Review article: hepatobiliary complications associated with total parenteral nutrition.

Parenteral nutrition is often associated with hepatobiliary complications. Hepatic steatosis, intrahepatic cholestasis and biliary sludge are the most frequent. Cholestasis predominates in infants, steatosis in adults, and biliary sludge in both. Other less frequent complications are steatohepatitis and gallstones. All hepatobiliary complications are more likely to occur after extended periods of total parenteral nutrition, and are prevented by the concomitant consumption of nutrients by the enteral route. The pathogenic causes are multiple and only partially known. They include lack of gastrointestinal stimuli for biliary secretion and gall-bladder motility, abnormalities in bile acid metabolism, the presence of sepsis, and the potentially unfavourable effects of individual components in the total parenteral nutrition formulae, including an excess of calories. Each potential mechanism and its clinical relevance is discussed in this review.

Impaired human gallbladder lipid absorption in cholesterol gallstone disease and its effect on cholesterol solubility in bile.

BACKGROUND & AIMS: The role of the gallbladder in gallstone pathogenesis is still unclear. We examined the effects of gallbladder mucosal lipid absorption on lipid composition and cholesterol crystallization in bile. METHODS: The in vitro-isolated, intra-arterially perfused gallbladder model was used (1) to compare the absorption rates of lipids from standard bile by gallbladders obtained from 7 patients with cholesterol gallstones and 6 controls; and (2) to measure the microscopic cholesterol crystal detection time in cholesterol-enriched pig bile before and after lipid absorption by the pig gallbladder. RESULTS: Control gallbladders, but not cholesterol gallstone gallbladders, significantly reduced cholesterol (P < 0.02) and phospholipid (P < 0.01) and increased bile salt (P < 0.01) molar percentages in bile over a 5-hour period by efficient and selective cholesterol and phospholipid absorption. A histomorphometric study of the epithelial cells showed significantly higher values for nuclear density (P < 0.01) and nuclear (P < 0.05) and cytoplasmic (P < 0.05) areas in the cholesterol gallstone than the control group. Sequential microscopy of cholesterol-enriched pig bile showed significantly shorter cholesterol filament (P < 0.01) and typical cholesterol plate (P < 0. 02) detection times before than after exposure of bile to the gallbladder lipid absorption. CONCLUSIONS: In cholesterol gallstone disease, the human gallbladder epithelium loses its capacity to selectively and efficiently absorb cholesterol and phospholipids from bile, even if it is hyperplastic and hypertrophic. This epithelial dysfunction eliminates the positive effect that the normal gallbladder exerts on cholesterol solubility in bile and might be a pathogenetic cofactor for cholesterol gallstone formation.

Plasma lipoproteins affect rate of cholesterol absorbed from bile by gallbladder: preliminary data.

BACKGROUND: The excessive accumulation of cholesterol absorbed from bile by the gallbladder impairs its contractility and favours gallstone formation. The total low plasma and high density lipoprotein cholesterol concentrations are associated with gallstone disease. AIMS: To investigate the effect of plasma lipoproteins on gallbladder cholesterol and phosphatidylcholine absorption from bile and to establish whether cholesterol absorption is Brefeldin A-sensitive. METHODS: Gallbladder mucosa lipid absorption rates were measured using: 1) in vitro isolated intra-arterially perfused pig gallbladder model with and without plasma lipoproteins perfusing the vascular tree; 2) human gallbladder fragments mounted in Ussing chambers with plasma lipoproteins at different concentrations in the serosal side; 3) pig gallbladder fragments mounted in Ussing chambers in the presence and absence of Brefeldin A. RESULTS: Total lipoproteins and high density lipoprotein significantly increased the release of biliary cholesterol and phosphatidylcholine in plasma and significantly decreased the tissue accumulation of cholesterol absorbed from bile. The scavenger effect of plasma lipoproteins on cholesterol absorbed from bile was concentration dependent. Brefeldin A did not influence gallbladder absorption of biliary cholesterol. CONCLUSIONS: Biliary cholesterol is absorbed by gallbladder mucosa via a Brefeldin-insensitive pathway and is removed by plasma lipoproteins.

The human gallbladder increases cholesterol solubility in bile by differential lipid absorption: a study using a new in vitro model of isolated intra-arterially perfused gallbladder.

