**5.3 Exchange arthroplasty**

424 Recent Advances in Arthroplasty

Immunofluorescence microscopy (IFM) is another novel nonculture technique for diagnosing prosthetic joint infections. In immunofluorescence microscopy, samples are mixed with monoclonal antibodies (MAb) to specific antigens on bacterial cell walls. Samples are then incubated with a second antibody conjugated with a fluorescent dye. The bacteria are then visualised using fluorescence microscopy(Tunney, et al. 1999). As with FISH, IFM can be used to assess the biofilm structure and can detect multiple pathogens (McDowell & Patrick 2005, Tunney, et al. 1999). In the study by Tunney et al IFM was performed on the sonicate fluid from explanted prostheses using Mab for both

Optimal treatment of prosthetic joint infections involves the eradication of infection whilst maintaining function of the joint and patient quality of life (Zimmerli, et al. 2004). However there are no large, multi-centred, randomised prospective studies of treatment strategies to guide recommendations. The successful treatment of prosthetic joints is contingent on the elimination of the biofilm dwelling microorganism. The two mainstay methods of achieving this are through either surgical removal of the prosthesis or through use of biofilm active

The surgical strategies used to treat arthroplasty infections include: resection arthroplasty, one-stage or two-stage exchange procedures, amputation and debridement and retention. Resection arthroplasty entails the removal of all foreign material including cement, resection of devitalised tissue and bone and may or may not involve arthrodesis. Exchange procedures involve resection arthroplasty with reimplantation of a new joint prosthesis performed at the time of removal of the infected prosthesis (one-stage exchange); or delayed by a variable period of time while antibiotic therapy is administered (two-stage exchange). Debridement and retention of the prosthesis usually involves open arthrotomy, removal of all infected and necrotic bone, exchange of liners and lavage of the joint (Giulieri, et al. 2004, Matthews, et al. 2009, Rand, et al. 1986, Steckelberg & Osmon 2000, Trampuz & Zimmerli

A number of factors influence the surgical approach selected for an individual patient, these include a patient's general health and fitness for anaesthesia, condition of the prosthesis and bone stock, the causative agent, the timing of the infection relative to the prosthesis insertion, the availability of effective antibiotics and clinicians' and patient preference.

Administration of antibiotic therapy without surgical management is not routinely recommended, as it is rarely associated with successful cure. Early studies of antibiotic therapy alone for prosthetic joint infections had disappointing results with successful outcomes in as little as 8-15% of patients (Bengtson, et al. 1989, Canner, et al. 1984). The confounding factor when analysing these poor results is that biofilm active antibiotics, were not used. Treatment with biofilm active antibiotics alone including rifampicin and ciprofloxacin for three to six months has yielded successful outcomes in highly selected patients; those presenting with early infections (less than one year following implant),

**5.2 Systemic antibiotic therapy without surgical debridement** 

antibiotics in conjunction with surgical debridement and retention of the prosthesis.

Propionibacterium and Staphylococcus species. (Tunney, et al. 1999).

**5. Treatment** 

**5.1 Treatment goal** 

2008, Zimmerli, et al. 2004).

Interpretation of the current literature describing the outcomes of patients having one- and two-stage exchange procedures is challenging owing to the heterogeneity of the patient populations, the causative organisms and the surgical techniques (including use of antibiotic impregnated cement), the differences in the duration of patient follow up and probable publication bias. The greatest concern with one-stage exchange procedures is the implantation of the prosthesis into an infected field with subsequent reinfection of the revised arthroplasty. In one-stage exchange, reported success rates range from 38-100%; but there is significant variability in the definition of success which includes freedom from infection, freedom from pain or simply the presence of a functional joint (Callaghan, et al. 1999, Jamsen, et al. 2009, Steckelberg & Osman 2000). In examination of the outcomes of onestage exchange revision hip arthroplasty, 80% (range 57-92%) of patients have been reported to remain infection free after one-stage exchange without the use of antibiotic cement (Steckelberg & Osman 2000). When antibiotic impregnated cement was used, 88% (range 76- 100%) of patients have been reported to remain infection free at follow up (Callaghan, et al. 1999, Jackson & Schmalzried 2000, Langlais 2003, Steckelberg & Osman 2000). Results for one-stage exchange in knee arthroplasty revision are in general worse than for hips, with only 65% (range 57-100%) of patients remaining free of recurrence of infection at follow up (Steckelberg & Osmon 2000). On the basis of these results, one-stage exchange of an infected prosthesis is rarely advised for prosthetic knee infections(Trampuz & Zimmerli 2008, Zimmerli, et al. 2004). There are, however some advantages with one-stage exchange; patients undergo a single operation and generally require a shorter period of hospitalisation in total.

