Perioperative Radiotherapy and Survival After Surgical Treatment of Nonmetastatic Retroperitoneal Sarcoma

Abstract

Background: Among patients with nonmetastatic retroperitoneal sarcoma, the use of radiotherapy (RT) use and its association with survival following surgical treatment remain unknown.

Patients and methods: Patients with surgically treated, nonmetastatic retroperitoneal well-differentiated, dedifferentiated liposarcoma, and leiomyosarcoma were identified within the SEER database (2004-2020). We used estimated annual percentage change (EAPC) analyses, multivariable logistic regression (LR) models, Kaplan-Meier survival plots, and multivariable Cox regression (MCR) models to evaluate treatment patterns and outcomes.

Results: Perioperative radiotherapy (RT) was delivered to 96 (16%) of 582 patients with well-differentiated liposarcoma, 211 (29%) of 739 patients with dedifferentiated liposarcoma, and 247 (35%) of 703 patients with leiomyosarcoma. Over time, the rate of perioperative RT use decreased from 27.3% to 5.4% for well-differentiated liposarcoma (EAPC, -3.5%; P=.02), but remained stable for dedifferentiated liposarcoma (P=.9) and leiomyosarcoma (P=.055). In multivariable LR, dedifferentiated liposarcoma (odds ratio [OR], 1.99; P<.001) and leiomyosarcoma (OR, 2.83; P<.001) independently predicted higher RT rates. In all 3 retroperitoneal sarcoma histologic subtypes, perioperative RT was not associated with lower cancer-specific mortality in either the Kaplan-Meier plots or the MCR models (all P>.05).

Conclusions: The rate of perioperative RT use was lowest in well-differentiated retroperitoneal liposarcoma. Conversely, perioperative RT was used at approximately twice that rate in dedifferentiated liposarcoma and leiomyosarcoma. These findings are discordant with the most recent guideline recommendations but should be interpreted in the context of the historical nature of the investigated cohort, which predates the publication of those guidelines. Perioperative RT was not associated with differences in cancer-specific mortality across any of the 3 examined retroperitoneal sarcoma histologic subtypes. Unfortunately, local recurrence, surgical margin status, and metastatic progression rates could not be addressed.