/2006 $2.00+.40 1577

Table I. Histology of tumors. Histology No. Percent Liposarcoma 31 26.5 Leiomyosaroma 23 19.7 Malignant Schwannoma 20 17.1 MFH 18 15.4 Chondrosarcoma 5 4.3 Rhabdomyosarcoma 5 4.3 Fibrosarcoma 4 3.4 Myxoid sarcoma 2 1.8 Pleomorph sarcoma 2 1.8 PNET 2 1.8 Neurofibrosarcoma 1 1 Ganglioneuroblastoma 1 1 Synovial sarcoma 1 1 Not classified 2 1.8 Total 117 100 MFH=malignant fibrous histiocytoma; PNET=primitive neuroectodermal tumor. Table II. Staging according to UICC. Stage No. Percent Ia 1 0.9 IIa 56 47.9 III 43 36.8 IV 17 14.5 Total 117 100 Table III. Grading of recurrent tumors. G1 G2 G3 Total R0 1 1 3 5 R1 10 7 11 28 R2 7 3 8 18 Total 18 11 22 51 Results Operative treatment. All 117 patients presented in this study underwent surgical resection. The primary tumor was resected in 70 (59.8%) patients in our institutions and 47 (40.2%) patients in other hospitals, subsequently presenting in our hospitals with recurrent disease. In 20 (17.1%) patients, initial surgery resulted in an R0 status, while 54 resections (46.2%) were rated R1 and 41 (35%) R2, respectively. The percentage of patients in each group being almost identical in our hospitals and the other hospitals (R0: 20% vs. 14.9%; R1: 44.3% vs. 44.7%; R2: 32.9% vs. 36.2%). The R status of 2 patients could not be obtained. Histology. Leiomyosarcoma (23, 19.6%), liposarcoma (31, 26.5%), malignant fibrous histiocytoma (MFH, 18, 15.4%) and malignant peripheral nerve sheet sarcoma (MPNST) (20, 17.1%) accounted for 92 of the 117 RSTS (78.6%). The other histological entities were rare (Table I). Thirty sarcomas (25.6%) were rated well-differentiated by the pathologist (G1), 34 (29.1%) were of moderate (G2) and 52 (44.4%) were of poor differentiation (G3). The grading of 1 patient was not available. Staging. One hundred and fifteen of the 117 patients had a tumor of more than 5 cm in diameter (T2); only 2 tumors were T1 (<5 cm). In 11 patients, lymph node metastases were resected (9.4%). Seventeen patients presented with distant metastasis at the time of resection of the primary tumor (14.5%). The tumor staging according to the UICC classification is listed in Table II. Post-operative treatment. Post-operative adjuvant/palliative therapy was administered after interdisciplinary consultation based on individual decisions. Post-operative radiation therapy was performed in 18 patients (15.4%), 4 with R0, 12 with R1 and 2 with R2 resections. In 21 patients, radiotherapy was performed after the first or second recurrence. Fourteen patients (12%) received adjuvant chemotherapy for their primary disease, while 33 other patients received chemotherapy either for local recurrence or distant metastases. Local recurrence. Thirty-three of the 74 patients (44.6%) with initial R0 or R1 resection developed local recurrence (5 out of 20 R0 (25%) and 28 out of 54 R1 (52.3%)). Eighteen out of the 41 patients (43.9%) with a tumor resection rated as R2 had growth of their residual tumor mass diagnosed at a later stage of their disease. The grading of recurrent resected tumors is listed in Table III. The grading and resection quality were significant prognostic factors for local tumor recurrence. Local recurrence was more likely to occur for high-grade tumors (G3) than for G1 and G2 tumors (23 of 52 vs. 30 of 65 patients, respectively; p=0.0104, Figure 1). In the multivariate analysis, the R status remained the sole independent prognostic factor (p=0.003, exp. 2.508, 95% confidence interval [CI] 1.37 to 4.59). Metastasis. In addition to the 20 patients with initial metastatic disease, 36 patients (30.8%) developed metastases after a median of 33 months. Distant metastases were predominantely found in the liver (14, 38.9%) and lungs (11, 30.6%). Univariate analyses revealed positive regional lymph nodes at initial resection (p<0.0001) and the chemotherapy 1578

Alldinger et al: Prognosis of Retroperitoneal Soft Tissue Sarcomas Figure 1. Time to local recurrence depending on tumor grade. Figure 2. Metastasis depending on histology. applied (p<0.0245) to be predictive factors. The development of distant metastases was found for all histological entities but liposarcoma, for which no distant metastasis was detected in either the liver or lungs (p<0.0001, Figure 2). Two patients, however, presented with either regional lymph nodes metastases or peritoneal metastases at initial resection. Both patients died within 7 months after the operation. One patient with liposarcoma developed lymph node metastases 13 months after operation of the primary tumor. Multivariate analysis revealed positive regional lymph nodes (N1, p<0.0001), administered chemotherapy (p=0.023) and a histological entity other than liposarcoma (p=0.005) as risk factors for the development of metastasis. Patient status, last follow-up and survival. Of all 117 patients, 