Leiomyosarcoma: One Disease or Distinct Biologic Entities Based on Site of Origin?

Transcription

1 2015;111: Leiomyosarcoma: One Disease or Distinct Biologic Entities Based on Site of Origin? DAVID J. WORHUNSKY, MD, 1 MIHIR GUPTA, BS, 1 SEPIDEH GHOLAMI, MD, 1 THUY B. TRAN, MD, 1 KRISTEN N. GANJOO, MD, 2 MATT VAN DE RIJN, MD, 3 BRENDAN C. VISSER, MD, 1 JEFFREY A. NORTON, MD, 1 AND GEORGE A. POULTSIDES, MD, MS 1 * 1 Department of Surgery, Stanford University Medical Center, Stanford, California 2 Department of Medical Oncology,, Stanford University Medical Center, Stanford, California 3 Department of, Pathology, Stanford University Medical Center, Stanford, California Background: Leiomyosarcoma (LMS) can originate from the retroperitoneum, uterus, extremity, and trunk. It is unclear whether tumors of different origin represent discrete entities. We compared clinicopathologic features and outcomes following surgical resection of LMS stratified by site of origin. Methods: Patients with LMS undergoing resection at a single institution were retrospectively reviewed. Clinicopathologic variables were compared across sites. Survival was calculated using the Kaplan Meier method and compared using log-rank and Cox regression analyses. Results: From 1983 to 2011, 138 patients underwent surgical resection for LMS. Retroperitoneal and uterine LMS were larger, higher grade, and more commonly associated with synchronous metastases. However, disease-specific survival, recurrence-free survival, and recurrence patterns were not significantly different across the four sites. Synchronous metastases (HR 3.20, P < 0.001), but not site of origin, size, grade, or margin status, were independently associated with worse DSS. A significant number of recurrences and disease-related deaths were noted beyond 5 years. Conclusions: Although larger and higher grade, retroperitoneal and uterine LMS share similar survival and recurrence patterns with their trunk and extremity counterparts. LMS of various anatomic sites may not represent distinct disease processes based on clinical outcomes. The presence of metastatic disease remains the most important prognostic factor for LMS. J. Surg. Oncol. 2015;111: ß 2015 Wiley Periodicals, Inc. KEY WORDS: sarcoma; leiomyosarcoma; outcome assessment INTRODUCTION Leiomyosarcoma (LMS) comprises approximately 25 percent of all soft tissue sarcomas (STS) [1 3]. In general, LMS can be differentiated from other forms of STS by the presence of smooth muscle actin and desmin on immunohistochemistry [4,5]. Importantly, gastrointestinal stromal tumors (GIST) are now recognized as a biologically and clinically distinct tumor compared to LMS, primarily based on their expression of the KIT (CD117) protein [3,6].LMS can arise from any area of smooth muscle in the body, with the retroperitoneum, uterus and extremity being the most common anatomic sites. As with most STS, complete surgical resection offers the best chance for cure. In addition to complete surgical resection, other clinicopathologic features have been found to be associated with LMS disease prognosis, including age, depth, size, grade and metastatic spread [7 10]. The impact of primary site of origin on LMS behavior and prognosis remains a matter of debate. Studies reporting LMS outcomes have had variable inclusion criteria, with some studies excluding visceral tumors, [9 12] others excluding cutaneous tumors [11,12], others excluding uterine tumors [8], and others including uterine tumors [13]. A recent study investigating clinicopathologic differences between uterine and non-uterine LMS found no evidence to support that these two entities represent distinct disease processes on a clinical level [7]. Similarly, molecular studies evaluating potential differences between uterine and non-uterine LMS also have failed to demonstrate any conclusive differences between the two in terms of molecular profiling or DNA copy number changes [14,15]. The goal of our study was to assess the clinicopathologic features, recurrence patterns, and survival outcomes for patients with LMS and to determine the impact of site of origin on clinical outcome. PATIENTS AND METHODS Patients with LMS who underwent surgical resection at Stanford University Medical Center between 1983 and 2011 were identified using the Stanford Cancer Center Research Database. Patients with LMS of any anatomic site and who underwent resection of their primary tumor or of recurrent disease were included in the present study. Patients with GIST were excluded. Clinical data were obtained by review of the patients electronic medical record after obtaining institutional review board approval. Patient demographics, clinicopathologic data, and outcomes were collected. Anatomic site was divided into four broadly defined categories: retroperitoneal (including intraabdominal), uterine, extremity, and trunk (including head and neck as well as chest, abdominal and pelvic walls external to the body cavities). Cutaneous tumors were included in either the extremity or trunk categories. Depth of the tumor was categorized as superficial or deep to the superficial Sources of Funding: None *Correspondence to: George A. Poultsides, MD, MS, Stanford University Medical Center, Division of Surgical Oncology, 300 Pasteur Drive, Suite H3680, Stanford, CA Fax: gpoultsides@stanford.edu Received 26 October 2014; Accepted 28 February 2015 DOI /jso Published online in Wiley Online Library (wileyonlinelibrary.com). ß 2015 Wiley Periodicals, Inc.

