Search

Objective: Our previous coeliac disease genome-wide association study (GWAS) implicated risk variants in the human leucocyte antigen (HLA) region and eight novel risk regions. To identify more coeliac disease loci, we selected 458 single nucleotide polymorphisms (SNPs) that showed more modest association in the GWAS for genotyping and analysis in four independent cohorts. Design: 458 SNPs were assayed in 1682 cases and 3258 controls from three populations (UK, Irish and Dutch). We combined the results with the original GWAS cohort (767 UK cases and 1422 controls); six SNPs showed association with p

Serum immunoglobulins were measured in one hundred and eighty-four healthy controls. One hundred and fifty-nine of these were also HLA typed. IgG levels did not differ with sex or HLA-B8 status. IgM levels did not differ with HLA-B8 status but were significantly higher in females than males. IgA levels were lower in females than males; they were also lower in HLA-B8 positive compared with HLA-B8 negative individuals; the lowest IgA levels were found in HLA-B8 positive females. The IgA variation may be relevant to the higher incidence of certain disorders among HLA-B8 positive individuals and among females.

PUBLISHED, * Joint Senior Authors, We acknowledge funding from Coeliac UK (to DAvH); the Coeliac Disease Consortium (an innovative cluster approved by the Netherlands Genomics Initiative and partly funded by the Dutch Government, grant BSIK03009 to CW); the European Union (STREP 036383); the Netherlands Organization for Scientific Research (VICI grant 918.66.620 to CW); the Science Foundation Ireland; the Irish Health Research Board; Hitachi Europe Ltd.; and the Wellcome Trust (GR068094MA Clinician Scientist Fellowship to DAvH; New Blood Fellowship to RMcM; support for the work of RMcG and PD). We thank the Barts and The London Genome Centre for genotyping support; J Swift, P. Kumar, D.P. Jewell, S.P.L. Travis, K. Moriarty for collection of UKGWAS and additional samples. We acknowledge use of DNA from the British 1958 Birth Cohort collection, funded by the UK Medical Research Council grant G0000934 and the Wellcome Trust grant 068545/Z/02.. We thank A. Monsuur for patient recruitment, G. Meijer and J. Meijer for histology review, K. Duran for DNA extraction, H. van Someren and F. Mulder for clinical database management, M. Plateel and the Genotyping facilities at UMCG and UMC Utrecht for technical assistance (the Netherlands). We thank M. Abuzakouk, R. McLoughlin, K. Brophy, C. Feighery, J. McPartlin for sample collection. Irish Control DNA was supplied by Irish Blood Transfusion Service/Trinity College Dublin Biobank. We thank the Wellcome Trust Centre for Human Genetics, University of Oxford for provision of computing facilities and G. McVean for recombination rate data. We thank all celiac and control individuals for participating in this study.

Chromosomal region 2q33 encodes the immune regulatory genes, CTLA4, ICOS and CD28, which are involved in regulation of T-cell activity and has been studied as a candidate gene locus in autoimmune diseases, including coeliac disease (CD). We have investigated whether an association exists between this region and CD in the Irish population using a comprehensive analysis for genetic variation. Using a haplotype-tagging approach, this gene cluster was investigated for disease association in a case-control study comprising 394 CD patients and 421 ethnically matched healthy controls. Several SNPs, including CTLA4_CT60, showed association with disease; however, after correction for multiple-testing, CTLA4-658C/T was the only polymorphism found to show significant association with disease when allele, genotype, or carrier status frequency were analysed (carrier status (Allele C), P = 0.0016). Haplotype analysis revealed a haplotype incorporating the CD28/CTLA4 and two 5' ICOS polymorphisms to be significantly associated with disease (patients 24.1%; controls 31.5%; P = 0.035), as was a shorter haplotype composed of the CTLA4 markers only (30.9 vs 34.9%; P = 0.042). The extended haplotype incorporating CD28/CTLA4 and 5' ICOS is more strongly associated with disease than haplotypes of individual genes. This suggests a causal variant associated with this haplotype may be associated with disease in this population.