Consensus recommendations for one-stage exchange suggest that it should only be considered where there is minimal soft tissue damage and where less virulent organisms are involved (Hirakawa, et al. 1998, Jackson & Schmalzried 2000, Miley, et al. 1982, Trampuz & Zimmerli 2008, Zimmerli, et al. 2004). The presence of sinus tract is considered a relative contraindication for one-stage exchange. Ideally the causative agent should be known prior to resection arthroplasty and treatment commenced preoperatively(Zimmerli, et al. 2004).

In two-stage exchange procedures, reimplantation is delayed for a variable length of time from 2 weeks to several months. Spacers impregnated with antibiotic are commonly inserted to maintain limb length and improve patient mobility during that interval (Leunig, et al. 1998). Antibiotics with activity against the isolated pathogen are administered for at least 6 weeks. Tissue samples are often routinely taken from the periprosthetic tissue at the time of reimplantation for microbiological culture to assess the efficacy of the interim treatment (Insall, et al. 1983, Wilson, et al. 1990, Windsor, et al. 1990). In infections with

Infection in Primary Hip and Knee Arthroplasty 427

2006a, Crockarell, et al. 1998, Marculescu, et al. 2006b). High failure rates have also observed with specific organisms, only 12% of patients with *Staphylococcus aureus* infection experiencing successful treatment outcomes (Deirmengian, et al. 2003, Marculescu, et al. 2006a). Gram-negative infections, polymicrobial infections and culture negative prosthetic joint infections were also associated with higher rates of recurrence of infection following debridement and retention (Berbari, et al. 2007, Hsieh, et al. 2009a, Marculescu & Cantey 2008). However, successful outcomes have been reported in the setting of early or haematogenous infections (Morrey, et al. 1989, Wasielewski, et al. 1996). Tsuyakama et al reported a successful outcome in 71% of patients with early infections treated with debridement of the prosthetic joint followed by four weeks of parenteral antibiotics with an

Newer treatment strategies evolved as understanding of the role of biofilm in the pathogenesis of prosthetic joint infections increased. In vivo studies using guinea pig tissue-cage animal model by Widmer et al demonstrated the clinical utility of rifampicin in chronic biofilm infections (Widmer, et al. 1990). Further studies also identified that quinolones retained activity in the presence of biofilms (Schwank, et al. 1998, Widmer, et

The addition of rifampicin to antimicrobial regimens has led to a significant improvement in success rates reported in the treatment of gram-positive prosthetic joint infection; in many instances comparable to that reported for two-stage exchange. In our experience combination treatment including rifampicin has resulted in successful treatment in up to 90% of patients (Aboltins, et al. 2007); however successful outcomes are associated with several factors. Higher success rates are reported where the causative organism is a staphylococcus species and when antibiotic therapy is continued for a protracted period; 12 months or greater, (Choong, et al. 2007, Widmer, et al. 1992). In contrast the rate of success is significantly diminished when a fistula is present, particularly in knee arthroplasty where successful outcomes have been reported in only 45-69% of patients.

The only randomised double-blinded control trial examining the role of rifampicin in the treatment of prosthetic device staphylococcal infections was conducted from 1992 through 1997. The study involved 33 patients with orthopaedic device infections and duration of symptoms less than one year. Patients were randomised to receive rifampicin 450mg and ciprofloxacin 750mg (twice daily) or ciprofloxacin and placebo. Rifampicin/ciprofloxacin combination was successful in all patients compared to 58% of patients who received ciprofloxacin alone (Zimmerli, et al. 1998). Subsequent studies corroborated these results with success rates of greater than 85% of patients treated with debridement and retention and rifampicin containing antibiotic treatment (Berdal, et al. 2005, Byren, et al. 2009, Rao, et al. 2003). The main limitations with the use of rifampicin are the high likelihood of generation of resistance when used without a second antibiotic and the hepatic and gastrointestinal toxicities (John, et al. 2009, Widmer, et al. 1990). Therefore careful, regular follow up of patients is necessary and the management these antibiotics should involve

collaboration between Infectious Diseases Physicians and Orthopaedic Surgeons.

