We selected 12 consecutive CD patients on GFD for at least one year, at follow-up in our Celiac Clinic, and meeting the following selection criteria: strict compliance with the GFD, absence of symptoms, reconstitution of villous structure and negative tissue transglutaminase (t-TG) and/or antiendomysial (AMA) antibodies during GFD. Compliance with the GFD was assessed as previously described  using a 4 point Likert scale that includes no dietary indiscretions (score 1), 1 serving with gluten per month (score 2), < 4 servings per month (score 3) or = > 4 servings per month (score 4). We also selected 7 CD patients freshly diagnosed with villous atrophy and positive CD related serology and on a gluten containing diet. Twelve asymptomatic healthy subjects selected among the health care professionals in our Institution volunteered in the study as normal controls.
CD patients on GFD entered a single blind cross over study involving challenge with 3 proteins (random order): rice (MyProtein, Cent Ltd, Northwick, UK) as atoxic control, pure gluten (Amygluten, Tereos Syral, Marckolsheim, France) as toxic control, and Tm (Triticum monococcum ssp monococcum, cultivar “Monlis”, CRA, Rome, Italy) as investigational protein. Challenge with different proteins was carried out on 3 separate occasions on day 0, 14 and 28. The primary endpoints were the effect of challenge on the urinary recovery L/R ratio as a measure of intestinal permeability, and the effect on symptoms. Patients were instructed to report any symptom experienced during the challenge, which was graded for severity according to the WHO toxicity grading scale as mild (grade 1), moderate (grade 2), severe (grade 3) and life threatening (grade 4) . We used a 2.5 g protein dose for the challenge, corresponding approximately to one slice of bread and to the dose used by others [16–18] for “proof of the concept” challenge studies.
The timetable in the study was as follows: fasted patients were admitted to a day-case Unit, asked to empty the urinary bladder and immediately after they were asked:
t 0 to eat a gluten free pudding (BiAglut VAN, Heinz, Italy) with dispersed 2.5 g cereal protein
t 2 h to drink a solution of 5 g lactulose + 1 g rhamnose in 60 ml water
t 2-7 h to collect urine and to record symptoms
At t 7 h urine volume was measured, a sample was retained, frozen and stored until analysis. Urinary samples were analyzed by using HPLC  for L and R concentration and for calculation of the L/R ratio in one batch for each patient. The normal urinary L/R, for our laboratory, calculated in 40 healthy controls is 0.045. All analyses were carried out at the Burlo-Garofolo Paediatric Hospital in Trieste (Italy) under the supervision of one of the authors (T.N.).
The same protocol used for CD patients on GFD challenged with cereal proteins was also adopted, with the exclusion of protein challenge, to measure intestinal permeability in 7 CD patients on gluten containing diet and in 12 healthy controls. Five CD patients on GFD and 5 healthy controls were also studied with the same protocol on 2 occasions, one day apart, to test for reproducibility of results.
Results were expressed as mean ± SD. Paired or unpaired t-test and χ2 test were used as appropriate to compare continuous and categorical variables. The statistical analysis was carried out using GraphPad Prism 5 statistical package (GraphPad Software, San Diego, Ca, USA). The study was approved by the Ethics Committee of Spedali Civili of Brescia on February 5th, 2008 and was given the number n2008-000697-20 in our national registry of clinical trials (EudraCT-AIFA). Patients and controls signed a written informed consent to the study.