Soft tissue sarcoma is rare, accounting for less than 1% of all cancers, and retroperitoneal sarcoma (RPS) accounts for about 15% of all soft tissue sarcomas (1). RPS is usually asymptomatic and often detected after the tumor has grown large because there is no anatomical barrier around the area (2). Although surgery to address large tumors is often challenging, complete surgical resection is the only potential curative treatment for patients with localized RPS (3). The 10-year recurrence rate for extremity sarcoma is 25%, while the five-year recurrence rate for RPS is as high as 50% (4). Unlike extremity sarcoma, the tumor-related mortality of RPS is mainly due to local recurrence (LR) in the absence of distant metastasis (5). Therefore, en bloc resection with adjacent structures is recommended to increase the safety margins of RPS treatment (2). The optimal extent of resection is considered to encompass both the attainment of good oncologic outcomes and the avoidance of serious complications. Resection of the kidney and colon is accepted as a treatment for RPS due to the relatively low morbidity, but resection of the pancreas is more controversial given the accompanying high risk of morbidity and death. Li et al. reported that pancreaticoduodenectomy (PD) is a feasible way to achieve complete resection for right-sided RPS despite the many major complications that may occur (6). Bagaria et al. also reported that distal pancreatectomy (DP) for primary RPS can achieve complete resection with acceptable morbidity and oncologic outcomes (7). On the other hand, Flacs et al. reported that pancreatic resection was associated with significant postoperative morbidity and mortality, and PD should be avoided whenever possible (8).

In this study, we aimed to evaluate postoperative outcomes of DP for RPS abutting the pancreas in the left upper quadrant (LUQ) area.

This retrospective study was approved by the institutional review board (IRB) of our institution (IRB no. 2021-05-125), who also waived the need for informed consent.

In this study, among 86 patients with tumors located in the LUQ and abutting the pancreas, 33 (33/86; 38.4%) patients underwent DP. The rates of complications and severe complications were similar between the DP and non-DP groups. POPF occurred in six patients (6/33; 18.2%) in the DP group, but all cases were grade B. Fourteen patients (14/33; 42.4%) showed microscopic pancreatic invasion in the DP group. Resection of the pancreas itself was not a risk factor associated with OS and LR. However, microscopic pancreatic invasion was a risk factor associated with LR but not OS in the DP group. Preoperative CT findings could not discriminate the presence of microscopic pancreatic invasion.

Complete resection achieving negative margins in RPS improves survival (13). However, the extent of the appropriate resection range is still being studied. Singer et al. (14) reported that contiguous organ resection without nephrectomy was a risk factor associated with death from tumor. Thus, they reported that nephrectomy may be necessary to achieve complete resection in retroperitoneal LPS. Similarly, we reported in a previous study that nephrectomy had a beneficial effect on disease-free survival in retroperitoneal LPS (15). Bonvalot et al. (16) observed that complete compartmental resection more significantly reduced the recurrence rate than simple complete resection or contiguously involved organ resection. However, in their study, the pancreas was not included in compartmental resection unless it was involved. The primary reason to consider resection of the pancreas is because the incidence of complications is high. In a systematic review in 2005, there were studies in which the morbidity rate was reported to exceed 40%, with the highest rate (of 64%) reported after DP by Knaebel et al. (17). However, Korrel et al. (18) recently found that the severe complication rate was 21.3% to 34.9% after DP, which was lower than in previous reports.

