SEMINARS IN LIVER DISEASE-VOL.
I 1, NO. 1, 1991
Primary Sclerosing Cholangitis: A Progressive Disease?
The natural history of patients with primary sclerosing cholangitis (PSC) is sometimes variable and often unpredictable, and studies concerning the frequency and rate of disease progression have led to conflicting results.' Reports emphasizing diminished survival and signs of disease progression in the majority of patients with PSC contrast with studies in which patients have been observed with quiescent disease and long-term survival. Consequently, defining the prognosis of PSC in individual patients is difficult, and disagreement continues as to when and if intervention with experimental therapies is appropriate. In this article, we will review published data regarding the natural history of patients with PSC, data that we interpret as consistent with the conclusion that PSC is most often a progressive disease that will, without intervention, frequently result in premature death from liver failure. We will also speculate on reasons for different results among various centers evaluating the long-term outcome of patients with PSC. Finally, we will describe methods to estimate the survival of an individual patient with PSC using multivariate Cox regression analysis. These survival models are important for appropriate patient selection and determination of optimal timing for liver transplantation.'
HISTORICAL PERSPECTIVE The difficulty in establishing the natural history of PSC relates in part to our lack of knowledge of the pathogenesis of PSC and the evolution of uniform criteria for its d i a g n 0 ~ i s . j . ~ PSC was considered a rare disease before the discovery and widespread application of endoscopic retrograde cholangiography (ERC) and percutaneous transhepatic cholangiography. It has been estimated that only 100 documented cases of PSC existed in the English literature before 1980.' Originally, the diagnosis was es-
From the Division of Gastroenterology. M o w Clinic crnd Mow) Foundation. Rochester, Minnesutc~ Reprint requests: Dr. Porayko. Mayo Clinic. 200 First Street, SW, Rochester, MN 55905
18
tablished with laparotomy by palpation of a fibrotic common bile duct.' The common bile duct was also biopsied for histologic confirmation and to exclude the presence of bile duct ~ a r c i n o m a .Symptomatic ~ patients were often treated by one of several types of bile duct surgery in an attempt to enhance biliary drainage. However, the lack of controlled clinical trials and the frequent occurrence of bacterial cholangitis make it impossible to assess what effect surgery had on the natural history of the disea~e.~.' In 1964, Holubitsky and McKenziel" formulated the following criteria for the diagnosis of PSC: (1) Absence of previous operative trauma to the biliary system; (2) absence of calculi in the gallbladder and common bile duct; (3) sclerosis and stenosis involving all or almost all of the extrahepatic ducts; and (4) exclusion of malignancy involviig the biliary system. Applying these criteria, the authors were able to identify only 25 acceptable cases of PSC out of the more than 100 reported in the literature at the time. They emphasized that the diagnosis of PSC carried a poor prognosis; in fact, two of the four cases re~ortedin~theirseries died of liver failure 8 and 10 years after diagnosis. The diagnostic criteria for PSC were modified several more times after the Holubitsky publication, frequently with more stringent requirements."-l3 Some investigators raised questions about the need for such rigid criteria, suggesting that many cases of PSC were being inappropriately eliminated from study.14 These differences in opinion led to confusion. Recognition of the relationship between PSC and inflammatory bowel disease (IBD) altered the perception of the disease and provided an additional diagnostic Identifying patients having both IBD and persistent liver enzyme abnormalities led to earlier diagnosis and increased detection of PSC.18 Patients with IBD became a target population for investigations of associated liver disease, and, as a result, a variety of hepatic histologic abnormalities associated with IBD were described. The term "pericholangitis" was often used as a disease-specific description (rather than a morphologic finding) to categorize such patients.22With more frequent use of ERC to evaluate hepatic abnormalities found in patients with IBD, many patients with pericholangitis on liver biopsy were indeed actually found to have PSC." This introduced the concept that chronic
'""
Copyright 0 1991 by Thieme Medical Publishers, Inc., 381 Park Avenue South, New York, NY 10016. All rights reserved
Downloaded by: NYU. Copyrighted material.
