Diagnostic Testing Algorithms for Celiac Disease


Originally posted: November 7, 2022


Melissa Snyder, Ph.D.

Associate Professor of Laboratory Medicine and Pathology
Division chair, Clinical Biochemistry
Mayo Clinic, Rochester, Minnesota



Hello, everyone. My name is Melissa Snyder, and I am the co-director of the Antibody Immunology Laboratory at Mayo Clinic in Rochester, Minnesota. I’m so glad you are able to join me for a brief discussion about celiac disease and the role of diagnostic testing algorithms.


Before beginning the presentation, I will disclose that I have served as a member of the Strategic Advisory Committee for Werfen, which is an in vitro diagnostic company that manufactures and sells kits for celiac serology testing.

Introduction to celiac disease

The clinical symptoms of celiac disease result from damage to the small intestinal villi, caused by an inflammatory response by the patient’s immune system. In the figure to the left, you see a biopsy of a normal small intestine, with intact villae. In the middle and right figures, you see the partial and total villous atrophy that can occur in celiac disease. For celiac disease to develop, an individual must have both the genetic susceptibility and the proper environmental exposure. The environmental exposure is ingestion of gluten, a cereal grain protein from wheat, barley, and rye. The genetic component of celiac disease are specific alleles of the human leukocyte antigen complex, namely HLA-DQ2 and HLA-DQ8. 

Clinical manifestations of celiac disease

The clinical symptoms associated with celiac disease vary widely. In the context of the gastrointestinal symptoms, patients may present with diarrhea, weight loss, steatorrhea, or abdominal pain, just to name a few. Because of the villous atrophy in the small intestine, patients with celiac disease may not be able to absorb nutrients from their food. As a result, patients may show symptoms of malabsorption, including iron-deficient anemia, various vitamin deficiencies, hypoproteinemia, or hypocalcemia. Patients with celiac disease may show manifestations that appear to have little to do with the GI system. Reports in the literature have demonstrated associations between celiac disease and ataxia, infertility, arthralgia, dermatitis herpetiformis, and hyposplenism. Lastly, celiac disease has been shown to occur more frequently in patients with other autoimmune diseases such as type I diabetes, autoimmune thyroid disease, and autoimmune liver disease, and in patients with specific chromosomal abnormalities. The point to stress here is that there are a wide variety of clinical symptoms or other scenarios in which testing for celiac disease may be warranted.

Diagnosis of celiac disease

The diagnosis of celiac disease can be based on a review of clinical symptoms and co-existing conditions, an intestinal biopsy with evidence of villous atrophy, serology testing for endomysial, tissue transglutaminase, and deamidated gliadin antibodies, and evaluation for genetic risk factors including family history and presence of HLA-DQ2/DQ8 alleles. Treatment for celiac disease is implementation of a gluten-free diet. The goal of this treatment is to remove the environmental exposure of dietary gluten. With a successful gluten-free diet, the patient should begin to see resolution of their clinical symptoms, which is often accompanied by reconstitution of the intestinal villae and conversion to a negative antibody serology.

Celiac-specific antibodies

As we saw on the previous slide, there are several classes of antibodies which are useful for the diagnosis of celiac disease. Endomysial antibodies (EMAs) are so named because they were originally identified as binding to the endomysium, which is the connective tissue that surrounds smooth muscle fibers. This testing is still performed by indirect immunofluorescence using a smooth muscle substrate. Ultimately, it was demonstrated that EMAs were binding specifically to tissue transglutaminase. Anti-TTG antibodies is the nomenclature for antibodies detected by binding specifically to tissue transglutaminase, which is an enzyme responsible for deamidating glutamine residues to glutamic acid. Patients with celiac disease may also develop antibodies against gliadin, which are peptides produced from digestion of gluten. The first immunoassays developed tested for antibodies against unmodified gliadin. However, these assays were inferior to the TTG antibody and EMA assays and are no longer recommended. Currently, the preferred gliadin antibody assays use deamidated gliadin; deamidated gliadin is formed in vivo through the activity of tissue transglutaminase (TTG). TTG and deamidated gliadin antibodies are detected using antigen-specific immunoassays, including plate-based enzyme immunoassays, bead-based multiplex, and chemiluminescent immunoassays. Lastly, it is important to note that testing for each of the antibodies listed on this slide can involve assessing for either IgA or IgG isotypes. 

