U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Coeliac Disease: Recognition, Assessment and Management. London: National Institute for Health and Care Excellence (NICE); 2015 Sep. (NICE Guideline, No. 20.)

Cover of Coeliac Disease

Coeliac Disease: Recognition, Assessment and Management.

Show details

6Evidence for non-responsive and refractory coeliac disease

6.1. Causes of non-responsive coeliac disease

6.1.1. Review questions

  1. What are the potential causes of non-responsive coeliac disease (NRCD)?
  2. In people with confirmed refractory coeliac disease (RCD), what investigative procedures should be undertaken?

NRCD can manifest where persons with coeliac disease do not experience symptomatic or histological improvement after excluding gluten from their diet, or when the symptomatic respite afforded by gluten exclusion dissipates and people again become symptomatic. Understanding the potential causes for NRCD is highly important in order to identify and rectify the causes of increased symptoms and investigate the potential for refractory coeliac disease.

RCD is often deconstructed into separate sub-classifications and prognosis varies substantially depending on subtype. Subtype is determined by the presence (RCD II) or absence (RCD I) of aberrant small intestinal intraepithelial lymphocytes (IEL’s). People with RCD II have a proportion of aberrant intra epithelial lymphocytes (IEL’s) that lose surface expression of CD3 and CD8, which is frequently associated with the presence of a monoclonal IEL population. This subtype is associated with a significantly reduced survival expectancy compared to those with RCD I, and this is predominantly driven by the increased risk of enteropathy associated T-cell lymphoma (EATL). Survival rate of those who develop EATL is poor, with a 2 year survival estimated between 15% – 20%. Investigative procedure which allow for the sub-classification of RCD, and monitoring for development of lymphoma and other enteropathies is of great importance to the clinical management of people with this condition.

6.1.2. Methods

The aim of this review question was to determine:

  • The proportion of differing causes of persistent symptoms in people with a confirmed diagnosis of coeliac disease who have been advised to exclude gluten from their diet
  • The clinical utility of investigative tests in people with refractory coeliac disease

NRCD is defined as a continuation of symptoms and/or signs of coeliac disease (CD) despite reporting being on a gluten-free diet (GFD). There are no current consensus criteria for the precise definition of RCD. We have defined RCD in this review as the persistence or later development of severe villous atrophy in people with CD despite a strict gluten-free diet, where adherence/inadvertent gluten ingestion and potential concomitant conditions have also been ruled out.

Studies reviewed were only included if the population examined was people in whom a biopsy-confirmed diagnosis of coeliac disease had been made. Studies considered for question 6.1.1 (a) investigated people who were deemed to have non-responsive coeliac disease (NRCD), and the potential causes for this non-responsiveness. Studies considered for question 6.1.1 (b) focused on investigative procedures for the sub-classification and monitoring of people with confirmed RCD.

Cohort studies were considered to be the highest quality evidence available and are graded as high in a modified GRADE framework. The quality for each outcome could be downgraded due to risk of bias in terms of methods, indirectness in terms of population, tests and outcomes used, inconsistency between studies, or imprecision in terms of outcomes.

For full details of the review protocol please see Appendix C.

Included Studies

A systematic search was conducted (see Appendix C) which identified 1859 references. The references were screened on their titles and abstracts and full text papers of 39 references were obtained and reviewed against the inclusion and exclusion criteria in the review protocol (see Appendix C).

Twenty seven studies were excluded as they did not meet the eligibility criteria such as inappropriate study design (case-control studies with healthy or symptom-free uncomplicated CD, or other patient comparator groups), not a primary study (descriptive narrative, opinion, etc.). A detailed list of excluded studies and reasons for their exclusion is provided in Appendix F.

The 12 remaining published papers did meet the eligibility criteria and were included. Data were extracted into detailed evidence tables (see Appendix D).

6.1.3. Evidence review

6.1.3.1. Causes of nonresponsive coeliac disease

Four studies (Dewar et al., 2012; Leffler et al., 2007; Abdulkarim et al., 2002; Van Weyenberg et al., 2013) contributed data to this analysis. Samples sizes ranged from 48 to 113 adults with NRCD. The mean age of study participants ranged from 42 to 63 years. The study participants were on a gluten-free diet for between 6 months and 6 years.

6.1.3.2. Subtyping of RCD into RCD type I and RCD type II

Three studies (Daum et al., 2009; Arguelles-Grande et al., 2013; Malamut et al., 2009) contributed data to this analysis. Samples sizes ranged from 14 to 73 adults with confirmed RCD. The mean age of study participants ranged from 48 to 67 years. Only participants on a strict gluten-free diet were included in all studies at the time of testing.

