An update on coeliac disease from the NHS England National Centre for Refractory Coeliac Disease

Key points

• Coeliac disease is common, with a prevalence of around 1%; many patients are undiagnosed and can present with subtle symptoms to a wide variety of medical specialties. Diagnosis can be made by positive serology (IgA-tTG/IgAEMA) in conjunction with confirmatory histology (Marsh 3a or above), but during the COVID-19 pandemic, a serologybased diagnosis can be made in select individuals.

• Management is with a lifelong strict gluten free diet, requiring the input of a specialist dietitian.

• Persisting symptoms on a gluten free diet requires reevaluation, occurring in up to a third of individuals, with repeat duodenal biopsies being the only effective method to assess for ongoing villous atrophy.

• Refractory coeliac disease affects a small subset of patients with coeliac disease and may require treatment with steroids and immunomodulators.

Clinical presentation

Despite an increase in incidence of CD, many remain undiagnosed.³ Diagnosis can be delayed, and symptoms may be present for a mean of 11 years before diagnosis.⁴ This delay in diagnosis is associated with a prolonged and significant decrease in quality of life.

Historically, it was thought that CD solely presented with signs and symptoms of malabsorption (such as diarrhoea, weight loss and steatorrhoea).⁵ While individuals may present like this, known as classical CD, it is now known that individuals can present in a multitude of ways.⁵ Individuals may present without symptoms and signs of malabsorption, known as non-classical CD.⁵ These can be gastrointestinal (GI; such as constipation and abdominal pain) but can also can be extra-intestinal (such as osteoporosis, dermatitis herpetiformis, neurological manifestations, infertility, deranged liver function tests and thyroid dysfunction).⁵

Presenting features may be subtle (such as fatigue, recurrent or severe mouth ulcers, or haematinic deficiencies).⁶ A history of autoimmune disease should be sought (such as type 1 diabetes and thyroid dysfunction), as well as a family history of CD, with the prevalence being around eight times higher in first degree relatives.^3,6 In addition, individuals presenting with irritable bowel syndrome (IBS) should be tested for CD.⁶ CD should also be considered in some chromosomal disorders such as Down’s syndrome and Turner’s syndrome.⁷

Diagnosis

Individuals with suspected CD must be on a gluten containing diet prior to both serological and histological testing. They should be advised to consume at least 10 g of gluten daily (around four slices of bread) for 6 weeks prior to testing. This may be challenging for those with significant ongoing symptoms with gluten; in this scenario a shorter gluten challenge for 2 weeks may be considered, although the detection of villous atrophy is variable, reported between 26% and 68% using this approach.^8,9

For serological testing, immunoglobulin A (IgA) tissue transglutaminase (tTG) testing should be performed in the first instance.⁶ IgA-tTG has a high sensitivity and specificity for the diagnosis of CD, reported as mean of 94% and 97%, respectively.¹⁰ Alternatively, IgA endomysial antibody (EMA) testing can also be used, and is recommended when IgA-tTG is weakly positive.⁶ While IgA-EMA can also be used in view of its high sensitivity and specificity (reported at >86% and around 100%, respectively), it has limitations, being a qualitative test open to inter-observer variability as well as being labour intensive to analyse.¹⁰ When assessing serology, IgA levels should be checked, owing to the fact of the high prevalence of IgA deficiency for individuals with CD, reported at 2.6%.³ In individuals with IgA deficiency, IgG-tTG serology should be performed.³ A small minority of individuals (3–5%) have seronegative CD but positive histology. These individuals respond to a gluten free diet (GFD), and may present at an older age, with classical symptoms being more common.¹¹ However, it is worth noting there are several other causes for seronegative villous atrophy (such as due to non-steroidal anti-inflammatory drugs, Helicobacter pylori, common variable immunodeficiency, tuberculosis and autoimmune enteropathy).^3,11

If an individual has positive coeliac serology, a diagnosis of CD is confirmed histologically by taking duodenal biopsies from the duodenal bulb and distal duodenum, with four biopsies doubling the diagnostic rate.³ Histological features of CD include an increase in intraepithelial lymphocytes, crypt hyperplasia and villous atrophy (Fig 1).¹² While duodenal biopsies remain the cornerstone to confirm diagnosis in individuals with adult CD, the European Society for Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) guidelines suggest a diagnosis of CD can be made in selected children and adolescents with high tTG titres (>10× upper limit of normal (ULN)), compatible human leukocyte (HLA) typing and positive IgA-EMA serology, without the requirement for a duodenal biopsy.¹³ This approach is now being evaluated in the adult population, with a recent large multicentre study demonstrating that a high tTG titre (>10× ULN) had a strong positive predictive value for identifying CD, reported between 95.2% and 100%.¹⁴ In light of the COVID-19 pandemic, current interim British Society of Gastroenterology guidelines suggest a no biopsy approach for adults under the age of 55 years with high tTG titres (>10× ULN) and positive IgA-EMA provided they have no alarm symptoms.¹⁵ A no biopsy approach in selected groups of adults maybe the future for diagnosis globally, but hurdles remain, including the challenge of tTG assay standardisation.³

• Download figure
• Open in new tab
• Download powerpoint

Fig 1.