The investigation of newer agents for the treatment of prosthetic joint infection is ongoing. In guinea pig foreign-body infection model, John et al assessed the activity of newer agents including linezolid and daptomycin, alone and in combination with rifampicin. In this study

average follow up of 3.8 years (Tsukayama, et al. 1996).

al. 1991).

(Drancourt, et al. 1993).

difficult-to-treat micro-organisms such as MRSA, resistant enterococci and fungi, current consensus guidelines recommends prolonged interval between removal and reimplantation without the use of a spacer (Trampuz & Zimmerli 2008, Zimmerli, et al. 2004)

Two-stage exchange, in general, has a higher success rate compared to one-stage with rates of 63-100% (Colyer & Capello 1994, Haleem, et al. 2004, Jamsen, et al. 2009, Woods, et al. 1983). In hip arthroplasty two-stage exchange without the use of antibiotic impregnated cement, 81% of patients (range 53-100%) remain free of recurrent infection increasing to 93% (73-100%) when antibiotic cement is used (Laffer, et al. 2006, Langlais, et al. 2006, Steckelberg & Osman 2000). In knee arthroplasty infections, 84% (38-100%) of patients in whom antibiotic cement is not used, and 88% (63-100%) of patients in whom antibiotic impregnated cement is used remain infection free following two stage exchange (Bengtson, et al. 1989, Grogan, et al. 1986, Hanssen, et al. 1994, Insall, et al. 1983, Morrey, et al. 1989, Rand, et al. 1986, Wang & Chen 1997, Wasielewski, et al. 1996, Wilson, et al. 1990, Windsor, et al. 1990, Woods, et al. 1983).

A number of factors potentially influence treatment outcomes in two-stage exchange procedures. Polymicrobial infection, infection with virulent organisms including *Staphylococcus aureus* and methicillin resistant coagulase negative Staphylococcus species, the presence of rheumatoid arthritis and a history of prior multiple revisions have all been shown to associated with lower rates of success in two-stage exchanges (Hirakawa, et al. 1998, Lim, et al. 2009, Mittal, et al. 2007). Current consensus guidelines recommend twostage exchange in chronic infections with moderately or severely damaged tissue or if a sinus tract is present (Trampuz & Zimmerli 2008, Zimmerli, et al. 2004).

### **5.4 Resection arthroplasty**

Resection arthroplasty and amputation are generally reserved for patients with refractory infections particularly where there is severe loss of bone stock or where functional improvement following revision is unlikely (Trampuz & Zimmerli 2008, Zimmerli, et al. 2004). Whilst rates of recurrence of infection are low, patients have worse functional outcomes and up to 80% of patients report residual pain following resection (Morrey, et al. 1989).

#### **5.5 Debridement and retention of the prosthesis**

Debridement and retention of the prosthesis is an attractive treatment option for many patients given that it is the least invasive, with a lower surgical morbidity, and is generally associated with good functional outcomes (Trampuz & Zimmerli 2008, Zimmerli, et al. 2004). Open arthrotomy and debridement is recommended when attempting retention of the prosthesis as poorer results are reported with arthroscopic 'washout' compared to open 'washout'(Laffer, et al. 2006). Following debridement, patients should receive biofilm active antibiotics generally for a longer duration than with surgical exchange or resection.

