Preoperative Thoracolumbar Mobility (ΔTLK) Predicts Early Upper Instrumented Vertebra-Adjacent Fracture after Lower Thoracic-Pelvic Fusion in Elderly Patients with Adult Spinal Deformity.

INTRODUCTION

To determine whether preoperative thoracolumbar mobility predicts early upper instrumented vertebra-adjacent fracture (eUIVAF) after lower thoracic (LT)-pelvic fusion for adult spinal deformity (ASD) in elderly patients, in whom fracture-type proximal junctional failure is most prevalent.

METHODS

Seventy-four ASD patients aged ≥40 years who underwent corrective fusion spanning ≥6 vertebrae (2018-2024) by a single surgeon were analyzed. Sagittal parameters were measured on preoperative standing radiographs and supine computed tomographic (CT) images. ΔTLK and ΔTK were defined as angular changes in thoracolumbar (T10-L2) and thoracic (T5-T12) kyphosis between standing and supine positions. eUIVAF was defined as a new vertebral fracture at or adjacent to the UIV within 3 months. Given that all eUIVAFs occurred in patients aged ≥70 years who underwent LT-pelvic fusion (n=40), this elderly LT-pelvic subgroup constituted the primary analytic cohort for group comparisons, logistic regression, and receiver operating characteristic (ROC) analyses.

RESULTS

eUIVAF occurred in 14 of 74 patients (18.9%), exclusively among those aged ≥70 years with LT-pelvic fusion. In this subgroup, ΔTLK was greater in patients with eUIVAF than in those without (17.6°±8.0° vs 7.2°±7.0°, p<0.001), whereas ΔTK did not differ significantly. In a multivariable model including ΔTLK and ΔTK, only ΔTLK remained associated with eUIVAF (odds ratio, 1.27; 95% confidence interval, 1.09-1.47). ROC analysis for ΔTLK showed excellent discrimination (area under the curve, 0.90), with an optimal cutoff of 12°.

CONCLUSIONS

In elderly patients undergoing LT-pelvic fusion, preoperative ΔTLK was a strong radiographic predictor of eUIVAF. ΔTLK ≥12° identified a high-risk group, although this ROC-derived cutoff from a small single-center cohort without internal validation remains exploratory and requires confirmation in larger multicenter studies. Given that ΔTLK can be obtained from routine preoperative standing radiographs and supine CT, incorporating it into preoperative planning may help tailor fixation range, construct design, and correction magnitude in elderly ASD patients.