Predictive Autoantibodies Before the Diagnosis of Type I Diabetes in Adults

Technical Summary: Predictive Autoantibodies Before the Diagnosis of Type I Diabetes in Adults

Problem Statement
While Type 1 Diabetes (T1D) is well-characterized in pediatric populations, adult-onset T1D (AOT1D) remains clinically underrecognized despite recent epidemiologic data suggesting it may be more common than childhood-onset T1D. A critical gap exists in understanding the natural history of islet autoimmunity in adults. Specifically, the temporal emergence, trajectories, and predictive value of islet autoantibodies prior to the clinical diagnosis of AOT1D are poorly defined compared to pediatric cohorts. This study addresses the need to characterize the prodromal phase of AOT1D to determine if a prolonged preclinical window exists that could facilitate risk stratification.

Methodology
The authors conducted a retrospective study utilizing the United States Military Health System (USMHS) electronic medical records (EMRs) and the Department of Defense Serum Repository (DoDSR).

• Cohort Selection: From an initial pool of 3,332 potential cases identified via ICD codes, the study focused on 169 individuals with confirmed AOT1D (defined by the presence of islet autoantibodies at diagnosis) and 40 age- and sex-matched healthy controls.
• Sample Collection: The study analyzed 803 longitudinal serum samples, comprising 643 prediagnostic samples from the AOT1D cohort and 160 samples from controls. Samples were collected between 1970 and 2019, with most cases having samples within two years of diagnosis.
• Assay Technique: Autoantibodies were measured using Luciferase Immunoprecipitation Systems (LIPS) assays against five antigens: GAD65 (GADA), IA-2, IA-2β, and two ZnT8 isoforms (ZnT8-R and ZnT8-W). Notably, insulin autoantibodies (IAA) were not assessed due to assay unavailability.
• Blinding and Analysis: Testing was performed in a blinded manner by a single investigator. Positivity thresholds were established based on the mean plus three standard deviations of the 160 healthy control samples. Statistical analyses included Mann–Whitney U tests for group comparisons and temporal grouping of samples into 5-year intervals preceding diagnosis.

Key Contributions and Results
The study provides the first detailed longitudinal characterization of autoantibody emergence in AOT1D using pre-diagnostic serum samples.

• High Prevalence of Pre-diagnostic Autoimmunity: 85% (144/169) of AOT1D cases were seropositive for at least one autoantibody prior to clinical diagnosis. Among the 643 prediagnostic samples, IA-2 autoantibodies were the most prevalent (50%), followed by GADA (46%), IA-2β (34%), ZnT8-R (27%), and ZnT8-W (15%).
• Extended Preclinical Phase: Seropositivity was detected decades before diagnosis. The proportion of seropositive subjects increased progressively as diagnosis approached: 38% were positive >20 years prior, rising to 44% (15–20 years), 59% (10–15 years), 73% (5–10 years), and 91% within 5 years of diagnosis. The earliest detected case showed seropositivity 27 years prior to diagnosis.
• Autoantibody Patterns and Timing:
  • Multiplicity: Among seropositive individuals, positivity for two or more autoantibodies was the most common pattern (50%).
  • Temporal Sequence: Isolated GADA positivity appeared earliest (median onset 7.9 years before diagnosis). In contrast, multiple-autoantibody positivity, IA-2/IA-2β positivity, and ZnT8 positivity emerged later, with median onsets of 4.6, 4.5, and 1.9 years before diagnosis, respectively. These differences in timing were statistically significant.
  • Trajectories: Distinct patterns emerged, including persistent seropositivity lasting over a decade (often with multiple antibodies) and isolated early GADA that either remained restricted or progressed to multiple antibodies. A minority of cases remained seronegative until shortly before diagnosis.

Significance and Claims
The paper claims that these findings extend observations from pediatric T1D to adults, demonstrating that AOT1D is frequently preceded by a prolonged period of progressive autoimmune activity. The authors assert that:

1. AOT1D has a long preclinical phase: Autoimmunity often begins decades before clinical onset, highlighting a significant window for potential risk stratification.
2. Distinct Autoantibody Kinetics: GADA is typically the earliest marker in adults, whereas IA-2 and ZnT8 appear closer to disease onset, suggesting different rates of disease progression associated with specific autoantibody profiles.
3. Clinical Implications: The identification of a prolonged preclinical phase supports the potential utility of targeted screening strategies in adults, particularly focusing on GADA and IA-2. The authors suggest that identifying adults during these early stages could be crucial for applying disease-modifying therapies (such as teplizumab) to preserve residual β-cell function, though they note that such applications require further evaluation.

The study acknowledges limitations, including the lack of insulin autoantibody data, incomplete genetic and environmental data, and variable sampling intervals, which may limit the precise determination of seroconversion timing in some individuals. However, the authors maintain that the use of longitudinal serum samples provides a unique and valuable view of the natural history of AOT1D.