An assay of Staphylococcus epidermidis biofilm responses to therapeutic agents.

Implant-associated infections tend to become persistent, resisting host defences and antibiotic therapy. Routine clinical laboratory testing of bacterial isolates in the fluid phase for antibiotic susceptibility may not be predictive of therapeutic outcomes and therefore a number of antibiotic regimes have been formulated empirically. The resistance of implant-associated infections has been ascribed to the production by bacteria, when adherent to the implant surface, of a shielding matrix of polymerized carbohydrates protecting the enclosed bacteria from immune defences and antibiotics. This complex of surface, bacteria and matrix is termed a biofilm. We describe a technique to evaluate the efficacy of antimicrobial agents directed against biofilm-forming strains of Staphylococcus epidermidis (a major cause of implant infections) utilizing standardized biofilm preparations formed on glass. The impact of the antimicrobial agents was assessed quantally using the end-point of permanent cessation of metabolic activity (cell death) of the entire biofilm determined by the loss of ability to reduce 2,3,5-triphenyltetrazolium chloride to a visible red intracellular precipitate of formazan. The relative rate of action of differing antimicrobial agents could be determined by the minimum period of exposure of the biofilm to the agents that is required to bring about sterilization, the clinically relevant marker. A wide range of antimicrobial substances could be evaluated, including chemical disinfectants, immunoreactive substances, antibiotics; singly and in combination; and the modifying effects of interaction with non-antibacterial therapeutic agents and other environmental factors. The technique is simple, inexpensive, reproducible and readily adaptable to the clinical situation where evaluation of therapeutic regimes for individual cases of prosthetic device-associated infection is required routinely with despatch and ease of interpretation.