For a substantial proportion of patients diagnosed with type 2 diabetes, islet autoimmunity plays an important role in disease onset, risk of complications, and progression to pancreatic ?-cell failure. However, advances in understanding the heterogeneity within type 2 diabetes have not yet been harnessed to improve clinical diabetes care. In this proposal, we will use state-of-the-art assays and novel statistical techniques in Look AHEAD (Action for Health in Diabetes), a well-characterized longitudinal cohort in order to develop a precision medicine approach to islet autoimmunity in type 2 diabetes care. We will measure islet autoantibody assays currently available for clinical use (antibodies to GAD65, insulinoma antigen 2, and zinc transporter 8), and novel assays that are critical for determining clinical significance and could be widely available in the near future. Novel assays include the GAD65 antibody epitope specificity and IgG subclass, which are key determinants of immunologic activity. We will also measure an antibody to novel epitopes of insulinoma antigen 2 which has been shown to greatly expand its sensitivity and improve diabetes classification. With these rich islet autoantibody data, we will be able to identify patterns of islet autoimmunity that more accurately predict a patient?s diabetes phenotype than traditional measures. We will accomplish this in Look AHEAD, a U.S.-based randomized trial investigated the effects of an intensive lifestyle intervention versus self-directed care among 5,145 overweight/obese adults with type 2 diabetes. Participants have been followed for >15 years with detailed longitudinal ascertainment of clinical and laboratory data, diabetes complications, and mortality. Our pilot data has identified a 6.4% prevalence of traditional islet autoantibodies and associations with mortality and other outcomes. Informed by our work in precision medicine for other disease models, we will use advanced statistical techniques and interactive visualizations to develop a platform for clinicians to interpret the significance of islet autoimmunity. Bayesian hierarchical models will jointly estimate patterns of autoimmunity and associations with a patient?s health trajectory, and we will develop a visualization tool to present these predictions for clinicians. We have developed such visualization tools for prostate cancer, scleroderma, and other disease states which are currently implemented in Johns Hopkins Precision Medicine Centers of Excellence.
Our specific aims are:
Aim 1) To identify, among Look AHEAD participants, islet autoantibody signatures (patterns of antibody positivity, epitope specificity, titer, and IgG subclass) and determine their association and clustering with clinical features;
Aim 2) Determine the impact of islet autoimmunity on outcomes;
and Aim 3) To develop clinically useful a data visualization to display these findings to clinicians. Through these aims, we will achieve our overall goal to advance the epidemiologic knowledge and statistical tools for understanding islet autoimmunity in type 2 diabetes so that it can be used for precision diabetes care.
Type 2 diabetes affects 30.3 million people in the U.S., and the reasons why some people with diabetes suffer worse diabetes complications are not well understood. For many people with diabetes, differences in the immune system may play an important role in how diabetes progresses. This research project will measure immune system differences in blood samples from people with diabetes to understand how these differences affect their diabetes complications and outcomes so that we can develop better, more individualized treatment approaches.
