Soft-Tissue Sarcomas

, , ,

The soft-tissue sarcomas are a group of rare but anatomically and histologically diverse neoplasms. This is due to the ubiquitous location of the soft tissues and the nearly three dozen recognized histologic subtypes of soft-tissue sarcomas.

Overview

The soft-tissue sarcomas are a group of rare but anatomically and histologically diverse neoplasms. This is due to the ubiquitous location of the soft tissues and the nearly three dozen recognized histologic subtypes of soft-tissue sarcomas. It is estimated that in 2016 approximately 12,310 new cases of soft-tissue sarcoma will be identified in the United States, and 4,990 patients will die of the disease. The age-adjusted incidence is 3.3 cases per 100,000 persons.

Epidemiology

Sarcomas constitute less than 1% of all cancers, but 15% of cancers in children, although bone tumors predominate in children. There are familial syndromes in which sarcomas are common, but such genetic conditions are rare.

Gender

There is a slight male predominance, with a male-to-female ratio of 1.1:1. Some sarcomas, such as desmoplastic small round cell tumor, demonstrate a distinct male predominance.

Age

The age distribution in adult soft-tissue sarcoma studies is as follows: 20.7% of patients are < 40 years of age; 27.6% are 40 to 60 years old; and 51.7% of patients are > 60 years.

Race

Studies in large cohorts of patients demonstrate that the race distribution of soft-tissue sarcomas mirrors that of the American population (86% Caucasian, 10% African American, 1% Asian American, and 3% other). Some sarcomas, such as Ewing sarcoma, are distinctly unusual in African-American and Asian populations.

Geography

Besides issues of ethnicity and the frequency of specific cancers, there are not strong data to indicate a geographical bias in the development of sarcomas.

Etiology and Risk Factors

In the majority of soft-tissue sarcoma cases, no specific etiologic agent is identifiable. However, a number of predisposing factors have been recognized, which are discussed below.

Radiation Therapy

Soft-tissue sarcomas are recognized to originate in radiation fields following therapeutic irradiation for a variety of malignancies. Frequently, they are seen in the lower-dose regions at the edge of the radiation target volume. By definition, radiation-induced sarcomas arise no sooner than 3 years after radiation therapy and often develop decades later; the median time from radiation to development of a soft-tissue sarcoma is 8 to 10 years. The majority of these sarcomas are high-grade lesions (90%), and high-grade undifferentiated pleomorphic sarcoma is a predominant histology. Osteosarcoma, angiosarcoma, and other histologic subtypes have also been reported.

Chemical Exposure

Exposure to various chemicals in specific occupations or situations has been linked with the development of soft-tissue sarcoma. These chemicals include the phenoxyacetic acids (forestry and agriculture workers), chlorophenols (sawmill workers), vinyl chloride (individuals working with this gas, used in making plastics and as a refrigerant), and arsenic (vineyard workers).

Chemotherapy

Soft-tissue sarcomas have been reported after previous exposure to alkylating chemotherapeutic agents, most commonly after treatment of pediatric acute lymphocytic leukemia. The drugs implicated include cyclophosphamide, melphalan, procarbazine, nitrosoureas, and chlorambucil. The relative risk of sarcoma appears to increase with cumulative drug exposure.

Chronic Lymphedema

Soft-tissue sarcomas have been noted to arise in the chronically lymphedematous arms of women treated with radical mastectomy for breast cancer (Stewart-Treves syndrome). Lower-extremity lymphangiosarcomas have also been observed in patients with congenital lymphedema or filariasis complicated by chronic lymphedema.

Trauma and Foreign Bodies

Although a recent history of trauma is often elicited from patients presenting with soft-tissue sarcoma, the interval between the traumatic event and diagnosis is often short; thus, a causal relationship is highly unlikely. Chronic inflammatory processes, however, may be a risk factor for sarcoma. Foreign bodies, such as shrapnel, bullets, and implants, have also been implicated.

Signs and Symptoms

Signs and symptoms of soft-tissue sarcoma depend, in large part, on the anatomic site of origin. Due to the ubiquitous location of the soft tissues, these malignancies may arise at any site in the body where soft tissues are located. Since 50% of soft-tissue sarcomas arise in an extremity, the majority of patients present with a palpable soft-tissue mass. Pain at presentation is noted in only one-third of cases.

Extremity and Superficial Trunk

Extremity and superficial trunk sarcomas account for 60% of all soft-tissue sarcomas. The majority of patients present with a painless primary soft-tissue mass. Lipomas are at least 100 times more common than soft-tissue sarcomas; however, any growing lesion or even a deep-seated fatty lesion should be biopsied to rule out a sarcoma.

Retroperitoneum

Retroperitoneal sarcomas account for 15% of all soft-tissue sarcomas. Most patients (80%) present with an abdominal mass, with 50% of patients reporting pain at presentation. Due to the considerable size of the retroperitoneum and the relative mobility of the anterior intra-abdominal organs, these tumors often grow to substantial size before the patient's nonspecific complaints are evaluated or even before an abdominal mass is noted on physical examination.

Viscera

Visceral soft-tissue sarcomas, which comprise 15% of all soft-tissue sarcomas, present with signs and symptoms unique to their viscus of origin. For example, gastrointestinal stromal tumors (GISTs) present with GI symptoms that are usually indistinguishable from those of the more common adenocarcinomas, such as anemia, melena, abdominal pain, or weight loss. Similarly, uterine leiomyosarcomas frequently present with painless vaginal bleeding, such as that often noted in patients with more common uterine malignancies.

Head and Neck

Head and neck sarcomas comprise 10% of all soft-tissue sarcomas. Although generally smaller than sarcomas in other sites, they may present as a palpable mass or with important mechanical problems related to compression or invasion of adjacent anatomy (eg, orbital contents, airway, or pharynx). In addition, their proximity to critical anatomy can pose management difficulties due to compromise in the delivery of both surgery and radiotherapy.

Pathology

Histopathologic Classification

As a consequence of the wide spectrum of soft tissues, a variety of histologically distinct neoplasms have been characterized. The current histopathologic classification is based on the putative cell of origin of each lesion. Such classification based on histogenesis is reproducible for the more differentiated tumors. However, as the degree of histologic differentiation declines, it becomes increasingly difficult to determine a potential cellular origin, and such tumors are usually described by their architecture or cellular morphology, for example, clear cell sarcoma.

In addition, many of these tumors appear to have the ability to dedifferentiate. This process results in a variety of overlapping patterns, making uniform classification difficult. Experienced soft-tissue pathologists frequently disagree as to the cell of origin of an individual tumor. Comparative studies have demonstrated concordance in histopathologic diagnosis in only two-thirds of cases. Malignant fibrous histiocytoma, now termed undifferentiated pleomorphic sarcoma, used to be the most common histologic subtype of soft-tissue sarcoma. However, in one study, reanalysis histologically, immunohistochemically, and ultrastructurally allowed reclassification of most tumors to a specific line of differentiation. GIST is now recognized as the most common form of soft-tissue sarcoma, if all small and relatively inconsequential lesions removed as very early tumors are counted.

Staging and Prognosis

AJCC/UICC Staging System

The relative rarity of soft-tissue sarcomas, the anatomic heterogeneity of these lesions, and the presence of more than 50 recognized histologic subtypes of variable grade have made it difficult to establish a functional system that can accurately stage all forms of this disease. The staging system of the American Joint Committe on Cancer (AJCC) and the International Union Against Cancer (UICC), now in its 7th edition (2010, with publication of the 8th edition planned for mid 2016), is the most widely employed staging classification for soft-tissue sarcomas (Table 1). In the 7th edition, the staging for GIST has been separated from staging of other sarcomas for the first time (Table 2). All soft-tissue sarcoma subtypes are included, except desmoid tumors (deep fibromatosis), Kaposi sarcoma, and infantile fibrosarcoma. In keeping with the FNCLCC (Fédération Nationale des Centres de Lutte Contre la Cancer) sarcoma grading system, three distinct histologic grades are recognized, based on the degree of differentiation, mitotic activity, and necrosis.

Histologic grade and tumor size are the primary determinants of clinical stage. In the 6th version of the AJCC staging system, tumor size was further substaged as “a” (a superficial tumor that arises outside the investing fascia) or “b” (a deep tumor that arises beneath the fascia or invades the fascia). While these data are collected in AJCC/UICC version 7, they are not used to stage the tumor in version 7.

TABLE 1: AJCC version 7 staging for soft-tissue sarcomas

The AJCC/UICC system is designed to optimally stage extremity tumors but is also applicable to the trunk and head and neck. It is more difficult to employ for retroperitoneal or visceral (GI or other viscera) tumors, since the designation of superficial or deep is meaningless here, as all are deeply seated tumors in these anatomic sites.

Anatomic site is itself an important determinant of outcome. Patients with retroperitoneal, head and neck, and visceral sarcomas have an inferior overall prognosis compared with patients with extremity tumors. Although the anatomic site is not incorporated as a specific component of any current staging system, outcome data should be reported on a site-specific basis.

A retrospective review of 369 patients with high-grade soft-tissue sarcoma of the extremities treated with postoperative radiation therapy was conducted to evaluate the influence of tumor site on local control and complications. The tumor site was upper extremity in 103 patients (28%) and lower extremity in 266 patients (72%). With a median follow-up of 50 months, the 5-year actuarial rates of local control, distant relapse–free, and overall survival for the entire population were 82%, 61%, and 71%, respectively. The 5-year local control rates in patients with upper extremity lesions vs lower extremity lesions were 70% and 86%, respectively (P = .0004). On multivariate analysis, upper extremity site (P = .001) and positive resection margin (P = .02) were significant predictors of poor local control.

TABLE 2: AJCC version 7 staging for GIST

Prognostic Factors

Understanding relevant clinicopathologic prognostic factors is important in treatment planning for patients with soft-tissue sarcoma. Several reports document the adverse prognostic significance of tumor grade, anatomic site, tumor size, and depth relative to the investing fascia (for extremity and body wall tumors). Patients with high-grade lesions, large (T2) sarcomas, a nonextremity subsite, or deep tumor location are at increased risk for disease relapse and sarcoma-specific death.

