Recent Advances in the Evaluation of Antimicrobial Materials for Resolution of Orthopedic Implant-Associated Infections In Vivo

Abstract

While orthopedic implant-associated infections are rare, revision surgeries resulting from infections incur considerable healthcare costs and represent a substantial research area clinically, in academia, and in industry. In recent years, there have been numerous advances in the development of antimicrobial strategies for the prevention and treatment of orthopedic implant-associated infections which offer promise to improve the limitations of existing delivery systems through local and controlled release of antimicrobial agents. Prior to translation to in vivo orthopedic implant-associated infection models, the properties (e.g., degradation, antimicrobial activity, biocompatibility) of the antimicrobial materials can be evaluated in subcutaneous implant in vivo models. The antimicrobial materials are then incorporated into in vivo implant models to evaluate the efficacy of using the material to prevent or treat implant-associated infections. Recent technological advances such as 3D-printing, bacterial genomic sequencing, and real-time in vivo imaging of infection and inflammation have contributed to the development of preclinical implant-associated infection models that more effectively recapitulate the clinical presentation of infections and improve the evaluation of antimicrobial materials. This Review highlights the advantages and limitations of antimicrobial materials used in conjunction with orthopedic implants for the prevention and treatment of orthopedic implant-associated infections and discusses how these materials are evaluated in preclinical in vivo models. This analysis serves as a resource for biomaterial researchers in the selection of an appropriate orthopedic implant-associated infection preclinical model to evaluate novel antimicrobial materials.

Graphical Abstract

1. INTRODUCTION

Each year, >1.35 million primary arthroplasties, including >750 000 knee, >500 000 hip, and >100 000 shoulder prostheses, are performed in the United States, with procedure volumes for each of the mentioned subtypes increasing steadily year over year.^1,2 While implant-associated orthopedic infections occur in approximately 1–3% of primary arthroplasty patients, clearance of infections is nontrivial.³ For example, patients may be subjected to prolonged antibiotic therapy, repetitive revision surgeries resulting in functional impairment, and the possibility of permanent handicap (e.g., amputation).³ Infection recurrence following revision surgeries occurs in 9–50% of patients, depending upon the type of bacteria involved, if the infection is acute or chronic, and the type of revision surgery (e.g., one- versus two-stage revision, irrigation and debridement with implant retention).^4–7 Due to the rising age and longevity of the general population, the number of implant-associated infection revision procedures has been projected to increase by 43–182% from 2014 to 2030.⁸ Implant-associated infections also pose a substantial socioeconomic burden, costing the healthcare system an additional ~$13 000 per patient.⁹

Implant-associated orthopedic infections can arise through several means pre- and postoperatively. For example, these infections can be initially acquired through mechanisms such as a breach of sterility during surgery (i.e., surgical site infections), acute trauma (i.e., open fractures),^10,11 hematogenous seeding,^12–14 or postsurgical infiltration of skin micro-flora.¹⁵ An individual patient’s risk of developing an implant-associated orthopedic infection can be exacerbated by underlying comorbidities, such as obesity or diabetes,^3,16,17 or by failure to observe proper wound care practices during the postoperative period.¹⁵

Both biodegradable and nondegradable biomaterials used in orthopedic implants (e.g., metals, ceramics, polymers, etc.) have the potential to serve as a nidus for infection and bacterial attachment if the surgical site is infiltrated with bacteria.¹⁸ When implanted materials are infected, the typical course of treatment involves determining the susceptibilities of the causative pathogen(s), perioperative systemic antibiotics, and surgical debridement and removal and replacement of nondegradable devices.¹⁹ However, when biomaterials are effectively designed and engineered with antimicrobial properties, they can be applied to manage or treat implant-associated infections locally. Specifically, biomaterials have historically played a pivotal role in the prevention and treatment of implant-associated orthopedic infections. Since 1972, antibiotics have been directly incorporated into poly(methyl methacrylate) (PMMA) bone cement as a measure against orthopedic implant-associated infections. Antibiotic-laden PMMA thereby provides a dual functionality, mechanical stability (e.g., fixing a prosthesis or serving as a temporary spacer) and antimicrobial activity.²⁰ Nevertheless, incorporation of antibiotics into PMMA bone cement, for example, results in insufficient elution of the drug to effectively maintain therapeutic concentrations necessary to eradicate infections.^21,22 Consequently, there has been growing interest to develop materials that provide a more controlled and consistent antimicrobial activity for orthopedic implant applications. Materials that have been recently developed for prevention and mitigation of orthopedic implant infections include stimuli-responsive (e.g., to temperature, electrical/magnetic fields, microwaves, etc.) and nanocomposite materials.^23–30

While many of the antimicrobial materials have shown promise in initial in vitro studies, translating these systems to clinically relevant orthopedic animal models presents several challenges. There are species-specific differences in the manifestation of bacterial infections in animal models as well differences in bacterial infections between animal models and humans.³¹ Additionally, in clinical practice, there are patient comorbidities that can increase the risk of developing implant-associated infections (e.g., diabetes, obesity, etc.), and due to the difficulty of recapitulating these comorbidities in animal models these factors have only recently been explored in preclinical models.^32,33 Technological advancements (e.g., 3D-printing, bacterial genomic sequencing, etc.^33,34) have enabled the development of more clinically relevant and reproducible in vivo models that may facilitate and improve understanding of the development and treatment of implant-associated infections moving forward. Furthermore, advances in real-time in vivo imaging modalities (e.g., positron emission tomography (PET), fluorescence) and radioactive tracers and probes offer the potential to monitor the efficacy of antimicrobial orthopedic materials and progression of infection and inflammation noninvasively.^35–37

This Review highlights the advantages and limitations of antimicrobial materials used in conjunction with orthopedic implants for implant-associated infection applications (section 3) and how these materials are evaluated in preclinical in vivo models for the prevention or treatment of infection (sections 4 and 5). Our analysis is intended to serve as a resource to help guide biomaterials researchers in the selection of an appropriate implant-associated infection preclinical model for evaluation of novel antimicrobial materials.

2. METHODS

To construct this analysis, a literature search was performed in PubMed with the following keywords (July 2021): “orthopedic implant infection model” (1050 results), “orthopedic implant infection animal model” (401 results), “materials prevent orthopedic implant infection” (616 results), and “materials treat orthopedic implant infection” (1321 results). The results were screened, and the biomaterials and preclinical in vivo models highlighted in this Review were selected based upon their publication date (within the past decade), if they developed a material to address implant-associated infections (n = 45), and if they either established a novel orthopedic implant-associated infection model (n = 28) or used an orthopedic implant-associated infection model to evaluate the ability of biomaterials to prevent or treat infection (n = 88).

3. ADVANTAGES AND LIMITATIONS OF ANTIMICROBIAL BIOMATERIALS USED FOR PREVENTING/TREATING IMPLANT-ASSOCIATED INFECTIONS

Antimicrobial materials surveyed in this Review comprise those that combat infections directly and indirectly. Specifically, materials that exhibit bactericidal or bacteriostatic activity through the release of antimicrobial agents and materials that repel the adherence of bacteria (e.g., nanotopography, surface treatments) are included.^38–41 This Review primarily focuses on the material-level properties of the antimicrobial biomaterials used in conjunction with orthopedic implants and does not consider the design and properties of the orthopedic implant device/component itself (e.g., Kirschner wires, screws, etc.) unless it has been modified to exhibit antimicrobial activity (e.g., nanotopography, surface coating, etc.). For the context of this Review, an orthopedic antimicrobial biomaterial is defined as a material that has been designed to work in conjunction with permanent orthopedic implants to provide localized antimicrobial therapy and support the function of the implant (e.g., materials with antimicrobial activity alone and those with antimicrobial activity and osteoconductive properties). Tables 1–4 highlight the advantages and limitations of a selection of antimicrobial biomaterials from the past decade that have been developed to address orthopedic implant-associated infections.

Table 1.

Summary of Advantages and Limitations of Metallic Antimicrobial Biomaterials from the Past Decade Designed for Orthopedic Implant-Associated Infection Applications

metals
composition	tested in vivo?	antimicrobial activity	advantages	limitations	ref
titanium oxide nanocoating on titanium	no	bacterial clearance of 80–90% CFUs, inhibit adhesion 80% bacteria	(1) intrinsic activity	(1) leach metal particles (cytotoxicity/genotoxicity)	49, 50
			(2) long-lasting
			(3) decreased resistance
titanium oxide coated titanium and titanium hydride powder coated on porous titanium	no	inhibit adhesion 80% bacteria	(1) intrinsic activity	(1) leach metal particles (cytotoxicity/genotoxicity)	51
			(2) long-lasting
			(3) decreased resistance
silver and copper nanoparticle coated titanium oxide on Ti6Al4V	no	inhibit adhesion 100% bacteria	(1) intrinsic activity	(1) leach metal particles (cytotoxicity/genotoxicity)	52
			(2) long-lasting
			(3) decreased resistance
silver and zinc nanoparticle coated titanium oxide on Ti6Al4V	no	inhibit adhesion 100% bacteria	(1) intrinsic activity	(1) leach metal particles (cytotoxicity/genotoxicity)	53
			(2) long-lasting
			(3) promotes osseointegration
titanium–copper-oxide coated Ti6Al4V	no	2log10 decrease in bacterial adhesion	(1) intrinsic activity	(1) leach metal particles (cytotoxicity/genotoxicity)	56
			(2) long-lasting
fluorine- and phosphorus doped nanostructured Ti6Al4V	yes	not evaluated, used to detect presence of infection	(1) intrinsic activity	(1) leach metal particles (cytotoxicity/genotoxicity)	40
			(2) long-lasting
			(3) promotes osseointegration
magnesium alloy	yes	limited in vivo, requires modification	(1) intrinsic activity	(1) limited duration of activity	59
				(2) antimicrobial activity in vitro superior to in vivo

Table 4.

