Effects of regional limb perfusion technique on concentrations of antibiotic achieved at the target site: A meta-analysis.

Intravenous regional limb perfusions (RLP) are widely used in equine medicine to treat distal limb infections, including synovial sepsis. RLPs are generally deemed successful if the peak antibiotic concentration (Cmax) in the sampled synovial structure is at least 8-10 times the minimum inhibitory concentration (MIC) for the bacteria of interest. Despite extensive experimentation and widespread clinical use, the optimal technique for performing a successful perfusion remains unclear. The objective of this meta-analysis was to examine the effect of technique on synovial concentrations of antibiotic and to assess under which conditions Cmax:MIC ≥ 10. A literature search including the terms "horse", "equine", and "regional limb perfusion" between 1990 and 2021 was performed. Cmax (μg/ml) and measures of dispersion were extracted from studies and Cmax:MIC was calculated for sensitive and resistant bacteria. Variables included in the analysis included synovial structure sampled, antibiotic dose, tourniquet location, tourniquet duration, general anesthesia versus standing sedation, perfusate volume, tourniquet type, and the concurrent use of local analgesia. Mixed effects meta-regression was performed, and variables significantly associated with the outcome on univariable analysis were added to a multivariable meta-regression model in a step-wise manner. Sensitivity analyses were performed to assess the robustness of our findings. Thirty-six studies with 123 arms (permutations of dose, route, location and timing) were included. Cmax:MIC ranged from 1 to 348 for sensitive bacteria and 0.25 to 87 for resistant bacteria, with mean (SD) time to peak concentration (Tmax) of 29.0 (8.8) minutes. Meta-analyses generated summary values (θ) of 42.8 x MIC and 10.7 x MIC for susceptible and resistant bacteria, respectively, though because of high heterogeneity among studies (I2 = 98.8), these summary variables were not considered reliable. Meta-regression showed that the only variables for which there were statistically significant differences in outcome were the type of tourniquet and the concurrent use of local analgesia: perfusions performed with a wide rubber tourniquet and perfusions performed with the addition of local analgesia achieved significantly greater concentrations of antibiotic. The majority of arms achieved Cmax:MIC ≥ 10 for sensitive bacteria but not resistant bacteria. Our results suggest that wide rubber tourniquets and concurrent local analgesia should be strongly considered for use in RLP and that adequate therapeutic concentrations (Cmax:MIC ≥ 10) are often achieved across a variety of techniques for susceptible but not resistant pathogens.