OBJECTIVE: To investigate whether postoperative T4-L1PA mismatch (T4 pelvic angle minus L1 pelvic angle) predicts mechanical revision within 5 years after multilevel fusion to the upper lumbar and lower thoracic spine, versus other alignment parameters.
SUMMARY OF BACKGROUND DATA: T4-L1PA mismatch is an established predictor of mechanical failure after upper thoracic fusion to the sacropelvis. Its value in lower thoracic spine and upper lumbar constructs remains unclear.
METHODS: We retrospectively analyzed 204 patients who underwent posterior spinal fusion of 3 to 10 instrumented levels (upper instrumented vertebra [UIV] T8-L3) with distal fixation to the sacrum or ilium (2006-2024). Alignment was assessed preoperatively, at the first postoperative timepoint, and at 1 year: T4-L1PA mismatch, PI-LL mismatch, sagittal vertical axis, L1PA deviation, lordosis distribution index, and distal lordosis. The primary outcome was mechanical revision within 5 years. Cox regression, ROC, Kaplan-Meier, and combined-risk analyses were performed.
RESULTS: Thirty-three patients (16.2%) underwent mechanical revision within 5 years. On first postoperative radiographs, T4-L1PA mismatch predicted revision (hazard ratio [HR] 1.10, P=0.012), as did PI-LL (HR 1.03, P=0.027). Adjusted for PI-LL, age, and levels fused, T4-L1PA remained significant (HR 1.09, P=0.028) while PI-LL did not (P=0.070). At 1 year (n=148, 16 events), T4-L1PA had an HR of 1.17 (P=0.001; PI-LL P=0.171). Other parameters were non-significant. Five-year cumulative revision incidence rose from 6% with neither parameter malaligned to 19% with one and 39% with both (log-rank P=0.003).
CONCLUSIONS: Postoperative T4-L1PA mismatch was the most consistent alignment predictor of mechanical revision over 5 years in thoracolumbar and upper lumbar fusions (UIV T8-L3) to the sacropelvis and remained independently prognostic at 1 year, with most revisions attributable to proximal junctional failure. PI-LL contributed early risk but weakened over time. Combined assessment identified a high-risk subgroup.
STUDY DESIGN: Retrospective single-center cohort study.
Read the article: PubMed · Publisher (DOI)