In this study, we first developed and validated a new in vitro isolated, intra-arterially perfused, gallbladder model and then applied the method to investigate the absorption of biliary lipids by the gallbladder wall and the effect of this process on the composition of human bile. Oxygenated and glucose-added buffer was perfused through the cystic artery to maintain organ viability. A standard pooled natural bile, radiolabeled with H3-cholesterol and C14-palmitoyl-linoleoyl-phosphatidylcholine, was instilled in the lumen via a cystic duct catheter. Changes in bile volume and lipid concentrations were monitored at time intervals to evaluate the disappearance of lipids from bile caused by gallbladder absorptive function. Organ viability was demonstrated by stable lactate dehydrogenase (LDH) organ release and oxygen consumption throughout the experiments. In the pig, disappearance rates of lipids from bile were similar in vitro and in vivo, demonstrating the validity of the isolated in vitro model for functional studies. By applying our in vitro isolated preparation to the human gallbladder, we found that 23% of cholesterol and 32% of phosphatidylcholine, but only 9% of bile salts, disappeared from bile in 5 hours. As a consequence, at the end of the experiments, cholesterol (P < .05) and phospholipid (P < .05) molar percentages were significantly reduced, while the bile salt (P < .05) molar percentage was significantly increased with respect to values at the beginning of the studies. Our findings are of pathophysiological relevance and support the concept that the human gallbladder modifies the relative composition of biliary lipids in such a way as to increase cholesterol solubility in bile.

Human gallbladder mucosal function: effects on intraluminal fluid and lipid composition in health and disease.

Gallbladder mucosal absorption of fluid during fasting is a well-known process. Indirect in vivo and recent in vitro evidence for physiologically relevant gallbladder absorption of cholesterol and phospholipids from bile has been observed in humans. The present study explored and compared by indirect means the relative efficiences of human gallbladder mucosal absorption of fluid and lipids in health and disease. Biliary lipids and pigment content were measured in fasting gallbladder bile samples obtained from gallstone-free controls and from four study groups: multiple and solitary cholesterol gallstone patients, and morbidly obese subjects with and without gallstones. Bile salts and pigment content were significantly greater in gallstone-free controls than in all other disease study groups. This was interpreted as evidence of more effective gallbladder mucosal fluid absorption in nonobese gallstone-free controls compared to that in all other groups. Correlation plot analyses of biliary lipids showed lower concentrations of phospholipids than expected from the index bile salt concentrations. The same was found for cholesterol concentrations but only in supersaturated samples. These findings were much more pronounced in gallstone free-controls and were accordingly interpreted as evidence of more efficient gallbladder absorption of both phospholipids and cholesterol in controls compared with that found in each of the disease study groups. Moreover, impaired gallbladder mucosal function, while invariably associated with cholesterol gallstone disease, was not found to be a necessary consequence of the physical presence of stones. It is concluded that efficient gallbladder mucosal absorption of both fluid and apolar lipids from bile is a normal physiological process that is often seriously impaired in the presence of either cholesterol gallstone disease or at least one of its precursor forms.

Variations in pigment and carbohydrate content of gallbladder bile affect accurate quantitation of total protein when using the fluorescamine method.

BACKGROUND: Despite solute dilution and reduced total lipid concentrations, an unexplained increase in protein concentration has been reported to occur in the gallbladder bile of cholesterol gallstone patients. METHODS: Solutes in gallbladder bile from gallstone-free controls and from four study groups were measured using standard methods. Total proteins were measured using amino acid analysis and a conventional fluorescamine method. RESULTS: Bile salts and pigment content were greater in gallstone-free controls than in all other study groups, including morbidly obese gallstone-free subjects. Total biliary protein concentration, as determined by amino acid analysis in the gallstone-free control group was higher than in non-obese gallstone patients with multiple stones and in morbidly obese gallstone-free subjects. Total biliary proteins as measured with fluorescamine, however, did not show intergroup differences. A major problem of the conventional fluorescamine assay is shown to be an artefact arising from the high pigment content of the more concentrated samples. CONCLUSIONS: Very dilute gallbladder bile samples are often found in the presence of gallstone disease. This also occurs in morbidly obese subjects, even in the absence of gallstones. Although the contribution of protein secretion/absorption by the gallbladder can also be relevant, especially in the presence of morbid obesity, the protein concentration in gallbladder bile, when accurately measured, generally parallels the concentrations of non-absorbed biliary solutes, reflecting the efficiency of fluid absorption. Measurement of biliary proteins by the conventional fluorescamine method is unreliable in clinical studies in which intergroup differences in pigment content are commonly present.

