Patterns of chemotherapy administration in high-risk soft tissue sarcoma and impact on overall survival

Background: Conflicting data exist on the benefit of chemotherapy in the management of high-risk soft tissue sarcoma (STS). Use of chemotherapy may be dependent on patient, tumor, and facility characteristics. This study sought to identify these factors and compare survival between treatment groups. Patients and Methods: Patients with stage III STS were identified from the National Cancer Data Base (NCDB) from 1998 to 2012. Multiple logistic regression analysis was used to determine factors that influenced the probability of receiving chemotherapy. In a subset of patients, we determined the relationship between chemotherapy use and overall survival, using Kaplan-Meier curves and Cox regression analysis with propensity score adjustment. We also examined the effect of chemotherapy by histologic subgroup using interaction models. Results: A total of 16,370 patients were included (N=5,377 for survival analysis). Patients who were younger than 40 years; male; privately insured; earned a higher income; had no comorbidities; had synovial sarcoma, angiosarcoma or "other" histology; and whose tumors were high-grade, greater than 10 cm, or from the lower extremity were significantly more likely to receive chemotherapy. Median unadjusted overall survival (OS) in the nonchemotherapy and chemotherapy groups was 51.3 and 82.7 months, respectively (P<.001). On adjusted analysis, the survival benefit remained significant (hazard ratio [HR], 0.85; P=.004). The benefit was particularly strong in the undifferentiated pleomorphic sarcoma (UPS) group on adjustment, with a median OS of 49.1 and 77.8 months for nonchemotherapy versus chemotherapy, respectively (HR, 0.77; P=.02). Conclusions: In addition to expected tumor and patient factors, histology, location of primary tumor, and socioeconomic status are associated with receipt/nonreceipt of chemotherapy in stage III STS. Chemotherapy use was associated with improved OS in the overall population, and specifically in the UPS subgroup.