Early studies of debridement and prosthesis retention strategies to treat prosthetic joint infection were disappointing with recurrence of infection at 2 years reported in 69% of patients (Brandt, et al. 1997). Poor outcomes have been reported when symptoms are present greater than 8 days, when a sinus tract is present and with late chronic infections; In some instances all patients with late infection experienced treatment failure (Berbari, et al.

difficult-to-treat micro-organisms such as MRSA, resistant enterococci and fungi, current consensus guidelines recommends prolonged interval between removal and reimplantation

Two-stage exchange, in general, has a higher success rate compared to one-stage with rates of 63-100% (Colyer & Capello 1994, Haleem, et al. 2004, Jamsen, et al. 2009, Woods, et al. 1983). In hip arthroplasty two-stage exchange without the use of antibiotic impregnated cement, 81% of patients (range 53-100%) remain free of recurrent infection increasing to 93% (73-100%) when antibiotic cement is used (Laffer, et al. 2006, Langlais, et al. 2006, Steckelberg & Osman 2000). In knee arthroplasty infections, 84% (38-100%) of patients in whom antibiotic cement is not used, and 88% (63-100%) of patients in whom antibiotic impregnated cement is used remain infection free following two stage exchange (Bengtson, et al. 1989, Grogan, et al. 1986, Hanssen, et al. 1994, Insall, et al. 1983, Morrey, et al. 1989, Rand, et al. 1986, Wang & Chen 1997, Wasielewski, et al. 1996, Wilson, et al. 1990, Windsor,

A number of factors potentially influence treatment outcomes in two-stage exchange procedures. Polymicrobial infection, infection with virulent organisms including *Staphylococcus aureus* and methicillin resistant coagulase negative Staphylococcus species, the presence of rheumatoid arthritis and a history of prior multiple revisions have all been shown to associated with lower rates of success in two-stage exchanges (Hirakawa, et al. 1998, Lim, et al. 2009, Mittal, et al. 2007). Current consensus guidelines recommend twostage exchange in chronic infections with moderately or severely damaged tissue or if a

Resection arthroplasty and amputation are generally reserved for patients with refractory infections particularly where there is severe loss of bone stock or where functional improvement following revision is unlikely (Trampuz & Zimmerli 2008, Zimmerli, et al. 2004). Whilst rates of recurrence of infection are low, patients have worse functional outcomes and up to 80% of patients report residual pain following resection (Morrey, et al.

Debridement and retention of the prosthesis is an attractive treatment option for many patients given that it is the least invasive, with a lower surgical morbidity, and is generally associated with good functional outcomes (Trampuz & Zimmerli 2008, Zimmerli, et al. 2004). Open arthrotomy and debridement is recommended when attempting retention of the prosthesis as poorer results are reported with arthroscopic 'washout' compared to open 'washout'(Laffer, et al. 2006). Following debridement, patients should receive biofilm active antibiotics generally for a longer duration than with

Early studies of debridement and prosthesis retention strategies to treat prosthetic joint infection were disappointing with recurrence of infection at 2 years reported in 69% of patients (Brandt, et al. 1997). Poor outcomes have been reported when symptoms are present greater than 8 days, when a sinus tract is present and with late chronic infections; In some instances all patients with late infection experienced treatment failure (Berbari, et al.

without the use of a spacer (Trampuz & Zimmerli 2008, Zimmerli, et al. 2004)

sinus tract is present (Trampuz & Zimmerli 2008, Zimmerli, et al. 2004).

**5.5 Debridement and retention of the prosthesis** 

et al. 1990, Woods, et al. 1983).

**5.4 Resection arthroplasty** 

surgical exchange or resection.

1989).

2006a, Crockarell, et al. 1998, Marculescu, et al. 2006b). High failure rates have also observed with specific organisms, only 12% of patients with *Staphylococcus aureus* infection experiencing successful treatment outcomes (Deirmengian, et al. 2003, Marculescu, et al. 2006a). Gram-negative infections, polymicrobial infections and culture negative prosthetic joint infections were also associated with higher rates of recurrence of infection following debridement and retention (Berbari, et al. 2007, Hsieh, et al. 2009a, Marculescu & Cantey 2008). However, successful outcomes have been reported in the setting of early or haematogenous infections (Morrey, et al. 1989, Wasielewski, et al. 1996). Tsuyakama et al reported a successful outcome in 71% of patients with early infections treated with debridement of the prosthetic joint followed by four weeks of parenteral antibiotics with an average follow up of 3.8 years (Tsukayama, et al. 1996).