Caveats of serology testing for celiac disease 

Although serologic testing is important in establishing a diagnosis of celiac disease, we must appreciate its limitations. One issue is selective IgA deficiency. Selective IgA deficiency is defined as the absence of detectable IgA in the presence of normal IgG and IgM production. Although relatively rare, it is more common in patients with celiac disease compared with the general population. In celiac diagnostic testing, the IgA isotypes are more sensitive and specific compared to the IgG isotypes. It is for this reason that the IgA isotype antibodies are preferred. However, for patients with selective IgA deficiency, testing for the IgG isotype antibodies is necessary. The other issue that can impact the utility of the serology testing is the effect of a gluten-free diet. In a patient with celiac disease, removal of gluten from the diet leads to “down regulation” of the inflammatory immune response, ultimately leading to reduced autoantibody production. This is useful when monitoring patients with celiac disease, as decreasing reactivity of the celiac-specific autoantibodies is consistent with a favorable response to the gluten-free diet. However, if a patient is already following a gluten-free diet before the diagnosis of celiac disease has been established, serology testing becomes less useful as there is a risk of a false negative result.

Genetic testing for celiac disease

Another component sometimes useful in the evaluation of a patient with suspected celiac disease is assessing for the presence of the genetic risk factors HLA-DQ2 and HLA-DQ8. These HLA alleles play a role in the pathogenesis of celiac disease, by binding to and presenting deamidated gliadin peptides to T-cells and initiating T-cell activation. Given this mechanistic role, it probably isn’t too surprising to learn that HLA-DQ2 is present in 90%-95% of patients with celiac disease, while HLA-DQ8 is found in the remaining 5%-10%. Because HLA-DQ2 and HLA-DQ8 are detected in virtually all patients with celiac disease, it might appear that genetic testing would be the preferred diagnostic test for this disorder. Unfortunately this is not the case, because 30%-40% of the general population of the United States is positive for HLA-DQ2 or HLA-DQ8. So what does this mean for the utility of HLA typing for celiac disease? The power of the HLA testing lies in a negative result. If a patient is negative for both HLA-DQ2 and HLA-DQ8, we can exclude celiac disease as a diagnosis. In contrast, if the patient is positive for either HLA-DQ2 or HLA-DQ8, we can only say that the patient has the genetic susceptibility for celiac disease, and he or she may or may not develop the disease in his/her lifetime.

Test performance and utility

So, let’s summarize the test performance and clinical utility of the various laboratory tests used for the diagnosis of celiac disease. The IgA isotypes for TTG and deamidated gliadin consistently have shown the best combination of sensitivity and specificity. EMA IgA generally demonstrates excellent specificity. However, because EMA is performed by indirect immunofluorescence, this testing can have some analytical challenges for the laboratory. If we consider the IgG isotypes for TTG and deamidated gliadin, we find that they are probably most appropriate for patients with a selective IgA deficiency. And lastly, for HLA-DQ2 and HLA-DQ8, we find these to be most useful as a “rule-out test” to exclude celiac disease as a diagnosis, as well as being the only laboratory test not affected by removal of gluten from a patient’s diet.

Laboratory testing algorithms 

Given the variety of tests that are available for the diagnosis of celiac disease, choosing the tests that are most appropriate for a given patient — not to mention interpreting the results — can be a challenge. The clinical labs at Mayo, working closely with our GI colleagues, have established several algorithms to aid in the diagnosis of celiac disease. 

The first cascade is the Celiac Disease Serology Cascade, or CDSP. This algorithm is applicable to most patients and incorporates a reflex approach to serologic testing. The second cascade is the Celiac Disease Comprehensive Cascade and has the test ID CDCOM. This algorithm includes both serologic and genetic testing. Then we have the Celiac Disease Comprehensive Cascade for Patients on a Gluten-Free Diet, or CDGF. This cascade only performs serology in the context of a positive genetic test. Now, let’s review each algorithm, beginning with the Serologic Cascade.

