Radiotherapy and surgery—An indispensable duo in the treatment of retroperitoneal sarcoma



The authors undertook a systematic review to designate the role that radiotherapy (RT) might play in the treatment of retroperitoneal sarcomas. Correlating with recent literature, the objective of this review was to evaluate whether there was enough evidence for the authors to develop an institutional treatment protocol concerning the use of RT in the treatment of retroperitoneal sarcoma. Furthermore, this was a call for surgeons to talk to radiation oncologists before performing surgery. The 2 objectives of this review were: 1) to determine the benefit of RT in terms of local control and/or survival in the treatment of retroperitoneal sarcomas and 2) to discover the optimal timing of RT in the treatment sequence. A computerized literature search was performed in the PubMed database, the Cochrane Library database, and reference lists; and journals also were searched by hand to identify all retrospective and prospective reports published since 1998 relating to RT treatment of adult retroperitoneal sarcoma. Mainly, analyses were sought that were based on a 5-year local control rate (LCR), 5-year disease-free survival, and 5-year overall survival (OS). If only 2 years follow-up were available, then the authors also noted this outcome. Toxicity data were collected and analyzed separately. The synthesis of the literature was based on 9 prospectively nonrandomized studies and 10 retrospective studies that, together, reviewed a total of 1426 patients. The 5-year LCR varied from 27% to 62%, and the results from other reports fell in between those values. The 5-year OS rate ranged from 12% to 90%, and complete resection and tumor grade were the most important prognostic factors in most studies. This review resulted in 7 recommendations concerning the use of RT in the treatment of retroperitoneal sarcoma. The authors concluded that there is good evidence from multiple single-institutions studies that RT improves the LCR in patients with retroperitoneal sarcoma. Until now, there has not been a translation of this approach into survival benefit. The current results indicated that preoperative external-beam RT followed by radical surgery seems to be the preferred sequence, and adding intraoperative RT is a safe procedure for dose escalation in the upper abdomen. Cancer 2011;. © 2011 American Cancer Society.

Soft tissues sarcomas (STS) are rare malignant tumors that develop from mesenchymal tissue. They usually reside in muscle, fat, and connective tissues. STS can be located anywhere in the body, although the extremities (50%), trunk (25%), retroperitoneal space (15%), and head and neck region (10%) are the most frequent sites of occurrence.1, 2 Staging for STS is performed according to the American Joint Committee on Cancer (AJCC) staging system. In 2002, Kattan and colleagues published a nomogram that was able to predict the 12-year sarcoma-specific survival rate based on age, histology, grade, location, depth, and size of the tumor. This nomogram has a good predictive value and is widely used in the United States ( accessed March 4, 2011.3-5

Retroperitoneal STS

Within the retroperitoneal STS (RSTS) group, the 2 predominant histologic subtypes are liposarcoma and leiomyosarcoma.6, 7 Approximately 50% of RSTS are high-grade tumors. Liposarcomas, however, more often are low-grade to intermediate-grade tumors. From a clinical point of view, there is mostly a slow onset of vague symptoms during which tumors can become a bulky mass and often become fixed to vital structures (Fig. 1).5, 7, 8

Figure 1.

These T1-weighted magnetic resonance images of the abdomen from a 60-year-old man reveal a large, lobulated mass in the right abdomen. The images depict (Top, Middle) the tumor in the coronal plane and (Bottom) the tumor in the axial plane. The tumor is highly heterogeneous with irregular peripheral enhancement and central necrosis, and it causes displacement of and pressure on the right liver lobe, the inferior caval vein, and the right kidney.

