Implant‐associated osteomyelitis: Development, characterisation, and application of a porcine model
2021; Wiley; Volume: 129; Issue: S141 Linguagem: Inglês
10.1111/apm.13125
ISSN1600-0463
Autores Tópico(s)Infective Endocarditis Diagnosis and Management
ResumoThe author would like to thank the Independent Research Fund Denmark (Technology and Production) and the EU Horizon 2020 research programme for funding the research presented in this doctoral dissertation. In particular, I would like to thank Professor in Veterinary Pathology Henrik Elvang Jensen who participated in all parts of the experimental studies. Thanks for your inspiration, guidance and enthusiasm. My thanks also go to all the staff at the Pathobiological Sciences Section for being great colleagues and for the many happy times during work and social events. A special thanks to laboratory technicians Elizabeth Petersen and Betina Andersen for all your help in the histo-lab. I apologize for all the challenging and problematic issues regarding histological preparation of bones. Many thanks also to technicians Dennis Brok and Frederik Andersen for their skilful technical support and help producing images. I would also like to thank the animal technicians at the Department of Experimental Medicine for their excellent care of the pigs and help with anaesthesia, blood sampling and inoculation procedures. Thank you to all my research assistants during the past 5 years, including Kristine Dich-Jørgensen, Anne Sofie Boym Johansen, Nicole Lind Henriksen and Amalie Blirup-Plum. Thank you for your hard work, including endless hours spent evaluating and scoring histological slides, and thank you for bringing enthusiasm and fun into our daily routine. Thanks also to Expert in Experimental Surgery Janne Koch, Associate Professor in Veterinary Microbiology Bent Aalbæk and to Professor in Chronic Infections and Biofilm Thomas Bjarnsholt. You have all been indispensable for the research described in this dissertation. Thomas, I enjoyed our biofilm discussions and I am grateful for all your support. Thanks to everyone in Professor and Orthopedic Surgeon Kjeld Søballe's research group at Aarhus University Hospital for our collaboration on antimicrobial penetration into infected bone tissue. Mikkel Tøttrup, Pelle Hanberg and Mats Bue, it has been a pleasure to work with you. Finally, thanks to all my friends and family for their never ending support and trust, and thanks to my wonderful husband and daughters, Kim, Alberte and Frida, for your love and patience. Worldwide, there has been an ongoing increase in the number of bone infections that lead to amputations and lifelong disability, affecting millions of people every year. Therefore, research investigating the prevention, diagnosis and treatment of bone infections is vitally important. However, to develop effective new approaches and techniques for managing bone infections, preclinical testing and evaluations using reliable animal models are necessary. A novel porcine model of implant-associated osteomyelitis (IAO) in humans was recently developed. The model was based on female pigs, and osteomyelitis was induced by inoculation of Staphylococcus aureus bacteria into a predrilled tibial cavity (2 × 20 mm). Following inoculation, a steel implant (2 × 15 mm) was inserted into the cavity. The animals were euthanized 5 days after inoculation. Control pigs were exposed to the same surgical procedure and inoculated with sterile saline. The success rate of the model was 100%; that is, all the pigs inoculated with bacteria developed osteomyelitis. Bone lesions similar to those found in human patients with osteomyelitis developed in the porcine IAO model, and the inoculated bacteria were detected both within peri-implant bone tissue and on the surface of the implants. Thus, peri-implant bone tissue may serve as a reservoir for biofilm (bacterial aggregates surrounded by an extracellular matrix) shortly after surgical contamination. Biofilms are extremely tolerant to antibiotics and are reportedly the main reason for human bone infection treatment failure. This new porcine IAO model was used to develop a biofilm staining technique, which combined histochemistry (HC) and immunohistochemistry (IHC). The new staining technique allowed the bacterial cells and extracellular matrix to be visualized simultaneously: Alcian blue pH3 stained the carbohydrates of the extracellular matrix blue, and IHC treatment with an antibody specific for S. aureus coloured the bacterial cells red. The new staining technique could also be used reliably on human bone tissue with chronic staphylococcal osteomyelitis. Two microdialysis studies were performed using the porcine IAO model. These studies found that systemically administered antibiotics (cefuroxime and vancomycin) showed significantly less penetration into the tissue surrounding infected bone implants than into healthy bone tissue. Furthermore, this reduced antimicrobial penetration was correlated with the progression of peri-implant