Effect of glenoid component compression timing on the cement-bone interface of pegged all-polyethylene glenoid components: a micro-computed tomography study

Abstract

Background: Glenoid loosening remains a principal failure mode after anatomic total shoulder arthroplasty. Cement pressurization improves fixation; however, the optimal timing of axial compression during polymethylmethacrylate curing remains unknown. We hypothesized that brief-early compression, followed by undisturbed curing would yield a more favorable cement-bone morphology than continuous or intermittent compression.

Methods: Fifteen ovine scapulae received three-pegged all-polyethylene glenoid components (Next Health Products, Ankara, Turkey) using standard-viscosity polymethylmethacrylate. Specimens were randomized to continuous compression (70 N for 10 min), brief-early compression (70 N for 2 min, then none), or intermittent compression (70 N; 2 min on, 1 min off, 3 min on, 1 min off, and 3 min on) groups (n = 5/group). An axial load was applied using a universal testing machine and custom seating jig; no off-axis or cyclic loading was introduced. High-resolution micro-computed tomography with blinded analysis quantified cement-bone interdigitation volume, cement porosity, cement-bone contact area, penetration depth, and peri-implant trabecular bone mineral density. Group comparisons used one-way or Welch analysis of variance, Holm-Bonferroni correction across the 4 prespecified interface outcomes, and Tukey or Games-Howell tests for pairwise comparisons. Bone mineral density was analyzed exploratorily.

Results: Omnibus tests showed significant group effects for interdigitation volume (P = .002), cement porosity (P < .001), and cement-bone contact area (P = .017), whereas penetration depth did not differ between the groups (P = .475) despite these differences. Brief-early and continuous compression produced greater interdigitation than intermittent compression (brief-early vs. intermittent mean difference [MD] +273 mm³, 95% confidence interval [CI] 119-428; P = .001; continuous vs. intermittent MD +162 mm³, 95% CI 7-317; P = .040), whereas brief-early versus continuous compression was not significant (P = .176). For cement porosity, intermittent compression yielded more porous mantles than brief-early (MD +1.5%, 95% CI 0.9-2.2; P < .001) and continuous (MD +0.9%, 95% CI 0.3-1.6; P = .006); continuous versus brief-early was not significant (P = .091). Cement-bone contact area was greatest with brief-early and lowest with intermittent; only the brief-early versus intermittent comparison reached significance (MD +20 mm², 95% CI 8-32; P = .004).

Conclusion: In this ovine glenoid model, brief-early compression followed by undisturbed curing produced a more favorable cement-bone micro-architecture than an intermittent pattern and a numerically more favorable profile than continuous compression without altering penetration depth; however, the brief-early versus continuous comparison did not reach statistical significance. These micro-computed tomography findings support brief-early compression for pegged all-polyethylene glenoids, but remain hypothesis-generating pending mechanical validation.

Keywords: Anatomic total shoulder arthroplasty; biomechanics; bone-cement interface; cement pressurization; glenoid component; micro-computed tomography; ovine scapulae; polymethylmethacrylate.