Patient factors influencing acute gluten reactions and cytokine release in treated coeliac disease

Abstract

Background: Patients with coeliac disease (CD) commonly report a variety of adverse symptoms to gluten, but descriptions of the symptomatic response in the literature may have been confounded by the presence of food components such as fermentable carbohydrates (FODMAPs) causing symptoms of irritable bowel syndrome independent of gluten. In recent unmasked and masked low FODMAP gluten challenge studies in small groups of treated CD patients, nausea and vomiting were shown to be the key symptoms associated with serum interleukin (IL)-2 release. Our objective was to utilise a large and diverse cohort of people with CD undertaking a standardised gluten food challenge to characterise the demographic, genetic and clinical factors influencing the severity and timing of acute gluten reactions and IL-2 production.

Methods: A total of 295 adults treated for CD were observed for 6 h after an unmasked food challenge consisting of 10 g vital wheat gluten (low in FODMAPs) in 100 ml water. Assessments included patient-reported outcomes, serum IL-2 and adverse events. Responses were analysed according to patient characteristics, HLA-DQ genotype, duodenal histology and response to a second gluten challenge.

Results: Peak symptom severity was at 3 h (median severity 5/10). Peak IL-2 was at 4 h (median 4 pg/ml, range undetectable to 1028 pg/ml). Older age, older age at diagnosis, HLA-DQ2.5 positivity and homozygosity for HLA-DQB1*02 were each significantly associated with IL-2 elevations after gluten. Patients positive for HLA-DQ2.5, DQ8, DQ2.2 or DQ7 showed elevated IL-2 after gluten. Patient factors were not significantly associated with severity of digestive symptoms, but symptoms were correlated to one another and serum IL-2. Gluten challenge after 5 months caused more vomiting and higher IL-2 levels, but responses correlated with the first.

Conclusions: Gluten-induced symptoms and cytokine release is common in adults with treated CD. Age, genetics and previous response to gluten predict these acute reactions to gluten challenge. Structured symptom assessment and serum IL-2 after standardised gluten challenge may inform on patient diagnosis, the role of gluten in symptomatology and the need for adjunctive treatment.

Trial registration: ClinicalTrials.gov , NCT03644069 Registered 21 May 2018.

Keywords: Coeliac disease; Cytokine; Food challenge; Gluten; Patient-reported outcome.

Fig. 1

Study outline

Fig. 2

Six-hour time series after gluten for patient-reported outcome scores (modified CeD PRO and Global Symptom Survey). Profiles are for all patients (a–c) and for patients whose worst global gastrointestinal score over 6 h was “no symptoms” (d–f), “mild symptoms” (g–i), “very mild symptoms” (j–l), “moderate symptoms” (m–o) and “severe symptoms” (p–r), “very severe symptoms” (s–u). Points indicate mean ± standard error of the mean for scores rated 0 to 10

Fig. 3

Serum IL-2 elevation in 295 patients over 6 h after gluten challenge. a Median and interquartile range of serum IL-2 concentrations at 2-h intervals after gluten; significance tested by Wilcoxon matched-pairs signed rank test. b The frequency distribution of peak serum IL-2 concentrations in patients after gluten. c Median and interquartile range of serum IL-2 concentrations at 4 h and peak hourly numerical rating for global severity of digestive symptoms (GloSS). d Median and interquartile range of serum IL-2 concentrations 4 h and peak hourly severity of nausea. e Median and interquartile range of serum IL-2 concentrations at 4 h and whether patients experienced vomiting after gluten challenge. f Mean and standard error of the mean of serum IL-2 concentrations after gluten according to gluten reaction severity (peak severity descriptor of digestive symptoms rated each hour in the GloSS). g Peak serum IL-2 concentrations after gluten according to villous height to crypt depth ratio (VH:CrD) in the second part duodenal biopsies indicating the presence of villous atrophy (VH:CrD ≤ 2) (n = 39 patients) or absence (n = 13). h Paired serum IL-2 concentrations at 4 h after two separate food challenges with 6-g gluten protein. Data for 36 patients who received placebo treatment during the RESET CeD Study and had both a screening (unmasked) and a second (masked) challenge 20 to 22 weeks later. The gluten challenges used the same format administering the equivalent of 6-g gluten protein in 10-g vital wheat gluten mixed in water