Long-term radiographic outcomes of expandable versus static cages in transforaminal lumbar interbody fusion

OBJECTIVE Potential advantages of using expandable versus static cages during transforaminal lumbar interbody fusion (TLIF) are not fully established. The authors aimed to compare the long-term radiographic outcomes of expandable versus static TLIF cages. METHODS A retrospective review of 1and 2-level TLIFs over a 10-year period with expandable and static cages was performed at the University of California, San Francisco. Patients with posterior column osteotomy (PCO) were subdivided. Fusion assessment, cage subsidence, anterior and posterior disc height, foraminal dimensions, pelvic incidence (PI), segmental lordosis (SL), lumbar lordosis (LL), pelvic incidence-lumbar lordosis mismatch (PI-LL), pelvic tilt (PT), sacral slope (SS), and sagittal vertical axis (SVA) were assessed. RESULTS A consecutive series of 178 patients (with a total of 210 levels) who underwent TLIF using either static (148 levels) or expandable cages (62 levels) was reviewed. The mean patient age was 60.3 +/- 11.5 years and 62.8 +/- 14.1 years for the static and expandable cage groups, respectively. The mean follow-up was 42.9 +/- 29.4 months for the static cage group and 27.6 +/- 14.1 months for the expandable cage group. Within the 1-level TLIF group, the SL and PI-LL improved with statistical significance regardless of whether PCO was performed; however, the static group with PCOs also had statistically significant improvement in LL and SVA. The expandable cage with PCO subgroup had significant improvement in SL only. All of the foraminal parameters improved with statistical significance, regardless of the type of cages used; however, the expandable cage group had greater improvement in disc height restoration. The incidence of cage subsidence was higher in the expandable group (19.7% vs 5.4%, p = 0.0017). Within the expandable group, the unilateral facetectomy-only subgroup had a 5.6 times higher subsidence rate than the PCO subgroup (26.8% vs 4.8%, p = 0.04). Four expandable cages collapsed over time. CONCLUSIONS Expandable TLIF cages may initially restore disc height better than static cages, but they also have higher rates of subsidence. Unilateral facetectomy alone may result in more subsidence with expandable cages than using bilateral PCO, potentially because of insufficient facet release. Although expandable cages may have more power to induce lordosis and restore disc height than static cages, subsidence and endplate violation may negate any significant gains compared to static cages.