There have been few studies on the outcomes of patients undergoing DP in RPS. Keung et al. (19) retrospectively reviewed 43 cases from a single-center cohort and reported outcomes after DP in patients with non-pancreatic retroperitoneal tumors; notably, 39.5% (17/43) of their patients were indicated for RPS. In their study, the rates of complications and severe complications were 65.1% (28/43) and 23.3% (10/43), respectively. Also, POPF occurred in 32.6% (14/43) of cases, all of which were grade B (no grade C). Flacs et al. (8) retrospectively reviewed 50 cases in a dual-center study and reported outcomes of pancreatic resection for RPS. In their study, 86% (43/50) of patients underwent DP, and rates of complications and severe complications were 64% (32/50) and 28% (14/50), respectively. Also, POPF occurred in 14% (7/50) of cases, with 12% (6/50) of cases being grade B and 2% (1/50) of cases being grade C. The Trans-Atlantic Australasian Retroperitoneal Sarcoma Working Group (TARPSWG) retrospectively reviewed 280 cases in a multicenter study and reported outcomes after DP for RPS (7). In their study, rates of complications and severe complications were 62.5% (175/280) and 39.6% (111/280), respectively. Also, POPF occurred in 23.6% of cases—18.6% being grade B and 5% being grade C. In a recent meta-analysis, a total of 8,864 patients who underwent DP were included, and POPF occurred in 20.4%, with 20.2% and 1.6% being grades B and C, respectively (20). In our study, rates of complications and severe complications in the DP group were 57.6% and 18.2%, respectively, which were higher than those of 43.4% and 13.2% in the non-DP group, respectively, but not significantly so. These results suggest that the fear of greater complications in DP may not be warranted. Also, POPF occurred in 18.2% of cases, all of which were grade B. Among grade B cases, five patients required percutaneous drainage and one patient needed a pancreatic drain for more than three weeks after surgery. Also, among 27 patients without POPF, there were eight biochemical leaks. However, there were no cases of organ failure, reoperation, or death. Notably, this study achieved similar or better results relative to those of previous studies, and our results suggest that DP can be safely performed in patients with a pancreas-abutting mass located in the LUQ with respect to complications. However, performing PD in RPS is still controversial. Flacs et al. (8) reported that mortality occurred in 40% of patients who underwent PD. Thus, they suggested that PD should be avoided whenever possible. However, only five of their patients underwent PD, and the study sample size was small. On the other hand, Li et al. (6) reviewed the details of 27 patients who underwent PD for RPS and reported that rates of severe complications, POPF, and mortality were 40.7% (11/27), 29.6% (8/27), and 3.7% (1/27), respectively. Thus, they concluded that PD in RPS is a feasible way to achieve complete resection.

Studies have shown that macroscopic complete resection improves OS and local control (21, 22). Also, Bonvalot et al. (16) reported that microscopic positive margins decreased OS by 1.7-fold and increased the recurrence rate by 3.4-fold. When only patients with confirmed microscopic pancreatic invasion were analyzed among all patients with resected pancreas, Fairweather et al. (23) confirmed invasion in 42.9% (3/7), Flacs et al. (8) in 34% (17/50), and the French Sarcoma Group (24) in 15.2% (5/33), respectively. In addition, the TARPSWG analyzed only patients who underwent DP and confirmed microscopic pancreatic invasion in 38.2% (107/280) (7). Similarly, in our study, microscopic pancreatic invasion was confirmed in 42.4% (14/33) of patients who underwent DP. The TARPSWG reported that microscopic pancreatic invasion did not affect either LR or OS, but the R1 pancreatic margin was associated with LR (7). However, microscopic pancreatic invasion was deemed a risk factor associated with LR when multivariate analysis was performed in our study.

Due to widespread availability, CT has become increasingly useful in the diagnosis and preoperative evaluation of almost all anatomical diseases. However, diagnostic challenges still remain in the precise localization and detection of adjacent organ invasion in patients with RPS. Furthermore, achieving success becomes even more challenging when it comes to determining the presence or absence of microscopic organ invasion. We hypothesized that, when greater lengths of RPS are in contact with the pancreas, when the pancreas is displaced by the tumor, or when the pancreas is in contact with the enhancing solid portion of the RPS, the possibility of microscopic pancreatic invasion may increase. However, none of these findings were significant predictors for microscopic pancreatic invasion. Thus, our results show that, even using state-of-the-art CT imaging, it is difficult to preoperatively predict microscopic pancreatic invasion by RPS.

This retrospective study had some limitations. First, this was a single-center cohort of patients referred to a tertiary medical center and, due to our relatively small sample size and only a single ethnicity, the results may be misestimated. Also, the inclusion period of this study was quite long, which causes selection bias. During this period, the therapeutic approach changed and clinician experience varied.

In conclusion, DP is reasonable for treating pancreas-abutting RPS located at the LUQ when both complications and complete resection are considered.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Samsung Medical Center. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

KK acquired, analyzed, and interpreted data and wrote the manuscript. KL designed the project, analyzed and interpreted data, and wrote the manuscript. JL analyzed and interpreted data and wrote the manuscript. SJ and SL acquired data. JH and JK analyzed and interpreted data. JP interpreted data. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.