29 (24.8%) patients were still alive after a median of 71 (range: 17 195) months, 19 of them (65.5%) without evidence of disease. Eighty-one of our patients (69.2%) died of tumor disease after a median of 45 (range 1 208) months. The mean survival for all patients was 45 (range 1 208, median 28.5) months; 35 patients (30%) lived after 5 years and 11 patients (9.4%) 10 years after resection. Prognostic factors predicting survival. Tumor grading was a significant prognostic factor for overall survival. Patients with a tumor of poor differentiation had a poorer outcome than patients with a well-differentiated tumor (p<0.0001, Figure 3). Additional prognostic factors in the univariate analyses were positive distant metastases at initial diagnosis (M1, p<0.0001) and positive regional lymph node metastases (N1, p<0.0001). Accordingly, the UICC tumor stage correlated with survival (p<0.0001, Figure 4). Complete or marginal resections were associated with a higher overall survival compared to R2 resected tumors (p<0.0001, Figure 5). Adjuvant radiotherapy improved survival (p=0.0413), as did adjuvant chemotherapy (p=0.0220). Interestingly, patients suffering from local recurrence of the resected tumor at one time in their disease course did have a better outcome than patients who never had local recurrence (p=0.0019). Resection of the recurrent tumor was of benefit for the patients (p=0.0033). This might be due to a selection bias for those patients whose recurrent sarcoma was not appropriate for recurrent resection in spite of regular follow-up visits. Patients with MFH and malignant peripheral nerve sheath sarcoma (MPNST) had a worse outcome than other patients (p=0.0071). In the multivariate analysis, UICC stage III (p<0.001), M1 (p=0.019), R2 resection (p<0.001), the absence of local recurrence (p<0.001) and the histological entities MFH and MPNST (p=0.027) proved to be independent factors predicting worse survival. Discussion One hundred and seventeen patients resected for primary retroperitoneal STS are presented. The ratio of men and women in our population was nearly 1:1 (59 men and 58 1579

Figure 3. Survival depending on tumor grade. Figure 4. Survival depending on UICC stage. women), reflecting the populations of other studies on STS with a mean relation of 1.1:1 (1, 2, 8, 9). Sixty-five percent of patients underwent gross total resection at the initial operation. However, only those patients where the tumor could be resected without opening of the tumor capsula were rated RO. STS grows in a pseudo capsula that is frequently opened in the course of the operation (10), the reason being an unexperienced surgeon or a technically difficult operation. In retroperitoneal localization of STS it is often impossible to resect the tumor without damaging this capsula. In this case, even if the surgeon is convinced that the tumor has been completely resected, we believe that the operation should be rated R1. We did not see a difference in survival or local recurrence between patients rated R0 and those rated R1. There was a higher incidence of local recurrence and a shorter survival time for patients with R2 resection. This emphasizes the importance of local control. While for carcinomas a consensus concerning the classification and description of tumors has been established and observed by pathologists throughout the world, correct sarcoma classification requires a specialized and experienced pathologist. Any STS that cannot be classified without any doubt should be co-evaluated by a specialized patholgist. In our population, almost half of the tumors (44.4%) were poorly-differentiated, 29% moderately- and 25.6% were well-differentiated. Only 17 patients presented with Figure 5. Survival depending on resection quality (R status). metastases at the time of the operation of their primary tumor. Forty patients developed metastases during followup. Interestingly, liposarcoma, in contrast to all other 1580

Alldinger et al: Prognosis of Retroperitoneal Soft Tissue Sarcomas histological entities, caused no distant metastases. The same phenomenon was described by Gronchi et al. (11). Local tumor recurrence did not depend on histology, however, but on tumor grading and resection quality only. Patients had a shorter survival with MFH and MPNST. The rate of R1 or R2 resections did not vary significantly between the different histotypes. The role of chemotherapy in STS remains controversial. Since response rates to chemotherapy of only 20 to a maximum of 60% are seen, patients are often exposed to sometimes severe adverse effects without a therapeutic outcome. The therapeutic options are even worse in the case of retroperitoneal sarcoma where, in most cases, radiation therapy cannot be administered due to the risk of bowel damage (12). Pre-operative radiation with an intraoperative