Fourth Department of Internal Medicine and Institute of Clinical Biochemistry and Laboratory Diagnostics, Charles University of Prague, Prague, Czech Republic (L. Vítek); Department of Medicine, Harvard Medical School and Harvard Digestive Diseases Center, Gastroenterology Division, Brigham and Women's Hospital, Boston, MA (M. C. Carey).
Libor Vítek, 4th Department of Internal Medicine and Institute of Clinical Biochemistry and Laboratory Diagnostics, Charles University of Prague, U Nemocnice 2, Praha 2, 128 08, Prague, Czech Republic. Tel.: 004-202-2496-2534; fax: 004-202-2492-3524; e-mail: email@example.com
In contrast to bile salts, which undergo a highly efficient enterohepatic circulation with multiple regulatory and physiologic functions, glucuronic acid conjugates of bilirubin are biliary excretory molecules that in health do not have a continuing biologic life. Intestinal absorptive cells are devoid of recapture transporters for bilirubin conjugates, and their large size and polarity prevent absorption by passive diffusion. However, unconjugated bilirubin, the β-glucuronidase hydrolysis product of bilirubin glucuronides can be absorbed passively from any part of the small and large intestines. This can occur only if unconjugated bilirubin is kept in solution and does not undergo rapid bacterial reduction to form urobilinoids. Here we collect, and in some cases reinterpret, experimental and clinical evidence to show that in addition to the well-known occurrence in newborns, enterohepatic cycling of unconjugated bilirubin can reappear in adult life. This happens as a result of several common conditions, particularly associated with bile salt leakage from the small intestine, the most notable ileal dysfunction resulting from any medical or surgical cause. We propose that when present in excess, colonic bile salts solubilize unconjugated bilirubin, delay urobilinoid formation, prevent calcium complexing of unconjugated bilirubin and promote passive absorption of unconjugated bilirubin from the large intestine. Following uptake, reconjugation, and resecretion into bile, this source of ‘hyperbilirubinbilia’ may be the important pathophysiological risk factor for ‘black’ pigment gallstone formation in predisposed adult humans.
The enterohepatic circulation (EHC) of bile salts (Fig. 1) is a highly efficient internal circulation of water-soluble hydroxylated derivatives of cholanoic acid, the principal catabolic products of cholesterol and the major solutes of bile . Bile salts are potent steroidal detergents that solubilize biliary as well as dietary lipids, thereby facilitating their hepatic secretion and intestinal absorption, respectively. Maintenance of bile salts within this circulation is governed by sodium-coupled co-transporters on hepatic sinusoidal membranes as well as apical plasma membranes of ileal enterocytes . In contrast, the natural glucuronic acid conjugates of bilirubin IXα, the principal catabolic product of Fe-protoporphyrin IX, do not recycle enterohepatically, as intestinal cells lack transporters for their recapture. However, resorption of unconjugated bilirubin (UCB) can occur passively by nonionic diffusion from any part of the small and large intestine . This resorption and recycling of UCB occurs commonly in human newborns and results from (i) deficiency in the intestinal microflora capable of reducing UCB to urobilinoids , (ii) high β-glucuronidase activity in human milk, which hydrolyzes bilirubin conjugates to UCB  and (iii) bile salt malabsorption  associated with intestinal immaturity.
Here we collect and critically collate a large body of literature, including our own work, to show that enterohepatic recycling of UCB can re-emerge in adult life. This recycling of UCB may be caused by a number of common pathophysiological conditions, which have in common principally, but not exclusively, increased bile salt loss into the large intestine (Fig. 2). We have marshalled evidence to show that the clinical manifestations of recycling UCB from the intestines are distinct in adults compared with neonates. In adults, recycling UCB molecules elevate biliary levels of bilirubin conjugates, simply because of physiologic maturity of uptake, conjugation, transport and secretory mechanisms of the adult liver. In contrast, because of immaturity of these hepatic steps as well as the lack of urobilinoid production in the neonatal intestinal lumen, recycling UCB contributes to hyperbilirubinaemia in newborns. The adult syndrome would be otherwise benign, were it not for the knowledge that sustained ‘hyperbilirubinbilia’, even of modest proportions, places patients at risk for ‘black’ pigment (calcium bilirubinate) gallstones .
