Abstract
Background context: Cage subsidence is a complication of interbody fusion associated with poor clinical outcomes. 3D-printed titanium interbody cages allow for the alteration of features such as stiffness and porosity. However, the influence of these features on subsidence and their biological effects on fusion have not been rigorously evaluated. Purpose: This 2-part study sought to determine how changes in 3D-printed titanium cage parameters affect subsidence using an in vitro bone model (Part 1) and biological fusion using an in vivo sheep model (Part 2). Study design: Biomechanical foam block model. Methods: In Part 1 of this study, 9 implant types were tested (8 per implant type). The implant types included 7 3D-printed titanium interbody cages with various surface areas, porosities, and surface topographies, along with 1 standard polyetheretherketone (PEEK) cage and 1 solid titanium cage. Subsidence testing was performed in a standardized foam block model using 2 different densities of foam. Digital imaging correlation was used to determine the relative vertical displacement of the interbody cage-foam block construct. Results: Subsidence decreased as the surface contact area with the bone model increased (all p <=.01). Increased porous surface topography increased subsidence, while a nonporous surface significantly decreased subsidence (all p<.001). Subsidence did not differ based on changes in implant porosity (all p >=.35) or material property/modulus (all p >=.19). Subsidence was significantly decreased with the higher density foam (p<.001). The stiffness of the implant was affected by porosity (all p<.02) and smooth surface topography (p=.01) but not by lumen size (all p >=.15). Stiffness did not differ between porous titanium and PEEK implants (p=.96), which were both less stiff than solid titanium implants (both p<.001). Surface area negatively correlated with subsidence (r=-0.786, p=.012) but was not correlated with stiffness (r=0.560, p=.12). Conclusions: Implant surface area and surface topography greatly influenced interbody subsidence. Apparent stiffness, implant porosity, and material property did not affect subsidence in this in vitro model. Higher foam density also led to lower subsidence than low-density foam. Biological response in the in vivo setting likely also influences clinical subsidence, which is evaluated in the companion study (Part 2).