Celiac disease is an autoimmune disorder induced by gluten in genetically susceptible individuals. It can result in intraintestinal and extraintestinal manifestations of disease including diarrhea, weight loss, anemia, osteoporosis, or lymphoma. Diagnosis of celiac disease is made through initial serologic testing and then confirmed by histopathologic examination of duodenal biopsies. Generally celiac disease is a benign disorder with a good prognosis in those who adhere to a gluten-free diet. However, in refractory disease, complications may develop that warrant additional testing with more advanced radiologic and endoscopic methods. This article reviews the current strategy to diagnose celiac disease and the newer modalities to assess for associated complications.
Key points
- •
Patients who should be tested for celiac disease include those with classic gastrointestinal manifestations or those deemed at high risk based on genetic susceptibility.
- •
Diagnosis of celiac disease is usually initiated by serologic testing with anti-tTG, anti-DGP, or EMA and confirmed with duodenal biopsy.
- •
Generally, patient nonresponse to gluten-free diet is typically caused by either unintentional gluten exposure or by a secondary cause, such as inflammatory bowel disease or small intestinal bacterial overgrowth.
- •
Patients with refractory celiac disease may benefit from further radiologic or endoscopic evaluations including MRE, CT enterography/enteroclysis, capsule endoscopy, or device-assisted enteroscopy to evaluate for complications including ulcerative jejunoileitis or malignancy.
Introduction
Celiac disease (CD) is an autoimmune disorder induced by gluten in genetically susceptible individuals characterized by intraepithelial lymphocytosis, crypt hyperplasia, and villous atrophy of the small bowel. It is a chronic inflammatory state that heals on exclusion of gluten-containing foods from the diet. The prevalence of CD is about 1% of the general population worldwide. Gluten from wheat, barley, and rye are enriched in glutamines and prolines, which undergo partial digestion in the small bowel resulting in peptide derivatives that are deamidated by tissue transglutaminase, which renders them immunogenic to those with CD. Active CD can result in intestinal and extraintestinal manifestations of disease including diarrhea, weight loss, anemia, osteoporosis, arthritis, hepatitis, or malignancy. Some patients are also asymptomatic.
Diagnosis of CD is generally initiated through serologic testing with antitissue transglutaminase IgA antibodies (anti-tTG), gliadin-derived peptide antibodies IgA/IgG (anti-DGP), endomysial IgA antibodies (EMA), and/or antigliadin antibodies (AGA). Given the lower sensitivity and specificity of AGA tests for CD, the EMA, anti-tTG, and anti-DGP have largely replaced other serologic testing. Following positive serologic testing, diagnosis should generally be confirmed by histopathologic examination of duodenal biopsies.
Generally CD is a benign disorder with a good prognosis in those patients that can adhere to a gluten-free diet. However, in those with refractory disease, complications may develop, which warrant additional testing with more advanced radiologic and endoscopic methods including magnetic resonance enterography/enteroclysis (MRE), PET/computed tomography (CT), capsule endoscopy, and device-assisted enteroscopy.
Introduction
Celiac disease (CD) is an autoimmune disorder induced by gluten in genetically susceptible individuals characterized by intraepithelial lymphocytosis, crypt hyperplasia, and villous atrophy of the small bowel. It is a chronic inflammatory state that heals on exclusion of gluten-containing foods from the diet. The prevalence of CD is about 1% of the general population worldwide. Gluten from wheat, barley, and rye are enriched in glutamines and prolines, which undergo partial digestion in the small bowel resulting in peptide derivatives that are deamidated by tissue transglutaminase, which renders them immunogenic to those with CD. Active CD can result in intestinal and extraintestinal manifestations of disease including diarrhea, weight loss, anemia, osteoporosis, arthritis, hepatitis, or malignancy. Some patients are also asymptomatic.
Diagnosis of CD is generally initiated through serologic testing with antitissue transglutaminase IgA antibodies (anti-tTG), gliadin-derived peptide antibodies IgA/IgG (anti-DGP), endomysial IgA antibodies (EMA), and/or antigliadin antibodies (AGA). Given the lower sensitivity and specificity of AGA tests for CD, the EMA, anti-tTG, and anti-DGP have largely replaced other serologic testing. Following positive serologic testing, diagnosis should generally be confirmed by histopathologic examination of duodenal biopsies.
