QUANTIFICATION OF HUMAN CYTOPLASMIC ISLET-CELL ANTIBODIES WHICH CROSS-REACT WITH MOUSE PANCREAS - A FOLLOW-UP-STUDY IN TYPE-1 (INSULIN-DEPENDENT) DIABETIC-PATIENTS AND IN 1ST-DEGREE RELATIVES

We studied the heterogeneity of cytoplasmic islet-cell antibodies for cross-reaction with mouse pancreas in 31 recent-onset Type 1 (insulin- dependent) diabetic patients and 31 first-degree relatives with islet- cell autoantibodies detected on human pancreas. Only six Type 1 diabet ic patients displayed islet-cell antibodies binding to human pancreas but not to mouse pancreas. Among 15 first-degree relatives displaying such antibodies which did not react with mouse pancreas, including one identical twin and one subject with polyglandular autoimmunity, none developed diabetes or even lost acute insulin response to intravenous glucose after 5 years of follow-up. By contrast, 14 of 20 (70%) of the Type 1 diabetic patients with islet-cell antibodies detected on human pancreas, and five first-degree relatives who progressed to a loss of acute insulin response to glucose and then to either Type 1 diabetes or glucose intolerance, also displayed antibodies reactive with mouse islets. Surprisingly, islet-cell antibodies were detectable on mouse p ancreas but not on human pancreas in four Type 1 diabetic patients and in one relative who progressed to diabetes. In the five relatives who progressed to metabolic abnormalities, islet-cell antibody titres on mouse pancreas, quantified by the fluorescence intensity per islet at each serum dilution, progressively increased concomitantly with the lo ss of acute insulin response to glucose, whereas islet-cell antibody t itres on human pancreas remained stable. The usefulness of such quanti fication was also validated by the fact that antibody titres on mouse pancreas were decreased after 3 months (p < 0.01) in recent-onset Type 1 diabetic patients, while titres on human pancreas were not. Our res ults confirm that the use of mouse pancreas, combined with the convent ional assay on human pancreas, reveals the heterogeneity of islet-cell autoantibodies. The presence of cross-species reactive islet-cell aut oantibodies in subjects at risk may improve the predictive value, indi cating relatives at lower risk whose antibodies are unable to bind to mouse islets. It could also allow the identification of subjects who p rogress to the disease without ever displaying islet-cell autoantibodi es detectable on human pancreas. The increase of antibody titres on mo use pancreas during subject follow-up could be indicative of the worse ning of the course during the prediabetic phase. Finally, islet-cell a utoantibodies detected on mouse pancreas may be more transient after t he onset of diabetes than is the more complex mixture of antibodies de tected on human tissue.