CLINICAL LIVER, PANCREAS, AND BILIARY TRACT. Epidemiology of Primary Biliary Cirrhosis in Victoria, Australia: High Prevalence in Migrant Populations

Transcription

1 GASTROENTEROLOGY 2004;127: CLINICAL LIVER, PANCREAS, AND BILIARY TRACT Epidemiology of Primary Biliary Cirrhosis in Victoria, Australia: High Prevalence in Migrant Populations SIDDHARTH SOOD, PAUL J. GOW, JOHN M. CHRISTIE, and PETER W. ANGUS Department of Gastroenterology, Austin Hospital, Heidelberg, Victoria, Australia Background & Aims: The prevalence of primary biliary cirrhosis (PBC) reported in different countries varies widely, indicating that genetic or environmental factors may be important in the etiology of the disease. The aim of this study was to examine this issue further by determining the overall prevalence of PBC in one state in Australia and to examine the prevalence among different migrant groups within this population. Methods: Thorough case-finding methods were used to identify all cases of PBC in Victoria, Australia. Age-adjusted prevalence rates among different migrant groups were examined. Results: A total of 249 cases were identified, giving a prevalence of 51 cases per million. This is significantly higher than the rate documented in a 1991 Victorian study. Prevalence in the 3 largest migrant groups was greater than that of Victoria as a whole (141, 200, and 208 cases per million in British, Italian, and Greek migrants, respectively). In women older than 40 years, previous studies have documented a prevalence of 940 cases per million in women in the United Kingdom; however, the prevalence was 344 cases per million in British-born immigrants to Victoria and 160 cases per million in Australian-born women. Conclusions: The current prevalence of PBC in Victoria is higher than previously reported, but the age-adjusted prevalence in those born in Victoria remains significantly lower than in the United Kingdom and is less than in migrant communities. These findings suggest that Victorians may be relatively protected from developing the disease and add further weight to the suggestion that environmental factors may play a role in the etiology of PBC. The prevalence of primary biliary cirrhosis (PBC) seems to vary in different parts of the world, with particularly high levels reported in studies performed in northern Europe. 1 3 In 1991, a case-finding study of PBC undertaken in Victoria, Australia, 4 documented a prevalence of 19.1 cases per million population, a level much lower than found in parts of northern Europe and England. 2,3,5 A study conducted in the Canadian population found a similarly low prevalence compared with European data. 6 Considering the shared ethnic and genetic background of the Australian, Canadian, and British populations, it has been suggested that these regional differences point to the involvement of an environmental agent in the etiology of the disease. 7 However, alternative explanations may be that they are largely due to differences in local interest and awareness of the condition or variation in the case-finding methods used in different studies. Clearly further studies using standardized methodologies are required to address this issue. In addition, recent data from populations in northern England, where repeated studies of the prevalence of PBC have been performed using the same case-finding approach, have suggested an increasing prevalence of the disease. 1 However, it is unknown whether this trend is apparent in other parts of the world, where a much lower prevalence of the disease has previously been recorded. The aim of the present study was to document the current prevalence of PBC in the state of Victoria, Australia, and to compare it with the prevalence of the disease documented in this population in Although the majority of the population in this state is Australian born and of Anglo-Celtic background, Victoria also has a large and growing immigrant population that comes from many racial backgrounds. Therefore, a second aim of the study was to investigate the epidemiology of PBC in the major ethnic groups in Victoria because this could provide evidence supporting the role of genetic or environmental factors in the etiology of the disease. Patients and Methods Criteria for the Diagnosis of PBC The diagnosis of PBC was based on the following criteria. Definite PBC was defined as the presence of compat- Abbreviations used in this paper: AMA, antimitochondrial antibody; PBC, primary biliary cirrhosis by the American Gastroenterological Association /04/$30.00 doi: /j.gastro

