Retroperitoneal Sarcomas: Prognostic Factors and Outcomes of a Series of Patients Treated at a Single Institution

Abstract

Retroperitoneal sarcomas (RPS) are rare heterogeneous tumors arising in the retroperitoneum with unique biological and behavioral patterns that are thought to be closely linked to histology. The aim of the study was to audit our results and analyze various clinico-pathological factors including surgical excision, histology, and their implications on the recurrences and survival outcomes in RPS. Retrospective analysis of patients treated at a tertiary referral center in India from March 2008 to July 2017 was performed. The clinico-pathological variables were analyzed for their association with tumor recurrence and survival with special emphasis on histological subtype. The primary outcome was overall survival (OS). One hundred consecutive patients operated for RPS were analyzed. Of these, 27 were operated for recurrent tumors. Liposarcomas (LPS) and leiomyosarcomas (LMS) constituted 50% (n = 50) and 30% (n = 30) of patients respectively. Complete tumor excision was achieved in 83%, with 43% patients undergoing adjacent organ resection. At a median follow-up of 25.3 months, the median disease-free survival (DFS) and overall survival (OS) were 30 months and 87.8 months respectively. On multivariate analysis, tumor grade was the only factor to significantly affect survival (p = 0.001 for DFS and 0.005 for OS). There was no difference in survival outcomes between infiltrative and adhesive tumors with respect to adjacent organ invasion (p = 0.361 for OS). Tumor grade remains an important prognostic factor affecting disease-free and overall survival in retroperitoneal sarcomas irrespective of tumor size, site, and histology.

Introduction

Retroperitoneal sarcomas (RPS) constitute 15% of all soft tissue sarcomas [1]. They are rare [2], heterogeneous group of tumors with variable biological behavior and outcomes. The peak incidence of RPS is in the 5th decade, although any age can be affected [3]. Liposarcomas (LPS) and leiomyosarcomas (LMS) represent the most common histological types. Complete surgical excision remains the standard of care in order to achieve long-term cure [4].

Patterns of recurrences are closely related to histology and grade, with low-grade tumors showing a predominantly local recurrence pattern. On the contrary, high-grade tumors often fail systemically, the most common site being the lung. Ten to fifteen percent of patients present with synchronous metastases [5] and have a dismal outlook. Local recurrences in low-grade tumors tend to be asymptomatic; these recurrences manifest only when the tumor attains a considerable size. Owing to the anatomical site and the presence of critical structures surrounding the tumor, re-excision may also be inadequate/ incomplete. Multiple recurrences due to inadequate/incomplete surgeries worsen the overall outcome and increase the risk of tumor dissemination [3]. This underscores the principle of first attempt at resection being the best attempt at complete resection.

Both patient and tumor-related prognostic factors influence outcomes. Patient factors include age and gender; tumor-related factors include tumor size, histology, grade, location, and extent of local and distant disease.

The aim of this study is to audit our results and analyze factors affecting recurrence and survival in resected patients of RPS at a single tertiary referral center in India. Over the years, there has been a steady rise in the number of cases of RPS being operated as shown in Fig. 1.

Fig. 1.

Bar graph showing a steady increase in the number of operated patients at our institution over the years

Materials and Methods

This was a retrospective analysis of a prospectively maintained database of patients undergoing retroperitoneal sarcoma excision with a curative intent at Tata Memorial Hospital from March 2008 to July 2017. Patients with gastrointestinal stromal tumors (GIST), desmoids, and visceral and pediatric sarcomas were excluded. All patients underwent a thorough clinical examination with emphasis on pertinent details regarding the type of surgical procedure and recurrence-free intervals. Outside specimens were reviewed by a group of expert sarcoma pathologists at our institute; in cases where establishing a histological diagnosis deemed difficult, these were subjected to a multidisciplinary meeting involving the clinicians, pathologists, and radiologists. Tumors were categorized as upfront or recurrent respectively.

