Pedicle subtraction osteotomy: a comprehensive analysis in 104 patients. Does the cause of deformity influence the outcome?

OBJECTIVE The clinical outcomes and complications of patients who underwent pedicle subtraction osteotomy (PSO) for various diagnoses were compared. More specifically, the purpose was to identify if outcomes differed between patients with flat-back syndrome after lumbar fusion (FBS-LF) versus patients who underwent surgery for adult spinal deformity (ASD). METHODS A retrospective analysis of 104 patients who underwent a PSO for sagittal plane imbalance was performed. There were 28 patients with FBS-LF and 76 patients with various forms of ASD. Outcome was measured using visual analog scale (VAS)-back, VAS-leg, Oswestry Disability Index (ODI) (range 0-100 for all scales), and EQ-5D scores (range 0-1). Patients also rated their global outcomes as much better, better, unchanged, or worse at follow-up. The minimum follow-up was 1 year (range 1-4 years). Clinical outcomes and complications were compared between the 2 groups of patients. RESULTS The most common level of PSO was L-3 and L-2; 100 single and 4 double PSOs were performed. The average local correction by PSO itself was 27.2°. The sagittal vertical axis (SVA) improved from a mean preoperative value of 74 ± 23 mm to 49 ± 20 mm at the final follow-up. The VAS-back, ODI, and EQ-5D scores improved significantly for the entire group by 33, 16, and 0.31 points, respectively. In total, 57% of patients reported that they were "much better" or "better" than before surgery. Preoperatively, as well as postoperatively, the FBS-LF patients reported significantly worse VAS scores. According to VAS-back results, the ASD group improved by 34 points compared with 29 points in FBS-LF patients. ODI scores were similar between the 2 groups preoperatively but improved significantly more in the ASD group (18 points) compared with the FBS-LF group (13 points). The EQ-5D scores improved from 0.07 to 0.35 in FBS-LF patients, and from 0.21 to 0.56 on average in ASD patients. Similarly, a "much better" or "better" outcome compared with before surgery was reported by 72% of patients in the ASD group compared with 24% of patients in the FBS-LF group (p < 0.001). The overall reoperation rate was 31%: 46% of patients in the FBS-LF group compared with 25% of patients in the ASD group. There were 19 (18%) dural tears, 14 (13.5%) surgical site infections, 12 (11.5%) instances of pseudarthrosis, 15 (14%) proximal junctional failures, and 2 distal junctional failures. The 12 (11%) neurological complications were dominated by partial weakness of hip flexion and knee extension, and all but 2 of these were temporary. CONCLUSIONS PSO is a safe and effective method for correcting sagittal plane imbalance due to multiple etiologies. The authors found patient satisfaction to be high, and health-related quality of life was greatly improved by the procedure in patients with ASD. In contrast, in FBS-LF patients, a suboptimal outcome was observed and the cautious use of PSO seems warranted in this subset of patients.

The impact of spinopelvic morphology on the short-term outcome of pedicle subtraction osteotomy in 104 patients.

OBJECTIVE Pedicle subtraction osteotomy (PSO) is commonly performed for correction of spinal sagittal plane deformities. The PSO results in complex, multiple changes of the spinopelvic alignment. The influence of the variability of individual pelvic morphology has not been fully analyzed in previous outcome studies of sagittal imbalance. The aim of this study was to define radiological variables affecting the outcome after PSO in adult spinal deformities, with special emphasis on the variability of pelvic morphology. METHODS Clinical and radiographic outcomes were analyzed in a retrospective analysis of 104 patients who underwent a PSO at a single center. The radiographic variables studied were sagittal vertical axis (SVA), T1SPI (T-1 spinopelvic inclination), lumbar lordosis (LL), thoracic kyphosis (TK), pelvic incidence (PI), pelvic tilt (PT), and sacral slope (SS). To control for the individual variation of pelvic morphology, the LL/PI, PT/PI, and SS/PI ratios were calculated. Clinical outcome was assessed using the visual analog scale for pain, Oswestry Disability Index, and EQ-5D preoperatively and at a minimum 1-year follow-up. Correlation coefficients were calculated between each individual radiographic variable and the outcome measures. The importance of LL mismatch to TK, reflecting the importance of a harmonious spine, was analyzed by comparing the outcome of patients with a TK+LL+PI ≤ 45° to those with a sum > 45°. RESULTS SVA and T1SPI demonstrated the strongest correlation with the clinical outcome scores (r = 0.4-0.5, p < 0.001). LL correlated weakly with the clinical outcome (r = 0.2-0.3, p < 0.003). Mismatch of LL to PI, however, did not correlate significantly with the outcome. Similarly, only weak and inconsistent correlation was observed between PT, SS, PT/PI, SS/PI, and functional outcome. Patients with a TK+LL+PI ≤ 45° had a significantly lower ODI score (33 vs 44) and a significantly higher EQ-5D score (0.64 vs 0.40) than patients with a sum > 45° (LL is a negative value). CONCLUSIONS PSO resulted in a substantial correction of sagittal imbalance and improved outcome in most patients in this study. Correction of the global sagittal balance appears to be a necessary precondition for a good outcome. A harmonious spine with a TK and an LL of similar magnitude seems to add to a positive outcome.