Injuries of the cervical spine in patients with ankylosing spondylitis: experience at two trauma centers

Abstract

Object: The cervical spine in a patient with ankylosing spondylitis (AS) (Bechterew disease) is exposed to maximal risk due to physical load. Even minor trauma can cause fractures because of the spine’s poor elasticity (so-called bamboo spine). The authors conducted a study to determine the characteristics of cervical fractures in patients with AS to describe the standard procedures in the treatment of this condition at two trauma centers and to discuss complications of and outcomes after treatment. Methods: Between 1990 and 2006, 37 patients were surgically treated at two institutions. All patients were examined preoperatively and when being discharged from the hospital for rehabilitation. Single-session (11 cases) and two-session anterior–posterior (13 cases), anterior (11 cases), posterior (two cases), and laminectomy (one case) procedures were performed. The injury pattern, segments involved, the pre- and postoperative neurological status, and complications were analyzed. Preoperative neurological deficits were present in 36 patients. All patients experienced improvement postoperatively, and there was no case of surgery-related neurological deterioration. In patients in whom treatment was delayed because of late diagnosis, preoperative neurological deficits were more severe and improvement worse than those treated earlier. The causes of three deaths were respiratory distress syndrome due to a rigid thorax and cerebral ischemia due to rupture of the vertebral arteries. There were 12 perioperative complications (32%), three infections, one deep venous thrombosis, five early implant failures, and the three aforementioned fatalities. There were no cases of epidural hematoma. In all five cases in which early implant failure required revision surgery, the initial stabilization procedure had been anterior only. A comparison of complications and the outcomes at the two centers revealed no significant differences. Conclusions: The standard intervention for these injuries is open reduction, anterior decompression and fusion, and anterior–posterior stabilization; these procedures may be conducted in one or two stages. Based on the early implant failures that occurred exclusively after single-session anterior stabilizations (five of 10—a failure rate of 50%), the authors have performed only posterior and anterior procedures since 1997 at both centers. Diagnostic investigations include computed tomography scanning or magnetic resonance imaging of the whole spine, because additional injuries are common. The causative trauma may be very slight, and diagnosis may be delayed because plain radiographs can be initially misinterpreted. In cases in which diagnosis is delayed, patients present with more severe neurological deficits, and postoperative improvement is less pronounced than that in patients in whom a prompt diagnosis is established. Because of postoperative pulmonary and ischemic complications, the mortality rate is high. In the present series the mortality rate was lower than the mean rate reported in the literature.