bone lesions. Bone lesions that extended to a depth of >3 mm from the implant cavity into the bone tissue showed almost no antimicrobial penetration. The reduced penetration was due to suppurative and necrotic bone inflammation and expansion of the implant cavity. Consequently, diffusion of antimicrobial agents from the capillary system into the implant cavity was hampered because it had to cover a greater distance. Bone inflammation also had a negative impact on the efficacy of a single dose of locally administered gentamicin. Different doses of gentamicin were added to the inoculum 1 min prior to inoculation into the porcine IAO model. Due to the development of acute inflammation including vasodilatation and increased vascular permeability, only high doses of gentamicin (>1000× minimum inhibitory concentration) were effective at the implant surface. This shows that the prophylactic concentration of locally administered antimicrobial agents cannot be evaluated solely by using in vitro assays. A systematic review of all contemporary large, non-rodent animal models of bone infection (i.e. goats, sheep, dogs, pigs and rabbits) was performed. Overall, it was found that experimental design was poorly reported, methods were of poor quality and the pathological parameters used varied significantly. Therefore, study template guidelines as a standard for reporting on animal models of bone infection was established. It was apparent that the animal species per se was one of the most important study design parameters. Based on a narrative review, we found that there were many advantages in using pigs for modelling bacterial biofilm infections in humans because comparable infections also occur spontaneously in pigs. Currently, there are many problems associated with treating patients who have chronic biofilm-based infections, including bone infections. Furthermore, the increasing prevalence of antimicrobial resistance means that new treatments for infectious diseases are urgently required. The novel porcine IAO model described here is a valuable and reliable tool for investigating new prophylactic strategies and treatment regimens for bone infections in humans. Flere og flere patienter får diagnosen knogleinfektion. En knogleinfektion fører til operation, indlæggelse, flere ugers antibiotikabehandling og i yderste konsekvens amputation. De fleste knogleinfektioner opstår, fordi der kommer bakterier ind i knoglen under en operation. Det kan være under indsættelsen af en ny kunstig hofte, eller ved korrektion af et benbrud hvor der anvendes osteosyntese. Knogleinfektioner er enormt svære at behandle, fordi bakterierne danner biofilm, dvs. de går i dvale, og danner en beskyttende matrix omkring sig. Biofilmdannelse beskytter bakterierne mod kroppens eget immunforsvar og det antibiotikum som gives i behandlingsøjemed. Den stigende udvikling af antibiotikaresistens udgør en alvorlig trussel mode vores evne til at bekæmpe infektioner, og derfor er forskning indenfor forebyggelse, diagnosticering og behandling af knogleinfektioner mere relevant end nogensinde før. For at udvikle nye tiltag og teknikker med klinisk relevans, er det helt afgørende, at disse testes og evalueres i velkarakteriserede dyremodeller, som efterligner den humane patologi. Denne afhandling beskriver udviklingen, karakteriseringen og anvendeligheden af en ny grisemodel for implantat-relateret osteomyelitis hos menneseker. Grisemodellen er baseret på, at der bores en lille kavitet i højre tibia hvori der inokuleres saltvand eller Stafylokokkus aureus bakterier. Efter inokuleringen indsættes et lille implantat (2x15mm). Grisene aflives efter 5 dage og alle dyr inokuleret med bakterier udvikler komparative knoglelæsioner. De inokulerede bakterier genfindes på både det indsatte implantat og et stykke inde (1 cm) i det omgivende væv. Modellen demonstrerer dermed, at det knoglevæv som omgiver et implantat kan udgøre et biofilmreservoir allerede korttid efter kirurgisk kontaminering. Den nye grisemodel er blevet brugt til, at udvikle en ny farvemetode der kan synliggøre biofilmdannelse i vævsnit ved brug af almindelig lysmikroskopi. Princippet i den nye metode er en kombination af almindelig histokemi og immunohistokemi. Ved at kombinere en protokol for Alcian Blue pH3 med immunhistokemi baseret på et S. aureus specifikt antistof, kunne både matrix og bakterier i biofilmaggregater visualiseres i to forskellige farver. Den nye farvemetode er også anvendt på humant væv med kronisk osteomyelitis. Knogleinfektioner kræver lang tids antibiotikabehandling og de eksisterende doseringsprotokoller bygger på farmakokinetiske studier af rask knoglevæv. Baseret på grisemodellen for implantat-relateret osteomyelitis blev det demonstreret, at penetrationen af systemisk indgivet antibiotika (cefuroxime og vancomycine) til inficerede knogleimplantater, er signifikant reduceret sammenlignet med normalt rask knoglevæv. Den reducerede antibiotika penetration var korreleret til udbredelsen af patologiske forandringer i