Primary therapy for RSTS is surgical resection with removal of all gross disease while, if possible, sparing adjacent viscera not invaded by tumor. However, en bloc resection of contiguous organs, such as kidney, adrenal gland, pancreas, spleen, and colon, often is necessary, because there are few fascial planes in the retroperitoneal space. The completeness of surgical resection is a predictor of long-term survival.9 Unfortunately, obtaining a microscopically complete resection remains a heavy burden even in the hands of experienced surgeons.8, 10-13 After undergoing macroscopically complete surgery alone, local recurrences develop in 20% to 75% of patients, and these can occur even after 5 years of follow-up.1, 8, 10, 14-19 The estimated survival rate at 5 years varies between 12% and 75%,6, 13, 20, 21 and the cause of death for these patients usually is local recurrence, which is in sharp contrast to outcomes for patients with sarcoma of the extremities.8, 22 There is much controversy concerning the role of radiotherapy (RT) in the treatment of RSTS. Opponents argue that the lack of an overall survival (OS) benefit renders RT a treatment of little value.6, 9, 22, 23 Defenders argue that RT is associated with a significant improvement in local control and even an OS benefit compared with surgery alone.24-27 A major challenge for radiation oncologists is to combine the delivery of an effective radiation dose with an acceptable rate of radiation side effects.19 The objective of this review was to define the exact role of RT in the management of RSTS.


Literature Data

Data from the literature were obtained by entering the following search terms in Medline: retroperitoneal soft tissue sarcoma. The National Library of Medicine search engine was consulted—a service of the US National Institutes of Health—for the referral of articles in the PubMed database. The time span of the literature search covered from January 1998 to January 2011. Only peer-reviewed, A1 publications dealing with patient outcome were selected. RT planning studies; “short communications”; publications without details on RT timing, dose, or technique; publications on pediatric tumors; and publications concerning desmoid tumors, dermatofibrosarcoma protuberans, Kaposi sarcoma, round-cell sarcoma, and angiosarcoma were excluded, because these histopathology types also are excluded from the AJCC staging system. Rhabdomyosarcoma and primitive neuroectodermal tumor were not included because of their well known radiosensitivity, which was the same reason we excluded round cell sarcoma.28 Series that did not analyze RSTS separately from other intra-abdominal or extra-abdominal sarcomas were excluded. Publications dealing with recurrent RSTS were included. We divided the articles according to surgery and timing of RT. The relation between a set of variables and OS was studied. These variables included, among others, histologic type and grade, tumor extent, expertise of the surgical team, completeness of resection, distant metastases, and neurovascular or bone involvement. Toxicity from surgery, chemotherapy, external-beam RT (EBRT), intraoperative RT (IORT), or any combination often was pooled because of difficulty identifying precise contributing factors separately for each treatment.


Surgery Alone

Data from the literature on the use of surgery alone in RSTS are scarce. Patients with RSTS should receive treatment preferentially at high-volume centers29, 30 (see Table 1). In general, there are 2 “surgery schools”: School 1 defends aggressive surgery, including reoperation and debulking in unresectable RSTS2, 23, 34, 35; whereas School 2 does not take this aggressive approach.24, 36, 37 Subgroup analysis from the study by Gronchi et al demonstrated a benefit from aggressive surgery only for patients who received additional RT.24, 26 A median OS of 33 to 49 months was reported among patients who underwent surgery alone,31, 38 and the recurrence-free survival rate dropped to only 23%.32 Depending on resection status, the 5-year local recurrence-free survival (LRFS) rate varied between 55% and 62%. The postoperative mortality rate was 3%, and postoperative complications that required reintervention occurred in 24 patients (12%).33 Sampath et al observed that surgery alone was not sufficient as a treatment modality.26 Several other reports lacked critical details regarding the number of patients who underwent surgery as single treatment modality. In general, there is a clear trend toward using RT—preferably in the preoperative setting—to reduce the local recurrence rate23, 24, 26, 39-41 and, in some publications, even improvement of OS.24, 41

Table 1. Published Data on Patients Who Underwent Surgery as the Only Treatment Modality
AuthorNo. of PatientsFollow-Up, mo5-Year LRFS, %5-Year DFS, %5-Year OS, %
  • LRFS indicates local recurrence free survival; DFS, disease-free survival; OS, overall survival; NR, not reported.

  • a


  • b

    Stoeckle et al analyzed data from 165 patients, including 150 who underwent surgical resection. A complete excision was achieved by 94 of 145 patients with nonmetastatic disease; of those 94 patients, only 34 did not receive radiotherapy as adjuvant treatment.