2 Leiomyosarcoma Outcomes by Site of Origin 809 muscular fascia (applies to trunk or extremity tumors). Size was defined as the maximal dimension of the tumor on final pathology. Tumor grade was classified as high based on the French Federation of Cancer Centers Sarcoma Group (FNCLCC) system that incorporates tumor differentiation, mitotic count and tumor necrosis [16]. All grading data were from the original pathology reports reviewed at our institution. Surgical resection margins were defined as R0 (microscopically negative), R1 (microscopically positive), or R2 (gross residual disease). Survival data were determined from the electronic medical record and the Stanford Cancer Center Research Database (that includes survival data from the California Cancer Registry). Categorical data were represented as count (percentage) and compared using the Fisher s exact test. Continuous data were represented as median (range) and compared using analysis of variance (ANOVA). Disease-specific survival (DSS) was calculated from the date of initial operation to the date of death from disease. Recurrence-free survival (RFS) was calculated from the date of initial operation to the date of first relapse. Kaplan Meier survival analysis was used to calculate survival probabilities, and differences between survival curves were evaluated using the log-rank test. Recurrence patterns were categorized by site of first recurrence (local, distant or both). Univariate and multivariate analyses were performed using Cox regression analysis and reported as hazard ratio (HR) and 95% confidence interval (CI). A P-value less than 0.05 was considered statistically significant. All statistical analyses were performed using SPSS (version 21.0, IBM). RESULTS Between 1983 and 2011, 138 patients underwent surgical resection for LMS at our institution. This cohort includes 93 (67%) patients who had their initial operation at our institution, 22 (16%) who underwent definitive surgical resection at our institution for persistent disease shortly after incomplete resection at an outside institution, and 23 (17%) who had resection of recurrent or metastatic disease at our institution following primary resection at another institution (after a median interval of 35 months). Tumors were located in the retroperitoneal/intraabdominal region (n ¼ 49, 35%), uterus (n ¼ 41, 30%), extremities (n ¼ 26, 19%), and trunk (n ¼ 22, 16%). This included 3 (2%) patients with cutaneous tumors. Clinicopathologic features of the 138 patients are shown in Table I. Patients with LMS of the retroperitoneum and uterus were more likely to present with synchronous metastases and to have larger and higher grade tumors. Of 33 patients with synchronous metastases, 30 (91%) received some form of treatment for their metastatic disease. This included fifteen (45%) patients who underwent surgical metastasectomy in a combined or staged fashion. Six (18%) patients received neoadjuvant chemotherapy and two (6%) received neoadjuvant chemoradiation. Finally, 21 (64%) received adjuvant therapy, 11 (33%) with chemotherapy and 10 (30%) with radiotherapy. Radiotherapy was aimed at the primary tumor in seven patients, at the metastatic disease in three patients, and at both in two patients. Survival analysis was performed on 105 patients presenting without metastatic disease. After a median follow-up of 39 (1 355) months, there was no significant difference in DSS across the four sites of origin (Fig. 1). Though there was a trend toward worse survival among LMS of the retroperitoneum and uterus, this difference did not reach statistical significance (P ¼ 0.31). Similarly, there was no significant difference in RFS across the four sites of origin (Fig. 2,P ¼ 0.14). A careful review of the survival curves (Figs. 1 and 2) reveals that a significant number of recurrences and disease-related deaths occurred past the 5-year mark following resection. Multivariate analysis was performed on the entire cohort (n ¼ 138) to identify prognostic factors important for DSS. With covariates including high grade, positive resection margin (R1 or R2), size >5 cm, and site of origin, only the presence of synchronous metastases was an independent predictor of DSS (HR 3.20, 95%CI , P < 0.001; Table II). When a similar analysis was performed on the 105 patients without synchronous metastases, no significant independent predictors of DSS were identified, including site of origin (Table III). Of the 105 patients presenting without synchronous metastases, 56 (53%) experienced disease recurrence. Recurrence patterns (local, distant or both) did not differ among the four anatomic sites (Fig. 3). Distant recurrence (n ¼ 34, 61%) was the most common pattern of failure, with lung (n ¼ 27, 79%) being the most common site, followed by bone (n ¼ 8, 24%) and liver (n ¼ 7, 21%). DISCUSSION In the present study, we report on 138 patients who underwent surgical resection of LMS over a 28-year period at a single institution with special attention to stratifying outcomes by site of origin. Our data suggest that although retroperitoneal and uterine LMS tend to be larger and of higher grade compared to their extremity and trunk counterparts, survival outcomes (both DSS and RFS) as well as recurrence patterns TABLE I. Clinicopathologic Features of 138 Patients With Leiomyosarcoma Who Underwent Surgical Resection Characteristic All Patients (n ¼ 138) Retroperitoneal (n = 49) Uterine (n = 41) Extremity (n = 26) Trunk (n = 22) P Age, median (range) 54 (17 91) 57 (35-91) 51 (35 86) 60 (17 88) 55 (27 84) Female, No. (%) 99 (72) 35 (71) 41 (100) 8 (31) 15 (68) <0.001 Definitive operation at our institution, No. (%) 115 (83) 42 (86) 30 (73) 23 (89) 20 (91) Disease Extent at Presentation, No. (%) Local 86 (62) 26 (53) 20 (49) 19 (73) 21 (96) Recurrent 19 (14) 8 (16) 8 (20) 2 (8) 1 (5) Synchronous Metastases 33 (24) 15 (31) 13 (32) 5 (19) 0 Size, median (range), cm a 7.5 ( ) 11.0 ( ) 9.0 ( ) 3.3 ( ) 6.5 ( ) <0.001 Size, No. (%) a T1 (5 cm) 43 (32) 9 (18) 5 (13) 19 (73) 10 (45) <0.001 T2 (>5 cm) 92 (68) 40 (82) 33 (87) 7 (27) 12 (55) High grade, No. (%) b 77 (58) 29 (60) 29 (74) 8 (33) 11 (50) Deep to investing fascia, No. (%) c 124 (91) 49 (100) 41 (100) 16 (64) 18 (82) <0.001 Negative (R0) resection margin, No. (%) b 102 (77) 31 (65) 31 (80) 23 (92) 17 (81) R0 or R1 resection margin, No. (%) b 129 (97) 44 (92) 39 (100) 25 (100) 21 (100) a Tumor size not available for 3 patients. b Tumor grade and margin status not available for 5 patients. c Tumor depth not available for one patient.

3 810 Worhunsky et al. Fig. 1. Disease-specific survival after surgical resection for 105 patients with leiomyosarcoma without metastatic disease at presentation (P ¼ 0.31). are similar across the four sites of origin. Additionally, only the presence of synchronous metastatic disease is an independent predictor of DSS on multivariate analysis. Though further studies are needed, these data may call into question the notion that LMS of various anatomic sites represent clinically distinct disease processes. Though LMS is a relatively common form of soft tissue sarcoma, few series in the literature analyze the natural history of LMS based on site of origin. Analysis is especially limited by the fact that many older reports included GIST and LMS together due to similar gross and microscopic features (prior to the immunohistochemical basis for diagnosing GIST) [6,13]. Furthermore, some of the larger studies to date have had varying inclusion criteria and some have excluded patients with LMS of the retroperitoneum, uterus, or skin [8 11]. Conflicting data also exist as to whether LMS of various sites of origin exhibit differential chemosensitivity, molecular signatures, and staging features. Studies of chemotherapy outcomes for uterine versus non-uterine LMS have shown variable but not statistically different response rates [17,18]. In addition, studies on the molecular differences between LMS of various anatomic sites remain contradictory. A study on LMS genetic profiling by Italiano et al found retroperitoneal LMS to Fig. 2. Recurrence-free survival after surgical resection of 105 leiomyosarcoma patients without metastatic disease at presentation (P ¼ 0.14).

