Periprosthetic joint (PJI) infection following total joint arthroplasty of the hip or knee is one of the most devastating complications that can affect this otherwise successful surgical procedure. Although the risk of PJI after primary total joint arthroplasty is reported to be 1%1,2, recent analyses suggest that the incidence of PJI is increasing3,4, culminating in a predicted incidence of >65,000 new cases of PJI annually in the United States by 20305. The introduction of bacteria into the periprosthetic space occurs either intraoperatively6-8 or postoperatively through hematogenous dissemination9. Once inside the joint, bacteria such as Staphylococcus aureus rapidly adhere to the implant surface and establish an extracellular biofilm that protects them from the immune response, antibiotic therapy10, and even mechanical debridement11. The capability of bacteria to thrive within the periprosthetic space is reflected in the failures associated with current PJI treatments. Treatment of acute PJI through irrigation and debridement has a failure rate of 29% to 92%12-15. The gold-standard treatment for chronic PJI, 2-stage revision, fails to eradicate infection in approximately 25% of cases16-18.
Although pioneering in vitro investigations have improved our comprehension of initial bacteria-implant interaction and biofilm formation, poor outcomes of current treatments reflect a pressing need to develop a representative PJI animal model that permits quantitative analysis of both biofilm formation and periprosthetic bacterial loads. Current animal models involve the insertion of a loose, biofilm-coated intramedullary metal wire into a mouse femur19,20 or drilling a screw into the extra-articular distal femoral metaphysis of a rabbit21. Such approaches do not model a clinically relevant etiology of PJI, do not use clinically representative implants, and do not recreate the unique biological environment whereby an arthroplasty implant separates the relatively avascular articular space from the well-vascularized metaphyseal bone. In response to such limitations, we developed a mouse model that successfully reproduces the clinical, radiographic, and microbiological features of clinical PJI. We present the model along with quantitative and qualitative assessments of ambulatory status, biofilm formation, radiographic changes, soft-tissue damage, and bacterial loads both on the implant surface and within periprosthetic tissues.
Materials and Methods
All animal experiments followed protocols approved by the Hospital for Special Surgery (HSS) animal care and use committee.
S. aureus Strain and Mouse Implant Production
S. aureus Xen 36 (PerkinElmer) is derived from a clinical isolate of parental strain S. aureus ATCC (American Type Culture Collection) 49525 and is resistant to kanamycin. For all studies, Xen 36 was cultured overnight in tryptic soy broth (TSB) media containing 200 μg/mL of kanamycin to remove contaminants.
For in vivo and in vitro investigations, we utilized a 3-dimensionally printed mouse-sized tibial implant composed of the same Ti-6Al-4V alloy utilized in human components22. The implant has a smooth tibial plateau (2.0 × 1.5-mm oval with a mean surface roughness of 8.4 μm) and a 2.0 mm-long intramedullary stem textured with 40-μm diameter titanium spheres. We previously demonstrated that implanted mice could bear weight immediately postoperatively and that osseointegration in response to intermittent parathyroid hormone (PTH) began by day 7 post-implantation22. Prior to insertion, implants were passivated with 25% nitric acid for 24 hours, in accordance with ASTM F86 (Standard Practice for the Surface Preparation and Marking of Metallic Surgical Implants)23, neutralized with phosphate buffered saline (PBS) solution, and then autoclaved.
In Vitro Biofilm Formation
The effect of bacterial concentration and time on biofilm formation in vitro was quantified using a crystal violet assay24,25. The mouse implants were placed in wells of a 96-well plate containing a 250-μL suspension of 104, 105, or 106 colony forming units (CFUs)/mL in TSB. Wells with TSB alone acted as a control. Plates were incubated in a shaking incubator at 37°C and removed at 1, 3, and 24 hours. At each of these time points, the implants were rinsed twice, placed in wells containing 250 μL of 0.01% crystal violet (C0775; Sigma Aldrich) for 15 minutes, rinsed again, and allowed to dry. Crystal violet adherent to biofilm on implants was solubilized with 250 μL of 33% glacial acetic acid (A6283; Sigma Aldrich) and then transferred to a 96-well optical well plate (164588; Thermo Fisher Scientific). Solubilized crystal violet was measured using a multimode microplate reader (Infinite 200 PRO; Tecan Schweiz) at an absorbance of 590 nm26. All experiments were performed in quadruplicate and were repeated 3 times.