Susceptibility to sarcoidosis and coeliac disease has been linked to the class II haplotype HLA-DR3, DQ2, and an association between the two disorders has been suggested. As a pilot study, we have sought to determine the prevalence of coeliac disease in a cohort of Irish patients with sarcoidosis.Prospective, case-controlled study.One hundred and two sarcoid patients (47 males, 55 females) from the west of Ireland and 105 (52 males, 53 females) healthy, ethnically matched, controls underwent interview and screening for coeliac disease and human leucocyte antigen typing by serology. Those with elevated anti-gliadin IgA (AGA) and/or positive endomysial antibody (EMA) were offered small intestinal biopsy.Three (3%) sarcoid patients had a prior diagnosis of coeliac disease. A further 12 (12%) patients and four (4%) controls had elevated AGA (P = 0.047), of whom three and one, respectively, had positive EMA. Small intestinal biopsy in 11 patients and three controls confirmed coeliac disease in one individual each, giving a prevalence of coeliac disease in patients compared with controls of 4/102 (4%) versus 1/105 (1%) (P = 0.21). Sensitivity and specificity of EMA and elevated AGA in sarcoid patients was 100% and 50%, and 50% and 9%, respectively. Of the four affected sarcoid patients, three carried HLA-DR3, DQ2 and one carried DR5 (12), DR7, DQ2.We have demonstrated a moderately increased prevalence of coeliac disease in Irish patients with sarcoidosis, which we feel justifies future screening of our sarcoid population. Estimation of EMA is recommended and should be restricted to those with susceptible haplotypes.

In addition to the well-established association of coeliac disease (CD) with HLA-DQ (6p21) and possibly CTLA4 (2q33), there is considerable evidence for a susceptibility locus on chromosome 5q, which contains many potential candidates for inflammatory disease, including a cluster of cytokine genes in 5q31. CD cases and controls were genotyped for four single-nucleotide polymorphism (SNP) markers that together characterize >90% of the haplotype variation at the IBD5 locus encoding, among others, the SLC22A4 gene. IBD5 and SLC22A4 map to 5q31 and have recently been associated with Crohn's disease and rheumatoid arthritis. Haplotype frequencies do not differ significantly between CD cases and controls in the Irish population, and therefore the chromosome 5 CD susceptibility locus most likely lies elsewhere on 5q.

In a control population of 127 healthy adults in the West of Ireland, the possession of the HLA-antigens associated with idiopathic haemochromatosis (IH), i.e. HLA-A3, B7 and B14 was not accompanied by higher levels of serum iron, total iron binding capacity (TIBC) or transferrin saturation (TS%). HLA-A3 does not appear to confer enhanced ability to absorb dietary iron.

Immunoglobulin deficiency, especially deficiency of IgA, has been described in patients with celiac sprue (CS). Our study was performed in an area of high prevalence of CS to determine the prevalence of immunodeficiency states in patients with CS, to examine their clinical characteristics, response to treatment, and HLA phenotypes compared with a group of age- and sex-matched persons with CS but without immunoglobulin deficiency. Fourteen of 604 patients with CS were identified as being selectively deficient in IgA, whereas one had common variable immunodeficiency. At diagnosis, anemia was present in 8 of 14 IgA-deficient patients compared with 10 of 42 controls (p = 0.047), whereas abdominal pain was more common in controls with CS. Autoimmunity and recurrent infection were more prevalent in the IgA-deficient group. Response to gluten-free diet was similar in both groups in terms of histologic structure and recovery of intestinal brush-border enzyme activity. IgA-deficient participants with CS had no increased risk of associated malignancy or lymphoma. HLA phenotypes were similar in both groups. The prevalences of selective IgA deficiency and common variable immunodeficiency in this series of patients with CS are 2.31 in 100 and 0.16 in 100, respectively. Although this group is unique in character, close follow-up coupled with conscientious compliance with a gluten-free diet, remains the mainstay of treatment for these patients.

Celiac disease is a strongly heritable gastrointestinal illness that is an especially important model of the genetically complex multifactorial immune mediated diseases. The HLA component of celiac disease (a specific HLA-DQ heterodimer)is largely established and is relatively uncomplicated, and the environmental component (gluten and related grain storage proteins in the diet) is remarkably well understood. Previous family studies of celiac disease suggested that there is at least one important non-HLA locus. This locus may be a stronger genetic factor than HLA, and it apparently has a recessive mode of inheritance. We used a three step genome screening protocol to identify loci that contribute to celiac disease in the western counties of ireland, a region with the highest prevalence of celiac disease in the world. The most significant of several possible non-HLA loci that we found was on chromosome 6p about 30 cM telomeric from HLA. It has a multipoint maximum lod score of 4.66 (compared with 4.44 for HLA-DQ) and appears to have a recessive mode of inheritance. Our study localizes and provides strong evidence for linkage of at least one non-HLA locus to celiac disease and may serve as a prototype for an efficient approach to screening the human genome for loci that contribute to complex diseases.

