L5-S1 Transforaminal Lumbar Interbody Fusion is Associated With Increased Revisions Compared With L4-L5 Transforaminal Lumbar Interbody Fusion at Two Years.

STUDY DESIGN

Retrospective cohort study.

OBJECTIVE

The aim of this study was to compare outcomes in patients undergoing 1-level transforaminal lumbar interbody fusion (TLIF) at L4-S1.

BACKGROUND

TLIF is frequently performed at L4-S1 to treat degenerative lumbar pathologies. However, the native alignment and biomechanics differ across L4-L5 and L5-S1, and there is limited data regarding comparative radiographic outcomes.

PATIENTS AND METHODS

Patients who underwent 1-level TLIF at L4-L5 or L5-S1 at a single academic institution were identified. Baseline demographics, procedural characteristics, change in postoperative spinopelvic alignment patient-reported outcome measures, and 2-year postoperative surgical complications were compared. Multivariate regression analyses, accounting for age, sex, Charlson Comorbidity Index, and body mass index, were also performed.

RESULTS

Across the 175 included patients, 125 had L4-L5 TLIF and 50 had L5-S1 TLIF. The mean age was 57.8 years, 56.6% were females, the mean Charlson Comorbidity Index was 0.9, and the mean follow-up was 26.7 months. In the hospital, the 2 cohorts were not statistically different with regard to estimated blood loss and length of stay. Two years postoperatively, multivariate linear regression analyses revealed that L5-S1 TLIF achieved 6.0° higher correction in L4-S1 lordosis ( P = 0.012) than L4-L5 TLIF. At the same time, however, L5-S1 patients undergoing TLIF experienced significantly higher rates of pseudoarthrosis (8.0% vs. 1.6%, P = 0.036) and subsequent spine surgery (18.0% vs. 7.2%, P = 0.034), specifically for pseudoarthrosis (6.0% vs. 0.0%, P = 0.006), with this cohort having 8.7 times higher odds of subsequent spine surgery for pseudoarthrosis ( P = 0.015) than L4-L5 patients undergoing TLIF on multivariate logistic analyses. Patient-reported outcome measures, in contrast, were not different across the 2 cohorts.

CONCLUSIONS

Although L5-S1 TLIF yielded good radiographic correction, it was associated with higher rates of subsequent spine surgery for pseudoarthrosis compared with L4-L5 TLIF. These findings may be related to differences in native segmental alignment and biomechanics across the L4-L5 and L5-S1 motion segments and are important to consider during surgical planning.

LEVEL OF EVIDENCE

Level IV.