Summary: Delivering antibiotics (or other drugs) at effective and safe concentrations to unstable, critically ill hospitalized patients is a daunting problem – for example, for all six antibiotics that are the focus of this work, pharmacokinetic studies reveal truly wide discrepancies between predicted and actual active concentrations. While important strides have been made to help physicians make decisions that are most likely to help, and least likely to harm, an important cornerstone for such approaches is still missing: There are no clinically validated devices that can rapidly, accurately and precisely measure concentrations of drugs in a manner that would address the long turn- around-times and prohibitive cost of the central-laboratory-based approach to high-frequency therapeutic drug monitoring. We approach the problem systematically from the bottom up by building and validating at each step each of the components needed to overcome such barriers. As our first challenge (and our first Aim), we focus on high-quality receptors – well characterized and validated on their own – which could then be widely used and implemented by other groups in different formats, by simply being ordered “off the shelf”. We identified oligonucleotide-based receptors (aptamers) as such reagents. Building on our vast preliminary results, we delineate a process that will lead to isolation of sets of receptors, with each candidate validated on its own in a fluorescent format against gold-standard analytical methods in patient-derived fluids, under conditions that allow these simple sensors to be applied directly in mix-and-measure formats (e.g., dialysis effluents, ultrafiltered sera, and extracts from standard SPE columns). Our second challenge, exacerbated by the highly variable nature of samples collected from critically ill patients, is cross-reactivity and deviations from standardized conditions. We address these by identifying pairs of aptamers with orthogonal properties for each antibiotic: One of the aptamers will be used in biosensor modules (Aim 2) under strictly controlled conditions in conjunction with a commonly used nanomaterial (graphene) and validated on dialysis effluents and extracts from SPE columns (here, biosensors can be used without pre- purification) against gold-standard chromatographic methods. The other aptamer from the pair, from an unrelated family, will be comprehensively validated as the affinity component of extraction modules (Aim 3) on spiked commercial samples of sera and actual samples from critically ill patients. Through this approach, we will provide rigorously characterized, standardized components for analysis of polymyxins, fluoroquinolones, daptomycin, linezolid, and beta lactams, which will enable us (and others) to combine components into devices, either to be used at the bedside, or as cartridges in automatized analyzers. In either case, these will facilitate routine high-frequency drug monitoring outside of large, academic hospital settings. We expect to demonstrate one design of multi-modular devices by the end of this funding period.

Narrative: The sudden and profound physiological changes in critically ill patients cause dramatic variations of concentrations of antibiotics that we actually deliver, away from those that are effective and, often, towards those that are harmful. In theory, this challenge can be addressed by high-frequency therapeutic drug monitoring, which would allow us to adjust delivered doses of antibiotics as these concentrations change – yet, in practice, there are no clinically validated analytical methods that can perform suitably rapid and cost-effective quantification of drugs. In this project, we systematically develop modular components, which, individually or combined, will enable precise and accurate measurements of antibiotic concentrations in real-time at the bedside.