A Novel Device for Rapid and Noninvasive Volatile Metabolite-based Screening and Diagnosis of Multiple Disease States - Project Summary/Abstract: Humans emit an array of volatile organic compounds (VOCs) as part of normal metabolism. There are metabolic shifts in many disease states, and animals with highly sensitive olfactory systems can be trained to identify patients with certain diseases based on their characteristic scent. We seek to translate detection of these unique VOCs emanating through the skin to a more robust, standardized, and mechanized platform, readily adaptable for the screening, diagnosis, and monitoring of a variety of human disease states with distinct pathophysiologies. We hypothesize that systemic metabolic derangements in many high burden infectious, inflammatory, metabolic, malignant, psychiatric, and neurologic diseases can be identified via skin emissions. We will adapt a novel, portable gas chromatography- differential mobility spectrometry (GC-DMS) device that rapidly examines volatile samples directly at the point of care for the assessment of these skin volatile metabolite signatures. This device is highly sensitive, allowing a comprehensive and biologically representative assessment of the landscape of human volatile emissions, is more robust to confounding and environmental factors than many other gas sensing devices, has a long track record of successful use in various real-world sensing applications, and is already undergoing commercial development, which will greatly facilitate its rapid development for the skin VOC-based diagnosis of these 20 disease states. We propose further development and rigorous evaluation of this scent-based diagnostic approach to these disease states, (1) integrating additional process analytical technologies (PATs) to ensure instrumental accuracy and precision between samples and devices through repeated, high-volume patient testing over time and modifying the device inlet for skin volatile analysis, (2) identifying and validating GC- DMS signatures and the corresponding set of skin volatile metabolites that distinguish individuals with and without each of these 20 disease states using machine learning methods, integrating automated detection of these signatures on the GC-DMS device, and (3) developing a wearable sensor for at least one disease with a simple volatile signature, determining the test characteristics of this wearable diagnostic device. A lack of reproducibility is a major issue plaguing the field of volatile metabolite analysis, often due to confounding and study design flaws. We will make every effort to minimize and eliminate any sources of confounding, bias, and extraneous variability as we develop and evaluate this scent-based diagnostic approach, to yield generalizable signatures for each disease state. Ultimately, successful completion of these aims will yield rapid, noninvasive skin volatile metabolite assays for the screening, diagnosis, and monitoring of each of these diseases, with a clear, binary (yes/no) assessment of whether the individual has evidence of one or more of these diseases, facilitating early administration of appropriate therapeutics in these patients and mitigating the clinical consequences of each of these diseases.
