Interventional procedure overview of vertebral body tethering for idiopathic scoliosis in children and young people

Improvement in scoliotic curve

In a meta-analysis of 24 studies (n=1,280: 1,278 patients with idiopathic scoliosis and 2 patients with syndromic scoliosis), the pooled mean Cobb angle of the main thoracic curve was 46.0^° (95% CI 42.3^° to 50.0^°; 10 studies) in patients who had anterior VBT and 53.3^° (95% CI 52.8^° to 53.9^°; 14 studies) in patients who had PSF preoperatively (Shin 2021). Of the studies with a follow up of 36 months or more after operation (number of studies not reported), the mean main thoracic curve was corrected to 22.5^° (95% CI 14.1^° to 30.9^°) for anterior VBT and 22.7^° (95% CI 19.6^° to 25.8^°) for PSF. In the same meta-analysis, the pooled mean Cobb angle of the lumbar curve was 28.7^° (95% CI 25.6^° to 32.0^°; 9 studies) for anterior VBT and 30.9^° (95% CI 29.2^° to 32.5^°; 5 studies) for PSF preoperatively and corrected to 18.0^° (95% 3.5^° to 32.5^°) and 15.2^° (13.3^° to 17.1^°) at a follow up of 36 months or more (number of studies not reported).

In a non-randomised comparative study of 49 patients with thoracic idiopathic scoliosis, the anterior VBT and PSF groups were similar in upper (31^°±9^° compared with 30^°±6^°, p=0.633) and main thoracic curves (53^°±8^° compared with 54^°±7^°; p=0.444) preoperatively (Newton 2020). After operation, both groups obtained statistically significant corrections in upper and main thoracic curves at the final follow up (anterior VBT, 24^°±8^° before operation compared with 34±8^° at a mean follow up of 3.4 years; PSF, 13^°±5^° before operation compared with 12±4^° at a mean follow up of 3.6 years; all p<0.001). Comparative analysis between groups showed that the anterior VBT group had statistically significantly more spinal deformity than the PSF group at immediate postoperation and final follow up (immediate postoperation, 36% correction compared with 78% correction, p<0.001; final follow up, 43% correction compared with 69% correction, p=0.002).

In a non-randomised comparative study of 43 patients with adolescent idiopathic scoliosis, the mean magnitude of the main thoracic-thoracolumbar curve was statistically significantly corrected from 48.2^° before operation to 9.1^° at a mean follow up of 39.4 months in patients who had VBT (p<0.001; Pehlivanoglu 2021). In patients who had PSF, the main thoracic-thoracolumbar curve was also statistically significantly corrected from 48.8^° to 9.7^° during the same period (p<0.001).

In a non-randomised comparative study of 130 patients with idiopathic early onset scoliosis, the median major scoliosis curve was statistically significantly smaller in the VBT group (50^°, range 43.5^° to 58^°) compared with the MCGR group (64.5^°, range 55^° to 75^°) and the PSF group (63^°, range 57^° to 72^°) preoperatively (p<0.0005; Mackey 2021). After operation, the median major scoliosis curve was statistically significantly corrected to 28^° (range 21^° to 35^°) in patients who had VBT at a median follow up of 3 years, 42^° (range 34.4^° to 54.5^°) in patients who had MCGR at a median follow up of 2.9 years and 29^° (range 22^° to 36^°) in patients who had PSF at a median follow up of 3.6 years (all p<0.0005). Statistically significant difference was found in scoliosis curve correction between groups (VBT, 41% correction; MCGR, 27% correction; PSF, 52% correction; p<0.0005).

In a case series of 120 patients with idiopathic scoliosis who had anterior VBT, the mean thoracic scoliosis was 51.2^° preoperatively and statistically significantly improved to 26.9^° at immediate postoperation (p<0.05; Abdullah 2021). After 1 year, the curve was further corrected to 23.0^°, which also statistically significantly improved compared with the immediately postoperative value (p<0.01). This statistically significant improvement did not happen at 2-year follow up (27.5^°) compared with immediate postoperation (p=0.64).

