Background/aims: Development of primary biliary cirrhosis in the close relatives of patients with the disease has been reported in several series, suggesting a genetic component to disease susceptibility. In this study we set out to calculate, as accurately as possible, the prevalence of familial primary biliary cirrhosis in a geographically-based population. Using local population prevalence data, we have also calculated the first-degree relative, sibling and offspring relative risks of overt primary biliary cirrhosis development.
Methods: All patients with definite or probable primary biliary cirrhosis in the city of Newcastle-upon-Tyne, England, were identified by an exhaustive case-finding search and were prospectively interviewed by a single investigator. Full details of family pedigree and familial primary biliary cirrhosis history were obtained.
Results: One hundred and seventy-three patients were identified, with 160 participating in the study. Thirteen reported a family history of primary biliary cirrhosis. In three cases, both relative pairs were within the study group. The prevalence of a positive family history of primary biliary cirrhosis was therefore 10/157 (6.4% [95% Confidence Interval 2.6-10.2%]), 8/10 cases occurring in first-degree relatives. The patients had a total of 1118 first-degree relatives (live or dead) and 468 siblings. The first-degree relative prevalence of primary biliary cirrhosis was 0.72% [0.2-1.2%] (siblings 0.41% [-0.2-1.0%]). The offspring prevalence was 1.2% [0.04-2.4%], (2.3% [0.1-4.5%] for daughters). The sibling relative risk (lambda(s)) was 10.5.
Conclusions: The overall prevalence of definite or probable primary biliary cirrhosis in the first-degree relatives of existing patients is <1%. The risk of disease is not, however, uniform, the highest prevalence being seen in the daughters of patients. Suspicion of disease should therefore be highest in this relative group. The calculated lambda(s) for primary biliary cirrhosis in this geographically-based study is significantly lower than previous estimates from case-note-derived case series, but similar to values seen in other autoimmune diseases.