Summary of key evidence on vertebral body tethering for idiopathic scoliosis in children and young people

Study 1 Newton PO (2020)

Study details

Study type	Non-randomised comparative study (retrospective)
Country	US (single centre)
Recruitment period	2011 to 2016
Study population and number	n=49 (anterior VBT, n=23; PSF, n=26) Patients with thoracic idiopathic scoliosis
Age and sex	Anterior VBT: mean 12 years; 70% (16/23) female PSF: mean 13 years; 88% (23/26) female
Patient selection criteria	Inclusion criteria for anterior VBT: patients with idiopathic scoliosis who had anterior VBT with a minimum follow up of 2 years (82% of available cases). Inclusion criteria for PSF: patients identified by matching the demographic characteristics of the anterior VBT group and included patients with primary thoracic idiopathic scoliosis, curve magnitude of 40^° to 67^°, Risser stage of ≤1, age of 9 to 15 years at the time of the surgical procedure, no prior spine surgery, surgery between 2011 and 2016, and minimum follow up of 2 years.
Technique	Anterior VBT: a right-side thoracoscopic approach with single-lung ventilation was used. The pleura overlying the convex lateral aspect of the thoracic vertebrae were divided, and the segmental vessels were ligated prior to the placement of pronged washers and bicortical vertebral body screws. Tensioning of the cord was done sequentially from proximal to distal, with the greatest tension placed at the apical segments. PSF: segmental pedicle screws were used.
Follow up	Anterior VBT: mean 3.4 years (range 2 to 5 years) PSF: mean 3.6 years (range 2 to 7 years)
Conflict of interest/source of funding	Conflict of interest: one or more of the authors indicated that the author had a relevant financial relationship in the biomedical arena outside the submitted work and indicated that the author had a patent and/or copyright, planned, pending, or issued, broadly relevant to this work. Funding: none

Study design issues: This retrospective study compared outcomes for patients with thoracic idiopathic scoliosis between a group of patients who had anterior VBT and a matched cohort of patients who had PSF and instrumentation.

Outcome data included radiographic, and perioperative outcomes, complications, tether breakages (indicated by ≥6^o increase of angulation between adjacent screws on any 2 postoperative radiographs), revision procedures, clinical deformity measurements, and SRS-22 outcomes. All radiographs were measured by trained research staff, who were not involved with the surgical or clinical care of the patients and were not blinded to non-radiographic clinical data. The number of patients having anterior VBT with an outcome that was considered a clinical success (defined as having a curve of <35^° and no PSF done or indicated at the final follow up) and the number of patients with curves of <50^° (≥50^° is considered classically within the surgical range because of the risk of progression) at the time of the final follow up were recorded.

Study population issues: Preoperatively, the 2 groups were similar in age, Risser stage and height but the anterior VBT group was statistically significantly less skeletally mature in terms of triradiate cartilage and Sanders stage.

Other issues: There was a potential for selection bias for patients in the anterior VBT group because the data for this group were collected retrospectively, and, despite the best attempts at matching the PSF group, there were minor differences, particularly in skeletal maturity. Patients were also skeletally mature at the time of the final follow up. In addition, there was no available data on the amount of tension applied to the tethers at each level, which might play a role in the amount of correction achieved.

Key efficacy findings

Number of patients analysed: 49

Clinical success:

Anterior VBT group: n=12
PSF group: n=17

Length of hospital stay: 5.0±1.3 days in the anterior VBT group and 4.9±1.2 days in the PSF group (p=0.7).

Radiographic and clinical deformity measurements for the anterior VBT group

	Visit type				P value
Variable	Preoperative	First postoperative	FU before any revision	Final FU or before PSF	Preoperative vs. first postoperative	Preoperative vs. pre-revision	Preoperative vs. final	First postoperative vs. pre-revision	First postoperative vs. final	Pre-revision vs. final
Upper thoracic curve (^°) (n=22)	31±9 (16 to 52)	24±9 (10 to 42)	23±13 (1 to 59)	24±13 (1 to 59)	<0.001	0.001	0.012	0.99	0.99	0.99
Main thoracic curve (^°) (n=23)	53±8 (41 to 67)	34±8 (16 to 50)	30±21 (-12 to 62)	33±18 (-5 to 62)	<0.001	<0.001	<0.001	0.99	0.99	0.99
Lumbar curve (^°) (n=22)	34±8 (20 to 57)	26±7 (7 to 48)	26±11 (7 to 48)	29±11 (12 to 58)	<0.001	0.027	0.084	0.99	0.99	0.99
T2 to T12 kyphosis (^°) (n=22)	25±12 (6 to 54)	23±12 (8 to 62)	20±12 (-2 to 52)	19±13 (-2 to 52)	0.098
Thoracic angle of trunk rotation (^°) (n=5)	7±3 (5 to 9)	7±1 (6 to 7)	8±4 (5 to 10)	8±4 (5 to 10)	NA
Coronal imbalance (C7-central sacral vertical line) (cm)	1.1±1 (0.1 to 3.5)	1.2±1 (0.3 to 5)	1.3±1 (0 to 3.7)	1.4±1 (0 to 3.7)	0.91
Shoulder height difference (cm)	0.7±0.8 (0 to 2)	0.9±0.4 (0.5 to 1.5)	0.7±0.7 (0 to 2)	0.8±0.7 (0 to 2)	0.55

FU, follow up

Radiographic and clinical deformity measurements in the PSF group

	Visit type			P value
Variable	Preoperative	First postoperative	Final follow up	Preoperative vs. first postoperative	Preoperative vs. final follow up	First postoperative vs. final follow up
Radiographic measurements (n=26)
Upper thoracic curve (^°)	30±6 (17 to 42)	13±5 (2 to 23)	14±6 (2 to 24)	<0.001	<0.001	0.99
Main thoracic curve (^°)	54±7 (40 to 66)	12±4 (5 to 21)	16±6 (6 to 31)	<0.001	<0.001	<0.001
Lumber curve (^°)	34±9 (19 to 62)	13±8 (3 to 34)	12±6 (3 to 26)	<0.001	<0.001	0.99
T2 to T12 kyphosis (^°)	25±12 (3 to 51)	31±8 (18 to 44)	29±8 (16 to 50)	0.01	0.12	0.64
Thoracic angle of trunk rotation (^°) (n=15)	17±4 (10 to 22)	6±4 (0 to 14)	6±3 (0 to 10)	<0.001	<0.001	0.99
Lumbar angle of trunk rotation (^°) (n=7)	11±7 (0 to 21)	1±3 (0 to 7)	2±3 (0 to 7)	0.03	0.05	0.03
Coronal imbalance (C7-central sacral vertical line) (cm)	1.1±1 (0 to 3)	1.4±1 (0 to 3.8)	0.7±0.5 (0 to 1.8)	0.99	0.1	0.026
Shoulder height difference (cm)	1.4±1 (0 to 4)	1.2±1 (0 to 3)	1.2±1 (0 to 3)	0.99	0.99	0.99

Between-group comparisons of the radiographic and clinical measures of deformity

	Anterior VBT group	PSF group	P value
Preoperative
Upper thoracic curve (^°)	31±9 (16 to 52)	30±6 (17 to 42)	0.633
Main thoracic curve (^°)	53±8 (41 to 67)	54±7 (40 to 66)	0.444
Lumbar curve (^°)	34±8 (20 to 57)	34±9 (19 to 62)	0.791
T2 to T12 kyphosis (^°)	24±12 (6 to 54)	25±12 (3 to 51)	0.79
Thoracic angle of trunk rotation (^°)	1±4 (5 to 24)	17±3 (10 to 22)	0.109
Lumbar angle of trunk rotation (^°)	6±5 (0 to 20)	8±5 (0 to 21)	0.213
Coronal imbalance (C7-central sacral vertical line) (cm)	1±1 (0.1 to 3.5)	1±1 (0 to 3)	0.81
Shoulder height difference (cm)	1±1 (0 to 3.5)	1.5±1 (0 to 4)	0.26
First postoperative
Upper thoracic curve (^°)	24±8 (10 to 42)	13±5 (2 to 23)	<0.001
Correction of thoracic curve (%)	34±8 (15 to 50)	12±4 (5 to 21)	<0.001
Main thoracic curve (^°)	36±11 (18 to 65)	78±8 (57 to 90)	<0.001
Lumbar curve (^°)	26±7 (14 to 41)	13±8 (3 to 34)	<0.001
T2 to T12 kyphosis (^°)	22±12 (6 to 54)	31±8 (18 to 44)	0.004
Thoracic angle of trunk rotation (^°)	10±3 (5 to 24)	6±4 (0 to 14)	0.008
Lumbar angle of trunk rotation (^°)	6±1 (0 to 20)	1±3 (0 to 7)	0.016
Coronal imbalance (C7-central sacral vertical line) (cm)	1.2±1 (0.2 to 5)	1.4±1 (0 to 3.8)	0.47
Shoulder height difference (cm)	1±1 (0 to 2.7)	1.2±1 (0 to 3.1)	0.19
Final follow up
Upper thoracic curve (^°)	24±13 (1 to 59)	14±6 (2 to 24)	0.001
Main thoracic curve (^°)	33±18 (-5 to 62)	16±6 (6 to31)	<0.001
Correction of thoracic curve (%)	43±38 (-3 to 154)	69±13 (34 to 89)	0.002
Lumbar curve (^°)	30±12 (12 to 58)	12±6 (3 to 26)	<0.001
T2 to T12 kyphosis (^°)	19±13 (-2 to 52)	29±8 (16 to 50)	0.001
Thoracic angle of trunk rotation (^°)	11±5 (2 to 22)	6±3 (0 to 10)	<0.001
Lumbar angle of trunk rotation (^°)	6±5 (0 to 20)	3±3 (0 to 11)	0.021
Coronal imbalance (C7-central sacral vertical line) (cm)	1.3±1 (0 to 3.7)	0.7±0.5 (0 to 1.8)	0.012
Shoulder height difference (cm)	1±1 (0 to 3)	1.3±1 (0 to 3)	0.39

Between-group comparisons of SRS-22 scores at the time of the final follow up

SRS-22 domain	Anterior VBT group (n=12)	PSF group (n=22)	P value
Pain	4.4±0.6 (3.6 to 5.0)	4.4±0.4 (3.2 to 5.0)	0.903
Mental health	4.3±0.6 (3.0 to 5.0)	4.0±0.7 (3.2 to 5.0)	0.279
Self-image	4.1±0.7 (2.8 to 5.0)	4.4±0.6 (2.8 to 5.0)	0.244
Satisfaction	4.3±0.7 (3.0 to 5.0)	4.7±0.5 (3.0 to 5.0)	0.053
General function	4.3±0.4 (3.4 to 4.8)	4.3±0.4 (3.4 to 4.6)	0.748
Total	4.2±0.4 (3.4 to 4.9)	4.4±0.4 (3.6 to 4.9)	0.29

Key safety findings

Revision procedures in the anterior VBT group

ID	Time to first revision (year)	Reason for first revision	First revision procedure	Time to second revision (year)	Reason for second revision	Second revision procedure
A3	2.7	Overcorrection	Tether removal		NA	NA
A6	3.7	Broken tether with progression	PSF T3 to L3		NA	NA
A10	1.7	Progression of lumbar curve	Removal of tether at distal 2 levels	2.8	Broken tether with progression	PSF T3 to L3
A11	2.5*	Broken tether with progression	Retethered		NA	NA
A12	1.2	Overcorrection	Tether removal	3.1	Progression	PSF T9 to L3
A13	2.1	Progression of lumbar curve	Lumbar curve tethered		NA	NA
A22	2.1	Overcorrection	Tether replaced with less tension		NA	NA

*Tether revised at outside institution.

Broken tether: n=12 (52%) in the anterior VBT group.

3 of these patients had revision linked to progression of the curve because of tether breakage. Of the identified broken tethers, the majority broke at 1 level, but ranged from 1 to 3 levels. The breakages happened at T8/T9 (6), T9/T10 (2), T10/T11 (4), T11/T12 (2), T12/L1 (2), and L1/L2 (1). Among the 12 patients with broken tethers, 3 had breakages at the apex only, 1 had breakage above the apex, 6 had breakages below the apex, and 2 had breakages both at and below the apex.

