Influence of the dietary history on the prevention of celiac disease: continuation of the PreventCD study (PreventCD2).

Rationale: Coeliac disease (CD) is a chronic disorder caused by intolerance to
gluten and affects as much as 1-3% of the Europeans (5 million people). It is
treated with a gluten free diet (GFD), but prevention would be even more
beneficial. In 2007, the multicenter European PreventCD project was initiated,
from now on referred to as PreventCD. The purpose of PREVENTCD is to
investigate the influence of infant nutrition on the development of CD and
related autoimmune phenomena, as well as how genetic, immunological, and
environmental factors relate to this development (www.preventcd.com). The
hypothesis of the food intervention study in PreventCD-1 was that it would be
possible to induce tolerance for gluten in genetically predisposed children by
introducing small quantities of gluten during the period of breastfeeding. The
results (cohort age 3 to 6.6 years) have been published in October 2014 in the
New England Journal of Medicine 1. The most important findings were that
development of CD was not influenced by introducing small quantities of gluten
at 4 months of age, and that it is not influenced by the presence or duration
of breastfeeding. Moreover, we learned that CD develops already at a very young
age, more often in girls and more often in children homozygous for HLA-DQ2.

In PreventCD-2, the children have been followed to the age of 12 years to
evaluate the effect of the gluten intervention in the initial randomized study
(PREVENTCD-1)1. In PreventCD-2, (data at June 2019 (confidential unpublished
results) a total of 699 from the 944 original cohort of children are included
in the follow-up, 563 of them with TGA determinations during the last 3 years.
Mean age 10,1 years (range 8,4-12,0), 52% of them boys. 135 Children had been
diagnosed with CD and 158 small bowel biopsies had been performed. The mean age
at CD diagnosis is 4,3 years (SD 2,2) and 53% of the cases had symptoms, being
the most common abdominal pain (42.3%), diarrhoea (35.7%) and failure to
thrive/abdominal distention (both 30%). The cumulative incidence of CD at 9
years of age is 16,9% with incidences at 3, 5 and 7 years of 5,1%, 10.4% and
15,4% respectively. CD development significantly more frequent in girls at all
ages. The frequency of CD development was significantly higher in those
children homozygous for HLA-DQ2 (DR3-DQ2/DR3-DQ2, DR3-DQ2/DR7-DQ2), with a
cumulative incidence at 8 years of 33.2%. However, and intriguingly, this
effect of HLA-DQ2 homozygosity is only present in girls and not in boys.As well
as at the age of 3 years, the intervention with early introduction of gluten
did not have a significant effect on the development of CD at the age of 8
years (figure 4), neither for boys or for girls. It is essential to increase
the follow-up up to the age of 18 years to give insight in the natural history
and risk for the disease in this unique high risk population for CD , among
others to clarify, during and after puberty, the intriguing differences between
males and females

Primary: to evaluate the effect of the intervention of PreventCD at age 18
years in terms of frequency of CD and CD autoimmunity.

Observational study.

Nature and extent of the burden and risks associated with participation,
benefit and group relatedness: Participating families are well aware of the
fact that their child has an increased risk to develop CD and have experience
with participating in PreventCD. They are able to judge whether the burden of
annually obtaining blood and faecal samples for PreventCD-2 is in balance with
the burden from the work-up that their child would undergo whenever he/she is
suspected to have CD. In PreventCD, approximately, 45% of the children
diagnosed with CD were asymptomatic or had only mild symptoms. It is likely
that also in PreventCD-2, we will detect CD before the start of clinical
manifestations, allowing timely initiation of a GFD and preventing adverse
effects on the child*s growth, bone density, development, and health. Small
bowel biopsies will only be performed when medically indicated and not just for
the purpose of the study.