2 August 2004 PREVALENCE OF PBC IN VICTORIA, AUSTRALIA 471 ible liver histology, cholestatic liver function tests, and an antimitochondrial antibody (AMA) titer of at least 1:40. Patients were defined as having probable PBC when 2 of the 3 diagnostic criteria were fulfilled. Patients not fulfilling the criteria of definite or probable PBC were excluded. Study Population and Period The census date applied to this study was June 30, Patients not fulfilling the criteria for the diagnosis of definite or probable PBC at this date were excluded from analysis. The study population is the state of Victoria, Australia. Patient postcodes (zip codes) were collected to verify residence within Victoria. Population statistics were obtained from the Australian Bureau of Statistics. The population of Victoria at the time of the study was 4.88 million, with 72% of the population residing in Melbourne. Data regarding specific immigrant populations were obtained from the Australian Bureau of Statistics 2001 national census Case-Finding Methods Multiple case-finding methods were used in this study. These were largely the same as those used in the 1991 Victorian prevalence study. 4 Firstly, a list of gastroenterologists, hepatologists, and general physicians in Victoria was obtained from the Gastroenterological Society of Australia and the Royal Australasian College of Physicians. An initial mailing was sent to all 302 physicians who were identified in this way. For each patient with PBC under their care, the physicians were asked to complete a questionnaire detailing the patient s demographic information, country of birth, date of presentation, treatment, and clinical condition. In addition, a second questionnaire was sent to those physicians who were caring for the initial cohort of 84 patients identified in the 1991 study to collect clinical follow-up information. Nonrespondents were contacted by telephone calls, s, and further letters to encourage their participation. The database of the Victorian Liver Transplant Unit, based at the Austin Hospital (Heidelberg, Victoria, Australia), was also surveyed. This unit is responsible for all liver transplants performed in the state of Victoria and has maintained a database of all patients referred for consideration for liver transplantation since its inception in Any patients with a diagnosis of PBC had their medical records reviewed and a questionnaire completed. Patients who had received a liver transplant (n 17) before the census date were excluded from analysis. The third method of patient identification was a medical record search for cases of PBC between January 1, 1990, and June 30, 2002, at Victoria s 6 major teaching hospitals. In contrast to the previous study, 4 and in an attempt to match the case-finding technique used in recent studies of the prevalence of PBC in the United Kingdom, 1 a search to identify all patients in whom a positive AMA result had been obtained was conducted at these hospitals from January 1, 1997, to June 30, However, at 2 hospitals (Monash Medical Center and the Western Hospital), due to changes in hospital computer systems, records for AMA-positive patients were only available from 1999 onward. The previously described questionnaire was completed for each patient who was identified as having PBC via the AMA or hospital medical records search. To determine the adequacy of the period of the AMA search, a small pilot study was conducted at the Austin Hospital. With the assistance of the information technology department, the previous hospital computer system was searched for positive AMA results from an extra 2 years (i.e., for a total of 7 years). This only identified one patient in addition to the 31 identified at the hospital overall. This allowed us to accept 5 years as a sufficient and practical time period for which to search AMA results. To prevent duplication of patients, initials, date of birth, and postcode were collected. All data were entered into a secure Microsoft Access 2000 database. The project had ethical approval from relevant institutional ethics committees. Statistical Analysis Rates are presented as mean SD unless otherwise specified. Prevalence rates among different patient groups were compared by the 2 test. To avoid the potential confounding influence of age, the method of indirect standardization was used. P 0.05 was considered statistically significant. Results Case Ascertainment A total of 262 practitioners (87%) responded to the physician survey and were treating 114 patients. A further 8 patients were identified from the Victorian Liver Transplant Unit database. The hospital medical record and AMA search uncovered another 127 patients. Forty-six percent of patients identified in Victorian hospitals (n 59) were found through the search for positive AMA results, 42% (n 53) through discharge coding, and 12% (n 15) through both methods. Thus, 249 cases of PBC were identified in the current study (172 definite PBC and 77 probable PBC), giving an overall prevalence of 51.0 cases per million. This is more than 2.5 times higher than the prevalence recorded in the 1991 Victorian study (19.1 cases per million). 