Tumour necrosis factor-α promoter polymorphisms in primary biliary cirrhosis

Background/aims: The incidence of primary biliary cirrhosis (PBC) is increased in the close relatives of patients, suggesting that genetic factors play a role in disease susceptibility. Decreased in vitro production of tumour necrosis factor (TNF)-alpha has been reported in PBC patients, suggesting a potential aetiological role for this cytokine. The aim of this study was to examine two biallelic polymorphisms in the promoter region of the TNF-alpha gene, which may play a role in the control of TNF-alpha secretion, as candidate susceptibility loci in PBC. Methods: The polymorphisms at positions -238 and -308 in the TNF-alpha promoter region were analysed by polymerase chain reaction in 168 unrelated PBC patients and 145 local unrelated, geographically matched normal individuals. All PBC subjects were also genotyped for HLA DR8, a previously identified susceptibility locus in PBC. Results: The -308 TNF1/TNF1 genotype was seen in a similar proportion of PBC patients (66%) and controls (60%). However, this genotype was found significantly more frequently in the 95 PBC patients with more advanced disease (histological stage III/IV) (77%) than in either controls (p<0.01, OR=2.2 [1.2-4.0]) or the PBC patients with earlier disease (38/73 (52%), p=0.001 OR 3.1 [1.6-5.9]). Linkage between TNF -308 and HLA DR8 was not seen. No association was found between PBC and the biallelic -238 TNF-alpha polymorphism, either in the whole PBC population or the histological Stage III/IV subgroup. Conclusions: Our study provides no evidence for involvement of the TNF-alpha -308 or -238 promoter polymorphisms in genetic predisposition to PBC. However. the significantly increased frequency of the -308 TNF1/TNF1 genotype seen in 95 patients with more advanced disease raises the possibility that this allele may be linked to disease progression rather than susceptibility The finding of different allele frequencies in PBC patients in different disease subgroups emphasises the importance of clinical phenotype/case-mix in the design of disease association studies.