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

INTRODUCTION. Currently used routine imaging of the lower back provides limited predictive value for the severity of low back pain (LBP) or future outcomes. Although muscles of the lumbar spine thought to play a crucial role in patient with lumbar disorders, the quantity and quality of paraspinal muscles have been rarely studied [1] and, to the best of our knowledge, the amount epimuscular fat (EF) in the lumbar spine has never been investigated. This retrospective study aimed to: 1. examine the presence and extent of EF in the lumbar spine and its association with spinal levels (L1-L5) and 2. assess the possible correlations between EF and subjects' BMI, age, sex, muscle characteristics, and LBP. 

METHODS. Baseline T2 axial MRI images from L1 to L5 of 27 adult patients (n=16 females, n=11 males; mean age: 62.3±15.1) with back disorders waiting for surgical treatment were retrospectively collected from an institutional clinical database and manually segmented by an expert investigator. Cross-Sectional Area (CSA), expressed in mm² and Fat Infiltration (FI), expressed in percentages, of psoas major, quadratus lumborum, erector spinae, and multifdus were analyzed. Further, EF was manually segmented at each lumbar level and expressed in mm². LBP scores were extracted by the Core Outcome Measures Index (COMI) and ranged from 0 with "no pain" to 10 with "worst pain that I can imagine" [2]. Normal distribution for each outcome measures was tested with the Shapiro-Wilk test. To determine differences in EF among each lumbar level a Kruskal-Wallis test followed by Dunn's multiple comparisons was performed. The existence of a correlation between EF and the other outcome variables was tested by the means of the Pearson's correlation index, or by the Spearman index for non-normally distributed variables. The level of statistical significance was set at p<0.05. All statistical procedures were performed using GraphPad Prism 9.5.1.

RESULTS. The presence of EF was detected in n=21 subjects (77.7%) at L1, n=25 (92.5%) at L2, n=26 (96.2%) at L3, and n=27 (100%) at L4 and L5. Further, the general content of EF was higher at L5 (251.5 mm²) and L4 (249.7 mm²) compared with L1 (45.2 mm²) and L2 (81.3 mm²) with p<0.001 in each comparison and, further EF was higher at L3 (147.7 mm²) compared with L1 (p=0.002; Figure 1a). No difference in EF between genders was detected. Positive correlations between EF and BMI (r²=0.47; p<0.001) and FI (r²=0.28; p=0.005) were detected (Figure 1b). Further, EF also showed a significant positive correlation with LBP (r²=0.18; p=0.042). On the other hand, EF did not show any association with age and CSA.

DISCUSSION. Higher frequency and values of EF were observed at L4-L5 than at L1-L2. Subjects with greater BMI and FI in the paraspinal muscles had higher EF values. Further, EF was positively associated with LBP scores. The functional implications of EF and its possible prognostic value for the outcome of surgical treatment in back disorders warrant further investigations.