Summary of Advantages and Limitations of Composite Antimicrobial Biomaterials from the Past Decade Designed for Orthopedic Implant-Associated Infection Applications

combination materials/composites
composition	tested in vivo?	antimicrobial activity	advantages	limitations	ref
zirconium nitride coated ceramic covered Co–Cr–Mo	no	log10 decrease bacterial CFUs	(1) intrinsic activity	(1) short range of antimicrobial action (only prevent bacterial attachment to surface)	149
			(2) long-lasting
Titanium–niobium-nitride coated titanium	no	4-fold decrease in bacterial adhesion	(1) intrinsic activity	(1) leach metal particles (cytotoxicity/genotoxicity)	159
			(2) long-lasting
nanostructured silver-substituted fluorhydroxyapatite-titanium oxide coated titanium	no	bacterial clearance of 100% CFUs	(1) intrinsic activity	(1) leach metal particles (cytotoxicity/genotoxicity)	146
			(2) long-lasting
			(3) promote osseointegration
carboxymethyl chitosan/hyaluronic-acid-catechol conjugated vascular endothelial growth factor functionalized titanium	no	inhibit adhesion of 46–84% bacteria	(1) intrinsic activity	(1) short range of antimicrobial action (only prevent bacterial attachment to surface)	160
			(2) promotes osseointegration
zinc, cerium, selenium substituted hydroxyapatite/poly(sorbitol sebacate glutamate) coated titanium	yes	1 day activity zone of inhibition (in vitro)	(1) intrinsic activity	(1) leach metal particles (cytotoxicity/genotoxicity)	147
			(2) promotes osseointegration
silver hydroxyapatite coating on titanium	yes	10–20% decrease in bacterial biofilm adhesion	(1) intrinsic activity	(1) leach metal particles (cytotoxicity/genotoxicity)	148
			(2) long-lasting
			(3) promotes osseointegration
silver doped nano calcium phosphate coated Ti6Al4V	yes	significant reduction in bacterial adhesion relative to controls	(1) intrinsic activity	(1) leach metal particles (cytotoxicity/genotoxicity)	157
			(2) long-lasting
			(3) promotes osseointegration
Eudragit coated Ti6Al4V	yes	15 days activity zone of inhibition (in vitro)	(1) pH-triggered drug delivery	(1) require antibiotics	161
				(2) limited duration of activity
mesoporous silica microparticles in porous stainless steel	no	2–3log10 decrease in bacterial adhesion	(1) tunable drug delivery properties	(1) require antibiotics	162
				(2) limited duration of activity
copper-nanoparticle coated sulfonated poly(ether ether ketone)	yes	35-fold decrease in bacterial adhesion	(1) intrinsic activity	(1) leach metal particles (cytotoxicity/genotoxicity)	114
			(2) long-lasting
cationic liposomes in calcium sulfate	yes	bacterial clearance of 100% CFUs	(1) promote osseointegration	(1) require antibiotics	28
				(2) limited duration of activity
chitosan bonded borate bioglass particles	yes	81–87% clearance of infection in vivo	(1) promotes osseointegration	(1) require antibiotics	30
			(2) injectable	(2) limited duration of activity
brushite calcium phosphate functionalized poly(ether ether ketone)	Yes	inhibit adhesion 100% bacteria	(1) promotes osseointegration	(1) require antibiotics	113
				(2) limited duration of activity
cyclodextrin microparticles in PMMA	No	10–60 days activity zone of inhibition (in vitro)	(1) can be repeatedly filled with drug locally	(1) require antibiotics	29, 123–125
			(2) unaffected by bacterial biofilm

3.1. Metals.

A variety of metals and alloys have been used in preclinical models for the prevention and treatment of orthopedic implant-associated infections (Table 1). While many of the metals frequently utilized in orthopedic implants are not inherently antimicrobial (e.g., titanium and titanium alloys (Ti6Al4V),^42–44 cobalt–chromium–molybdenum (Co–Cr–Mo alloys),⁴⁵ stainless steel (316L),^46–48 etc.), surface modifications (e.g., atomic layer and plasma electro-deposition^49–53) can be used to coat the implant with antimicrobial metal nanoparticles (e.g., silver,^52–54 copper,^52,55,56 zinc,^53,57 magnesium,^58,59 gold, etc.).

Due to their intrinsic antimicrobial activity, coatings with silver, gold, copper, zinc, or magnesium are particularly advantageous because they can provide long-term broad-spectrum antimicrobial activity without the use of antibiotics (i.e., offer a decreased risk of stimulating bacterial resistance^53,60,61) and are gradually released over a prolonged period (30 days).⁶² Generally, smaller nanoparticles (~10–15 nm diameter) have improved biocompatibility, can penetrate bacterial cells, and have more potent antimicrobial activity than that of larger nanoparticles.^63,64 The geometry of nanoparticles also influences their antimicrobial activity (e.g., nanoparticles with sharp edges have increased antimicrobial activity).⁶⁵

There are several proposed mechanisms of action for the antimicrobial activity of metallic nanoparticles including production of intracellular reactive oxygen species and hydrogen peroxide damaging bacterial proteins, damage to the bacterial cell membrane via electrostatic penetration (e.g., cations bind to anionic lipopolysaccharides in membrane of bacteria cell), and prevention of the replication of bacterial DNA.^66,67 The amount of reactive oxygen species produced, for example, can be used to evaluate the antimicrobial activity of metallic nanoparticle implant coatings.^52,68 Alternatively, antimicrobial properties can be imparted to certain metals, such as titanium, when they are exposed to aerobic environments oxidizing the surface (TiO₂).^69,70 The surface of metallic implants can also be modified to have a nanotopography architecture.^38–40 A rough textured implant surface increases the surface area that the bacterial cells are exposed to, enhancing antimicrobial activity.^38,39 Certain surface modifications to metals including plasma electrolytic oxidation⁵³ and phosphorus doping⁴⁰ have been shown to enhance osseointegration of the antimicrobial materials. Plasma electrolytic oxidation is attractive due to its ease of use, high deposition rate,⁷¹ and ability to create multifunctional surfaces (e.g., antimicrobial and osteoconductive).⁵³ Phosphorus doping provides a bottle-shaped nanostructure that assists in osseointegration.⁴⁰

Nevertheless, antimicrobial metallic materials have several limitations. Specifically, they can release metallic ions and particles that can enter the lymphatic and circulatory system and cause damage when they accumulate in tissues (e.g., genotoxicity)^72–77 or are excreted (e.g., nephrotoxicity).^78,79 Ions from Co–Cr–Mo alloys are particularly problematic due to their carcinogenicity.⁸⁰ Additionally, metallic ions generally have a lack of specificity and have been shown to exhibit cytotoxicity to eukaryotic cells in high concentrations.⁵³ Therefore, in some cases, different metallic nanoparticles in the coating (e.g., silver and zinc, etc.) have been combined to offer comparable antimicrobial activity to individual metals while reducing cytotoxicity by decreasing the concentration of nanoparticles of a single composition.⁵³ Furthermore, since many metallic antimicrobial materials are nondegradable, they have the potential to harbor bacterial biofilms (i.e., dense cluster of sessile bacteria encapsulated in extracellular polymeric substance with protein, lipids, and teichoic acids⁸¹), which typically result in invasive debridement, removal, and replacement of the implant.

3.2. Polymers.

Natural and synthetic polymeric materials have been developed to prevent and treat orthopedic implant-associated infections (Table 2). While a number of polymeric materials have intrinsic antimicrobial activity (e.g., silicone nanotopography that can repel bacteria,^41,82 chitosan,⁸³ hyaluronic acid,^23,84,85 etc.), it is often necessary to incorporate antibiotics into polymeric materials to impart antimicrobial activity.^{23,24,83,86–105} The rate of enzymatic/hydrolytic degradation of biodegradable polymers such as chitosan,^24,83,99 hyaluronic acid,^23,85 alginate,^85,105 poly(lactic acid) derivatives (e.g., poly[l-lactic] acid, poly[d,l-lactide-co-lactide], poly[d-l-lactide]),^{87–89,94–96} polycaprolactone,^87,98,100 and poly(lactic-co-glycolic acid) (PLGA)^{35,98,100,101,104,106} dictates the rate of antibiotic release from the implanted material and subsequent duration of antimicrobial activity. An advantage of biodegradable polymeric materials is that their rate of degradation can be tailored through several means including their molecular weight and degree of crystallinity.¹⁰⁷ Specifically, for chitosan, several factors including the degree of deacetylation and molecular weight can influence the rate of degradation.¹⁰⁸ If chitosan has a higher degree of deacetylation (processed in alkaline conditions), it will have a greater hydrophilicity and have an increased degradation rate.¹⁰⁸ The degradation rate and hydrophilicity of hyaluronic acid can be influenced by the chemistry of the cross-linking agent used, for example,¹⁰⁹ whereas the degradation rate of hydrolytically cleavable copolymers, such as PLGA, for example, can be increased by incorporating a greater fraction of the relatively hydrophilic component poly(glycolic acid) at the molecular level.¹¹⁰ Additionally, naturally derived polymers (e.g., chitosan, hyaluronic acid, and alginate) exhibit excellent biocompatibility.¹¹¹

Table 2.

Summary of Advantages and Limitations of Polymeric Antimicrobial Biomaterials from the Past Decade Designed for Orthopedic Implant-Associated Infection Applications

polymers: natural and synthetic
composition	tested in vivo?	antimicrobial activity	advantages	limitations	ref
chitosan sponges	yes	7–21 days activity zone of inhibition	(1) tunable delivery kinetics	(1) may require antibiotic supplement	83
			(2) intrinsic activity	(2) limited duration of activity
			(3) biocompatibility
collagen sponges	yes	3–4log10 decrease bacterial CFUs	(1) biocompatibility	(1) require antibiotics	86
				(2) limited duration of activity
poly(lactic acid) diol and poly(caprolactone) diol coated polyethylene terephthalate	yes	up to 66 days activity (in vitro) and 20 days activity (in vivo) zone of inhibition	(1) tunable delivery kinetics	(1) require antibiotics	87
				(2) limited duration of activity
poly(l-lactic)acid coated poly(d,l-lactide-co-lactide)	no	not evaluated	(1) tunable delivery kinetics	(1) require antibiotics	88
				(2) limited duration of activity
poly(d-l-lactide) coating	no	inhibit adhesion 60% bacteria	(1) tunable delivery kinetics	(1) require antibiotics	89
				(2) limited duration of activity
micropatterned silicone	yes	inhibit adhesion of >90% bacteria	(1) intrinsic activity	(1) short range of antimicrobial action (only prevent bacterial attachment to surface)	41
			(2) long-lasting
xerogel coating (silane-based) on silicone	yes	82% reduction implant infection (in vivo)	(1) intrinsic activity	(1) limited duration of activity	135
sulfonated poly(ether ether ketone)	yes	bacterial clearance of 80–100% CFUs	(1) promote osteogenesis	(1) limited duration of activity	112
			(2) durability
cross-linked PEG	yes	bacterial clearance of 100% CFUs	(1) injectable (minimally invasive application)	(1) limited duration of activity	120
poly (ethylene imine)/poly(sodium-4-styrenesulfonate)/poly(allylamine hydrochloride)	no	inhibit adhesion 80% bacteria	(1) local drug synthesis	(1) cytotoxicity of poly(ethylenimine)	121

Nondegradable polymeric materials such as poly(ether ether ketone),^97,112–119 poly(ethylene glycol) (PEG),^{90–92,98,104,120} poly(ethyleneimine),^121,122 cyclodextrin,^29,123–125 poly(methyl methacrylate) (PMMA),^{29,34,86,118,119,123–134} polyethylene terephthalate,^37,87 silicone,^41,135 and polyesters^87,93 have the potential to serve as long-term antimicrobial materials. Since the release of antibiotics from nondegradable materials is not associated with the rate of degradation of the material, these materials primarily rely on diffusion and affinity-based interactions to dictate the release kinetics of antibiotics. Poly(ethyleneimine) is a versatile polymer that has been used in layer-by-layer coatings of orthopedic implants to regulate drug release,¹²¹ and in its quaternized form it is intrinsically antimicrobial.¹²² Nevertheless, it is important to consider that poly(ethyleneimine) has been shown to exhibit cytotoxic effects and the mechanisms of cytotoxicity are still being investigated.¹³⁶ Poly(ether ether ketone) and PMMA are particularly amenable for orthopedic load-bearing applications due to their mechanical properties (ability to resist wear following long-term cyclical loading).^137,138 In terms of prevention of implant-associated infections, degradable antimicrobial polymers are advantageous relative to nondegradable antimicrobial polymers. Specifically, if a bacterial biofilm is to form on a nondegradable material, complete eradication of the infection often requires a full surgical debridement and replacement¹³⁹ and biodegradable materials have been shown to have a decreased risk of developing bacterial biofilms (e.g., decreased surface area for attachment during degradation, etc.).¹³⁹ However, there are some nondegradable materials such as insoluble cross-linked cyclodextrin that has been shown to retain its ability to be repeatedly filled with antibiotics even in the presence of a bacterial biofilm.¹⁴⁰

