Statistics from Altmetric.com
Long term exposure to gluten in coeliacs,1 and coeliac disease (CD) diagnosis after 16 years of age2 may induce type 1 diabetes (T1D) and other autoimmune disorders. Increased prevalence of CD among diabetics and their relatives is well documented.3 Early introduction of gluten to children at high risk for T1D produces T1D associated islet autoantibodies.4 Similarly, in the absence of overt clinical symptoms of T1D, some coeliac children produce diabetes autoantibodies in a gluten dependent manner.5 In diabetics, intestinal challenge with gluten produces mucosal recruitment of lymphocytes,6 similar to that in CD patients.7 In diabetics, however, there is no production of CD related anti-tissue transglutaminase antibodies (anti-tTG).6
We have used a phage display assay8 to show that in CD patients, production of anti-tTG is limited to the intestine. Here, we monitored the effects of a gluten free diet (GFD) on anti-tTG antibody synthesis in the intestinal mucosa of a diabetic adult and a boy at high risk of diabetes, both carrying HLA DQ2/DQ8, but lacking serum anti-tTG. Intestinal specimens from both subjects and samples of peripheral blood lymphocytes were used to make phage-antibody libraries8 to look for lymphocytes synthesising anti-tTG antibodies.
Patient No 1 was a 35 year old man who had T1D for 20 years. During 1998–2001, serum anti-tTG responses were negative, and clinical control of T1D was good (mean glycosylated haemoglobin 6.8% (range 8.1–6.1) but the patient developed surgically treated diabetic retinopathy and microalbuminuria, with an average albumin excretion rate (AER) of 230 μg/min, despite treatment with angiotensin converting enzyme inhibitors. In 2001, “burning” epigastric pain appeared with abdominal distension. Duodenal biopsy and number of intraepithelial lymphocytes were normal.
Patient No 2 was a two year old boy at risk of CD and T1D (diabetic father and coeliac brother) who tested positive for two of six HLA T1DM specific genotypes (DR1 *0301, DQA1*0501, DQB1*0201 and DRB1*0401, DQA1*0301, DQB1*0302). Tests for anti-tTG serum antibodies were negative while anti-islet cell antibodies (ICA) became positive at 20 months. Informed of the potential risks, the child’s parents consented to intestinal biopsy to detect possible silent CD. Duodenal biopsy and number of intraepithelial lymphocytes were normal.
In both subjects, positive tTG antibody clones (table 1) were isolated only from the intestinal lymphocyte libraries. Two control subjects aged 10 and 45 years, suffering from Helicobacter pylori gastritis and with no family history of CD or T1D, tested negative for intestinal anti-tTG clones. The diabetic adult and the child’s parents agreed to a GFD for 12 months, after which laboratory tests and biopsy were repeated. In the adult, control of diabetes was unchanged but AER was markedly improved (20 μg/min). The boy tested negative to ICA. In both subjects, biopsies were normal, and analysis of new phage antibody libraries showed complete elimination of anti-tTG clones in the adult and 90% reduction in both positivity and diversity in the child (table 1). Both patients remain on a GFD; AER is still normal in the diabetic adult and the child is still negative for ICA.
In the context of genetic predisposition to gluten intolerance, in line with Maki’s data on the gradual development of CD in diabetics,9 we found a gluten dependent immunological response, apparently only in the intestine. In the boy, reduced response to tTG and elimination of ICA after GFD may have been due to very early intervention, indicating temporary protection from the disease. In the diabetic adult, reduction of microalbuminuria may have indicated that while a GFD is of little benefit to the pancreas, improvements can still be obtained in other organs.
In conclusion, at risk subjects with HLA DQ2/8 may develop intestinal anti-tTG antibodies on extended exposure to gluten. Similar larger scale studies are needed to prove that gluten is harmful in these subjects and confirm the benefits of a GFD.
The paper was founded in part by MURST COFIN 2004060237/3 and RFI 149/03 IRCCS “Burlo Garofolo”.
Conflict of interest: None declared.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.