An updated overview of spectrum of gluten-related disorders: clinical and diagnostic aspects

Celiac disease (CD) is a common GRD in which genetic and environmental factors as well as gluten intolerance are the main causes of innate and adaptive immune responses [14‐18]. CD is characterized by small intestine mucosal lesions, subtotal, or total intestinal villi atrophy and nutrient malabsorption [19]. The global prevalence of CD is estimated at 1–2% in the general population and 0.3–2.9% in children [20, 21].

CD can be associated with a wide spectrum of manifestations, including intestinal and extra-intestinal symptoms or it can even be asymptomatic [7]. Common intestinal features include chronic and persistent diarrhea, malabsorption, abdominal pain, weight loss, and steatorrhea. Atypical and extra-intestinal manifestations include hepatic hypofunction, iron deficiency anemia, hair loss, osteoporosis, growth retardation, epilepsy, psychiatric disorders, mouth ulcers, muscle weakness, fatigue, arthropathy, delayed onset of puberty in children and infertility in adults [7, 22‐25].

Conditions associated with CD include genetic disorders such as Downs syndrome, Turners syndrome and Williams syndrome; autoimmune disorders including type 1 diabetes mellitus (DM1), inflammatory bowel disease (IBD), autoimmune thyroid disorders, autoimmune hepatitis; neurological disorders like ataxia and epilepsy [26, 27].

CD can present at any age after the introduction of gluten to the diet [28]. Children under 2 years old typically present with gastrointestinal symptoms and failure to thrive. Older children and adults typically present with symptoms that are mostly nonspecific and atypical [29‐31]. The differences in the clinical presentation of CD may be due to immunological factors, the age of onset, the duration and the extent of disease, degree of small intestinal mucosal inflammation, gender, and family history [26].

The diagnosis of CD is challenging and it should be considered when patients present with either intestinal or extra-intestinal symptoms, such as bloating or fatigue. CD is more common in patients with a family history of CD and DM1 than in the general population, even in the absence of gastrointestinal symptoms [32]. The correct diagnosis of CD requires a combination of clinical, serological, and histopathological evaluations [33]. It is recommended that patients with a clinical presentation of CD should undergo serological tests [34]. Several antibodies can be used in CD detection such as anti-tissue transglutaminase (Anti-tTG), anti-endomysial (EMA), and anti-deamidated gliadin peptides (Anti-DGP), IgA and IgG antibodies [32, 35]. Anti-tTG antibodies are the most common serologic markers for CD diagnosis that have 96–98% sensitivity and 88–100% specificity [36, 37]. IgA-tTG is the recommended serological test for the detection of CD. As IgA-deficiency affects 2–3% of CD patients and leads to false-negative results, total IgA levels also need to be measured. In the presence of IgA-deficiency, IgG antibody-based tests (IgG-tTG and/or IgG-DGP) should be used [32, 37]. High tTG-antibody levels (> 5 times the upper normal unit (ULN)) is suggestive of a diagnosis of CD. IgA-EMA antibody-based tests have a high sensitivity and specificity for the diagnosis of CD. These tests can be used as additional and confirmatory serological tests for the initial diagnosis of celiac disease, in conjunction with the measurement of anti-tTG antibodies. Unfortunately, IgA-EMA antibody tests are expensive and user-dependent [37, 38]. An antibody test followed by a small intestinal biopsy evaluation is the gold standard for the definitive diagnosis of CD [36, 39]. Adequate small intestinal sampling is essential in this regard and studies have concluded that at least one biopsy from the duodenal bulb and at least four biopsies from the distal duodenum are needed for an accurate diagnosis [40, 41]. Biopsy appearances in keeping with CD include scalloping, villous flattening, and fissuring of mucosal folds. Small intestinal appearances can become more pronounced with disease progression [17, 40, 41]. According to the Marsh classification, the intestinal biopsy changes are classified on the extent of increased intraepithelial lymphocytes, crypt hyperplasia, and villous atrophy [42]. Corazza and Oberhuberhave also proposed modifications to the Marsh classification [43, 44] (Table 1), and these proposed modifications have been challenged, Marsh et al. [45]. CD diagnostic tests (serologic and endoscopic tests) should be performed when the patient is on a gluten-containing diet to avoid false-negative results [46]. Human leukocyte antigens (HLA-DQ2/DQ8) are the most important genetic risk factors for celiac disease [47]. HLA typing can be used when the results of the serological and histopathological tests are inconclusive and the diagnosis of CD is uncertain [48, 49]. As almost all patients with CD have HLA-DQ2 or HLA-DQ8, the absence of these HLAs makes the diagnosis of CD very unlikely [47, 50, 51]. HLA-DQ2 is found in 90–95% of CD patients, and the other 5% have HLA-DQ8 variant [52]. The European Society for Pediatric Gastroenterology, Hepatology, and Nutrition (ESPGHAN) guidelines suggest that the diagnosis of CD in pediatrics can be made without biopsy evaluation, if the patient has a high level of anti-tTG antibodies (> 10 times ULN) along with testing positive for EMA-antibodies and HLA-DQ2/DQ8 haplotypes. Biopsy evaluation is advised as being essential if either the EMA or HLA results do not support a diagnosis of CD. The guidance also suggests that if patients have symptoms suggestive of CD, but their anti-tTG level is < 10-time ULN, or when the patients are asymptomatic, but the anti-tTG level is > 10-time ULN, endoscopic intestinal biopsy is advisable [53]. In addition to the ESPGHAN guidelines, Rubio et al. [32] (Fig. 1) and Mayo medical laboratories have given an algorithm for the diagnosis of CD (Fig. 2) [7]. There are emerging methods for the diagnosis of CD. These methods include video capsule endoscopy (VCE), biochemical tests such as measurement of intestinal fatty acid-binding protein (I-FABP), radiology methods and intestinal permeability tests that can provide additional information about the appearance and function of the small intestinal mucosa, increasing the detection and evaluation of CD [54].

