• Aucun résultat trouvé

How transfer rates generate Gd-BOPTA concentrations in rat liver compartments: implications for clinical liver imaging with hepatobiliary contrast agents

N/A
N/A
Protected

Academic year: 2022

Partager "How transfer rates generate Gd-BOPTA concentrations in rat liver compartments: implications for clinical liver imaging with hepatobiliary contrast agents"

Copied!
9
0
0

Texte intégral

(1)

Article

Reference

How transfer rates generate Gd-BOPTA concentrations in rat liver compartments: implications for clinical liver imaging with hepatobiliary

contrast agents

PASTOR, Catherine

Abstract

Following the injection of hepatobiliary contrast agents, MRI detects all molecules included in a region of interest but cannot estimate true concentrations in sinusoids, interstitium, hepatocytes or bile canaliculi. The aim of the study was to measure true concentrations in hepatocytes and to show how transfer rates across sinusoidal and canalicular membranes generate these concentrations. We perfused livers isolated from normal rats with 200 μM Gd-DTPA and Gd-BOPTA and measured clearances from sinusoids to liver and from hepatocytes to bile canaliculi or back to interstitium. We detected Gd-BOPTA with a gamma probe and determined true concentrations in each liver compartment knowing their liver volumes. No pharmacokinetic modelling was applied. Gd-BOPTA clearance from sinusoids to liver (2.5 ± 0.4 mL/min) was 50 times higher than that of Gd-DTPA (0.05 ± 0.02 mL/min) when portal flow rate was 30 mL/min (p 

PASTOR, Catherine. How transfer rates generate Gd-BOPTA concentrations in rat liver

compartments: implications for clinical liver imaging with hepatobiliary contrast agents. Contrast Media & Molecular Imaging , 2016, vol. 11, no. 4, p. 291-298

DOI : 10.1002/cmmi.1691 PMID : 27060676

Available at:

http://archive-ouverte.unige.ch/unige:84627

Disclaimer: layout of this document may differ from the published version.

(2)

How transfer rates generate Gd-BOPTA concentrations in rat liver compartments:

implications for clinical liver imaging with hepatobiliary contrast agents

Catherine M. Pastor

a,b

*

Following the injection of hepatobiliary contrast agents, MRI detects all molecules included in a region of interest but cannot estimate true concentrations in sinusoids, interstitium, hepatocytes or bile canaliculi. The aim of the study was to measure true concentrations in hepatocytes and to show how transfer rates across sinusoidal and canalicular membranes generate these concentrations. We perfused livers isolated from normal rats with 200μM Gd-DTPA and Gd-BOPTA and measured clearances from sinusoids to liver and from hepatocytes to bile canaliculi or back to interstitium. We detected Gd-BOPTA with a gamma probe and determined true concentrations in each liver compartment knowing their liver volumes. No pharmacokinetic modelling was applied. Gd-BOPTA clearance from sinusoids to liver (2.5 ± 0.4 mL/min) was 50 times higher than that of Gd-DTPA (0.05 ± 0.02 mL/min) when portal flow rate was 30 mL/min (p<0.0001). Gd-BOPTA clearance from sinusoids to liver was always superior to hepatocyte clearance, and hepatocyte Gd-BOPTA concentrations measured by the probe increased over time. Gd-BOPTA concentrations reached 439 ± 83μM in hepatocytes and 15 × 700 ± 3100μM in bile canaliculi, while concentrations in sinusoids were 200μM. Gd-BOPTA true concentrations in hepatocytes depend on the simultaneous clearances from sinusoids to hepatocytes and from hepatocytes to bile canaliculi and back to sinusoids. The study better defines how signal intensities are generated when hepatobiliary contrast agents are injected in clinical imaging.

Copyright © 2016 John Wiley & Sons, Ltd.

Keywords:liver MRI; hepatobiliary contrast agent; hepatocytes; transfer rates; concentrations

1. INTRODUCTION

Over the past years, several publications have investigated the transporter-mediated hepatic pharmacokinetics of hepatobiliary contrast agents to detect and characterize focal lesions and to score liver function in chronic diseases (1–3). For liver magnetic resonance imaging (MRI), two hepatobiliary contrast agents (Gd-BOPTA, MultiHance®, Bracco Imaging, Milan, Italy, and Gd- EOB-DTPA, Primovist®, Bayer Pharma, Berlin, Germany) are available (4). In the liver, these agents distribute into sinusoids and interstitium before entering hepatocytes across organic anion transporting polypeptides (OATPs), to reach across the canalicular transporter multiple resistance-associated protein 2 (Mrp2) bile canaliculi (Fig. 1) (5–8). When transported inside hepatocytes, Gd-EOB-DTPA may return to sinusoids across the sinusoidal MRP3 transporters (9). Efflux transporters of Gd- BOPTA back to sinusoids are unknown.

