Retroperitoneal Liposarcoma: Treatment Outcome in Primary and Secondary Surgical Intervention
Approximately 30% to 40% of all retroperitoneal soft tissue tumors are sarcomas, with liposarcoma prevailing in approximately 50% of these cases. Retroperitoneal liposarcomas typically show a high rate of local recurrence and late distant metastases. The aim of our retrospective analysis was to investigate the efficacy of treatment in our patients with liposarcoma. Thirty-four consecutive patients underwent surgery in our clinic between October 2004 and November 2017. Liposarcomas arising from the mesenteric or abdominal adipose tissue or the pelvis were excluded. Of 34 patients, 23 (67.6%) presented with primary and 11 (32.4%) with recurrent disease. In 7 of the 34 patients (20.6%), a radical resection (R0) could be achieved, and in 27 patients (76.5%) resection was marginal (R1). Time to recurrence was not affected by neoadjuvant radiotherapy. Patients who underwent systemic chemotherapy followed by radiotherapy developed earlier recurrences (P = 0.003) than patients with neoadjuvant radiotherapy only. Overall survival was significantly better in the neoadjuvant group (P 0.045) Combining surgical resection with neoadjuvant radiation treatment showed survival benefits in primary but not recurrent disease. Repeated surgery remains a valid approach in carefully selected patients, but all patients should be referred to a center of expertise in multimodal treatment approaches for retroperitoneal liposarcomas.Background
Methods
Results
Conclusions
The retroperitoneum is the anatomic space between the dorsal skeletal muscles and the ventral visceral organs. It reaches from the diaphragm cranially to the pelvis. The organs are localized inside this space, embedded in plenty of adipose tissue, which sustains and protects the organs. The inferior vena cava divides the retroperitoneum into a right and a left compartment and is an important structure in defining resectability and radical resection of tumors.
Benign lesions, at less than 5%, are rare in the retroperitoneum, whereas approximately 30% to 40 % of all diagnosed retroperitoneal soft tissue tumors are sarcomas. Sarcomas in the retroperitoneum are reported to have an annual incidence of 4 to 5 cases per 100,000 people in Europe.1 Because of the local growth pattern, diagnosis of retroperitoneal sarcoma is often established late, frequently by chance, and often because of late symptoms caused by compression of adjacent structures or because of significant increase in abdominal girth, which often makes radical resection difficult. Nevertheless, curative surgery for these tumors is the preferred treatment option.
In order to initiate the correct treatment, histology should be obtained, even if images appear to be diagnostic. Retroperitoneal liposarcoma (RPLS) is the most common form, diagnosed in about 50% of all sarcomas, followed by leiomyosarcoma (21%) and other rare histologic subtypes.2 These 2 most common histopathologic types have a different biologic behavior regarding local recurrence and the potential for distant metastases.
Whereas leiomyosarcomas are highly aggressive, having a 5-year survival rate between 0 and 20%2 and carrying the risk of distant metastases developing in more than 50% of cases, liposarcoma are much more prone to local recurrence—rates approach nearly 100 %—and only rarely present with distant metastases, most commonly to the lungs. Therefore, the most important tumor burden in liposarcoma is locoregional, and surgery with curative intent is the mainstay of locoregional control.
Liposarcomas can be subdivided into 4 types: well-differentiated, dedifferentiated, myxoid, and pleomorphic, with most being well-differentiated and dedifferentiated. The last 2 types should probably be regarded as related entities because the latter often arises within the former.3,4 Clinical distinctions between the various categories within the types of liposarcoma remain important because of their different modes of biologic behavior. However, a significant proportion (10%–20%) will dedifferentiate during the course of a mean period of 7 to 8 years, with the capacity to metastasize, irrespective of the extent of dedifferentiation.5
Based on the high local recurrence rate in patients with RPLS, a few recently published studies have demonstrated the efficacy of radiotherapy alone as well as the benefit of combining chemotherapy and radiotherapy to improve the sensitivity of liposarcomas to radiotherapy and to provide systemic coverage.6,7 Furthermore, there have been limited data on the role of pseudoneoadjuvant treatment prior to resection of recurrent RPLS.
The purpose of our retrospective study was to investigate treatment in our patients with a diagnosis of primary or locally recurrent RPLS. We examine the incidence of recurrence, the therapy in the recurrence situation, and overall and recurrence-free survival.
