Poor Sensitivity of Fecal Gluten Immunogenic Peptides and Serum Antibodies to Detect Duodenal Mucosal Damage in Celiac Disease Monitoring

Abstract

A lifelong gluten-free diet (GFD) is the only current treatment for celiac disease (CD), but strict compliance is complicated. Duodenal biopsies are the “gold standard” method for diagnosing CD, but they are not generally recommended for disease monitoring. We evaluated the sensitivity and specificity of fecal gluten immunogenic peptides (GIPs) to detect duodenal lesions in CD patients on a GFD and compared them with serum anti-tissue transglutaminase (tTG) IgA antibodies. A prospective study was conducted at two tertiary centers in Spain on a consecutive series of adolescents and adults with CD who maintained a long-lasting GFD. Adherence to a GFD and health-related quality of life were scored with validated questionnaires. Mucosal damage graded according to the Marsh–Oberhüber classification (Marsh 1/2/3) was used as the reference standard. Of the 97 patients included, 27 presented duodenal mucosal damage and 70 had normal biopsies (Marsh 0). The sensitivity (33%) and specificity (81%) of GIPs were similar to those provided by the two assays used to measure anti-tTG antibodies. Scores in questionnaires showed no association with GIP, but an association between GIPs and patients’ self-reported gluten consumption was found (p = 0.003). GIP displayed low sensitivity but acceptable specificity for the detection of mucosal damage in CD.

Determination of gluten immunogenic peptides for the management of the treatment adherence of celiac disease: A systematic review

“…Other authors also found a significant association between patient self-reported gluten consumption and GIP detection. GIP was detected in 52.9% of patients who suspected gluten intake in the last month in comparison to 16.3% for those who were not aware of any gluten exposure[ 45 ].…”

“…Other authors reported a weaker association between evident gluten exposure, as measured by GIP detection in urine and stool, and persistent villous atrophy[ 29 , 30 , 45 ]. In the DOGGIE BAG study referred to above, at least one GIP-positive sample was found in 7 of 11 patients with normal serology, and 2/3 of them had persistent villous atrophy (Marsh IIIa) after 24 mo of follow-up.…”

“…Further, 2/3 of patients with GIP-negative results for all samples were Marsh 0-I, showing significant mucosal recovery[ 29 , 30 ]. A possible explanation for the discordance described before is that occasional or low gluten exposure may be sufficient for GIP detection but not enough to induce mucosal damage in some patients with CeD[ 23 , 29 , 30 , 45 ]. A delay of intestinal mucosal recovery[ 29 , 30 , 45 ] and improved compliance of participants during the study was also considered[ 29 , 30 ].…”

“…A possible explanation for the discordance described before is that occasional or low gluten exposure may be sufficient for GIP detection but not enough to induce mucosal damage in some patients with CeD[ 23 , 29 , 30 , 45 ]. A delay of intestinal mucosal recovery[ 29 , 30 , 45 ] and improved compliance of participants during the study was also considered[ 29 , 30 ]. Besides, Laserna-Mendieta et al [ 45 ] reported a low sensitivity but acceptable specificity of GIP for the detection of mucosal damage, however for the calculation they included in the same group Marsh 1 patients ( n = 21), who do not present mucosal damage but lymphocytic infiltration, and Marsh 2 and 3 patients ( n = 6).…”

“…A delay of intestinal mucosal recovery[ 29 , 30 , 45 ] and improved compliance of participants during the study was also considered[ 29 , 30 ]. Besides, Laserna-Mendieta et al [ 45 ] reported a low sensitivity but acceptable specificity of GIP for the detection of mucosal damage, however for the calculation they included in the same group Marsh 1 patients ( n = 21), who do not present mucosal damage but lymphocytic infiltration, and Marsh 2 and 3 patients ( n = 6). Moreover, in this study, weak GIP-positive results were discarded for analysis and duodenal biopsies were performed in two hospitals with different pathologists, entailing a risk of interobserver variability[ 63 , 64 ].…”