MICHAEL K. PORAYKO, M.D., NICHOLAS F. LaRUSSO, M.D., and RUSSELL H. WIESNER, M.D.
LaRUSSO, WIESNER
hepatitis associated with ulcerative colitis ("pericholangitis") probably represented one component in the disease spectrum of PSC.23Microscopic or small duct PSC (abnormalities on liver biopsy without cholangiographic changes) seems to be more appropriate terminology than pericholangitis and is now widely accepted. The natural history of patients with small duct PSC remains poorly defined. Although endoscopic and transhepatic cholangiography clearly had the greatest impact on the frequency of diagnosis of PSC, particularly in patients with IBD and abnormal liver tests, operative cholangiography had already revealed the characteristic radiologic features of PSC. Thorpe et all7 reported that the appearance of "beading" was diagnostic and emphasized the use of operative cholangiography to confirm the diagnosis. They also emphasized that cholangiography could demonstrate the extent of biliary involvement. Krieger and colleagues24added to these observations by describing decreased aborization of the biliary tree and a lack of marked dilation of the intrahepatic bile ducts proximal to extrahepatic biliary strictures. These authors also noted that patchy or isolated intrahepatic involvement of PSC could only be recognized radiographically. Endoscopic retrograde cholangiopancreatography (ERCP) became technically feasible in the early 1970s, and one of the first accounts of its usefulness in the diagnosis of PSC was reported by Elias et al in 1974.25 Gradually, PSC changed from being a surgically defined disease to that of a radiologically defined disease. Abnormalities involving the gallbladder, pancreatic duct, and isolated segments of the biliary tree were found in many patients with PSC.2h-2R The definition of PSC was expanded to include patients with isolated intrahepatic disease. The number of cases of PSC diagnosed per year increased two- to threefold, primarily because of the increased availability of ERCP.2yEndoscopic cholangiography also made possible the detection and follow-up of patients with asymptomatic disease. Even patients without typical cholestatic liver enzyme patterns (such as normal serum levels of alkaline phosphatase) could be diagnosed cholangiographically under appropriate clinical c i r c u m s t a n ~ e s . ~ ~ ~ " An important relationship between PSC and the development of cholangiocarcinoma was confirmed using serial cholangiography."." The development of bile duct carcinoma in patients with preexisting PSC suggested that malignancy may develop as a result of the chronic inflammatory condition. Therefore, what was once considered to be an exclusionary criteria for the diagnosis of PSC (cholangiocarcinoma) was now considered to be within the spectrum of complications developing during the course of PSC. The present criteria for diagnosis of PSC (Table 1) can now be easily understood considering the evolution of changes that have occurred over the last 25 years. Laparotomy is no longer required to make a diagnosis of PSC, and bile duct biopsy is reserved for cases of suspected cholangio~arcinoma.~ Currently, endoscopic or percutaneous cholangiography with adequate filling of the entire biliary tree is the most important diagnostic procedure to establish the diagnosis. Typical radiologic
19
TABLE 1. Criteria for Diagnosis of PSC I . Presence of typical cholangiographic abnormalities of PSC (involving bile ducts segmentally or extensively) 2. Compatible clinical, biochemical, and hepatic histologic findings (recognizing that they are frequently nonspecific) 3. Exclude the following in most instances: Biliary calculi (unless related to stasis) Biliary tract surgery (other than simple cholecystectomy) Congenital abnormalities of the biliary tract AIDS-associated cholangiopathy Ischemic stricturing Bile duct neoplasms (unless PSC previously established) Exposure to irritant chemicals (such as floxuridine. formalin) Evidence of another type of liver disease, such as primary biliary cirrhosis or chronic active hepatitis
features have been described in detail in other publicat i o n ~ . ~Alternative ~ . ~ ~ diagnostic imaging techniques such as hepatic ultrasonography and technetium-99m iminodiacetic acid hepatobiliary scintigraphy can be useful in detecting PSC when cholangiographic studies are incomplete or cannot be ~ t i l i z e d . ~Since ' clinical, biochemical, and hepatic histologic findings in PSC patients are, for the most part, nonspecific, cholangiography has become the gold standard for the diagnosis of this syndrome. " Given the growing number of conditions that have been recognized to cause sclerosing cholangitis (that is, secondary sclerosing cholangitis), it is important to exclude each of these etiologies carefully before making a diagnosis of the primary variety of sclerosing cholangitis, since the treatment and prognosis of these entities is so variable. Confusion can also result in patients with coexisting diseases that appear to develop as a result of PSC (such as cholangiocarcinoma or biliary stone disease). Patients with an unclear history of hepatobiliary surgery may also be a source of diagnostic confusion. Therefore, when investigating the natural history of PSC or when performing controlled clinical trials of medical therapy, adherence to strict diagnostic criteria is important so that patients with secondary sclerosing cholangitis mimicking PSC are not included. This historical overview emphasizes the dramatic differences in definition and diagnostic recognition that have occurred since the first description of PSC. These differences have obviously had an enormous effect on our ability to study the natural history of this syndrome. In fact, many studies prior to the use of ERCP remain difficult to interpret, since all patients studied did not meet the strict diagnostic criteria that are applied today.