Neuhaus 2005315826NRNR45
Nishimura 201032822435.4/10aNR62
Strauss 2010332002954.668.6NR
Stoeckle 20011834b4723NR44

Postoperative RT

The advantage of postoperative RT is that surgery can be performed without delay, and histology and tumor grade can be assessed in detail.42 Postoperative RT usually starts 4 to 6 weeks after surgery, allowing enough time for appropriate wound healing. However, it has been suggested that the delay between surgery and RT could result in cellular repopulation of neoplastic clones. In most institutions, postoperative RT is mainly considered for patients who have high-grade disease, microscopic positive margins (R1 resection), or very large tumors and close surgical margins independent of tumor grade.

The French Cancer Center Federation Sarcoma Group performed a multivariate analysis of 145 patients with primary RSTS. In total, 123 patients underwent surgery, and a complete excision was achieved in 65%. Postoperative RT was received by 89 patients at a mean dose of 48 grays (Gy). Fifty-one patients in that study received chemotherapy.18 Lewis et al prospectively followed 500 patients with RSTS at the Memorial Sloan-Kettering Cancer Center (MSKCC), including 321 patients with resectable disease (primary setting, n = 231; recurrent setting, n = 90). In total, 173 of those patients (62%) received “some form of RT.” Local control was better with a higher dose, and a threshold of 55 Gy was considered the minimal dose.16 The delivery of 55 Gy, however, was hampered by the presence of small bowel loops, which tended to “fall” into the resection bed and become fixed by postoperative adhesions.7, 16 Gilbeau et al treated 45 consecutive patients who had primary RSTS with curative intent in a multimodality approach. In that study, 28 patients received EBRT, 14 patients received both intraoperative electron RT (IOERT) and EBRT, and 3 patients received IOERT alone. IOERT consisted of a mean dose of 15 Gy as a boost to the tumor bed or to sites at which surgery seemed to be microscopically incomplete. The median follow-up at the time of our analysis was 53 months.43 Zlotecki et al reported outcomes and associated complications for 40 patients with RSTS who underwent surgery and received either postoperative RT (n = 25) or preoperative RT (n = 15). The maximum dose of postoperative RT in that study was 50 Gy at 1.8 Gy per fraction.13


Details on the local control rate (LCR), disease-free survival (DFS), and OS, along with patient and treatment characteristics, are provided in Table 2. The 5-year OS rate in the trial by Stoeckle et al was 49% (median follow-up, 58 months). There was a significant 3.4-fold reduction in the 5-year local recurrence rate for patients who received postoperative RT (P = .002). Omission of RT was the most important prognostic factor for local recurrence in multivariate analysis.18 Lewis et al reported that patients who were treated for primary disease had a median survival of 72 months versus 28 months for those who were treated for a local recurrence. The LRFS rate at 2 years and 5 years was 81% and 59%, respectively. The analysis of DFS, metastasis-free survival, and disease-specific survival was confined to 231 patients who had primary disease. Clear conclusions concerning the therapeutic effect of RT in the preoperative, intraoperative, or postoperative setting were not possible, because RT as a variable was not prospectively randomized.16 In the study by Gilbeau et al, postoperative RT to 49 Gy resulted in a 5-year locoregional control rate of 40%. Recurrences were mostly “in-field.” The authors suggested that 49 Gy may have been too low to obtain sufficient disease control.43

Table 2. Published Data on Patients Who Received Postoperative Radiotherapy as the Only Radiation Modality or in Combination With Other Radiation Modalities or Chemotherapy
 Postoperative EBRT 
StudyAloneWith Preoperative EBRTWith IORTWith CTResults
AuthorNo. of PatientsNo.Dose, GyNo.Dose, GyNo.Dose, GyNo.Follow-Up, mo5-Year LC, %5-Year DFS, %5-Year OS, %
  • EBRT indicates external-beam radiotherapy; IORT, intraoperative radiotherapy; CT, chemotherapy; Gy, grays; LC, local control, DFS, disease-free survival; OS, overall survival; ND, not determined; NR, not reported.

