Interactions Between Chronic Liver Disease and Inflammatory Bowel Disease

Liver disease was first recognized in association with idiopathic inflammatory bowel disease (IBD) over a century ago (1). The introduction of better diagnostic tools such as colonoscopy and endoscopic retrograde cholangiopancreatography during the last 25 years has strengthened the association between IBD and chronic hepatobiliary dysfunction, especially primary sclerosing cholangitis (PSC). IBD [ulcerative colitis (UC) and Crohn's disease (CD)] and PSC are both chronic conditions that can result in substantial morbidity and mortality and require coordinated medical and surgical interventions (2,–7). Since the conditions frequently coexist and the presence of one can alter the natural history or the management of the other, the study of the interaction between these illnesses is of great importance. We review the epidemiology, clinical features, treatment, and prognosis of hepatobiliary diseases associated with IBD as well as the unique clinical features of IBD associated with PSC. Hepatobiliary disease, once thought to be a rare extraintestinal manifestation of UC and CD, has become increasingly recognized in recent decades (3,–9). Measurement of the true prevalence of liver dysfunction in IBD remains unclear because of imprecise definitions, differences in diagnostic aggressiveness, and differences in IBD cohorts studied. For example, if “hepatobiliary disease” is defined as a single abnormal hepatic biochemical parameter, even if transiently elevated, the prevalence of “hepatobiliary disease associated with IBD” may be higher than 50%. On the other hand, persistently elevated hepatic biochemistries are less common, ranging from 3 to 50%, with an average of 12%, across numerous studies (10,–23) (Table 1). Estimates from population-based studies, which theoretically better represent the true spectrum of a given illness, range from 5 to 16% (11,13,–14,19,20,23), whereas studies from referral centers provide higher estimates (10,12,15,16). Several series detected a high prevalence of histologic hepatobiliary abnormalities in consecutive IBD patients undergoing intestinal resection; however, such series are prone to selection bias in that only patients with IBD severe enough to warrant resection are being studied (18,–22). Moreover, transient hepatobiliary abnormalities that seemingly have no impact on long-term prognosis have been noted in IBD patients during flares of acute colitis (19), and some of the pathologic findings in the surgical series may be the histologic correlates of such colitis-related rises in hepatic biochemistries. In summary, population-based estimates of the prevalence of hepatobiliary dysfunction of 4.5-15.6% likely represent minimal estimates, whereas studies of operative liver biopsies are likely to provide maximal estimates of prevalence. Prevalence of hepatobiliary disease and primary sclerosing cholangitis (PSC) in inflammatory bowel disease (IBD) Prevalence of hepatobiliary disease and primary sclerosing cholangitis (PSC) in inflammatory bowel disease (IBD) The most common hepatobiliary complications of IBD, in order of frequency and importance, are PSC, small duct PSC (“pericholangitis”), cholangiocarcinoma (carcinoma of the bile ducts), fatty infiltration of the liver, cholelithiasis, autoimmune hepatitis, and primary biliary cirrhosis (PBC). PSC is a chronic, cholestatic liver disease characterized by inflammation and fibrosis of the bile ducts (3,–7). This idiopathic disorder, with features of immune and genetic derangements, is strongly associated with IBD, especially UC, and may progress to biliary cirrhosis and hepatic failure. The only effective treatment at present is orthotopic liver transplantation (OLT) (24,25). The prevalence of PSC among patients with UC ranges from 2 to 26% (10,–23) (Table 1). Again, the wide variance in prevalence figures can be explained by definitions for both PSC and IBD, diagnostic aggressiveness, and the nature of the UC cohort studied. For example, the 26% prevalence of PSC among hospitalized UC patients at the Rikshospitalet in Oslo, Norway, between 1986 and 1991 is likely due to the fact that this facility is a national referral center for hepatology (17). Population-based studies from Scandinavia are fairly consistent in their estimates; between 2 and 7.6% of UC patients develop PSC (13,14,19,20,23). In general, the prevalence of PSC in patients with CD is much lower (˜1%), although a recent Finnish study is a notable exception (23). Owing to the extreme rarity of disease, very few direct population-based measurements of PSC prevalence have been performed. In Spain, where the prevalence of IBD is much less than that in the United States, a multicenter group estimated a prevalence of 2.24 cases of PSC per 10 × 106 persons in 1988 (26). A multicenter Swedish study measured a prevalence of PSC of 3.7% of UC patients in an