RECURRENCE PATTERNS AND SURVIVAL FOR PATIENTS WITH INTERMEDIATE- AND HIGH-GRADE MYXOFIBROSARCOMA

doi:10.1016/j.ijrobp.2010.08.042 Int. J. Radiation Oncology Biol. Phys., Vol. 82, No. 1, pp. 361 367, 2012 Copyright Ó 2012 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/$ - see front matter CLINICAL INVESTIGATION Sarcomas and Pediatric Cancer RECURRENCE PATTERNS AND SURVIVAL FOR PATIENTS WITH INTERMEDIATE- AND HIGH-GRADE MYXOFIBROSARCOMA KARL E. HAGLUND, M.D., PH.D.,* 1 CHANDRAJIT P. RAUT, M.D., M.SC., y{ ALESSANDRA F. NASCIMENTO, M.D., z QIAN WANG, PH.D., { SUZANNE GEORGE, M.D., x{ AND ELIZABETH H. BALDINI, M.D., M.P.H. k{ *Harvard Radiation Oncology Program and Departments of y Surgery, z Pathology, x Medical Oncology, and k Radiation Oncology, Brigham and Women s Hospital, Boston, MA; and { Center for Sarcoma and Bone Oncology, Dana-Farber Cancer Institute, Boston, MA Purpose: Myxofibrosarcoma (MFS) is a rare sarcoma with a predilection for multiple local recurrences (LR), for which optimal treatment has not been defined. We reviewed our experience to determine the impact of surgery and radiation therapy (RT) on pattern of recurrence, limb salvage, and overall survival (OS). Methods and Materials: Between 1995 and 2005, 36 patients with localized intermediate- or high-grade MFS were treated at our institution. Data on clinicopathologic features, treatments, and patient outcomes were reviewed and analyzed. Results: Median age was 72.5 years (range, 42 96 years). Median tumor size was 7.5 cm, and 34 tumors (94%) were high grade. All patients underwent surgery at our institution, including re-resections in 20 patients (56%) after initial surgery elsewhere. Margins were microscopically positive in 9 patients (25%). RT was given to 28 patients (78%) pre - and/or postoperatively. After a median follow-up of 3.5 years (range, 0.4 12.4 years), 11 patients (31%) developed LR. There were no significant predictors for LR on univariate analysis, including margin status or use of RT. Limb salvage was ultimately achieved in only 5 of 11 LRs (45%) because of multiple subsequent LRs. Distant recurrence (DR) occurred in 6 patients (17%). Median and 4-year OS were 96 months and 65%, respectively. Seven patients (19%) died of tumor-related causes, 6 of whom had DRs. On univariate analysis, tumor size was associated with OS. Conclusions: Despite aggressive surgery and RT, intermediate- and high-grade MFS are associated with a high rate of LR that adversely affects limb preservation. More aggressive local treatment strategies are necessary. Ó 2012 Elsevier Inc. Myxofibrosarcoma, surgery, radiation therapy, amputation, extremity sarcoma. INTRODUCTION Myxofibrosarcoma (MFS) is a relatively rare histologic subtype of soft tissue sarcoma (STS), but is a common sarcoma in the extremities of elderly patients (1, 2). Formerly classified as the myxoid variant of malignant fibrous histiocytoma (3), MFS was recognized as a pathologic entity distinct from malignant fibrous histiocytoma in the 2002 World Health Organization (WHO) criteria (4). An unfortunate clinical feature of MFS is its propensity for multiple, relentless local recurrences (LRs), even with low grade lesions (1 3, 5). Low-grade lesions have relatively low metastatic potential, but intermediate- and high-grade lesions demonstrate incrementally increasing rates of distant metastases (1, 2, 6). In addition, 15% to 38% of LRs of MFS progress to higher histologic grade with an attendant increase in metastatic potential (1, 2, 6 8). The current standard of care for local treatment of extremity STS favors limb preservation over amputation (9). Most patients with extremity STS undergo wide or radical resection combined with preoperative or postoperative radiation therapy (RT). Using this approach, local control rates of 80% or greater are often achieved in modern series (9 21) This general approach to managing extremity STS has also been applied to MFS in particular. However, given reported LR rates in MFS of 55% to 65% (1 3, 8), it remains unclear whether this standard approach for MFS is optimal (22). Because MFS is relatively uncommon, there are no large series or randomized trials to guide treatment Reprint requests to: Elizabeth H. Baldini, M.D., M.P.H., Department of Radiation Oncology, Brigham and Women s Hospital, 75 Francis Street ASB1-L2, Boston, MA 02115. Tel: (617) 732-7906; Fax: (617) 264-5242; E-mail: ebaldini@lroc.harvard.edu Presented in part at the Cutaneous Tissue Oncology Society Annual Meeting, November 5-7, 2009 Miami, FL. 361 1 DR. Haglund is currently at the National Institutes of Health, National Cancer Institute, Radiation Oncology Branch, Bethesda, MD. Conflict of interest: none. Received May 5, 2010, and in revised form Aug 17, 2010. Accepted for publication Aug 17, 2010.