In a case series of 112 patients with idiopathic scoliosis who had anterior VBT, the mean coronal Cobb angle was statistically significantly corrected from 50.8^°±10.2^° preoperatively to 26.6^°±10.1^° at time of first erect radiograph and 25.7^°±16.3^° at a mean follow up of 37 months (all p<0.001; Rushton 2021).

In a case series of 50 patients with idiopathic scoliosis who had anterior VBT, the mean Cobb angles of the proximal thoracic and main thoracic curves were 22.8^°±8.8^° and 49.4^°±8.5^° preoperatively (Miyanji 2021). After operation, the mean Cobb angles were corrected to 16.8^°±8.8^° and 27.1^°±8.5^° at immediate postoperation, 12.4^°±8.1^° and 22.1^°±8.9^° at 1-year follow up and 13.1^°±7.5^° and 24.9^°±9.5^° at a mean follow up of 2.1 years.

In a case series of 31 patients with adolescent idiopathic scoliosis who had VBT, the mean magnitude of the main thoracic curve improved from 47^°±7.6^° preoperatively to 21.8^°±6.4^° in first erect radiographs after operation (Alanay 2020).

In a case series of 57 patients with adolescent idiopathic scoliosis who had anterior VBT, the mean thoracic Cobb angle was 40.4^°±6.8^° before operation and was statistically significantly corrected to 19.3^°±8.4^° at first erect, 12.6^°±7.2^° at 1-year follow up, 13.8^°±8.9^° at 2-year follow up and 18.7^°±13.4^° at a mean follow up of 55.2 months (all p<0.0001; Samdani 2021). The mean proximal thoracic and lumbar curves were 25.0^°±5.7^° and 23.7^°± 6.1^°o before operation and were statistically significantly corrected to 17.9^°±6.4^° and 15.7^°±8.4^° (p<0.05) at a mean follow up of 55.2 months.

In a case report of 1 patient with adolescent idiopathic scoliosis, preoperative thoracic and lumbar curves were 25^° and 22^°, respectively (Rathbun 2019). After VBT, thoracic and lumbar curves improved to 22^° and 3^° at 2 weeks and then changed to 21^° and 6^° about 1 year.

Change in kyphotic or lordotic curves

In the meta-analysis of 24 studies (n=1,280), the pooled mean thoracic kyphosis was 24.3^° (95% CI 17.8^° to 30.8^°; 8 studies) in patients who had anterior VBT and 23.0^° (95% CI 20.7^° to 25.2^°; 8 studies) in patients who had PSF preoperatively. Of the studies with a follow up of 36 months or more after operation (number of studies not reported), the mean thoracic kyphosis changed to 22.5^° (95% CI 12.0^° to 33.0^°) for anterior VBT and 24.5^° (95% CI 21.9^° to 27.1^°) for PSF. In the same meta-analysis, the pooled mean lumbar lordosis was 52.0^° (95% CI 46.2^° to 57.9^°; 5 studies) for anterior VBT and 47.2^° (95% CI 28.1^° to 66.3^°; 5 studies) for PSF preoperatively and changed to 55.1^° (95% CI 51.3^° to 58.8^°) and 46.1^° (95% CI 25.0^° to 67.1^°) respectively at a follow up of 36 months or more (number of studies not reported).

In the non-randomised comparative study of 49 patients, the anterior VBT and PSF groups were similar in preoperative T2 to T12 kyphosis (anterior VBT, 24^°±12^°; PSF, 25^°±12^°; p=0.79; Newton 2020). After operation, there were statistically significant differences in T2 to T12 kyphosis between groups at the first postoperative and final follow up time points (immediate postoperation, 22^°±12^° in the anterior VBT group compared with 31^°±8^° in the PSF group, p=0.004; final follow up, 19^°±13^° in the anterior VBT group compared with 29^°±8^° in the PSF group, p=0.001).