Serology Cascade

The Serologic Cascade begins with total IgA quantitation. All further testing reflexes automatically within the lab, based on the total IgA result. The total IgA result is classified as normal, or within the age-adjusted reference range, as low, being still detectable but below the reference range, or as deficient, or undetectable by our nephelometric assay. All samples with a normal total IgA result automatically reflex to a TTG-IgA antibody. For all samples testing positive or negative, no further testing is required. The final report includes the total IgA and TTG-IgA results, along with an interpretive comment. However, if the TTG-IgA result falls into the equivocal range, then EMA and deamidated gliadin-IgA testing is performed. These results, along with the total IgA and TTG-IgA results, are in the final report. On the other side of the cascade, for those individuals who have no detectable IgA, TTG and deamidated gliadin testing are performed, but only the IgG isotypes. These results are released as part of the final report, along with the total IgA quantitation. Finally, for those individuals with low but detectable IgA, testing for TTG and deamidated gliadin, both IgA and IgG isotypes, is performed. This cascade is designed to include all testing necessary to identify patients who may have celiac disease and in whom a biopsy would be suggested. It is not applicable to patients who have been following a gluten-free diet, due to the possibility of a false negative serology test result.

Comprehensive Cascade

The Comprehensive Cascade includes HLA typing and a reflex of serology tests. The Comprehensive Cascade begins with both total IgA and HLA-DQ typing. All further testing reflexes automatically based on the total IgA result similar to the Serology Cascade and independent of the HLA result. The IgA results are classified as normal, low, or deficient. For normal total IgA, a TTG-IgA is performed. For positive and negative results, no further testing is required. If the TTG-IgA is weakly positive, EMA and deamidated gliadin-IgA are performed, the results of which are included in the final report. For individuals with selective IgA deficiency, testing for the IgG isotype for TTG and deamidated gliadin antibodies is performed, followed by the release of the final report. For a low total IgA result, both isotypes for TTG and deamidated gliadin are performed.

Comprehensive Cascade for Patients on a Gluten-Free Diet

For a patient who has instituted a gluten-free diet in whom the diagnosis of celiac disease has not been confirmed, the Comprehensive Cascade for Patients on a Gluten-Free Diet may be appropriate. In this algorithm, only the HLA-DQ typing is performed initially. For those individuals who have neither the DQ2 nor DQ8 alleles, celiac disease is virtually excluded as a diagnosis. At this point, testing for celiac disease should stop, and other potential diagnoses related to the patient’s clinical presentation should be evaluated. On the other hand, a positive result for DQ2 or DQ8 means only that celiac disease is a possible diagnosis. At this point, testing for all the serologic tests is performed. Depending upon how long the patient has been following the gluten-free diet, and how strict the diet is, some of these serologic tests may provide a positive result. In that case, the results of all laboratory testing are consistent with celiac disease and a biopsy would be indicated. If all results are negative, celiac disease has not been completely ruled out, since this could simply be a reflection of a successful gluten-free diet. At this point, the clinician must determine how likely the diagnosis of celiac disease is and if further evaluation, such as a gluten challenge, should be considered.


To summarize, we know that laboratory testing is important in the evaluation of patients with suspected celiac disease. However, the number of available tests, serologic and genetic, can be confusing. Mayo Clinic Laboratories offers three laboratory reflex algorithms for celiac disease, which target the most appropriate testing for the individual patient. Each cascade has a specific utility. The Celiac Disease Serology Cascade is the most widely applicable algorithm and uses targeted serology testing for identification of patients in whom celiac disease is a possible diagnosis. The Celiac Disease Comprehensive Cascade includes serologic and genetic testing and is designed for the small subset of patients in whom HLA-DQ2 and HLA-DQ8 typing is desired. The Celiac Disease Comprehensive Cascade for Patients on a Gluten-Free Diet relies on genetic testing to exclude celiac disease in patients who have initiated a gluten-free diet prior to a confirmed diagnosis.

One final point to mention is that all celiac testing offered by Mayo Clinic Laboratories is also available as individually orderable tests. The testing algorithms are most useful for diagnostic evaluation, while the individual tests are most appropriate for monitoring patients and their response to a gluten-free diet.

Thank You

I hope this presentation has provided you with useful information regarding laboratory testing for celiac disease and has helped to clarify the many options available for diagnostic testing. Thank you for your participation.


Contact us: mcleducation@mayo.edu

Image credit: Shutterstock/SciePro

MCL Education

This post was developed by our Education and Technical Publications Team.