  • a

    Chemotherapy was received by 51 patients as neoadjuvant treatment.

  • b

    In this series, another 3 patients received IORT alone.

  • c

    After preoperative EBRT.

  • d

    Adjuvant postoperative chemotherapy was receivd by 7 patients.

  • e

    Complete resection (R0) vs R1 resection, respectively.

Stoeckle 2001181458950NoneNoneNDa47522949
Lewis 19981623166NDNDND1722859NR54
Gilbeau 200243452849None14b15115340NR60
Zlotecki 200513402550NonecNoneNDd3465NR69 vs 12e


Lewis et al observed a 30-day postoperative mortality rate of 4%. Causes of death in that study were bleeding, sepsis, myocardial infarction, and multisystem organ failure.16 Overall, acute postoperative RT-induced enteritis is present in 30% to 80% of patients, and grade 3 and 4 enteritis is present in approximately 6% to 80%.13, 43 Zlotecki et al reported more side effects in the postoperative RT group with more acute enteritis compared with the preoperative group (80% vs 36%; P = .098). Postoperative, pulsed-dose-rate brachytherapy using iridium-192 caused severe morbidity in >50% of patients and, in particular, caused high rates of duodenitis and gastric outlet obstruction, eventually leading to toxic death in 2 patients.1, 44 Therefore, postoperative brachytherapy should be reserved exclusively for pelvic treatments.

IORT Alone or Combined With Other Forms of RT

The rationale for using IORT is the accurate and safe escalation of dose to a better visualized tumor bed. In addition, normal structures can be displaced or shielded.7, 8 There are 2 methods for delivering IORT: with electrons (IOERT) or with the use of brachytherapy catheters (high-dose rate IORT [HDR-IORT]). Table 3 summarizes the results from these treatment policies.

Table 3. Published Data on Patients Who Received Intraoperative Radiotherapy as the Only Radiation Modality or in Combination With Other Radiation Modalities or Chemotherapy
StudyAloneWith EBRTWith Postoperative BTWith CTResults
AuthorNo. of PatientsNo.Dose, GyNo.EBRT/IORT Dose, GyNo.Dose, GyNo.Follow- Up, mo5-Year LC, %5-Year DFS, %5-Year OS, %
  • IORT indicates intraoperative radiotherapy; EBRT, external-beam radiotherapy; BT, brachytherapy; CT, chemotherapy; Gy, grays; LC, locoregional control; DFS, disease-free survival; OS, overall survival; ND, not defined; NR, not reported.

  • a

    The dose ranged from 8.75 to 30 Gy.

  • b

    EBRT was received preoperatively by 53 patients, postoperatively by 12 patients, and both preoperatively and postoperatively by 12 patients.

  • c

    EBRT was received preoperatively by 7 patients and postoperatively by 15 patients.

  • d

    EBRT was received preoperatively by 50 patients and postoperatively by 32 patients.

  • e

    Thirty-nine patients received adjuvant chemotherapy, and 17 received concurrent chemotherapy.

  • f

    Sixty-two of 103 patients underwent complete tumor resection and received radiotherapy regimens, including 14 who received IORT plus EBRT, 27 who received EBRT, and 21 who did not receive any form of RT.

  • g

    Twenty-four patients received chemotherapy when distant disease or positive lymph nodes were detected.

Petersen 200217871015a77b47.6/15None1042592948
Gieschen 20014537None2045/10-20NoneNone38593850
Alektiar 20001032712-152550.4/12-152140-160433625545
Bobin 20034624None22c45-50/15None553NR2856
Ballo 20074783None18d50-55/15NoneNDe4740 (at 10 y)39 (at 10 y)NR
Krempien 2006486722154545/15NoneNone30402864
Pierie 200649103None14f10-20NoneNDg27NRNR48
Dziewirski 201050, 515722203450NoneNone4065NR50

