Breath Test Identifies Different Kinds of Lung Cancer

E-mail Print PDF


A recent abstract study has revealed that a breath test for unpredictable organic compounds accurately detected lung cancer in high risk patients and typified its subtype.

The test had C-statistics of 0.824 to 0.874 for an assortment of cancer subtypes compared with controls; 0.8 or better on the 1-point scale is considered strong concordance, Peter Mazzone, MD, MPH, of the Cleveland Clinic, and peers discovered.

In fact, the test was even more effective at differentiating between adenocarcinoma from squamous cell carcinoma, with a C-statistic of 0.896, the group reported at the American College of Chest Physicians meeting.

The test was deemed a "colorimetric sensor assay" for its use of pigments on an expendable chip that changed color in the presence of certain explosive compounds, in this case, compounds formerly associated to lung cancer.

Other groups have also been working on systems to single out lung cancer, with “artificial nose” sensors that record and measure fluctuations in electrical resistance when particles attach to nano-sized sensor ranges similarly reported to differentiate between cancer subtypes.

“Like all biomarkers for lung cancer so far, validation is needed in large clinical trials to prove that the tests not only work but are clinically useful,”said Gerard A. Silvestri, MD, of the Medical University of South Carolina in Charleston.

Yet, he was still optimistic that exhaled breath could hold the answers to either prescreening to select truly high-risk individuals for low-dose CT scan programs or to triage suspicious nodules detected on scans, to reduce damage from false positives.

"It's incredible. I don't understand why it hasn't taken off. If you could imagine the least invasive way to get an answer for patients, instead of having bronchoscopy or surgery or multiple biopsies, it would make so much sense,” said Silvestri.

Mazzone cited that one main reason maybe lack of funding.

“The technology is in the "valley of death" where it will fade away without getting picked up by companies to commercialize it,” he said.

breath test finds lung cancer

"The one advantage this biomarker could have over the others is once the models are developed, it's breathing in a machine and the answer comes up on the screen as opposed to sending a blood test away for a couple-thousand-dollar test, to come back a week or 2 later.”

His group's study included 236 high-risk patients over the age of 40 from two hospital centers, who had undefined lung nodules 4 to 20 mm in diameter that were either suspected as lung cancer awaiting biopsy, or biopsy-proven, but untreated lung cancer.

The final version had patients inhale through a filter eliminated outside volatile organic compounds. Then, triggered by end-tidal carbon dioxide, the device caught only the alveolar portion of the exhale so the final 4 L of breath passed across the sensor array.Mazzone's group utilized tests on a separate set of 288 patients to optimize the system, making changes to the filters, sensors, and processing.

The sensor array was imaged various times during and following the breath, with the colors changed to numerical vectors for red, green, blue, and ultraviolet spectrums.

Predicated on the pathology report or standard clinical follow-up, 81 of the confirmation set patients had cancer, with a variety of stages and subtypes fitting the expected allotment.

The breath test results concurred quite well with pathology or clinical findings. The C-statistics, also known as the area under the receiver operating characteristic curve, were:

0.828 for non-small cell lung cancer

0.824 for lung adenocarcinoma

0.874 for lung squamous cell carcinoma

Those numbers were an improvement over the former version of the test, which had C-statistics of 0.811 to 0.849, Mazzone noted.

“The volatile organic compound profile can be considered a candidate biomarker but still has many steps to take before reaching the clinic, including technical validation of reliable performance and clinical validation in multicenter studies,” he said.

Silvestri praised the group’s careful, methodical approach, however still maintained concerns that the required large-scale trial could be a success or total failure.

"Sometimes when we see preliminary data from biomarkers it doesn't pan out when we start looking at a diverse population. They can look so promising and not pan out,” he said.

These signals are relayed buying clomid online safe which then is by a number of such as medial preoptic and paraventricular nulcei.