[Antibiotic release by tricalcic phosphate bone implantation. In vitro and in vivo pharmacokinetics of different galenic forms].

The purpose of this study was to evaluate the load and the release of antibiotics obtained with an implant made of a macroporous beta tricalcium phosphate ceramic (beta-TCP). Two parameters have been assessed: macroporosity and external shape (beads and parallelepipeds). In vitro, the ceramic beads were soaked in a Vancomycin-aqueous-solution, and the load of the antibiotic was then evaluated: it was 9.3% of the weight of the 40%-porosity beads and 4.6% of the weight of the 20%-porosity beads. The release has been evaluated by elution in phosphate-buffered-saline (PBS). With a 20% porosity, 12 beads (6.3 mm, 279 +/- 38 mg) demonstrated a short an massive release which ended within the 32 first hours. On the opposite, the release was sustained until the third week for the 40%-porosity beads (6.9 mm, 353 +/- 25 mg), while only one third of the load was released during the first 24 hours. A macroporosity of 40% of the ceramic could allow a deep incorporation of the antibiotic in the beads and thus decrease the rate of release. The in vivo study compared the bone concentrations of antibiotics obtained after implantations of either parallelepipedical or spherical devices in the distal femurs of 14 sheep. The bone concentrations of Gentamicin obtained with parallelepipeds until the end of the third week were from 5 to 10 times the minimum inhibitory concentration of this antibiotic for staphylococci. On the opposite, beads achieved only low concentrations of Vancomycin and nearly no detectable Gentamicin in the bone. We hypothesize a negative effect of the fibrous tissue which fills the gaps between the beads, and which could impair the diffusion of the antibiotics into the bone.