Correlation analysis of the PI-LL mismatch according to the pelvic incidence from a database of 468 asymptomatic volunteers.

Abstract

Previous studies on adults with degenerative scoliosis (ADS) have been fixed the threshold of PI-LL mismatch less than 10° for achieving good clinical outcomes. Recent studies discussed that PI-LL mismatch should consider individual pelvic incidence (PI) and should be set first in a normal population. The purpose of this study is to assess the variability of PI-LL mismatch according to PI in an asymptomatic population. Full-body low dose stereoradiographic evaluation was done in a multi-ethnic cohort of 468 asymptomatic adult volunteers. Patients were clustered in three groups depending on individual PI values: PI < 45°, 45° < PI < 60° and PI > 60°. 3D measurements were performed using a commercially available 2D/3D modeling software to establish a correlation of PI with other spinopelvic parameters. ANOVA and Tukey's HSD for post-hoc analysis were used to determine the differences between the three groups. In our asymptomatic population, the mean value of PI-LL mismatch is -5.4° ± 10.7°. Clusterization of the population reveals significant differences in the distribution of L1S1 lordosis, pelvic tilt and PI-LL with positive linear correlation according to PI values. As an interestingly result, PI-LL mismatch is equal to 0° when PI is around 64°. The present study demonstrated that PI-LL mismatch is negative in an asymptomatic population (-5.4° ± 10.7°) and the value should be customized to each patient to be able to restore the appropriate lordosis in ADS. The PI-LL mismatch is given by the formula PI-LL = -28.5 + 0.44 × PI.