Medical complications:

Anterior VBT group: atelectasis with pulmonary oedema treated with positive airway pressure that resolved by postoperative day 6, pain radiating down the leg 3 years postoperatively that resolved with physical therapy (relationship to procedure unclear), and Horner syndrome with asymmetric pupils remaining at the time of this retrospective review (exact data were not reported).
PSF group: no postoperative complications.

Study 2 Pehlivanoglu T (2020)

Study details

Study type	Non-randomised comparative study (retrospective)
Country	Turkey (single centre)
Recruitment period	2016 to 2019
Study population and number	n=43 (VBT, n=21; PSF, n=22) Patients with adolescent idiopathic scoliosis
Age and sex	VBT: mean 11.1 years, 71% (15/21) female PSF: mean 10.9 years, 73% (16/22) female
Patient selection criteria	VBT: patients with Risser ≤2 - Sanders ≤4, Age: 9 to 14, thoracic-thoracolumbar curves (40° to 60°), and a minimum of 3-year follow up. PSF: To compare the functional outcomes of the VBT group, an age-gender-instrumented level and minimum follow-up duration matched PSF group was selected from the patients operated by the same surgical team.
Technique	VBT: VBT was applied to thoracolumbar levels (T5 to L3) using thoracoscopic approach and to lumbar levels using mini-retroperitoneal approach. The most proximal and most distal instrumented levels of VBT group were T5 and L3. PSF: Posterior segmental spinal fusion was applied to thoracolumbar levels. The most proximal and most distal instrumented levels of PSF group were T3 and L3.
Follow up	Mean 39.4 months (range 36 to 48 months)
Conflict of interest/source of funding	None

Analysis

Study design issues: This retrospective, comparative study compared clinical and functional outcomes of VBT and PSF in age-gender-instrumented level and minimum follow-up duration matched patients to assess if VBT was superior compared to PSF in terms of functional outcomes and health-related quality of life.

The functional evaluation, which was done in the last follow up, included lumbar range of motion, anterior (hand to feet distance)-lateral (pre–post-bending distance) lumbar bending flexibility (cm); flexor (evaluated with modified Kraus-Weber test) and extensor (evaluated with modified Biering-Sörensen test) endurances (sec) of trunk, and average motor strength of trunk muscles. Patient reported outcomes included SRS-22 and SF-36 MCS/PCS scores.

Study population issues: Patients in the VBT group had an average of 9.3 levels (range 6 to 11) of tethering. They had main thoracic-thoracolumbar curves with an average preoperative major curve magnitude of 48.2^°. Patients in the PSF group had an average of 10.1 levels (range 8 to 12) of fusion. They had main thoracic-thoracolumbar curves with an average preoperative major curve magnitude of 48.8^°. The curve flexibility was higher in the VBT group and curve magnitudes were higher in the PSF group at baseline, even if the patients were selected age-gender-follow-up duration and instrumented level matched.

Key efficacy findings

Number of patients analysed: 43

Radiographic data

	VBT group (n=21)	PSF group (n=22)
Average preoperative major curve magnitude (^°)	48.2	48.8
Average major curve magnitude at the last follow up (^°)	9.1 (p<0.001)	9.7(p<0.001)
Average preoperative coronal balance (cm)	1.9 (-0.6 to 2.4)	1.8 (-0.4 to 2.2)
Average coronal balance (cm) at the last follow up	0.8 (-0.6 to 1.2)	0.3 (-0.1 to 0.7)
Average preoperative sagittal balance (cm)	1.2 (-0.4 to 1.7)	1.0 (-0.6 to 1.4)
Average sagittal balance (cm) at the last follow up	0.4 (-0.3 to 1.1)	0.3 (-0.2 to 0.8)

Lumbar range of motion and lumbar bending flexibility at the last follow up

	Average lumbar ROM (^o)				Average lumbar bending flexibility (cm)
	Flexion	Extension	Lateral bending	Rotation	Anterior	Lateral
VBT group	78.2	34.6	34.4	45.4	3.7	22.4
PSF group	58.1	19.4	18.3	24.1	23.4	11.3
P value	<0.001	<0.001	<0.001	<0.001	<0.001	0.003

Trunk endurances and motor strength of trunk at the last follow up

	Average trunk endurance(s)		Average motor strength of trunk muscles
	Flexion	Extension	Extensor-Ant/Lat/Obl flexor
VBT group	65.1	60.8	4.7
PSF group	19.2	28.7	3.2
P value	<0.001	<0.001	0.003

SRS-22 scores

		Average preoperative SRS-22 scores	Average SRS-22 scores at the last follow up
VBT group	Function	3.4	4.8
	Pain	3.6	4.9
	Self-image	2.8	4.8
	Mental health	3.7	4.9
	Satisfaction	2.5	4.9
	Total	3.2	4.9
PSF group	Function	3.2	4.1
	Pain	3.5	4.1
	Self-image	2.9	3.3
	Mental health	3.6	3.9
	Satisfaction	2.6	3.6
	Total	3.2	3.8
P value		>0.05	<0.001

SF-36 scores

	Average preoperative SF-36 scores		Average SF-36 scores at the last follow up
	MCS	PCS	MCS	PCS
VBT group	52.7	46.8	56.9	57.2
Fusion group	52.3	47.1	52.3	53.1
P value	>0.05	>0.05	<0.001	<0.001

Key safety findings

Safety data were not reported.

Study 3 Mackey C (2021)

Study details

Study type	Non-randomised comparative study (registry)
Country	International (multiple centres)
Recruitment period	VBT: 2013 to 2017 MCGR: 2013 to 2018 PSF: 2007 to 2017
Study population and number	n=130 (VBT, n=37; MCGR, n=51; PSF, n=42) Patients with idiopathic early onset scoliosis
Age and sex	VBT: median 11.3 years; 97.3% (36/37) female MCGR: median 9.6 years; 68.6% (35/57) female PSF: median 10.9 years; 80.9% (34/42) female
Patient selection criteria	Inclusion criteria: ambulatory children with idiopathic early onset scoliosis, ages 8 to 11 years at index surgery, index surgery of single PSF, MCGR or VBT and minimum follow up of 2 years.
Technique	VBT, MCGR or PSF was done. The most proximal UIV and most distal LIV were T5 and L3 for VBT, C7 and L4 for MCGRs and C5 and L5 for PSF.
Follow up	All groups: Mean 3.4±1.5 years VBT: median 3.0 years (range 2.1 to 3.6 years) MCGR: median 2.9 years (range 2.4 to 3.9 years) PSF: median 3.6 years (range 2.2 to 5.1 years)
Conflict of interest/source of funding	None

Analysis

Study design issues: This retrospective review of prospective data from an international multicentre spine registry compared outcomes of VBT versus MCGR versus single PSF in 8- to 11-year-old patients with idiopathic early onset scoliosis.

Major curve size was measured using the Cobb method. T2 to T12 Kyphosis and Proximal Junctional Angle (PJA) was measured. Spinal growth was assessed using thoracic height (T1 to T12) and spinal height (T1 to S1). Complications, additional surgeries and unplanned surgeries were recorded. Repeat surgical events were classified as either planned or unplanned returns to the operating room. Patient's health-related quality of life was assessed using the validated EOSQ-24, which was administered to families before their index surgery and during follow up.

Radiographic parameters were recorded pre-operatively and at most recent follow up. For VBT and MCGR patients undergoing final fusion, radiographic parameters at time of final fusion were included as most recent follow up. Unplanned returns were because of curve progression, hardware failure, pulmonary complications, pain, infection, or pseudoarthrosis. Planned returns included definitive spinal fusion events (in patients who previously had VBT /MCGR) and elective wound modifications (scar revisions).

Study population issues: VBT patients had higher median age at index surgery compared to MCGR and PSF patients (p<0.0005). Fewer patients in the VBT group had open triradiate cartilage (41% VBT compared with 96% MCGR compared with 65% PSF; p<0.0005) and when evaluating Risser stage 96% of MCGR patients fell into Risser Stage 0, compared with 82% of VBT and 82% of PSF patients (p=0.025). The VBT group included more females (97%) compared to the MCGR (69%) and PSF groups (81%) (p<0.0005).

Other issues: There were differences between centres including number of patients treated, surgeon experience and decision-making.

Key efficacy findings

Number of patients analysed: 130

Radiographic data

	VBT	MCGR	PSF	P value
Scoliosis curve
Major scoliosis angle, preoperative (^°), median (IQR)	50 (43.5 to 58)	64.5 (55 to 75)	63 (57 to 72)	<0.0005
Major scoliosis angle, recent follow up (^°), median (IQR)	28 (21 to 35)	42 (34.4 to 54.5)	29 (22 to 36)	<0.0005
p-value pre-operative to recent follow up within surgical group	<0.0005	<0.0005	<0.0005
Major scoliosis angle pre-operative to recent follow up (%), mean±SD	41.1±22.4	27.4±23.9	52.2±19.9	<0.0005
Kyphosis and proximal junctional angle
Kyphosis T2 to T12, preoperative (^°), mean±SD	26.1±12.3	34.7±16.3	35.9±13.1	0.010
Kyphosis T2 to T12, recent follow up (^°), mean±SD	25.0±13.0	34.2±12.0	25.8±11.5	0.002
p-value pre-operative to recent follow up within surgical group	0.522	0.887	0.022
Proximal junctional angle recent follow up (^°), median (IQR)	8 (5 to 12)	5.6 (3 to 12.2)	7.5 (4.7 to 12.8)	0.436
Thoracic height (T1 to T12)
T1 to T12 pre-operative (cm), median (IQR)	21 (19.8 to 22.4)	19 (17.9 to 21)	21.5 (19.1 to 22.6)	0.004
T1 to T12 recent follow up (cm), median (IQR)	24.4 (22.9 to 25.6)	22.3 (21 to 23.5)	24.7 (23.4 to 23.6)	<0.0005
T1 to T12 pre to recent follow up (cm), median (IQR)	2.8 (1.9 to 4.4)	3.3 (1.5 to 6.2)	3.6 (2.9 to 5.3)	0.182
T1 to T12 pre to recent follow up (%), median (IQR)	13.7 (8.4 to 19.6)	16.8 (7.5 to 36)	16.6 (12.4 to 29.1)	0.346
Spinal height (T1 to S1)
T1 to S1 preoperative (cm), median (IQR)	34.2 (31.9 to 35.6)	31.2 (29 to 35.1)	35.6 (32.9 to 36.9)	0.008
T1 to S1 recent follow up (cm), median (IQR)	39.3 (37.1 to 40.1)	36.1 (33.9 to 38.7)	39.6 (37.1 to 43.9)	<0.0005
T1 to S1 pre-operative to recent follow up (cm), median (IQR)	4.3 (3.7 to 6.4)	5.0 (2.7 to 8.7)	5.0 (3.2 to 8.2)	0.736
T1 to S1 pre-operative to recent follow up (%), median (IQR)	12.8 (10.1 to 19.7)	16.3 (7.8 to 29.7)	14.1 (9.3 to 22.9)	0.620

