British scientists want to use fingerprints to diagnose breast cancer painlessly and non-invasively in the future. They weren’t interested in the pattern of the papillary ridges, but in the sweat on them. The team led by Simona Francese from Sheffield Hallam University in Sheffield looked for the characteristic protein profiles of early-stage breast cancer, metastatic breast cancer and benign breast cancer as part of a small pilot study with 15 women.
The five patients in each group first stroked a thin aluminum plate several times with three fingers. The researchers then sent the sweat samples taken through a MALDI spectrometer. MALDI stands for matrix-assisted laser desorption/ionization. It ionizes samples using a laser beam, transfers them to a vacuum, and then identifies them by their weight.
Machine learning sweat samples
The researchers used some of the samples to train three different machine learning methods. They then used the remaining samples to test how accurately the methods could correctly determine the three breast cancer diagnoses. The best method was 97.8 percent accurate across all three groups. She correctly identified all samples from patients with early and metastatic breast cancer as such and not incorrectly as benign. They published their results in the journal Nature.
Every year, 2.3 million people are diagnosed with breast cancer worldwide, and more than 600,000 people die from it. Traditional screening and diagnostic methods have so far been mammograms followed by biopsies, which are considered the gold standard methods. While they are effective, mammograms, for example, expose patients to a certain dose of radiation, and both can be painful. This often deters even women with symptoms from examinations. In addition, mammography in younger women under the age of 40 is less accurate in detecting suspicious changes because their breast tissue is denser.
Next, the team wants to conduct much larger studies, ideally with several hundred patients, to validate the method. If everything goes well, it would still be a few years before it could be used in practice.
(vs.)