NATURAL HISTORY-THE
PROBLEM
Based on prospective follow-up of large numbers of patients with PSC, there appears to be growing evidence that PSC is frequently a progressive syndrome leading to significant complications related to chronic c h o l e ~ t a s i s . ~ ~ However, not all investigators have found such relentless disease progression and evidence for diminished survival in PSC. Some have raised the question of whether subsets of PSC patients may have an improved prognosis,
Downloaded by: NYU. Copyrighted material.
PRIMARY SCLEROSING CHOLANGITIS-PORAYKO,
20
SEMINIARS IN LIVER DISEASE-VOLUME
I I , NUMBER 1 , 1991
disease (patients who were asymptomatic) had better survival than those patients who were symptomatic at diagnosis. Of the 70 patients enrolled in the D-penicillamine study, serial prospective follow-up showed progression of liver disease based on clinical, biochemical, and histologic data had occurred in 80% within 3 years of study entry (Fig. 2). In another large study, Martin et a144identified 178 patients with PSC who had been treated and followed at the Lahey Clinic Medical Center over a period of 40 years. These authors emphasized the number and variety of palliative surgical and radiologic procedures that had been performed in their patients. These included 233 biliary tract operations, 15 portosystemic shunts, and 16 EVIDENCE SUGGESTING PSC IS orthotopic liver transplants. Despite initial improvement A PROGRESSIVE DISEASE of symptoms and signs of PSC liver disease in a majority of their patients after operation, the overall mortality reInvestigators at the Mayo Clinic have published mained high at 58% during a median time from diagnosis several studies on the natural history of PSC, suggesting to death of 10.4 years. Liver failure, bleeding, and sepsis disease progression and a poor prognosis in most paaccounted for 75 of 103 (73%) of these deaths. An adtients. The larger of these studies was recently reported ditional 13 of 47 survivors had progressed to the point by Wiesner et al" after reviewing and analyzing data colof being considered for liver transplantation. Only 2 of lected on 174 patients. It must be noted that 70 of these their 28 asymptomatic patients died of liver disease durpatients took part in a controlled clinical trial evaluating ing follow-up. Interestingly, there were no significant the therapeutic effects of D-penicillamine, a drug that differences in survival among patient subgroups based was later found to have no effect on the course of this on the presence or absence of IBD, or the distribution of disease.43The study was initiated and survival analysis biliary -tract disease (such as extrahepatic, intrahepatic, began at the time when all criteria for diagnosis were or diffuse disease). fulfilled for the first time at the Mayo Clinic. This served A poor prognosis for patients with PSC was further as a uniform and easily identifiable starting point from documented in studies from several other institutions. In which to assess the natural history. During the study, 54 an early report (1980) from the Royal Free Hospital in (31%) patients died as a result of underlying liver disease England, it was concluded that PSC carried a variable or the development of cholangiocarcinoma; an additional 17 (10%) were referred for liver transplantation. The prognosis; yet 38% of their patients died after a mean of 7 years from the onset of symptoms.45In a later study mean duration of follow-up was 6 . 