¶ Executive Summary:
Clear Recommendation:
Based on eight concordant RCTs comparing intravenous (IV) to oral therapy(Greenberg, Tice et al. 1987, Gentry and Rodriguez 1990, Mader, Cantrell et al. 1990, Gentry and Rodriguez-Gomez 1991, Gomis, Barberan et al. 1999, Schrenzel, Harbarth et al. 2004, Euba, Murillo et al. 2009, Li, Rombach et al. 2019) (Figure 3) and nine RCTs in which oral therapy was predominantly used in both arms,(Lipsky, Baker et al. 1997, Lipsky, Itani et al. 2004, Lazaro-Martinez, Aragon-Sanchez et al. 2014, Bernard, Dinh et al. 2015, Tone, Nguyen et al. 2015, Lora-Tamayo, Euba et al. 2016, Benkabouche, Racloz et al. 2019, Gariani, Pham et al. 2020, Bernard, Arvieux et al. 2021) we recommend oral antibiotic therapy with a drug/dose used in published studies as a reasonable option for osteomyelitis of any type (i.e., hematogenous, prosthetic, and contiguous, the latter including vertebral and DFO) for patients who:
- Are clinically stable (hemodynamically and at the site of infection, e.g., no spinal instability);
- Have adequate source control (i.e., not requiring further procedural drainage and without persistent bacteremia);
- Are likely to absorb oral medications from a functioning gastrointestinal (GI) tract;
- Have an available regimen used in published osteomyelitis studies to cover likely target pathogens; and
- Have no psychosocial reasons that preclude the safe use of oral therapy.
There is no required minimum duration of IV lead-in; patients may be switched to oral therapy when all the above criteria are met, even at the empiric therapy stage. Specific drug options and doses are discussed in the detailed review section (Tables 4 and 5 ).
¶ Overall Summary:
¶ Observational Data for Osteomyelitis Including DFO and PJI
The historical basis of requiring IV therapy for the treatment of osteomyelitis was the relatively poor outcomes achieved with parenteral penicillin and aminoglycosides in the 1940s and 1950s.(Waldvogel, Medoff et al. 1970) During that early period in the history of antibiotics, the only drugs that were available for oral administration had limited bioavailability and/or spectra of activity (e.g., sulfanilamide, erythromycin, tetracycline). By the time advanced oral formulations of penicillin derivatives and other modern antibiotics became available in the late 1950s and 1960s, medicine had already long adopted a traditional, non-evidence-based culture of requiring IV-only therapy for osteomyelitis.
However, extensive pharmacology studies in the modern era have demonstrated that numerous oral antibacterial agents, including clindamycin, the fluoroquinolones, fosfomycin, fusidic acid, linezolid, metronidazole, rifampin, and TMP-SMX, achieve levels in bone with standard oral dosing that are well above the MICs of susceptible pathogens (as discussed in Section 4d). Oral amoxicillin may also penetrate into bone to achieve peak levels well above the MICs for sensitive gram-positive pathogens, at least when administered at a 1-2 g dose.(Thabit, Fatani et al. 2019) The data for other oral b lactams and tetracyclines are less clear regarding the reliability of exceeding target MICs in bone. However, clinical experience, discussed below, suggests that these drugs may be effective in some cases.
Concordant with the pharmacology data, more than 40 observational studies have demonstrated that oral administration of antibiotics resulted in treatment success rates for osteomyelitis similar to those historically experienced with IV therapy (Table 4).(Greenberg, Newman et al. 2000, Senneville, Poissy et al. 2007, Spellberg and Lipsky 2012, Joel, Graham et al. 2014, Bonnaire, Vernet-Garnier et al. 2021, Suzuki, Mosher et al. 2022) These studies evaluated a wide variety of patients with adult osteomyelitis, including long bone, vertebral, skull based, DFO, and PJI.