In the case series of 120 patients, the mean global kyphosis was 28.5^°±11.2^° preoperatively and changed to 27.4^°±11.4^° at immediate postoperation and 29.2^°±12.5^° at 2 years (Abdullah 2021). During the same period, T5 to T12 kyphosis was 16.0^°±11^°, 16.9^°±10.8^° and 17.0^°±11.8^°, respectively. No statistically significantly differences were found between preoperation and 2-year follow up in global kyphosis and T5 to T12 kyphosis (all p>0.05).

In the non-randomised comparative study of 130 patients, the mean kyphosis T2 to T12 was statistically significantly lower in the VBT group (26.1^°±12.3^°) compared with the MCGR group (34.7^°±16.3^°) and the PSF group (35.9^°±13.1^°) preoperatively (p=0.010; Mackey 2021). After operation, the mean kyphosis T2 to T12 changed to 25.0^°±13.0^° in the VBT group (p=0.522) and 34.2^o±12.0^o in the MCGR group (p=0.887) and statistically significantly reduced to 25.8^°±11.5^° in the PSF group (p=0.022) at the most recent follow up. Statistically significant difference was found in the mean kyphosis T2 to T12 between groups at the most recent follow up (p=0.002).

In the case series of 112 patients, the mean thoracic kyphosis T5 to T12 statistically significantly increased from 18.6^°±11.4^° preoperatively to 21.4^°±13.0^° (p=0.004) at a mean follow up of 37 months (Rushton 2021). During the same period, the mean lumbosacral lordosis L1 to S1 changed from -55.9^°±10.5^° to -56.2^°±11.4^° (p=0.86).

In the case series of 31 patients, the median changes between preoperation and a mean follow up of 27 months for thoracic kyphosis were 4^°, 3^°, 0^° and 0^° in Sanders 2, 3, 4 to 5, and 6 to 7 groups, respectively, and for lumbar lordosis were 4^°, 0^°, -3^° and -2^° (Alanay 2020). There were no statistically significant differences in Sagittal Cobb angle changes between groups (all p>0.05).

In the case series of 57 patients, the mean thoracic kyphosis T5 to T12 changed from 15.5^°±10.0^° preoperatively to 17.0^°±10.1^° at the first erect measurement (p=0.40) and 19.6^°±12.7^° at a mean follow up of 55.2 months (p<0.05; Samdani 2021). No statistically significantly change was found in lumbar lordosis between the preoperative and 55.2-month measurements (51.9^°±11.4^° compared with 54.4^°±11.8^°, p=0.10).

Range of motion and muscle endurance

In the meta-analysis of 24 studies (n=1,280), the mean thoracic rotation was 13.7^° (95% CI 12.1^° to 15.2^°; 6 studies) in patients who had anterior VBT and 15.4^° (95% CI 12.4^° to 18.4^°; 3 studies) in patients who had PSF preoperatively (Shin 2021). After operation, thoracic rotation changed to 8.4^° (95% CI 1.0^° to 15.7^°) and 13.0^° (95% CI 3.3^° to 22.6^°) respectively at a follow up of 36 months or more (number of studies not reported).

In the non-randomised comparative study of 49 patients, the anterior VBT and PSF groups were similar in preoperative trunk rotation (thoracic angle of trunk rotation, 15^°±4^° compared with 17^°±3^°, p=0.109; lumbar angle of trunk rotation, 6^°±5^° compared with 8±5, p=0.213; Newton 2020). After operation, there were statistically significant differences in trunk rotation between groups at immediate postoperation and final follow up (thoracic angle of trunk rotation - immediate postoperation, 10^°±3^° in the anterior VBT group compared with 6^°±4^° in the PSF group, p=0.008; final follow up, 11^°±5^° in the anterior VBT group compared with 6^°±3^° in the PSF group, p<0.001; lumbar angle of trunk rotation - immediate postoperation, 6^°±1^° in the anterior VBT group compared with 1^°±3^° in the PSF group, p=0.016; final follow up, 6^°±5^° in the anterior VBT group compared with 3^°±3^° in the PSF group, p=0.021).