The most useful pharmacokinetic parameters to obtain following the injection of hepatobiliary contrast agents are true concentra- tions generated in each liver compartment according to transfer rates between two adjacent compartments. Liver tissue is divided in sinusoids, interstitium, hepatocytes and bile canaliculi. In clinical MRI, concentrations from each compartment cannot be estimated and there are averaged by a single signal intensity measured in the entire region of interest (ROI). Transfer rates between two liver compartments are not measured but are estimated by pharmacoki- netic modelling (10–15). However, it is important to measure true

concentrations inside each compartment, because concentrations across membranes partly regulate transfer rates across them (16).

In liver tissue, signal intensities given by MR systems do not esti- mate true concentrations. In the common bile duct, signal intensi- ties estimate true concentrations because the structure has a single compartment. In portal and hepatic veins, signal intensities estimate true concentrations once the volume of blood cells has been subtracted. The aim of the study was to measure true concen- trations in hepatocytes and show how transfer rates across

* Correspondence to: C. Pastor, Département de Radiologie, Hôpitaux Universitaires de Genève, Switzerland. E-mail: catherine.pastor@hcuge.ch

a C. M. Pastor

Centre de Recherche sur lInflammation U1149 INSERM and University Paris-Di- derot, Paris, France

bC. M. Pastor

Département de Radiologie, Hôpitaux Universitaires de Genève, Switzerland

Abbreviations used:MRI, magnetic resonance imaging; OATP, organic anion transporting polypeptide; Mrp2, multiple resistance-associated protein 2; ROI, region of interest; KHB, KrebsHenseleit bicarbonate; Gd-DTPA, gadopentetate dimeglumine; Gd-BOPTA, gadobenate dimeglumine; CPV, concentration in por- tal vein,μM or nmol/mL; CHV, concentration in hepatic vein,μM or nmol/mL;

C1ELIMRC2, elimination rate from C1 to C2, nmol/min;C1CLC2, clearance from C1 to C2, mL/min (sinusoid) or g/min (liver);C1ERC2, extraction ratio from C1 to C2, %.

Received: 22 October 2015, Revised: 8 January 2016, Accepted: 17 February 2016, Published online in Wiley Online Library

(wileyonlinelibrary.com) DOI: 10.1002/cmmi.1691

(3)

sinusoidal and canalicular membranes generate these concentra- tions. No pharmacokinetic modelling was applied.

2. METHODS

2.1. Experimental groups

In a previous study (17), we published how Gd-BOPTA accumu- lates in rat livers in two experimental groups: normal Sprague- Dawley rats (B200,n= 5) and rats lacking Mrp2 (B200TR,n= 3).

We returned to raw data of these two groups to analyse and calculate all parameters available in the present manuscript (Fig. 1A). Thus, no additional rats were used in the present study.

2.2. Isolated and perfused rat livers

Rat livers were isolated and perfused with the experimental protocol previously published (17,18). Briefly, two catheters were introduced into portal vein and right atrium. The hepatic artery was not perfused. The system included a reservoir, pump, heating circulator, bubble trap, filter and oxygenator. Livers were perfused with a Krebs–Henseleit bicarbonate (KHB) solution

± contrast agents using a non-recirculating system, livers being always perfused by fresh solutions. The portal flow rate was 30 mL/min. The common bile duct was cannulated to measure Gd-BOPTA bile concentrations (μM) every 5 min and bile flow rates (μL/min). Samples were collected from hepatic veins every 5 min to measure Gd-BOPTA hepatic vein concentrations (μM).

Figure 1. (A) Gd-BOPTA transport in compartments: sinusoids (SIN), interstitium (INT), hepatocytes (HC) and bile canaliculi (BC). Hepatocyte A illus- trates Gd-BOPTA concentrations while Hepatocyte B illustrates Gd-BOPTA transfer rates. Clearance from SIN to LIVER (SINCLLIVER, mL/min) and extraction ratio (SINERLIVER, %) are measured during Gd-BOPTA perfusion (from 50 to 75 min). Uptake rate from INT to HC (INTUPTRHC,μM/min) is measured using true Gd-BOPTA hepatocyte concentrations. Elimination rate from HC to BC (HCELIMRBC, nmol/min), clearance from HC to BC (HCCLBC, g/min), elimination rate from HC back to INT (HCELIMRINT, nmol/min) and clearance from HC back to INT (HCCLINT, g/min) interfere with Gd-BOPTA concentrations in each compartment. (B) Protocol of perfused solutions. All livers were successively perfused with Gd-DTPA (200μM, 10 min), KHB solution (35 min), Gd-BOPTA (perfusion period, 30 min, 200μM) and KHB solution (rinse period, 30 min). Livers were perfused with a constant portal flow rate (30 mL/min). We dis- tinguish a drug perfusion period when Gd-BOPTA accumulates in all liver compartments and a rinse period during which Gd-BOPTA leaves hepatocytes for bile canaliculi or sinusoids. To ensure that Gd-BOPTA measured in hepatic veins originates from hepatocytes, parameters of efflux back to sinusoids are calculated from 85 to 105 min after a 10 min rinse period (corresponding to a 300 mL rinse volume, dashed grey line).