Gieschen et al published their experience in 37 patients (29 patients with primary disease and 8 patients with recurrent disease) at the Massachusetts General Hospital with primary or recurrent RSTS. The results from 20 patients who received preoperative RT (median dose, 45 Gy/1.8 Gy per fraction), underwent resection, and received IOERT were compared retrospectively with the results from 17 patients who received preoperative RT, underwent resection, and received no IOERT.45 Petersen and colleagues from the Mayo Clinic reported the results from 87 patients (43 patients with primary disease and 44 patients with recurrent disease) who received RT at a median dose of 15 Gy IOERT, including 77 patients who also received EBRT (median dose, 48.6 Gy). RT was delivered preoperatively in 53 patients, postoperatively in 12 patients, and preoperatively and postoperatively in 12 patients. Ten patients received chemotherapy.17 An update of the work by Petersen et al was presented in 2008 at a conference of the International Society of Intraoperative Radiation Therapy. ( meeting/resoconti/ISORT2008%20Auditoriumpercent20 Presentations/presentat.htm accessed March 4, 2011).

Ballo et al reported on 83 patients (60 patients with primary disease and 23 patients with recurrent disease) who received EBRT (n = 63), EBRT with IOERT (n = 18), EBRT with a perioperative brachytherapy boost (n = 1), and brachytherapy alone (n = 1) after R0/R1 resection. The median IORT dose was 15 Gy. Thirty-nine of those patients received adjuvant chemotherapy, and 17 patients received concomitant chemotherapy.47 Pierie et al compared EBRT (40-50 Gy in the preoperative setting if possible) with or without IOERT in patients with high-grade tumors and/or microscopically positive margins after macroscopically complete (R0) resection.49


Details on LCR, DFS, and OS as well as patient and treatment characteristics are listed in Table 3.


One randomized, prospective study that was published in 1993 compared the combination of IOERT (20 Gy) and postoperative EBRT (35-40 Gy) with postoperative EBRT alone (50-55 Gy) in patients with completely resected RSTS. The LCR at 5 years was better in the combination group (60% vs 20%).52 Gieschen et al reported a 83% LCR and an OS rate of 74% in patients who received preoperative EBRT followed by gross tumor resection and IOERT at a dose of 10 to 20 Gy, depending on their resection status. Those authors retrospectively compared the results with those from 16 patients who received preoperative RT followed by resection without IOERT. In patients who underwent complete resection, the receipt of IOERT resulted in a significantly better OS rate (74% vs 30%; P = .04) without affecting local control (83% vs 60%; P = nonsignificant).45

Petersen et al reported 2-year and 5-year rates OS of 83% and 48%, respectively, in 87 patients who underwent maximal surgical resection and received IOERT, which was combined with EBRT in 77 patients. The median follow-up in that trial was of 3.5 years, and the 2-year and 5-year estimated LCR was 84% and 59%, respectively. OS was associated inversely with tumor size (>10 cm). An update of the Mayo Clinic data included 226 patients and confirmed earlier conclusions. It is noteworthy that the reported 5-year LCR was 69% compared with 59% in a previous publication from those authors.17 For the 26 patients who completed IOERT (15 Gy) and EBRT (45 Gy) after R0 resection at Heidelberg, the 5-year and 10-year OS rates were both 80%, and the 5-year and 10-year LCR was as high as 100%. The only factor that had a significant impact on OS was resection status, with an almost 40% advantage observed among patients who underwent R0 resection (5-year OS: 87% [R0] vs 50% [R1/R2]).48 In a multivariate analysis of 103 patients reported by Pierie et al, IOERT (10-20 Gy) plus EBRT in high-risk patients was correlated significantly with both OS (P = .027) and recurrence-free survival (P = .048).49