4 TABLE II. Univariate and Multivariate Analysis of Disease-Specific Survival for All Study Patients Leiomyosarcoma Outcomes by Site of Origin 811 Univariate Analysis Multivariate Analysis Characteristic HR (95% CI) P HR (95% CI) P High Grade 2.00 ( ) ( ) 0.47 Positive Margin (R1 or R2) 1.82 ( ) ( ) 0.08 Size >5 cm 2.21 ( ) ( ) 0.26 Site of origin Retroperitoneal 1.0 (Reference) 1.0 (Reference) Uterine 1.09 ( ) ( ) 0.59 Extremity 0.67 ( ) ( ) 0.94 Trunk 0.45 ( ) ( ) 0.67 Synchronous metastases 4.42 ( ) < ( ) <0.001 TABLE III. Univariate and Multivariate Analysis of Disease-Specific Survival for Patients Presenting Without Synchronous Metastases Univariate Analysis Multivariate Analysis Characteristic HR (95% CI) P HR (95% CI) P High grade 1.43 ( ) ( ) 0.75 Positive margin (R1 or R2) 1.10 ( ) ( ) 0.82 Size >5 cm 2.26 ( ) ( ) 0.25 Site of origin Retroperitoneal (Reference) (Reference) Uterine 1.26 ( ) ( ) 0.48 Extremity 0.53 ( ) ( ) 0.70 Trunk 0.65 ( ) ( ) 0.66 more commonly overexpress genes involved in muscle differentiation, while extremity LMS were characterized by overexpression of genes mainly involved in extracellular matrix, wounding, and adhesion pathways [19]. Svarvar and colleagues used comparative genomic hybridization to reveal two genetic clusters that significantly segregated uterine and non-uterine LMS [15]. Conversely, using a combination of gene expression profiling, array comparative genomic hybridization, and immunohistochemistry, Beck et al failed to identify anatomic site of LMS as a factor predictive of DSS on univariate or multivariate modeling [14]. Finally, although many clinicians utilize the modified International Federation of Gynecology and Obstetrics (FIGO) staging system for uterine LMS compared to the American Joint Committee on Cancer (AJCC) soft tissue sarcoma staging system, the ability of either staging system to provide an accurate prognosis for uterine LMS (especially differentiating between stage II and III patients) has been shown to be limited [20]. Farid and colleagues [7] recently evaluated the natural history, outcomes and prognostic factors for LMS, specifically to determine Fig. 3. Recurrence patterns for 105 patients presenting without synchronous metastases (P ¼ 0.812).

5 812 Worhunsky et al. whether differences existed between uterine and non-uterine LMS. In keeping with our findings, they reported no statistically significant differences in survival (overall or recurrence-free) between the two groups, with only the presence of synchronous metastases being predictive of worse survival. Our results are also consistent with other recent LMS series. Gladdy et al [8], in one of the larger series to date (353 patients with primary resectable LMS), found that although retroperitoneal location was associated with worse DSS compared to extremity or trunk tumors on univariate analysis, only grade and size (but not site of origin) were independent predictors of DSS and distant recurrence on multivariate analysis. Similarly, Abraham and colleagues [21], reporting on a series of 115 patients with somatic LMS (including soft tissue, retroperitoneal, vascular, and bone, but excluding uterine, gastrointestinal, and cutaneous forms), found that tumor grade, as well as depth and synchronous metastases (but not site of origin) were independent predictors of DSS. One possible limitation of our study is that it retrospectively spanned a 28-year period, during which significant changes occurred in the care of patients with LMS. These changes include the increasing ability to perform more complicated resections (such as with en bloc vascular resection and reconstruction), refinement of pathologic criteria to determine grade, evolving efficacy of systemic therapy, and willingness to more aggressively treat metastatic disease in selected patients [22,23]. As a result, our study cohort is fairly heterogeneous and our results may be further confounded by the fact that 17% of our patient population had their primary tumor resection at an outside institution (but had a metachronous resection at our institution for locally recurrent or distant metastatic disease). Nonetheless, our intention was to study a diverse cohort of patients with a rare but apparently unified pathologic diagnosis, examining the impact of primary site of origin on the natural history of a heterogeneous tumor entity. Another limitation of the present study is its relatively small sample size, which may explain why tumor grade and size were not shown to be independent predictors of survival, as has been previously demonstrated. As LMS is a relatively rare disease, the problem of sample size is not unique to our study. Despite this sample size limitation, our univariate and multivariate analyses quite clearly showed no association between site of origin and survival after LMS resection, a finding that