Oral antibiotic treatment of staphylococcal bone and joint infections in adults

J Antimicrob Chemother 2014; 69: 309 322 doi:10.1093/jac/dkt374 Advance Access publication 26 September 2013 Oral antibiotic treatment of staphylococcal bone and joint infections in adults Baek-Nam Kim 1, Eu Suk Kim 2 and Myoung-Don Oh 3 * 1 Department of Internal Medicine, Inje University Sanggye-Paik Hospital, Seoul, Republic of Korea; 2 Department of Internal Medicine, Seoul National University Bundang Hospital, Seongnam, Republic of Korea; 3 Department of Internal Medicine, Seoul National University College of Medicine, Seoul, Republic of Korea *Corresponding author. Tel: +82-2-2072-2945; Fax: +82-2-762-9662; E-mail: mdohmd@snu.ac.kr Bone and joint infections, especially implant-associated infections, are difficult to cure. Long-term antibiotic therapy, combined with appropriate surgery and the removal of prostheses, is required. The most common causative organisms in bone and joint infections are staphylococci. Oral agents are often used after an initial course of parenteral antibiotic treatment. However, it is unclear which oral regimens are most effective in staphylococcal bone and joint infections. We review various oral antibiotic regimens and discuss which regimens are effective for this indication. Keywords: antimicrobial treatment, Staphylococcus, osteomyelitis, infectious arthritis, orthopaedic fixation devices, joint prosthesis Introduction Bone and joint infections, especially implant-associated infections, are difficult to treat. The most common organisms causing bone and joint infections are staphylococci, including Staphylococcus aureus and coagulase-negative staphylococci. 1 4 To cure these infections, long-term antibiotic treatment combined with appropriate surgery and removal of the implant is necessary. Despite the paucity of large prospective randomized clinical trials evaluating the efficacyand safetyof oral therapy, and the heterogeneity of bone and joint infections, recent systematic reviews show that oral therapy is as effective as parenteral therapy provided that the microorganisms are susceptible to the agents used. 5 7 Oral antibiotic agents, alone or combined, are used for long-term treatment after the initial therapy with parenteral agents. 8 Previous reviews on the treatment of bone and joint infection have recommended certain oral agents for long-term therapy, although limited data on these antibiotic regimens have been reported. 1 4,9 13 The most recent clinical practice guidelines published by the Infectious Diseases Society of America recommend a combination of rifampicin and fluoroquinolone as a first-line regimen when an agent active against biofilms is needed, and include several other antibiotics for the oral therapy of prosthetic joint infection (PJI). 14 However, the background of these recommendations on what oral antibiotics should be used has not been clearly indicated, except for the combination of rifampicin and fluoroquinolone. We summarize here the available evidence on the choice of oral antibiotics for staphylococcal bone and joint infection. To determine effective oral antibiotic regimens, we reviewed the literature on pharmacokinetic characteristics, animal models and clinical studies of oral agents against staphylococcal bone and joint infections. Search strategy and selection criteria We searched Medline for articles published in English. No other restrictions were applied. The last search was done on 10 May 2012. The search algorithm was staphylococcus AND{osteomyelitis OR arthritis OR [(bone OR joint OR orthopedic OR implant) AND infection]} AND names or classes of antibiotics used in the treatment of staphylococcal infections. By searching the reference lists of the retrieved articles, we also identified relevant articles published in other languages and included them if appropriate. Each article was assessed for its clinical relevance and the quality of its methodology. Preference was given to randomized comparative trials focusing on oral antibiotic treatment regimens for staphylococcal bone and joint infections. However, the majority of the articles were case reports or observational studies, and only a few were randomized clinical trials. Therefore, although relevant clinical trials or animal studies were preferred, this review is mainly based on experimental models, historical observational studies, non-randomized clinical trials and the guidelines of expert societies. Commencing oral antibiotic therapy In the management of bone and joint infections, the selection of antibiotic regimens and the duration of antibiotic therapy vary depending on the clinical setting and the treatment approaches available. 9 Usually, an initial short course of intravenous treatment is given to reduce the bacterial burden and thereby minimize the risk of emergence of resistance to oral agents. 15 Intravenous therapy is administered for the first 2 4 weeks, followed by longterm oral therapy to complete the treatment. 9 Recent evidence indicates that an early switch to oral therapy is effective in patients # The Author 2013. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com 309

with PJIs. 16 In that publication, intravenous therapy for 10 14 days was followed by a switch to oral antibiotics either for 6 8 weeks or for up to 3 months, depending on the type of infection and the clinical conditions. An even shorter course of parenteral therapy of less than 7 days before oral switching was recently used for children with acute haematogenous osteomyelitis. 17,18 The limited evidence relating to the treatment of chronic osteomyelitis in adults suggests that the method of antibiotic administration (oral versus parenteral) does not affect the rate of disease remission provided the bacteria are susceptible to the antibiotic used. 7 Although oral antibiotic therapy is increasingly being shown to have promise for treating chronic osteomyelitis in adults, more evidence from comparative trials with adequate statistical power is necessary. Duration of oral antibiotic therapy The optimum duration of antibiotic treatment for bone and joint infections remains unknown because this has never been studied in prospective randomized studies. 6,7 For osteomyelitis, a total duration of 4 6 weeks of antibiotic therapy (after the last major debridement surgery) is generally recommended. 19 The suggested duration of oral antibiotic therapy in patients with implant retention or a one-stage exchange, based on a controlled trial in patients with orthopaedic implant-associated infection, is 3 months for hip prostheses and 6 months for knee prostheses. 