C. M. PASTOR

(4)

2.3. Drug preparation and protocol of perfused solutions (Fig. 1B)

Gadopentetate dimeglumine (Gd-DTPA, Magnevist®, Bayer Pharma) and gadobenate dimeglumine (Gd-BOPTA, MultiHance®, Bracco Imaging) were successively perfused. Contrast agents were labelled by adding153GdCl3to a 0.5 M Gd-BOPTA or Gd- DTPA solution (1 MBq/mL). Contrast agents were diluted in KHB solution to obtain a 200μM concentration. Livers were succes- sively perfused with 200μM DTPA for 10 min, KHB solution for 35 min, 200μM Gd-BOPTA (perfusion period, 30 min) and KHB solution (rinse period, 30 min). We distinguish a drug perfusion period when Gd-BOPTA accumulates in all liver compartments, and a rinse period during which Gd-BOPTA leaves hepatocytes into bile canaliculi or back to sinusoids.

2.4. Liver Gd-DTPA and Gd-BOPTA concentrations

To quantify liver concentrations, a gamma scintillation probe that measures radioactivity every 20 s was placed 1 cm above a liver segment. To transform radioactivity count rates into concentrations, the Gd-BOPTA concentration in each liver (end of experiment) was measured by an activimeter (Isomed 2000, Canberra, Saint-Quentin-en-Yvelines, France) and re- lated to the last count rate measured by the probe. Gd-DTPA and Gd-BOPTA concentrations in the common bile duct and hepatic veins were measured every 5 min with a gamma coun- ter (Canberra). Concentrations are expressed as μM in great vessels, bile duct and livers. In the liver, we consider 1 g to be close to 1 mL.

2.5. Gd-BOPTA concentrations in bile canaliculi and hepa- tocytes (Fig. 1)

Probe bile canaliculus concentrations were calculated by multi- plying bile concentrations measured in common bile duct by 0.43%, the volume of bile canaliculi in liver tissue according to Blouin et al. (19). Throughout the manuscript, concentrations measured by the probe are designated ‘probe concentrations’ while concentrations inside each compartment are named‘true concentrations’. Knowing Gd-BOPTA probe concentrations in bile canaliculi and the extracellular compartment (given by Gd-DTPA liver probe), we subtracted both values from Gd-BOPTA total liver probe concentrations to obtain Gd-BOPTA probe concentrations in hepatocytes. Finally, because the probe counts Gd-BOPTA in a 78% volume of hepatocytes (19), we must increase the values to 100% to obtain true hepatocyte concentrations.

2.6. Transfer rates across membranes (Fig. 1)

Elimination from sinusoids to liver (SINELIMRLIVER) was measured by (CPV CHV) × 30 (constant portal blood flow, mL/min), and sinusoidal clearance to liver (SINCLLIVER) was SINELIMRLIVER/CPV, whereCPVis the Gd-BOPTA concentration in the portal vein and CHVis the Gd-BOPTA concentration in the hepatic vein. We mea- sured the Gd-BOPTA sinusoidal extraction ratio to liver (SINERLIVER) as 100 × [CPV CHV]/CPV. Gd-DTPA SINELIMRLIVER, SINCLLIVER and

SINERLIVERwere also calculated.

To quantify Gd-BOPTA uptake rate from interstitium into hepatocytes or INTUPTRHC (μM/min), we measured the slope of the relation between true Gd-BOPTA concentrations in hepatocytes and time (min). The parameter was measured over 2 min at the very beginning of the perfusion period, after

a delay of 1 min to assure Gd-BOPTA distribution within the extracellular space. This short period of measurement pre- cludes underestimation of hepatocyte concentrations associ- ated with early Gd-BOPTA bile excretion. A single value is available.SINUPTRLIVERinμM/min was also measured, because it corresponds to liver enhancement measured in clinical imaging.

