Hepatic transport of the magnetic resonance imaging contrast agent gadobenate dimeglumine in the rat

Hepatic Transport of the Magnetic Resonance Imaging Contrast Agent Gadobenate Dimeglumine in the Rat Christoph de Haen, PhD, Vito Lorusso, PhD, Franco Luzzani, PhD, Piero Tirone, PhD

Rationale and Objectives. Gadobenate dimeglumine is a new octadenrate gadolinium (III) complex salified with meglumine. The compound is currently under evaluation as an intravenously administered paramagnetic contrast agent for magnetic resonance (MR) imaging. We investigated the mechanisms involved in the biliary excretion of gadobenate ion, the contrast-effective ion in gadobenate dimeglumine. Methods. Biliary and urinary excretion of gadobenate ion injected intravenously to rats at 0.25 mmol/kg was studied following pretreatment with bromosulfophthalein (BSP) disodium salt, sodium taurocholate (TC), or oxyphenonium bromide (OP) and at various times after common bile duct ligation. Gadobenate ion was assayed by high-pressure liquid chromatography in bile and urine. Plasma bilimbin levels after duct ligation were measured by colorimetric assay. Results. The 90-min excretion of gadobenate ion into bile accounted for 35.5 -+ 3.7% and excretion into urine for 45.7 -+ 3.5% of the injected dose (mean _+ SD). Pretreatment with BSP reduced recovery of the compound in bile to less than 1% and increased urinary excretion to 65.6 + 4.7%. Gadobenate dimeglumine had a substantial choleretic effect that was completely abolished by pretreatment with BSP. Pretreatment with TC and OP did not change the biliary or urinary excretion of gadobenate ion. Surgical cholestasis led to a massive increase in plasma bilirubin levels from 3.9 -+ 2.2 (day of surgery) to 129 _+ 37 btmol/L (4 days after common bile duct ligature) and decreased 6-hr cumulative biliary excretion of gadobenate ion from 45 --+- 16% to 5.3 -+ 4.2% of the injected dose. Urinary excretion increased correspondingly from 49 _+ 15% to 83 -+ 12%. Conclusion. The transport of gadobenate ion from plasma to bile occurs in the rat mainly through the BSP/bilirubin transport systems. Key Words. Magnetic resonance imaging contrast media; gadolinium; liver magnetic resonance imaging; liver transport systems. adobenate dimeglumine is a new octadentate chelate of gadolinium (III) [1-3], salified with meglumine, that is currently under evaluation as an intravascularly administered paramagnetic contrast agent for magnetic resonance (MR) imaging. In Tl-weighted MR images, gadobenate dimeglumine

G 232

From the Research and Development Division, Magnetic resonance, Bracco S.p.A., Milan, Italy. Address reprint requests to C. de Ha6n, PhD, Research and Development Division, Preclinical Research, Bracco S.p.A., Via E. Folli 50, 1-20134 Milan, Italy. Acad Radio11995;2:232-238 © 1995, Association of University Radiologists