knoglevævet omkring implantatet. I de tilfælde hvor læsionerne udbredte sig mere end 3 mm fra implantatet, var der næsten ingen penetration. Den reducerede penetration blev forårsaget af en purulent og nekrotisk inflammation, hvilket ødelage knoglevævet og skabte en kavitet omkring implantat. Derfor kan det antages, at diffusion af antibiotika fra kapillærerne og indtil implantatet er blevet nedsat delvist pga. en øget diffusionsafstand. Udover den store effekt på penetrationen af antibiotika, blev det også påvist, at det inflammatoriske respons har stor indflydelse på effekten af lokal indgivet antibiotika. Forskellige doser af gentamicin blev iblandet det bakterielle inokulum et minut før inokulering i grisemodellen. På grund af udviklingen af akut inflamamtion, med vasodilation og øget vaskulær permeabilitet, var det kun de meget høje gentamicin doser (>1000× MIC) som kunne fastholde en baktericid koncentration på implantatets overflade. Derfor kan det konkluderes, at det for knogleinfektioner, er utilstrækkeligt at basere effekten af antibiotika udelukkende på studier af raskt væv og på in vitro assays. I et stort systematisk review af alle ikke-gnaver modeller for knogleinfektioner blev det påvist, at der de sidste 10 år er sket en markant stigning i brugen af større forsøgsdyr som fx får og grise. Det blev også påvist, at afrapporteringen af studiedesign ofte var meget mangelfuld, og at den metodologiske kvalitet var meget lav. Dette førte til udviklingen af et set retningsliner for standard afrapportering ved udvikling og anvendelse af dyremodeller for knogleinfektioner. Det systematiske review viste, at dyrearten er en af de mest afgørende design parameter. Baseret på en gennemgang af alle grisemodeller for bakterielle infektioner hos mennesker, blev det konkluderet, at grisemodeller er særdeles fordelagtige som modeldyr for biofilm baseret infektioner hos mennesker, fordi grise (slagtesvin) udvikler komparative infektioner spontant. Der er store problemer med behandlingen af patienter som lider af kroniske infektioner inklusive knogleinfektioner. Det er tydeligt, at den nye grisemodel for implantat-relateret osteomyelitis er et brugbart og pålideligt redskab, til at studere nye profylaktiske strategier og behandlingsregimer. Worldwide, there has been an ongoing increase in the number of bone infections that lead to amputations and lifelong disability, affecting millions of people every year [1, 2]. Consequently, bone infections constitute a substantial economic burden in terms of patients, physicians, hospitals and healthcare systems [1, 2]. When applied to patients, the term 'bone infection' can include prosthetic joint infections (PJIs), fracture-related infections (FRIs), implant-associated osteomyelitis (IAO), chronic osteomyelitis (CO), osteomyelitis in children and diabetic foot osteomyelitis (DFO) [3]. Presently, DFO is the leading cause of lower extremity amputations, and it is estimated that a lower limb is lost every 30 s due to DFO [4, 5]. The increased number of bone infections is mainly due to an increase in the size of the elderly population, an increased prevalence of diabetes and an increase in the number of joint prostheses and bone fixation implants being used [1]. The persistence of the problem and the unsatisfactory proportion of positive treatment outcomes implies that the current prophylaxis and treatment strategies are incomplete despite best practice [2]. Therefore, future research must focus on prevention, diagnosis and treatment of bone infections. However, the development of effective new approaches and techniques will depend on preclinical testing and evaluations performed in reliable animal models. In this dissertation, a novel porcine IAO model is presented. The publications referred to (I–IX) describe the development, characterization and application of this model. Chronic bacterial infections are generally caused by biofilm-forming bacteria [6]. Therefore, it is surprising that most reports regarding bacterial biofilms are based on in vitro observations, because these laboratory findings cannot be extrapolated into clinical settings (VIII) [7]. Consequently, there are many problems associated with treating patients who have chronic biofilm-based infections, including bone infections. In addition, the increasing prevalence of antimicrobial resistance means that new treatments for chronic infectious diseases are urgently required [8]. The porcine IAO model has generated new relevant in vivo observations regarding biofilms (II, VI, VIII). In particular, studies have shown that biofilms do not simply involve artificial surface attachment; therefore, the old dogma used to describe biofilms in terms of a 'race for the surface' seems to be clinically inadequate (II, VI, VIII). Biofilm formation also affects tissues, and research into both tissue and implant biofilms is equally important for evaluating chronic bone infections. Promising new approaches to prevent biofilm formation in orthopaedic research include the following: 