Prognostic factors and nomogram construction for primary retroperitoneal myxoid/round cell liposarcoma: an analysis of population-based data

Abstract

Background: It has been reported that the prognosis for myxoid/round cell liposarcoma (MLPS/RCLPS) is inconsistent across different sites. However, there are neither prognostic studies nor predictive models that focused on MLPS/RCLPS of retroperitoneal origin. Methods: Utilizing the Surveillance, Epidemiology, and End Results database, we selected 171 primary retroperitoneal MLPS/RCLPS cases from the period between 2000 and 2019. Prognostic factors influencing disease-specific survival were identified through Cox regression analysis. These independent prognostic factors were then used to construct a DSS nomogram prediction model. The accuracy and reliability of this nomogram were evaluated using the concordance index (C-index) and calibration plots. Furthermore, we categorized patient prognosis using an X-tile based on the nomogram score. Results: The observed 5-year and 10-year DSS rates for all patients were 64.0% (95% CI: 56.2% – 71.8%) and 47.1% (95% CI: 38.1% – 56.1%), respectively. The patient cohort had a median age of 64 years, ranging from 24 to 92 years, with a slight male predominance (n = 92, 53.8%) over females (n = 79, 46.2%). Distant metastases were diagnosed in 24 patients (14%). The distribution of MLPS and RCLPS was 89.5% and 10.5%, respectively. In terms of treatment, adjuvant radiotherapy was administered to 33 patients (19.3%), neoadjuvant radiotherapy to 9 patients (5.3%), and chemotherapy to 20 patients (11.7%), while a significant majority (83.6%) underwent surgical procedures. Independent prognostic factors for DSS included age (HR = 1.039, P < 0.001), marital status (P = 0.029), history of previous tumors (HR = 0.257, P = 0.007), presence of metastatic disease (HR = 2.206, P = 0.027), and surgical treatment (HR = 0.490, P = 0.036). A nomogram prediction model was constructed to forecast 1-, 5-, and 10-year DSS rates, with a C-index of 0.739. Calibration plots demonstrated a strong correlation between the nomogram’s predictions and actual observations. Based on the prediction model, patients were stratified into three groups, and significant differences in prognosis were observed between these groups. Conclusion: A poorer prognosis is associated with retroperitoneal-derived MLPS/RCLPS than with other sites. The nomogram prediction model we built can be used to assist patients in consulting with their doctors and selecting patients for clinical trials. Relevance for Patients: Our study highlights the unique challenges and prognosis variations in retroperitoneal myxoid/RCLPS. The developed nomogram serves as a valuable tool for patients, aiding informed discussions with doctors and guiding decisions on treatment and clinical trial participation.