Antibiotic-tolerant bacterial biofilms are notorious in causing PJI. Antibiotic loaded calcium sulfate bead (CSB) bone void fillers and PMMA cement and powdered vancomycin (VP) have been used to achieve high local antibiotic concentrations; however, the effect of drainage on concentration is poorly understood. We designed an in vitro flow reactor which provides post-surgical drainage rates after knee revision surgery to determine antibiotic concentration profiles. Tobramycin and vancomycin concentrations were determined using LCMS, zones of inhibition confirmed potency and the area under the concentration-time curve (AUC) at various time points was used to compare applications. Concentrations of antibiotcs from the PMMA and CSB initially increased then decreased before increasing after 2 to 3 h, correlating with decreased drainage, demonstrating that concentration was controlled by both release and flow rates. VP achieved the greatest AUC after 2 h, but rapidly dropped below inhibitory levels. CSB combined with PMMA achieved the greatest AUC after 2 h. The combination of PMMA and CSB may present an effective combination for killing biofilm bacteria; however, cytotoxicity and appropriate antibiotic stewardship should be considered. The model may be useful in comparing antibiotic concentration profiles when varying fluid exchange is important. However, further studies are required to assess its utility for predicting clinical efficacy.

Introduction: Bacterial biofilms are an important virulence factor in chronic periprosthetic joint infection (PJI) and other orthopedic infection since they are highly tolerant to antibiotics and host immunity. Antibiotics are mixed into carriers such as bone cement and calcium sulfate bone void fillers to achieve sustained high concentrations of antibiotics required to more effectively manage biofilm infections through local release. The effect of antibiotic diffusion from antibiotic-loaded calcium sulfate beads (ALCS-B) in combination with PMMA bone cement spacers on the spread and killing of Pseudomonas aeruginosa Xen41 (PA-Xen41) biofilm was investigated using a large agar plate model scaled for clinical relevance. Methods: Bioluminescent PA-Xen41 biofilms grown on discs of various orthopedic materials were placed within a large agar plate containing a PMMA full-size mock spacer unloaded or loaded with vancomycin and tobramycin, with or without ALCS-B. The amount of biofilm spread and log reduction on discs at varying distances from the spacer was assessed by bioluminescent imaging and viable cell counts. Results: For the unloaded spacer control, PA-Xen41 spread from the biofilm to cover the entire plate. The loaded spacer generated a 3 cm zone of inhibition and significantly reduced biofilm bacteria on the discs immediately adjacent to the spacer but low or zero reductions on those further away. The combination of ALCS-B and a loaded PMMA spacer greatly reduced bacterial spread and resulted in significantly greater biofilm reductions on discs at all distances from the spacer. Discussion: The addition of ALCS-B to an antibiotic-loaded spacer mimic increased the area of antibiotic coverage and efficacy against biofilm, suggesting that a combination of these depots may provide greater physical antibiotic coverage and more effective dead space management, particularly in zones where the spread of antibiotic is limited by diffusion (zones with little or no fluid motion).