OBJECTIVE

Hemivertebrae in the cervicothoracic junction in the pediatric population are treated conventionally with a two-rod instrumentation pattern. However, the increase in complexity, severity, and immaturity of osseous malformation in the cervicothoracic spine presents additional challenges in construct planning. This study aims to introduce an integrated instrumentation strategy named the sequential correction technique in the treatment of congenital cervicothoracic scoliosis caused by hemivertebra (CTS-HV) and evaluate its feasibility and treatment effects.

METHODS

We retrospectively analyzed a consecutive series of patients with CTS-HV who underwent posterior-only HV resection with sequential correction technique from March 2018 to November 2023. This technique employed multiple rods, each being designated for a specific task, to sequentially perform surgical maneuvers involving osteotomy closure, torticollis correction, and implant integration. Individualized adjustments on instrumentation configuration involving rod number, rod type (whole, segmental, or satellite), cervical anchor choice, and connector placement could be made according to the severity of CTS and cervical pedicle dysplasia. Radiographic deformity parameters of the head-neck-shoulder complex were measured preoperatively, postoperatively, and at the latest follow-up. One-way repeated measures analysis of variance and Bonferroni correction were used to compare data at different time points. Additionally, any complications that occurred intraoperatively and during follow-up would be recorded.

RESULTS

Twenty-two pediatric and adolescent patients were recruited with a mean age of 8.3 ± 3.7 years. The ratio for the location of the resected CTS-HVs were C6 (4.6%), C7 (13.6%), T1 (31.8%), T2 (9.1%), T3 (27.6%), and T4 (13.6%). All patients were instrumented with screw-hook hybrid constructs, of which 3-rod and 4-rod constructs accounted for 81.8% and 18.2%, respectively. The cervicothoracic scoliosis, T1 tilt, neck tilt, clavicular angle, head tilt, and head shift were all significantly corrected from 53.1° ± 11.4°, 25.3° ± 10.1°, 19.6° ± 9.3°, 4.5° ± 3.1°, 10.7° ± 8.3°, and 21.8 ± 18.0 mm preoperatively to 20.8° ± 7.6°, 14.4° ± 7.2°, 7.3° ± 6.5°, 2.3° ± 2.6°, 4.4° ± 2.5°, and 9.8 ± 8.8 mm postoperatively (all p < 0.05). No significant correction loss was observed at the final follow-up (all p > 0.05). The incidences of intraoperative dural tear and iatrogenic Horner's syndrome were both 4.6%. Transitory bilateral nerve root paralysis causing upper limb dysfunction occurred in 1 patient. Additionally, 3 patients suffered severe distal curve progression with trunk tilt and were surgically revised with instrumentation extending to the stable zone. No implant-related complications were observed.

CONCLUSIONS

This modified sequential correction technique possesses the merits of easy rod installation, satisfying torticollis correction, good symmetry and verticality of the entire instrumentation, and high fixation rigidity with multi-rod constructs across the cervicothoracic junction. Thus, it is endowed with great application values in the treatment of CTS.