Several polymeric antimicrobial biomaterials offer the advantage of stimuli-responsive properties that enable control of gelling and release of antibiotics. For example, chitosan functionalized materials have been formulated to be thermosensitive and gel following injection, enabling minimally invasive application.^23,24 Additionally, polymers including PEG, PLGA, and alginate have been formulated to gel in situ.^85,104,105

3.3. Ceramics.

A range of ceramic materials have been used in preclinical orthopedic antimicrobial applications including calcium sulfate,^28,141–145 hydroxyapatite,^{144,146–148} β-tricalcium phosphate,¹⁴⁵ zirconium nitride,¹⁴⁹ borate bioactive glass,³⁰ and calcium phosphate^113,150 (Table 3). Several ceramics are intrinsically antimicrobial, such as calcium phosphate and bioactive glass;^151,152 however, for orthopedic implant infection applications, ceramics often incorporate antibiotics. For degradable ceramic materials (e.g., calcium phosphate, calcium sulfate, hydroxyapatite, etc.), the rate of antibiotic release is strongly dependent upon the rate of degradation. Degradation of calcium phosphate materials including hydroxyapatite and β-tricalcium phosphate is mediated by physical, chemical, and biological factors including crystallinity, porosity, pH, and ionic substitutions.¹⁵³ An advantage of these materials is that their physical properties can be tailored to obtain the desired degradation and subsequent release kinetics of incorporated antibiotics.^153,154 Recently an injectable formulation of bioactive glass has been developed to enable noninvasive application of the antibiotic carrier for treatment of orthopedic implant infections.³⁰ An advantage of ceramic orthopedic antimicrobial materials is that they often have crystalline structures that mimic the structure of bone and exhibit excellent osteoconductive properties.^151,155 Calcium phosphate, for example, releases calcium and phosphate ions that bind collagen and promote bone in-growth to the implant.¹⁵⁶ Despite their excellent osteoconductive properties and tunable drug release, many ceramic antimicrobial materials are not amenable for load-bearing orthopedic applications. For example, bioactive glass, hydroxyapatite, and calcium phosphate can be brittle.¹⁵¹

Table 3.

Summary of Advantages and Limitations of Ceramic Antimicrobial Biomaterials from the Past Decade Designed for Orthopedic Implant-Associated Infection Applications

ceramics
composition	tested in vivo?	antimicrobial activity	advantages	limitations	ref
calcium sulfate (dihydrate/hemihydrate)	yes	not evaluated	(1) promote osseointegration	(1) require antibiotics	141–143
				(2) limited duration of activity
biphasic nanohydroxyapatite/calcium sulfate	yes	bacterial clearance of 100% CFUs	(1) promote osseointegration	(1) require antibiotics	144
				(2) Limited duration of activity
β-tricalcium phosphate/calcium sulfate	no	5–40 days activity zone of inhibition (in vitro)	(1) promote osseointegration	(1) require antibiotics	145
				(2) limited duration of activity

3.4. Composites.

Antimicrobial orthopedic composite materials are composed of a combination of metal, polymer, and ceramic materials (Table 4). Ceramics including zirconium nitride,¹⁴⁹ hydroxyapatite,^146–148 and calcium phosphate¹⁵⁷ have been used as coatings on titanium, titanium alloy, and cobalt–chromium–molybdenum implants in implant-associated infection models. An advantage of composite materials is that they can incorporate the beneficial properties of both materials to create a dual-functioning system. For example, when osteoconductive ceramic materials such as hydroxyapatite are combined with intrinsically antimicrobial silver, the composite can simultaneously provide antimicrobial activity while promoting osseointegration of the implant.^{146–148,157}

An advantage of composite materials is that they can be designed to address drug delivery limitations of individual materials. Specifically, antibiotics are generally incorporated directly into cement materials (e.g., PMMA bone cement, calcium sulfate, borate bioglass) to impart antimicrobial activity.^{28–30,123–125} Nevertheless, antibiotic release kinetics from nondegradable cements (e.g., PMMA) are suboptimal as a small amount of antibiotic is only released from the surface of the PMMA (~10% of incorporated drug) and the rest remains entrapped permanently.¹⁵⁸ Antibiotic release from degradable cements (e.g., calcium sulfate and borate bioglass) is dependent upon the rate of degradation of the material.³⁰ To enable a more controlled release of antibiotics from cements, polymers such as cyclodextrin^29,123–125 and cationic liposomes²⁸ have been incorporated. Specifically, insoluble antibiotic-filled cyclodextrin microparticles have been incorporated into PMMA bone cement and have been shown to enable a more prolonged and consistent release of antibiotics at therapeutically relevant levels.^29,123–125 Cationic liposomes have been incorporated into calcium sulfate to locally increase antibiotic concentration at the implant infection site.²⁸ Nevertheless, depending upon their composition, composite materials may still have the limitations of individual materials (e.g., leach cytotoxic metallic ions, risk for formation of bacterial biofilms on nondegradable materials).

4. EVALUATING THE EFFICACY OF BIOMATERIALS TO PREVENT/TREAT IMPLANT-ASSOCIATED INFECTIONS IN VIVO

4.1. Considerations in Establishing Clinically Relevant Preclinical Implant-Associated Infection Models.

When establishing an orthopedic implant-associated infection in vivo, there are several factors that must be considered. For example, the inoculum used to generate the infection should be relevant to the animal species and region of the body, and this must be determined through a series of preliminary studies. An excessively large inoculum can cause sepsis and death while an insufficient inoculum will quickly be cleared.^34,163,164 The duration of time following the inoculation of the pathogen in which the infection is established in the model must also be considered since this is dependent upon the species used in the model and differs from clinical presentation.³¹ Additionally, the virulence of the bacterial strain used in the model and the culture conditions (i.e., log phase versus stationary phase growth) impact the concentration of the inoculum required to establish an infection.¹⁶⁵ Furthermore, the aim of the in vivo model (i.e., acute versus chronic infection) dictates the selection of the bacterial species used. Specifically, if the model is studying acute infection, bacteria with a high virulence will be used (e.g., S. aureus), whereas if the model is studying chronic infection low virulence bacteria will be used (e.g., coagulase-negative staphylococci).¹⁶⁶ It is also important to consider that bone remodeling kinetics are dependent upon the species used in the in vivo model and that these kinetics differ from those observed in humans.¹⁶⁷ As a result, the time course for acute and chronic implant-associated infections (i.e., duration of experimental study) varies across animal species and humans. To improve the clinical relevancy of the infection in vivo, it is also critical that the location and placement of the implanted material allow for weight-bearing locomotion.¹⁶⁸

Table 5 provides a summary of in vivo models from the past decade that have been used to establish and study implant-associated infections in a variety of species (e.g., zebrafish,¹⁶⁹ mice,^{32,33,115,163,170–175} rats,^{12,36,116,117,128,176–180} rabbits,^{34,134,181–183} pigs,^184,185 etc.). Generally, these models involve placing either a stainless steel, titanium, or titanium–aluminum (Ti6Al4V) alloy implant in the animal’s tibia (proximal tibia)¹⁷³ or femur (lateral femoral condyle).¹⁷⁹ Once implanted, a bacterial culture is inoculated into the intra-articular space near the implant.¹⁷³ Implant-associated infection models have been established using a range of bacteria including Staphylococcus aureus and methicillin-resistant S. aureus (MRSA),^{32–34,36,115,117,128,134,169–182,184,185} Staphylococcus epidermidis,¹¹⁶ Escherichia coli,^163,180 Pseudomonas aeruginosa,¹⁶³ Propionibacterium acnes,^116,183 and Streptococcus agalactiae.¹⁷² Bioluminescent strains of bacteria (e.g., S. aureus, E. coli) are frequently used to enable in vivo tracking and imaging of the extent and evolution of the infection over time.^{33,87,90–92,97,118,119,126,127,163,173–175,179,186–190}

Table 5.

Summary of Preclinical In Vivo Models from the Past Decade Used to Establish and Study Implant-Associated Infections