Newer treatment strategies evolved as understanding of the role of biofilm in the pathogenesis of prosthetic joint infections increased. In vivo studies using guinea pig tissue-cage animal model by Widmer et al demonstrated the clinical utility of rifampicin in chronic biofilm infections (Widmer, et al. 1990). Further studies also identified that quinolones retained activity in the presence of biofilms (Schwank, et al. 1998, Widmer, et al. 1991).

The addition of rifampicin to antimicrobial regimens has led to a significant improvement in success rates reported in the treatment of gram-positive prosthetic joint infection; in many instances comparable to that reported for two-stage exchange. In our experience combination treatment including rifampicin has resulted in successful treatment in up to 90% of patients (Aboltins, et al. 2007); however successful outcomes are associated with several factors. Higher success rates are reported where the causative organism is a staphylococcus species and when antibiotic therapy is continued for a protracted period; 12 months or greater, (Choong, et al. 2007, Widmer, et al. 1992). In contrast the rate of success is significantly diminished when a fistula is present, particularly in knee arthroplasty where successful outcomes have been reported in only 45-69% of patients. (Drancourt, et al. 1993).

The only randomised double-blinded control trial examining the role of rifampicin in the treatment of prosthetic device staphylococcal infections was conducted from 1992 through 1997. The study involved 33 patients with orthopaedic device infections and duration of symptoms less than one year. Patients were randomised to receive rifampicin 450mg and ciprofloxacin 750mg (twice daily) or ciprofloxacin and placebo. Rifampicin/ciprofloxacin combination was successful in all patients compared to 58% of patients who received ciprofloxacin alone (Zimmerli, et al. 1998). Subsequent studies corroborated these results with success rates of greater than 85% of patients treated with debridement and retention and rifampicin containing antibiotic treatment (Berdal, et al. 2005, Byren, et al. 2009, Rao, et al. 2003). The main limitations with the use of rifampicin are the high likelihood of generation of resistance when used without a second antibiotic and the hepatic and gastrointestinal toxicities (John, et al. 2009, Widmer, et al. 1990). Therefore careful, regular follow up of patients is necessary and the management these antibiotics should involve collaboration between Infectious Diseases Physicians and Orthopaedic Surgeons.

The investigation of newer agents for the treatment of prosthetic joint infection is ongoing. In guinea pig foreign-body infection model, John et al assessed the activity of newer agents including linezolid and daptomycin, alone and in combination with rifampicin. In this study

Infection in Primary Hip and Knee Arthroplasty 429

approach between infectious Diseases Physicians and Orthopaedic Surgeons. Results at our

Aalto, K., Osterman, K., Peltola, H. & Rasanen, J. Change in Erythrocyte sedimentation rate

Aboltins, C.A., Dowsey, M.M., Buising, K.L., Peel, T.N., Daffy, J.R., Choong, P.F. & Stanley,

Aboltins, C.A., Page, M.A., Buising, K.L., Jenney, A.W.J., Daffy, J.R. & Choong, P.F.M.

Ahnfelt, L., Herberts, P., Malchau, H. & Andersson, G.B.J. Prognosis of total hip

Amann, R. & Fuchs, B.M. Single-cell identification in microbial communities by improved

Amorena, B., Gracia, E., Monzon, M., Leiva, J., Oteiza, C., Perez, M., Alabart, J.L. &

Bauer, T.W., Parvizi, J., Kobayashi, N. & Krebs, V. Diagnosis of periprosthetic infection. *J* 

Bengtson, S. & Knutson, K. The infected knee arthroplasty. *Acta Orthopaedica Scandinavica* 62,

Bengtson, S., Knutson, K. & Lidgren, L. Treatment of Infected Knee Arthroplasty. *Clinical* 

Berbari, E.F., Hanssen, A.D., Duffy, M.C., Steckelberg, J.M., Ilstrup, D.M., Harmsen, W.S. &

Berbari, E.F., Marculescu, C., Sia, I., Lahr, B.D., Hanssen, A.D., Steckelberg, J.M., Gullerud,

Berbari, E.F., Osmon, D.R., Duffy, M.C., Harmssen, R.N., Mandrekar, J.N., Hanssen, A.D. &

Osmon, D.R. Risk factors for prosthetic joint infection: case-control study. *Clin Infect* 

R. & Osmon, D.R. Culture-negative prosthetic joint infection. *Clinical Infectious* 

Steckelberg, J.M. Outcome of prosthetic joint infection in patients with rheumatoid arthritis: the impact of medical and surgical therapy in 200 episodes. *Clinical* 

replacement. *Acta Orthopaedica Scandinavica* 61, 2-26 (1990).

