Antimicrobial concentrations in the cornea and aqueous humour: a meta-analysis

Aim To interpret the likely clinical susceptibility of isolates from microbial keratitis (MK), we performed a meta-analysis of published data that measured the concentrations of topically applied antimicrobials in the cornea or aqueous humour. We then correlated these values with the in vitro minimum inhibitory concentration (MIC). Methods We searched PubMed to identify studies reporting aqueous and/or corneal concentrations of 53 topically applied ocular antimicrobials, spanning the following classes: beta-lactams, glycopeptides, aminoglycosides, chloramphenicol, lincosamides, macrolides, oxazolidinones, steroidal antimicrobials, tetracyclines, diaminopyrimidines, sulphonamides, lipopeptides and polymyxins. Two clinicians independently screened the abstracts and extracted data from studies meeting the inclusion criteria, including participant species, antimicrobial concentration, dosing regimens, epithelial status and measurement methods. Concentrations were standardised to mg/L. The data were stratified by applied concentration, dosing regime and species. First quartile concentrations (EQ1) were extrapolated to provide a conservative estimate and tabulated practice resource for clinicians treating MK. Results We screened 7247 publications. 81 publications were included in the meta-analysis, comprising data on the aqueous and/or corneal concentrations of 28 antimicrobials. Bioassay was the most frequently used method for quantifying antimicrobial concentrations (25 studies), followed by liquid chromatography and fluorescence assays (18 studies each), mass spectrometry (12 studies) and radioactivity and colourimetric assays (3 studies each). Conclusion We provide a practical resource for clinicians to assess whether the expected EQ1 of an antimicrobial in the cornea is above the in vitro MIC of the pathogen. This reduces reliance on systemic break-point concentrations. This enables standardised guidelines for evidence-based antimicrobial treatment decisions for MK. All data relevant to the study are included in the article or uploaded as supplementary information. Not applicable.