models to establish and study implant infections
animal model	implant/location	pathogen	study duration	key evaluation	summary	ref
fish – zebrafish	none	S. aureus SH1000-derived	3 weeks	(1) histology	bone fracture model in zebrafish	169
		2.5 × 109 CFU		(2) confocal imaging
mice – BALB/C	titanium in femur	S. aureus ATCC 29213	4 weeks	(1) CFU count	model to study implant infection biofilms	170
		2 × 103–5 × 106 CFU		(2) collect blood
				(3) X-rays
				(4) SEM biofilm
				(5) histology
mice – BALB/C	titanium in femur	S.aureus ATCC 29213	4 weeks	(1) CFU count	model for histology scores of implant infection	171
		1 × 103 CFU		(2) X-rays
				(3) histology
mice – BALB/C	stainless steel in tibia	S. aureus USA300	2 weeks	(1) CFU count	first animal model of S. agalactiae implant infection	172
		5 × 105 CFU		(2) micro-CT
		S. agalactiae COH1		(3) histology
		5 × 105 CFU		(4) TRAP quantify
mice – diabetic NOD/ShiLtJ CD1	stainless steel in femur	S. aureus ATCC 25923	4 weeks	(1) CFU count	diabetic model of implant infection	32
		1 × 103 CFU		(2) collect blood
				(3) micro-CT
				(4) SEM biofilm
				(5) histology
mice – C57BL/6 and BALB/C	titanium or poly(ether ether ketone) in femur	S. aureus JAR 06.01.31	1 week	(1) CFU count	influence of implant material on implant infection	115
		9 × 105 CFU		(2) histology
mice – C57BL/6	titanium in femur	P. aeruginosa Xen 41	3 weeks	(1) CFU count	Gram-negative implant infection model	163
		1 × 103–1 × 105 CFU		(2) X-rays
		E. coli Xen 14		(3) histology
		1 × 103–1 × 105 CFU		(4) In vivo BLI
				(5) PET imaging
				(6) Flow cytometry
mice – C57BL/6	Ti6Al4V in tibia	S. aureus Xen 36	2 and 6 weeks	(1) CFU count	clinically relevant load-bearing model of periprosthetic joint infection	173
		3 × 105 CFU		(2) collect blood
				(3) X-rays
				(3) SEM biofilm
				(4) gait analysis
mice – C57BL/6	titanium in tibia	S. aureus Xen 36	12 weeks, disrupt microbiota	(1) CFU count	role of the gut microbiota on periprosthetic joint infection	33
		1 × 102 CFU		(2) collect blood
				(3) X-rays
			5 days, postsurgery	(4) gait analysis
				(5) bacterial sequencing
mice – C57BL/6	stainless steel in humerus	S. aureus Xen 36	6 weeks	(1) CFU count	model of implant infection in humerus	174
		1 × 103 CFU		(2) X-ray
				(3) histology
				(4) in vivo BLI
mice – C57BL/6	stainless steel in tibia	S. aureus Xen 40	2 weeks	(1) CFU count	quantitative model of implant biofilm	175
				(2) SEM biofilm
				(3) in vivo BLI
rats – Sprague–Dawley	Ti6Al4V in tibia	S. aureus ATCC 25923	6 weeks	(1) CFU count	model of implant infection in metaphysis	176
		1 × 103–1 × 106 CFU		(2) collect blood
				(3) X-rays
				(4) histology
rats – Sprague–Dawley	stainless steel in tibia	S. aureus ATCC 49230	4 weeks	(1) CFU count	model of implant infection dependent on bacterial inoculation	177
		1 × 102–1 × 106 CFU		(2) collect blood
				(3) X-rays
				(4) histology
rats – Sprague–Dawley	metal alloy and high density poly ethylene in femur/tibia	S. aureus clinical MN8 and UAMS-1	6 weeks	(1) CFU count	model of knee implant infection	178
		1 × 102–1 × 104 CFU		(2) collect blood
				(3) X-rays
				(4) histology
rats – Sprague–Dawley	titanium in femur	S. aureus ATCC 49230	6 weeks	(1) CFU count	model of implant infection, hematogenous osteomyelitis complication	12
		1 × 104–1 × 109 CFU		(2) micro-CT
				(3) histology
rats – Sprague–Dawley	titanium and ultrahigh molecular weight polyethylene in tibia/femur	S. aureus Xen 29	4 weeks	(1) CFU count	longitudinal infection model with two implant components that enables monitoring of postsurgery recovery	179
		2 × 107 CFU		(2) X-rays
				(3) micro-CT
				(4) collect blood
				(5) histology
				(6) gait analysis
rats – Sprague–Dawley	3D printed Ti6Al4V + PMMA in tibia	S. aureus ORI16_C02N	4 weeks	(1) CFU count	model of knee implant infection	128
		2 × 104 CFU		(2) X-rays
				(3) collect blood
				(4) micro-CT
				(5) histology
rats – Sprague–Dawley	stainless steel in femur	S. aureus ATCC 29213	1 week	(1) collect blood	model of effect of ethylenediaminetetraacetic acid to decrease implant infection	180
		1 × 107 CFU		(2) histology
		E. coli ATCC 25922
		1 × 107 CFU
rats – Sprague–Dawley	none	S. aureus	2 weeks	(1) CFU count	68Ga-citrate for PET imaging of implant infections	36
		3 × 108 CFU		(2) peripheral quantitative computed tomography
				(3) histology
				(4) 68Ga-citrate-chloride PET/CT imaging
rats – Wistar	poly(ether ether ketone) in tibia	S. epidermidis Epi103.1	4 weeks	(1) CFU count	longitudinal micro-CT imaging of implant infection with aerobic and anaerobic pathogens	116
		1 × 106 CFU		(2) time lapse micro-CT
		P. acnes Type IA, IB
		1 × 106 CFU
rats – Wistar	poly(ether ether ketone) and titanium in tibia	S. aureus JAR 06.01.31	4 weeks	(1) CFU count	model of morphological bone changes near infected implant	117
		3 × 107 CFU		(2) in vivo micro-CT
				(3) histology
				(4) pull-out testing
rabbits – New Zealand White	3D printed stainless steel + PMMA in tibia	S. aureus ATCC 29213	1 week	(1) CFU counts	3D printed custom implant for animal models, improve recovery	34
		5 × 106 CFU		(2) collect blood
				(3) erythrocyte sedimentation rate
rabbits – New Zealand White	3D printed stainless steel + PMMA in tibia	S. aureus ATCC 29213	1 week	(1) gait analysis	3D printed custom implant for animal models–improve recovery	134
		5 × 106 CFU		(2) crystal violet biofilm stain on implant
rabbits – New Zealand White	stainless steel in femur	S. aureus ATCC 25923	3 weeks	(1) X-rays	model of implant infection following open fracture	181
		1 × 106 CFU		(2) micro-CT
				(3) SEM biofilm
				(4) histology
rabbits – New Zealand White	TiAl6V4 in tibia	S. aureus ATCC 49230	6 weeks	(1) CFU count	model of early implant infections in rabbits	182
		4 × 105 CFU		(2) X-rays
				(3) histology
rabbits – New Zealand White	stainless steel in tibia	P. acnes 3 × 107 CFU	4 weeks	(1) CFU count	model of anaerobic species implant infection	183
				(2) collect blood
				(3) histology
pigs – Danish Landrace	stainless steel in tibia	S. aureus 4F9 (porcine)	5 days	(1) collect blood	low-inoculum porcine model of implant infection	184
		1 × 102–1 × 104 CFU		(2) CT imaging
pigs – Yorkshire-Landrace cross	none	S. aureus UAMS-1	11 and 15 days	(1) CFU count	pig model of hematogenous infection	185
		1 × 104 CFU		(2) collect blood
				(3) CT
				(4) histology

Once the infection is established, a series of evaluations are carried out to analyze the extent of the implant-associated infection pathogenesis. Clinically, in periprosthetic joint infections (PJIs), for example, a variety of analyses have been employed.^191–193 For example, measurement of erythrocyte sedimentation rate, serum C-reactive protein, white blood cell count in synovial fluid, and radiographic imaging have all been utilized as metrics for implant-associated infections clinically,^191,194,195 and many of these analyses translate to in vivo models. Over the duration of a study, blood is collected to monitor the evolution of the infection through erythrocyte sedimentation rate, serum amyloid A (a factor that, in the mouse, is more sensitive than serum C-reactive protein),^33,173 white blood cell count, and presence of inflammatory cells.^{32,34,179,185} To measure and track the extent of the infection during and at the end of the study, in vivo bioluminescence imaging (BLI) of bacteria and colony forming unit (CFU) counts in excised bone and synovial tissue and on removed implants have been analyzed.^{12,32,33,36,115–117,128,134,163,170–179,182,183,185} Histology (e.g., hematoxylin and eosin, Gram stain, tartrate-resistant acid phosphatase (TRAP)) is performed on excised bone and synovial tissue to supplement CFU counts and to identify osteoclasts.^{12,32,36,115,117,128,163,169–172,174,176–179,182,183,185} Scanning electron microscopy (SEM) and crystal violet staining are used to further examine the architecture and presence of bacterial biofilm on the excised implant.^{32,170,173,175,181} Radioactive tracers (e.g., ¹⁸F-fluoro-deoxy-glucose, ⁶⁸Ga-citrate-chloride) are used in combination with positron emission tomography (PET) to enable visualization of inflammation at the implant infection site.^36,163 A combination of X-ray and microcomputed tomography (micro-CT) imaging and gait analysis studies are applied to identify possible osteolytic regions surrounding the implant and to determine if the infection has a deleterious impact on the ambulation of the animals.^{12,32,33,116,117,128,134,163,170–174,176–179,181,182,184,185}

Models highlighted in Table 5 have a variety of goals including those that strive to study implant-associated bacterial biofilms,^170,175 establish new methodologies for improving in vivo imaging of implant-associated infections,^36,116 evaluate hematogenous implant-associated infections,^12,185 and study a range of aerobic and anaerobic pathogens (e.g., S. agalactiae, P. acnes, E. coli, P. aeruginosa, etc.).^163,172,183

4.2. Considerations in Evaluating Activity of Antimicrobial Materials in Preclinical Models.

Prior to incorporating antimicrobial orthopedic biomaterials in implant-associated infection in vivo models, the properties of the materials can be first evaluated in in vivo models without tibial or femoral implants (e.g., materials can be placed into dorsal subcutaneous pouches).^{37,41,59,83,87,135,144,148,161} Table 6 provides a summary of preclinical in vivo models that have been used within the past decade to evaluate antimicrobial activity, drug release kinetics, and degradation of antimicrobial orthopedic materials.

Table 6.

Summary of Preclinical In Vivo Models from the Past Decade Used to Evaluate the Antimicrobial Activity and Properties of Implanted Biomaterials for Orthopedic Implant-Associated Infection Applications

models to evaluate antimicrobial activity/properties of implanted materials
animal model	implant/location	pathogen	antimicrobial agent	study duration	key evaluation	summary	ref
mice – BALB/C	magnesium in subcutaneous pouch	P. aeruginosa PAO1 CTX::lux	none	8 days	(1) CFU count	degradable magnesium implant to modulate host immune response	59
					(2) histology
					(3) in vivo BLI
					(4) SEM biofilm
mice – BALB/C	polyester-poly urethane in subcutaneous pouch	S. aureus Xen 29	levofloxacin (local, sustained)	4 weeks	(1) CFU count	antibiotic polymer coatings to prevent implant infections	87
		1 × 106 CFU			(2) activity of residual drug in implant
mice – BALB/C	polyethylene terephthalate in subcutaneous pouch	P. aeruginosa PsAer-9	none	1 week	(1) in vivo fluorescence	near-infrared fluorescence probes to image implant inflammation and infection	37
					(2) CFU count
mice – C57B1/6	PEG hydrogel in femur	S. aureus UAMS-1	none	1 and 5 weeks	(1) CFU count	lysostaphin hydrogels to prevent implant infections	120
		1.55 × 108 CFU			(2) micro-CT
		S. aureus USA 300			(3) histology
		3.43 × 108 CFU
mice – C57B1/6	none	MRSA USA 300 1 × 106 CFU	None	2 weeks	(1) micro-CT	poly(propylene sulfide) nanoparticles to prevent implant infections	198
					(2) histology
					(3) in vivo BLI
rats – Sprague–Dawley	chitosan sponge in subcutaneous pouch	none	vancomycin, ciprofloxacin, cefuroxime (local, sustained)	6 weeks	(1) collect blood	degradable antibiotic chitosan sponges to prevent implant infections	83
					(2) HPLC detect drug in plasma
					(3) in vivo degradation
					(4) tissue drug concentration
rats – Sprague–Dawley	micropatterned silicone in subcutaneous pouch	S. aureus ATCC 6538	none	<1 week	(1) CFU count	micropatterned implant to prevent implant infection	41
		1 × 104 CFU
rats – Sprague–Dawley	silicone in subcutaneous pouch	S. aureus ATCC 29213	none	1–2 weeks	(1) CFU count	nitric oxide releasing implant coatings to prevent implant infection	135
		1 × 108 CFU			(2) SEM biofilm
					(3) histology
rats – Sprague–Dawley	silver hydroxyapatite coated titanium in subcutaneous pouch	MRSA UOEH6	none	1 week	(1) quantify biofilm	composite implant coating to mitigate biofilm	148
		10 × 108 CFU
rats – Sprague–Dawley	Eudragit particle coated Ti6Al4V in subcutaneous pouch	none	vancomycin (local, sustained)	1 week	(1) collect blood	antibiotic coated implant to prevent infection	161
					(2) LC-MS detect drug in serum
rats – Wistar	hydroxyapatite/calcium sulfate in subcutaneous pouch	none	rifampin (local, sustained)	4 weeks	(1) collect blood	antibiotic bioactive ceramic for treating implant infections	144
					(2) HPLC detect drug in serum and bone
rats – Wistar	titanium in tibia	none	none	1–4 weeks	(1) histology	hydroxyapatite and metal coating for promoting implant osseointegration	147
					(2) X-rays
					(3) forced swimming test
rats – Wistar	stainless steel in femur	S. aureus ATCC 29213	vancomycin, moxifloxacin (systemic)	3 weeks	(1) collect blood	evaluation of tissue concentration of antibiotics for prevention of implant infection	196
		1 × 108 CFU			(2) HPLC detect drug in serum
					(3) minimum inhibitory concentration drugs
rabbits – New Zealand White	Ti6Al4V in femur	P. aeruginosa Pa11	none	5 weeks	(1) CFU count	method to detect presence of implant infection through urine with biomaterials	40
		1 × 106 CFU			(2) X-rays
					(3) histology
					(4) micro-CT
					(5) measure Al in urine
rabbits – New Zealand White	calcium sulfate in tibia	none	gentamicin, vancomycin, tobramycin (local, sustained)	4–12 weeks	(1) X-rays	calcium sulfate to promote osseointegration of implants	141
					(2) micro-CT
					(3) histology
rabbits – New Zealand White	calcium sulfate in tibia	none	vancomycin (local, sustained)	4 weeks	(1) collect blood	calcium sulfate with bone morphogenic protein-2 to promote osseointegration and prevent implant infection	143
					(2) HPLC detect drug in serum and bone
					(3) histology
rabbits – New Zealand White	calcium sulfate in tibia	none	gentamicin, vancomycin, tobramycin (local, sustained)	4 weeks	(1) X-rays	antibiotic calcium sulfate to prevent implant infection	142
					(2) collect blood
					(3) HPLC detect drug in serum and bone
rabbits – New Zealand White	calcium phosphate granules in tibia	MRSA clinical	vancomycin, tobramycin (local, sustained)	4 weeks infection	(1) CFU count	biodegradable antibiotic cement for treatment of MRSA implant infection	197
		5 × 107 CFU		6 weeks antibiotics	(2) X-rays
					(3) SEM
					(4) histology
					(5) HPLC detect drug

