Introduction. Osteosarcoma is the most common sarcoma of bone. Pelvic osteosarcoma presents a significant therapeutic challenge due to potential late symptom onset, metastatic dissemination at diagnosis, and inherent difficulties of wide surgical resection secondary to the complex and critical anatomy of the pelvis. The rates of survival are well reported for osteosarcoma of the appendicular skeleton, but specific details regarding presentation and survival are less known for osteosarcoma of the pelvis. Methods. The Surveillance, Epidemiology, and End Results (SEER) program was queried for primary osteosarcoma of the bony pelvis from 2004 to 2015. Cases with Collaborative Staging variables (available after 2004) were analyzed by grade, histologic subtype, surgical intervention, tumor size, tumor extension, and presence of metastasis at diagnosis. The 2-, 5-, and 10-year survival rates were assessed with respect to these variables. The SEER database was then queried for age, tumor size, surgical intervention, metastasis at time of presentation, and survivorship data for patients with primary osteosarcoma of the upper extremity, lower extremity, vertebrae, thorax, and face/skull, and rates for all anatomic locations were then compared to patients with primary pelvic osteosarcoma. Results. A total of 292 cases of pelvic osteosarcoma were identified from 2004 to 2015 within the database, representing 9.8% of cases among all surveyed primary sites. The most common histologic subtype was osteoblastic osteosarcoma (69.9%), followed by chondroblastic osteosarcoma (22.3%). The majority of cases were high-grade tumors (94.3%), of size >8 cm (72.0%), and with extension beyond the originating bone (74.0%). For the entire pelvic osteosarcoma group, the 2-, 5-, 10-year survival rates were 45.6%, 26.5%, and 21.4%, respectively, which were the poorest among surveyed anatomic sites. The 5-year overall survival was an abysmal 5.3% for patients with metastatic disease at diagnosis, and 37.0% for non-metastatic pelvic osteosarcoma treated with surgery and chemotherapy. When compared to other locations, pelvic osteosarcoma had higher rates of metastatic disease at presentation (33.5%), larger median tumor size (11.0 cm), and older median age at diagnosis (47.5 years). While over 85% of patients with tumors at the extremities received surgery, only 47.4% of pelvic osteosarcomas in this cohort received surgical resection—likely influenced by larger tumor size, sacral involvement, frequency of metastasis, older age, or delayed referral to a sarcoma center. Conclusion. This study clarifies presenting features and clinical outcomes of pelvic osteosarcomas, which often present with large, high-grade tumors with extracompartmental extension, high likelihood of metastatic disease at diagnosis, and a potential limited ability to be addressed surgically. The survival rates of primary osteosarcoma of the pelvis are poor and are lower than osteosarcomas from other anatomic locations. While acknowledging the influence of metastasis, tumor characteristics, and advanced age on the decision to undergo surgical excision of a pelvic osteosarcoma, the rates of surgical resection are low and highlight the importance of understanding appropriate conditions for oncologic resection of pelvic sarcomas.
Analysis of tumor size enhances predictive power of CA19-9 for overall survival following surgical resection in pancreatic adenocarcinoma
