ORIGINAL ARTICLE. Adult Soft Tissue Ewing Sarcoma or Primitive Neuroectodermal Tumors

Adult Soft Tissue Ewing Sarcoma or Primitive Neuroectodermal Tumors Predictors of Survival? Robert C. G. Martin II, MD; Murray F. Brennan, MD ORIGINAL ARTICLE Background: Ewing sarcoma (ES) is the second most common primary osseous malignancy in childhood and adolescence. The improvement in survival is primarily associated with the combination of surgery and chemotherapy. Hypothesis: Little is known about the outcome of adults with soft tissue ES or primitive neuroectodermal tumors (PNET). Certain prognostic factors from soft tissue sarcomas (tumor size, tumor location, margin status, and initial presentation) in adults ( 16 years) with ES/PNET will help to identify factors associated with outcome. Methods: Between July 1, 1982, and June 30, 2000, we identified 59 adult patients with primary soft tissue ES/ PNET. Clinicopathologic factors were correlated with the end points studied: patient factors, tumor factors, pathologic factors, status of surgical margins, adjuvant chemotherapy, and radiation therapy. Results: There were 41 male and 18 female patients, with a median age of 27 years (range, 16-72 years). Median tumor size was 8 cm, with all lesions being high grade. The most common site was the trunk (n=22), with an even distribution of retroperitoneal, pelvis, buttock, and lower extremity (all n=5). The median follow-up was 29 months (range, 6-222 months), with local recurrence identified in 13 patients (22%), with a median time to recurrence of 15 months (range, 5-200 months). Overall 5-year survival was 60%. Initial presentation was the only predictor of longterm survival, with primary tumor only presentation having a 5-year survival of 60% (median not reached) compared with primary tumor plus metastatic disease having a 5-year survival of 33% (median, 17 months) (P=.02). Conclusion: Initial presentation of disease represents the only predictor of survival identified in this small group of adult patients with ES/PNET. Arch Surg. 2003;138:281-285 From the Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, NY. FROM THE first description of a diffuse small cell tumor of the bone by Dr James Ewing in 1921, 1 the diagnosis of Ewing sarcoma (ES) has been debated because of a lack of consistency in description of a certain cell of origin. The specific translocation t(11;22) (q24;q12) identified in ES 2 and primitive neuroectodermal tumors 3 (PNET) allowed for consistent criteria for the diagnosis of these tumors. These studies lead to the identification of 2 genes fused by this translocation, EWS and FLI1. 4-6 Ewing sarcoma is the second most common primary osseous malignancy in childhood and adolescence, 7 and most of the success in survival with multimodality treatment has been in that age group. Survival rates in patients with childhood nonmetastatic ES/PNET have improved from 10% in 1967 8 to 50% 9 in 2000. 10,11 The improvement in survival is primarily associated with the combination of surgery and chemotherapy. Little has been published about the outcome of adults with extraosseous (soft tissue) ES/PNET. Four studies to date, regarding both skeletal and extraskeletal adult ES/PNET, have been published evaluating survival and predictors of survival. Two studies 12,13 have demonstrated age as a poor predictor of outcome, and 2 studies 14,15 have not seen age as an adverse risk factor. The aims of this study are to present our experience with soft tissue ES/PNET in adults and to attempt to identify predictors of survival. METHODS Between July 1, 1982, and June 30, 2000, we identified 59 patients ( 16 years of age) with primary soft tissue ES/PNET from a prospective database at Memorial Sloan-Kettering Cancer Center (MSKCC). All cases were rereviewed by a senior pathologist, who confirmed that soft tissue ES/PNET had been evaluated. The primary end points that were evaluated were time to local recurrence, distal metastasis, and disease-specific mortality rate. The following clinicopathologic factors 281

