Edgardo Smecuol, MD Small Bowel Section; Clinical Gastroenterology Unit Department of Medicine Dr. C. Bonorino Udaondo Gastroenterology Hospital Buenos Aires, Argentina
	Julio C. Bai, MD Head of the Department of Medicine Dr. C. Bonorino Udaondo Gastroenterology Hospital Buenos Aires, Argentina Professor; Gastroenterology Chair; Universidad del Salvador, Buenos Aires, Argentina

Celiac disease (CD) is a life-long chronic systemic disorder characterized by an immunologically mediated enteropathy induced by the dietary gluten, which improves with the exclusion of the toxic protein from diet. The disorder affects genetically susceptible individuals carrying the human leukocyte antigen (HLA) class II DQ2 and/or DQ8.

Some decades ago, CD had been considered a rare disorder basically due to the fact that diagnosis was based primarily on the presence of  gastrointestinal (GI) symptoms, such as diarrhea, accompanied by signs of malabsorption and weight loss. However, recent epidemiological studies have revealed that CD is much more common than previously thought. Thus, the prevalence of this condition is shown to be close to 1% in Europe, the Americas, North Africa, the Middle East, and India ^2,3. Furthermore, recent research has also found evidence that the increasing rates of CD patients detected worldwide may reflect a true rise in disease prevalence ⁴.

In the clinical context, the extraordinary expansion in the number of patients diagnosed of CD correlates with a notably wide variety of clinical manifestations of the disorder ranging from severe GI symptoms and malnutrition (very rare today) to silent cases that are diagnosed  incidentally or in relatives of already-diagnosed patients. Indeed, the majority of cases currently diagnosed are characterized by extra-intestinal manifestations, including dermatitis herpetiformis, chronic anemia, gastroesophageal reflux symptoms, hypertransaminasemia, osteopenia or osteoporosis, puberal delay, aphthous stomatitis, short stature, cerebellar ataxia, infertility, etc. ¹. However, the prevalence of different clinical forms is variable according to the geographic area considered and the degree of expertise of the medical community. This diversity of symptoms represents a challenge to the professionals not familiar with CD. Many patients are diagnosed as a result of screening in population at risk or by incidental endoscopic findings of characteristic duodenal mucosal features in patients undergoing upper endoscopy for reasons  other than suspected CD ⁵. It is interesting to note that more that 50% of the newly diagnosed cases from screening relatives of CD patients  are completely silent and, otherwise, would not have required consultation for medical attention ³. In addition, compared with two decades ago, a greater proportion of patients are being diagnosed later in the adulthood ⁵.

Importance of Detection

Symptomatic CD patients are affected in the daily life, while introduction of a gluten-free diet (GFD) most often induces significant  improvement of symptoms, abnormal biochemical measures, and impaired quality of life. Concern remains about the long-term consequences of CD and whether maintaining a life-long GFD is necessary for all CD patients. However, understanding of the outcomes of CD is complicated by a number of factors:

• The broad clinical spectrum of the disease determines that a large number of subjects have minimal or no symptoms;
• The bulk of current scientific knowledge has been derived from a minority of patients (mostly symptomatic), but the natural history of undiagnosed CD is unclear; and
• The influence of the degree of adherence to the GFD which might contribute to the prevention or treatment of complications.

Notably, the ratio of patients with undetected CD to those with diagnosed disease is approximately 8:1, which seems to be higher in some countries ^1,3. Many published studies exploring the prevalence of CD-associated adverse events or complications have been limited by the sample population of patients, choice of control groups, ascertainment bias, etc.

CD has been associated with increased mortality and malignancy compared with the general population (all–cause mortality: 30% to 40% increment; cancer mortality: 60% to 80% increment) ⁶. The risk of malignancies in CD patients, in particular lymphomas, has been estimated at 1.3 times higher than that of the general population, which is lower than data reported in the 1970s and 1980s. Recent studies have raised concerns that the risk of malignancy in people with undetected CD may be high. Interestingly, Ludvigson et al. ⁷ found that patients with mild enteropathy not defined as CD have increased mortality compared with the control population. Finally, the effect of the GFD on preventing of CD-related mortality or malignancies remains undefined.

