**4. Diagnosis**

The diagnosis of CD may require genetic and serological tests and a duodenal biopsy.

### **4.1 Genetic risk markers**

The main genetic risk factor for CD is the presence of HLA-DQ2 and -DQ8 heterodimers, which are identified in 90% and 5–7% of patients with CD, respectively [7]. Since these alleles are found in 30–40% of the general population (HLA-DQ2 being the most common) [39], their absence is important due to their negative predictive value (NPV). Therefore, the HLA-DQ2/HLA-DQ8 test plays an important role in CD diagnosis and is recommended in the following situations [40]—(a) exclusion of the disease, especially in patients who have started GFD; (b) in situations of uncertain diagnosis due to negative serology, but histology suggestive of CD; (c) to differentiate siblings in whom it is intended to ensure that it is unlikely that they will develop the disease from those who will need monitoring; (d) in subjects with autoimmune diseases and other diseases in which CD should be investigated.

A negative result for HLA-DQ2/HLA-DQ8 means a very low probability of developing the disease. Therefore, this test can be used to support the diagnosis of CD, since it has a high NPV, allowing exclusion with 99% certainty [41]. However, it has little positive predictive value (PPV) (only around 12%), so its determination has no diagnostic value in situations with elevated antibodies directed against tTG and should be reserved as second-line in patients with diagnostic doubt [42, 43].

## **4.2 Specific serum antibodies**

Various serological tests have been developed to detect CD—antigliadin antibodies (anti-AGA), antibodies against deaminated gliadin peptides (anti-DGP), antiendomysia antibodies (anti-EMA), and anti-transglutaminase antibodies (anti-tTG). Serological tests are important for two reasons—(1) they select patients in whom duodenal biopsy should be indicated to confirm clinical suspicion, and (2) they confirm the diagnosis in cases in which enteropathy has been observed [43].

Anti-AGA has been used for decades and is reasonably safe when the probability of suffering from CD is very high. However, it has been shown that these antibodies present variability in their diagnostic precision, due to the fact that they have low sensitivity and specificity; therefore, they should not be included in routine tests for the diagnosis of CD [41, 44].

Anti-EMA has a relatively low sensitivity (80–90%), but its specificity is close to 100%. However, they require more complex laboratory techniques and depend on the experience of the laboratory staff, remaining as a second-line test adequate to confirm clinical suspicion [1].

### *Celiac Disease, Management, and Follow-Up DOI: http://dx.doi.org/10.5772/intechopen.104652*

Anti-tTG IgA has a sensitivity and specificity of 95 and 90%, respectively [41, 45]. Anti-DGP has shown good precision, although lower than anti-tTG IgA, so an isolated positive result for IgA and/or IgG-DGP in patients at low risk for CD, predicts the disease only in 15%, being in the rest of the cases false positives. Therefore, in a first approximation, anti-tTG are the preferred antibodies for the diagnosis of CD according to the ESPGHAN diagnostic criteria [46, 47]. Anti-DGP is considered less sensitive or specific for the detection of CD compared to anti-tTG and anti-EMA. However, these last two antibodies are less sensitive in children under 2 years of age. It should also be taken into account that anti-tTG can be negative in 5–16% of patients with histologically confirmed CD [48]. Therefore, there is no serological test with perfect sensitivity and specificity [44]. In case of general IgA deficiency, which is observed in 2–3% of patients with CD, the IgGbased test (anti-DGP IgG and anti-tTG IgG) should be performed. IgG anti-tTG has diagnostic utility in patients with selective IgA deficiency (IgA < 0.07 mg/dl). Regarding anti-DGP IgG, there is no evidence of greater efficacy compared to antitTG IgG or anti-EMA IgG [41].

### **4.3 Intestinal biopsy**

Duodenal biopsy of the small intestine is a key point in the diagnosis of CD. A distinctive pattern of histological abnormalities has been identified in this disease, including partial or total villous atrophy, elongated crypts, decreased villus/crypt ratio, increased crypt mitotic index, increased crypt density of intraepithelial lymphocytes (IELs), and infiltration of plasma cells in the lamina propria. An increase in IELs tends to be located at the tips of the villi and are usually CD8+ [37]. The presence of a diffuse and uniform infiltrate of these lymphocytes is the most sensitive finding, but it is not specific to CD. A count of at least 25 IELs/100 enterocytes represents a definitive increase in IELs [49, 50]. Immunohistochemical studies have shown that the increase in IELs represents an expansion of cytotoxic T cells alpha-beta and gamma-delta. Gamma-delta T cells are observed in 1–10% of the normal small intestinal mucosa but increase in patients with CD, where they may represent 15–30% of all IELs [1]. In addition, the absence of the brush border can be identified, as well as alterations in epithelial cells.

There are three grading systems to establish the severity of histological damage proposed by Marsh, Oberhuber [51], and Corazza-Villanaci [52]. Marsh system, with three types of grades, was replaced in 1999 by Oberhuber [51], which proposes a better standardization with six types [51]. In 2007, a new, simpler classification was published by Corazza-Villanacci [52]. These classifications are qualitative and subjective [1, 37]. Marsh-Oberhuber classification is used by most pathologists both for diagnosis and to ensure regression of the lesion after GFD [1]. Generally, six stages are distinguished—type 0 without lesion, type 1 (infiltrative lesion), type 2 (crypt hyperplasia), type 3 (villi atrophy: 3a: partial; 3b: subtotal; 3c: total) [51]. Furthermore, these lesions are not pathognomonic for CD, and there is a wide spectrum of diseases that can produce indistinguishable microscopic lesions.

Currently, it is considered that, in patients with high levels of antibodies, the diagnosis could be based on the combination of symptoms, antibodies determination, and genetics, omitting in this case the duodenal biopsy [11, 46], unlike what was established in the previous ESPGHAN guidelines for the diagnosis of CD. However, confirmation of CD by biopsy is considered the gold standard in the diagnosis of CD in certain types of patients.

The biopsy can be used to diagnose and monitor, but CD is a burden for patients. Therefore, less invasive and objective biomarkers are required to assess the disease. In addition, in certain patients, a challenge with gluten is necessary to make a correct diagnosis of CD. Based on this, Leonard et al. [53] investigated the ability of different biomarkers to diagnose CD after provocation. These biomarkers could, complement or replace histology in the diagnosis of CD. These authors evaluated traditional diagnostic techniques, such as biopsy, antibodies, symptomatology, as well as different biomarkers to measure the response to two levels of gluten exposure, studying interleukin-2 (IL-2), the tetramer test, and the dot enzyme-linked immunosorbent assay (Enzyme-Linked ImmunoSpot Assay, ELISpot), among others. Results showed that the measurement of IL-2 in plasma might be the first and most sensitive marker for the evaluation of gluten exposure in patients with CD. This study provides a framework for the rational design and selection of biomarkers in future gluten challenge studies with the goal of incorporating them into clinical practice.