We begin with a summary of current thinking concerning elevated levels of biliary bilirubins and the molecular pathogenesis of pigment gallstone formation. We then classify and highlight the pathogenetic causes of enterohepatic cycling of UCB, and describe our hypothesis for each category in this classification, followed by a list of both the disease states as well as the dietary conditions that may result in both bile salt malabsorption and, as a result, enterohepatic cycling of bilirubin. We close with some suggestions for heightened suspicion at the clinical level of the syndrome and speculate on clinical methodology for its noninvasive diagnosis, plus an end note on how recycling of UCB might be prevented.
Elevated biliary bilirubin levels and pigment gallstones
Pigment gallstones have an estimated prevalence rate of 20–25% among patients undergoing cholecystectomy in the United States . By international consensus, they are classified trivially as ‘black’ or ‘brown’. Black gallstones are composed principally of calcium bilirubinates in which the pigment becomes polymerized and chemically degraded. In contrast, the pigment salt in brown stones remains generally unaltered chemically . These stones result from chronic infection of a partially obstructed biliary tree and their pathogenesis is probably unrelated to enterohepatic cycling of bilirubin. Consequently they are not considered further in this review.
Black pigment gallstone formation is associated with any condition associated with biliary hypersecretion of bilirubin . Hence, they are caused by any type of haemolytic anaemia and ineffective or ‘shunt’ erythropoiesis. However, they are also associated with other pathophysiological states such as Gilbert's syndrome  where the total biliary output of bilirubins is unchanged but there is higher proportions of the bilirubin monoglucuronides (BMG) in bile from decreased hepatic bilirubin UDP-glucuronosyltransferase (UDPGT) activity . Interestingly, biliary BMG secretion is also increased in biles of otherwise healthy patients with black pigment gallstones [9,10]. Most likely this is secondary to increased UCB concentrations delivered to the conjugating system of the liver. Moreover, biliary BMG is the predominant biliary bile pigment in most animal models of pigment gallstone formation [11,12]. Compared with BDG, BMG is more easily deconjugated by β-glucuronidase, a hepatic lysosomal and biliary epithelial enzyme that functions in mildly acidic media [10,13,14]. If the biliary concentrations of UCB and Ca2+ are such that the ion-product of Ca(HUCB)2, the mono-acid salt, exceeds its solubility product , then the physical-chemical conditions are met for pigment supersaturation of bile. This has been documented in gallbladder biles of humans and animal models with ‘black’ pigment gallstones, as well as in ‘biliary sludge’. It is possible that calcium bilirubinate plays some, as yet undefined, role in the earliest events of cholesterol gallstone formation , as the monoacid salt of calcium bilirubinate is found invariably in trace amounts in the cores of ‘pure’ cholesterol stones .
Passive enterohepatic cycling of bilirubin
The methodology for determining an EHC of bile pigments was validated when radiolabelled UCB became available in the early 1960s. Through intraduodenal instillation of radiolabelled UCB with rapid bile sampling for radiolabelled conjugates, an EHC of UCB was confirmed in rats (16) and humans . Subsequent studies  demonstrated that an EHC occurs exclusively with UCB and not with conjugated bilirubins. Therefore, intestinal deconjugation of bilirubin glucuronides is a necessary antecedent for EHC of bilirubin. As with other soluble, high polar and bulky amphiphiles, the apical membranes of enterocytes act as a barrier to the passive absorption of bilirubin glucuronides . In contrast, UCB is highly insoluble in physiologic fluids because of a conformational change at pH values < 8·3 resulting in the involvement of all its polar groups, in six internal H-bonds . By removing these polar functions from water contact, bilirubin becomes a ‘nonpolar’ hydrophobic molecule, therefore the solution state of the molecule with respect to pH will predict whether an EHC can occur. As solubilization of UCB is crucial for passive intestinal absorption, a solubilization mechanism must be available in the lumen of the intestine to promote its resorption. Keeping UCB in solution by the action of bile salts may even enhance the passage of weakly polar bilirubin across the unstirred water layer to the mucosal surface as shown for bile acids by Wilson and Dietschy . As listed below, this is likely to be provided at the site of bilirubin deconjugation by bile salts either in monomer or micellar concentrations (Fig. 2). Because of bilirubin's varying polarity, the pH of the colonic contents may also play an important role. For example, a fall in colonic pH secondary to severe pancreatic bicarbonate deficiency or as a result of loss of a functioning chloride-bicarbonate exchanger in the ileum might have competing effects, i.e. increasing the proportion of bilirubin diacid available for passive diffusion, but decreasing the proportion of bilirubin anions precipitable as calcium salts.
