Background: Nonstructural curves are defined in the Lenke classification system for adolescent idiopathic scoliosis as bending out to <25 degrees. A caveat in the original paper states, however, that if the difference in Cobb magnitude between the major and minor curves is <5 degrees, then the minor curve should be considered structural, regardless of its Cobb magnitude. It is unclear whether following this rule affects patient outcomes.
Methods: A multicenter retrospective study using a prospectively collected database was performed on surgical adolescent idiopathic scoliosis patients, with 2 years of minimum follow-up. All the lumbar curves measured <25 degrees on bending x-ray. Curves (major and minor) that were within <5 degrees of each other were analyzed in 2 treatment subsets: nonselective fusion (NS) and selective thoracic fusion (STF1), and compared with similar selective fusion cases with false double major curves with a lumbar curve between 5 and 10 degrees less in magnitude than the thoracic curve magnitude (STF2). Preoperative and 2-year postoperative radiographic and SRS-22 parameters were compared.
Results: Of 58 patients, there were 14 NS, 11 STF1, and 33 STF2. NS had larger preoperative curves than STF1, but achieved better lumbar correction (58%) than STF1 (32%) or STF2 (41%) (P=0.004). STF1 tended to have more preoperative coronal imbalance than STF2 (−2.26 vs. −1.12 cm) (P=0.066) but were similar at 2 years. Preoperative thoracic rib prominence measures were similar for all groups, but NS had significantly worse rib prominence than STF1 (NS: 8.4 vs. STF1: 4.4, P=0.046) at 2 years. There were no differences among the 3 groups in SRS-22 scores preoperatively and at 2 years.
Conclusions: Almost half of the cases with curves within <5 degrees of each other did not follow the 5-degree rule, suggesting that there is variability among surgeons in their definition of what is truly structural. Selective thoracic fusion cases behaved similarly regardless of whether or not the curves were within 5 degrees of each other. The results of our analysis of the 5-degree caveat challenge its utility as a criterion for defining structural curves.
Level of Evidence: Level II.