0 years (range, 2.7 to (1988) from the same hospital, the clinical courses of 70 15.5 years) with no patients lost to follow-up. Median patients were reviewed and evaluated after a mean folsurvival from the time of Mayo Clinic diagnosis was low-up of 8.3 years; clinical deterioration occurred in 11.9 years and both symptomatic and asymptomatic pa53% of symptomatic patients and 59% of asymptomatic tients had reduced survival when compared to an age, patients. Twenty-eight patients developed hepatic failure sex, and race matched United States population (Fig. 1). and 22 eventually died within a mean survival time from As may be expected, those patients with earlier stage diagnosis to death of approximately 7 years; 10-year survival was estimated to be 55%.4h Lebovics et from the Mount Sinai Medical Matched U.S. population1 100 7 Center in New York, retrospectively analyzed the records of all patients with PSC who had been identified by '-l Asymptomatic a single physician within one institution. A total of 38 patients were followed for a mean duration of 75 months. Seventeen (45%) patients had a poor outcome Symptomatic defined as death, hepatic failure, or referral for liver transplantation. Eleven patients (29%) died after a mean --'+ follow-up of 103 months; nine deaths were related to liver failure. The Kaplan-Meier estimated 10-year survival was near 60%. The authors emphasized that a lack of symptoms at presentation did not portend improved survival or a prolonged course. Indeed, five of nine pa0 tients who had been asymptomatic at the time of diag0 1 2 3 4 5 6 7 8 nosis had a poor outcome,and four ultimately died, on Time, years average, approximately 9 years after diagnosis. FIG. 1. Kaplan-Meier estimated survival curves of asymptomatic and symptomatic PSC patients. For comparison, the Taub et from the Cleveland Clinic retrospecsurvival curve of the United States population, matched for tively reviewed records from 83 patients with PSC folage, sex, and race to the asymptomatic population, is also lowed for a mean duration of 48 months. Twenty-nine shown. For difference in survival: asymptomatic vs control, p percent of these patients either died or required liver
I 1, NO. 1, 1991
Primary Sclerosing Cholangitis: A Progressive Disease?
The natural history of patients with primary sclerosing cholangitis (PSC) is sometimes variable and often unpredictable, and studies concerning the frequency and rate of disease progression have led to conflicting results.' Reports emphasizing diminished survival and signs of disease progression in the majority of patients with PSC contrast with studies in which patients have been observed with quiescent disease and long-term survival. Consequently, defining the prognosis of PSC in individual patients is difficult, and disagreement continues as to when and if intervention with experimental therapies is appropriate. In this article, we will review published data regarding the natural history of patients with PSC, data that we interpret as consistent with the conclusion that PSC is most often a progressive disease that will, without intervention, frequently result in premature death from liver failure. We will also speculate on reasons for different results among various centers evaluating the long-term outcome of patients with PSC. Finally, we will describe methods to estimate the survival of an individual patient with PSC using multivariate Cox regression analysis. These survival models are important for appropriate patient selection and determination of optimal timing for liver transplantation.'
