Glucose tolerance and betacell function in islet

0 downloads 0 Views 521KB Size Report
Pediatric Diabetes 2013: 14: 341–349 ... Andersson C, Carlsson A, Cilio C, Cedervall E, Ivarsson S-A, Jonsdottir B, ... tolerance test (OGTT) were performed 8–65 d apart. ... were not associated with glucose metabolism, except for an increased ... type in T1D is DQ 2/8, found in nearly 30% of children ..... 58 (33–130).
Pediatric Diabetes 2013: 14: 341–349 doi: 10.1111/pedi.12023 All rights reserved

© 2013 John Wiley & Sons A/S

Pediatric Diabetes

Original Article

Glucose tolerance and beta-cell function in islet autoantibody-positive children recruited to a secondary prevention study Andersson C, Carlsson A, Cilio C, Cedervall E, Ivarsson S-A, Jonsdottir B, ˚ Elding Larsson H, for the ¨ Jonsson B, Larsson K, Neiderud J, Lernmark A, DiAPREV-IT study group. Glucose tolerance and beta-cell function in islet autoantibody-positive children recruited to secondary prevention. Pediatric Diabetes 2013: 14: 341–349. Aims: Children with type 1 diabetes (T1D) risk and islet autoantibodies are recruited to a secondary prevention study. The aims were to determine metabolic control in relation to human leukocyte antigen (HLA) genetic risk and islet autoantibodies in prepubertal children. Methods: In 47 healthy children with GADA and at least one additional islet autoantibody, intravenous glucose tolerance test (IvGTT) and oral glucose tolerance test (OGTT) were performed 8–65 d apart. Hemoglobin A1c, plasma glucose as well as serum insulin and C-peptide were determined at fasting and during IvGTT and OGTT. Results: All children aged median 5.1 (4.0–9.2) yr had autoantibodies to two to six of the beta-cell antigens GAD65, insulin, IA-2, and the three amino acid position 325 variants of the ZnT8 transporter. In total, 20/47 children showed impaired glucose metabolism. Decreased (≤30 μU/mL insulin) first-phase insulin response (FPIR) was found in 14/20 children while 11/20 had impaired glucose tolerance in the OGTT. Five children had both impaired glucose tolerance and FPIR ≤30 μU/mL insulin. Number and levels of autoantibodies were not associated with glucose metabolism, except for an increased frequency (p = 0.03) and level (p = 0.01) of ZnT8QA in children with impaired glucose metabolism. Among the children with impaired glucose metabolism, 13/20 had HLA-DQ2/8, compared to 9/27 of the children with normal glucose metabolism (p = 0.03). Conclusion: Secondary prevention studies in children with islet autoantibodies are complicated by variability in baseline glucose metabolism. Evaluation of metabolic control with both IvGTT and OGTT is critical and should be taken into account before randomization. All currently available autoantibody tests should be analyzed, including ZnT8QA.

Cecilia Anderssona , Annelie Carlssonb , Corrado Cilioa , Elisabeth Cedervallc , Sten-Anders Ivarssona , ¨ Berglind Jonsdottira , Bjorn d ¨ , Karin Larssone , Jonsson ˚ Jan Neiderudc , Ake a Lernmark , Helena Elding Larssona and for the DiAPREV-IT study group† a Department of Clinical Sciences, Lund ˚ University/CRC, Skane University ¨ Sweden; Hospital SUS, Malmo, b Department of Paediatrics, Lund ˚ University, Skane University Hospital SUS, Lund, Sweden; c Department of Paediatrics, Helsingborg Hospital, Helsingborg, Sweden; d Department of Paediatrics, Ystad Hospital, Ystad, Sweden; and e Department of Paediatrics, Kristianstad Hospital, Kristianstad, Sweden † See Appendix.

Key words: diabetes mellitus – FPIR – GAD65 autoantibodies – glucose tolerance – HLA genotype – IA-2 autoantibodies – insulin autoantibodies – secondary prevention – T1D – zinc transporter – ZnT8 Corresponding author: Cecilia Andersson, Department of Clinical Sciences, Lund University/CRC, ¨ Jan Waldenstroms gata 35, ˚ Skane University Hospital SUS, ¨ SE-20502 Malmo, Sweden. Tel: +46 40 39 11 12; fax: +46 40 39 19 19; e-mail: [email protected]

341

Andersson et al. Submitted 21 September 2012. Accepted for publication 4 January 2013