15 In patients with fracture fixation devices, it is recommended that the duration of oral antibiotic therapy be 3 months when the device is retained and 6 weeks when all devices have been removed. 10 Long-term oral suppressive antibiotics may be considered in selected cases, particularly if it is not possible to remove the device. 10 For vertebral osteomyelitis, the recommended total duration ranges from 4 6 weeks to 3 months. 11 Prolonged antibiotic treatment is recommended in patients with undrained abscesses or spinal implants. 11 For arthritis, a 2 3 week course of therapy is suggested. 12 Selection of oral antibiotic agent Several factors should be considered when selecting oral antibiotics to treat bone and joint infections. These include the type of infection, the extent of debridement when applicable, the antibiotic susceptibility of the pathogen, antibiotic penetration into the bone and joint tissues, oral bioavailability and cost. The drug(s) selected must have activity against the isolated organism and have a low risk for the development of adverse reactions and drug drug interactions. The presence of a foreign body may also be one of the most important factors in choosing the antibiotic regimen. In the presence of a foreign body, there are slow-growing or adherent organisms in biofilms, against which antibiotic efficacy is diminished. 20 Therefore, for managing staphylococcal bone and joint infections, especially implant-associated infections, an optimal antibiotic agent should have activity against surface-adherent, slow-growing and biofilm-associatedpathogens. 21 It isnotable that standard antibiotic susceptibility tests, which evaluate drug efficacy on freely growing bacteria inthelogarithmicgrowthphase, are not reliable in predicting the outcome of implant-associated bone and joint infections. 22 Their only use is to exclude antibiotic agents without in vitro efficacy. 4 Antibiotic bone penetration is also an important factor to consider, especially in the treatment of orthopaedic implant-associated infection, because it often has bone sequesters and an established biofilm. 23 Regarding data on bone penetration, readers are advised to refer to an excellent review by Landersdorfer et al. 24 Monotherapy versus combination therapy Whether monotherapy or combination therapy is more effective in staphylococcal bone and joint infections, especially implantassociated infections, remains unanswered. In a retrospective cohort study of implant-associated infections caused by methicillinresistant S. aureus (MRSA) (44% with prosthetic joints and 56% with osteosynthesis devices), 35% of the patients experienced treatment failure, and monotherapy (hazard ratio 4.4, 95% CI 1.2 16.3; P¼0.025) was an independent predictor of treatment failure. 25 Most of the combination therapy regimens contained rifampicin. These findings suggest that combination therapy with rifampicin should be considered for patients with MRSA implant-associated infection, especially when implant removal is not feasible. It should be noted that the primary aim of antibiotic combination therapy is to decrease the risk of emergence of resistance to a companion drug or to provide synergistic or additive antibacterial activity. Monotherapy Rifampicin Rifampicin has excellent oral bioavailability (70% 90%) and potent antistaphylococcal activity. It is also able to penetrate biofilms and has good activity in them. It can eradicate adherent and stationary-phase staphylococci with MICs 10 100 times higher than those for proliferative-phase organisms. 22 The efficacy of rifampicin in staphylococcal bone and joint infections has been proven in many animal models (Table 1). 26 44 As a single agent, it is more active than fusidic acid or ciprofloxacin against MRSA retrieved from device-associated biofilm infections. 45 Therefore, it is a critically important antibiotic in the treatment of bone and joint infections, especially where implants are retained. Resistance to rifampicin develops readily as a result of single point mutations in the DNA-dependent RNA polymerase gene. 46 Rifampicin-resistant mutants were recovered at a frequency of around 10 8 in rifampicin monotherapy, whereas they were not recovered (frequency,10 211 ) in combination therapy with rifampicin/fusidic acid. 47 Therefore, an adequate companion drug must be used to prevent the emergence of rifampicin resistance. Nevertheless, rifampicin resistance may still emerge when the inoculum of bacteria is high or surgical drainage is inadequate. 48 Fluoroquinolones Fluoroquinolones are active against staphylococci in vitro, but are less active, compared with rifampicin, against adherent staphylococci, which these agents can rarely eradicate when given alone. 49 They exhibit high bone to serum concentration ratios, and bone concentrations are higher than the MIC 90 for the causative organisms. 50 52 Newer fluoroquinolones (such as levofloxacin, moxifloxacin, gatifloxacin and gemifloxacin) tend to have lower MICs for Gram-positive pathogens than do older fluoroquinolones (such as ciprofloxacin and ofloxacin) and have a higher barrier to the emergence of resistance. 53 Older fluoroquinolones when used alone in staphylococcal infections tend to select resistant mutants. 54 To prevent the emergence of resistance, older fluoroquinolones are recommended in combination with other agents. 4 310

JAC Table 1. Summary of animal studies of treatment for staphylococcal bone and joint infections or implanted-associated infections Antibiotic regimen Findings Rifampicin-containing rifampicin/ciprofloxacin effective 26,27 rifampicin/fleroxacin effective 28 rifampicin/pefloxacin effective 27 rifampicin/levofloxacin effective 29 better than rifampicin/linezolid 30 antagonistic 31 rifampicin/linezolid effective 32 rifampicin/trimethoprim/sulfamethoxazole effective 33 rifampicin/clindamycin effective 34 rifampicin/minocycline effective, but less so than rifampicin/vancomycin 35 rifampicin/azithromycin effective 36 rifampicin/clarithromycin effective 36 rifampicin/nafcillin effective 36 Fluoroquinolone levofloxacin effective, 37 but less so than nafcillin in MSSA osteomyelitis 38 effective at high dose 39 moxifloxacin effective 40 ciprofloxacin less effective than cefuroxime in chronic S. aureus osteomyelitis 41 gatifloxacin comparable to nafcillin in MSSA osteomyelitis 42 Oxazolidinone linezolid ineffective in MSSA osteomyelitis, whereas cefazolin was active 43 Lincosamide clindamycin effective 44 Macrolide azithromycin ineffective 34 MSSA, methicillin-susceptible S. aureus. In the case of newer fluoroquinolones, monotherapy has been proven to be effective in decreasing bacterial counts in the bone and joint fluid and in biofilms in animals with implant-associated staphylococcal infections. 