Use of pelvic incidence to predict proximal junctional failure in long-segment spinal fusion.

Abstract

The objective of this study was to evaluate the impact of pelvic incidence (PI) on spinopelvic correction and complication rates following surgical correction of adult spinal deformity. The authors conducted a retrospective analysis of a prospectively maintained database and identified 204 patients who underwent long-segment thoracolumbar fusions with pelvic fixation and a minimum of 1-year follow-up for deformity correction defined by the Scoliosis Research Society (SRS)-Schwab criteria. Patients were dichotomized into lower (108 patients) and higher (96 patients) PI groups on the basis of the mean value of 55°. Patient demographic characteristics, spinopelvic parameters, and complications were compared using univariate, multivariate, and mixed-effect analyses. Patients with higher PI demonstrated significantly lower rates of proximal junctional failure (PJF) than the lower PI group (9% vs 21%, p = 0.019). In particular, lower rates of PJF were found in patients with an upper thoracic upper instrumented vertebra (UIV) (0% vs 15%, p = 0.014). Multivariate analysis revealed that lower PI (OR 4.3, 95% CI 1.5-15.4, p = 0.006), lower thoracic UIV (OR 6.5, 95% CI 1.8-24.0, p = 0.005), and reoperation status (OR 23.8, 95% CI 5.8-97.8, p = 0.001) independently predicted PJF while controlling for significant differences. Bayesian mixed-effects modeling confirmed lower PI (OR 6.25, 95% credible interval [CrI] 1.52-50.0, posterior probability = 0.996) and lower thoracic UIV (OR 7.03, 95% CrI 1.6-76.0, posterior probability = 0.997) as significant predictors of PJF while controlling for random effects between patients. Patients with lower PI are at an increased risk of PJF following long-segment fusion. Preoperative PI morphology may be used to inform surgical planning and reduce the risk of junctional pathologies.