Diagnostic Challenges of Celiac Disease
Click CC to turn on closed captioning.
Published: June 2010Print Record of Viewing
This Clinical Insight program covers several cases of celiac disease and the appropriate diagnostic testing for each situation.
Presenter: Dr. Joseph Murray
- Practicing gastroenterologist
- Professor of Medicine and Immunology at Mayo Clinic
- Director of the Mayo's Celiac Disease Program.
Questions and Feedback
Contact us: .
Welcome to Mayo Medical Laboratories' Hot Topics. These presentations provide short discussions of current topics and may be helpful to you in your practice.
Our speaker for this Clinical Insight program is Dr. Joseph Murray, a practicing gastroenterologist and Professor of Medicine and Immunology at Mayo Clinic, as well as Director of Mayo Clinic’s Celiac Disease Program. Dr. Murray will discuss several cases of celiac disease and the appropriate diagnostic testing for each situation. Thank you, Dr. Murray.
Case 1: Negative Serology for Celiac Disease
The first case. A 45-year-old female patient with a 6-month history of diarrhea and bloating, is found to be anemic. She has a family history of a brother who has dermatitis herpetiformis. A tissue transglutaminase IgA test is performed and is negative. An endoscopy is performed and an endoscopic biopsy shows partial villous atrophy.
The next logical step is one of the following: 1. Treat her for bacterial overgrowth; 2. Give her iron and loperamide, an antidiarrheal agent; 3. Order an endomysial IgA, indirect immunofluorescence assay (IFA) test; or 4. Measure total IgA levels.
The correct answer in this case would be to measure total IgA levels.
Limitations of Serology
There are limitations of serology for celiac disease. A significant limitation is that of IgA deficiency. Three to 5% of patients with celiac disease are IgA deficient, and up to 10% of patients with IgA deficiency have celiac disease. Another limitation of serology is that it is less sensitive for partial villous atrophy. And, very important, is the effect of a prior gluten-free diet. Patients who have reduced or eliminated gluten from their diet may cause their serologic test to become negative, and this may occur very quickly. So it is very important when considering testing for celiac disease to ensure that the patient has been on a normal, gluten-containing diet before testing.
A solution in this regard would be to start the serologic strategy for the detection of celiac disease with a measurement of total IgA. This strategy is incorporated into both the Celiac Disease Comprehensive Cascade, as well as the Celiac Disease Serology Cascade, that use a total IgA level to drive the most logical subsequent testing to maximize the accuracy of the serologic approach to the detection of celiac disease.
Case 2: Symptomatic Malabsorption
The second case. This is a case of a 58-year-old woman with diarrhea and a 6-month history of steatorrhea and bloating. She is anemic, she has lost over 25 pounds, and her endoscopic biopsy shows partial villous atrophy.
The next logical step would be: 1. To make a definite diagnosis of celiac disease at this point and start a gluten-free diet; 2. Do HLA testing; 3. Order gliadin antibody levels; or 4. Order a choice of a Comprehensive Celiac Disease Cascade or a Serology Cascade.
In this case, either the Comprehensive Cascade or the Serology Cascade provide logical approaches to the further evaluation of this patient.
This illustrates the circumstance where a patient has had a presumptive diagnosis of celiac disease made by biopsy first, without prior serologic testing. In this case, the patient had a duodenal biopsy with villous atrophy. The next most important stage would be to undertake testing for serology to provide important supportive evidence for the diagnosis of celiac disease. Again, we deal with the issue of, is this patient IgA deficient? And, the primary test in most circumstances is the tissue transglutaminase IgA level. It is important to recognize that a proportion of patients who have an abnormal biopsy do not have celiac disease, but rather have some other cause. Combining serologic testing, often with the HLA celiac-specific susceptibility testing, can be a very powerful tool for discriminating between celiac disease and those who do not have celiac disease in this circumstance.
Comparison of Serological Tests
When we compare the different available serologic tests in terms of their sensitivity and specificity, as well as technical difficulty and cost, there are significant choices to be made. The top row is that of the human tissue transglutaminase antibodies. Generally speaking, the sensitivity is high and the specificity is reasonably high. The technical difficulty of this test is low. The endomysial antibody is an indirect immunofluorescence assay, which is a variable sensitivity that is somewhat interpreter and laboratory dependent. Its specificity is quite high, approaching 99% to 100%. Its technical difficulty is greater.