31,37,40,42 Fluoroquinolone-resistant mutants did not emerge during monotherapy with levofloxacin or moxifloxacin in animal models. 31,37,55 Newer fluoroquinolones such as moxifloxacin are at least as active against staphylococci as b-lactam and glycopeptide antibiotics in animal models of arthritis and chronic implant-associated osteomyelitis. 40,56 Fluoroquinolones have been used as single agents against chronic osteomyelitis caused by staphylococci (Table 2). 57 59 In a recent study of orthopaedic implant-associated staphylococcal infections, the overall cure rate with moxifloxacin monotherapy (400 mg/ day for 3 months) was 82.6%, and the cure rate for patients retaining implants was 71.4%. 59 In that study, 77% of the patients underwent surgery and the implant was retained in 43.8%. Of the eight patients who relapsed, six had microbiologically confirmed disease, and all the organisms recovered were susceptible to fluoroquinolone. The investigators in that study suggested that fluoroquinolone monotherapy might be a suitable option for the long-term treatment of bone and joint infections caused by fluoroquinolone-susceptible staphylococci. However, as fluoroquinolone monotherapy may induce resistance, we think that this approachshouldbe employed onlywhen alternative regimens are not available. Fusidic acid Fusidic acid is a bacterial protein synthesis inhibitor with antibiotic activity against staphylococci, including methicillin-resistant organisms. 60 It has good penetration into infected bone and joints, 61 and although it is less effective than rifampicin, it has activity against staphylococcal biofilms. 45 Fusidic acid-resistant strains occur naturally at a rate of between 10 26 and 10 28 cfu. 60 The rate of emergence of resistance reached 5.1% with fusidic acid monotherapy while it was less than 1% with fusidic acid combination therapy. 62 A higher rate (15%) of resistance was noted in patients with chronic osteomyelitis who received prolonged courses of fusidic acid monotherapy. 60 Therefore fusidic acid monotherapy is not a rational option for staphylococcal bone and joint infections, although there are old reports describing experience with it. 63,64 The use of fusidic acid is generally restricted to oral maintenance treatment in combination with other agents such as rifampicin or a fluoroquinolone. 65 Fusidic acid, combined with other agents, has been used for more than 40 years for various staphylococcal bone and joint infections, including acute and chronic osteomyelitis, arthritis and other orthopaedic infections. 66,67 Linezolid Linezolid has antibiotic activity against a wide spectrum of Grampositive organisms. It does not cause cross-resistance to antibiotic agents of other classes, and it does not require its dose to be adjusted according to renal and hepatic function. 68 The concentration of linezolid in bone and joint fluid is high enough to treat infections. 69 When 600 mg of linezolid was given orally every 12 h over 48 h, its mean concentration in cancellous bone 90 min after the final dosewas at leasttwice the MIC 90 (4 mg/l) for staphylococci. 70 311

Table 2. Summary of clinical studies of fluoroquinolone monotherapy for staphylococcal bone and joint infections Overall clinical cure rate Note Reference a Antibiotic Number of patients Type of infection Organism (n) failure in 6 with S. aureus (3 treated with ciprofloxacin and 3 treated with ofloxacin) and 1 with S. epidermidis (treated with ofloxacin) 66.7% in staphylococci 39 chronic osteomyelitis S. aureus (19), S. epidermidis (2) and Gram-negative pathogens (18) Dellamonica ciprofloxacin, et al. 57 ofloxacin, pefloxacin chronic osteomyelitis various (S. aureus in 8) 77% (100% in S. aureus) Gentry and ciprofloxacin 31, compared with a Rodriguez 58 combination of b-lactam and aminoglycoside relapse in 8 (6 microbiologically confirmed, all fluoroquinolone susceptible) MSSA (33), CoNS (15) 82.6% (71.4% with implant retention) San Juan moxifloxacin 46 (evaluable) implant-associated et al. 59 infection CoNS, coagulase-negative staphylococci; MSSA, methicillin-susceptible S. aureus. a Only studies for which the post-treatment follow-up period was more than 1 year are shown. In one animal study, it was not effective for the treatment of chronic S. aureus osteomyelitis. 43 Thrombocytopenia and anaemia may occur due to durationdependent reversible myelosuppression, especially in patients receiving linezolid for more than 2 weeks. 71 Irreversible peripheral neuropathy may also occur with prolonged treatment. 72 Such adverse reactions, along with its high cost, are major obstacles to the wide use of this potent drug in staphylococcal bone and joint infections. Linezolid has been used for a variety of bone and joints infections in humans (Table 3). Even though the conditions 73 77 varied between trials, clinical cure rates were 55% 100%. Trimethoprim/sulfamethoxazole Trimethoprim/sulfamethoxazole has been used to treat staphylococcal bone and joint infections in children, 78 as well as in adults. 79 Rates of trimethoprim/sulfamethoxazole resistance amongs. aureus isolates are highly variable and increasing. 80 Time kill studies indicate that trimethoprim/sulfamethoxazole is rapidly bactericidal against MRSA at concentrations four times the MIC. 81 Trimethoprim penetrates bone at about 50% of serum levels, while sulfamethoxazole penetration is somewhat lower (15%). 82 Concentrations of trimethoprim in synovial fluid approach serum levels, whereas sulfamethoxazole does not as readily penetrate into the synovial fluid. 83 Oral trimethoprim/sulfamethoxazole alone has been used to treat staphylococcal bone and joint infections. 84 In a recent study, oral trimethoprim/sulfamethoxazole at high doses was used as an alternative to conventional parenteral therapy in patients with staphylococcal orthopaedic implant-associated infections. 85 However, trimethoprim/sulfamethoxazole treatment is not suitable for infections with abscesses or a lot of necrotic tissue because the availability of exogenous thymidine may bypass the double biosynthetic blockade of trimethoprim/sulfamethoxazole. 