RISK FACTORS AND MANAGEMENT OF PSEUDOARTHROSIS FOLLOWING POSTERIOR SEGMENTAL SPINAL INSTRUMENTATION FOR IDIOPATHIC SCOLIOSIS

Daniel J. Sucato, MD, MS,
Bryan Kaiser, MD
Texas Scottish Rite Hospital, Dallas, TX, USA

INTRODUCTION:
Segmental posterior spinal instrumentation provides improved 3–dimensional correction and more stable fixation than previous systems. Although these attributes should reduce the incidence of pseudoarthrosis, this has not been previously studied. The objectives of this study were: to evaluate the incidence of pseudoarthrosis following segmental spinal instrumentation for adolescent idiopathic scoliosis (AIS), identify construct patterns predisposing to pseudoarthrosis and evaluate the management of this complication.

METHODS:
A retrospective review was performed of 965 patients who had posterior spinal instrumentation and fusion for AIS between 1986 and 1997. Of these, those who had surgically–documented pseudoarthroses were included for analysis. Radiographs were reviewed to determine loss of correction in the coronal and sagittal planes, hardware failure and hardware patterns present at each pseudoarthrosis level.

RESULTS:
The initial instrumentation used was TSRH in 10 patients and CD in 6 with supplemental iliac crest graft (ICG) in all but one patient in which rib graft was used. Patients with pseudoarthrosis had pain as the presenting symptom at an average of 22 months after the initial surgery. There were 16 patients (1.7%) who had 22 levels (1.4 /patient) in which a pseudoarthrosis was present at the time of revision surgery. At the time of revision, loss of correction in the coronal plane of the thoracic and lumbar curves was 14.8 and 7.1 degrees, respectively with an average change in the trunk shift of 8.1 mm to the right. In the sagittal plane, thoracic kyphosis increased 10.4 degrees and an increase in negative sagittal balance averaged 13 mm. 8 of 22 pseudoarthroses occurred between T10 and L2. At each pseudoarthrosis, an average of 1.6 hardware segments (hook(s) or crosslink) were present, with 11 crosslinks being the most common. There were 2 acute infections and 3 late infections. Hardware failure consisted of 2 rod fractures, and 5 loose hooks. The time to reoperation averaged 29.8 months (range 9–60) and included ultimate reinstrumentation in 10 patients supplemented with ICG in 5, rib in 2 and local graft from the fusion mass in 9. Two patients who had not been reinstrumented required reoperation with instrumentation to obtain complete healing of the pseudoarthrosis.

CONCLUSIONS:
Pseudoarthrosis following segmental spinal instrumentation for AIS is relatively rare and is best treated with compression instrumentation and bone grafting. The development of a pseudoarthrosis is greatest at the thoracolumbar junction, when infection is present, and with "overcrowding" of a single motion segment with hardware. Careful planning of hook and crosslink placement is necessary, together with meticulous decortication and abundant bone grafting at these levels.