Prediction of Proximal Junctional Kyphosis and Failure after Corrective Surgery for Adult Spine Deformity: an MRI-Based Model Combining Bone and Paraspinal Muscle Quality Metrics

Zhen Tian,Jie Li,Hui Xu,Yanjie Xu,Zezhang Zhu,Yong Qiu,Zhen Liu
DOI: https://doi.org/10.1016/j.spinee.2024.08.010
IF: 4.297
2024-01-01
The Spine Journal
Abstract:Background Proximal junctional kyphosis (PJK) and proximal junctional failure (PJF) are common complications observed after adult spinal deformity (ASD) surgery and major cause for unplanned reoperations. In addition to spinal alignment, osteoporosis and paraspinal muscle (PSM) degeneration are reportedly indispensable factors that account for PJK/PJF. Purpose To investigate the utility of the preoperative risk assessment model using MRI-based skeletomuscular metrics in predicting PJK and/or PJF(PJK/PJF) after ASD correction. Study Design Retrospective Case-Control study. Patient Sample Consecutive series of 149 patients at a single academic institution. Outcome Measures MRI based measurements of vertebral bone quality at upper instrumented vertebra (VBQ-U) score and fat infiltration rate (FI%) of paraspinal muscle (PSM). Methods We performed a retrospective analysis of patients with ASD who underwent ≥5-segment fusion. The vertebral bone quality (VBQ) scoring system was used to assess the bone quality. The PSM quality including FI% and cross-sectional area (CSA) were evaluated. Multivariate logistic regression was performed to determine potential risk factors of PJK/PJF. Results Of 149 patients who underwent ASD surgery, PJK/PJF was found in 45(30.2%). Mean VBQ-U scores were 3.45 ± 0.64 and 3.00 ± 0.56 for patients with and without PJK/PJF (P<0.001). Mean FI% of PSM(L3/L4) was 27.9 ± 12.8 and 20.7 ± 13.3 for patients with and without PJK/PJF (P<0.001). On multivariate analysis, the VBQ-U score and FI% of PSM were significant independent predictors of PJK/PJF. The AUC for the novel risk assessment model is 0.806, with a predictive accuracy of 86.7%. Conclusion In patients undergoing ASD correction, paraspinal muscle and vertebral bone quality significantly outweigh radiographic alignment parameters in predicting PJK/PJF. The MRI-based risk assessment model offers a valuable tool for early assessing individualized risk information for PJK/PJF.
What problem does this paper attempt to address?