Effect of taurohyodeoxycholic acid on biliary lipid secretion in man: preliminary report.

Taurohyodeoxycholic acid and tauroursodeoxycholic acid were infused intraduodenally at a rate of 0.8 g/h for three hours in 3 cholecystectomized T-tube patients. Biliary lipid secretion and bile acid composition were evaluated before and after replacement of the endogenous bile acid pool with the two bile acids. As compared to basal values (2.78 +/- 1.67 mM/l), taurohyodeoxycholic acid induced a greater increase in the biliary concentration of phospholipids (4.12 +/- 1.23 mM/l) as compared to tauroursodeoxycholic acid (3.14 +/- 0.98 mM/l). Biliary cholesterol concentration after taurohyodeoxycholic acid (1.89 +/- 0.63 mM/l) was unchanged as compared to the pretreatment period (1.98 +/- 0.58 mM/l), while it decreased significantly after tauroursodeoxycholic acid (0.85 +/- 0.08 mM/I). Biliary cholesterol secreted per unit of bile acid was greater during taurohyodeoxycholic acid than during tauroursodeoxycholic acid, while the opposite was observed for the secretion of phospholipids.

Effect of glucagon on intracellular pH regulation in isolated rat hepatocyte couplets.

To elucidate mechanisms of glucagon-induced bicarbonate-rich choleresis, we investigated the effect of glucagon on ion transport processes involved in the regulation of intracellular pH (pHi) in isolated rat hepatocyte couplets. It was found that glucagon (200 nM), without influencing resting pHi, significantly stimulates the Cl-/HCO3- exchange activity. The effect of glucagon was associated with a sevenfold increase in cAMP levels in rat hepatocytes. The activity of the Cl-/HCO3- exchanger was also stimulated by DBcAMP + forskolin. The effect of glucagon on the Cl-/HCO3- exchange was individually blocked by two specific and selective inhibitors of protein kinase A, Rp-cAMPs (10 microM) and H-89 (30 microM), the latter having no influence on the glucagon-induced cAMP accumulation in isolated rat hepatocytes. The Cl- channel blocker, NPPB (10 microM), showed no effect on either the basal or the glucagon-stimulated Cl-/HCO3 exchange. In contrast, the protein kinase C agonist, PMA (10 microM), completely blocked the glucagon stimulation of the Cl-/HCO3- exchange; however, this effect was achieved through a significant inhibition of the glucagon-stimulated cAMP accumulation in rat hepatocytes. Colchicine pretreatment inhibited the basal as well as the glucagon-stimulated Cl-/HCO3- exchange activity. The Na+/H+ exchanger was unaffected by glucagon either at basal pHi or at acid pHi values. In contrast, glucagon, at basal pHi, stimulated the Na(+)-HCO3- symport. The main findings of this study indicate that glucagon, through the cAMP-dependent protein kinase A pathway, stimulates the activity of the Cl-/HCO3- exchanger in isolated rat hepatocyte couplets, a mechanism which could account for the in vivo induced bicarbonate-rich choleresis.

Acute ethanol hepatotoxicity is modulated by bile salt hydrophilic-hydrophobic properties.

The influence of the hydrophobic-hydrophilic properties of bile salts (BS) on acute ethanol hepatotoxicity was investigated. Bile flow, biliary BS secretion and enzyme (LDH,AST) release in the perfusate were measured before and after exposure to low (0.1%) or high (1%) doses of ethanol in in vitro isolated livers perfused with 1 microM/min taurocholate (TCA), tauroursodeoxycholate (TUDCA) or taurodeoxycholate (TDCA). Ethanol promotes a rapid decrease of basal bile flow and BS secretion in TCA-perfused livers [-28% of basal values with 0.1% (N = 6), and -35% with 1% ethanol (N = 6)]. Bile flow and BS secretion were minimally decreased by ethanol in livers perfused with a hydrophilic BS (TUDCA) [-8% decrease of basal values with 0.1% ethanol (N = 6), and -10% with 1% ethanol (N = 9); p < 0.02 vs TCA-perfused livers]. In contrast, when livers were perfused with a hydrophobic BS (TDCA), ethanol showed a higher cholestatic effect than either TCA- or TUDCA-perfused livers. Enzyme release in the perfusate was not modified by 0.1% ethanol, while 1% ethanol promoted a 4-5 fold increase in LDH and AST release in the perfusate of TCA-perfused livers with respect to a mere 2-fold increase in TUDCA-perfused livers and a 6-7 fold increase in TDCA perfused livers (p < 0.03). In conclusion, we showed that TUDCA almost completely counteracts the cholestatic and cytolitic effects promoted by ethanol in the isolated perfused rat liver.

