INTRODUCTION — Sarcomas are malignant tumors that arise from skeletal and extraskeletal connective tissues, including the peripheral nervous system. The majority of soft tissue sarcomas (STS) present in the extremities; however, many other sites can be affected, including the retroperitoneum, chest wall, head and neck, and subcutaneous tissues.
The clinical features, evaluation, and initial treatment of STS arising in the retroperitoneum will be reviewed here. A general discussion of the diagnostic evaluation and staging of STS in general, treatment for locally recurrent retroperitoneal soft tissue sarcoma (RPS), local and systemic therapies for metastatic disease, treatment of desmoid tumors (which may present in intraabdominal and abdominal wall sites), treatment of gastrointestinal stromal tumors (mesenchymal tumors that usually present in the wall of the gastrointestinal tract but occasionally present in extravisceral abdominal locations), and treatment of solitary fibrous tumors are presented elsewhere:
●(See "Management of locally recurrent retroperitoneal sarcoma".)
●(See "Surgical treatment and other localized therapy for metastatic soft tissue sarcoma".)
●(See "Systemic treatment of metastatic soft tissue sarcoma".)
●(See "Solitary fibrous tumor".)
EPIDEMIOLOGY — RPS are relatively uncommon, constituting only 10 to 15 percent of all soft tissue sarcomas [1]. In a population-based series from the Surveillance, Epidemiology, and End Results (SEER) database, the average annual incidence of RPS was approximately 2.7 cases per million population [2].
Patients usually present in their 50s, although the age range is broad [3-5]. The frequency is approximately equal in men and women.
ANATOMY, HISTOLOGY, AND DIFFERENTIAL DIAGNOSIS — The retroperitoneum is the space just posterior to the peritoneal cavity and anterior to the paraspinous musculature (eg, psoas major, psoas minor, quadratus lumborum, obturator internus, pyriformis). Superiorly, the retroperitoneum is bordered by the diaphragm, and inferiorly, it forms a natural extension to the pelvis (pelvic diaphragm) [6]. The structures of the retroperitoneum include the kidneys, adrenal glands, and perirenal fat bilaterally, the aorta and its major branches (eg, renal arteries), the inferior vena cava and its major tributaries (eg, renal veins), and the bilateral iliac vessels (common, internal, external arteries/veins), as well as the duodenum and pancreas. The ascending and descending colon are considered partially retroperitoneal (figure 1).
Approximately 80 percent of the neoplasms that arise within the retroperitoneal space are malignant. Furthermore, the majority of patients who present with a primary retroperitoneal, extravisceral, unifocal soft tissue mass will be found to have a sarcoma.
Histologic types — In adults, the most common histologic types of RPS are liposarcoma and leiomyosarcoma, followed by undifferentiated/unclassified soft tissue sarcoma (STS; a subset of which are pleomorphic undifferentiated sarcomas). Historically, undifferentiated/unclassified STS were included within the older term "malignant fibrous histiocytoma." A variety of other histologic types may be observed, but they are much less common in the retroperitoneum than in other primary sites. Approximately one-half of all RPS are high-grade tumors, although this varies according to histology. The majority of retroperitoneal liposarcomas are low- to intermediate-grade lesions. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Histopathology'.)
Among children, the most common histologic types of RPS are extraskeletal Ewing sarcoma/primitive neuroectodermal tumor (PNET), alveolar rhabdomyosarcoma, and fibrosarcoma [7]. The remainder of this review will focus on the tumor types that are most common in adults. (See "Clinical presentation, staging, and prognostic factors of Ewing sarcoma" and "Rhabdomyosarcoma in childhood and adolescence: Clinical presentation, diagnostic evaluation, and staging".)
Liposarcomas — Among the different variants of liposarcoma, well-differentiated (low-grade) liposarcomas are the most common, followed by dedifferentiated liposarcomas. Myxoid, round cell, and pleomorphic liposarcomas are uncommonly found in the retroperitoneum.
Microscopically, well-differentiated liposarcomas consist of a background of adipocytes that contain scattered lipoblasts, each with a single atypical nucleus surrounded by large intracytoplasmic vacuoles. There is often an inflammatory cell infiltrate. Hypercellular well-differentiated liposarcomas may be confused with dedifferentiated liposarcomas [8].
Well-differentiated liposarcomas have no potential to metastasize; as such, they are referred to as atypical lipomatous tumors when they arise in the body wall/trunk or extremity [9]. However, given their propensity for local recurrence in the retroperitoneum/mediastinum and spermatic cord, these same tumors are referred to as well-differentiated liposarcomas in these locations [9].
Dedifferentiated liposarcomas are defined by the presence of sharply demarcated regions of nonlipogenic sarcomatous tissue within a well-differentiated tumor [9]. They may be difficult to distinguish from undifferentiated/unclassified "pleomorphic" sarcomas [9-12].
Well-differentiated and dedifferentiated liposarcomas have morphologic and cytogenetic similarities [9,13] but significantly different biologic behavior. Dedifferentiated liposarcomas are high-grade tumors. Compared with well-differentiated low-grade liposarcomas, they have higher local recurrence rates, the potential to metastasize (20 to 30 percent distant recurrence rate versus 0 percent for well-differentiated liposarcomas), and a sixfold higher risk of death [14,15]. It is thought that high-grade dedifferentiated liposarcomas arise from well-differentiated liposarcomas and that well-differentiated liposarcomas can recur as the more aggressive dedifferentiated subtype.
Leiomyosarcomas — Leiomyosarcomas of the retroperitoneum generally arise from the inferior vena cava, its tributaries, or any small vessel. They often present as a mass or, occasionally, with unilateral or bilateral lower extremity swelling, and they are usually of a large size when diagnosed.
The clinical issues presented by a large leiomyosarcoma differ from those of a well-differentiated/dedifferentiated liposarcoma. Lung (more than liver) metastases are sometimes observed at the time of presentation or appear relatively quickly after diagnosis. Leiomyosarcomas may also arise from the wall of the gastrointestinal tract or from the uterus. In this situation, they are visceral rather than retroperitoneal, and they have a greater risk of peritoneal spread and metastasis to the liver. For these reasons, follow-up after complete resection of a leiomyosarcoma of the abdomen or retroperitoneum should involve imaging of the chest as well as the abdomen and pelvis on a regular schedule, as noted for other RPS. (See 'Posttreatment follow-up' below.)
Others — Other less common histologic types include malignant peripheral nerve sheath tumor, undifferentiated sarcoma, synovial sarcoma, solitary fibrous tumor, and desmoplastic small round cell tumor [16]. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma" and "Solitary fibrous tumor".)
Differential diagnosis — Rarely, lymphoma, a primary germ cell tumor, or metastatic testicular cancer presents as a retroperitoneal mass. Neoplasms originating in the duodenum, pancreas, adrenal glands, or kidney can usually be distinguished from extravisceral retroperitoneal soft tissue masses based upon radiographic appearance. (See "Clinical manifestations, diagnosis, and staging of testicular germ cell tumors" and "Extragonadal germ cell tumors involving the mediastinum and retroperitoneum".)
Other benign and malignant tumors, such as schwannomas and paragangliomas, should be considered in the differential diagnosis of a retroperitoneal mass, particularly one located in the midline adjacent to the aorta or vena cava.
Two typically benign processes, Castleman disease (angiofollicular lymph node hyperplasia) and retroperitoneal fibrosis, arise within the retroperitoneum and can mimic STS both clinically and radiographically. (See "Unicentric Castleman disease" and "Clinical manifestations and diagnosis of retroperitoneal fibrosis".)
CLINICAL PRESENTATION — RPS typically produce few symptoms until they are large enough to compress or invade surrounding structures. Most cases come to medical attention as an incidentally discovered abdominal mass in an asymptomatic or minimally symptomatic patient. Most tumors are already large and locally advanced at the time they are first detected (median size at diagnosis is approximately 15 cm [3]).
Some patients present with pain or signs and/or symptoms that are related to mass effect of the tumor on surrounding structures. Local invasion or compression of retroperitoneal neurovascular structures can result in lower extremity edema and neurologic or musculoskeletal symptoms that may be referred to the lower extremities. Gastrointestinal symptoms, such as early satiety, obstruction, and/or bleeding, can also occur.
Serous ascites due to portal vein compression has been described [17,18]. Rarely, patients with high-grade, rapidly expanding tumors may experience flu-like symptoms and present with fevers and leukocytosis [17,19].
Rarely, leiomyosarcomas present with paraneoplastic hypoglycemia, which is usually secondary to tumor production of "big" insulin-like growth factor 2 (IGF-2) [17,18,20]. (See "Nonislet cell tumor hypoglycemia".)