Health-related quality of life

EOSQ-24 domains, median (IQR)	VBT (n=10)	MCGR (n=30)	PSF (n=16)	P value
General health pre-operation	75 (75 to 87.5)	75 (75 to 87.5)	81.3 (65.6 to 87.5)	0.995
General health postoperation	62.5 (62.5 to 87.5)	75 (62.5 to 100)	93.8 (65.6 to 100)	0.103
Pre- to post-operative within surgical group	0.211	0.460	0.241
Pain/discomfort pre-operation	56.3 (37.5 to 62.5)	75 (50 to 100)	75 (62.5 to 100)	0.041
Pain/discomfort post-operation	75 (50 to 100)	75 (50 to 100)	75 (50 to 96.9)	0.963
Pre- to post-operation within surgical group	0.023	0.381	0.306
Pulmonary function pre-operation	93.8 (87.5 to 100)	100 (75 to 100)	100 (87.5 to 100)	0.734
Pulmonary function post-operation	100 (75 to 100)	100 (87.5 to 100)	100 (87.5 to 100)	0.889
Pre- to post-operation within surgical group	1.000	0.624	0.828
Transfer pre-operation	100 (50 to 100)	100 (100 to 100)	100 (100 to 100)	0.108
Transfer post-operation	100 (100 to 100)	100 (72 to 100)	100 (100 to 100)	0.155
Pre- to post-operation within surgical group	0.313	0.065	1.000
Physical function pre-operation	95.8 (83.3 to 100)	100 (83.3 to 100)	100 (93.8 to 100)	0.315
Physical function post-operation	100 (91.7 to 100)	95.8 (75 to 100)	100 (93.8 to 100)	0.178
Pre- to post-operation within surgical group	0.500	0.661	1.000
Daily living pre-operation	100 (100 to 100)	100 (68.8 to 100)	100 (100 to 100)	0.013
Daily living post-operation	100 (87.5 to 100)	100 (75 to 100)	100 (100 to 100)	0.028
Pre- to post-operation within surgical group	0.250	0.555	1.000
Fatigue/energy pre-operation	87.5 (62.5 to 87.5)	100 (75 to 100)	100 (75 to 100)	0.184
Fatigue/energy post-operation	75 (62.5 to 100)	71.9 (75 to 100)	93.8 (87.5 to 100)	0.241
Pre- to post-operation within surgical group	0.816	0.140	0.725
Emotion pre-operation	75 (50 to 75)	62.5 (62.5 to 87.1)	81.3 (65.6 to 87.5)	0.102
Emotion post-operation	93.8 (75 to 100)	81.3 (50 to 100)	87.5 (75 to 100)	0.280
Pre- to post-operation within surgical group	0.020	0.062	0.090
Parental impact pre-operation	70 (55 to 80)	75 (65 to 95)	90 (80 to 95)	0.017
Parental impact post-operation	87.5 (75 to 95)	90 (75 to 96.3)	100 (85 to 100)	0.094
Pre- to post-operation within surgical group	0.020	0.206	0.005
Child satisfaction pre-operation	75 (50 to 75)	75 (50 to 100)	100 (75 to 100)	0.037
Child satisfaction post-operation	75 (75 to 100)	75 (50 to 100)	100 (75 to 100)	0.203
Pre- to post-operation within surgical group	0.219	0.873	1.000
Parent satisfaction pre-operation	62.5 (50 to 75)	75 (50 to 100)	100 (75 to 100)	0.107
Parent satisfaction post-operation	75 (75 to 100)	75 (50 to 100)	100 (75 to 100)	0.067
Pre- to post-operation within surgical group	0.063	0.563	0.781

Post-operative assessment was on average 3.0±1.3 years after index surgery.

Key safety findings

Complications

	VBT (n=37)	MCGR (n=51)	PSF (n=42)	P value
Complications	27.0% (n=10)	60.8% (n=31)	14.3% (n=6)	<0.0005
# complications	15	45	9
Minor complications	10.8% (n=4)	17.6% (n=9)	4.8% (n=2)	0.151
# minor complications	5	11	3
Intra-operative complication	0	5.9% (n=3)	4.8% (n=2)	0.443
# intra-operative complication	0	3	2
Planned surgeries	8.1% (n=3)	31.4% (n=16)	0	<0.0005
# planned surgeries	3	16	0
Unplanned surgeries	16.2% (n=6)	21.6% (n=11)	7.1% (n=3)	0.154
# unplanned surgeries	7	17	4
Hardware issue	13.5% (n=5)	19.6% (n=10)	2.4% (n=1)	0.023

Time to planned or unplanned surgery

	Time to unplanned surgery (n=126 with 20 events)		Time to planned or unplanned surgery (n=126 with 36 events)
	HR* (95% CI)	P value	HR *(95% CI)	P value
VBT	4.5 (0.8 to 24.5)	0.084	7.1 (1.4 to 36.4)	0.019
MCGR	5.6 (1.1 to 28.4)	0.038	21.0 (4.8 to 92.5)	<0.001
PSF (reference)	1.00		1.00

* HR adjusted for age, gender and pre-operative major cobb

Reasons for unplanned revisions:

VBT: curve progression (n=3), hardware failure (n=3) and pulmonary complication (n=1)
MCGR: hardware migration (n=9), hardware failure (n=5), removal of hardware because of pain (n=2) and infection (n=1)
PSF: hardware failure (n=1), infection (n=3), removal of hardware because of pain (n=1) and re-instrumentation after pseudarthrosis (n=1)

Study 4 Abdullah A (2021)

Study details

Study type	Case series (Paediatric spine study group registry)
Country	Canada (3 centres)
Recruitment period	2015 to 2018
Study population and number	n=120 Patients with idiopathic scoliosis
Age and sex	Mean 12.6 years; 89% female; median Risser 1
Patient selection criteria	Inclusion criteria: idiopathic scoliosis, immature skeleton with Risser score 0 to 3, Cobb angle 40^° to 70^°, and main thoracic curve.
Technique	All patients had a thoracoscopic approach using 4 (64.1%) or 5 (35.9%) portals. Dynesys instrumentation (Zimmer Biomet Company, Warsaw, Indiana) was used for all surgeries. The most frequent upper instrumented vertebra was T5 (55.8%), T6 (35.8%), T7 (6.7%), and T4 (1.7%). The most frequent lowest instrumented vertebra was T12 (59.2%), T11 (23.3%), L1 (14.2%), L2 (2.5%), and T10 (0.8%).
Follow up	2 years
Conflict of interest/source of funding	Conflict of interest: 2 authors had nothing to disclose and other declared conflicts of interest. Funding: none.

Analysis

Follow-up issues: A total of 175 patients were treated with anterior VBT and 166 of these had the potential for 2-year follow up. Complete data was available for 120 patients who had main thoracic curve and only thoracic tether were included in this study.

Study design issues: This study determined perioperative morbidity associated with anterior VBT for idiopathic scoliosis. This study reported intraoperative and postoperative efficacy outcomes and follow-up data included complications, scoliosis curves and kyphosis measurements at 1 and 2 years. Unplanned return to the operating room was defined as a postoperative complication that could not be treated without an additional anaesthesia.

Study population issues: At baseline, the median Risser stage was 1 (Risser 0: 47.5%, Risser 1: 25.0%, Risser 2: 16.7% and Risser 3: 10.8%). The mean arm span was 162.0 cm (95% CI 159.2 to 164.8 cm). 99.2% of patients had a right thoracic curve and 0.8% of patients had a left thoracic curve.

Key efficacy findings

Number of patients analysed: 120

Perioperative data

Anterior VBT	Number of patients	Mean±SD	Range	95% CI
Surgical time (minutes)	99	215.3±78.9	111 to 472	199.6 to 231.0
Anaesthesia time (minutes)	86	303.5±76	207 to 480	287.2 to 319.8
Estimated blood loss (ml)	101	200±135.3	20 to 700	173.2 to 226.7
Estimated blood loss %	100	6.4±4	0.6 to 20	5.6 to 7.2
Postoperative hospital stay (days)	103	4.5±1.3	2 to 9	4.3 to 4.8
ICU stay (days)	73	0.2±0.5	0 to 2	0.1 to 0.3
Upper instrumented vertebra (UIV)	T5: 55.8%	T7: 6.7%	T6: 35.8% (minimum)	T4: 1.7%
Lower instrumented vertebra (LIV)	T12: 59.2%	L2: 2.5%	T11: 23.3% to L1: 14.2%	T10: 0.8%

Radiographic data

	Number of patients	Mean±SD	Range	95% CI
Pre-op scoliosis (^°)	112	51.2±7.8	40 to 70	49.7 to 52.7
Pre-op bending (^°)	104	32±11.5	9 to 63	29.8 to 34.2
Curve flexibility (%)	96	38.2±17.8	6.1 to 77.5	34.6 to 41.8
Post-op scoliosis (^°)	120	26.9±8.9	6 to 53	25.3 to 28.5
1-year post-op scoliosis (^°)	118	23.0±10.4	-11 to 50	21.1 to 24.8
2-year post-op scoliosis (^°)	104	27.5±11.6	-5 to 52	25.2 to 29.7
Pre-op global kyphosis (^°)	113	28.5±11.2	2 to 64	26.4 to 30.6
Post-op global kyphosis (^°)	120	27.4±11.4	2 to 56	25.4 to 29.5
1-year post-op global kyphosis (^°)	118	28.7±12.2	6 to 65	26.5 to 31.0
2-year post-op global kyphosis (^°)	94	29.2±12.5	6 to 67	26.6 to 31.8
Pre-op T5 to T12 kyphosis (^°)	113	16.0±11	-23 to 52	13.9 to 18
Post-op T5 to T12 kyphosis (^°)	120	16.9±10.8	-7 to 44	14.9 to 18.8
1-year post-op T5 to T12 kyphosis (^°)	117	17.5±11.8	-14 to 61	15.3 to 19.6.0
2-year post-op T5 to T12 kyphosis (^°)	79	17.0±11.8	-10 to 50	14.4 to 19.7
Pre-operative standing height (cm)	118	154.4±8.3	137 to 179	152.9 to 155.9
Post-operative standing height (cm)	112	156.1±8.5	137.7 to 180.9	154.5 to 157.7

Preoperative mean main thoracic scoliosis was 51.2^° (95% CI 49.7^° to 52.7^°) with correction to 32.0^° (95% CI 29.8^° to 34.2^°) on bending radiographs and mean curve flexibility of 38.2% (95% CI 34.6% to 41.8%). The mean global kyphosis was 28.5^° (95% CI 26.4^° to 30.6^°) and the mean T5 to T12 kyphosis was 16.0^° (95% CI 13.9^° to 18.0^°).

Postoperative compared with preoperative standing height: t=14.4, df=110, p<0.01

Scoliosis:

Immediate postoperative compared with preoperative standing radiographs: t=30, df=111, p<0.01
Immediate postoperative compared with preoperative bending radiographs: t=5.1, df=103, p<0.01
1-year compared with immediate postoperation: t=6.6, df=117, p<0.01
2-year compared with immediate postoperation: t=-0.47, df=103, p=0.64

Global kyphosis: 2 years after surgery compared with preoperation: t=0.35, df=110, p=0.73

T5 to T12 kyphosis: 2 years after surgery compared with preoperation: t= -0.36, df=78, p=0.72

Key safety findings

Complication data

Complication	Description	Classification according to		Unplanned return to operative room
Complication	Description	Smith	Modified Clavien-Dindo-Sink system	Unplanned return to operative room
Early complications that developed during hospitalisation
Pneumothorax	Developed on POD 3 and patient was transferred back to the ICU for 1 day, treated with BiPAP and Chest tube. The complication almost completely resolved after 3 days, and the chest tube was removed (POD 6) and patient discharged the day after (POD 7).	1	4a	No
Complications that developed after discharge and within 90 days from surgery
Superficial wound infection	Developed after 1 week. Treated as outpatient with local wound care and oral antibiotics. Resolved by 2 weeks post-operative.	1	2	No
Pneumonia	Developed 2 weeks after surgery. The patient visited the ER with shoulder pain. CXR showed small right sided atelectasis with small effusion, treated as outpatient with oral antibiotics. Resolved by 6 weeks postoperation.	1	2	No
CSF leak	It was discovered when the patient presented with acute headaches at 3 weeks postoperation. The patient was managed with blood patch injection treatment as an inpatient. Resolved.	1	3	No
Pleural effusion	Discovered 3 weeks postoperatively. Patient was admitted and managed with chest tube for 4 days during and oral antibiotics. Resolved.	1	3	No
Pleural effusion	Discovered 2 months postoperatively. Patient was admitted for 1 week and managed with chest tube and antibiotics. Resolved.	1	3	No
Late complications that developed between 3 months and 1 year
CSF leak	Discovered 4 months after surgery when the patient presented with headaches. Treated with blood patch injection in the clinic. The condition did not resolve. It was discovered to have screw threads in the spinal canal. The patient underwent planned surgical revision the screw and a lumbar drain installed for the CSF leak. Resolved.	2a	3	Yes
Right arm and shoulder paraesthesia	Developed 6 months after surgery with paraesthesia and neuropathic pain in the right arm and breast area. Symptoms progressed to sharp pain along her arm. Not resolved by 1-year follow up.	1	2	No
Right arm and shoulder paraesthesia	Right shoulder and posterior arm paraesthesia with pain inferior to right clavicle developed 1 year after surgery. Treated as an outpatient and not yet resolved by 1 year postoperative.	1	2	No
Compensatory curve add on	Lumbar curve add on, initially observed, then surgical management was chosen in the form of tethering the lumbar curve.	2a	3	Yes
Late complications 1 to 2 years postoperation
Curve progression	Moderate pain in the lumbar spine and left hip developed around 1 year after the initial surgery which was thought to be a result of lumbar curve progression. The patient underwent PSFI.	3	3	Yes
Keloid	Developed keloid and underwent scar revision and steroid injection.	2a	3	Yes
Right leg weakness	Experienced weakness in the right leg, with right knee give away, experienced pain in the medial aspect of her right thigh and was unable to move her right leg at certain times and has been to the ER as a result of these complications, scheduled for MRI.	1	2	No
Cable failure	Discovered broken tether at 2-year follow-up (2 years and 3 months) from surgery, the tether was broken at three different places and the patient underwent tether replacement.	2a	3	Yes
Cable failure	The tether cable failed at T8 to T9, however the patient was treated conservatively as the curve did not progress.	1	2	No
Cable failure	The tether cable failed at T10 to T11 at 2 years follow up (2 year and 2 months from surgery), leading to curve progression, patient treated with PSF.	3	3	Yes
Cable failure	The tether cable broken between T10 and T11 discovered during 2-year follow up (2 years and 4 months from surgery) and was treated with observation.	1	2	No
Overcorrection	Overcorrection and progression discovered at 2-year follow up (2 years and 4 months from surgery), requiring cable removal. The patient underwent cable removal which was complicated by haemothorax of 1,700 ml as a result of bleeding from T4 segmental artery which required return to the OR to control the bleeding.	3	3	Yes
Overcorrection	Overcorrection discovered during follow up (1 year and 6 months), treated with removal of the cable.	3	3	Yes