Patients and Methods
Patients and data collection
During the study period between October 2004 and November 2017, we analyzed the data of all patients who underwent surgical resection for a primary or recurrent RPLS in our clinic. A total of 34 patients with abdominal liposarcoma met inclusion criteria. Patient clinical information and operative and histologic reports were reviewed retrospectively from the archives of the University Hospital Tübingen and from telephone inquiries to their primary physicians and treating oncologists. Wherever available we also used the information provided to us by the referring hospitals for the patients who experienced recurrence after external treatment. Ethics approval of the University of Tübingen was given (project number 550/2015B02). Because of the retrospective nature of the study there is no possibility for written consent (Table 1).
Definitions
Primary disease was defined as presentation to our hospital at the time of initial diagnosis (including 1 patient who underwent R2 resection at an outside hospital and who was referred for urgent completion). Recurrent disease was defined as presentation to our hospital with locally recurrent disease after initial (R0/R1) resection.
Beginning in 2011, we introduced standard preoperative biopsy in all patients on first presentation to confirm the histologic diagnosis and to define the need for neoadjuvant therapy using a core needle biopsy with a 12- to 16-gauge diameter. In cases of G2/G3 liposarcoma and patients with recurrent events, (pseudo)neoadjuvant radiotherapy or chemotherapy, or both, was requested before aggressive radical surgery. Patients were only withheld neoadjuvant treatment if they experienced severe comorbidities and/or were older than 70 years.
The overall goal of surgical resection in primary RPLS was gross resection of the tumor with en bloc removal of infiltrated organs and/or retroperitoneal musculature to achieve negative margins on pathology, which were defined as tumor cells within 1 mm of the stained tumor's surface.
The line between a left- or right-sided RPLS was by definition the inferior vena cava to differentiate between a left and a right retroperitoneal compartment. Bilaterally extending sarcomas or sarcomas exceeding the inferior vena cava line at presentation were defined as not radically resectable.
Tumor size was determined as the maximum tumor diameter by our pathologists and radiographic investigation. Adequate surgical resection was evaluated based on a pathology and surgical report. Microscopically clear margins were defined if the tumor was not within 1 mm or more of the edge of the inked surgical margin. According to the World Health Organization classification primary histology was classified into 4 distinct subtypes: well-differentiated, dedifferentiated, myxoid/round cell, and pleomorphic. The grading was documented according to the French Federation of Cancer Centers Sarcoma Group grading systems in grade 1, 2, 3, or 4.2
Data analysis
Patient follow-up was done at our sarcoma center or at specialist outpatient clinics and physician practices at 6-month intervals. Disease status (alive, dead of disease, or dead of other causes without evidence of recurrence) was documented. Local recurrence was defined as any clinical and radiologic evidence of tumor regrowth at the primary site in the retroperitoneal space. Distant metastases were diagnosed by radiologic confirmation of sarcoma at distant sites.
Disease-free survival and overall survival were also recorded. Recurrence-free survival was defined from date of primary complete resection to date of any recurrence, local or distant. Overall survival was defined from the date of primary resection to date of death.
Statistics
All statistics were performed with IBM SPSS Statistics for Windows, version 26.0 (IBM Corp, Armonk, New York). Comparison between groups was carried out by χ2 test or Fisher exact test for nominal variables and Mann-Whitney U test or Kruskal-Wallis test for continuous variables, as appropriate. A probability of less than 0.05 was considered to be statistically significant. All P values reported are results of 2-sided testing. Where needed, Bonferroni correction was applied.
Results
During the study period, 34 patients with RPLS matched the inclusion criteria. Median age at presentation was 59 years (range, 35–81 years). The sex distribution was almost equal (16 female, 18 male). A total of 23 patients (67.6%) presented to the University Hospital Tübingen with recently diagnosed liposarcoma. One of these 23 patients had an incomplete resection at an outside hospital and was admitted for completion of resection. The other 11 patients (32.4%) presented to our hospital with locally recurrent disease after initial resection at another hospital. Of the total of 11 patients with recurrent sarcoma, 7 patients were admitted for the first episode of recurrence, 3 for the second, and 1 for the third recurrence. Overall mean size of the LPS during primary resection was 21 cm (range, 5–44 cm). The mean size of all primary lesions presented to our hospital was 25 cm (range, 10–44 cm).
The clinical goal in primary RPLS was complete radical resection of the tumor in the right or left compartment with en bloc removal of closely associated organs. In recurrent disease, our desired surgical treatment was radical resection, as organ sparing as possible. Within the patient total (n = 34), a radical resection (R0) could be achieved in 7 patients (20.6%). Table 2 presents the therapeutic procedures of the study population. All 34 patients had surgical intervention as part of the therapy.