4 By using the indirect method of standardization to correct for age, this rate represented a significant increase in prevalence since 1991 (P ). Of the 249 cases, there were 224 women and 25 men with the disease in Victoria (female-to-male ratio, 9:1). The mean age of patients at the census date was years (range, years) and is shown in Figure 1. Seventy-two percent of patients (n 180) were diagnosed while living in Melbourne, 20.5% (n 51) while

3 472 SOOD ET AL. GASTROENTEROLOGY Vol. 127, No. 2 Table 1. Prevalence of PBC in Australian-, British-, Italian-, and Greek-Born Migrants to Victoria Country of birth No. of patients Prevalence (cases/million) Australia Britain Italy Greece a Refers to method of indirect standardization to compare with the overall Victorian PBC community. P a Figure 1. Age of all 249 patients with PBC at census date. living outside of Melbourne in Victoria, 0.8% (n 2) while interstate, and 1.6% (n 4) while living overseas. There was no significant difference in disease prevalence between those who presented while living in Melbourne and those in the rest of the state (P 0.12). The location at presentation was unknown for 4.8% (n 12) of patients. Prevalence in Immigrant Groups Of the 249 cases, data regarding country of birth were available for 236 patients (94.8%). Of these, only 51.7% (n 122) were born in Australia. Although overseas-born persons comprise 29% of the Victorian population, they make up 46% of our entire cohort with PBC; this apparent difference remained when the prevalence in the 2 groups was adjusted for age (P ). We examined the specific prevalence of PBC in the 4 ethnic groups with the most patient numbers. The prevalence was 37.2 cases per million in Australian-born persons, cases per million in Italian immigrants, cases per million in Greek immigrants, and cases per million in British immigrants. The prevalence of PBC in the Italian and Greek communities was significantly higher than that of the overall Victorian PBC community after correcting for age (P 0.05 for both groups) (Table 1). The prevalence of PBC in the British immigrant community was also statistically greater than the prevalence in the overall population after correcting for age (P 0.02) (Table 1). Although age-specific prevalence rates are not available from published British data, 1 to facilitate comparison between the U.K.-born Victorian community and their native population in Britain, rates for women older than 40 years (the major at-risk group) were compared. The most recently published prevalence of PBC in women older than 40 years in Britain is cases per million. 1 In U.K.-born women older than 40 years now living in Victoria, the prevalence was lower at cases per million (P ). However, the prevalence in this group was significantly higher than the rate in Australian-born women older than 40 years in the current study, which was cases per million (P ). These data are shown in Figure 2. The age of arrival (to Australia) was only available in 50% (n 57) of the overseas-born patients. The mean age at immigration to Australia in this patient group was years. Ursodeoxycholic Acid Seventy percent (n 175) of Victoria s PBC population were being treated with ursodeoxycholic acid at the census date. The dose (in mg kg 1 day 1 )was only available for 52% of patients undergoing therapy. A majority (52%) of these patients were being treated with a dose of ursodeoxycholic acid considered to be suboptimal ( 13 mg kg 1 day 1 ). 11 Clinical Presentation Fifty percent (n 125) of the patients were asymptomatic at diagnosis, with 82.4% (n 103) of these patients remaining asymptomatic at the census date, a mean of years later (range, 0 19 years). The most common symptoms at presentation were fa- Figure 2. Prevalence (cases per million) of PBC in women older than 40 years: Australian born, U.K.-born migrants to Victoria, and U.K. women living in the United Kingdom. 1