Study 5 Baroncini A (2021)

Study details

Study type	Case series
Country	US (single centre)
Recruitment period	2017 to 2019
Study population and number	n=90 Patients with adolescent idiopathic scoliosis having single or bilateral VBT
Age and sex	Mean 14.6 years; 89% (80/90) female; mean Risser 2.3
Patient selection criteria	Inclusion criteria: patients diagnosed with adolescent idiopathic scoliosis and showing flexible curves (<30^o on bending or traction X-rays). Exclusion criteria: patients with radiculopathy or other neurological symptoms deriving from spine pathology and/or previous spine surgery uncontrolled chronic disease; infections; malignancy; pregnancy; any blood anomalies; immunodeficiency; and other omitted criteria that could influence the results of this study.
Technique	Under general anaesthesia, patients were positioned on the side with the convex side of the curve facing up. Thoracic curves down to L1 were approached with video-assisted thoracic surgery, with 1 or 2 intercostal incisions, and 1 or 2 thoracoscopic portals. Lumbar curves were instrumented through a mini-retroperitoneal approach. For patients needing a bilateral correction, this was conducted in 1 stage with lumbar instrumentation performed first. After suturing and dressing the wounds, patients were repositioned for thoracic instrumentation. The screws were placed bicortically following anatomic landmarks and under antero-posterior fluoroscopic control. The screws were connected with a polyethylene tether and curve correction was performed. Lastly, a chest drainage was placed and removed when the output was less than 200 ml over 24 hours. For some patients, a double tether was added in lumbar curves to prevent tether ruptures or disc releases were done at the apex of thoracic curve to increase flexibility
Follow up	Range 6 to 24 months
Conflict of interest/source of funding	Conflict of interest: 2 authors, none; one author, consultant, paid lectures – globus medical. Funding: none

Analysis

Follow-up issues: VBT was done in 91 patients and 1 patient was excluded from the study: for 1 girl with bilateral scoliosis, after successful instrumentation of the lumbar curve, anterior, dynamic correction of the thoracic curve had to be abandoned because of diffuse pleural scarring and a second-time thoracic spondylodesis was done instead. Thus, 90 patients were included in the analysis.

Study design issues: This study investigated the intraoperative data and complications of the first 90 patients who had VBT, with the aim of defining the learning curve. This study was conducted according to the strengthening the reporting of observational studies in epidemiology: the STROBE statement.

All surgeries were done by 1 spinal deformity, (US) fellowship trained senior surgeon who had limited previous experience with anterior approaches to the thoracic and lumbar spine.

Study population issues: Of the 90 patients, most were skeletally immature (Risser ≤4 and/or Sanders ≤7) - Risser 2.3±1.7, Sanders 5.1±2. The mean height was 161±10 cm, the mean weight was 52.7±9.8 kg and mean BMI 20.3±3.11 kg/m². Before surgery, the mean Cobb angle of the instrumented curves was 55.7°±13.5°.

Other issues: The main limitation was the bias created by performing disc releases in selected patients. This study has not yet reached a plateau in intubation time, surgical duration and hospitalisation length so a longer observation period would be useful to measure the plateau values for these parameters. Radiologic data was not analysed, as no influence of the learning curve on scoliosis correction was observed. This was probably because of the heterogeneity of the treated curves. Further radiologic studies with a longer follow up are needed to analyse the results of VBT.

Key efficacy findings

Number of patients analysed: 90

First 20 patients (8 thoracic, 4 lumbar and 8 double instrumentations):

The mean Cobb angle of the instrumented curves was 58°±11° before surgery and 24°±13° at the first standing X-ray.

Last 20 patients (5 thoracic, 4 lumbar and 11 double instrumentations):

The mean Cobb angle was 58°±11° before surgery and 24°±12° at the first standing X-ray.

Within 6 months from surgery, none of the patients needed pain medication and all had returned to the daily activities and sports they were participating to before VBT.

Mean intubation time per screw: 33.1±7.6 minutes (range 15.4 to 61.6 minutes)

First 20 patients: 439.2±52.8 minutes
Last 20 patients: 358.4±83.4 minutes
p=0.0007

Mean surgical duration per screw: 21.3±5.7 minutes (9.4 to 44.8 minutes)

First 20 patients: 390±267.3 minutes
Last 20 patients: 163±57.7 minutes
p=0.0006

Mean estimated blood loss per screw: 21.3±18.2 ml (range 100 to 6.6 ml)

First 20 patients: 286.5±86 ml
Last 20 patients: 188.8±54.3 ml
p=0.0001

Four patients had a transfusion from cell-salvage and 3 blood transfusions were done: all these patients were among the first 10 patients.

Mean hospitalisation length: 8.3±3.1 days (range 4 to 14 days)

First 20 patients: 9.3±2.1 days
Last 20 patients: 7.8±1.6 days
p=0.01

There was evidence of a negative correlation between growing number of patients and the intubation time (ρ= -0.57; p<0.0001), surgical duration (ρ= -0.55; p<0.0001), total estimated blood loss (ρ= -0.66; p<0.0001), hospitalisation length (ρ= -0.32; p=0.002).

Key safety findings

No intraoperative complications, neuromonitoring anomalies or screw misplacements were reported.

Pulmonary complications within 6 weeks after surgery: n=6 (5 complications happened after double VBT and 1 after thoracic VBT)

Recurrent, right-side pleural effusion at 2 to 6 weeks after surgery: n=3

Two patients had chest-tube reinsertion and 1 had exploratory thoracotomy because of a bleeding from the right pulmonary ligament, but this lesion could not be repaired, and a chest-tube was reinserted.

Chest infection: n=1

This patient was admitted to the local hospital 2 weeks after surgery because of fever and dyspnoea and had antibiotics for 2 weeks.

Minor pulmonary embolism after a 24-hour flight: n=1

This patient had intramuscular low-molecular-weight heparin therapy for 1 month.

Persistent atelectasis of the lower left lobe: n=1

This happened on the second postoperative day after thoracic VBT from the right side. Since she did not tolerate non-invasive ventilation, intubation was needed for 3 days. The symptoms resolved after bronchoscopic removal of a large mucus accumulation.

All patients recovered well and without sequelae. Four of the reported complications were observed within the first 25 operated patients. No further complications were observed beyond 6 weeks after surgery.

Tether rupture: n=28

Because of a high rate of tether rupture, double tether was adopted and most in lumbar curves. Of the 28 patients, 3 patients needed revision surgery for loss of correction, the second tether might provide more stability to the construct. Data regarding the rupture rate with double tether was not available.

Study 6 Rushton PRP (2021)

Study details

Study type	Case series
Country	Canada (2 centres)
Recruitment period	2012 to 2018
Study population and number	n=112 (116 primary tethering procedures) Patients with idiopathic scoliosis
Age and sex	Mean 12.7 years; 93% (104/112) female; mean Risser 0.5
Patient selection criteria	Inclusion criteria: skeletally immature patients with progressive major main thoracic and/or lumbar curves ≥40^o
Technique	Thoracic tethers were done thoracoscopically and thoracolumbar/lumbar tethers needed a mini-open approach. L1 and commonly L2 were accessed via a rib sparing thoracotomy with incision of the posterior aspect of the diaphragm and development of a plane anterior to psoas to allow instrumentation. Lower lumbar levels were accessed through a second incision via an anterolateral retroperitoneal approach. Single screws (Zimmer Dynesys, Winterthur, Switzerland) were placed in each vertebra aside from in double tethers when the inflection vertebra was typically instrumented from both sides. Levels were typically instrumented Cobb to Cobb and tether tensioned to bring the tilted discs into neutral alignment where possible.
Follow up	Mean 37 months (range 15 to 64 months)
Conflict of interest/source of funding	Funding: none Relevant financial activities outside the submitted work: consultancy, grants, royalties.

Analysis

Follow-up issues: A total of 114 patients with idiopathic scoliosis had tethering procedures, 2 were lost to follow up and excluded from the study. Two patients had fusion procedures within 2 years post anterior VBT (15 and 19 months) hence excluded from further follow up, the remaining patients had a follow up of more than 2 years.

Study design issues: This study evaluate the clinical, radiographic, and perioperative outcomes and complication rates to determine the efficacy of anterior VBT in a prospective multicentre cohort of skeletally immature patients with idiopathic scoliosis. Cases were considered a success if at follow up they had not undergone (or awaiting) fusion and their tethered curve(s) was <35°. In addition to conventional radiographic measures the angulation between upper and lower instrumented vertebra was measured in coronal and sagittal planes and termed 'instrumented Cobb'.

Study population issues: Of the 104 females, 22 were recently postmenarchal. All patients were skeletally immature, Risser 0.5±0.9 (range 0 to 3). 51% of patients had previous bracing treatment.

Of the 116 tethering procedures 108 were done as a single stage, of which 104 were thoracic tethers and 4 were lumbar tethers. Eight tethers were done on 4 patients via a planned staged approach undergoing a lumbar tether followed by a thoracic tether 1 to 2 days later. Overall, 108 thoracic tethers were done and 8 lumbar tethers.

Other issues: This study had several limitations: 1) preoperative Sanders score was only available for 33 patients; 2) lumbar and thoracic tethers were analysed together, so further work is needed to determine if their responses to anterior VBT differ; and 3) there were no patient reported outcome measures.

