Major complications and comparison between 3-column osteotomy techniques in 105 consecutive spinal deformity procedures

Abstract

Study design: A retrospective review.

Objective: To characterize the risk factors for the development of major complications in 3-column osteotomies and determine whether the presence of a major complication affects ultimate clinical outcomes.

Summary of background data: Three-column spinal osteotomies, including pedicle subtraction osteotomy (PSO) and vertebral column resection (VCR), are common techniques to correct severe and/or rigid spinal deformities.

Methods: Two hundred forty consecutive PSO (n = 156) and VCR (n = 84) procedures in 237 patients were performed at a single institution between 1995 and 2008. Of these, 105 patients (87 PSOs, 18 VCRs) had complete preoperative and minimum 2-year postoperative clinical outcomes data available for analysis. Using established criteria, we reported complications as major or minor and further stratified complications as surgical versus medical and permanent versus transient. Risk factors for complications and their effect on Scoliosis Research Society (SRS) clinical outcomes at baseline and at 2 years or more were assessed.

Results: Major medical and surgical complications occurred at similar rates in both PSOs and VCRs (38%, 33 of 87 vs. 22%, 4 of 18; P = 0.28). Overall, 24.8% (26 of 105) experienced major surgical complications (3 permanent) and 15.2% (16 of 105) experienced major medical complications (4 permanent). Patients with PSO were older (53 vs. 29 yr; P < 0.001), had greater estimated blood loss (1867 vs. 1278 mL; P = 0.02), and showed a trend toward fewer fused levels (10.1 vs. 12.2; P = 0.06). Risk factors for major complications included preoperative sagittal imbalance of 40 mm or more (P = 0.01), age 60 years and older (P = 0.01), and the presence of 3 or more medical comorbidities (P = 0.04). Both groups improved significantly from baseline in SRS subscores; however, patients with PSO started off worse but improved more than VCRs in both the pain (+1.0 vs. +0.1; P < 0.001) and function (+0.6 vs. +0.2; P = 0.01) domains, with no differences in final satisfaction (4.1 vs. 4.3; P = 0.54). PSO and VCR patients with no complications had slightly higher satisfaction scores than patients with minor-only complications, major transient complications, and major permanent complications. There were no significant differences among the groups with respect to change in SRS subscores from baseline, and all complication groups improved significantly from baseline (P = 0.04).

Conclusion: Major complications occurred in 35% of 3-column osteotomies and at similar rates for both PSO (38%) and VCR (22%) procedures. The presence of a major complication did not affect the ultimate clinical outcomes at 2 years or more.