A Prospective Analysis of Spontaneous Changes of the Vertebral Rotation of Secondary Curves After Selective Anterior Fusion of Idiopathic Scoliosis

Objective: Prospective digital radiometric and rasterstereographic rotation analysis of idiopathic thoracic or thoracolumbar/lumbar scoliosis before and after selective anterior correction and fusion.

Methods: Fifteen patients with idiopathic thoracic scoliosis (group 1) and 27 patients with idiopathic thoracolumbar/lumbar scoliosis (group 2) were prospectively evaluated. All patients underwent selective anterior dual rod instrumentation of the thoracic or the thoracolumbar/lumbar curve. Segmental rotation was analyzed by means of digitized radiographs according to method of Drerup and by rasterstereographic threedimensional back-surface analysis preoperatively and on average 1 year postoperatively. Curve corrections were measured in Cobb angle.

Results: In group 1 the thoracic Cobb angle was corrected from 66° preoperatively to 27° postoperatively with a spontaneous lumbar curve correction from 48° to 24°. Using digital radiometric vertebral rotation measurement an average spontaneous derotation of the lumbar spine of 14% (p=0.036) was found in group 1 as well as an average derotation of 48% (p=0.001) of the instrumented thoracic curve. In the rasterstereography the lumbar hump was spontaneously corrected by on average 55% (p=0.001) with an average correction of the rib hump of 46% (p=0.001).

In group 2 the lumbar Cobb angle was corrected from 54° preoperatively to 19° postoperatively with a spontaneous thoracic curve correction from 36° to 26°. Using digital radiometric rotation measurement the patients in group 2 showed a spontaneous increase of rotation in the thoracic spine by on average 37% (p=0.003). Average derotation in the instrumented lumbar curves was 61 % (p<0 .001) on average. By rasterstereography an average increase of the rib hump 39% (p="0.003)" was found. There correction lumbar 68% (p<0.001).

Discussion: Derotation of the instrumented thoracic curves was less than that of the instrumented lumbar curves. A spontaneous radiometric derotation of 14% was found in the not instrumented secondary lumbar curves. However, both digital radiometric rotation measurement as well as rasterstereographic back shape analysis demonstrated a spontaneous increase of malrotation in the not instrumented thoracic spine.

The results show that in contrast to anterior fusion of thoracic curves the cosmetic outcome after anterior thoracolumbar/lumbar fusion may be limited due to insufficient correction and even deterioration of the thoracic hump. Patients with primary lumbar/thoracolumbar curves and a preoperatively prominent rib hump should be informed about the risk of cosmetic deterioration after anterior selective fusion.