362 I. J. Radiation Oncology d Biology d Physics Volume 82, Number 1, 2012 decisions. We reviewed our experience with patients with localized intermediate- or high-grade MFS treated at the Dana-Farber/Brigham and Women s Cancer Center (DF/ BWCC) to evaluate patterns of recurrence, limb salvage, and survival rates. METHODS AND MATERIALS Study cohort After approval from the Dana-Farber Cancer Institute Institutional Review Board, all 109 patients with MFS 18 years or more of age and evaluated at DF/BWCC between 1995 and 2005 were identified from the Brigham and Women s Hospital (BWH) Department of Pathology diagnostic database. Of those, 32 cases were pathology consultations only and were excluded. An additional 37 patients who did not return for scheduled clinical follow-up at DF/BWCC for more than 4 months after diagnosis were excluded. Finally, 3 patients presenting with locally recurrent disease and 1 patient with low-grade MFS were excluded, because our aim was to report on outcomes for patients with intermediate- and highgrade primary MFS. The remaining 36 patients had primary, localized, intermediate- or high-grade MFS with minimum follow-up of 4 months, and constituted the study cohort. Data collection Clinicopathologic and follow-up data were collected for the study cohort. Data recorded included subject demographics, age at diagnosis, site of primary tumor, type of definitive surgery (wide local excision, radical resection, amputation), invasive procedures before definitive surgery (biopsy, enucleation, piecemeal or marginal excision), maximal tumor dimension, tumor depth, microscopic surgical margin status, closest margin, tumor grade, American Joint Committee on Cancer (AJCC), 6th Edition, staging (23), use of preoperative and/or postoperative RT, radiation dose in Gray (Gy), use of preoperative and/or postoperative chemotherapy, duration of follow-up, number and timing of LRs and their treatment, number and timing of distant recurrences (DRs) and their treatment, and date of death. Histopathology In all 36 cases, the pathologic diagnosis of MFS was made by experienced sarcoma pathologists, including 1 of the authors (A.N.). In general, the diagnosis of MFS was based on the presence of (1) alternating hypocellular, myxoid areas and hypercellular, fibrous areas; (2) pleomorphism; (3) curvilinear vessels, most readily identified in the myxoid areas; and (4) aggregation of neoplastic cells around curvilinear vessels. Immunostains were used to exclude other entities that might mimic MFS. A three-tiered system was used to assign MFS grade. Intermediate-grade MFS had solid areas present in varying degrees, no necrosis, and moderate nuclear pleomorphism. High-grade lesions had a prominent solid component (sometimes with very few myxoid areas), variable amounts of necrosis, and pronounced nuclear atypia or pleomorphism. Treatment All subjects in the study cohort had definitive surgery at DF/ BWCC by experienced sarcoma surgeons. Limb-sparing, function-sparing surgery was attempted in all patients. However, 1 patient required amputation as part of the definitive initial operation. For the remaining patients, wide excision with a minimum of a 1-cm radial margin extending to and including the fascia of the underlying muscle was attempted. Some patients required more radical composite tissue resections as deemed appropriate by the surgeon. In all, 28 patients received RT. For the 14 patients treated with external beam RT at DF/BWCC, patients underwent CT simulation, and RT was administered using a three-dimensional conformal technique. Typical RT fields included 5-cm proximal and distal margins and 2- to 3-cm radial margins, unless there was an anatomic boundary such as bone or an intact fascial plane. Patients were treated with 1.8- or 2-Gy fractions to total doses of 