is consistent with two recent publications. This highlights the need for a robust multi-institutional study on LMS patients to elucidate the role of clinicopathologic factors in predicting biologic behavior and long-term outcome. CONCLUSIONS We found notable clinicopathologic differences between retroperitoneal and uterine LMS compared with extremity and trunk tumors, likely a result of the former being diagnosed at a later stage. However, after controlling for other clinicopathologic variables, survival outcomes and recurrence patterns are similar for LMS anywhere in the body, with synchronous metastases being the only independent predictor of disease-specific survival. As a result, this study suggests that LMS of different anatomic sites may not be biologically distinct tumors, though further studies are clearly needed to elucidate whether distinct clinical and molecular signatures exist across LMS of different anatomic sites. REFERENCES 1. Stiller CA, Trama A, Serraino D, et al.: Descriptive epidemiology of sarcomas in Europe: report from the RARECARE project. Eur J Cancer 2013;49: Ferrari A, Sultan I, Huang TT, et al.: Soft tissue sarcoma across the age spectrum: a population-based study from the Surveillance Epidemiology and End Results database. Pediatr Blood Cancer 2011;57: Toro JR, Travis LB, Wu HJ, et al.: Incidence patterns of soft tissue sarcomas, regardless of primary site, in the surveillance, epidemiology and end results program, : An analysis of 26,758 cases. Int J Cancer 2006;119: Brennan MF, Antonescu CR, Maki RG: Management of Soft Tissue Sarcoma. New York: Springer; Miettinen M, Fetsch JF: Evaluation of biological potential of smooth muscle tumours. Histopathology 2006;48: Katz SC, DeMatteo RP: Gastrointestinal stromal tumors and leiomyosarcomas. J Surg Oncol 2008;97: Farid M, Ong WS, Tan MH, et al.: The influence of primary site on outcomes in leiomyosarcoma: a review of clinicopathologic differences between uterine and extrauterine disease. Am J Clin Oncol 2013;36: Gladdy RA, Qin LX, Moraco N, et al.: Predictors of survival and recurrence in primary leiomyosarcoma. Ann Surg Oncol 2013;20: Miyajima K, Oda Y, Oshiro Y, et al.: Clinicopathological prognostic factors in soft tissue leiomyosarcoma: a multivariate analysis. Histopathology 2002;40: Svarvar C, Bohling T, Berlin O, et al.: Clinical course of nonvisceral soft tissue leiomyosarcoma in 225 patients from the Scandinavian Sarcoma Group. Cancer 2007;109: Farshid G, Pradhan M, Goldblum J, et al.: Leiomyosarcoma of somatic soft tissues: a tumor of vascular origin with multivariate analysis of outcome in 42 cases. Am J Surg Pathol 2002;26: Gustafson P, Willen H, Baldetorp B, et al.: Soft tissue leiomyosarcoma. A population-based epidemiologic and prognostic study of 48 patients, including cellular DNA content. Cancer 1992;70: Clary BM, DeMatteo RP, Lewis JJ, et al.: Gastrointestinal stromal tumors and leiomyosarcoma of the abdomen and retroperitoneum: a clinical comparison. Ann Surg Oncol 2001;8: Beck AH, Lee CH, Witten DM, et al.: Discovery of molecular subtypes in leiomyosarcoma through integrative molecular profiling. Oncogene 2010;29: Svarvar C, Larramendy ML, Blomqvist C, et al.: Do DNA copy number changes differentiate uterine from non-uterine leiomyosarcomas and predict metastasis?. Mod Pathol 2006;19: Neuville A, Chibon F, Coindre JM: Grading of soft tissue sarcomas: from histological to molecular assessment. Pathology 2014;46: Oosten AW, Seynaeve C, Schmitz PI, et al.: Outcomes of first-line chemotherapy in patients with advanced or metastatic leiomyosarcoma of uterine and non-uterine origin. Sarcoma 2009;2009: Pautier P, Floquet A, Penel N, et al.: Randomized multicenter and stratified phase II study of gemcitabine alone versus gemcitabine and docetaxel in patients with metastatic or relapsed leiomyosarcomas: a Federation Nationale des Centres de Lutte Contre le Cancer (FNCLCC) French Sarcoma Group Study (TAXOGEM study). Oncologist 2012;17: Italiano A, Lagarde P, Brulard C, et al.: Genetic profiling identifies two classes of soft-tissue leiomyosarcomas with distinct clinical characteristics. Clin Cancer Res 2013;19: Zivanovic O, Leitao MM, Iasonos A, et al.: Stage-specific outcomes of patients with uterine leiomyosarcoma: a comparison of the international Federation of gynecology and obstetrics and american joint committee on cancer staging systems. J Clin Oncol 2009;27: Abraham JA, Weaver MJ, Hornick JL, et al.: Outcomes and prognostic factors for a consecutive case series of 115 patients with somatic leiomyosarcoma. J Bone Joint Surg Am 2012;94: Song TK, Harris EJ, Jr., Raghavan S, et al.: Major blood vessel reconstruction during sarcoma surgery. Arch Surg 2009;144: Fueglistaler P, Gurke L, Stierli P, et al.: Major vascular resection and prosthetic replacement for retroperitoneal tumors. World J Surg 2006;30: