Predicting antifungal concentrations that select for resistance: an enhanced approach to establish environmental thresholds

• Data censoring of MIC data was considered to determine PNECRs for antifungals. • Over 300,000 MIC datapoints were collated, where clinical data was overrepresented. • PNECRs were determined for 13 clinical antifungals for both water and soil. • PNECRs were not determined for agricultural antifungals due to a lack of data. • Risk quotients suggest current MECs could pose a risk for AFR development. Antifungal resistance (AFR) is an emerging threat. Understanding the concentrations at which antifungals select for resistance is critical for guiding policy to minimise risks. This study aimed to determine predicted no effect concentrations for resistance (PNECRs) for antifungals in water and soil. PNECRs for water (PNECR water ) were derived from species sensitivity distributions fitted using a Maximum Likelihood Estimation approach to estimate the lower 5th percentile Hazard Concentrations (HC5s) from censored species/compound level MIC data and applying a 10-fold assessment factor. PNECRs ranged from 5.67x10 -4 (clotrimazole) to 7.94 µg/L (nystatin). PNECRs derived using standard methodologies that do not account for censoring are always higher, and therefore less conservative for environmental protection, than when considering censoring. PNECRs for soil (PNECR soil ) were derived by applying soil partitioning coefficients to PNECR water for each antifungal, thereby providing an estimate for the bulk soil concentration needed to achieve the PNECR water in soil pore water. These ranged from 2.26x10 -6 (voriconazole) to 2.16 mg/kg (nystatin). Risk quotients were generated from measured environmental concentrations, and 6.54% for water (n = 200) and 12.5% for soil (n = 1) were over 1, suggesting selection for AFR could be occurring. This type of data generation and analyses will inform discussions about targeted mitigation strategies to reduce the risk of selection for AFR, however, PNECR estimations can be improved with increased data for certain compounds, particularly agricultural fungicides. Preventing an increase in resistance is critical for reducing the risk posed to human health from exposure to environmental AFR.