Celiac disease (CD) has one of the strongest class II HLA associations of any human illness. We used DNA-RFLP typing to study the class II HLA genotypes of celiac disease patients from the West of Ireland, the geographic area with the highest rate of celiac disease in the world. We confirmed the high frequency of HLA-DR3 in this population, and we were also able to demonstrate the additional risk of developing celiac disease imparted by HLA-DR7. This was done by clearly distinguishing DR7,DQ2 haplotypes from DR7,DQ9 haplotypes, and by "subtraction analysis" of haplotype frequencies. As reported in other populations, most of the patients without DR3 were heterozygous for DR7 and DR11 or 12 (DR5), or had DR4. We used PCR-RFLP and direct sequencing of amplified DNA to examine HLA-DR4 subtypes. The frequency of HLA-DR4 was markedly decreased in patients compared with controls (p = 0.000001) and there was a significant alteration of DR4 subtypes of the patients compared with controls (p = 0.0227). Moreover, all of the CD patients (5 of 5) with DR4 had a haplotype associated with the DQB1*0302 allele compared with only 11 of 23 control subjects with DR4. Our results in this population with exceptionally high risk of CD strongly support the DQ heterodimer hypothesis and suggest that the recently described sequence difference between the DQB1*02 alleles of DR3 and DR7 may contribute to a synergistic increased risk when these haplotypes are inherited together. In addition, our findings suggest a role for HLA-DQ in DR4-associated CD.

Celiac sprue (CS) is frequently complicated by malignancy, most commonly small intestinal lymphoma. Our study was performed in an area with a high prevalence of CS to define the clinical features, response to treatment, and outcome of this tumor. Of a total of 31 lymphomas complicating CS identified, 30 case records and 24 tumor specimens were reviewed. Overall 1-year survival was 9 of 29 (31%) and 5-year survival 3 of 27 (11%). Seven previously diagnosed celiac patients developed lymphoma ; length on gluten-free diet ranged from 12 to 252 months (median 44 months). In this group, presentation was nonspecific, diagnosis difficult, and survival poor (lymphoma diagnosed in life in four of seven, mean survival 2.25 months). Twenty-three patients had CS and lymphoma diagnosed during the same illness. In this group, 14 of 23 presented with a surgical emergency and were treated with tumor resection and chemotherapy. Nine are disease-free and alive or died of another cause after 10-196 months (mean 74 follow-up). Celiac-associated lymphoma is a frequent, difficult to diagnose, and commonly fatal complication of CS. An aggressive diagnostic approach, including laparoscopy, is recommended. Long-term survival can be expected in a significant number of these patients and in our series was almost exclusively confined to those treated with chemotherapy.

Background Recent whole genome analysis and follow-up studies have identified many new risk variants for coeliac disease (CD, gluten intolerance). The majority of newly associated regions encode candidate genes with a clear functional role in T-cell regulation. Furthermore, the newly discovered risk loci, together with the well established HLA locus, account for less than 50% of the heritability of CD, suggesting that numerous additional loci remain undiscovered. Linkage studies have identified some well-replicated risk regions, most notably chromosome 5q31 and 11q23. Methods We have evaluated six candidate genes in one of these regions (11q23), namely CD3E, CD3D, CD3G, IL10RA, THY1 and IL18, as risk factors for CD using a 2-phase candidate gene approach directed at chromosome 11q. 377 CD cases and 349 ethnically matched controls were used in the initial screening, followed by an extended sample of 171 additional coeliac cases and 536 additional controls. Results Promotor SNPs (-607, -137) in the IL18 gene, which has shown association with several autoimmune diseases, initially suggested association with CD (P < 0.05). Follow-up analyses of an extended sample supported the same, moderate effect (P < 0.05) for one of these. Haplotype analysis of IL18-137/-607 also supported this effect, primarily due to one relatively rare haplotype IL18-607C/-137C (P < 0.0001), which was independently associated in two case-control comparisons. This same haplotype has been noted in rheumatoid arthritis. Conclusion Haplotypes of the IL18 promotor region may contribute to CD risk, consistent with this cytokine's role in maintaining inflammation in active CD.

