Optimization of 3D-printed titanium interbody cage design. Part 2: An in vivo study of spinal fusion in sheep.

Background context: 3D-printed titanium cage designs can incorporate complex, porous features for bone ingrowth and a greater surface area for minimizing subsidence. In a companion study (Part 1), we determined that increased surface area leads to decreased subsidence; however, it remains unclear how increasing the cage surface area, resulting in a smaller graft aperture, influences fusion.

Purpose: We evaluated the effects of surface area of 3D-printed titanium cages and the use of autologous bone grafts on spinal fusion in sheep.

Study design: In vivo large animal study in 12 sheep.

Methods: Interbody fusion was performed in 12 adult sheep at 24 levels (L2-3 and L4-5) using 3D-printed titanium cages with bilateral pedicle screw fixation. The cage designs varied in aperture: standard (low endplate surface area), small (medium endplate surface area), or none (high endplate surface area). These cages were packed with autologous iliac crest bone grafts (ICBG). A fourth group was implanted without bone grafts, using the no-aperture cage. Fusion was evaluated at 16 weeks via manual palpation, microcomputed tomography (microCT), histology, and histomorphometry.

Results: Standard, small, and no-aperture cages packed with ICBG resulted in high fusion rates (80%, 100%, and 83%, respectively) at 16 weeks by manual palpation, and these results were not significantly different. Implantation without ICBG was associated with a significantly lower fusion rate (33%, p<.05). Histological, histomorphometry, and microCT results supported the findings obtained by manual palpation; findings from these modalities showed new bone spanning the vertebral endplates in the spines graded as fused by manual palpation.

Conclusions: Similar fusion results for standard, small, and no-aperture cage designs packed with ICBG suggest that aperture size does not influence fusion results in the sheep model. However, without ICBG grafting, fusion was significantly decreased, suggesting that graft material is necessary to predictably obtain fusion in this model. When the in vitro subsidence data (companion study, Part 1) is considered with the in vivo fusion data described here, porous 3D-printed titanium cages with maximal surface endplate contact and bone grafting perform favorably, resulting in low subsidence and high fusion rates.

Clinical significance: 3D-printed porous titanium interbody cages are novel devices with increasing clinical use. The study results show that the aperture size of the interbody cage did not influence fusion in a large animal (sheep) model. The use of bone graft material was the most important variable affecting fusion. These data suggest that the clinical use of 3D Ti cages without graft material should be avoided.