Sarcoma-specific nomograms

Kattan and colleagues from Memorial Sloan Kettering Cancer Center developed a sarcoma-specific nomogram for estimation of sarcoma-specific 12-year survival. The nomogram takes into account pretreatment clinicopathologic factors, including anatomic site, histologic subtype, tumor size, histologic grade, tumor depth, and patient age. It is based on prospectively collected data and has been validated in a population of 2,136 patients with sarcoma. The sarcoma nomogram may be useful for patient stratification for clinical trials and for risk assessment and treatment planning for individual patients. Similar nomograms have been generated for a variety of specific sarcoma subtypes, such as GIST, synovial sarcoma, or the family of liposarcomas, and are validated for both 4-grade and 3-grade staging systems. Since the AJCC criterion of deep vs superficial tumor is not of consequence in staging retroperitoneal sarcomas, in which all tumors are deep, anatomic site–specific nomograms, can also be useful in more accurately predicting patient outcomes. A multi-institutional nomogram for retroperitoneal sarcoma has been developed and validated.

Risk assessment in GIST

Beyond the cumbersome AJCC/UICC staging system for GIST (Table 2), there are other ways to assess tumor risk. Perhaps the most useful of these is the risk-stratification strategy from the Armed Forces Institute of Pathology (AFIP). Tumors are classified by anatomic site, size, mitotic rate, and risk of recurrence (Table 3). This simple method allows one to discuss with a patient the risk of a particular GIST with respect to mortality or the potential benefit of adjuvant imatinib. For any particular size and mitotic count, GISTs of gastric origin have better prognosis than those arising at other sites. What is notable from the data in Table 3 is that even large gastric GISTs, one of the most common scenarios for diagnosis, have a low (≤ 12%) mitotic rate; thus, for the average patient with smaller primary tumors in the most common site of GIST, adjuvant imatinib can be avoided. A single-institution nomogram for primary GIST has also been developed and validated to predict outcome without use of adjuvant imatinib. The nomogram employs tumor size, mitotic count, and site of origin as the key prognostic variables. KIT and PDGFRA mutation status can also have an impact on the choice of adjuvant imatinib, and should be examined routinely to help in clinical decision making.

TABLE 3: AFIP risk stratification strategy

Prognostic factors for local vs distant recurrence

Unlike other solid tumors, the adverse prognostic factors for local recurrence of a soft-tissue sarcoma (other than GIST) differ from those that predict distant metastasis and tumor-related mortality. In other words, patients with a constellation of adverse prognostic factors for local recurrence are not necessarily at increased risk for distant metastasis or tumor-related death.

This concept has been validated by an analysis of the Scandinavian Sarcoma Group prospective database. In 559 patients with soft-tissue sarcomas of the extremities and trunk treated with surgery alone, inadequate surgical margin was found to be a risk factor for local recurrence but not for distant metastasis. Therefore, staging systems that are designed to stratify patients for risk of distant metastasis and tumor-related mortality using these prognostic factors (such as the AJCC/UICC system) do not stratify patients for risk of local recurrence.

Screening and Diagnosis

Currently, there are no screening tests for soft-tissue sarcomas. Since the majority of patients with soft-tissue sarcoma have lesions arising in the extremities or superficial trunk, most of the comments here apply to soft-tissue lesions in those sites. A separate algorithm is usually employed for the evaluation of a primary retroperitoneal mass or visceral sarcoma.

Physical Examination

Physical examination should include an assessment of the size of the mass and its mobility relative to the underlying soft tissues. The relationship of the mass to the investing fascia of the extremity (superficial vs deep) and nearby neurovascular and bony structures should be noted. Site-specific neurovascular examination and assessment of regional lymph nodes should also be performed.

Biopsy

Any soft-tissue mass in an adult extremity should be biopsied if it is symptomatic or enlarging, is > 5 cm, or has persisted beyond 4 to 6 weeks.

Percutaneous approaches

Percutaneous tissue diagnosis can usually be obtained by percutaneous core biopsy for histology. The needle track should be placed in an area to be excised or that can be encompassed in adjuvant radiotherapy fields if they are to be used. In most instances, when an experienced histopathologist examines the specimen, a diagnosis of malignant soft-tissue sarcoma can be made. Fine needle aspiration (FNA) is often viewed as a suboptimal method of establishing an initial diagnosis of soft-tissue sarcoma. Histology is usually preferred to cytology because more tissue is obtained, which allows for a more accurate delineation of tumor type and grade. In an era in which tissue is increasingly important for identifying molecular characteristics of the tumor for diagnosis or prognosis, several needle passes with a core biopsy are frequently obtained for diagnosis. FNA is a less-invasive means to confirm recurrence, however. Percutaneous tissue diagnosis is preferred over open biopsy to facilitate subsequent treatment planning and permit surgical resection to be performed as a one-stage procedure.

Open biopsy

In some cases, an adequate histologic diagnosis cannot be secured by percutaneous means. Open biopsy is indicated in these instances, with the exception of relatively small superficial masses, which can be easily removed by excisional biopsy with clear margins. Biopsies should be incisional and performed with a longitudinal incision parallel to the long axis of the extremity. This approach facilitates subsequent wide local excision of the tumor, and the incisional scar results in minimal difficulties in wound closure. It also facilitates inclusion of any scars within the area of the tumor in adjuvant radiation fields without the excessive morbidity of large-field radiotherapy planning. The incision should be centered over the mass at its most superficial location. Care should be taken not to raise tissue flaps. Meticulous hemostasis should be ensured after the biopsy to prevent dissemination of tumor cells into adjacent tissue planes by hematoma.

Retroperitoneal or intra-abdominal mass

Biopsy of primary retroperitoneal soft-tissue masses is not required for radiographically resectable masses. There are no data to suggest that percutaneous core-needle biopsy increases the risk of local recurrence or tumor rupture. The circumstances under which percutaneous or preoperative biopsy of retroperitoneal masses should be strongly considered include:

• tissue diagnosis for radiographically unresectable disease

• clinical suspicion of lymphoma or germ-cell tumor

• tissue diagnosis for neoadjuvant treatment, including radiotherapy and/or chemotherapy

• suspected metastases from another primary tumor.

For GISTs, when the diagnosis is uncertain, endoscopic ultrasound–guided FNA is a reasonable approach that avoids the risk of tumor rupture from a percutaneous approach.

Primary Tumor Imaging

Despite one study showing equivalence in detection of sarcoma by computed tomography (CT) or magnetic resonance imaging (MRI), most surgeons recommend MRI of the primary site for extremity tumors. CT scan or MRI is used for truncal or abdominal/visceral tumors. MRI may be more helpful for fixed organ tumors (rectum or liver) rather than for tumors of the other viscera. More invasive studies, such as angiography and cavography, are almost never required for the evaluation of soft-tissue sarcomas. The role of positron emission tomography (PET) scan in sarcoma management is not well defined. For example, in metstatic disease, higher quality images can be obtained with a contrast CT than with a PET/CT, which uses less optimal CT for registration of images with PET images. There may be selected cases in which PET can be useful, particularly in patients who cannot tolerate IV contrast for CT scan or MRI. In particular, if the lack of IV contrast uptake as a sign of a responding tumor is taken into account, contrast-enhanced CT scans appear to yield results similar to PET scans in patients with GIST.

Imaging for Metastatic Disease

Cost-effective imaging to exclude the possibility of distant metastatic disease depends on the size, grade, and anatomic location of the primary tumor. In general, patients with low-grade soft-tissue sarcomas < 10 cm or intermediate-/high-grade tumors < 5 cm in diameter require only a chest x-ray for satisfactory staging of the chest. This reflects the fact that these patients are at comparatively low risk of presenting with pulmonary metastases. In contrast, patients with very large (≥ 10 cm) low-grade tumors or high-grade tumors ≥ 5 cm should undergo more thorough staging of the chest by CT.

Patients with retroperitoneal and intra-abdominal visceral sarcomas should undergo single-modality imaging of the liver to exclude the possibility of synchronous hepatic metastases. The liver is a common site for a first metastasis from these lesions. Only rarely does PET appear to be indicated, in particular for patients with contrast allergies who cannot be staged definitively preoperatively with CT or MRI.

Treatment

Treatment of Localized Disease

Surgical resection is the cornerstone of therapy for patients with localized disease. After 1980, there was a gradual shift in the surgical management of soft-tissue sarcoma of the extremities away from radical ablative surgery, such as amputation or compartment resection, toward limb-sparing approaches combining wide local resection with preoperative or postoperative radiotherapy. The development of advanced surgical techniques (eg, microvascular tissue transfer, bone and joint replacement, and vascular reconstruction) and the application of multimodality approaches have allowed most patients to retain a functional extremity without any compromise in survival.

Surgery

The surgical approach to soft-tissue sarcomas depends on careful preoperative staging with MRI or CT for lesions of the extremities and a percutaneous histologic diagnosis and assessment of tumor grade. In most instances, preoperative imaging studies allow for accurate prediction of resectability. The surgical approach to soft-tissue sarcomas is based on an awareness that these lesions tend to expand and compress tissue planes, producing a pseudocapsule. However, some histologic subtypes, such as myxofibrosarcoma or desmoid tumors, may be more infiltrative, extending with microscopic projections beyond the palpable extent of the tumor.

Wide local resection. Wide local resection encompassing a rim of normal tissue around the lesion has led to improvements in local tumor control, with local recurrence rates of approximately 30% in the absence of adjuvant therapies. However, studies indicate that carefully selected patients with localized, small (T1), low-grade soft-tissue sarcomas of the extremity can be treated by wide resection alone, with local recurrence rates of < 10%.

A prospective trial examined local recurrence rates for T1 primary soft-tissue sarcomas of the trunk and extremities. Patients underwent function-preserving surgery. Postoperative radiation was delivered for microscopically positive margins (R1 resection). No radiation was delivered if margins were negative (R0 resection). A total of 88 patients were evaluated in this trial. A total of 16% had R1 resection and received adjuvant radiation therapy, whereas 84% had R0 resection and did not receive radiation therapy. With a median follow-up of 75 months, isolated local recurrence was observed in six patients in the R1 arm (43%) and six patients in the R0 arm (8%). The 5- and 10-year local recurrence rates in the R0 arm were 7.9% and 10.6%, respectively, and the sarcoma-specific death rates were 3.2% at both 5 and 10 years in the R0 arm.