Distant metastases (most commonly to the lung and liver) are present at the time of diagnosis in approximately 10 percent of cases [21].
DIAGNOSTIC EVALUATION AND STAGING — Given that lymphomas and germ cell tumors may occasionally present as a retroperitoneal mass, the initial history should always include questions as to symptoms related to lymphoma (eg, fever, night sweats, weight loss). The physical examination should include palpation of all nodal basins and, in men, a careful testicular examination. A normal testicular exam does not exclude the diagnosis of testicular cancer metastatic to the retroperitoneal nodes; thus, an ultrasound of the testicles should be strongly considered in the evaluation of a young man with a newly diagnosed retroperitoneal mass.
Laboratory studies should include measurement of lactate dehydrogenase (LDH; an elevated level may be suggestive of lymphoma), alpha-fetoprotein (AFP), and beta-human chorionic gonadotropin (beta-hCG; which if elevated, raises the suspicion for a germ cell tumor).
Radiographic evaluation — Radiographic imaging is a key component of the evaluation of a patient with a retroperitoneal mass. The preferred diagnostic studies are a contrast-enhanced computed tomography (CT) scan of the abdomen and pelvis to evaluate the primary site as well as a chest CT to rule out metastatic disease to the lungs. The lungs are the first site of metastasis in the majority of cases. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Pattern of spread'.)
In most cases, CT is less sensitive to motion artifact than magnetic resonance imaging (MRI), and it better defines the anatomic relationship of the tumor to other abdominal organs. It can also detect metastasis to the liver or peritoneum. The radiographic appearance of the primary tumor on a CT scan can also offer clues as to the histologic subtype and grade, which may guide decisions as to the need for a pretreatment biopsy as well as treatment.
MRI with gadolinium is reserved for patients with an allergy to iodinated contrast agents or if there is equivocal muscle, bone, or foraminal involvement on CT. MRI may also be useful for delineating disease in the pelvis. For patients in whom preoperative radiation therapy is being considered, MRI is useful for assessing local tumor extent and surrounding edema, which are optimally included in the treatment volume [22]. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'MRI and CT'.)
A role for positron emission tomography (PET) with fluorodeoxyglucose (FDG) in the initial staging evaluation is not established. PET scanning can achieve whole body imaging, and it is widely considered to be more sensitive than CT for the detection of occult distant metastases in a variety of solid tumors. One reported benefit in sarcomas is to detect sites of extrapulmonary metastatic disease. However, the risk is so low with most soft tissue sarcomas (STS) that routine use of PET for this purpose is unlikely to change the therapeutic plan. As a result, neither PET nor integrated PET/CT is routinely recommended as a component of the initial staging evaluation for STS, including those arising in the retroperitoneum. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Radiographic studies'.)
Criteria for unresectability — Radiographic findings that indicate unresectability include [23]:
●Extensive vascular involvement (aorta, vena cava, and/or iliac vessels), although involvement of the vena cava and iliac veins is a relative, rather than absolute, contraindication as these vessels can often be ligated or replaced with interposition grafts.
●Peritoneal implants.
●Distant metastases that are not potentially resectable for cure. (See "Surgical treatment and other localized therapy for metastatic soft tissue sarcoma".)
●Involvement of the root of the mesentery (specifically, the superior mesenteric vessels).
●Spinal cord involvement.
Need for biopsy — Clinicians should have a low threshold to proceed with a percutaneous core needle biopsy as the very small risk of complications and the unsubstantiated fear of tumor seeding of the biopsy tract are far outweighed by the information obtained from the biopsy. The need for and scope of surgery can vary dramatically based upon histology, and this information can be vital in terms of developing a rational multidisciplinary approach to the patient, which may entail referral to a center specializing in the treatment of the given tumor.
A biopsy is clearly indicated if the diagnosis is in doubt or if preoperative (neoadjuvant) therapy is planned (eg, for a locally advanced unresectable lesion or in the case of radiographic suspicion of focal dedifferentiation [eg, focal nodular/water density areas within an otherwise fat-containing retroperitoneal mass] [24]). At least some data support the view that percutaneous core needle biopsy of RPS is safe and does not increase the risk of local recurrence or adversely impact overall survival [25,26]. Tumor seeding of the biopsy tract is rare (only 2 suspected out of 547 cases in one report [27]). (See 'Preoperative therapy' below.)
However, initial surgery without biopsy is an acceptable alternative if the radiographic diagnosis is likely benign lipoma or well-differentiated liposarcoma, or when the diagnosis of STS seems certain based on cross-sectional imaging and the patient is not a candidate for preoperative therapy.
Staging — RPS are staged using the Tumor, Node, Metastasis (TNM) system as defined by the combined American Joint Committee on Cancer (AJCC)/Union for International Cancer Control (UICC). The most recent (eighth edition, 2017) version has separate T stage classifications and prognostic stage groupings for RPS (table 1) [28].
The earlier (seventh edition) AJCC staging system for STS at all sites was not in widespread use for RPS because it did not account for the prognostic importance of disease site or histology, two major prognostic indicators. In addition, its ability to discriminate outcome was limited [29]. Several studies have found no prognostic role for tumor size in RPS [3,4,10,11,30-32]. Furthermore, most patients with resectable RPS present with large lesions (T2 (table 2)) without nodal or distant metastases. The new AJCC staging system includes a separate staging system for RPS, but T stage is still stratified according to size.
Given that histologic grade, the presence or absence of metastatic disease, and achieving macroscopic (gross) total resection are the major determinants of survival for patients with RPS, alternative staging systems have been proposed, one of which is depicted in the table (table 3), but none are in widespread use. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Staging'.)
TREATMENT
Importance of multidisciplinary evaluation and management — Due to the rarity of these tumors and the complexity of treatment, evaluation and management should be carried out in a center with multidisciplinary expertise in the treatment of sarcomas. The multidisciplinary team approach by dedicated subspecialists at a sarcoma referral center optimizes treatment planning, minimizes the duplication of diagnostic studies, reduces the time to implementation of the definitive therapeutic protocol, and improves clinical outcomes in patients with RPS [33].
Overview of the approach to multidisciplinary treatment — There is no "one size fits all" approach to treatment of RPS. Surgical resection has traditionally been the only potentially curative approach. Because of the typically large size and anatomic complexity of these tumors at diagnosis, a resection with microscopically negative margins (R0 resection) is often not achieved, and locoregional recurrence is common, especially for low-grade tumors. Most sarcoma specialists would define complete resection of RPS as R0/R1, with ideally negative microscopic margins (R0), but with acceptance of the possibility of positive microscopic margins. Yet, the role of adjunctive therapy (radiation therapy [RT], chemotherapy, either given preoperatively or postoperatively) continues to be debated, and there is no consensus as to the best approach for all patients.
Decisions about therapy should be made by sarcoma experts in a multidisciplinary tumor board, taking into account not only histology and size, but also tumor location, and whether the surgeon feels that there is a risk for positive margins at the time of resection.
However, even among centers of excellence, there is widespread disagreement as to the optimal way to approach RPS and a paucity of high-level evidence to support any approach. This is reflected in the widely disparate recommendations from expert groups. Neither the National Comprehensive Cancer Network (NCCN) [34] nor the European Society of Medical Oncology (ESMO) [35] provides guidance in selecting treatment for individual patients with RPS. However, guidelines from ESMO do mention chemotherapy as an appropriate option for chemosensitive histologies, such as synovial sarcoma, and choosing RT for solitary fibrous tumor due to its radiosensitivity. (See 'Recommendations from expert groups' below.)
There are many acceptable treatment alternatives, and clinical practice is variable among different institutions. We tend to base our treatment approach on a multidisciplinary assessment of resectability, histology, and the likely pattern of recurrence:
●For all patients, we suggest participation in ongoing clinical trials, where available.
●If protocol therapy is not available or participation is not feasible, the following reflects our general treatment principles:
●Debulking surgery (ie, partial resection of a tumor leaving behind grossly positive margins [R2 resection]) should be avoided for most patients, with the possible exception of large unresectable well-differentiated liposarcomas, for which palliative debulking surgery may improve symptoms and prolong survival. (See 'Role of debulking' below and "Surgical resection of retroperitoneal sarcoma", section on 'Palliative resection'.)
●For most patients, if the preoperative staging evaluation suggests a high likelihood of grossly positive margins, preoperative therapy (RT, chemotherapy) is appropriate. (See 'Preoperative therapy' below.)