6.1.3.3. Sensitivity and specificity of investigative procedures to detect enteropathy-associated–cell lymphoma (EATL) and ulcerative jejunitis (UJ)

One primary study with one publication (Daum et al., 2007), and two primary studies with two publications (Hadithi et al., 2006, 2007; Van Weyenberg et al., 2011, 2013) contributed to this analysis. Samples size ranged from 14 to 68 adults with a confirmed diagnosis of RCD. Mean age of participants ranged from between 48 to 63 years. Only participants on a strict gluten-free diet were included in all studies at the time of testing.

6.1.3.4. Cumulative survival at 5 years post diagnosis of RCD

Four studies (Daum et al., 2009; Malamut et al., 2009; Van Weyenberg et al., 2011; Arguelles Grande, 2013) contributed to this analysis. Sample sizes ranged from 12 to 67 participants with RCD type I and 6 to 43 participants with RCD type II. Mean age of participants ranged from 50 to 56 years.

6.1.3.5. Predictive factors of EATL development in patients with RCD

Two studies (Liu et al., 2009; Malamut et al., 2009) contributed data to this analysis. Sample sizes were 41 and 57 participants with RCD, respectively. Mean age of participants at diagnosis of RCD ranged from 48 to 63 years.

6.1.3.6. Predictive factors for clinical worsening in patients with RCD

One study (Arguelles Grande et al., 2012) contributed to this analysis. This study examined 73 participants with RCD, with a mean age of 56 years. The mean time since RCD diagnosis was 5 years.

6.1.4. Health economic evidence

An economic evaluations filter was applied to the search protocol for this research question with the aim of finding economic evaluations that explored potential causes of non-responsive coeliac disease or the cost effectiveness of investigations for individuals diagnosed with refractory coeliac disease.

The search identified 113 references. The references were screened on their titles and abstracts and none of the studies met the inclusion criteria.

No cost–utility analyses were found to address selection criteria.

6.1.5. Evidence statements

6.1.5.1. The potential causes of nonresponsive coeliac disease (NRCD)

High quality evidence from 3 studies of 148 participants with NRCD suggested that between 10% to 12% of patients presenting with NRCD could be accounted for by a misdiagnosis of coeliac disease.

High quality evidence from 4 studies of 165 adults with NRCD suggested that between 36% to 82% of patients presenting with NRCD could be accounted for by noncompliant or inadvertent gluten ingestion.

High quality evidence from 3 studies of 148 adults with NRCD suggested that between 6% to 11% of patients presenting with NRCD could be accounted for microscopic colitis.

High quality evidence from 3 studies of 148 adults with NRCD suggested that between 6% to 14% of patients presenting with NRCD could be accounted for by bacterial overgrowth.

High quality evidence from 3 studies of 148 adults with NRCD suggested that between 7% to 12% of patients presenting with NRCD could be accounted for by lactose intolerance.

High quality evidence from 2 studies of 117 adults with NRCD suggested that between 6% to 7% of patients presenting with NRCD could be accounted for by inflammatory colitis.

High quality evidence from 2 studies of 149 adults with NRCD suggested that between 2% to 12 % of patients presenting with NRCD could be accounted for by pancreatic insufficiency.

High quality evidence from 3 studies of 148 adults with NRCD suggested that between 9% to 18% of patients presenting with NRCD could be accounted for by true refractory coeliac disease (RCD).

6.1.5.2. Investigative procedures in patients with confirmed refractory coeliac disease: Change to clinical management

6.1.5.2.1. Aberrant T-cell receptor gene rearrangement (TCR) by polymerase chain reaction (PCR)

High quality evidence from 3 studies of146 adults with RCD showed that aberrant T-cell receptor gene rearrangement (TCR) assessed by polymerase chain reaction (PCR) has between 97% to 100% sensitivity and 100% specificity to discriminate RCD type II from RCD type I.

6.1.5.2.2. Immunohistochemistry to detect aberrant IEL immunophenotype

High quality evidence from 3 studies of 146 adults with RCD showed that immunohistochemistry to detect an aberrant CD3+ CD8− immunophenotype has between 56% to 100% sensitivity, and 100% specificity to discriminate RCD type II from RCD type I.

6.1.5.3. Investigative procedures in patients with confirmed refractory coeliac disease: Detection of lymphoma

6.1.5.3.1. Positron emission tomography with 2-deoxy-2-[fluorine-18] fluoro-D-glucose (F-18-FDG-PET)

High quality evidence from a study of 30 adults with RCD reported 18-F FDG PET to have a sensitivity of 100% and specificity of 90% (95% CI: 79 to 100) to detect the presence of EATL in people with RCD.

6.1.5.3.2. Computerised tomography (CT)

Low quality evidence from 2 studies of 37 adults with RCD reported abdominal CT to have a sensitivity of between 50% (95% CI: 36 to 69) and a specificity of 76% (95% CI: 36 to 100) to detect EATL in people with RCD.

6.1.5.3.3. Magnetic resonance (MR) enteroclysis

Low quality evidence from a single study of 28 adults with RCD and uncomplicated CD reported MR enteroclysis to have a sensitivity of 88% (95% CI: 47 to 99) and a specificity of 97% (95% CI: 87 to 99) to detect EATL in people with RCD.