Gliadin IgA and gliadin IgG tests have been around for over 30 years. However, they are limited by quite variable sensitivity and often quite low specificity and, while they are simple to do, they are really not an appropriate test for screening for celiac disease. Their inclusion in panels for testing for celiac disease do not add much to sensitivity, but greatly detract from the specificity and, hence, are no longer recommended in most guidelines that address the detection of celiac disease.
A newer development in serologic testing for celiac disease involves the use of deamidated gliadin peptide antibodies. These tests are a substantial improvement over the earlier generation gliadin antibody tests and, at the Mayo Clinic, have replaced the earlier gliadin antibody tests as a choice for testing for celiac disease. Their sensitivity is perhaps a little less than that of tissue transglutaminase and their specificity is comparable. The combination of both tests probably adds to the sensitivity without substantially reducing the specificity.
So in summary, human tissue transglutaminase ELISA testing, especially for the IgA isotype, is probably the single best test for accuracy in terms of celiac disease. Endomysial antibody testing and deamidated gliadin antibody testing also play a role in the testing for celiac disease; however, gliadin IgA and gliadin IgG antibodies have no role at this time in the testing for celiac disease.
Histology has long been considered the gold standard for the diagnosis of celiac disease and is a required part of the confirmation step for celiac disease diagnosis. However, there is a significant spectrum of damage that occurs in gluten-sensitive enteropathy.
In this illustration, we illustrate what has been termed the Marsh Classification of gluten-sensitive enteropathy lesions. While lesion number 3 is the classic lesion of celiac disease (where there is at least partial villous atrophy, crypt hyperplasia, and infiltration of the surface epithelium with lymphocytes characteristic for celiac disease) there is, however, a greater spectrum of disease.
Patients may have milder disease so called Marsh 1 or Marsh 2, where there are subtle changes in inflammation or slight degree of crypt hyperplasia but intact villi. These present greater challenges for diagnosis not only because they may not be recognized as part of the spectrum of gluten-sensitive enteropathy, but also have a broader differential diagnosis and frequently have other disorders that cause these changes.
Luckily, class 4 on the right, the hypoplastic lesion, is quite rare and often occurs in the context of very complicated celiac disease. There are a few patients who have been described who have Marsh 0 in the context of serologic positivity for celiac disease. This circumstance has been termed latent or potential celiac disease and patients who have positive serology, especially by tissue transglutaminase antibodies or endomysial antibodies, should be followed over time as they may be at increased risk of developing celiac disease in the future.
Not all that flattens is celiac disease. This is a case of a 65-year-old male with a history of diarrhea and weight loss. The tissue transglutaminase antibody and endomysial antibodies were performed and were negative. The patient had no response to a gluten-free diet. His intestinal biopsies showed quite substantial change with inflammation, partial villous atrophy, marked crypt hyperplasia and, this photo micrograph illustrates, the dense lymphoplasmacellular infiltrate. And, to the untrained eye, one would say this looks very much like celiac disease. However, the patient’s lack of response and negative serology at the time of initial diagnosis should alert the clinician that this may not be celiac disease and there may be an alternative explanation.
In this case, the patient underwent further serologic testing and this illustrates the immunofluorescence pattern of antigoblet cell antibodies, that is suggestive of the condition of autoimmune enteropathy. This patient, had these antigoblet cell antibodies detected as part of the antienterocyte antibody test. Additionally, HLA genetic susceptibility typing for celiac disease identified the patient carried the genes that encoded DQ6 and not the genes that encoded the celiac susceptibility HLA types of now, or previously called, DQ2 or DQ8. This patient responded to steroids and Azathioprine for his disease that was not celiac disease.
There are many reasons why biopsies may look like celiac disease, but are not; the condition of graft-versus-host disease, recent chemotherapy, Sjögren syndrome. Crohn disease is a rare cause of such pathologic change and often is quite obvious to the endoscopist, or by other means of patient investigation. Bacterial overgrowth, while is thought to be relatively common, is unlikely to cause injury of the degree seen in celiac disease. In young children, other food protein intolerances can cause histologic changes in the small intestine, but this is rare in adults.
We again must be concerned, especially when biopsies have been poorly oriented, of the concept of “flattened biopsies” meaning the patient does not have a flat epithelium, but rather the biopsies have been mal-oriented or flattened so the villi present are not readily seen. Certain drugs can cause injury to the intestine, nonsteroidal anti-inflammatories, for one example. There has also been the condition of self-limited enteritis described. Tropical sprue should be suggested by a travel history to parts of the world where tropical sprue is endemic. Other conditions such as combined variable immunoglobulin deficiency, autoimmune enteropathy, as we have just illustrated, and even the rarer condition of nongranulomatous enterocolitis, can also mimic the pathologic changes of celiac disease.
