Lumbar, thoracic, and bilateral vertebral body tethering: do surgical outcomes differ?

VBT is increasingly being used in the lumbar spine to preserve mobility. Surgeons have noted differences in complications and outcomes among single thoracic (T), single lumbar (TL/L), and bilateral (B) VBT, but evidence has been limited to small, separate case series. We aim to comprehensively assess complications and outcomes of vertebral body tethering (VBT) for adolescent idiopathic scoliosis (AIS) in the pediatric population. We hypothesize that the three approaches will have similar curve correction and surgical outcomes, and rates of major complications and reoperations.

A retrospective review was conducted on consecutive AIS patients undergoing VBT between 2015 and 2022 at a tertiary center, all performed by the same surgical team. Data were analyzed at preoperative, first erect (FE), and final follow-up (minimum 2 years postoperative). Inter-screw angles, indicating suspected tether breakage, were measured using a validated deep-learning model.

133 patients (93 T, 20 B, 20 TL/L) were included with a mean follow-up of 2.8 ± 0.5 years. Groups were demographically similar, differing only in instrumented levels (p < 0.001) and preoperative curve bending (p = 0.027). Operative time was longest in B (355.0 ± 21.3 min; p < 0.0001). EBL varied but was not significantly affected by approach after adjusting for levels (p = 0.272). TL/L showed the best curve correction at FE (23 ± 4° vs. T:31 ± 1°, B:28 ± 1°; p = 0.0026). At final follow-up, curve magnitudes were similar (T:28 ± 8°, B:28 ± 10°, TL/L:23 ± 10°; p = 0.089). No significant differences emerged in suspected tether breakage (T:31.2%, B:40%, TL/L:45%; p = 0.068), major complications (T:3.2%, B:10%, TL/L:5%; p = 0.254), or reoperations (T:10.8%, B:10%, TL/L:5.0%; p = 0.903).

In this large series of consecutive VBT cases, no significant differences emerged in tether breakage, major complications, or reoperations among the three approaches. Bilateral tethering required the longest operative time, while TL/L demonstrated superior initial curve correction. Final outcomes were similar, supporting the efficacy and safety of all approaches.