50 Gy preoperatively and 63 to 66 Gy postoperatively. Three patients received low-dose rate iridium-192 brachytherapy boost at the time of resection to a dose of 18 to 20 Gy. Data analysis Data were analyzed for associations between clinicopathologic characteristics and survival outcomes by univariate analysis. Local recurrence-free survival (LRFS) and overall survival (OS) were estimated using the Kaplan Meier method. LRFS interval was defined as the time between diagnosis and the first LR or last follow-up, whichever came first. OS interval was defined as the time between diagnosis and death or last follow-up, whichever came first. Median survival times estimated by the Kaplan Meier method are reported with 95% confidence intervals (CI). Comparisons of the difference in distributions were assessed by the log-rank test. RESULTS Subjects Between 1995 and 2005, 36 adult patients with intermediate- or high-grade primary, nonmetastatic MFS underwent definitive treatment at the DF/BWHCC and had at least 4 months of follow-up. Table 1 summarizes the demographic data and tumor characteristics at presentation. Median follow-up was 3.5 years (range, 0.4 12.4 years). As expected, the most common site of presentation was in the lower extremities (n = 22, 61%). There were two (6%) intermediate-grade and 34 (94%) high-grade tumors. No lymph node metastases were noted on initial presentation. Treatment Details of treatment are shown in Table 2. All subjects underwent definitive resection of their primary tumor at DF/BWCC in the form of a wide local excision (n =8, 22%), radical resection (n = 27, 75%), or amputation (n = 1, 3%). Amputation of the hand and distal forearm was performed in 1 patient as part of the initial definitive resection, as detailed below. In all, 20 patients (56%) had invasive procedures at outside institutions before definitive surgery at DF/BWCC. These procedures included open biopsy (n = 9) and marginal resection (n = 11). Radiation therapy was administered to 28 patients (78%), either preoperatively (n = 5), postoperatively (n = 19), or both (n =4;Table 2). Of the patients, 14 received RTat DF/BWCC and 14 at outside institutions. Radiation doses were known for 2 of 5 patients who received preoperative only RT, and were 50 Gy and 50.5 Gy, respectively. RT doses were known

Recurrence patterns and survival in myxofibrosarcoma d K. E. HAGLUND et al. 363 Table 1. Characteristics of patients with localized myxofibrosarcoma (n = 36) Median age at diagnosis (y) 72.5 (range, 42 96) Sex n (%) Male 21 (58%) Female 15 (42%) Primary site Lower extremity 22 (61%) Proximal 16 (44%) Distal 6 (17%) Upper extremity 10 (28%) Proximal 4 (11%) Distal 6 (17%) Trunk 4 (11%) Grade High 34 (94%) Intermediate 2 (6%) AJCC T stage T1a 8 (22%) T1b 3 (8%) T2a 1 (3%) T2b 24 (67%) Stage grouping II 12 (33%) III 24 (67%) Largest tumor dimension (cm)* 7.5 (range, 0.7 22.0) Abbreviation: AJCC = American Joint Committee on Cancer. * Tumor dimension not known for 1 patient. for 10 of 16 patients who received postoperative RT without brachytherapy. The median dose was 63 Gy (range, 59.4 70 Gy). Among 3 patients who received postoperative RT and brachytherapy, the total combined doses were known for 2 and were 59.8 and 63.6 Gy. Finally, the RT dose was known for all 4 subjects who received both preoperative and postoperative RT, and the median combined preoperative and postoperative dose was 64 Gy (range, 63 66.2 Gy). Among the 8 patients (22%) who did not receive RT, 1 patient developed LR within 2 months of resection before postoperative radiation could begin, 2 patients had no residual tumor on re-excision, and 5 patients had a final median histologic margin of 1.5 cm (range, 0. -3.5 cm). Chemotherapy was administered to only 1 patient as part of the initial definitive treatment. Two cycles of