The need for adjuvant irradiation in small (< 5 cm), high-grade lesions has been studied. A retrospective review of 204 patients with stage IIB soft-tissue sarcoma of the extremity treated at Memorial Sloan Kettering Cancer Center yielded a total of 57% of patients who did not receive adjuvant radiation therapy, whereas 43% received either brachytherapy or external-beam radiation therapy (EBRT). With a median follow-up of 67 months, there was no significant difference in 5-year local tumor control, distant relapse–free survival, or disease-specific survival when adjuvant irradiation was delivered.

Further studies will be required to define which subsets of patients with primary extremity sarcoma can be treated by wide excision surgery alone. Preoperative or postoperative radiotherapy should be employed for patients with primary T1 sarcomas in whom a satisfactory gross surgical margin cannot be attained without compromise of functionally important neurovascular structures.

Limb-sparing surgery plus irradiation. Limb-sparing surgery employing adjuvant irradiation to facilitate maximal tumor local control has become the standard approach for large (T2) soft-tissue sarcomas of the extremities. In most centers, more than 90% of patients are treated with limb-sparing approaches. Amputation is reserved as a last-resort option for local tumor control or when there are no limb salvage options, and is used with the knowledge that it does not affect survival. This approach was validated in a prospective National Cancer Institute (NCI) study, in which patients with a limb-sparing surgical option were randomized to receive limb-sparing surgery with postoperative radiation therapy or amputation. Both arms of the study included postoperative therapy with doxorubicin, cyclophosphamide, and methotrexate.

Surgical procedure. The planned resection should encompass the skin, subcutaneous tissues, and soft tissues adjacent to the tumor, including the previous biopsy site and any associated drain sites. The tumor should be excised with a 2- to 3-cm margin of normal surrounding tissue whenever possible; wider margins may be considered for some histologies such as myxofibrosarcoma, which can have microscopic extensions along fascial or neurovascular bundle tissue planes. Since good adjuvant approaches are available to facilitate local tumor control, the surgeon sometimes compromises this ideal margin rather than attempting resection of adjacent, possibly involved bone or neurovascular structures that would result in significant functional loss. In the rare circumstance of gross involvement of neurovascular structures or bone, en bloc resection and reconstruction can be performed.

Metal clips should be placed at the margins of resection to facilitate radiation field planning, when and if external irradiation is indicated. Drain sites should be positioned close to the wound to allow inclusion in radiation therapy fields. As noted earlier, avoidance of transverse incisions greatly facilitates the ability to include the tissues at risk in the radiation target volume without unduly large fields.

Regional lymphadenectomy. Given the low (2% to 3%) prevalence of lymph node metastasis in adult sarcomas, there is no role for routine regional lymphadenectomy. Patients with embryonal rhabdomyosarcoma and epithelioid sarcoma have an increased incidence of lymph node metastasis; these patients should be carefully examined and are good candidates for sentinal lymph node biopsy. Clinically apparent lymphadenopathy should be treated with therapeutic lymphadenectomy. A recent analysis suggested that select patients undergoing lymphadenectomy, particularly in the absence of systemic metastases, may have a 5-year survival rate superior to the survival expected for patients with AJCC stage IV disease, as was defined in AJCC version 6. As a result, the AJCC staging system (version 7) for sarcomas suggests that patients with lymph node metastasis should be considered in a different category from patients with overt blood-borne metastases.

Radiotherapy

Radiation therapy is usually combined with surgical resection in managing soft-tissue sarcomas of the extremities.

Preoperative irradiation. Preoperative irradiation has a number of theoretic and practical advantages: (1) Smaller radiation portals can be utilized, as the scar, hematomas, and ecchymoses do not need to be treated. (2) Preoperative irradiation may produce tumor encapsulation, facilitating surgical resection from vital structures. (3) It is easier to spare a strip of skin and thereby reduce the risk of lymphedema. (4) The size of the tumor may be reduced, thus decreasing the extent of surgical resection. (5) Lower radiation doses can be utilized, as there are fewer relatively radioresistant hypoxic cells.

Preoperative irradiation also has several drawbacks, however. They include (1) the inability to precisely stage patients based on pathology due to downstaging and (2) increased problems with wound healing. Studies of preoperative irradiation from the University of Florida, the University of Texas MD Anderson Cancer Center, and Massachusetts General Hospital demonstrated local tumor control rates of 90% using doses of approximately 50 Gy. Survival depended on the size and the grade of the primary tumor. Distant metastases were the primary pattern of failure.

Postoperative irradiation. A number of retrospective reports, as well as a randomized trial from the NCI, have demonstrated that limb-sparing surgery plus postoperative irradiation produces local tumor control rates comparable to those achieved with amputation. Five-year local tumor control rates of 70% to 90%, survival rates of 70%, and limb-preservation rates of 85% can be expected.Equivocal or positive histologic margins are associated with higher local recurrence rates, and, therefore, adjuvant external-beam irradiation should be considered in all patients who have sarcoma of the extremities with positive or close microscopic margins in whom reexcision is impractical. Postoperative doses of 60 to 65 Gy should be used.

Interstitial therapy. Interstitial therapy with iridium-192 is used at some institutions as a radiation boost to the tumor bed following adjuvant external-beam irradiation. In a randomized trial, the 5-year local tumor control rate was 82% in patients who received adjuvant brachytherapy vs 69% in those treated with surgery alone. On subset analysis, the local tumor control rate was found to be 89%, vs 66% for patients with high-grade lesions. This study and additional studies have indicated that brachytherapy has no impact on local tumor control for low-grade lesions.

FIGURE 1A: Kaplan-Meier plots for probability of local recurrence in the National Cancer Institute of Canada Clinical Trials Group phase III trial.

FIGURE 1B: Kaplan-Meier plots for probability of overall survival in the National Cancer Institute of Canada Clinical Trials Group phase III trial.

If an implant alone is used, the dose is 40 to 45 Gy to a volume that includes all margins; when a boost is combined with additional external-beam irradiation, a dose of 20 to 25 Gy is utilized. Some data suggest a higher rate of wound complications and a delay in healing when implants are afterloaded prior to the third postoperative day. Although some centers load implants sooner, this step must be performed with caution and strict attention to the incision site.

Over a 15-year period, 202 patients with high-grade sarcoma of the extremities underwent complete gross resection and adjuvant brachytherapy to a median dose of 45 Gy, delivered over 5 days. With a median follow-up of 61 months, the 5-year local tumor control, distant relapse-free survival, and overall survival rates were 84%, 63%, and 70%, respectively. These rates compared favorably with data on external-beam irradiation. Morbidity of brachytherapy was considered acceptable, with reoperation rates of 12%, bone fractures in 3%, and nerve damage in 5%.

Comparison of irradiation techniques. Comparable local tumor control results (90%) are obtained with preoperative, postoperative, and interstitial techniques, although rates of wound complications are higher with preoperative techniques. Brachytherapy can offer a number of advantages. When brachytherapy is employed as the sole adjuvant (ideally in R0 cases), the entire local therapy (surgery plus radiation) is completed in a 10- to 12-day period. This compares favorably to the 3-month period reqiured for surgery, postoperative recovery (6-8 weeks), and external beam radiation treatment (6 weeks). Generally, smaller volumes can be irradiated with brachytherapy, which could improve functional results. However, smaller volumes may not be appropriate, depending on the tumor size, grade, and margin status.

The National Cancer Institute of Canada (NCIC) Clinical Trials Group published 3-year median follow-up results of a randomized phase III trial comparing preoperative and postoperative radiotherapy for limb soft-tissue sarcoma (Figures 1A, 1B). Wound complications were observed in 31 of 88 patients (35%) in the preoperative group and 16 of 94 patients (17%) in the postoperative group (difference, 18% [95% CI, 5–30]; P = .01). Tumor size and anatomic site were also significant risk factors in multivariate analysis. Local tumor control was identical in both arms of the trial. Five-year outcomes have been reported, and no difference in metastases, cause-specific survival, or overall survival was noted. Because preoperative radiotherapy is associated with a greater risk of wound complications than postoperative radiotherapy, but less late fibrosis and edema, the choice of regimen for patients with soft-tissue sarcoma should take into account the timing of surgery and radiotherapy and the size and anatomic site of the tumor.

Regardless of the technique employed, local control is a highly achievable and worthwhile endpoint, as demonstrated in a retrospective study of 911 patients treated by various techniques at Memorial Sloan Kettering Cancer Center. Of the 116 patients who developed local recurrence, 38 patients subsequently developed metastases and 34 patients died. Metastases after local recurrence were predicted in patients with high-grade or large (> 5 cm) tumors.

Treatment recommendations. Adjuvant radiotherapy should be employed for virtually all high-grade sarcomas of the extremities and larger (≥ 5 cm) low-grade lesions. If small (T1) lesions can be resected with clear margins, radiotherapy may be omitted. Postoperative therapy with either external-beam irradiation (with or without an interstitial implant boost) or an implant alone will achieve a high likelihood of local tumor control and, therefore, limb preservation. Preoperative irradiation, although equally efficacious, does carry a higher wound complication rate than the postoperative approach. For smaller tumors under 5 cm in more difficult anatomic sites such as the head and neck, consideration can still be given to neoadjuvant or adjuvant radiation therapy, given the implications of local recurrence in these anatomic sites.

Primary radiation therapy

Several studies of radiation therapy alone in the treatment of unresectable or medically inoperable soft-tissue sarcomas have reported 5-year survival rates of 25% to 40% and local tumor control rates of 30%. Local tumor control depends largely on the size of the primary tumor. Radiation doses should be at least 65 to 70 Gy, if delivery of such doses is feasible. The tumor's location may be particularly important in determining this dose because of the potential for damage to critical structures (eg, the spinal cord) with the higher doses normally used. Image-guided techniques with EBRT may make it possible to deliver an even higher dose to such tumor beds, as can particulate radiation such as proton beam therapy.

Radiation therapy in retroperitoneal sarcomas

Only 50% of patients with retroperitoneal sarcomas are able to undergo complete surgical resection. Of patients undergoing complete resection, well over half will develop local recurrence. This significant local failure rate suggests a potentially important role for adjuvant treatment in all patients with retroperitoneal sarcomas. However, the role of radiation therapy for retroperitoneal sarcomas remains controversial due to the rarity of the tumor, the paucity of data, the retrospective nature of available studies, the low doses of radiation used in many studies, and the lack of consistent policies for determining the indications for radiation therapy.

