Session Item

Patients with large, high-grade extremity and trunk (ET) soft-tissue sarcomas (STS) are at high risk for distant recurrence and death. The integration of chemotherapy (C) to standard of care neoadjuvant radiotherapy (RT) remains controversial even for these patients due to conflicting trial results and toxicity concerns. This study examines the impact of adding C to neoadjuvant RT on overall survival (OS) in high risk ET-STS.

The National Cancer Data Base (NCDB) was queried for patients ≥18 years with high risk (≥5 cm + FNCLCC grade 2-3) ET-STS (histology per NRG DT001) who received neoadjuvant RT and limb sparing surgery from 2006-2014. Patients were stratified based upon receipt of C (RT and CRT cohorts).

Overall survival (OS) for RT vs CRT cohorts was analyzed using the Kaplan-Meier (KM) method, log-rank test & Cox proportional hazards models. Propensity score-matched analysis (PSM) was employed to account for potential treatment selection bias between cohorts.

Mean inpatient stay, 30-day unplanned readmission rate and 90-day mortality were calculated as surrogates for toxicity.

A total of 884 patients were identified: 639 (72.3%) in the RT cohort & 245 (27.7%) in the CRT cohort.

Patient cohorts were well-balanced in key socioeconomic (race, sex, insurance, income & education) & clinical characteristics (tumor site, grade, size & surgical margin status). Patients in the CRT arm were more likely to be younger (≤50) (44.9% vs 20.8%), have a Charlson-Deyo [CD] score of 0 (85.7% vs 79.3%) & have synovial sarcoma histology (19.6% vs 2.8%).

The unadjusted 5-year KM OS was significantly higher in the CRT vs RT cohort: 72.0% vs 56.1% on univariate (p<0.0001) & multivariate analysis (Hazard Ratio [HR]: 0.57; 95% Confidence Interval [CI]: 0.41-0.78; p<0.001) even after adjusting for age, race, income, CD score, histology, tumor size, tumor grade & primary site (lower extremity; upper extremity; trunk).

PSM identified evenly matched cohorts of 212 patients each with respect to age, income, CD score, histology, grade, tumor size & primary site. The adjusted 5-year KM OS was significantly higher in the CRT vs RT cohort: 69.8% vs 55.4%. The addition of neoadjuvant C remained prognostic for OS on PSM analysis (HR: 0.56 [0.39-0.83], p=0.003).

There were no statistical differences in the mean surgical inpatient stay (3.8 vs 4.3 days), 30-day unplanned readmission rate (5.0% each), or 90-day mortality (2.0% vs 1.2%) between the RT vs CRT cohorts.

The addition of chemotherapy to neoadjuvant RT was associated with improved OS in patients with high risk ET-STS in the NCDB. Additionally, toxicity surrogates were similar between RT vs CRT cohorts. These findings warrant further investigation in a prospective randomized trial.