Generally CD is a benign disorder with a good prognosis in those patients that can adhere to a gluten-free diet. However, in those with refractory disease, complications may develop, which warrant additional testing with more advanced radiologic and endoscopic methods including magnetic resonance enterography/enteroclysis (MRE), PET/computed tomography (CT), capsule endoscopy, and device-assisted enteroscopy.
Pathogenesis
CD develops in genetically susceptible individuals who are exposed to gluten. The clinical presentation of CD can vary greatly including the age of onset, presenting symptoms, the level of antibody titers, and a range of histopathologic findings, which can likely be explained by the interaction between genetic predisposition and environmental exposure.
Genetic Predisposition
Sibling studies in CD have demonstrated a disease concordance of about 80% in monozygotic twins and less than 20% in dizygotic twins indicating a genetic link. The major genetic determinants in CD involve the HLA, which is estimated to contribute to about 36% of the hereditability between siblings.
HLA-DQ molecules are made up of two subunits, α and β, which are encoded by two different genes of the class II MHC molecule: HLA-DQA1 and HLA-DQB1, respectively. In CD, it has been found that 90% of patients carry the alleles DQA1*05 and DQB1*02 , which make up the HLA-DQ2 heterodimer. More specifically, they tend to have the HLA-DQ2.5 variant, which involves the DQA1*05:01 and DQB1*02:01 genes in cis configuration on the DR3 haplotype. This molecule has a high affinity for the peptides that are formed from incomplete digestion of gluten, which results in their presentation and resultant intestinal inflammation. HLA influence on CD susceptibility also demonstrates a dose effect. Homozygous HLA-DQ2 individuals, for example, may have an increased risk for CD and enteropathy-associated T-cell lymphoma (EATL).
Of the 10% who have not inherited the HLA-DQ2.5 molecule ( DQA1*05:01 and DQB1*02:01 alleles), most have inherited the DQA1*03 and DQB1*03:02 alleles of the HLA-DQ8 molecule. In addition, there are also non-HLA genetic factors that play a role in the development of disease. In Western countries about 40% of the general population possess one or both of the HLA-DQ2/HLA-DQ8 heterodimers, yet only 1% of individuals develop CD. This indicates that there must be other genetic and environmental factors that contribute to the development of disease. Through genome-wide association studies, several different non-HLA alleles associated with risk of CD have been discovered. Currently there are about 40 loci outside of HLA that have been determined through genome-wide association studies that have been found to either protect or predispose to CD, although they contribute little when compared with HLA.
Environmental Exposure and Trigger Factors
In addition to genetic predisposition, patients with CD need to be exposed to gluten to develop the disease. Gluten is the storage protein for the cereal grains of wheat, rye, and barley. There has been some thought that the timing of gluten introduction, the amount of gluten exposure, and breastfeeding patterns may influence the development of CD. Large amounts of gluten exposure without breastfeeding may increase the risk of future CD, although data are conflicting. There has also been some work looking to evaluate if other factors, such as gastrointestinal infection, surgery, or certain drugs, may be the trigger for development of CD.
Immunology
Gluten proteins are incompletely digested by the gastric, pancreatic, and intestinal brush border proteases. The remaining peptides pass through the epithelial barrier of the small bowel and enter the lamina propria through transcellular and paracellular mechanisms. This triggers the innate and adaptive immune response in patients with CD leading to intestinal inflammation.
Gluten peptides were initially thought to activate an innate immune response. However, nongluten, proteins such as the wheat amylase-trypsin inhibitor, have been demonstrated to activate macrophages, monocytes, and dendritic cells via toll-like receptor 4, which is the receptor for bacterial lipopolysaccharide. The innate response is manifested by increased expression of interleukin-15 by enterocytes, which leads to activation of intraepithelial lymphocytes (IEL) that express the natural killer T-cell receptors.
The adaptive response occurs within the lamina propria, where gluten-reactive CD4 + T cells recognize gluten peptides presented on HLA-DQ2.5/HLA-DQ8 molecules. This occurs because tissue transglutaminase is released by inflammatory and endothelial cells in response to mechanical irritation and inflammation. It cross-links with gluten proteins and deamidates gliadin, thereby altering the charge and conformation of gliadin peptides. These changes greatly increase the binding affinity of gliadin to HLA-DQ2 or HLA-DQ8 molecules, leading to T-cell stimulation. This activates the antigen presenting cells leading to auto antibody formation (tTG), proinflammatory cytokines including interferon-γ, and subsequent tissue injury, leading to crypt hyperplasia and villous blunting.