1) modification of implant surfaces to prevent bacterial adhesion, 2) coating of implants so that they can elute high concentrations of antibiotics locally (without causing systemic toxicity), 3) new drugs directed against bacterial adhesion molecules or quorum sensing and 4) the development of vaccines against biofilm-forming bacteria [9]. All of these new technologies for combating osteomyelitis may be tested using the porcine IAO model. During the past 5 years, the porcine IAO model has been applied in several studies [10-13]. In brief, the model was used as the basis of an EU HORIZON 2020 Research and Innovation Programme project (Novel Marine Biomolecules Against Biofilm [NoMoRFilm] no. 634588) to identify new antibiotics in micro-algae. As a result, one promising anti-biofilm compound was isolated, synthesized, chemically characterized and evaluated for anti-infective properties in the porcine model. In addition, the NoMorFilm project also resulted in the development of a new surface-coating technique for orthopaedic implants. This technique provides a coating that can bind and release high concentrations of antibiotics. An EU patent application based on the new coating technique is currently under review and describes very successful test results obtained using the porcine IAO model. In addition to the prophylactic coating studies, the porcine model was also recently used to evaluate the therapeutic impact of a gentamicin-loaded biodegradable bone void filler, following limited or extensive debridement of osteomyelitis lesions [12]. Currently, the porcine IAO model is also being used to investigate the molecular orchestration of bacterial bone infections. Gene expression analyses of porcine and human infected bone biopsies have shown that bone tissue can mount and sustain a local acute phase response (i.e. an extra hepatic acute phase response) [11]. In addition, research has shown that the classical receptor activator of nuclear factor κB (RANK)–RANK-ligand (RANKL) pathway is not responsible for bone loss resulting from bacterial osteomyelitis [13]. In particular, the molecular studies demonstrated upregulation of specific inflammatory genes, which had no murine homologues; that is, the corresponding genes are not present in mice. This dissertation consists of six chapters. Chapter 1 provides background information regarding IAO in humans. Chapter 2 focuses on previously studied non-rodent animal models of bone infections and the advantages of using pigs for modelling osteomyelitis. In Chapter 3, the development of the new porcine IAO model is described. Chapter 4 focuses on a new staining technique for in situ visualization of biofilms and biofilm formation in spontaneous porcine bacterial infections. Chapter 5 describes the application of the porcine IAO model in studies investigating antimicrobial penetration into infected bone tissue and the prevention of biofilm formation. Finally, Chapter 6 summarizes the main conclusions from the publications on which this dissertation is based. In elective trauma surgery, bone infections occur at a rate of 1%–5% after closed fractures, and at a rate of 3%–50% after open fractures [14]. In 2004, it was estimated that two million fracture fixation procedures were performed each year in the United States leading to approximately 100,000 cases of IAO annually [15]. Risk factors for the development of infections include smoking and comorbidities such as diabetes, immunosuppression and chronic infections at other sites [16]. Fractures occur most frequently in the feet and hands, followed by the long bones, that is the humerus, radius, ulna, femur, tibia and fibula (Fig. 1) [16]. If IAO occurs after fracture fixation of a long bone, this may delay healing (Fig. 1) or lead to permanent functional loss or even amputation of the affected limb [2, 16]. The term 'osteomyelitis' refers to inflammation of the bone marrow, and the term 'osteitis' refers to inflammation of the entire bone including the cortex [14]. Commonly, osteomyelitis is the preferred term and this is used synonymously for both bone infection conditions in patients [14]. IAO may be classified based on the time from surgery to occurrence of the infection. 'Early infections' are seen within 2 weeks after bone fixation, 'delayed infections' between the third and the 10th week, and 'late infections' occur more than 10 weeks after implantation [2, 16]. 