Susceptibility to celiac disease in Northern Europe is associated with the human leukocyte antigens (HLA) B8, DR3 and DQ2, which exist together on an extended haplotype. The strong predominance of this haplotype within the Northern European celiac populations, together with the linkage disequilibrium which occurs between these loci, does not allow identification of the gene(s) primarily associated with disease susceptibility. Studies from Southern Europe using both serology and examination of restriction fragment length polymorphisms (RFLP) have demonstrated associations with DR3, DR7 and DQ2, suggesting that the DQ locus is primarily involved. We investigated 43 celiac patients and 41 healthy controls from Rome, Italy, using sequence-specific oligonucleotide (SSO) probes, in conjunction with gene amplification by the polymerase chain reaction (PCR), to determine alleles at the DRB, DQA1, DQB1 and DPB1 loci: 19% of celiac patients possessed the alleles DRB1*0301 DRB3*0101, 33% DRB1*0301 DRB3*0201 and 33% of celiac patients were heterozygous for DRB1*1101-1201/DRB1*0701. The strongest association with celiac disease susceptibility was the combination of alleles DQA1*0501 DQB1*0201 (91% celiac patients vs. 12% controls; p = 0.000002). There was no additional susceptibility associated with alleles at the DPB locus. This study confirms the hypothesis that susceptibility is associated with a particular combination of DQ alleles and the ethnic variation in DR frequencies is secondary to linkage disequilibrium with these DQ alleles.

Aims-To highlight the pitfalls in the diagnosis of coeliac disease and to make recommendations for its diagnosis and the management of refractory cases with equivocal histology. Methods-Six patients, referred since 1989 with a diagnosis of coeliac disease based on duodenal biopsy specimens taken at endoscopy, and who failed to respond to a gluten-free diet were studied. All patients were subjected to peroral jejunal biopsy. Morphometric analysis of villus height:crypt depth ratios, surface enterocyte cell heights, and intraepithelial counts was used to aid in the assessment of equivocal histology. Results-Subsequent small intestinal biopsy specimens both taken when the patients were following a gluten-free diet and after gluten challenge were normal in all cases. Morphometric analysis and intraepithelial counts were normal.

The NHS has introduced the two week wait scheme to detect upper gastrointestinal cancers at an early stage and improve survival rates The aim of this study was to assess the impact of this scheme and changes in endoscopy waiting times on tumour stage and resection rates over a four year period. Data were analysed prospectively for all patients diagnosed with oesophagogastric cancer between September 1998 and September 2002 and from those referred under the two week wait scheme since its introduction in 2000. Of those tumours diagnosed by this scheme (15%) only 5% were early disease (stage 1 or 2). Patients with early cancer, mainly diagnosed by routine gastroscopy, do not present with symptoms meeting the two week wait criteria. An increase in the resection rates for early disease will most probably be seen with a reduction in routine endoscopy waiting times.

Fifty-two British and 29 Greek idiopathic membranous nephropathy (IMN) patients were analysed for DRB, DQA1, DQB1 and DPB1 gene polymorphism using second exon amplification and sequence-specific oligonucleotides (SSO). In addition 100 British and 92 Greek controls were analysed. A highly significant increased frequency of the DRB1*0301 allele was found in IMN patients from Britain (80%), when compared to controls (27%, OR 10.6, P = 0.000004). A lower frequency of DRB1*0301 was observed in Greek IMN patients (33%), but this was just significant before correction, when compared to Greek controls (15%, OR 3, P = 0.02). The DRB3 allele most often associated with DRB1*0301 was DRB3*0101 (OR 4.2, P = 0.00025) in British patients and DRB3*0201/2 (OR 11, P = 0.006) in Greek patients. In Greek IMN patients a decrease in DR16 was found (OR 0.08, P = 0.004), and the overall incidence of DR2 was significantly lowered when both sets of IMN patients were combined (OR 0.21, chi 2 17.6, P = 0.00013). The incidence of DQA1*0501 was raised in both Greek (96% vs. 66%, OR 9.7, chi 2 6.9, P = 0.009) and British IMN (85% vs. 45%, OR 7.4, chi 2 20, P = 0.00007) patients. This gives some support to a proposal for a major role for this allele in IMN.(ABSTRACT TRUNCATED AT 250 WORDS)