4 August 2004 PREVALENCE OF PBC IN VICTORIA, AUSTRALIA 473 Figure 3. Clinical status of patients at census date. tigue (29.9%), followed by pruritus (20.4%), right upper quadrant discomfort (15.6%), and jaundice (4.7%). At the census date, 79.1% (n 197) of patients with PBC were noncirrhotic, 14.1% (n 35) had compensated cirrhosis, and 6.8% (n 17) had decompensated cirrhosis. Figure 3 represents the clinical status of the patients at the census date. Follow-up Study The 1991 study identified 84 patients with PBC. 4 Of these, full clinical follow-up information was only available for 38 (45%). Nine patients from the original cohort had undergone liver transplantation, and all were alive and well at the census date. These patients underwent transplantation a mean of years after presentation. A further 9 patients died from liver failure. A total of 18 patients (21% of the original cohort) are therefore known to have progressed to end-stage liver disease in the 11 years since Seven patients from the original study were alive at the census date. Five patients were well, one patient had compensated cirrhosis, and another patient had decompensated cirrhosis and was awaiting liver transplantation. Seven patients had died from non liver-related causes. Discussion We have documented the prevalence of PBC in Victoria to be 51.0 cases per million. This is more than 2.5 times the prevalence recorded in the 1991 Victorian study. 4 Even when the data are corrected for age (to take into account an aging Victorian population), the increase in prevalence since 1991 remains significant. The finding of an increasing prevalence of PBC supports the observations from Newcastle, England. 1 However, as with the Newcastle study, it is not possible to determine whether the large increase in cases reflects a true increase in the prevalence of the disease or whether it is due to other factors, such as a change in case-finding methods. By conducting a very limited search of AMApositive results in the 1991 study 4 (only 2 hospitals were searched and no new patients identified), some cases of PBC were almost certainly missed. Indeed, in contrast to the previous study, 24% of all patients in the current study were identified through the search of AMA results alone. It is also likely that increased screening for PBC (by increased physician awareness or more routine testing of liver biochemistry and AMA) played a part in the apparent increase in the prevalence of the disease. This is supported by the fact that one half of the current patient population were diagnosed while asymptomatic, compared with just 36% in the 1991 study. Another contributing factor to the increased prevalence of the disease is that the proportion of the population in the at-risk group (women older than 40 years) is increasing in Australia. As in most of the developed world, the mean age of the population in this country is increasing 12 and more people are in the age group in which PBC is most common. Furthermore, if the disease is more common in certain immigrant groups, those who came to Australia in the postwar immigration boom are now reaching the age at which they are at highest risk of having the disease. There is also the chance that the use of ursodeoxycholic acid and better management of patients who have chronic liver disease is improving survival and thus affecting prevalence. Victoria s ethnic diversity was apparent in the findings of this study, which identified patients with PBC from 36 different countries. Although overseas-born persons comprise 29% of the Victorian population, they make up 46% of our cohort with PBC; this apparent difference remained when the prevalence in the 2 groups was adjusted for age (P ). By itself, this finding suggests that Australian-born persons are in some way protected from the disease. When considering that 70% of the Australian-born population are of Anglo-Celtic descent 13 and thus of similar genetic background to the U.K. population, the striking difference in prevalence between Victoria s 51 cases per million and the U.K. rate of 251 cases per million 1 suggests this may be due to the effects of environmental rather than genetic factor(s). It is possible that this finding reflects differences in the efficiency of case finding in the 2 countries; however, the only major methodological difference between our study

