GADIA2-combi determination as first-line screening for improved prediction of type 1 diabetes in relatives.

Abstract

A new radiobinding assay for the simultaneous detection of antibodies to GAD and the tyrosine phosphatase IA2 has been recently described in patients with newly diagnosed type 1 diabetes. Here we assessed sensitivity and predictive value of this GADIA2-combi test in first-degree relatives of type 1 diabetic patients compared with islet cell antibody (ICA) and insulin autoantibody (IAA) screening. Of 1,606 relatives, 77 (4.8%) had elevated GADIA2-combi titers above the 99th percentile of 105 nondiabetic control subjects, and results were confirmed by testing these samples for GAD antibody (GADA) and tyrosine phosphatase IA2 antibody (IA2A) in the single antibody test (29 GADA+/IA2A+, 44 GADA+/IA2A-, and 4 IA2A+/GADA-). A further 9 of 1,606 relatives had detectable ICA (1) or IAA (8), but they were negative in the GADIA2-combi assay as well as in the single test for GADA or IA2A. Twenty-four relatives progressed to IDDM within a median follow-up time of 5.6 years (range 0.5-8.2). The sensitivity of antibody determination in relatives with progression to IDDM was 92% for the GADIA2-combi assay, 96% for the combined testing of IAA and GADIA2-combi antibodies, and 83, 67, 67, and 79%, respectively, for GADA, IA2A, IAA, or ICA testing alone. The cumulative life-table risk of antibody-positive relatives was related to GADIA2-combi titers (5-year risk: >50 U, 51% [95% CI 30-73]; >10 to 50 U, 12% [1-24]; <10 U, 0.17% [0-0.5]; P=0.0001) and on the presence of IA2A in addition to GADA (5-year risk: GADA+/IA2A+, 47% [25-68]; GADA+/IA2A-, 15% [2-28]; P=0.006). In those with detectable antibodies, risk was not associated with age (<15 vs. >15 years) or relation to proband (offspring, sibling, parent). Relatives with GADIA2-combi antibodies >10 U and the additional presence of IAA had a slightly higher diabetes risk than relatives without IAA (5-year: IAA+, 46% [23-68]; IAA-, 19% [6-32]; P=0.07). Furthermore, low first-phase insulin release after intravenous glucose tolerance test was associated with risk in relatives with GADIA2-combi antibodies (P=0.01). These results indicate that the GADIA2-combi test is a valuable marker for first-line screening and risk assessment of type 1 diabetes in relatives. It can be used for venous as well as capillary blood samples.

Diabetes Antibody Standardization Program: evaluation of assays for autoantibodies to glutamic acid decarboxylase and islet antigen-2

“…Although used in fewer laboratories, the ELISA for IA-2A also improved and achieved AUC equivalent to those of in-house RIA, although results of the two assay formats correlated less well than for GADA. A small number of combined assays for GADA and IA-2A were included in DASP 2005 and, as expected, some achieved high sensitivity, confirming the value of this approach for initial screening [14][15][16], particularly since one of these assays used the ELISA format. In addition, the IA-2 clone used was a determinant of differences in assay performance; the relatively small numbers of RIA using the IA-2bdc giving significantly lower sensitivity and AUC than those using IA-2ic.…”

Karlsburg Type I diabetes risk study of a general population: frequencies and interactions of the four major Type I diabetes-associated autoantibodies studied in 9419 schoolchildren

“…Our results suggest that the co-occurrence of Type I diabetes-associated AAbs detected in persons of a general population has a similar predictive value for developing diabetes as described for first-degree relatives [9,17,25]. In both cohorts, the diabetic subjects' and AAb-positive schoolchildren's levels of each AAb were significantly (p < 0.05) increased by AAb co-occurrence, which is consistent with data from first-degree relatives of Type I diabetic patients [7,8] (Table 1).…”

Type 1 diabetes risk assessment: improvement by follow-up measurements in young islet autoantibody-positive relatives

“…This study was performed on 141 of the 180 autoantibodypositive relatives from the Bart's-Oxford (BOX; n=67) and the Munich family studies (n=74) [12,13] previously used to define the risk stratification models [11]. Relatives for this study were selected on the basis of the availability of followup samples to allow testing of all antibody characteristics (titre, IgG subclasses of IAA and IA-2A, antibodies to IA-2β ).…”