To detect the presence and concentration of eluted antibiotics from the implanted biomaterial, blood can be collected both throughout the experiment and terminally in conjunction with hard and soft tissues surrounding the material.^{83,142–144,161,196,197} The antibiotic can then be extracted from the serum and tissues, and high performance liquid chromatography (HPLC) and liquid chromatography–mass spectrometry (LC-MS) can be used to quantify the systemic and local concentration of the antibiotics.^{83,142–144,161,196,197} Following euthanasia, residual drug remaining entrapped in the implanted material can be extracted and evaluated for its concentration and antimicrobial activity in the zone of inhibition assays.⁸⁷ Alternatively, the antimicrobial activity of the material can be evaluated indirectly through the quantification of the extent of infection remaining throughout the study. Specifically, techniques such as in vivo BLI can be used in conjunction with bacteria that have been genetically modified to express the lux operon (e.g., Xen strains) such that the extent of the infection can be imaged in real time.^59,87,198 Additionally, the extent of infection remaining can be evaluated at the termination of the study through removal of the material and SEM imaging of the morphology of the bacterial biofilm and CFU counts of adherent bacteria.^{37,40,41,59,87,120,135,148,197} To evaluate the degradation of implanted materials in vivo, the material can be explanted with surrounding subcutaneous tissue at intermediate time points throughout the study and characterized for its morphology using SEM, molecular weight using gel permeation chromatography, glass transition temperature using differential scanning calorimetry, and residual weight.^83,199

4.3. Considerations in Evaluating Mechanical Strength of Antimicrobial Materials in In Vivo Models.

When orthopedic antimicrobial materials are used in load-bearing applications (e.g., antibiotic-laden PMMA bone cement implanted around arthroplasty components), it is critical that the system is mechanically robust and integrates with native bone tissue.²⁰⁰ While American Society for Testing and Materials (ASTM) and International Organization for Standardization (ISO) standards have been well-established for evaluating compressive strength, 3- and 4-point bending strength, impact and fatigue strength, and intrusion of PMMA bone cement,^201,202 there is currently no standardized evaluation protocol for interfacial shear strength of the PMMA bone cement/bone interface. Since interfacial shear strength is a metric reflective of the ability of PMMA bone cement to integrate with trabecular bone and poor interlock can result in failure of the material, it is a crucial factor to consider.²⁰³ Several groups have developed in vivo/ex vivo push-out and pull-out tests to quantify the interfacial shear strength of PMMA bone cement and other composites, nevertheless, there are inherent challenges in the standardization of these testing methods.^204–210 In both preclinical and clinical settings, age, sex, size and type of defect, the composition, method of preparing, and amount of PMMA bone cement, and the setting time, temperature, humidity, and moisture/fluid composition prior to testing are all factors which must be considered to standardize interfacial shear strength testing methodologies.²¹¹ In general, it has been historically challenging to compare results of the mechanical properties of PMMA bone cement across studies due to a lack of standardization and detail of storage and preparation conditions of the samples.²¹¹

Pull-out testing is frequently used in implant-associated infection in vivo models where the metallic or polymeric implant integrates with native bone for a period, dependent upon the animal species and age and the rate of bone remodeling.²¹² Upon completion of the study, tibial/femoral tissue is excised and fixed in PMMA bone cement and uniaxial tensile loading is used to determine the force required to remove the osseointegrated implant from the bone.^{205,213–215} Push-out testing serves an analogous function to pull-out testing. Despite the similar goals of pull- and push-out studies it is challenging to compare interfacial shear strength values across studies. Specifically, a wide range of loading rates have been used in pull- and push-out testing, depending upon the species, from 0.0083 to 0.5 mm/s, which can influence the resultant interfacial pull-out strength.^{206,207,213–216} Calculation of interfacial shear strength from raw load–displacement plots is also highly variable. For example, some studies define interfacial shear strength as the peak force on the load–displacement plots when the implant is fully removed from the bone or the bone is fractured, while others interpret interfacial shear strength as the force resulting when the implant is only slightly displaced in the bone.^206,213,215 Interfacial strength can also vary based upon several factors including the geometry of the bone and the implant placement within the bone (e.g., contact area of trabecular versus cortical bone).^207,209 Prior studies have demonstrated the key role that bone density plays in influencing interfacial shear strength; thus, if a greater percentage of the implanted material is in contact with cortical bone relative to trabecular bone, the interfacial strength will be increased.²⁰⁹ Interfacial shear strength also depends upon the type of material used in the implant, the duration that it is implanted in the animal (in vivo versus solely ex vivo), and the species and relative mobility of the animal (i.e., relative mechanical loading that the implant will experience).

Nevertheless, due to the heterogeneous nature of clinical cases, for example, patients that present with comorbidities (e.g., diabetes, osteoarthritis, etc.) that interfere with bone matrix deposition and healing, it is challenging to compare outcomes from preclinical interfacial shear strength studies to what would occur clinically and bone remodeling kinetics differ across animal species.^167,217,218 As a result, the osseointegration of the implanted biomaterial can be evaluated using histology in conjunction with biomechanical testing.^32,179,185 Infiltration of inflammatory cells (e.g., macrophages, neutrophils, monocytes, foreign body giant cells, etc.) in tissue surrounding the implanted material is indicative of osteolysis as inflammatory cells release cytokines (e.g., tumor necrosis factor-alpha (TNF-α), interleukins (IL-1, IL-6), etc.) that recruit osteoclasts and promote osteolysis and loss of fixation of the implant.^{32,179,185,219} Additionally, radiographs and micro-CT can be used to image osteolysis.^179,185,220

5. INCORPORATING ANTIMICROBIAL MATERIALS INTO IMPLANT-ASSOCIATED INFECTION PRECLINICAL MODELS

5.1. Preclinical Models to Evaluate Prevention of Implant-Associated Infections.

In vivo models have been established to enable evaluation of the ability of antimicrobial biomaterials to prevent orthopedic implant-associated infections in a variety of species including mice,^{38,90–93,186–188,221–224} rats,^{94–99,112–114,122,129,189,225–234} rabbits,^{23,26,100–102,150,157,190,235–240} sheep,^{103,104,241–244} and goats^130,245 (Table 7). In these models, a metallic component is placed in either the tibia or femur and the prophylactic antimicrobial material is placed at the time of bacterial inoculation to enable evaluation of the ability of the material to prevent the development of implant-associated infection. Throughout the duration of the study, the extent of infection can be monitored through the use of in vivo BLI in conjunction with luminescent bacteria (e.g., Xen strains).^{90–93,100,186–189} The ability of the antimicrobial implant to prevent the infection can be determined via quantification of the bacteria remaining following euthanasia using CFU counts on the surface of the implant, and the morphology of bacterial biofilms can be analyzed using SE M.^{23,26,38,90–100,102–104,114,129,130,150,157,188,190,216,221,222,227–234,236–244} Additionally, the concentration of antimicrobial agent (e.g., antibiotics or metallic ions) used to prevent infection can be evaluated through collection of blood and soft tissues surrounding the implanted antimicrobial material. Specifically, the antibiotic and metallic ions (e.g., silver, titanium) can be extracted from the serum and tissues and quantified using HPLC (for antibiotics) and inductively coupled plasma mass spectrometry (for metallic ions).^241,242,245

Table 7.

Summary of Preclinical In Vivo Models from the Past Decade Used to Evaluate the Prevention of Orthopedic Implant-Associated Infections with Antimicrobial Biomaterials