6.1.5.3.4. Double balloon enteroscopy

High quality evidence from a single study of 35 adults with RCD reported double balloon enteroscopy to have a sensitivity and specificity of 100% to detect EATL in people with RCD.

6.1.5.3.5. Capsule endoscopy (CE)

Very low quality evidence from 2 studies of 48 adults with RCD reported capsule endoscopy to have a sensitivity of between 50% (95% CI: 19 to 100) and a specificity of 100% to detect EATL in people with RCD.

6.1.5.4. Health-related outcomes: Cumulative survival at 5 years post diagnosis

High quality evidence from 4 studies of 112 participants reported that people with a diagnosis of RCD type I have between 95% (95% CI: 76 to 100) cumulative survival rate at 5 years post diagnosis of RCD.

High quality evidence from 4 studies of 68 participants reported that participants with a diagnosis of RCD type II have a 53% (95% CI: 12 to 94) cumulative survival rate at 5 years post diagnosis of RCD.

6.1.5.5. Health-related outcomes: predictive factors for lymphoma

6.1.5.5.1. Aberrant immunophenotype

Moderate quality evidence from 2 studies of 98 participants with RCD reported that an aberrant immunophenotype has a significant predictive value (OR: 4.59; 95% CI: 0.51 to 20.7) for the development of lymphoma in people with RCD.

6.1.5.5.2. Age

Moderate quality evidence from studies of 98 participants with RCD reported age to have no significant predictive value (OR: 1.13; 95% CI: 0.9 to 1.7) for the development of lymphoma in people with RCD.

6.1.5.5.3. Ulcerative jejunitis

Moderate quality evidence from a single study of 57 people with RCD reported the presence of ulcerative jejunitis to have no significant predictive value (OR: 1.8; 95% CI: 0.7 to 4.7) for the development of lymphoma in people with RCD.

6.1.5.5.4. Gender

Moderate quality evidence from a single study of 41 people with RCD reported gender to have a no predictive value (OR: 2.17; 95% CI: 0.45 to 10.44) for the development of lymphoma in people with RCD.

6.1.5.5.5. Persistent monoclonality

Moderate quality evidence from a single study of 41 people with RCD reported the presence of persistent monoclonality to have significant predictive value (OR: 3.6; 95% CI: 0.6 to 21.6) for the development of lymphoma in people with RCD.

6.1.5.5.6. Persistent concurrent aberrant immunophenotype and monoclonality

Moderate quality evidence from a single study of 41 people with RCD reported the presence of persistent monoclonality to have significant predictive value (OR: 9; 95% CI: 0.51 to 48.75) for the development of lymphoma in people with RCD.

6.1.5.5.7. Persistent >80% CD3+ CD8− IEL’s

Moderate quality evidence from a single study of 41 people with RCD reported the presence of persistent >80% CD3+ CD8− IEL’s to have significant predictive value (OR: 21.33; 95% CI: 2.94 to 154.6) for the development of lymphoma in people with RCD.

6.1.5.5.8. Persistent concurrent >80% CD3+ CD8− IEL’s and monoclonality

Moderate quality evidence from a single study of 41 people with RCD reported the presence of persistent concurrent >80% CD3+ CD8− IEL’s and monoclonality to have significant predictive value (OR: 45.33; 95% CI: 4.05 to 506.86) for the development of lymphoma in people with RCD.

6.1.5.6. Health-related outcomes: predictive factors for clinical worsening in RCD

6.1.5.6.1. Age ≥ 50 years

Moderate quality evidence from a single study of 73 people with RCD reported an age of greater than or equal to 50 years to have no predictive value (OR: 1.55; 95% CI: 0.8 to 3.0) for clinical worsening in RCD.

6.1.5.6.2. Monoclonality

Moderate quality evidence from a single study of 73 people with RCD reported monoclonality to have significant predictive value (OR: 4.33; 95% CI: 1.7 to 10.98) for clinical worsening in RCD.

6.1.5.6.3. Severe villous atrophy

High quality evidence from a single study of 73 people with RCD reported the presence of severe villous atrophy to have no predictive value (OR: 1.54; 95% CI: 0.25 to 0.8) for clinical worsening in RCD.

6.1.5.6.4. Aberrant IEL immunophenotype

Moderate quality evidence from a single study of 73 people with RCD reported the presence of an aberrant immunophenotype to have significant predictive value (OR: 3.01; 95% CI: 1.5 to 6.01) for clinical worsening in RCD.

6.1.5.6.5. Non-EATL lymphoma

Moderate quality evidence from a single study of 73 people with RCD reported the presence of non-EATL lymphoma to have no predictive value (OR: 2:76; 95% CI: 0.8 to 9.19) clinical worsening in RCD.