Restriction fragment length polymorphism analysis, using a single restriction enzyme TaqI-multiple-probe system for HLA-DRB1-DQB1 and -DQA1, was used to determine HLA-DR and -DQ frequencies in 56 unrelated celiac patients and 47 unrelated controls from the west of Ireland. In addition, HLA-DPB1 allelic frequencies were determined in the same group of patients and controls by using the technique of enzymatic DNA amplification of the polymorphic second exon of HLA-DPB1 genes in conjunction with sequence-specific oligonucleotide probing. The results suggest that HLA-DQ rather than HLA-DR is more important in conferring susceptibility to celiac disease. Furthermore, no association between HLA-DP and celiac disease was found in this study.

Although coeliac disease (CD) is strongly associated with the HLA alleles B8 and DR3, the genetic basis of this illness remains obscure. Recent studies show that at least two unlinked loci are involved. Most studies agree on recessivity at the HLA-unlinked locus but differ with respect to dominance or recessivity at the HLA-linked disease susceptibility locus. To address this controversy, we examined the association of CD with HLA in 39 families from the West of Ireland. Previous studies have shown that the prevalence of CD and the frequencies of the HLA antigens associated with it are higher in this population than in most others. Analysis of the data revealed a significant excess of concordant sib-pairs with two HLA haplotypes in common and an excess of discordant pairs with no haplotype in common. Chi-square tests confirmed a highly significant association between HLA-B8 and CD. Both heterozygotes and homozygotes for B8 had a significantly increased risk of CD. The risk for homozygotes was slightly higher than for heterozygotes, although not significantly so. The segregation ratio for disease occurrence among sibs of probands was estimated to be 0.185 when neither parent is affected. We estimated a gene frequency of 0.003 for the disease allele (C) at the HLA-linked locus and of 0.648 for the disease allele (d) at the HLA-unlinked locus. Assuming that CCdd homozygotes are always affected and that only carriers of C who are homozygous dd can be affected, the disease was found to be completely penetrant in Ccdd heterozygotes. These results support dominance at the HLA-linked locus conferring susceptibility to CD. Possible reasons for the discrepancy between the West of Ireland and other populations are discussed.

Hypertransaminasaemia is a common abnormality found in up to 40% of untreated coeliac patients, which resolves with the institution of a gluten-free diet. A much rarer occurrence is the association of chronic liver disease with coeliac disease. Primary biliary cirrhosis, primary sclerosing cholangitis, and chronic autoimmune hepatitis have all been recognized in coeliac patients. More recently, the occurrence of coeliac disease in patients with chronic liver disease is found to be 10 times greater than that in the general population. Of particular note are the 13 patients reported to date with liver failure and coeliac disease. Among those patients commenced on a gluten-free diet an improvement in liver function was observed. The relationship between coeliac disease, a gluten-free diet, and the course of chronic liver disease needs to be clarified.

Genetic predisposition to coeliac disease (CD) is determined primarily by alleles at the HLA-DQB locus, and evidence exists implicating other major histocompatibility complex-linked genes (6p21) and the CTLA4 locus on chromosome 2q33. In addition, extensive family studies have provided strong, reproducible evidence for a susceptibility locus on chromosome 5q (CELIAC2). However, the gene responsible has not been identified. We have assayed genetic variation at the IL4, IL5, IL9, IL13, IL17B and NR3C1 (GR) loci, all of which are present on chromosome 5q and have potential or demonstrated involvement in autoimmune and/or inflammatory disease, in a sample of 409 CD cases and 355 controls. Thirteen single nucleotide polymorphisms were chosen on the basis of functional relevance, prior disease association and, where possible, prior knowledge of the haplotype variation present in European populations. There were no statistically significant allele or haplotype frequency differences between cases and controls. Therefore, these results provide no evidence that these loci are associated with CD in this sample population.

To the Editor: Case 15-1996 (May 16 issue)1 describes a typical case of celiac disease in an adult in whom a complicating T-cell lymphoma led to death. The chronically ill and malnourished conditio...

