SEATTLE — The use of breath analysis for exhaled carbonyl compounds as a lung cancer screening tool is comparable to the results of the National Lung Screening Trial, according to data presented at the American Association for Thoracic Surgery Annual Meeting.
“We make this conclusion primarily because our stage and histology distribution are similar to that of the NLST,” Erin M. Schumer, MD, of the University of Louisville, said during a presentation. “The breath analysis process is simple, less expensive and limits radiation exposure.”
Erin M. Schumer
Schumer and colleagues conducted a retrospective analysis of 435 patients who were categorized as having lung cancer (n = 162), benign disease (n = 86), or controls (n = 187).
The researchers collected 1 L of breath in an inert bag from a single exhalation by all patients and analyzed exhaled carbonyl compounds as a screening modality for lung cancer. More than 97% of patients in the lung cancer group had a positive breath analysis vs. 82.6% of the benign group and 32.6% of controls.
Among lung cancer patients, breath analysis determined that 64.6% had early-stage cancer, while in the NLST study (n = 1,060), 57.1% were classified with early-stage cancer.
The sensitivity with at least one elevated carbonyl cancer marker was 97.5% for all patients with lung cancer and 96.1% for patients with early-stage disease. The specificity increased with the number of elevated cancer markers, rising from 51.6% for one marker to 94.5% for three elevated markers in the carbonyl concentrations (2-butanone, 3-hydroxy-2-butanone and 2-hydroxyacetaldehyde). A large number of false-positives were attributed to a lower threshold.
Schumer praised the potential benefits of using breath analysis, saying it “may supplant CT scan as the initial screening modality due to high sensitivity.” She also said more research is needed.
“We need a prospective, multicenter trial to better assess the false-positive rate in the at-risk population,” Schumer said. “With further data evaluation, we anticipate a more optimal method of limiting false-positives.” – by Ryan McDonald
Schumer EM, et al. LB3. Presented at: the American Association for Thoracic Surgery Annual Meeting; April 25-29, 2015; Seattle.
Disclosure: Schumer reports no relevant financial disclosures.