●In other cases of apparently resectable RPS, the role of preoperative RT is controversial, and clinical practice varies widely. At some institutions, including that of some of the authors and editors associated with this review, preoperative RT is offered to all patients with larger well-differentiated liposarcomas (although there are no evidence-based guidelines on the appropriate tumor size to consider for initial RT, and decision making must be individualized), and for those with intermediate- and high-grade tumors with histologies that are not chemotherapy sensitive. Those who favor preoperative RT argue that these patients tend to die from local recurrence and not distant disease, unless they dedifferentiate, and that combined therapy optimizes local control, which was shown in the randomized phase III STRASS trial.
On the other hand, at other institutions, including those of some of the authors and editors associated with this review, preoperative RT is rarely utilized unless the preoperative staging evaluation suggests a high likelihood of a grossly positive margin. Proponents of this approach argue that the risk for local recurrence is based more on biology than on specific treatments, and they point to the lack of benefit from preoperative RT versus surgery alone (using a composite endpoint of abdominal recurrence-free survival) in the randomized phase III STRASS trial.
In our view, either approach is acceptable. (See 'Preoperative radiation therapy' below.)
●A role for concurrent administration of preoperative RT and chemotherapy (chemoradiation) in patients with RPS is not established, and for most patients, we suggest not pursuing this approach. In our view, given the lack of data from prospective trials that preoperative chemoradiation is more effective than preoperative RT and the potential for treatment-related toxicity, this approach should only be used for RPS in the context of a clinical trial by clinicians experienced with a chemoradiation approach that has been validated in a prior clinical trial. (See 'Concurrent chemoradiation' below.)
●Although this is another controversial area, for patients with intermediate- and high-grade tumors, and locally advanced tumors with histologies that are chemotherapy sensitive (eg, synovial sarcoma, myxoid/round cell liposarcoma) and/or when the risk of distant metastatic disease is high (eg, leiomyosarcoma of the inferior vena cava or large dedifferentiated liposarcoma), we tend to favor neoadjuvant chemotherapy, with or without preoperative RT. The choice of chemotherapy regimen should be based upon the results of small series in which preoperative chemotherapy was used for patients with RPS. Benefit of histology-driven treatment has not been shown in any localized soft tissue sarcoma (STS), including RPS. In patients with large, high-grade leiomyosarcomas and dedifferentiated liposarcomas, the addition of neoadjuvant chemotherapy to surgery is being investigated in a randomized trial (STRASS-2) conducted by the European Organisation for Research and Treatment of Cancer (EORTC) [36]. (See 'Neoadjuvant chemotherapy' below.)
●Although not used in the United States, preoperative, regional, externally delivered deep wave hyperthermia combined with systemic chemotherapy is an option for preoperative therapy for high-risk primary tumors in parts of the world where it is available (mainly Germany). Whether this approach is superior to surgery plus preoperative RT is unknown and will require a randomized trial. (See 'With regional hyperthermia' below.)
●The use of adjuvant therapy following resection of an RPS without neoadjuvant therapy depends on the tumor grade and completeness of resection:
•We do not suggest adjuvant RT for a low-grade grossly completely resected (R0/R1) tumor. Postoperative RT could be considered for patients with a high- or intermediate-grade tumors who are at risk for local recurrence; however, in practice, most patients are just observed because it is rarely possible to deliver postoperative adjuvant RT with acceptable morbidity. (See 'Management of microscopically positive margins' below and 'Adjuvant radiation therapy' below.)
•Following an R2 resection, residual tumor frequently abuts structures or organs that were judged to be not easily or safely resected. For these patients, a small-field postoperative boost dose of RT is a reasonable approach if bowel and other dose-limiting normal tissues can be adequately spared (which might require placement of omentum or a tissue spacer at the time of surgery to displace bowel from the area of residual tumor). Another option, if the surgeon anticipates a grossly incomplete resection, is intraoperative radiation therapy (IORT) for an additional dose of 10 to 15 Gy to areas of residual microscopic or gross disease. (See 'Adjuvant radiation therapy' below and 'Role of intraoperative radiation therapy' below.)
•Adjuvant chemotherapy cannot be considered a standard approach for STS at any site, including the retroperitoneum, and we do not suggest its use outside of the context of a clinical trial. (See 'Adjuvant chemotherapy' below.)
●Given that late recurrences are common with RPS, long-term follow-up to at least 10 years is mandatory. (See 'Posttreatment follow-up' below.)
Surgery
Tumor resection — Surgical resection has traditionally been the only potentially curative treatment for localized RPS. Preoperative evaluation, preparation, and the conduct of the surgery are reviewed separately. (See "Surgical resection of retroperitoneal sarcoma".)
The ability to perform a complete surgical resection (R0/R1) at the time of initial presentation is the most important prognostic factor for survival [3,5,17,37-42]. The usual reasons for unresectability are extensive vascular involvement or the presence of multiple peritoneal implants. (See 'Criteria for unresectability' above and "Surgical resection of retroperitoneal sarcoma", section on 'Retroperitoneal exploration'.)
The primary oncologic goal is microscopically negative (R0) resection. However, the large size of most RPS, coupled with the inability to obtain wide margins due to anatomic constraints, makes this goal difficult to achieve. In clinical practice, for many resections, the tumor is grossly removed but the margins are microscopically positive (ie, R1 resection [40,41]). In a review of four large series of patients with RPS from institutions with extensive experience in management of this disease [3-5,43], complete resection rates were between 50 and 67 percent [44]. In the largest of these series, the "complete" (R0 plus R1) and R0 resection rates were 80 and 58 percent, respectively [5]. The implications and management of a microscopically positive margin are discussed below. (See 'Management of microscopically positive margins' below and "Surgical resection of retroperitoneal sarcoma", section on 'Resection of primary disease'.)
Resection of adjacent organs, such as the small bowel, colon, or kidney, is often required to achieve complete resection [18,45]. Liberal en bloc resection of adjacent viscera may allow a subset of patients who might otherwise have been considered unresectable to achieve wide, macroscopically negative surgical margins. The most common organs removed, in order of frequency, are the kidneys, colon, spleen, and pancreas [4]. Despite the frequent need for organ resection, only a minority of the resected organs have histopathologic organ invasion (25 percent in one series [46]). Although this practice results in lower rates of microscopically positive margins and better local control [3,4,47-49], it is not clear that additional resection (eg, compartmental resection) will translate into improvements in sarcoma-specific survival or whether this approach is preferred over preoperative RT followed by surgery (see 'Preoperative radiation therapy' below). Guidelines are available from two consensus groups that attempt to standardize the approach to front-line extended surgery for RPS in order to optimize outcomes [50,51]. (See "Surgical resection of retroperitoneal sarcoma", section on 'Tumor resection'.)
The same surgical considerations apply to RPS that involves or originates from major blood vessels [52]. Involvement of vascular structures is common. In one series, 24 of 141 patients with RPS had major blood vessel involvement at diagnosis and underwent aggressive surgery including resection and reconstruction of a major blood vessel [52]. The perioperative mortality rate was 4 percent, and the five-year overall survival rate was 66 percent. For leiomyosarcomas arising in the inferior vena cava, the need for postresection vascular reconstruction is debated [53-55]. Below the renal veins, the vena cava is often amenable to ligation, as many of these patients have developed extensive venous collaterals. (See "Surgical resection of retroperitoneal sarcoma", section on 'En bloc resection'.)
If the surgeon anticipates a margin-positive resection intraoperatively, IORT should be considered. (See 'Role of intraoperative radiation therapy' below.)
Alternatively, clips should be placed within high-risk areas for recurrence so that these sites can be considered for boosting to higher RT dose levels postoperatively. If small bowel can be displaced from the high-risk area with omentum or a prosthetic spacer, this may facilitate postoperative RT. (See 'Adjuvant radiation therapy' below.)
Role of debulking — Debulking surgery (ie, partial resection of a tumor leaving behind grossly positive margins) should be avoided for most patients, with the possible exception of large unresectable well-differentiated liposarcomas, for which palliative debulking surgery may improve symptoms and prolong survival. (See "Surgical resection of retroperitoneal sarcoma", section on 'Palliative resection'.)
There is no survival benefit for incomplete resection (a debulking procedure) in patients with unresectable RPS [18,44]. In a review of the largest retrospective series from four expert institutions, the median survival for patients undergoing incomplete resection was not significantly better than that for those with categorically unresectable disease [44].
However, in select patients with unresectable well-differentiated retroperitoneal liposarcoma, incomplete surgical resection can result in prolonged survival (median survival 26 versus 4 months in one retrospective series [56]) as well as palliation of symptoms when compared with patients who are not resected at all. The decision to operate must be individualized.
