Open Access Research article

Thoracic pedicle classification determined by inner cortical width of pedicles on computed tomography images: its clinical significance for posterior vertebral column resection to treat rigid and severe spinal deformities—a retrospective review of cases

Ying Zhang, Jingming Xie*, Yingsong Wang, Ni Bi, Zhi Zhao and Tao Li

Author Affiliations

Department of Orthopaedics, 2nd Affiliated Hospital of Kunming Medical University, #374 Dianmian Road, Kunming, Yunnan Province 650101, P.R. China

For all author emails, please log on.

BMC Musculoskeletal Disorders 2014, 15:278  doi:10.1186/1471-2474-15-278

Published: 13 August 2014



Posterior vertebral column resection (PVCR) is an effective alternative for treating rigid and severe spinal deformities. Accurate placement of pedicle screws, especially apically, is crucial. As morphologic evaluations of thoracic pedicles have not provided objective criteria, we propose a thoracic pedicle classification for treating rigid and severe spinal deformities.


A consecutive series of 56 patients with severe and rigid spinal deformities who underwent PVCR at a single institution were reviewed retrospectively. Altogether, 1098 screws were inserted into thoracic pedicles at T2-T12. Based on the inner cortical width of the thoracic pedicles, the patients were divided into four groups: group 1 (0–1.0 mm), group 2 (1.1–2.0 mm), group 3 (2.1–3.0 mm), group 4 (≥3.1 mm). The proportion of screws accurately inserted in thoracic pedicles for each group was calculated. Statistical analysis was also performed regarding types of thoracic pedicles classified by Lenke et al. (SPINE 35:1836-1842, 2010) using a morphological method.


There were statistically significant differences in the rates of screws inserted in thoracic pedicles between the groups (P < 0.008) except groups 3 and 4 (P > 0.008), which were then combined. The accuracies for the three new groups were 35.05%, 65.34%, and 88.32%, respectively, with statistically significant differences between the groups (P < 0.017). Rates of screws inserted in thoracic pedicles classified by Lenke et al. (SPINE 35:1836-1842, 2010) were 82.31%, 83.40%, 80.00%, and 30.28% for types A, B, C, and D, respectively. There was no statistically significant difference (P > 0.008) between these types except between type D and the other three types (P < 0.008).


The inner cortical width of thoracic pedicles is the sole factor crucial for accurate placement of thoracic pedicle screws. We propose a computed tomography-based classification of the pedicle’s inner cortical width: type I thoracic pedicle: absent channel, inner cortical width of 0–1 mm; type II: presence of a channel of which type IIa has an inner cortical width of 1.1–2.0 mm and type IIb a width of ≥2.1 mm. The proposed classification can help surgeons predict whether screws can be inserted into the thoracic pedicle, thus guiding instrumentation when PVCR is performed.

Scoliosis; Thoracic vertebrae; Pedicle; Computed tomography; Posterior approach; Vertebral column resection