Key efficacy findings

Number of patients analysed: 112 (116 procedures)

Operative data

Variable		Value	P
Vertebrae tethered	Thoracic tether	7.3±0.7 (6 to 9)	<0.001
Vertebrae tethered	Lumbar tether	5.3±0.5 (5 to 6)	<0.001
Operating time (minutes)	Thoracic tether	232±102 (110 to 585)	0.03
Operating time (minutes)	Lumbar tether	295±65 (163 to 387)	0.03
Blood loss (ml/kg)	Thoracic tether	5.0±3.3 (0.5 to 18.8)	0.61
Blood loss (ml/kg)	Lumbar tether	5.1±2.5 (1.3 to 9.4)	0.61
Length of stay (days)	Single stage thoracic/lumbar	4.7±1.4 (3 to 11)	0.001
Length of stay (days)	Planned 2 stage	10.5±4.0 (7 to 16)	0.001

Radiographic and surface measurements

	Preoperation	First erect	1 year	Most recent follow up
Coronal plane
Tethered curve Cobb (^o)	50.8±10.2 (31 to 81)	26.6±10.1 (-3 to 61)^a	23.1±12.4 (-37 to 57)^b	25.7±16.3 (-32 to 58)^cd
Tethered curve correction (%)		47.7±16.2 (7 to 107)	55.1±22.7 (5 to 184)^b	49.6±30.5 (-16 to 159)^c
Instrumented Cobb (^o)		22.7±10.6 (-1 to 57)	15.1±12.6 (-18 to 55)^b	19.0±15.8 (-29 to 57)^c
Untethered minor curve Cobb (^o)	31.0±9.5 (3 to 57)	20.3±10.3 (0 to 52)^a	18.1±10.8 (-22 to 50)^b	18.4±14.2^e (-13 to 62)^d
Untethered minor curve correction (%)		34.3±33.8	43.5±31.3^b	41.6±60.9
Sagittal plane
Thoracic kyphosis T5 to T12 (^o)	18.6±11.4 (-8 to 47)	18.8±11.8 (-12 to 45)	18.6±12.3 (-14 to 55)	21.4±13.0 (-14 to 66)^cd
Lumbosacral lordosis L1 to S1 (^o)	-55.9±10.5 (-99 to -28)	-54.0±10.9 (-88 to -30)^a	-55.7±10.9 (-87 to -24)^b	-56.2±11.4 (-83 to -24)
Thoracic tether instrumented sagittal Cobb (^o)		16.6±12.2 (-12 to 50)	16.7±13.0 (-13 to 54)	17.1±13.1 (-12 to 51)
Lumbar tether instrumented sagittal Cobb (^o)		-10.5±12.9 (-35 to 5)	-10.8±11.5 (-24 to 6)	-8.8±10.2 (-20 to 9)
Surface measurements
Rib hump (^o)	14.1±4.8 (0 to 26)	8.1±4.3 (0 to 22)^a	8.6±4.7 (0 to 25)	8.8±5.4 (0 to 22)^d
Lumbar prominence (^o)	3.6±4.7 (0 to 17)	2.2±3.6 (0 to 16)^a	2.4±3.6 (0 to 15)	2.5±4.4 (0 to 18)^d

Changes from preoperation to first erect: ^adenotes significance; tethered curve Cobb / tethered curve correction/ untethered minor curve Cobb/ rib hump all p<0.001, lumbosacral lordosis p=0.03 (non-significant: thoracic kyphosis p=0.58, lumbar prominence p=0.052).
Change from first erect to 1 year: ^bdenotes significance; tethered curve Cobb/ tether curve correction/ instrumented Cobb/untethered minor curve Cobb/ untethered minor curve correction all p<0.001, lumbosacral lordosis p=0.04 (non-significant: thoracic kyphosis p=0.50, thoracic tether instrumented sagittal Cobb p=0.80, lumbar tether instrumented sagittal Cobb p=0.67, rib hump p=0.40, lumbar prominence p=0.91).
Change 1 year to most recent follow up: ^cdenotes significance; tethered curve Cobb/ tethered curve correction / instrumented Cobb all <0.001, thoracic kyphosis p=0.002 (non-significant; untethered minor curve Cobb p=0.37, untethered minor curve correction p=0.60, thoracic kyphosis p=0.26, lumbosacral lordosis p=0.45, thoracic tether instrumented sagittal Cobb p=0.80, lumbar tether instrumented sagittal Cobb p=0.14, rib hump p=0.55, lumbar prominence p=0.52).
Change preoperation to most recent follow up: ^ddenotes significance; tethered curve Cobb/ untethered minor curve Cobb/ rib hump all p<0.001, thoracic kyphosis p=0.004, lumbar prominence p=0.03 (non-significant; lumbosacral lordosis p=0.86).
^e1 case had delayed lumbar curve tethering between 1 year and MRF, thus excluded and n=107 for this time point.

Comparison of those considered successful and unsuccessful

		Success (n=80)	Failure (n=32)	P
Preoperative
Age at surgery		12.7±1.5 (8.2 to 16.7)	12.5±1.0 (10.2 to 14.7)	0.54
Sex (% female)		94%	91%	0.56
Females premenachal (%)		77%	82%	0.52
Mass		45.2±11.1 (24.5 to 81.1)	44.8±11.6 (30.8 to 78.0)	0.32
BMI		18.9±3.7 (12.9 to 31.7)	18.9±3.5 (14.1 to 27.7)	0.92
Risser	Mean	0.6±0.9 (0 to 3)	0.4±0.8 (0 to 3)	0.29
Risser: number of cases	0	53	25	0.66
	1	13	2
	2	10	3
	3	4	1
Sanders score (n=33)		3.5±1.3 (2 to 6)	2.8±0.8 (2 to 4)	0.69
Major coronal Cobb		48.3^o±8.9 (31 to 70)	56.5^o±11.1 (35 to 81)	<0.001
Major Cobb group (% success for group)	30 to 39^o	10 (83%)	2	0.03
	40 to 49^o	36 (78%)	10
	50 to 59^o	25 (73%)	9
	>60^o	9 (45%)	11
Flexibility		42.3%±19.2 (0 to 100)	34.6%±20.9 (2 to 91)	0.046
Intraoperative
Tether location
Thoracic		74	30	N/A
Lumber		3	1
Dual		3	1
Postoperative
Tether breakage^e		19/80 (24%)	17/32 (53%)	0.003
Follow up (months)		35.6±8.7 (24 to 61)	41.7±10.9 (15 to 64)	0.001

^eThree patients had confirmed tether breakage (2 verified at tether replacement and 1 at fusion) and 33 patients had radiographs suggestive of tether breakage (2 of these had 2 sites of suspected breakage). Thus 32% of patients (36/112) had a confirmed/suspected tether breakage. The most common sites for breakage were T9/10, T10/11 and T11/12 occurring in 11,13 and 7 cases respectively. Tether breakage was noted on radiographs taken mean 31 months (12 to 43 months) postoperatively.

Key safety findings

Overall complications: 22% (n=25) of patients who had 28 complications.

Revision: 13% (n=15) of patients who needed 18 revision operations.

Complications and revision further operations

Nature of complication/revision operation			Number cases	Number revision operations
Perioperative	Pulmonary	Atelectasis (needing admission to intensive care for respiratory support)	4
		Haemothorax	2	1
		Pneumonia	2
		Pneumothorax	1
	GI	Clostridium difficile infection	1
	Infection	Superficial wound	1
	CSF leak (presenting with orthostatic headaches and vomiting)		2*	1
Prompting revision of tether**	Overcorrection (loosening tether)		5	5
	Tether breakage (replaced)		2	2
	Adding on (extension of tether)		1	1
Prompting fusion	Inadequate curve correction with no apparent tether breakage		3	3
	Inadequate correction tethered curve with tether breakage +/- fusion of progressive untethered lumbar curve		2	3***
	Adding on		2	2

* One related to a T12 screw narrowly breaching the posterior wall needing revision and the other without obvious cause was treated conservatively.

** At follow up, 6 patients had fusion operations with 1 awaited (6%). One patient subsequently needed revision for distal junctional failure.

***One patient needed subsequent revision of fusion for distal junctional failure.

Study 7 Miyanji F (2021)

Study details

Study type	Case series (retrospective; multicentre scoliosis registry)
Country	Canada (2 centres)
Recruitment period	2013 to 2017
Study population and number	n=50 Patients with idiopathic scoliosis
Age and sex	Mean 11.9 years; 92% (46/50) female
Patient selection criteria	Inclusion criteria: Patients with Lenke type 1 and Lenke type 2 juvenile idiopathic scoliosis/adolescent idiopathic scoliosis who had anterior VBT; a minimum of 2-year follow up. Exclusion criteria: patients who transitioned to primary fusion before 2 years postoperatively.
Technique	Anterior VBT was done.
Follow up	Mean 2.1 years
Conflict of interest/source of funding	None

Analysis

Follow-up issues: A total of 50 patients were included in the final case series. Although 55 patients with Lenke type 1 or 2 idiopathic scoliosis who had anterior VBT with 2-year follow up were identified, 5 patients were not included in the analysis. The reasons for exclusion were because of inadequate view of the shoulders on radiographs taken at 2-year follow up (n=4) and conversion to PSF before the 2-year time point (n=1) to address the remaining deformity.

Study design issues: This retrospective case series evaluated shoulder balance following anterior VBT of the spine for Lenke type 1 and 2 idiopathic scoliosis and reported the prevalence of postoperative shoulder imbalance in patients having anterior VBT for idiopathic scoliosis.

The primary outcome measure was shoulder balance assessed at the immediate, 1-year (10 to 18 months), and 2-year (22 to 30 months) postoperative follow-up visits. Radiographic shoulder height was used as the primary radiographic parameter to assess shoulder balance. Radiographic shoulder height was defined as the linear distance between the superior horizontal reference line, which passes through the intersection of the soft-tissue shadow of the shoulder and a line drawn vertically up from the acromial clavicular joint of the cephalad shoulder, and the inferior horizontal reference line, constructed in a similar fashion over the caudal acromioclavicular joint.

Secondary outcome measures included other surrogate radiographic markers of shoulder balance, such as clavicular angle and T1 tilt angle assessed at the preoperative, immediate postoperative, 1-year, and 2-year postoperative follow-up visits. Clavicular angle was defined as the angle between a horizontal line and a tangential line connecting the most cephalad points of each clavicle. T1 tilt angle was defined as the angle between a horizontal line and a line along the upper endplate of T1. Coronal balance was determined and defined as the horizontal distance between the central sacral vertical line and the C7 plumb line (C7CSVL) in the standing posteroanterior radiograph.

Study population issues: Of the 50 patients, 43 (86%) patients had Lenke type 1 and 7 (14%) patients had Lenke type 2 curves. Preoperatively, absolute radiographic shoulder height averaged 15.6±10.5 mm. 35 (70%) had acceptable shoulder balance, 14 (28%) had moderate shoulder imbalance, and 1 (2%) had severe shoulder imbalance. All patients with shoulder imbalance preoperatively had right-sided elevation (i.e., radiographic shoulder height ≤ -2 cm). 4 patients with acceptable shoulder balance preoperatively had slight left-sided elevation (i.e., 1 cm ≤ radiographic shoulder height <2 cm). Risser 0 was in 66% (33/50) of patients, Risser 1 in 20% (10/50), Risser 2 in 12% (6/50) and Risser 3 in 2% (1/50).

Other issues: This study had several limitations. Although a patient may have balanced shoulders radiographically, an imbalance may exist clinically and vice versa. Unfortunately, clinical photos were not available to verify shoulder imbalance in this study and the results should be considered preliminary. It was also important to note that most patients (66%) were Risser 0 and 30% of these patients had open triradiate cartilages. It was difficult to predict curve behaviour in these particularly skeletally immature patients. Furthermore, when stratifying by Lenke type, the Lenke 2 group suffered from small sample size. Finally, all patients with preoperative shoulder imbalance included in this study suffered from a right-sided imbalance, so, the effect of anterior VBT on left-sided shoulder imbalance cannot be concluded from this study.

Key efficacy findings

Number of patients analysed: 50

Radiographic results

	Preoperative	Immediate postoperation	1-year postoperation	2-year postoperation
Radiographic parameter (mean±SD [min to max])
Proximal thoracic curve (^o)	22.8±8.8 (7.7 to 43.4)	16.8±8.8 (3 to 36)	12.4±8.1 (0.2 to 33.4)	13.1±7.5 (1.2 to 30)
Main thoracic curve (^o)	49.4±8.5 (32 to 75)	27.1±8.5 (14 to 52)	22.1±8.9 (4 to 41)	24.9±9.5 (6 to 46)
TL curve (^o)	32.7±6.9 (18 to 46)	25±7 (11 to 42)	21.4±7.6 (10 to 40)	21.3±7.9 (10 to 39)
*Clavicular angle (^o)	-1.9±3 (-7.8 to 3.6)	1±2.3 (-6.3 to 6.4)	1.5±2.4 (-4.1 to 6.5)	0.9±2.6 (-5 to 7.4)
IClavicular angle1 (^o)	2.8±2.2 (0 to 7.8)	1.9±1.6 (0 to 6.4)	2.3±1.7 (0 to 6.5)	2.2±1.7 (0 to 7.4)
*T1 angle (^o)	-1.1±5.9 (-12.4 to 10.7)	3.1±5.2 (-7.8 to 14.2)	2.4±4.8 (-8.9 to 12.1)	2.8±5.6 (-9.2 to 18.2)
IT1 angle1 (^o)	4.8±3.5 (0 to 12.4)	4.9±3.6 (0 to 14.2)	4.2±3.3 (0 to 12.1)	4.7±4.2 (0 to 18.2)
*Shoulder height (mm)	-12.3±14.2 (-41.1 to 13.4)	3.4±12.3 (-24.3 to 28.4)	6.9±12.2 (-24 to 29.6)	2.6±13.8 (-21.4 to 32.4)
IShoulder height1 (mm)	15.6±10.5 (0 to 41.1)	9.6±8.2 (0 to 28.4)	11.5±7.8 (0 to 29.6)	11.3±8.3 (0 to 32.4)
C7CSVL (mm)	6±17.2 (-33.6 to 45)	0.6±13.5 (-24.3 to 30.3)	1±12.9 (-23 to 27.5)	-0.9±14.4 (-37 to 35)
IC7CSVL1 (mm)	13.7±11.9 (0 to 45)	10.6±8.3 (0 to 30.3)	10±8 (0 to 27.5)	11.4±8.7 (0 to 37)
Shoulder balance
Acceptable balance ≤2 cm	70.0% (n=35)	82.0% (n=41)	70.0% (n=35)	84.0% (n=42)
2 cm ≤ moderate imbalance <4 cm	28.0% (n=14)	10.0% (n=5)	12.0% (n=6)	16.0% (n=8)
Severe imbalance ≥4 cm	2.0% (n=1)	0% (n=0)	0% (n=0)	0% (n=0)
Missing radiograph	0% (n=0)	8.0% (n=4)	18.0% (n=9)	0% (n=0)

*Measurement where negative value indicates right side up/left side down.