Role of intraoperative radiation therapy — If the surgeon anticipates a grossly incomplete resection, IORT for an additional dose of 10 to 15 Gy to areas of residual microscopic or gross disease is reasonable.
Nonrandomized studies of IORT in patients undergoing surgery for RPS suggest high rates of local control, recurrence-free survival (RFS), and overall survival [57-62]. A single prospective trial has been completed, in which patients with completely resected RPS were randomly assigned to postoperative high-dose RT (50 to 55 Gy) alone or IORT (20 Gy) in combination with postoperative low-dose external beam radiation therapy (EBRT; 35 to 40 Gy) [63].
Median survival time was similar in the two groups (45 versus 52 months). Use of IORT was associated with significantly fewer locoregional recurrences (6 of 15 [40 percent] versus 16 of 20 [80 percent]) and fewer complications of disabling enteritis (2 of 15 versus 10 of 20) but a higher frequency of radiation-related peripheral neuropathy (9 of 15 versus 1 of 20).
At our institution, we have used a treatment policy of preoperative RT, resection, and IORT boost for RPS since 1982. The latest update of our experience included 37 patients with primary or recurrent RPS, 20 who received preoperative RT (median dose 45 Gy) followed by resection and IORT, and a nonrandomized group of 17 patients who underwent preoperative RT followed by attempted resection without IORT [61]. Gross total resection was possible in 29 patients, 16 of whom also received IORT. The five-year actuarial overall survival, disease-free survival, and local control rates for all 37 patients were 50, 38, and 59 percent, respectively. When the 29 completely resected patients were considered, patients receiving IORT as a component of therapy had a significantly better overall survival (74 versus 30 percent) and local control rate (83 versus 61 percent) than did those not undergoing IORT. Complications, which were seen in 4 of 20 patients receiving IORT, included three patients with neuropathy and three ureteral injuries. These findings suggest that care must be taken if these structures are in the intraoperative radiation treatment volume.
Despite high rates of local control, systemic failure remains a significant problem despite preoperative RT and IORT, especially for higher grade tumors [64].
Preoperative therapy — Preoperative therapy is a reasonable approach for a patient whose initial imaging suggests that complete (ie, R0/R1) resection may not be feasible. In other cases of apparently resectable RPS, the role of preoperative RT is controversial, and clinical practice varies widely. At some institutions, preoperative RT is offered to all patients with larger well-differentiated liposarcomas (although there are no evidence-based guidelines on the appropriate tumor size to consider for initial RT, and decision making must be individualized), and for those with intermediate- and high-grade tumors with histologies that are not chemotherapy sensitive. On the other hand, at other institutions, preoperative RT is rarely utilized unless the preoperative staging evaluation suggests a high likelihood of a grossly positive margin. In our view, either approach is acceptable. For patients with intermediate- and high-grade, or locally advanced tumors with histologies that are chemotherapy sensitive (eg, synovial sarcoma, myxoid/round cell liposarcoma) and/or when the risk of distant metastatic disease is high (eg, leiomyosarcoma of the inferior vena cava or large dedifferentiated liposarcoma), we tend to favor neoadjuvant chemotherapy with preoperative RT. (See 'Neoadjuvant chemotherapy' below.)
Preoperative radiation therapy — The rationale for delivery of RT prior to surgery, with or without IORT at the time of resection, is that it may permit the safe delivery of higher RT doses than are possible in the postoperative setting, where RT doses to the tumor bed are often limited by the large field size and the proximity and tolerance of surrounding radiosensitive normal structures, such as the liver and bowel. In fact, many multidisciplinary sarcoma groups do not routinely offer postoperative RT to patients with resected RPS because of significant concerns about the narrow therapeutic ratio. (See 'Adjuvant radiation therapy' below.)
The theoretical advantages of preoperative, as compared with postoperative, RT for RPS include the following [65,66]:
●The main advantage of preoperative RT is that the gross tumor volume (GTV) can be precisely defined for radiation treatment planning, allowing accurate targeting of the radiation volume around the tumor.
●The tumor itself can act to displace small bowel from the high-dose radiation treatment volume, resulting in safer and less toxic treatment.
●Higher RT doses can be delivered to the actual tumor field since bowel adhesions to the tumor are less likely compared with the postoperative setting.
●The risk of intraperitoneal tumor dissemination at the time of the operation may be reduced by preoperative RT.
●Radiation is considered to be biologically more effective in the preoperative setting.
●It is possible that an initially unresectable tumor may be converted to one that is potentially resectable for cure [53,67].
These advantages may result in an improvement in the therapeutic ratio when RT is administered preoperatively. Preoperative RT has been found to be very well tolerated [68].
The benefits (similar tumor control and a lower incidence of late irreversible side effects) of preoperative, as compared with postoperative, RT in extremity STS have been demonstrated in a randomized trial. (See "Overview of multimodality treatment for primary soft tissue sarcoma of the extremities and superficial trunk", section on 'Choosing between preoperative and postoperative RT'.)
The benefit of preoperative RT for RPS has been addressed in retrospective single-institution series [53,57,61,64,69-73], a few prospective nonrandomized trials [74-76], a propensity score-matched analysis of data derived from the National Cancer Database [77], and a single randomized controlled trial [78]. Most data support improved local control with acceptable rates of acute and long-term toxicity, but there is no clear evidence that long-term outcomes (including survival) are improved by the addition of preoperative RT.
The following reflects the range of findings:
●Two prospective uncontrolled studies from Princess Margaret Hospital and the University of Texas MD Anderson Cancer Center are informative because the acute toxicities of preoperative RT were prospectively recorded separately from other toxicities [74-76].
•Investigators at Princess Margaret Hospital prospectively treated 41 patients with preoperative RT (median dose 45 Gy) combined with a postoperative brachytherapy boost (with afterloading catheters placed on the surgical resection bed after extirpation; median dose 25 Gy) [74]. Although the median radiation volume was relatively large, the EORTC/Radiation Therapy Oncology Group (RTOG) acute upper gastrointestinal and lower gastrointestinal/pelvis radiation toxicity [79] scores were ≤2 (grade 2 defined as anorexia with ≤15 percent weight loss from pretreatment baseline, nausea and/or vomiting requiring antiemetics, abdominal pain requiring analgesics/diarrhea requiring medication, mucous discharge not necessitating sanitary pads, or rectal or abdominal pain requiring analgesics) in all patients who underwent resection. Furthermore, no patient was hospitalized for acute toxicity, and there were no treatment interruptions or required cessations of treatment because of acute toxicity.
Postoperative brachytherapy to the upper abdomen was associated with substantial toxicity (eg, duodenitis and gastric outlet obstruction), prompting investigators to limit subsequent use of brachytherapy to the lower abdomen. The two-year overall and disease-free survival rates for patients with resected RPS were 88 and 80 percent, respectively.
•Long-term results from preoperative RT are available in a combined report of the experience from both Princess Margaret Hospital and the University of Texas MD Anderson Cancer Center [76]. For the 54 patients with primary RPS who underwent R0 or R1 resection after preoperative RT, the five-year local RFS, disease-free survival, and overall survival rates were 60, 46, and 61 percent, respectively.
●A survival benefit from preoperative RT relative to surgery alone was also suggested in a case-control propensity score-matched analysis of 9068 patients with RPS diagnosed between 2003 and 2011 and reported to the National Cancer Database [77]. In the matched dataset comparing outcomes in 563 patients receiving preoperative RT with those in 1126 patients treated with surgery alone, who were matched for demographic and histopathologic data, median overall survival was significantly longer with preoperative RT (median overall survival 110 versus 66 months, hazard ratio [HR] for death 0.70, 95% CI 0.59-0.82).
●On the other hand, long-term benefit for preoperative RT was called into question by the randomized phase III multicenter EORTC-62092 (STRASS I) trial [78].
The STRASS trial randomly assigned 266 patients with previously untreated primary RPS (histologically documented, localized, operable, and suitable for RT, with a World Health Organization [WHO] performance status score and American Society of Anesthesiologists score of ≤2) to surgery alone or preoperative RT followed by surgery [78]. The RT was administered as 50.4 Gy in 28 daily fractions using either three-dimensional conformal RT or intensity-modulated RT (IMRT). Histologic subtypes included leiomyosarcoma (n = 38), well-differentiated liposarcoma (n = 88), dedifferentiated liposarcoma (n = 105), other liposarcoma (n = 5), and other histologies (n = 29). The primary endpoint was investigator-assessed abdominal recurrence-free survival (aRFS). Abdominal recurrence was defined by one of the following events: local (abdominal) or distant progressive disease during preoperative RT, tumor or patient becoming inoperable, peritoneal metastasis found at surgery, macroscopic residual disease left at surgery, or local relapse after a macroscopically complete resection.