preoperative mesna, doxorubicin, and ifosfamide were administered for a high-grade, 22-cm primary MFS of the thigh before radical resection and low-dose-rate brachytherapy and external beam RT. Local recurrences LRs were observed in 11 (31%) patients. Median time to first LR was 21 months (range, 6 48 months). The 2- and 4- year actuarial LRFS rates were 75% and 60%, respectively (Fig. 1). LRs were observed in 7 of 27 patients (26%) with margin-negative resections, and in 4 of 9 patients (44%) with final positive microscopic margins (p = 0.30). Among patients with negative microscopic margins, 6 of 15 patients with margins less than 1 cm had LR, whereas none of the 6 patients with margins 1 cm or greater had LRs. The closest margin was unknown in 6 patients, 1 of whom had LR. Table 2. Initial treatment for patients with localized myxofibrosarcoma (n = 36) Treatment n (%) Surgery 36 (100%) Initial surgery at DF/BWHCC 16 (44%) Initial surgery at an outside hospital 20 (56%) Open biopsy elsewhere 9 (25%) Marginal resection elsewhere 11 (31%) Definitive surgery performed at 36 (100%) DF/BWCC Wide local excision 8 (22%) Radical resection 27 (75%) Amputation 1 (3%) Radiation therapy 28 (78%) Preoperative only 5 (14%) Median preoperative dose (Gy)* 50 Gy (50 50.5) Postoperative only 19 (53%) Median postoperative dose (Gy) y 62.5 (range, 34.2 70) Both preoperative and postoperative 4 (11%) doses Median combined dose (Gy) z 64 (range, 63 66.2) Chemotherapy 1 (3%) Abbreviation: DF/BWCC = Dana-Farber/Brigham and Women s Cancer Center. * Median dose and range were calculated from the available preoperative radiation therapy dose data for 2 of 5 patients. y Median dose and range were calculated from the available postoperative radiation therapy dose data for 12 of 19 patients. z Median dose and range were calculated from the available combined preoperative and postoperative dose for all 4 patients. LRs were observed in 10 of 28 patients (36%) who received RTand in 1 of 8 (13%) who did not (p = 0.21); the latter developed a rapid LR before RT could be administered. This patient underwent radical resection of a high-grade, 18-cm thigh MFS with negative microscopic margins, but developed concurrent thigh LR and ipsilateral groin regional recurrence (RR) 2 months after surgery. The mean and median tumor sizes of those who neither received RT nor developed LR were 5.0 and 4.5 cm (range, 1.5 9.7 cm). Six of 7 were high grade, and all had negative surgical margins. Fig. 1. Kaplan Meier plot of local recurrence free survival as a function of time (solid line). Dashed lines indicate 95% confidence intervals.

364 I. J. Radiation Oncology d Biology d Physics Volume 82, Number 1, 2012 Among the 10 patients who developed LR and had received RT, 4 received RT at DF/BWCC and 6 received RT at outside institutions. Review of RT treatment films for those treated at DF/BWCC revealed that LRs occurred within the irradiated field. Among these 4 patients, 2 patients had microscopically positive margins, and 2 had closest margins of 0.1 and 0.3 cm, respectively. Treatments for first LRs in 11 patients included amputation (n = 1), radical resection alone (n = 3), external beam RT alone (n = 2), radical resection with brachytherapy (n = 2), external beam RT followed by radical resection and chemotherapy (n = 1), and chemotherapy followed by radical resection (n = 1). One patient died of unknown causes the day after being diagnosed with LR. Univariate analysis did not reveal any statistically significant association between the development of LR and age, tumor grade, maximal initial tumor dimension, AJCC stage, type of surgery (radical vs. wide excision), re-excision vs. initial excision at BWH, administration of RT, timing of RT, tumor location, and surgical margin status. Multiple local recurrences and