Transcript Slide 1

Progress of the Prospective
study - Medicel
Prof. Dr. Giuseppe Magazzu
Prof. Dr. Dušanka Mičetić-Turk
and all
Background
 Celiac disease (CD)has increased at
an unexpected rate in the last two
decades not only in Europe but also in
Mediterranean regions.
 The consequences of unrecognized
CD can be severe particularly with
regard to mortality and morbidity
among these patients.
The aims of study:
 To identify the factors that may
influence the onset of CD in
Mediterranean basin
 To identify clinical and laboratory data
which can predict the diagnosis of the
disease
 To determine where the new ESPGHAN
diagnostic criteria could be applied and
what % of patients can omitt the biopsy
Service and survey design:
 Prospective cohort observational survey
 Duration: April 2013 to March 2014
 Setting: 13 CD centers from 12 Mediterranean
countries:Albania, Algeria, Croatia, Egypt,
Greece, Italy,Malta, Montenegro, Morocco,
Slovenia, Tunisia, Turkey
 Sample: unselected new cases referred for
suspected CD
Patients
 Data of patients were collected and
put into a web-based database of
Medicel Mediterranean Network
(www.medicel.unina.it/)
Which data were collected?
•
•
•
•
•
•
•
•
Pregnancy duration and outcome
Birth weight
Breastfeeding
Age of gluten introduction into the diet
Number of previous admissions to hospital
Clinical symptoms (main, second and third)
Familiarity
Associated diseases
Which data were collected?
•
•
•
•
•
•
Antitransglutaminase antibodies IgA (t-TG) as
Nx upper limit of normal (ULN)
Endomysial antibodies (EMA)
HLA-DQ2/DQ8
Histology (Marsh-Oberhuber classification)
DNA, sera and tissue frozen samples were
collected for further studies
Follow-up to confirm uncertain case
Statistics
 Crosstabs and stepwise statistics
were generated by SPSS and t-Test,
Odds Ratio (OR) and Positive
Predictive Value (PPV) were
estimated for variables.
Results of the study
PATIENTS ENROLLED INTO THE STUDY
Country Total
CD
NCD
Albania
Algeria
Croatia
31
31
9
28
21
8
3
10
1
Egypt
Greece
Italy-Me
Italy-Na
20
35
270
339
15
29
100
196
5
6
170
143
Malta
10
Monten. 8
Morocco 85
9
8
49
1
0
36
Slovenia 61
Tunisia
41
Turkey
104
31
20
23
30
21
81
Results
 1044 patients enrolled
 537 CD
 507 NCD
 491 NCD
 16 potential CD (15 EMA positive)
Results
 Duration of gestation and
breastfeeding were significantly
longer in CD
 NCD had a heavier birth weight
 No difference was found for age of
gluten introduction and number of
previous hospital admissions
Results
Symptoms
PPV
(95% CI)
OR
(95% CI)
Anorexia/lack of
appetite
73% (62-81)
2.74 (1.64-4.60)
Abdominal
distention
68% (60-76)
2.30 (1.55-3.42)
Anemia
63% (53-71)
1.71 (1.11-2.63)
Vomit
54% (48-62)
1.50 (1.07-2.10)
Abdominal pain
63% (53-71)
0.66 (0.50-0.87)
Contipation
37% (30-44)
0.51 (0.35-0.74)
Results
 PPV of tTGA > 10 x ULN: 0.951
(95% CI:0.922-0.969); LR+ = 21.1
(95%CI: 13.0-34.37).
 66.9% tTGA > 10 x ULN might avoid
intestinal biopsy (ESPGHAN GL)
 4.6% with tTGA > 10 x ULN (EMA
positive) had a potential CD.
Decision threshold
Pre-test
probability of a
symptom
e.g. anorexia
73%(62-81)
PPV of tTGA > 10
x ULN: 0.951
(95% CI:0.9220.969); LR+ =
21.1 (95%CI:
13.0-34.37).
Conclusion
 This large prospective study provides for the
first time PPV, that is the pre-test probability,
of symptoms.
 Based on tTGA diagnostic accuracy obtained in
own laboratory, single centers can make
clinical decision on if and when intestinal
biopsy may be omitted, according to the new
ESPGHAN guidelines.
 There is a chance to meet potential CD even in
the presence of a tTGA > 10 x ULN value.
Conclusion
 The results of the study raises many
questions, the most important:
• Is it ever justifiable to place a child
with gastrointestinal symptoms on a
gluten-free diet without clearly
making a diagnosis of CD?
Conclusion
 The results of the study raises many questions, the most
important:
•
•
•
Do we know the natural history of potential
CD?
We have to determine which factors most
influence the onset of CD in Mediterranean
basin
Are repeat screenings necessary in NCD group,
because CD can present at any age, and if so,
at what intervals?
Conclusion
 The results of the study raises many questions, the most
important:
•
In further evaluation we have to clearly
identify:
- which symptoms can be used to identify
children for screening and diagnostics?
- which laboratory data predict the diagnosis of
CD?