The study drugs in the majority of these reports were fluoroquinolones, with or without adjunctive rifampin.(Giamarellou and Galanakis 1987, Gilbert, Tice et al. 1987, Greenberg, Kennedy et al. 1987, Hessen, Ingerman et al. 1987, Lesse, Freer et al. 1987, Nix, Cumbo et al. 1987, Scully and Neu 1987, 1988, Ketterl, Beckurts et al. 1988, Dellamonica, Bernard et al. 1989, Peterson, Lissack et al. 1989, Drancourt, Stein et al. 1993, Drancourt, Stein et al. 1997, Greenberg, Newman et al. 2000, Senneville, Poissy et al. 2007, Barberan, Aguilar et al. 2008) However, second in frequency of study has been TMP-SMX, which has been shown in at least seven observational studies of adults to be associated with excellent treatment success rates, again with or without adjunctive rifampin.(Saengnipanthkul, Pongvivat et al. 1988, de Barros, Calapodopulos et al. 1992, Sanchez, Matamala et al. 1997, Stein, Bataille et al. 1998, Javaloyas de Morlius and Monreal Portella 1999, Nguyen, Pasquet et al. 2009, Deconinck, Dinh et al. 2019) While the majority of these infections were caused by staphylococci, outcomes were similar in the remaining infections caused by streptococci or gram-negative bacilli. TMP-SMX has also been shown to be safe and effective for treating osteomyelitis in children.(Messina, Namtu et al. 2011) Linezolid, clindamycin, and fosfomycin (the latter with an oral formulation available outside of North America; not studied with the sachet powder formulation) have also been studied as oral therapy, with reasonably high cure rates,(Pontifex and McNaught 1973, Fernandez-Valencia, Saban et al. 1976, Meissner, Haag et al. 1989, Birmingham, Rayner et al. 2003, Joel, Graham et al. 2014, Courjon, Demonchy et al. 2017, Theil, Schmidt-Braekling et al. 2020, Bonnaire, Vernet-Garnier et al. 2021) and clindamycin has been studied widely in children, including in an RCT,(Peltola, Paakkonen et al. 2012) with favorable outcomes.
¶ Randomized Controlled Trials of Osteomyelitis Including DFO and PJI
Concordant with the observational data, eight RCTs in adults have unanimously demonstrated similar efficacy outcomes of oral vs. IV therapy in more than 1,300 patients with osteomyelitis, including with vertebral or long bone osteomyelitis and PJI (Figure 3).(Greenberg, Tice et al. 1987, Gentry and Rodriguez 1990, Mader, Cantrell et al. 1990, Gentry and Rodriguez-Gomez 1991, Gomis, Barberan et al. 1999, Schrenzel, Harbarth et al. 2004, Euba, Murillo et al. 2009, Li, Rombach et al. 2019, Wald-Dickler, Holtom et al. 2021) Furthermore, in the largest RCT, patients receiving oral therapy reported better mobility, self-care, and activity levels, and less pain, discomfort, anxiety, and depression than patients receiving IV therapy.(Li, Rombach et al. 2019) Finally, cost was reduced by more than $3,500 for oral vs. IV therapy. Four of these RCTs explicitly included orthopaedic implants/PJIs, which constituted more than half of the enrolled population in the largest RCT.(Gentry and Rodriguez 1990, Mader, Cantrell et al. 1990, Euba, Murillo et al. 2009, Li, Rombach et al. 2019) Vertebral osteomyelitis was also included in the largest RCT, comprising approximately 10% of the enrolled patients.(Li, Rombach et al. 2019)
In addition, nine RCTs have been published in which oral antibiotics constituted the large majority of therapy in both arms for the treatment of osteomyelitis, with excellent outcomes.(Lipsky, Baker et al. 1997, Lipsky, Itani et al. 2004, Lazaro-Martinez, Aragon-Sanchez et al. 2014, Bernard, Dinh et al. 2015, Tone, Nguyen et al. 2015, Lora-Tamayo, Euba et al. 2016, Benkabouche, Racloz et al. 2019, Gariani, Pham et al. 2020, Bernard, Arvieux et al. 2021) These RCTs compared different durations of oral therapy or different oral antibiotic agents, and included patients with vertebral osteomyelitis, DFO, and PJI, with only short IV-lead in periods before patients were switched to oral therapy.
For example, in an RCT comparing 6 vs. 12 weeks of therapy for 359 patients with vertebral osteomyelitis, more than 90% were treated with oral antibiotics in both the 6- and 12-week arms.(Bernard, Dinh et al. 2015) Treatment success rates were 91% in both arms and did not differ between patients treated with <1 week or >1 week of IV lead-in antibiotics.
In an RCT of patients with PJI, <1 week of other antibiotics were administered before patients were switched to oral levofloxacin at 750 mg plus rifampin 600 mg once daily. The per protocol (n = 44) treatment success rates were 92% and 95% in the short vs. long therapy arm at a median of 862 days of follow-up.(Lora-Tamayo, Euba et al. 2016) In a second RCT of PJI involving 123 patients, a mean of only four days of IV antibiotic lead-in was administered before patients were switched to oral therapy, which consisted of a variety of antibiotics, including a relatively even blend of quinolones, clindamycin, doxycycline, amoxicillin-clavulanate, and TMP-SMX.(Benkabouche, Racloz et al. 2019) Treatment success rates were 94% and 95% in the 4 vs. 6 week therapy arms. In a third, large RCT of 384 patients with PJI, the median duration of IV therapy administered was only nine days, with the remainder of therapy of 6 vs. 12 weeks of therapy being administered orally, with a wide variety of agents.(Bernard, Arvieux et al. 2021) Treatment success rates were 83% and 93% in the 6- vs. 12-week arm at a median of two years of follow-up. Thus, not only do RCTs comparing outcomes of oral vs. IV therapy confirm oral efficacy with PJIs with little IV lead-in, but very high rates of treatment success have been seen in RCTs comparing durations of therapy predominantly administered orally.