Type 1 diabetes (T1D) is one of the most common serious chronic diseases in children, with an increasing incidence worldwide. In Sweden alone, two children per day are diagnosed with T1D. The increase in T1D is greatest in children below 5 yr of age (1–4). Human leukocyte antigen (HLA)-risk genotypes and islet autoantibodies may be used to identify children at risk for T1D. The HLADQA1*0501-DQB1*0201(DQ2) and HLA-DQA1 *0301-DQB1*0302 (DQ8) haplotypes are associated with the highest risk for T1D. About 90% of children with T1D in Sweden have at least one of these HLADQ haplotypes (5), compared with about 45% in the general population (5). The most common HLA genotype in T1D is DQ 2/8, found in nearly 30% of children developing T1D compared to only 3% of the healthy population (5). Still, the risk of developing the disease during a lifetime is only 7% also with this genotype (6). Islet autoantibodies against GAD65 (GADA), insulinoma-associated protein 2 (IA-2A), insulin (IAA), or the three different variants at position 325 of the Zinc transporter 8 (ZnT8RA, Arginin 325 Zinc transporter 8 autoantibody; ZnT8WA, Tryptophan 325 Zinc transporter 8 autoantibody; and ZnT8QA, Glutamine 325 Zinc transporter 8 autoantibody) are immunological markers of T1D risk (7, 8). At the clinical onset of T1D, 95% have one or several of GADA, IA-2A, ZnT8A, or IAA (5, 9, 10). Among first-degree relatives and children with high risk HLA-DQ, the number of islet autoantibodies appears to be associated with an increased risk for clinical onset of T1D and is the best known predictor of T1D (7, 11–14). While a single autoantibody may not be predictive and often represents nonprogressive beta-cell autoimmunity, the appearance of multiple islet autoantibodies is a marker of a progressive autoimmune process (11, 12, 15–17). In studies of family members of T1D patients, 60–80% of individuals with two or more islet autoantibodies developed clinical diabetes over the next 5–15 yr, and population-based studies indicate that the risk is similar in the general population (18). As the autoimmune process proceeds, but before clinical symptoms of T1D, a gradual metabolic decompensation may be detected. A rising hemoglobin A1c (HbA1c) within the normal limits has been reported (19). Gradually increasing 2-h glucose values in oral glucose tolerance test (OGTT) have also been reported (20, 21) and repeated OGTTs may be a tool of evaluating remaining beta-cell function in children with an ongoing autoimmune process (22). Still, the

342

first-phase insulin response (FPIR) in the intravenous glucose tolerance test (IvGTT) was reported to predict diabetes in young children (23). A combination of parameters for glucose metabolism (OGTT, IvGTT, HbA1c, and fasting plasma glucose) with the number of islet autoantibodies may still be needed to increase the precision to predict T1D in children. In this study, we test the hypothesis that the number of islet autoantibodies is the best predictor for impaired glucose metabolism in prepubertal children with GAD65 autoantibodies and at least one additional islet autoantibody. The aims were to determine metabolic control (OGTT and HbA1c) and beta-cell function (IvGTT) in relation to age and body mass index (BMI) as well as HLA genetic risk and islet autoantibodies.

Subjects and methods Subjects Children and adolescents with genetic risk of T1D and GAD65 autoantibody and at least one more islet autoantibody, followed prospectively in the Diabetes Prediction in Sk˚ane (DiPiS) and The Environmental Determinants of Diabetes in the Young (TEDDY) studies or screened for autoantibodies in TrialNet, were eligible for the secondary prevention study Diabetes Prevention – Immune Tolerance (DiAPREV-IT). A total of 50 children aged 4–18 were included in the DiAPREV-IT study between May 2009 and January 2012. Of these, 47 were prepubertal and included in this analysis.

Methods Study design. At baseline in the DIAPREV-IT study a total of 54 children were screened. Four children did not fulfill the inclusion criteria. Although they had previously been positive, three children did not have either GADA or at least one more autoantibody at screening. One family withdrew the consent. In this study, another three children were excluded as they were not prepubertal and outliers in age and hence the reference values differed. Among the included children in this study 25 were boys and 22 were girls. The median age of the 47 children was 5.1 yr (range 4.0–9.2 yr). Body mass index. The Swedish mean ±1, 2, 3 standard deviation (SD) reference ranges of BMI were used (24). The child’s height and weight were measured and the SD of BMI was plotted on the curves as shown (24). Pediatric Diabetes 2013: 14: 341–349

Multiple autoantibodies and glucose metabolism

Plasma glucose. Venous plasma glucose was deter® ¨ mined in a bedside HemoCue (Angelholm, Sweden) glucometer. Normal range: fasting venous plasma glucose