Lamellar bodies coexist with vesicles and micelles in human gallbladder bile. Ursodeoxycholic acid prevents cholesterol crystal nucleation by increasing biliary lamellae.

The aggregative forms of lipids in human gallbladder bile and their relation to cholesterol crystallization are controversial. Using combined chemical, gel-chromatographic, optical/electron microscopic and quasielastic light-scattering methods, we investigated this issue in native gallbladder bile obtained from nine untreated cholesterol gallstone patients and eight cholesterol gallstone patients treated for 1 week with 600 mg/day of ursodeoxycholic acid. Bile obtained at cholecystectomy was ultracentrifuged for 2 h at 150,000 g to obtain isotropic samples. The conventional cholesterol crystal observation time was 3.1 +/- 4.1 (SD) days in controls and 19.0 +/- 1.9 days in the ursodeoxycholic acid-treated group (p < 0.001). Bile was analyzed by high-resolution gel-chromatography using 7 mM sodium taurocholate in the elution buffer. Biliary lipids eluted in four chromatographic zones: zone #I, corresponding to the column void volume, contained only minimal amounts of lipids; zone #II (apparent m.w. 100-220 kDa) comprised 29.1 +/- 12.4% of biliary cholesterol in the untreated group and 8.3 +/- 4.3% in the ursodeoxycholic acid-group (p < 0.001). At negative staining electron microscopy, this region was composed of roundish vesicles ranging from 7 to 20 nm in diameter. Zone #III (apparent m.w. 50-100 kDa) carried 59.1 +/- 2.1% of cholesterol in untreated patients and 81.2 +/- 9.5% in ursodeoxycholic acid-rich biles, respectively (p < 0.001). At negative staining electron microscopy, this region was composed of lamellar stacks of variable length, usually with 5 nm interspaces and up to 30 nm in width. In ursodeoxycholic acid-rich biles, lamellae often appeared in the form of concentric fingerprint-like images. Quasielastic light-scattering measurements in this region were compatible with the size estimates obtained at electron microscopy. Zone #IV (apparent m.w. 6-50 kDa) carried 11.8 +/- 9.4% and 11.6 +/- 9.0% of cholesterol, respectively (not significant). Since this region comprised a considerable fraction of endogenous bile salts and had no distinct morphological structures, it was interpreted as mixed micelles. The cholesterol crystal observation time showed a significant inverse correlation (r = -0.85, p < 0.001) with percent cholesterol carried by vesicles (zone #II) and a direct correlation (r = 0.86, p < 0.001) with percent cholesterol carried by lamellar bodies (zone #III). Vesicles and lamellae identical to those observed in isolated gel-chromatographic fractions were observed also on direct electron microscopic examination of unfractionated isotropic native biles. Similar findings were observed also in matched model biles.(ABSTRACT TRUNCATED AT 400 WORDS)

Functional and ultrastructural features of ethanol/bile salts interaction in the isolated perfused rat liver.

