Chang-yun Hue et al., JCI, 117 (December 2007)

B cells banished.
An antibody that depletes B cells kept the pancreas cells of one mouse looking healthy (top), whereas the control animal sustained immune attacks on the organ (bottom).

Antibody Therapy Shows Promise for Diabetes

An antibody therapy that's approved to treat certain cancers and arthritis has helped prevent--and even beat back--type 1 diabetes in mice. The news is heartening to researchers, who had already launched a human trial of the therapy, rituximab. But it also raises concerns that diabetes patients and their doctors will start trying the drug before it's been shown to work in humans.

Type 1 diabetes occurs when the immune system attacks and destroys insulin-producing cells in the pancreas. For years, T cells were considered enemy number one, because they commit the actual attacks. But more recently, scientists have eyed another potential culprit: B cells, which may be setting the T cells off by presenting them with antigens, proteins that stimulate the immune system. The drug rituximab, made by Genentech, is an antibody that depletes B cells and has been shown to combat rheumatoid arthritis, another autoimmune disease. With that in mind, an international network of researchers successfully lobbied for a clinical trial of the drug in type 1 diabetes, even though mouse studies were lacking. That trial, begun last year, has enrolled 82 people--the youngest being 8 years old--and will take another year or so to finish.

The absence of mouse studies was "a concern in our scientific community," says immunologist Li Wen of Yale University. With that in mind, she, along with Yale immunologist Mark Shlomchik and their colleagues, genetically engineered a mouse model to test the drug. Wen's animals are predisposed to diabetes and have the human version of CD20, the molecule rituximab targets, on the surface of their B cells. Wen and her colleagues were reluctant to work directly with Genentech and acquire rituximab from them (companies often impose restrictions on publication and other matters, she says), so instead they designed a rituximab-like antibody to test.

The researchers tested the drug in mice of various ages, including 4- and 9-week-old animals that didn't have diabetes and older mice within 6 days of diagnosis. In the animals that were still healthy, about 70% of those receiving the antibody therapy were diabetic by 35 weeks of age, compared to nearly 100% of those that received a placebo. This translates to a 10- or 15-week delay in developing diabetes--the equivalent of 10 or 15 years in humans, says Wen. In the 14 mice with established diabetes, five stopped needing insulin for more than 2 months, some for up to 5 months, the researchers report in the 3 December issue of the Journal of Clinical Investigation.

As the animals regained their B cells, Wen and her colleagues found that those cells were better behaved than their B-cell predecessors that were wiped out by the drug. For example, the new B cells were less likely to produce autoantibodies that often precede type 1 diabetes, suggesting that the drug had somehow recalibrated the immune system.

Mark Pescovitz, a transplant immunologist at Indiana University, Indianapolis, and head of the rituximab clinical trial, is heartened by the results. "[I'm] even more confident that our study will succeed," he says. Still, Pescovitz worries that the new mouse findings will cause physicians to jump the gun. "I would hope that [doctors] aren't going to run out there with their next diabetic and throw rituximab at them, but there's nothing to stop them," he says. The drug is not without side effects such as infections and rashes, and although Pescovitz is impressed with the mouse work, he notes that many drugs that work in mice fare less well in people.

Related sites

Posted in Biology