At a median follow-up of 43.1 months, the composite endpoint aRFS was not significantly improved by the addition of RT to surgery (median 4.5 versus 5.0 years, HR 1.01, 95% CI 0.71-1.44). Overall survival rates were also comparable at three years (84.6 percent [95% CI 76.5-90.1] with surgery versus 84 percent [95% CI 76.3-89.4] with preoperative RT) and five years (79.4 percent [95% CI 69.1-86.5] with surgery alone and 76.7 percent [95% CI 66.9-84] with preoperative RT). Median overall survival was not reached in either group, but was not significantly different (HR 1.16, 95% CI 0.65-2.05). However, local recurrence rates were almost twofold higher with surgery alone (47 of 127 [37 percent] versus 23 of 118 [19.5 percent]), and the difference was even more striking when the liposarcoma subgroup was assessed (30 versus 11 percent). Treatment-related serious adverse events were noted in 24 percent of the RT group (including one death from a gastropleural fistula) compared with 10 percent of those receiving surgery alone.
Of the 133 patients treated with preoperative RT, 19 progressed on RT, three of whom developed distant metastases and thus did not undergo what would have been noncurative surgery. Fifteen of those 19 patients had local progression but subsequently were able to undergo a macroscopically complete resection. These results led the Data Monitoring Committee to propose a sensitivity analysis whereby local progression while receiving RT was no longer regarded as an event for the patients who were subsequently able to achieve a complete resection (first analysis) and regardless of operability (second analysis). In both of these analyses, the addition of RT still did not improve RFS compared with surgery alone. Exploratory (not preplanned) subgroup analysis by sarcoma subtype and grade suggested that preoperative RT might improve outcomes in liposarcoma and in low-grade RPS, but not for leiomyosarcoma and higher grade RPS. Notably, there were only 31 high-grade sarcomas in the trial population.
The authors concluded that based on their results, preoperative RT should not be considered as standard of care for retroperitoneal sarcoma. However, some disagree with this conclusion, citing the following arguments:
•STRASS used a nonstandard definition of aRFS that included development of distant metastases and local "progression" during preoperative RT (which can in fact occur before RT is even started and probably has little clinical relevance if the opportunity for an R0/R1 resection is not lost, which seemed to be the case in STRASS based upon the sensitivity analysis).
•The median follow-up was relatively short (median 43 months), as local recurrences can develop after five years.
•Not enough emphasis was placed in the analysis on well-differentiated/dedifferentiated liposarcoma. This histology has a relatively greater chance of local-regional recurrence, and lower risk of metastatic disease compared with other subtypes of sarcomas found in the retroperitoneum, such as leiomyosarcoma of the vena cava or its branches. The well-differentiated/dedifferentiated liposarcoma patients, therefore, are most likely to benefit from preoperative RT, a contention that is supported by the subset analysis of the STRASS trial.
●On the other hand, others argue that despite the higher rate of local recurrence with surgery alone, the overall survival was similar in both groups.
At the very least, the issue of benefit from preoperative RT remains unsettled.
Intensity-modulated radiation therapy — IMRT has the potential to further improve the therapeutic index by permitting dose escalation to the area of the tumor while minimizing the dose to normal tissues at risk for radiation toxicity [80-82]. An innovative strategy combines IMRT with a novel target volume concept for preoperative treatment of RPS [80]. IMRT was used to deliver RT (50 Gy in 25 daily 2 Gy fractions) to a preoperative clinical target volume (CTV) that was limited to the contact area between the tumor mass and the posterior abdominal wall. All 18 patients completed the planned treatment with acceptable acute toxicity and underwent successful resection without major complications. With early follow-up (median 27 months), only two patients failed locally, and one developed distant metastases.
The authors concluded that this strategy was feasible, well tolerated, and associated with better radiation sparing of critical structures without compromising the rate of resectability. Longer follow-up is needed to assess the ultimate impact on local control and survival.
In one series, investigators were able to perform selective preoperative radiation dose escalation to the retroperitoneal margin of an RPS in 16 patients using IMRT to deliver 45 Gy in 25 fractions (1.8 Gy per fraction) to the entire tumor and 57.5 Gy in 25 fractions (2.3 Gy per fraction) to the boosted retroperitoneal margin [66]. Treatment morbidity was acceptable, and the two-year actuarial local control rate was 80 percent.
Radiation treatment planning — Due to the complexity of treatment, preoperative RT should only be carried out in a center with expertise in the treatment of sarcomas. An international expert panel has published treatment guidelines for preoperative RT for RPS, including suggested radiation target volumes [83]. In a subsequent project, 12 sarcoma radiation oncologists contoured preoperative target volumes (GTV, CTV, and high-risk CTV) for two RPS cases to assess levels of contouring agreement and to determine whether the guidelines were practical, feasible, and reproducible. GTV and CTV were contoured with a high level of agreement, but agreement for high-risk CTV was only moderate [84]. To facilitate consistent implementation of a "high-risk boost," further clarification of the "high-risk" volume is needed. Given that collaboration with surgical oncologists is critical in order to determine those regions of the tumor at high risk for positive margins, radiation oncologist and surgical oncologist teams were asked to contour high-risk GTVs together to serve as a basis for contour agreement analysis. The level of agreement of RPS high-risk GTV boost volumes between sarcoma radiation oncologist and surgical oncologist teams was substantial to moderate, with differences most striking in regions abutting visceral organs [85].
Concurrent chemoradiation — A role for concurrent administration of preoperative RT and chemotherapy (chemoradiation) in patients with RPS is not established, and for most patients, we suggest not pursuing this approach. In our view, given the lack of data from prospective trials that preoperative chemoradiation is more effective than preoperative RT and the potential for treatment-related toxicity, this approach should only be used for RPS in the context of a clinical trial by clinicians experienced with a chemoradiation approach that has been validated in a prior clinical trial.
Preoperative concurrent chemoradiation has been used in some patients with large high-grade or locally recurrent extremity STS, particularly if limb salvage is an issue. (See "Overview of multimodality treatment for primary soft tissue sarcoma of the extremities and superficial trunk", section on 'Is there a role for chemoradiation?' and "Treatment of locally recurrent and unresectable, locally advanced soft tissue sarcoma of the extremities", section on 'Neoadjuvant chemoradiation'.)
There has been interest in this approach for RPS because of the high risk of metastatic disease, particularly with leiomyosarcomas and undifferentiated pleomorphic sarcomas. Although randomized trials are not available, at least three small studies showed that doxorubicin-based or ifosfamide-based chemoradiation is safe in this setting but may not be feasible in all patients [75,86,87]:
●In one of these reports, 6 of 23 patients treated preoperatively with doxorubicin, ifosfamide, and cisplatin plus concurrent RT (28 Gy) had a complete pathologic response (ie, no residual tumor at exploration) [87].
●The second report was a phase I study of weekly doxorubicin for four to five weeks (4 mg/m2 as an initial bolus once weekly, followed by a four-day continuous infusion of 4 mg/m2 per day with concurrent EBRT in escalating doses) [75]. Four to eight weeks after chemoradiation, radiographic restaging was performed, and patients with localized disease underwent resection and IORT.
Chemoradiation was successfully completed by all 35 patients, with four requiring hospital admission during or in the immediate postchemoradiation period. At the highest RT doses (50.4 Gy), there were two patients with grade 3 or 4 nausea; otherwise, treatment was well tolerated. An R0 or R1 resection was possible in 26 of the 29 patients who had surgery; interval progression precluded surgery in six patients. Long-term outcomes were not reported.
●A more recent phase I/II study from Italy evaluated the feasibility, safety, and activity of a combination of high-dose long-infusion ifosfamide (HLI) and RT as preoperative treatment for resectable RPS [86]. Patients received three cycles of HLI (14 g/m2) and 50.4 Gy of RT prior to surgery. The primary endpoint was three-year RFS. Between 2003 and 2010, 83 patients were enrolled, and the main histologic subtypes were dedifferentiated liposarcoma (n = 26, 31 percent), leiomyosarcoma (n = 14, 17 percent), and well-differentiated liposarcoma (n = 13, 16 percent). All of the preoperative therapy was completed by 60 patients (the major cause of treatment interruption was hematologic toxicity), and 79 underwent surgery. At a median follow-up of 4.8 years, three-year RFS and overall survival rates were 56 and 74 percent, respectively. The authors concluded that this treatment approach was safe but not feasible in all patients.