boost, combined or not with chemotherapy, has been administered (13-16), and a study on hyperthermic intraperitoneal intra-operative chemotherapy after cytoreductive surgery had been conducted by Rossi et al. (14), but these approaches remain experimental, and no long-time survival benefit for the patients has been shown to date (14). In our study, the administered chemotherapy was a risk factor for metastasis and worse survival, the reason of course being a selection bias towards patients with an aggressive tumor in an advanced tumor stage. The observation that patients presenting with recurrent disease do have a better prognosis than patients who never have a recurrent tumor parallels the finding that men suffering from prostate cancer live longer than men who never have this cancer diagnosed (17); patients living long enough to develop a local recurrence may have a less aggressive tumor than patients who die from their tumor before the manifestation of local recurrence. Like other authors (2, 13, 18), we showed a correlation of survival time with tumor grading and staging. Complete resection is a crucial prognostic factor. The benefit of clean resection margins cannot be proved statistically. Local control remains the most significant challenge in the management of retroperitoneal sarcomas. Though the rates of complete resection and overall survival are improving according to the data published, no new surgical options are available, underlining the need to implement adjuvant therapies to further improve outcome. New therapies could be based on the expression of biological markers like CD44 or vascular endothelial growth factor, which have been shown to have an impact on the progression and prognosis of STS (19, 20) and should be evaluated in the future. References 1 Malerba M et al: Primary retroperitoneal soft tissue sarcomas: results of aggressive surgical treatment. World J Surg 23(7): 670-675, 1999. 2 Bautista N, Su W and O'Connell TX: Retroperitoneal softtissue sarcomas: prognosis and treatment of primary and recurrent disease. Am Surg 66(9): 832-836, 2000. 3 Wittekind C, Bootz F and Meyer HJ: TNM, Klassifikation maligner Tumoren. Springer Verlag, Berlin, 2003. 4 Kaplan E and Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53: 457-481, 1958. 5 Walker S and Duncan D: Estimation of the probability of an event as a function of several independant variables. Biometrika 54: 167-179, 1967. 6 Mantel N: Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 50: 163-170, 1966. 7 Cox D: Regression models and life-tabels (with discussion). J R Stat Soc (B) 34: 187-220, 1972. 8 Peiper M et al: Survival in patients with primary soft-tissue sarcomas treated within 6 years. J Cancer Res Clin Oncol 124(3-4): 199-206, 1998. 9 Pirayesh A et al: The management of retroperitoneal soft tissue sarcoma: a single institution experience with a review of the literature. Eur J Surg Oncol 27(5): 491-497, 2001. 10 Zornig C, Peiper M and Schroder S: Re-excision of soft tissue sarcoma after inadequate initial operation. Br J Surg 82(2): 278-279, 1995. 11 Gronchi A et al: Retroperitoneal soft tissue sarcomas: patterns of recurrence in 167 patients treated at a single institution. Cancer 100(11): 2448-2455, 2004. 12 McGinn CJ: The role of radiation therapy in resectable retroperitoneal sarcomas. Surg Oncol 9(2): 61-65, 2000. 13 Gilbeau L et al: Surgical resection and radiotherapy for primary retroperitoneal soft tissue sarcoma. Radiother Oncol 65(3): 137-143, 2002. 14 Rossi CR et al: Hyperthermic intraperitoneal intraoperative chemotherapy after cytoreductive surgery for the treatment of abdominal sarcomatosis: clinical outcome and prognostic factors in 60 consecutive patients. Cancer 100(9): 1943-1950, 2004. 15 Sindelar WF et al: Intraoperative radiotherapy in retroperitoneal sarcomas. Final results of a prospective, randomized, clinical trial. Arch Surg 128(4): 402-410, 1993. 16 Bussieres E et al: Retroperitoneal soft tissue sarcomas: a pilot study of intraoperative radiation therapy. J Surg Oncol 62(1): 49-56, 1996. 17 Harris R and Lohr KN: Screening for prostate cancer: an update of the evidence for the U.S. Preventive Services Task Force. Ann Intern Med 137(11): 917-929, 2002. 18 Stoeckle E et al: Prognostic factors in retroperitoneal sarcoma: a multivariate analysis of a series of 165 patients of the French Cancer Center Federation Sarcoma Group. Cancer 92(2): 359-368, 2001. 19 Potti A et al: Immunohistochemical detection of HER-2/neu, c-kit (CD117) and vascular endothelial growth factor (VEGF) overexpression in soft tissue sarcomas. Anticancer Res 24(1): 333-337, 2004. 20 Peiper M et al: CD44s expression is associated with improved survival in soft tissue sarcoma. Anticancer Res 24(2C): 1053-1056, 2004. Received September 20, 2005 Revised December 7, 2005 Accepted February 1, 2006 1581