Table 1. Patient Demographics Demographics No. of Patients Sex Male 41 Female 18 Age, y* 30 21 30 38 Presentation Primary 46 Primary and metastatic 7 Local recurrence 4 Metastatic 2 Tumor volume, ml 100 26 100 15 Unknown 18 Size, cm 5 15 6-10 19 10 22 Unknown 3 Anatomic site of primary Trunk 22 Buttock 5 Pelvis 5 Retroperitoneum 5 Lower extremity distal 5 Lower extremity proximal 5 Upper extremity distal 1 Upper extremity proximal 7 Visceral/genitourinary 3 Head and neck 1 Primary treatment Surgical resection 52 Limb-sparing resection 39 Amputation 4 Resection outside institution 9 Biopsy alone 7 *The median age of the patients was 27 years (range, 16-72 years). The median tumor size was 8 cm (range, 2.2-19 cm). were correlated with the end points studied: patient factors (sex, age, symptoms), tumor factors (initial presentation [primary vs recurrence] and tumor depth), pathologic factors (tumor size, tumor volume, histologic subtype, histologic grade, status of surgical margins), and treatment factors (chemotherapyrelated response and use of radiation therapy). Primary tumor was defined as a localized lesion treated with definitive surgical therapy. Tumor size was defined as the maximum diameter of the tumor with either radiographic imaging or pathologic analysis. Tumor size was classified as small ( 5 cm), intermediate (5-10 cm), or large ( 10 cm). Tumor depth was defined as superficial or deep relative to the superficial fascia. Macroscopic margins were defined at the time of surgery, and microscopic margins were defined histopathologically. Complete resection was defined as the absence of gross residual disease following surgical excision of the tumor. Local recurrence was defined as a documented tumor being within or contiguous to the previouslyexcisedfield3ormoremonthsfollowingprimarytherapy. Surgical treatment decisions were based on reviewed pathological material obtained from biopsy or prior resections, as well as size, grade, and location of the primary lesion. The goal of surgical resection was the compete removal of the primary tumor with negative gross and microscopic margins. Not all patients were re-reviewed for the genetic translocation t(11;22) (q24;q12), but since 1997, this practice has been routine. Patients who presented with metastatic disease were thoroughly evaluated with computed tomography of the chest, abdomen, and pelvis to evaluate the extent of disease. Most patients received doxorubicin hydrochloride and ifosfamide based chemotherapy, and 50% of the patients received radiation therapy. The administration of adjuvant chemotherapy was based on prognostic factors thought to predict an increased risk of local or distant disease at the discretion of the Multidisciplinary Soft Tissue Sarcoma Group at MSKCC or as part of a clinical trial. If a diagnostic biopsy specimen showed ES/PNET, then neoadjuvant chemotherapy was administered. Most patients whose preoperative diagnosis was not diagnostic of ES/PNET or whose definitive surgical resection revealed ES/PNET received postoperative adjuvant chemotherapy. Responses to chemotherapy were defined as no response ( 50% decrease in tumor size), partial response ( 50% size reduction), and complete response (based on no identifiable tumor in the resected pathological specimen). Radiation therapy was also used in the adjuvant setting based on presumed prognostic risk factors of local recurrence. Techniques used to deliver radiation therapy included external beam radiation therapy (n=24) and brachytherapy (n=3) with iridium Ir 192 seed implantation at 5 days following surgery. Total adjuvant radiation dose ranged from 3000 to 5040 cgy. Statistical analysis was performed by the 2 test for associations among categoric variables. Time to local recurrence and distant recurrence, recurrence-free survival, and disease-related mortality were all calculated from the date of diagnosis. Deaths that resulted from disease were treated as an end point for diseasespecific survival. Relapse-free survival was calculated from time of prior operation to first local or distant disease recurrence. The rate of recurrence or death was estimated using the method of Kaplan-Meier. Univariate influence of prognostic factors on study end points was analyzed using the log rank test. Statistical analysis was performed