PURPOSE OF THE STUDY: The anterior approach to the thoracic and lumbar spine is used with increasing frequency for various indications. With the advent of prosthetic intervertebral disc replacement, its use has become even more frequent and has often been associated with serious complications. The aim of this study was to evaluate vascular complications in patients who underwent anterior spinal surgery of the thoracic and lumbar spine. MATERIAL: We performed a total of 531 operations of the thoracolumbar spine from the anterior approach. In 12 cases, after exposure of the body of the first or second thoracic vertebrae, we employed the Smith-Robinson technique to expose the cervical spine. We used sternotomy in six, posterolateral thoracotomy in 209, the pararectal retroperitoneal approach in 239, anterolateral lumbotomy in 58 and the transperitoneal approach in seven patients. The aim of surgery was somatectomy in 190 patients and discectomy in 341 patients. Sternotomy and transperitoneal approaches were carried out by a thoracic or vascular surgeon and all the other procedures were done by the first author. The indications for spinal surgery included an accident in 171, tumor in 56, spondylodiscitis in 43 and a degenerative disease in 261 patients. METHODS: All patients indicated for anterior spinal surgery were examined by conventional radiography in two projections, and this was completed by CT sagittal and frontal reconstructions of the affected region. Most patients also underwent MR imaging. The Smith-Robinson approach was used for exposure of T1 or T2. Sternotomy was indicated for treatment of T2-T4 and also T1 in the patients with a short, thick neck. Access to T3-L1 was gained by posterolateral thoracotomy, in most cases performed as a minimally invasive transpleural procedure. For access to the lumbar spine we usually used the retropleural approach from a pararectal incision or lumbotomy. We preferred the pararectal retroperitoneal approach in L2-S1 degenerative disease, L5 fractures, and L5-S1 spondylodiscitis. We carried out lumbotomy in patients with trauma, tumors and L1-L4 spondylodiscitis. The transperitoneal approach from lower middle laparotomy was used only in tumors at L5 or L4. For treatment of trauma and degenerative disease of the lumbar spine we preferred less invasive procedures, and for tumors and spondylodiscitis we used more extensive exposure because of the difficult terrain. The patients were followed up for 2 to 96 months (average, 31.4 months) after anterior spinal surgery. RESULTS: In 12 patients treated by the Smith-Robinson procedure and in six patients undergoing sternotomy, neither early nor late signs of any injury to major blood vessels or internal organs were recorded. The 209 patients with posterolateral thoracotomy were also free from any signs of vascular injury, but trauma to the thoracic duct was recorded in one case. We found injury to major blood vessels in three patients in the group treated by the pararectal retroperitoneal procedure. In the total of 531 anterior spinal surgery procedures this accounts for 0.56 %; of the 304 lumbar operations and 239 pararectal retroperitoneal operations it is 0.99 % and 1.26 %, respectively. In one patient the vascular injury was associated with trauma to the ureter. DISCUSSION: In our group of 531 patients we found a higher risk of vascular injury when the L4-L5 segment was treated, when less invasive surgery was used or when spinal anatomy was altered due to tumor or spondylodiscitis. All the complications were recorded in the first 250 patients. It should be emphasized that, because in five patients, the planned anterior spondylodesis would have been associated with high risk due to altered anatomy of the bifurcation of the aorta, these patients were treated by dorsal instrumented spondylodesis. We also avoided the anterior approach for revision spinal surgery and used the posterior approach instead. Vascular complications were treated in cooperation with a vascular or cardiac surgeon. In the most serious case, if a sophisticated cardiosurgical technique had not been immediately available, the patient would probably have died. CONCLUSIONS: The technique of anterior approach is safe only in the hands of experienced spinal surgeons with long experience. In institutions where anterior spinal surgery is not a routine method it is advisable to involve a vascular or cardiac surgeon. However, the most important point is to know when not to operate.