A 75-year-old woman from an outside hospital was referred because of continued signs and symptoms of celiac disease (gluten-sensitive enteropathy) that persisted despite self-reported adherence to a gluten-free diet. The patient reported excessive gas, bowel distension, a 15-pound weight loss over the past few years, insomnia, and a rash over her lower extremities. The patient had required hospitalizations, intravenous fluids, and continuing therapy with corticosteroids for 6 months.
A diagnosis of celiac disease had been made 6 years previously, based on (a) typical gastrointestinal signs and symptoms with negative stool cultures and Clostridium difficile toxin assay, (b) positive serology for celiac disease, (c) unremarkable colonoscopy with normal random biopsy results, and (d) small-bowel biopsy results showing evidence of villous blunting with increased chronic inflammatory cells. At that time, the patient’s laboratory results included antigliadin antibody (AGA) IgG 0.8 AU (<10 AU), anti-AGA IgA 1.1 AU (<5 AU), anti–tissue transglutaminase (tTG) IgA 9.2 AU (<4 AU), and normal total IgA and IgA antiendomysial antibody (EMA) values. A computed tomographic scan was negative for lymphoma, and an upper gastrointestinal series and small-bowel follow-through barium x-ray were normal. Endoscopic biopsy results obtained during the previous 2 years showed continued villous atrophy with intraepithelial lymphocytes. Shortly before the patient’s referral, repeat biopsies showed villous blunting with increased chronic inflammation, findings confirmed by a gastrointestinal pathologist at our institution.
The patient, a pleasant, frail-looking, elderly woman in no acute distress, was retired and married with 2 adult children. She denied smoking and alcohol use and had no family history of celiac disease, liver disease, or colon cancer. Her medical history was remarkable for placement of a carotid artery stent 5 years earlier. Physical examination was unremarkable except for the presence of a maculopapular rash inconsistent with dermatitis herpetiformis and with dependent distribution over the lower legs.
The patient’s blood pressure was 133/59 mmHg, pulse 51 beats/min, temperature 36.5 °C, and weight 59.4 kg. Laboratory results since her referral included vitamin B12 245 ng/L [reference interval (RI), 251–911 ng/L], iron 370 ug/L (RI, 400–1450 μg/L), anti-tTG IgA 13 AU (RI, 0–20 AU), and 5′nucleotidase 22.1 U/L (RI, 4.0–11.5 U/L).
The patient met with a nutritionist and implemented recommended dietary changes to eliminate gluten. Her symptoms temporarily improved, with a return to normal bowel function, but after a short time her symptoms recurred. Results of further tests excluded conditions known to complicate or coexist with celiac disease, including bacterial overgrowth, microscopic colitis, and lactose intolerance. Because the patient’s symptoms were refractory to treatment and required prolonged, continued use of corticosteroid therapy, esophagogastroduodenoscopy with duodenal biopsies was performed, and formalin-fixed small-bowel biopsy tissue samples were sent to the molecular diagnostic laboratory for additional testing.
Celiac disease is a T-cell driven, multifactorial chronic inflammatory disorder of the small intestine characterized by mucosal inflammation, villous atrophy, and crypt hyperplasia; it has a prevalence of approximately 1% in the population. Among autoimmune diseases, celiac disease is unique in that an environmental trigger (gluten) and an autoantigen (tissue transglutaminase) have been identified (1).
The main dietary sources of gluten are wheat, rye, barley, and oats, but the gluten in oats has not been found to contribute to celiac disease. Gluten is broken down into smaller peptides by gastric acid and digestive enzymes. In the intestine, tTG converts glutamine to glutamic acid, thereby increasing the affinity of the binding of gluten peptides in the cleft of HLA class II molecules. The modified peptides are inappropriately recognized by helper T cells, perhaps because of molecular mimicry of microbial peptides The identity of these immunogenic peptides has been determined (2)(3)(4). Most individuals with celiac disease express HLA-DQ2 (95% of patients), and the others typically express HLA-DQ8. The presence of HLA-DQ2 and/or DQ8 alone, however, is not sufficient for disease, which is thought to involve other contributing factors such as additional genetic loci, stress, inflammation, and infection. The key treatment for celiac disease is lifelong adherence to a strict gluten-free diet.
diagnosis of celiac disease
The diagnosis of celiac disease is based on concordance of serological tests, small-bowel biopsy, and resolution of symptoms on withdrawal of gluten from the diet (5)(6). Serological testing for celiac disease includes anti-EMA IgA and anti-tTG IgA and IgG. Anti-AGA tests are no longer recommended because of their lower sensitivity and specificity [anti-AGA IgA sensitivity 75%–95%, specificity 80%–95% (7); anti-AGA IgG sensitivity 57%–100%, specificity 47%–94% (8)].
In the anti-EMA test, antibodies from the patient’s serum bind to connective tissue surrounding smooth muscle cells of either monkey esophagus or human umbilical cord and are detected by immunofluorescence. Identification of EMA as tTG led to the development of anti-tTG immunoassays. The first assays used guinea pig tTG; current assays use human tTG, either native (from erythrocytes) or recombinant. The anti-tTG test is easier to perform and more cost-effective than the anti-EMA test, and results are objective and quantitative. The test approaches 100% specificity and >90% sensitivity for celiac disease in a variety of clinical settings and populations (see Table 1⇓ ) (7). A test for antibodies against a 9 amino-acid deamidated gliadin peptide has recently become commercially available, but few studies of its diagnostic accuracy have been published.