Perhaps because of the long-standing, general acceptability of oral step-down therapy for DFO (dating back to the early to mid-1980s in observational studies), there have not been published RCTs comparing IV-only to oral therapy specifically for DFO. However, there have been RCTs that compared various oral regimens in both arms for treating DFO. Amoxicillin-clavulanate, clindamycin, or cephalosporins or fluoroquinolones with or without metronidazole or rifampin have been typically used in such studies, as described below.
For example, in 1997, Lipsky et al. randomized 88 evaluable patients with DFI to treatment with IV to oral step-down therapy with ofloxacin vs. ampicillin-sulbactam followed by amoxillin-clavulanate.(Lipsky, Baker et al. 1997) Twenty-six (30%) of these patients had osteomyelitis. The mean duration of IV therapy was approximately one week, followed by a mean of approximately two weeks of oral therapy for both arms. There was no significant difference in long-term treatment success between the two arms (85% vs. 83%), indicating that both options were reasonable.
In 2004, Lipsky et al. conducted another RCT comparing linezolid vs. ampicillin/sulbactam or amoxicillin-clavulanate in patients with DFI.(Lipsky, Itani et al. 2004) Treating physicians were able to decide the route (oral or IV) of administration, and this could change according to their judgment over the treatment course.(Lipsky, Itani et al. 2004) Of 361 study patients, 77 (21%) had osteomyelitis. The mean durations of treatment were 8 vs. 10 days IV, and 17 vs. 17 days total, for linezolid vs. the b lactam arms. The success rates for the osteomyelitis patients were 61% (27/44) vs. 69% (11/16) for the linezolid vs. b lactam arms.
In 2014, Lazaro-Martinez et al. randomized 46 patients with DFO to treatment with 90 days of antibiotics without surgical management vs. conservative surgical debridement (defined as removal of infected bone without amputation) plus 10 days of antibiotic therapy.(Lazaro-Martinez, Aragon-Sanchez et al. 2014) All antibiotics were administered orally, consisting of twice daily ciprofloxacin 500 mg, amoxicillin-clavulanate 875/125 mg, or TMP-SMX 160 mg/800 mg. Overall outcomes were excellent, as treatment success was achieved in 86% (19/22) of patients treated with 10 days of antibiotics vs. 75% (18/24) of patients treated with 90 days of antibiotics. No difference in outcome was described by type of antibiotic.
Finally, two RCTs of longer vs. shorter therapy for patients with DFO were based on the use of oral therapy, with or without IV lead-in, with a wide variety of agents, including amoxicillin-clavulanate, quinolones, TMP-SMX, doxycycline, linezolid, with or without adjunctive rifampin.(Tone, Nguyen et al. 2015, Gariani, Pham et al. 2020) Treatment outcomes were good in both arms of both studies, with overall 39/64 (61%) and 50/69 (72%) patients respectively achieving successful clinical outcomes.
Particularly striking is the absence of any contrary published data, whether observational or RCT, that demonstrates superior outcomes with IV therapy.(Wald-Dickler, Holtom et al. 2021) The numerous studies of oral therapy cannot necessarily encompass every conceivable iteration of osteomyelitis (body site, organism, resistance profile, type of foreign implant, patient factors, etc). Nevertheless, the data demonstrating safety and efficacy of oral therapy are sufficiently robust and concordant across numerous patient and disease types, including vertebral osteomyelitis, DFO, and PJI, that combined with the absence of any data indicating superior outcomes with IV therapy, they provide a reasonable basis for determining that oral therapy is a generally acceptable option when patients become clinically stable and can tolerate oral medications.
¶ Published Antibiotic Selection
As for the observational data, the majority of the RCTs comparing oral to IV therapy studied fluoroquinolones with or without rifampin therapy, while one trial also specifically studied TMP-SMX plus rifampin.(Euba, Murillo et al. 2009) However, as mentioned, in other RCTs comparing various oral regimens, myriad antibiotic types were administered. Overall, fluoroquinolones and TMP-SMX have the most published data. Clindamycin is also reasonable to consider based on observational studies, its inclusion in the largest oral vs. IV RCT as an option, its use in studies comparing different oral regimens for PJI, and extensive pediatric data.(Li, Rombach et al. 2019) Metronidazole is also a reasonable option for anaerobic coverage based on published observational data.(Spellberg and Lipsky 2012)
Unfortunately, there is a high rate of relapse (in some studies more than 50%), typically with emergence to resistance, among osteomyelitis caused by staphylococci treated with fluoroquinolone monotherapy.(Zimmerli, Widmer et al. 1998, Greenberg, Newman et al. 2000, Spellberg and Lipsky 2012) Thus, it seems prudent to avoid fluoroquinolone monotherapy in treating staphylococcal osteomyelitis. An alternative agent (e.g., TMP-SMX, clindamycin, linezolid), or addition of rifampin to fluoroquinolone therapy, are reasonable options.