5 474 SOOD ET AL. GASTROENTEROLOGY Vol. 127, No. 2 and contemporary studies in the United Kingdom is that the results of AMA testing are centralized in the United Kingdom, making access to AMA test results much easier. In our view, this difference would be unlikely to explain the marked difference in prevalence between Victoria and the United Kingdom. Importantly, as outlined below, when looking at the data from our study, in which the same case-finding methods were used to identify prevalence in all ethnic groups, when adjusted for age the British-born population in Victoria still has a markedly increased prevalence compared with the rest of the population. To our knowledge, the prevalence of PBC in migrant populations has not been specifically addressed in previous studies. Our findings in the British migrant population were of particular interest in this regard because there have been carefully conducted and repeated studies of PBC epidemiology in the United Kingdom with which it was possible to make comparisons and there is a large population of British-born migrants in the Victorian population. Our finding that the British-born population in Victoria had a lower prevalence of PBC in the at-risk population (women older than 40 years) than has been recorded in the United Kingdom (although still significantly higher than in the Victorian population) suggests that these people have been relatively protected from developing the disease once they migrated to Victoria. The prevalence of PBC in a number of other migrant groups was also significantly increased compared with the overall Victorian community. It has been suggested that PBC is more common in the north than in the south of Europe. 14 However, the prevalence in the Greek and Italian Victorian communities (both large expatriate groups) was as high as that documented in British-born persons. There have been few studies examining the prevalence of PBC in Asia There is evidence that the prevalence of this disease in Japan may be similar to that in the United Kingdom. 18 The disease has also been reported in other Asian populations, including a small number of cases in first-generation Asian immigrants living in the United Kingdom. 19 We identified 15 patients from Asia and Oceania, including several patients from China and Vietnam (both large migrant communities). However, the numbers are insufficient to speculate on the prevalence of the disease in these areas. In a review that addressed the possible geographic variation in the epidemiology of PBC, many studies were criticized for their lack of clarity of case definition and case finding methods, and accurate determination of study period, study area, and population at risk. 20 In an attempt to avoid these concerns, we strictly adhered to case criteria that are widely accepted for PBC. We also applied a specific census date in our defined study area. The accuracy of our data is supported by several observations. The rate of patients asymptomatic at presentation in this current study (50%) is comparable to the rates found in other studies. 1,3 Furthermore, it is encouraging that the proportion of patients who were diagnosed while living in metropolitan Melbourne (72%) is the same as the proportion of the Victorian population who reside in this city (72%). This finding makes it unlikely that confining our medical record search to the state s major hospitals, all of which are located within Melbourne, produced any major bias in the data. However, the lack of a central immunology database such as that found in Britain, combined with the incomplete physician response rate, suggests that in our study, as in most others, some cases of PBC are likely to have been missed. 21 It is accepted that, used in the optimal dose (13 15 mg kg 1 day 111,22 ), ursodeoxycholic acid has the ability to delay progression of PBC and the need for liver transplantation However, earlier studies conducted in Britain and the United States have recorded relatively low uptake of treatment, with only 30% 40% of patients receiving the drug. 26,27 Our results have shown a higher rate of treatment in Victoria, with more than two thirds of patients receiving the drug. However, it is of concern that many patients are not receiving an adequate dose. Increasing physician education may be necessary to improve the dose of the drug being administered to patients. In conclusion, although we have documented an apparent increase in the prevalence of PBC in Victoria since 1991, the prevalence remains among the lowest recorded anywhere in the world where thorough case-finding epidemiologic studies have been performed. There is considerable geographic variation in the reported prevalence of PBC; however, this is the first study that we are aware of in which migrant groups were observed to have a different prevalence than is found in both their native and adopted communities. We believe this finding provides further support to the suggestion that an environmental factor(s) contributes to the development of PBC. References 1. James OF, Bhopal R, Howel D, Gray J, Burt AD, Metcalf JV. Primary biliary cirrhosis once rare, now common in the United Kingdom? Hepatology 1999;30: Danielsson A, Boqvist L, Uddenfeldt P. Epidemiology of primary biliary cirrhosis in a defined rural population in the northern part of Sweden. Hepatology 1990;11: Lofgren J, Jarnerot G, Danielsson D, Hemdal I. Incidence and