limb preservation Among the 11 patients who developed a LR, 6 patients had a single LR, 3 patients had 2 LRs, 1 patient had three LRs, and 1 patient had four LRs (Table 3). One of 6 patients with a single LR required amputation. Ultimately, all 5 patients who had two or more LRs required amputation. Therefore, the limb preservation rates were 45% (5/11) among patients developing LR and 83% (30/36) for the entire cohort. Regional recurrences Two patients who developed LRs also experienced RRs (Table 3). In both cases, the primary tumor site was the thigh, and the RR was in the ipsilateral groin. In one case, amputation was performed before a subsequent groin recurrence. In the other case, radical resection was performed for the LR with concurrent groin dissection for the RR. Pathologic examination revealed that neither of the two RRs appeared to occur in lymph nodes. Although complete replacement of nodal tissue remains a possibility, MFS typically does not metastasize to lymph nodes, and soft tissue metastases were the presumed diagnoses. Not included in the statistical analysis of LRs was the patient described above with a 5.3-cm MFS of the hand, whose initial surgery at DF/BWCC was amputation with negative margins. This patient had multiple RRs in the absence of any LRs, the first of which occurred in the ipsilateral axillary adipose tissue 11 months after initial surgery. The second RR occurred 4 months after wide excision and postoperative doxorubicin. Including this case in the analysis yields a total locoregional recurrence rate of 33% (12/36). Distant recurrences and overall survival Six of 36 patients (17%) developed metastatic disease with a median time to first DR of 21 months (range, 4.8 97 months). The metastatic sites included lung (n = 5), bony pelvis (n = 2), liver (n = 1), xiphoid (n = 1), and brain (n = 1). Median DR-free survival was not achieved. Two of six DRs occurred after a preceding LR, whereas four of six DRs occurred without a preceding LR. During the study period, 15 of 36 patients (42%) died. The 2- and 4-year actuarial OS rates were 82% and 65%, respectively (Fig. 2). Median OS was 96 months (52 149 months, 95% confidence interval [CI]). The cause of death was disease progression in 7 patients, other non-mfs etiologies in 4 patients, and unknown in 4 patients. Univariate analysis for clinicopathologic factors associated with OS revealed a statistically significant association with maximal initial tumor dimension. The 4-year OS was Table 3. Characteristics of the 11 patients who developed local recurrence No of LRs y No of RRs z Primary site Maximal tumor dimension Grade/ T stage Margin status Adjuvant treatment Current limb status Metastatic status Survival status (Time [mo]) x 1 1 Thigh 22 cm High/T2b Negative None Intact Metastatic DOD (20 m.) Disease 1 Thigh * High/T2b Positive Pre/Postop. Amputation D-Other (114 m.) RT 1 Forearm 3 cm High/T1a Negative Postop. RT Intact D-Other (75 m.) 1 Forearm 8 cm High/T2b Negative Postop. RT Intact Alive (75 m.) 1 Arm 9 cm High/T2b Negative Preop. RT Intact D-Unk (33 m.) 1 Thigh 16 cm High/T2b Negative Postop. RT Intact D-Unk (9 m.) 2 1 Thigh 22 cm High/T2b Positive Postop. RT Amputation Metastatic DOD (44 m.) Disease 2 Forearm 2.5 cm High/T1a Positive Postop. RT Amputation D-Unk (96 m.) 2 Leg 7.5 cm High/T2b Negative Postop. RT Amputation D-Unk (35 m.) 3 Forearm 6 cm High/T2a Negative Postop. RT Amputation Alive (88 m.) 4 Thigh 19 cm High/T2b Positive Postop. RT Amputation Alive (149 m.) Abbreviations: DOD = dead of disease; D-Other = dead of other causes; D-Unk = dead of unknown causes. * Maximal tumor dimension unknown for one subject. y LR, local recurrence. z RR, regional recurrence. x For Survival Status, Time refers to the interval in months between diagnosis and death or date of last follow-up.