At the MSKCC, a phase 1/2 study was initiated to treat patients with maximal tumor resection, HDR-IORT (12-15 Gy), and postoperative EBRT (45-50.4 Gy at 1.8 Gy per fraction). Thirty-two patients were treated, and the median follow-up was 33 months. Four patients in that study received preoperative chemotherapy for reasons that were not mentioned. In 7 patients, tissue expanders were placed during surgery to increase the safety of postoperative EBRT. In the remaining 25 patients, either a history of RT or postoperative complications made the use of tissue expanders impossible. This multimodality approach resulted in 5-year LCR and OS rates of 66% and 45%, respectively. Univariate analysis revealed that RT was the only factor that significantly decreased the risk of local recurrence (P < .02). Alektiar et al reported a 5-year DFS rate of 55% that did not change with presentation (primary vs recurrent) or grade (low vs high).10, 12 Ballo et al observed no difference in the LCR depending on the receipt of IORT (5-year LCR: 46% without IORT vs 51% with IORT). Concerning the whole group, the 10-year disease-specific survival, DFS, and distant metastasis-free survival rates were 44%, 39%, and 67%, respectively. Further analysis of RT-related treatment characteristics failed to reveal any improvement in local control with higher dose of RT (≥50 Gy), the use of concurrent chemotherapy, or specific timing of RT treatment (preoperative vs postoperative).10, 47 Dziewirski et al analyzed the results from 57 patients who underwent surgery and received HDR-IORT at a dose of 20 Gy in a single fraction. Sixty percent of those patients also received additional adjuvant EBRT. More than 75% of patients (n = 65) presented with recurrent disease. At a median follow-up of 40 months, the 5-year OS and local recurrence-free survival rates were 50% and 65%, respectively. Patients who received the combination of HDR-IORT and EBRT had a significant improvement in local recurrence-free survival compared with patients who received HDR-IORT alone.50, 51



IOERT induces severe complications in up to 37% of patients; and neuropathy, hydronephrosis, gastrointestinal fistula, small bowel obstruction, and vaginal and ureteroarterial fistula are the most frequent.17, 45, 49 To limit the risk of grade 3 neuropathy, the RT dose should kept below 15 Gy.46


Late severe complications have been observed in 10% to 18% of patients; and the most frequently reported is gastrointestinal obstruction, followed by fistula formation, peripheral neuropathy, hydronephrosis, wound complications, abscess, bleeding, and hydroperitoneum.10, 47, 50, 51 Ballo et al did not use Radiation Therapy Oncology Group criteria for toxicity, but all 5 of their patients (10% at 5 years) who developed RT-related complications had received postoperative EBRT.47

Preoperative RT Alone or in Combination With Other Forms of RT

There are several theoretical advantages for preoperative RT compared with postoperative RT. First and foremost, the tumor often displaces the bowel out of the field of radiation and acts as a natural tissue expander, allowing treatment with a lower dose to the adjacent normal tissue.19, 53 Second, the often-present pseudocapsule of the tumor may thicken and become acellular, which eases resection and decreases the risk of local recurrence. Third, and probably most relevant, preoperative RT may sterilize the operative field against seeding of microscopic tumor emboli during surgical manipulation.54 Finally, the improved oxygenation compared with the postoperative setting theoretically increases radiosensitivity and lowers the biologically effective radiation dose required to kill the tumor cells.1, 7 Preoperative RT, however, may be a double-edged sword. Patients who initially are diagnosed with probable resectable tumors may be better off when surgery is not delayed.

