Abstract
Introduction
One of the most deleterious complications after pedicle subtraction osteotomy (PSO) is proximal junctional failure (PJF), typically leading to reoperations with unpredictable outcome. Various measures to prevent it have been suggested, all, however, with limited success. The purpose of this study was to correlate the occurrence of PJF to preoperative clinical as well as postoperative radiological factors, to define preventive strategies.
Patients and Methods
This retrospective study included 104 consecutive patients with a kyphotic deformity and sagittal imbalance operated with a PSO. There were 35 men and 71 women, mean age 58 (range, 18–80 years). There were 28 patients with flat back syndrome after lumbar fusion (FBS-LF), 21 degenerative scoliosis (DS), 20 posttraumatic kyphosis (PTK), 16 flat back syndrome after scoliosis surgery (FBS-SF), 9 adult idiopathic scoliosis (AISA), 6 neuromuscular scoliosis (NS), 3 congenital scoliosis (CS), and 1 ankylosing spondylitis (AS). A total of 83 (80%) patients were fused to the thoracolumbar or thoracic level and 21 (19%) patients were fused to the upper thoracic spine. The patients were followed up 12 to 36 months postoperatively with radiographs and reoperations documented for PJF.
Results
There were 18 patients with PJF requiring revision surgery (17%). In 10 patients PJF occurred before 6 months, in 4 between 6 and 12 months, and in 4 after 12 months. Mean time to PJF was 9 months. There were 16 PJFs of the bony type (fracture) and 2 of the ligamentous type. Mean age among those with PJF was 66 years as compared with 55 years in non-PJF. Female proportion was 77% in PJF and 66% in non-PJF. Patients fused to the thoracolumbar junction had a PJF in 28% of cases compared with 10% of patients with a high thoracic fusion (p < 0.001). The PJFs were distributed according to diagnosis as follows: FBS-LF 36%, AISA 22%, PTK 15%, DS 10%, and FBS-SF 6% with no PJF in other diagnoses. In patients with a previous osteoporotic fracture PJF occurred in 40%, as compared with 10% in patients with no osteoporotic fracture (p = 0.002). Sagittal vertical axis was 97 mm in PJF and 50 mm in the non-PJF (p = 0.015). In patients not corrected to a SVA < 50 mm the incidence of PJF was 25%, as compared with patients corrected to an SVA < 50 mm with an incidence of 17% (p = 0.117). PJF was more common in patients with a mismatch between pelvic incidence and lumbar lordosis > 15 degrees, and also in patients with a mismatch of pelvic incidence and sacral slope. The average local correction by the PSO was 31 degrees. There was no correlation between the magnitude of correction and the occurrence of a PJF.
Conclusions
A PJF is a common complication to a PSO. We identified the following risk factors; previous lumbar fusion, preoperative osteoporotic fracture, a PI-LL mismatch of > 15 degrees and upper instrumented vertebra in thoracolumbar region, with previous lumbar fusion being the most common risk factor, found in 55% of patients with a PJF. The results of the study add to the difficult selection of patients suitable for a PSO for correction of sagittal imbalance.