Recurrence patterns and survival in myxofibrosarcoma d K. E. HAGLUND et al. 365 Fig. 2. Kaplan Meier plot of overall survival as a function of time (solid line). The dashed lines indicate the 95% confidence intervals (upper confidence interval does not reach a median value). 85% for patients with a maximal tumor dimension less than 7.5 cm (n = 17) and 47% for patients with a maximal tumor dimension of 7.5 cm or greater (n = 18; p = 0.04; Fig. 3). There was a trend toward better OS with lower stage grouping Actuarial 4-year OS was 100% for patients with Stage II disease (n = 11) and 52% for those with Stage III disease (n = 25; p = 0.06). DISCUSSION We report here outcomes after treatment of intermediateand high-grade MFS at our institution. It is important to reiterate that although myxofibrosarcoma (MFS) and malignant fibrous histiocytoma (MFH) were previously grouped together, there is now clear recognition that the two pathologic diagnoses demonstrate different clinical behaviors (3). Fig. 3. Kaplan Meier plot of overall survival by tumor size. Solid line indicates overall survival for patients with primary tumors less than 7.5 cm in maximal dimension (n = 17). Dashed line corresponds to overall survival for patients with primary tumors 7.5 cm or greater in maximal dimension (n = 18; p = 0.04). We found a crude LR rate of 30% and an actuarial 4-year LR rate of 40%. These high LR rates are similar to those reported by other investigators, but are higher than those reported for other histologic subtypes of STS (1, 3, 5 8, 22). Studies that included all MFS tumor grades have reported LR rates of 44 61% (2, 3, 8). Angervall et al. (1) reported that 6 of 10 patients with Grade III/IV MFS and 7 of 11 patients with Grade IV/IV MFS experienced LR after surgery alone, for an overall LR rate of 62%. A more contemporary series reported by Lin et al. (8) showed an LR rate of 44%. Limited data on the use of RT is included in these studies. We also observed RRs arising some distance from the primary site of disease in 3 patients (8%), often in lymph node basins. We are unable to identify any other series specifically reporting RRs, although investigators may have grouped such recurrences with LRs in their analyses. Among the patients in this study who developed any LR (11/36), approximately half (5/11) developed multiple LRs. Focusing primarily on low-grade MFS, Huang et al. (5) reported that of 28 LRs, 15 (54%) developed multiple recurrences. Similarly, 27 of 61 patients with MFS described by Lin et al. (8) developed LRs, 13 of whom (48%) went on to develop multiple LRs. Mentzel et al. (2) described a similar rate of second or subsequent LRs of 42%. Therefore, it appears that approximately half of patients with MFS who develop a LR will ultimately develop two or more subsequent LRs. Our data suggest that this group of patients who develop multiple LRs will be more likely to require amputation. In the initial randomized trial of limb-sparing therapy for multiple histologic types of STS reported by Rosenberg et al. (9), the rate of salvage amputation was not reported, but only 4 of 27 patients (15%) in the limb-sparing group experienced LR. All of those had pathologically positive surgical margins. In our cohort, amputation was ultimately required after attempted limb preservation in 6 of 11 patients with LRs, only 4 of whom had positive surgical margins (Table 3). All patients with multiple LRs ultimately required amputation. Only 1 of 8 patients who did not receive RT developed LR. Therefore, comparisons could not be performed from our data to determine whether RT could be safely omitted in certain patients. However, several studies have identified associations between surgical margin status in MFS and LR, as well as with OS (6) and disease-specific survival (8). Like a number of other investigators (2, 3, 5), we were unable to identify any clinicopathologic characteristics associated with LR. However, it is interesting to note that none of the 6 patients with resection margins >1 cm developed LRs. It is not surprising that positive or close resection margins would be associated with an increased risk of LR for MFS, as the association between margin status and LR has been well documented in other sarcoma subtypes (9, 12, 21, 24 30). No association between LR and administration of RT was found, which is not surprising, given that the majority (78%) of patients in this series received RT. Although RT treatment