Clinical features
The clinical features of CD range from classical symptoms, to nonclassical and symptomatic, to asymptomatic. Classical symptoms generally include those resulting from malabsorption including diarrhea, steatorrhea, weight loss, and growth restriction in children. Nonclassical and symptomatic patients tend to have either some gastrointestinal symptoms, such as abdominal pain or constipation, or may have extraintestinal symptoms. Age of onset of disease is at any age when there is exposure to gluten, and the presentation can vary depending on age of presentation. Children who present with CD early tend to present with more severe disease manifested as growth problems and recurrent abdominal pain, and less commonly (about 10%) with diarrhea. Older teenagers and adults, however, often have subtle symptoms that may be misdiagnosed as irritable bowel syndrome.
Gastrointestinal Manifestations
The classic gastrointestinal manifestations of CD result from villous atrophy in the small intestine leading to malabsorption. This leads to the development of diarrhea, steatorrhea, weight loss, and growth failure in children. Malabsorption also results in many of the common complications secondary to nutrient loss including iron-deficiency anemia, neurologic disorders from vitamin B deficiencies, and osteopenia from vitamin D deficiency.
As serologic testing has improved, more adults are being diagnosed with less severe symptoms. Many adults develop irritable bowel syndrome–type symptoms including minor gastrointestinal complaints, such as abdominal pain, boating, constipation, or mild diarrhea. Often, however, these individuals are diagnosed because of nutritional deficiencies resulting in iron-deficiency anemia, osteoporosis, or extraintestinal manifestations as outlined next.
CD has also been linked to other gastrointestinal disorders that range from a mild increase in transaminases to liver failure, liver cancer, and pancreatic cancer.
Extraintestinal Manifestations
In addition to expected manifestations as a result of villous atrophy, there have been many other disorders linked to CD. These include iron-deficiency anemia, neuropsychiatric disease, lymphoma, arthritis, and metabolic bone disease.
Iron-deficiency anemia is commonly reported in patients with CD even in those without malaborptive or gastrointestinal symptoms. In asymptomatic patients with iron-deficiency anemia, the prevalence of CD was found to range from 2.3% to 5.0%, whereas in those individuals with gastrointestinal symptoms and iron-deficiency anemia the prevalence ranged from 10.3% to 15%.
Neurologic or psychiatric diseases have been described in patients with CD including headache, peripheral neuropathies, ataxia, dysthymia, depression, anxiety, and epilepsy.
CD has also been associated with the development of lymphoma. Although classically associated with the development of EATL, patients are at a greater risk of developing other types of lymphoma including intestinal and extraintestinal non-Hodgkin lymphoma. Studies have demonstrated that the standardized incidence ratio of non-Hodgkin lymphoma in patients with CD compared with the general population ranges between 2.7% and 6.3%.
CD has also been linked to several different autoimmune disorders including type 1 diabetes mellitus and autoimmune thyroid disease. One study found that autoimmune diseases occurred in 14% of patients with CD compared with only 2.8% of control subjects. The risk also was found to increase with duration of gluten exposure. One argument for screening patients for CD is that it may also reduce the risk of development of other disorders, although the data on this are limited.
Traditional approach to testing
Based on the range of patient presentations from asymptomatic to malaborptive symptoms, and the associated intestinal and extraintestinal manifestations that may develop, it is necessary to know when and which patients to test for CD. The American College of Gastroenterology Clinical Guideline published in the American Journal of Gastroenterology in 2013, outlined those individuals who should be tested for CD ( Box 1 ).
- 1.
Patients with symptoms, signs, or laboratory evidence suggestive of malabsorption, such as chronic diarrhea with weight loss, steatorrhea, postprandial abdominal pain, and bloating, should be tested for CD—strong recommendation, high level of evidence.
- 2.
Patients with symptoms, signs, or laboratory evidence for which CD is a treatable cause should be considered for testing for CD—strong recommendation, moderate level of evidence.
- 3.
Patients with a first-degree family member who has a confirmed diagnosis of CD should be tested if they show possible signs or symptoms or laboratory evidence of CD—strong recommendation, high level of evidence.
- 4.
Consider testing of asymptomatic relatives with a first-degree family member who has a confirmed diagnosis of CD—conditional recommendation, high level of evidence.
- 5.
Celiac disease should be sought among the explanations for elevated serum aminotransferase levels when no other cause is found—strong recommendation, high level of evidence.
