Periprosthetic joint infection (PJI) is one of the most dreaded complications after arthroplasty surgery; thus numerous approaches have been undertaken to equip metal surfaces with antibacterial properties. Due to its antimicrobial effects, silver is a promising coating for metallic surfaces, and several types of silver-coated arthroplasty implants are in clinical use today. However, silver can also exert toxic effects on eukaryotic cells both in the immediate vicinity of the coated implants and systemically. In most clinically-used implants, silver coatings are applied on bulk components that are not in direct contact with bone, such as in partial or total long bone arthroplasties used in tumour or complex revision surgery. These implants differ considerably in the coating method, total silver content, and silver release rates. Safety issues, such as the occurrence of argyria, have been a cause for concern, and the efficacy of silver coatings in terms of preventing PJI is also controversial. The application of silver coatings is uncommon on parts of implants intended for cementless fixation in host bone, but this option might be highly desirable since the modification of implant surfaces in order to improve osteoconductivity can also increase bacterial adhesion. Therefore, an optimal silver content that inhibits bacterial colonization while maintaining osteoconductivity is crucial if silver were to be applied as a coating on parts intended for bone contact. This review summarizes the different methods used to apply silver coatings to arthroplasty components, with a focus on the amount and duration of silver release from the different coatings; the available experience with silver-coated implants that are in clinical use today; and future strategies to balance the effects of silver on bacteria and eukaryotic cells, and to develop silver-coated titanium components suitable for bone ingrowth.

Cite this article: Bone Joint J 2021;103-B(3):423–429.

Uncemented implants combining antimicrobial properties with osteoconductivity would be highly desirable in revision surgery due to periprosthetic joint infection (PJI). Silver coatings convey antibacterial properties, however, at the cost of toxicity towards osteoblasts. On the other hand, topological modifications such as increased surface roughness or porosity support osseointregation but simultaneously lead to enhanced bacterial colonization. In this study, we investigated the antibacterial and osteoconductive properties of silver-coated porous titanium (Ti) alloys manufactured by electron beam melting, rendering a macrostructure that mimics trabecular bone.

Trabecular implants with silver coating (TR-Ag) or without coating (TR) were compared to grit-blasted Ti6Al4V (GB) and glass cover slips as internal controls. Physicochemical characterization was performed by X-ray diffraction (XRD) and energy dispersive X-rays (EDX) together with morphological characterization through electron scanning microscopy (SEM). Bacterial adherence after incubation of samples with Staphylococcus (S.) aureus and S. epidermidis strains harvested from PJI patients was quantitatively assessed by viable count after detachment of adherent bacteria by collagenase/dispase treatment. Primary human osteoblasts (hOB) were used to investigate the osteoconductive potential by lactate dehydrogenase (LDH) and alkaline phosphatase (ALP) activity. Cell morphology was investigated by fluorescence microscopy after staining with carboxifluorescein diacetate succinimidyl ester (CFDA-SE) and 4′,6-diamidino-2-phenylindole (DAPI).

The trabecular implants depicted a porosity of 70% with pore sizes of 600µm. The amount of silver analyzed by EDX accounted for 35%wt in TR-Ag but nil in TR. Silver-coated TR-Ag implants had 24% lower S. aureus viable counts compared to non-coated TR analogues, and 9% lower compared to GB controls. Despite trabecular implants, both with and without silver, had higher viable counts than GB, the viable count of S. epidermidis was 42% lower on TR-Ag compared to TR. The percentage of viable hOB, measured by LDH and normalized to controls and area at 1 day, was lower on both TR-Ag (18%) and on TR (13%) when compared with GB (89%). However, after 1 week, cell proliferation increased more markedly on trabecular implants, with a 5-fold increase on TR-Ag, a 3.4-fold increase on TR, and a 1.7-fold increase on GB. Furthermore, after 2 weeks of hOB culture, proliferation increased 20-fold on TR-Ag, 29-fold on TR, and 3.9-fold for GB, compared to 1 day. The osteoconductive potential measured by ALP illustrated slightly higher values for TR-Ag compared to TR at 1 day and 2 weeks, however below those of GB samples. Cell morphology assessed by microscopy showed abundant growth of osteoblast-like cells confined to the pores of TR-Ag and TR.

Overall, our findings indicate that the silver coating of trabecular titanium exerts limited cytotoxic effects on osteoblasts and confers antimicrobial effects on two PJI-relevant bacterial strains. We conclude that improving material design by mimicking the porosity and architecture of cancellous bone can enhance osteoconductivity while the deposition of silver confers potent antimicrobial properties.

The success of biomaterials lies in the direct interaction with the host tissue. Calcium phosphates (CaP) stand as an alternative graft material for bone regeneration due to their similar composition to natural bone. Few studies have focused on the early stages of bone-like material remodeling by osteoclasts (OC), though the CaP fate is to be resorbed and then replaced by new bone. Instead, to understand how osteoclasts modify the CaP surface and initiate resorption, so as to influence subsequent osteoblast activities and bone formation, is mandatory.