Surgical Technique and Bacterial Inoculation
Pilot experiments were performed to determine the intra-articular volume of the mouse knee following implant insertion27. In 12-week-old C57BL/6 mice (Jackson Laboratory), a 2-μL injection of methylene blue administered after implant insertion and arthrotomy closure remained within the intra-articular space without spreading into the medullary canal or femoral vein. Introduction of bacteria into the joint space reproduced the hypothesized clinical etiology of PJI through intraoperative contamination6-8,28.
For the in vivo PJI model, 25 twelve-week-old C57BL/6 mice underwent unilateral proximal tibial implant insertion through a previously described surgical technique22. Following press-fit implant insertion and arthrotomy closure, a gas-tight syringe (65 RN; Hamilton) was utilized to administer a 2-μL intra-articular injection of bacteria. To produce PJI, 20 mice received an injection of 3 × 105 CFUs, a quantity shown previously to reliably cause osteomyelitis in mice29,30. CFU dosing was verified through parallel syringe injections placed directly in triplicate onto kanamycin-containing agar plates, which were cultured overnight with direct visual quantitation of CFUs31. The remaining 5 animals served as controls and received a 2-μL injection of sterile saline solution. Following skin closure, implant position was verified with use of radiographs (MX-20; Faxitron). Mice were provided with subcutaneous buprenorphine (0.05 mg/kg) for 48 hours postoperatively and not restricted in activity.
Weight-Bearing Activity and Radiographic Changes
Three infected and 3 control mice were randomly selected and placed in an empty cage at 2 and 6 weeks postoperatively. Weight-bearing was assessed using slow-motion video recording software (iPhone 6; Apple) and graded for each mouse as being full (3 points), partial (2 points), toe-touch (1 point), or non-weight-bearing (0 points) by an observer blinded to the mouse group allocation. The weights of all mice were recorded following surgery and weekly.
Lateral radiographs of all operative knees were obtained at 2 and 6 weeks postoperatively. Radiographs were graded using a modified version of an established quantitative radiographic score32. We developed this modified animal PJI scale on the basis of accepted radiographic findings in human PJI33. Points were assigned for periosteal reaction, peri-implant radiolucency, metaphyseal fracture, and implant migration. All radiographs were assessed by 2 blinded observers, and the inter-rater reliability was calculated.
Immune Response, Periprosthetic Bacterial Load, and Implant Biofilm
Blood was collected at 2 and 6 weeks postoperatively from all animals via submandibular bleeding. Blood samples were centrifuged and serum was collected. Serum was analyzed for circulating levels of serum amyloid A (SAA), an acute phase reactant that is found in both mice and humans34 and rapidly upregulates within 1 to 2 days in response to infection34,35. C-reactive protein is not elevated in mice with infection or inflammation35.
Operative legs were disinfected with povidone-iodine, and both the skin and foot were removed. A second scalpel was then used to incise the medial retinaculum and enter the joint space. Soft-tissue and bone damage was graded with use of the Rissing scale, which quantitatively evaluates wound erythema, bone destruction, and purulent exudate36,37.