Vol. 2, No~3, March 1995

enhances the signal of liver parenchyma in rodents [4-6] and in humans [7]. Signal enhancement of the bile is complex. At a low concentration of gadobenate ion, the contrast-effective ion in gadobenate dimeglumine, the bile signal is also enhanced. However, transiently, gadobenate ion may accumulate in bile to concentrations at which its effects on T2 relaxation predominate. Consequently, in T1weighted images the bile signal may even be abolished [8]. After intravenous administration of gadobenate dimeglumine to rats, gadobenate ion is excreted in unaltered form both in the urine and in the rat bile, the extent in the bile reaching approximately 40% in 6 hr at a dose of 0.5 mmol/kg [3]. In rabbits and humans, biliary excretion is much more modest, without detriment to the MR signal enhancement of the liver. In any case, the biliary excretion of gadobenate ion remains above that observed with gadopentetate ion. The latter compound to a minimal extent also appears in bile because of some permeability of the tight junction complexes that hold hepatocytes and bile ductulus epithelial cells together, and because of some piggyback transcytosis. To be transported actively from blood to bile, small to midsized organic molecules have to pass through the endothelial fenestrae (diameter -1200 A) of hepatic sinusoids into the space of Disse. After diffusion through the loose extracellular matrix in this space [9], the molecules must be recognized by proteinaceous transport systems in the sinusoidal plasma membrane that mediate entry into hepatocytes. Four transport systems have been described so far, named for the substrate classes they recognize: the bile acid transporter, the bilirubin transporter, the organic cation transporter, and the fatty acid transporter. The first three are promiscuous. In particular, the bilirubin transporter is also responsible for the uptake of bromosulfophthalein (BSP). Moreover, the bilirubin transporter has been implicated in the transport of a number of other anionic drugs, imaging agents for nuclear medicine, and dyes [10-12]. To pass from the hepatocytic cytoplasm into the bile, molecules must traverse the bile canalicular plasma membrane, usually against a concentration gradient. Adenosine triphosphate-dependent pumping, exocytosis, and membrane-potential-driven facilitated transport contribute to this process, depending on the molecule. A number of proteinaceous canalicular transporters for organic molecules have been identified [13]. These, together with pumps and exchangers of inorganic ions, are also important in bile formation. Among them, the canalicular transporter for organic anions such as BSP

GADOBENATE ION HEPATIC T R A N S P O R T

appears to be distinct from that for bile acids [14] and from that for bilirubin diglucuronide [15]. Preliminary results from our laboratories have shown that BSP can inhibit the enhancement of hepatic MR signal intensity by gadobenate ion [16]. To determine which transport systems are involved in the transit of gadobenate ion from blood plasma to bile in rats, we used two experimental approaches. The first experiment tested the capacity of known exogenous substrates of the various transporters to inhibit the biliary excretion of gadobenate ion. The second experiment measured the biliary and urinary excretion of gadobenate ion at enhanced plasma levels of the endogenous substrate bilirubin, which were brought about by common bile duct ligature. MATERIALS AND METHODS Compounds Details on the preparation of the gadolinium complex are provided elsewhere [1, 2]. The product was formulated as a sterile and apyrogenic 0.25 M aqueous solution. BSP disodium salt was purchased from Fluka AG (Buchs, Switzerland), taurocholic acid sodium salt (TC) and oxyphenonium bromide (OP) from Sigma Chemical Co. (St. Louis, MO). The formulations of BSP, OP, and TC used in the study were the following: BSP disodium salt, 2.2% w/v solution in distilled water; OP, 0.8% w/v solution in sterile physiological saline; TC, 2.4% w/v solution in sterile physiological saline. Sterile physiological saline was purchased from SIFRA (Verona, Italy), fentanyl-droperidol (Leptofen) from Carlo Erba (Milan, Italy). All other reagents were analytical grade. Animals Male Sprague-Dawley rats (Charles River Italia, Calco, Italy; body weight = 230-410 g) were used. Five animals per group were used in BSP, TC, and OP competition experiments; five to eight animals per group underwent bile duct ligation and were used in bilirubin competition studies. EXPERIMENTAL DESIGN BSP, TC, and OP Competition Study The rats, which had been fasted overnight, were anesthetized with 1.2 ml/kg of intramuscular fentanyldroperidol. The common bile duct was cannulated with an Intramedic PE50 polyethylene catheter (Becton-