6.1.5.6.6. Presence of focal proximal erythema on capsule endoscopy

Low quality evidence from a single study of 48 people with RCD reported the presence of focal proximal erythema on capsule endoscopy to have significant predictive value (OR: 6.7; 95% CI: 1.2 to 38.7) for worsening clinical outcome.

6.1.5.6.7. Absence of progression of capsule to distal intestina during capsule endoscopy

Low quality evidence from a single study of 48 people with RCD reported the absence of progression of the capsule to distal intestina during capsule endoscopy to have significant predictive value (OR: 16.5; 95% CI: 1.2 to 224.9) for worsening clinical outcome.

6.1.6. Evidence to recommendations

Relative value of different outcomesAccurate diagnosis of nonresponsive and refractory coeliac disease was of the highest importance to the GDG, as these two conditions have very important and different implications. Mortality is a primary outcome of concern for refractory disease. Identifying methods of subtyping of refractory disease into type I and type II was also highlighted as a highly important outcome of interest for the GDG, as those with RCD II have a poorer prognosis and require closer monitoring.
Identifying the cause of symptoms in nonresponsive CD was also highlighted as a key outcome of concern for the GDG in order to be able to address and rectify ongoing symptoms and improve the quality of life of patients.
Trade-off between benefits and harmsIdentifying the proportion of differing causes of NRCD
Non responsive coeliac disease was found to be most often associated with exposure to gluten. The group discussed potential reasons for this and considered that for those who were incidentally ingesting gluten, increased education about the GFD is needed. For this reason, the group thought it was important that people should be referred to a specialist dietitian to review diet and any potential consumption of gluten which may be causing ongoing problems. When gluten ingestion has been rules out, investigation for other potential causes highlighted in this review, such as IBS or microscopic colitis, should be undertaken.
Patients who are experiencing ongoing symptoms often have a poor quality of life, and the group discussed that it was important that further investigation was undergone as quickly as possible in order to identify any potential causes before considering diagnosis of true refractory coeliac disease. The GDG also recognised that true refractory coeliac disease was a condition only diagnosed in adults, whereas any children continuing to have symptoms are almost always inadvertently ingesting gluten or have a contributing comorbid condition.
Change to clinical management
The importance of establishing the certainty of diagnosis was discussed as potentially problematic. In the absence of accurate biopsy orientations, a re-biopsy may be required. In order for the biopsy to be accurate, a gluten-challenge may need to be undertaken. This is often distressing and problematic for people. It can be difficult for people to understand when they are told to make sure that all gluten is excluded from the diet and then recommended to resume eating gluten for the purposes of diagnostic testing. The group felt that the benefit of making certain the diagnosis in order to identify causes of ongoing problems would outweigh the potential discomfort of patients while they underwent a gluten challenge for the 6 week period recommended before investigations.
Patient outcome at follow up
The group raised the notion that in making a referral to a specialist centre to follow-up refractory CD patients would be particularly beneficial to both people with RCD and treating clinicians. This is because numbers of patients with confirmed RCD are so small that referral of these people to a specialist centre will allow the few centres to build up sufficient expertise in this area, and to ensure correct and appropriate immunology tests are done to determine the diagnosis and subtype. At present, very few centres will have this expertise.
Health-related quality of life
Health-related quality of life was particularly discussed in terms of the impact that ongoing symptoms and potential mortality and cancer risk has on patients with RCD. Investigative procedures which assess subtyping of RCD into type I and type II were discussed as particularly important in order to understand and communicate prognosis to patients. As RCD II is associated with an increased risk of EATL, investigative procedures for this were also highlighted as essential for monitoring and assessment. The utility of these procedures was discussed as problematic, as the evidence suggests, because these have low sensitivity and specificity to detect EATL. Harms of these procedures were also discussed in terms of their invasive nature, which could cause these patients, who are likely very unwell, further discomfort. The capsule endoscopy method in particular was discussed as potentially harmful to patients who do have EATL or another major intestinal obstruction due to the capsule not being able to pass through and thus requiring surgery for its removal. The group discussed the very poor prognosis of those diagnosed with EATL and agreed that the most important consideration for any patient was to diagnose or exclude EATL as soon as possible.
Economic considerationsNo economic evidence was found
Quality of evidenceOverall the evidence identified for causes of non-responsive coeliac disease was high. The evidence for investigations for refractory coeliac disease, however, varied from low to high quality. This is because populations that were not truly refractory were included in some of the studies, and because the confidence intervals for estimated sensitivity and specificity were often very wide, reflecting the lack of power to detect clinically meaningful results.
Other considerationsThe GDG discussed and agreed that currently there was no clear consensus over how non-responsive and refractory coeliac disease should be defined and diagnosed. The presence of symptoms is often cited as necessary for recognition; however it was recognised by the GDG that some people may not be symptomatic, such as those detected through case-finding or incidentally and thus would not show clinical remission. It was also recognised that clinical remission may not reflect histological remission (see 5.4). The group expressed concern that as such, refractoriness may be over diagnosed when the most likely cause is inadvertent exposure to gluten. The input of a specialist dietitian was highlighted as essential in order to make certain that gluten has been completely excluded from the diet.
The GDG acknowledged that because RCD populations are rare it is difficult to obtain adequate sample sizes to detect true sensitivity and specificity estimates of imaging procedures to detect malignancy.
The GDG also acknowledged that computerised tomography (CT) was a poor tool to use for detection of malignancy, but may be useful in conjunction with other imaging modalities. The limitations of capsule endoscopy were also discussed in terms of the presence of lesions such as ulcerative jejunitis and malignancy being contraindications for the capsule to successfully pass through the intestines. If the capsule becomes lodged in the intestine, an operation is required to remove it and this can be distressing for people with RCD. Some members of the group stated that in certain circumstances, capsule endoscopy may be of some, albeit limited, utility.

