Objective: To examine existing literature and pool the data to determine the relative odds ratio of "adding-on" (AO) based on various reported criteria for lower instrumented vertebra (LIV) selection in Lenke type 1A and 2A curves.

Methods: Using electronic databases, studies reporting on AO and LIV selection in Lenke type 1A and 2A curves were identified. Studies were excluded if they failed to meet the following criteria: ≥ 30 patients, Lenke type 1A or 2A curves, thoracic-only fusions, and inclusion of outcome differences in AO and non-AO groups. Review articles, letters, and case reports were excluded.

Results: Six studies were identified reporting on 732 patients with either Lenke type 1A or 2A curves treated with thoracic-only fusions. Five different landmarks were used for LIV selection in these studies including the stable vertebra (SV) -1, end vertebra (EV) +1, neutral vertebra (NV), touched vertebra (TV), and substantially touched vertebra (STV) versus nonsubstantially touched vertebra (nSTV) +1. The pooled odds ratios of AO for choosing LIV at levels above the afore landmarks (i.e. , ending the construct "short") versus at the landmarks were 2.59 (SV-1), 2.43 (EV+1), 3.05 (NV), 3.40 (TV), and 4.52 (STV/nSTV+1), all at 95% confidence interval.

Conclusions: Five landmarks shared a similar characteristic in that the incidence of AO was significantly higher if the LIV was proximal to the chosen landmark. In addition, choosing STV/(nSTV+1) as the LIV have the lowest absolute risk of AO and the greatest risk reduction. If additional levels were fused (i.e. , LIV distal to the landmark), there was no statistically significant benefit in further reducing the risk of AO. Selection of the optimal LIV is a complex issue and spine surgeons must balance the risk of AO with the need for motion preservation in young patients.