233

DE

HAI~N ET

AL,

Dickinson, Parsippany, NJ). Another PE50 polyethylene catheter was inserted into the urinary bladder. The jugular vein was cannulated with a PE50 polyethylene catheter for infusion of saline, BSP, or TC solutions. Solution of OP instead was administered into the jugular vein as a bolus injection. Bile was collected for 30 rain to evaluate basal flow. Subsequently, 4 ml of saline (control animals), 4 ml of BSP, or TC were infused at a rate of 2 ml/hr (Injector Model 50-4928 Harvard, Dover, MA). This rate corresponds to an infusion of 2.7 and 7.8 [,tmol.kg-l.min-1 for BSP and TC, respectively. OP was administered as a bolus at a dose of 18.7 btmol/kg; thereafter, animals were infused with saline at the same rate as with the other drugs. After infusing the animals for 30 min with BSP, TC, or saline, gadobenate dimeglumine (0.25 mmol/kg) was injected into the tail vein at a rate of 6 ml/min. While continuing the infusions, bile was collected immediately before product administration and every 15 rain up to 90 min thereafter. Urine was collected twice, once for 30 min before the administration and once from 0 to 90 min after the administration of gadobenate dimeglumine.

Vol. 2, No. 3, March 1995

Gadobenate Ion Assay The assay of gadobenate ion in the bile and urine samples collected during the studies was p e r f o r m e d by reversed-phase ion-pair high-pressure liquid chromatography with ultraviolet detection at 210 n m [17].

Bilirubin Assay Total (free, albumin-bound, and conjugated with glucuronic acid) bilirubin in plasma was measured by a standard colorimetric assay based on the reaction of bilirubin with 2,5-dichlorophenyl diazonium salt and absorbance m e a s u r e m e n t at 550 nm (total bilirubin assay no. 19717, E. Merck, Darmstadt, Germany).

Histology At the end of the experiments, the animals were exsanguinated via the catheter inserted in the carotid artery. The liver was excised and processed through paraffin wax. After the organ was sectioned and stained with hematoxylin and eosin, it was examined microscopically.

Statistics Bilirubin Competition Study The animals were anesthetized with 1.2 ml/kg of intramuscular fentanyl-droperidol. The c o m m o n bile duct near the opening into the d u o d e n u m was cannulated with a 1-cm-long PE50 polyethylene catheter that was connected to a 10-cm-long Silastic catheter (0.64 m m inside diameter, 1.19 m m outside diameter, D o w Coming, Midland, MI) closed at the free end with a silk wire ligature, thus producing biliary obstruction. The closed catheter was introduced into a pouch made surgically between skin and abdominal musculature, and the abdomen was closed. Control animals (sham-operated) were subjected to all surgical manipulations except duct ligation and pouch formation. One, 4, 8, and 28 days after duct ligation, the Silastic catheter tip was cut for the collection of bile. The urinary bladder and carotid artery were catheterized (PE50 polyethylene catheter) for the collection of urine and blood samples. Bile was collected for 6 hr immediately after the intravenous administration of gadobenate dimeglumine (0.25 mmol/kg). Likewise, urine was collected from 0 to 6 hr after c o m p o u n d administration. Five minutes before treatment, blood samples (0.5 ml) were withdrawn into heparinized tubes for the assay of total bilirubin in plasma.

234

The data were analyzed statistically by analysis of variance (ANOVA). Where a comparison b e t w e e n a selected control group and all other groups was. required, Dunnett's t test was used. In the bilirubin competition study, ANOVA was performed after transforming the data by raising to a suitable p o w e r (0.1 for bilirubin data, 0.9 for biliary and urinary g a d o b e n a t e dimeglumine data) to ensure homogeneity of the variances and normalize the distribution of the data [18]; multiple comparison between experimental groups was done b y Scheff~'s test. A value less than .05 was considered statistically significant.