For 50% of the infected mice, the implant was removed and the proximal aspect of the tibia, the distal aspect of the femur, and knee extensor mechanism were cut into small pieces and placed into sterile, 5-mL centrifuge tubes (Eppendorf) containing 2 mL of 4°C PBS solution. The contents were homogenized for 5 minutes (Bullet Blender; Next Advance) at setting 16, diluted in 4°C TSB, and then plated for CFU assay to quantify the number of bacteria within the periprosthetic tissues. To quantify the number of living bacteria adherent to the implant within biofilm, the removed implants were placed individually in a 1.5-mL Eppendorf tube containing 200 μL of 4°C PBS solution, vortexed for 30 seconds, and sonicated for 15 minutes in a water bath (Biosonic UC125, 52 kHz, 137 watts; Coltène) followed by an additional 30 seconds of vortexing38,39. The resulting TSB suspensions were then diluted, plated in triplicate, and cultured overnight, and CFUs were counted.
For the remaining 50% of the infected mice, the implant was left in place and the proximal 3 cm of the tibia was harvested, fixed in glutaraldehyde plus 4% paraformaldehyde, and decalcified using 10% ethylenediaminetetraacetic acid (EDTA). Implants were carefully removed and their surfaces were examined by a single, experienced observer blinded to treatment, using scanning electron microscopy (SEM) (Zeiss Auriga field emission SEM and Gatan digital camera system). S. aureus was identified as spherical structures with the following features: no surface deformities, organized in pairs or clusters, and approximately 1 μm in diameter40. Biofilm was visually defined as any accumulation of S. aureus covered with an extracellular fibrin mesh or lattice40. Host leukocytes were identified as spherical objects larger than 2 μm that were in proximity to bacteria, not covered by any extracellular material, and not appearing adherent to the implant surface.
All continuous variables were assessed for normality using the Shapiro-Wilk test. Parametric pairwise comparisons were performed using the Student t test when appropriate. Parametric group comparisons were performed using a 1-way analysis of variance with Tukey post hoc comparisons. Inter-rater reliability was determined through calculation of the Cohen kappa statistic, with interpretation according to guidelines established by Landis and Koch41. Potential correlational relationships between variables were identified through calculation of the Spearman rank correlation coefficient. For all comparisons, a p value of <0.05 was considered significant.
In vitro experiments revealed that bacterial biofilm could be detected on the surface of mouse implants as early as 3 hours post-incubation with 106 CFUs of S. aureus Xen 36. At 24 hours, all implants inoculated with Xen 36 showed significant (p = 0.006) biofilm formation compared with that observed for controls (Fig. 1).
All mice survived surgery and the postoperative period with no signs of systemic illness. Infected mice had significantly elevated serum SAA levels at both 2 and 6 weeks postoperatively compared with controls (p < 0.01) (Fig. 2). Weight did not differ between the 2 groups. After 2 days, all animals could partially bear weight on the operative leg. Blinded assessment at both 2 and 6 weeks revealed plantigrade, full weight-bearing gait in the control animals (a score of 3 of 3). Conversely, infected mice were partially weight-bearing at 2 weeks (mean score [and standard deviation] of 1.6 ± 0.57 of a maximum score of 3) and were toe-touch weight-bearing at 6 weeks (0.6 ± 0.57) (Video 1).
An analysis of radiographs showed clear differences between the infected and control mice. Postoperatively, all implants were appropriately seated within the tibial canal. At 2 and 6 weeks, the infected mice exhibited tibial periosteal reaction, peri-implant radiolucency, retrograde implant migration, and fracture of the anterior tibial metaphysis (Fig. 3). This result was in contrast with that of the control animals, which exhibited a stable, osseointegrated implant (Fig. 3). Radiographic scores were significantly higher for the infected mice at 2 weeks (3.1 ± 0.77 compared with 0.2 ± 0.45, respectively; p < 0.001), 4 weeks (3.6 ± 0.51 compared with 0.8 ± 1.79, respectively; p < 0.001), and 6 weeks (3.1 ± 0.77 compared with 0.2 ± 0.45, respectively; p < 0.001). A nonsignificant decline in radiographic scores from the 4 to 6-week interval among the infected animals was due to bone bridging and healing of the tibial metaphyseal fracture (Fig. 3-D). The assessment of interobserver reliability revealed substantial agreement for peri-implant radiolucency and implant migration (κ = 0.615 and 0.802, respectively) and moderate agreement for periosteal reaction and metaphyseal fracture (κ = 0.601 and 0.594, respectively).
