Pk-pd parameters ratio of antibotics released from antibiotic-loaded spacers in drainage fluids.

Objectives: Infection remains a difficult problem in total hip or knee replacement. Polymethylmethacrylate (PMMA) cements impregnated with aminoglycosides and/or vancomycin are currently utilised as local antibiotic carriers in orthopaedic prosthetic infections. Few data are available on antibiotics release in vivo at the moment of implant as well as their antimicrobial inhibitory activity. Methods: The local and systemic release of gentamicin (G) from temporary spacers (hip and knee) loaded with 2.5% gentamicin (SpacerG®) was studied in 13 patients undergoing two-stage revision surgery. We currently combine the use of spacers with systemic therapy for the treatment of prosthetic infections. Vancomycin (V) was parenterally (1 g, b.i.d.) or locally administered (cement, 2.5%). The concentrations of G and V were determined in drainage fluid (DF) and serum collected in the first 24 hours after implantation of spacers. Microbiological method and FPIA were utilised. Their inhibitory activity was evaluated against multiresistant clinical isolates (S. aureus, S. epidermidis, S. hominis, S. haemolyticus, E. coli, P. aeruginosa) by bactericidal titer and continuous turbidimetric recording of bacterial growth (Bioscreen, Labsystem). Results: The release of G from spacers in the infection site seems prompt and effective, attaining high local concentrations. Serum levels of G were very low. Parenteral V showed good penetration in the infection site. Higher levels in drainage fluids were attained by local administration. G and V were present in the implant site at inhibitory concentrations. Drainage bactericidal levels presented great intersubject variability. Highest concentrations of DFs (from 80% to 20%) inhibited bacterial growth of all strains tested for 24 h, 10% for 12 h. G and V in combination showed an additive effect against multiresistant strains when tested at the same concentration of bactericidal titer; lower levels of G and V inhibited bacterial strains for 10−14 h. DFs maintained their inhibitory activity also after filtration and proteins precipitation. Conclusions: G and V are released from cement at inhibitory concentrations in the infection site for prolonged periods. V presents a good and prompt penetration in the infection site after i.v. administration. The combination of G and V exerts an effective effect against multiresistant bacteria. PK/PD parameters ratio seems good and can be considered a predictive value for therapeutic efficac