Dermatitis Herpetiformis (DH) is a chronic, autoimmune, and recurrent cutaneous-intestinal disorder identified in genetically susceptible individuals, which is often associated with CD [67, 68]. Anti-tTG antibodies that are produced in response to gluten exposure can also recognize epidermal transglutaminase (ETG). ETG is structurally homologous to tTG and is the main antigen in DH [67]. Deposition of IgA antibodies in the superficial papillary dermis of DH patients causes vesiculobullous, pruritic, and localized lesions. DH affects the extensor surfaces such as elbows, buttocks, knees, and scapular areas [67, 69]. DH is prevalent in Scandinavian countries and the UK and typically presents in patients aged between 15 and 40 years. Males are affected more than females, with a ratio of 3:2, but interestingly it is more prevalent in females than males under the age of 20 years [67, 70].

DH patients rarely present gastrointestinal symptoms, although all of them have gluten sensitivity. Approximately two-thirds of patients have some degree of villous atrophy, and one third has intraepithelial lymphocytosis [67].

DH is associated with a wide range of autoimmune diseases, such as T1DM, pernicious anemia, Addison’s disease, vitiligo, alopecia areata, and rheumatoid arthritis, systemic lupus erythematosus, Sjogren ‘s syndrome and thyroid abnormalities [71].

Skin biopsy evaluation is advised in patients with clinical manifestations that are suggestive of DH and should be taken close to, but not from any vesicles. Direct immunofluorescence (DIF) should be performed on skin biopsy specimens [72]. The classic histopathological finding of DH is a sub-epidermal cleft rich in neutrophils and eosinophils that presents at the dermal papillae [67, 72, 73]. For patients with clinical presentations of DH, but negative DIF, other confirmatory tests (such as anti-tTG antibody level) can be applied. Both CD and DH patients have raised tissue transglutaminase specific auto-antibodies level in serum and small bowel mucosa [69, 74]. An additional serological test used in the diagnosis of DH is the measurement of serum EMA-antibody levels; this has sensitivity and specificity of at least 90% [67, 75]. Recently, some researchers have introduced the anti-deamidated gliadin peptide antibody (anti-DPG) as a possible marker of DH. Anti-DPG has a high sensitivity (between 84 and 90%), and it can be useful in the diagnosis of cases with suspected DH with negative anti-tTG [76, 77]. ETG is the key auto-antigen in DH, therefore, it can efficiently differentiate DH from other dermatological diseases. Its sensitivity and specificity have been reported between 52 and 90%, and 93 and 100% respectively, nevertheless ETG is currently not approved as a diagnostic marker of DH [73, 78‐80]..

DH is treated with a GFD combined with pharmacological treatment including sulfones, such as dapsone, and sulfonamides [81]. Dapsone is an anti-inflammatory agent that downregulates neutrophil chemotaxis, reduces the release of leukotrienes and prostaglandins, and thus prevents tissue damage [82]. Possible side effects of dapsone include hematological disturbances such as methemoglobinemia and agranulocytosis [83].