Preoperative irradiation. The advantages of preoperative radiotherapy have already been discussed for soft-tissue sarcomas of the extremities. In the retroperitoneum, an additional advantage is that bowel is frequently displaced significantly by the tumor. In contrast to the postoperative setting, the bowel being treated is also unlikely to be tethered by adhesions from prior surgery. These features significantly offset acute toxicity of large-field intra-abdominal radiotherapy (eg, nausea, vomiting, and diarrhea) as well as the potential for late-onset bowel toxicity. Conformal techniques capable of sparing normal tissues are also more easily applied in the preoperative setting, when the tumor can be visualized and the target area more readily defined. An ongoing, international phase III study (the European Organisation for Research and Treatment of Cancer [EORTC] 62092-22092, the STRASS trial) is evaluating the role of radiation therapy, randomizing a planned 256 patients to surgery alone or preoperative radiation therapy followed by surgery.

Intraoperative irradiation. In a prospective trial from the NCI, 35 patients with completely resected retroperitoneal sarcomas were randomized to receive either intraoperative electron-beam irradiation (IORT) followed by low-dose (30 to 40 Gy) postoperative EBRT or high-dose postoperative EBRT (35 to 40 Gy plus a 20-Gy boost). Absolute local recurrence rates were significantly lower in the IORT group (P < .05), but disease-specific and overall survival rates did not differ between the two groups.

Similarly, a nonrandomized series from the Massachusetts General Hospital has suggested improved local tumor control with IORT for patients with retroperitoneal sarcoma. In 16 patients who underwent irradiation, complete gross resection, and IORT, overall survival and local tumor control rates were 74% and 83%, respectively. These numbers diminished to 30% and 61%, respectively, in the 13 patients treated with irradiation and complete gross resection without IORT. Although these local tumor control results are encouraging, IORT remains investigational and cannot be advocated on a routine basis at this time.

Postoperative irradiation. Two-year local tumor control rates of 70% have been reported with the addition of postoperative irradiation. However, irradiation of the retroperitoneum/abdomen in doses that have effected local tumor control in soft-tissue sarcoma of the extremities (50 to 65 Gy) is usually associated with significant GI toxicity. Obviously, the incidence of GI toxicity depends on the exact fields and technique used. However, as most retroperitoneal sarcomas are > 10 to 15 cm, the radiation fields employed are generally also quite large, and bowel is often located and/or tethered in the high-risk area. Three-dimensional treatment planning and conformal techniques can now be utilized to maximize the radiation dose to the tumor bed while minimizing the dose to the surrounding normal tissues.

Isolated limb perfusion

Recent studies have evaluated the role of isolated limb perfusion (ILP) in the management of sarcomas of the extremities. These studies have generally been extrapolations from protocols initially designed to treat locally advanced melanoma. The agents most commonly employed for ILP have been melphalan and tumor necrosis factor-alpha, with or without interferon-gamma (IFN-γ-1b [Actimmune]). The results of the largest series of ILP in patients with locally advanced soft-tissue sarcoma of the extremities were reported by Eggermont and colleagues. TNF-α has now been approved in Europe for ILP in patients with locally advanced intermediate-high grade soft-tissue sarcomas of the extremities.

The Netherlands Cancer Institute published its results in patients with unresectable soft-tissue sarcoma of the extremities who were perfused with melphalan and TNF-α. A total of 49 patients were treated and followed for a median of 26 months. One patient died shortly after perfusion, but 31 patients (63%) were able to undergo resection of the tumor. Based on clinical and pathologic grounds, an overall response was seen in 31 patients (63%), and a complete response was seen in 4 patients (8%). A total of 28 patients (57%) had local tumor control with limb preservation. Toxicity was frequent but usually mild, with post-treatment edema being a common complication.

Role of Adjuvant Chemotherapy

The striking success of combined-modality therapy in children with osteogenic sarcoma, rhabdomyosarcoma, and the Ewing sarcoma family of tumors has provided the stimulus for the use of aggressive combined-modality approaches in adults. The literature is replete with reports of the apparent benefit of combined-modality therapy in patients with resectable soft-tissue sarcoma, yet most series are either retrospective or small nonrandomized trials.

Preoperative chemotherapy

Preoperative chemotherapy has been adopted at a variety of centers for patients with large high-grade sarcoma. The specific regimens employed have evolved over the years but generally contain both an anthracycline and ifosfamide.

Aside from theoretic considerations, there are several pragmatic reasons to favor preoperative over postoperative treatment. First, a reduction in the size of a large lesion may permit surgical resection with less morbidity. Second, compliance may be better with preoperative therapy. One observation that supports the neoadjuvant approach is that response to preoperative chemotherapy, whether pathologic or radiographic, predicts improved tumor control and survival.

Neoadjuvant chemotherapy has been explored in a prospective randomized trial initiated by the EORTC. The trial was open to patients who had a sarcoma measuring at least 8 cm (of any grade), a primary or recurrent intermediate- to high-grade sarcoma of any size, or a locally recurrent or inadequately excised Actimmune. In spite of these broad eligibility criteria, accrual was slow, and the trial was closed after only 150 patients entered.

Patients were randomized to receive either immediate surgery, followed by radiation therapy for close or positive margins, or 3 cycles of chemotherapy with doxorubicin (50 mg/m2 by IV bolus) plus ifosfamide (5 g/m2 by 24-hour continuous infusion) with mesna (Mesnex). Among the 134 eligible patients, over 80% had primary tumors of the extremities, but only 4% had grade 2 or 3 lesions > 8 cm. Among 49 patients evaluable for response, 29% had major objective responses, including four complete responses. Only 18% had progression of disease before surgery. Chemotherapy was generally well tolerated and never prevented surgery. With a median follow-up of 7.3 years, the estimated 5-year survival rate was similar for both groups.

Trials have explored the role of neoadjuvant chemotherapy and radiation therapy to decrease the rate of distant failure and possibly impact survival. A study reported from Massachusetts General Hospital enrolled patients with high-grade soft- tissue sarcomas (8 cm or larger). Patients were treated with 3 cycles of preoperative chemotherapy consisting of MAID (mesna, doxorubicin [Adriamycin], ifosfamide, dacarbazine) interdigitated with 44 Gy of radiation therapy. This regimen was followed by surgical resection and 3 cycles of postoperative MAID chemotherapy. In cases with positive surgical margins, an additional 16 Gy of radiation therapy was delivered. This regimen resulted in a significant improvement in 5-year freedom from distant metastasis (75% vs 44%; P = .0016) when compared with historic control patients. Additionally, 5-year disease-free and overall survival rates were 70% vs 42% (P = .0002) and 87% vs 58% (P = .0003) for the MAID and control groups, respectively. There was a 29% rate of wound healing complications in the MAID group.

These data have been extended in a follow-up study of similar interdigitated chemotherapy/radiation therapy in a phase II study from the Radiation Therapy Oncology Group. In this study, 66 patients with primary high-grade soft-tissue sarcoma ≥ 8 cm in diameter received a modified MAID regimen plus granulocyte colony-stimulating factor (G-CSF, filgrastim [Neupogen]) and radiation therapy, followed by resection and postoperative chemotherapy. Preoperative radiotherapy and chemotherapy were successfully completed by 89% and 79% of patients, respectively. Grade 4 hematologic and nonhematologic toxicities affected 80% and 23% of patients, respectively. Two patients developed acute myelogenous leukemia (AML) following therapy. Delayed wound healing was noted in 31%. The estimated 3-year survival, disease-free survival, and local tumor control rates were 75%, 55%, and 79%, respectively.

The MD Anderson Cancer Center conducted a phase I trial to define the maximum tolerated dose of continuous-infusion doxorubicin administered with preoperative radiation therapy to a dose of 50 Gy. In total, 27 patients with intermediate- or high-grade sarcomas were enrolled in the trial. The maximum tolerated dose of doxorubicin was 17.5 mg/m2/week. Twenty-six patients underwent surgery, and all had a macroscopic complete resection (R0 or R1). Two patients had a pathologic complete response. These studies suggest that further investigation of a preoperative approach combining chemotherapy and radiation therapy is warranted. The lack of randomized data regarding the addition of chemotherapy to radiation therapy for extremity sarcomas limits the ability to apply these treatment regimens outside the setting of a study.

Postoperative chemotherapy

A number of published trials have compared postoperative chemotherapy with observation alone in adults who had undergone resection of a primary or recurrent soft-tissue sarcoma. Most of these trials included fewer than 100 patients, and even the largest trial had inadequate statistical power to detect a 15% difference in survival. Other flaws confound the interpretation of many of the studies. Some trials included low-risk patients with small and/or low-grade sarcomas. In some trials, patient ineligibility rates were as high as 20%, and in none of the trials published before 2000 was ifosfamide part of the combination evaluated.

In five of the six trials in which doxorubicin monotherapy was studied, including one study limited to patients with uterine sarcoma, a significant improvement in survival could not be demonstrated. Among the trials of combination chemotherapy, most used the combination known as CYVADIC (cyclophosphamide, vincristine, doxorubicin [Adriamycin], dacarbazine). A significant survival advantage was seen in only one combination chemotherapy trial.

Nonetheless, some of the trials showed a trend or a statistically significant improvement in disease-free survival among patients who were administered adjuvant chemotherapy, especially among those with high-grade sarcomas of the extremities. Analyses of the pooled results of the published literature are consistent with this observation.

Ifosfamide-containing trials. Only one trial included in a meta-analysis by the Sarcoma Meta-Analysis Collaboration (SMAC) used an ifosfamide-containing regimen; that trial involved only 29 patients. An attempt to conduct a large prospective trial of postoperative chemotherapy with the MAID regimen in the United States failed because of insufficient patient accrual.

