The management of post-traumatic bone infections relies on antibiotic therapy and surgical debridement. Antibiotic concentration in infected bone is a major determinant of response to medical treatment.

The aim is to assess glycopeptides, fluoroquinolones and carbapenems diffusion in infected human bone, since they are widely used for treating bone infections.

Twenty-four patients with septic pseudoarthrosis undergoing surgical debridement and treated with glycopeptides/fluoroquinolones/carbapenems iv for > 1 week were studied. Plasma and bone specimens were collected intraoperatively at a mean of 4.8h after antibiotic administration. Antibiotic concentrations were measured by the HPLC-UV method.

Five patients received vancomycin: mean bone concentrations were 2.4mg/L in cortical and 7.1mg/L in cancellous bone, with a bone/plasma extraction of 12% and 36%, respectively.

Nine patients were treated with teicoplanin: bone concentrations were 2.5mg/L for cortical and 8.3mg/L for cancellous bone (14% and 46% of plasma levels).

Five patients received a fluoroquinolone. Ciprofloxacin concentrations were 1.8mg/L in cortical bone and 30.2mg/L in cancellous and newly formed bone (respective bone/plasma ratios 1.06 and 8.4). Levofloxacin concentrations were 0.3 and 2.69mg/L in cortical and cancellous bone, with diffusion rates of 12% and 108%, respectively.

Five patients received a carbapenem. Imipenem diffusion rates were respectively 7.5% and 58.3% for cortical and cancellous bone (bone concentrations 0.09 mg/L and 0.7 mg/L). Meropenem levels were 1.2 mg/L and 5.2 mg/L in cortical and cancellous bone, with respective diffusion rates of 3.6% and 15%.

Both glycopeptides provided concentrations exceeding the MIC of infecting agents, with satisfactory bone diffusion. Fluoroquinolones, especially ciprofloxacin, displayed excellent diffusion. Ciprofloxacin concentrations in cancellous and new bone were far higher than in plasma, suggesting accumulation into highly vascularized tissue. Imipenem had better diffusion than meropenem, but bone levels were under the MIC of susceptible agents. Glicopeptides and fluoroquinolones appear excellent options for bone infections, while carbapenems should be a second choice treatment.

Antibiotic concentration in infected bone is a major determinant of clinical response. As glycopeptides and fluoroquinolones are widely used for the treatment of bone infections, aim of our study was to assess their diffusion in infected human bone. Patients with a posttraumatic septic pseudoarthrosis undergoing debridement of infected tissue, who received a glycopeptide or a fluoroquinolone for > 1 week, were studied. Plasma and bone specimens were collected intraoperatively for phamacokinetic and microbiologic assays at a mean of 4.1h after antibiotic administration. Bone samples were crushed and concentrations were measured by HPLC-UV method. Overall plasma exposure was also determined with daily sampling. 16 patients were studied. 6 patients received iv vancomycin 1 g bid over a 1-hr infusion Bone cultures grew E. faecalis, MRSA and MRSE (MIC < 2 mg/L). Mean plasma concentration of vancomycin at time of osteotomy was 19.8 mg/L. Mean bone concentrations were 2.4 mg/L in cortical and 7.1 mg/L in cancellous bone, with a mean bone extraction of 12 % and 36 %, respectively. 4 patients were treated with iv teicoplanin 10/mg/Kg for MRSA infection (MIC < 2 mg/L). Mean bone concentrations were 8.9 mg/L and 37 mg/l respectively for cortical and cancellous bone, respectively corresponding to 6% and 25% of plasma levels. Six patients were treated with a fluoroquinolone. 3 patients received iv ciprofloxacin 400mg bid and E. coli grew from bone samples(MIC = 0.5 mg/L). Mean Plasma concentration of ciprofloxacin at the time of osteotomy was 3.6 mcg/mL. Mean bone concentrations were 1.7 mg/L in cortical bone and 30.2 mg/L in cancellous and newly formed bone, with respective bone/ plasma ratios of 0.5 and 8.4. 3 patients were administered iv levofloxacin 500mg qd and Enterobacter spp. were isolated (MIC = 1 mg/L). Mean plasma concentration at the time of surgery was 2.5 mcg/mL. Mean bone concentrations were 0.3 and 2.69 mcg/mL in cortical and cancellous bone, respectively. To our knowledge this is the first study that compares different antibiotic’s concentration in infected bone with the same dosing procedure. Both vancomycin and teicoplanin provided mean bone concentrations exceeding the susceptibility breakpoint of the infecting agents. Higher and constant glycopeptides plasma levels may be required for preventing recurrencies in bone infections. Only ciprofloxacin provided cortical bone concentrations higher than the susceptibility breakpoint of the infecting agent, and similar to those reported in non-infected bone. Ciprofloxacin concentration in cancellous bone and in bony callus were far higher than those detected in plasma, which may be related to an augmented vascularization and/or selective accumulation of fluoroquinolones into regenerating bone, as observed in children’s cartilage growth plate. Ciprofloxacin may be therefore preferred to levofloxacin.