Computed tomography-derived assessment of respiratory impairment in thoracic scoliosis: Comparison between idiopathic and non-idiopathic etiologies

BACKGROUND Thoracic scoliosis causes complex three-dimensional deformities of the thoracic cage that can impair lung mechanics and airway geometry, leading to restrictive ventilatory dysfunction. Respiratory impairment differs by etiology, with non-idiopathic scoliosis often showing more severe and persistent deformity. Conventional pulmonary function testing provides global functional assessment but is frequently limited in patients with non-idiopathic scoliosis. Advances in chest computed tomography (CT) allow quantitative evaluation of lung volume and airway morphology, offering an alternative approach to assess respiratory function and the relationship to spinal deformity. AIM To investigate CT-derived airway morphology and respiratory function in thoracic scoliosis, focusing on etiologic differences and structure-function relationships. METHODS This retrospective observational study included 53 patients ≤ 30 years of age who underwent corrective surgery for thoracic scoliosis and had preoperative and postoperative CT examinations. Patients were classified into idiopathic (n = 25) or non-idiopathic scoliosis (n = 28). Spinal deformity parameters, lung volumes, predicted total lung capacity percentage (TLC%), and airway dimensions were quantified using CT. Group comparisons were performed using the Wilcoxon rank-sum test, correlations were determined using Spearman’s rank correlation, and pre- and postoperative changes were determined using the Wilcoxon signed-rank test. RESULTS Patients with non-idiopathic scoliosis had significantly more severe spinal deformities and a lower TLC% compared to patients with idiopathic scoliosis (median TLC%: 39% vs 64%, P < 0.001). The Cobb angle was negatively correlated with TLC% (ρ = -0.52, P = 0.004), lung volumes, and multiple airway parameters in patients with non-idiopathic scoliosis. Surgical correction significantly improved spinal alignment in both groups. CT-derived lung volume parameters did not show significant postoperative changes. In contrast, selective improvement was observed in the left bronchus. CONCLUSION CT-derived analysis showed differences in respiratory impairment between idiopathic and non-idiopathic scoliosis. Surgical correction improved alignment without immediate lung volume improvement, with selective changes in the left bronchus.