Radiographic results for Lenke type 1 compared with Lenke type 2 curves

	Lenke 1 (n=43), 86.0%)	Lenke 2 (n=7, 14.0%)	P value
Radiographic parameter (mean±SD)
Preoperative proximal thoracic curve (^o)	21.1±7.8	34.2±6.5	<0.001
2-year postoperative proximal thoracic curve	11.9±7.1	19.6±7	0.012
Preoperative main thoracic curve (^o)	49±9	54±8	0.171
2-year postoperative main thoracic curve	24.5±10	27.7±5.5	0.419
Preoperative TL curve (^o)	33±7	33±5	0.792
2-year postoperative TL curve	21±7.8	23.1±9.4	0.536
Preoperative Ishoulder height1 (mm)	3±2.3	2.4±1.8	0.571
2-year postoperative Ishoulder height1	2.2±1.6	2.6±2.5	0.514
Preoperative IClavicular angle1 (^o)	5±3.7	3.7±2.5	0.372
2-year postoperative IClavicular angle1	4.2±4	7.4±5	0.066
Preoperative IT1 tilt angle1 (^o)	15.7±11	14.8±7	0.829
2-year postoperative IT1 tilt angle1	10.9±7.9	13.7±10.8	0.419
Preoperative shoulder balance
Acceptable balance ≤2 cm	67.4% (n=29)	85.7% (n=6)	0.328
2 cm ≤ moderate imbalance <4 cm	30.2% (n=13)	14.3% (n=1)	0.657
Severe imbalance ≥4 cm	2.3% (n=1)	0% (n=0)	1.000
2-year postoperative shoulder balance
Acceptable balance ≤2 cm	86.0% (n=37)	71.4% (n=5)	0.310
2 cm ≤ moderate imbalance <4 cm	14.4% (n=6)	28.6% (n=2)	0.310
Severe imbalance ≥4 cm	0% (n=0)	0% (n=0)	-

In particularly skeletally immature patients (i.e., Risser 0 with open triradiate cartilage), mean |radiographic shoulder height| was 15.6±12.9 mm preoperatively and 12.0±10.1 mm at 2 years postoperatively (p=0.500). In more mature patients (i.e., Risser 0 with closed triradiate cartilage or > Risser 0), mean |radiographic shoulder height| was 15.6±10.0 mm preoperatively and 11.1±7.9 mm at 2 years postoperatively (p=0.028).

Key safety findings

Safety data were not reported.

Study 8 Alanay A (2020)

Study details

Study type	Case series (retrospective)
Country	Turkey
Recruitment period	2014 to 2018
Study population and number	n=31 Patients with adolescent idiopathic scoliosis
Age and sex	Mean 12.1 years; 94% (29/31) female; median Risser 0
Patient selection criteria	Inclusion criteria: patients having a Lenke 1 or 2 pattern, and more than or equal to a follow up of 12 months.
Technique	In lateral decubitus position, video-assisted thoracoscopy was done. Three visualisation and 3-to-4 working ports were used for T5 to L1. For Lenke 1Ar curves, L2 and L3 screws were inserted through a mini-open retroperitoneal approach. Hydroxyapatite-coated mono-axial screws were placed with a bicortical purchase. Staples were not used in first several patients as the necessary transportal insertion instruments were not available. While applying translation and derotation, the tether was tensioned. Slighter tension was applied at the UIV and UIV-1 disc and greater tension was applied at the apex. The movement of the screws during tensioning and the change in disc wedging guided the tensioning. Tension applied at the LIVþ1 and LIV disc was judged by the curve pattern under fluoroscopy and was mostly slight, sometimes even left slack. A chest tube was placed before closure.
Follow up	Mean 27.1 months (range 12 to 62 months)
Conflict of interest/source of funding	None

Analysis

Follow-up issues: Twenty-one patients (67.7%) had more than or equal to 2-year follow up. Four patients (12.9%) had a follow up of 18 months. Remaining 6 patients (19.4%) had a follow up of 12 months. Patients were followed up at the day before discharge (namely first erect), postoperatively at 6 weeks, at 3, 6, 12, 18 and 24 months, and each year thereafter.

Study design issues: This retrospective study reported the follow-up curve behaviours for patients in different Sanders groups who had similar curves preoperatively that were intraoperatively corrected with a similar strategy, to form a basis for clinical decision-making and surgical planning. Outcomes included perioperative data, radiographic and clinical measurements, overcorrection, pulmonary and mechanical complications, readmission, and reoperations.

To evaluate the height gained within the instrumented segment, UIV-LIV vertical height was measured. For both body height and UIV-LIV height, the difference from preoperative to 6 weeks was considered operative gain. The increase from 6 weeks onwards was considered follow-up gain. UIV-LIV follow-up height gain was divided by the number of instrumented levels to calculate gain-per-level.

Study population issues: Preoperatively, 14 patients were Lenke 1A, while 10 were 1B, 6 were 1Ar, and 1 was 2C. The median Sanders stage was 3 (1 to 7). The median Risser score was 0 (0 to 4). Of the female patients, 19 (66%) were premenarchal while the rest were on average 11.9±7.3 (1 to 24) months postmenarchal. All demographic and perioperative variables were similar between Sanders groups except age, Risser sign and postmenarchal status.

Key efficacy findings

Number of patients analysed: 31

A mean of 7.5±0.8 levels was tethered. UIV was T5 in 17 (54.8%) patients, while it was T6 in 11 (35.5%), T7 in 2 (6.5%) patients, and T8 in 1 (3.2%) patient. LIV was T11 in 13 (41.9%) patients, while it was T12, L1, L2, and L3 in 8 (25.8%), 6 (19.4%), 1 (3.2%), and 3 (9.7%) patients, respectively.

Sanders stage at the final follow up:

Sanders 7 or 8: 74.2% (n=23)
Sanders 6: 19.4% (n=6)
Sanders 4 and 3B: 6.1% (n=2, both had a revision surgery)

The median Risser score was 5 (2 to 5). Of the female patients, all but 2 (93%) were postmenarchal. Of the 10 patients that had less than 2-years follow up, 7 (70%) were Sanders 7, and the remaining 3 (30%) were Sanders 6, at their final follow up.

Preoperative and follow-up total body height and UIV-LIV vertical height measurements, and amounts of operative and follow-up gain in Sanders groups

	Sanders 2	Sanders 3	Sanders 4 to 5	Sanders 6 to 7
Height, cm, median (range)
Preoperative	145 (130 to 168)	157 (145 to 178)	160 (153 to 163)	157 (152 to 166)
6 weeks	148 (133 to 170)	158 (147 to 181)	161 (153 to 164)	158 (153 to 167)
Last follow up	162 (149 to 181)	164 (153 to 186)	164 (156 to 166)	159 (153 to 168)
Operative gain^a	2 (1 to 3)	2 (0 to 3)	1 (0 to 2)	1 (0 to 2)
Growth gain^b	13 (9 to 16)	5 (3 to 8)	3 (2 to 4)	1 (0 to 2)
UIV-LIV vertical height, mm, median (range)
Preoperative	153.8 (119 to 181)	152.0 (129 to 201)	142.4 (135 to 162)	151.6 (127 to 191)
6 weeks	157.9 (127 to 187)	159.1 (134 to 209)	145.4 (139 to 165)	157.3 (132 to 198)
Last follow up	177.5 (136 to 204)	169.1 (140 to 223)	150.7 (143 to 169)	159.0 (133 to 197)
Operative gain^c	4.0 (0.3 to 8)	7.2 (1.5 to 18)	3.1 (2 to 4.3)	5.7 (1.2 to 10)
Growth gain	19.7 (9.3 to 25.9)	10 (2.4 to 14.8)	5.3 (4.3 to 6.9)	1.7 (-1.3 to 4)
Growth gain per level^d	2.4 (1.3 to 3.7)	1.3 (0.3 to 2.1)	0.7 (0.5 to 1)	0.2 (-0.1 to 0.6)

^aX² (3) =5.355, p=0.148

^bX² (3) =25.231, p<0.001; Sanders 2 to 4/5, p=0.010; Sanders 2 to 6/7, p<0.001; Sanders 3 to 6/7, p=0.002

^cX² (3) =3.995, p=0.075

^dX² (3) =21.4, p<0.001; Sanders 2 to 4/5, p=0.038; Sanders 2 to 6/7, p<0.001; Sanders 3 to 6/7, p=0.00

Mean operative height gain: 1.5±0.9 cm for all patients

Mean UIV-LIV follow-up height gain: 9.2±7.1 mm for all patients

Preoperative and follow-up Cobb angle measurements for upper thoracic, main thoracic and thoracolumbar/lumbar curves, and bending flexibility, and operative, follow up and total correction percentages in Sanders groups

	Sanders 2	Sanders 3	Sanders 4 to 5	Sanders 6 to 7
Upper thoracic curve degree, median (range)
Preoperative	30 (14 to 37)	24 (14 to 55)	23 (17 to 44)	27 (20 to 31)
First erect	17 (6 to 28)	17 (8 to 37)	17 (15 to 36)	15 (6 to 27)
6 months	15 (5 to 27)	16 (3 to 40)	19 (12 to 32)	18 (15 to 29)
12 months	14 (4 to 18)	14 (3 to 31)	18 (13 to 33)	16 (5 to 27)
Last follow up	11 (1 to 19)	13 (2 to 31)	17 (12 to 36)	15 (5 to 24)
Follow-up correction	9 (4 to 17)	1 (-4 to 9)	5 (0 to 6)	1 (-3 to 5)
Percentage, median (range)
Bending flexibility %	77 (60 to 90)	78 (35 to 96)	59 (41 to 84)	70.5 (37 to 80)
Operative correction %^e	24 (20 to 71)	33 (19 to 73)	18 (10 to 55)	30 (0 to 70)
Follow-up correction %^f	26 (19 to 71)	3 (-13 to 50)	0 (0 to 29)	5 (-10 to 23)
Total correction %	70 (46 to 93)	44 (17 to 86)	29 (18 to 55)	44 (201 to 75)
Main thoracic curve degree, median (range)
Preoperative	46 (35 to 51)	50 (36 to 68)	43 (38 to 51)	47 (39 to 51)
First erect	24 (16 to 33)	24 (10 to 34)	21 (12 to 26)	20 (8 to 29)
6 weeks	17 (13 to 32)	27 (15 to 33)	20 (17 to 26)	24 (14 to 26)
6 months	15 (10 to 24)	22 (5 to 31)	19 (16 to 21)	22.5 (10 to 30)
12 months	10 (2 to 14)	16.5 (0 to 25)	17 (8 to 20)	19 (9 to 28)
Last follow up	-3 (-5 to 3)	17 (-6 to 280)	15 (6 to 20)	19 (9 to 27)
Follow-up correction	29 (18 to 30)	9 (-4 to 23)	6 (3 to 7)	0 (-2 to 6)
Percentage, median (range)
Bending flexibility %	80 (30 to 100)	71 (52 to 94)	82 (70 to 88)	70 (63 to 94)
Operative correction %^g	51 (13 to 63)	52 (27 to 80)	47 (40 to 72)	55 (40 to 83)
Follow-up correction %^h	59 (37 to 79)	17 (-0.8 to 48)	14 (6 to 18)	0 (-4 to 12)
Total correction %	109 (92 to 111)	71 (41 to 112)	63 (53 to 86)	55 (44 to 81)
Toracolumbar/lumbar curve degree, median (range)
Preoperative	26 (12 to 35)	32 (22 to 42)	28 (22 to 37)	32 (19 to 37)
First erect	13 (2 to 23)	16 (1 to 36)	15 (2 to 28)	17 (12 to 30)
6 months	8 (-5 to 6)	16 (0 to 29)	15 (1 to 29)	15 (9 to 24)
12 months	5 (-12 to 18)	11.5 (0 to 34)	15 (0 to 20)	14 (2 to 23)
Last follow up	-8 (-18 to 20)	4 (-5 to 31)	12 (-6 to 16)	14 (5 to 23)
Follow-up correction	13 (3 to 31)	11 (-5 to 23)	8 (-4 to 18)	6 (-1 to 12)
Percentage, median (range)
Bending flexibility %	150 (65 to 242)	100 (57 to 178)	89 (61 to 191)	105 (75 to 258)
Operative correction %ⁱ	63 (0 to 83)	50 (14 to 96)	32 (18 to 93)	37 (13 to 62)
Follow-up correction %^j	89 (9 to 142)	33 (-15 to 66)	27 (-14 to 53)	24 (-3 to 33)
Total correction %	131 (41 to 225)	85 (26 to 121)	57 (27 to 122)	50 (28 to 86)