6.1.7. Recommendations & research recommendations

19.

Consider the following actions in people with coeliac disease who have persistent symptoms despite advice to exclude gluten from their diet:

  • review the certainty of the original diagnosis
  • refer the person to a specialist dietitian to investigate continued exposure to gluten
  • investigate potential complications or coexisting conditions that may be causing persistent symptoms, such as irritable bowel syndrome, lactose intolerance, bacterial overgrowth, microscopic colitis or inflammatory colitis.
20.

Diagnose refractory coeliac disease if the original diagnosis of coeliac disease has been confirmed, and exposure to gluten and any coexisting conditions have been excluded as the cause of continuing symptoms.

21.

Refer people with refractory coeliac disease to a specialist centre for further investigation

6.2. Pharmacological interventions

6.2.1. Review question

What is the effectiveness of pharmacological treatments for people with refractory coeliac disease?

There is currently very little known about the optimal pharmacological treatment strategy for patients with refractory coeliac disease. Current clinical practice is to treat with steroids such as prednisone and monitor clinical response, however further information and guidance on this and other potential treatment strategies is needed in order to ensure that patients with this condition receive the best pharmacological management.

6.2.2. Methods

The aim of this review was to establish the effectiveness of pharmacological interventions for the symptoms of refractory coeliac disease.

A systematic search was conducted which recovered a total of 928 results (after duplicates were removed). Results were screened on title and abstract, and a total of 68 full-text papers were ordered to be assessed for eligibility against the inclusion and exclusion criteria specified in the review protocol. Two additional studies were identified from cross-referencing narrative reviews. These were not retrieved from our search because the abstract and title did not report the drug used in the studies.

Of the 71 full-text articles obtained, 57 were excluded for reasons specified in Appendix F. Overall, 14 small case series were selected for inclusion, of which only 4 studies included more than 30 people with refractory coeliac disease (see summary table X). Only 6 studies were prospective, while the remaining 8 were retrospective reports of people treated at particular institutions for refractory coeliac disease.

Among the 14 studies included, there was a potential for overlap of participants reported in 3 pairs of the studies which were included (see footnotes on tables below). The GDG acknowledged that this was likely due to the rarity of this condition. The later publications included longer follow-up of participants included in earlier studies but tended to be retrospective and report on the outcomes of all people treated at a centre with a larger number of pharmacological treatments or combinations of these treatments. These later studies also tended to report some outcomes across all people treated at a centre, regardless of which drug or combinations of drugs they received.

It was not possible to pool results due to the heterogeneous nature of the included studies, owing to factors such as small differences in definitions applied to refractory coeliac disease and differences in the reporting of outcomes (such as different definitions of clinical, immunological and histological response),. A summary of the results is included in table X below.

Randomised controlled trials are the optimal study design to answer this review question. However, due to the rarity of true refractory coeliac disease, the GDG was aware of only case series in this field with very small numbers of participants per study. As such, a modified GRADE assessment of individual studies was applied for this question.

The outcomes specified in the review protocol were extracted, where available, and included:

  • Resolution of gastrointestinal and non-gastrointestinal symptoms
  • Complications of coeliac disease
  • Adverse effects
  • Health-related quality of life
  • Impact on carers
  • Serological response
  • Histological response

The reporting of these outcomes varied across the included studies and no study reported on the impact of pharmacological treatments on health-related quality of life or on carers. Full evidence tables are found in Appendix D.

6.2.3. Evidence review

6.2.3.1. Immunosuppressants (azathioprine, cyclosporine, cladribine, tioguanine, methotrexate)

Six small case series in up to 22 adults examined the effectiveness of immunosuppressive therapy to treat the symptoms of refractory coeliac disease in a group of people with biopsy-confirmed coeliac disease and a diagnosis of refractory coeliac disease.

6.2.3.2. Corticosteroids (prednisone, prednisolone, budesonide)

Six small case series in up to 22 adults examined the effectiveness of corticosteroids to treat the symptoms of refractory coeliac disease in a group of people with biopsy-confirmed coeliac disease and a diagnosis of refractory coeliac disease.

