Skip to content
Brian Feldman

Type 1 diabetes: Developing an early warning system

Type 1 diabetes starts out as a sneak attack by bad-actor antibodies. But scientists at Stanford and UCSF have developed an early-warning system.

Type 1 diabetes starts out as a sneak attack. For unknown reasons, the immune system makes antibodies against insulin and other proteins in the insulin-producing islet cells in the pancreas, targeting these healthy cells for destruction. Clinical trials have suggested that, if the attack is stopped early with immune-modulating therapies, full-blown type 1 diabetes might be prevented.

But there's been a big hurdle in getting this early-intervention strategy to work: You need to figure out who has the bad-actor antibodies before too much damage occurs.

"Identifying these patients as early as possible is a really high priority for the field," said Brian Feldman, MD, PhD, an endocrinologist who has been studying the problem. Existing tests are expensive, require a lot of technical expertise to run, and don't do a good job of finding the earliest antibodies the immune system produces.

Now, Feldman's team at Stanford and the University of California-San Francisco has published a proof-of-concept paper in PLOS ONE showing that they have a new method to find the bad-actor antibodies much earlier than traditional testing. (Feldman was at Stanford until recently, when he moved to UCSF to accept an endowed professorship position.)

The work expands on the team's prior success at building an inexpensive nanotech-based microchip for diagnosing type 1 diabetes. In the new study, Feldman and his colleagues adjusted the microchip to detect immunoglobulin M (IgM) antibodies against insulin. IgM antibodies are the first to appear in the autoimmune response that precedes full-blown type 1 diabetes. The body later switches to making a different class of antibodies, IgG, which are detected by traditional tests.

The researchers used the microchip on blood samples collected as part of TrialNet, a large study following people who have close relatives with type 1 diabetes. They tested blood samples collected every six months for several years from six individuals who started out without type 1 diabetes, but went on to develop the disease. As a control, the researchers also used the microchip to test blood samples from eight healthy volunteers.

In three of the six people who developed type 1 diabetes, the new test found anti-insulin antibodies much earlier than traditional tests: For two subjects, the new method found the bad antibodies a year earlier, and in one person, the bad antibodies were detected four years earlier than via the traditional test. The microchip did not register any false positives on blood from healthy volunteers.

"The biggest surprise was how far ahead we could detect these antibodies," Feldman said. Before the study, the team had hoped that, at best, they might find bad-actor antibodies a few months in advance.

If the result holds up in larger trials, the extra warning time could make a big difference for the success of immune-modulating preventive strategies for diabetes. "The earlier you apply these next-generation therapies, the better they work," Feldman said.

The researchers have lots of ideas for what to do next. Although the current study only looked for antibodies against insulin, three other types of antibodies also presage type 1 diabetes; the microchip technology could easily be adapted to look for all four at once. The researchers would also like to greatly expand the number of people studied, and try the microchip method for people who don't have relatives with type 1 diabetes, since most individuals who develop the disease don't have an affected family member.

A big advantage of the microchip is its low-cost, easy-to-use design, Feldman said. "We're excited about that because we have an interest in making state-of-the-art diagnostics much more available, both in our own country and globally," he said.

Photo by Norbert von der Groeben

Popular posts