CORR Insights®: Is Implant Coating With Tyrosol- and Antibiotic-loaded Hydrogel Effective in Reducing Cutibacterium (Propionibacterium) acnes Biofilm Formation? A Preliminary In Vitro Study

Where Are We Now?

Treating orthopaedic infections can be difficult because many of the common pathogens form bacterial biofilms, which reduce the effectiveness of antibiotics even when the isolated bacteria are susceptible to those drugs [7]. In-vitro studies [1, 4] have found that achieving even modest reductions in bacteria in biofilms may require dosages that exceed those that can be achieved through systemic administration. Local release of antibiotics at the site of the implants by materials such as bone cement, absorbable bone-void fillers [4], and hydrogel coatings [3] offer the potential to achieve such dosages. In addition, because of concerns regarding antibiotic resistance, researchers are looking for novel agents with antibiofilm efficacy.

The study by Tsikopoulos and colleagues [9] assessed the ability of hydrogels impregnated with tyrosol, as well as conventional antibiotics coated on the surface of titanium (Ti) coupons, simulating an orthopaedic implant surface, to prevent biofilms formed by Cutibacterium (Propionibacterium) acnes (C. acnes). Tsikopoulos and colleagues [9] speculated that tyrosol could potentially protect shoulder implants from C. acnes. Although C. acnes strains generally are susceptible to antibiotics like vancomycin, they become highly tolerant when they form biofilms [2, 5]. The team reasoned that if tyrosol was incorporated into absorbable hydrogels and then applied to the surface of orthopaedic implants prior to implantation it would have antibiofilm activity localized at the surface. For comparison, the team used rifampicin and vancomycin in hydrogels and when applied as solutions. The team found that tyrosol had antibacterial activity against C. acnes but the minimum inhibitory and bactericidal concentrations (MIC and MBC) were high, at 9 and 35 mg/mL, respectively. However, when tyrosol was in the hydrogel (9 mg/mL) there was no inhibition of planktonic growth. Surprisingly, neither tyrosol, vancomycin, nor rifampicin inhibited biofilm formation when loaded into the hydrogel, even though they were in high concentrations. Taken together these results suggest that either the release of kinetics did not allow the antimicrobial agents to build up to MIC or MBC levels or the hydrogel had an antagonistic effect.

These data contradict in vitro findings by a different group of researchers who showed that the Defensive Antibacterial Coating (DAC®; Novagenit Srl. Mezzolombardo, Italy) hydrogels loaded with various antibiotics, including vancomycin and rifampicin, did reduce biofilm [3]. Drago and colleagues [3] conducted release kinetic assays and used multiwall plates with disks, as well as a number of full-size implant models and a range of Gram-positive and negative prosthetic joint infection (PJI) pathogens but not C. acnes. It is not clear, therefore, whether differences between these studies might be due to the challenge organism(s) or differences in the experimental procedures. It is also not clear whether tyrosol incorporated in DAC on the surface of implants could potentially control biofilms formed from other common PJI pathogens.

Where Do We Need To Go?

The titles of the current study [9] and of Drago and colleagues [3] ask similar questions regarding the efficacy of implant coated antimicrobial loaded DAC hydrogels to control biofilms, but come to opposite conclusions. The most obvious difference is in the choice of challenge organisms, although it is difficult to see why C. acnes would be so much harder to kill than the other pathogens, when MICs, MBCs, and minimum biofilm eliminating concentration values were similar, or less, than many of the strains tested. Another possibility could be differences in the assays themselves. System-specific factors such as contact time scales, residence times, liquid volumes, surface areas within the system, inoculum density and growth phase, maturity of the biofilm, and media type likely all come into play. In addition, there are no standardized internal controls (strain type and antibiotic) that can be used as comparative yard sticks, as there are for standard clinical MIC and MBC methods. It is important to ask in the setting of an in vitro test what constitutes a meaningful biofilm reduction. Here, 80% is used as a threshold. This was taken from a previous study [8], although there was no rationale provided. Probably because there are little or no data directly correlating in vitro efficacy with clinical outcome when it comes to antibiofilm technologies, leaving researchers guessing as to what might be clinically relevant. As the difficulties for conventional antibiotic therapies to treat orthopaedic infections continue to be recognized, there is a concomitant drive to develop new technologies and agents specifically targeting biofilms. The recognition that treating biofilm-associated infections presents different challenges than treating acute planktonic infections by surgeons, researchers (both in biology and materials science and engineering), and commercial companies is a good thing; necessity is the mother of invention.

How Do We Get There?

The pathway to in-vitro standardized methods to directly compare the efficacy of different technologies for treating biofilm-associated orthopaedic infections is not straightforward. Unlike acute infections in which the bacteria tend to be homogeneous and rapidly growing, biofilm bacteria contain multiple phenotypes, including dormant and persister populations. There is large variation in the pathogens, types of surfaces (materials, textures, scale of surface features), and environments (body fluid components, shear stresses, mass transfer conditions) with which biofilms are associated. Thus, the properties of the biofilms are likely to be quite different and difficult to recapitulate in a few simple standard growth models. Similarly, there is much variation in antibiofilm technologies and in many cases a test system has to be tailored to be compatible with a particular mode of action. Just as the recognition of biofilms has led to growth in the development of antibiofilm technologies, so will the recognition for the needs for developing systems that more closely capture the salient features of the infecting biofilm environment, whether these be physical, biological, or both. I concur with Malone and colleagues [6] who believe we need to develop a framework for biofilm testing which includes a decision tree to provide direction as to which standard method should be used to answer a particular research question. In vitro standard methods for orthopaedics need to reflect more closely the in vivo conditions and be compatible with the mechanism of action of the therapeutic technology being tested. It is important to incorporate clinical factors such as relevant volumes, surface areas, shear stresses, pH, oxygen tension, viscosities, drainage rates, fluid exchange, nutrient availability, exposure times, and interaction with host cells into the model as much as possible. Conflict between the balance of simplicity required for a rapid screen and complexity is inevitable. A good in vitro model should be as simple as need be, yet still be predictive of clinical outcome.

Three areas of basic research which will drive the field forward are: (1) mapping the biofilm to more precisely pinpoint the anatomic location of the infection; (2) better understanding of the local mechanical, chemical and biological milieu at those locations, and (3) the development of models which reflect the relevant features of those locations. Mapping the biofilm is restricted because currently there are no biofilm specific targets and the size scale of biofilms is often below that of medical imaging modalities. Development of biofilm contrast agents for imaging would be an enormous breakthrough. There is great untapped potential to learn from cancer tumor imaging, through forging collaborations between radiologists, tumor biologists, orthopaedic surgeons, and biofilm microbiologists. With respect to the local biofilm environment, it is possible that some of these anatomical parameters can be determined from current medical imaging modalities while others may be determined from clinical data such as periprosthetic pharmacokinetics of antibiotics and antimicrobials released locally. Still, others may require cadaveric joint or animal research. Lastly, deveopments in tissue engineering incorporating host cells, and hard and soft tissue scaffolds has great potential for the development of models. While there are many challenges to overcome, there are also many exciting opportunities that will require collaboration between bioengineers, material scientists, anatomists, physiologists, surgeons, immunologists, and biofilm microbiologists.