Although these results seem promising, questions remain as to whether any combined chemoradiation approach is more effective than preoperative RT alone. The lack of data from prospective trials is a major hindrance to progress in this field. In 2003, the RTOG initiated a multicenter phase II trial of preoperative combined modality therapy (doxorubicin plus ifosfamide plus EBRT) followed by resection with an intraoperative or postoperative RT boost for intermediate- or high-grade primary or recurrent RPS. Unfortunately, the trial was closed because of poor patient accrual. (See "Treatment of locally recurrent and unresectable, locally advanced soft tissue sarcoma of the extremities", section on 'Neoadjuvant chemoradiation'.)
Neoadjuvant chemotherapy — Although data are limited, neoadjuvant (preoperative) chemotherapy appears to be safe and occasionally induces a radiographic response, which may impact surgical therapy in a few patients [88-91]. Although responders tend to do better than nonresponders, whether this reflects an impact of chemotherapy, disease biology, or patient selection remains uncertain [92].
If this approach is chosen, the optimal regimen has not been established, and the choice should be based on the small published studies of RPS [75,86,87]. Although widely used for advanced metastatic STS, at least one trial conducted in patients with high-risk STS of the extremities or trunk wall showed that histology-tailored chemotherapy did not provide benefit over standard chemotherapy (an anthracycline plus ifosfamide) when used in the neoadjuvant setting for localized disease [93]. (See "Adjuvant and neoadjuvant chemotherapy for soft tissue sarcoma of the extremities", section on 'Histotype-driven therapy'.)
With regional hyperthermia — Although not used in the United States, preoperative, regional, externally delivered deep wave hyperthermia combined with systemic chemotherapy is an option for high-risk primary tumors in parts of the world where it is available (mainly Germany). Whether this approach is superior to surgery plus preoperative RT is unknown and will require a randomized trial.
At least some data from Europe suggest that systemic chemotherapy combined with regional hyperthermia may represent an alternative treatment technique for patients with locally advanced, unresectable, high-grade STS of the extremities and trunk, as well as those with a locally recurrent tumor in an irradiated field [89,94]. Although different devices have been used to deliver regional hyperthermia, typically, this approach aims to raise tumor temperatures to 42°C for 60 minutes on days 1 and 4 of each chemotherapy cycle [95,96].
Benefit for this approach was suggested in a trial (EORTC study 62961) in which 341 patients with locally recurrent (n = 37), incompletely resected or resected with a surgical margin <1 cm (n = 142), or grade 2 or 3 primary STS ≥5 cm (n = 162) of the extremity (43 percent) or a non-extremity site (56 percent, with the majority in the pelvis or abdomen) were randomly assigned to four courses of systemic chemotherapy with or without regional hyperthermia, followed by aggressive local therapy (surgery and/or RT) and four additional courses of chemotherapy with or without regional hyperthermia [95]. Chemotherapy consisted of 21-day cycles of etoposide (125 mg/m2 on days 1 and 4), ifosfamide (1500 mg/m2 per day on days 1 through 4), and doxorubicin (50 mg/m2 on day 1 only), while regional hyperthermia was performed by exposing the affected body part to temperatures between 40 and 43 degrees for 60 minutes on days 1 and 4 of each chemotherapy course.
The objective response rate from preoperative chemotherapy was significantly higher in the hyperthermia group (29 versus 13 percent), although the number of patients who had definitive tumor resection as a component of local treatment was similar (two-thirds of both groups). At a median follow-up of 34 months, disease-free survival (median 32 versus 18 months) and two-year local progression-free survival (76 versus 61 percent) significantly favored the regional hyperthermia arm. In subset analysis, the higher rate of two-year local progression-free survival with hyperthermia was statistically significant in those with non-extremity sarcomas (64 versus 45 percent, p = 0.012) but not in those with extremity sarcoma (92 versus 80 percent).
Overall survival was not significantly better in the hyperthermia group when all patients were analyzed (at four years, 59 versus 57 percent), but in the 269 patients who completed the full treatment, those who received hyperthermia were 44 percent less likely to die during follow-up than those assigned to chemotherapy alone.
Adjuvant therapy
Adjuvant radiation therapy — Following resection of an RPS without neoadjuvant therapy, we do not suggest adjuvant RT for a low-grade, completely resected (margin-negative) tumor. Postoperative RT could be considered for patients with R0/R1 resected high- or intermediate-grade tumors that are at risk for local recurrence; however, in practice, most patients are just observed because it is rarely possible to deliver postoperative adjuvant RT with acceptable morbidity. (See 'Management of microscopically positive margins' below and 'Preoperative radiation therapy' above.)
Following an R2 resection, residual tumor frequently abuts structures or organs that cannot be easily or safely resected. For these patients, a small-field postoperative boost dose of RT is a reasonable approach if the surgeon can move the bowel, which is a dose-limiting normal tissue, away from the field of radiation. This might require placement of omentum or a tissue spacer at the time of surgery to displace bowel from the area of residual tumor. Another option, if the surgeon anticipates a grossly incomplete resection, is IORT for an additional dose of 10 to 15 Gy to areas of residual microscopic or gross disease. (See 'Role of intraoperative radiation therapy' above.)
Benefit of RT — In contrast to extremity STS, in which the most common site of first recurrence for intermediate- or high-grade tumors is a distant site, the primary pattern of treatment failure after resection of most RPS (specifically well-differentiated liposarcoma, the most common sarcoma in this anatomic location) is local-regional. Five-year local recurrence rates after complete resection of an RPS are between 41 and 50 percent, and local recurrence is the site of first failure in 90 percent of cases [31]. These high relapse rates have prompted investigation of combined modality treatment approaches. (See 'Outcomes and prognostic factors' below.)
While adjuvant RT can be considered, it most frequently cannot be administered following resection because normal tissue in the tumor bed precludes safe administration of postoperative RT. Hence, increasingly, preoperative RT is being chosen for large high-grade, intermediate-grade, or locally advanced RPS. (See 'Preoperative therapy' above.)
For patients who do not receive preoperative RT, there are no randomized trials of surgery with and without postoperative (adjuvant) EBRT. In retrospective uncontrolled series, the addition of postoperative RT reduces the risk of local recurrence and lengthens the recurrence-free interval; it has been more difficult to demonstrate a survival benefit [3,31,38,44,48,74,97-100]:
The benefits of adjuvant RT can be illustrated by the following reports:
●The largest uncontrolled single-institution experience included 145 French patients who presented with localized RPS [3]. The median tumor size was 15 cm (range 2 to 70 cm), 31 percent presented with neurovascular or bone involvement, 39 percent had grade 3 lesions, and 30 percent had a liposarcoma subtype. Complete (macroscopic) excision had been carried out in 94, and 60 of these patients received postoperative RT (median dose 50 Gy).
Among the patients who underwent complete excision, the five-year actuarial local RFS was significantly greater for the 60 irradiated patients compared with the 34 who did not receive it (55 versus 23 percent). In univariate analysis, five-year overall survival was modestly but significantly better in those patients receiving RT (89 of the 145 patients, 52 versus 44 percent). The toxicities of postoperative RT were not reported.
●A survival benefit from adjuvant RT was also suggested in a case-control propensity score-matched analysis of 9068 patients with RPS diagnosed between 2003 and 2011 and reported to the National Cancer Database [77]. In the matched dataset comparing 2196 patients who had received postoperative RT with 2196 patients who had undergone surgery alone (matched for demographic and histopathologic data as well as surgical margin status and extent of surgical resection), median follow-up time was 54 months for the postoperative RT group and 47 months for the surgery alone group. Median overall survival was significantly better in the RT group (89 [95% CI 79-100] versus 64 [95% CI 59-69] months).
Management of microscopically positive margins — Microscopically positive margins increase the risk for local recurrence [39,41,101], but whether they adversely influence survival is less clear. In some series [39,40,101], patients who have an R1 resection have higher rates of distant recurrence and inferior survival as compared with those undergoing R0 resections, but others have failed to show a relationship between microscopically positive margins and inferior survival [15,102]. As an example, the University of Texas MD Anderson Cancer Center reviewed its experience with 666 patients who had an R0 or R1 resection for STS, 24 originating in the retroperitoneum [102]. Within the entire cohort, those who underwent reresection followed by RT had significantly higher rates of local RFS (85 versus 78 percent), distant RFS (81 versus 69 percent), and disease-specific survival (83 versus 73 percent) than did those who underwent postoperative irradiation without reresection. When the analysis was limited to patients with RPS, the same local recurrence effect was seen, but there was no significant impact of the positive margins on either distant recurrence or survival.