*Journal of Clinical Microbiology* 36, 2932-2939 (1998).

*Orthopaedics and Related Research* 245, 173-178 (1989).

*Bone Joint Surg Am* 88-A, 869-882 (2006).

and c-reactive protein after total hip arthorplasty. *Clinical Orthopaedics and Related* 

P.A. Gram-negative prosthetic joint infection treated with debridement, prosthesis retention and antibiotic regimens including a fluoroquinolone. *Clin Microbiol Infect*

Treatment of staphylococcal prosthetic joint infections with debridement , prosthesis retention and oral rifampicin and fusidic acid. *Clinical Microbiology and* 

fluorescence in situ hybridization techniques. *Nature Reviews Microbiology* 6, 339-348

Hernandez-Yago, J. Antibiotic susceptibility assay for Staphylococcus aureus in biofilms developed in vitro. *Journal of Antimicrobial Chemotherapy* 44, 43-55 (1999). Atkins, B.L., Athanasou, N., Deeks, J.J., Crook, D.W.M., Simpson, H., Peto, T.E.A., Lardy-

smith, P.M.C. & Berendt, A.R. Prospective Evaluation of Criteria for Microbiological Diagnosis of Prosthetic-Joint Infection at Revision Arthroplasty.

institution attest to the success of such a protocol.

*Research* 184, 118-120 (1984).

*Infection* 13, 586-591 (2007).

17, 862-867 (2011).

(2008).

301-311 (1991).

*Dis* 27, 1247-1254 (1998).

*Diseases* 45, 1113-1119 (2007).

*Infectious Diseases* 42, 216-223 (2006a).

**7. References** 

neither daptomycin nor linezolid had activity against adherent MRSA when used as monotherapy. When used in combination with rifampicin, daptomycin at a dose of 30mg/kg (corresponding to a dose of 6mg/kg in humans) cured 67% of cage infections. At this dose, no cases of rifampicin resistance emerged. Results were less encouraging for linezolid; even in combination with rifampicin, linezolid failed to cure any cage infection. Resistance to rifampicin emerged in 8% of cage infections treated with rifampicin-linezolid combinations (John, et al. 2009).

For gram-negative infections, ciprofloxacin has been shown to be effective in guinea pig tissue cage models(Widmer, et al. 1991). In a study of 28 patients with bone and joint infections secondary to gram-negative bacilli combination therapy with cefepime and fluoroquinolone obtained a cure in 79% of patients. However only 5 patients in this cohort had a prosthetic joint infection, two were treated with debridement and retention and only one of which was cured (the second patient died from a cause unrelated to the infection)(Legout, et al. 2006). In prosthetic joint infection secondary to gram negative bacilli, debridement and retention has yielded a success rate as low as 27% (Hsieh, et al. 2009b). This contrasts with our results where by infection free survival at 2 years was 94% in gram-negative infections when fluoroquinolone was used in conjunction with debridement and retention (Aboltins, et al. 2011). Again this is in the setting of short duration of symptoms (median 7 days) and prolonged oral antibiotic treatment (median 12 months).

The duration of antibiotic after debridement and retention varies in reported clinical studies ranging from six months to greater than 4 years. In a study by Laffer et al there was no difference in outcome in patients receiving three to six months of antibiotics compared with greater than six months (91% v 87% success). In this study patients were followed up for a median duration of 28 (range, 2–193) months and 55% of infections were caused by Staphylococcus species(Laffer, et al. 2006). In accordance with consensus guidelines, debridement and retention of the prosthetic joint should be considered in patients with a short duration of symptoms in the absence of implant loosening and soft tissue damage where antibiotics with biofilm activity are available (Laffer, et al. 2006, Matthews, et al. 2009, Trampuz & Zimmerli 2008, Zimmerli, et al. 2004).
