Chris JJ Mulder
Department of Gastroenterology
Fellow in Gastroenterology
Department of Gastroenterology
In a small subgroup of coeliac disease patients, the clinical and histological abnormalities of the disease persist or recur after improvement, despite long-term strict adherence to a a gluten-free diet (GFD). This nonresponsiveness leaves a poorly understood syndrome, known as refractory coeliac disease (RCD). A specific definition of RCD is lacking in the literature; we define it as malabsorption in the presence of persisting or recurring severe inflammatory infiltration of the epithelium and lamina propia, hyperplasia of crypts and (partial) villous atrophy, despite strict adherence to a GFD.
Which patients are at risk for RCD?
Patients’ symptoms and signs of malabsorption usually improve within weeks after gluten withdrawal. The histological normalisation of the intestinal mucosa may take more time.(1,2)
In children, mucosal abnormalities are thought to normalise within 6–12 months. Recently we found that histological recovery in adults is incomplete in a substantial subgroup of patients.(3) After two years, only 60% of patients showed normalisation of villi, and after five years mucosal inflammation with at least partial villous atrophy was still present in 10% of our patients. This delayed histological recovery was irrespective of the clinical course in the majority of patients. The clinical relevance of this finding with respect to RCD is unclear. We speculate that RCD may appear in a subgroup of these patients with persisting histological abnormalities.
RCD: diagnosis and treatment
RCD is a rare disorder that mimics coeliac disease but is refractory to a GFD and is without initial evidence of overt lymphoma. A workup for RCD must be simple and reproducible.(4) Diagnosis is based on nonresponsiveness to GFD that is not due to a developing lymphoma, ulcerative jejunoileitis, collagenous sprue or other diseases.
Recently it was suggested that RCD and enteropathy- associated T-cell lymphoma (EATL) could be distinguished from gluten-responsive disease by intraepithelial lymphocyte (IEL) phenotyping and by polymerase chain reaction for T-cell receptor (TCR)-γ gene rearrangements carried out on DNA extracted from duodenal biopsies.(5–7) Although morphologically normal, IELs are phenotypically distinct from those in patients with active coeliac disease who subsequently responded to a GFD and healthy controls. Intracytoplasmic-positive CD3 receptors, an absence of CD4, CD8 or surface TCRs, and restricted rearrangement of the TCR-gamma gene are characteristic. Reduced expression of CD8 and TCR-β on IELs and detection of clonality of the TCR-gamma gene may be markers for the presence of an EATL. Reduced expression of T-cell antigens on IEL is also found in RCD without definite diagnosis and without clonality of the TCR-gamma gene.
HLA (human leukocyte antigen) phenotyping helps to identify noncoeliac disease patients – those with non-DQ2 or non-DQ8 phenotypes should be considered not to be affected by gluten-sensitive enteropathy.
Taking into consideration that adherence to a “GFD” is difficult in our society, we advise patients with RCD to stick to a diet that is as gluten-free as possible.(4)
Protein-losing enteropathy may be a problem in some patients. In general, prognosis is poor and treatment is complicated and disappointing. Parenteral nutrition allows correction of colloid osmotic pressure and restores deficits. After parenteral support, a GFD can be reintroduced alongside careful monitoring of intestinal function and histology.
Albumin substitution may play an important role in the treatment of RCD, although no data on albumin infusions in these patients are available. A decrease in albumin below 20g/l may contribute to the clinical deterioration seen in severe RCD. We start to give weekly or monthly albumin infusions (200ml; 20% human albumin), aimed at achieving a serum albumin level of 30g/l. Blood transfusions are given if haemoglobin decreases below 6.0mmol/l.
Iron and folic acid must be supplied, if necessary intravenously, since both micronutrients are normally absorbed at the epithelial surface of the proximal small intestine.
If osteopenia is present, vitamin D and calcium must be supplied and intravenous bisphosphonates should be considered. If dietary review reveals any gluten ingestion the diet should be corrected. Although avoidance of products containing gluten is straightforward, the complex properties of modern processed food mean that the ongoing advice of trained dieticians is required.
Immunosuppressive treatment in RCD
The immune mechanism involved in the pathogenesis of CD is believed to involve aberrations in mucosal and intraepithelial T-cell populations and locally increased production of cytokines such as interferon, interleukin-6 (IL-6) and tumour necrosis factor (TNF). In RCD, an unexplained sustained stimulation of T-cell cytotoxic activity seems to occur. Treatment aiming to moderate immunological activity is largely empirical and often ineffective, with steroids and immunosuppression being the principal therapeutic options.(4) An overview of current immunosuppression strategies in RCD is given.
