OBJECT: Non-small cell lung carcinomas with spinal and brachial plexus involvement have traditionally been considered to be Stage IIIb lesions and therefore unresectable. Advances in spinal surgery, the application of magnetic resonance (MR) imaging, and improvements in neoadjuvant therapy require a reassessment of the potential for complete resection. METHODS: The authors conducted a retrospective review of all procedures involving the resection of superior sulcus tumors with spinal or brachial plexus involvement performed between 1985 and 1999. Assessment or resectability and operative planning were based on an MR imaging classification scheme in which the extent of spinal involvement was considered. Class A tumors involved the periosteum of the vertebral body (VB) (16 patients); Class B, distal neural foramen without epidural compression (eight patients); Class C, proximal neural foramen with epidural compression (four patients); and Class D, bone involvement (VB or posterior elements) with or without epidural involvement (14 patients). Brachial plexus involvement was present in 21 patients, including 17 with T-1 nerve root only and four with C-8 or lower-trunk infiltration. Complete tumor resection was achieved in 27 patients and incomplete resection in 15. Complications occurred in 14 patients, two of which were related to instrumentation failures. The overall median survival was 1.44 years. The median survival for the complete and incomplete resection groups were 2.84 and 0.79 years, respectively (p = 0.0001). There was no statistical difference in survival among classification groups. CONCLUSIONS: Complete tumor resection of superior sulcus tumors is possible in selected patients in whom involvement of the spinal column and/or brachial plexus is present. Preoperative MR imaging is essential for evaluation of the spine and surgical planning. Survival and cure are dependent on complete resection, regardless of the extent of spinal involvement.
OBJECT: Non-small cell lung carcinomas with spinal and brachial plexus involvement have traditionally been considered to be Stage IIIb lesions and therefore unresectable. Advances in spinal surgery, the application of magnetic resonance (MR) imaging, and improvements in neoadjuvant therapy require a reassessment of the potential for complete resection. METHODS: The authors conducted a retrospective review of all procedures involving the resection of superior sulcus tumors with spinal or brachial plexus involvement performed between 1985 and 1999. Assessment or resectability and operative planning were based on an MR imaging classification scheme in which the extent of spinal involvement was considered. Class A tumors involved the periosteum of the vertebral body (VB) (16 patients); Class B, distal neural foramen without epidural compression (eight patients); Class C, proximal neural foramen with epidural compression (four patients); and Class D, bone involvement (VB or posterior elements) with or without epidural involvement (14 patients). Brachial plexus involvement was present in 21 patients, including 17 with T-1 nerve root only and four with C-8 or lower-trunk infiltration. Complete tumor resection was achieved in 27 patients and incomplete resection in 15. Complications occurred in 14 patients, two of which were related to instrumentation failures. The overall median survival was 1.44 years. The median survival for the complete and incomplete resection groups were 2.84 and 0.79 years, respectively (p = 0.0001). There was no statistical difference in survival among classification groups. CONCLUSIONS: Complete tumor resection of superior sulcus tumors is possible in selected patients in whom involvement of the spinal column and/or brachial plexus is present. Preoperative MR imaging is essential for evaluation of the spine and surgical planning. Survival and cure are dependent on complete resection, regardless of the extent of spinal involvement.