Intended for healthcare professionals


Primary biliary cirrhosis: epidemiology helping the clinician

BMJ 1996; 312 doi: (Published 11 May 1996) Cite this as: BMJ 1996;312:1181
  1. Jane V Metcalf,
  2. Denise Howel,
  3. Oliver F W James,
  4. Raj Bhopal
  1. The PBC Support Group may be contacted at the British Liver Trust, Central House, Central Avenue, Ransomes Europark, Ipswich IP3 9QG.
  2. Wellcome research fellow Department of Medicine, University of Newcastle upon Tyne, Newcastle upon Tyne NE2 4HH
  3. Senior lecturer in statistics Department of Public Health and Epidemiology
  4. Professor and head Department of Medicine
  5. Professor and head Department of Public Health and Epidemiology

    Epidemiology improves our knowledge of disease

    Recent epidemiological and clinical research is changing our thinking on primary biliary cirrhosis, a chronic cholestatic liver disease probably of autoimmune origin. Until the early 1970s primary biliary cirrhosis was perceived as a disease that was rare, presented with persistent jaundice, and usually progressed to death from liver failure.1 Epidemiological and clinical research based on both registers of cases in defined populations and, more commonly, clinical case series has transformed that picture. The disease is much commoner than was believed,2 3 4 as many as half of all patients may have no symptoms of liver disease at diagnosis,5 and both the onset of complications and death can be delayed by treatment with ursodeoxycholic acid6 and, when necessary, liver transplantation. Good clinical and health care strategies will depend on further integration of epidemiological and clinical research to provide better information on the frequency and natural course of the disease and to generate hypotheses about causation.

    For both epidemiological and clinical purposes a widely accepted and validated case definition is necessary. In primary biliary cirrhosis the triad of results from liver function tests suggesting cholestatic liver disease, the presence of antimito-chondrial antibody on indirect immunofluorescence, and compatible or diagnostic liver histology provide such criteria.7 A possible case may be regarded as having any two of these. Recent descriptive studies using these criteria have indicated an apparent dramatic increase in the prevalence of primary biliary cirrhosis in some regions but with considerable geographical variation. This variation is not only global but also extends to adjacent, apparently similar, countries such as Denmark, where prevalence is low, and Sweden, where it is high.2 8 9 North east England has seen an apparent tenfold increase in prevalence over the past 20 years.3 4

    It is still unclear to what extent these geographical and temporal variations, seen in a few areas where data are available, are due to true variations in prevalence or to variations in case registration, diagnostic activity, access to expertise in hepatology, or even local interest in the disease. These variations are important, both clinically, for health care planning, and for research. If there are true geographical variations in prevalence then different levels of awareness and investigation are needed according to local patterns. If, however, the geographical variations are artefactual then clinicians in some areas are missing cases. A genuine rise in disease frequency requires planning now for the increased health resources needed in the future.

    Most published studies have described case series and have not used consistent methods for case finding or screened a well defined population over a defined period. The use of an epidemiological framework10 to establish whether prevalence is truly increasing and to study geographical variation will require collaborative studies between clinicians and epidemiologists both locally and internationally, using registers based on internationally agreed protocols.

    Evidence of true temporal and geographical variations in the prevalence of primary biliary cirrhosis will enable us to generate new hypotheses about its cause. Here analytical epidemiology can be combined with laboratory and clinical observations to help establish the cause. Many putative causes and risk factors for primary biliary cirrhosis have been suggested over the years. These have included an unknown factor associated with the water supply,11 urinary tract infections caused by Escherichia coli, drugs (particularly phenothiazines), a history of obstetric complications or gynaecological interventions, disease onset after childbirth or anaesthetic,12 and, most recently, an atypical mycobacterial infection (a response possibly associated with immune mimicry).13 None of these has been confirmed as having a direct or indirect causal relation.

    Recent genetic studies have suggested that possession of one of the HLA class II antigens associated with the DRW8 locus may confer increased susceptibility to the disease, and it may be no coincidence that this relatively rare tissue type is more prevalent in populations thought to have the highest prevalence of the disease.14 At present a best guess as to cause is that a number of possible trigger factors may initiate progress of the disease in a susceptible individual. Furthermore, familial primary biliary cirrhosis (usually mother-daughter pairs) is described in 1-4% of cases,15 16 and other autoimmune diseases such as thyroid disease and rheumatoid arthritis are commonly found in patients with primary biliary cirrhosis. Surprisingly, there have been no well conducted case-control or cohort studies to examine these or other possible risk factors, and these hypotheses need testing. An understanding of the cause of primary biliary cirrhosis will enable clinicians to develop preventive programmes and rational treatment policies to evolve and answer patients' questions about the disease.

    In conclusion, recent descriptive epidemiological studies have shown that primary biliary cirrhosis is by no means rare, certainly in northern Europe and particularly among middle aged women. Its rising prevalence, together with the fact that it may be present with few symptoms for months or years and that its cardinal symptom of lethargy is non-specific, should alert physicians and family doctors to its possible diagnosis in any patient with a positive test result for antimitochondrial antibody, a highly specific serological marker for the disease. Our evolving understanding of the clinical course of the disease and an increasing range of treatment options mean that purchasers and providers of health care need to be aware of their changing obligations to these patients, including giving them access to the highly successful patient support group (the PBC Support Group). Continuing collaboration between epidemiologists and clinicians will provide important clues about the cause and pathogenesis of this enigmatic disease and develop our knowledge of its prevention and management.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.
    16. 16.
    View Abstract