A retrospective study.
To assess the mechanisms and the independent risk factors associated with proximal junctional kyphosis (PJK) in patients treated surgically for adult spinal deformity with long fusions to the sacrum.
Summary of Background Data.
The occurrence of PJK may be related to preoperative and postoperative sagittal parameters. The mechanisms and risk factors for PJK in adults are not well defined.
Consecutive patients who underwent long instrumented fusion surgery (≥6 vertebrae) to the sacrum with a minimum of 2 years of follow-up were retrospectively studied. Risk factors included patient factors, surgical factors, and radiographical parameters such as thoracic kyphosis (TK), lumbar lordosis (LL), sagittal vertical axis, pelvic tilt, and pelvic incidence.
Ninety consecutive patients (mean age, 64.5 yr) met inclusion criteria. Radiographical PJK occurred in 37 of the 90 (41%) patients with a mean follow-up of 2.9 years. The most common mechanism of PJK was fracture at the upper instrumented vertebra (UIV) in 19 (51%) patients. Twelve (13%) patients with PJK were treated surgically with proximal extension of the instrumented fusion. Preoperative TK more than 30°, preoperative proximal junctional angle more than 10°, change in LL more than 30°, and pelvic incidence more than 55° were identified as predictors associated with PJK. Achievement of ideal global sagittal realignment (sagittal vertical axis <50 mm, pelvic tilt <20°, and pelvic incidence-LL <±10°) protected against the development of PJK (19% vs. 45%). A multivariate regression analysis revealed changes in LL more than 30°, and preoperative TK more than 30° were the independent risk factors associated with PJK.
Fracture at the UIV was the most common mechanism for PJK. Change in LL more than 30° and pre-existing TK more than 30° were identified as independent risk factors. Optimal postoperative alignment of the spine protects against the development of PJK. A surgical strategy to minimize PJK may include preoperative planning for reconstructions with a goal of optimal postoperative alignment.
Level of Evidence: 3