All patients underwent a computed tomography (CT) scan of the abdomen and thorax. The type of treatment (surgical/neoadjuvant therapy) was based on the size and location of the tumor and the ability to achieve R0 resection. Patients whose tumors were deemed such that a safe R0 resection was not feasible were given neoadjuvant therapy (radiation or chemotherapy). Chemotherapy administered was doxorubicin and ifosfamide–based combination. Reassessment after neoadjuvant therapy was performed 4 weeks after last chemotherapy/radiation. Where the tumor was found to be attached/infiltrating adjacent organs, organ excision was done wherever feasible. Histopathology reports were analyzed for tumor size, histology, nodal involvement (if any), margins, grade, and adjacent organs being microscopically involved or not. Tumor size/diameter was calculated on the surgical specimen excised. In case where multiple tumor masses were removed, the diameter of the largest mass was taken into account. Histology was in accordance with the World Health Organization (WHO) classification. Tumor grade was classified based on differentiation, necrosis, and mitotic rate per high-powered field as per the FNCLCC (Fédération Nationale des Centres de Lutte Contre Le Cancer) system [6]. Each adjacent organ removed was counted for except for the kidney and ureter which was considered as single organ. Margins were evaluated both grossly and microscopically and categorized as clear, microscopically positive, or grossly positive. Post-operative morbidity and mortality were graded as per Clavien-Dindo grading system [7].

Patients received adjuvant therapy based on histology, grade, tumor size, and margin status. The decision to administer adjuvant therapy was taken in multidisciplinary clinic. All patients were followed up every 3 months for the first 2 years, every 6 months for the subsequent 3 years, and annually thereafter. At each visit, a complete physical examination was done with imaging of the abdomen and chest. Local recurrence (LR) was defined as recurrences occurring within the retroperitoneum and distant recurrence (DR) was defined as those occurring in the lung, liver, peritoneal lining, and nodes.

Statistical Analysis

Data were analyzed using the Statistical Package for the Social Sciences (SPSS) ver. 25 software (IBM Corporation, NY; formerly SPSS Inc. Chicago, IL). Descriptive statistics including median, frequency, and percentage for categorical variables was used. The Kaplan-Meier method with Cox logistic regression was used to calculate survival. Statistical significance was defined as p value < 0.05. Univariate and multivariate analyses were used to identify prognostic variable which would affect recurrences and survival. Overall survival (OS) was defined as the time period from the date of surgical resection until the date of last follow-up or death while disease-free survival (DFS) was defined as the time period from the date of surgical resection until the development of recurrent disease. Time to tumor recurrence was defined as the date of documented pathologic or radiographic recurrence of disease from the date of previous definitive treatment. The primary outcome was OS and secondary outcome was recurrence rate.

Results

Patient Demographic, Surgical, and Pathological Details

A total of 100 patients were included in the study. The various demographic, surgical, and pathological details are depicted in Table 1. In upfront and recurrent tumors, R0 resection rates were 83% and 77% respectively. The histological subtypes are shown in Table 2. Grade III tumors in particular demonstrated the following histologies: LMS 73.3% (n = 22/30), LPS 44% (n = 22/50), spindle cell sarcoma 85.7% (n = 6/7), and pleomorphic sarcoma 100% (n = 4/4).

Table 1.

Patient demographic, surgical, and pathological parameters

Parameter		Values
Type of tumors at presentation	Upfront	73 (73%)
	Recurrent	27(27%)
Gender	Males	60 (60%)
	Females	40 (40%)
Age (years)	Median	52
	Range	21–81
ECOG status	0	16 (16%)
	1	84 (84%)
Symptoms at presentation	Pain in abdomen	46%
	Abdominal lump	26%
Prior biopsy	Yes	72 (72%)
	-Upfront + recurrent	50 + 22
	No	28 (28%)
Blood loss (median)		1350 ml
Median hospital stay (range)		9 (6–67) days
Resection status	R0	83 (83%)
	R1	08 (8%)
	R2	09 (9%)
Post-operative complications (29%)	II	2 patients
	III	16 patients
	IV	5 patients
	V	1 patient
Median tumor diameter (range)		15 cm (3.5–60 cm)
Tumor grade	I	31 (31%)
	II	13 (13%)
	III	56 (56%)
pT status	T1b	13 (13%)
	T2b	87 (87%)
pN status	N0/Nx	97 (97%)
	N+	03 (3%)
Margins	Negative	66 (66%)
	Positive	34 (34%)
Neoadjuvant therapy (11%) (n = 11)	Chemotherapy	7 (7%)
	Radiotherapy	1 (1%)
	Both (sequential/concurrent)	3 (3%)
Adjuvant therapy (63%) (n = 63)	Chemotherapy	09 (9%)
	Radiotherapy	50 (50%)
	Concurrent chemoradiation	04 (4%)

Table 2.