We investigated whether bile salts (BS) with different hydrophobic-hydrophilic properties interact with ethanol on bile secretion, enzyme (aspartate transaminase [AST], lactate dehydrogenase [LDH]) release in the perfusate, liver ultrastructure, and vesicular exocytosis in the isolated perfused rat liver. Ethanol (0.1 or 1%) promoted a rapid decrease of bile flow and BS secretion in livers perfused with taurocholate (TCA), the physiologic BS in the rat (-28% decrease of baseline values with 0.1% and -34% with 1% ethanol). The inhibitory effect of ethanol on bile flow and BS secretion was significantly (P < .02) attenuated by perfusing liver with the hydrophilic BS, tauroursodeoxycholate (TUDCA), and it was exacerbated (P < .02) by perfusion with the hydrophobic BS, taurodeoxycholate (TDCA). The release of AST and LDH in the perfusate was unaffected by 0.1% ethanol, but increased threefold to fivefold by 1% ethanol in TCA-perfused livers. This cytolitic effect of ethanol was not observed in TUDCA-perfused livers, but it was enhanced (P < .03) by perfusion with TDCA. No ultrastructural abnormalities were found in either TCA- or TUDCA-perfused livers, with or without 1% ethanol. Only minimal changes were found in livers perfused with TDCA alone, but, in the presence of TDCA, 1% ethanol induces marked mitochondrial damage. The biliary excretion of the fluid phase marker horseradish peroxidase was inhibited by ethanol, an effect reversed by TUDCA (P < .02) and exacerbated by TDCA (P < .04). In conclusion, this study demonstrates that hydrophilic BS such as TUDCA counteract the inhibitory effect of ethanol on bile secretion and vesicular exocytosis as well as the ethanol-induced cytolitic effect in the isolated perfused rat liver. In the presence of hydrophobic BS such as TDCA, the exposure to ethanol promotes a marked inhibition of bile secretion and vesicular exocytosis as well as prominent mitochondrial damage.

Effect of taurohyodeoxycholic acid, a hydrophilic bile salt, on bile salt and biliary lipid secretion in the rat.

Taurohyodeoxycholic acid is a natural 6 alpha-hydroxylated bile acid with an apparent hydrophilicity intermediate between those of tauroursodeoxycholic and taurocholic acids. We investigated in the rat the hepatobiliary metabolism, choleretic properties, and biliary maximum secretory rate (SRmax) of taurohyodeoxycholic in comparison with these two bile salts. Each compound was infused intravenously, at a rate increased in a stepwise manner from 100 to 300 nmol/min/100 g body wt, in bile salt-depleted bile fistula rats. The three bile salts appeared rapidly starting with the infusion and increased to represent more than 95% of the total bile salts. No apparent biliary metabolites were formed. All the bile salts caused a dose-dependent increase in bile flow and biliary lipid output. The absolute increase in bile flow was lower in rats infused with taurohyodeoxycholic acid, yet the volume of bile formed per nanomole of secreted bile salt was 13.8 nl for taurohyodeoxycholic, 6.4 nl for tauroursodeoxycholic acid, and 10.9 nl for taurocholic. The SRmax values were 1080, 3240, and 960 nmol/min/100 g, respectively. At all infusion rates, taurohyodeoxycholic acid caused a greater (P < 0.001) secretion of biliary lecithin compared to the other bile salts. There were no significant differences in the biliary secretion of cholesterol and proteins. Electron microscopy showed the recruitment of vesicles and lamellar bodies around and within bile canaliculi. In conclusion, taurohyodeoxycholic promotes a biliary lecithin secretion greater than expected from physicochemical predictions, representing a novel secretory property with potential pharmacological relevance.

Development and validation of a quantitative assay for cholesterol crystal growth in human gallbladder bile.

Crystal observation time is a rough estimate of the first microscopic appearance of cholesterol monohydrate crystals in an isotropic bile, and does not provide information on crystal growth kinetics. We have developed a method for quantitating cholesterol crystal growth in gallbladder bile. Crystals were separated from other biliary particles by ultracentrifugation on a discontinuous NaBr gradient, after bile density adjustment to d = 1.060 g/ml. More than 95% of crystals, both of native or synthetic source, floated in the density range 1.045-1.055. This density fraction was collected and crystal mass was measured by photometric turbidity, after calibration with suspensions of different-sized cholesterol crystals. The recovery of crystals added to original bile samples averaged 96.0 +/- 2.8%. Contamination with vesicles, which may potentially interfere with the turbidimetric assay, was excluded by gel-chromatography. The method was sequentially applied, until the 20th day of incubation, to biles obtained at surgery from patients with (A, n = 6) or without cholesterol gallstone (B, n = 4), and from gallstone patients pretreated for 1 week with oral ursodeoxycholic acid (C, n = 5). Crystal growth curves greatly differed, being much steeper in group A and almost flat in patients receiving ursodeoxycholic acid. The mean percent mass of biliary cholesterol in crystalline form at the 20th day was 19.2 +/- 13.5%, 1.2 +/- 0.8% and 2.7 +/- 1.1% in A, B and C, respectively (A vs. B: P = 0.014; A vs. C: P = 0.008). We conclude that the method allows a precise estimate of cholesterol crystal growth and can be usefully applied to human gallbladder biles.