^eX² (3) =2.357, p=0.502

^fX² (3) =6.103, p=0.107

^gX² (3) =0.357, p=0.949

^hX² (3) = 16.502, p<0.001; Sanders 2 to 6/7, p<0.001

ⁱX² (3) =1.952, p=0.582

^jX² (3) =6.295, p=0.098

Mean MT curve magnitude for all patients: preoperative, 47^o±7.6^o; first erect, 21.8^o±6.4^o

Preoperative and follow-up Cobb angle measurement for thoracic kyphosis, and lumbar lordosis, and amount of operative and follow-up change, and preoperative and last follow-up hump measurements and correction percentages in Sanders groups

	Sanders 2	Sanders 3	Sanders 4 to 5	Sanders 6 to 7
Thoracic kyphosis degree, median (range)
Preoperative	34 (17 to 59)	25 (15 to 41)	31 (28 to 34)	22 (15 to 48)
First erect	28 (5 to 57)	30 (12 to 44)	29 (18 to 32)	28 (14 to 43)
6 months	31 (10 to 37)	30 (15 to 44)	28 (19 to 32)	32 (21 to 44)
12 months	32 (15 to 35)	30.5 (21 to 39)	24 (20 to 37)	25 (14 to 41)
Last follow up	30 (6 to 40)	30 (21 to 46)	28 (18 to 33)	29 (14 to 38)
Operative change^k	-7 (-12 to -2)	0 (-15 to 11)	-3 (-10 to -2)	0 (-6 to 17)
Follow-up change^l	4 (-17 to 11)	3 (-15 to 12)	(-11 to 10)	0 (-6 to 6)
Lumbar lordosis degree, median (range)
Preoperative	55 (52 to 70)	61 (38 to 91)	61 (37 to 72)	65 (56 to 66)
First erect	52 (36 to 60)	48 (27 to 78)	51 (41 to 69)	53 (44 to 67)
6 months	56 (50 to 68)	55 (41 to 68)	61 (45 to 65)	55.5 (48 to 68)
12 months	53 (45 to 75)	56.5 (34 to 88)	52 (44 to 63)	60 (48 to 67)
Last follow up	53 (40 to 75)	55 (34 to 88)	52 (44 to 63)	60 (47 to 70)
Operative change^m	-10 (-18 to 4)	-9 (-21 to -2)	-4 (-19 to 4)	-8 (-18 to 1)
Follow-up changeⁿ	4 (-4 to 18)	0 (-9 to 11)	-3 (-10 to 17)	-2 (-3 to 15)
Hump degree, median (range)
Preoperative^o	13 (12 to 16)	12 (8 to 21)	13 (6 to 17)	12 (11 to 16)
Last follow up	5 (3 to 8)	7 (1 to 10)	9 (2 to 10)	5 (3 to 10)
Correction %^p	67 (38 to 80)	50 (27 to 93)	36 (30 to 67)	58 (25 to 73)

^kX²(3) =6.255, p=0.100

^lX²(3) =3.997, p=0.262

^mX²(3) =0.817, p=0.845

ⁿX²(3) =1.955, p=0.582

^oX²(3) =0.975, p=0.807

^pX²(3) =3.999, p=0.262

Key safety findings

Overcorrection: 19.4% (n=6)

Overcorrections were greater than or equal to –10^o: n=2 (both patients needed a tether release. Bottom 3 levels were released in 1 and completely removed for the other.)
Overcorrection was -3^o at 2 years: n=1 (a tether breakage and curve increase to 20^o was noted on the 3^rd year.)
Overcorrections were less than -10^o: n=3 (no intervention was considered necessary.)

Pulmonary complications: 12.9% (n=4)

Atelectasis: n=2 (these events resolved with intensive physical therapy.)
Chylothorax: n=1 (this event resolved with dietary precautions.)
Pleural effusion: n=1 (the patient needed readmission within 3 days but a chest tube was not reinserted.)

Mechanical complications: 19.4% (n=6)

Overcorrection and mechanical complications overlapped in 3 patients, adding up to 9 patients (29%) where there was an adverse event related to either implantation or correction. Overcorrection was accompanied by UIV loosening and LIV pull-out in 2 separate cases. There were 1 UIV pull-out, and 1 UIV migration, both of which were stable during follow up and did not need a reintervention. A 51^o main thoracic curve that was corrected to 16^o at first year and had a gradual increase to 25^o from 18 to 24 to 36 months, which was considered as loss of a previously achieved correction.

Overall, 1 patient (3.2%) had a readmission and 2 (6.5%) needed reoperation. Occurrence of pulmonary complications was similar in Sanders groups (p=0.804), while mechanical complications and overcorrection was higher in Sanders 2 (p=0.002 and p=0.018).

Study 9 Samdani AF (2021)

Study details

Study type	Case series (FDA-IDE study)
Country	US (single centre)
Recruitment period	2011 to 2015
Study population and number	n=57 Patients with adolescent idiopathic scoliosis
Age and sex	Mean 12.4 years; 86% (49/57) female; median Risser grade 0
Patient selection criteria	All skeletally immature patients (a Risser grade of 0 through 4 and a Sanders stage of <5) with Lenke type-1A or 1B curves between 30^o and 65^o who had anterior VBT, with a clinical visit at a minimum follow up of 2 years.
Technique	Anterior VBT was done.
Follow up	Mean 55.2 months (range 30.4 to 77.6 months)
Conflict of interest/source of funding	This study was funded by Zimmer Biomet.

Analysis

Follow-up issues: Of the 57 patients, 56 (98.2%) patients were available for analysis of the primary end point of the thoracic Cobb angle at the most recent visit.

Study design issues: This prospective review of a retrospective dataset presented the results from the first U.S. FDA-IDE study on anterior VBT. An outside independent radiologist carried out a comprehensive radiographic analysis to include major and minor Cobb angels, shoulder balance, coronal and sagittal balance, kyphosis, and lumbar lordosis.

Study population issues: With respect to preoperative skeletal maturity, the median Risser grade was 0 (range 0 to 4) and 96% (55/57) of patients had a Risser grade of ≤2. The median Sanders stage was 3 (range 2 to 5).

Key efficacy findings

Number of patients analysed: 57

Operative characteristics

Variable	Findings
Surgical approach
Thoracoscopic access	56.1% (n=32)
Mini-thoracotomy	0
Thoracoscopic and mini-thoracotomy	43.9% (n=25)
Changes in MEPs and SSEPs from baseline
Yes	0
No	100% (n=57)
Duration of surgery (minute)	223.2±79.4 (range 80 to 412)
Blood loss (mL)	105.7±85.9 (range 10 to 440)
Duration of chest tube (day)	2.3±1.1 (range 1 to 5)
Days in ICU	1.5±0.7 (range 1 to 5)
Length of stay (day)	4.8±1.4 (range 3 to 9)
Levels tethered	7.5±0.6 (range 6 to 9)

One patient needed a blood transfusion.

Coronal radiographic outcomes

Visit	Main thoracic curve (^o)	Proximal thoracic curve (^o)	Lumbar curve (^o)
Preoperative	40.4±6.8 (n=56)	25.0±5.7 (n=14)	23.7±6.1 (n=45)
First erect	19.3±8.4 (n=42)	20.1±5.5 (n=16)	15.2±6.1 (n=35)
12 months	12.6±7.2 (n=48)	16.1±6.4 (n=19)	8.8±6.7 (n=42)
24 months	13.8±8.9 (n=46)	15.3±6.9 (n=22)	11.4±7.7 (n=42)
Most recent (a mean follow up of 55.2 months)	18.7±13.4 (n=56)	17.9±6.4 (n=27)	15.7±8.4 (n=49)
Comparison of outcomes	P value
Preoperative compared with first erect	<0.0001
Preoperative compared with most recent	<0.0001
First erect compared with most recent	0.90

Radiographic outcomes

Visit	T5 to T12 Kyphosis (^o)	Lumbar lordosis (^o)	Sagittal balance (mm)	Coronal balance (mm)	Total vertebral thoracic spine length (mm)
Preoperative	15.5±10.0 (n=55)	51.9±11.4 (n=55)	2.6±26.7 (n=55)	2.4±14.8 (n=56)	245.3±19.8 (n=56)
First erect	17.0±10.1 (n=39)	50.9±10.6 (n=38)	3.8±29.1 (n=38)	-2.5±15.6 (n=41)	250.7±16.4 (n=42)
12 months	17.9±11.1 (n=42)	52.4±11.6 (n=40)	-3.7±28.2 (n=39)	-8.4±13.0 (n=46)	258.5±19.1 (n=47)
24 months	17.8±11.9 (n=39)	53.0±10.3 (n=39)	-3.4±26.5 (n=39)	-7.5±11.7 (n=45)	263.0±17.9 (n=46)
Most recent (a mean follow-up of 55.2 months)	19.6±12.7 (n=56)	54.4±11.8 (n=55)	-8.8±30.1 (n=55)	-0.7±15.8 (n=55)	271.5±19.4 (n=56)
Comparison of outcomes	P value	P value	P value	P value	P value
Preoperative compared with first erect	0.40	0.46	0.99	0.07	0.11
Preoperative compared with most recent	0.0023	0.10	0.08	0.31	<0.0001
First erect compared with most recent	0.05	0.0126	0.08	0.59	<0.0001

Clinical trunk rotation based on thoracic sociometer readings: 13.6^o±3.9^o preoperatively; 6.3^o±3.0^o at 2-year follow up; 8.6^o±4.9^o at the most recent follow up.

Should balance: Clinically unlevel shoulders were reported in 54% (28/52) of patients preoperatively and in 25% (14/55) at the latest follow up.

SRS-22 scores: 4.5±0.4 at the most recent follow up (the preoperative SRS-22 scores were not obtained).

Self-image scores: 4.4±0.7 at the most recent follow up.

Measured and predicted lung function at the time of follow up

Follow-up period	No. of patients evaluated	FEV1	Percent of predicted FEV1	FVC	Percent of predicted FVC	FEV1/FVC	Percent of predicted FEV1/FVC
Preoperative	42	2.29	83.0	2.67	83.1	0.86	99.6
1 year	1	3.56	78.2	4.85	92.5	0.73	83.6
2 years	3	2.17	70.5	2.66	74.7	0.81	94.4
3 years	30	2.77	77.7	3.19	77.7	0.88	100.4
4 years	26	2.82	75.4	3.24	75.6	0.88	100.0
5 years	17	2.76	71.8	3.22	74.3	0.86	96.5
6 years	10	2.66	69.0	3.24	74.6	0.82	91.0
Last visit at ≥2 years for patients with preoperative lung function data*	42	2.77	74.9	3.17	74.1	0.88	100.9

*The mean follow up at the last visit was 50.1 months (range 30 to 76.6 months).

At the last follow-up evaluation, the average Risser grade was 4.2±0.9, with 93% of patients having a Risser grade of 4 or 5. The Sanders stage averaged 7.5±0.9, confirming that most patients had attained skeletal maturity.