Dissection of the knees of infected animals euthanized at 2 weeks exhibited distorted soft-tissue planes, purulent intra-articular material, eroded peri-implant bone, and a loose implant (mean modified Rissing scale score of 2.83 ± 0.38 of a maximum score of 4.0) (Fig. 4). Among the animals euthanized at 6 weeks, dissection showed widespread bone destruction and a loose implant (modified Rissing scale score of 2.64 ± 0.67). Conversely, the controls exhibited an osseointegrated implant and significantly lower mean modified Rissing score (0.6 ± 1.34; p = 0.004).
S. aureus Xen 36 were successfully retrieved from the periprosthetic space in the infected animals. At 2 weeks, bacterial counts in periprosthetic tissues had increased tenfold (mean of 4.46 ± 3.5 × 106 CFUs). By 6 weeks, bacterial counts were similar to those of the initial inoculations (2.53 ± 3.4 × 105 CFUs). Sonication of implants identified viable, adhesive S. aureus Xen 36 on the implant surface at 2 weeks (1.13 ± 0.9 × 103 CFUs) and 6 weeks (6.93 ± 2.7 × 102 CFUs). The control animals did not yield any positive S. aureus Xen 36 cultures from within the periprosthetic tissues or the implant surface.
Scores and bacterial counts were assessed for significant correlational relationships. For mouse ambulation, a significantly negative correlation was identified with respect to radiographic scores at 6 weeks (r = −0.973; p = 0.005) and soft-tissue/bone damage scores (r = −0.913; p = 0.030). Furthermore, radiographic scores at 2 and 6 weeks were significantly positively correlated with soft-tissue/bone damage scores (r = 0.917, p = 0.04 and r = 0.918, p = 0.028, respectively). Periprosthetic bacterial counts had positive but not significant correlations to 6-week radiographic scores (r = 0.315; p = 0.543).
SEM examination of implants from infected animals revealed the presence of S. aureus clusters within a biofilm (Fig. 5). At 2 weeks, bacterial biofilm was found adjacent to large titanium beads (Fig. 5-C) with abundant linear fibrin fibers connecting bacteria (Fig. 5-D). By 6 weeks, fibrinous tissue was adherent to implants (Fig. 5-E), with multiple specimens showing S. aureus biofilm both on the surface of the implant and densely packed within the walls of peri-implant tissue (Fig. 5-F). In 1 specimen, large numbers of leukocytes were observed on the implant surface tissue (Fig. 5-G) with remaining S. aureus being surrounded by leukocytes (Fig. 5-H).
This novel animal model offers several important advances in the translational representation of clinical PJI. As seen clinically, infected mice initially walked similarly to the controls but then had difficulty loading the leg; developed radiographic features of septic loosening29; had elevated acute inflammatory markers; and produced the intra-articular purulence, bone loss, and implant loosening that we see intraoperatively. These findings as well as positive bacterial cultures from both the implant and periprosthetic material are consistent with definitions of PJI from the Musculoskeletal Infection Society42 and the international consensus on PJI43. In addition to its translational appeal, this model presents a total of 8 unique quantitative measures (Fig. 6), making it, to our knowledge, the most comprehensive assessment of PJI to date compared with previous models19,20,21,44,45. Furthermore, we have demonstrated that such measures are interrelated, specifically that leg function is significantly correlated with radiographic signs of PJI and visual signs of tissue destruction. Although a specific relationship between function and bacterial counts was not established, future investigations with larger animal sizes and varying bacterial doses will assist in determining whether such a relationship exists.
This model also provides the ability to downsize a representative PJI model to the mouse. This development permits future investigations in genetically modified mice modeling pertinent clinical conditions that contribute to PJI risk, including obesity (metabolic syndrome), diabetes, and inflammatory arthritis.