Non-malignant complications of CD include a modest increase in fracture risk as well as neurological and/or psychiatric conditions.  Reproductive problems are present in a minority of cases and may have been overestimated. Contrary to the uncertainties of malignancy risks, the available literature has convincingly demonstrated a positive effect of early diagnosis of CD and prompt GFD treatment on the prevention and improvement of these nonmalignant complications. Therefore, early detection for CD and long-term strict adherent to GFD treatment have been strongly recommended by experts ^8,9.

Diagnosing Celiac Disease: Past and Present

Until the 1950s, the diagnosis of CD was based on clinical observations focused on malabsorptive features. The peroral intestinal biopsies, introduced in 1956, marked a significant change in CD diagnosis. Since then, histological assessment of intestinal mucosa, with evidence of characteristic gluten-dependent mucosal damage, is considered the gold standard for CD diagnosis.

In last two decades, upper GI endoscopy has gained prominence in mucosal sample collection, which is less invasive than peroral biopsy ¹⁰. Today, endoscopy is widely used for duodenal sampling. In addition, the endoscopic procedure allows the incidental observation of typical duodenoscopic features that are highly predictive of the disease. However, endoscopic markers alone are not very sensitive (<50% in low-risk  populations) for CD diagnosis; while apparently normal duodenal mucosa cannot exclude CD ¹⁰.

Histological damage is considered characteristic but not pathognomonic of CD, as similar lesions are seen in several other disorders. According to criteria developed by the European Society of Pediatric Gastroenterology and Nutrition, gluten-dependence of the intestinal mucosal damage should be established by histological improvement in response a GFD and recurrence of enteropathy after reintroduction of the offensive protein (the “Interlaken criteria”). This diagnostic algorithm requiring three consecutive biopsies was challenged by experts and rejected by patients and physicians. Thus, it was suggested that the first biopsy alone is highly predictive of CD (>95%). Meanwhile, the advent of CD-related serology tests, introduced more the 25 years ago, contributed to the revision of CD diagnostic criteria ^1,5. Increasingly sensitive  and specific serologic tests provide more options for diagnosis of the disorder. Currently, serological tests are considered important surrogate markers for gluten-dependent damage ^9,10.

In current practice, the diagnosis of CD hinges on the diagnostic intestinal biopsy and the concomitant presence of a positive CD-specific serology. A second (post-treatment) biopsy should be reserved for patients in whom the first biopsy and serologic test were inconclusive (e.g. seronegative enteropathy) or for patients on a strict GFD who have less than expected outcomes. Gluten challenge should be reserved for patients on a GFD without an unequivocal diagnosis. Genetic typing (HLA DQ2 and DQ8) can be useful tools for excluding CD in those whose tests remain inconclusive ¹⁰.

Characteristic Enteropathy

As stated above, the diagnosis of CD requires histological examination of small bowel biopsies. The disease has an almost invariable gradient of lesions, more pronounced in the proximal intestine (duodenum) while minor or not apparent distally. Patients with severe CD may have damage in the ileum. Therefore, duodenal sampling is considered sufficient to establish a diagnosis of CD; only exceptional cases require more distal biopsies. It is important to note that duodenal biopsies have some limitations (i.e., as a result of size and orientation of biopsy that may  affect the diagnostic accuracy. Moreover, the patchy nature of mucosal damage in CD should be considered. Biopsy samples taken from proximal duodenum (above Vater's papilla) may have artifacts (e.g., stretching of villi) produced by submucosal Brunner glands, which may be falsely interpreted as flat mucosa. However, a recent prospective study ¹¹ estimated that 13% of new CD patients are diagnosed based only on biopsies obtained in the duodenal bulb. Evidence shows that larger-sized and more biopsy samples (≥4) produce more accurate diagnosis as they minimize the bias of the patchy distribution of lesions.

Certain techniques during the endoscopic procedure can help physicians obtain better samples for accurate evaluation. For example, duodenoscopy using videoendoscopes with or without enhancing procedures (e.g. water immersion, vital staining, magnification, etc.) is  helpful for detecting endoscopic markers and orientating sampling.

Under light microscopy, the most characteristic histological findings in patients consuming a gluten-containing diet are:

• blunted or atrophic villi,
• crypt hyperplasia,
• mononuclear cell infiltration of the lamina propria,
• epithelial changes, including structural abnormalities in epithelial cells and intraepithelial lymphocyte infiltration.

