Abstract: BACKGROUND: Bone grafting is one of the important steps in the treatment of thoracolumbar burst fractures. Because the fracture involves the spinal canal or is accompanied by spinal cord nerve damage, severe fracture bleeding and other factors, minimally invasive bone grafting for thoracolumbar burst fractures is restricted. At present, the minimally invasive treatment of thoracolumbar burst fractures is limited to percutaneous screw fixation under the tunnel. Minimally invasive percutaneous bone grafting of injured vertebrae is rarely reported, and percutaneous precise bone grafting under the endplate has not yet been reported. 
OBJECTIVE: To investigate the clinical effect of subcutaneous endplate bone graft support reduction combined with percutaneous pedicle screw short-segment fixation in the treatment of A3+B2 thoracolumbar burst fractures. 
METHODS: From June 2017 to December 2021, 90 patients with A3+B2 type asymptomatic thoracolumbar burst fracture were randomly divided into 3 groups according to admission time. In group A, 33 patients received the bone graft funnel accurately placed through the pedicle channel by percutaneous puncture under C-arm fluoroscopy, bone graft support reduction under the fracture endplate, percutaneous pedicle screw fixation. In group B, 30 patients received multifissure intermuscular approach through pedicle bone graft support reduction combined with pedicle screw fixation. In group C, 27 patients received percutaneous pedicle screw short-segment fixation under postural reduction. All patients were followed up for at least 18 months after surgery. The clinical data of the three groups, including preoperative, postoperative and last follow-up Cobb angle, anterior edge height ratio and visual analog scale pain score, were compared and analyzed. 
RESULTS AND CONCLUSION: (1) There were no significant differences in age, sex, injury segment and causative factors among the three groups (P > 0.05). (2) All patients at follow-up had no neurological impairment, no obvious lumbar posterior deformity or intractable low back pain. (3) The operation time of group C was less than that of group A and group B (P < 0.05). Intraoperative blood loss was less in group A and group C than in group B (P < 0.05). (4) There were no significant differences in the anterior edge height ratio and Cobb angle among the three groups (P > 0.05). Postoperative data in groups A and B were better than that in group C. At last follow-up, group A and group B outperformed group C (P < 0.05). The height and Cobb angle of the vertebral body lost in the three groups were smaller in groups A and B than those in group C (P < 0.05). (5) Visual analog scale pain score was better in groups A and C than that in group B after surgery (P < 0.05). There was no significant difference in visual analog scale pain score among the three groups at last follow-up (P > 0.05). (6) In group C, there was one case of loose internal fixation and displacement in 1 month after surgery, and the vertebral height was lost again with back pain, and after strict bed rest for 6 weeks, the vertebral height loss was not aggravated, the pain was relieved, and the internal fixation was removed after 1 year, and the height loss at the last follow-up was not aggravated. There were no cases of failure of internal fixation in groups A and B. (7) It is indicated that subcutaneous endplate bone graft support reduction combined with percutaneous pedicle screw short-segment fixation in the treatment of A3+B2 thoracolumbar burst fracture has the advantages of less trauma, less bleeding and light postoperative pain symptoms, and the effect of injury vertebral reduction and height maintenance is the same as the reduction through pedicle bone grafting support and short segment fixation with pedicle screws through the multifidus space approach.

Key words: thoracolumbar burst fracture, subcutaneous endplate bone graft, support reduction, pedicle screw, short-segment fixation, minimally invasive