Stabilisation of fractured thoracic and lumbar spine with Cotrel-Dubousset instrument

Journal of Orthopaedic Surgery, Jun 2003 by Moon, M S, Choi, W T, Moon, Y W, Kim, Y S, Moon, J L

ABSTRACT

Objective. To evaluate the effectiveness of 2 different types of Cotrel-Dubousset instrument systems in stabilising thoracolumbar and lumbar spine fractures.

Methods. Between January 1989 and December 1993, 45 fractures in 42 patients with unstable fracture or fracture dislocation of the thoracolumbar and lumbar spines were randomly assigned to 2 surgical treatments with Cotrel-Dubousset instrumentation-using either a long segment (Group 1) or a short segment (Group 2)-and short posterolateral fusion.

Results. Consolidation of the fractured vertebral body and posterolateral fusion were achieved at a mean time of 4.5 months; fusion rates were 75% in Group 1 and 83% in Group 2. The average collapses of anterior vertebral body height in Group 1, in the immediate postoperative period and at the final follow-up, were 15% and 17%, respectively; and in Group 2, the figures were 16% and 24%, respectively. The correction of vertebral height and kyphosis at the last follow-up were lost more in Group 2 (5.7[degrees]) than in Group 1 (4.4[degrees]). There were neurological recoveries in 6 of the 9 cases of incomplete paraplegics, including complete recovery in 5, and one-Frankel grade increase in one. There were 15 instrument failures in 12 patients, including screw breakage in 3 Group 1 cases and 6 Group 2 cases. The plug dislodged in 3 Group 1 cases, and the hook dislodged in 3 Group 2 cases. In other words, instrument failures were more common in Group 2.

Conclusion. Cotrel-Dubousset stabilisation of the fractured spine achieves fracture consolidation, but does not maintain the restored height and sagittal curve completely until fusion. The long rod and short fusion construct was more effective for all fracture types than was the short rod and fusion construct, although it leads to wider immobilisation of normal segments.

Key words: Cotrel-Dubousset; spinal fracture; spinal fusion

INTRODUCTION

The goals of surgical management of unstable fractures of the spine include attainment of the normal spinal anatomy, as well as maintenance of reduction, decompression of the neurological structures, and early mobilisation.1-3 Although immediate reduction and stabilisation of a fractured spine by using hook or pedicular screw instrumentation can be achieved postoperatively, loss of reduction can occur during follow-up. Therefore, maintenance of the reduced fracture and restored spinal curvature has been a main concern even after instrumented stabilisation.1,2,4-6 Some surgeons have stressed the application of the load-sharing principle in the surgical treatment of the fractured spine on the basis of numerical instability indices.7

In this study, we evaluated the 2 types of Cotrel-Dubousset instrumentation constructs for thora-columbar and lumbar spine fractures.

MATERIALS AND METHODS

We studied 45 fractures in 42 patients who were treated at the Catholic University Affiliated Hospitals, Seoul, Korea, between January 1989 and December 1993 for unstable fracture or fracture dislocation of the thoracolumbar and lumbar spines. In all, there were 33 men and 9 women, with a mean age of 34.5 years (range, 18-57 years). All patients were treated with Cotrel-Dubousset instrumentation systems and posterolateral fusion, and they were followed up for 32 to 72 months. The causes of injury were as follows: falls from a height (n=30), slip injuries on stairs (n=3), hang-glider crashes (n=3), traffic injuries (n=3), and industrial injuries (n=3); 21 patients had associated fractures in the calcaneus (n=12), tibia (n=6), and pelvis (n=3).

Surgeries were performed, on average, 9.9 days after injury (range, 1-24 days). In all patients without fracture dislocation, the fractures were reduced by the contoured rods to apply the lordotic distraction principle, a deformity correction method by rotating the contoured rods into opposite direction, without primary vertical distraction or compression force. In Group 1 cases, reduction involved the 2 vertebrae above the fracture site (compression hook attached to the second upper vertebra and pedicle screw attached to the first upper vertebra) and the first lower vertebra (combined pedicle screw and compression hook) [Fig. I]. In Group 2 cases, the vertebrae immediately above and below the fractured vertebra were fixed posteriorly with pedicle screws, and were fused posterolaterally with autogenous iliac bone graft (Fig. 2). Anterior surgery was not performed to stabilise the anterior column additionally in any of the patients.

The anterior vertebral height and local kyphotic angle were measured radiographically before and after surgery, and at the last follow-up. Degree of bone loss in patients older than 55 years was assessed preoperatively on plain X-rays, by using Seville's method8 to rule out the poor surgical indication. The osteoporosis index9 was Grade I in most cases, and did not correlate with implant failure (Table 1). In addition, computed tomography was performed for all patients to classify the fracture type, and to assess the neural canal encroachment by the retropulsed middle column fragment.