ABSTRACT
BACKGROUND CONTEXT: Proximal junctional failure (PFJ) is a common and dreaded complication of adult spinal deformity. Previous research has identified parameters associated with the development of PJF and the search for radiographic and clinical variables continues in an effort to decrease the incidence of PFJ. The lordosis distribution index (LDI) is a parameter not based on pelvic incidence. Ideal values for LDI have been established in prior literature with demonstrated association with PJF. PURPOSE: The purpose of this study is compare PJF and mechanical failure rates between patients with ideal and nonideal LDI cohort. STUDY DESIGN: This is a retrospective, single-center case-controlled study. PATIENT SAMPLE: Adult patients who underwent surgical treatment for spinal deformity as defined by the SRS-Schwab criteria between 2001 and 2016 were included. Furthermore, fusion constructs spanned at least four vertebral segments with the upper instrumented vertebra (UIV) T9 or caudal. Patients who were under the age of 18, those with radiographic data less than 1 year, and those with neoplastic or trauma etiologies were excluded. Prior thoracolumbar spine surgery was not an exclusion criterion. OUTCOME MEASURES: The outcome measures were physiologic in nature: The primary outcome was defined as PFJ. The International Spine Study Group (ISSG) definition for PJF was used, which includes postoperative fracture of the UIV or UIV+1, instrumentation failure at UIV, PJA increase greater than 15° from preoperative baseline or extension of the construct needed within 6 months. Secondary outcomes included extension of the construct after 6 months or revision due to instrumentation failure, pseudarthrosis or distal junctional failure. METHODS: A portion of this project was funded through National Institute of Health Grant 5UL1TR001067-05. The authors have no conflict of interest related to this study. The records of patients meeting the inclusion criteria were reviewed. Clinical and radiographic data were extracted and analyzed. Univariate cox proportional hazard models were used to identify factors associated with mechanical failure and included in a multivariate Cox proportional hazards model. RESULTS: There were 187 patients that met the inclusion criteria. Univariate analysis demonstrated the number of levels fused, instrumentation to the sacrum or pelvis, PI-LL difference between pre- and postoperative states, T1-SPI, T9-SPI, and postoperative LDI (treated as a continuous variable). When LDI was treated as a categorical variable using an LDI cutoff of less than 0.5 for hypolordotic, 0.5 to 0.8 for aligned and greater than 0.8 for hyperlordotic, there was no difference in failure rates between the two groups. CONCLUSIONS: Lumbar lordosis is an important parameter in adult deformity. However, the LDI is an imperfect variable and previously developed categories did not show differences in failure rates in this cohort.
