S. Z. Thalji1, M. Hembrook1, S. Tsai1, T. C. Gamblin1, C. Clarke1, M. Bedi2, J. Charlson3, C. G. Ethun4, T. B. Tran5, G. Poultsides5, V. P. Grignol6, J. H. Howard6, J. Tseng7, K. K. Roggin7, K. Chouliaras8, K. Votanopoulos8, D. Cullinan9, R. C. Fields9, K. A. Vande Walle10, S. Ronnekleiv-Kelly10, H. Mogal1 1Medical College Of Wisconsin,Department Of Surgery,Milwaukee, WI, USA 2Medical College Of Wisconsin,Department Of Radiation Oncology,Milwaukee, WI, USA 3Medical College Of Wisconsin,Department Of Medical Oncology,Milwaukee, WI, USA 4Emory University School Of Medicine,Division Of Surgical Oncology, Winship Cancer Institute,Atlanta, GA, USA 5Stanford University,Department Of Surgery,Palo Alto, CA, USA 6Ohio State University,Department Of Surgery,Columbus, OH, USA 7University Of Chicago,Department Of Surgery,Chicago, IL, USA 8Wake Forest University School Of Medicine,Department Of Surgery,Winston-Salem, NC, USA 9Washington University,Department Of Surgery,St. Louis, MO, USA 10University Of Wisconsin,Department Of Surgery,Madison, WI, USA
Objective: While outcomes of patients with Retroperitoneal Sarcomas (RPS) who have microscopically or macroscopically positive margins after curative-intent resection has been studied, few studies have focused on outcomes in patients undergoing surgery with a palliative intent. This study aims to define common indications for and to elucidate factors that determine outcomes after palliative-intent resection of RPS.
Patients and
Methods: Using the retrospective 8-institution United States Sarcoma Collaborative (USSC) database, all patients who underwent resection of a primary or recurrent RPS (including intraabdominal or distant metastasis) with palliative intent between 2000 and 2016 were identified. Overall Survival (OS) was estimated by the Kaplan-Meier method and compared using the log-rank test. Multivariable logistic regression and Cox-proportional hazards models of patient- and treatment-related factors were constructed to determine their effect on postoperative complications and OS.
Results: Of a total of 3,088 patients, 70 patients underwent 87 distinct palliative-intent surgeries for RPS. Median age was 62.3 years (IQR 46.7–71.5) and 57% (n=50) were female. Most common indications for palliative surgery included pain (n=23, 26%), bowel obstruction (n=18, 21%), bleeding (n=7, 8%), and infection (n=2, 2%). Dedifferentiated liposarcoma (n=18, 21%) and leiomyosarcoma (n=17, 20%) were the predominant tumor types. Median OS for the entire cohort was 10.69 months (IQR 3.91-23.23). R2 resection status was the only factor independently associated with incidence of postoperative complications (OR 4.42, CI 1.01-19.38, p=0.049). The presence of complications (HR 3.25, CI 1.35-7.81, p=0.008) and high-grade histology (HR 4.45, CI 1.19-16.62, p=0.026) were associated with increased mortality. While OS was not independently affected by resection status, in patients who underwent R2 resections, the development of postoperative complications significantly reduced survival (p=0.042) (figure 1).
Conclusions: The occurrence of postoperative complications and high grade tumor biology rather than resection status determines survival in patients undergoing palliative-intent resections for RPS. Palliative-intent R2 resections should be cautiously performed for RPS given the higher-incidence of post-operative complications which may significantly lower survival.