Sintered hydroxyapatite (s-HA) and biomimetic hydroxyapatite with two different microstructures (b-HA-C, coarse and b-HA-F, fine) discs (1500×250 µm²) were produced from the same reagents [1]. Tissue culture polystyrene (TCPS) was used as control. Precursor human OC from buffy coats were seeded on ceramic substrates [6·10⁶cells/cm²] and supplemented with RANKL-containing osteoblast supernatant as differentiation medium over 21 days. Cell interaction with the biomaterials was investigated in terms of OC adhesion and differentiation, with gene expression, tartrate-resistant acid phosphatase (TRAP) and Hoechst staining for OC maturation. Cell culture supernatants were analyzed for ionic exchange, namely Ca and P, due to biomaterials or cells. Osteoclasts morphology was evaluated using SEM at 21 days. Innovatively, focused ion beam (FIB) was used to evaluate biomaterial structure beneath the OC to further investigate the resorption effects. To this aim, selected OC were cut cross-sectioned using a Gallium ion beam at an acceleration of 30KV, followed by a coarse milling at 10nA and a deposition of platinum to achieve a fine milling at 500pA.

Clear differences in cellular behavior were noted relative to the different substrate microstructures. Control TCPS and s-HA showed similar TRAP-positive staining and gene expression for mature OC. Several resorption pits with partial dissolution of the equiaxial grains of s-HA were noticed. b-HA substrates also showed attached and differentiated TRAP-positive OC, but gene expression resulted lower than control and s-HA. However, morphological evaluation with SEM-FIB interestingly showed early stages of osteoclast-mediated degradation on b-HA-F, i.e.an increased surface roughness in the substrate underlying cells. B-HA-C also showed attached and mature OC with a scarce degradation activity

FIB technique has been applied to cell-seeded CaP and shown as a viable method to investigate OC morphology and resorption. Though gene expression showed similarities for both biomimetic substrates, substrate morphology observed underneath OC was significantly different. b-HA-F showed early stages of OC mediated degradation underneath well spread cells similar to those seen on s-HA. No resorptive activity was found on b-HA-C even though gene expression values were similar to b-HA-F: both the acute ion exchange and the surface tortuosity on b-HA-C could explain the difficulty with the resorptive process by OC. In conclusion focused ion beam technique complements SEM imaging and may disclose changes in the inner structure of materials due to cell/material interactions.

The regenerative potential of bone grafts is tightly linked to the interaction of the biomaterial with the host tissue environment. Hence, strategies to confer artificial extracellular matrix (aECM) cues on the material surface are becoming a powerful tool to trigger the healing cascade and to stimulate bone regeneration. The use of glycosaminoglycans (GAGs), such as heparin, as aECM components has gained interest in the last years as a strategy to improve biological response. Calcium phosphates (CaP) are extensively used as bone grafts, however no studies have investigated the effect of GAG functionalisation on their surface. Some authors have focused on the effects of GAGs on osteoblastic cells, however, little work has been performed on the interaction with osteoclasts (OC), and still the reported effects are controversial [1]. The aim of this study was to investigate the effect of heparin on osteoclastic fate in terms of adhesion and differentiation.

Sintered CaP (β-TCP) and biomimetic CaP (calcium-deficient hydroxyapatite, CDHA) discs were synthesized at 1100 ºC and at 37ºC, respectively. Heparinisation was achieved though silane coupling (APTES) followed by amidation in the presence of EDC/NHS to covalently link heparin. The osteoclast response of heparinised (H) vsnon-heparinised substrates was studied using human monocytes as OC precursors. Tissue culture plastic (TCPS) was used as a control sample. Cell densities were 6·10⁶and 3·10⁶cells/cm²for biomaterials and TCPS, respectively. Cell cultures were supplemented every 3 days with 25% supernatant of osteoblast-like cell line as a source of RANKL, as well as other stimulating factors [2]. Tartrate-resistant acid phosphatase and Hoechst staining were used to evaluate OC adhesion, differentiation and morphology at different time points from seeding on the surfaces (14–21–28 days).

OC precursors showed adhesion on all substrates. β-TCP and β-TCP-H hosted higher number of OC precursors which might be related to the smoother sintered surface of the materials. Oppositely, the high roughness of CDHA and CDHA-H hamper the adhesion of OC, hence a lower number of cells was observed on heparin-coated and uncoated biomimetic apatites. However, the maturation of OC precursors was found to take place at earlier times (14days) on biomimetic substrates compared to sintered ones. TCPS, CDHA, CDHA-H and β-TCP-H showed clearly differentiated OC at 14 days, as revealed by TRAP positivity and multinuclearity. Interestingly, CDHA-H and β-TCP-H induced the highest multinuclearity among all differentiated OC. Both heparinised substrates point at an enhancing effect of heparin on OC maturation.

OC precursors are able to differentiate on β-TCP and CDHA substrates, a process enhanced when heparin functionalisation is performed on the materials surface. In our hands heparinisation is promoting OC differentiation at early time points, similarly to TCPS control. Interestingly, heparin substrates induced larger TRAP positive-OC and higher multinuclearity in the mature OC than TCPS control. As pointed out by Irie et al., heparin might interact through the RANKL/OPG ratio [3], thus inhibiting OPG activity and enhancing RANKL which triggers OC maturation.