Patterns of Extrapulmonary Metastases in Sarcoma Surveillance

SIMPLE SUMMARY: Soft tissue sarcomas are cancerous tumors that can cause death by spreading throughout the body (metastasizing), most commonly to the lungs. After initial removal of the tumor, patients undergo periodic chest imaging for 10 years to monitor for metastases, but experts disagree about whether the abdomen/pelvis should be imaged as well. Our goal was to determine when, where, and how frequently metastases occur in the abdomen/pelvis; this would help answer the question of whether these areas should be imaged as part of surveillance. We found that 7% of patients treated for soft tissue sarcoma developed abdomen/pelvis metastases. Compared to lung metastases, abdomen/pelvis metastases appeared later and over a broader period of time. We were not able to predict which patients are most likely to develop abdomen/pelvis metastases. Understanding these patterns will help experts develop surveillance guidelines that will identify metastatic cancer without subjecting patients to more scans than necessary. ABSTRACT: Soft tissue sarcomas (STS) most commonly metastasize to the lungs. Current surveillance guidelines variably recommend abdominal and pelvic imaging, but there is little evidence to support this. We sought to determine the proportion of initial pulmonary versus extrapulmonary metastases, the time to development of each, and factors to identify patients that would benefit from abdominopelvic surveillance. We retrospectively reviewed 382 patients who underwent surgical treatment for STS at a single institution. Of the 33% (126/382) of patients who developed metastases, 72% (90/126) were pulmonary, 22% (28/126) were extrapulmonary, and 6% (8/126) developed both simultaneously. Initial extrapulmonary metastases occurred later (log rank p = 0.049), with median 11 months (IQR, 5 to 19) until pulmonary disease and 22 months (IQR, 6 to 45) until extrapulmonary disease. Pulmonary metastases were more common in patients with high grade tumors (p = 0.0201) and larger tumors (p < 0.0001). Our multivariate analysis did not identify any factors associated with initial extrapulmonary metastases. A substantial minority of initial metastases were extrapulmonary; these occurred later and over a broader time range than initial pulmonary metastases. Moreover, extrapulmonary metastases are more difficult to predict than pulmonary metastases, adding to the challenge of creating targeted surveillance protocols.