models to evaluate prevention of implant-associated infections
animal model	implant/location	pathogen	antimicrobial agent	study duration	key evaluation	summary	reference
mice – BALB/C	titanium in femur	S. aureus Xen 29	none	4 weeks	(1) collect blood	hydroxyapatite coating with ionic silver to prevent implant infections	186, 187
		1 × 108 CFU			(2) in vivo BLI
					(3) histology
mice – BALB/C	stainless steel in femur	MRSA ATCC 43300	linezolid and MR-5 bacteriophage (local, sustained)	3 weeks	(1) CFU count	antibiotic and bacteriophage implant coatings for infection prevention	221
		1 × 105–1 × 108 CFU			(2) gait analysis
					(3) histology
					(4) X-rays
mice – BALB/C	titanium in femur	S. aureus ATCC 25923	none	2 weeks	(1) SEM biofilm	nanocomposite tantalum oxynitride-silver coating to prevent implant infection	222
		5 × 105 CFU
mice – BALB/C	silicon nitride in tibia	MRSA USA 300	none	2 weeks	(1) SEM biofilm	role of surface topography of silicon nitride to prevent implant infection	38
		1 × 105 CFU			(2) histology
mice – C57Bl/6	stainless steel tibia	S. aureus ATCC 43300	none	2 weeks	(1) micro-CT	aspirin to facilitate healing of implant infections	223
		1 × 106 CFU			(2) histology
mice – C57Bl/6	stainless steel in L4 spine and femur	S. aureus Xen 36	vancomycin, tigecycline (local, sustained)	4 and 6 weeks	(1) CFU count	PEG-poly(propylene sulfide) coating to prevent implant infections	90–92
		1 × 103 CFU			(2) in vivo BLI
					(3) X-rays
mice – C57Bl/6	stainless steel or titanium in femur	S. aureus Xen 36	vancomycin, daptomycin, tigecycline (systemic)	1 week	(1) CFU count	model to compare efficacy of prophylactic antibiotics for implant infections	188
		1 × 104 CFU			(2) X-rays
					(3) in vivo BLI
					(4) SEM biofilm
mice – Lys-EGFP	stainless steel in femur	S. aureus ALC-2906	rifampin, minocycline (local, sustained)	2 weeks	(1) CFU counts	antibiotic polyesteramide coating to prevent implant infections	93
		5 × 102 CFU			(2) in vivo BLI
					(3) histology
					(4) SEM biofilm
mice – Kunming	titanium in femur	MRSA USA 300	none	4 weeks	(1) CFU count	graphdiyne-titanium oxide nanofiber composite implant prevent implant infections	224
		1 × 107 CFU			(2) histology
					(3) SEM
rats – Sprague–Dawley	stainless steel in tibia	S. aureus ATCC 49230	gentamicin (local, sustained)	6 weeks	(1) CFU count	gentamicin in poly(d-l-lactide) coating to prevent implant infections	94–96
		1 × 102– 1 × 103 CFU			(2) collect blood
					(3) X-rays
					(4) histology
rats – Sprague–Dawley	titanium in femur	S. aureus ATCC 25923	none	6 weeks	(1) micro-CT	dimethylaminododecyl methacrylate and hydroxyapatite coated implant to prevent infection and promote osteogenesis	225
		1 × 108 CFU			(2) histology
					(3) biocompatibility
rats – Sprague–Dawley	magnesium and titanium in femur	MRSA ATCC 43300	none	8 weeks	(1) CFU count	magnesium to prevent MRSA implant infections	226
		1 × 106 CFU			(2) X-rays
					(3) micro-CT
					(4) SEM of biofilms
rats – Sprague–Dawley	titanium in femur	S. aureus ATCC 29213	none	4 weeks	(1) CFU counts	layer-by-layer coating (silver nanoparticles + poly-l-glutamic acid and polyallylamine to prevent implant infection	227
		1 × 106 CFU			(2) micro-CT
					(3) histology
rats – Sprague–Dawley	titanium in femur	S. aureus ATCC 49230	gentamicin palmitate (local, sustained)	6 weeks	(1) CFU count	antibiotic coated implant for infection prevention	228
		1 × 102 CFU			(2) X-rays
					(3) collect blood
					(4) histology
rats – Sprague–Dawley	TiAl6V4 in tibia	S. aureus ATCC 49230	gentamicin (local, sustained)	4 weeks	(1) CFU count	antibiotic coated titanium oxide implant for infection prevention and osseointegration	229
		1 × 104 CFU			(2) X-rays
					(3) histology
rats – Sprague–Dawley	titanium in femur	S. aureus ATCC 25923	none	1–12 weeks	(1) CFU count	poly(glycidyl methacrylate) and quaternized poly (ethylene imine) functionalized titanium implants with alendronate for infection prevention and osseointegration	122
		1 × 109 CFU			(2) micro-CT
					(3) histology
					(4) pull-out test
rats – Sprague–Dawley	titanium in tibia	S. aureus ATCC 25923	none	4 weeks	(1) CFU count	poly-l-lysine functionalized implants for infection prevention and osseointegration	230
		1 × 104 CFU			(2) micro-CT
					(3) histology
rats – Sprague–Dawley	titanium in femur	S. aureus ATCC 25923	gentamicin (local, sustained)	6 weeks	(1) CFU count	antibiotic nanotube implants for infection prevention and osseointegration	231
		1 × 105 CFU			(2) X-rays
					(3) histology
rats – Sprague–Dawley	poly(ether ether ketone) in tibia	S. aureus Xen 29	gentamicin (local, sustained)	1–8 weeks	(1) CFU count	nanolayered antibiotic and bone morphogenic protein-2 implant coating for infection prevention and osseointegration	97
		5 × 105 CFU			(2) micro-CT
					(3) histology
					(4) pull-out testing
rats – Sprague–Dawley	titanium in femur	S. aureus ATCC 29213	none	1 week	(1) CFU count	hydrophobic triethoxysilane implant coating for infection prevention	232
		1 × 106 CFU			(2) histology
rats – Sprague–Dawley	PMMA in femur	MRSA clinical	teicoplanin (local, sustained)	3 weeks	(1) CFU count	antibiotic and calcium sulfate PMMA to prevent implant infections	129
		1 × 108 CFU			(2) X-rays
					(3) histology
rats – Sprague–Dawley	stainless steel in femur	S. aureus ATCC 43300	none	6 weeks	(1) micro-CT	sodium butyrate modified implant to mitigate infection and promote osteogenesis	112
		1 × 104 CFU			(2) histology
rats – Sprague–Dawley	copper-poly(ether ether ketone) in femur	MRSA ATCC 43300	none	4 weeks	(1) CFU count	composite implant to prevent MRSA infection	114
		1 × 106 CFU			(2) micro-CT
					(3) histology
					(4) biocompatibility
rats – Sprague–Dawley	poly(ether ether ketone) in femur	S. aureus ATCC 6538	gentamicin (local, sustained)	4–6 weeks	(1) X-rays	antibiotic composite implant to promote osseointegration and prevent infection	113
		1 × 104 CFU			(2) micro-CT
					(3) histology
rats – Sprague–Dawley	gallium–strontium magnesium alloy in femur	S. aureus ATCC 43300	none	5 days	(1) CFU count	metal alloys for intrinsic prevention of implant infections	233
		1 × 105 CFU			(2) histology
					(3) SEM
rats – Sprague–Dawley	Pro Osteon 500R + PEG + poly (capro-lactone) + PLGA in tibia	S. aureus ATCC 49230	rifampin, vancomycin (local, sustained)	10 weeks	(1) CFU count	antibiotic putty to prevent implant infection	98
		1 × 108 CFU			(2) X-rays
					(3) micro-CT
					(4) histology
rats – Sprague–Dawley	3D printed titanium in tibia	S. aureus ATCC 49230	vancomycin (local, sustained)	4 weeks	(1) CFU count	chitosan silver particle antibiotic implant coatings to prevent infection	99
		1 × 106 CFU			(2) micro-CT
					(3) collect blood
					(4) histology
rats – Wistar	titanium in femur	S. aureus Xen 40	none	6 weeks	(1) CFU count	vitamin E phosphate coating to improve implant integration	189
		3 × 104 CFU			(2) collect blood
					(3) micro-CT
					(4) histology
					(5) in vivo BLI
rats – Wistar	Ti6Al4V in tibia	S. aureus ATCC 25923	none	6 weeks	(1) CFU count	hydroxyapatite and silver coated implant infection model	234
		1 × 102–1 × 103 CFU			(2) X-rays
					(3) histology
rabbits – Dutch Belted	titanium in femur	S. aureus SAP231	linezolid, rifampin (local, sustained)	1 week	(1) CFU count	PLGA and poly(caprolactone) nanofiber coated implant infection model	100
		1 × 104 CFU			(2) in vivo BLI
rabbits – New Zealand White	stainless steel in tibia	MRSA ATCC 43300	none	6 weeks	(1) micro-CT	titanium with nanothick calcium oxide MRSA implant infection model	246
		1 × 104 CFU			(2) histology
rabbits – New Zealand White	titanium in femur	MRSA ATCC 43300	none	6 weeks	(1) CFU count	silver ion calcium phosphate ceramic nanopowder implant coating infection model	150
		5 × 102 CFU			(2) X-rays
					(3) histology
rabbits – New Zealand White	Porous tantalum in radius	S. aureus ATCC 49230	tobramycin (local, sustained)	2 weeks	(1) X-rays	antibiotic PLGA microspheres in porous implant to prevent infections	101
		2 × 106 CFU			(2) histology
rabbits – New Zealand White	titanium in tibia	S. aureus ATCC 10832	tobramycin (local, sustained)	4 weeks	(1) CFU count	antibiotic-periapatite coated implants to prevent infection and promote osseointegration	236
		1 × 103–1 × 105 CFU			(2) collect blood
					(3) histology
					(4) erythrocyte sedimentation rate
rabbits – New Zealand White	stainless steel in humerus	S. aureus JAR 060131	gentamicin (local, sustained)	1 week	(1) CFU count	injectable antibiotic thermoresponsive hyaluronic acid implant coating to prevent infections	23
		2 × 106 CFU			(2) X-rays
					(3) collect blood
					(4) histology
rabbits – New Zealand White	titanium and stainless steel in humerus	S. aureus JAR 060131	none	4 weeks	(1) CFU count	impact of implant topography on infection prevention	237
		2 × 103–2 × 105 CFU			(2) X-rays
rabbits – New Zealand White	titanium in femur	MRSA clinical isolate	vancomycin (local, sustained)	12 weeks	(1) CFU count	antibiotic hydrogel implant coating for infection prevention	102
		5 × 104–5 × 106 CFU			(2) collect blood
					(3) histology
rabbits – New Zealand White	titanium in femur	S. aureus ATCC 6538P	none	1 week	(1) CFU count	antimicrobial fusion peptide implant coating for infection prevention and osseointegration	238
		1 × 108 CFU			(2) micro-CT
					(3) histology
rabbits – New Zealand White	titanium in tibia	S. aureus ATCC 25923	none	4 weeks	(1) CFU count	molybdenum disulfide/polydopamine – RGD implant coating for infection prevention	239
		2 × 103 CFU			(2) micro-CT
					(3) histology
rabbits – New Zealand White	Ti6Al4V in femur	MRSA ATCC 43300	none	10 weeks	(1) CFU count	silver and ceramic implant coating for infection prevention	157
		5 × 104 CFU			(2) collect blood
					(3) histology
rabbits – New Zealand White	magnetic iron oxide particles with carbon nanotubes in tibia	MRSA CCTCC 16465	gentamicin (local, sustained)	2 weeks	(1) CFU count	bacterial capturing implant infections with magnetic microwave activated composites	26
		1 × 106 CFU			(2) collect blood
					(3) histology
					(4) MRI
rabbits – New Zealand White	calcium phosphate in tibia	S. aureus Xen 29	tobramycin (local, sustained)	4 weeks	(1) CFU count	antibiotic calcium phosphate for prevention of implant infections	190
		1 × 107 CFU			(2) collect blood
					(3) X-rays
					(4) histology
rabbits – New Zealand White	chitosan and calcium phosphate in tibia	S. aureus	moxifloxacin (local, sustained)	4 weeks	(1) CFU count	composite antibiotic material for prevention of implant infection	240
		1 × 106 CFU			(2) histology
sheep	stainless steel in tibia	S. aureus ATCC 6538	cefazolin (local, sustained)	7 days	(1) CFU count	antibiotic filled steel implants for infection prevention in sheep	241, 242
		6.6 × 106 CFU			(2) collect blood
					(3) histology
					(4) antibiotic distribution
sheep – Rambouillet	titanium in femur	MRSA clinical	trialkyl norspermidine-biaryl (local, sustained)	4 and 24 weeks	(1) CFU count	antimicrobial compound active release implant coating for infection prevention	243
		2 × 108 CFU			(2) X-rays
					(3) SEM
					(4) histology
sheep – Columbia Cross	hydroxyapatite calcium carbonate and PLGA in femur	S. aureus ATCC 49230	tobramycin (local, sustained)	12 weeks	(1) CFU count	antibiotic bone void filler to prevent implant infection	103
		5 × 105 CFU			(2) collect blood
					(3) micro-CT
					(4) histology
sheep – Dorset Cross	titanium plate in tibia	S. aureus ATCC 25923	vancomycin (local, sustained)	12 weeks	(1) CFU count	antibiotic modified implant surface to prevent infection	244
		2 × 106 CFU			(2) X-rays
					(3) SEM biofilm
					(4) histology
					(5) gait analysis
sheep – English Mule	PLGA and PEG in femur	S. aureus F2789	gentamicin, clindamycin (local, sustained)	2 and 13 weeks	(1) CFU count	injectable and biodegradable antibiotic gel to prevent implant infection	104
		2 × 106 CFU			(2) collect blood
					(3) micro-CT
					(4) histology
goat	stainless steel in tibia	S. aureus ATCC 25923	none	5 weeks	(1) X-rays	titanium oxide and siloxane implant coating to prevent infection	245
		2 × 104 CFU			(2) micro-CT
					(3) histology
					(4) silver and titanium in organs
goat – Spanish	PMMA and calcium sulfate in tibia	S. aureus ATCC 29213	tobramycin (local, sustained)	3 weeks	(1) CFU count	evaluation of ability of commercial antibiotic composites for prevention of implant infection	130
		2 × 106 CFU

Models for the prevention of implant-associated infection included in Table 7 serve a variety of functions including those that evaluate systemic delivery of drugs,^188,223 evaluate the effect of the surface properties and topography of the implant,^38,232,237 and evaluate materials that provide both antimicrobial activity and promote osseointegration^{97,112,113,122,186,187,190,225,229–231,234,236,238} and those that evaluate intrinsic antimicrobial materials (e.g., without antibiotics).^{112,122,150,157,227,230,233,234,238,245,246}

5.2. Preclinical Models to Evaluate Treatment of Implant-Associated Infections.

Orthopedic implant-associated infection in vivo models that incorporate antimicrobial biomaterials have been developed to treat established infections following an initial debridement in a variety of species including mice,^{118,119,126,127,247–249} rats,^{25,27,85,86,105,131,250–255} rabbits,^{24,28,30,35,106,132,133,235,256–258} and dogs,²⁵⁹ for example (Table 8). Models that evaluate treatment of implant infections generally involve an initial implantation of a tibial or femoral implant and bacterial inoculation. Following inoculation, the implant-associated infection is allowed to develop for a period (dependent upon the species). Once the infection has been established (typically ≥ 7 days), the surgical site is debrided and an antimicrobial biomaterial is implanted. The efficacy of the antimicrobial material to treat and clear established implant-associated infections is evaluated through terminal CFU counts on implants and surrounding tissue, morphology of bacterial biofilms on implanted materials (SEM), and histology of tissue surrounding the infected implant.