HISTORICAL PERSPECTIVE The difficulty in establishing the natural history of PSC relates in part to our lack of knowledge of the pathogenesis of PSC and the evolution of uniform criteria for its d i a g n 0 ~ i s . j . ~ PSC was considered a rare disease before the discovery and widespread application of endoscopic retrograde cholangiography (ERC) and percutaneous transhepatic cholangiography. It has been estimated that only 100 documented cases of PSC existed in the English literature before 1980.' Originally, the diagnosis was es-
From the Division of Gastroenterology. M o w Clinic crnd Mow) Foundation. Rochester, Minnesutc~ Reprint requests: Dr. Porayko. Mayo Clinic. 200 First Street, SW, Rochester, MN 55905
18
tablished with laparotomy by palpation of a fibrotic common bile duct.' The common bile duct was also biopsied for histologic confirmation and to exclude the presence of bile duct ~ a r c i n o m a .Symptomatic ~ patients were often treated by one of several types of bile duct surgery in an attempt to enhance biliary drainage. However, the lack of controlled clinical trials and the frequent occurrence of bacterial cholangitis make it impossible to assess what effect surgery had on the natural history of the disea~e.~.' In 1964, Holubitsky and McKenziel" formulated the following criteria for the diagnosis of PSC: (1) Absence of previous operative trauma to the biliary system; (2) absence of calculi in the gallbladder and common bile duct; (3) sclerosis and stenosis involving all or almost all of the extrahepatic ducts; and (4) exclusion of malignancy involviig the biliary system. Applying these criteria, the authors were able to identify only 25 acceptable cases of PSC out of the more than 100 reported in the literature at the time. They emphasized that the diagnosis of PSC carried a poor prognosis; in fact, two of the four cases re~ortedin~theirseries died of liver failure 8 and 10 years after diagnosis. The diagnostic criteria for PSC were modified several more times after the Holubitsky publication, frequently with more stringent requirements."-l3 Some investigators raised questions about the need for such rigid criteria, suggesting that many cases of PSC were being inappropriately eliminated from study.14 These differences in opinion led to confusion. Recognition of the relationship between PSC and inflammatory bowel disease (IBD) altered the perception of the disease and provided an additional diagnostic Identifying patients having both IBD and persistent liver enzyme abnormalities led to earlier diagnosis and increased detection of PSC.18 Patients with IBD became a target population for investigations of associated liver disease, and, as a result, a variety of hepatic histologic abnormalities associated with IBD were described. The term "pericholangitis" was often used as a disease-specific description (rather than a morphologic finding) to categorize such patients.22With more frequent use of ERC to evaluate hepatic abnormalities found in patients with IBD, many patients with pericholangitis on liver biopsy were indeed actually found to have PSC." This introduced the concept that chronic
'""
Copyright 0 1991 by Thieme Medical Publishers, Inc., 381 Park Avenue South, New York, NY 10016. All rights reserved
Downloaded by: NYU. Copyrighted material.
MICHAEL K. PORAYKO, M.D., NICHOLAS F. LaRUSSO, M.D., and RUSSELL H. WIESNER, M.D.
LaRUSSO, WIESNER
hepatitis associated with ulcerative colitis ("pericholangitis") probably represented one component in the disease spectrum of PSC.23Microscopic or small duct PSC (abnormalities on liver biopsy without cholangiographic changes) seems to be more appropriate terminology than pericholangitis and is now widely accepted. The natural history of patients with small duct PSC remains poorly defined. Although endoscopic and transhepatic cholangiography clearly had the greatest impact on the frequency of diagnosis of PSC, particularly in patients with IBD and abnormal liver tests, operative cholangiography had already revealed the characteristic radiologic features of PSC. Thorpe et all7 reported that the appearance of "beading" was diagnostic and emphasized the use of operative cholangiography to confirm the diagnosis. They also emphasized that cholangiography could demonstrate the extent of biliary involvement. Krieger and colleagues24added to these observations by describing decreased aborization of the biliary tree and a lack of marked dilation of the intrahepatic bile ducts proximal to extrahepatic biliary strictures. These authors also noted that patchy or isolated intrahepatic involvement of PSC could only be recognized radiographically. Endoscopic retrograde cholangiopancreatography (ERCP) became technically feasible in the early 1970s, and one of the first accounts of its usefulness in the diagnosis of PSC was reported by Elias et al in 1974.25 Gradually, PSC