Unfortunately, no randomized controlled trial has compared preoperative RT with postoperative RT or with “no RT.” Our search resulted in 5 nonrandomized studies and 1 trial that was stopped early. The details are presented in Table 4.28 Pawlik et al combined data from 2 prospective trials, resulting in data on 72 patients.1 In their analysis, patients with low-grade RSTS were excluded because of the assumed difference in natural history. The 2 participating centers were the University of Toronto (Princess Margaret Hospital [PMH] and Mount Sinai Hospital) and The University of Texas MD Anderson Cancer Center (MDACC). At PMH, 41 patients received preoperative EBRT with postoperative, pulsed, high-dose-rate brachytherapy (median dose, 25 Gy) received by 21%. No single patient received chemotherapy, and the majority of patients (>70%) received a median preoperative RT dose of 45 Gy.44, 53 Thirty-five patients who were treated at MDACC also received a fixed-dose continuous-infusion of doxorubicin as radiosensitizer concurrent with an escalated dose of EBRT (50.4 Gy if possible).1, 53 IOERT to 15 Gy was administered to the tumor bed in approximately 30% of patients. This decision was made at the discretion of the treating surgeon.12 At the University of Alabama, intensity-modulated RT (IMRT) delivered a dose of 45 Gy in 25 fractions. At the same time, a simultaneous integrated boost of 12.5 Gy was delivered to the “margins at risk” with the use of IMRT.19 In an effort to reduce the bowel-related toxicity of RT, White et al used an abdominal tissue expander in association with preoperative EBRT. Two weeks after tissue expander insertion, EBRT at a dose of 45 to 50 Gy was commenced in 36 patients. In 11 patients for whom surgery was not possible, palliative RT was administered.55 The American College of Surgeons Oncology Group (ACOSOG) started a phase 3 randomized trial (ACOSOG Z9031) in 2004 to compare preoperative RT and surgery with surgery alone. The primary endpoints of that study were OS and progression-free survival. The trial closed prematurely because of poor accrual.

Table 4. Published Data on Patients Who Received Preoperative Radiotherapy as the Only Radiation Modality or in Combination With Other Radiation Modalities or Chemotherapy
StudyPreop EBRT AlonePreop EBRTand Postop BTPreop EBRT and IORTPostop EBRT and CTResults
AuthorNo. of PatientsNoDose, GyNo.Dose, GyNo.Dose, GyNo.Follow- Up, mo5-Year LC, %5-Year DFS, %5-Year OS, %
  • Preop indicates preoperative; EBRT, external-beam radiotherapy; Postop, postoperative; BT, brachytherapy; IORT, intraoperative radiotherapy; CT, chemotherapy; Gy, grays; LC, locoregional control; DFS, disease-free survival, OS, overall survival; NR, not reported.

  • a

    Nineteen patients received preoperative EBRT plus BT, 2 patients received BT alone, and 2 patients received BT and postoperative EBRT.

  • b

    Eleven patients received palliative RT versus 25 patients who received EBRT and underwent surgery.

Pisters 200712351350.40221535NRNRNRNR
Jones 20024446214519a250019NR80 (at 2 years)NR
Tzeng 200619161657.50002880 (at 2 years)NRNR
White 2007553825b46.500057NR80 (primary)74 (90% primary)
Gieschen 20014537174502010-20None38593850


Details on LCR, DFS and OS as well as patient and treatment characteristics can be found in Table 4. Pawlik et al pooled the data from Jones et al and Pisters et al and reported that (macroscopically) complete resection was achieved after preoperative RT by 75% of patients. At a median follow-up of 40 months, the 2-year and 5-year local recurrence-free survival rates were 79% and 60%, respectively. The 5-year OS rate was 61%, the median OS had not been reached and exceeded 60 months at the time of this publication, and the median DFS was 33 months. Others have confirmed these results.19, 55 These data compared favorably with historic controls and with series on “surgery alone,” especially considering the 25% recurrence rate and the high-grade histology in all tumors.1 Jones et al published their patient series separately and reported 2-year OS and DFS rates of 88% and 80%, respectively. Patients who had low-grade tumors fared significantly better compared with patients who had high-grade tumors, but only in the absence of R0 resection. Both an RT dose ≥45 Gy and primary disease were associated with a trend toward longer DFS and longer OS.44

Zlotecki et al (Table 2) compared preoperative and postoperative EBRT. Preoperative EBRT resulted in a nonsignificant increase in local control and OS. However, there was an important selection bias in their study. The patients who were scheduled for preoperative EBRT had significantly smaller tumors, and these tumors are predictors of achieving better local control. In addition, the follow-up in the preoperative group was significantly shorter than that in the postoperative group (17 months vs 42 months). The increase in the LCR with preoperative RT disappeared at 5 years (66% vs 65%, respectively).13