86 In summary, trimethoprim/sulfamethoxazole as a monotherapy can be a therapeutic option for patients with staphylococcal bone and joint infections, particularly to complete an extended course of therapy after the bacterial burden has been reduced. 84 Lincosamides Clindamycin has been used either alone or in combination for the long-term oral therapy of staphylococcal bone and joint infections. It has excellent bone penetration and oral bioavailability. However, a recent literature review shows that it has only the same or a slightly higher bone penetration than b-lactams. 24 It is not generally recommended if the target organism is erythromycin resistant, because a single-step resistance mutation due to the MLS B mechanism can occur. 87 Staphylococcal isolates that are clindamycin susceptible, but erythromycin resistant, should be tested for inducible clindamycin resistance using the D-test. 87 D-test-positive organisms are likely to develop resistance to clindamycin during treatment. 88 Clindamycin has been successfully used against staphylococcal bone and joint infections, especially in children, 89,90 but rarely in adults. 91 The use of clindamycin in MRSA osteomyelitis is suggested for two situations: extended courses of oral clindamycin, which may be used in patients who display osteomyelitis that is refractory to other agents, and community-acquired MRSA 312

JAC Table 3. Summary of clinical studies of linezolid monotherapy for staphylococcal bone and joint infections Reference a Number of patients Type of infection Organism (n) Overall clinical cure rate Note MSSA 3, MRSA 19, VRE 18 90.0% for osteomyelitis Broder et al. 73 40 osteomyelitis, skin and soft-tissue infections Bassetti et al. 74 20 PJI MRSA 14, MRCoNS 5, enterococci 1 80% 55% Aneziokoro et al. 75 20 osteomyelitis MRSA 8, MSSA 1, CoNS 5, enterococci 3, streptococci 1, others 2, culture negative or not done 3 98.0% (49/50) in remission 16 treated with initial linezolid followed by long-term suppression with/without linezolid Rao and Hamilton 76 53 PJI 23, osteomyelitis 25, others 5 MRSA 21, MSSA 6, MRCoNS 17, MSCoNS 2, enterococci 7 100% monomicrobial 9+polymicrobial 13, including MRSA 10 and VRE 5 Vercillo et al. 77 22 (14 evaluable) implant-associated osteomyelitis (18 with fracture fixation implants, 4 with arthroplasty implants) CoNS, coagulase-negative staphylococci; MRCoNS, methicillin-resistant coagulase-negative staphylococci; MSCoNS, methicillin-susceptible coagulase-negative staphylococci; MSSA, methicillin-susceptible S. aureus; VRE, vancomycin-resistant enterococci. a Only studies for which the post-treatment follow-up period was more than 1 year are shown. infections, which often retain susceptibility to clindamycin. 92 A recent guideline suggests that clindamycin can be considered for the treatment of bone and joint infections caused by susceptible staphylococci where inducible MLS B resistance has been excluded. 93 Streptogramins Pristinamycin is an oral streptogramin antibiotic consisting of two structurally unrelated, but synergistic, compounds, pristinamycin IA and pristinamycin IIA. It has been available in Europe for over 30 years for treating respiratory tract, skin and soft-tissue infections caused by susceptible Gram-positive bacteria. Pristinamycin is mainly active against Gram-positive bacteria including erythromycin-resistant staphylococci and MRSA. 94 Recent clinical studies have also demonstrated the usefulness of pristinamycin for bone and joint infections. 95,96 In one study, oral pristinamycin achieved cure or suppression in 21 of 22 patients with staphylococcal bone and joint infections. 96 It appears to be a well-tolerated, effective oral alternative agent for treating difficult-to-treat bone and joint infections caused by staphylococci, particularly where there is intolerance of or resistance to rifampicin or fusidic acid. 95,96 Tetracyclines Tetracyclines are not as widely used for treating staphylococcal infections as they once were. Long-acting tetracyclines such as doxycycline and minocycline have good oral bioavailability and tissue penetration, and better antistaphylococcal activity than tetracycline. 97 The antistaphylococcal activity of minocycline is better than that of doxycycline in vitro, but clinical superiority has not been demonstrated. 98 Clinical data on the use of tetracyclines, singly or in combination, against staphylococcal bone and joint infections are very sparse. 97 A recent review does not support their use as monotherapy in cases of osteomyelitis because of the present insufficiency of clinical data. 97 Given the pharmacokinetic advantages of long-acting tetracyclines, we think that their efficacy as monotherapy for staphylococcal bone and joint infections needs to be determined. Macrolides Erythromycin has low bone penetration. 24 It poorly penetrates biofilms produced by MRSA. 99 In contrast, azithromycin has a long half-life in serum and tissues, and its bone concentrations are higher than its serum concentrations. 100 However, azithromycin was ineffective as a single drug against experimental staphylococcal osteomyelitis despite concentrations in bone that markedly exceeded the MIC. 34 Macrolides should not be used as monotherapy in staphylococcal bone and joint infections. Fosfomycin Fosfomycin has excellent in vitro activity against many Grampositive and Gram-negative organisms, including methicillinresistant staphylococci. Only low rates of adverse events, mainly mild gastrointestinal distress, have been reported. 101 It achieved clinically relevant concentrations in cortical bone, cancellous bone and post-osteomyelitis sequestra. 102 It has been shown to achieve levels in bone tissue well above the expected MICs for 313