An Italian cooperative group conducted a trial in which patients 18 to 65 years old with high-grade (> 5 cm) or any recurrent sarcoma of the extremities were randomized to receive postoperative chemotherapy or observation alone. The treatment consisted of 5 cycles of epirubicin at 60 mg/m2 on days 1 and 2, plus ifosfamide at 1.8 g/m2 on days 1 to 5. Filgrastim was used to support the granulocyte counts during therapy. The trial had been planned to accrue 200 patients but was interrupted after accrual of 104 patients, when an interim analysis showed a significant survival advantage for the chemotherapy-treated group. At 36 months after the last randomization, with a median follow-up of 59 months, median overall survival among the patients who received adjuvant chemotherapy was 75 months, vs 46 months for control patients (P = .03). In a longer-term follow-up analysis, survival was not improved in an intention-to-treat analysis, although 5-year overall survival rates still favored the patients receiving chemotherapy.

An analysis of adjuvant chemotherapy using doxorubicin and ifosfamide was conducted by the EORTC (study 62931). This study examined surgery and adjuvant radiation therapy vs the same local therapy and adjuvant chemotherapy with 5 cycles of doxorubicin (75 mg/m2) and ifosfamide (5 g/m2) every 21 days. A total of 351 patients were randomly assigned over 8 years (1995–2003) to 5 cycles of adjuvant doxorubicin-ifosfamide chemotherapy or not, in addition to appropriate primary therapy for the sarcoma. Overall survival was not different between groups (hazard ratio [HR] = 0.94; 95% CI, 0.68–1.31; P = .72) nor was relapse-free survival (HR = 0.91; 95% CI, 0.67–1.22; P = .51). The 5-year overall survival was 66.5% (58.8–73) in the chemotherapy group and 67.8% (60.3–74.2) in the control group. Eighty percent of people assigned chemotherapy completed all five cycles. No deaths due to toxic effects were recorded.These data temper some of the enthusiasm regarding adjuvant chemotherapy as demonstrated in a positive Italian study of epirubicin and ifosfamide. Since the time of the original Italian study, other randomized studies have been performed. They do not indicate a benefit for chemotherapy but were underpowered to detect small differences in outcome.

A 2008 meta-analysis by Pervaiz et al included more ifosfamide-based studies, although it was published before the large randomized EORTC 62931 study. This work demonstrated for the first time in a meta-analysis the overall survival benefit of adjuvant chemotherapy. In this updated meta-analysis, 18 trials, representing 1,953 patients, were examined. The odds ratio (OR) for local recurrence was 0.73 in favor of chemotherapy (P = .02). In terms of overall survival, use of doxorubicin-based therapy without ifosfamide had an OR of 0.84, which was not statistically significant (P = .09). However, the OR for doxorubicin/ifosfamide-based therapy was 0.56 (P = .01) in favor of chemotherapy. The updated meta-analysis confirmed the SMAC meta-analysis in that there was efficacy of adjuvant chemotherapy for resected soft-tissue sarcoma with respect to local recurrence, distant recurrence, overall recurrence, and overall survival. The authors concluded that benefits were further improved with the addition of ifosfamide to doxorubicin-based regimens but must be weighed against the toxicity of the addition of ifosfamide.

Analyses of other collected prospective data regarding adjuvant chemotherapy from large referral centers have also yielded conflicting data. In two analyses of patients with synovial sarcoma and one involving myxoid/round cell liposarcoma, chemotherapy appeared to improve overall survival. In a large analysis of two prospective databases, patients receiving chemotherapy initially had superior survival but then suffered inferior survival compared with those who received no adjuvant chemotherapy. Notably, patients were not randomized as part of their treatment. Given the fact that this was a registry instead of a randomized study, there was by definition a bias to treat patients who had higher-risk tumors with chemotherapy, although this did not appear to correlate with a specific single variable in the analysis. The data from the most recent meta-analysis appear to be consistent with data from prior studies when taken as a whole. If there is a benefit in terms of overall survival with the use of adjuvant chemotherapy, it is modest, perhaps akin to that of patients with resected non–small-cell lung cancer, and thus the risks and benefits of this toxic therapy should be discussed with patients on an individual basis.

Treatment recommendations

• Multidisciplinary treatment planning should precede the initiation of any therapy. An experienced multidisciplinary team should evaluate pathologic material and imaging studies and coordinate the integration of surgical resection, irradiation, and systemic therapy.

• Ideally, patients should be offered participation in clinical trials. Unfortunately, there are no active trials in the United States that will definitively answer the most important questions. Thus, a decision to treat must be made on an individual basis.

• Preoperative chemotherapy may be considered for fit, high-risk patients after a discussion of the risks and potential benefits; older patients, especially those with cardiac or renal disease, are not optimal candidates for such treatment.

• Patients who do not receive preoperative chemotherapy may still be offered postopertive treatment. Adjuvant anthracycline/ifosfamide combinations improve relapse-free survival in selected patients and can be considered for the treatment of those with tumor size > 5 cm, deep tumor location, and high histologic grade. Overall survival was superior in a 2008 meta-analysis of patients receiving anthracycline-ifosfamide-based therapy, and this should be the standard combination to consider if adjuvant therapy were to be administered.

• For patients who opt for preoperative or postoperative chemotherapy, a regimen that includes doxorubicin (60 to 75 mg/m2) or epirubicin (120 mg/m2) plus ifosfamide (9 to 10 g/m2), given for a total of 4 to 6 cycles, is a reasonable choice for patients younger than age 60.

• Outside the context of a clinical trial, it is difficult to recommend concurrent doxorubicin and radiation therapy, or closely spaced doxorubicin-based chemotherapy with radiation therapy, owing to the observed risk of second malignancies such as AML in clinical trials to date. The authors recommend doxorubicin-ifosfamide therapy (AIM) rather than the same drugs with dacarbazine (MAID), owing to the lack of activity of dacarbazine in most sarcomas in the metastatic setting.

Treatment of Local Recurrence

Despite optimal multimodality therapy, local recurrence develops in 10% to 50% of patients, with a median local recurrence-free interval of ~24 months. Local recurrence rates are a function of the primary site and are highest for patients with retroperitoneal and head and neck sarcomas, for which adequate surgical margins are difficult to attain. In addition, high-dose adjuvant irradiation of these sites is often limited by the relative radiosensitivity of surrounding structures. These factors result in local recurrence rates of 40% for retroperitoneal sarcomas and up to 50% for head and neck sarcomas, which are substantially higher than the 10% rate typically seen for extremity sarcomas. A large retrospective analysis of patients with high-grade sarcoma of the extremities was reported from UCLA. Local recurrence required amputation in 38% of cases and was associated with a threefold decrement in survival. This finding accentuates the need for adequate local therapy for sarcomas presenting primarily as well as for multidisciplinary management of local recurrence.

Reoperation

Following staging evaluation, patients with isolated local recurrence should be considered for possible reoperation. The results of reoperation in this setting are good, with two-thirds of patients achieving long-term survival.

Adjuvant radiation therapy

If no prior radiation therapy has been employed, adjuvant irradiation (50 to 65 Gy) should be used before or after surgery for locally recurrent disease. Radiation therapy (EBRT or brachytherapy) should be considered in patients for whom previous radiation doses were subtherapeutic or the previous radiation field design permitted additional treatment. Reports from several referral centers suggest that patients who develop local recurrence following previous full-dose irradiation represent a difficult local tumor control challenge. A report from Memorial Sloan Kettering Cancer Center suggests that limb-sparing surgery combined with adjuvant brachytherapy may produce excellent local tumor control and function in this group.

ILP

Ongoing clinical investigations are defining the role of isolated limb perfusion, or ILP, for the management of patients with locally recurrent sarcoma. ILP is approved in Europe for treatment of otherwise unresectable extremity sarcomas (see section on “Isolated limb perfusion”).

Treatment of Limited Pulmonary Metastasis

Thoracotomy and metastasectomy

The most common site of metastatic disease involvement of soft-tissue sarcoma is the lungs. Rates of 3-year survival following thoracotomy for pulmonary metastasectomy range from 23% to 42%. This fact, combined with the limited efficacy of systemic therapy, is the basis for the recommendation that patients with limited pulmonary metastases and no extrapulmonary disease should undergo thoracotomy and metastasectomy. Appropriate patient selection for this aggressive therapeutic approach to metastatic disease is essential. The following are generally agreed upon criteria: (1) the primary tumor is controlled or controllable; (2) there is no extrathoracic metastatic disease; (3) the patient is a medical candidate for thoracotomy; and (4) complete resection of all disease appears to be possible.

Preresection chemotherapy

Chemotherapy is sometimes recommended before resection of pulmonary metastases. Although occasional patients may have tumor shrinkage to a degree that an unresectable tumor becomes resectable, there are no convincing data that chemotherapy impacts favorably on patient survival. It is also notable that there are no randomized data on which to base this judgment.

Chemotherapy for Unresectable Locally Advanced or Metastatic Disease

Single agents

Doxorubicin. Early trials of doxorubicin reported major responses in approximately 30% of patients with advanced soft-tissue sarcoma. In more recent randomized series, however, the rate of response has been closer to 15%. Subset analysis of patients with soft-tissue sarcoma from a broad phase II trial in which patients were randomized to receive various doses of doxorubicin demonstrated a steep dose-response relationship; patients treated with doses below 60 mg/m2 rarely responded. Whether dose intensification of doxorubicin is associated with improved survival remains an open question (see section on “Intensifying chemotherapy”).

Pegylated liposomal doxorubicin has demonstrated limited activity in phase II trials, especially in patients whose disease is refractory to standard doxorubicin. In a randomized comparison among 95 previously untreated patients, however, the response rates to pegylated liposomal doxorubicin (50 mg/m2 every 4 weeks; 10%) and to standard doxorubicin (75 mg/m2 every 3 weeks; 9%) were similar, with no significant difference in time to disease progression or survival. Response rates improved to 14% and 12%, respectively, when GIST cases were excluded.

Ifosfamide. In a randomized phase II trial conducted by EORTC, 18% of patients treated with ifosfamide (5 g/m2) experienced major responses, in contrast to 12% of patients treated with cyclophosphamide (1.5 g/m2), despite the greater myelosuppression with the latter agent. In a large American phase II trial, 17 of 99 patients with soft-tissue sarcoma responded to ifosfamide (8 g/m2). All of the patients had been treated previously with doxorubicin-based therapy, suggesting a degree of non–cross-resistance.