The distinguishing feature of our model is the effective recreation of the periprosthetic environment, which can be described as including an articular space that is hypovascular and immune-privileged46,47, a metal arthroplasty implant capable of bearing weight and with a surface that can osseointegrate, and an intramedullary space that is hypercellular and vascularized. Since the implant is the only factor than can be modified by the investigator, utilizing one that bears load and can achieve initial stability is critical. Implant loading and stability have direct effects on both biofilm formation and immunological activity9,48. Unstable arthroplasty implants do not osseointegrate and instead promote fibroblast activity, chronic neutrophil and lymphocyte inflammation, and osteoclast-mediated bone resorption48,49. Furthermore, unstable implants create shear forces that interfere with bacterial adhesion and biofilm density and strength50. Having an implant that bears load is also important, as serial loading affects bacterial biofilm density and strength51.
Previous PJI models ultimately have limited translational appeal. In mice, investigators used a smooth, stainless steel wire inserted retrograde into the femur19,20,44,45,52. Such an implant does not bear weight, is unstable, and is not representative of any clinical device. Alternatively, rabbits have been fitted with an extra-articular, transcortical screw21. While easy to reproduce, this approach does not access the medullary canal and uses an implant that does not encounter any tensile or rotational loads. A slightly more appealing manifestation in the rabbit utilizes a stemmed, silicone-elastomer implant in the toe. While appearing similar to a human tibial implant, silicone-elastomer bears no similarity to the metal alloys and cross-linked polyethylene inserts used clinically21. Given these limitations, it is clear how challenging it is to recreate the periprosthetic environment in an animal. In contrast, the approach outlined in the current study meets all of the major criteria for a suitable translational model.
Our approach does have limitations. The dose of bacteria utilized, while similar to that of other animal PJI models, is much higher than the exposure occurring clinically (<100 CFUs per m3)28. We acknowledge this shortcoming but argue that the dose utilized in this study is similar to that of other models in the literature19,20,40 and is necessary to convincingly produce features of PJI. A second limitation is that SEM could not be used in our model to quantify biofilm coverage because the implant has a roughened surface. Furthermore, SEM could not be used to visualize bacteria under the fibrous surface, making it inferior to implant sonication in determining the number of adherent bacteria. Finally, we did not compare pullout strength between the infected and control groups, as such testing would disturb the implant surface bacteria and interfere with bacterial counting. This variable will be assessed in additional studies to determine what relationship exists between bacterial load, the interruption of osseointegration, and subsequent loss of implant stability.
In summary, the present model of PJI is, to our knowledge, the most translationally relevant model to date and has the potential to improve our fundamental knowledge of bacterial infection around a weight-bearing implant. The surgical technique is reproducible, and the described quantitative assays are comprehensive yet not costly and can be readily performed in any reasonably equipped laboratory.
Investigation performed at the Hospital for Special Surgery, New York, NY
Disclosure: No external funding was received for this study. The Disclosure of Potential Conflicts of Interest forms are provided with the online version of the article (http://links.lww.com/JBJS/A154).
1. Jämsen E, Huhtala H, Puolakka T, Moilanen T. Risk factors for infection after knee arthroplasty. A register-based analysis of 43,149 cases. J Bone Joint Surg Am. 2009 ;91(1):38–47.
2. Phillips JE, Crane TP, Noy M, Elliott TS, Grimer RJ. The incidence of deep prosthetic infections in a specialist orthopaedic hospital: a 15-year prospective survey. J Bone Joint Surg Br. 2006 ;88(7):943–8.
3. Bozic KJ, Kurtz SM, Lau E, Ong K, Vail TP, Berry DJ. The epidemiology of revision total hip arthroplasty in the United States. J Bone Joint Surg Am. 2009 ;91(1):128–33.
4. Bozic KJ, Kurtz SM, Lau E, Ong K, Chiu V, Vail TP, Rubash HE, Berry DJ. The epidemiology of revision total knee arthroplasty in the United States. Clin Orthop Relat Res. 2010 ;468(1):45–51. Epub 2009 Jun 25.