Transfer rates from hepatocytes to bile canaliculi or interstitium were also measured.HCELIMRBC(nmol/min) was CBC× bile flow, where CBC is Gd-BOPTA concentration in bile canaliculi.

HCELIMRINT(nmol/min) wasCHV× portal flow rate, whereCHV is Gd-BOPTA concentration in the hepatic vein. HCCLBC

(HCELIMRBC/CHC) and HCCLINT (HCELIMRINT/CHC) were also caculated, whereCHCis Gd-BOPTA concentration in hepatocytes.

2.7. Statistics

Parameters are means ± SD. To compare the evolution of con- centrations over time, we use two-way ANOVA with multiple comparisons between groups within each time-point (Prism 6, GraphPad, La Jolla, CA, USA). Kruskal–Wallis tests compared mean values between the two experimental groups.

3. RESULTS

3.1. Pharmacokinetic parameters of Gd-DTPA

In both groups, liver Gd-DTPA probe concentrations reached steady concentrations (62 ± 10μM) within 2 min (p= 0.9; Fig. 2 and Table 1). The true concentrations in sinusoids are around 200μM, while the probe measures 20μM (200 × 0.10) in sinusoids, sinusoidal volume being 10% of tissue volume (19).

Consequently, the probe would measure in interstitium 42μM (liver–sinusoid probe concentrations). Two different concentra- tions in sinusoids and interstitium does not fit with the common knowledge of a single concentration in the extracellular volume.

Whether the extracellular volume has one or two concentrations is impossible to define with our experimental model. However, an important finding is that concentrations measured by the probe are not the true compartmental concentrations. This point

Figure 2. Gd-DTPA liver probe concentrations measured using the gamma probe placed over the livers for 25 min. Two groups were stud- ied: normal rats (B200, black symbols, n= 5) and rats lacking Mrp2 (B200TR, grey symbols,n= 3). Livers were perfused with 200μM DTPA (0 to 10 min) and KHB (10 to 25 min, horizontal bar). Concentrations are similar over time in the two groups (p= 0.9). Extracellular space is mostly cleared of DTPA after 2 min.

(5)

is important to mention because transfer rates between two adjacent compartments rely on true concentrations.

DTPASINELIMRLIVER,SINCLLIVERandSINERLIVERwere 9±5 nmol/min, 0.05 ± 0.02 mL/min and 0.15 ± 0.08% respectively in the B200 group and 9 ± 6 nmol/min, 0.04 ± 0.04 mL/min and 0.15 ± 0.13%

in the B200TR group (Table 2). During the rinse period, the extra- cellular space was mostly cleared of contrast agent within 2 min (Fig. 2).

3.2. Gd-BOPTA concentrations in liver, common bile duct and great vessels

We first illustrate Gd-BOPTA probe concentrations in compart- ments that are available in clinical liver MRI (Fig. 3). In the B200 group, liver probe concentrations were much higher (maximal concentrations 473 ± 56μM, Fig. 3A and Table 1) than those measured during Gd-DTPA perfusion. Gd-BOPTA concen- trations increase rapidly over 10 min. Thereafter, increase of liver Gd-BOPTA concentrations was slower. When Gd-BOPTA perfusion was replaced by a KHB perfusion (rinse period), Gd- BOPTA liver concentrations steadily decreased (Fig. 3A and Table 1).

Gd-BOPTA in common bile duct (Fig. 3B and Table 1) was pres- ent 5 min after the start of drug perfusion (first sampling), and maximal true concentrations reached 15 700 ± 3100μM (end of perfusion period), a concentration 78 times higher than that perfused in portal veins (200μM). True Gd-BOPTA concentrations in hepatic veins remained around 186 ± 1μM at steady state (Fig. 3C and Table 1).

In rats lacking Mrp2, liver probe concentrations linearly increased during the perfusion period (Fig. 3A), Gd-BOPTA accu- mulating in the absence of transfer into bile canaliculi.

Concentrations were much higher in livers lacking Mrp2 than in normal livers (Fig. 3A and Table 1), while Gd-BOPTA true con- centrations in common duct were tiny (Fig. 3B and Table 1).

When Gd-BOPTA perfusion was replaced by a KHB solution, an early and small decrease in liver concentrations was observed, corresponding to the rinse of extracellular space. Then, Gd- BOPTA remained trapped inside hepatocytes until the end of the protocol.

3.3. Gd-BOPTA transfer rates between two compartments From sinusoids to liver, values differed in the two groups for

SINELIMRLIVER (p= 0.03), SINCLLIVER (p= 0.03) and SINERLIVER

(p= 0.02), values being slightly higher in B200 than B200TR groups (Table 2). The three parameters were high 5 min after the start of Gd-BOPTA perfusion and then decreased to a steady state. At steady state, the Gd-BOPTA liver extraction ra- tio was 7 ± 1 %, explaining why Gd-BOPTA hepatic vein true concentrations remained high and close to that of the portal vein (Fig. 3C).

