Introduction; Thoracolumbar (TL) burst fractures accounts for 20% of all spine fractures. The complexity of the injury pattern complicates treatment decisions. Dural tears and entrapped cauda equina (ECE) should be assessed preoperatively because any reduction maneuver results in crushing the entrapped nerve. However, presence of ECE may be difficult to determine unambiguously by clinical and radiological examination before surgical treatment. This study aimed to identify the radiological features of ECEs and to provide an index to predict the occurrence of ECEs preoperatively.

Methods; This retrospective observational study included 143 patients who underwent surgical treatment for thoracolumbar burst fractures at the Hokkaido Spinal Injury Center. The relationship between the occurrence of ECE and the spinal canal occupancy rate (OR) of the vertebral posterior wall bone fragments (PWF) (OR–PWF) in the axial view of computed tomography was investigated. We investigated the differences between the ECE-positive and ECE-negative groups in terms of the demographic data, the injured vertebral levels, the presence and type of lamina fracture, and the OR–PWF. 

Results; One hundred one patients (103 vertebrae) fulfilled the inclusion criteria for this study (74 males and 27 females, mean age: 48.4 ± 19.4 years). Average follow-up period was 51.2 ± 40.0 months (median: 42.5 months). The preoperative MRI showed that ECE was suspected in 9 cases, of which 7 cases actually occurred. There were no significant differences in age, height, or BMI according to the presence or absence of ECE, but weight was significantly heavier in the ECE-positive group than in the ECE-negative group. The levels of the injury in the ECE-positive group were at L2 in 3 cases, at L3 in 3 cases and at L4 in 1 case. The distribution of the vertebral level of lumbar burst fracture occurrence was highest in L1, followed by L2, 3, 4, and 5. There were 28 cases of complete type lamina fractures (27.2 %) and 60 cases of incomplete type lamina fractures (58.3 %) on axial view, and 68 cases of whole type lamina fractures (66.0 %) and 19 cases of partial type lamina fractures (18.4 %) on coronal view. One case was unclassifiable. In 14 cases, no lamina fractures occurred. In the ECE-positive group, all cases had lamina fractures; 4 complete type lamina fractures and 3 incomplete type lamina fractures. On the coronal view, all cases were whole type lamina fractures. The OR–PWF was significantly higher in the ECE-positive group than in the ECE-negative group (P = 0.03).

Discussion; This study was the first to clarify the relationship between the degree of protrusion of the PWF and ECE. In addition, it was shown that burst fractures with ECE were more likely to occur at the L2 and L3 levels. OR–PWF was considered to be a highly accurate predictor of ECE occurrence. Correct preoperative evaluation of ECE and cauda equina salvage prior to any reduction maneuver is essential for proper treatment of TL burst fractures.