5. Kurtz SM, Lau E, Watson H, Schmier JK, Parvizi J. Economic burden of periprosthetic joint infection in the United States. J Arthroplasty. 2012 ;27(8)(Suppl):61-5.e1. Epub 2012 May 2.
6. Peersman G, Laskin R, Davis J, Peterson MGE, Richart T. Prolonged operative time correlates with increased infection rate after total knee arthroplasty. HSS J. 2006 ;2(1):70–2.
7. Panahi P, Stroh M, Casper DS, Parvizi J, Austin MS. Operating room traffic is a major concern during total joint arthroplasty. Clin Orthop Relat Res. 2012 ;470(10):2690–4.
8. Waked WR, Simpson AK, Miller CP, Magit DP, Grauer JN. Sterilization wrap inspections do not adequately evaluate instrument sterility. Clin Orthop Relat Res. 2007 ;462:207–11.
9. Murdoch DR, Roberts SA, Fowler Jr VG Jr, Shah MA, Taylor SL, Morris AJ, Corey GR. Infection of orthopedic prostheses after Staphylococcus aureus bacteremia. Clin Infect Dis. 2001 ;32(4):647–9. Epub 2001 Feb 7.
10. Brady RA, Leid JG, Calhoun JH, Costerton JW, Shirtliff ME. Osteomyelitis and the role of biofilms in chronic infection. FEMS Immunol Med Microbiol. 2008 ;52(1):13–22. Epub 2007 Dec 11.
11. Black CE, Costerton JW. Current concepts regarding the effect of wound microbial ecology and biofilms on wound healing. Surg Clin North Am. 2010 ;90(6):1147–60.
12. Deirmengian C, Greenbaum J, Lotke PA, Booth RE Jr, Lonner JH. Limited success with open debridement and retention of components in the treatment of acute Staphylococcus aureus infections after total knee arthroplasty. J Arthroplasty. 2003 ;18(7)(Suppl 1):22–6.
13. Tsukayama DT, Estrada R, Gustilo RB. Infection after total hip arthroplasty. A study of the treatment of one hundred and six infections. J Bone Joint Surg Am. 1996 ;78(4):512–23.
14. Hanssen AD, Rand JA. Evaluation and treatment of infection at the site of a total hip or knee arthroplasty. Instr Course Lect. 1999;48:111–22.
15. Crockarell JR, Hanssen AD, Osmon DR, Morrey BF. Treatment of infection with débridement and retention of the components following hip arthroplasty. J Bone Joint Surg Am. 1998 ;80(9):1306–13.
16. Azzam K, McHale K, Austin M, Purtill JJ, Parvizi J. Outcome of a second two-stage reimplantation for periprosthetic knee infection. Clin Orthop Relat Res. 2009 ;467(7):1706–14. Epub 2009 Feb 18.
17. Haleem AA, Berry DJ, Hanssen AD. Mid-term to long-term followup of two-stage reimplantation for infected total knee arthroplasty. Clin Orthop Relat Res. 2004 ;428:35–9.
18. Fehring TK, Calton TF, Griffin WL. Cementless fixation in 2-stage reimplantation for periprosthetic sepsis. J Arthroplasty. 1999 ;14(2):175–81.
19. Pribaz JR, Bernthal NM, Billi F, Cho JS, Ramos RI, Guo Y, Cheung AL, Francis KP, Miller LS. Mouse model of chronic post-arthroplasty infection: noninvasive in vivo bioluminescence imaging to monitor bacterial burden for long-term study. J Orthop Res. 2012 ;30(3):335–40. Epub 2011 Aug 11.
20. Bernthal NM, Stavrakis AI, Billi F, Cho JS, Kremen TJ, Simon SI, Cheung AL, Finerman GA, Lieberman JR, Adams JS, Miller LS. A mouse model of post-arthroplasty Staphylococcus aureus joint infection to evaluate in vivo the efficacy of antimicrobial implant coatings. PLoS One. 2010 ;5(9):e12580.