Minimal Disease: Uncertain Histology
The next case. A 65-year-old female patient with dyspepsia and a 6-month history of epigastric pain and bloating. Is not anemic. However, a duodenal biopsy performed shows increased intraepithelial lymphocytes without villous atrophy.
The next logical step could be: 1. Put her on a gluten-free diet, presuming that she has celiac disease; 2. Give her a drug called budesonide, a powerful, topically active steroid; 3. Order a Celiac Disease Comprehensive Cascade; or 4. Treat her for Crohn disease.
In this case, rather than assuming that this patient has celiac disease, or treating her with an anti-inflammatory or treating for another disease, that’s Crohn disease, the next logical test would be to undertake the Celiac Disease Comprehensive Cascade.
This condition also known as lymphocytic duodenosis is really a pathologic finding on duodenal biopsies. It is not a diagnosis in itself. It is characterized by raised intraepithelial lymphocytosis. About 20% of patients with this finding may actually have celiac disease, but most do not. It can also be seen in patients with known celiac disease who are not compliant with their diet.
Anti-inflammatories, autoimmune disease, inflammatory bowel disease, giardiasis and a host of other conditions can mimic this change.
Patient Presenting on Gluten-free Diet
What about the patient who presents on a gluten-free diet? A 25-year-old, male patient with chronic fatigue, bloating, and abdominal pain, reads an article about celiac disease. He starts on the gluten-free diet and feels better immediately. Six-months later he presents to your office for a diagnosis.
The next logical step could be one of these: 1. Tell him to continue the gluten-free diet; 2. Start a gluten challenge and biopsy him after 2 days; 3. Order a tissue transglutaminase IgA IFA test; or 4. Do HLA typing for celiac genes.
Well, we have several choices. In this case, the HLA typing may be a worthy choice. So let’s consider these. Should he stay on a gluten-free diet? Well, he is obviously not happy on a gluten-free diet. A gluten challenge is a very reasonable and standard option; however, biopsying after 2 days is not long enough to generate enough pathologic damage to be certain of the diagnosis and a much more prolonged challenge is usually necessary. A TTg-IgA test is typically not accurate in patients who have been on a gluten-free diet for as long as he has. However, genetic testing for celiac susceptibility genes does not change with introduction of a gluten-free diet.
What About Patients on Gluten-free Diet?
So, what about those patients on a gluten-free diet? They are often unhappy; this is not an easy diet to follow. Serology and biopsies can normalize. HLA typing might help. A challenge may not be acceptable to some patients who simply will not eat gluten. Other patients were too ill to risk a rechallenge with gluten. And, for patients with severe neurologic disease, elderly patients with severe nutritional impairment, are usually patients in which a gluten challenge is not recommended. And, of course, one must always remember that why argue with success if the diet itself is nutritionally adequate, if the patient is neither suitable nor willing to undergo a sufficient gluten challenge for accurate testing.
Gluten Challenge Testing
Gluten-challenge testing requires adequate gluten intake for long enough to develop gut lesions. Four slices of whole wheat bread daily for 4 weeks usually suffices. However, it will make patients ill and they can become quite symptomatic. There are some patients who are delayed responders and if a patient has had no symptoms develop by 4 weeks, I will usually perform serology. If it becomes positive at that point, we’ll proceed to biopsy. If it’s negative we will wait longer, for both serology and histology change to occur. Typically, I will not wait longer than 6 months, however.
Celiac Disease and HLA Risk
The theoretical application of HLA typing to celiac disease is as follows. The large green circle illustrates the genetic type of the general population. The yellow circle illustrates those individuals who carry the gene pairs that can encode for the HLA type that makes celiac disease possible. But only the small red dot illustrates those who actually have celiac disease. The strength of this approach is really in its negative predictive value in those who do not reside within the yellow circle. So, those in the green area who do not have these gene pairs are very unlikely to ever get celiac disease. Whereas those contained within the yellow circle, including the red dot, have the genetic susceptibility for celiac disease, but as one can, see most individuals who carry the genes will not get celiac disease.
Genetic Tests-Big Limitation
So that brings to our notice the big limitation of this type of genetic testing. First of all, having the HLA type does not equal the disease. Most people with the at-risk types will not have celiac disease. Two-thirds of family members of people with celiac disease will carry the at-risk types, but most of these don’t get the disease. Another category of individuals who are at risk for celiac disease are type 1 diabetics, also known as juvenile onset diabetes. And fully 50% of those individuals carry the same HLA type as risk for celiac disease, but only 6% of them will get celiac disease.