area where the prevalence of UC was 170 cases per 100,000 persons in 1988, yielding a rough prevalence of 5 cases of PSC per 100,000 (13). Using the prevalence of UC in Olmsted County, MN, U.S.A., as a guide (150 cases per 100,000 persons in 1991; E. V. Loftus, Jr., unpublished data), the prevalence of PSC in the United States is estimated to be between 3 and 11 cases per 100,000 persons. The true prevalence of PSC may actually be higher since a substantial minority of PSC patients do not have associated IBD (27,–43) (see below). Between 60 and 70% of PSC patients are males (27,–43). Diagnosis of the disease is most frequently made in the fourth decade of life, although studies examining first onset of hepatic biochemical abnormalities suggest the true onset of disease may be, on average, a decade earlier (19). A number of recent pediatric case series of PSC demonstrate that the onset of disease may be as early as 18 months of age (36,44). In Western countries, between 60 and 100% of PSC cases are associated with IBD (27,–38,41,–43) (Table 2; also see below). Interestingly, the normally increased male/female ratio was reversed in one study of PSC patients without IBD, indirectly suggesting that PSC may be a heterogeneous group of diseases, but this remains to be confirmed (32). Other diseases that have been associated with PSC in case reports and small case series include chronic pancreatitis (23,45) (see below), autoimmune hepatitis (46,–48) (see below), celiac disease (49,–52), dermatitis herpetiformis (53), diabetes mellitus (54,55), autoimmune hemolytic anemia (56,57), glomerulonephritis (58,–60), polymyositis (61), and a variety of thyroid disorders (62,–64). It is of interest to note that most if not all of these associated disorders are thought to have autoimmune mechanisms. The finding of familial cases of PSC suggests that genetic factors may play a role in pathogenesis (65). Prevalence of inflammatory bowel disease (IBD) in primary sclerosing cholangitis (PSC) Prevalence of inflammatory bowel disease (IBD) in primary sclerosing cholangitis (PSC) Two case control studies have noted an epidemiologic association between the development of PSC and an environmental factor, tobacco smoking (66,67). As in UC (68), there appears to be an inverse relationship between smoking and illness; in other words, there are a paucity of current smokers and an excess of never smokers among PSC patients when compared with controls as well as with the general population (66,67). The magnitude of the inverse relationship appeared stronger in the U.S. study (66), where smoking is less prevalent, compared with that in the study from the Netherlands (67) (estimated odds ratios of PSC in current smokers relative to never smokers, 0.13 and 0.37, respectively). Although much of the association between smoking and PSC can be explained by the strong association between PSC and UC, both studies examined the association between smoking and illness in the subgroup of PSC patients without IBD (66,67). The point estimates of the odds ratios of illness in current smokers still suggested an inverse relationship; however, neither study reached statistical significance at the 5% level (66,67). The mechanism by which smoking might exert a protective effect against the development of UC or PSC is unclear, although a number of alterations in rectal blood flow, intestinal permeability, colonic mucus production, colonic mucosal eicosanoids, and cell-mediated and humoral immunity have been noted (69,–76). Moreover, two randomized clinical trials have shown that transdermal nicotine for treatment of active UC resulted in clinical improvement (77,78). Whether nicotine would be beneficial for the treatment of PSC is unknown. The cause of PSC remains undetermined. A host of hypothetical causal mechanisms have been proposed, chronic bile or other small to the bile and chronic with hepatitis or or For a variety of most of these have been to the cause of PSC has on factors because of the of familial cases and the association of PSC with other autoimmune with a variety of autoimmune disorders are to frequently which are in immune The and to be strongly associated with PSC to of PSC patients may the compared with of controls the is in of PSC patients compared with of controls In may be associated with PSC, although are in PSC patients the that play a role in Several features of PSC suggest that immune factors are operative in the development of infiltration and of the biliary increased of of the and elevated of immune and their abnormal by the to a colonic in patients with UC but not in controls to biliary and colonic against the and have been noted frequently in patients with PSC and UC than in patients with UC The finding of in a of PSC and UC patients with and without the associated not only to in humoral immunity as a causal but also a to a common mechanism in or other of these in a when the are in and are