Complications and outcomes after three-column osteotomy: a 5-year follow-up

Three-column spinal osteotomies (3-CO) are powerful techniques used to correct spinal deformity. These surgeries are associated with an elevated risk of complications. While outcomes have been reported in the literature with 2 years follow-up, longer-term outcomes of three-column osteotomies remain understudied.

This study aims to examine patient reported outcomes and complications for three-column osteotomies at 5 years.

Retrospective review of a prospectively collected spinal deformity cases database.

All consecutive adult patients at a single spine surgery center treated with either a pedicle subtraction osteotomy (PSO) or vertebral column resection (VCR) for spinal deformity, and with a minimum 5-year follow-up were included.

Visual-analog scale (VAS) for back pain score (0 to 10), Oswestry Disability Index (ODI) score (0 to 100), number of complications, revision rate, sagittal balance, lumbar lordosis at preoperative and at 5-year visit.

Data was extracted from a prospectively collected spinal deformity surgery database continuously updated since 2002 with data from operative reports, patients' medical visit notes and patients' self-reported VAS and ODI questionnaires completed at each office visit. Radiographic measurements were performed on standing full-length spine radiographs taken at preop and 5-year visits. Descriptive statistics, comparison of means and proportions among groups, and a logistic regression analysis were conducted using the statistical software package SPSS version 28. Statistical significance was set at p<.05.

Of 127 consecutive adult patients with minimum of 5-year follow-up undergoing a 3-CO posterior spinal surgery for spinal deformity were identified and included in the study, 79 (62%) were treated with PSO, and 48 (38%) with VCR. Both PSO and VCR groups demonstrated significant improvements in VAS (PSO preop: 6.7, 5-year: 4.6, p<.001; VCR preop: 7.1, 5-year: 5.2, p<.001), and ODI (PSO preop: 52.9, 5-year: 45.4, p<.001; VCR preop: 57.5, 5-year 43.0, p<.001) that were maintained at 5 years. Major and minor complications occurring within 5 years were not statistically different between the PSO and VCR groups (major: 59.5% vs 56.3%, p=.85; minor: 45.6% vs 37.5%, p=.46). The rate of revision surgery within 5 years was 39.4% overall. Of the fifty patients requiring revision, 37.5% were for nonunion, 27.1% instrumentation failure, 14.6% proximal junctional kyphosis (PJK), 12.5% vertebral fracture, 6.3% motor weakness, and 2.1% infection. Improvements in lumbar lordosis were maintained at 5 years in both the PSO (29.9° vs 47.2°, p<.001) and VCR (34.6° vs 48.5°, p<.001) groups while sagittal balance maintained significant improvement in the VCR group (9.5 cm vs 6.3 cm, p=.008) but not the PSO (11.4 cm vs 9.3 cm, p=.065).

Patients undergoing three-column osteotomies had a major complication rate of 57.5% and a minor complication rate of 42.5% after 5 years. Improvements in lumbar lordosis were maintained at 5-year follow-up, but sagittal balance was only maintained in the VCR group. Despite these radiographic findings, both VCR and PSO groups maintained significant clinical improvements in both VAS and ODI scores at 5-year follow up.