We have classified three conditions that render intestinal resorption of UCB possible (Table 1):
3. Augmented solubilization ofUCB in the large intestine
Bile salt malabsorption with higher bile salt concentrations in the proximal large intestine; prevention of bilirubin from complexing with intraluminal Ca2+
Ileal disease, injury, resection or bypass Bile acid therapy to effect gallstone dissolution High cholesterol or carbohydrate diets Alcohol abuseImmaturity of ileal apical sodium-coupled bile salttransporter-ASBT (neonatal jaundice) Dysfunctional mutations of ASBT causing mild to moderate spillage of bile salts ?Therapy with bile acid resorption inhibitors
1) Higher rates of bilirubin deconjugation in the small intestine. The proposed causes include prolonged intestinal transit time, colonization with β-glucuronidase-producing bacteria, or a supply of β-glucuronidase from extraintestinal sources.
2) Deficiency of bacteria capable of reducing UCB in the large intestine. Typically, this category is physiologic in neonates and possibly acquired in some adults receiving broad-spectrum antibiotics for gut ‘sterilization’ or as an adverse effect of prolonged systemic antibiotic therapy.
3) Augmented solubilization of UCB in the large intestine. We propose that high bile salt concentrations in the proximal large intestine increase solubilization of UCB, delay urobilinoid formation and also prevent calcium bilirubinate formation.
Once resorbed, UCB returns to the liver in the portal circulation, to be taken up, reconjugated and resecreted into bile. A major proportion does ‘spill over’ into the systemic circulation because of its low ≈ 30% single-pass hepatic extraction  and hence may in some cases induce mildly elevations in serum bilirubin levels. Indeed this has been confirmed by Berlin and Berk  who demonstrated a close relationship between serum bilirubin levels and bilirubin turnover in patients with haemolytic anaemia and decreased hepatic bilirubin clearance rates.
Proximal intestinal deconjugation of bilirubin conjugates
In 1906, Gilbert and Herscher  observed increased plasma bilirubin levels in humans undergoing prolonged calorie restriction. Many suggestions to account for this observation were proposed and only recently was the mechanism whereby intestinal resorption of bilirubin is promoted by fasting clarified experimentally . It was shown in both Wistar and hyperbilirubinemic Gunn rats that decreased gut motility caused a diminution in fecal outputs of bile pigments. As a consequence, serum bilirubin levels increased with expansion of the intestinal bile pigment pool . To prove directly the enteric origin of elevated blood bilirubin levels, Wistar rats were injected intracaecally with UCB in amounts corresponding to the intestinal bilirubin that accumulated during 48 h of fasting, and significant increases in serum bilirubin resulted. Although biliary outputs in these animals were not measured, the entire data set, especially increased biliary outputs in fasted animals, strongly suggest an appreciable contribution of EHC of bilirubin to increased biliary bilirubin secretion rates . The concept of enhanced enterosystemic circulation of bilirubin during fasting was further supported by experiments demonstrating that hepatic transport mechanisms for bilirubin were unimpaired by calorie restriction .
Prolonged intestinal transit times inducing an EHC of bilirubin  might contribute to mild jaundice in patients on long-term total parenteral nutrition (TPN)  along with other pathogenic, especially mechanical, contributing factors . Decreased gut motility may also play a role in the unconjugated hyperbilirubinaemia of premature and low birth weight infants . In healthy neonates fed 2-hourly, peristalsis is promoted and evacuation of bowel contents is associated with significant lowering of serum bilirubin concentrations compared with newborns fed less frequently .
Breast-milk jaundice was first described in detail in the mid 1960s  as a syndrome of unconjugated hyperbilirubinaemia that peaked during the first 10–20 days postpartum. Invariably, within the first couple of months of life jaundice disappears despite continued breast-feeding. Although it was suggested that inhibitors of hepatic bilirubin glucuronidation are present in human milk , it now appears that the high levels of β-glucuronidase in human breast milk is likely to be the responsible factor . Because of immature hepatic bilirubin UDPGT activity , the principal bilirubin conjugates in neonates are monoglucuronides. Therefore, proximal luminal hydrolysis is likely to produce augmented amounts of UCB in the small intestine. The other contributing factor is the larger intestinal UCB pool in neonates, because of defective intestinal reduction of bilirubin from delayed colonization with the appropriate microflora . In addition, the ileal apical sodium-dependent bile salt transporter (ASBT) is underexpressed during early neonatal life  and its immaturity augments bile salt spillage from the small intestine, thereby promoting an EHC of bilirubin from the colon. Because hepatic uptake, hepatocellular conjugation, transport, and secretory mechanisms for bilirubin in the newborn are underdeveloped , an EHC of UCB is manifested principally by enterosystemic circulation of bilirubin and hyperbilirubinaemia rather than hyperbilirubinbilia. Because of maturation of the functionality of these uptake, conjugatory, transport and secretory steps with passage of time, breast-milk jaundice disappears spontaneously usually within the first 3 months of life.
