Article Text


IL-1 gene cluster and TNFA−307 polymorphisms in the risk of perforated duodenal ulcer
  1. J B Guerra,
  2. G A Rocha,
  3. A M C Rocha,
  4. C M de Castro Mendes,
  5. I E B Saraiva,
  6. C A de Oliveira,
  7. D M M Queiroz
  1. Laboratory of Research in Bacteriology, Faculty of Medicine, UFMG, Belo Horizonte, Brazil
  1. Correspondence to:
    Professor D M M Queiroz
    Laboratory of Research in Bacteriology, FMUFMG Av Alfredo Balena, 190/4026-30130-100, Belo Horizonte, Brazil; dqueiroz{at}

Statistics from

Helicobacter pylori virulence markers have been associated with duodenal ulcer (DU) but there are few studies evaluating host factors such as cytokine polymorphisms and, to the best of our knowledge, no study has evaluated these polymorphisms as risk factors for perforated DU. We investigated associations among interleukin 1 (IL-1) cluster and tumour necrosis factor α (TNFA)−307 polymorphisms, and DU and perforated DU in a non-Caucasian population. We included 223 patients with DU, 29 patients with perforated DU, and 541 blood donors. H pylori status was investigated by culture, preformed urease test, stained smear, polymerase chain reaction (PCR), and the 13C-urea breath test. cagA status was assessed by PCR. In the blood donors, H pylori status and cagA status were determined by serology. IL-1B−511/−31, IL-1RN, and TNFA−307 polymorphisms were genotyped by PCR, PCR/restriction fragment length polymorphism, or PCR/confronting two pair primers.1 Data were analysed in logistic models. The loci did not deviate significantly from the expected Hardy-Weinberg distribution in the control group. IL-1B−511T and IL-1B−31C polymorphic alleles were in almost complete linkage disequilibrium in all three groups (p<10−6). We thus restricted further analyses to IL-1B−31. No polymorphism remained associated with non-complicated DU after correcting for age and sex, but the IL-1RN2 carrier showed a trend towards increasing DU risk (p = 0.06; odds ratio (OR) 1.43 (95% confidence interval (CI) 0.99–2.05)). Regarding perforated DU, in the multivariate analysis, IL-1B−31C and TNFA−307A alleles remained inversely associated with the disease, even after inclusion of confounding factors (table 1). cagA status remained the strongest factor associated with either uncomplicated (p = 0.00; OR 4.29 (95% CI 2.63–6.98)) or perforated DU. The other polymorphisms were not associated with perforated DU (table 1).

Table 1

 Univariate and multivariate analysis of the cytokine loci between patients with perforated duodenal ulcer (n = 29) and all blood donors (n = 539), and between patients with perforated duodenal ulcer (n = 29) and Helicobacter pylori positive blood donors (n = 369)

Although morbidity from peptic DU has greatly decreased since early studies on H pylori infection,2 little change was observed regarding perforated DU, as measured by surgical interventions in emergency services.3 Knowing who, among all H pylori infected subjects, will develop a perforated DU is therefore an important issue in treatment.

Garcia-Gonzales and colleagues4 and Zambon and colleagues,5 evaluating Spanish and Italian populations, respectively, did not find associations between single IL-1 polymorphisms and DU. Conversely, Furuta and colleagues6 found that IL-1RN allele 2 and IL-1B−511T/T were protective factors for DU in a Japanese population.

In this investigation, in accordance with previous studies,4,5 no role could be established for IL-1B−511T or IL-1B−31C alleles in non-complicated DU. However, IL-1B−31C and TNFA−307A carriage was negatively associated with perforated DU. Thus the same IL-1B and TNFA polymorphisms which were associated with atrophy and increased gastric carcinoma risk in Caucasian populations7–9 were found to be inversely associated with perforated DU.

The mechanism by which overproduction of IL-1β and TNF-α due to IL-1B−31 and TNFA−307 polymorphisms protects from DU perforation may not differ from that associated with gastric carcinoma. The prevailing mechanism is probably inhibition of gastric acid production. Consequently, bacteria spread to the corpus where they accentuate the inflammation, lowing acid production, with the net effect of diminishing the risk of DU perforation.

Even though our results are biologically plausible, several factors may contribute to geographical specificities, as already seen in studies on other gastrointestinal diseases.7,8,9,10 Also, we have previously demonstrated1 in our population that the distribution of the inflammatory alleles at IL-1 loci is intermediate between Asians and Caucasians.1

In conclusion, one of the questions that motivated the studies associating host cytokine polymorphisms with H pylori associated diseases was the possibility of explaining why some infected individuals develop gastric carcinoma, others peptic ulcer, and the majority remain otherwise without complications. These polymorphisms may play a role in the genesis of H pylori associated diseases but are probably insufficient to completely answer this question. Our study demonstrated independent inverse associations between IL-1B−31C and TNFA−307A polymorphic alleles and perforated DU, but no association with non-complicated DU.


View Abstract


  • Conflict of interest: None declared.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.