[Delayed indications for additive ventral treatment of thoracolumbar burst fractures : What correction loss is to be expected].

introductionThere is a general consensus that unstable vertebral body fractures of the thoracolumbar junction with a B type fracture or a high load shear index need to be surgically stabilized, primarily by a dorsal approach. The authors believe that there are indications for an additional ventral spondylodesis in cases of reduction loss or a relevant intervertebral disc lesion in magnetic resonance imaging (MRI) 6 weeks after dorsal stabilization. However, in cases of unstable vertebral fractures it remains unclear if a delayed anterior spondylodesis will lead to unacceptable loss of initial reduction. MATERIAL AND

methodsA total of 59 patients were included in this study during 2013 and 2014. All patients suffered from a traumatic vertebral fracture of the thoracolumbar junction and were initially treated with a dorsal short segment stabilization. All vertebral body fractures had a load shear index of at least 5 or were B type fractures. An x-ray control was carried out after 2 and 6 weeks and MRI was additionally performed after 6 weeks. An additional ventral spondylodesis was recommended in patients showing a reduction loss of at least 5° and in patients with relevant intervertebral disc lesions. The extent of the reduction loss was analyzed. Other parameters of interest were the fracture level, fracture classification, patient age and surgical technique (e.g. implant, index screw, laminectomy and cement augmentation).

resultsThe patient collective consisted of 23 women and 36 men (average age 51 years ± 17 years). The mean reduction loss was 5.1° (± 5.2°) after a mean follow-up of 60 days (± 56 days). The reduction loss was significantly higher when polyaxial implants were used compared to monoaxial dorsal fixators (10.8° versus 4.0°, p < 0.001). There was a significantly higher reduction loss in those patients who received a laminectomy (11.3° versus 4.3°, p = 0.01) but there were no significant differences if an index screw was used (4.5° versus 5.3°). Additionally, there was a significantly lower reduction in the subgroup of patients 60 years or older who were stabilized using cement-augmented screws (3.9° versus 11.3°, p = 0.02). The mean reduction loss was 2.8° (± 2.5°) in patients treated with a monoaxial implant, cement-augmented if 60 years or older and without laminectomy (n = 39). There was no significant correlation between reduction loss and the other parameters of interest, such as fracture morphology with classification according to the working group on questions of osteosynthesis (AO) and McCormack or fracture level.

conclusionDelayed indications for an additional ventral spondylodesis in patients with unstable thoracolumbar vertebral fractures and initial dorsal stabilization will cause no relevant reduction loss if monoaxial implants are used and laminectomy can be avoided. Additionally, cement augmentation of the pedicle screws seems to be beneficial in patients 60 years of age or older.