Histological subtypes on final pathology, both in upfront and recurrent tumors

Histology	Number	Percent
Liposarcoma	50	50
-WDLPS	29	29
-DDLPS	21	21
Leiomyosarcoma	30	30
Spindle cell sarcoma	7	7
Myofibroblastic sarcoma	2	2
Myxoid sarcoma	5	5
Pleomorphic sarcoma	4	4
Desmoplastic round cell	1	1
Follicular dendritic cell sarcoma	1	1
Total	100	100

WDLPS Well differentiated liposarcoma, DDLPS Dedifferentiated liposarcoma

Contiguous organ resections were performed in 43% of patients and vascular resections in 7%. Additional 7% of patients underwent both multivisceral and vascular resections in the same sitting. The kidney was the most common organ resected (n = 33) followed by the large bowel (n = 22) and adrenal (n = 15) (Table 3). The total number of organs resected was 72 with median number of organ resected being 2. Sixty-eight patients (68%) had tumors adherent to the adjacent organs while 40 patients (40%) had tumor invasion into the adjacent organs on final histological confirmation that also included 6 patients with inferior vena cava (IVC) sarcomas. Adjacent organ resections were evaluated with respect to histology, grade of the tumor, and quadrant in which the tumor was predominantly located (Table 4). It was found that high-grade tumors, tumors in the right lower quadrant, and those having dedifferentiated liposarcoma and leiomyosarcoma histologies had a higher necessity for adjacent organ resection.

Table 3.

Adjacent organ removed during surgery: adhesion alone versus infiltration on pathology

Organ removed (n)	Adhesion	Infiltration on final HPR
Kidney/ureter (33)	18	Renal capsule: 1
		Renal parenchyma: 3
		Renal pelvis and hilum: 4
		Ureteric wall: 2
		Perirenal/periureteric fat positive: 5
Spleen (10)	9	Spleen capsule: 1
Pancreas (8)	6	Parenchyma: 2
Small intestine (13)	5	Serosa involved: 3
		Muscularis propria involved: 4
		Subserosal involvement: 1
Adrenal (15)	14	1
Colon* and rectum (22)	14	Full thickness involved: 5
		Serosa involved: 2
		Pericolonic fat/mesentery involved: 1
Appendix (10)	9	1
Liver (3)	2	1
Vessels only (5)	1	Adventitia involved: 3
		Vessel wall infiltrated: 1
Ovary and tube (1)	-	1
Gall bladder (6)	6	-
Stomach (2)	2	-
Lung (1)	1	-
Diaphragm (1)	1	-
Testis and scrotum (1)	1	-
Iliac bone (1)	1	-

HPR histopathology report. *Both right and left colon

Table 4.

Relationship of adjacent organ removal with various parameters including grade, location, and histology

Parameter	Infiltration	Adhesion	Total
1 organ removed 2 organs removed 3 organs removed 4 organs removed 5 organs removed Vessel alone removed	6 11 5 12 1 5	15 9 6 2 0 0	21 20 11 14 1 5
RUQ RLQ LUQ LLQ	11 12 8 9	6 11 7 8	17 23 15 17
Grade I Grade II Grade III	12 4 24	7 3 22	19 7 46
WDLPS DDLPS LMS Myxoid sarcoma UPS Spindle cell sarcoma Sarcoma NOS FDCT Myofibroblastic	10 11 12 1 2 2 - 1 1	7 9 7 3 3 1 1 - 1	17 20 19 4 5 3 1 1 2

RUQ Right upper quadrant, RLQ Right lower quadrant, LUQ Left upper quadrant, LLQ Left lower quadrant, UPS Undifferentiated pleomorphic sarcoma, FDCT Follicular dendritic cell tumour

Treatment Details

Neoadjuvant Treatment

Eleven patients (11%) received pre-operative treatment. Five patients received ifosfamide and Adriamycin–based chemotherapy. Two patients received vincristine and cyclophosphamide–based chemotherapy.

