RESUMEN
OBJECTIVES: The aetiology of primary sclerosing cholangitis (PSC) is unknown, and the role of micro-organisms has been studied only to a limited extent. We tested the hypothesis that past or persisting infection with common viruses or atypical bacteria might play a role in genetically susceptible hosts. DESIGN: Case-control study. METHODS: Serological screening for antibodies against 22 viruses as well as Chlamydia spp. and Mycoplasma pneumoniae was carried out in 41 well-established PSC patients. All 5110 sera tested in 1997 for these micro-organisms at our laboratory served as a background reference group. Subsequently, Chlamydia anti-lipopolysaccharide (LPS) antibodies were determined by enzyme-linked immunosorbent assay (ELISA) in the PSC group and in three race-matched control groups (inflammatory bowel disease (IBD) group, n = 35; non-IBD patients group, n = 39; healthy blood donor group, n = 40). Subtyping in Chlamydia trachomatis and C. pneumoniae serotypes by specific anti-major outer membrane protein (MOMP) assays was carried out in the four groups. Immunohistochemical staining using specific markers for chlamydiae was carried out on liver biopsies of 14 PSC patients. RESULTS: There was a markedly elevated seroprevalence of Chlamydia-LPS antibodies compared with the 1997 reference group. The odds ratios (ORs) for the presence of immunoglobulin G, immunoglobulin M and immunoglobulin A antibodies for the PSC patients versus the control group were 2.4 (95% confidence interval (CI) 1.1 to 5.4), 1.9 (95% CI 0.9 to 4.0) and 6.7 (95% CI 3.0 to 17.0), respectively. All other micro-organisms tested showed normal antibody profiles that did not differ from the 1997 reference group. The seroprevalence of Chlamydia-anti-LPS antibodies was elevated markedly in the PSC patients compared with the IBD, non-IBD and blood donor groups. The outcomes in the C. trachomatis and C. pneumoniae anti-MOMP assays did not correlate with the anti-LPS-positive PSC sera. The actual presence of Chlamydia bodies in liver tissue could not be demonstrated. CONCLUSION: Our findings suggest an association between PSC and (previous) infection with Chlamydia.