RESULTS The inhibition of biliary excretion of g a d o b e n a t e ion was studied in competition experiments with TC, BSP, and OP, which are the substrates for the bile acid, bilirubin, and organic cation transporters, respectively. The values for cumulative biliary and urinary excretion of gadobenate ion, expressed as percentage of administered dose, are shown in Figure 1. TC and OP did not affect the excretion pattern of gadobenate ion. In contrast, BSP reduced the amount of c o m p o u n d recovered in bile to almost undetectable levels (less than 1%). In

Vol. 2, No. 31 March 1995

GADOBENATE

80

o

ION HEPATIC

TRANSPORT

100

J

80

'.~

6o

60

0)

mm;

40

"0

Urine

,J 40 o

20

0

i

-60

Bile

-30

i

i

i

i

0

30

60

90

Time (rain) FIGURE 1. Ninety-minute cumulative biliary and urinary excretion of gadobenate after IV bolus injection of gadobenate dimeglumine (0.25 mmol/kg) in rats with coadministration of saline (C]), sodium taurocholate ( ~ ) (7.8 #mol.kg-l.min -1, constant infusion rate, starting 30 min before gadobenate dimeglumine injection), oxyphenonium bromide ( [ ] ) (18.7 ~mol.kg -1 , bolus injection 30 min before gadobenate dimeglumine injection followed by saline infusion) and bromosulfophthalein (11) (2.7 #mol-kg-l-min -1, constant infusion rate, starting 30 min before gadobenate dimeglumine injection). Data are expressed as means + standard deviations from five animals. *p <.01 versus controls.

• OControl rats; O-obromosulfophthalein (BSP)-treated rats. The infusions of saline (control) and BSP (BSP-treated) were begun at -30 min and continued throughout the experiment• The IV bolus injection of gadobenate dimeglumine (0,25 mmol/kg) was administered at 0 min, Data are means _+standard deviations from five animals, *p < .05 versus basal.

compensation, urinary excretion of the compound was significantly enhanced. The effect of gadobenate dimeglumine on the biliary flow of animals perfused with saline (control) or BSP solution is shown in Figure 2. In agreement with earlier findings [19], BSP per se was not choleretic. After intravenous injection of gadobenat e dimeglumine (0.25 mmol/kg), bile flow was increased in the mean by 31 btl.min q . k g -1 and remained significantly above basal values for at least 90 min. The choleretic effect of gadobenate dimeglumine was completely abolished by BSP infusion.

In control (sham-operated) rats, liver histology was normal, with only minimal muhifocal inflammatory cell infiltration. In rats with total biliary obstruction, minimal bile duct hyperplasia and slightly increased cellular mitosis were already noticeable after 4 days. The hyperplastic reaction progressively increased. These observations agree with those of others [20] and confirm the success of surgery and maintenance of duct ligature throughout the intended period. Common bile duct ligature in rats led to a marked increase in plasma bilirubin levels (Table 1). In sham-

FIGURE 2. Effect of gadobenate dimeglumine on rat biliary flow.

/

TABLE 1: Bilirubin Plasma Levels and (0 to 6 hr) Cumulative Biliary and Urinary Excretion of Gadobenate Ion in Normal Rats and in Rats with Ligature of the Bile Duct i

Animals

N o n o p e r a t e d (5) a S h a m - o p e r a t e d (8) Fully operated Day 1 after surgery (5) D a y 4 after surgery (8) Day 8 after surgery (8) Day 28 after surgery (6)

G a d o b e n a t e Ion % of Injected Dose

Total Bilirubin (~mol/I) 5.3 _+ 4.2 b 3.9 + 2.2 82 129 102 30

_+ 13 c + 37 c + 58 c + 23 c

Bile

Urine

45 + 16

49 +_ 15

12.8 4.9 9.3 42

+ + + +

5.6 c 3.9 c 9.3 c 10

80.7 83 80 44

+ + + +

8.8 c 12 c 12 c 11

Gadobenate dimeglumine was given at a dose of 0.25 mmoi/kg. aNumber of animals in parentheses. bData are means + standard deviations. Cp < .01 vs sham-operated (Scheffe's test).

