The Dwyer procedure in the treatment of idiopathic scoliosis. A 10-year follow-up review of 21 patients.

Despite the well-known Dwyer procedure, developed in 1969, comprehensive reports on its use, with long-term follow-up, are relatively scarce. The purposes of this study were to detect eventual late complications and to compare late results with postoperative angular curve correction. This article reports on 21 children operated on between October 1972 and October 1975 and reviewed with a minimum follow-up of 10 years (10 other patients were lost to follow-up after 5 years). Patients had idiopathic lumbar or thoracolumbar curves (average curve, 56 degrees). Results are discussed with a special reference to longitudinal observation. There is a great correction of the instrumented curve (postoperative, 5 degrees), but a loss of correction of 10 degrees is generally observed, prevented by a complete immediate correction or even hypercorrection. The upper curve, noninstrumented, also shows improvement (mean preoperative, 38 degrees; postoperative, 22 degrees; 10 years, 22 degrees) but re-equilibration cannot be predicted. Pseudarthrosis of one intervertebral space occurs frequently, and may cause failure of the cable with a loss of correction of 10-20 degrees. Kyphosis (or simple loss of lumbar lordosis) is commonly observed but should be balanced with correction of rotation. The following conclusions were made: morbidity is not severe, despite the advanced surgical technique. The technique is difficult and has a direct consequence on the quality of results; pseudarthrosis is a frequent complication, followed by important loss of correction; indications should be discussed carefully in idiopathic lumbar and thoracolumbar curves. It is still too early to advocate either anterior instrumentation of Zielke (VDS) or segmental posterior instrumentation (C. D. Luque) because of short-term follow-up.