Journal of Clinical Pediatrics ›› 2024, Vol. 42 ›› Issue (12): 1025-1031.doi: 10.12372/jcp.2024.23e1154

• Original Article • Previous Articles     Next Articles

Role of anti-tissue transglutaminase IgA antibody titer and HLA-DQ typing in the diagnosis of type 1 diabetes mellitus with celiac disease

HAO Huiling(), ZHANG Gaixiu, FENG Mei   

  1. Department of Endocrinology, Shanxi Children's Hospital, Taiyuan 030013, Shanxi, China
  • Received:2023-12-05 Published:2024-12-15 Online:2024-12-02

Abstract:

Objective To investigate the predictive value of anti-tissue transglutaminase IgA (IgA-anti-tTG) antibody titers and human leukocyte antigen (HLA) -DQ typing in the occurrence of type 1 diabetes mellitus (T1DM) associated with celiac disease (CD). Methods All T1DM children<18 years old who visited the endocrinology department from November 2019 to November 2023 and healthy children who underwent physical examination during the same period were selected as the study objects. All participants underwent serological analysis and detection of CD-related HLA alleles (DQ2 and DQ8). All children with positive IgA-anti-tTG antibody underwent duodenal pathological biopsy (Marsh grade). The correlation between IgA-anti-tTG titer and Marsh classification was analyzed. ROC curve was used to evaluate the diagnostic value of IgA-anti-tTG in predicting CD. Results A total of 577 patients with T1DM were enrolled, of whom 314 (54.4%) were females with a median age of 10.0 (5.0 to 16.0) years. The control group consisted of 150 patients, including 72 females (48.0%), with a median age of 9.5 (6.0 to 17.0) years. In the T1DM group, 77 patients (13.3%) had a positive IgA-anti-tTG antibody response (>18 IU/L), of whom 60 (77.6%) had positive duodenal histopathology. Nineteen children were in Marsh 2 stage, 4 were in Marsh 3a stage, 18 were in Marsh 3b stage, 19 were in Marsh 3c stage, and all of whom were diagnosed with CD. The remaining 17 patients had negative duodenal histopathology, and after other tests, they were finally diagnosed with wheat allergy or gluten sensitivity. Duodenal pathological biopsy was also performed in 46 children in the T1DM group who were negative for IgA-anti-tTG antibody but still had glut-dependent symptoms. Among them, 21 patients were found to be Marsh stage 2 or 3 and were also diagnosed as CD. Finally, a total of 81 children were identified as having T1DM combined with CD. Spearman correlation analysis showed a significant positive correlation between IgA-anti-tTG antibody level and Marsh stage in 81 children with T1DM and CD (rs=0.76, P<0.001). The area under ROC curve (AUC) of IgA-anti-tTG antibody titers predicting CD-positive was 0.83 (95%CI: 0.72-0.93). The optimal cutoff threshold for IgA-anti-tTG antibody titer was 87.65 IU/L, and its sensitivity and specificity for predicting CD were 83.19% and 92.20%. The serum IgA-anti-tTG antibody reaction in the control group was negative. There was no significant difference in HLA-DQ typing distribution between T1DM group with CD and without CD (P>0.05). Conclusions The titer of IgA-anti-tTG antibody has a certain value in the diagnosis of T1DM children complicated with CD, and HLA typing may be an auxiliary screening method for serum antibody detection results.

Key words: type 1 diabetes mellitus, anti-tissue transglutaminase IgA, celiac disease, duodenal biopsy, human leukocyte antigen