In the non-randomised comparative study of 43 patients, the mean lumbar ranges of motion in flexion, extension, lateral bending and rotation were statistically significantly superior in the VBT group compared with the PSF group at a mean follow up of 39.4 months (flexion, 78.2^° compared with 58.1^°, p<0.001; extension, 34.6^° compared with 19.4^°, p<0.001; lateral bending, 34.4^° compared with 18.3^°, p<0.001; rotation, 45.4^° compared with 24.1^°, p<0.001; Pehlivanoglu 2021). During the same period, the mean lumbar anterior and lateral bending flexibility were also statistically significantly superior in the VBT group compared with the PSF group (anterior, 3.7 cm compared with 23.4 cm, p<0.001; lateral, 22.4 cm compared with 11.3 cm, p=0.003). In the same study, the mean flexor and extensor trunk endurance were statistically significantly higher in the VBT group than the PSF group at a mean follow up of 39.4 months (flexor, 65.1 seconds compared with 19.2 seconds, p<0.001; extensor, 60.8 seconds compared with 28.7 seconds, p<0.001). The mean motor strength of the trunk extensor and anterior–lateral–oblique flexor muscles was statistically significantly superior in the VBT group than the PSF group (4.7 compared with 3.2, p=0.003).

In the case series of 57 patients, the mean trunk rotation based on thoracic sociometer reading was 13.6^°±3.9^° preoperatively and changed to 6.3^°±3.0^° at 2 years after operation and 8.6^°±4.9^° at a mean follow up of 55.2 months (Samdani 2021).

Improvement in thoracic and lumbar prominence

In the case series of 112 patients, the mean rib hump was statistically significantly corrected from 14.1^°±4.8^° preoperatively to 8.8^°±5.4^° at a mean follow up of 37 months (p<0.001; Rushton 2021). During the same period, the mean lumbar prominence statistically significantly improved from 3.6^°±4.7^° to 2.5^°±4.4^° (p=0.03).

In the case series of 31 patients, the median thoracic hump corrections between the preoperative and last follow up values were 67%, 50%, 36% and 58% in Sanders 2, 3, 4 to 5, and 6 to 7 groups, respectively (Alaney 2020). There was no statistically significant difference in hump angle between groups (p>0.05).

Shoulder height and balance

In the non-randomised comparative study of 49 patients, the mean shoulder height differences were 1±1 cm in the anterior group and 1.5±1 in the PSF group preoperatively and changed to 1± 1 cm and 1.3± 1 cm at the final follow up (Newton 2020). No statistically significant differences were found within and between groups (all p>0.05).

In the case series of 50 patients, the mean absolute shoulder height was 15.6±10.5 mm preoperatively and changed to 9.6±8.2 mm at immediate postoperation, 11.5±7.8 mm at 1 year and 11.3±8.3 mm at 2 years (Miyanji 2021). Preoperatively, there were 70% of patients (35/50) with acceptable shoulder balance, 28% (14/50) with moderate shoulder imbalance and 2% (n=1) with severe shoulder imbalance. At 2 years after operation, these proportions changed to 84% (42/50) of patients with acceptable shoulder balance and 16% (8/50) with moderate shoulder balance.

In the case series of 57 patients, clinically unlevel shoulders were reported in 54% of patients (28/52) preoperatively and reduced to 25% (14/55) at a mean follow up of 55.2 months (Samdani 2021).

Pulmonary function

In the case series of 57 patients, 42 patients had pulmonary function tests at both the preoperative evaluation and the 2-year follow-up evaluation or later (Samdani 2021). The mean FEV1 and FVC statistically significantly increased from 2.29 litres and 2.67 litres preoperatively to 2.77 litres and 3.17 litres at a mean follow up of 50.1 months (p<0.01). Based on normal subjects, the percentages of predicted FEV1 and FVC decreased from 83% and 83% to 75% and 74% respectively.