Following R1 resection, reresection of the tumor bed is typically not practical or feasible without excessive morbidity. For these patients and for those who refuse further surgery, postoperative RT can improve local tumor control and the opportunity for long-term RFS. However, movement of viscera into the tumor bed after resection increases the risk of radiation-associated morbidity, and these patients are frequently just observed, reserving RT for use (generally in conjunction with salvage surgery with or without IORT) if the patient develops a local recurrence of tumor [1,97]. (See "Management of locally recurrent retroperitoneal sarcoma" and "Surgical resection of retroperitoneal sarcoma", section on 'Resection of recurrent disease'.)
It is often difficult to deliver postoperative RT because the bowel and other organs fall into the resection cavity. Although newer techniques such as IMRT and proton beam RT make it more feasible, the therapeutic ratio is still more favorable with preoperative RT, and this approach is generally preferred. Nevertheless, it is reasonable to consider use of postoperative RT if it can be delivered safely. (See 'Preoperative therapy' above.)
Adjuvant chemotherapy — Adjuvant chemotherapy cannot be considered a standard approach for STS at any site, including the retroperitoneum.
The benefit of adjuvant chemotherapy following surgical resection of an STS at any site is controversial. Multiple randomized trials have been undertaken, with disparate results. An updated meta-analysis suggested that use of an optimal adequately dosed anthracycline/ifosfamide-containing regimen significantly prolongs survival, but the analysis did not include the two largest trials, both conducted in Europe and both testing the value of an anthracycline- and ifosfamide-containing regimen [103]. A pooled analysis of both trials indicated no benefit from this approach [104]. Approximately 10 percent of the patients entered into these trials had central tumors.
Adjuvant chemotherapy in the management of STS is discussed in detail elsewhere. (See "Adjuvant and neoadjuvant chemotherapy for soft tissue sarcoma of the extremities".)
OUTCOMES AND PROGNOSTIC FACTORS — RPS have a substantially less satisfactory outcome than soft tissue sarcomas at other sites, such as the extremities or trunk [3-5,37,38,44,69,102,105]. Several factors contribute to the poor outcome and high rate of recurrence [105]:
●RPS are often large at diagnosis and anatomically situated such that wide resection is often not achievable.
●Even with complete resection, retroperitoneal liposarcomas tend to do worse than extremity liposarcomas, independent of tumor size, grade, or surgical margin.
●The surrounding normal tissues (liver, kidney, stomach, intestines, spinal cord) have relatively low tolerance for radiation therapy. As a result, radiation dose levels must be kept below those typically employed for extremity sarcomas.
In contrast to extremity sarcomas, the majority of first recurrences are local [31]. Eventually, distant metastases develop in 20 to 30 percent [21,44,106]. The main sites of distant metastases are the liver and lungs.
Local recurrence rates are higher with higher grade (poorly differentiated) histology (table 4), with liposarcoma histology, and in patients with positive resection margins. Delayed recurrences are more commonly seen in RPS than in extremity sarcomas [106,107]:
●In several series, local control rates range from 41 to 55 percent at five years but are only 18 to 40 percent at 10 years [21,31,38,108].
●In a report of 198 adult patients with RPS who were eligible for ≥5 years of follow-up, 40 percent of patients who were alive and disease free at five years recurred by 10 years [31].
Thus, long-term follow-up is mandatory. (See 'Posttreatment follow-up' below.)
Five-year disease-specific survival rates range from 20 to 69 percent [3-5,18,38,106,108-112]. The most important predictive factor for survival is tumor resectability, as illustrated by the following:
●In two surgical series in which the resectability rate was 50 percent or less, five-year survival rates were 20 and 36 percent, respectively [38,111]. In one of these series, only 14 percent of patients were still alive at 10 years, and only 9 percent remained free of locoregional recurrence [38].
●By contrast, those series reporting high rates of surgical resectability have better survival and fewer local recurrences [3,5,39,43,108,113-115]. As an example, in a series of 87 consecutive patients undergoing complete resection for RPS at a single institution, the five-year local recurrence rate was 25 percent [108]. Survival rates at 5 and 10 years were 66 and 57 percent, respectively.
However, long-term survival is still possible in patients who undergo a margin-positive resection of RPS. In a retrospective analysis of data on 12,028 patients with RPS diagnosed from 1998 to 2011 and reported to the National Cancer Database, 384 (3.2 percent) had a grossly incomplete (R2) resection [112]. Of the 272 patients with available long-term survival data, 64 (24 percent) survived five years or more.
After resectability, the next most important prognostic factor is histologic grade of differentiation [11,30,116]. In a single-center series of 183 patients with truncal soft tissue sarcoma or RPS, high/intermediate-grade histology was associated with a five- to sixfold increased risk of death compared with low-grade differentiation [30].
Others emphasize the importance of histologic subtype, with well-differentiated liposarcoma having the most favorable outcomes and leiomyosarcoma, pleomorphic sarcoma/malignant fibrous histiocytoma, malignant peripheral nerve sheath tumor, and dedifferentiated liposarcoma having the least favorable outcomes [10,11,15,21,106,116-119]. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Prognostic factors'.)
Nomograms have been developed and validated to more accurately predict postoperative survival based upon these and other features (figure 2) [116,118,120-125]. None of the site-specific nomograms is available online. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma", section on 'Prognostic tools'.)
RECOMMENDATIONS FROM EXPERT GROUPS — Consensus-based guidelines for initial treatment of RPS are available from several groups:
●Consensus-based guidelines for management of RPS from the National Comprehensive Cancer Network (NCCN) [34] include the following:
•For potentially resectable tumors, surgery, preoperative radiation therapy (RT), or preoperative chemotherapy are options. If an initial nonsurgical approach is chosen, pretreatment biopsy is suggested.
•Following complete surgical resection (ie, microscopically complete [R0] or grossly complete [R1]), the guidelines suggest that postoperative RT should not be administered routinely, with the exception of highly selected patients (not defined) and unless a local recurrence would cause undue morbidity. For patients undergoing an R1 resection, they suggest a postoperative boost of 10 to 16 Gy if preoperative RT was given. For grossly incomplete (R2) resection margins (ie, gross residual disease), reresection is recommended if technically feasible.
●On the other hand, consensus-based guidelines from the European Society for Medical Oncology (ESMO) consider that the value of preoperative treatments (RT, chemotherapy, regional hyperthermia, or combinations) in potentially resectable RPS is not established but may be relevant for a technically unresectable/borderline resectable RPS that could potentially be rendered resectable by downsizing [35]. For resected tumors, they state that postoperative RT may be an option in well-defined areas felt to be at high risk for local recurrence, but that the value of adjuvant chemotherapy is not established.
●The Trans-Atlantic Retroperitoneal Sarcoma Working Group (TARPSWG) was established in 2013 to evaluate the current evidence to define best practices in the evaluation and treatment of RPS, and they published a consensus document outlining their approach to primary RPS [51]. They state that neoadjuvant therapy (chemotherapy, chemotherapy combined with regional, externally delivered deep wave hyperthermia, external beam RT, or combined RT and chemotherapy) is safe for well-selected patients and may be considered after careful review by a multidisciplinary sarcoma tumor board. On the other hand, postoperative RT after complete resection was considered to be of no proven value.
●Guidelines for use of preoperative RT in RPS are also available from an international consensus group [83]; they state that the role of preoperative RT for RPS has not been proven.
POSTTREATMENT FOLLOW-UP — There are no randomized trials evaluating different surveillance strategies following completion of treatment for RPS. Surveillance recommendations for soft tissue sarcoma after surgical resection are available from several groups, two of which are specific for RPS (table 5) [126].
Our recommendations are in keeping with consensus-based guidelines from the National Comprehensive Cancer Network (NCCN) [34]:
●For completely resected RPS as well as those resected with positive margins, physical examination with abdominal/pelvic imaging every three to six months for two to three years, then every six months for the next two years, and then annually.
●A specific recommendation is not made by the NCCN for periodic chest imaging. However, given the higher rate of lung metastases with retroperitoneal and visceral leiomyosarcomas and pleomorphic undifferentiated sarcomas (as compared with other histologies), we obtain imaging of the chest, in addition to the abdomen and pelvis, on a regular schedule in these cases. Some clinicians routinely perform surveillance chest imaging for all patients with large high-grade tumors, regardless of histology.
Given that late recurrences are not uncommon with RPS, long-term follow-up to at least 10 years is mandatory. (See 'Outcomes and prognostic factors' above.)