21. Craig MR, Poelstra KA, Sherrell JC, Kwon MS, Belzile EL, Brown TE. A novel total knee arthroplasty infection model in rabbits. J Orthop Res. 2005 ;23(5):1100–4.
22. Yang X, Ricciardi BF, Dvorzhinskiy A, Brial C, Lane Z, Bhimani S, Burket JC, Hu B, Sarkisian AM, Ross FP, van der Meulen MC, Bostrom MP. Intermittent parathyroid hormone enhances cancellous osseointegration of a novel murine tibial implant. J Bone Joint Surg Am. 2015 ;97(13):1074–83.
23. ASTM International. ASTM F86-04, standard practice for surface preparation and marking of metallic surgical implants. ASTM International, West Conshohocken, PA, 2005.
24. Koseki H, Yonekura A, Shida T, Yoda I, Horiuchi H, Morinaga Y, Yanagihara K, Sakoda H, Osaki M, Tomita M. Early staphylococcal biofilm formation on solid orthopaedic implant materials: in vitro study. PLoS One. 2014 ;9(10):e107588.
25. Bauer J, Siala W, Tulkens PM, Van Bambeke F. A combined pharmacodynamic quantitative and qualitative model reveals the potent activity of daptomycin and delafloxacin against Staphylococcus aureus biofilms. Antimicrob Agents Chemother. 2013 ;57(6):2726–37. Epub 2013 Apr 9.
26. O’Toole GA. Microtiter dish biofilm formation assay. J Vis Exp. 2011 ;47:47.
27. Yang X, Craveiro V, Carli A, Shirley MB, Adebayo O, van der Meulen MCH, Ross FP, Bostrom MPG. Refining Intra-articular and intramedullary injection techniques to establish joint, bone and periprosthetic joint infection in mice. Presented as a poster exhibit at the Orthopaedic Research Society 2016 Annual Meeting; 2016 Mar 5-8; Orlando, FL.
28. Stocks GW, O’Connor DP, Self SD, Marcek GA, Thompson BL. Directed air flow to reduce airborne particulate and bacterial contamination in the surgical field during total hip arthroplasty. J Arthroplasty. 2011 ; 25(5):771–6. Epub 2010 Sep 18.
29. Li D, Gromov K, Søballe K, Puzas JE, O’Keefe RJ, Awad H, Drissi H, Schwarz EM. Quantitative mouse model of implant-associated osteomyelitis and the kinetics of microbial growth, osteolysis, and humoral immunity. J Orthop Res. 2008 ;26(1):96–105.
30. Varoga D, Tohidnezhad M, Paulsen F, Wruck CJ, Brandenburg L, Mentlein R, Lippross S, Hassenpflug J, Besch L, Müller M, Jürgens C, Seekamp A, Schmitt L, Pufe T. The role of human beta-defensin-2 in bone. J Anat. 2008 ;213(6):749–57.
31. Coenye T, Nelis HJ. In vitro and in vivo model systems to study microbial biofilm formation. J Microbiol Methods. 2010 ;83(2):89–105. Epub 2010 Sep 8.
32. An YH, Friedman RJ. Animal models of orthopedic implant infection. J Invest Surg. 1998 ;11(2):139–46.
33. Bauer TW, Parvizi J, Kobayashi N, Krebs V. Diagnosis of periprosthetic infection. J Bone Joint Surg Am. 2006 ;88(4):869–82.
34. Cray C, Zaias J, Altman NH. Acute phase response in animals: a review. Comp Med. 2009 ;59(6):517–26.
35. Pepys MB, Hirschfield GM. C-reactive protein: a critical update. J Clin Invest. 2003 ;111(12):1805–12.
36. Rissing JP, Buxton TB, Weinstein RS, Shockley RK. Model of experimental chronic osteomyelitis in rats. Infect Immun. 1985 ;47(3):581–6.