Through a series of well-designed studies, M. Marsh ¹² interpreted the spectrum of mucosal lesions as representing the progression of   damage induced by gluten and categorized the CD histological modifications as ranging from normal mucosa (an initial event defined as  Marsh's type 0) (e.g. 30% of cases with well-established dermatitis herpetiformis) to complete flat villi. The Marsh classification has been widely used in clinical practice. Abnormal villi structure is considered the destructive type of lesion (Marsh type III), which is highly predictive of CD independent of other subrogate markers of gluten dependency. In 1999, Rostami et al. ¹³ created subcategories within the Marsh type III: A (partial villous atrophy), B (subtotal villous atrophy), and C (total villous atrophy). Damage of different intensity in the same sample or  in different samples obtained at the same time is often observed.

In the middle of the spectrum of CD mucosal damage is the presence of subtle abnormalities, defined by Marsh as type I, normal villi with increased intraepithelial lymphocytes, or type II, hyperplastic damage with increased intraepithelial lymphocyte density and crypt hyperplasia.  Because these subtle abnormalities can be detected in many other disorders as well (for example, 10% of cases with only increased IEL density  are finally diagnosed of CD) ¹², patients with minimal enteropathy need additional examination.

Conventional diagnosis of CD requires an intestinal damage type III A or worse ¹⁴. There is a general agreement among experts that, as other surrogate markers of gluten-dependence become available, new standards will be developed for diagnosing this disorder.

Celiac Disease-Specific Serology

CD-specific serological tests have gained importance over time and have been used for more than 20 years as tools for screening suspected CD patients who require an intestinal biopsy. Immunoglobulin A (IgA) and IgG antigliadin antibodies (AGA) addressing crude gliadin as a substrate were the first tests introduced. Later studies demonstrated disappointing sensitivity and specificity of AGA, making them unsuitable for screening CD ^15-17. Subsequently, the detection of IgA antibodies against the endomysium (EmA) of monkey esophagus was shown to be highly sensitive and specific. Thus, the immunofluorecence assay became an essential tool in the serological algorithm for CD screening. In  addition, the test assesses gluten dependence of the characteristic histological findings. Further research identified the ubiquitous enzyme tissue transglutaminase as the autoantigen reacting with EmA, leading to the development of the tissue transglutaminase antibody (tTG) test, which uses the ELISA platform ¹⁵. As the performance of new tests was demonstrated, the AGA serology test has become obsolete ⁹.

More recently, a new generation of antigliadin antibody assays using synthetic deamidated gliadin peptides as substrates has shown equivalent performance to the conventional tTG test ¹⁶. These antibodies are IgA and/or IgG based. Specifically, IgGbased tests are useful for detecting CD in selective IgA-deficient cases. In addition, combining different isotypes of antibodies in a single assay can add sensitivity to screening algorithms while a second assay substantially increases diagnostic performance ¹⁷.

Thus, a number of serological markers are now available for use for two purposes: for selecting patients appropriate for biopsies, or to confirm a diagnosis of CD. Nevertheless, the accuracy and reliability of serological tests have been established in studies conducted in research settings  under experimental conditions and may not reflect the accuracy in clinical practice.

Because of the invasiveness and complexity of obtaining histological diagnosis of CD in clinical practice, where misdiagnosis (over- and under-interpretation of enteropathy) is relatively common, the usefulness of serology tests as the sole diagnostic method for CD has also been explored. Based on the high positive predictive values of serological tests, some authors have suggested that intestinal biopsy is no longer mandatory for diagnosing CD in at least some patients ¹⁷. Various approaches to finding a blood test-based CD diagnosis are being  investigated. If confirmed, the diagnosis of CD will become more convenient for patients and more practical and accurate in general practice.

Who Should Be Tested for CD?

Testing for CD should be considered in a variety of clinical situations for the following types of at-risk populations: 1) persons with  gastrointestinal symptoms including chronic diarrhea, malabsorption, weight loss, and abdominal distension; 2) individuals without other explanation for signs and symptoms including persistently elevated serum aminotransferases, short stature, delayed puberty, iron-deficiency  anemia, recurrent misconception, osteoporosis at a young age, infertility; 3) conditions associated with a high risk for CD, including type 1 diabetes mellitus, autoimmune disorders, endocrinopathies, first- and second-degree relatives of individuals with CD (probably the most  relevant at risk population), patients with Turner, Down, or Williams syndromes. Other conditions for which CD testing may also be considered include irritable bowel syndrome, persistent aphthous stomatitis, peripheral neuropathy, cerebellar ataxia and other neurological manifestations, and dental enamel hypoplasia ^1,5,9,10.