common pathogens in diabetic patients with foot infections. 103 However, in an animal model of MRSA osteomyelitis, 22.2% (2/9) of rats were positive for MRSA in bone after 4 weeks of fosfomycin treatment. 104 Studies of oral fosfomycin are very rare. This drug is a promising option, but the clinical efficacy of oral fosfomycin, alone or in combination with other antibiotics, should be adequately evaluated for the treatment of bone and joint infections. b-lactam antibiotics b-lactams, which inhibit cell wall synthesis, are inactive against biofilm-associated staphylococci, but are active when combined with rifampicin. 45 Cefalotin combined with rifampicin was more effective than a rifampicin/tetracycline combination in young biofilms of Staphylococcus epidermidis and at high concentrations, whereas the opposite was the case at lower concentrations in aged biofilms. 105 Because of their limited bioavailability, oral dosing with b-lactams is unlikely to achieve adequate bone levels. 106 In one study, however, oral cefadroxil achieved adequate antibiotic concentrations in uninfected bone. 107 Oral amoxicillin/clavulanate and some first-generation cephalosporins fulfilled the pharmacokinetic/ pharmacodynamic requirements for clinical efficacy, especially in children. 108 Amoxicillin/clavulanate was as effective as flucloxacillin and clindamycin by subcutaneous injection in an experimental rat model of staphylococcal osteomyelitis. 109 Oral antibiotic therapy with b-lactams has been successful against childhood bone and joint infections, 110 112 which are usually haematogenous in origin and heal more rapidly than adult bone and joint infections. In one retrospective study involving 29 children with acute osteomyelitis treated withshort-course parenteral antibiotics followed by oral cefalexin, none suffered treatment failure or complications at a 6 month follow-up. 113 In a recent prospective study comparing oral first-generation cephalosporins (cefradine, cefalexin and cefadroxil) with oral clindamycin after intravenous therapy for the first 2 4 days for paediatric osteoarticular infections, the two regimens had equal efficacy and safety profiles. 90 Oral b-lactam/b-lactam inhibitors or first-generation cephalosporins are recommended for paediatric staphylococcal osteoarticular infections in a French guideline. 114 In summary, some oral b-lactams can be used for the follow-up treatment of staphylococcal osteoarticular infections in children after short-term intravenous therapy. 17 Only limited studies are available on the use of oral b-lactams against bone and joint infections in adults as opposed to children. 6 Oral b-lactams are not generally recommended against staphylococcal bone and joint infections in adults, particularly with implant retention. 1 However, some experts recommend oral switching with b-lactams in adults with acute osteomyelitis. 115 117 The most recent Infectious Diseases Society of America practice guidelines also recommend oral antistaphylococcal penicillinsorfirst-generation cephalosporins in combination with rifampicin as long-term oral therapy in PJIs caused by methicillin-susceptible staphylococci, albeit based on expert opinion without strong evidence. 14 Combination therapy Rifampicin-containing combinations Rifampicin combinations in vitro showed antagonism or indifference against staphylococci more frequently than synergism. 118 However, in animal models of osteomyelitis, they led to significant reductions in positive bone cultures and in cfu per gram of bone. 119 A recent large multicentre study also found that combination therapy with rifampicin protected against treatment failure in staphylococcal PJIs managed with debridement and retention. 120 Recent retrospective analyses found that the emergence of rifampicin resistance during rifampicin combination therapy was associated with previous surgical revisions, a high initial bacterial load and previous inadequate rifampicin therapy. 121 Combination therapy with rifampicin is most promising for the treatment of osteomyelitis and prosthetic device-related infections caused by staphylococci, but more definitive data are needed. 119,122 It is of note that rifampicin, alone or in combination, is not recommended for chronic suppression. 14 Of the oral combination agents available so far, the rifampicin/fluoroquinolone combination has been the most frequently used against bone and joint 15,123 133 infections (Table 4). (1) Rifampicin/fluoroquinolone combinations Although combinations of rifampicin and a fluoroquinolone yielded a variety of results against staphylococci in vitro, these results may not correlate with outcomes in animal models or clinical infections. 134 The rifampicin/fluoroquinolone combination seemed more effective than rifampicin alone in experimental MRSA osteomyelitis, 27 whereas rifampicin/levofloxacin was antagonistic in a rib cage model of S. aureus foreign body infection. 31 Rifampicin/fluoroquinolone combinations are useful especially for implant-associated infections or osteomyelitis caused by staphylococci. 135 Rifampicin/ fluoroquinolone combination therapy was associated with better outcomes than other antibiotic regimens for patients with PJIs due to S. aureus. 136 In particular, the rifampicin/ciprofloxacin combination is the only regimen whose efficacy has been proven for staphylococcal bone and joint infections associated with stable orthopaedic devices in a randomized comparative trial. 15 In that study, after initial debridement, either rifampicin/ciprofloxacin combination therapy or ciprofloxacin monotherapy was given after a 2 week intravenous course of flucloxacillin or vancomycin. The oral therapy lasted for 3 months for patients with hip prostheses and internal fixation devices, and 6 months for patients with knee prostheses. All 12 patients in the rifampicin/ciprofloxacin combination group remained cured after more than 24 months of follow-up. In contrast, only seven (58%) of the 12 patients in the ciprofloxacin monotherapy group remained cured. Rifampicin-containing combination regimens with ofloxacin and fleroxacin have been evaluated in staphylococcal orthopaedic implant-associated infections, 123 osteomyelitis associated with diabetic foot ulcers 124 and acute bone and joint infections. 137 One study reported an overall success rate of 74% after 6 9 months of rifampicin/ofloxacin treatment, with a figure of 62% in patients without prosthesis removal. 123 Rifampicin combinations with newer fluoroquinolones, such as levofloxacin, moxifloxacin or gemifloxacin, have been studied more often than combinations with older fluoroquinolones in bone and joint infections. 125,126,138 140 The overall cure rates of the rifampicin/levofloxacin combination were 72% in staphylococcal implant-associated infections 139 and 96% in staphylococcal spondylodiscitis. 140 Considering the antibiotic spectrum and pharmacokinetic/pharmacodynamic characteristics of the newer 314