Furthermore, fluoroquinolones may be less desirable than other available agents due to increasing reports of various potentially serious toxicities,(2019, Kuula, Viljemaa et al. 2019) and the need to preserve them as oral step-down agents for broad gram-negative bacterial coverage. Thus, alternative options for gram-positive bacteria may often be preferred and, if the etiologic organism is likely susceptible to TMP-SMX, the latter may be a reasonable option for gram-negative osteomyelitis as well. Of course, TMP-SMX has also been associated with a variety of potentially serious adverse events (e.g., allergic reactions, hepatitis, hyperkalemia, etc.).(Ho and Juurlink 2011)
Both linezolid and metronidazole (more commonly the former) may cause irreversible neuropathies after >4 weeks of therapy, and these agents should generally be avoided for prolonged therapeutic periods if other agents or safer treatment options are available. If these agents are to be administered for more than 2-3 weeks, patients should be counseled regarding these potential side effects. Hematologic side effects of linezolid should be monitored for after two weeks of therapy.
There is less published experience with most oral b lactams, doxycycline, fosfomycin, and fusidic acid for the treatment of osteomyelitis. Amoxicillin-clavulanate is an exception; it has been widely used in the treatment of DFO, including in RCTs, as described above. Fosfomycin (again not studied with the sachet powder formulation available in North America) and fusidic acid have favorable bone PK and resulted in favorable outcomes in several published observational studies (Table 4). Furthermore, doxycycline is standard of care for treatment of some forms of atypical osteomyelitis, such as those caused by Coxiella burnetii and Brucella spp.(Pourbagher, Pourbagher et al. 2006, Colmenero, Ruiz-Mesa et al. 2008, Ghanem-Zoubi, Karram et al. 2021) While this experience may not be directly extrapolatable to pyogenic osteomyelitis, it does support the concept that doxycycline can get into bone adequately to cure infection.
Pharmacological considerations may be less favorable for most b lactams (excepting amoxicillin) and doxycycline. Nevertheless, in the largest RCT, oral penicillins and doxycycline were each administered to more than 10% of patients, and outcomes were not described to differ in these patients.(Li, Rombach et al. 2019) Thus, limited data for oral b lactams, doxycycline, fosfomycin, and fusidic may suggest potential usefulness. Such data are not yet robust enough to enable a recommendation for or against their use, except for amoxicillin-clavulanate, which has resulted in high success rates in several RCTs for DFO. Nevertheless, given their spectra of activity, high potency against susceptible pathogens, and limited data in an RCT, other b lactam agents may be considered for individual patients.
All these oral options, including linezolid, are generic and relatively inexpensive. Hence, cost is generally not a relevant factor for selection among them (while it is decidedly a relevant factor favoring oral options over IV agents).
¶ Published Antibiotic Dosing (Table 5)
In osteomyelitis studies, doses of ciprofloxacin have typically ranged from 500 to 750 mg twice daily (the latter especially for Pseudomonas).(Giamarellou and Galanakis 1987, Gilbert, Tice et al. 1987, Greenberg, Kennedy et al. 1987, Greenberg, Tice et al. 1987, Hessen, Ingerman et al. 1987, Lesse, Freer et al. 1987, Nix, Cumbo et al. 1987, Scully and Neu 1987, Peterson, Lissack et al. 1989, Gentry and Rodriguez 1990, Mader, Cantrell et al. 1990, Greenberg, Newman et al. 2000) Many of the fluoroquinolone studies were of ofloxacin, which has been replaced clinically by its active L-enantiomer, levofloxacin. A dose of 750 mg of levofloxacin once daily is reasonable for osteomyelitis.(Senneville, Poissy et al. 2007, El Helou, Berbari et al. 2010, Lora-Tamayo, Euba et al. 2016) TMP-SMX has generally been studied at approximately 7.5-10 mg/kg per day of trimethoprim in divided doses (e.g., two DS tablets twice daily for a 70 kg adult).(Sanchez, Matamala et al. 1997, Stein, Bataille et al. 1998, Euba, Murillo et al. 2009, Nguyen, Pasquet et al. 2009, Lazaro-Martinez, Aragon-Sanchez et al. 2014) It has not been established that these higher doses of TMP-SMX are necessary to affect cure; in two studies doses of 4-6 mg/kg per day of TMP-SMX were administered,(Saengnipanthkul, Pongvivat et al. 1988, Deconinck, Dinh et al. 2019) but cure rates were notably lower in one of them.(Saengnipanthkul, Pongvivat et al. 1988) Attention to renal function and potassium are important with higher doses of TMP-SMX and concomitant exposure to other agents which are potassium-sparing (e.g., angiotensin-converting enzyme inhibitors and angiotensin receptor blockers) should be evaluated and considered.(Fralick, Macdonald et al. 2014)