The PMH and MDACC trials recorded acute toxicity profiles prospectively and separately from other toxicities.1 Only 4 of 35 patients (11%) reportedly required hospitalization during or soon after combined treatment with RT and doxorubicin. The low toxicity rate from this regimen was attributed to tumor-associated bowel displacement away from the RT target volume.12 In series of patients who did not receive chemotherapy, the toxicity profile was very mild, and hospital admission reportedly was rare.19, 44, 55


The treatment of RSTS remains a challenge because of their location, large size, and often aggressive clinical behavior. Surgery alone is curative in a relatively small number of patients. Local recurrence is the major cause of mortality and continues to occur after 5 years of follow-up, as stated previously by the National Cancer Institute Sarcoma Progress Review Group.20-22, 24-28, 44, 45, 47, 48, 52-55 This is in contrast to the cause of death from extremity sarcomas, which almost always is a consequence of metastases. This difference may be explained by differences in anatomic location and tumor biology.28 The 3 recurrent RSTS prognostic factors for OS in all previously described studies were margin status after surgery, histologic type (liposarcoma vs others), and high-grade tumor.

Any adjuvant therapy has to be both tolerable and effective in improving local control. For example, the single-institution experience reported by Stoeckle et al indicated a significant doubling in the LCR when patients received postoperative RT.18 Population-based studies, however, revealed that only 25% of patients actually receive a form of adjuvant RT. This percentage has not changed since 1973.28, 56 Overall, the use of RT is low, and RT often is omitted from treatment schedules because of a lack of level I evidence; concerns about RT-induced toxicity; and complexity concerning timing, dose and treatment planning, and delivery.56 The lack of randomized trials makes evidence-based decision making even more difficult and feeds the opponents of RT. Until now, we could not identify a single randomized controlled trial that compared surgery alone versus surgery combined with RT in whatsoever form.

Most reports deal with only a few patients who are treated nonuniformly over a time span of several decades. The use of combinations of surgery and irradiation with different schedules makes the effect of RT more difficult to determine. Nevertheless, there are some practical tips and tricks that can be derived from published data. Side effects appear to be less pronounced with preoperative RT.44, 47 The addition of IORT has been evaluated in multiple series and sometimes produces promising disease control rates, although high toxicity rates have been reported, with peripheral neuropathy the most frequent and severe.1, 11, 44, 45 Concerning postoperative RT, there may be a dose response and better local control rates with doses >55 Gy.14 Modern technologies, such as IMRT, tomotherapy, and intensity-modulated arc therapy, should be the technologies of choice, because they combine dose escalation with sparing of the organs at risk and, consequently, are less toxic.25, 57, 58 Even more promising are the results obtained with carbon ions, which have produced OS and local control rates at 5 years of 50% and 69%, respectively, with no severe intestinal toxicity.59

Author's Recommendation

We hope that this literature overview can serve as a basis for making recommendations with regard to the place of RT in the treatment of RSTS. We propose the following conclusions:

  • 1Adding external RT with or without IORT to surgery improves local control and may be associated with improved OS.10, 18, 24, 26
  • 2The administration of EBRT in the preoperative setting may be the preferred sequence, because it reduces the amount of normal tissue irradiated and, consequently, diminishes gastrointestinal toxicity.44, 47
  • 3Adding intraoperative RT improves local control even for minimal microscopic residual disease. The combination of preoperative therapy with an intraoperative boost may favorably effect local recurrence and 5-year OS.45, 49-51
  • 4The use of postoperative RT is bounded by bowel toxicity because of small bowel migration into the tumor bed. Consequently, there are more complications in postoperative treatment.13 If postoperative RT is applied, then modern technology like IMRT and tomotherapy are used preferentially to safely achieve the threshold dose of 55 Gy.7 However, the implementation of this strategy needs further investigation.59, 60
  • 5The use of postoperative brachytherapy should be restricted to the lower abdomen.44
  • 6Although surgeons still are the gatekeepers of treatment, RSTS should be discussed and treated in a multidisciplinary setting. The presence of a radiation oncologist is vital.
  • 7Multicenter, randomized, controlled trials are urgently needed.


No specific funding was disclosed.


The authors made no disclosures.