Correlation between multifidus muscle and proximal junctional kyphosis after long-segment instrumentation for lumbar degenerative disease

Background: The purpose of this study was to use magnetic resonance imaging (MRI)-based assessments of multifidus muscle to predict development of proximal junctional kyphosis (PJK). Methods: We retrospectively reviewed 62 patients with lumbar degenerative disease (LDD) who underwent fusion of L5 and all the upper instrumented vertebrae (UIVs) in the thoracolumbar spine (T9-L2), with ≥2 years’ follow-up, at a single center between 2008 and 2014. All patients underwent standing radiographic imaging preoperatively, immediately postoperatively, and at final follow-up, as well as preoperative supine MRI. The patients were divided into PJK and non-PJK groups and subdivided into four groups by the UIVs (T9-T12 vs L1-L2). The demographic data included age, sex and body mass index. The MRI-based multifidus assessment was performed using ImageJ version 1.46. Subsequently, the relative functional cross-sectional area (rFCSA) was calculated. Sagittal parameters were measured within the whole spine, using Surgimap version 2.0. Results: PJK was observed in 22 of 62 patients (35%). Average followup was 34 months. There was a significant difference in rFCSA between the PJK and non-PJK groups (P=0.026). Patients who developed PJK comprised 37.5% of the L1-L2 group and 31.8% of the T9-T12 group, but there was no significant difference in the incidence of PJK. The L1-L2 group showed significant differences in age, rFCSA, lumbar lordosis, and global sagittal alignment (P=0.047, 0.001, 0.041, 0.016, respectively), between the PJK and non-PJK groups. However, there was no difference in the T9-T12 group. Conclusions: rFCSA measured using MRI at the level of L4-L5 is an effective tool in predicting PJK after long-segment fusion for LDD, especially UIVs in the lumbar spine.