Splenic function was assessed using 'pitted' erythrocyte counts in 61 first degree relatives of patients with coeliac disease. 'Pitted' erythrocyte counts were normal in 12 parents, but were raised in 20% of 49 siblings and/or children of coeliac patients. First degree relatives had higher 'pitted' erythrocyte counts than normal controls (p = 0.002). The counts were lower in coeliac relatives than in age matched coeliacs (p = 0.0001), but no difference was present between the relatives and coeliac patients whose small bowel mucosa was morphologically normal. Considerable interfamily variation was found in 'pitted' erythrocyte counts, both in the coeliac patients and first degree relatives, and the pattern tended to 'run true' within families. The genetic factor influencing splenic function in coeliac disease is not HLA-linked but seems to be associated with a second, probably recessive, gene influencing the inheritance of coeliac disease.

Splenic function was quantitatively assessed using "pitted" erythrocyte counts in 177 patients with gluten-sensitive enteropathy. Hyposplenism was found to be a common, but not inevitable, complication of gluten-sensitive enteropathy that fluctuated with disease activity. Splenic function improved after withdrawal of gluten from the diet and a close relationship was demonstrated between hyposplenism and the morphology of the small intestine. The severity of the hyposplenism increased with advancing age and prolongation of exposure to dietary gluten. Splenic function did not vary with the HLA-A, -B, and -DR antigens, but, due to the presence of HLA-B8 and -DR3 in the vast majority of patients, a role for this haplotype in the causation of the hyposplenism cannot be excluded.

Organ culture of small intestinal mucosa from coeliac and non-coeliac patients is described. Morphological preservation of structure was observed and an increase in the levels of alkaline phosphatase and a decrease in the lactic dehydrogenase from coeliac mucosa was noted. Similar trend in non-coeliac mucosa were not statistically significant. The initial elevted lactic dehydrogenase activity in coeliac mucosa has not previously been reported.

This paper examines the issue of non-attendance at outpatient clinics at the Regional Hospital, Galway, from the viewpoint of the patients who include both urban and rural residents. The results of a questionnaire survey of outpatients attending general and specialist medical and surgical clinics illustrate that very substantial costs are incurred and long periods of time are spent travelling by many patients. Females, and married females in particular, experience special difficulty in keeping appointments. Non-attendance increases as the cost of transport increases but many patients seriously underestimate the real cost of travel. Patients who have been attending over long periods of time have the worst record of non-attendance. It is recommended that any reorganisation of hospital outpatient systems in rural areas such as Western Ireland should take account of the particular needs of widely dispersed populations.

SERUM antibodies to Epstein-Barr viral antigens were studied in 56 patients with treated coeliac disease and 53 healthy controls. Four individuals with a malignancy were among the group studied. Twenty-one patients (37.5%) showed unusual antibody response patterns which were broadly divisible into two groups: (1) those with elevated antiviral capsid antigen (VCA), antibodies associated with positive anti-early antigen (EA) responses, and (2) those with low anti-Epstein-Barr nuclear antigen (EBNA) responses. While the results indicate that some patients with coeliac disease have an immunological abnormality with respect to EBV, there is as yet no direct evidence that this results in malignancies.

A study of various aspects of access to outpatients clinics in rural Ireland is reported. A dispersed catchment area is served by the Regional Hospital, Galway; 50 per cent of respondents live 20 miles from the hospital; in 40 per cent the round trip will take more than six hours and in 50 per cent the cost of travel will be more than IR £5. Transport‐related problems are cited as the major cause for previously missed appointments, especially in subjects expected to attend four or more times per year. The highest rate of defaulting was found in those patients who had been attending outpatient clinics for several years with a long interval between appointments. Housewives with elderly or young dependants constitute an occupational group with difficulty in keeping appointments.

Endoscopic dye-spraying enables the visualization. in vivo, of the surface morphology throughout the duodenal cap. Examination of the duodenal bulb after dye-spraying in 19 untreated coeliac patients revealed the absence of villi and their replacement with a mosaic pattern. Villi were present in all 35 non-coeliac patients examined. Mucosal biopsies taken at endoscopy demonstrate that the morphological, histological and biochemical changes found in coeliac disease are similar in the proximal duodenum and in the jejunum.

Normal healthy individuals that were HLA-DR3 positive were found to have significantly higher total lymphocyte counts in peripheral blood compared to healthy individuals that were HLA-DR3 negative.

