Celiac disease, one of the most common autoimmune diseases in the world today, often goes undiagnosed due to asymptomatic patients and many nonspecific symptoms. To diagnose a patient with celiac disease, a positive intestinal biopsy is typically necessary. But when physicians need to decide on whether or not to order a biopsy, they have many tools at their disposal.

Many serologic assays now exist that exploit the presence of particular antibodies in celiac patients after eating gluten. The most common serological tests look for immunoglobulin A (IgA) antibodies against tissue transglutaminase (tTG), endomysial tissue (EMA) and deamidated gliadin (dGDN). However, many patients with celiac disease also suffer from IgA deficiency, so immunoglobulin G (IgG) versions of these assays can also be used. The tTG-IgA and EMA-IgA tests have high pooled sensitivities of about 90% and high pooled specificities of around 98%. With these precise screening tests available, one would think that most intestinal biopsies ordered to confirm or rule out celiac disease would have a high rate of confirming the disease. However, a recent paper looked at just that and found otherwise.

The article, published by BMC Gastroenterology on November 9th of 2013, is titled “Limited utilization of serologic testing in patients undergoing duodenal biopsy for celiac disease.” The authors, Homer WIland, Walter Henricks, and Thomas Daly, three medical professionals from Cleveland Clinic’s pathology department, observed the pathology lab’s information system for 6 months, looking at all biopsy submissions in the database that contained the words “celiac”, “gluten”, or “sprue”.  Then they searched electronic medical records for any serology results for the patients found in the database.

In total, the authors found 1432 duodenal biopsies in which celiac disease was at least part of the reason for the biopsy. The results they found were astonishing. Only 455 of the 1432 (31.8%) patients had celiac disease serology results. Of those, 387 were negative and 68 were positive.

For those patients who did have positive serology,  51.5% of the biopsies were negative for celiac disease. Compared to those patients who had no serology done who had a negative rate of 84.5% or those who had negative serology who had a negative rate of 95.3% and the benefits of using serological tests prior to biopsy is very clear. The p-value for comparing the negative rate of celiac disease for both of the latter groups to the former was less than 0.001 in both cases, using Fisher’s exact test. This indicates that the difference in negative rate between the first group who had positive serology and the other two groups was very significant.

The authors wondered then, why was serology so rarely used before ordering a biopsy? To find out, they went back to the source of the data, the pathology database. After looking over the data, the authors realized that the vast majority (88%) of the 1432 duodenal biopsies were accompanied by at least one other biopsy location. The most common locations of these multi-site workups were the duodenum and the stomach (26%) or the duodenum, the stomach, and the esophagus (24%).

Even for patients with positive serology indicating celiac disease, multiple biopsy sites were common, albeit less so than for those with negative serology (2.5 vs 2.9 sites on average, respectively).

Wiland, Henricks, and Daly then looked at the diagnostic yield of locations other than the duodenum. They found that most of the patients (58%) who had positive serology also had “findings noted in at least one non-duodenal site.”

The authors draw several conclusions from this retrospective study. First, that serology is a very useful tool for lower-risk patients to optimize cost-effective treatment and diagnosis of celiac disease, which has been established in previous studies. But second, that using serology prior to duodenal biopsy has not been widely adopted.

Instead, most biopsies were not really targeted towards celiac disease, even though celiac disease was put as the reason the duodenal biopsy was done. The duodenal biopsies were just “part of a multi-site sampling strategy for a larger GI workup”. And that kind of strategy will always have a low diagnostic yield.

Even for these large multi-site biopsy strategies, serological results can still be very useful to guide physicians on which bodily locations to sample. Patients with positive serology for celiac disease were more likely to have other findings in non-duodenal sites.

This paper contributes to a growing body of work that suggests serology be used prior to duodenal biopsy for ruling out celiac disease but shows that this approach has not been very well adopted clinically.

If you think you or a family member may be suffering from celiac disease, you can try our Symptoms Checklist, found on our website. After completing it, you will be emailed a printable copy of the information that you can bring to your doctor to discuss your options. If you need a doctor familiar with celiac disease, you can use our Healthcare Practitioner Directory, also found on our website, which will provide you with a free listing of physicians, dietitians, mental health professionals, and allied health providers in all 50 states who treat patients with gluten-related disorders.

The article discussed in this paper can be found here or in the November 2013 issue of BMC Gastroenterology.