Slow or absent bacterial reduction of bilirubin in the intestine
Bilirubin reduction to urobilinoids (see nomenclature of hydroderivatives of bilirubin in ) by the intestinal microflora represents the normal catabolic pathway for tetrapyrrole degradation. Because of poor colonization with bilirubin-reducing microflora in neonates, only negligible amounts of urobilinoids are present in the distal intestinal lumen or faeces . As a result, nearly all UCB formed is available for resorption provided it is maintained in solution . The role of the intestinal microflora in homeostasis of bilirubin and its metabolites was directly demonstrated in Gunn rats whose intestinal microflora was eradicated by broad-spectrum antibiotics. As a consequence of the dearth of fecal urobilinoids, serum bilirubin levels increased significantly. Once the intestine was colonized with a bacterial strain that reduced bilirubin quantitatively, serum levels of UCB declined paralleling reappearance of urobilinoids in the stool .
Increased intestinal solubilization of unconjugated bilirubin
Because of intramolecular hydrogen bonding , UCB molecules are highly insoluble in aqueous solution. In native bile, the UCB formed in small amounts is solubilized principally by bile salt micelles ; concomitantly most bile calcium is bound effectively to mixed bile salt-lecithin-cholesterol micelles . Several factors may increase UCB's solubility in the intestinal lumen, either by preventing bilirubin precipitation with ionized calcium and/or mechanisms directly solubilizing UCB in intestinal contents.
Preventing insoluble salt formation by bilirubinate and ionized calcium
Preventing UCB from forming sparingly soluble salts with dianionic calcium [37,38] leads to higher UCB solubilization, thereby promoting its intestinal resorption. Although cecum and colonic concentrations of calcium especially in the presence of varying bile salt concentrations are unknown, bile salts may bind calcium to form soluble calcium complexes  thereby facilitating UCB's solubilization and resorption . Another possibility promoting UCB absorption is that variable degrees of fat malabsorption ensue with severe intestinal bile salt loss. The increased fatty ‘acid soap’ concentrations in the proximal small intestine can in turn remove calcium from solution as salts of fatty acids. Presumably, the removal of Ca2+ ions from solution could result in increased UCB concentrations in the lumina of small and large intestines, as was suggested for colonic absorption of oxalic acid in enteric hyperoxaluria .
Influence of distal intestinal bile salt concentrations
With bile salt malabsorption, cecal bile salt levels become elevated by bile salt spillage from the small intestine. As bilirubin conjugates are hydrolyzed at the same site, the resulting UCB molecules become solubilized by both bile salt monomers and micelles , and thus passive intestinal absorption is promoted . Bile salt malabsorption occurs typically from ileal disease or resection as well as chemically from oral bile acid therapy . Moreover, it has been claimed  that in the distal small intestine, bile salts facilitate bilirubin's deconjugation by bacteria by enhancing β-glucuronidase diffusion through bacterial plasma membranes. Because bile salts per se may influence the activity of β-glucuronidase substantially , the available evidence suggests that there must be a net balance in the gut lumen in favour of deconjugation. As EHC of bilirubin commonly results in black pigment gallstone formation in the setting of bile salt malabsorption, specific pathogenic mechanisms will now be discussed.
Association of pigment cholelithiasis with enterohepatic cycling of bilirubin
Several lines of evidence support the proposal that ‘black’ pigment gallstones can be caused by an acquired EHC of bilirubin. Increased prevalence of black pigment cholelithiasis is known to occur in Crohn's disease [5,42,43], ileal resection [5,44] or bypass , oral bile acid administration [46–50], high cholesterol [51,52] and carbohydrate-containing diets , total parenteral nutrition [54,55], alcoholism  and cystic fibrosis . We show now that the common denominator in all of these appears to be hyperbilirubinbilia caused by bile salt malabsorption with resorption of UCB from the distal small and large intestines (Table 2).