Adjuvant Treatment

Sixty-three patients (63%) received adjuvant treatment. In patients with grade III tumors, adjuvant chemotherapy, radiation, and concurrent chemoradiotherapy were administered in 8 patients, 33 patients, and 3 patients respectively. Twelve patients with grade III tumors, however, received no adjuvant therapy. The reasons included old age (n = 6), doubtful compliance (n = 1), large tumor at presentation (n = 2), prior concurrent chemoradiation (CTRT) (n = 1), and post-operative complications (n = 2). Four patients with grade II tumors received radiation/chemoradiation. When the two main histological subtypes were considered, 26 patients with LPS received adjuvant therapy (52%) which included radiotherapy in 25 patients. However, most patients with LMS received some form of adjuvant treatment which included RT (n = 18; 60%), chemotherapy (n = 3; 10%), and concurrent chemoradiation (n = 2; 6.67%).

Oncological Outcomes

The median follow-up time was 25.3 months (range 1.5 to 117 months). The median DFS was 30 months (range 25 to 35.8 months) and median OS was 87.8 months. Thirty-seven of seventy-three (50.68%) patients in the upfront setting and 18/27 (66.67%) patients in the recurrent setting relapsed at a median time of 30.2 months and 29.1 months respectively. The corresponding DFS at 5 years were 32.3% and 20.1% (Fig. 2). Among the relapsed patients, 35 were local recurrences and 20 distant recurrences (9 patients lung, 9 patients liver, and 7 patients other sites). Seventeen of seventy-three (23.29%) patients and 2/27 (7.4%) patients in the upfront and recurrent settings died of disease at a median time of 62.8 months (upfront) and not reached (recurrent) respectively. The corresponding OS at 5 years were 67.1% and 83.5% respectively (Fig. 3).

Fig. 2.

Kaplan-Meier curves depicting disease-free survival (DFS) in upfront and recurrent tumors (blue line: upfront; red line: recurrent)

Fig. 3.

Kaplan-Meier curves depicting overall survival (OS) in upfront and recurrent tumors (blue line: upfront; red line: recurrent)

Relationship of survival with respect to various prognostic factors is shown in Table 5. The univariate analyses of gender, age at presentation, tumor size, tumor grade, histological subtype, type of presentation (upfront versus recurrent tumors), surgical margins, presence of contiguous organ resection, and the use of adjuvant treatment were analyzed with regard to DFS and OS. It was found that tumor grade was the main factor associated with both DFS and OS (p = 0.001 and 0.005 respectively) (Figs. 4 and 5). The remaining factors did not affect DFS or OS significantly (Tables 5 and 6). Among the patients undergoing adjacent organ removal, there was no difference between those with adhesion and infiltration with respect to OS (p = 0.361) and DFS (p = 0.139).

Table 5.

Relationship of various prognostic variables on survival: Only tumor grade showed significance

Variables		Outcomes				p value (OS)
		A-D	A+D	D+D	Total
Grade	Grade I	17	12	2	31	0.005
	Grade II	6	4	3	13
	Grade III	26	16	14	56
Resection status	R0	40	26	17	83	0.718
	R1	5	2	1	8
	R2	4	4	1	9
Histology	WDLPS	14	10	3	27	0.15
	DDLPS	8	10	4	22
	LMS	17	6	7	30
	Myxoid sarcoma	3	2	1	6
	UPS	2	1	1	4
	Spindle cell sarcoma	4	-	3	7
	Others	1	3	-	4
Presentation	Upfront tumors	37	19	17	73	0.159
	Recurrent tumors	12	13	2	27
T status	T1b (< 5 cm)	9	2	2	13	0.433
	T2b (> 5 cm)	40	30	17	87
Gender	Males	33	17	10	60	0.966
	Females	16	15	9	40
Adjacent organ removal	Infiltration	15	17	8	40	0.361
	Adhesion	13	7	7	27
	None	21	8	4	33
Total		49	32	19	100

A-D alive without disease, A+D alive with disease, D+D dead due to disease

Fig. 4.