Modified Marsh criteria for classification of coeliac disease.¹⁷ Stage 3 (lower row) denotes histology consistent with diagnosis of coeliac disease. a) Stage 0: normal. b) Stage 1: increase in intraepithelial lymphocytes. c) Stage 2: increase in intraepithelial lymphocytes and crypt hyperplasia. d) Stage 3a: increase in intraepithelial lymphocytes, crypt hyperplasia and partial villous atrophy. e) Stage 3b: increase in intraepithelial lymphocytes, crypt hyperplasia and subtotal villous atrophy. f) Stage 3c: increase in intraepithelial lymphocytes, crypt hyperplasia and total villous atrophy.

The main genetic factors predisposing to CD lie in the HLA regions, with the HLA class II genes HLA-DQA1 and HLA-DQB1 being key.⁷ HLA typing is not used routinely in the diagnosis of CD, but may have utility in its exclusion, such as in individuals self-treated on a GFD who are unwilling to undergo a gluten challenge to have serological testing. HLA typing has a high negative predictive value for CD of greater than 99% but up to 40% of the general population may have a positive result, limiting its general use for diagnosis.¹³

Management and monitoring

Once a diagnosis of CD is confirmed, physicians should strongly recommend a GFD and support the individual to understand the reasons for this. This involves the strict dietary exclusion of wheat, rye and barley, ideally under the guidance of a specialist dietitian.³ Individuals may be able to consume pure oats in small amounts, although gluten cross contamination is possible.³

Current National Institute for Health and Care Excellence (NICE) guidelines suggest annual follow-up for individuals with CD, which is commonly performed in primary care, although follow-up remains variable.⁶ Individuals appear to be keen to be followed up by a dietitian with a physician being available if required. At review, symptoms, osteoporotic risk, micronutrient deficiencies (iron, folate, B12 and vitamin D) and dietary adherence should be assessed.^3,6 Some individuals with CD may have splenic hypofunction, with pneumococcal vaccination recommended for all.¹⁶

Up to a third of individuals with CD may have persisting symptoms despite being on a GFD, known as non-responsive CD (NRCD).¹⁷ The commonest cause is ongoing gluten exposure, which may be inadvertent or intentional.¹⁷ Adherence to a GFD can be challenging, with reported adherence in individuals with CD being between 42 and 91%.³ Currently, there are no effective non-invasive markers of GFD adherence, with point-of-care testing, dietary adherence questionnaires and serology having a poor sensitivity for the detection of villous atrophy.³ In view of this, repeat duodenal biopsies remains the gold standard for assessment.³ This assessment remains crucial in individuals with persisting symptoms, as mucosal healing is important to reduce the long-term complications in CD (such as osteoporosis and malignancy).³ Alternative diagnosis for persisting symptoms should also be considered (such as IBS, pancreatic exocrine insufficiency, microscopic colitis, lactose/fructose intolerance and thyroid dysfunction; Fig 2).¹⁷

• Download figure
• Open in new tab
• Download powerpoint

Fig 2.

Diagnosis for persisting symptoms.

A small minority of individuals with CD may have persisting villous atrophy despite strict adherence to a GFD. If the duration of GFD management has been greater than 12 months, and other diagnosis have been excluded, a diagnosis of refractory CD (RCD) may be made.³ The prevalence of RCD is rare, reported in less than 4% of individuals with CD.³ Techniques such as flow cytometry on histological samples can be used to distinguish between type 1 RCD (RCD 1) and type 2 RCD (RCD 2), with RCD 2 having a poor prognosis of 50% at 5 years in comparison with 80–100% at 5 years in RCD 1.³ Ideally, these individuals should be managed at specialist centres, where therapies such as corticosteroids and immunosuppressants may be employed.^3,17 It is worth noting that these therapies do not prevent the development of complications such as enteropathy-associated T-cell lymphoma, but may delay its onset.¹⁷ NHS England have selected Sheffield as the national centre for RCD (lead centre contact: david.sanders1{at}nhs.net), with Cambridge as the member centre.

Future developments

Currently, adherence to a GFD remains challenging. In view of this, portable gluten detection devices are now commercially available to assist individuals with inadvertent gluten exposure.¹⁸ However, with these devices being able to detect levels of gluten less than 20 parts per million, there are potential limitations with this approach of unknown clinical significance.¹⁸ As a result, this approach may lead to an excessive obsession with healthy eating (known as orthorexia nervosa), with further research required in this area to assess the utility of these devices.³

A promising area for the assessment of GFD adherence includes the use of gluten immunogenic peptides (GIPs), either via urine or faeces.³ It has been suggested that this approach may be a useful way to indirectly assess mucosal healing, although further research is required.¹⁹ In addition, GIPs are only transiently present for a few days post-gluten ingestion, which may limit its use if individuals alter their behaviour prior to testing.¹⁹

While the GFD remains the cornerstone for the management of CD, there has been a search for novel medical treatments for CD, such as zonulin inhibitors and vaccination.²⁰ However, most trials are currently within the pre-clinical phase, with few in phase II or III, and it is possible that medical treatments may play a useful role in the future as an adjunct to a GFD rather than replacement.²⁰