A total of 10 of the 34 patients underwent neoadjuvant treatment. They had all presented to our center with untreated primary disease. Five of those patients had percutaneous radiation (± ifosfamide sensitizing), and 5 had percutaneous radiochemotherapy following a median of 4 (range, 3–5) cycles of doxorubicin/ifosfamide. Median radiation dose for these patients combined was 50.4 Gy (range, 42.2–50.4 Gy).
Three other patients received 5 cycles of doxorubicin/ifosfamide of adjuvant chemotherapy after primary resection according to the Interdisciplinary German Sarcoma group (IAWS) protocols.
Of the patients who developed recurrence but had not undergone radiation therapy, 4 received pseudoneoadjuvant radiation therapy; in 3 patients it was combined with pseudoneoadjuvant doxorubicin/ifosfamide chemotherapy. For 3 patients it was the first recurrence, and for 1 it was the second. All 4 underwent repeat laparotomy afterwards. Three were marginally resectable (R1) and 1 was not.
When irresectable disease was found, most patients underwent palliative chemotherapy. Most of them were treated with varying cycles of doxorubicin and ifosfamide, 1 patient received gemcitabine, and a few had second-line treatment with pazopanib, olaratumab, and/or ecteinascidin.
Histologic workup
Regarding histologic workup, the sarcomas in patients with primary disease (n = 23) were well differentiated (G1) in 3 (13%), moderately differentiated (G2) in 12 (52.2%), and poorly differentiated (G3) in 8 (34.8%) cases. In the referred patients with recurrent disease (n = 11), there were 4 well differentiated (G1; 36.3%), 5 moderately differentiated (G2; 45.4%), and 2 poorly differentiated (G3; 18.1%). Except for 1 patient with sarcoma recurrence, all patients were staged T2b. According to the World Health Organization classification, primary histology was classified as well differentiated in 7 of the 34 patients (20.6%; primary sarcoma n = 3, recurrent sarcoma n = 4), dedifferentiated in 20 (58.8%; primary sarcoma n = 15, recurrent sarcoma n = 5), and myxoid/round cell in 2 patients (5.9%). One of the patients exhibited a pleomorphic subtype. In 4 patients the pathology report was inconclusive of which subtype was present and slides/tissue was no longer available.
Surgical and survival outcome
Of the 34 patients studied in this analysis 7 patients had complete (R0) resections; 27 were considered R1 by the pathologist; and 29 patients developed recurrent disease, whereas 5 did not. Local recurrence was the same between the patients that had complete resection and R1 resections (85% each; Fisher exact test 1.000).
Time to recurrence, which was defined as the time between the initial resection and the first date new LPS lesions were diagnosed on imaging or by histology (if imaging was inconclusive) was a mean of 27 months (StD ±27 months; range, 7–156 months). There was no statistical difference between the time to recurrence in the neoadjuvant and the surgery only group (Mann-Whitney U test 0.664; Fig. 1).
Citation: International Surgery 106, 3; 10.9738/INTSURG-D-19-00011.1
We also performed subanalyses comparing only the patients who had been pretreated with chemotherapy followed by radiotherapy to those who only underwent radiotherapy. Primary inoperable patients who underwent chemotherapy followed by radiotherapy had significantly earlier tumor recurrence (Fig. 2).
Citation: International Surgery 106, 3; 10.9738/INTSURG-D-19-00011.1
Tumor recurrence was found in 29 of 34 patients (85%). The mean time to recurrence was 27 months (StD ± 32 months; range, 2–156 months). There was no difference in the rate of recurrence with regard to the resection margins (R0, 6/7 [85%] versus R1, 4/27 [85%]; P 1.000). The time to recurrence appeared to be longer for R0 resections; however, the difference was not statistically significant (40 ± 58 versus 23 ± 22 months, P 0.893).
The rate of recurrence was 7 of 10 patients (70%) who underwent neoadjuvant treatment prior to resection and 22 of 24 patients (92%) primarily treated surgically (P 0.138). The time to recurrence was 22 months in the neoadjuvant group (StD ± 23 months; range, 2–67 months) compared with 28 months in the other group (StD ± 34 months; range, 3–156 months; P 0.628).
Of the 29 patients who developed recurrence 22 underwent a second laparotomy. Of these patients, 19 were resectable: 7 complete (R0) and 12 marginal (R1).