Kaplan-Meier curves depicting relationship of tumor grade on OS

Fig. 5.

Kaplan-Meier curves depicting relationship of tumor grade on DFS

Table 6.

Relationship of various prognostic factors on tumor recurrence patterns

Variables		Recurrences			p value (DFS)
		Local	Distant	Total
Grade	Grade I	15	2	17	0.001
	Grade II	3	4	7
	Grade III	17	14	31
Resection status	R0	29	18	47	0.392
	R1	3	1	4
	R2	3	1	4
Histology	WDLPS	14	-	14	0.48
	DDLPS	12	2	14
	LMS	3	12	15
	Myxoid sarcomas	2	1	3
	Spindle cell sarcoma	2	1	3
	UPS	1	1	2
	Others	-	3	3
Presentation	Upfront	17	20	37	0.55
	Recurrent	18	-	18
Gender	Males	21	8	29	0.611
	Females	14	12	26
Adjacent organ removal	Infiltration	16	10	26	0.139
	Adhesion	11	3	14
	None	8	7	15
Total		35	20	55

Out of the 67 patients with grade II and grade III tumors, 39 patients received radiotherapy. In the radiotherapy group, the 5-year OS was 55.9% (median OS, not reached) while in the no-radiotherapy group, it was 24.9% (median OS, 33.58 months) with a p value of 0.069. The corresponding 5-year DFS in the radiotherapy and no-radiotherapy groups respectively were 27.1% (median DFS, 26.35 months) and 22.5% (median DFS, 12.52 months) with a p value of 0.12.

Pattern of recurrence and prognostic factors are tabulated in Table 6. When comparing LPS and LMS histologies, local recurrences were commonly associated with well-differentiated LPS (n = 14) while distant recurrences were associated with LMS (n = 12). The overall local and distant recurrence rates in LPS and LMS respectively were 92.86%, 7.14%, 20%, and 80%. Grade III tumors were associated with higher local as well as distant recurrences. Patients who needed adjacent organ removal had a higher risk of local recurrences. Although there was no difference in survival between LPS and LMS histologies, local recurrences were higher in the former whereas distant failures were predominant in the latter.

Discussion

Surgery for RPS has undergone a paradigm shift over the years from marginal resections to multivisceral and compartmental excisions. Historically, marginal resections were commonly performed, resulting in margin-positive resections with higher local recurrence rates that paved way for wider resection margins, at the cost of removal of adjacent organs involved grossly by the tumor. The Transatlantic Retroperitoneal Sarcoma Working Group (TARPSWG) [8] published their one-decade-long (2002–2011) data on 1007 patients undergoing resection for RPS. Patients were subjected to radical surgeries that included multivisceral resections. Majority of the patients had at least 2 organs being removed. This led to higher R0 resections and improved outcomes.

In our study, adjacent organ to achieve margin negativity was 42.9%, with kidney being the most common organ removed followed by small/large bowel. Right lower quadrant tumor location, grade III tumors, and histologies like LMS and DDLPS were factors associated with a higher contiguous organ removal during surgery.

Prognosis of patients with adjacent organ removal is related to tumor directly infiltrating the particular organ or just abutting it. Hogg et al. [9], Wan Jie et al. [10], and Fairweather et al. [11] found that such infiltrating tumors had poorer outcomes compared with abutting tumors. Furthermore, Fairweather’s study [11] also correlated histology with invasion and reported that dedifferentiated or high-grade tumors were more likely to be infiltrative in nature. In our study, there was no difference in survival with adhesive or infiltrative tumors. A Resected Organ Score (ROS) has been introduced to account for differences in morbidity in patients undergoing organ resection [12]. Organs such as the pancreas (especially head) and duodenum, whose resections can often lead to serious morbidity, are assigned higher scores. Organs such as the kidney, and appendix whose removal is not so morbid, are assigned lower scores. Furthermore, patients who undergo nephrectomy as part of multivisceral resections are not at a higher risk of developing chronic kidney disease compared with the general population [13].