366 I. J. Radiation Oncology d Biology d Physics Volume 82, Number 1, 2012 films were available for only 4 of 10 patients with LR who received RT, review of those films suggests that the recurrences were within the radiation treatment fields. In our cohort, maximal tumor dimension greater than 7.5 cm (the median size in this series) was associated with worse OS. This association of increased tumor size and worse survival is well documented for STS (28, 31, 32). We also identified a trend (p = 0.06) toward worse survival with higher AJCC stage. Other factors reported to be associated with worse survival in MFS include tumor grade (8) and mitotic rate (8), or, more specifically, the presence of more than 20 mitoses per high-powered field (5). Distant metastasis free survival was reported to be associated with mitotic activity and margin status in one study (8), with a second study confirming an association with tumor necrosis (5). Six of 36 (17%) of our patients developed metastatic disease during the follow-up period, which is similar to findings reported by others (2, 5, 8). A larger cohort may have revealed statistical associations that were not captured here. The retrospective nature of our study also suggests caution in interpretation of the results. Nevertheless, the overall outcomes that we observed are noteworthy and in agreement with the observations of others. Specifically, compared with other STS histologies, intermediate- and high-grade MFS is associated with a higher rate of LR, multiple LRs, and ultimate need for amputation. Of greatest concern is the observed MFS amputation rate of 17%. By contrast, the overall amputation rate for STS of all histologies is only about 5% (33, 34). Achieving a high rate of limb preservation in patients with extremity MFS who experience LR is challenging, emphasizing the importance of optimizing multimodal therapy for LC at initial presentation. Because of the strong propensity for intermediate- and high-grade MFS to develop local recurrences, about half of which are multiple, we feel that initial aggressive local therapy is imperative. Furthermore, the prevention of a LR may increase the probability of limb preservation. We believe that surgery should be performed with a goal of attaining greater than 1-cm resection margins whenever possible and that RT should be considered for all patients diagnosed with intermediate- or high-grade MFS. Future studies should address strategies to intensify local treatment, such as radiation dose escalation and the potential role of radiation sensitizers and/or systemic chemotherapy. REFERENCES 1. Angervall L, Kindblom LG, MerckMyxofibrosarcoma C. A study of 30 cases. Acta Pathol Microbiol Scand [A] 1977; 85A:127 140. 2. Mentzel T, Calonje E, Wadden C, et al. Myxofibrosarcoma. Clinicopathologic analysis of 75 cases with emphasis on the low-grade variant. Am J Surg Pathol 1996;20: 391 405. 3. Weiss SW, Enzinger FM. Myxoid variant of malignant fibrous histiocytoma. Cancer 1977;39:1672 1685. 4. Mentzel T, van den Berg E, Molenaar W. Myxofibrosarcoma. In: Fletcher C, Unni K, Mertens F, editors. WHO classification of tumors pathology and genetics, tumors of soft tissue and bone. Lyon: IARC Press; 2002. p. 102 103. 5. Huang HY, Lal P, Qin J, et al. Low-grade myxofibrosarcoma: A clinicopathologic analysis of 49 cases treated at a single institution with simultaneous assessment of the efficacy of 3-tier and 4-tier grading systems. Hum Pathol 2004;35:612 621. 6. Merck C, Angervall L, Kindblom LG, et al. Myxofibrosarcoma. A malignant soft tissue tumor of fibroblastichistiocytic origin. A clinicopathologic and prognostic study of 110 cases using multivariate analysis. Acta Pathol Microbiol Immunol Scand Suppl 1983;282:1 40. 7. Willems SM, Debiec-Rychter M, Szuhai K, et al. Local recurrence of myxofibrosarcoma is associated with increase in tumour grade and cytogenetic aberrations, suggesting a multistep tumour progression model. Mod Pathol 2006;19:407 416. 8. Lin CN, Chou SC, Li CF, et al. Prognostic factors of myxofibrosarcomas: Implications of margin status, tumor necrosis, and mitotic rate on survival. J Surg Oncol 2006;93:294 303. 9. Rosenberg SA, Tepper J, Glatstein E, et al. The treatment of soft-tissue sarcomas of the extremities: Prospective randomized evaluations of (1) limb-sparing surgery plus radiation therapy compared with amputation and (2) the role of adjuvant chemotherapy. Ann Surg 1982;196:305 315. 