using JMP software (SAS Institute Inc, Cary, NC). P.05 was considered statistically significant. RESULTS During this period, 4677 adult patients were seen at MSKCC for evaluation of soft tissue sarcomas. Fifty-nine (1%) of these adult patients were seen for treatment of ES/PNET. There were 41 male and 18 female patients, with a median age of 27 years (range, 16-72 years) (Table 1). Median tumor size was 8 cm, with all lesions being high grade and 57 lesions being deep to the superficial fascia. The most common site was the trunk (n=22), with an even distribution of retroperitoneum (n=5), pelvis (n=5), buttock (n=5), and lower extremity (n=5) (Table 1). Most patients presented with primary disease alone (n=46, 78%), of which 3 patients presented with unresectable disease and received chemotherapy. Four patients (7%) presented with local recurrent disease alone, with 3 patients undergoing resection of the recurrence. Seven patients (12%) presented with both primary and metastatic disease, with 6 patients undergoing resection of their primary tumor. Two patients (3%) presented with metastatic disease after previously undergoing treatment of their primary tumor. Most patients (n=52, 88%) underwent complete resection of their primary tumor, with 4 requiring amputation. Chemotherapy was given to 55 patients (93%), with 27 (46%) receiving neoadjuvant therapy. All patients with recurrent disease and 42 patients (91%) with primary dis- 282

Table 2. Adjuvant Therapy Prognostic Factors No. of Patients External Beam Radiation Therapy (n = 24) Brachytherapy (n=3) Microscopic margins Positive 6 1 Negative 18 2 Size, cm 5 5 1 6-10 9 0 10 8 2 Age, y 30 13 3 30 11 0 Site of lesion Buttock 2 1 Upper extremity proximal 5 0 Upper extremity distal 1 0 Lower extremity proximal 2 1 Lower extremity distal 1 0 Pelvis 2 0 Retroperitoneum 3 0 Trunk 8 1 Chemotherapy Neoadjuvant 12 1 Adjuvant 11 2 ease received chemotherapy. In the 4 patients who did not receive chemotherapy, 2 had extremity lesions smaller than 5 cm, and 2 had lesions smaller than 10 cm and refused therapy. Three of these patients remain free of disease, with 1 patient with a retroperitoneal lesion who refused therapy dying from recurrent disease at 20 months. Five patients (8%) had a complete pathologic response, with 15 (25%) having some form of partial response. One patient demonstrated no evidence of response on pathologic evaluation; one patient did not undergo resection, and in 5 patients the pathologic analysis of response was not recorded. Radiation therapy was given in 27 patients, 18 with primary disease and 9 with recurrent disease. Most received adjuvant external beam radiation (Table 2). An even distribution of patients in regard to margin status, size of primary tumor, age, and site of primary tumor was seen in this patient population. Nine patients had microscopic positive margins, 7 received radiation therapy, and 2 patients with retroperitoneal primary tumors did not. The median follow-up was 29 months (range, 6-222 months), with local recurrences identified in 13 patients (22%) at the time of last follow-up, with a median time to recurrence of 15 months (range, 5-200 months) in these 13 patients. Differences in local recurrence were not found when age, size of primary tumor, site of primary tumor, sex, microscopic margin, type of chemotherapy (adjuvant vs neoadjuvant), or radiation therapy were evaluated. When the 27 patients who had adjuvant radiation therapy where evaluated, 8 (30%) had a local recurrence, 3 in the primary disease group and 5 in the recurrent group. Only 2 of these 8 patients had positive microscopic margins at initial resection. One patient had 0 20 40 60 80 100 120 140 160 Figure 1. Survival by presentation (P=.03). primary and the other had local recurrence at initial presentation. In the 32 patients who did not undergo adjuvant radiation therapy, there were 5 local recurrences (16%). Radiation therapy did not significantly reduce local recurrence in this nonrandomized study. Distant disease recurrences were identified in 15 patients (25%) at the time of last follow-up, with a median time to distant recurrence of 11.5 months (range, 6-68 months). The