The prevalence of in PSC and without ranges between and with an average of across studies of the in prevalence appears to be to The to these do not to with disease the against which are remains and if role the play is unclear A high prevalence of is in patients with UC and without the prevalence of in CD patients is much lower studies have differences in between PSC and UC in in PSC UC increased for PSC at higher of and of treatment of in UC only but not PSC patients A number of have detected a higher than prevalence of in of both PSC and UC suggesting that may represent a genetic of disease Other have been to this and is not if this differences in or differences in and genetic causal mechanisms immune factors likely play a role in PSC pathogenesis as A in especially with a in the has been noted The abnormal ratio has been detected in from colonic of patients with PSC in the are increased in PSC but the ratio in the has not from that of controls of and blood group on biliary and increased of have been number of patients are are being of abnormal hepatic biochemistries in the clinical in an IBD such as and are common (3,–7). and frequently (3,–7). acute cholangitis may be the of especially in patients have or of the biliary (3,–7). the of is by features of such as or (3,–7). The is abnormal in most and are the most common noted (3,–7). are as are and (3,–7). The of PSC patients have an elevated although rare have been noted The level is elevated two or the of In PSC may present with a than a biochemical (36,44). The is frequently elevated in patients on (3,–7). in may represent disease, development of a biliary or cholangiocarcinoma (3,–7). in and in the of PSC to hepatic (3,–7). Liver which is abnormal in PSC, may in the but the histologic abnormalities are The histologic features include hepatitis with or without fibrosis and and with The finding of cholangitis is but is on Liver appears to be most for (Table of primary sclerosing cholangitis of primary sclerosing cholangitis or remains the for of PSC The most common findings are of the bile ducts with or ducts to a can be in the or or both in the of may with and the histologic abnormalities but with may have small duct PSC or some have in the (see below). The prevalence of abnormalities of chronic pancreatitis in PSC with an average of across studies case series of PSC a with hepatic and years of (3,–7). As PSC recognized and cases to be in an and from of (27,–43). A recent of PSC patients from the an of at 5 years and at 10 years the disease is in the patients and patients liver transplantation A recent multicenter study from an of although patients at a better than at years The most common of in patients to be cholangiocarcinoma and hepatic failure. Several of PSC on have been and of these have that and histologic of PSC are of of PSC has been can be two at complications of PSC and at the of can be to chronic and to PSC. may result in and with The cause of is Although an of bile not to the patients with the and The and to be the that may be due to an in which are increased in A host of other have been to and in PSC is due to bile in the small common include chronic pancreatitis and celiac disease, both of which may be associated with PSC the clinical significance of of and was The prevalence of such in PSC patients undergoing for may be Whether patients with remains disease, not to is also frequently in patients with disease and is not that are to PSC include the development of biliary acute or and cholangiocarcinoma (3,–7). cholangitis most frequently in patients have of the biliary patients may from long-term since cholangitis may of PSC. or in to of PSC patients and can be of endoscopic or or are that may to of biliary cirrhosis be from bile duct be can be or with frequently with for be especially for with severe disease or because may with liver transplantation in the is one of the most complications of PSC Although the once thought to be because of is that bile duct can in the of PSC. studies of PSC cohorts have a between and over an average of years with associated IBD may be at higher Diagnosis is in the of PSC with bile duct frequently have of bile but this finding is not such as and may be in suggesting the presence of cholangiocarcinoma but less in the presence of PSC The prognosis is even in patients to have at The of patients with cholangiocarcinoma is 5 or may be of in cases or endoscopic biliary may provide at disease of PSC have been have not been in a randomized however, not in an with and there is that the disease in PSC clinical trials of and have not clinical although there appeared to be hepatic biochemical in patients on and Several trials have examined the of in general, biochemical have but clinical have not are but trials have not been is the treatment of for PSC Several centers have in excess of patients at the between and was V. Loftus, Jr., unpublished PSC patients may be at increased