- 6.
Patients with type I diabetes mellitus should be tested for CD if there are any digestive symptoms, or signs, or laboratory evidence suggestive of celiac disease—strong recommendation, high level of evidence.
As outlined in Box 1 , testing should be performed on those with signs or symptoms of malabsorption including diarrhea, weight loss, or vitamin deficiencies. Given the genetic hereditability, all children and any symptomatic patients with first-degree family members with CD should be tested, and an argument can be made for also testing subclinical adults. CD should also be tested in patients with elevated serum aminotransferase levels without other cause, or in patients with autoimmune diseases, such as type 1 diabetes mellitus with symptoms or consistent laboratory findings.
Serologic Testing
Serologic testing for diagnosis should be performed on patients while on a gluten-containing diet. Noninvasive screening for CD includes serologic testing with EMA, AGA, antitissue anti-tTG, and anti-DGP. Given the lower sensitivity and specificity of AGA for CD, the EMA, anti-tTG, and anti-DGP tests have largely replaced AGA testing.
Following AGA, IgA antibodies against the endomysium (EMA) of monkey esophagus was discovered as being highly sensitive and specific in the diagnosis of CD. IgA EMA has been evaluated by many studies and the pooled sensitivity was found to be 97.4% with a specificity of 96.1%. However, there was some variation in sensitivity in the studies with one reporting the sensitivity as 75%. The IgA EMA can also be performed using human umbilical cord as substrate with pooled sensitivity and specificity of 90.2% and 99.6%, respectively.
Further research identified the enzyme tissue transglutaminase as the autoantigen that reacts with EMA, which led to the development of enzyme-linked immunosorbent assays that detect anti-tTG. Most commercial tests for IgA tTG use human-recombinant or red-cell derived tTG as a substrate with mixed-age population pooled estimates of sensitivity and specificity of 90.2% and 95.4%, respectively.
Most EMA and anti-tTG testing are IgA-based tests, and therefore a total IgA should be measured to exclude IgA deficiency. Selective IgA deficiency has a prevalence of about 1.7% to 3% in patients with CD, which is 10 to 15 times more common than in the general population.
More recently, anti-DGP have also been studied, which are IgG- or IgA-based and can therefore be used for testing in patients with IgA-deficiency. DGP IgA and DGP IgG antibodies have been shown to have a sensitivity of 94% and specificity of 99%, and a sensitivity of 92% and 100%, respectively.
There is continued debate to determine if noninvasive testing could be used as the only testing required for diagnosis of CD. In children the use of two separate serologic tests if positive greater than 10 times the upper limit of normal with positive genetic markers for HLA-DQ2 and HLA-DQ8 has been suggested as a possible testing algorithm, thus avoiding the need for upper endoscopy with biopsy. However, 2% to 3% of people with CD have negative results in serologic tests, have low serologic titers, or have fluctuating titers, thus upper endoscopy with biopsy is still the gold standard for confirmation.
Genetic Testing
HLA genotyping in the diagnosis of CD, specifically to look for HLA-DQ2.5 and HLA-DQ8, is useful for its negative predictive value. Less than 1% of patients with CD are negative for HLA-DQ2 and HLA-DQ8. The positive predictive value, however, is low because a large proportion of individuals without CD carry either HLA-DQ2.5 or HLA-D8. Specifically, the prevalence of DQ2 in the general population is around 30% to 40%, and the prevalence of DQ8 in the general population is around 5% to 10%.
HLA genotyping has therefore been found to be helpful in patients with suspected CD who fail to respond to a gluten-free diet, to help rule out the disease. HLA typing can also be useful in patients who have self-diagnosed themselves with CD and are on a gluten-free diet at the time of presentation, because serologic testing and biopsy may be less accurate. HLA typing can also help to rule out CD in high-risk patients, such as patients with first-degree relatives with CD, thus minimizing further testing.
Endoscopic Evaluation
Endoscopy with biopsies should be performed while the patient is on a gluten-free diet to make the diagnosis of CD. Upper endoscopy allows for the direct observation of gross mucosal changes of the small bowel including scalloping, reduction of villous folds, and nodularity ( Fig. 1 ). Although endoscopic markers are helpful, they are not sensitive or specific and biopsy is required. Chromoendoscopy using indigo carmine or methylene blue may be helpful in enhancing visualization of patchy areas for biopsy.