General dosing of oral clindamycin has been 600 mg thrice daily, with escalation to 900 mg thrice daily or 600 mg four times daily in larger patients.(Bouazza, Pestre et al. 2012, Courjon, Demonchy et al. 2017, Bonnaire, Vernet-Garnier et al. 2021) Linezolid has been studied at 600 mg twice daily.(Birmingham, Rayner et al. 2003, Lipsky, Itani et al. 2004, Nguyen, Pasquet et al. 2009, Joel, Graham et al. 2014) If rifampin is to be administered, 600 mg once per day may be preferred to 300 mg twice daily or 450 mg once daily both due to dramatically superior AUC PK,(Acocella 1983, Peloquin, Jaresko et al. 1997, Ruslami, Nijland et al. 2007, Chirehwa, Rustomjee et al. 2016) the fact that the PK/PD driver that best correlates with the drug’s antimicrobial activity is total AUC/MIC,(Jayaram, Gaonkar et al. 2003, Hirai, Hagihara et al. 2016, Stott, Pertinez et al. 2018) and the simpler regimen of once vs. twice daily (discussed at length in Section 4e). Whether dosing rifampin at higher levels (900-1,200 mg per day in divided doses) alters efficacy is unclear, although it may increase toxicity.
As discussed in question 4, section 3, caution should be taken if rifampin is to be administered with linezolid, due to a substantial pharmacokinetic interaction, lowering linezolid levels, which may be associated with a higher clinical failure rate.(Gandelman, Zhu et al. 2011, Tornero, Morata et al. 2016, Okazaki, Tsuji et al. 2019) Similarly, co-administration with rifampin may lower clindamycin and fusidic acid, although there is no evidence that clinical failures are more likely with these pairings.(Aboltins, Page et al. 2007, Peel, Buising et al. 2013, Bernard, Kermarrec et al. 2015, Curis, Pestre et al. 2015, Pushkin, Iglesias-Ussel et al. 2016, Zeller, Magreault et al. 2021)
Amoxicillin-clavulanate has been dosed at either 500/125 mg thrice daily(Lipsky, Baker et al. 1997, Lipsky, Itani et al. 2004) or 875/125 mg twice daily(Lipsky, Itani et al. 2004, Lazaro-Martinez, Aragon-Sanchez et al. 2014) in individual studies for DFO (and some studies gave the option of either(Lipsky, Itani et al. 2004)), with no apparent distinction in outcomes. There are few published data for amoxicillin-clavulanate for osteomyelitis outside the context of DFO.
Of note, dosing of antibiotics should be adjusted for renal function or other clinical factors based on prescribing recommendations for each drug.
¶ Patient Selection Criteria for Oral Therapy
Reasonable clinical criteria can be applied to select patients eligible for oral therapy. The duration of IV therapy prior to initiation of oral therapy has varied in the RCTs. In some studies, no IV lead-in was administered per the study protocol.(Euba, Murillo et al. 2009, Lazaro-Martinez, Aragon-Sanchez et al. 2014, Tone, Nguyen et al. 2015, Gariani, Pham et al. 2020) In the largest studies, a mean of only 9 or 10 days of IV therapy were administered before switching to oral agents for multiple subsequent weeks of therapy.(Li, Rombach et al. 2019, Bernard, Arvieux et al. 2021) Thus, the cumulative data do not indicate that it is necessary to begin with IV therapy, nor for how long to administer it, before switching to oral therapy.
Nevertheless, patients who are clinically unstable (e.g., hemodynamically unstable, spinal instability, etc) should generally receive IV therapy, due to concerns about the ability to administer and absorb oral regimens and the desire to achieve more rapid therapeutic levels. Patients who will require procedural source-control typically require inpatient care, and often require withholding oral intake to prevent aspiration during procedures. Therefore, these patients are likely better suited to receive IV therapy. While RCTs have demonstrated efficacy of oral therapy for bacteremia and endocarditis,(Spellberg, Chambers et al. 2020, Wald-Dickler, Holtom et al. 2021) persistence of bacteremia on therapy portends a poor prognosis, may indicate source control is needed, and likely necessitates ongoing inpatient care to ensure bacteria clear from the blood. Hence, IV therapy may be preferred until clearance of bacteremia. Finally, there may be psychosocial reasons why IV therapy is preferred in individual patients. For example, patients who are unlikely to be willing to take oral therapy, and/or who are otherwise likely to be in need of skilled nursing care, may benefit from IV therapy to help justify the higher level of care.