6.2.3.3. Aminosalicylates (mesalamine/mesalazine)

One very small case series in four adults examined the effectiveness of mesalamine to treat the symptoms of refractory coeliac disease in a group of people with biopsy-confirmed coeliac disease and a diagnosis of refractory coeliac disease.

6.2.3.4. Combinations (corticosteroids + immunosupressants or multiple immunosupressants)

Five small case series in up to18 adults examined the effectiveness of corticosteroids and immunosupressants in combination to treat the symptoms of refractory coeliac disease in a group of people with biopsy-confirmed coeliac disease and a diagnosis of refractory coeliac disease.

One very small case series in 6 adults examined the effectiveness of both prednisone and mesalamine in combination to treat the symptoms of refractory coeliac disease in a group of people with biopsy-confirmed coeliac disease and a diagnosis of refractory coeliac disease. One very small case series in 10 adults examined the effectiveness of cladribine plus pre-treatment with azathioprine or prednisone to treat the symptoms of refractory coeliac disease in a group of people with biopsy-confirmed coeliac disease and a diagnosis of refractory coeliac disease.

6.2.3.5. Cytokine modulators (unspecified anti-TNF α)

One very small case series in 4 adults examined the effectiveness of anti-TNF to treat the symptoms of refractory coeliac disease in a group of people with biopsy-confirmed coeliac disease and a diagnosis of refractory coeliac disease.

6.2.4. Health economic evidence

An economic evaluations filter was applied to the search protocol for this research question with the aim of finding economic evaluations that explored the cost effectiveness of pharmacological therapy for refractory coeliac disease.

The search identified 63 references. The references were screened on their titles and abstracts and none of the studies met the inclusion criteria.

No cost–utility analyses were found to address selection criteria.

6.2.5. Evidence statements

Overall, the studies included were very low quality and contained no control group. Additionally, it is likely that there was an overlap of people reported in some of the included studies.

6.2.5.1. Immunosuppressants (azathioprine, cyclosporine, cladribine, tioguanine, methotrexate)

Very low quality evidence from 6 small case series (n=22 or less) reported that immunosuppressants may reduce symptoms and villous atrophy in some people with refractory coeliac disease, but also reported some adverse effects of the treatment.

6.2.5.2. Corticosteroids (prednisone, prednisolone, budesonide)

Very low quality evidence from 6 small case series (n=47 or less) reported that corticosteroids alone may reduce symptoms and villous atrophy in some people with refractory coeliac disease.

6.2.5.3. Aminosalicylates (mesalamine/mesalazine)

Very low quality evidence from 1 very small case series (n=4) reported that mesalamine may reduce symptoms in some people with refractory coeliac disease.

6.2.5.4. Combinations (corticosteroids + immunosupressants or multiple immunosupressants)

Very low quality evidence from 5 small case series (n=18 or less) reported that corticosteroids and immunosupressants in combination may reduce some symptoms and villous atrophy in some people with refractory coeliac disease.

Very low quality evidence from 1 very small case series (n=6) reported that both prednisone and mesalamine may reduce some symptoms in some people with refractory coeliac disease.

Very low quality evidence from 1 very small case series (n=10) reported that cladribine plus pre-treatment with azathioprine or prednisone reduces symptoms and villous atrophy in a small proportion of people with refractory coeliac disease.

6.2.5.5. Cytokine modulators (unspecified anti-TNF α)

Very low quality evidence from 1 very small case series (n=4) reported that anti-TNF α may reduce symptoms in some people with refractory coeliac disease.

