Case of the month from Hungary

A 12 years old boy was referred to our center for a second opinion.

He was diagnosed with new-onset type-1 diabetes mellitus in November of 2020. Routine screening for coeliac disease serum antibodies yielded positive results in a University Hospital on the 1st of December 2020 as follows:

Transglutaminase-2 specific serum IgA antibodies (TGA-IgA)  >100U  (>10xULN, cut-off value 5 U/ml, antigen: human recombinant TG2 with valine at position 224, natural form) and the same sample was also positive for endomysial antibodies (EMA-IgA), titre not reported.

He had weight and height within normal ranges and did not have enteral symptoms or complaints.

Since this was the time when the second wave of COVID-19 peaked in Hungary, the mother was asked by the paediatric gastroenterologist to perform a rapid coeliac test at home. The reported result (Biocard) was said to be positive upon which the diagnosis of coeliac disease was confirmed in a non-invasive way, and the family was instructed to start a gluten-free diet in addition to the diabetes diet.

The boy was next seen at the same gastroenterology outpatient unit in March 2021, after the pandemic restrictions had been lifted. His general condition was good.  The mother told that they never started the gluten-free diet and even did not decrease gluten intake, since the Biocard test only gave a very faint line according to their view. The mother presented a photo on which, in fact, a negative Biocard test result was seen. The boy consumed his prescribed carbohydrate meal with normal gluten-containing bread even at the GI unit. The family declared that they do not accept the diagnosis of coeliac disease based on the previous results.

The blood sample taken in March 2021 showed negative TGA-IgA and TGA-IgG values with the same ELISA test as done initially. HLA-DQ typing result was DQ2/X.

Now the doubt has arisen about the initial coeliac disease diagnosis and the following possibilities were checked: i/ possible sample mix-up at the initial diagnosis, ii/decrease of ingested gluten amounts, iii/possible damage of the villous structure in the gut despite current seronegativity.

Before heading for endoscopy, a new blood sample was drawn in April and was sent to another laboratory that uses a different TGA test (more sensitive capture ELISA). The results were as follows:

Serum TGA-IgA: 19.8 U/ml (normal < 3 U/ml), control test without TG2 antigen: 0.52 U, thus the positive result was TG2-specific.

Serum EMA-IgA: positive at the titre of 1:10, EMA-IgG negative (<1:2.5)

Upper endoscopy was done in June 2021, which showed Marsh IIIA lesion. The histology samples (5 tissue fragments) were also sent to another pathology lab for a second opinion and this also confirmed the presence of partial villous atrophy, villus/crypt ratio 2, high IEL (45/100 enterocytes), and according to the classification of Marsh, grade IIIA lesion. In this way, the diagnosis of coeliac disease has been secured and now is based on additional evidence. A strict gluten-free diet was prescribed.

Points to discuss:

• Main message: after a doubtful initial medical history with TGA/EMA positivity ever before and HLA-DQ2 positivity, checking gut condition is mandatory for a correct opinion/diagnosis

• Importance of family consultation before relying on a non-invasive diagnosis, possibilities to detect non-willingness to accept
• Was the initial non-invasive diagnosis correct in this case? High TGA-IgA plus EMA positivity in lab, second, independent TGA-IgA said to be positive on a rapid test
• Observer-related problems with Biocard/ rapid test execution and interpretation. Since the initial lab test was a correct positive, why was the photo negative?
• Differences in TGA-IgA levels in different laboratories, some TGA-IgA determinations are not sufficiently sensitive, in doubtful cases use another testing method
• How can we exclude sample mix-ups – in this case, the third lab test (April) confirmed initial TGA-IgA/EMA positivity, and in fact, when getting the original lab report from the March sampling, EMA-IgA was still reported positive, even is TGA-IgA was below the normal cut-off. Always read the original lab report, not the secondary interpretation by the GI department
• In case of discrepantly negative TGA result compared to the initial TGA+ one: wait or test or biopsy?
• Pitfalls in telemedicine – it is understandable that the peak of the second wave of COVID-19 pandemic was not the best time for the child with diabetes mellitus to come to the ward and to have the biopsy. However, simple solutions, like non-invasive CD diagnosis, might sometimes go wrong.

In summary: how can we explain the whole story? Why did the antibodies decline? Was the initial TGA-IgA value really high? [Probably yes, when thinking that this was a not very sensitive detection method, and ULN was still 6-7x in April]

Most probably gluten intake did decrease even if the diet was not complete.