Collectively, therefore, it is reasonable to consider administration of oral antibiotics for the treatment of osteomyelitis when the patient meets all of the below criteria(Spellberg, Chambers et al. 2020, Wald-Dickler, Holtom et al. 2021):
- Clinically stable (e.g., hemodynamically stable, no spinal instability, etc);
- Does not require (or no longer requires) procedural source control and without persistent bacteremia;
- Is likely to absorb oral medications from a functioning GI tract;
- There is an available regimen used in published studies to cover likely target pathogen(s); and
- There are not psychosocial (e.g., need to justify specific levels of care, adherence concerns, etc.) reasons that preclude the safe use of oral therapy
¶ Tables and Figures:
¶ RCTs comparing long-term success oral vs. IV (Figure 3)
¶ Treatment Success Rates in Observational Studies of Oral Treatment (Table 4)
|
Drug |
Dose/Duration |
Follow up |
Cure* |
Comment |
Ref. |
||
| Fluoroquinolones | |||||||
| Ciprofloxacin |
500-750 mg PO bid x 3-4 months |
1 yr |
81% (30/37) |
All cured patients had foreign material removed; 1/3 underwent debridement |
(Nix, Cumbo et al. 1987) |
||
| Ciprofloxacin |
750 mg PO bid x 3-4 months |
6 mo |
91% (21/23) |
Cure defined as resolved or improved |
(Lesse, Freer et al. 1987) |
||
| Ciprofloxacin |
750 mg PO bid x 3 months |
7-21 mo |
65% (13/20) |
15/20 previously failed therapy; 3 patients with sternal osteomyelitis; cured only 7/13 Pseudomonas; all debrided |
(Gilbert, Tice et al. 1987) |
||
| Ciprofloxacin |
750 mg PO bid x 2-4 months |
1-17 mo |
77% (17/22) |
4 of the non-cured infected with Pseudomonas; 20 debrided |
(Hessen, Ingerman et al. 1987) |
||
| Ciprofloxacin |
750 mg PO bid x 1-6 months |
0-22 mo |
48% (14/29) |
7/12 Pseudomonas & 4/9 S. aureus cured |
(Greenberg, Kennedy et al. 1987) |
||
|
Ciprofloxacin or Nafcillin, Clindamycin, or Gentamicin |
750 mg PO bid x 12-64 d
varying dose & durations |
25-39 mo |
11/14 (79%) ciprofloxacin vs. 10/12 (83%) IV therapy |
Not randomized; patients were sequentially enrolled in the two arms |
(Mader, Cantrell et al. 1990) |
||
| Ciprofloxacin |
200 mg IV bid, then 750 mg PO bid |
? |
67% (6/9) |
Unknown duration of treatment; 5/7 Pseudomonas cured |
(Giamarellou and Galanakis 1987) |
||
| Ciprofloxacin |
200 mg IV bid, then 750 mg PO bid |
? |
83% (10/12) |
Unknown duration of treatment |
(Scully and Neu 1987) |
||
| Ciprofloxacin |
500-1500 mg PO bid x 0.5-18 months |
? |
65% (22/34) |
20/28 Pseudomonas eradicated microbiologically |
(1988) |
||
|
Pefloxacin
Ofloxacin
Ciprofloxacin |
400 mg IV q 12 h x 4 doses, then 400 mg PO q 12 h 200 mg PO q8-12 h 500-750 mg PO q 12 h All for 3-6 mo |
? |
76% (29/38) |
All cured patients had foreign material removed; 1/3 underwent debridement; 88% (15/17) treatment success for gram-negative bacteria vs. 67% (14/21) for gram-positive |
(Dellamonica, Bernard et al. 1989) |
||
| Ofloxacin |
200 mg PO tid x 4-6 weeks |
>6 mo |
85% (98/115) |
3/15 Pseudomonas and 5/74 S. aureus failed; 113 debrided |
(Ketterl, Beckurts et al. 1988) |
||
| Ciprofloxacin |
750-1000 mg PO bid x 3 mos |
12 mo |
61% (19/31) |
No benefit from higher dose; all had soft tissue, but not bone, debrided |
(Peterson, Lissack et al. 1989)† |
||
|
Ciprofloxacin
Lomefloxacin Levofloxacin |
750 BID
800 BID 500 once daily |
Variable, most > 1 year |
2/5 (40%)
5/7 (71%) 9/15 (60%) |
6 patients infected with S. aureus and 1 Pseudomonas relapsed |
(Greenberg, Newman et al. 2000) |
||
|
Ofloxacin + Rifampin |
200 mg PO tid
300 mg PO tid both for 6-9 mo |
> 60 mo |
35/49 (71%) |
All infections of prostheses with Staphylococcus |
(Drancourt, Stein et al. 1993) |
||
|
Levofloxacin + Rifampin |
500 mg PO qd
600 mg PO qd both for >6 weeks |
> 6 mo |
18/25 (72%) |
All had prosthetic bone implants; mean duration of therapy 5 months for those cured and 2.6 months for those who failed to be cured |
(Barberan, Aguilar et al. 2008) |
||
|
Rifampin + (Ofloxacin or Fusidic acid) |
900 mg PO qd
200 mg PO tid
500 mg PO tid x 5 d, then PO bid, both for >6 mo |
Average 24 mos (range 12-36 mo) |
11/20 (55%)
11/22 (50%) |
All patients had orthopaedic implants, only 14 of which were removed; patients were assigned to treatment arm by year of birth (ofloxacin for even years, fusidic acid for odd years) |
(Drancourt, Stein et al. 1997) |
||
|
Rifampin + Fluoro-quinolone vs. Other |
When used, rifampin at 20 mg/kg divided bid (not to exceed 1800 mg/d) |
Average 44 +/- 32 mos |
37/39 (98%)
vs.
