AESKULISA and H-TTG/DGP screen for celiac disease diagnosis.

Serum testing is recommended as the first step for celiac disease (CD) diagnosis and IgA class anti- transglutaminase (tTG) is actually the test of choice, although limited in sensitivity in patients with total IgA deficit. New screen tests for CD are now available, which allow to recognize IgA and IgG antibodies anti different CD associated antigens in one step. The aim of this study was to verify the clinical performance of two screen tests, one provided by Inova (IL, Italy) which identifies IgA and IgG anti DGP (deamidated gliadin peptides) and tTG (h-tTG/DGP), and another provided by Aesku.Diagnostics (Grifols, Italy), which identifies IgA and IgG anti neo-epitopes of tTG, resulting from the cross linking of tTG and gliadin specific peptides (Aeskulisa). We retrospectively analysed a series of 500 children (173 M, 327 F) consecutively subjected to EGD for abdominal symptoms. Based on duodenal histological findings, CD was diagnosed (n=232) or ruled out (n=268). In all sera collected immediately before endoscopy and stored at -20 °C. h-tTG/DGP and Aeskulisa screen tests were performed using an automated ELISA system (DSX, Technogenetics, Italy). The areas under the ROC curves ± SE were 0.96±0.09 for h-tTG/DGP and 0.94±0.011 for Aeskulisa. 32.66 U/mL and 45.1 U/mL were identified as best cut-off (delta positive rate method) for h-tTG/DGP and Aeskulisa respectively. The following results were found for h-tTG/DGP and Aeskulisa respectively: 94.8% and 94.3% sensitivity; 89.1% and 85.8% specificity; 90.7% and 88.3% PPV; 93.8% and 93.0% NPV. On considering CD children only, both h-tTG/DGP and Aeskulisa screen test results were correlated with the degree of duodenal atrophy (Marsh Oberhuber criteria), the highest levels being recorded among those with type IIIa-c than those with type I-II (F=18.8, p<0.0001 and F=14.7, p<0.0001 respectively). In children categorized on the basis of age quartiles we obtained the following results: the areas under the ROC curve of h-tTG/DGP and Aeskulisa were 0.97±0.018 and 0.95±0.023 in the 0-4 yrs group; 0.967±0.013 and 0.96±0.017 in the group 5-8 yrs; 0.94±0.029 and 0.92±0.038 in the group 8.5-11 yrs; 0.947±0.020 and 0.93±0.023 in the group > 11 yrs. In conclusion the new screen tests appear sensitive and specific enough to be proposed as first step in the flow-chart of CD diagnosis in children of any age.