Volatile Metabolite-Based Detection of Clostridium difficile Infection



 DESCRIPTION (provided by applicant): Clostridium difficile infection (CDI) is a major cause of nosocomial diarrhea and mortality worldwide, increasing dramatically in incidence and severity over the past decade and rivaling or exceeding methicillin-resistant Staphylococcus aureus as the most common healthcare-associated infection in the US. A critical barrier to reducing the nosocomial spread of CDI is the delay in timely identification of CDI cases, with subsequent delays in initiation of appropriate antimicrobial therapy and implementation of contact precautions and isolation measures to reduce horizontal CDI transmission. Despite advances in rapid laboratory-based CDI testing, there are still delays of 2-5 days between the onset of clinical symptoms and stool sampling and another 0.5-3.5 days between sample collection and actionable test results. Anticipating a delay in obtaining diagnostic results, many providers prescribe CDI antimicrobial therapy empirically, resulting in frequent and unnecessary treatment of patients who ultimately test negative for CDI. There is clearly an unmet need for diagnostic methods that can reduce this interval between onset of clinical symptoms and accurate identification of CDI. To address this unmet need, we propose a novel approach to CDI diagnosis based on detection of the volatile metabolome of the altered CDI microbiome. C. difficile flourishes in the antibiotic- disrupted intestinal microbiome and emits a distinctive odo detectable by human and canine olfaction. In a pilot study, we assessed the volatile metabolome of stool specimens from patients with suspected CDI using analytical chemistry techniques and discerned major differences in the stool volatome of patients with and without CDI. We propose two aims executed in parallel, to test the hypothesis that the altered CDI intestinal microbiome has a unique volatile metabolite profile, distinct from the profile of patiens with other causes of antibiotic-associated diarrhea, which can be used to identify patients with CDI in the hospital setting. Using thermal desorption GC-MS/MS to capture and identify these trace volatile metabolites and supervised learning methods - support vector machines, prediction analysis for microarray analysis, and random forests - to analyze this high-dimensional data matrix, we will deconvolute the complex volatome of CDI, (1) identifying the characteristic volatile metabolite profile of stool samples from hospitalized patients with CDI, compared to patients with other antibiotic-associated diarrhea, and (2) defining the CDI volatile metabolite profile in ambient air samples from the inpatient environment of patients with CDI, compared to patients with other antibiotic-associated diarrhea. The ultimate objective of this research is to derive and validate the volatile metabolic signature of CDI in stool samples and in the patient's environment, laying the groundwork for a novel CDI assay that can be coupled to a point-of-care gas sensor system for the bedside identification of CDI, reducing delays in diagnosis and decreasing nosocomial transmission of this common, highly morbid, and life- threatening infection.