• Hongda Bao, Renaud Lafage, Barthelemy Liabaud, Jonathan Elysée, Bassel G Diebo, Gregory Poorman, Cyrus Jalai, Peter Passias, Aaron Buckland, Shay Bess, Thomas Errico, Lawrence G Lenke, Munish Gupta, Han Jo Kim, Frank Schwab, and Virginie Lafage.
• Nanjing Drum Tower Hospital, Nanjing University, Nanjing, China.
• Eur Spine J. 2018 Feb 1; 27 (2): 397-405.

PurposeA comprehensive understanding of normative sagittal profile is necessary for adult spinal deformity. Roussouly described four sagittal alignment types based on sacral slope, lumbar lordosis, and location of lumbar apex. However, the lower limb, a newly described component of spinal malalignment compensation, is missing from this classification. This study aims to propose a full-body sagittal profile classification in an asymptomatic population based on full-body imaging.MethodsThis is a retrospective analysis of a prospective single-center study of 116 asymptomatic volunteers. Cluster analysis including all sagittal parameters was first performed, and then ANOVA was performed between sub-clusters to eliminate the non-significantly different parameters. This loop was repeated until all parameters were significantly different between each sub-cluster.ResultsThree types of full-body sagittal profiles were finalized according to cluster analysis with ten radiographic parameters: hyperlordosis type (77 subjects), neutral type (28 subjects), and compensated type (11 subjects). Radiographic parameters included knee angle, pelvic shift, pelvic angle, PT, PI-LL, C7-S1 SVA, TPA, T1 slope, C2-C7 angle, and C2-C7 SVA. Age was significantly different across compensation types, while BMI and gender were comparable. Age-matched subjects were randomly selected with 11 subjects in each type. ANOVA analysis revealed that all parameters but PT and C2-C7 angle remained significantly different.ConclusionsThe current three compensation types of full-body sagittal profiles in asymptomatic adults included significant changes from cervical region to knee, indicating that subjects should be evaluated with full-length imaging. All three types exist regardless of age, but the distribution may vary.

32 keywords...

hide…