• Increasing ifosfamide dose-Responses to ifosfamide (≥ 12 g/m2) have been observed in patients whose disease progressed while receiving lower doses, supporting the concept of a dose-response relationship. In a randomized trial, the response to 9 g/m2 of ifosfamide (17.5%) was superior to the 3% response observed among patients treated with 5 g/m2. The reason for the low response to the lower dose was unclear. In a subsequent trial by the same investigators, the response to 12 g/m2 was only 14%, however. Among 45 evaluable patients enrolled in a Spanish phase II trial of ifosfamide (given by continuous infusion over 6 days), the response rate was 38%, but 47% of patients developed febrile neutropenia and 32% had grade 3 neurotoxicity. At MD Anderson Cancer Center, ifosfamide (14 g/m2 given by continuous infusion over 3 days) yielded responses in 29% of 37 patients with soft-tissue sarcoma and 40% of 37 patients with bone sarcoma. Also within that report was a small cohort of patients in whom the response to the same total dose of ifosfamide was higher when the drug was given by an intermittent bolus rather than a continuous infusion; this finding led the authors to suggest that bolus therapy is more efficacious than continuous infusion. Pharmacokinetic studies, however, have shown no difference between a 1-hour infusion and bolus injection of ifosfamide with respect to the area under the concentration-time curve (AUC) for serum ifosfamide or its metabolites or the levels of ifosfamide metabolites in urine. In an EORTC phase II trial, ifosfamide (12 g/m2 given as a 3-day continuous infusion every 4 weeks) yielded a response rate of 17% among 89 chemotherapy-naive patients and of 16% among 25 previously treated patients. Ifosfamide doses as high as 14 to 20 g/m2 have been given with hematopoietic growth factor support; reported response rates are high, but neurologic and renal toxicities often are dose-limiting. The available data suggest that synovial sarcoma is particularly sensitive to ifosfamide.

Dacarbazine. The activity of dacarbazine in soft-tissue sarcoma has been recognized since the 1970s and was confirmed in a formal phase II trial. This marginally active agent has been used mostly in doxorubicin-based combinations. In particular, patients with leiomyosarcoma respond better to dacarbazine than do patients with other sarcoma subtypes.

Ecteinascidin. Ecteinascidin (ET-743, trabectedin), a novel compound derived from a marine organism, has demonstrated promising activity as well. In phase I trials, trabectedin demonstrated activity in heavily pretreated patients with advanced sarcoma. Three phase II trials of trabectedin (at 1.5 mg/m2 over 24 hours every 3 weeks) in refractory non-GIST soft-tissue sarcoma have been reported.

On the basis of randomized phase II data, ecteinascidin (Yondelis) was approved in Europe for use in refractory soft-tissue sarcomas, and was also approved in the United States in October 2015. Myxoid/round cell liposarcoma appears to be the subtype with greatest sensitivity to this novel agent. Care must be taken in its administration to follow liver function tests, which predict for rhabdomyolysis and death if appropriate dose reductions are not made.

TABLE 4: Chemotherapy regimens for soft-tissue sarcoma

Other agents. Gemcitabine has demonstrated modest activity in several phase II trials, although results of a recent Southwest Oncology Group (SWOG) trial were disappointing. Taxanes, vinca alkaloids, and platinum compounds have demonstrated only marginal activity, however. It should be noted that the taxanes, gemcitabine, and vinorelbine have been observed to be active in angiosarcoma, especially cases involving the scalp and face.

Combination chemotherapy

Combination chemotherapy regimens have been used widely in the management of patients with soft-tissue sarcoma (Table 4). High response rates have been reported in a number of single-arm phase II trials. Most combination regimens include an anthracycline (either doxorubicin or epirubicin) plus an alkylating agent, dacarbazine, or both agents. Overall response rates are higher in these single-arm trials than when the same regimens are tested in larger, randomized studies.

CYVADIC and doxorubicin/dacarbazine regimens. Combinations of doxorubicin with other agents have not proved to be superior to doxorubicin alone in terms of overall survival. Also, for over a decade, the CYVADIC regimen was widely accepted as the standard of care. In a prospective, randomized trial, however, CYVADIC was not superior to doxorubicin alone.

Doxorubicin (or epirubicin) plus ifosfamide. Combinations of doxorubicin (or epirubicin) plus ifosfamide have consistently yielded responses in over 25% of patients in single-arm trials. In sequential trials conducted by the EORTC, doxorubicin at 75 mg/m2 plus ifosfamide (5 g/m2) was superior to doxorubicin at 50 mg/m2 plus ifosfamide (5 g/m2). A prospective randomized EORTC trial with 314 patients compared the two regimens. There was no difference in response rate or overall survival, but disease progression-free survival favored the more intensive regimen. The strategy of intensifying the dosing of ifosfamide within the context of combination chemotherapy was explored in a randomized phase II trial. This study included patients with localized disease treated with four cycles of preoperative chemotherapy as well as patients with metastatic disease. Overall, there was no survival benefit for patients treated with doxorubicin (60 mg/m2) plus 12 g/m2 of ifosfamide over those treated with doxorubicin (60 mg/m2) plus 6 g/m2 of ifosfamide. Also, there was no advantage to the patients with localized disease in terms of disease-free survival.

Combination chemotherapy vs single-agent doxorubicin. Combination chemotherapy has been compared with single-agent doxorubicin in eight randomized phase III trials. Two trials were limited to patients with uterine sarcoma. Some of these studies showed superior response rates with combination chemotherapy, but none of the trials found a significant survival advantage. Kaplan-Meier plots of survival are virtually superimposable within each trial and from trial to trial. A meta-analysis confirmed the higher response rate when ifosfamide is added to other agents and showed no benefit at 1 year of combination therapy over single agents, likely due to the lack of synergy between anthracyclines and ifosfamide. Patients can achieve equal benefit by sequential use of such agents rather than by combinations. It should be emphasized that approximately 20% to 25% of patients entered into such trials are alive 2 years after therapy was initiated. Complete responses are uncommon and do not appear to translate into prolonged survival.

Gemcitabine plus docetaxel. In a phase II study of 34 patients with unresectable leiomyosarcoma, mostly uterine in origin, 53% responded to a combination of gemcitabine (given by 90-minute infusion) plus docetaxel, with G-CSF support. An additional 20% had stable disease. Almost half of the patients had disease progression after anthracycline-based therapy. The median time to disease progression was 5.6 months, and grade 3 or 4 toxicity was uncommon. The activity of the gemcitabine-docetaxel combination was confirmed in a variety of other sarcoma subtypes in another study, which also confirmed the rationale for the sequence used in the study in vitro. Conversely, gemcitabine-docetaxel was examined in the adjuvant setting, in a phase II study, with no evidence of long-term benefit over historical controls.

A prospective, randomized trial comparing gemcitabine-docetaxel with gemcitabine alone in a spectrum of histologic types of sarcoma has been completed. The response rates were 8% for gemcitabine alone and 16% for gemcitabine-docetaxel. Time to disease progression and overall survival were superior with gemcitabine-docetaxel (17.9 months vs 11 months). Although this is one of the few studies in metastatic sarcoma to show a survival advantage, enthusiasm is tempered by toxicity, causing treatment discontinuation in up to 50% of patients after 6 months of gemcitabine-docetaxel chemotherapy. These data suggest that a dose reduction, of the docetaxel in particular, is needed in the off-study use of the therapy.

In the authors' experience, weekly administration of lower doses of each agent is more tolerable than the large day-8 dose of docetaxel and remains an active regimen. The randomized study previously noted that the response rate was higher in patients with high-grade undifferentiated pleomorphic sarcoma than in patients with leiomyosarcoma.

Gemcitabine plus dacarbazine. In an under-referenced randomized phase II study from Spain, the combination of gemcitabine (at 1,800 mg/m2, 10 mg/m2/min) and dacarbazine (at 500 mg/m2, both every 2 weeks) was superior to dacarbazine (1,200 mg/m2 every 3 weeks) alone, both in terms progression-free survival and overall survival. A total of 113 patients were studied. progression-free survival at 3 months was 56% for gemcitabine plus DTIC vs 37% for DTIC alone (P = .001). Median progression-free survival time was 4.2 months vs 2 months (P = .005), and median overall survival was 16.8 months vs 8.2 months (P = .014); both favored the treatment arm of gemcitabine plus DTIC. Gemcitabine plus DTIC was also associated with a higher objective response or stable disease rate than was observed with DTIC alone (49% vs 25%; P = .009). In the authors' experience, this regimen is more useful in leiomoysarcoma than in other histologies, although responses are seen with other diagnoses, such as undifferentiated pleomorphic sarcoma. It is also the authors' experience that this combination is less toxic than gemcitabine in combination with docetaxel.

Kinase-targeted agents in non-GIST sarcomas. Several of the commercially available tyrosine kinase inhibitors have been examined for activity in the setting of metastatic disease. In general, response rates have been low.

Exceptions to this statement may include a 15% response rate of patients with synovial sarcoma to pazopanib (Votrient), a 14% response rate of patients with angiosarcoma to sorafenib (Nexavar) (particularly those induced by therapeutic radiation) and similar activity of bevacizumab (Avastin) in angiosarcomas, and reports of the utility of imatinib in patients with the rare sarcoma subtype dermatofibrosarcoma protuberans. There appears to be at least modest activity of sunitinib (Sutent) in alveolar soft-part sarcoma, clear-cell sarcoma, and solitary fibrous tumor/hemangiopericytoma, and of cediranib (Recentin) in patients with metastatic alveolar soft-part sarcoma.

Activity of pazopanib was confirmed in a randomized phase III study reported by van der Graaf et al, comparing placebo to pazopanib at 800 mg orally daily (see section on recent approvals). Median progression-free survival was 20 weeks, compared with 8 weeks on placebo. These findings have led to regulatory approval of pazopanib for soft-tissue sarcomas other than liposarcoma or GIST that have gotten worse despite prior therapy.

Intensifying chemotherapy. Hematopoietic growth factors have facilitated the evaluation of dose-intensive chemotherapy in patients with sarcoma. The nonhematologic toxicities (cardiac, neurologic, and renal) of the agents most active in soft-tissue sarcoma prevent dramatic dose escalation.

