BIOPSY IN COELIAC DISEASE DIAGNOSIS: WHEN IS IT NECESSARY?

Duodenal biopsy is considered the Gold Standard for Coeliac Disease (CD) diagnosis at the moment. As the ESPGHAN guidelines published in 2012 state, CD can be diagnosed without biopsy in patients with typical symptoms and tTG-IgA levels which are ten times over the reference values. The aim of this study is to establish the association between tTG-IgA levels and anatomic lesions, by redefining the biopsy role in the CD diagnosis. The research focuses on Positive Predictive Value of serological tests, also in patients without typical clinical presentation. In this retrospective study, clinical and laboratory data have been analysed in 95 pediatric patients, assessed at the Pediatric Unit in Siena from 2005 to 2011. Eightysix patients have been included, with typical symptoms, monosymptomatic or asymptomatic, all with tTG-IgA >9U/ml and a biopsy report. Patients have been divided into 5 groups (A-E) according to their tTG-IgA levels: A, >100 UI/ml; B, 80- 99 UI/ml; C, 60-79 UI/ml; D, 40-69 UI/ml; E, 9-39 UI/ml. A Positive Predictive Value (PPV) for each group has been evaluated. 78,9 of the patients included reported a positive biopsy. Biopsy was positive in 100 of the patients in groups A, B, C, with a Positive Predictive Value of 100%. PPV was 75% in group D and 78,7% in group E. In group A patients were divided in 3 subgroups: typical symptoms (67.4%), monosymptomatic (16.2%) and asymptomatic (16.2%). All of them (100%) had a positive biopsy report. PPV was 100% in typical symptoms subgroup as well as in monosymptomatic and asymptomatic patients. This study confirms the ESPGHAN Recommendation, concluding that in patients with typical symptoms and tTg-IgA ten times over the reference values the biopsy could be avoided. These results open the question whether the biopsy can be also avoided in patients with tTG-IgA ten times over the reference values even if monosymptomatic or asymptomatic. This approach could be followed by 50% of biopsies avoided in our case series.