A Novel Pedicle Channel Classification Describing Osseous Anatomy

Study Design. Prospective clinical series. Objective. To determine how many thoracic scoliotic pedicles have cancellous versus cortical versus absent channels. Summary of Background Data. Although morphologic evaluations of thoracic pedicles have been well reported, the results do not practically reflect clinical findings during actual pedicle screw placement. We propose a novel pedicle channel classification describing the osseous anatomy encountered during pedicle probe insertion. Methods. We noted 4 pedicle types in 53 consecutive scoliosis patients. Type A: pedicle probe smoothly inserted without difficulty; the morphology is described as a "Large Cancellous Channel." Type B: pedicle probe inserted snugly with increased force; described as a "Small Cancellous Channel." Type C: pedicle probe cannot be manually pushed but must be tapped with a mallet down the pedicle into the body; described as a "Cortical Channel." Type D: pedicle probe cannot locate a channel thus necessitating a "juxtapedicular" screw position; described as a "Slit/Absent Channel." The average age at time of surgery was 23.4 ± 16.7 years. Diagnoses included idiopathic scoliosis (n = 38) and syndromic scoliosis (n = 15). The average main thoracic Cobb angle was 73° ± 26°. Evaluation of pedicle morphology of the 4 types was also performed in 21 consecutive cases of adolescent idiopathic scoliosis using preoperative computed tomography images. Results. A total of 1021 pedicles with screws placed were evaluated. The average percent per type was as follows: 61.0% type A; 29.2% type B, 6.8% type C, and 3.0% type D. On the convexity, 98.2% of pedicles were type A or B versus 81.5% on the concavity (P < 0.05). There were significant differences between adolescent versus adult idiopathic scoliosis (P = 0.007), and syndromic scoliosis versus adult idiopathic scoliosis (P = 0.017) regarding pedicle morphologic proportions. There was a significant tendency toward a decrease in the proportion of type A pedicles, an increase in the proportion of type B pedicles as the Cobb angle increased (P < 0.0001). Evaluation based on 312 thoracic pedicles in 21 consecutive adolescent idiopathic scoliosis patients using preoperative computed tomography axial images confirmed assumptions of the 4 pedicle types. Conclusion. We propose a classification for pedicle channels describing the osseous anatomy encountered during pedicle probe insertion. Based on the classification, surprisingly, we found during surgery that 90% of thoracic pedicles had a cancellous channel, whereas 7% had a cortical channel and only 3% had an absent channel.