6.2.6. Evidence to recommendations

Relative value of different outcomesThe GDG felt that all outcomes of interest for this question were equally valuable and important to address.
Trade-off between benefits and harmsThe GDG was concerned with the state of the evidence on the pharmacological treatment for refractory coeliac disease. The evidence was of very low quality (see ‘quality of evidence’ below) and on very small numbers of participants.
As a result of the limited and low quality evidence, the GDG did not feel it was possible to make recommendations about treatment based on the evidence. They were also concerned that many clinicians are not experienced in dealing with people with this rare condition and that other factors such as nutritional management may play a role Consequently, the GDG felt that advice should be sought from specialists with experience in managing people with this rare condition.
Despite feeling that they could not make recommendations about pharmacological management based on the evidence, the GDG was cognisant that people with refractory coeliac disease are typically in very poor condition and felt it was important that these people received some treatment, even while waiting for advice from a specialist. However, they found this difficult because of the very poor evidence to support the use of any pharmacological treatment for refractory coeliac disease so made this decision based on their clinical expertise.
While the GDG was concerned about encouraging the use of drugs for which there is little evidence, they felt that because there was not enough evidence to support a change in clinical practice, they should recommend the use of prednisone or prednisolone. However, they expressed concerns about the long-term effects of these drugs so recommended that they only be used while clinicians are awaiting advice from a specialist. The GDG felt that people with suspected refractory coeliac disease should not be started on prednisolone or prednisone without the expectation of an early review. Due to the poor clinical condition of these people, the GDG felt that it was reasonable to use these drugs for short term management given that these people would be under close and frequent clinical review.
Furthermore, regarding the use of immunosupressants, the group felt that these drugs should not be prescribed without advice and follow-up of a specialist because prescribing these drugs requires experience and specific expertise, particularly with the associated risk of adverse events.
The GDG felt mesalazine was promising as an option and that there is less potential for harm related to this drug, but there is very little evidence to make recommendations about its use. They felt that further research into its use should be supported.
They also felt that budesonide should not be prescribed routinely as it is not commonly used and there is little evidence on its use.
Overall, the GDG felt that there is a need for high-quality evidence to inform the management of patients with refractory coeliac disease.
Economic considerationsNo economic evidence on pharmacological management of coeliac disease was found.
Quality of evidenceThe available evidence on the pharmacological management of refractory coeliac disease is observational and of very low quality. The evidence is based on case series with no control or case reports (though the later were excluded from the guideline) and are, therefore, likely to be subject to bias. In addition, the included case series are predominantly retrospective studies where it was unclear why the participants were allocated to different drugs or different combinations of drugs. Furthermore, for most studies, it is unclear if participants were recruited consecutively so reporting bias is a possibility. Many of the studies were retrospective reports of all people with refractory coeliac disease treated at specific centres and the focus of the study was not specifically on the efficacy of specific treatments rather the study reported some outcomes for all participants regardless of treatment. Consequently, this made it difficult to attribute the affect that each individual treatment had on participant outcomes.
Other considerationsThe GDG was aware that prednisolone is the only drug that is licensed for coeliac disease unresponsive to gluten withdrawal.

6.2.7. Recommendations & research recommendation

22.

Consider prednisolone for the initial management of the symptoms of refractory coeliac disease in adults while waiting for specialist advice.

6.3. Nutritional interventions

6.3.1. Review question

What is the effectiveness of nutritional management or nutritional support for people with refractory coeliac disease?

It is currently unknown whether any specific dietary interventions beyond the gluten-free diet can alleviate symptoms of refractory coeliac disease. This chapter sought to investigate the clinical efficacy of any such dietary interventions for this population.

6.3.2. Methods

The aim of this question was to determine whether people with refractory coeliac disease would benefit from additional nutritional support or nutritional management.

A systematic search was conducted which identified 2161 references studies. After removing duplicates the references were screened on their titles and abstracts and 38 references were obtained and reviewed against the inclusion and exclusion criteria (for the full review protocol please see Appendix C).

All 38 studies were excluded as they did not meet the inclusion criteria, such as, inappropriate study population (e.g. participants did not have refractory coeliac disease), the intervention wasn’t one of interest or not primary research study (e.g. expert opinion). A list of excluded studies and reasons for their exclusion is provided in Appendix F.

6.3.3. Evidence review

No evidence that met the inclusion and exclusion criteria for this question was found.

6.3.4. Health economic evidence

An economic evaluations filter was applied to the search protocol for this research question with the aim of finding economic evaluations that explored the cost-effectiveness of nutritional support or nutritional management of people with refractory coeliac disease.

The search identified 173 references. The references were screened on their titles and abstracts and none of the studies met the inclusion criteria.

No cost–utility analyses were found to address selection criteria.

6.3.5. Evidence statements

No evidence that met the inclusion and exclusion criteria for this question was found.

6.3.6. Evidence to recommendations

Relative value of different outcomesNo evidence was included to answer this review question. The GDG based its decisions on the knowledge and experience of the group.
Trade-off between benefits and harmsThe GDG did not feel able to make a recommendation on this topic. True refractory coeliac disease is quite rare (as most people are found to have either been still being exposed to gluten, or are found to have another condition), therefore they could not be sure that additional nutritional support or nutritional management would be of benefit. The GDG also noted that as stipulated in section 5.4 (Monitoring of people with CD), all people with coeliac disease, regardless of whether they are refractory to treatment, should be offered access to specialist dietetic advice and that as such these patients should at least be getting some dietetic support.
Economic considerationsNo health economic evidence was found and this question was not prioritised for de novo modelling.
Quality of evidenceNo evidence was available.

6.3.7. Recommendations & research recommendations

No recommendations were made for this review question.

6.4. Autologous stem-cell transplants

6.4.1. Review question

What is the effectiveness of autologous stem cell transplant for people with refractory coeliac disease?

Stem cell research in refractory coeliac disease is a very new and potentially promising area of clinical medicine. This chapter sought to investigate the clinical efficacy of native stem cell replacement to treat refractory coeliac disease.

6.4.2. Methods

The aim of this review question was to determine the efficacy of chemotherapy followed by transplantation of native stem cells (from the patient’s own body) for the treatment of confirmed refractory coeliac disease.