Table 2. Clinical conditions associated with ‘black’ pigment gallstones putatively caused by enterohepatic recycling of unconjugated bilirubin
Impaired ileal bile salt resorption as a result of chronic inflammation of the ileum
Ileal resection or bypass
Removal or bypass of active bile salt resorption sites in the distal ileum
Therapy with unconjugated bile acids
Excess free and conjugated bile salts in the proximal large intestine secondary to oral administration of bile acids
High cholesterol intake
Excess bile salts in the large intestine as a result of down-regulation of active bile salt resorption transporters in the distal ileum
Excess bile salts in the proximal large intestine due to starch-induced bile salt sequestration or osmotically induced diarrhea
Total parenteral nutrition
Decreased intestinal motility with increased bilirubin conjugate hydrolysis and resorption of unconjugated bilirubin from the small intestine
Defects in intestinal motility and bile salt resorption because of ethyl alcohol-induced damage to ileal mucosa, and bile salt malabsorption as a result
Cause of bile salt malabsorption not determined. Possibly resulting from bile salt binding to undigested carbohydrates and proteins, as well as precipitation of glycine conjugates from hyperacidity of the upper small intestine. However, in vivo functional down-regulation of the ileal apical sodium-dependent bile salt transporter is possible
Ileal Crohn's disease, resection or bypass
In humans as well as animal models, ileal disease, bypass, or resection represent a major risk factor for gallstones [5,42,43]. As bile salt malabsorption is the rule in ileal dysfunctional states , this may lead to shrinkage of the bile salt pool. It was believed previously that an increase in relative biliary cholesterol concentration occurred with bile salt malabsorption, thereby elevating cholesterol saturation of bile . However, studies of adults with partial ileal resection, ileectomy or Crohn's disease [60,61] have demonstrated the opposite: they lower biliary cholesterol saturations compared with matched healthy controls. These results are consonant with animal studies by Pitt et al. who showed that following ileal resection in prairie dogs, cholesterol saturation indexes of gallbladder bile remained essentially unchanged, whereas pigment gallstones formed frequently. Moreover, in the majority of ileectomized animals, there were significant increases of bilirubin conjugates as well as microscopic calcium bilirubinate precipitates in gallbladder bile .
Over three decades ago, Heaton and Read  noted a 32% prevalence of gallstones in patients with disorders of the distal ileum, compared with 12% in controls. Plain radiographs of the abdomen revealed that 57% of patients carried radioopaque calculi compared with 18% in the control group, suggesting a higher prevalence of noncholesterol stones in patients with ileal dysfunction . Magnuson and colleagues  verified by chemical analysis that calcium bilirubinate was the principal component of gallstones in Crohn's disease, and Dawes and colleagues  were the first to demonstrate that patients with Crohn's disease exhibited significantly increased levels of monoconjugated and unconjugated bilirubins in their gallbladder biles. These findings were confirmed  and enrichment of gallbladder bile with bile pigments was most pronounced in patients with Crohn's disease who had had an ileectomy. Moreover, ileal dysfunction increases the risk for black pigment gallstone formation in paediatric patients especially those treated with TPN . This ‘proof-of-principle’ was corroborated by a recent observation by Dawson et al. who found a new dysfunctional mutation in the ileal ASBT gene, which apparently gave rise to ‘idiopathic’ black pigment gallstones in a single patient, thereby supporting the putative importance of bile salt malabsorption in the pathogenesis.
The important longitudinal studies of Buchwald et al. probably adds the greatest weight to the hypothesis that bile salt malabsorption induces an EHC of bilirubin. This surgical treatment of hypercholesterolaemia embodied a large cohort of patients who, following initial myocardial infarction, underwent a partial ileal bypass to control secondary cardiovascular mortality. As levels of low density lipoprotein cholesterol fell significantly; major degrees of bile salt malabsorption must have resulted from the intestinal bypass. Concomitantly, the incidence of gallstones increased significantly, even within the first few postoperative years. Moreover, since increased incidence of calcium oxalate kidney stones also occurred, the concordance of both diseases suggests that the gallstones were due to hyperbilirubinbilia and therefore both the kidney and gallbladder stones apparently had similar enteral hyperabsorptive mechanisms.
