Surgical outcomes of adolescent idiopathic scoliosis using fulcrum bending radiographs to determine curve rigidity

Abstract

<h3>Background</h3><p>Spinal curve flexibility can be assessed reliably by fulcrum bending (FB) radiograph for surgical planning. However, it remains unknown how curve correctability of rigid curves are different from flexible curves. This study aims to investigate the relationship of curve flexibility and surgical outcomes in adolescent idiopathic scoliosis, to investigate for any differences of surgical outcomes in rigid and non-rigid curves, and to explore whether postoperative changes were different in rigid curves.</p><h3>Methods</h3><p>This retrospective study included patients diagnosed with AIS, aged ≤18 years and underwent posterior spinal fusion with >2 years of postoperative follow-up at two affiliated hospitals between 2010 and 2022. Radiographic parameters were collected, including coronal Cobb angle of major curve, curve type, FB flexibility, fusion length, and implant density. Relationship of flexibility and correction rate was tested. Rigid curves (FB flexibility <50 %) were compared with non-rigid curves, in terms of curve correction rate, FB correction index (FBCI) and loss of correction at postoperative 2 years.</p><h3>Results</h3><p>A total of 190 patients (83.7 % females) were included. Preoperative major curve was 60.2° (SD 10.9°) with 64.4 % (SD 17.2 %) flexibility, 21.6 % (n = 41) of the patients had rigid curves. FB flexibility correlated with curve correction rate (immediate postoperative <em>r</em>_<em>s</em>: 0.281, <em>p</em> < 0.001). After covariate adjustment, correction rate for rigid and non-rigid curves were 69.4 % and 75.5 %, with mean difference −6.0 % (95 %CI −10.6 % to −1.4 %, <em>p</em> = 0.013). Rigid curves had higher FBCI (175.5 % vs 110.4 %; mean difference: 65.1 %, 95 %CI 53.2 % to 77.0 %, p < 0.001). Fewer rigid curves demonstrated loss of correction, with comparable changes in correction rates (−2.5 % vs −4.6 %, <em>p</em> = 0.124).</p><h3>Conclusion</h3><p>Spinal curve flexibility significantly correlated with curve correction rate and FBCI. Rigid curves demonstrated ∼70 % correction rate, all with FBCI >100 %, indicating curve correction achieved more than estimated preoperatively. Despite rigid curves had lower curve correction rate, correction lost at 2-year follow-up was comparable between rigid and non-rigid curves.</p>