SPECIAL CONSIDERATIONS DURING THE COVID-19 PANDEMIC — The COVID-19 pandemic has increased the complexity of cancer care. Important issues include balancing the risk from delaying cancer treatment versus harm from COVID-19, minimizing the number of clinic and hospital visits to reduce exposure whenever possible, mitigating the negative impacts of social distancing on delivery of care, and appropriately and fairly allocating limited health care resources. Additionally, immunocompromised patients are candidates for a modified vaccination schedule (figure 3) and the early initiation of COVID-directed therapy. Specific guidance for decision-making for cancer surgery on a disease-by-disease basis is available from the American College of Surgeons, from the Society for Surgical Oncology, and from others. These issues and other recommendations for cancer care during the COVID-19 pandemic are discussed separately. (See "COVID-19: Considerations in patients with cancer".)
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Soft tissue sarcoma".)
SUMMARY AND RECOMMENDATIONS
●Anatomy and histology – Retroperitoneal soft tissue sarcomas (RPS) constitute between 10 and 15 percent of all soft tissue sarcomas (STS). The most common histologic types are liposarcoma and leiomyosarcoma. (See 'Anatomy, histology, and differential diagnosis' above.)
●Clinical presentation – RPS typically produce few symptoms until they are large enough to compress or invade surrounding structures; most commonly, they are discovered in asymptomatic patients as an incidental abdominal mass. (See 'Clinical presentation' above.)
●Diagnostic evaluation
•The preferred diagnostic workup is a computed tomography (CT) scan of the abdomen and pelvis to evaluate the primary site and a chest CT to rule out metastatic disease to the lungs. Magnetic resonance imaging (MRI) is reserved for patients with an allergy to iodinated contrast agents, in cases where there is equivocal muscle, bone, or foraminal involvement on CT, for pelvic tumors, and if preoperative radiation therapy (RT) is being considered. (See 'Radiographic evaluation' above.)
•Clinicians should have a low threshold to proceed with a percutaneous core needle biopsy as the very small risk of complications and the unsubstantiated fear of tumor seeding of the biopsy tract are far outweighed by the information obtained from the biopsy. A preoperative biopsy is always indicated if the diagnosis is in doubt, if preoperative therapy is being considered as an alternative to upfront surgery, or if the lesion is suspected of having a high-grade component based upon the results of cross-sectional imaging. (See 'Need for biopsy' above.)
●Treatment options – There is no "one size fits all" approach to treatment of RPS. Surgical resection has traditionally been the only potentially curative approach. Because of the typically large size and anatomic complexity of these tumors at diagnosis, a resection with microscopically negative margins (R0 resection) is often not achieved, and locoregional recurrence is common, especially for low-grade tumors such as well-differentiated liposarcomas. Yet, the role of adjunctive therapy (RT, chemotherapy; either given preoperatively or postoperatively) continues to be debated, and there is no consensus as to the best approach for all patients. (See 'Overview of the approach to multidisciplinary treatment' above.)
Due to the rarity of these tumors and the complexity of treatment, evaluation and management should ideally be carried out in a center with multidisciplinary expertise in the treatment of sarcomas in the context of a multidisciplinary tumor board. However, even among centers of excellence, there is widespread disagreement as to the optimal way to approach RPS and a lack of high-level evidence to support any approach. This is reflected in the disparate recommendations from expert groups. (See 'Recommendations from expert groups' above.)
●Treatment approach – There are many acceptable treatment alternatives, and clinical practice is variable among different institutions. We tend to base our treatment approach on a multidisciplinary assessment of resectability, histology, and the likely pattern of recurrence:
•For all patients, we suggest participation in ongoing clinical trials, where available.
•If protocol therapy is not available or participation is not feasible, the following reflects our general treatment principles:
•Debulking surgery (ie, partial resection of a tumor leaving behind grossly positive margins [R2 resection]) should be avoided for most patients, with the possible exception of large unresectable well-differentiated liposarcomas, for which palliative debulking surgery may improve symptoms and prolong survival. (See 'Role of debulking' above and "Surgical resection of retroperitoneal sarcoma", section on 'Palliative resection'.)
•For most patients, if the preoperative staging evaluation suggests a high likelihood of grossly positive margins, preoperative therapy (RT, chemotherapy) is appropriate. (See 'Preoperative therapy' above.)
•In other cases of apparently resectable RPS, the role of preoperative RT is controversial, and clinical practice varies widely. At some institutions, including that of some of the authors and editors associated with this review, preoperative RT is offered to all patients with larger well-differentiated liposarcomas (although there are no evidence-based guidelines on the appropriate tumor size to consider for initial RT, and decision making must be individualized), and for those with intermediate- and high-grade tumors with histologies that are not chemotherapy sensitive. Those who favor preoperative RT argue that these patients tend to die from local recurrence and not distant disease, unless they dedifferentiate, and that combined therapy optimizes local control, which was shown in the randomized phase III STRASS trial.
On the other hand, at other institutions, including those of some of the authors and editors associated with this review, preoperative RT is rarely utilized unless the preoperative staging evaluation suggests a high likelihood of a grossly positive margin. Proponents of this approach argue that the risk for local recurrence is based more on biology than on specific treatments, and they point to the lack of benefit from preoperative RT versus surgery alone (using a composite endpoint of abdominal recurrence-free survival) in the randomized phase III STRASS trial.
In our view, either approach is acceptable. (See 'Preoperative radiation therapy' above.)
•A role for concurrent administration of preoperative RT and chemotherapy (chemoradiation) in patients with RPS is not established, and for most patients, we suggest not pursuing this approach outside of the context of a clinical trial (Grade 2C). (See 'Concurrent chemoradiation' above.)
•Although this is another controversial area, for patients with intermediate- and high-grade tumors, and locally advanced tumors with histologies that are chemotherapy sensitive (eg, synovial sarcoma, myxoid/round cell liposarcoma) and/or when the risk of distant metastatic disease is high (eg, leiomyosarcoma of the inferior vena cava or large dedifferentiated liposarcoma), we tend to favor neoadjuvant chemotherapy, with or without preoperative RT. The choice of chemotherapy regimen should be based upon the results of small series in which preoperative chemotherapy was used for patients with RPS. Benefit of histology-driven treatment has not been shown in any localized STS, including RPS. (See 'Neoadjuvant chemotherapy' above.)
•Although not used in the United States, preoperative, regional, externally delivered deep wave hyperthermia combined with systemic chemotherapy is an option for preoperative therapy for high-risk primary tumors in parts of the world where it is available (mainly Germany). Whether this approach is superior to surgery plus preoperative RT is unknown and will require a randomized trial. (See 'With regional hyperthermia' above.)
•The use of adjuvant therapy following resection of an RPS without neoadjuvant therapy largely depends on the tumor grade and completeness of resection:
-Following resection of an RPS without neoadjuvant therapy, we do not suggest adjuvant RT for a low-grade, completely resected (margin-negative) tumor (Grade 2C). Postoperative RT could be considered for patients following a complete surgical resection (R0/R1) for high- or intermediate-grade tumors at risk for local recurrence; however, in practice, most patients are just observed because it is rarely possible to deliver postoperative adjuvant RT with acceptable morbidity. (See 'Management of microscopically positive margins' above and 'Adjuvant radiation therapy' above.)
-If the surgeon anticipates a grossly incomplete resection (ie, R2 resection), a small-field postoperative boost dose of RT is a reasonable approach if the surgeon can move the bowel so that it can be adequately spared from the field of radiation. Another option, is intraoperative radiation therapy for an additional dose of 10 to 15 Gy to areas of residual microscopic or gross disease. (See 'Adjuvant radiation therapy' above and 'Role of intraoperative radiation therapy' above.)
-Adjuvant chemotherapy cannot be considered a standard approach for STS at any site, including the retroperitoneum, and we do not suggest its use outside of the context of a clinical trial (Grade 2B). (See 'Adjuvant chemotherapy' above.)
●Prognostic factors – The main prognostic factors for long-term outcome are resectability, histologic grade of differentiation, and histologic subtype. (See 'Outcomes and prognostic factors' above.)
●Posttreatment follow-up – Given that late recurrences are not uncommon with RPS, long-term follow-up to at least 10 years is mandatory. Our recommendations for posttreatment follow-up are consistent with consensus-based guidelines from the National Comprehensive Cancer Network (NCCN) [34] (see 'Posttreatment follow-up' above):
•Physical examination with abdominal/pelvic imaging every three to six months for two to three years, then every six months for the next two years, and then annually.
•For retroperitoneal and visceral leiomyosarcomas and pleomorphic undifferentiated sarcomas, we recommend imaging of the chest, in addition to the abdomen and pelvis, on a regular schedule.
ACKNOWLEDGEMENT — The UpToDate editorial staff acknowledges Thomas F DeLaney, MD, who contributed to earlier versions of this topic review.