Oral administration of bile acids
Chronic administration of the cholesterol gallstone ‘dissolving’ bile acids, chenodeoxycholic acid (CDCA) and ursodeoxycholic acid (UDCA)  was often associated with formation of calcified shells on gallstones , and dissolution failure from this cause occurred in 10–15% of subjects [46–50]. As calcification was first described in UDCA-treated patients, it was postulated that UDCA was most likely precipitating as the calcium salt of its less soluble glycine conjugate . However, gallstone calcification also occurred with similar frequencies during CDCA therapy  and even following administration of the sodium salt of tauroursodeoxycholate , which forms highly soluble calcium salts . In contrast, in vitro studies suggested that glycoursodeoxycholate and glycochenodeoxycholate might actually be protective against calcium precipitation in bile [37,66] by virtue of prolonged metastability of their supersaturated solutions. Chemical analysis of radiolucent gallstones that failed to dissolve with CDCA or UDCA suggested that either calcium bilirubinate  or calcium carbonate was deposited on their surfaces . Moreover, CT evaluation  of the calcium content of stones that failed to dissolve on oral dissolution therapy revealed signal attenuation consistent with gallstone calcification. Increased biliary secretion of bilirubin , calcium  and bicarbonate  were all suggested to explain acquired calcification during cholelitholytic therapy. However, the fact that UDCA and CDCA induce endogenous bile salt malabsorption in humans  is consistent with the findings of Capron et al. who showed appreciable increases of serum UCB levels during CDCA treatment without any signs of haemolysis or liver disease . Recently, Fickert and colleagues  demonstrated that the canalicular multidrug-resistance-related protein, isoform 2 (MRP2), the bilirubin conjugate export pump, was indeed overexpressed in mice fed UDCA. Most likely this up-regulation is secondary to increased levels of bilirubin conjugates being secreted by the liver secondary to enterohepatic cycling of UCB.
High cholesterol diets
Mendéz-Sánchez et al. demonstrated biliary hypersecretion of bilirubin conjugates in rats and mice fed large amounts of dietary cholesterol. This is consistent with malabsorption of bile salts from down-regulation of the ileal ASBT . In 1968, Dam et al. showed that 1% cholesterol added to a 10% fat diet lowered the prevalence of cholesterol gallstones in Syrian hamsters, but increased the frequency of ‘black’ pigment gallstones. Dietary cholesterol (0·5–1%) was also shown to be essential for development of pigment gallstones in several experimental animals including the domestic dog . Based on these results it seems likely now that an EHC of bilirubin and hyperbilirubinbilia occurs from cholesterol-induced bile acid malabsorption .
Epidemiologically, high dietary intake of carbohydrates by rural Japanese have been implicated in the pathogenesis of pigment gallstones . Moreover, in prairie dogs, carbohydrate-rich diets lead to formation of pigment biliary sludge and black pigment gallstones . Several speculative explanations have been offered to explain this association, but in light of several considerations, it is likely that the responsible factor is an EHC of UCB induced by the carbohydrate-rich diet. For example, excess dietary carbohydrates are known to cause bile salt malabsorption. Bile salt malabsorption has been induced in mice and hamsters by a diet rich in nondigestible starch and β-cyclodextrin . Moreover, both taurine and glycine conjugates and unconjugated bile salts bind with high affinity to cyclodextrins in vitro. In fact, serum cholesterol levels in rats can be decreased by oral administration of amylase-resistant starch , consistent with intestinal bile salt malabsorption and compensatory up-regulation of hepatic de novo bile salt synthesis. Malabsorbed bile salts resulting from adsorption to the polymer chains of polysaccharides in the small intestine would be released therefrom in the cecum because of rapid bacterial fermentation of the sugars by the anaerobic enteric microflora. In turn, the bile salts would contribute to enhanced solubilization of UCB in the colonic lumen. Similar results have been found with diets replete in refined sugars: high sucrose diets increase fecal bile salt loss, possibly because of induction of more rapid intestinal transit times . Moreover, hypertonic solutions of sucrose may also increase the permeability of the intestinal mucosa transiently . For the same reasons, these considerations may account for the observation that in individuals with Gilbert's syndrome, feeding refined carbohydrate diets results in elevated plasma bilirubin levels .
Total parenteral nutrition
During TPN, an EHC of bilirubin seems likely to contribute to pigment gallstone formation for several reasons . Biliary sludge containing bile pigment precipitates is observed frequently in the gallbladder  and its incidence is related to the duration of TPN. The pigment cholelithogenic effect of TPN is also manifested in paediatric patients , especially those with concomitant ileal disease. In all of these situations, increased levels of bilirubin conjugates and UCB are found in bile, without any evidence of haemolysis. Apart from decreased gut motility as a cause of recycling of UCB, prolonged TPN may also cause ileal atrophy and down-regulation of ileal ASBT, complicating the scenario with added bile salt malabsorption . An exceptionally high prevalence of TPN-associated pigment cholelithiasis occurs in subjects with massive small bowel resections, which strongly underscores the putative role of bile salt malabsorption in its pathogenesis . Although a possible EHC of bilirubin has never been investigated during TPN, indirect evidence allows us to infer that it plays an appreciable role in pathogenesis of TPN-induced pigment gallstone formation.
