PROJECT SUMMARY / ABSTRACT
Acute respiratory distress syndrome (ARDS) is common, costly, and responsible for high mortality and long-
term morbidity. Unfortunately, our understanding of patient susceptibility and the underlying pathobiology of
ARDS is incomplete, resulting in imprecise ARDS diagnosis and the inability to accurately track its progression.
Exhaled breath contains hundreds of volatile organic and inorganic compounds (VOCs and VICs), many of
which are related to inflammatory and metabolic processes in the lungs and other organs. The goal of the
proposed project is to refine and test a novel, portable micro-gas chromatography (micro-GC) device that can
non-invasively and serially sample the lower respiratory tract for VOCs and VICs at the point of care. The
overarching hypothesis is that exhaled VOCs/VICs will exhibit dynamic patterns that can indicate the presence
of ARDS and track the disease trajectory. The team has developed a preliminary portable micro-GC device
and used it to measure exhaled VOCs, and they discovered VOC patterns distinguishing ARDS from other
causes of acute hypoxic respiratory failure in a small group of adult patients. To advance the device for clinical
use, they must now validate these preliminary findings and obtain robust time series data to identify the breath
biomarkers that enable disease trajectory monitoring. They will also evaluate the added benefit of measuring
VICs within the system, which they expect will improve device accuracy by providing additional inflammation
markers. In Aim 1, they will validate VOC breath biomarkers that identify ARDS patients and predict changes
in clinical status and outcomes. First, they will quickly refine their preliminary GC devices to enhance clinical
utility in exhaled VOC measurement. Then they will prospectively recruit 400 patients with acute hypoxic
respiratory failure and longitudinally measure VOCs and obtain other clinical and physiologic data for up to 10
days. They will also develop algorithms for GC signal analysis and test the algorithms that use VOC patterns to
identify ARDS. They will further validate algorithms that use dynamic changes in VOC patterns to predict
clinical trajectory. In Aim 2, they will Identify additional VIC breath biomarkers using new VIC detection
modules and develop enhanced breath signatures using both VOCs and VICs to detect ARDS and monitor its
trajectory. Once they integrate the new VIC detection modules into the existing GC devices, they will collect
VICs in patients recruited into the ongoing prospective longitudinal (in addition to VOCs). They will again
leverage machine learning approaches to develop and validate algorithms that identify ARDS from VOC/VIC
patterns, compare with algorithms using VOCs alone, and track dynamic changes to predict changes in clinical
status and outcomes. The innovative point-of-care gas analyzer will bring molecular diagnostics to the bedside,
enabling more accurate ARDS diagnosis that allows for earlier initiation of treatments to improve outcomes, as
well as novel ARDS trajectory monitoring to inform prognosis and downstream critical care decision-making.
