Use of gluten immunogenic peptide (GIP) assays to provide objective confirmation of gluten ingestion will be one way to improve management of the patient with non-responsive coeliac disease.

Dr Jason Tye-Din, head of the Coeliac Research Lab at the Walter and Eliza Hall Institute, told AGW 2018 that neither symptoms nor serology were reliable enough in detecting transgressions of the gluten free diet.

Yet deliberate or inadvertent gluten intake was one of the most likely causes of non-responsive coeliac disease.

A recent meta-analysis found people with coeliac disease on a gluten free diet were inadvertently exposed to more than 200 mg/d of gluten, based on stool and urine GIP tests.

Expert dietetic evaluation and consideration of psychological drivers of adherence could also help reduce gluten contamination.

Dr Tye-Din said the work up for non-responsive coeliac disease should also back track to review the initial serology and histology and confirm the patient had been given the correct diagnosis.

Up to 10% of patients with non-responsive coeliac disease sent for tertiary evaluation did not have coeliac disease, he said.

“The positive predictive value of villous atrophy for coeliac disease is not 100%.”

Patchy disease, biopsy orientation, and ultra-short coeliac disease were potential confounders.

Dr Tye-Din said differential diagnoses included infective conditions such as tropical sprue, H. pylori, Giardia lamblia, small intestinal bacterial overgrowth (SIBO) and Whipple’s disease.

Immune-related causes of villous atrophy included common variable immune deficiency (CVID), autoimmune enteropathy, Crohn’s disease and cow’s milk protein intolerance while drugs such as olmesartan and NSAIDs could also be involved.

And there were the basics to consider.

“Beware the patient who is seronegative at diagnosis. Seronegative coeliac disease is unusual with modern, sensitive serology,” he said. “Beware also the ‘empiric trial of a gluten free diet’.”

HLA typing had a useful negative predictive value while a gluten challenge and repeat biopsy, with multiple spaced samples, may be required.

Dr Tye-Din said true refractory coeliac disease probably represented less than 20% of patients with non-responsive disease.

A reasonable expectation of time for symptom improvement on a gluten free diet was weeks to months while serological and histological improvements could take months to years.

“HLA-DQ2-5 homozygosity, the gene dose effect, supports the notion that some patients are biologically more sensitive to gluten,” he added.

A recent US study in refractory coeliac disease suggested that most patients managed with open capsule budesonide could achieve a clinical (92%) and histologic (89%) improvement.

Anti-IL-15 and JAK inhibitors were also promising.