Non-invasive imaging may help predict type 1 diabetes and response to treatment in humans

A key obstacle to early detection of type 1 diabetes - as well as to rapid assessment of the effectiveness of therapeutic intervention - has been the lack of direct, non-invasive technologies to visualize inflammation in the pancreas, an early manifestation of disease. Instead, clinicians have had to await overt symptoms before diagnosing an individual, by which time destruction of the insulin-producing beta cells of the pancreas has already progressed significantly.

Recent proof-of-principle experiments by Joslin Diabetes Center and Massachusetts General Hospital (MGH) researchers, however, offer hope that physicians may one day be able to identify individuals with preclinical type 1 diabetes, and to assess the effectiveness of therapies much earlier than is now possible. Findings of the study will be published in the September issue of the Journal of Clinical Investigation.

Type 1 diabetes is an autoimmune disease in which the body's immune system mistakenly attacks its own insulin-producing beta cells and eventually kills them. Early in this process, white blood cells called T lymphocytes infiltrate the islets of the pancreas (an inflammatory condition known as insulitis), causing the blood vessels to become leaky. Over time, this infiltration of lymphocytes destroys the beta cells, leading to high blood glucose and full-blown diabetes. Today, the only accurate method for detecting the progression or regression of insulitis is a biopsy of the pancreas, which is almost never performed because it is an invasive and potentially risky procedure.

"The most exciting aspect of this study is that it demonstrates that we can, at least in mice, use a non-invasive imaging method to predict at a very early time whether a drug will stop the progression of diabetes or not. In fact, the drug we used in these proof-of-principle experiments is analogous to one currently being tried in humans with diabetes, and so far showing great promise," said Diane Mathis, Ph.D., who led the study together with Christophe Benoist, M.D., Ph.D., also from Joslin, and Ralph Weissleder, M.D., Ph.D., of MGH.

Drs. Mathis and Benoist head Joslin's Section on Immunology and Immunogenetics, hold William T. Young Chairs in Diabetes Research at Joslin, and are Professors of Medicine at Harvard Medical School. Other investigators in the study included Stuart Turvey, M.D., Ph.D., formerly of Joslin, who is now at the University of British Columbia and British Columbia Children's Hospital, both in Vancouver, Canada; Maria Denis, Ph.D., a former Joslin research fellow who now works at the BSRC Alexander Fleming Institute of Immunology in Greece, and Eric Swart and Umar Mahmood, M.D., Ph.D., from MGH.

In this study, the Joslin and MGH researchers used a new imaging technique to reveal the otherwise undetectable inflammation of pancreatic islets in recently diagnosed diabetic mice. As T lymphocytes invade the pancreas, blood vessels swell, become more permeable, and leak fluid – as well as small molecules carried in the fluid – into surrounding tissues. In previous experiments, the researchers demonstrated that this leakage can be detected with the help of magnetic nanoparticles (MNP) and magnetic resonance imaging (MRI). After being injected intravenously, these MNPs, which are minute particles of iron oxide, travel through the blood vessels of the body including the pancreas. If pancreatic vessels have become leaky from inflammation, the magnetic particles spill into nearby tissues, where they are "eaten" by scavenger cells called macrophages. Thus, the MNPs become concentrated at the inflamed site and can be spotted by high-resolution MRI.

In their recent study, the researchers applied the MRI-MNP technique to determine whether they could predict which mice would develop autoimmune diabetes and monitor the effectiveness of immune therapy aimed at reversing diabetes. The goal of this study was to gather data on mouse models that could guide the safe application of the technique in human patients with, or at risk of, type 1 diabetes.

Results of this study suggest that the MRI-MNP imaging technology may be helpful in identifying people at immediate risk of developing autoimmune diabetes, but most of all for early prediction of response to therapy, which might be very useful for reducing the time and cost of clinical trials. "Because the results in mice looked so good, and because our MGH colleagues have already successfully used essentially the same drug on many patients with prostate cancer," said Dr. Benoist, "we have been able to move relatively quickly into clinical trials." Dr. Turvey added: "We hope to know soon whether we can use this drug and imaging technique to monitor pancreas inflammation in humans."