Table 4. Summary of clinical studies of rifampicin-containing combination therapy for staphylococcal bone and joint infections Antibiotic regimen Number of patients Type of infection Organism (n) Overall clinical cure rate Reference a Review Rifampicin/ fluoroquinolone flucloxacillin (1 g every 6 h) or vancomycin (1 g every 12 h) for 2 weeks, then rifampicin (450 mg bid)+ciprofloxacin (750 mg bid) versus ciprofloxacin only for 3 6 months rifampicin (900 mg qd)+ofloxacin (200 mg tid) for 6 9 months without removal of the device rifampicin (600 mg bid) + ofloxacin (200 mg tid) for a median of 6 months rifampicin (600 mg qd)+levofloxacin (500 mg qd) rifampicin+levofloxacin or moxifloxacin for a median of 7 months (dosage unknown) 33 orthopaedic device-related infection 47 (evaluable) orthopaedic device-related infection 20 episodes in 17 patients with diabetic foot infection osteomyelitis MSSA 26, MSSE 5, MRSE 2 100% versus 58% Zimmerli et al. 15 S. aureus 26, CoNS 21 74% (81% for hip, 69% for knee and 69% for osteosynthesis device) monomicrobial 15+polymicrobial 5, including staphylococci 18 76.5% at a median of 22 months follow-up Drancourt et al. 123 Senneville et al. 124 11 PJI S. aureus 5, CoNS 6 91.9% Soriano et al. 125 7 infected orthopaedic implant 4, osteomyelitis 3 MSSA 5, streptococci 1, MSCoNS 1 86% (6/7) at a 19 month follow-up Frippiat et al. 126 Rifampicin/fusidic acid vancomycin iv, then rifampicin+fusidic acid for MRSA rifampicin (900 mg qd)+fusidic acid (500 mg tid for 5 days, then 500 mg bid) versus rifampicin (900 mg qd) + ofloxacin (200 mg tid) for 6 9 months, with removal of the prosthesis or implant, if necessary a b-lactam or glycopeptide iv for a median of 12 days, then rifampicin (300 mg bid)+fusidic acid (500 mg tid) for a median of 12 months; all with surgical debridement and prosthesis retention 20 spondylitis MRSA 6, MSSA 7, CoNS 3, others 4 67% in MRSA Torda et al. 127 46 (23 versus 23) orthopaedic implant S. aureus 28, CoNS 18 55% versus 50% Drancourt et al. 128 infection 20 PJI MRSA 10, MSSA 7, MRCoNS 1, MRSA+MSSA 1, MSSA+MSCoNS 1 90% Aboltins et al. 129 315 Rifampicin/ linezolid rifampicin [10 mg/kg (maximum 900 mg) bid]+linezolid (600 mg bid) versus rifampicin (10 mg/kg bid)+sxt (8/40 mg/kg/day) salvage with rifampicin (300 mg tid) +linezolid (600 mg bid) without implant removal 56 infected orthopaedic device 36, chronic osteomyelitis 20 MRSA 21, MSSA 11, MRCoNS 18, MSCoNS 3, others 23 49 PJI MRSE 22, MRSA 6, culture-negative 17, not described 4 89.3% versus 78.6% (P¼0.47) 69.4% at a 24 month follow-up Nguyen et al. 130 Gomez et al. 131 Continued JAC

Table 4. Continued Overall clinical cure rate Reference a Antibiotic regimen Number of patients Type of infection Organism (n) Euba et al. 132 all MSSA 89% versus 91% at a 10 year follow-up 28 versus 22 chronic osteomyelitis (orthopaedic implant 35.7% versus 45.2%) Rifampicin/SXT rifampicin (600 mg qd)+sxt (240/ 1200 mg bid) for 8 weeks versus cloxacillin (iv for 6 weeks, then orally for 2 weeks) 7 PJI S. aureus 3, CoNS 4 71.4% Soriano et al. 125 rifampicin (600 mg qd)+clindamycin (300 mg tid) Rifampicin/ clindamycin 100% Czekaj et al. 133 MSSA 12, MRSA 5, CoNS 3 (co-infection of S. aureus and Streptococcus agalactiae in 1 case) 20 bone and joint infections including implant-associated infections (10) empirical antibiotics (amoxicillin/ clavulanate, cefuroxime, vancomycin) iv for a median of 6 days (0 14 days), then rifampicin (600 mg bid)+clindamycin (600 mg tid) orally for a median of 45 days (35 90 days) bid, twice daily; CoNS, coagulase-negative staphylococci; iv, intravenously; MRCoNS, methicillin-resistant coagulase-negative staphylococci; MRSE, methicillin-resistant S. epidermidis; MSCoNS, methicillin-susceptible coagulase-negative staphylococci; MSSA, methicillin-susceptible S. aureus; MSSE, methicillin-susceptible S. epidermidis; qd, once daily; SXT, trimethoprim/sulfamethoxazole; tid, three times daily. a Only studies for which the post-treatment follow-up period was more than 1 year are shown. fluoroquinolones, these combinations with rifampicin are promising against staphylococcal bone and joint infections. (2) Rifampicin/fusidic acid combination The efficacy of rifampicin/fusidic acid was equivalent to that of rifampicin/ofloxacin in a prospective trial of staphylococcal implant-associated infections. 128 Success with rifampicin/fusidic acid was also reported in staphylococcal PJIs treated with debridement and prosthesis retention. 129,141 Rifampicin/fusidic acid has been used against staphylococcal bone and joint infections after initial effective control with vancomycin. 116,142 However, the addition of fusidic acid did not prevent the emergence of resistance to rifampicin in a recent case report. 142 In summary, rifampicin/ fusidic acid offers an alternative option to rifampicin/fluoroquinolone for the oral treatment of staphylococcal orthopaedic implant-associated infections either when the patient is intolerant of fluoroquinolones or when the isolate is resistant to these drugs. (3) Rifampicin/linezolid combination In a rat model of staphylococcal foreign body-associated infection, rifampicin/linezolid had a protective effect against the development of rifampicin resistance. 32 In that study, its efficacy was similar to that of rifampicin/vancomycin. However, it was less effective than rifampicin/levofloxacin in another study. 30 Because linezolid monotherapy is not able to eradicate adherent bacteria, 30 it would be prudent not to use linezolid alone in implant-associated infections caused by MRSA. Case series have shown that rifampicin/ linezolid and rifampicin/trimethoprim/sulfamethoxazole were equally effective (89.3% versus 78.6%; P¼0.47) in patients with boneand jointinfectionscaused by resistantgram-positive cocci. 130 Rifampicin/linezolid was less frequently associated with anaemia than linezolid alone or rifampicin in combination with other drugs (9.3% versus 44.0% versus 52.0%, respectively; P,0.01). 143 However, the rates of thrombocytopenia were similar (44.2% versus 48.0% versus 57.7%), as were those of peripheral neuropathy. A recent study indicated that rifampicin may decrease serum concentrations of linezolid. 144 Increased extrarenal linezolid metabolism by rifampicin, resulting in lower serum concentrations of linezolid, has been suggested to be responsible for the lower frequency of haematological adverse events with rifampicin/linezolid. 145 In summary, because of the risk of severe adverse reactions during long-term therapy, rifampicin/linezolid can be considered for staphylococcal bone and joint infections only when no alternative regimen is available. 143 (4) Rifampicin/trimethoprim/sulfamethoxazole combination Rifampicin/trimethoprim has been shown to prevent the emergence of resistance to rifampicin or to trimethoprim, which is often seen in monotherapy with either agent. 146 In a rabbit model of chronic osteomyelitis due to S. aureus, rifampicin/trimethoprim was significantly more effective in sterilizing infected bones than either agent alone. 33 In a randomized trial of efficacy in chronic S. aureus osteomyelitis, overall cure rates with oral rifampicin/trimethoprim/sulfamethoxazole versus intravenous cloxacillin were 89% versus 91% at a 10 year follow-up. 132 Rifampicin/trimethoprim/sulfamethoxazole is recommended as an initial antibiotic regimen for acute and non-complicated 316