most common site of distant recurrence was lung (n=8) followed by bone (n=3). Prognostic factors were evaluated for predictors of distant recurrence and age, size, site of primary tumor, sex, microscopic margin, radiation therapy, or type of chemotherapy (adjuvant vs neoadjuvant) and were not predictive of distant recurrence. The overall 5-year survival rate in this group of patients was 60%. The overall disease-free interval in this patient population was 24.4 months. Initial presentation was the only predictor of long-term survival, with primary tumor only presentation having a 5-year survival rate of 60% (median not reached) compared with primary tumor with metastatic disease having a 5-year survival rate of 33% (median, 17 months) (P=.02) (Figure 1). When other prognostic factors (age, sex, size, tumor volume [ 100 vs 100 ml], microscopic margin, radiation therapy, and type of chemotherapy) were evaluated, none were predictive of outcome. All 3 tumor size criteria ( 5 cm, 6-10 cm, and 10 cm) demonstrated similar censored events, with similar long-term survival after 40 months (Figure 2). This was also seen when tumor volume was evaluated (Figure 3). The timing of initial chemotherapy also demonstrated similar survival (Figure 4). When the predictors of survival were evaluated in only the 46 patients who presented with primary disease, the results were similar. Evaluation of all prognostic factors failed to identify any factor associated with survival outcome. COMMENT Primary Tumor Only (n = 46) Metastatic Disease (n = 9) Classically, ES/PNET has been described as a highly aggressive neoplasm that occurs most commonly in children and young adults. 7 Long-term survival has been 283

5 cm (n = 15) 6-10 cm (n = 19) >10 cm (n = 22) Neoadjuvant Chemotherapy (n = 27) Adjuvant Chemotherapy (n = 28) 0 20 40 60 80 100 120 140 160 0 50 100 150 200 250 Figure 2. Survival by tumor size (P=.35). 0 10 20 30 40 50 60 70 80 90 100 Figure 3. Survival by tumor volume (P=.76). <100 ml (n = 26) >100 ml (n = 15) achieved with both aggressive local and adjuvant therapy in the form of multidrug chemotherapy. This rare soft tissue sarcoma in the adult patient (median age, 27 years) had an overall 5-year survival rate of 60%. This report also demonstrates an outcome similar to the 2 most recent reports of ES/PNET in the adult population (Table 3), although in these prior reports, the combined number of soft tissue adult ES/PNET was small (n=27). The trunk was the most common site of primary disease in this study of soft tissue ES/PNET; however, there was an even distribution of all other sites (Table 1). Site of primary tumor was not a predictor of survival primarily because of small numbers but was also not a predictor of local recurrence or distant recurrence. In evaluating the other reported cases of soft tissue ES/PNET in the adult population, no site predominated. 12-15 Although ES/ PNET is uncommon, the trunk may be predictive of outcome; however, the small numbers prevent any site of initial presentation from being predictive of outcome. Figure 4. Survival by timing of initial chemotherapy (P=.88). Tumor volume, which has been demonstrated to be a significant factor of prognosis in a large cooperative study in children 16 and a recent study in adults, 15 was also evaluatedinthisstudy.tumorvolumewasnotapredictorofoverall survival in this patient population (Figure 3). The primary reason for the lack of difference between the 2 groups is the small number of patients who had a tumor volume greater than 100 ml (n=15). The number of patients with tumor volume greater than 100 ml is far less than the number in the only study 15 in adults in which this was a predictor of outcome. Because the number of ES/PNET cases that are diagnosed increases with the standardization in pathologic evaluation, tumor volume may become a predictor of outcome in the adult population, although one would expect smaller volumes with earlier diagnosis. The size of the primarytumor, whichhasbeenareproduciblefactorinother soft tissue sarcomas, was also not a predictor in this group of patients (Figure 2). Again, the small number of patients and the favorable overall