Oral Presentation 50th International Society for the Study of the Lumbar Spine Annual Meeting 2024

INTRODUCTION

Proximal junctional kyphosis (PJK) and failure (PJF) are complications affecting up to 39% of fusion surgery patients with adult spinal deformity [1] and leading to revision surgery in 2-17% of all cases [2-4]. Poor spinopelvic sagittal alignment is one of the important risk factors, potentially predictive of complications [5]. Sagittal imbalance was shown to be moderately-to-strongly associated with spinal loads [6], and elevated forces acting repetitively on the proximal segment could contribute to the development of mechanical complications in the junctional region. However, the link between proximal segment loads and PJK/PJF has not been explored yet. The hypothesis of this combined clinical and computational study is that early postoperative loads at adjacent segment, as related to sagittal alignment, are greater in patients who later develop PJK or PJF.

METHODS

A sample of N=70 patients with adult spinal deformity treated with T10-Pelvis fusion and followed-up for at least 2 years was retrospectively retrieved from a prospectively collected multi-centric database. Sagittal alignment (individual vertebrae C2-L5, sacrum and pelvis) was annotated in postoperative (<6 weeks) radiographs. Alignment data, body weight and height were used to modify a previously established validated generic musculoskeletal model of the spine [7] with fusion [8] to construct patient-specific spine models. Inverse-static simulations of upright standing posture were performed to predict postoperative loads at the adjacent segment, T9/T10.

RESULTS

Thirteen (19%) patients developed PJK/PJF within the course of the follow-up. There were no statistically significant differences between PJK and no-PJK groups in terms of age, body mass, body height, or sagittal parameters, except for Global Alignment and Proportion (GAP) score (7.7±3.5 vs. 4.8±3.3, p<.01, effect size Cohen’s d=0.88). Patients who developed PJK had higher simulation-predicted postoperative compressive forces at the T9/T10 adjacent segment (0.81 ± 0.21 vs. 0.54 ± 0.16, p <.001, d=1.58) and slightly lower anterior shear forces (-0.003 ± 0.58 vs. 0.030 ± 0.061, p <.05, d=0.56), normalized to body weight. Diagnostic performance of postoperative compressive forces in discriminating between future PJK and no-PJK, was very good (ROC-AUC=0.85; Accuracy=76%, Sensitivity=92%, Specificity=72%, for cut-off value NormCompr≥1.075) and better than or comparable to GAP score (ROC-AUC=0.73; Accuracy=74%, Sensitivity=69%, Specificity=75%, for cut-off value GAP≥7).

DISCUSSION

Biomechanical simulations predicted that patients who developed PJK or PJF at the follow-up, had elevated compressive forces and more posterior direction of load at the proximal segment shortly after surgery. This suggests postoperative loads, as related to spinal alignment, might be involved in the development of proximal junctional disorders. Daily exposure to elevated compressive loads might compromise the integrity of spinal tissues, leading to progressive formation of vertebral compressive fractures or dehydration and accelerated degeneration of the intervertebral disc. While the diagnostic potential of biomechanical analysis of postoperative loads is encouraging, the gained improvement compared to simpler alignment analysis methods, such as GAP score, must be balanced against the technical complexity. Future analyses should include other fusion scenarios than T10-Pelvis and accounting for mechanical properties of tissues in order to increase results generalizability and improve accuracy of PJK/PJF predictions.