changed from being a surgically defined disease to that of a radiologically defined disease. Abnormalities involving the gallbladder, pancreatic duct, and isolated segments of the biliary tree were found in many patients with PSC.2h-2R The definition of PSC was expanded to include patients with isolated intrahepatic disease. The number of cases of PSC diagnosed per year increased two- to threefold, primarily because of the increased availability of ERCP.2yEndoscopic cholangiography also made possible the detection and follow-up of patients with asymptomatic disease. Even patients without typical cholestatic liver enzyme patterns (such as normal serum levels of alkaline phosphatase) could be diagnosed cholangiographically under appropriate clinical c i r c u m s t a n ~ e s . ~ ~ ~ " An important relationship between PSC and the development of cholangiocarcinoma was confirmed using serial cholangiography."." The development of bile duct carcinoma in patients with preexisting PSC suggested that malignancy may develop as a result of the chronic inflammatory condition. Therefore, what was once considered to be an exclusionary criteria for the diagnosis of PSC (cholangiocarcinoma) was now considered to be within the spectrum of complications developing during the course of PSC. The present criteria for diagnosis of PSC (Table 1) can now be easily understood considering the evolution of changes that have occurred over the last 25 years. Laparotomy is no longer required to make a diagnosis of PSC, and bile duct biopsy is reserved for cases of suspected cholangio~arcinoma.~ Currently, endoscopic or percutaneous cholangiography with adequate filling of the entire biliary tree is the most important diagnostic procedure to establish the diagnosis. Typical radiologic
19
TABLE 1. Criteria for Diagnosis of PSC I . Presence of typical cholangiographic abnormalities of PSC (involving bile ducts segmentally or extensively) 2. Compatible clinical, biochemical, and hepatic histologic findings (recognizing that they are frequently nonspecific) 3. Exclude the following in most instances: Biliary calculi (unless related to stasis) Biliary tract surgery (other than simple cholecystectomy) Congenital abnormalities of the biliary tract AIDS-associated cholangiopathy Ischemic stricturing Bile duct neoplasms (unless PSC previously established) Exposure to irritant chemicals (such as floxuridine. formalin) Evidence of another type of liver disease, such as primary biliary cirrhosis or chronic active hepatitis
features have been described in detail in other publicat i o n ~ . ~Alternative ~ . ~ ~ diagnostic imaging techniques such as hepatic ultrasonography and technetium-99m iminodiacetic acid hepatobiliary scintigraphy can be useful in detecting PSC when cholangiographic studies are incomplete or cannot be ~ t i l i z e d . ~Since ' clinical, biochemical, and hepatic histologic findings in PSC patients are, for the most part, nonspecific, cholangiography has become the gold standard for the diagnosis of this syndrome. " Given the growing number of conditions that have been recognized to cause sclerosing cholangitis (that is, secondary sclerosing cholangitis), it is important to exclude each of these etiologies carefully before making a diagnosis of the primary variety of sclerosing cholangitis, since the treatment and prognosis of these entities is so variable. Confusion can also result in patients with coexisting diseases that appear to develop as a result of PSC (such as cholangiocarcinoma or biliary stone disease). Patients with an unclear history of hepatobiliary surgery may also be a source of diagnostic confusion. Therefore, when investigating the natural history of PSC or when performing controlled clinical trials of medical therapy, adherence to strict diagnostic criteria is important so that patients with secondary sclerosing cholangitis mimicking PSC are not included. This historical overview emphasizes the dramatic differences in definition and diagnostic recognition that have occurred since the first description of PSC. These differences have obviously had an enormous effect on our ability to study the natural history of this syndrome. In fact, many studies prior to the use of ERCP remain difficult to interpret, since all patients studied did not meet the strict diagnostic criteria that are applied today.
NATURAL HISTORY-THE
PROBLEM
Based on prospective follow-up of large numbers of patients with PSC, there appears to be growing evidence that PSC is frequently a progressive syndrome leading to significant complications related to chronic c h o l e ~ t a s i s . ~ ~ However, not all investigators have found such relentless disease progression and evidence for diminished survival in PSC. Some have raised the question of whether subsets of PSC patients may have an improved prognosis,
Downloaded by: NYU. Copyrighted material.