'Late infections' can be acute if they are caused by a haematogenous infection; otherwise, they represent a chronic 'delayed infection' or the recurrence of an incorrectly treated 'early infection' [16]. Clinically, osteomyelitis is also frequently classified as either acute or chronic. CO is subjectively defined as a bone infection with symptoms that persist for at least 6 weeks, the presence of bone pathology in an image, or infections that require major interventions due to sequestra or deformities [17]. In adults, CO is often associated with foreign bodies, for example implants [17]. Therefore, 'delayed' or 'late' IAO may also be characterized/described as CO. Osteomyelitis in the long bones may also be classified using the Cierny–Mader staging system, which correlates with treatments and prognoses [18, 19]. The terms 'acute' and 'chronic' are not used in this classification system and the stages are dynamic, reflecting the anatomy and pathophysiology of the disease [18, 19]. The Cierny–Mader system classifies osteomyelitis into four stages [18, 19]. Stage 1, or medullary, osteomyelitis is confined to the medullary cavity of the bone. Stage 2, or superficial, osteomyelitis involves only the cortical bone. Stage 3, or localized, osteomyelitis usually involves both cortical and medullary bone. In this stage, the bone remains stable, and the infection does not involve the entire bone diameter. Stage 4, or diffuse, osteomyelitis involves the entire diameter of the bone, with loss of stability. The Cierny–Mader system includes a second dimension and classifies the host as A, B, or C [18, 19]. An A-host is a patient with no systemic or local compromising factors [18]. The B-host is affected by systemic or local factors that affect the immune response, metabolism and local vascularity [18]. C-hosts are patients who are so severely compromised that they cannot have surgery [18]. The presence of pyogenic bacteria within bone tissue and on the surface of bone implants will lead to a suppurative inflammatory reaction dominated by neutrophils (II, III, V) [17, 19]. Bone lysis will be induced by proteolytic enzymes released from the inflammatory cells and the activation of matrix metalloproteinases [20]. Furthermore, the activation of osteoclasts will also contribute to bone matrix resorption [20]. Within the trabecular bone tissue, suppuration will increase the intraosseous pressure and cause compression of blood vessels resulting in thrombosis, which leads to ischaemic bone necrosis (i.e. trabecular sequestra; II, III, V) [17, 19]. If the exudate enters and spreads within the cortical Volkmann's and Haversian canals, this will lead to a compromised medullar and periosteal blood supply and generate separated dead cortical segments (i.e. sequestra) [17, 19]. Regardless of location, a bone sequestra provide a perfect substrate for bacterial growth [17, 19]. Eventually, an infectious bone lesion will contain granulation tissue and fibrosis, which may be surrounded by osteoid-producing osteoblasts [17, 19]. Osteogenic stem cells in the periosteum will differentiate into osteoblasts and form a sheath of vital bone, called an involucrum, which surrounds the dead cortical bone [17, 19]. The involucrum is irregular and is often perforated with openings through which pus may spread into the surrounding soft tissues and ultimately drain to the skin surface via a sinus tract. The density of the involucrum may gradually increase and eventually almost forms a new diaphysis [17, 19]. Most cases of IAO are caused by Staphylococci, and Staphylococcus aureus and Staphylococcus epidermidis (Fig. 1) are the most common aetiologies [2, 16, 21]. Other pathogens that are frequently responsible for IAO include Streptococci, coagulase-negative staphylococci, virulent gram-negative bacteria and Propionibacterium acnes [2, 16, 21]. Early infections are often caused by highly virulent pathogens such as S. aureus, Group A Streptococcus and gram-negative bacteria [2, 16, 21]. The three major routes by which bacterial pathogens can access bone tissue are: 1) direct, 2) haematogenous or 3) contiguous. [16] Direct or exogenous infections are caused by bacterial contamination of the bone tissue during trauma and surgery or during the perioperative period [16]. Haematogenous infections occur due to bacterial seeding following bacteraemia, and contiguous infections are a result of progression from arthritis or soft tissue infections to the adjacent bone fixation device [16]. Both bone tissue exposed to surgery and the inserted bone implants are highly susceptible to infections [19, 22]. If the infecting bacteria are not eliminated rapidly, they will adhere to the bone tissue and/or the surface of the implant (II, III, VI, VIII) [19, 22]. This process of bacterial attachment is mediated by physical factors, for example surface tension, hydrophobicity and electrostatic interactions or specific adhesion to plasma proteins such as fibronectin and fibrin [22-24]. Normally, healthy bone tissue is highly resistant to infection. Thus, the increased susceptibility to infection is a direct consequence of orthopaedic procedures. Orthopaedic surgery to implant plates, screws or prostheses, correct fractures or debride IAO will cause tissue damage and bleeding of exposed bone ends (Figs 1C and 2B). Immediately thereafter, the bone tissue and any inserted implant will be covered by plasma proteins (Figs 2 and 3). Interestingly, the plasma protein coating plays a more important role in bacterial surface adhesion to implants than the implant material (e.g. steel or titanium) [22]. If the infecting bacteria