The high prevalence of coeliac disease in children in the West of Ireland previously reported is confirmed by recent studies (Figure 2.1)1–3. When the number of coeliac births is compared with the total number of births over the period 1960–78, 1 coeliac, presenting in childhood, is born per 555 births in families domiciled in County Galway. So far, no child born in County Galway in 1964 has developed symptoms leading to the diagnosis of coeliac disease. However, to date 17 children born in 1972 have been diagnosed. A reliable estimate of The Genetics of Coeliac Disease the frequency of childhood coeliac disease for the years 1960–78 can be made as during that period the Regional Hospital, Galway (RHG) was the only referral centre for the paediatric patients from County Galway.

The sera of 69 index coeliac patients, 121 of their first-degree relatives, and 104 controls were screened for the presence of reticulin antibodies. Among the untreated coeliac patients 75% of adults and 93% of children had reticulin antibody in their serum. Reticulin antibody was not present in any adequately treated coeliac patient. Of the first-degree relatives, 21 were reticulin antibody positive; 17 of these were biopsied and 12 were shown to have coeliac disease. Sixty-five of the coeliac relatives who did not have reticulin antibodies in their sera were biopsied and two had coeliac disease. Of the 68 relatives and 63 controls with normal biopsies, five of the relatives and four of the controls were reticulin antibody positive.

Features that might distinguish uncomplicated coeliac disease from enteropathy associated T cell lymphoma were investigated. Of 76 patients with coeliac disease, 71 (93%) had raised levels of alpha gliadin antibody and all responded clinically and histologically to treatment with a gluten free diet. In contrast, none of 16 patients with enteropathy associated T cell lymphoma had raised levels of alpha gliadin antibody, and treatment with a gluten free diet resulted in histological improvement in one and transient clinical improvement in six patients. The ratio of women to men was 2.2:1 in the group with coeliac disease and 1:1.6 in the patients with enteropathy associated T cell lymphoma. Thus patients with enteropathy associated T cell lymphoma do not display a humoral immune response to wheat protein (alpha gliadin), rarely respond to a gluten free diet, and are often men. Patients with uncomplicated coeliac disease usually have raised levels of alpha gliadin antibody, always respond to a gluten free diet, and are frequently women. These findings suggest the presence of two separate forms of enteropathy: one is benign and sensitive to wheat protein whereas the other runs a malignant course.

Follow-up studies on 36 children, in whom celiac disease (gluten-sensitive enteropathy) was established by gluten challenge, were carried out after management on gluten-free diets for a mean of six years. Evaluations included measurement of height and weight, which for the group approximated normal distributions, and histologic examination of the duodenal or jejunal mucosa. Mucosal morphology was regarded as normal in 16, and there were minimal changes in 20. Epithelial cell height was within the normal range in all of the children. Interepithelial lymphocytes were within normal range in the majority and lymphoid cells in the lamina propria were not different from those in control subjects. Mucosal lactase was significantly lower in patients than in control subjects in the duodenum and the jejunum, whereas sucrase and alkaline phosphatase values were significantly lower in the jejunum but not in the duodenum. Low content of mucosal lactase and increased numbers of interepithelial lymphocytes may be sensitive indicators of persisting ingestion of gluten in mucosa that is otherwise normal or approximately so in appearance.

1. Gliadin from which carbohydrate was removed by treatment with carbohydrase from Aspergillus niger was fed to three coeliac patients in remission. 2. Xylose absorption, mucosal morphology and brush-border enzymes were used to assess the toxicity of the carbohydrase-treated gliadin. 3. Gliadin treated with carbohydrases did not damage the intestinal mucosa of the coeliac patients. 4. The primary structure of the gliadin proteins was not altered by the enzyme treatment.

HLA, A, B, and DR typing was done on seven of 10 patients with lymphoma, and six patients with carcinoma, all of whom had a flat small intestinal mucosa. Sixty-nine per cent (nine of 13) had HLA B8 and 71% (five of seven) had DR3. The corresponding levels for the local coeliac population are 76% and 84% respectively and the local non-coeliac population 43% and 44%. Two of the patients with lymphoma had a child with coeliac disease. The histological type of the lymphoma was malignant histiocytosis of the intestine in seven, histiocytic lymphoma in two, and was not classifiable in one. The similarity of the HLA type suggests that the flat mucosa in both groups is due to coeliac disease.