of chronic as well as biliary Whether these represent PSC, bile duct or cholangitis due to the bile duct is unclear years the appeared in the IBD to the chronic cholestatic hepatitis in IBD As diagnostic that of these patients the features of PSC a without PSC and these patients still as a study that some patients with duct PSC years Moreover, the age and of patients to of PSC appears that may be a of PSC the small bile ducts only Although the clinical of this group of patients is not as well characterized as with duct PSC, appears that such patients are at for biliary cirrhosis and cholangiocarcinoma In a recent study of 18 small duct PSC patients from Oslo, there no but of may have been (17). with UC are at increased of bile duct a case series and review of UC patients with PSC, a (see are reports of bile duct in CD and only a minority noted to have PSC Whether cholangiocarcinoma in IBD patients in the of PSC is The of PSC is to in the presence of bile duct Liver from the of can be in the presence of PSC cases of cholangiocarcinoma in IBD patients without PSC do not liver We strongly that IBD patients are at increased of bile duct because of PSC or small infiltration of the liver was in early reports of liver disease in IBD patients but has in recent due to A recent population-based study of liver disease among UC patients in that of UC patients with persistently elevated hepatic biochemistries thought to have fatty infiltration of the liver, although this was in only of the group (19). The of fatty infiltration may be to the of but are with CD are at increased for studies have a prevalence of to much higher than that of the general population This increased prevalence has been in patients with and has been to of the The which is the primary of bile is frequently or in The of the bile to and other than inflammation or resection may be A recent of CD patients that of CD, and factors for cholelithiasis, but of inflammation or resection was not in CD patients appears to be UC patients with have no elevated of but the in patients undergoing with may be as high as hepatitis, as autoimmune chronic active hepatitis, is an idiopathic inflammation of the liver characterized by hepatitis with hepatitis as and autoimmune hepatitis in of UC the frequency of UC in autoimmune hepatitis may be as high as 16% compared with autoimmune hepatitis patients without UC, patients with both conditions less well to and progress to cirrhosis frequently This may be due in to the finding that to of patients with autoimmune hepatitis and UC have features of PSC PSC and autoimmune hepatitis histologic features, and can be to the two conditions on liver The finding of cholangitis is of PSC be in patients with autoimmune hepatitis and associated UC if have cholestatic features, of cholangitis on liver or a to is a chronic cholestatic liver disease characterized by and of small bile Several case reports of UC and however, the association remains and is not if has frequently than would be by PSC patients have been or be with IBD, UC The prevalence of IBD in PSC ranges from to 100% in Western countries, with an average of (27,–38,41,–43) (Table In and where IBD is less prevalent, PSC patients are to have IBD reports from Scandinavia suggested a very high prevalence of IBD in PSC and was that IBD was in PSC numerous reports that even with diagnostic of the colonoscopy with a of UC or CD not be made in a substantial minority of patients Moreover, recent population-based studies from have detected PSC without IBD in of all PSC patients The prevalence of UC in PSC ranges from to (27,–43) (Table examined and the of patients to have colitis to the The prevalence of CD in PSC between and with an average of (27,–43) (Table patients have colitis or but not the IBD in PSC patients is or (Table It is not if the variance in prevalence of IBD is to a true in prevalence or a of IBD is to from most reports of IBD and PSC which or histologic features to the of The of patients are with IBD PSC (27,–43). the can be made and PSC may be years IBD or even In a of patients at the between and 1988, the of UC at the of PSC was 10 but PSC was from years to years the UC few studies have examined the clinical features and of IBD in patients with PSC. detected IBD in of PSC patients colonoscopy and but of patients noted to have rectal but to suggest CD, such as disease, colonic or on the such patients are as UC with rectal than Crohn's of the differences in the prevalence in IBD may be due to differences in of such In the of the colitis is or A number of centers have noted that the colitis can be and that studies of the have because of disease histologic of IBD in of PSC patients without bowel and or no endoscopic that studies a of in patients with colonoscopy with be in PSC even if bowel are On the other hand, is to colitis in all especially in the of of to in some case series IBD and of are the most for and the most study