40/59 (68%) |
All had S. aureus prosthetic joint infections; 29 patients received rifampin in combination with non-quinolone antibiotics; in multi-variate analysis rifampin-quinolone combination had an odds ratio of 0.4 (0.17-0.97) for failure |
(Senneville, Joulie et al. 2011) |
||
|
Rifampin + Levofloxacin (prospective) vs. Historical cohort with variable antibiotics without vs. with Rifampin |
Prospective rifampin at 900 mg PO qd x 3-6 mos |
? |
13/14 (93%) vs. 34/56 (63%) vs. 21/31 (68%) |
All had retained prosthetic joints; by multivariate analysis, hazard ratio for treatment failure 1.0 for historical cohort without rifampin, 0.55 (0.25-1.26) for historical cohort with rifampin, 0.11 (0.01-0.84) for prospective rifampin cohort, p = 0.03. |
(El Helou, Berbari et al. 2010) |
||
|
Other Agents |
|||||||
|
Rifampin + Various other antibiotics |
600 mg PO qd x 6 mos |
Variable |
50% (7/14) |
All cases refractory to prior therapy |
(Norden, Fierer et al. 1983) |
||
| Linezolid |
600 mg PO q12 h |
? |
60% (45/89) |
Compassionate use program |
(Birmingham, Rayner et al. 2003) |
||
| Linezolid |
600 mg PO bid |
Variable |
77% (17/22) |
Post arthroplasty (10), ortho trauma (8), other (4) |
(Joel, Graham et al. 2014) |
||
| Clindamycin |
50-150 mg PO q 6 h x mean 16 weeks |
Variable |
42% (5/12) |
|
(Pontifex and McNaught 1973) |
||
| Clindamycin |
600 mg tid |
1 year |
67% (31/46) |
Combined with rifampin (37), fusidic acid (4), or fluoroquinolone (4), including 40% of patients with prosthetic infections |
(Bonnaire, Vernet-Garnier et al. 2021) |
||
| Clindamycin |
600 mg tid or qid by body weight +/- other antibiotics |
3-6 weeks |
83% (111/133) |
Clindamycin alone (31), with rifampin (27), levofloxacin (61), other (51) |
(Courjon, Demonchy et al. 2017) |
||
| TMP-SMX |
1-2 DS tab PO bid |
? |
83% (5/6) |
None had debridement |
(Craven, Pugsley et al. 1970) |
||
| TMP-SMX |
1 DS tab PO bid x 4-8 weeks |
11-70 mos |
45% (30/66) |
55% of patients had debridement |
(Saengnipanthkul, Pongvivat et al. 1988) |
||
|
TMP-SMX + Rifampin |
3.5 mg/kg (TMP) PO bid 600-1200 mg PO qd both x mean 5 weeks |
6 mo to 5 yrs |
100% (27/27) |
All patients had debridement |
(Sanchez, Matamala et al. 1997) |
||
|
TMP-SMX +/- Rifampin |
DS PO BID
300-450 mg PO bid both for median 10 weeks |
2 years |
82% (28/34) |
10 patients had debridement, all of whom were cured |
(Javaloyas de Morlius and Monreal Portella 1999) |
||
| TMP-SMX |
5 mg/kg (TMP) PO bid x 6-9 mos |
24-75 mos |
67% (26/39) |
11 patients had device removed |
(Stein, Bataille et al. 1998) |
||
| TMP-SMX |
Dose unclear, treated for 6 mos |
12-60 mos |
98% (59/60) |
All patients had debridement |
(de Barros, Calapodopulos et al. 1992) |
||
| TMP-SMX |
4-6 mg/kg (TMP) PO |
6-7 wks |
78% (40/51) |
76% with prosthetic infections, 47% caused by gram-negative bacteria |
(Deconinck, Dinh et al. 2019) |
||
|
(TMP-SMX or Linezolid) + Rifampin |
8 mg/kg (TMP) PO 600 mg bid
10 mg/kg bid all given iv x 1 week and then oral |
≥12 mos |
89% (37/41)
79% (29/38) |