10. Karakousis CP, Emrich LJ, Rao U, et al. Feasibility of limb salvage and survival in soft tissue sarcomas. Cancer 1986;57: 484 491. 11. Suit HD, Mankin HJ, Wood WC, et al. Treatment of the patient with stage M0 soft tissue sarcoma. J Clin Oncol 1988;6: 854 862. 12. Williard WC, Hajdu SI, Casper ES, et al. Comparison of amputation with limb-sparing operations for adult soft tissue sarcoma of the extremity. Ann Surg 1992;215:269 275. 13. Suit HD, Russell WO, Martin RG. Sarcoma of soft tissue: Clinical and histopathologic parameters and response to treatment. Cancer 1975;35:1478 1483. 14. Lindberg RD, Martin RG, Romsdahl MM, et al. Conservative surgery and postoperative radiotherapy in 300 adults with soft-tissue sarcomas. Cancer 1981;47:2391 2397. 15. Scaife CL, Pisters PW. Combined-modality treatment of localized soft tissue sarcomas of the extremities. Surg Oncol Clin N Am 2003;12:355 368. 16. Yang JC, Chang AE, Baker AR, et al. Randomized prospective study of the benefit of adjuvant radiation therapy in the treatment of soft tissue sarcomas of the extremity. J Clin Oncol 1998;16:197 203. 17. O Sullivan B, Davis AM, Turcotte R, et al. Preoperative versus postoperative radiotherapy in soft-tissue sarcoma of the limbs: A randomised trial. Lancet 2002;359:2235 2241. 18. Pisters PW, Harrison LB, Leung DH, et al. Long-term results of a prospective randomized trial of adjuvant brachytherapy in soft tissue sarcoma. J Clin Oncol 1996;14:859 868. 19. Fein DA, Lee WR, Lanciano RM, et al. Management of extremity soft tissue sarcomas with limb-sparing surgery and postoperative irradiation: Do total dose, overall treatment time, and the surgery-radiotherapy interval impact on local control? Int J Radiat Oncol Biol Phys 1995;32:969 976. 20. Singer S, Corson JM, Gonin R, et al. Prognostic factors predictive of survival and local recurrence for extremity soft tissue sarcoma. Ann Surg 1994;219:165 173. 21. Zagars GK, Ballo MT, Pisters PW, et al. Prognostic factors for patients with localized soft-tissue sarcoma treated with conservation surgery and radiation therapy: An analysis of 1225 patients. Cancer 2003;97:2530 2543.

Recurrence patterns and survival in myxofibrosarcoma d K. E. HAGLUND et al. 367 22. Manoso MW, Pratt J, Healey JH, et al. Infiltrative MRI pattern and incomplete initial surgery compromise local control of myxofibrosarcoma. Clin Orthop Relat Res 2006;450:89 94. 23. American Joint Committee on Cancer. Cancer staging manual. 6th ed. Philadelphia: Lippincott-Raven; 2002. 24. LeVay J, O Sullivan B, Catton C, et al. Outcome and prognostic factors in soft tissue sarcoma in the adult. Int J Radiat Oncol Biol Phys 1993;27:1091 1099. 25. Sadoski C, Suit HD, Rosenberg A, et al. Preoperative radiation, surgical margins, and local control of extremity sarcomas of soft tissues. J Surg Oncol 1993;52:223 230. 26. Herbert SH, Corn BW, Solin LJ, et al. Limb-preserving treatment for soft tissue sarcomas of the extremities. The significance of surgical margins. Cancer 1993;72:1230 1238. 27. Tanabe KK, Pollock RE, Ellis LM, et al. Influence of surgical margins on outcome in patients with preoperatively irradiated extremity soft tissue sarcomas. Cancer 1994;73:1652 1659. 28. Pisters PW, Leung DH, Woodruff J, et al. Analysis of prognostic factors in 1,041 patients with localized soft tissue sarcomas of the extremities. J Clin Oncol 1996;14:1679 1689. 29. Trovik CS, Bauer HC, Alvegard TA, et al. Surgical margins, local recurrence and metastasis in soft tissue sarcomas: 559 Surgically-treated patients from the Scandinavian Sarcoma Group Register. Eur J Cancer 2000;36:710 716. 30. Alvegard TA, Sigurdsson H, Mouridsen H, et al. Adjuvant chemotherapy with doxorubicin in high-grade soft tissue sarcoma: A randomized trial of the Scandinavian Sarcoma Group. J Clin Oncol 1989;7:1504 1513. 31. Gaynor JJ, Tan CC, Casper ES, et al. Refinement of clinicopathologic staging for localized soft tissue sarcoma of the extremity: A study of 423 adults. J Clin Oncol 1992;10:1317 1329. 32. Rooser B, Berg NO, Ranstam J, et al. Prediction of survival in patients with high-grade soft tissue sarcoma. Int Orthop 1990; 14:199 204. 33. McKee MD, Liu DF, Brooks JJ, et al. The prognostic significance of margin width for extremity and trunk sarcoma. J Surg Oncol 2004;85:68 76. 34. Swallow CJ, Catton CN. Local management of adult soft tissue sarcomas. Semin Oncol 2007;34:256 269.