outcome in these patients may prevent size from being a predictor of outcome. Radiation therapy, which has been demonstrated to successfully prevent local recurrence in other soft tissue sarcomas, was used in 27 patients. The small number of patients who received radiation therapy and the limited rate of recurrence in these patients prevent us from effectively examininganyfactorsthatwoulddemonstrateabenefitfrom radiation therapy. This patient population demonstrated a local recurrence of 17% (10/59). This small local recurrence rate further emphasizes that the use of adjuvant radiation therapy should be used judiciously in this patient populationbecauseofalackofdefinitivebenefitinthisstudy. Multidrug chemotherapy has been the mainstay of therapy in ES/PNET, with most reports using an induction chemotherapy approach. With the older age group, the soft tissue presentation, and the uncommon diagnosis in adults, most of the patients in this study did not receive induction chemotherapy. In addition, the variations in chemotherapy treatment in this population evolved during the study, with no set treatment protocol in this small subset of adult soft tissue sarcomas. With this variance in timing, we were able to evaluate the sur- 284

Table 3. Reported Cases of Adult Ewing Sarcoma or Primitive Neuroectodermal Tumors Source, y Years of Study No. of Patients Primary Tumor Only LR DR DSS OS Primary Tumor Sinkovics et al, 12 1980 NS 50 34 2 18 27 mo 29 mo Bone (n = 43), soft tissue (n = 7) Siegel et al, 13 1988 1975-1985 16 11 3 5 10 mo 34 mo Bone (n = 11), soft tissue (n = 5) Verrill et al, 15 1997 1988-1994 34 33 NS NS 5 y (39%) 5 y (63%) Bone (n = 13), soft tissue (n = 21) Bacci et al, 14 2000 1983-1995 23 23 4 6 5 y (53%) 5 y (59%) Bone (n = 21), soft tissue (n = 2) Present study 1983-2000 59 46 13 15 24.4 mo 5 y (59%) Soft tissue (n = 59) Abbreviations: DR, distant recurrence; DSS, disease-specific survival; LR, local recurrence; NS, not stated; OS, overall survival. vival differences in the 27 patients receiving neoadjuvant chemotherapy and the 28 patients receiving adjuvant chemotherapy. The neoadjuvant chemotherapy group had 5 patients with pathologically confirmed complete responses, and all 5 patients have remained free of disease, with a median follow-up of 36 months. The remaining 22 patients had varying response rates but demonstrated no difference in survival when compared with patients who did not receive neoadjuvant therapy. A few small series have confirmed the need for aggressivelocaltherapyandmultidrugadjuvanttherapy, withvarying degrees of long-term survival (Table 3). The first report ofadulteswaspresentedbysinkovicsetal. 12 Theydescribed 50 patients with a mean age of 21 years (range, 16-36 years), with most patients presenting with primary disease in bone. Allpatientsreceivedchemotherapy, witha27-monthdiseasefree interval of 35% in the patients with primary disease only and an overall survival rate of 19% in patients who presented with metastatic disease. Their outcome represents the worst overall survival in this age group and was thought to be directly related to the age of the patients. Siegel et al 13 described 16 patients with a median age of 22 years (range, 17-50 years), with most presenting with primary disease in bone (n=11) (Table 3). All patients were treated with chemotherapy, and most received radiation therapy. This group obtained a median survival of 34 months. Again, age of the patients was believed to be the reason for the poor overall survival. Verrill et al 15 were the first to evaluate pediatric chemotherapy protocols in adult patients. This study included 34 patients with a median age of 23 years (range, 16-48 years), with most having disease within the soft tissue (n=21). They demonstrated a much improved 5-year survival rate of 63% and found that only initial presentation (primary vs metastatic) was a predictor of outcome. Age was not a predictor of poor outcome in these patients when compared with the published pediatric series. Bacci et al 14 confirmed this survival success in adults with their report of 23 cases of ES of the bone in patients older than 39 years. All patients received neoadjuvant