Key safety findings

Revision: n=7 (12.3%). Of these 7 patients, 5 patients had a tether release for overcorrection and 2 had the tether extended for adding-on. In 1 patient whose curve had overcorrected, the tether release did not stop the curve overcorrection. Subsequently, this patient had PSF approximately 4.6 years after the original tether surgery.

No patients experienced a postoperative neurologic deficit.

Study 10 Rathbun JR (2019)

Study details

Study type	Case report
Country	US
Recruitment period	Not reported
Study population and number	n=1 Patient with adolescent idiopathic scoliosis
Age and sex	13 years; female
Patient selection criteria	Not reported
Technique	VBT with video-assisted thoracoscopic surgery was done – bilateral VBT with PET cored placement on the right T6 to T11.
Follow up	1 year
Conflict of interest/source of funding	None

Analysis

Study design issues: This case report presented a case of primary ureteral injury following VBT with subsequent ureteral erosion and stricture formation requiring definitive ureteral reconstruction.

Study population issues: The patient had VBT for growth correction after a trial of bracing for Cobb angles of 42° in the thoracic and lumbar spine each. Before operation, the patient was classified as Sanders score 6 with Lenke 2C modifier, and the traction films showed reduction of curves to 25° and 22°, respectively.

Key efficacy findings

Number of patients analysed: 1

The patient had successful placement of the device and desired curve reduction:

Preoperation: thoracic and lumbar curves which were 25^o and 22^o, respectively.
2 weeks after operation: thoracic and lumbar curves which were 22.1^o and 3.07^o, respectively.
About 1 year after operation: thoracic and lumbar curves which were 21^o and 6^o, respectively.

The patient progressed to Sanders stage 7 at 1 year after operation.

Length of hospital stay: 15 days after the procedure

The patient was in the hospital for 15 days after the procedure while recovering with resolving pneumothorax and awaiting safe chest tube removal.

Key safety findings

Mid-ureteral injury following VBT with subsequent erosion and stricture formation needing definitive ureteral reconstruction:

Approximately a month after discharge, the patient returned with complaints of left flank and abdominal pain and had CT scan revealing large fluid collection in the left retroperitoneal space and left hydroureteronephrosis. A drain was placed, and fluid was confirmed to be urine. Urology attempted endoscopic retrograde ureteral stent placement, but was unable to pass beyond the L2 screw, revealing a mid-ureteral injury. A couple days later, she had exploratory laparotomy with left ureterolysis, ureteroureterostomy, renal mobilisation with interposition of Gerota's fascia for tissue coverage, and stent placement. She was discharged home on the 4^th postoperative day.

Endoscopic evaluation 6 weeks after primary repair showed ureteral narrowing with calcification on the polyethylene terephthalate cord in the ureteral lumen distal to the original repair. Prior to stent replacement, holmium laser lithotripsy was performed to remove encrustation of this cord. High-grade obstruction of the left kidney was confirmed on functional renal imaging.

Despite successful proximal primary repair, a ureteral stricture was the result of this erosion of the polyethylene terephthalate cord into the ureter. With stricture recurrence and high-grade obstruction, decision for second attempt definitive repair involved ureteral replacement with buccal mucosal graft with omental wrapping approximately 7 months after VBT. Her postoperative course was uncomplicated, and she was discharged on post-operative day 5 with stent removal 3 months later.

At most recent follow up (approximately 1 year after initial surgery), CT urogram showed 2 small ureteral diverticula with no hydronephrosis and appropriate left ureteral drainage. She is asymptomatic, and kidney function remains normal.

Study 11 Shin M (2021)

Study details

Study type	Meta-analysis
Country	Not reported for individual studies
Recruitment period	Publication year: anterior VBT, 2017 to 2019; PSF, 2013 to 2017
Study population and number	n=1,280 (24 studies; anterior VBT in 10 studies [n=211], PSF in 14 studies [n=1,069]) Patients with scoliosis
Age and sex	Anterior VBT: mean 12.4 years; 66.4% (140/211) female; mean Risser score 0.4 PSF: mean 14.2 years; 41.6% (445/1,069) female; mean Risser score 1.4
Patient selection criteria	Inclusion criteria: anterior VBT and/or PSF procedures; Lenke 1 or 2 curves; an age of 10 to 18 years for >90% of the patient population; <10% non-AIS scoliosis aetiology; and follow up of 1 year or more. Exclusion criteria: case reports; studies investigating anterior vertebral body stapling, Harrington rods, and anterior fusion.
Technique	Anterior VBT or PSF
Follow up	Anterior VBT: mean 33.7 months (range 14.4 to 49.5 months) PSF: mean 46.9 months (range 21.2 to 86.4 months)
Conflict of interest/source of funding	None

Analysis

Study design issues: This meta-analysis compared postoperative outcomes between patients with adolescent idiopathic scoliosis having anterior VBT and PSF. The primary objective was to compare complication and reoperation rates at available follow-up times. Secondary objectives included comparing mid-term SRS-22 scores, and coronal and sagittal-plane Cobb angle corrections.

This review followed the preferred reporting items for systematic reviews and meta-analyses guidelines. A systematic review of the literature was conducted for outcome studies following anterior VBT and/or PSF procedures. Reports from annual meetings of the SRS and in PubMed MEDLINE, Scopus, and Embase were included in the search. All studies underwent review by 2 independent reviewers. A single-arm, random-effects meta-analysis was carried out. To provide the most conservative estimates for curve measurements, worst-case imputation was used when averages for curve measurements were provided but SDs were not.

Study population issues: There were 10 case series, 12 retrospective cohort studies, 1 prospective cohort study, and 1 randomised controlled trial. Most anterior VBT studies were case series. Mean preoperative Sanders scores were 3.1 and 3.7 for anterior VBT and PSF patients, respectively. Of the 24 studies, 1 study included 2 patients with syndromic scoliosis who had anterior VBT, and 23 studies selected patients with idiopathic scoliosis. Mean preoperative flexibility was 44.6^o for anterior VBT and 39.2^o for PSF. The average number of vertebrae fused or tethered was 7.3 for patients who had anterior VBT and 10.2 for patients who had PSF.

Other issues: This meta-analysis had limitations. First, nearly all of the AVBT articles were case series, limiting the study design to a single-arm meta-analysis. Thus, the data in this study stem from heterogenous patient populations, and the potential for confounding abounds. Second, the lack of Cobb-angle-correction SDs precluded accurate estimations of the variability of correction rates. Third, SDs for curve measurements were frequently unreported in the PSF group, which necessitated imputation. Fourth, authors only assumed 0 complications and reoperations for studies that explicitly reported so. Therefore, it is possible that some patients who had PSF had unreported complications.

Key efficacy findings

Number of patients analysed: 1,280

Aggregate postoperative outcomes in anterior VBT and PSD studies

Outcomes	Pooled mean (95% CI)
Outcomes	Anterior VBT	PSF
Main thoracic curve (^o)
Preoperative	46.0 (42.3 to 50.0)	53.3 (52.8 to 53.9)
First erect	24.9 (20.1 to 29.8)	16.6 (12.8 to 20.3)
12 to <24 months	24.6 (17.8 to 31.4)	13.3 (8.7 to 17.8)
≥24 to <36 months	21.5 (8.3 to 34.7)	21.9 (17.4 to 26.4)
≥36 months	22.5 (14.1 to 30.9)	22.7 (19.6 to 25.8)
Compensatory lumbar curve (^o)
Preoperative	28.7 (25.6 to 32.0)	30.9 (29.2 to 32.5)
First erect	19.3 (16.6 to 22.4)	9.9 (8.1 to 11.7)
12 to <24 months	16.5 (11.2 to 21.7)	Insufficient data
≥24 to <36 months	13.2 (8.4 to 18.0)	10.7 (8.0 to 13.5)
≥36 months	18.0 (3.5 to 32.5)	15.2 (13.3 to 17.1)
Thoracic kyphosis (^o)
Preoperative	24.3 (17.8 to 30.8)	23.0 (20.7 to 25.2)
First erect	22.1 (16.5 to 27.7)	31.0 (27.9 to 33.3)
12 to <24 months	25.0 (13.4 to 36.6)	Insufficient data
≥24 to <36 months	23.0 (19.6 to 26.4)	17.9 (15.1 to 20.7)
≥36 months	22.5 (12.0 to 33.0)	24.5 (21.9 to 27.1)
Lumbar lordosis (^o)
Preoperative	52.0 (46.2 to 57.9)	47.2 (28.1 to 66.3)
First erect	46.5 (40.1 to 52.8)	Insufficient data
12 to <24 months	56.0 (47.2 to 64.9)	Insufficient data
≥24 to <36 months	52.7 (48.6 to 56.8)	46.3 (42.3 to 50.3)
≥36 months	55.1 (51.3 to 58.8)	46.1 (25.0 to 67.1)
Thoracic rotation (^o)
Preoperative	13.7 (12.1 to 15.2)	15.4 (12.4 to 18.4)
First erect	10.0 (8.7 to 11.2)	6.0 (4.5 to 7.5)
12 to <24 months	8.1 (5.9 to 10.3)	Insufficient data
≥24 to <36 months	6.9 (4.8 to 8.9)	8.07 (5.0 to 11.1)
≥36 months	8.4 (1.0 to 15.7)	13.0 (3.3 to 22.6)
Postoperative SRS-22 self-image	4.27 (4.0 to 4.56)	4.23 (4.07 to 4.40)
Postoperative SRS-22 total	4.36 (4.06 to 4.65)	4.30 (4.17 to 4.43)

No significant publication bias was detected.

Reporting of pre- and post-operative results

Characteristic	Anterior VBT, studies (patients)	PSF, studies (patients)
Pre-operative Risser score	8 (139)	5 (342)
Pre-operative Sanders score	7 (134)	1 (26)
Pre-operative flexibility	4 (34.6)	4 (265)
Complications	10 (211)	9 (610)
Reoperations	10 (211)	4 (312)
Main thoracic Cobb angle	10 (211)	14 (1,069)
Compensatory lumbar curve angle	9 (158)	5 (515)
Thoracic kyphosis angle	8 (187)	8 (662)
Lumbar lordosis angle	5 (94)	5 (421)
Thoracic rotation	6 (109)	3 (127)
SRS self-imaging	2 (76)	7 (616)
SRS total	2 (76)	7 (616)

Key safety findings

Aggregate postoperative outcomes in anterior VBT and PSD studies

Outcomes	Pooled mean (95% CI)
Outcomes	Anterior VBT	PSF
Complication rate (%)	26.0 (12.0 to 40.0)^a	2.0 (0.0 to 4.0)^b
<36 months	11.8 (4.4 to 18.6)	1.0 (0.0 to 2.4)
≥36 months	25.2 (19.1 to 31.7)	2.9 (0.5 to 5.3)
Reoperation rate (%)	14.1 (5.6 to 22.6)	0.6 (0.0 to 2.3)
<36 months	2.9 (0.0 to 8.4)	1.3 (0.0 to 1.7)
≥36 months	24.7 (10.7 to 38.7)	1.8 (0 to 5.4)
Conversion to PSF (%)	1.4 (0 to 4.5)	Not applicable

^aI²=86.14%; 10 studies with 211 patients

^bI²=19.21%; 9 studies with 610 patients

Characteristics of complications

Complication	Anterior VBT	PSF
Tether breakage	7.5% (n=17)	-
Overcorrection	7.5% (n=17)	0.15% (n=2)
Pulmonary	4.8% (n=11)	0.08% (n=1)
Neurological	0.88% (n=2)	0.46% (n=6)
Infection	0.44% (n=1)	0.31% (n=4)
Adding-on	3.2% (n=7)	0.30% (n=2)
Screw pullout/loosening	0.44% (n=1)	0.46% (n=6)
Other^c	4.8% (n=11)	0.46% (n=6)
Overall	26.0% (n=67)	2.0% (n=27)

^c Indicates complications mild in nature, reported as 'other' or with a frequency of 1 across all studies.

Reasons for reoperation

Reason	Anterior VBT	PSF
Tether breakage/pedicle screw loosening	3.2% (n=7)	0.15% (n=1)
Overcorrection	7.7% (n=17)	0 (0)
Adding-on	3.2% (n=7)	0.30% (n=2)
Other (rib hump deformity)	0 (0)	0.15% (n=1)
Overall	14.1% (n=31)	0.6% (n=4)

How are you taking part in this consultation?

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