Chronic alcohol abuse and alcoholic cirrhosis
Although alcoholic cirrhosis and pigment cholelithiasis are strongly associated epidemiologically, chronic alcoholism per se is not widely accepted as a predisposing factor for biliary sludge  and pigment gallstones . In fact, more than two-thirds of alcohol abusers with pigment gallstones did not appear to have coexisting cirrhosis .
It remains speculative whether an EHC of bilirubin may play any role in alcoholism-related pigment gallstones. Nonetheless, chronic alcohol ingestion by laboratory animals leads to morphological and functional alterations of the small intestinal mucosa, which may compromise Na+-coupled bile salt transport . Furthermore, intravenous infusions of ethanol augment biliary UCB secretion in rabbits and humans even without evidence of accompanying haemolysis , and which might also contribute to pigment stone pathogenesis. In addition, the frequent prevalence of folate deficiency in chronic alcohol abusers may cause macrocytosis and possibly mild pandysfunction of the intestinal epithelia . In patients with exocrine pancreatic insufficiency , bile salt malabsorption may also contribute to an EHC of UCB from the binding of bile salts to undigested polysaccharides and proteins and their desorption when the dietary polymers are anaerobically fermented in the large intestine. Therefore, the high prevalence of pigment gallstones in chronic alcoholism may, in part, be related to a wide variety of pathogenic causes, including but not limited to ethanol-induced bile salt malabsorption and an EHC of bilirubin.
The prevalence of cholelithiasis in cystic fibrosis is at least 10–30% compared with less than 5% in age-matched controls . It was considered earlier that the gallstones were cholesterol in type, attributable to excessive fecal bile salt loss, which occurs in at least one-third of patients with cystic fibrosis . Traditionally, bile salt malabsorption in cystic fibrosis was believed to occur from bile salt binding to undigested dietary nutrients [87, 88], and/or precipitation of glycine conjugates in the acidic proximal small intestine .
For the same reasons as indicated earlier for Crohn's ileitis and ileectomy, cholesterol supersaturation of bile  secondary to shrinkage of the bile salt pool was thought to be the cause of the so-called ‘cholesterol’ gallstones in this condition. However, a careful chemical analysis by Angelico and colleagues  demonstrated that the major chemical component of gallstones in cystic fibrosis patients is calcium bilirubinate and not cholesterol. In light of this evidence it seems likely that bile salt malabsorption in cystic fibrosis patients induces an EHC of bilirubin, which may be responsible for the high pigment gallstone prevalence rates. Recent experimental studies in mice with mutations in the cystic fibrosis transmembrane conductance regulator gene have shown that bile salt malabsorption also occurs in these animal models . Moreover, more than half of cystic fibrosis mice acquire black pigment gallstones and all demonstrate hypersecretion of bilirubin conjugates into bile with up-regulation of MRP-2, the canalicular transport protein for bilirubin conjugates, thereby supporting an EHC of UCB as causing the stones and perhaps chronic liver disease .
Therapeutic uses of bile acid resorption inhibitors
A time-honored therapeutic approach to decreasing plasma low-density lipoprotein cholesterol levels is by increasing fecal bile salt loss by employing oral sequestrants (colestipol/cholestyramine). A more efficient bile salt binder colesevelan HCl (WelChol®, Sankyo Pharma Inc., New York, NY, USA) has appeared recently on the market . However, these traditionally used bile salt sequestrants do not engender an increased concentration of bile salt molecules in solution in colonic water. Rather, they produce the reverse and are not associated with an increased risk of gallstone formation. However, a new class of hypolipidaemic drugs is now emerging  that selectively inhibit ileal ASBT and cause bile salt malabsorption with subsequent hypocholesterolaemia. With these agents, there is likely to be increased bile salt concentrations in solution in the large intestine, which might solubilize UCB and facilitate its resorption with induction of an EHC of bilirubin. Only time and careful surveillance will tell whether excess colonic bile salts from deployment of these bile acid reabsorption inhibitors will have any clinically significant consequences.
