Islet autoantibody positivity is lost from slow progressors who remain diabetes free despite historical multiple antibody positive status.

Introduction: Most young children with multiple islet autoantibodies (mAabs) develop diabetes by 10 years of follow-up. Using data from several longitudinal studies of diabetes, we identified a group of slow progressors; children and adults who remain diabetes free for at least 10 years following mAab detection. The longitudinal islet autoantibody characteristics of these slow progressors were examined.

Methods: An index sample, found mAab positive with harmonised (GADA or IA-2A) and/or IASP-validated local assays was compared with the last available sample from each individual, taken at least 10 years later. Three individuals were diagnosed (14 days, 4 months, and 12 years) after this sample. Individuals (n=40) from the DAISY, BOX and Pittsburgh family studies had data suitable for analysing changes in antibody prevalence over time (McNemar’s Test) and with antibody, HLA combinations, age at index samples, or sex (Fisher’s exact). Antibody levels were compared for samples tested by the same method from the BOX and Pittsburgh studies (n=25, Wilcoxon Matched Pairs on z-score data).

Results: Positivity was lost in 28% of Slow Progressors with GADA (p=0.016), 67% of those with IAA (p=0.001), 55% of those with ZnT8A (p=0.009), and 33% of those with IA-2A (p>0.05). GADA were more likely to be lost if IAA were present in the index sample (p=0.007). IA-2A were more likely to be lost if mAab were originally detected before 10 years of age (p=0.017). The presence of HLA Class II DQ2 was weakly associated with more GADA loss (p=0.052). The number with mAabs decreased from 100% to 43% (p<0.001) and levels of GADA (p=0.043), IA-2A (p=0.011), and ZnT8A (0.007) declined between sampling.

Conclusion: Loss of all autoantibody types is typical of Slow Progressors. The humoral immune response is not continuing to amplify in these individuals, which suggests ongoing regulation of the aberrant immune response.