235

DE

HAI~N ET

AL.

operated animals, values did not differ from those of nonoperated animals. After bile duct ligation, the increment was already marked 1 day after surgical intervention. It reached its maximum after around 4 days, after which values progressively decreased. After 28 days the plasma levels of bilirubin were, however, higher than those of sham-operated animals. The amount of gadobenate ion recovered in bile with restored flow, expressed as percentage of the administered dose, was negatively correlated ( r = -.78, p < .01) with plasma bilirubin levels. Biliary gadobenate ion in animals with restored bile flow progressively decreased with increased duration of the bile duct ligation up to 4 days after surgery. Twenty-eight days after bile duct ligation, the amount of gadobenate ion excreted into bile had returned to levels similar to those measured in shamoperated animals despite persistence of slightly elevated bilirubin levels. Mirroring the biliary excretion pattern, urinary excretion of gadobenate ion in animals with ligated bile duct was enhanced relative to that in sham-operated animals to a maximum on day 4 (p < .01) and returned to essentially basal levels on day 28.

DISCUSSION

After intravenous administration of gadobenate dimeglumine, gadobenate ion is rapidly distributed in all the extracellular space both in animals and in humans. The rapid onset of an enhanced signal in Tl-weighted magnetic resonance images of the liver in excess of what can be accounted for by the concentration of gadobenate ion in blood and extracellular spaces [5] provides evidence of fast hepatocellular uptake. In animals, gadobenate ion is excreted massively (25-50% of the injected dose) into the bile [5]. Because signal enhancement is present even when the bile signal is completely abolished by excessive T2 relaxation at a high biliary concentration of gadobenate ion, signal enhancement must originate primarily inside hepatocytes [8]. The elimination of gadobenate ion in humans occurs more strongly by the urinary route (76-94% of the injected dose in 24 hr) and to a lesser extent through the biliary route (2-4% of the injected dose, as recovered in feces in 24 hr) [21]. In spite of the relatively low biliary excretion, significant liver signal enhancement is observed also in humans [7]. In isolated hepatocytes, BSP and bilirubin show mutually competitive uptake inhibition, which is not altered by physiologic concentrations of TC [12]. These findings indicate that BSP and bilirubin share a common sinusoi-

236

vol. 2, No. 3, March 1995

dal transport mechanism, which is different from that of TC. Tt~e fate of the organic anions inside the hepatocytes is complex and comprises weak interactions with specific cytoplasmic proteins. In the intact liver, organic anions are translocated to the pericanalicular spaces and appear in the bile [14]. Distinct carrier systems for bile acids and several organic anions and cations have been characterized in canalicular membranes [14]. Also, for the transport system on the canalicular side of hepatocytes, BSP and bilirubin compete, at least if the latter has been converted to the diglucuronide [15]. In our studies, we observed that the administration of BSP at concentrations known to be saturating its transit through hepatocytes [22] prevented and high plasma bilirubin levels drastically reduced the excretion of gadobenate ion into bile. TC and OP were without effect. This finding means that gadobenate ion uses the transport systems of BSP and bilirubin to pass from blood plasma to bile and excludes participation of the bile acid transporters and the cation transporters. Essentially the same results were recently reported with gadolinium-ethoxybenzyl-DTPA, a new hepatobiliary MR imaging contrast agent currently under clinical evaluation [23]. Although BSP inhibits excretion of gadobenate ion into bile completely (Fig. 1), liver MR signal enhancement by gadobenate ion is slowed down and reduced in intensity only by one third in the presence of BSP [16]. Thus, gado-" benate ion may enter hepatocytes in the presence of BSP, although it seems to do so more slowly. A possible interpretation is a competition between BSP and gadobenate ion for a common transporter in the sinusoidal plasma membrane. BSP and bilimbin are transported in that membrane by a transporter that primarily facilitates flux down a concentration gradient [24]. This transporter should therefore also allow gadobenate ion to return to plasma when plasma levels fall. Although the augmentation of water proton magnetic relaxivity of gadobenate dimeglumine by serum albumin [6] indicates a weak interaction between these molecules, gadobenate ion does not bind to plasma albumin to an extent measurable by equilibrium dialysis [2]. In contrast, BSP and bilirubin circulate tightly complexed with albumin. Gadobenate ion, as well as BSP and bilirubin, are rapidly extracted by the liver and excreted into the bile, indicating that, contrary to widely held belief, albumin binding is not a prerequisite for biliary excretion. In fact, certain organic anions are readily taken up by hepatocytes even in the absence of albumin [25], and increasing concentrations