A systematic search was conducted (see Appendix C) which identified 1035 references studies. After removing duplicates the references were screened on their titles and abstracts and 22 references were obtained and reviewed against the inclusion and exclusion criteria (for full review protocol see Appendix C).

Twenty studies were excluded as they did not meet the eligibility criteria such as inappropriate study population (e.g. patients were already diagnosed with cancer) or not primary research study (e.g. expert opinions). A list of excluded studies and reasons for their exclusion is provided in Appendix F.

The overall quality of the 2 included published papers (from one study) was of very poor quality with very low confidence in the effect estimates. This is due to the methodological issues of the study design such as non-randomised and non-comparative, prone to selection bias, poor reporting on data and analysis, unclear recruitment strategies.

Moreover, the study only focused on a specific subgroup of people with refractory coeliac disease (RCD) (i.e. RCD type II, unresponsive to cladribine therapy), and hence, this limited inconclusive evidence cannot be generalised to the overall RCD patient population.

6.4.3. Evidence review

Two papers from 1 primary study (Tack et al., 2011; Al-Toma et al., 2011) with a total of 18 adults (age < 70 years) with a biopsy-confirmed diagnosis of CD and a further diagnosis of refractory coeliac disease were included in this review. All participants had no response to 1 or 2 prior courses of cladribine for 5 consecutive days. Of the 18 participants, only 13 went through autologous stem cell transplantation (ASCT); 2 had unsuccessful leukapharesis and the remaining 3 progressed to enteropathy associated T-cell lymphoma (EATL).

6.4.4. Health economic evidence

An economic evaluations filter was applied to the search protocol for this research question with the aim of finding economic evaluations that examined the cost effectiveness of autologous stem cell transplants in people with refractory coeliac disease.

The search identified 62 references. The references were screened on their titles and abstracts and none of the studies met the inclusion criteria.

No cost–utility or cost-effectiveness analyses were found to address selection criteria.

6.4.5. Evidence statements

There was very limited inconclusive evidence from one very low quality case study on the effectiveness of autologous stem cell transplant for people with refractory coeliac disease.

6.4.6. Evidence to recommendations

Relative value of different outcomesThe GDG discussed and agreed that a person’s quality of life, resolution of gastrointestinal and non-gastrointestinal symptoms, and mortality (as a complication of coeliac disease e.g. development of T-cell lymphoma) are the critical outcomes of concern. However, the only included study (2 published papers) only reported 1 of the 3 critical outcomes which was mortality/survival from a very small study sample (13 participants).
Trade-off between benefits and harmsDue the very limited and very low quality data reported in the only included study, and the fact that autologous stem cell transplant services are very rare in the UK, the GDG felt that they were unable have a meaningful discussion about the benefits and harms of autologous stem cell transplant for people with refractory coeliac disease.
Economic considerationsNo published cost utility or cost-effectiveness analyses were found from the systematic searches that addressed the inclusion criteria.
The GDG further discussed that currently there are only approximately 3 services across the whole UK that could deliver autologous stem cell transplant, and that the cost impact of establishing more autologous stem cell transplant services would be very high. The GDG had made an assumption that based on the current available very limited low quality evidence and the high cost of autologous stem cell transplant services, autologous stem cell transplant would not be cost effective for people with refractory coeliac disease.
Quality of evidenceAfter the discussion of the only included study (2 published papers), the GDG agreed that the included study was of very low quality case series and they have a lot of uncertainties about the effect estimates reported in the study because i) very small and narrow-defined sample size (only 13 patients with specific type II RCD) and a proportion of the participants already had enteropathy associated T-cell lymphoma; ii) the data reported was non-comparative; iii) the outcomes reported were unclear.
Hence, the GDG agreed that this limited inconclusive evidence cannot be generalised to the overall population of people with refractory coeliac disease, and therefore the GDG was unable to make any recommendation regarding autologous stem cell transplant for people with refractory coeliac disease.

6.4.7. Recommendations & research recommendations

No recommendations were made.

Research recommendation

6. What is the clinical and cost-effectiveness of autologous stem cell transplant for the treatment of people with refractory coeliac disease?
Why this is important

Refractory celiac disease (RCD) is often very severe and requires additional therapeutic intervention besides a gluten-free diet (GFD). RCD type 2 is particularly associated with poor prognosis despite conventional therapeutic interventions (GFD and/or pharmacological treatments) with 5-year survival rates of 40–58%. Poor prognosis is largely explained by the much more frequent progression to overt enteropathy-associated T-cell lymphoma (EATL) in patients with RCD type 2. Due to the poor prognosis of RCD, the aim of this research recommendation is to determine, other than GFD and pharmacological treatments, how effective it is to treat refractory coeliac disease with chemotherapy followed by transplantation of autologous stem cells.

Copyright © 2015 Internal Clinical Guidelines Team.
Bookshelf ID: NBK343372

Views

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...