Table 8.

Summary of Preclinical In Vivo Models from the Past Decade Used to Evaluate the Treatment of Orthopedic Implant-Associated Infections with Antimicrobial Agents and Biomaterials

models to evaluate treatment of implant-associated infections
animal model	implant/location	pathogen	antimicrobial agent	study duration	key evaluation	summary	ref
mice – BALB/C	titanium in femur	S. aureusATCC	gentamicin, vancomycin (local, sustained)	6 weeks	(1) CFU counts	antibiotic calcium sulfate/hydroxyapatite spacer for infection treatment	247
		292131.35 × 105 CFU			(2) collect blood
					(3) X-rays
mice – BALB/C	poly(ether ether ketone) and titanium in femur	S. aureus Xen 36	vancomycin, rifampin (local, sustained)	1 week infection	(1) CFU count	3D printed calcium phosphate and PMMA antibiotic spacers for infection treatment	118
		8 × 104 CFU		3 weeks antibiotics	(2) in vivo BLI
					(3) X-rays
					(4) micro-CT
					(5) SEM biofilm
mice – BALB/C	poly(ether ether ketone) and titanium in femur	S. aureus Xen 36	vancomycin (local, sustained)	1 week infection	(1) CFU count	antibiotic PMMA spacers for infection treatment	119
		8 × 104 CFU		3 weeks antibiotics	(2) in vivo BLI
					(3) micro-CT
mice – BALB/C	titanium in femur	S. aureus Xen 36	gentamicin (local, sustained)	1 week infection	(1) CFU count	antibiotic PMMA spacers for infection treatment	126
		2.5 × 106 CFU		2 weeks antibiotics	(2) X-rays
					(3) in vivo BLI
mice – BALB/C	stainless steel in tibia	S. aureus UAMS-1	cefazoline, gentamicin, vancomycin, rifampin (systemic)	1 week infection	(1) CFU count	evaluation of combination antibiotics for treatment of implant biofilms	248
		2.5 × 105 CFU		2 weeks antibiotics	(2) SEM biofilm
					(3) histology
mice – C57Bl/6	titanium in femur	MRSA USA 300	linezolid, rifampin, vancomycin, daptomycin (systemic)	2 weeks infection	(1) CFU count	evaluation of combination antibiotics for treatment of implant infection	249
		1 × 103 CFU		6 weeks antibiotics	(2) in vivo BLI
					(3) X-rays
					(4) micro-CT
mice – C57Bl/6	titanium in tibia	S. aureus Xen 36	vancomycin (local, sustained)	2 weeks infection	(1) CFU count	antibiotic PMMA spacers for infection treatment, need for intravenous antibiotics	127
		3 × 105 CFU		4 weeks antibiotics	(2) collect blood
					(3) X-rays
rats – Sprague–Dawley	Kirschner wire in tibia	MRSA 40496/08	dalbavancin, vancomycin (systemic)	4 weeks	(1) CFU count	antibiotics for implant MRSA infection treatment	250
		1–5 × 106 CFU			(2) X-rays
rats – Sprague–Dawley	intramedullary implant in tibia	S. aureus EDCC 5055	none	6 weeks	(1) CFU count	use of CpG oligodeoxynucleotide to treat implant infections	251
		1 × 103 CFU			(2) histology
rats – Sprague–Dawley	Kirschner wire in tibia	MRSA N315	vancomycin (local, sustained)	2 weeks infection	(1) CFU count	antibiotic silicone cement for treatment of implant infection	260
		1 × 108 CFU		4 weeks antibiotics	(2) X-rays
					(3) histology
rats – Sprague–Dawley	PMMA in tibia	MRSA ATCC 25923	fusidic acid (local, sustained), teicoplanin (systemic)	3 weeks infection	(1) CFU count	antibiotic cement with systemic antibiotics for treatment of implant infection	131
		5 × 105 CFU		2 weeks antibiotics	(2) histology
rats – Sprague–Dawley	alginate-hyaluronic acid in femur	S. aureus KCTC1621	vancomycin (local, sustained)	2 weeks infection	(1) CFU count	in situ gelling hydrogel for treatment of implant infections	85
		1 × 104 CFU		3 and 6 weeks antibiotics	(2) micro-CT
					(3) collect blood
					(4) X-rays
rats – Sprague–Dawley	injectable alginate hydrogel in femur	S. aureus ATCC 6538	fosfomycin (local, sustained)	1 week infection	(1) CFU count	injectable antibiotic hydrogel and bacteriophage for treatment of implant infections	105
		5 × 104 CFU		1 week antibiotics	(2) SEM
					(3) histology
rats – Sprague–Dawley	silver nanoparticles in gelatin sponge in tibia	MRSA ATCC 43300	teicoplanin (local sustained)	3 weeks infection	(1) X-ray	silver nanoparticles for treatment of implant infections	252
		1 × 107 CFU		3 weeks antibiotics	(2) collect blood
rats – Wistar	stainless steel in femur	S. aureus ATCC 29213	moxifloxacin, flucloxacillin, rifampin, vancomycin (systemic)	1 week infection	(1) CFU count	evaluation of combination antibiotics for treatment of implant infection	253, 254
		1 × 108 CFU		2 weeks antibiotics
rats – Wistar	collagen sponge and PMMA beads in tibia	S. aureus ATCC 292113	gentamicin, cefazolin (local, sustained)	3 weeks infection	(1) CFU count	comparison of antibiotic carrier for treatment of implant infections	86
		5 × 106 CFU		2 or 4 weeks antibiotics
rats – Wistar	poly(d-l-lactide-co-glycolide-co-ε-capro-lactone) in tibia	S. aureus	ciprofloxacin (local, sustained)	4 weeks infection	(1) CFU count	antibiotic poly(d-l-lactide-co-glycolide-co-ε-caprolactone) implant coating to treat infections	255
		1 × 108 CFU		4 weeks antibiotics	(2) collect blood
rats – Wistar	magnetite encapsulated gelatin particles in tail vein	S. aureus	gentamicin (local, sustained)	2 weeks infection	(1) X-rays	magnetite antibiotic particles for treatment of implant infection	25
		1 × 109 CFU		2 weeks antibiotic	(2) collect blood
					(3) histology
rats – Long-Evans	titanium in humerus	MRSA NR0	vancomycin (systemic)	6 day infection	(1) CFU count	electrical stimulation with antibiotics to treat implant infection	27
		1 × 105 CFU		5 weeks antibiotics	(2) X-rays
					(3) histology
rabbits – New Zealand White	silicone elastomer in tibia	MRSA S271 ST20121238	daptomycin, rifampin, ceftaroline (systemic)	1 week infection	(1) CFU count	evaluation of combination antibiotics for treatment of MRSA implant infection	256, 257
		5 × 107 CFU		10 days antibiotics
rabbits – New Zealand White	stainless steel in tibia	MRSA EDCC 5443 EDCC 5398	vancomycin (local, sustained)	4 weeks infection	(1) CFU count	antibiotic PMMA spacer for two-stage implant infection revision	132
		1 × 105–1 × 107 CFU		4 weeks antibiotics	(2) X-rays
					(3) histology
rabbits – New Zealand White	Ti6Al4 V in femur	S. aureus ATCC 29213	gentamicin (local, sustained)	1 week infection	(1) CFU count	antibiotic PMMA chain for treatment of implant infection (copper and titanium oxide coated)	133
		1 × 105 CFU		2 weeks antibiotics	(2) collect blood
rabbits – New Zealand White	borate bioactive glass in tibia	MRSA ATCC 43300	vancomycin (local, sustained)	8 weeks	(1) CFU count	injectable borate bioactive glass for treating implant infection	30
		1 × 108 CFU			(2) collect blood
					(3) X-rays
					(4) histology
rabbits – New Zealand White	calcium sulfate in tibia	S. aureus ATCC 29213	gentamicin (local, sustained)	2 weeks infection	(1) CFU count	antibiotic calcium sulfate to treat implant infection	28
		2 × 106 CFU		2 weeks antibiotics	(2) X-rays
					(3) collect blood
rabbits – New Zealand White	PLGA in femur	S. aureus ATCC 65389	vancomycin (local, sustained)	2 weeks infection	(1) CFU count	monitor treatment of implant infections with 18F-FDG-PET imaging	35
		1 × 105 CFU		2 weeks antibiotics	(2) 18F-FDG-PET imaging
rabbits – New Zealand White	Hydroxyapatite–polyamino acid in tibia	S. aureus	vancomycin (local, sustained)	2 weeks infection	(1) CFU counts	antibiotic hydroxyapatite/polyamino acid for treatment of implant infection and osteogenesis	235
		1 × 108 CFU		6 weeks antibiotics	(2) X-ray
					(3) collect blood
					(4) histology
rabbits – New Zealand White	stainless steel with silver particles in femur	S. aureus ATCC 29213	none	3 weeks infection	(1) CFU count	silver nanoparticle implant coating to treat infections	258
		3 × 106 CFU		3 weeks treatment	(2) X-rays
					(3) histology
					(4) SEM
rabbits – New Zealand White	Hydroxyapatite–polyamino acid PLGA in tibia	S. aureus ATCC 25923	rifapentine (local, sustained)	4 weeks infection	(1) CFU count	antibiotic composite material for treatment of implant infection	106
		3 × 108 CFU		4 weeks antibiotics	(2) collect blood
					(3) X-rays
					(4) histology
					(5) gait analysis
rabbits – New Zealand White	chitosan thermosensitive hydrogel in tibia	S. aureus Kanin	vancomycin (local, sustained)	4 weeks infection	(1) collect blood	injectable thermosensitive hydrogel with quaternary ammonium chitosan for treatment of implant infections	24
		1 × 108 CFU		4–8 weeks antibiotics	(2) micro-CT
					(3) histology
dogs – Beagle	stainless steel in femur	S. aureus ATCC 29213	ciprofloxacin (local, sustained)	4 weeks infection	(1) CFU count	amylose starch antibiotic coated implants to treat infection	259
		3 × 108 CFU		6 weeks antibiotics	(2) X-rays
					(3) histology

Models for the treatment of orthopedic implant-associated infection included in Table 8 serve a variety of functions including those that evaluate the efficacy of antibiotic dosing regimens,^{27,248–250,253,254} antibiotic-alternative agents (e.g., silver),^251,252,258 cement spacers,^{28,86,118,119,126,127,131–133,235,247} and injectable implants^24,30,85,105 and use novel imaging modalities to monitor the infection in real time.³⁵