The mean follow-up for all patients was 68 months (StD ± 54 months; range, 7–236 months) since their primary resection. The neoadjuvant group had a median follow-up of 67 months (StD ± 55 months; range, 7–106 months), whereas the surgery group had a median follow-up of 74 months (StD ± 62 months; range, 13–236 months). During this study period 15 patients have died of their disease, 2 patients died of other diseases (stroke and metastatic rectal cancer), 14 are alive to date, and 3 are lost to follow-up. However, in all 3 cases their primary care physicians considered them dead, since they had not seen those patients in years, and the patients had experienced extensive tumor burden and/or been under palliative care when they had last seen them.
Figure 3 shows overall survival comparing pretreated patients to patients who underwent primary resection (with or without adjuvant chemotherapy), with a significant survival benefit for the patients who underwent neoadjuvant radiation (log rank P 0.045).We also compared all patients who had tumor resections, including nephrectomy, during their first surgical procedure (excluding open biopsy). This was true for 16 patients. The other 18 patients had marginal kidney-sparing sarcoma excisions removing the tumor as well as infiltrated mesentery/bowels. There was no difference in tumor recurrence (not shown) and no difference in survival among these 2 groups. There was also no difference in median tumor size, grading, or histologic subtype between these groups. Figure 4 shows overall survival of patients with marginal resections compared with radical resection.
Citation: International Surgery 106, 3; 10.9738/INTSURG-D-19-00011.1
Citation: International Surgery 106, 3; 10.9738/INTSURG-D-19-00011.1
Discussion
Because of the large extent of the retroperitoneum, retroperitoneal sarcomas typically develop undetected, silently growing, often without any clinical symptoms for a long time. The median size of our retroperitoneal soft tissue masses was 21 cm (range, 5–44 cm) for all our patients on their first presentation.
RPLSs are malignant tumors that rarely metastasize, and if they do, they do so only late in the course of the disease; however, nearly all of our patients have developed local recurrence. This is similar to the rates described in the literature. Local control is the most important determinant of survival because most patients with this type of sarcoma die in the absence of metastases.8 Surgery therefore remains the most effective treatment in achieving tumor-free and overall patient survival. Many authors have found a prolonged postoperative survival time in patients with tumor-negative margins.4–6 Nevertheless, incomplete surgical resection is frequent,9,10 and even complete surgical excision is associated with disappointing rates of local recurrence.11,12 In our collective we report 78% of R1 resections in patients primarily operated on in the sarcoma center, as well as 82% in those referred to us for recurrent disease. Furthermore, all patients in this study had an overall recurrence rate of 79.5%. Excluding the patients who had been referred for recurrent disease we still have to report 18 of 23 patients with tumor recurrence (78%), which is slightly higher than the roughly 60% described by Pawlik and Heslin.11 Neoadjuvant treatment did not really improve these numbers: 70% of the pretreated patients developed tumor recurrence. What is notable is the earlier recurrence in the patients who had received neoadjuvant chemotherapy prior to radiation. To our knowledge this has not been described elsewhere. Of course, larger numbers are highly necessary to get more insights into this observation.
The extent of surgery depends on whether the tumor is in the right or the left compartment. Patients experiencing right-sided RPLS undergo tumor resection with en bloc resection of the right hemicolon by an anterior approach and en bloc resection of the right kidney posteriorly. Complete access to the vena cava inferior is provided by this procedure. The duodenum and the head of the pancreas must then be removed by the Kocher maneuver, and the right liver lobe must be rotated as well after incision of the falciform and coronary ligament and the right triangular ligament. Frequently only a marginal resection between the sarcoma and the liver or the head of the pancreas and the duodenum is possible. Concerning morbidity, liver resection or pancreaticoduodenectomy, or both, are generally not recommended because of increased morbidity.13 Patients with RPLS in the left compartment undergo en bloc left hemicolectomy with the sarcoma and left kidney, as well as distal pancreatectomy and splenectomy. Resection and replacement of the abdominal aorta, inferior to the superior mesenteric artery, in patients with sarcoma on the left side was never necessary in our patients. Resection and replacement of the infrarenal inferior vena cava has been performed once in our institution. Over the years we have also developed from a more aggressive to a more conservative (i.e., organ-preserving) strategy. Nonetheless, there is an ongoing debate among surgical oncologists specializing in the field regarding what constitutes optimal surgery for retroperitoneal sarcoma, weighing the oncologic benefits of radical multivisceral resection with a possible increase in early and late morbidity and keeping in mind that radical surgery might not influence the inherent natural biology of the disease.14–16 Pollock et al17 support a possibly less aggressive surgical approach incorporating organ preservation, especially for patients with dedifferentiated sarcomas, which have a higher probability of developing distant metastases than well-differentiated sarcomas do. Tseng et al18 support the suggestions made by Neuhaus et al,19 who have pointed out that in some patients the entirety of retroperitoneal fat may be at risk of disease. Further investigation into the biology of this disease is needed.