20 patients with chronic osteomyelitis and 56 with orthopaedic implant infections; mean (range) treatment durations were 15 (1-53) weeks for TMP-SMX based therapy and 18 (8-36 weeks) for linezolid-based therapy; adverse event rates similar (46% vs. 43%), discontinuation rates similar (14% vs. 21%) |
(Nguyen, Pasquet et al. 2009) |
||
| Fosfomycin |
10g x 1, then 5 g tid |
5-28 days |
47% (29/60) |
Outcome defined as “very good”, mean 37 month follow up |
(Meissner, Haag et al. 1989) |
||
| Fosfomycin |
4 to 8 g per day |
IV or PO |
29/37 (78%) |
23 debrided |
(Fernandez-Valencia, Saban et al. 1976) |
||
| Fosfomycin |
8 to 16 g IV, then 2-4 g PO per day |
IV or PO |
99/99 (100%) |
39 debrided, started IV or IM, then transitioned to oral |
(Hernandez Casado 1977) |
||
| Fusidic acid |
Varied |
PO, varied |
73/80 (91%)‡ |
Review of numerous case reports and small case series |
(Wang, Tang et al. 2012) |
||
| Fusidic acid |
20 mg/kg |
PO |
19/20 (95%) |
15 received other antibiotics with fusidic acid, 5 fusidic acid alone |
(Coombs and Menday 1985) |
||
|
Diabetic Foot Osteomyelitis |
|||||||
| Clindamycin, Amoxicillin/ clavulanate, Metronidazole, Fusidic acid, Ciprofloxacin, |
Oral with some IV lead in |
Varied |
17/22 (77%) |
Varied treatment, varied durations |
(Venkatesan, Lawn et al. 1997) |
||
| Floxacillin, Amoxicillin/ clavulanate, Cephalo-sporins, Flouro-quinolones, Clindamycin, Metroniazole |
Varied |
Oral with IV lead in |
35/50 (70%) |
Treatment with a mean of 3 weeks IV followed by 6 weeks oral |
(Pittet, Wyssa et al. 1999) |
||
| Ofloxacin + Rifampin |
200 mg PO tid + 600 mg PO bid |
Oral |
13/17 (76%) |
Treated for 3 to 10 months |
(Senneville, Yazdanpanah et al. 2001) |
||
| Metronidazole, Flouroquinolones, TMP-SMX, Amoxicillin/ clavulanate, Clindamycin, Cephalexin |
Varied |
Oral (with some IV lead in) |
75/93 (82%) |
Culture guided antibiotics, mean duration 6 weeks |
(Embil, Rose et al. 2006) |
||
| Amoxicillin-clavulanate, Flouro-quinolones, Clindamycin, TMP-SMX, Rifampin |
Varied |
Oral (with some IV lead in) |
264/339 (78%) |
Numerous regimens used, however, amoxicillin-clavulanate was the most common (N = 301) |
(Gariani, Lebowitz et al. 2019) |
||
|
*Definition of cure varied among the studies. †This was a randomized study of ciprofloxacin at 750 mg vs. 1000 mg twice per day. Because no comparator therapy was utilized, it is included in the non-randomized study section. DS = double strength tablet. ‡Based on a literature review, total case numbers >80, but difficult to count precisely from the review. |
|||||||
¶ Summary of Oral Antibiotic Doses Used in Osteomyelitis (Table 5)
|
Drug |
Dose |
Comment/References |
|
Ciprofloxacin |
500-750 mg BID |
|
|
Levofloxacin |
750 mg once daily |
|
|
TMP-SMX |
7.5-10 TMP mg/kg/d divided twice or thrice daily (e.g., 2 DS tablets twice daily for a 70 kg adult) |
|
|
Clindamycin |
600 mg TID; 900 TID or 600 QID for larger patients |
|
|
Linezolid |
600 mg BID |
|
|
Amoxicillin/ Clavulanate |
500 mg TID or 875 mg BID |
|
|
Rifampin |
600 mg once daily |
|
|
Fosfomycin* |
4 to 16 g per day |
|
| *There are no published data for the treatment of osteomyelitis with the sachet powder oral formulation of fosfomycin available in the US | ||