6. RECENT ADVANCES IN ORTHOPEDIC ANTIMICROBIAL MATERIALS AND PRECLINICAL MODELS TO IMPROVE TREATMENT AND EVALUATION OF IMPLANT-ASSOCIATED INFECTIONS IN VIVO

6.1. Stimuli-Responsive Antimicrobial Materials.

Over the past decade, there have been numerous advances in the development of stimuli-responsive antimicrobial orthopedic materials that respond to temperature,^23,24 magnetic fields,^25,26 microwaves,²⁶ and electrical fields,²⁷ for example. Polymer hydrogels of hyaluronic acid and chitosan have been chemically modified to induce thermoresponsive properties.^23,24 Poly(N-isopropylacrylamide) has been engrafted onto hyaluronic acid to develop an injectable hydrogel that shifts from a sol to gel state when it reaches body temperature due to the conformational change of the poly(N-isopropylacrylamide) chains as they exceed their lower critical solution temperature.^23,261 Glycerol phosphate disodium salt has been used to induce thermoresponsive properties to chitosan.^24,262 Thermoresponsive polymers have been combined with antibiotics, thereby creating a biodegradable and biocompatible delivery system that can be applied minimally invasively to treat or prevent orthopedic implant infections. The thermoresponsive systems have been shown to effectively promote bone regeneration and clear infection in preclinical models.^23,24

Alternatively, several orthopedic antimicrobial materials have been developed that are responsive to magnetic fields and microwaves. Gelatin coated magnetite nanoparticles have been used in conjunction with an external neodymium magnet to guide the location of the injected particles to the target site (implant infection).²⁵ Additionally, materials that combine several stimuli (e.g., magnetic field, microwaves, temperature) have been engineered to target deep tissue orthopedic implant-associated infections.²⁶ Magnetic iron oxide (Fe₃O₄) nanoparticles have been combined with carbon nanotubes, antibiotics, and 1-tetradecanol for a “bacterial-capturing” stimuli-responsive system.²⁶ Specifically, iron oxide nanoparticles and carbon nanotubes generate heat when they are exposed to high microwaveocaloric therapy and 1-tetradecanol is thermoresponsive and allows for a controlled release of the encapsulated antibiotic at the elevated temperatures.²⁶

Additionally, an external treatment to titanium implants has recently been explored where a constant cathodic voltage is applied 7 days following implantation and establishment of infection.²⁷ This treatment has been shown to reduce bacterial adhesion to the surface of the implant by increasing the interfacial capacitance and decreasing the polarization resistance of the titanium.²⁶³ This treatment offers promising results where an external electrical stimulation may be used to noninvasively induce antimicrobial properties to the surface of titanium implants, without the need to chemically modify the surface of the implant.

6.2. Improved Recapitulation of Clinical Orthopedic Implant-Associated Infections In Vivo.

Widespread adoption of emerging technologies such as 3D-printing, bacterial genomic sequencing, and microbiome profiling have contributed to the development of orthopedic implant-associated infection in vivo models that more effectively recapitulate the clinical presentation of infections. Figure 1 highlights some of the recent advances in implant-associated infection in vivo models.

Figure 1.

Summary of recent advances in implant-associated infection preclinical in vivo models used to simulate the clinical environment of infections. Models have utilized emerging technologies of 3D-printing and bacterial genomic sequencing to improve the clinical relevancy of the model and to analyze the role of patient risk factors on the development of implant-associated infections. Figure created using BioRender.com.²⁶⁷

Decreasing costs of additive manufacturing over the past decade enable new implant designs that were previously impractical.^264,265 For example, 3D-printing has been used to design custom-fit orthopedic metallic in vivo implants in rabbits.^34,134 Micro-CT scans of the intact knee joint are used to design custom stainless steel tibial inserts.^34,134 Custom species-specific implants are particularly advantageous over implants traditionally used in preclinical models (non-species-specific) because they are more representative of clinical arthroplasties, may allow for more reproducible studies, and enable accelerated postsurgical loading of the implant to improve the evolution and treatment of implant-associated infections.^34,134 Nevertheless, the time intensive nature of custom printing implants that require initial micro-CT scans may hinder the progression of such models.

The 2018 International Consensus Meeting on Musculoskeletal Infection generated a list of “high priority” research questions related to implant-associated infections, half of which focus on modifiable patient factors.²⁶⁶ Animal models represent the best way of evaluating the contributions of patient factors to infection and have previously not been the focus of preclinical orthopedic infection models. Diabetes is a known risk factor for implant-associated infections due to the patient’s impaired innate immune system response and vasculopathy.³² Establishment of an effective diabetic implant-associated infection in vivo model can assist in the study of the progression and treatments of implant-associated infections for patients with an underlying comorbidity.³² More recently, the constituents of the gut microbiome have been implicated as a factor that can influence the development of implant-associated infections. Mice in which the composition of the gut microbiome had been modified throughout life were subjected to implantation of a titanium tibial component and bacterial inoculation.³³ Disruption of the gut microbiome resulted in a reduced ability to resist a small local inoculation of S. aureus as measured by CFU counts at the implant surface. Additionally, even when an infection was present, animals with an altered gut microbiome showed a reduced systemic response to implant-associated infection in the form of more muted changes in serum markers and immune cell populations following infection. These studies examining diabetes and the gut microbiome are just a few examples of ways in which animal models can be used to identify mechanisms linking patient factors to risk of implant-associated infection and ways of mediating that risk. Establishment of in vivo models that account for patient factors that more accurately portray clinical conditions can improve the evaluation of antimicrobial biomaterials in orthopedic implant-associated infection models.

6.3. Method to Assist in Standardization of Ex Vivo Mechanical Testing of Antimicrobial Materials.

Due to the variability of parameters used in existing pull- and push-out strength models, a standardized metric to evaluate interfacial shear strength would be desirable to ensure that antimicrobial arthroplasty materials are capable of withstanding shear forces. Ongoing work has focused on the development of a reproducible ex vivo push-out test method to calculate interfacial shear strength at the interface of bone and implanted PMMA bone cement. In this model, widely available cleaned/bleached bovine femoral tissue is used and machined into wafers (~4 mm thickness) with up to eight 5/32 in. diameter holes where PMMA is embedded (Figure 2). The bone wafer containing PMMA is then subjected to push-out testing under compressive loading using a specialized jig to ensure proper alignment of the pin (1/8 in. diameter) above each PMMA implant. Interfacial shear strength is defined as the peak force on the load–displacement plot. The setup of this push-out testing ex vivo model is particularly advantageous because many implants can be evaluated rapidly (up to 8 per waver) and the bone wafer has a uniform geometry, is reproducible, and can be customized in terms of the number of implants per wafer. Additionally, the bone wafer can be designed for bones of other large animal species (e.g., sheep). While this model may offer the possibility of assisting in the reproducibility of interfacial shear strength, it has several limitations. Specifically, the ex vivo model only provides basic information on the interdigitation of PMMA into excised bone and does not provide insight into the osseointegration of the PMMA upon bone remodeling in vivo. Additionally, this system may only be amenable for larger animal species.

Figure 2.

Setup of ex vivo PMMA bone cement push-out testing model using bovine femur tissue. Small, uniform wafers of bovine femur were machined from a cleaned/bleached femur segment (a) with eight 5/32 in. holes spaced 0.3 in. apart (b). Holes in the femur wafer were packed with PMMA materials, and a 1/8 in. steel pin was used to push embedded PMMA out of the femur wafer under compressive force at a loading rate of 20 mm/min (c). “Pushed-out” PMMA specimen from femur wafer (d). Representative raw force versus displacement plot of interfacial shear strength of PMMA pushed-out of a single 5/32 in. hole (e). Images were taken by the authors.

6.4. Enhanced In Vivo Imaging of Implant-Associated Infections.

To enhance the capabilities of imaging implant-associated infections real time in vivo, several radioactive tracers and probes have been developed. Radioactive tracers fiuorine-18-fluoro-2-deoxy-d-glucose (¹⁸F-FDG), gallium-68-labeled citrate (⁶⁸Ga-citrate), and gallium-68-labeled chloride (⁶⁸Ga-chloride) have recently been used to obtain high resolution images of orthopedic implant-associated infections^35,36 when combined with positron emission tomography and computed tomography. Both ¹⁸F and ⁶⁸Ga have been shown to have an increased uptake at the sites of bacterial infections. While ¹⁸F and ⁶⁸Ga have a similar uptake at bacterial infection sites, ¹⁸F has been shown to have a greater uptake in healing bone relative to ⁶⁸Ga; therefore, recent evidence has shown that ⁶⁸Ga imaging may be more amenable to use in implant-associated infection studies when bone healing is involved.³⁶ Additionally, near-infrared fluorescent molecular probes composed of the fluorophores H-sulfo-cyanine5 and diaminocyanine sulfonate have been developed to enable real-time imaging of inflammation and implant-associated infections noninvasively.³⁷ Reactive oxygen species are associated with inflammation, and H-sulfo-cyanine5 fluoresces when it reacts with reactive oxygen species, thereby enabling imaging of inflammatory sites,³⁷ whereas nitric oxide is released at bacterial infection sites, and diaminocyanine sulfonate fluoresces when it reacts with nitric oxide.³⁷ Real-time monitoring of implant-associated infections and inflammation in vivo can improve the analysis of the efficacy of antimicrobial orthopedic biomaterials.

7. CONCLUSIONS

The identification and design of an appropriate and clinically relevant in vivo orthopedic implant-associated infection model to evaluate the efficacy of antimicrobial materials is dependent upon several factors including the properties of the antimicrobial material in vivo. For example, the degradation, antimicrobial activity, and drug release kinetics of antimicrobial materials can influence the duration of the in vivo study and can first be evaluated in subcutaneous pouch implantation animal models.^{37,41,59,83,87,135,144,148,161} Furthermore, the bacterial inoculum is a key design feature of the model as it is specific to the animal species of interest, the virulence of the bacterial strain, and whether the model is of acute or chronic implant-associated infection.^{31,34,163–167} It is also critical to distinguish whether the antimicrobial material is intended to be used to either prevent or treat implant-associated infections as the structures of these models differ substantially (i.e., models to prevent infection involve placing antimicrobial materials at time of infection and models to treat infection involve establishment of infection and debridement prior to placing antimicrobial materials).^{23–25,38,94,118}

Several recent technological advances have assisted in the development of orthopedic implant-associated infection in vivo models that more effectively recapitulate the clinical presentation and monitor infections. These technologies include 3D-printing, bacterial genomic sequencing and microbiome profiling, and real-time in vivo imaging modalities.^33–37,134 These technologies have enabled the development of species-specific reproducible implants,^34,134 incorporation of patient risk factors for implant infection into the model (e.g., diabetes, gut microbiome constituents),^32,33 and real-time minimally invasive tracking of infection and inflammation in vivo.³⁷ Furthermore, the emergence of stimuli-responsive antimicrobial materials (e.g., responsive to temperature, magnetic/electric fields, microwaves, etc.) has resulted in the development of materials that can be applied minimally invasively for controlled gelling,^23,24 enable controlled release of antibiotics,²⁶ and enable noninvasive induction of antimicrobial activity after implantation.²⁷ Moving forward, emphasis can be placed on expanding imaging technologies of pathogens and antimicrobial agents (e.g., tracers and tags) in preclinical models to further improve the evaluation of the efficacy of novel antimicrobial biomaterials in clinically relevant implant-associated infection models.