Considering the number of patients in our study, the distribution in primary and recurrent sarcomas, and the varying treatment strategies, we can neither promote nor refuse the conclusions of Pollock et al.17 However, what we saw is that only 1 patient who underwent nephrectomy during the primary procedure showed actual infiltration of the kidney. Tumor recurrence and overall survival seemed uninfluenced, which is why we try to avoid abdominal compartment resections.
Nevertheless, the strategy in our center has been to achieve a complete resection with a wide microscopic margin along the maximum surfaces by removing additional and easily disposable organs while performing a marginal excision along critical structures. In 2012 E-Surge, a master class in sarcoma surgery, and the EORTC soft tissue and bone sarcoma group provided technical guidelines with a standardized approach to these tumors.20
In the case of recurrence, and regarding greater morbidity and the missing impact of extended resection on long-term survival, surgical resection is marginal and is more or less limited to the recurrent tumor. Hamilton et al21 provided data on patients they had treated neoadjvuantly in the recurrence setting. They provide data on 30 (26 RPLS) patients with recurrent or residual disease whom they treated with radiotherapy prior to subsequent resection. They reported a 5-year OAS of about 56% from the second resection.21 Of the patients who underwent pseudoneadjuvant radio(chemo)therapy at our center, nearly half were inoperable after completing the pretreatment, and the others certainly did not meet 60% 5-year OAS, and thus it is hard to deduct recommendations from our cases. However, the 5-year OAS after secondary resection was 50%. This was also the main criticism posed in a recent review by Bagaria et al.22
In sarcomas of the extremities, radiotherapy has shown marked improvements in local control.23 Previous analyses regarding morbidity24,25 have demonstrated a moderate rate of perioperative morbidity. Because RPLSs have the highest rate of local recurrence of all different histologic types of sarcoma, locally directed radiotherapy before surgery can improve the outcome of radical compartmental surgery.
Because neoadjuvant treatment was introduced at our center in 2011, the largest patient group in our retrospective study (21 out of 34 patients) underwent surgery without any additional treatment. Three patients prior to 2011 were given adjuvant chemotherapy (doxorubicin/ifosfamide), with mixed results. To date 9 of 10 patients who underwent neoadjuvant radio(chemo)therapy are alive. At a mean follow-up of 53 months only 1 death has been reported, whereas in the surgical group less than 60% of patients were alive after 50 months.
Few studies have assessed the long-term outcomes of modern radiotherapy for RPLS administered before surgery compared with surgery alone, and large trials are warranted. A larger cohort from Pennsylvania matched 174 patients who underwent neoadjuvant radiotherapy with a group of patients who underwent surgery only. They showed a small benefit for the patients who underwent neoadjuvant treatment,26 whereas Stucky et al7 showed better local control but no oncologic benefit for neoadjuvant radiation therapy in retroperitoneal soft tissue sarcomas (of which 68% were RPLS). Some also report a survival benefit only for well-differentiated RPLS.27 The results of the STRASS(EORTC) trial failed to demonstrate a benefit for preoperative radiotherapy for RPS. In the exploratory analysis, preoperative radiotherapy may benefit the LPS subgroup.28 Following the STRASS, another randomized study is in progress. STRASS 2 is going to address the role of neoadjuvant chemotherapy [Surgery With Our Without Neoadjuvant Chemotherapy in High Risk RetroPeritoneal Sarcoma (STRASS2)]. In a phase 2 study the effect of trabectedin in advanced RPLS and well-differentiated/dedifferentiated liposarcoma is under evaluation (EudraCT Number: 2012-005428-14).
Conclusions
The results summarized in this paper are based on a retrospective analysis of patients with RPLS at a sarcoma center. Prospective randomized studies concerning the management of these tumors are difficult to perform. The therapy of these sarcomas often is not evidence based and is performed according to retrospective data and personal experience. Even though 70% of our patients developed recurrent disease, neoadjuvant radiotherapy appears to have positive benefit on the outcome.
Probability of recurrence comparing pretreated patients (red) with patients who underwent primary resection (blue). Log rank: 0.636.
Time to recurrence comparing 5 patients each who underwent either neoadjuvant chemotherapy followed by radiotherapy (red) or neoadjuvant radiotherapy only (blue). Log rank test P = 0.003.
Overall survival comparing pretreated patients (red) to patients who underwent primary resection (blue). Log rank: 0.045.
Overall survival of patients with marginal resections (red) compared with radical resection (blue). Log rank: 0.646.
Contributor Notes