Vol. 2, No. 3, March 1995

of albumin can slow down biliary excretion of albumin binding anions at a fixed concentration [26]. Like many bilitropic substances, BSP and bilirubin undergo biotransformation within the cell prior to their biliary excretion. Gadobenate ion, however, is excreted unchanged into bile [5]. This feature and others, such as the weak or absent binding to albumin, are common to other hepatobiliary MR imaging contrast agents such as gadolinium-ethoxybenzyl-DTPA [27], Fe(III)-N-(3phenylglutaryl)desferrioxamine B (Fe-PGDF) [28], and scintiscanning agents [11]. All experimental evidence, obtained both in intact animals and in isolated hepatocytes, points to shared mechanisms of liver transit to bile for these agents. Extrahepatic biliary obstruction is expected to lead to hyperbilirubinemia [29]. In our study, we measured a progressive increase in total bilirubin plasma levels, with a maximum on day 4 after common bile duct ligation. Based on the above hypothesis about the transporters for gadobenate ion, biliary excretion of gadobenate ion on restoration of bile flow is expected to be considerably reduced. Our study confirmed this assumption. On day 28, bilirubin levels had decreased substantially. At that time the amount of gadobenate ion recoverable in bile and urine had returned to levels similar to those of control animals. Apparently, biliary excretion of gadobenate ion was little affected by bilirubin plasma levels that were only modestly elevated relative to those in control animals. This observation could be the result of effective competition of gadobenate ion with bilirubin for the transport system. Another possible explanation is that gadobenate ion could have entered the bile, mainly by the paracellular pathway through swollen tight junctions, without penetrating hepatocytes. Alterations in tight junction structure and function develop within a few days following bile duct ligation in rats, and even large proteins such as horseradish peroxidase can penetrate the tight junction from blood to bile, reflecting disruption of that permeability barrier [30]. If the degree of inhibition of biliary excretion of gadobenate ion as a function of the plasma concentration of bilirubin (Table 1) is analyzed for cooperative interaction of bilirubin with the rate-limiting transporter in the transit pathway, an apparent inhibitory concentrat.ion of 50% of about 60 I.tmol/l is obtained. This value is remarkably similar to the 75 I.tmol/1 value of the Michaelis-Menten constant K m characterizing the transport of bilirubin diglucuronide in the canalicular membrane [15]. The value is substantially higher than the value of the inhibi-

GADOBENATE

ION HEPATIC TRANSPORT

tion constant Ki = 10 gmol/l, describing inhibition by bilirubin of BSP uptake into isolated hepatocytes [31]. This finding suggests that biliary excretion of gadobenate ion is inhibited on the canalicular side by bilirubin after biotransformation to the diglucuronide. Thus, as for most hepatobiliary products, for gadobenate ion the rate-limiting step for biliary excretion seems to be on the bilecanalicular side of hepatocytes. Gadobenate dimeglumine had a choleretic effect that was completely abolished by BSP. Bile formation is believed to involve the pumping of osmolytes such as bile acids or glutathione across the bile-canalicular membrane followed by osmotic water recruitment from hepatocytes, through tight junctions from the space of Disse [32] and from ductular epithelium [33]. Gadobenate ion might therefore be choleretic by a similar mechanism, acting as an osmolyte actively pumped through the canalicular plasma membrane. BSP, or its conjugate with glutathione formed in hepatocytes, inhibits the appearance of gadobenate ion in the bile. Thus, BSP-mediated inhibition of excretion of gadobenate ion and inhibition of gadobenate ion-induced bile flow might involve the same mechanism. In conclusion, the results strongly suggest that biliary excretion of gadobenate ion occurs in rats mainly through specific transport systems of hepatocytes shared with BSP and bilirubin. ACKNOWLEDGMENTS

