D3.40 - Evaluation of wheat allergy diagnostic methods in children and potential cross-reactivity to other cereal grains

Complete avoidance of wheat and other cereal grains besides rice are common practices in parents if their child showed sensitisation (i.e. serum specific immunoglobulin E (sIgE) level ≥0.35 kUA/L) in Hong Kong. Unnecessary food restrictions may place children at risk of impaired growth. This study aimed to compare wheat sIgE against outcomes of double-blind, placebo-controlled food challenge (DBPCFC) and sIgE results for other common cereal grains.

81 children (51% male, median age 4.5 years, interquartile range 2.7 – 6.3) had bloods taken and underwent wheat DBPCFC to ascertain WA diagnosis as part of a 2-stage wheat oral immunotherapy trial (NCT06069492 under ClinicalTrials.gov). Serum sIgE of different cereal grains (wheat, rye, barley, oat, buckwheat) were measured by Phadia ImmunoCAP assays (ThermoFisher). For DBPCFC, wheat (active) or gluten-free (placebo) pasta, randomly assigned by dietitian, was divided into eight incremental doses at 15-minute intervals. The cumulative amount was 38.08 g (raw weight), equivalent to 4950 mg wheat protein. Challenge reactions were scored according to the PRACTALL guidelines.

61 children (75.3%) showed wheat sensitisation. 46 children (56.8%) failed wheat DBPCFC (i.e. WA) and 35 children (43.2%) passed DBPCFC. Their median wheat sIgE levels were 12.95 kUA/L (class 3) vs. 0.31 kUA/L (class 0) (p <.001)^a. Wheat sIgE class had significantly greater odds of failing DBPCFC (OR = 2.33, 95% CI [1.59, 3.41], p <.001)^b. Wheat sIgE showed good ability to predict DBPCFC outcomes with area-under-curve 0.85, 95% CI [0.76, 0.94] (p <.001)^c. Optimal cut-off point for wheat sIgE was determined to be 1.38 kUA/L (class 2), which yielded a sensitivity of 87% and a specificity of 74.3%. Wheat sIgE was associated with cross-reactivity to sIgE for rye (r_s = .96), barley (r_s = .93), oat (r_s = .83) and buckwheat (r_s = .49) (p <.001 for all)^d. 

^a analysed by Mann-Whitney U test with U = 265.5, z = -5.24 with a large effect size, r = .58. ^b analysed by binary logistic regression adjusted for age and gender. ^c analysed by receiver operator characteristic analysis. ^d analysed by Spearman’s rank-order correlations, n = 81

Immediate-type allergic reactions following wheat ingestion has modest reliability for predicting clinical WA. Oral food challenges should be performed in equivocal cases to ascertain WA diagnosis to avoid unnecessary food restrictions. A higher wheat sIgE cut-off may be considered when DBPCFC is not possible. Degree of cross-reactivity exists between wheat and rye, barley, oat and buckwheat in a descending order. Nutrition intervention by dietitians on food alternatives may improve growth of children with WA and need to eliminate other cereal grains concurrently. (Funded by Health and Medical Research Fund, Health Bureau of Hong Kong SAR [reference 10210336])