Shape fidelity, mechanical and biological performance of 3D printed polycaprolactone-bioactive glass composite scaffolds.

Direct ink writing (DIW) is a promising extrusion-based 3D printing technology, which employs an ink-deposition nozzle to fabricate 3D scaffold structures with customizable ink formulations for tissue engineering applications. However, determining the optimal DIW process parameters such as temperature, pressure, and speed for the specific ink is essential to achieve high reproducibility of the designed geometry and subsequent mechano-biological performance for different applications, particularly for porous scaffolds of finite sizes (total volume > 1000 mm3) and controlled pore size and porosity. The goal of this study was to evaluate the feasibility of fabricating Polycaprolactone (PCL) and bio-active glass (BG) composite-based 3D scaffolds of finite size using DIW. 3D-scaffolds were fabricated either as cylinders (10 mm diameter; 15 mm height) or cubes (5 × 5 × 5 mm3) with height/width aspect ratios of 1.5 and 1, respectively. A rheological characterization of the PCL-BG inks was performed before printing to determine the optimal printing parameters such as pressure and speed for printing at 110 °C. Microstructural properties of the scaffolds were analyzed in terms of overall scaffold porosity, and in situ pore size assessments in each layer (36 pores/layer; 1764 pores per specimen) during their fabrication. Measured porosity of the fabricated specimens-PCL: x¯ =46.94%, SD = 1.61; PCL-10 wt%BG: x¯ = 48.29%, SD = 5.95; and PCL-20 wt% BG: x¯=50.87%, SD = 2.45-matched well with the designed porosity of 50%. Mean pore sizes-PCL [x¯ = 0.37 mm (SD = 0.03)], PCL-10%BG [x¯ = 0.38 mm (SD = 0.07)] and PCL-20% BG [x¯ = 0.37 mm (SD = 0.04)]-were slightly fairly close to the designed pore size of 0.4 mm. Nevertheless there was a small but consistent, statistically significant (p < 0.0001) decrease in pore size from the first printed layer (PCL: 0.39 mm; PCL-10%BG: 0.4 mm; PCL-20%BG: 0.41 mm) to the last. SEM and micro-CT imaging revealed consistent BG particle distribution across the layers and throughout the specimens. Cell adhesion experiments revealed similar cell adhesion of PCL-20 wt% BG to pure PCL, but significantly better cell proliferation - as inferred from metabolic activity - after 7 days, although a decrease after 14 days was noted. Quasi-static compression tests showed a decrease in compressive yield strength and apparent elastic modulus with increasing BG fraction, which could be attributed to a lack of adequate mechanical bonding between the BG particles and the PCL matrix. The results show that the inks were successfully generated, and the scaffolds were fabricated with high resolution and fidelity despite their relatively large size (>1000 mm3). However, further work is required to understand the mechano-biological interaction between the BG particle additives and the PCL matrix to improve the mechanical and biological properties of the printed structures.

A new bone surrogate model for testing interbody device subsidence.

STUDY DESIGN: An in vitro biomechanical study investigating interbody device subsidence measures in synthetic vertebrae, polyurethane foam blocks, and human cadaveric vertebrae. OBJECTIVE: To compare subsidence measures of bone surrogates with human vertebrae for interbody devices varying in size/placement. SUMMARY OF BACKGROUND DATA: Bone surrogates are alternatives when human cadaveric vertebrae are unavailable. Synthetic vertebrae modeling cortices, endplates, and cancellous bone have been developed as an alternative to polyurethane foam blocks for testing interbody device subsidence. METHODS: Indentors placed on the endplates of synthetic vertebrae, foam blocks, and human vertebrae were subjected to uniaxial compression. Subsidence, measured with custom-made extensometers, was evaluated for an indentor seated either centrally or peripherally on the endplate. Failure force and indentation stiffness were determined from force-displacement curves. RESULTS: Subsidence measures in human vertebrae varied with indentor placement: failure forces were higher and indentors subsided less with peripheral placement. Subsidence measures in foam blocks were insensitive to indentor size/placement; they were similar to human vertebrae for centrally placed but not for peripherally placed indentors. Although subsidence measures in synthetic vertebrae were sensitive to indentor size/placement, failure force and indentation stiffness were overestimated, and subsidence underestimated, for both centrally placed and peripherally placed indentors. CONCLUSION: The synthetic endplate correctly represented the human endplate geometry, and thus, failure force, stiffness, and subsidence in synthetic vertebrae were sensitive to indentor size/placement. However, the endplate was overly strong and thus synthetic vertebrae did not accurately model indentor subsidence in human cadaveric vertebrae. Foam blocks captured subsidence measures more accurately than synthetic vertebrae for centrally placed indentors, but because of their uniform density were not sufficiently robust to capture changes generated from different indentor sizes/placements. The current bone surrogates are not accurate enough in terms of material property distribution to completely model subsidence in human cadaveric vertebrae.