Gluten ataxia (GA) is a type of cerebellar ataxia caused by exposure to gluten in sensitive and genetically susceptible individuals [84]. GA is an autoimmune disorder characterized by the presence of a cerebellar injury, affecting mainly Purkinje cells. In most cases of GA, there has been a previous diagnosis of CD or NCGS with digestive symptoms [84]. Several studies have suggested possible mechanisms for the development of GA in CD. Impaired intestinal absorption leading to vitamin E deficiency can cause spinocerebellar degeneration [85]. Malabsorption can also cause damage to the serotonin-containing neurons in the cerebellum, and brainstem [86]. Immunological and inflammatory processes may also be important in the etiology of GA. There is a cross-reactivity between antigenic epitopes located at the level of Purkinje cells and gluten-related antibodies. In susceptible individuals, anti-gliadin antibodies may have a clinically significant direct or indirect neurotoxic effect [87].. Hadjivassiliou et al. [84] estimated that GA accounts for approximately 15% of all ataxias and 40% of all idiopathic sporadic cerebellar ataxias. GA is more common in the USA and Europe than in Asia. It typically affects males and females aged over 50 years [88].

The clinical manifestations of GA are similar to those of other ataxias and include ocular signs like gaze-evoked nystagmus (84%), dysarthria (66%), upper limb ataxia (75%), lower limb ataxia (90%), gait ataxia (100%) and additional movement disorders such as myoclonus, chorea, palatal tremor and opsoclonus myoclonus [89]. GA is characterized by gradual onset of gait ataxia, associated with peripheral neuropathy. Occasionally, it can be rapidly progressive, similar to paraneoplastic cerebellar degeneration [90]. The absence of autonomic dysfunction helps differentiate these patients from patients with the cerebellar type of multiple system atrophy (MSA-C) [91].

The diagnosis of GA is supported by the presence of anti-gliadin, anti-tTG, and anti-TG6¹ (when available) antibodies in the serum. The optimum diagnostic strategy for patients with suspected GA remains uncertain. Published studies have suggested that the IgA anti-gliadin antibody is more specific than the IgG anti-gliadin antibody test [92‐97], but Hadjivassiliou et al. [89] reported that IgG anti-gliadin antibody is a better marker of gluten ataxia, because of its high sensitivity. Since the level of anti-gliadin antibodies is 5–12% in the general population, some clinicians believe that anti-gliadin antibody tests cannot be used for the diagnosis of GA [89]. Studies of GA patients have shown that anti-tTG antibodies are present in the brain, supporting a possible pathogenic role in the etiology of the condition. If CD serology is positive, then obtaining intestinal biopsies to look for evidence of CD should be considered [96]. Magnetic resonance imaging (MRI) can also be used for GA diagnosis. MRI studies of GA patients show the presence of moderate cerebellar atrophy in up to 60% of patients [87].

GA patients should be treated with a strict GFD. In addition, studies have shown that immunotherapy (steroid, intravenous immunoglobulins (IVIG)) can be an effective treatment for such patients [98]. As GA is a progressive disorder in which neurons and Purkinje cells are destroyed over time, the response to treatment depends on the time interval between the onset of the GA and treatment [99, 100].

Wheat allergy (WA) is one of the most common food allergies (since wheat provides 70% of dietary proteins), and should be considered as a serious health problem worldwide [101]. In contrast to CD, different wheat components such as water-soluble (albumin and globulin) and water-insoluble (glutenin and gliadin) proteins contribute to the development of wheat allergy [102‐104]. WA is more common in pediatric practice than adult medicine (the mean age of onset for WA is 5.5 years (3–16 years)) and the global prevalence of WA is reported at 0.5–1% [105‐108]. Wheat allergy as a subgroup of food hypersensitivity is categorized into two groups; IgE-Mediated and non-IgE-Mediated WA [108‐110].

Non-celiac gluten sensitivity (NCGS) refers to a reaction to gluten leading to intestinal and extra-intestinal manifestations that are not mediated by an allergic or immunologic response [132, 133]. The terms gluten sensitivity, gluten hypersensitivity, and non-celiac gluten intolerance also refer to this condition [133‐135]. NCGS is more common in adult females (F/M 6:1) and its prevalence is estimated at 0.6–13% of the general population [103, 136, 137].