Table 7 depicts the outcomes of various studies on RPS. Patient’s age, gender, extent of resection, margins, and tumor grade remain important prognostic variables across all studies. Three studies included patients from the National Cancer Database (NCDB) [17, 20, 23]. In our study, only tumor grade remained significant for both DFS and OS. However, there was a trend in the improvement in OS in the radiotherapy group (although not statistically significant) with no effect on DFS. This underscores the fact that, unlike in extremity sarcomas, there is limited benefit of adjuvant radiotherapy in retroperitoneal sarcomas and that striving for a margin-negative resection is the most important factor affecting outcomes.

Table 7.

Published literature on RPS

Author/year	Period	n	U/R	5-year OS	5-year DFS	LR/DR (%)	M:F	R0:R+ (%)	Grade I/II/III	Histology	Age (years)	NACT/NART	Adj RT	Adj CT	Death	T size	MVR (%)	F/U (months)
Hogg 2016 [9]	1997-2013	79	63/27	55.3%mean	24.8% mean	41/12.2	45:45	88.9:11.1	38.9/18.9/42.2%	WDLPS: 38% DDLPS: 29% LMS: 13.3%	60*	16.7%/1.1%	8.9%	3.3%	3 (postop)	20.5 cm mean	70 (≥ 1)	61*
Panda 2015 [1]	2008–2010	23	20/3	60% approx.	NA	39.1	13:10	61:39	-/-/43.5%	LPS Neuro-fibroma	39.1	NA	NA	NA	1 (post op) D+D: 1	10 cm median	26.1	24
Sandrucci 2018 [14] multi-institute	2006–2011	72/138	46/26	65% 31%(R+)	NA	NA	43:29	60:25	14/31/ 37.5%	LPS: 55.5% LMS: 14% Sarcoma, NOS: 11%	> 60 79%	NA	NA	NA	NA	< 10 cm: 30.5% > 10 cm: 69.5%	NA	85
Kim 2017 [13]	1990–2014	47	37/21	NA	NA	36/22	29:18	NA	NA	WDLPS: 30.2% DDLPS: 32.3% LMS: 24%	63	RT: 52% (overall) CT: 19.8% (overall)			D+D: 2	15 cm (median)	57.2 (k)	26.8
Fairweat-her 2018 [11]	2002–2011	99/118	118/0	34–62%	NA	48/22–46	52:47	48.3/35.6	28.8/15.3/40.7%	WDLPS: 12.7% DDLPS: 31.4% LMS: 21.1%	58–59	15.3%/21.2%	7.6%	4.2%	NA	15–17 cm (median)	84 (1)	33.6
Abdel-fateh 2016 [15]	1994–2010	131	131/0	48.7 months median	NA	NA	63:68	84.4/15.6	17.6/20.6/52.7%	LMS: 39.7% LPS: 38.2% Sarcoma, NOS: 7.6%	59	-/29%	48%	23.7% (incl. NACT)	NA	12.3 cm > 15 cm: 45.8%	51.3 (1–2)	NA
Wang 2018 [16]	2012–2017	25	17/8	83.2% (1 year)	91.3% (1 year)	4/-	17/8	92:8	16/40/44%	LPS: 68% UPS: 16% LMS: 12%	56	12% Both	NA	NA	8%	< 15 cm: 20% > 30 cm: 12%	96 (1)	11
Wan Jie 2017 [10]	2000–2014	85	55/30	45 months median	21 months median	59 (all)	40/45	50:50	42/24/34%	LPS: 72% LMS: 14%	55	1.1%/-	17.6%	9.4%	NA	16.5 cm	50 (1) 9.4 (>1)	46
Stahl 2016 [17] NCDB	1998–2011	4015	NA	64.7% 54.3% (R+)	NA	NA	1821/2194	64.6/35.4	II+III: 52.2%	LPS: 5 9.1%	60	-/25.4%	69% + 5.7% IORT	11.1% (incl. NACT)	35.4%	16 cm	NA	67
Maurice 2017 [18] NCDB	2004–2014	3141	NA	43.4–84.3 months median	NA	NA	1507/1634	67.6:32.4	28.95/9.85/41	LPS: 60.8% LMS: 26.3% NOS: 4.1%	61	2.55/11.35	NA	NA	NA	17 cm (median)	NA	31.1
Guiliano 2016 [19] SEER	2002–2012	2920	NA	58.4%	NA	4.6/14.6	1414/1506	NA	24.6/9.8/40.7	LPS: 46.9% LMS: 25% NOS: 14.3%	63	NA	21.6%	11.9%	33.6%	15 cm (median)	39.5	NA
Berger 2018 [20] NCDB	2004–2013	2762	NA	70.1–84.2 months median	NA	NA	1300/1462	52.3:34.5	11.7/7.6/20%	NA	63	−/9.9%	19.9%	NA	6.3% (90 days)	19.9 cm (mean)	NA	NA
Gronchi 2013 [12]	1999–2009	523	NA	56.8%	39.4%	24.5/17.8	280/243	90.8:9.2	28.1/23.3/48.6%	WDLPS: 23.1% DDLPS: 29.6% LMS: 17.6% UPS: 13.4%	57	CT: 39.6% (overall) RT: 36.9% (overall)			D+D: 139/171	16 cm (median)	34 (1) 57 (>1)	45
Tan 2016 [2]	1982–2010	675	NA	69% (DSS)	NA	38/24	299/376	56:44	III: 64%	LGLPS: 28% HGLPS: 32% LMS: 20% SFT: 5%	60	11%4%	4%	7%	DSD: 31%	17 cm (median)	58 (1) 27 (>1) 10 (v)	39.6
Rosa 2016 [21]	1984–2013	107	92/15	71%	65%	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	79.7
Amersi 2016 [22]	1985–2010	68	NA	55%	NA	NA	25/43	NA	II+III: 75%	NA	59*	NA	NA	NA	NA	12 cm (median)	43 10 (v)	NA
Present study/2019	2008–2017	100	73/27	62% 87.8 months median	30 months median	35/20	60/40	83:17	34/11/55	WDLPS: 29% DDLPS: 21% LMS: 30% Spindle cell: 7%	52	7%/1%	50%	9%	19%	15 cm (median)	50	25.3

