Abstract
BACKGROUND: A breath test for volatile organic compounds has identified biomarkers associated with breast cancer. We evaluated the potential clinical and economic benefits of a breath test to detect women at low risk for breast cancer by comparing its negative predictive value (NPV) to the NPV of screening mammography. METHODS: Sensitivity and specificity values for screening mammography were obtained from the Food & Drug Administration Mammography Quality Standards Act; Amendments to Part 900 Regulations Docket No. FDA-2013-N-0134. The high values were sensitivity = 79.0%, specificity = 88.9% and the low values were sensitivity = 66.0%, specificity = 88.9%. In two previous studies of 771 women undergoing mammography, breath testing identified breast cancer with sensitivity=84% and specificity = 68.6% in 178 asymptomatic women, and sensitivity=82% and specificity = 77% in 593 who were symptomatic. These values were projected to a hypothetical screening population of 100,000 asymptomatic women with average breast cancer prevalence of 450/100,000, in order to estimate the NPV and PPV (positive predictive value) for breath testing and screening mammography respectively. RESULTS: Breath test in asymptomatic women: NPV = 99.895% and PPV = 1.19%; in symptomatic women: NPV = 99.895% and PPV = 1.59%. For screening mammography, NPV = 99.83% and PPV = 2.82% (low values), increasing to NPV= 99.89% and PPV = 3.12% (high values). A negative breath test identified 68.3% of the screening population as having low risk of breast cancer, with NPV similar to mammography. Based on Medicare reimbursement rates, elimination of mammography in women with a negative breath test could reduce the annual cost of breast cancer screening by 38.9. CONCLUSIONS: In a hypothetical screening population, a negative breath test ruled out breast cancer with the same accuracy as a negative mammogram. A screening breath test could potentially eliminate the need for two thirds of all mammograms and reduce the costs of screening without increasing the risk of false-negative findings. If applied in clinical practice, this approach could potentially reduce the costs and burdens of breast cancer screening services, and benefit women by lessening the discomfort, anxiety, radiation exposure, and costs associated with mammography.