chemotherapy, with a 5-year survival rate of 53%. Chemotherapy response of the primary tumor was not reported or evaluated, but these investigators found no survival disadvantage in this older patient population when compared with ES that occurred in the pediatric population. The only reliable factor for long-term survival that has been seen in all series is the extent of disease at presentation. Patients who present with metastatic disease at their initial presentation have a far worse outcome than patients with primary disease alone. The favorable overall survival seen in this study is equivalent to 2 recently published studies 14,15 (Table 3) in which aggressive multimodality chemotherapy was used. In addition, the site of primary ES/PNET (bonevssofttissue) hasalsobeendemonstratedtobeequivalent in regard to disease-free survival and overall survival when this study of soft tissue ES/PNET is compared with the other reports of adult ES/PNET, which involved most patients with ES/PNET of bone (Table 3). With the pathologic standardization of ES/PNET, there hasbeenagreaterdetectionofthisdiseaseintheadultpopulation. Multidrug chemotherapy based on the pediatric protocols has become the standard therapy, with an identified long-term survival in the small number of patients who had a complete pathologic response. Primary site, size, and age older than 16 years were not adverse predictors of survival, with initial presentation of disease the only predictor of survival identified in this small group of patients. Accepted for publication October 25, 2002. Corresponding author and reprints: Murray F. Brennan, MD, Department of Surgery, Memorial Sloan-Kettering Cancer Center, 1275 York Ave, New York, NY 10021. REFERENCES 1. Ewing J. Diffuse endothelioma of bone. Proc N Y Pathol Soc. 1921;21:17-24. 2. Aurias A, Rimbaut C, Buffe D, Dubousset J, Mazabraud A. Chromosomal translocations in Ewing s sarcoma. N Engl J Med. 1983;309:496-498. 3. Whang-Peng J, Triche TJ, Knutsen T, Miser J, Douglass EC,Israel MA.Chromosome translocation in peripheral neuroepithelioma. N Engl J Med. 1984;311:584-585. 4. Dierick AM, Roels H, Langlois M. The immunophenotype of Ewing s sarcoma: an immunohistochemical analysis. Pathol Res Pract. 1993;189:26-32. 5. Ladanyi M, Bridge JA. Contribution of molecular genetic data to the classification of sarcomas. Hum Pathol. 2000;31:532-538. 6. Llombart-Bosch A, Carda C, Peydro-Olaya A, et al. Soft tissue Ewing s sarcoma: characterization in established cultures and xenografts with evidence of a neuroectodermic phenotype. Cancer. 1990;66:2589-2601. 7. Percy C, Young JL Jr, Muir C, et al. Histology of cancer: SEER population-based data, 1973-1987 introduction. Cancer. 1995;75:140-146. 8. Falk S, Alpert M. Five-year survival of patients with Ewing s sarcoma. Surg Gynecol Obstet. 1967;124:319-324. 9. Bacci G, Ferrari S, Bertoni F, et al. Prognostic factors in nonmetastatic Ewing s sarcoma of bone treated with adjuvant chemotherapy. J Clin Oncol. 2000;18:4-11. 10. Hayes FA, Thompson EI, Meyer WH, et al. Therapy for localized Ewing s sarcoma of bone. J Clin Oncol. 1989;7:208-213. 11. Nesbit ME Jr, Gehan EA, Burgert EO Jr, et al. Multimodal therapy for the management of primary, nonmetastatic Ewing s sarcoma of bone: a long-term follow-up of the First Intergroup study. J Clin Oncol. 1990;8:1664-1674. 12. Sinkovics JG, Plager C, Ayala AG, Lindberg RD, Samuels ML. Ewing sarcoma: its course and treatment in 50 adult patients. Oncology. 1980;37:114-119. 13. Siegel RD, Ryan LM, Antman KH. Adults with Ewing s sarcoma: an analysis of 16 patients at the Dana-Farber Cancer Institute. Am J Clin Oncol. 1988;11:614-617. 14. Bacci G, Ferrari S, Comandone A, et al. Neoadjuvant chemotherapy for Ewing s sarcoma of bone in patients older than thirty-nine years. Acta Oncol. 2000;39:111-116. 15. Verrill MW,Judson IR, Wiltshaw E, Thomas JM,Harmer CL, Fisher C. The use of paediatric chemotherapy protocols at full dose is both a rational and feasible treatment strategy in adults with Ewing s family tumours. Ann Oncol. 1997;8:1099-1105. 16. Paulussen M, Ahrens S, Dunst J, et al. Localized Ewing tumor of bone. J Clin Oncol. 2001;19:1818-1829. 285