NEW YORK – Using gas chromatography to measure the volatile organic compounds in breath, the University of Michigan and startup shop Blu Biotech are sniffing for biomarkers distinctive of COVID-19, colorectal cancer, lung cancer, acute respiratory distress syndrome (ARDS), and sepsis.
Researchers from the university's Weil Institute for Critical Care Research and Innovation and College of Engineering have scooped up about $3.5 million in funding since 2015 to develop the Micro GC instrument and tests, most of it from a two-year $2 million National Institutes of Health grant awarded in 2020 for developing a breath-based COVID-19 test. Blu Biotech, which has offices in New York, Michigan, and New Mexico, is seeking US Food and Drug Administration Emergency Use Authorization for the COVID-19 test and said in November it plans to develop the patented instrument and assay technology it licensed from the university into a series of commercial breath tests and regulatory approvals for sale in the US.
An agreement between the two parties gives Blu Biotech exclusive rights to sell the technology in the US, Canada, and Europe.
Blue Biotech CEO Robert Tavzel envisions physicians offering the firm's multiplexed breath tests with annual checkups, much like the routine blood tests conducted with physicals. The company so far has VOCs identified for colorectal and lung cancer and COVID-19, and it is working with U Michigan researchers to identify which VOCs are biomarkers of ARDS and sepsis. With that information, the company can run all those tests simultaneously, he said.
"If you give me a breath sample right now, I can run that sample through my machine and tell you, with one sample on one test run, if you have colorectal cancer, lung cancer, or COVID-19," he said.
Kevin Ward, executive director of the university's Max Harry Weil Institute for Critical Care Research and Innovation, said the prototype instrument developed at the university and the commercial version in development by Blu Biotech both identify signs of disease by measuring VOCs that prior research has linked to inflammation, such as alkanes, alkenes, ketones, and aldehydes, from among hundreds to thousands of VOCs produced by the body's vasculature and organs, such as pentane, isoprene, and ethane.
Ward said diseases outside of the lungs — such as colorectal cancer or sepsis — still produce VOCs that circulate in blood and emerge in breath when they reach the lungs.
The research team identified the VOC patterns characteristic of diseases using machine learning and further data analysis and incorporated those patterns into the algorithms applied to patient samples.
In addition to point-of-care uses, Ward said the Micro GC could be used to monitor a patient on a ventilator and provide updates every 30 minutes.
"Not only do you know that the patient has the disease, but you can use the technology to tell you if the therapy that you are providing to the patient is having the desired effect," he said.
Early testing shows promising results
With either the university's research version of the instrument or Blu Biotech's commercial prototype, a patient provides a sample by exhaling into a plastic collection bag for about one minute or via a connection from the instrument to a ventilator's exhalation port. The gas chromatography instrument runs a heating cycle before drawing in a small portion of that sample. Blu Biotech is working to reduce the current 20-30-minute time to results, which would be reported in a chromatogram chart of target VOC concentrations.
Blu Biotech plans to run the test analysis remotely using a cloud server that will receive the sample data and send back the results as well as use that server to deliver new testing capabilities through over-the-air software updates, Tavzel said.
Xudong Fan, who has led research on the prototype development and testing at the Weil Institute, said the commercial version of the instrument promises to let clinicians run a battery of tests with the sample collection bag as the only consumable. Tavzel said that while the university's prototype requires helium canisters to carry the sample through the instrument, the commercial version will use dehumidified ambient air.
In preliminary studies, VOC measurements using the university's prototypes had 96 percent predictive value for colorectal cancer, 96 percent for lung cancer, and 95 percent for COVID-19, according to Blu Biotech's website. The Weil Institute team has also started gathering data on applications further down the road and demonstrated in one study of 21 ARDS patients and 27 controls that its instrument and algorithm had a 94 percent positive predictive value for the disease, an 82 percent negative predictive value, and an overall accuracy of 87 percent, according to results published in 2019 in Analytical and Bioanalytical Chemistry. The team used multi-physician adjudication to evaluate the accuracy of the ARDS predictions.
In that study, the researchers used nine VOCs to discriminate between ARDS and non-ARDS respiratory failure. But the authors also said that a larger study is needed to validate the methods, and further research was needed to identify the VOCs relevant to lung and systemic pathologies to understand the underlying physiological conditions.
"One of the important advantages of breath analysis is the potential to noninvasively monitor the development of ARDS, the severity of ARDS (if present), and the resolution of ARDS," the authors wrote. "This would allow the technology to map the trajectory of the disease and potentially guide therapy and decision-making."
The Weil Institute researchers have also started an animal-use study on detection of sepsis in swine ahead of human trials, and Ward predicts the sepsis and ARDS tests are three years out. The team is planning further studies in humans to detect ulcerative colitis, and Ward sees potential to apply the technology for inflammatory bowel disease, Crohn's disease, and kidney failure.
He said positive results in a breath-based colorectal cancer screening test would support diagnosis along with confirmatory testing, whereas clinicians could consider positive results for other indications such as sepsis or ARDS along with other clinical findings.
Benjamin Singer, pulmonary and critical care specialist at the Northwestern University Canning Thoracic Institute, said the idea of using gas chromatography to detect VOCs in breath and link those levels with disease is a promising concept that needs further study before it could be implemented. It makes sense that as blood passes from the right side of the heart to the lungs, VOCs could cross from the blood into air spaces of the lung in volumes detectable in exhaled breath, he said.
"We have a lot to learn," Singer said. "These are what I would consider emerging technologies and, while they hold a lot of promise, we need to have reliable data before we start applying them at the bedside."
