Abstract
Introduction
Three-dimensional (3D) printing of porous titanium implants marks a revolution in orthopaedics, promising enhanced bony fixation whilst maintaining design equivalence with conventionally manufactured components. No retrieval study has investigated differences between implants manufactured using these two methods. Our study was the first to compare these two groups using novel non-destructive methods.
Materials and methods
We investigated 16 retrieved acetabular cups divided into ‘3D printed’ (n = 6; Delta TT) and ‘conventional’ (n = 10; Pinnacle Porocoat). The groups were matched for age, time to revision, size and gender (Table 1). Reasons for revision included unexplained pain, aseptic loosening, infection and ARMD. Visual inspection was performed to evaluate tissue attachment. Micro-CT was used to assess clinically relevant morphometric features of the porous structure, such as porosity, depth of the porous layer, pore size and strut thickness. Scanning electron microscopy (SEM) was applied to evaluate the surface morphology.
Results
Significant differences (p = 0.0002) were found for all morphometric parameters (Table 2). Microscopic analysis revealed uniform beads over the backside of conventional implants, due to the manufacturing technique (Figure 1a). Conversely, beads of random size were found on 3D printed implants, representing a by-product of the manufacturing process, where some starting powder particles are not completely fused together (Figure 1b).
The two groups showed similar tissue attachment (3D printed 76.9 ± 27.1%; conventional 73.8 ± 12.2%; p = 0.2635).
Conclusion
This was the first study to analyse retrieved 3D printed orthopaedic implants. Differences were found between these and conventional implants, but both literature and registry data do not suggest a short-mid-term clinical issue with 3D printed components.
Similar tissue on growth suggested a comparable behaviour with bone in situ. The key difference is the presence of the particles on 3D printed implants, whose clinical significance needs to be investigated.
For any figures or tables, please contact the authors directly.
