A 20-year retrospective analysis of CT-based pre-operative identification of pulmonary metastases in patients with osteosarcoma: A single-center review

Purpose Cooperative studies support complete metastasectomy in osteosarcoma (OS). Pre-operative CT is used to identify and quantify metastases and can facilitate minimally invasive techniques. Here we assess the accuracy of pre-operative CT compared to findings at thoracotomy and its change over time. Methods We reviewed OS thoracotomies performed at our institution from 1996 to 2015. The number of metastases identified on pre-operative chest CT was compared to the number of metastases seen on pathology (both metastases with viable cells and non-viable, osteoid-only metastases). Results Eighty-eight patients underwent 161 thoracotomies with a median of 14 days (range, 1–85) between CT and surgery, a median of 2 CT-identified lesions (range, 0–15), and a median of 4 resected lesions (range, 1–25). In 56 (34.8%) cases, more metastases were found surgically than were seen on CT, and among these, 34 (21.1%) had a greater number of viable metastases. There was poor overall correlation between CT and pathology findings (Kendall Tau-b = 0.506), regardless of CT slice thickness, decade of thoracotomy, or total number of CT-identified lesions. Conclusions CT accuracy in pre-operatively quantifying OS pulmonary metastases has not improved in recent decades. Consequently, we recommend an open technique with direct lung palpation for complete identification and resection of OS pulmonary metastases. Level of evidence Level IV, retrospective study with no comparison group.