TLIF with expandable vs static height cages: Defining guidelines based on pre-op disc height and lordosis

2015

Background
Like many new technologies, expandable TLIF/PLIF cages were introduced without research confirming where or why they might be more useful than current alternatives. No clinical or radiographic studies have been done comparing static height cages (SC) to expandable cages (EC).

Purpose
Compare clinical and radiographic difference after TLIF with SC vs EC, controlling for level, disc height, and disc angle. Study results will assist in defining the relative indications EC vs SC based on pre-op disc level, interspace height and angle.

Study Design
Prospective data retrospectively reviewed. 

Patient Sample
Consecutive surgical cohort of 157 adults with degenerative disease who underwent posterior spinal fusion (PSF) with TLIF at 1 or 2 levels, L2-S1, using either SC or EC, with minimum 2 year follow-up.

Outcome measures
Standing AP/Lateral radiographs: disc height and disc angle at TLIF levels.  Fusion at TLIF levels was defined as bridging bone across the interspace at 2 years, with no implant loosening.  Clinical:  Visual Analog pain scores (VAS), Oswestry Disability Index (ODI), and pain medication use. Mann Whitney, Wilcoxon signed rank test, t-test, and multivariate linear regression were used in statistical analysis.

Methods
VAS, ODI, pain med records recorded pre-op, 1 year, 2 years, and yearly. Radiographs measured pre-op, 1year, and 2years for: TLIF Disc Height (measured Anterior, Middle, Posterior), Disc Height category (Short: 0-8mm, Medium:9-12mm, Tall:>12mm), Segmental Lordosis (disc angle measured from endplates, categorized: Low:0-4°, Medium:5-11°, High:>12°), Lumbar Lordosis (T12-S1). Complications including xray evidence of adjacent level degeneration were recorded at each follow-up visit.

Results
Follow-up averaged 44months (24-68months). There was no difference SC (n=73, 103 cages) vs EC (n=84, 111 cages) groups for age (63years, range 20-86years), smoking (11% vs 10%), prior surgery (58% vs 49%); Surgery was similar for both groups: PSF averaged 2.4 levels, 1-level TLIF (57%), 2-level TLIF (43%). EC vs SC regional lordosis pre-op and 2 years was similar. Series complications: Nonunion-1, infection-3, neuro-3, revision surgery-16, without difference SC vs EC.  At 2 years, overall pre-op disc height improvements with surgery for all levels were noted with both cages(p<0.001), the greatest gain over preop using EC at L4-5(EC-4.7mm, SC-1.7mm) and L5-S1(EC-5.0mm, SC-0.7mm). Comparing within specific disc height categories (Short, Medium, Tall), EC increased height better than SC in each category. EC group had greater increase in segmental lordosis than SC (3.4° vs 0.3°).  Two-year interspace lordosis in High angled discs was maintained for EC but declined with SC (- 4.0°). By latest follow-up, radiographic degeneration at adjacent levels was similar: SC-35, EC-33. Both EC and SC groups improved VAS and ODI (p<0.01) after surgery without difference at 2 years.

Conclusions
Both EC and SC improved disc height and segmental lordosis at 2 years, though EC gains trended higher.  EC performed particularly well vs. SC in tall and highly lordotic discs, short discs, and for all disc heights and angles at L4-5 and L5-S1. Long-term segmental lordosis decreased in highly angled discs treated with SC. No difference in adjacent level degeneration was noted between groups at 4 years.  As this is the first clinical and radiologic study of EC, further study is warranted before firm conclusions can be made.