Diagnosing and preventing enterohepatic cycling of bilirubin
It is now well accepted that a ‘physiologic’ EHC of bilirubin in neonates can lead to augmented hyperbilirubinaemia. This occurs because of hepatic immaturity or rarely acquired dysfunction of the hepatic uptake, conjugation, and secretory systems. The hypothesis developed here is that hyperbilirubinbilia with subsequent pigment gallstone formation can be induced by enhanced EHC of bilirubin in susceptible adult patients by a variety of mechanisms all of which are based principally upon rather indirect evidence [5,24]. Further investigations are needed to elucidate the role of an EHC of UCB in adults in pigment gallstone pathogenesis, as well as possibly causing mild subclinical jaundice because of the low first-pass hepatic extraction and spill-over of recycled bilirubin into the systemic circulation. Such studies could be accompanied by assessment of the role of agents capable of interrupting an EHC of bilirubin and preventing hyperbilirubinaemia and hyperbilirubinbilia. Susceptible subjects would include those with bile salt malabsorption from ileal pathology or resection, patients on TPN or oral treatment with bile acids, alcohol abusers, those with exocrine pancreatic insufficiency and, possibly, polysaccharide-rich diets. Clearly if our hypothesis is validated directly, then safe, efficient nonabsorbed hydrogels with UCB binding affinity will need to be developed to prevent enterohepatic cycling of UCB.
Unfortunately, no simple diagnostic tool is currently available to identify subjects with an EHC of UCB and thereby the risk for hyperbilirubinbilia and ‘black’ pigment gallstones. At the research level, studies with radiolabelled or stable isotope-labelled bilirubin conjugates would be essential to prove an intestinal source of resorbed bilirubin and its role in mild hyperbilirubinaemia and hyperbilirubinbilia. Moreover, these studies would be important to directly prove and quantify an EHC of bilirubin in ileal dysfunctional states. At the clinical level, one is left with no choice but to rule in the diagnosis of EHC of UCB, only by excluding the two other major causes of hyperbilirubinbilia, which are haemolysis and shunt/dyserythropoiesis. Indirect diagnostic tests for excess biliary secretion of bilirubin conjugates (Fig. 3), such as a challenge with phenobarbital or nicotinic acid, which have been widely used in patients with unconjugated hyperbilirubinaemia, would be of little quantitative value, as biliary levels cannot easily be assayed except at the research level. A test of prolonged fasting might be useful for EHC-induced hyperbilirubinnaemia, as its effect is based on decreased gut motility; however, the specificity and sensitivity of such a test might be difficult to establish without investigations on a large cohort of humans.
Several approaches for the treatment of enteric hyperbilirubinaemia based on interruption of EHC of bilirubin have been deployed in infants , but in early life their exhibition and usefulness is required only on a temporary basis and the results are clearly visible. Among the classes of new experimental agents is administration of zinc salts, which decrease serum bilirubin levels in subjects with Gilbert's syndrome , but Zn is absorbed and potentially toxic. Interestingly, bilirubin-lowering effects without any significant elevation of serum Zn levels were recently reported in Gunn rats fed a nonabsorbable Zn methacrylate salt . It seems that only nonabsorbable agents have sufficient aesthetic appeal and safety, provided they are effective with low incidence of adverse effects, to be used chronically in adults. However, it is apparent that further investigation will be needed to prove the syndrome and to develop appropriate therapeutic approaches. Recently, two new Clostridial strains possessing bilirubin reductase activity have been characterized . This suggests that appropriate probiotic colonization of the large bowel might constitute a future therapeutic modality to reduce UCB to urobilinoids. As outlined in this review, newer mechanistic agents may need to be developed for preventing UCB resorption in a wide variety of clinical scenarios. Clearly, much more needs to be learned of the pathobiology and pharmacology of gut-induced hyperbilirubinbilia and perhaps its systematic reflection as mild hyperbilirubinaemia in the adult, before this new syndrome comes of age both diagnostically as well as therapeutically.
This work was supported in part by research grant GAUK nos. 203577 and 203578 (Granting Agency of Charles University, Prague), GACR no. 310/021436 (Granting Agency of the Czech Republic), and DK 36588, DK 34854 and DK 52911 from the National Institutes of Health (U.S. Public Health Service). The authors thank Jayanta Roy Chowdhury, MD, (Albert Einstein College of Medicine, Bronx, NY) for his critical reading of the manuscript and Ms Maya Kavtaradze for her editorial assistance.