The first celiac serology, AGA IgA, was developed in the early 1980s and revolutionized the diagnostic process of celiac disease.8
Prior to serologic studies, there was no screening test for celiac disease other than clinical suspicion, which was confirmed by small bowel biopsy. Shortly after the development of AGA, other serologic tests were introduced including tTG, antideaminated gliadin peptide antibodies, and antiendomysial antibodies.9
Although we recognize the significance of antiendomysial and antideaminated gliadin peptide antibodies, they are not routinely performed at our institution and were not included in the study.
Screening for celiac disease is recommended by the Amercian Gastroenterologic Association only for symptomatic patients. Although the prevalence of celiac disease has been estimated to be 1% in the general population, there is insufficient evidence to recommend celiac screening in the general population.10
Patients that are at high risk of celiac disease, such as those with first degree relatives of celiac disease, children or adolescents with short stature, patients with dermatitis herpetiformis, delayed puberty, type 1 diabetes mellitus, Down syndrome, persistent iron deficiency anemia, or osteoporosis should also be considered for serologic screening.11
There is a clear genetic predisposition for the development of celiac disease. Approximately 97% of patients with celiac disease share the major histocompatibility complex II class human leukocyte antigen DQ2 or DQ8 haplotype. Testing for these antigens may be considered in patients with equivocal small bowel histological findings.12
Ordering of human leukocyte antigens is not routinely performed and the results of this testing were not evaluated in this study.
The most significant finding in our study was the identification of five patients with biopsy confirmed celiac disease that had negative tTG but positive AGAs. Had these patients been screened using the American Gastroenterology Association's recommendation for tTG alone, they would have tested negative and would not have been referred for small bowel biopsy. However, positive AGA IgA and IgG antibodies with either negative tTG or untested tTG led to 61 negative small bowel biopsies and, therefore, screening with AGA will increase the number of small bowel biopsies performed. It has been reported that AGAs have a higher clinical significance in the pediatric population. Several studies have identified pediatric patients with celiac disease who were found to have positive AGA and negative tTG or antiendomysial antibodies, suggesting AGA may still be appropriate when screening this population.13,14
Of the five patients identified in our study, only one was under the age of 18 at the time of diagnosis. False negative tTG IgA testing has also been reported due to selective IgA deficiency. 1.7% of patients with celiac disease also has selective IgA deficiency and thus will have negative IgA screening antibodies.15
Of the five patients our study identified, 3 of them had normal IgA levels and thus their false negative tTG could not be attributed to a selective IgA deficiency; the other two patients were not tested for IgA deficiency.
Patients that had positive serologic testing that did not undergo small bowel biopsy were investigated to better determine the reason for not undergoing small bowel biopsy. We found that only 87 of 232 (38%) patients with positive serologic testing went on to have biopsy. While this percentage seems quite low, similar rate of biopsy have been described at other institutions; one study reported that only 39% of patients that screened positive in serologic testing had a small bowel biopsy.16
We further analyzed the 145 patients that did not undergo biopsy and determined that a little more than half (50.3%) were not referred to the gastroenterology service. This may represent primary care physicians treating empirically with a trial of gluten-free diet, patient refusal of referral, or failure to follow up on laboratory results. We also found about a third of patients (33.8%) that were seen by gastroenterology did not receive small bowel biopsy. This was due to a variety of reasons including patient refusal of biopsy, current treatment with gluten free diet (which may result in a false negative biopsy), or the sentiment among gastroenterologists that AGAs generate high number of false positive tests; the deferral of biopsy was often recommended in the pediatric population.
We acknowledge the limitations of our study. The retrospective nature of the study limits the information to that contained in patient medical records and does not reflect the number of patients placed on gluten-free diet without small bowel biopsy. Additionally, we are unable to determine whether serologic testing was performed on a gluten-free diet and thus impacted the results of these screening tests. It is also unclear how many primary care physicians and gastroenterologists discussed the benefits of small bowel biopsy but patients refused the procedure and opted for a trial of dietary modification.