Improvement in quality of life

In the meta-analysis of 24 studies (n=1,280), there was no statistically significant difference found in the postoperative SRS-22 self-image or total scores between patients who had anterior VBT and patients who had PSF (self-image, 4.27 [95% CI 4.0 to 4.56; 2 studies] compared with 4.23 [95% CI 4.07 to 4.40; 7 studies]; total score, 4.36 [95% CI 4.06 to 4.65; 2 studies] compared with 4.30 [95% CI4.17 to 4.43; 7 studies]; Shin 2021).

In the non-randomised comparative study of 49 patients, there were no statistically significant differences in any SRS-22 domain or total score between the anterior VBT group and the PSF group (pain, 4.4±0.6 compared with 4.4±0.4, p=0.903; mental health, 4.3±0.6 compared with 4.0±0.7, p=0.279; self-image, 4.1±0.7 compared with 4.4±0.6, p=0.244; satisfaction, 4.3±0.7 compared with 4.7±0.5, p=0.053; general function, 4.3±0.4 compared with 4.3±0.4, p=0.748; total score, 4.2±0.4 compared with 4.4±0.4, p=0.29; Newton 2020).

In the non-randomised comparative study of 43 patients, the mean total SRS-22 and SF-36 MCS/PCS scores were comparable between the VBT and PSF groups before operation (SRS-22, 3.2 compared with 3.2; SF-36 MCS, 52.7 compared with 52.3; SF-36 PCS, 46.8 compared with 47.1; all p>0.05; Pehlivanoglu 2021). At a mean follow up of 39.4 months, the mean total SRS-22 and SF-36 MCS/PCS scores were statistically significantly higher in the VBT group than the PSF group (SRS-22, 4.9 compared with 3.8; SF-36 MCS, 56.9 compared with 52.3; SF-36 PCS, 57.2 compared with 53.1; all p<0.001).

In the non-randomised comparative study of 130 patients, the median EOSQ-24 scores statistically significantly improved in pain/discomfort (56.3 at baseline compared with 75 at 3-year follow up, p=0.023), emotion (75 compared with 93.8, p=0.020) and parental impact domains (70 compared with 87.5, p=0.020) in patients who had VBT at a mean follow up of 3 years (Mackey 2021). For patients who had PSF, the median EOSQ-24 score statistically significantly increased in parental impact domain (90 compared with 100, p=0.005). For patients who had MCGR, there was no statistically significant improvement in any domain (all p>0.05).

In the case series of 57 patients, the mean scores for SRS-22 and self-image were 4.5±0.4 and 4.4±0.7 respectively at a mean follow up of 55.2 months (Samdani 2021). The preoperative SRS-22 scores were not obtained.

Length of stay

In the non-randomised comparative study of 49 patients, the mean length of hospital stay was 5.0±1.3 days in the anterior VBT group and 4.9±1.2 days in the PSF group (p=0.7; Newton 2020).

In the case series of 120 patients, the mean length of ICU stay was 0.2±0.5 days and the mean length of postoperative hospital stay was 4.5±1.3 days (Abdullah 2021).

In a case series of 90 patients with adolescent idiopathic scoliosis who had single or bilateral VBT, the mean length of hospital stay was 8.3±3.1 days (Baroncini 2021). The first 20 patients who had VBT stayed at hospital statistically significantly longer than the last 20 patients who had VBT (9.3±2.1 days compared with 7.8±1.6 days, p=0.01).

In the case series of 112 patients, the mean length of hospital stay was 4.7±1.4 days for patients who had single stage thoracic/lumbar tethering procedures and 10.5±4.0 days for patients who had planned 2 stage procedures (p=0.001; Rushton 2021).

In the case series of 57 patients, the mean length of ICU stay after operation was 1.5±0.7 days and the mean length of stay at hospital was 4.8±1.4 days (Samdani 2021). The length of stay might be confounded by the long distance travelled by families, with most families flying.

In the single case report, the patient stayed at hospital for 15 days after operation while recovering with resolving pneumothorax and awaiting safe chest tube removal.