Metabolomx has announced the results from a clinical study evaluating its first-generation colorimetric sensor array, which can detect and differentiate types of lung cancer in humans.

The sensor identifies the unique pattern of volatile organic compounds present in exhaled breath, and is found to detect the subtype of lung cancer with around 90% accuracy.

The study, conducted at the Cleveland Clinic, was intended to develop an exhaled breath biosignature of lung cancer using Metabolomx’s colorimetric sensor array and to determine the accuracy of breath biosignatures of lung cancer characteristics with and without the inclusion of clinical risk factors.

In the study, breath samples were drawn from 229 individuals, 92 with biopsy-proven, untreated lung cancer and 137 either at a risk for developing lung cancer or with indeterminate lung nodules, and reported that Metabolomx’s colorimetric sensor showed more than 80% accuracy in lung cancer detection, compared with a CT scan.

The trial reported around 81% accuracy for the detection of lung cancer without regard for subtype, 83% accuracy in discriminating patients with adenocarcinoma, 85% with squamous cell and 89% with small cell lung cancer, versus controls.

Metabolomx founder and CEO Paul Rhodes said the study results demonstrate the broad potential use of its first-generation colorimetric sensor array for breath analysis in the early detection of lung cancer.

“Detection of the metabolomic signature of lung cancer in exhaled breath is non-invasive, rapid, and inexpensive, and will become a valuable adjunct to help assess an indeterminate CT, and may come to have a central role in early detection and differentiation of lung cancer, while lowering costs to the healthcare system,” Rhodes added.

Dr Peter Mazzone of the Cleveland Clinic said that the research shows that breath testing may aid the diagnosis of lung cancer, as well as provide information that can help with treatment decisions, such as the type of lung cancer, its stage and prognosis.

”The accuracy of these non-invasive tests can be further augmented when combined with existing clinical predictors, such as health status and age,” Mazzone said.