Oral corticosteroids may result in some improvement.(8,9) Short-term corticosteroid therapy in coeliacs still consuming gluten has demonstrated clinical and mucosal responses in case reports. However, the use of steroid monotherapy remains controversial and increases the risk of osteopenia.
Most case studies report the resounding success of ciclosporin treatment in RCD, although one case report was negative about this treatment.(10–12) In our experience, we saw a response, including normalisation of villi, in less than 40% of patients.(13) Ciclosporin may be beneficial in a subgroup of RCD patients but should be used with caution owing to the adverse effects associated with longterm
Cytokines with inflammatory as well as regulatory activities are presumed to play a role in the initiation and perpetuation of autoimmune diseases. Recent data have shown that IL-10 downregulates both the secretion and transcription of IL-1β and TNFgamma in a dose-dependent manner in monocytes isolated from lamina propria mononuclear cells. However, a pilot study of IL-10 in RCD was disappointing.(14)
One case report of treatment with the anti-TNFα drug infliximab demonstrated a dramatic improvement in a female patient without abnormal aberrant T-cells.(15) The response was maintained for 18 months while on azathioprine therapy.
Azathioprine plus prednisolone
Data on the efficacy of this combination have not been published, except as a case report.(16) An initial dosage of prednisolone 40mg daily was tapered to a maintenance dose of 10mg daily after six weeks in combination with azathioprine 2mg/kg daily. Smallbowel biopsy was repeated after three months.(17) If possible, prednisolone is tapered to 2.5–5mg daily and biopsies are repeated after one year. The first results in RCD without aberrant T-cells are promising. Work from Argentina confirms this.(18) However, in RCD with intracytoplasmic CD(3)(+)-positive cells, all patients developed an EATL and died.(17)
As chemotherapy has not been extensively studied in RCD patients with aberrant cytoplasmic CD(3) (+) Tcells, we initiated a phase II study to evaluate the activity of cladribine (2-CdA) in RCD. 2-CdA seems effective in inducing clinical remission in about 70% of this subgroup of RCD (CJ Mulder; personal communication). However, the percentage of abnormal T-cells in the mucosa didn’t change in follow-up.
Increasing recognition of coeliac disease means the incidence of RCD is also rising. RCD without aberrant T-cells can be treated with a combination of prednisolone and azathioprine. RCD with aberrant T-cells such as cytoplasmic CD(3)(+) cells is in fact a premalignant condition, a window to EATL. More aggressive treatment with chemotherapeutic agents such as cladribine is indicated. Prospective multicentre studies are mandatory. Efforts of a European Working Group (Cellier, Daum, Mulder, et al) are promising to define the strategies of the years to come.
- Grefte JMM, et al. J Clin Pathol 1988;41:886-91.
- O’Mahony S, et al. Aliment Pharmacol Ther 1996;10:671-8.
- Wahab PJ, et al. Am J Clin Pathol 2002;118:459-63.
- Mulder CJJ, et al. Scand J GE 2000; Suppl 232:32-7.
- Bagdi E, et al. Blood 1999;94:260-4.
- Carbonnel F, et al. Blood 1998;92:3879-86.
- Cellier C, et al. Gastroenterology 1998:114:471-81.
- Wall AJ, et al. Gut 1970;11:7-14.
- Lloyds-Still JD, et al. J Pediatr 1972;81:1074-81.
- Longstreth GF. Ann Int Med 1993;119(10):1014-6.
- Eijsbouts AMM, et al. Eur J Gastroenterol Hepatol 1995:7(8):803-6.
- Gheorghe L, et al. Hepato-gastroenterology 1997;44:1342-5.
- Wahab PJ, et al. Aliment Pharmacol Ther 2000;14(6):767-74.
- Mulder CJJ, et al. Eur J Gastroenterol 2001;10:1183-9.
- Gillet HR, et al. Gastroenterology 2002;122(3):800-5.
- Hamilton JD, et al. Lancet 1976;i:1213-6.
- CJJ Mulder, et al. Gastroenterology 2001;120:A131.
- Maurino E , et al. Am J Gastroenterol 2002;97:2595-602.