JAC osteoarticular infections in infants and children in a French guideline. 114 However, it should not be used against bone and joint infections with necrotic tissue because folic acid antagonists do not have synergistic activity against S. aureus in the presence of the elevated thymidine concentrations present in damaged host tissues and fail to prevent the emergence of rifampicin resistance. 147 (5) Rifampicin/clindamycin combination Rifampicin/clindamycin was more effective than either agent alone in reducing the bacterial counts in bone in experimental staphylococcal osteomyelitis. 34 Reports of case series have described patients with staphylococcal bone and joint infections, especially implant-associated infections, who were successfully treated with this combination. 125,133 (6) Rifampicin/tetracycline combinations In a rabbit model of orthopaedic device-related infections due to S. epidermidis, rifampicin/minocycline yielded a cure rate of 70%, whereas rifampicin/vancomycin achieved 90% cure. 35 There are several anecdotal cases of staphylococcal bone and joint infections that were treated with rifampicin/tetracycline. 148,149 The combination of rifampicin with a tetracycline (minocycline or doxycycline) has been recommended in adults with staphylococcal bone and joint infections if there are no other treatment options. 117 (7) Rifampicin/macrolide combinations In animal models of staphylococcal osteomyelitis, rifampicin/ azithromycin or rifampicin/clarithromycin resulted in reductions of bacterial counts in bone similar to those seen with rifampicin/ clindamycin and rifampicin/nafcillin. 34,36,150 Based on these data, it is suggested that a rifampicin/macrolide combination can be an alternative oral regimen against staphylococcal bone and joint infections. 36 This still needs to be supported by more clinical experience. Fluoroquinolone-containing combinations In vitro studies on staphylococci gave diverse or inconsistent results with combinations of a fluoroquinolone and fusidic acid, 151,152 ciprofloxacin/moxifloxacin and clindamycin, 134,153 gemifloxacin/ moxifloxacin and trimethoprim/sulfamethoxazole, 153,154 ciprofloxacin and tetracycline, 134 moxifloxacin and doxycycline 153 and ciprofloxacin and erythromycin. 134 It has been suggested that a fluoroquinolone/fusidic acid combination could be used against MRSA bone and joint infections if the target organism is susceptible. 65 There is limited clinical experience of bone and joint infections treated with such a combination. 66 Clinical experience with other fluoroquinolone-containing combinations is lacking, except for one example of salvage treatments with the ciprofloxacin/clindamycin combination in staphylococcal bone and joint infections. 15 The value of fluoroquinolone-containing combinations needs to be established as they might be useful when the target organism is not susceptible to rifampicin. Fusidic acid-containing combinations Fusidic acid and erythromycin displayed in vitro synergism against staphylococci. 151 A few anecdotal cases have been reported of Table 5. Recommendations for oral antibiotic therapy of staphylococcal bone and joint infections in adults Condition Regimen a Note Susceptible to rifampicin Resistant to, or intolerant of, rifampicin Resistant to, or intolerant of, rifampicin and fluoroquinolones fusidic acid combined with erythromycin or trimethoprim/sulfamethoxazole for treating bone and joint infections. 66 In one case series, all 45 children with acute osteomyelitis and pyogenic arthritis (31 due to S. aureus) were cured with a combination of fusidic acid and erythromycin over 3 weeks. 155 Fusidic acid/clindamycin is suggested for adults with staphylococcal bone and joint infections in French guidelines. 156,157 Because evidence is scant, we cannot recommend the use of fusidic acid in combination regimens except for fusidic acid/rifampicin or fusidic acid/fluoroquinolone combinations. Conclusions rifampicin combined with another inhibitory agent (preferably a fluoroquinolone or fusidic acid) b fluoroquinolone combinations c monotherapy with an inhibitory agent d rifampicin/ fluoroquinolone should be considered as the first-line regimen for implant-associated infections a Suggested oral dosages for adults with normal renal function are as follows; cefadroxil, 500 1000 mg twice daily (bid); cefalexin, 500 mg four times daily (qid); ciprofloxacin, 500 750 mg bid; clindamycin, 300 600 mg qid; cloxacillin, dicloxacillin or flucloxacillin, 500 mg qid; doxycycline or minocycline, 100 mg bid; fusidic acid, 500 mg three times daily (tid); levofloxacin, 500 750 mg once daily (qd); linezolid, 600 mg bid; moxifloxacin, 400 mg qd; pristinamycin, 1 g bid or tid; rifampicin, 300 450 mg bid or 600 900 mg qd; trimethoprim/sulfamethoxazole (SXT), 160/800 mg bid or tid. b Oral agents available for rifampicin combination therapy include fluoroquinolones, fusidic acid, clindamycin, SXT, a tetracycline (doxycycline or minocycline), linezolid, antistaphylococcal penicillins and first-generation cephalosporins. Among the fluoroquinolones, we prefer newer ones with higher antistaphylococcal activity in vitro. c Oral agents available for fluoroquinolone combination therapy include fusidic acid, clindamycin, SXT and a tetracycline, but evidence supporting the use of these combination regimens is inadequate. Combination therapy is preferable to monotherapy. d Oral agents available for monotherapy include newer fluoroquinolones, clindamycin, SXT, pristinamycin and linezolid. Oral antistaphylococcal penicillins or first-generation cephalosporins can be considered in bone and joint infections without in situ implants. Monotherapy is inadequate against prosthetic joint infections and chronic osteomyelitis. 1 The type of infection, the presence of an implant and the treatment strategy should be considered when selecting antibiotics to treat bone and joint infections. Ideally, the antibiotic used, particularly in implant-associated infections, should have bactericidal activity 317