We thank Dr. Luigia Gozzini for a critical review of the manuscript. REFERENCES 1. Felder E, Uggeri F, Fumagalti L, Vittadini G (Bracco SpA, Milan, Italy). Paramagnetic chelates useful for NMR imaging. U.S. Patent 4,916,246 (CO7F13/00; 556/1 ) 10 April, 1990, IT 19236A/86 30 January 1986. 2. Davies A, de Hahn C. Gd-BOPTA/Dimeg. Drugs Fut1991;16:1001-1003. 3. Vittadini G, Felder E, Tirone P, Lorusso V. B-19036, a potential new hepatobiliary contrast agent for MR proton imaging. Invest Radiol 1988; 23(suppl 1):$246-$248. 4. Pavone P, Patrizio G, Buoni C, et al. Comparison of Gd-BOPTA with GdDTPA in MR imaging of rat liver. Radiology 1990;176:61-64. 5. Cavagna F, Tirone P, Felder E, de Hahn C. Hepatobiliary contrast agents for MRI. In: Ferrucci JT, Stark DD, eds. Liver imaging: current trends and new techniques. Boston, MA: Andover Medical Publishers, 1990:384-393. 6. Cavagna F, Tirone P, Felder E, de Hahn C. Gd-BOPTA/Dimeg, a new hepatobiliary contrast agent for magnetic resonance imaging. In: Rinck PA, Muller RN, eds. New developments in contrast agent research. Blonay, Switzerland: European Magnetic Resonance Forum, 1991:83-94. 7. Vogl TJ, Pegios W, McMahon C, et al. Gadobenate dimeglumine--a new contrast agent for MRI imaging: preliminary evaluation in healthy volunteers. AJR 1992; 158:88"7-892. 8. Cavagna F, Lorusso V, Maggioni F, Visentin G, de Hahn C. Mechanism of liver MRI signal enhancement by Gd-BOPTNDimeg. J Magn Reson Imaging 1993;3(P):$27 (abstr 411).

237

DE HAI~N E T AL. 9. Reid LM, FiorinoAS, Sigal SH, Brill S, Hoist PA. Extracellular matrix gradients in the space of Disse: relevance to liver biology. Hepatology 1992;15:1198-1203. 10. Berk PD, Potter BJ, Stremmel W. Role of plasma membrane ligand-binding proteins in the hepatoceilular uptake of albumin-bound organic anions. Hepatology 1987;7:165-176. 11. Okuda H, Nunes R, Vallabhajosula S, Strashun A, Goldsmith SJ, Berk PD. Studies of the hepatocellular uptake of the hepatobiliary scintiscanning agent 99mTc-DISIDA. J Hepato11986;3:251-259. 12. Berk PD, Potter BJ, Sorrentino D, et al. Characteristics of organic anion binding proteins from rat liver sinusoidal plasma membranes, in: Petzinger E, Kinne RK-H, Sies H, eds. Hepatic transport of organic substances. Berlin: Springer-Verlag, 1989:195-210. 13. Zimniak P, Awasthi YC. ATP-dependent transport system for organic anions. Hepatology 1993; 17:330-339. 14. Nathanson MH, Boyer JL. Mechanisms and regulation of bile secretion. Hepatology 1991; 14:551-566. 15. Adachi Y, Kobayashi H, Kurumi Y, Shouji M, Kitano M, Yamamoto T. Bilirubin diglucuronide transport by rat liver canalicular membrane vesicles: stim ulation by bicarbonate ion. Hepatology 1991; 14:1251-1257. 16. Lorusso V, Poggesi I, Cavagna F, de Hahn C, Tirone P. Gd-BOPTA/Dimeg: relationship between pharmacokinetics and MRI liver signal enhancement. Eur Radio11991;1(suppl):S218 (abstr 1234). 17. Lorusso V, Poggesi I, Arbughi T, Dal Fiume D, Tirone P. High-performance liquid chromatographic assay of magnetic resonance imaging contrast agent gadobenate (Gd-BOPTA 2-) in plasma, urine, and bile. J Chromatogr 1994;656:415-422. 18. Armitage P. Statistical methods in medical research. Oxford, England: Blackwell Scientific, 1971. 19. WareAJ, Carely MC, Combes B. Solution properties of sulfobromophthalein sodium (BSP) compounds, alone and in association with sodium taurocholate (TC). J Lab Clin Med 1976;87:443-456. 20. Slott PA, Liu MH, Tavoloni N. Origin, pattern, and mechanism of bile duct proliferation following biliary obstruction in the rat. Gastroenterology1990;99:466-477. 21. Lorusso V, Pirovano G, Tirone P, Rosati G. Pharmacokinetics and tolerance of Gd-BOPTNDimeg. J Magn Reson Imaging 1992;2(P):$18 (abstract 233).

