(W0930-12-71) 3D Printed Implants for the Treatment of Periprosthetic Joint Infections

Purpose: Periprosthetic joint infections (PIJs) occur in around 1% for all hip arthroplasties and up to 2% after knee replacement. PJIs have a negative impact on healthcare systems and patients. Currently, the most common treatments used to prevent PJIs are not fully effective and present some drawbacks. In this scenario, there is an unmet clinical need to find innovative solutions to reduce the risk of PJIs. Ideally, a combination of a broad spectrum antibacterial with an antifungal drug would be ideal to prevent PJIs. Also, drug delivery systems should not alter the biomechanical properties of the joint and should be ideally placed externally on top of the prosthesis with a millimetric thickness that does not hinder the surgical procedure or the comfort of the patient. The hypothesis underpinning this work is that a controlled drug release implant for parenteral administration, well adapted to the prosthesis shape with a complementary drug release profile (immediate and sustained release, respectively, to prevent both acute and delayed infections) would be ideal to prevent and treat PJIs. Hence, the objective of this work was to develop a personalized 3D printed implant loaded with vancomycin and amphotericin B (AmB) as board spectra antibacterial and antifungal drugs able to control the release overtime and well fitted to the prosthesis morphology.

Methods: 3D printed personalized implants were desiged on Tinkercad software matching the dimensions of hip prosthesis. Implants were printed using a Fuse Deposition Modeling (FDM) printer (Flashforge, USA) with a commercially available filament made of polyvinyl alcohol and polyethylenglycol (Hydrosupport). Drugs were loaded by passive diffusion followed by a drying step and a sterilization stage under UV radiation. Drug release profile and stability were evaluated as well as in vitro efficacy against a wide range of different species of Candida spp. and Staphylococcus spp., following the CLSI standards.

Results: 3D printed implants were printed with 1 mm in height fitting within the gap remaining between the hip and the femur bone (Fig. 1). Adhesive properties were optimal being required less thant 5 min to complete the whole placement and drying stage within the human hip prosthesis. The optimal time for drugs to cross inside the implant by passive diffusion was 4 h. The diffusion kinetics rates were 0.096%/h for vancomycin and 0.0042% for AmB (Fig. 2a). After drug loading, 3D printed implants were sterilized by UV radiation. No significant drug degradation was observed in any of the cases.

Vancomycin exhibited a faster release in physiological media compared to AmB. All drug content was release upon one hour of exposure to the media which concentration remained stable over a 48 h period. In contrast, AmB showed a sustained release over a 24 h period. A spring and parachute effect was observed for the latter drug, and drug concentration in the media decrease over time (Fig. 2b). The concentrations in the media of both drugs was maintained above the IC₅₀ for most Candida spp and Staphylococcus spp respectively.

In vitro efficacy showed that C. albicans, C. parapsilosis and C. glabrata were susceptible to AmB-loaded implants, with an inhibition halo greater than 15 mm, which indicated a good diffusion of the drug across the agar (Fig. 3a). However, a dose dependent efficacy was observed against C. Krusei. Similar results were obtained for those 3D printed implants loaded with vancomycin. An inhibition area greater than 15 mm were obtained when tested against (Fig. 3b) S. epidermidis and S. aureus.

Conclusion: 3D printed parenteral implants have been successfully engineered by FDM being perfectly fitted to the acetabular component of hip prosthesis. Implants were loaded with vancomycin and AmB remaining stable after sterilization by UV radiation. Both drugs showed a complementary release in physiological media with optimal in vitro efficacy against both bacterial and fungal strains typically responsible for biofilm formation in PJIs. This can be an excellent strategy for the personalized management of PJIs.

Acknowledgements:This work has been funded by the European Society of Clinical Microbiology and Infectious Diseases (ESCMID) with a Research Grant [year 2021, ID: 16306] to Dolores R. Serrano and also by Madrid Community and Fondo Social Europeo (PEJ-2019-AI/IND13859).


Figure 1. Personalised 3D printed implants. a) 3D printed implants. b) Implants adhered to the surface of the hip prosthesis after wetting.


Figure 2. Drug diffusion and release kinetics from 3D printed implants. a) Drug loading of vancomycin and AmB through passive diffusion. b) Release profiles of vancomycin and AmB in implants loaded with a single or both drugs.


Figure 3. Antimicrobial efficacy of the 3D printed implants. a) Inhibition halo against C. albicans, C. parapsilosis, C. glabrata and C. krusei. b) Inhibition halo against S. epidermidis and S. aureus. *, p < 0,05.