Breast cancer is one of the most well-studied cancers in modern medicine. Diagnostics can already differentiate between several sub-types of breast cancer, but new technology might make the tests even easier.
Breast cancer can be divided into four broad sub-groups. They are estrogen-positive (ER+), progesterone positive (PR+), human epidermal growth factor positive (HER2+), or triple-negative (PR-ER-HER2-) breast cancers. Usually, a staining technique is used to measure the indicator proteins' levels and determine the sub-typing, but that may soon change.
In a new study, a team from the National Taiwan University of Public Health examined if an "e-nose" could determine the sub-type of breast cancer in patients. The e-nose is a device that analyzes metabolites (small molecules produced during cellular metabolism) in the breath. These metabolites would be unique for breast cancer patients and could diagnose the subtype of cancer.
The problem was in translating the results from the e-nose into something we can understand. To do this, the team utilized machine learning. A program would gradually learn how to read the data, pick out the subtle differences in data that separate each cancer sub-type, and translate it into a profile that would fit either PR+, ER+, Her2+ triple-negative metabolite profiles.
The team gathered 833 patients over two years and found that their protocol could indeed diagnose a patient's breast cancer subtype. They gathered air from the lower respiratory tract by using a tube to get the best air for the test. Air that you breathe out gets "contaminated" as it runs through your throat. They did this before surgery began on most patients and could have results ready as fast as 30 minutes later. While identifying breast cancer sub-types isn't a new discovery, the speed at which this method could diagnose the sub-type is far quicker than previous methods.
The field of diagnostics is often overlooked in favor of cool and exciting drug technologies. Cancer is a diverse disease, with many types of cancer's having sub-types that each have their treatment regimen. Using something as simple as a patient's breath to diagnose a cancer's sub-type could save patients time, money, and an invasive procedure. It also brings to mind a future where a simple breath test can tell you everything you need about your body.
The study concludes, "This study used sensor array and machine learning algorithms to analyze breath samples from breast cancer patients. The results showed high accuracy and reliability in the discrimination of breast cancer and the molecular subtype. The novel breath test has great potential to develop a rapid breast cancer diagnostic tool during surgery."