OBJECTIVETo analyze the accuracy and safety of thoracic pedicle screws placement with interlaminar fenestration technique in severe kyphoscoliosis treatment.

METHODSTwenty-three cases of severe kyphoscoliosis between June 1996 and December 2007 underwent pedicle screw placement of thoracic vertebrae by interlaminar fenestration technique. Postoperative CT scan was performed in all cases; there were 9 males and 14 females with the averaging age 17.8 years old. The preoperative Cobb angle of the main thoracic curve was 97.3 degrees in average. The average kyphotic angle in main curve was 67.4 degrees . Patients who underwent screw placement by closing technique and postoperative CT scan in the same stage with severe kyphoscoliosis were selected as control group (B): there were 22 patients with the averaging age 17.2 years old. The preoperative Cobb angle of the main thoracic curve was 96.6 degrees in average. The average kyphotic angle in main curve was 62.1 degrees . The screw-related complications were analyzed and online measure and analyze the degree of screw penetration according to CT by statistics.

RESULTSThere were 209 thoracic pedicle screws being inserted in group A, no spinal cord and large blood vessel injury occurred, 5 patients occurred intraoperative pedicle fracture, dura lesion had 4 patients. Screws misplacement ratio was 8.6%, there were 11 screws located in the thoracic pedicle laterally, 6 screws located in medially, 1 screw presented a moderate anterior cortical perforation. The number of screws misplacement in upper and mild thoracic vertebrae were 15, occupied 83.3%; the number of screws misplacement in lower thoracic vertebrae were 3, occupied 16.7%. There was statistics difference (P < 0.05). Of the total 116 thoracic pedicle screws inserted on the convex side, the ratio of screw misplacement was 5.2%; 93 screws on the concave side, the ratio of screw misplacement was 12.9%, there was statistics difference (P < 0.05). There were 201 thoracic pedicle screws being inserted in group B, no spinal cord and large blood vessel injury occurred, 16 patients occurred intraoperative pedicle fracture, dura lesion had 7 patients. The ratio of screw misplacement was 22.4%, there were 24 screws located in the thoracic pedicle laterally, 11 screws located in medially, 10 screw presented a moderate anterior cortical perforation. The ratio of screw misplacement in Group B was higher than Group A (P < 0.05). All cases received 3.2 years and 3.4 years follow-up. There was no obvious loss of correction in coronal and sagittal plane at the latest follow-up.

CONCLUSIONSIt is technically demanding for placement of thoracic screw in the severe kyphoscoliosis treatment. The interlaminar fenestration technique can increase the accuracy and safety of thoracic pedicle screw placement.