The gold standard for diagnosis is histopathologic assessment of 4–8 mucosal biopsy specimens of the small bowel obtained while the patient is on a diet containing gluten.
causes of failure to respond to treatment
The case patient is among a small proportion of individuals with celiac disease whose illness does not respond to a gluten-free diet. The 3 main causes of treatment failure are (i) inadvertent or intentional failure to adhere to a strict gluten-free diet; (ii) other complicating or coexisting conditions such as small-bowel bacterial overgrowth, lactose intolerance, or microscopic colitis; and (iii) disease refractoriness to a gluten-free diet. The patient described here was compliant with the diet, and diagnostic testing revealed no evidence of lactose intolerance, microscopic colitis, small bowel bacterial overgrowth, ulcerative jejunitis, or lymphoma. These findings suggest a diagnosis of refractory celiac disease, which is characterized by persistent villous atrophy with an increase of intraepithelial lymphocytes in the small bowel while the patient is on a long-term gluten-free diet. In both responsive and refractory celiac disease, celiac antibodies usually decrease with dietary therapy (as observed in this case) and remain within reference intervals unless individuals are reexposed to gluten.
Two types of refractory celiac disease occur and are differentiated by the type of T-cell populations in the intestinal mucosa, which are polyclonal in type I disease and clonal in type II disease (9). Although the presence of this clonal T-cell population is termed “cryptic intraepithelial lymphoma,” this finding does not imply a diagnosis of a malignant process, although enteropathy-associated T-cell lymphoma develops in a subset of these patients.
tcrγ gene rearrangement
In the case patient, a T-cell receptor γ locus (TCRγ) gene rearrangement assay was the molecular test performed on the intestinal biopsy specimens to test for the presence of a clonal population of T cells.
Functional human TCRγ is encoded by the random rearrangement of 1 of 10 variable (V) segments and 1 of 5 joining (J) segments of the TCRγ gene. During T-cell maturation in the bone marrow, the (V) and (J) gene segments are randomly recombined to form a functional TCRγ chain.
The TCRγ chain gene segments are located on chromosome 7p14, and every T cell carries 2 alleles (paternal and maternal) of this locus. During T-cell development one or both alleles undergo rearrangement, so a clonal T-cell population carries 1 (monoallelic) or 2 (biallelic) rearranged TCRγ chain genes.
In the test for clonal T cells, pairs of specific PCR primers target the conserved regions flanking the (V) and (J) gene segments. Nonrearranged (V) and (J) segments in the germ line configuration are far apart, and therefore do not give rise to PCR products. A PCR product will arise only from rearranged (V) and (J) segments. Each specific (V) and (J) rearrangement produces a PCR product of a characteristic size. Individuals without celiac disease will have many different T cells, each with a specific TCR (a polyclonal population), leading to the formation of PCR products of many different sizes (Fig. 1A⇓ ). This test revealed that the intestinal biopsy specimen from the case patient contained biallelic TCR gene rearrangements (one of which is shown in Fig. 1B⇓ ). This finding of a predominant clonal T-cell population along with villous blunting observed by small bowel biopsy indicated a diagnosis of type II refractory celiac disease.
prognosis and treatment of type ii refractory celiac disease
The 5-year survival for type II refractory celiac disease is <50%, with the most common causes of death being T-cell lymphoma and infection. Treatment options include corticosteroids and immunosuppressive agents such as thiopurines and infliximab. There is concern that immunosuppressive therapy promotes progression to lymphoma, but no data confirm this risk. Therapies under investigation include antibody to IL-15, a cytokine that leads to enhanced enterocyte killing in celiac disease (10), and stem cell transplantation.
Given the results of the TCR studies, repeated confirmation of the patient’s strict adherence to a gluten-free diet, and the patient’s continued dependence on steroids for a period of several months, immunosuppressive therapy was initiated with 6-mercaptopurine with the goal of decreasing or eliminating the need for corticosteroid therapy. The patient was able to discontinue cortocisteroids and at the time of this report was doing well on 6-mercaptopurine alone.
Points to remember
Celiac disease has a prevalence of 1%, and diagnosis is based on concordance of serologic tests, small bowel biopsy, and resolution of symptoms upon withdrawal of gluten from the diet.
Wheat, rye, and barley are the main dietary sources of gluten that contribute to celiac disease. The gluten in oats does not seem to contribute to celiac disease.
Testing for anti–human tTG antibodies provides a sensitivity approaching 100%. Antigliadin antibody testing is no longer recommended because of low sensitivity and specificity.
Several possible explanations account for patient failure to respond to a gluten-free diet, including nonadherence, other coexisting conditions (such as small bowel bacterial overgrowth, lactose intolerance, and microscopic colitis), and the presence of type I or II refractory celiac disease.
An appropriate test for refractory celiac disease for which adherence to a gluten-free diet has been verified is a TCRγ gene-rearrangement study to determine if a prominent T-cell clone is present in the intestinal mucosa.
Grant/funding Support: L.M.M.’s postdoctoral training in clinical chemistry and laboratory medicine is supported by a Past Presidents’ Scholarship from the Van Slyke Foundation of the American Association for Clinical Chemistry. G.C.B. is supported by a Ruth L. Kirschstein National Research Service Award 1F32HL086046-01. We thank the Department of Pathology for additional support of L.M.M.
Financial Interests: None declared.
- © 2008 The American Association for Clinical Chemistry