Deformity Apex, Thoracolumbar Alignment, but not Cobb Angle, Determine Disability in Adult Major Thoracic Scoliosis: A Multi-Center Radiographic and Health Status Analysis

Purpose: While classification systems and prognostic parameters based on radiographic parameters have been outlined for adolescent idiopathic scoliosis such information is lacking for adult scoliotic deformities. Recent studies have reported radiographic criteria with significant correlation to clinical symptoms in the setting of adult scoliosis. The purpose of this study was to analyze correlation between frontal plane Cobb angle, thoracic scoliosis apex level, end level and thoracolumbar sagittal alignment with outcomes measures. This may lay the groundwork to a clinically useful classification.

Methods: This is a multi-center prospective study including 301 consecutive patients. Subjects were drawn from the Spinal Deformity Study Group (SDSG) database of adult patients with scoliosis of the thoracic spine with an apex at any level from T4 down to T11 (of degenerative or idiopathic origin). From this group 67 patients had single major thoracic curves and form the basis of this study. For all subjects radiographic analysis (from full-length standing films) included: frontal plane Cobb angle, apical level and lower end level of the thoracic deformity, sagittal plane thoracolumbar alignment (T11-L2). Completed health assessment questionnaires were available for all subjects (Oswestry Disability Index, Scoliosis Research Society instrument (SRS- 29)). Subjects were divided into groups by apical level of the thoracic scoliosis and further subdivided by degree of thoracolumbar alignment. Statistical comparison (t-test) amongst subgroups in terms of ODI and SRS function/pain scores was made.

Results: For the 67 patients average age was 49 years (SD 16 years), distribution by apical level was as follows: T4-T7 (4), T8-T9 (42), T10-T11 (21). The mean frontal plane Cobb angle in the study group was 39 degrees (SD 22 degrees), no significant correlation with curve severity to any of the health measures was noted (r<0 .011). Significant differences in SRS pain scores between apex T4-T7 and T10-T11 patients were noted (ODI, function, p<0.02). Thoracolumbar loss of lordosis was significantly correlated with lower function score (>150 kyphosis vs. any lordosis, p=0.019). Lower end level revealed trends in terms of ODI and SRS instrument scores but this did not reach statistical significance.

Conclusion: This study is a unique multi-center analysis of single major thoracic scoliosis in adults. The risk factors reaching statistical significance for greater “disability” (ODI, SRS-29) were a lower curve apex and loss of thoracolumbar lordosis. Frontal plane Cobb angle revealed no correlation with self-assessed health. This quantitative analysis may lay the foundation for a clinically significant classification approach to adult thoracic scoliosis.