Skip to content

Developing a faster, more accurate test for cystic fibrosis, without breaking a sweat

A new cystic fibrosis test could provide a more accurate, and easier, way to test newborns for the hereditary, lung-clogging disease.

Stanford clinicians and chemists are working together to develop a faster, more reliable pediatric test for cystic fibrosis, a genetic disorder that clogs the lungs and makes it hard to breathe. The new test could help doctors diagnose the disease faster and more reliably than the current gold-standard test, which could in turn help ease parents' uncertainty and aid in the development of new therapies for the disease, the researchers say.

Today, the best way to test newborns and young children suspected of being affected by cystic fibrosis is a sweat chloride test, which requires technicians to chemically induce sweat from a baby, sample it, and run test for the presence of higher concentrations of salt, a surrogate marker that lung cells aren't working properly. The test takes some time -- usually about three hours -- and in about 10% of infants and small children the test cannot be completed due to inadequate sampling. 

There is also a significant ambiguity in the results, said Carlos Milla, MD, a professor of pediatrics and a senior author of the new study that appears in PNAS. For every two infants whom the test clearly identifies as having cystic fibrosis, another three will be in a gray area where doctors can't say for sure whether a child will ever develop the disease.

"That's stressful for a family," Milla said.

Fortunately, cystic fibrosis testing got the attention of chemical physicist Richard Zare, PhD, a member of Stanford Bio-X, and his then graduate student Zhenpeng Zhou, the new paper's first author. The pair had been working on ways to infer health information from easy-to-obtain samples, such as perspiration gathered from fingerprints, so they got in touch with Milla to see if they could help.

Zhou and Zare's subsequent idea for cystic fibrosis testing had three key steps.

First, they would wipe a standard microscope slide across a baby's forehead, gathering a tiny bit of perspiration, without the need to stimulate sweating. Then they'd bombard that sample with a spray of tiny droplets of a solvent. The splash from those droplets enters a mass spectrometer, a device that can then provide information on the many hundreds of compounds found in a perspiration sample. Finally, drawing on samples from a group of people with cystic fibrosis and a control group, the team ran a machine learning algorithm to infer the key chemical differences that distinguish people with the disease from those without. 

In pilot tests, the researchers used those inferences to correctly identify whether someone had cystic fibrosis 98% of the time, with a test that is easier to run and takes only about two minutes.

If those results hold up in follow-up studies, Milla said, it could not only help diagnose cystic fibrosis more easily in newborns, it could also help researchers tracking the effectiveness of new cystic fibrosis treatments. If a treatment is working, he said, it should have an effect on many of the chemicals the body produces, and with the new test, researchers could see those changes that much faster. 

The results also suggest other simple tests for a variety of other conditions, Zare said. "It's faster, cheaper and it has the ability to do more than cystic fibrosis," he said -- all without breaking a sweat.

Image by bogitw

Popular posts

Category:
Biomedical research
How do the new COVID-19 vaccines work?

The Pfizer and Moderna COVID-19 vaccines are the first to use the RNA coding molecule to prompt our bodies to fight the virus. Here's how they work.