NCGS can cause a wide variety of symptoms including abdominal pain, diarrhea, weight loss, headache, fatigue, malaise, muscle pain, recurrent oral ulceration, and depression [138‐140]. Recent studies proposed that besides gluten, other components of wheat, such as poorly fermentable, poorly absorbed, short-chain carbohydrates, and wheat amylase-trypsin inhibitors may contribute to the development of NCGS [136, 137, 141‐143]. NCGS symptoms occur in a few hours or days after gluten ingestion, that resolve on a GFD and relapse after a gluten challenge [144, 145]. A significant proportion of NCGS patients are self-diagnosed and start a GFD without medial consultation [146, 147].

Currently, the diagnosis of NCGS is dependent upon a clinical assessment of symptoms and exclusion of WA and CD – the “Salerno Experts’ Criteria” (the probability of CD and WA must be ruled out) in patients on diets that contain gluten [146]. In other words, NCGS should be considered in patients with negative WA and CD tests (the small intestine of NCGS patients is usually normal, and serum tTG-antibodies and EMA-antibodies are negative) [148, 149]. HLA-DQ2/DQ8 haplotypes are found in approximately 50% of NCGS patients. These haplotypes are not required for the condition to develop, and HLA typing cannot be used to confirm or exclude a diagnosis of NCGS [1, 138]. As there is no specific NCGS diagnostic biomarker and test, a therapeutic trial of a GFD may be considered. Following a diagnosis of NGCS, patients may be asked to undergo a double-blind, placebo-controlled (DBPC) gluten challenge [140, 150]. Patients diagnosed with NCGS showed a significant worsening of symptoms after gluten consumption [151]. However, this exclusion diagnostic protocol remains cumbersome and is not easy to perform in daily clinical practice. There is an increasing need for having a clear diagnostic process that will lead to a definitive diagnosis in suspected NCGS individuals [144, 152]. To date, in various studies, the attempt has been made to identify the predictive pattern of NCGS. In the absence of a definitive test to diagnose NCGS, studies continue to focus on serum markers of wheat intolerance. Studies have shown that IgG anti-gliadin antibodies (IgG-AGA) are present in approximately 56% of NCGS cases and over 80% of CD cases, compared to 2–8% of the general population [146, 153, 154]. Duodenal biopsies from patients with NCGS are typically reported as normal, but detailed analysis suggests a mild increase in intraepithelial lymphocytes, increased expression of claudin 4, Toll-like receptor 2 (TLR2) and interferon-gamma (IFNγ) and increased goblet cell number in this condition [140, 155].

There are conflicting data on intestinal permeability in NCGS [156, 157]. In the primary study conducted by Sapone et al. [158] in 2011, gut permeability of CD and NCGS patients was determined using the urine lactulose/mannitol (LA/MA) test. The results of this study demonstrated a significantly lower small intestinal permeability in NCGS compared to CD patients and controls. The study also reported a high expression of claudin-4 mRNA, in duodenal biopsies from NCGS patients compared to the other patient groups, consistent with the finding of decreased intestinal permeability [158]. Hollon et al. [159] in their ex vivo study in 2015, evaluated changes in transepithelial electrical resistance (TEER) of tissue biopsies from active CD patients (ACD), CD patients in remission, patients with non-celiac gluten sensitivity and controls exposed to pepsin-trypsin digested gliadin (PT-gliadin). The results of their study demonstrated that gliadin exposure reduced TEER and increased intestinal permeability in all patient groups compared to controls. These results suggest that NCGS patients have abnormal intestinal permeability [159]. The study conducted by Uhde et al. [155] on individuals with wheat sensitivity in the absence of CD in 2016, also demonstrated the presence of enterocyte injury, increased intestinal permeability, and microbial products translocation in these patients. This study reported that the increased intestinal permeability was accompanied by an increase in the serum levels of the different biomarkers such as soluble CD14, lipopolysaccharide-binding protein (LBP), and fatty acid-binding protein 2 (FABP2). They proposed that these biomarkers, if validated in subsequent analysis, could be useful as possible NCGS diagnostic tools [155]. Furthermore, Barbaro et al. [160] in 2015 showed significantly high zonulin serum levels, a potential biomarker for monitoring changes in intestinal permeability, following gluten exposure. They noted that zonulin can contribute to NCGS pathophysiology and has a correlation with symptoms in NCGS patients [160]. In the past, patients with NCGS were frequently misdiagnosed. These patients were often believed to have an underlying psychiatric disorder [139, 161]. Unlike CD, NCGS is not considered a high risk for long-term complications or nutrient deficiencies. There is no need to screen the relatives of patients with NCGS [143, 162].

The only way to treat the NCGS is GFD adherence [163].