37. Shandley S, Matthews KP, Cox J, Romano D, Abplanalp A, Kalns J. Hyperbaric oxygen therapy in a mouse model of implant-associated osteomyelitis. J Orthop Res. 2012 ;30(2):203–8. Epub 2011 Aug 3.
38. Trampuz A, Piper KE, Jacobson MJ, Hanssen AD, Unni KK, Osmon DR, Mandrekar JN, Cockerill FR, Steckelberg JM, Greenleaf JF, Patel R. Sonication of removed hip and knee prostheses for diagnosis of infection. N Engl J Med. 2007 ;357(7):654–63.
39. Portillo ME, Salvadó M, Trampuz A, Plasencia V, Rodriguez-Villasante M, Sorli L, Puig L, Horcajada JP. Sonication versus vortexing of implants for diagnosis of prosthetic joint infection. J Clin Microbiol. 2013 ;51(2):591–4. Epub 2012 Nov 7.
40. Nishitani K, Sutipornpalangkul W, de Mesy Bentley KL, Varrone JJ, Bello-Irizarry SN, Ito H, Matsuda S, Kates SL, Daiss JL, Schwarz EM. Quantifying the natural history of biofilm formation in vivo during the establishment of chronic implant-associated Staphylococcus aureus osteomyelitis in mice to identify critical pathogen and host factors. J Orthop Res. 2015 ;33(9):1311–9. Epub 2015 May 18.
41. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977 ;33(1):159–74.
42. Parvizi J, Zmistowski B, Berbari EF, Bauer TW, Springer BD, Della Valle CJ, Garvin KL, Mont MA, Wongworawat MD, Zalavras CG. New definition for periprosthetic joint infection: from the Workgroup of the Musculoskeletal Infection Society. Clin Orthop Relat Res. 2011 ;469(11):2992–4.
44. Poultsides LA, Papatheodorou LK, Karachalios TS, Khaldi L, Maniatis A, Petinaki E, Malizos KN. Novel model for studying hematogenous infection in an experimental setting of implant-related infection by a community-acquired methicillin-resistant S. aureus strain. J Orthop Res. 2008 ;26(10):1355–62.
45. Niska JA, Meganck JA, Pribaz JR, Shahbazian JH, Lim E, Zhang N, Rice BW, Akin A, Ramos RI, Bernthal NM, Francis KP, Miller LS. Monitoring bacterial burden, inflammation and bone damage longitudinally using optical and μCT imaging in an orthopaedic implant infection in mice. PLoS One. 2012;7(10):e47397. Epub 2012 Oct 17.
46. Moser SA, Gilbert SR. Osteomyelitis. In: Pathobiology of human disease. McManus LM, Mitchell RN, editors. San Diego: Academic Press; 2014. p 799–814.
47. Revell CM, Athanasiou KA. Success rates and immunologic responses of autogenic, allogenic, and xenogenic treatments to repair articular cartilage defects. Tissue Eng Part B Rev. ;15(1):1–15.
48. Abu-Amer Y, Darwech I, Clohisy JC. Aseptic loosening of total joint replacements: mechanisms underlying osteolysis and potential therapies. Arthritis Res Ther. 2007;9(Suppl 1):S6.
49. Athanasou NA, Quinn J, Bulstrode CJ. Resorption of bone by inflammatory cells derived from the joint capsule of hip arthroplasties. J Bone Joint Surg Br. 1992 ;74(1):57–62.
50. Liu Y, Tay JH. Metabolic response of biofilm to shear stress in fixed-film culture. J Appl Microbiol. 2001 ;90(3):337–42.
51. Wijeyekoon S, Mino T, Satoh H, Matsuo T. Effects of substrate loading rate on biofilm structure. Water Res. 2004 ;38(10):2479–88.
52. Lovati AB, Drago L, Monti L, De Vecchi E, Previdi S, Banfi G, Romanò CL. Diabetic mouse model of orthopaedic implant-related Staphylococcus aureus infection. PLoS One. 2013 ;8(6):e67628.