Bioactivated lubricant-infused surfaces: A dual-action strategy for enhancing osseointegration and preventing implant-associated infections

Background

Implant-associated infections arise when bacteria adhere to implant surfaces and form biofilms; the authors cite infection rates of 1 to 5% for orthopedic implantation, which can exceed 20% for internal fixation of complex open fractures. Hydrated polymer coatings such as poly(ethylene glycol) and zwitterionic polymers are vulnerable to the mechanical damage of orthopedic surgery, and superhydrophobic surfaces lose repellency as trapped air dissolves. A lubricated orthopedic implant surface (LOIS) based on slippery liquid-infused porous surfaces repelled bacteria in vivo, but its repellency can also hinder osteoblast adhesion, which may make it unsuitable for implants such as intramedullary nails that require osseointegration.

Design and approach

Titanium is etched in sodium hydroxide to form a nanoporous surface, fully fluorinated by vapor deposition of trichloro(1H,1H,2H,2H-perfluorooctyl)silane (TPFS), and then briefly exposed to a secondary oxygen plasma that removes part of the TPFS. 3-Aminopropyltriethoxysilane (APTES) is deposited on the exposed sites, BMP-2 is coupled to its amine groups by carbodiimide chemistry, and a perfluoropolyether lubricant is infused. The fluorinated regions hold the lubricant that repels nonspecific adhesion, while the immobilized BMP-2 binds receptors on osteoblasts. The authors propose that immobilized BMP-2 spatially confines BMP-2 receptors and integrins, so that only cells that transduce BMP-2 signals, such as osteoblasts, spread and differentiate. A secondary plasma of 30 W for 5 s was chosen because it retained hydrophobicity while distributing amine groups over the surface.

How it was evaluated

Surfaces were characterized by scanning electron and atomic force microscopy, nanoindentation, water contact angle, fluorescence labeling, X-ray photoelectron spectroscopy and an ELISA-based estimate of immobilized BMP-2. BILS was compared with bare titanium and LOIS for contact and sliding angles of six liquids, albumin and fibrinogen adsorption, clotting of recalcified horse blood, and adhesion and biofilm formation of Pseudomonas aeruginosa and MRSA at 1.5 × 10^8 CFU/mL. MC3T3-E1 osteoblasts and NIH 3T3 fibroblasts were used for viability, adhesion, co-culture, coverage over 28 days, alkaline phosphatase (ALP) activity and calcium deposition. Custom titanium intramedullary nails (74 mm long, 6 mm and 4 mm diameters) were tested for coating stability and bending properties and implanted in femoral fracture models in 16 male New Zealand white rabbits divided into bare negative, bare positive, LOIS and BILS groups; all but the bare negative nails were incubated in MRSA for 12 h before surgery, and outcomes at 28 days included surgical-site inspection, soft-tissue histology, X-ray, tartrate-resistant acid phosphatase (TRAP) staining, micro-computed tomography and bone-implant contact. Pull-out tests used three additional rabbits (bare, LOIS and BILS) without bacterial incubation.

Key results

Etching for 6 h produced a root-mean-square roughness of about 84.12 nm, and full TPFS silanization gave a contact angle of about 150°. BMP-2 immobilization efficiency was about 92.43% at 100 ng/mL, falling to about 78% and 47% at 250 and 500 ng/mL, and the 100 ng/mL condition was selected to give a surface density of about 80 ng/cm2. BILS and LOIS showed no significant differences in contact and sliding angles for the six liquids, indicating that partial silanization did not compromise lubricant retention; the abstract reports sliding angles below 5°.

Albumin and fibrinogen coverage on both lubricated surfaces was below about 1%, and neither surface formed blood clots. After 6 h, LOIS and BILS carried no adherent P. aeruginosa or MRSA, and after 48 h they showed no biofilms. In the results section, colony counts on bare titanium were about 10^10-fold those on LOIS and BILS, with no significant difference between the two lubricated surfaces; the abstract states this difference as 10-fold.

Cell viability on BILS did not differ significantly from that on bare titanium. Fibroblasts adhered only slightly and remained rounded, whereas osteoblasts adhered and spread with strong vinculin expression, and in co-culture only osteoblasts spread. Osteoblast coverage on BILS reached about 75% over 28 days, against about 80% on bare titanium and significantly low coverage on LOIS; the abstract gives 30% after one day and up to 70% after three weeks for BILS. ALP activity on BILS was 1.20 ± 0.2 nmol at 7 days and 5.79 ± 0.9 nmol at 14 days, the latter not significantly different from bare titanium (6.33 ± 0.31 nmol), and calcium deposition on BILS at 21 and 28 days was almost similar to bare titanium, while LOIS showed no noticeable calcium.

Coated nails reduced albumin adsorption even after insertion into and removal from ex vivo rabbit femurs, and etching did not significantly change flexural modulus or strength. At 28 days, the MRSA-incubated bare nails produced severe inflammation with pus, whereas the LOIS and BILS groups showed no signs of inflammation on inspection or histology. Bone volume at the fracture site was significantly lowest in the bare positive group, and BILS showed significantly higher bone-implant contact than the bare positive and LOIS groups, and in the LOIS group a gap separated aggregated osteoblasts from the implant surface. In the separate pull-out test, the maximum force was about 5 N for LOIS, about 45 N for BILS and about 122 N for the uninfected bare nail.

Open questions

The authors observed no complete osseointegration with the host bone in any group and attribute this to the surgery, since thinner implants were used for the brittle rabbit femur. BILS showed lower pull-out force than the uninfected bare implant. The proposed physical mechanism of osteoblast adhesion, in which larger cells displace weakly held lubricant over immobilized BMP-2, remains a hypothesis that the authors state should be explored further.

How it connects to other research

A heparinized titanium implant delivering gentamicin and BMP-2 (Lee et al., Bone 2012) combines antibiotic release with an osteoinductive protein, whereas BILS relies on a lubricant layer to repel bacteria and on immobilized BMP-2 to select osteoblasts, without an antibiotic.