238

vol. 2, No. 3, March 1995 22. Hoener BA, Engelstad BL, Ramos EC, et al. Comparison of Fe-HBED and Fe-EHPG as hepatobiliary MR contrast agents. J Magn Reson Imaging 1991 ;1:357-362. 23. Clement O, Mehler A, Vexler V, Berthez~ne Y, Brasch RC. GadoliniumethoxybenzyI-DTPA, a new liver-specific magnetic resonance contrast agent: kinetic and enhancement patterns in normal and choleatatic rats. Invest Radio11992;27:612-619. 24. Berk PD, Stremmel W. Hepatocellular uptake of organic anions. In: Popper H, Schaffner F, eds. Progress in liver diseases, vol VIII. New York: Grune & Stratton, 1986:125-144. 25. Inoue M, Okajima K, Nagase S, Morino Y. Plasma clearance of sulphobromophthalein and its interaction with hepatic binding proteins in normal and analbuminemic rats: Is plasma albumin essential for vectorial transport of organic anions in the liver? Proc Natl Acad Sci USA 1983;80: 7654-7658. 26. Song CS, Beranbaum ER, Rothschild MA. The role of serum albumin in the hepatic excretion of iodipamide. Invest Radio11976;11:39-44. 27. Schuhmann-Giampieri G, Schmitt-Willich H, Press W-R, Negishi C, Weinmann H-J, Speck U. Preclinical evaluation of Gd-EOB-DTPA as a contrast agent in MR imaging of the hepatobiliary system. Radiology 1992;183: 59-64. 28. Hoener B-A, Tzika AA, Engelstad BL, White DL. Hepatic transport of the magnetic resonance imaging contrast agent Fe(lll)-N-(3-Phenylglutaryl)desferrioxamine B. Magn Reson Med 1991;17:509-515. 29. Blanckaert N, Fevery J. Physiology and pathophysiology of bilirubin metabolism. In: Zakim D, Boyer TD, eds. Hepatology: a textbook of liver disease, 2nd ed., vol. 1. Philadelphia: Saunders, 1990:254-302. 30. Boyer JL. Tight junctions in normal and cholestatic liver: does the paracellular pathway have functional significance? Hepatology 1983;3:614-617. 31. Stremmel W, Berk PD. Hepatocellular uptake of sulfobromophthalein (BSP) and bilirubin is selectively inhibited by an antibody to the liver plasma membrane BSP/bilirubin protein. J Clin Invest 1986;78:822-826. 32. Ballatori N, Truong AT. Glutathione as a primary osmotic driving force in hepatic bile formation. Am J Physio11992;263:G617-G624. 33. Alpini G, Lenzi R, Zhai N-R, et al. Bile secretory function of intrahepatic biliary epithelium in the rat. Am J Physiol 1989; 257:G124-G 133.