Phase I/II trials of dose-intense anthracycline/ifosfamide regimens with hematopoietic growth factor support have shown that doxorubicin (70 to 90 mg/m2) can be used in combination with ifosfamide (10 to 12 g/m2) in selected patients. Response rates as high as 69% have been reported. Although toxicity increases, often dramatically, with these relatively modest dose escalations, the clinical benefit in terms of survival or palliation in patients with metastatic disease remains uncertain. No randomized trial has demonstrated a survival advantage for patients treated with these more aggressive regimens. In one randomized trial, however, the French Federation of Cancer Centers Sarcoma Group (FFCCSG) demonstrated that, in comparison with standard doses, a 25% escalation in doses of doxorubicin, ifosfamide, and dacarbazine with G-CSF support did not improve outcome.

High-dose therapy with autologous stem-cell transplantation. Most trials are small and presumably involve highly selected patients. In one trial involving 30 patients with metastatic or locally advanced sarcoma accrued over 6 years, more than 20% were free of disease progression at 5 years after high-dose therapy with stem-cell rescue. Complete response to standard induction chemotherapy predicted superior 5-year survival. Based on these favorable results, the investigators suggested a prospective randomized trial examining this approach. Although some groups are still exploring this approach, the appropriateness of generalizing these results to most patients with soft-tissue sarcoma remains speculative. In Ewing sarcoma, the results of the Euro-EWING 99 study indicate there may be benefit from high-dose therapy with stem cell support for people with limited metastatic disease at presentation, though these data are not randomized.

Prognostic factors for response to therapy. Over the past 20 years, the EORTC has collected data on more than 2,000 patients with metastatic disease who participated in first-line anthracycline-based chemotherapy trials. Multivariate analysis of these data indicated that the patients most likely to respond to chemotherapy are those without liver metastases (P < .0001), younger patients, individuals with high histologic grade, and those with liposarcoma. In this Cox model, the factors associated with superior survival were good performance status, absence of liver metastases, low histologic grade, a long time to metastasis after treatment of the primary tumor, and young age.

More recently, these same investigators have reported that the observed response rate is superior in patients who have pulmonary metastases only, as compared with those who have metastases to the lungs and other sites or to other sites only. These findings highlight the danger of reaching broad conclusions based on extrapolations from small trials that include highly selected patients. The EORTC data are also consistent with the observation that patients with metastatic GIST rarely respond to standard chemotherapy regimens. This increasingly recognized observation has been used to explain the low response rates seen in some trials.

Targeted therapy for GISTs

Advances in our understanding of the biology of GIST, and the availability of an effective therapy for patients with advanced disease, have resulted in intense interest in this entity and rapid expansion of diagnosis of this disease. Because this entity had not been recognized, the incidence of GIST was underappreciated. GIST is the most common nonepithelial tumor of the GI tract, with an estimated annual incidence of 3,000 to 3,500 cases in the United States. Approximately 50% to 60% of GISTs arise in the stomach and 25% in the small bowel. Other sites include the rest of the GI tract, the omentum, mesentery, and retroperitoneum. These tumors may range in size from millimeters to huge masses. It is not clear how many of these GISTs become clinically relevant and how many are noted incidentally at the time of endoscopic ultrasonography or other abdominal procedures.

Earlier demonstrations of the activity of imatinib in metastatic disease were expanded in two parallel, multi-institution trials in which patients with GISTs were randomized to receive imatinib (400 mg or 800 mg daily). The results were very similar. In the American trial, among 746 registered patients, the overall response rate was 43% for patients treated with 400 mg and 41% for those treated with 800 mg. There were no differences in survival between the two arms. At 2 years, progression-free and overall survival rates in the 400-mg arm were 50% and 78%, respectively. In the 800-mg arm, the rate of progression-free survival at 2 years was 53%, and the rate of overall survival was 73%. In a large European-Australasian trial, 946 patients were randomized to receive imatinib (400 mg daily or twice a day). Among the 615 patients whose response could be evaluated, there was no difference in response frequency (43%) or survival between the two arms. Complete responses were seen in 3% and 2% of the lower-dose and higher-dose patients, respectively. Sixty-nine percent of patients whose disease was progressing on 400 mg of imatinib were allowed to cross over to the higher dose (800 mg). Further therapeutic activity was seen, with 26% of these patients free of disease progression at 1 year.

Among a group of 127 patients with advanced GISTs, activating mutations of KIT or PDGFRA were identified in 87.4% and 3.9% of patients, respectively. In patients harboring an exon 11 mutation of KIT, the partial remission rate was 83.5%, whereas in patients without a discernible mutation in KIT or PDGFRA, the partial remission rate was 9.1%. The presence of an exon 11 mutation in KIT correlated with clinical response, decreased risk of treatment failure, and improved overall survival.

Expert panels have developed guidelines for the evaluation and treatment of patients with GIST. These groups recommended 400 mg daily as the initial starting dose of imatinib for exon 11 KIT mutant GIST and 800 mg oral daily for exon 9 KIT mutant GIST. Dose escalation should be considered in patients who do not respond initially or who demonstrate unequivocal disease progression. Surgery remains the primary modality for treatment of primary GIST, and adjuvant imatinib has been approved for use in Europe and the United States on the basis of a large randomized clinical trial.

Sunitinib is an active agent in imatinib-refractory GIST. In both phase I/II and III studies, the response rate is on the order of 10%, with a greater than 60% chance of these patients remaining on treatment for 6 months or longer. Notably, the benefit was greatest in patients with the converse KIT genetic phenotype (exon 9 mutation or wild type KIT) to those who were sensitive to imatinib (exon 11 mutation). Nonetheless, imatinib remains the first line of therapy regardless of mutation type, because there is still a response rate seen for imatinib in patients with wild-type KIT or exon 9 KIT mutations and because imatinib is less toxic than sunitinib in its present schedule (4 weeks on at 50 mg oral daily, 2 weeks off). Other studies continue to evaluate the benefit of other small-molecule inhibitors of KIT.

The French Sarcoma Group (FSG) conducted a phase III randomized trial evaluating intermittent vs continuous imatinib therapy after completion of 1 year of continuous imatinib therapy. A total of 159 patients have enrolled in the trial. A partial or complete response was achieved in 52% of patients. Twenty-three patients were randomized to join the intermittent arm, and 23 were randomized to the continuous arm. After 3 months, five patients (21%) in the intermittent arm had evidence of disease progression, vs no patients in the continuous arm. Reintroduction of imatinib resulted in tumor control in all patients.

Assessment of response and treatment after disease progression on imatinib. The use of standard (RECIST [Response Evaluation Criteria in Solid Tumors]) response criteria in patients with GIST may be misleading. On CT or MR imaging, large tumor masses may become completely necrotic without a reduction in size for months in spite of dramatic clinical improvement. Indeed, such masses may actually increase in size. 18F-FDG (18F-fluorodeoxyglucose)–PET imaging may be useful in selected patients, since a response may be seen as early as 24 hours after a dose of imatinib. However, in general, with the understanding that loss of tumor vascularity represents a good radiological result, very little is to be gained with the use of PET vs IV contrast–enhanced CT scan or MRI. In patients who cannot have IV CT or MRI contrast, PET may provide an alternate means of following treatment efficacy over time. It should be noted that the survival of patients with stable disease parallels that of patients with major objective responses using RECIST criteria.

Surgery does not cure GIST that recurs after resection of primary disease and should be managed as metastatic disease. However, multimodality therapy should be considered in patients with limited sites of disease. Patients whose disease demonstrates partial response, stable disease, or even limited/unifocal progression on imatinib may benefit from aggressive cytoreductive surgery. Those with generalized/multifocal progression likely derive little, if any, benefit from surgery, and should not routinely be considered as surgical candidates. Identifying patients who may benefit from surgery while on sunitinib is more challenging. If patients undergo surgery for metastatic GIST, they should resume their prior tyrosine kinase inhibitor (TKI) therapy after surgery, as may experience rapid disease progression otherwise. Thus, TKI therapy should be continued indefinitely in such patients.

US Food and Drug Administration (FDA)-approved therapy for metastatic disease includes imatinib in first-line and sunitinib in second-line treatment. If there is disease progression on second-line sunitinib, regorafenib (Stivarga) is now approved in the United States and elsewhere for metastatic GIST. A phase III study tested regorafenib in a double-blind placebo controlled study. This study had a crossover design, so a survival advantage was neither anticipated nor observed. Progression-free survival was significantly better for patients randoized to regorafenib vs placebo (4.8 months vs 0.9 months; HR = 0.27; P < .0001). As a result, regorafenib is a new third-line standard of care for metastatic GIST.

Adjuvant imatinib for GIST

The American College of Surgeons Oncology Group (ACOSOG) phase III study of adjuvant imatinib examined imatinib for 48 weeks vs placebo. The study assessed patients with any GIST of at least 3 cm in maximum dimension. DeMatteo et al indicated that there was only a 3% chance of disease progression after the 48 weeks of therapy, in comparison to 17% in those who underwent surgery alone. Furthermore, there were certain mutational subtypes in which it was clear that 400 mg orally per day for 48 weeks did not improve progression-free survival over placebo, including patients with exon 9 mutations in KIT, so-called wild-type GIST, and patients with PDGFRA mutation D842V. Thus, if adjuvant therapy is to be given, patients with the more common exon 11 KIT mutations (or the rare PDGFRA mutant non-D842V) could be considered for therapy.

The successor to the Z9001 study that now defines the standard of care was the Scandinavian Sarcoma Group (SSG) XVII study, examining 1 year vs 3 years of imatinib for intermediate- to high-risk GIST. First reported by Joensuu et al in 2012 and more recently in 2015, the results showed for the first time not only a progression-free survival advantage to the longer course of therapy but also an overall survival advantage for longer exposure to imatinib (with 5-year survival rates of 92% for 3 years of imatinib vs 82% for 1 years of treatment). As a result, 3 years of imatinib is the present standard of care for high-risk GIST. Heat maps providing an accurate assessment of risk are now available, enabling more accurate discussion of the risk and potential benefit of imatinib in the adjuvant setting.

An ongoing phase II single-arm study (PERSIST-5 [NCI Clinical.Trials.gov identifier: NCT00867113]) is evaluating the role of 5 years of adjuvant imatinib following resection of primary GIST. The trial has completed accrual, and all patients will have completed 5 years of therapy by mid 2016.