of IBD in PSC by patients with UC controls on of that patients and hospitalized for colitis less frequently than the UC controls the was higher in the group because patients (see below). with UC are at increased of and of UC to be factors for In the last 5 a number of studies have PSC and as factors for in UC but PSC is an that associated with and of is from first noted that the prevalence of PSC was higher among a group of UC patients in a or than in the patients not detected a higher prevalence of cholestatic hepatic biochemical abnormalities in UC patients with but the nature of the was undetermined. a higher of in patients compared with patients with UC but controls not on of of these studies have been for small A case control study of U.S. a of and that PSC was as a frequently in IBD patients with a than in IBD patients without there was no control of well factors such as and of UC The that PSC is an for in UC from study from The of or in patients was much higher years of than in a control group of UC only patients on onset of UC, and Several reports have not that PSC of in UC patients a high of in patients years of but this appeared no and than the in UC patients with and no PSC on a cohort of PSC patients and detected a of and of and years of when the of was compared with that in patients from there was no a case control study among all UC patients with at between and patients without on of and and of UC PSC not to be an for for the factors (estimated odds ratio can the findings in these studies be that the differences in the studies from and the is that PSC is a for As the disease of colitis in PSC patients is and not In the group detected one case of and two cases of among their PSC patients without clinical of IBD, suggesting that these patients a of UC no that the of in patients with PSC and UC is at the high of the range in patients with UC over PSC that by and of a these differences may not be We that patients with biopsies from the of of UC than for years as is common in UC patients without PSC. Several centers have in patients have for PSC is that to might in UC have and months of transplantation to have the frequency of colonoscopy or even in patients are to A of patients with at the of patients with and 2 of 3 to of this cohort of patients an of of per and a of of at years V. Loftus, Jr., unpublished with an to to V. Loftus, Jr., unpublished We that colonoscopy be treatment of IBD in PSC patients is to treatment of IBD in The of and of for and immune in or patients Interestingly, for the treatment of PSC may the of IBD UC clinical disease in PSC patients in a randomized clinical of on an review of clinical by and a higher of patients with noted to be in or have clinical compared with patients a of PSC patients in a randomized clinical of no in clinical disease of UC between patients with study or The to liver transplantation may have beneficial on UC disease to patients with UC to or in all but two patients noted that among 25 patients with at the of disease was in the study of patients noted clinical improvement in in 5 patients The improvement in disease in patients is not since most of the in have been shown to have against IBD the development of IBD has been in the patients IBD an average of years Moreover, such to have or no effect on the clinical or of in have (see below). the role of immune factors and the mechanism of immune in IBD remains In the PSC UC is to medical may be The of complications is higher in patients than in UC patients In patients with chronic liver disease and to the development of at the of patients at frequently and treatment is or may be for treatment On the other hand, is not associated with the development of rectal the of in patients have such is high at 5 years This is higher than that in UC patients without PSC at 5 The association between PSC and the that may be a of UC the We that the of the of of which is with is the surgical treatment for UC in PSC hepatobiliary dysfunction is in of all patients with The most of liver disease is PSC, a chronic cholestatic liver disease characterized by inflammation and fibrosis of the biliary and genetic factors to play a role in although environmental factors such as smoking may the of cholangitis to acute liver or is the only effective treatment Other hepatobiliary conditions in IBD include or small duct PSC, autoimmune hepatitis, cholelithiasis, and fatty infiltration of the with PSC have UC, but Crohn's colitis has also been The colitis is in but in and rectal is with UC and PSC are at for Whether sclerosing cholangitis is an for and is unclear, since the presence of sclerosing cholangitis may be a for treatment of the colitis associated with sclerosing cholangitis is to that for IBD patients without liver have liver transplantation have colitis due to The presence of sclerosing cholangitis has for surgical treatment of since with may to which are to to a higher frequency of