Raed Dweik, chair of the Cleveland Clinic's Respiratory Institute, said exhaled breath is a rich matrix of components that can provide clues about a patient's state of health and disease. His research team has been using mass spectrometry with pattern recognition to develop breath-based tests for disease diagnosis and monitoring.
For breath-based testing instruments to be adopted by clinicians, they still need to become small enough to be truly portable, and their developers need to understand the connections between the compounds detected during testing and the biology of the disease, he said.
In 2019, Cleveland Clinic and Owlstone Medical announced they were establishing a Center for Early Disease Detection to assess the use of breath-based tests. Dweik's team is using mass spectrometry to develop a "breathprint" that might allow for the diagnosis and monitoring of diseases such as lung and kidney disease, and heart failure.
In recent years, using human breath has gained steam as a possible diagnostic sample, and the technology got a boost during the COVID-19 pandemic, with work on tests by companies including Breathonix and Canary Health Technologies. Rapid Biosensor Systems has been working on tuberculosis screening tests that combine breath sampling and evanescent-wave optical screening.
Blu Biotech's earliest devices will match the prototypes used in U Michigan's clinical testing. Tavzel said that model is easy to produce in small volumes, and he hopes to get them into hospitals of a high-volume market such as New York City, which could allow further testing and maintenance while the company tweaks the first commercial iteration.
The next version will have the same functions but with more precise microprocessor-based temperature and air pressure controls than the laboratory prototype's mechanical controls through analog pumps, he said. Blu Biotech is also in talks with manufacturers, especially BGM Engineering, a firm near Detroit that built ventilators early in the pandemic, he said.
Tavzel predicts he will know in early 2023 whether the company's testing data are sufficient for an FDA EUA of its COVID-19 test, but he also said the COVID-19 testing market is uncertain because of the unknown impact of virus variants, the inevitable end to the public health emergency declaration in the US, and the untested public perception of a breath-based test, especially how it is viewed in comparison with blood- or nasal swab-based tests. Blu Biotech is instead focusing its test development and market research efforts on cancer screening.
The firm has been in talks with hospital systems and university hospitals in New Jersey, New York, Louisiana, and New Mexico about starting a double-blind clinical trial for its colorectal cancer test in the first quarter of this year, and Tavzel sees potential to pursue trials at U Michigan. He said the goal is to start the trial at the institution with the most patients.
Tavzel expects little difficulty in securing FDA 510(k) premarket clearance for the Micro GC considering the agency's previous approvals of medical-use gas chromatography instruments. The company also plans to pursue the 510(k) clearance process for disease-specific tests, although he acknowledged that could be more difficult since few breath-based tests have gained approval.
Those that have include Otsuka Pharmaceutical's UBiT-IR300 Infrared Spectrometry System and Meridian Bioscience's BreathTek test. Tavzel said Blu Biotech intends to list both as predicate devices for its tests with the FDA.
Longtime pursuit of breath-based tests
Fan's team had been developing the gas chromatography instrument in the years prior to the pandemic, originally with a focus on diagnosing and monitoring patients with ARDS. According to the university, by January 2018, the team had developed a fully automated 9-pound prototype.
In December 2020, the National Institutes of Health awarded a two-year $1.9 million grant to Fan's team to develop a portable instrument capable of detecting and monitoring biomarkers of COVID-19 and lung injury. In the grant proposal, Fan said the researchers would build five automated and portable gas chromatography instruments that can return results in 20 minutes at parts per trillion-level sensitivity and identify breath biomarkers that distinguish people with COVID-19. The project is one of four funded through the NIH RADx Radical program's Screening for COVID-19 by Electronic-Nose Technology (SCENT) project.
Blu Biotech, which was founded in 2019, entered a patent license agreement in 2022 to produce and commercialize the device. The company's prototype is a shoebox-sized point-of-care instrument that can be operated without special training or staff certifications.
Fan said another firm, Guangzhou, China-based ChromX Health, licensed the same technologies in 2022 to develop their own breath analyzers for sale in China and nearby countries.
In addition to point-of-care uses, Tavzel sees potential that patients could submit breath samples collected at home if the company can stabilize them for shipment to a lab. He also recognizes that applying such a home-use test for colon cancer screening would pit the company against Exact Sciences, which uses home-collected stool samples for its Cologuard colorectal cancer screening tests.
Tavzel thinks Blu Biotech will have an advantage in being able to offer its tests in more places, from examination rooms to self-service testing stations inside drug stores.
Applying the technology to lung cancer becomes more complicated because the condition lacks a confirmatory test, Tavzel said. A bronchoscopy is only 60 percent to 65 percent accurate, he said, and radiographs and CT scans are less effective in detecting lung cancer than in detecting colorectal cancer, whereas he sees potential that Blu Biotech's test could become the standard for use ahead of radiation treatment.
"If we have a 95 percent accurate lung cancer test and the best confirmatory test that they have is 65 percent and it's a manual procedure, I don't know if we then become the standard," he said. "That would, of course, be the goal."
Tavzel also sees an absence of competition for sepsis and ARDS testing.
"Sepsis is tough," he said. "There's no test, it's very quick onset, and it can happen remotely; it can happen after surgery."
Fan said he recently began comparing the performance of Blu Biotech's prototype gas chromatography instrument against results from the device used in the Weil Institute's research.
"Once we establish such equivalency, then the device — the hardware — is ready to move forward for mass production, which is probably going to happen in about three months," Fan said.