Comparison of anterior and posterior instrumentation for correction of adolescent thoracic idiopathic scoliosis

Study Design. This was a prospective study of two cohort groups of patients (one group receiving anterior instrumentation and the other posterior inst rumentation) receiving treatment for thoracic idiopathic scoliosis. Objective. To present the 2-year postoperative results of a prospective mul ticenter study comparing the use of anterior instrumentation with that of p osterior multisegmented hook instrumentation for the correction of adolesce nt thoracic idiopathic scoliosis. Summary of Background Data. Despite reports of satisfactory results, proble ms have been reported with posterior system, including worsening of the lum bar curve after surgery and failure to correct hypokyphosis. Theoretically, the advantages of anterior instrumentation include prevention of lumbar cu rve decompensation by shortening the convexity of the thoracic curve. In ad dition, by removing the disc, better correction of thoracic hypokyphosis co uld be obtained. Methods. Seventy-eight patients who underwent an anterior spinal fusion usi ng flexible threaded rods and nuts (Harms-MOSS instrumentation, De Puy-Mote ch-Acromed, Cleveland, OH) were analyzed and compared with 100 patients who underwent posterior spinal fusion with multisegmented hook systems. Parame ters of comparison included coronal and sagittal correction, balance, dista l lumbar fusion levels, and complications. All patients had idiopathic thor acic curves of King Types II to V. The average age at surgery was 14 years in each group, the average preoperative curve 57 degrees, and the minimum d uration of follow-up for all patients 24 months. All data were collected pr ospectively and analyzed via Epi Info statistical analysis (Centers for Dis ease Control, Atlanta, GA). Results. Average coronal correction of the main thoracic curve was 58% in t he anterior group and 59% in the posterior group (P = 0.92). Analysis of sa gittal contour showed that the posterior systems failed to correct a preope rative hypokyphosis (sagittal T5 to T12 less than 20 degrees) in 60% of cas es, whereas 81% were normal postoperatively in the anterior group. However, hyperkyphosis (sagittal T5 to T12 greater than 40 degrees) occurred after surgery in 40% of the anterior group when the preoperative kyphosis was gre ater than 20 degrees. Postoperative coronal balance was equal in both group s. An average of 2.5 (range, 0-6) distal fusion levels were saved using the anterior spinal instrumentation according to the criteria used to determin ing posterior fusion levels in this study. Selective fusion of the thoracic curve (distal fusion level T11, T12, L1 was performed in 76 of 78 patients (97%) in the anterior group as compared with only 18 of 100 (18%) in the p osterior group. surgically confirmed pseudarthrosis occurred in 4 of 78 pat ients (5%) in the anterior group and in 1 of 100 patients (1%) in the poste rior group (P = 0.10). Loss of correction greater than 10 degrees occurred in 18 of 78 patients (23%) in the anterior group and in 12 of 100 patients (12%) in the posterior group (P = 0.01). Implant breakage occurred in 24 pa tients (31%) of the anterior group and in only 1 patients (1%) of the poste rior group. Conclusions. 1) Coronal correction and balance were equal in both the anter ior and posterior groups, even though the anterior group had the majority o f curves (97%) fused short or to L1, whereas only 18% were fused short or t o L1 in the posterior group. 2) In the anterior group there was a better co rrection of sagittal profile in those with a preoperative hypokyphosis less than 20 degrees. However, hyperkyphosis (with a mean of 54 degrees) occurr ed in 40% of those in the anterior group with a preoperative kyphosis of mo re than 20 degrees. 3) An average of 2.5 lumbar levels can be saved with an terior fusion and instrumentation according to the criteria used for choosi ng posterior fusion levels in this study. 4) Using the 3.2-mm flexible rod in this study, loss of correction, pseudarthrosis, and rob breakage were un acceptably higher in the anterior group than in the posterior group. Struct ural anterior support and a stronger implant are needed.