Conclusion

The diagnosis of CD is currently a challenge for gastroenterologists and other physicians. Increasing awareness and recognition of the clinical variability of the disorder in the medical community are essential for improving the accuracy of CD diagnosis. Moreover, clinicians should consider comorbid conditions associated with a higher risk of CD. Use of appropriate serologic tests, followed by intestinal biopsy, has proved to be a cost-effective algorithm for diagnosing CD. It is essential that a patient’s clinical, serological, and histological changes before and after the administration of the GFD be carefully evaluated if the patient’s histology and serology results are inconsistent or inconclusive.

Currently, duodenoscopy is the preferred procedure for obtaining duodenal biopsies. Physicians must be aware of the patchy distribution of the mucosal lesions and choose biopsies in areas with more severe damage. Histological assessment of biopsies carries potential risks of  misdiagnosis, especially in the community medical setting, which may be related to the poor quality of samples and/or assessment by general pathologists. CD-specific serology has a well-established place in the diagnostic armamentarium, but its role as a single tool has not yet been clearly defined. Finally, genotyping for HLA DQ2 and DQ8 may be valuable in equivocal cases when a negative result excludes CD.

References

1. Green PH, Cellier C. Celiac disease. N Engl J Med 2007; 357: 1731-43
2. Catassi C. The world map of celiac disease. Acta Gastroenterol Latinoam 2005; 35: 37-55
3. Fassano A, Catassi C. The global village of celiac disease and its evolution over time. E-WGN 2011; 15: 5-8
4. Lohi S, Mustalahati K, Kaukinen D, et al. Increasing prevalence of celiac disease over time. Aliment Pharmacol Ther 2007; 26: 1217-25
5. Di Sabatino A, Corazza GR. Coeliac disease. Lancet 2009: 373: 1480-93
6. Grainge MJ, West J, Card TR, et al. Causes of Death in People With Celiac Disease Spanning the Pre- and Post-Serology Era: A Population-Based Cohort Study From Derby, UK. Am J Gastroenterol. 2011 [Epub ahead of print].
7. Ludvigsson JF, Montgomery SM, Ekbom A, et al. Small-intestinal histopathology and mortality risk in celiac disease. JAMA. 2009; 302: 1171-8.
8. Roston A, Murray JA, Kagnoff MF. American Gastroenterological Association Institute Technical Review on the diagnosis and management of Celiac Disease. Gastroenterology 2006; 131: 1981-2002.
9. Bai JC, Zeballos E, Fried M, Corazza GR, Schuppan D, Farthing MJG, et al. Celiac Disease. WGO-OMGE Practice Guidelines. World Gastroenterology News 2005;10: S1-S8
10. Dickey W. Endoscopic marker for celiac disease. Nat Clin Pract Gastroenterol Hepatol 2006; 3: 546-51.
11. Gonzalez S, Gupta A, Cheng J, et al. Prospective study of the role of duodenal bulb biopsies in the diagnosis of celiac disease. Gastrointest Endosc. 2010; 72:758-65.
12. Marsh MN. Gluten, major histocompatibility complex, and the small intestine. A molecular and immunobiologic approach to the spectrum of gluten sensitivity ('celiac sprue'). Gastroenterology 1992; 102:330-54.
13. Rostami K, Kerckhaert J, Tiemessen R, et al. Sensitivity of antiendomysium and antigliadin antibodies in untreated celiac disease disappointing in clinical practice. Am J Gastroenterol 1999; 94: 888-94.
14. Working Group of the United European Gastroenterology Group in Amsterdam. When is a coeliac a coeliac? Eur J Gastroenterol 2001; 13: 1123-8
15. Rostom A, Dubé C, Cranney A, Saloojee N, Sy R, Garritty C, et al. The diagnostic accuracy of serologic tests for celiac disease: a systemic review. Gastroenterology 2005; 128: S38-S46
16. Sugai E, Vázquez H, Nachman F, et al. Accuracy of testing for antibodies to synthetic gliadin-related peptides in celiac disease. Clin Gastroenterol Hepatol. 2006; 4: 1112-7.
17. Sugai E, Moreno ML, Hwang HJ, et al. Celiac disease serology in patients with different pretest probabilities: is biopsy avoidable? World J Gastroenterol. 2010; 16: 3144-52.