6 August 2004 PREVALENCE OF PBC IN VICTORIA, AUSTRALIA 475 prevalence of primary biliary cirrhosis in a defined population in Sweden. Scand J Gastroenterol 1985;20: Watson RG, Angus PW, Dewar M, Goss B, Sewell RB, Smallwood RA. Low prevalence of primary biliary cirrhosis in Victoria, Australia. Melbourne Liver Group. Gut 1995;36: Myszor M, James OF. The epidemiology of primary biliary cirrhosis in north-east England: an increasingly common disease? QJMed 1990;75: Witt-Sullivan H, Heathcote J, Cauch K, Blendis L, Ghent C, Katz A, Milner R, Pappas SC, Rankin J, Wanless IR. The demography of primary biliary cirrhosis in Ontario, Canada. Hepatology 1990;12: Parikh-Patel A, Gold E, Mackay IR, Gershwin ME. The geoepidemiology of primary biliary cirrhosis: contrasts and comparisons with the spectrum of autoimmune diseases. Clin Immunol 1999; 91: Australian Bureau of Statistics. Basic community profile Melbourne. Canberra: Australian Bureau of Statistics, Australian Bureau of Statistics. Expanded community profile Victoria. Canberra: Australian Bureau of Statistics, Australian Bureau of Statistics. Australian demographic statistics June quarter. Canberra: Australian Bureau of Statistics, Lindor KD, Jorgensen R, Therneau TM, Smith C, Mahoney DW, Dickson ER. Comparison of three different doses of ursodeoxycholic acid in the treatment of primary biliary cirrhosis: a randomized trial (abstr). Hepatology 1997;26:A Australian Bureau of Statistics. Australian social trends 1999 population population projections: our ageing population. Canberra: Australian Bureau of Statistics, Price CA. Ethnic groups in Australia. Canberra: Australian Immigration Research Centre, Hamlyn AN, Macklon AF, James O. Primary biliary cirrhosis: geographical clustering and symptomatic onset seasonality. Gut 1983;24: Chan CY, Lee SD, Huang YS, Wu JC, Tsai YT, Tsay SH, Lo KJ. Primary biliary cirrhosis in Taiwan. J Gastroenterol Hepatol 1990; 5: Sasaki H, Inoue K, Higuchi K, Yasuyama T, Koyata H, Kuroki T, Yamamoto S, Ichida F. Primary biliary cirrhosis in Japan: national survey by the Subcommittee on Autoimmune Hepatitis. Gastroenterol Jpn 1985;20: Chainuvati T, Viranuvatti V, Damrongsak C, Hitanant S, Chearanai O. Primary biliary cirrhosis report of 3 cases in Thailand. J Med Assoc Thai 1973;56: Ohba K, Omagari K, Kinoshita H, Soda H, Masuda J, Hazama H, Tagawa M, Hata T, Nakamura H, Murata I, Kohno S. Primary biliary cirrhosis among atomic bomb survivors in Nagasaki, Japan. J Clin Epidemiol 2001;54: Anand AC, Elias E, Neuberger JM. End-stage primary biliary cirrhosis in a first generation migrant south Asian population. Eur J Gastroenterol Hepatol 1996;8: Metcalf J, James O. The geoepidemiology of primary biliary cirrhosis. Semin Liver Dis 1997;17: Triger DR. Primary biliary cirrhosis: is there an environmental contribution? J Gastroenterol Hepatol 1991;6: Van Hoogstraten HJ, De Smet MB, Renooij W, Breed JG, Engels LG, Den Ouden-Muller JW, Rijk MC, Smit AM, Zwertbroek R, Hop WC, van Berge Henegouwen GP, Schalm SW, van Buuren HR. A randomized trial in primary biliary cirrhosis comparing ursodeoxycholic acid in daily doses of either 10 mg/kg or 20 mg/kg. Dutch Multicentre PBC Study Group. Aliment Pharmacol Ther 1998;12: Gluud C, Christensen E. Ursodeoxycholic acid for primary biliary cirrhosis. Cochrane Database Syst Rev 2002:CD Poupon RE, Lindor KD, Cauch-Dudek K, Dickson ER, Poupon R, Heathcote EJ. Combined analysis of randomized controlled trials of ursodeoxycholic acid in primary biliary cirrhosis. Gastroenterology 1997;113: Goulis J, Leandro G, Burroughs AK. Randomised controlled trials of ursodeoxycholic-acid therapy for primary biliary cirrhosis: a meta-analysis. Lancet 1999;354: Prince M, Chetwynd A, Newman W, Metcalf JV, James OF. Survival and symptom progression in a geographically based cohort of patients with primary biliary cirrhosis: follow-up for up to 28 years. Gastroenterology 2002;123: Kim WR, Lindor KD, Locke GR III, Therneau TM, Homburger HA, Batts KP, Yawn BP, Petz JL, Melton LJ III, Dickson ER. Epidemiology and natural history of primary biliary cirrhosis in a US community. Gastroenterology 2000;119: Received January 29, Accepted April 22, Address requests for reprints to: Peter W. Angus, M.D., Department of Gastroenterology and Hepatology, Austin Hospital, Heidelberg, 3084, Victoria, Australia. peter.angus@austin.org.au; fax: (61)