Survival differences in pulmonary metastatic osteosarcoma between pediatric and adult institutions.
Daniel R Liesman, William J Kacey, Kelly R Bates, Yangruijue Ma, Ashley C Dodd, Seth D Goldstein, Terrance D Peabody, Ruohui Chen, Timothy B Lautz
Abstract
Open AccessPURPOSE: Prognosis for osteosarcoma with pulmonary metastases remains poor at all ages. Pediatric surgeons often employ a more aggressive approach to surgical disease clearance compared to adult surgeons, but the survival impact is unclear. We hypothesized that patients with relapsed pulmonary metastatic osteosarcoma cared for at pediatric institutions will show improved survival compared to patients cared for at adult institutions reflecting differences in treatment approach. METHODS: A retrospective chart review was performed at affiliated free-standing adult and pediatric hospitals. Characteristics were recorded of adult and pediatric patients with pulmonary relapse of osteosarcoma following primary tumor treatment. Adults over 60-years-of-age were excluded to minimize the effect of medical comorbidities. Overall survival was analyzed with Kaplan-Meier estimates and univariable and multivariable Cox regression. RESULTS: There were 28 pediatric and 24 adult patients. A greater number of pediatric patients had extremity tumors (82 % versus 46 %; p-value = 0.006). Timing of relapse was similar between groups. Care at a pediatric institution was associated with improved overall survival on univariable analysis (Hazard Ratio for mortality 0.48 [0.25, 0.92]; p-value = 0.028). Extremity tumor location, favorable response to neoadjuvant chemotherapy, and greater number of thoracic surgeries were also associated with improved survival. However, on multivariable analysis adjusting for favorable initial response and primary tumor site, institution was not independently predictive of survival. CONCLUSIONS: Patients treated at pediatric institutions had improved survival following pulmonary relapse. While institution alone was not independently predictive after adjustment in this cohort, these findings support further multi-institutional studies to identify modifiable factors that may influence outcomes.