are planktonic (single), then after adhering to the plasma protein-embedded bone tissue or implant, they will transform into a biofilm [6]. However, small aggregates of biofilm from the skin may also initiate the infection (Fig. 2A,C). A biofilm is defined as a cluster of bacterial cells embedded in an extracellular matrix, which is more tolerant towards antimicrobial agents and more capable of resisting the host immune defence mechanisms than planktonic bacteria [6, 25-27]. Bacteria growing in a biofilm will persist, whereas planktonic bacteria will be susceptible to the antibiotics used to treat them (VIII) [6]. This increased antibiotic tolerance of biofilm bacteria is mainly due to their lower metabolic activity and their supporting extracellular matrix, which can bind and inactivate antimicrobial agents (Fig. 3B) [6]. Biofilm bacteria are 1000- to 10,000-fold more tolerant of antibiotics than their planktonic counterparts [28]. A bone infection biofilm will also mature over time, increasing its antimicrobial tolerance [22]. Because of this biofilm maturation, the probability of successfully treating a bone infection decreases dramatically from 80–90% to 30–60% if treatment is initiated more than 3–4 weeks after infection [22]. Osteomyelitis research has shown that eradicating tissue-based biofilms is extremely difficult. Antimicrobial therapy alone is often unsuccessful in treating these infections, due to the high levels of antimicrobial tolerance acquired by mature biofilms, and implant removal and debridement of necrotic bone tissue may be necessary (Fig. 1) [16, 29]. Even after removal of the infected implant and extensive debridement, prolonged systemic antibiotic treatment over several weeks may be needed to render the bone culture negative [16]. Moreover, the main reason for recurrence of IAO is bacterial biofilm survival in the bone tissue (Fig. 2C) [29]. Non-rodent animals such as sheep, goats, dogs, pigs and rabbits are being used more frequently in preclinical orthopaedic research due to their relatively large bones, which are more accessible for the insertion of many orthopaedic devices than those of mice and rats [30]. This trend is also being reflected in the study of bone infections. Therefore, a comprehensive systematic review was carried out to provide an overview of previous studies on non-rodent (i.e. sheep, goats, dogs, pigs and rabbits) animal models of bone infections (IX). The aim was to identify differences in study design parameters, for example bacterial inoculation dose or infection time, among the five different animal species. The review also investigated the methodological quality of the studies and post-mortem recording and evaluation of pathology and microbiological factors. The review was based on a systematic search of two electronic databases (PubMed and Web of Science). This resulted in a total of 316 publications that fulfilled the inclusion criteria (i.e. experimental bacterial inoculation of animals in order to model any type of bacterial bone infection in humans; IX). Since 2005, rabbits, pigs and sheep have been used more frequently as experimental animals to study bone infections that affect patients (Fig. 4; IX). While the use of ovine and porcine models is an alternative to previously preferred models developed in dogs, the increasing use of rabbit models reflects a demand for easy handling, easy housing and low cost in vivo studies in a larger animal than a rodent [31]. Only a few differences were observed in the study design parameters among the five different animal species (e.g. sex, age, inoculation dose or time from inoculation to euthanasia; IX). However, porcine models, including the model described in this dissertation, were distinctive regarding a number of specific points. In particular, the mean bacterial inoculation dose was lower and the infection period (from inoculation to euthanasia) was shorter in pigs than in all other species (I–III, VII, VIII). Regardless of the animal species used or the type of bone infection modelled, the bacteria used most frequently for inoculation was S. aureus, although most of the studies did not report a specific identification code for the inoculated strain (IX). In general, studies on models that were based on traumatic inoculation directly into a specific bone reported using a significantly lower inoculation dose if an implant was inserted at the same time as the inoculum. Therefore, results from preclinical bone infection studies are consistent with those obtained in 1957 by Elek et al., who demonstrated that a foreign body can reduce the number of bacteria needed to establish an infection [32]. Elek et al. injected 7.5 × 106 colony forming units (CFU) of S. aureus into the skin of human volunteers, resulting in only 50% of the volunteers becoming infected and 100% of the cases being resolved. However, when Elek et al. inserted an implant and injected 100 CFU of S. aureus, all participants became infected and none of the infections were
Referência(s)