PRIMARY SCLEROSING CHOLANGITIS-PORAYKO,
20
SEMINIARS IN LIVER DISEASE-VOLUME
I I , NUMBER 1 , 1991
disease (patients who were asymptomatic) had better survival than those patients who were symptomatic at diagnosis. Of the 70 patients enrolled in the D-penicillamine study, serial prospective follow-up showed progression of liver disease based on clinical, biochemical, and histologic data had occurred in 80% within 3 years of study entry (Fig. 2). In another large study, Martin et a144identified 178 patients with PSC who had been treated and followed at the Lahey Clinic Medical Center over a period of 40 years. These authors emphasized the number and variety of palliative surgical and radiologic procedures that had been performed in their patients. These included 233 biliary tract operations, 15 portosystemic shunts, and 16 EVIDENCE SUGGESTING PSC IS orthotopic liver transplants. Despite initial improvement A PROGRESSIVE DISEASE of symptoms and signs of PSC liver disease in a majority of their patients after operation, the overall mortality reInvestigators at the Mayo Clinic have published mained high at 58% during a median time from diagnosis several studies on the natural history of PSC, suggesting to death of 10.4 years. Liver failure, bleeding, and sepsis disease progression and a poor prognosis in most paaccounted for 75 of 103 (73%) of these deaths. An adtients. The larger of these studies was recently reported ditional 13 of 47 survivors had progressed to the point by Wiesner et al" after reviewing and analyzing data colof being considered for liver transplantation. Only 2 of lected on 174 patients. It must be noted that 70 of these their 28 asymptomatic patients died of liver disease durpatients took part in a controlled clinical trial evaluating ing follow-up. Interestingly, there were no significant the therapeutic effects of D-penicillamine, a drug that differences in survival among patient subgroups based was later found to have no effect on the course of this on the presence or absence of IBD, or the distribution of disease.43The study was initiated and survival analysis biliary -tract disease (such as extrahepatic, intrahepatic, began at the time when all criteria for diagnosis were or diffuse disease). fulfilled for the first time at the Mayo Clinic. This served A poor prognosis for patients with PSC was further as a uniform and easily identifiable starting point from documented in studies from several other institutions. In which to assess the natural history. During the study, 54 an early report (1980) from the Royal Free Hospital in (31%) patients died as a result of underlying liver disease England, it was concluded that PSC carried a variable or the development of cholangiocarcinoma; an additional 17 (10%) were referred for liver transplantation. The prognosis; yet 38% of their patients died after a mean of 7 years from the onset of symptoms.45In a later study mean duration of follow-up was 6 . 0 years (range, 2.7 to (1988) from the same hospital, the clinical courses of 70 15.5 years) with no patients lost to follow-up. Median patients were reviewed and evaluated after a mean folsurvival from the time of Mayo Clinic diagnosis was low-up of 8.3 years; clinical deterioration occurred in 11.9 years and both symptomatic and asymptomatic pa53% of symptomatic patients and 59% of asymptomatic tients had reduced survival when compared to an age, patients. Twenty-eight patients developed hepatic failure sex, and race matched United States population (Fig. 1). and 22 eventually died within a mean survival time from As may be expected, those patients with earlier stage diagnosis to death of approximately 7 years; 10-year survival was estimated to be 55%.4h Lebovics et from the Mount Sinai Medical Matched U.S. population1 100 7 Center in New York, retrospectively analyzed the records of all patients with PSC who had been identified by '-l Asymptomatic a single physician within one institution. A total of 38 patients were followed for a mean duration of 75 months. Seventeen (45%) patients had a poor outcome Symptomatic defined as death, hepatic failure, or referral for liver transplantation. Eleven patients (29%) died after a mean --'+ follow-up of 103 months; nine deaths were related to liver failure. The Kaplan-Meier estimated 10-year survival was near 60%. The authors emphasized that a lack of symptoms at presentation did not portend improved survival or a prolonged course. Indeed, five of nine pa0 tients who had been asymptomatic at the time of diag0 1 2 3 4 5 6 7 8 nosis had a poor outcome,and four ultimately died, on Time, years average, approximately 9 years after diagnosis. FIG. 1. Kaplan-Meier estimated survival curves of asymptomatic and symptomatic PSC patients. For comparison, the Taub et from the Cleveland Clinic retrospecsurvival curve of the United States population, matched for tively reviewed records from 83 patients with PSC folage, sex, and race to the asymptomatic population, is also lowed for a mean duration of 48 months. Twenty-nine shown. For difference in survival: asymptomatic vs control, p percent of these patients either died or required liver