Recommendations for the treatment of metastatic sarcoma

• For patients with rapidly progressive disease or with symptoms, combination chemotherapy with an anthracycline/ifosfamide combination is indicated. For most patients, however, sequential single-agent therapy is less toxic and not inferior in terms of survival.

• The management of metastatic GIST involves imatinib as first-line therapy, and increasing doses of imatinib, when feasible, before changing to sunitinib. Some patients can be maintained with good responses to imatinib for more than 5 years.

• Regorafenib is a third-line option for metastatic GIST after failure of prior therapy.

• Three years of adjuvant imatinib is a standard of care for higher-risk GIST and is associated with a survival advantage. A simple rule of thumb regarding high-risk patients includes the rule of fives. High-risk tumors include gastric GISTs that are both 5 cm or more in size with at least five mitoses per 50 high-powered fields, and GIST from other sites are higher risk if they are either 5 cm or more in size or have at least five mitoses per 50 high-powered fields.

• Surgery is increasingly being performed at the time of best response (typically 6 to 9 months) and at the time of limited disease progression. There are no data that indicate that early surgery (at the time of best response) leads to superior survival than later surgery (at the time of limited disease progression). Surgery should generally be avoided in patients with multifocal progressive disease; a change in medical therapy is appropriate in this setting.

• The importance of histology relevant to selection of therapy is increasingly being appreciated. It is especially significant to distinguish GISTs from GI leiomyosarcomas. Patients with GIST that is progressive on standard therapy should be referred to subspecialty centers experienced in the multimodality management of this disease. Data regarding kinase-targeted agents will likely lead to the use of such agents for a limited number of rare sarcoma subtypes.

• Periods of watchful waiting may be appropriate for many patients with metastatic sarcoma who have no or only minimal symptoms.

Suggested Reading

Blay JY, Le Cesne A, Ray-Coquard I, et al: Prospective multicentric randomized phase III study of imatinib in patients with advanced gastrointestinal stromal tumors comparing interruption versus continuation of treatment beyond 1 year: The French Sarcoma Group. J Clin Oncol 25:1107–1113, 2007.

Cherix S, Speiser M, Matter M, et al: Isolated limb perfusion with tumor necrosis factor and melphalan for non-resectable soft tissue sarcomas: Long-term results on efficacy and limb salvage in a selected group of patients. J Surg Oncol 98:148–155, 2008.

Choi H, Charnsangavej C, Faria SC, et al: Correlation of computed tomography and positron emission tomography in patients with metastatic gastrointestinal stromal tumor treated at a single institution with imatinib mesylate: Proposal of new computed tomography response criteria. J Clin Oncol 25:1753–1759, 2007.

Corless CL, Ballman CV, Antonescu C, et al: Relation of tumor pathologic and molecular features to outcome after surgical resection of localized primary gastrointestinal stromal tumor (GIST): Results of the intergroup phase III trial ACOSOG Z9001. J Clin Oncol 28(suppl 15s): abstract 10006, 2010.

Corless CL, Ballman CV, Antonescu C, et al: Pathologic and molecular features correlate with long-term outcome after adjuvant therapy of resected primary GI stromal tumor: the ACOSOG Z9001 trial. J Clin Oncol 32:1563–1570, 2014.

Cormier JN, Huang X, Xing Y, et al: Cohort analysis of patients with localized, high-risk, extremity soft tissue sarcoma treated at two cancer centers: Chemotherapy-associated outcomes. J Clin Oncol 22:4567–4574, 2004.

Davis A, O'Sullivan B, Turcotte R, et al: Late radiation morbidity following randomization to preoperative versus postoperative radiotherapy in extremity soft tissue sarcoma. Radiother Oncol 75:48-53, 2005.

DeLaney TF, Spiro IJ, Suit HD, et al: Neoadjuvant chemotherapy and radiotherapy for large extremity soft-tissue sarcomas. Int J Radiat Oncol Biol Phys 56:1117–1127, 2003.

DeMatteo RP, Ballman KV, Antonescu CR, et al: Adjuvant imatinib mesylate after resection of localised, primary gastrointestinal stromal tumour: a randomised, double-blind, placebo-controlled trial. Lancet 373:1097–1104, 2009.

Demetri GD, van Oosterom AT, Garrett CR, et al: Efficacy and safety of sunitinib in patients with advanced gastrointestinal stromal tumour after failure of imatinib: A randomised controlled trial. Lancet 368:1329–1338, 2006.

Demetri GD, Reichardt P, Kang Y-K et al: Efficacy and safety of regorafenib for advanced gastrointestinal stromal tumours after failure of imatinib and sunitinib (GRID): An international, multicentre, randomised, placebo-controlled, phase 3 trial. Lancet 281:295–302, 2013.

Eggermont AM, Schraffordt Koops H, Klausner JM, et al: Isolated limb perfusion with tumor necrosis factor and melphalan for limb salvage in 186 patients with locally advanced soft tissue extremity sarcomas: The cumulative multicenter European experience. Ann Surg 224:756–764, 1996.

Eilber FC, Rosen G, Nelson SD, et al: High-grade extremity soft-tissue sarcomas: Factors predictive of local recurrence and its effect on morbidity and mortality. Ann Surg 237:218–226, 2003.

Feng M, Murphy J, Griffith KA, et al: Long-term outcomes after radiotherapy for retroperitoneal and deep truncal sarcoma. Int J Radiat Oncol Biol Phys 69:103–110, 2007.

Frustaci S, De Paoli A, Bidoli E, et al: Ifosfamide in the adjuvant therapy of soft tissue sarcomas. Oncology 65(suppl 2):80–84, 2003.

García-del-Muro X, Lopez-Pousa A, Maurel J, et al: Randomized phase II study comparing gemcitabine plus dacarbazine versus dacarbazine alone in patients with previously treated soft tissue sarcoma. J Clin Oncol 29:2528–2533, 2011.

Heinrich MC, Corless CL, Blanke CD, et al: Molecular correlates of imatinib resistance in gastrointestinal stromal tumors. J Clin Oncol 24:4764–4774, 2006.

Heinrich MC, Maki RG, Corless CL, et al: Primary and secondary kinase genotypes correlate with the biological and clinical activity of sunitinib in imatinib-resistant gastrointestinal stromal tumor. J Clin Oncol 26:5352–5359, 2008.

Hensley ML, Wathen K, Maki RG, et al: Adjuvant treatment of high-risk primary uterine leiomyosarcoma with gemcitabine/docetaxel, followed by doxorubicin: Results of phase II multicenter trial SARC005. Cancer 119:1555–1561, 2013.

Joensuu H, Eriksson M, Hatrmann J, et al: One vs three years of adjuvant imatinib for operable gastrointestinal stromal tumor: A randomized trial. JAMA 307:1265–1272, 2012.

Joensuu H, Eriksson M, Sundby K, et al: Adjuvant imatinib for high-risk GI stromal tumor: analysis of a randomized trial. J Clin Oncol November 2, 2015. [Epub before print]

Kraybill WG, Harris J, Spiro IJ, et al: Phase II study of neoadjuvant chemotherapy and radiation therapy in the management of high-risk, high-grade, soft tissue sarcomas of the extremities and body wall: Radiation Therapy Oncology Group Trial 9514. J Clin Oncol 24:619–625, 2006.

Ladenstein R, Potschger U, Le Deley MC, et al: Primary disseminated multifocal Ewing Sarcoma: Results of the Euro-EWING 99 trial. J Clin Oncol 28:3284–3291, 2010.

Le Cesne A, Blay JY, Judson I, et al: Phase II study of ET-743 in advanced soft tissue sarcomas: A European Organisation for the Research and Treatment of Cancer (EORTC) soft tissue and bone sarcoma group trial. J Clin Oncol 23:576–584, 2005.

Maki RG, D'Adamo DR, Keohan ML, et al: Phase II study of sorafenib in patients with metastatic or recurrent sarcomas. J Clin Oncol 27:3133–3140, 2009.

O'Sullivan B, Davis AM, Turcotte R, et al: Preoperative versus postoperative radiotherapy in soft-tissue sarcoma of the limbs: A randomized trial. Lancet 359:2235–2241, 2002.

Pervaiz N, Colterjohn N, Farrokhyar F, et al: A systematic meta-analysis of randomized controlled trials of adjuvant chemotherapy for localized resectable soft-tissue sarcoma. Cancer 113:573–581, 2008.

Pisters PW, Patel SR, Prieto VG, et al: Phase I trial of preoperative doxorubicin-based concurrent chemoradiation and surgical resection for localized extremity and body wall soft tissue sarcomas. J Clin Oncol 22:3375–3380, 2004.

Pisters PW, Pollock RE, Lewis VO, et al: Long-term results of prospective trial of selective use of radiation for patients with T1 extremity and trunk soft tissue sarcomas. Ann Surg 246:675–682, 2007.

Raut CP, Posner M, Desai J, et al: Surgical management of advanced gastrointestinal stromal tumors after treatment with targeted systemic therapy using kinase inhibitors. J Clin Oncol 24:2325–2331, 2006.

Riad S, Griffin AM, Liberman B, et al: Lymph node metastasis in soft tissue sarcoma in an extremity. Clin OrthopRelat Res Sep:129–134, 2004.

van der Graaf WT, Blay JY, Chawla SP, et al: Pazopanib for metastatic soft-tissue sarcoma (PALETTE): A randomised, double-blind, placebo-controlled phase 3 trial. Lancet 379:1879–1886, 2012.

Verma S, Younus J, Stys-Norman D, et al: Meta-analysis of ifosfamide-based combination chemotherapy in advanced soft tissue sarcoma. Cancer Treat Rev 34:339–347, 2008.

Wendtner C-M, Abdel-Rahman S, Krych M, et al: Response to neoadjuvant chemotherapy combined with regional hyperthermia predicts long-term survival for adult patients with retroperitoneal and visceral high-risk soft-tissue sarcomas. J Clin Oncol 20:3156–3164, 2002.

Zalcberg JR, Verweij J, Casali PG, et al: Outcome of patients with advanced gastro-intestinal stromal tumours crossing over to a daily imatinib dose of 800 mg after progression on 400 mg. Eur J Cancer 41:1751–1757, 2005.