against surface-adhering, slow-growing and biofilm-producing staphylococci. Investigations revealed that the in vitro results of antibiotic combinations, for example of rifampicin and another antibiotic, often do not correlate with the in vivo findings. 119 Animal studies and models of bone and joint infection have limitations such as a lack of debridement in the animals, high initial inocula, a lack of experience with recurrent or prolonged infection, and the infeasibility of long-term follow-up. 8 These findings and limitations raise a serious question as to whether in vitro or animal studies of the efficacy of antibiotic therapy have sufficient clinical relevance in the treatment of human bone and joint infections. In contrast, human studies, although mostly non-comparative clinical studies or case series, provide limited, but more useful, information on the choice of oral antibiotics in the treatment of staphylococcal bone and joint infections. Based on the available data as described above, we suggest oral agents for the treatment of staphylococcal bone and joint infections in adults, particularly those with chronic/ implant-associated infections, as shown in Table 5. First, rifampicin/fluoroquinolone should be considered as the first-line combination regimen, especially for implant-associated infections, because this combination has been the most extensively studied and its efficacy has been established. Rifampicin/fusidic acid can be used instead when the isolate is resistant to fluoroquinolones or the patient has adverse reactions to fluoroquinolones. When the aforementioned regimens cannot be used, rifampicin in combination with clindamycin, trimethoprim/sulfamethoxazole, a tetracycline (doxycycline or minocycline) or linezolid can be used. Oral antistaphylococcal penicillins or first-generation cephalosporins can be considered in bone and joint infections caused by methicillin-susceptible staphylococci. Second, combinations of fluoroquinolones with other antibiotics such as fusidic acid, clindamycin, trimethoprim/sulfamethoxazole or a tetracycline may be considered, but only when treatment fails or severe adverse reactions occur with the aforementioned antibiotic combinations. However, one should keep in mind that there is insufficient clinical experience with these combinations. Finally, when combinations cannot be employed, monotherapy with a newer fluoroquinolone, clindamycin, trimethoprim/sulfamethoxazole, pristinamycin or linezolid can be tried as a last resort. Funding This work was supported by the 2013 Inje University research grant (B.-N. K.). Transparency declarations None to declare. References 1 Lew DP, Waldvogel FA. Osteomyelitis. Lancet 2004; 364: 369 79. 2 Zimmerli W, Trampuz A, Ochsner PE. Prosthetic-joint infections. N Engl J Med 2004; 351: 1645 54. 3 Mathews CJ, Weston VC, Jones A et al. Bacterial septic arthritis in adults. Lancet 2010; 375: 846 55. 4 Widmer AF. New developments in diagnosis and treatment of infection in orthopedic implants. Clin Infect Dis 2001; 33 Suppl 2: S94 106. 5 Stengel D, Bauwens K, Sehouli J et al. Systematic review and meta-analysis of antibiotic therapy for bone and joint infections. Lancet Infect Dis 2001; 1: 175 88. 6 Lazzarini L, Lipsky BA, Mader JT. Antibiotic treatment of osteomyelitis: what have we learned from 30 years of clinical trials? Int J Infect Dis 2005; 9: 127 38. 7 Conterno LO, da Silva Filho CR. Antibiotics for treating chronic osteomyelitis in adults. Cochrane Database Syst Rev 2009; issue 3: CD004439. 8 Darley ES, MacGowan AP. Antibiotic treatment of Gram-positive bone and joint infections. J Antimicrob Chemother 2004; 53: 928 35. 9 Trampuz A, Zimmerli W. Antimicrobial agents in orthopaedic surgery: prophylaxis and treatment. Drugs 2006; 66: 1089 105. 10 Trampuz A, Zimmerli W. Diagnosis and treatment of infections associated with fracture-fixation devices. Injury 2006; 37 Suppl 2: S59 66. 11 Zimmerli W. Clinical practice. Vertebral osteomyelitis. N Engl J Med 2010; 362: 1022 9. 12 Shirtliff ME, Mader JT. Acute septic arthritis. Clin Microbiol Rev 2002; 15: 527 44. 13 Gouliouris T, Aliyu SH, Brown NM. Spondylodiscitis: update on diagnosis and management. J Antimicrob Chemother 2010; 65 Suppl 3: iii11 24. 14 Osmon DR, Berbari EF, Berendt AR et al. Diagnosis and management of prosthetic joint infection: clinical practice guidelines by the Infectious Diseases Society of America. Clin Infect Dis 2013; 56: e1 25. 15 Zimmerli W, Widmer AF, Blatter M et al. Role of rifampin for treatment of orthopedic implant-related staphylococcal infections: a randomized controlled trial. Foreign-Body Infection (FBI) Study Group. JAMA 1998; 279: 1537 41. 16 Darley ES, Bannister GC, Blom AW et al. Role of early intravenous to oral antibiotic switch therapy in the management of prosthetic hip infection treated with one- or two-stage replacement. J Antimicrob Chemother 2011; 66: 2405 8. 17 Paakkonen M, Peltola H. Antibiotic treatment for acute haematogenous osteomyelitis of childhood: moving towards shorter courses and oral administration. Int J Antimicrob Agents 2011; 38: 273 80. 18 Le Saux N, Howard A, Barrowman NJ et al. Shorter courses of parenteral antibiotic therapy do not appear to influence response rates for children with acute hematogenous osteomyelitis: a systematic review. BMC Infect Dis 2002; 2: 16. 19 Calhoun JH, Manring MM. Adult osteomyelitis. Infect Dis Clin North Am 2005; 19: 765 86. 20 Costerton JW, Stewart PS, Greenberg EP. Bacterial biofilms: a common cause of persistent infections. Science 1999; 284: 1318 22. 21 Sampedro MF, Patel R. Infections associated with long-term prosthetic devices. Infect Dis Clin North Am 2007; 21: 785 819. 22 Widmer AF, Frei R, Rajacic Z et al. Correlation between in vivo and in vitro efficacy of antimicrobial agents against foreign body infections. J Infect Dis 1990; 162: 96 102. 23 Sendi P, Zimmerli W. Antimicrobial treatment concepts for orthopaedic device-related infection. Clin Microbiol Infect 2012; 18: 1176 84. 24 Landersdorfer CB, Bulitta JB, Kinzig M et al. Penetration of antibacterials into bone: pharmacokinetic, pharmacodynamic and bioanalytical considerations. Clin Pharmacokinet 2009; 48: 89 124. 25 Ferry T, Uckay I, Vaudaux P et al. Risk factors for treatment failure in orthopedic device-related methicillin-resistant Staphylococcus aureus infection. Eur J Clin Microbiol Infect Dis 2010; 29: 171 80. 26 Zimmerli W, Frei R, Widmer AF et al. Microbiological tests to predict treatment outcome in experimental device-related infections due to Staphylococcus aureus. J Antimicrob Chemother 1994; 33: 959 67. 318