(P 217) PLANNED MARGINAL (R1) RESECTION IN PRIMARY RETROPERITONEAL SARCOMAS: RETROSPECTIVE SERIES ON 168 PATIENTS FROM A NATIONAL SPECIALIZED REFERRAL CENTER

Objective: Surgery is the standard treatment for retroperitoneal sarcomas (RPS). The recommended approach is to systematically resect adherent viscera even when not overtly invaded to increase margin. High morbidity associated with organ resection like major vessels may be avoided with locally marginal planned resection with unknown long-term results.
The objective was to assess the outcome of planned marginal resection (R1) on large vessels in a high-volume specialized referral center after primary RPS surgery.

Methods: All patients consecutively operated in our sarcoma center for a histologically proven primary well-differenciated liposarcomas (WDLPS), dedifferentiated liposarcomas with undifferentiated components (DDLPS) and leyomyosarcomas (LMS) between 2000 and 2018 were retrospectively identified. The R1 classification (defined with margin less than 1 mm) was subdivised in two: “vascular R1” if the contact of the tumor cells were near a vascular structure and “non-vascular R1” if the limits of resection were not a vascular structure.

Results: 168 patients among the 390 consecutive patients (43.1%) were identified. Among histological category, there were 42 WDLPS (25%), 100 DDLPS (59.5%) and 26 LMS (15.5%). The mean size of tumor was 20cm. There was 65 patients in the R0 group and 103 patients in the R1 group including 63 patients in the “non-vascular R1” group (61.2%) and 40 patients in the vascular R1 group (38.8%). The overall survival for entire population was 71.7% at 5 years. After a median follow up of 4.68 years, overall survival (OS) rates at 5 years were respectively 80% [69-92], 71% [57-88] and 60% [44-80], for R0, non-vascular R1 and vascular R1. Disease free survival rates at 3 years were respectively 65.1% [53.7-78.9], 56.1% [44.1-71.4] and 46.1% [32.1-66.0], for R0, non-vascular R1 and vascular R1. In multivariate analysis, psoas muscle resection, leiomyosarcoma histological subtype and male gender were prognostic factor (respectively 2.62 [1.35-5.09] p=0.004, 4.82 [1.57-14.75] p=0.005, 2.45 [1.19-5.06] p=0.01). Disease free survival was significantly reduced by a non-vascular R1 1.75[1.02-2.98], p = 0.04 or a vascular R1 HR=1.75[1.00-3.05], p=0.05 resection.

Conclusion: Among the 168 patients with retroperitoneal WDLPS, DDLPS and LMS analyzed, neither vascular R1 nor non-vascular R1 was statistically associated with a worse OS. Yet, after control of the confounding factors, they were associated with a shorter DFS. This results suggest two things: R1 resection was not as good as R0 resection in term of control of the disease and being R1 against a vascular structure is quite similar to other R1 in terms of recurrence. Psoas muscle invasion, male gender and LMS subtype were also identified as prognostic factor and should be taken in account in tailoring treatment strategy.

Overall survival in 168 patients who underwent compartmental resection for a retroperitoneal sarcoma according to marginality.pdf

Univariable and multivariable analyses of predictors of overall survival in 168 patients who underwent compartmental resection for retroperitoneal sarcomas.pdf

Disease free survival in 165 patients who underwent compartmental resection for a retroperitoneal sarcoma according to marginality.pdf

Factors of disease-free survival at univariate and multivariate analyses in 165 patients who underwent compartmental resection for RPS.pdf