(1) one organ resection, (>1) more than one organ resection, (v) vascular resection, NA not available, D+D dead due to disease, (k) nephrectomy, LPS liposarcoma, LMS leiomyosarcoma, UPS undifferentiated pleomorphic sarcoma, MPNST malignant peripheral nerve sheath tumor, LGLPS low-grade LPS, HGLPS high-grade LPS, NOS not otherwise specified sarcomas, WDLPS well-differentiated LPS, DDLPS dedifferentiated LPS, DSS disease-specific survival, U upfront tumors, R recurrent tumors, NACT/NART neoadjuvant chemotherapy/radiation, IORT intraoperative radiotherapy, LR local recurrence, DR distant recurrence, MVR multivisceral resections, DSD disease-specific death, NCDB National cancer, SEER Surveillance, Epidemiology, and End Results. *Mean value

While local recurrences remain significant for low-grade tumors, distant recurrences are common in higher grade tumors. Multiple local recurrences increase risk of distant tumor dissemination even in low-grade tumors. Most studies depict recurrence rates of 20–40%. In our study, the recurrence rates were 55% with a higher local recurrence pattern. Liposarcoma histology and grade III tumors showed higher local recurrence rates.

The strengths of this study include a single-center experience in managing these complex tumors with emphasis on surgical treatment and outcomes. The limitations of this study include the short follow-up time and a relatively small patient cohort.

Conclusions

RPS are rare tumors with ever-evolving treatment strategies. This study represents the largest single-center experience from India. There has been a steady increase in the number of patients undergoing multivisceral and compartmental resections to achieve macroscopic complete excision and minimize marginality. However, diligent pre-operative work up needs to be done to anticipate and reduce morbidity associated with multivisceral resections. Tumor grade remained an important prognostic factor affecting survival in our study. There is an urgent need for centralization in the treatment of RPS tumors with development of multidisciplinary treatment approach, in order to improve the survival outcomes beyond what can be achieved by surgery alone.

Compliance with Ethical Standards

Conflict of Interest

The authors declare that they have no conflict of interest.

Informed Consent

Informed consent was obtained from the patients for surgery.