BACKGROUND: No consensus has been reached on the optimal radiographic evaluation of thoracolumbar burst fractures. The role of MRI in the treatment decision is not fully documented. The objectives was to measure the agreement of MRI in detecting posterior ligamentous complex (PLC) and posterior longitudinal ligament (PLL) injury, and to determine whether the findings by MRI is correlated with the results of plain radiography and computed tomography (CT) scanning as well as neurological examination and with the treatment planning. METHODS: Sixty-one patients with acute thoracolumbar burst fracture were retrospectively reviewed for the presence of supraspinous ligament (SSL), interspinous ligament (ISL) or posterior longitudinal ligament (PLL) injury. The overall interobserver agreement between the three different observers was assessed by a kappa coefficient for multiple raters. The status of ligaments was correlated with the neurological function as assessed by Frankel scale and fracture severity as defined by the Load Sharing Classification. These patients were surgically treated according to the Load Sharing Classification and followed up for at least 5 years. RESULTS: The kappa coefficients for ISL or SSL injury ranged 0.601 to 0.736, representing substantial to almost perfect agreement, whereas the kappa coefficients for PLL injury were 0.441 to 0.574, representing moderate agreement. No significant difference (P > 0.05) of Frankel scale or load sharing score was found between patients with and without ligamentous injuries. Satisfactory results were achieved in all patients regarding the clinical and radiological assessment. CONCLUSIONS: MRI is reliable for detecting the ligamentous injury, especially PLC injury in thoracolumbar burst fractures but the ligamentous injury as shown by MRI is not correlated with the neurological function or fracture severity. As MRI finding is of little value in treatment planning of thoracolumbar burst fractures, MRI examination is not necessary to be used routinely for excluding occult ligamentous injury.
BACKGROUND: No consensus has been reached on the optimal radiographic evaluation of thoracolumbar burst fractures. The role of MRI in the treatment decision is not fully documented. The objectives was to measure the agreement of MRI in detecting posterior ligamentous complex (PLC) and posterior longitudinal ligament (PLL) injury, and to determine whether the findings by MRI is correlated with the results of plain radiography and computed tomography (CT) scanning as well as neurological examination and with the treatment planning. METHODS: Sixty-one patients with acute thoracolumbar burst fracture were retrospectively reviewed for the presence of supraspinous ligament (SSL), interspinous ligament (ISL) or posterior longitudinal ligament (PLL) injury. The overall interobserver agreement between the three different observers was assessed by a kappa coefficient for multiple raters. The status of ligaments was correlated with the neurological function as assessed by Frankel scale and fracture severity as defined by the Load Sharing Classification. These patients were surgically treated according to the Load Sharing Classification and followed up for at least 5 years. RESULTS: The kappa coefficients for ISL or SSL injury ranged 0.601 to 0.736, representing substantial to almost perfect agreement, whereas the kappa coefficients for PLL injury were 0.441 to 0.574, representing moderate agreement. No significant difference (P > 0.05) of Frankel scale or load sharing score was found between patients with and without ligamentous injuries. Satisfactory results were achieved in all patients regarding the clinical and radiological assessment. CONCLUSIONS: MRI is reliable for detecting the ligamentous injury, especially PLC injury in thoracolumbar burst fractures but the ligamentous injury as shown by MRI is not correlated with the neurological function or fracture severity. As MRI finding is of little value in treatment planning of thoracolumbar burst fractures, MRI examination is not necessary to be used routinely for excluding occult ligamentous injury.
Authors: Bharti Khurana; Luciano M Prevedello; Christopher M Bono; Erwin Lin; Steven T McCormack; Hamdi Jimale; Mitchel B Harris; Aaron D Sodickson Journal: Eur Spine J Date: 2018-08-03 Impact factor: 3.134