From the relationship between Gd-BOPTA hepatocyte true concentrations and time (Fig. 4A), we can measure INTUPTRHC

at the very beginning of Gd-BOPTA perfusion whenHCELIMRBC and HCELIMRINT are a minimum. INTUPTRHC were 44 ± 3 (B200 group) and 37 ± 2μM/min (B200TR group) (p= 0.14).INTUPTCLHC could not be calculated because we do not know the true Gd- BOPTA interstitial concentrations.INTUPTRLIVERvalues (Gd-BOPTA uptake measured from the relationship between liver probe con- centrations and time, similar to liver enhancement in clinical MRI) were 43 ± 5 (B200 group) and 36 ± 4 (B200TR group) (p= 0.10). Thus, signal intensity enhancement at the very Table 1. Gd-DTPA and Gd-BOPTA concentrations

Time* Gd-DTPA (B200) Gd-DTPA (B200TR) Gd-BOPTA (B200) Gd-BOPTA (B200TR)

Bile true concentrationsμM 10 0 0

50 1830 ± 750 0 ± 0

55 10’400 ± 2’400 44 ± 4

75 15’700 ± 3’100 813 ± 110

105 3’500 ± 640 1200 ± 160

Portal vein true concentrationsμM 0-10 200 200

45-75 200 200

Hepatic vein true concentrationsμM 50 158 ± 4 163 ± 3

75 186 ± 1 186 ± 2

Hepatocyte true concentrationsμM 50 197 ± 5 179 ± 18

55 277 ± 48 384 ± 31

75 439 ± 83 1223 ± 85

105 96 ± 32 1163 ± 63

Liver probe concentrationμM 10 62 ± 10 59 ± 20

50 230 ± 14 194 ± 11

55 320 ± 34 355 ± 15

75 473 ± 56 1014 ± 55

105 92 ± 23 915 ± 73

Bile canaliculi probe concentrationsμM 50 8 ± 3 0 ± 0

55 44 ± 11 0 ± 0

75 66 ± 20 4 ± 1

105 15 ± 3 5 ± 1

*Units in min. Gd-DTPA (200μM) is perfused from 0 to 10 min. Gd-BOPTA (200μM) is perfused from 45 to 75 min (perfusion period) and replaced by a rinse solution from 75 to 105 min (rinse period).

C. M. PASTOR

(6)

begining of Gd-BOPTA perfusion correctly measured Gd-BOPTA uptake rates from interstitium into hepatocytes.

We quantified hepatocyte Gd-BOPTA elimination rates and clearances into bile canaliculi (HCELIMRBCandHCCLBC) or back into interstitium (HCELIMRINTandHCCLINT) (Fig. 5 and Table 2).

HCELIMRBC and HCCLBC values were significantly different be- tween groups (p<0.0001).HCCLBC in the B200 group was not steady over time (p<0.006), suggesting that, besides cellular concentrations, other factors determine Gd-BOPTA transfer rates from hepatocytes into bile canaliculi. HCELIMRINT (p<0.0001) and HCCLINT (p<0.0001) over time also signifi- cantly differed between groups. At the end of the protocol, we calculated that Gd-BOPTA clearance from hepatocytes back to sinusoids represents 15% of Gd-BOPTA clearance from hepatocytes.

3.4. How transfer rates determine hepatocyte Gd-BOPTA true concentrations

Finally, Fig. 4B clearly explains how Gd-BOPTA transfer rates gen- erate Gd-BOPTA concentrations in hepatocytes. At the begining of perfusion, Gd-BOPTA SINELIMRLIVER was much higher than Gd-BOPTA HCELIMRBILE in the B200 group. Then, SINELIMRLIVER

decreased but remained higher thanHCELIMRBILEand hepatocyte concentrations continued to increase moderately. The absence

of Gd-BOPTAHCELIMRBILE clearly explains the high hepatocyte concentrations in B200TR group.

4. DISCUSSION

Without any pharmacokinetic modelling, we measure true concentrations in each compartment (except interstitium) and evidence how transfer rates generate concentrations in hepato- cytes. Gd-BOPTA true concentrations in hepatocytes depend on simultaneous clearances from sinusoids to hepatocytes and from hepatocytes to bile canaliculi and back to sinusoids. Clearly, biliary clearance of Gd-BOPTA should be kept in mind besides uptake clearance when analysing signal intensities in the hepatobiliary phase of clinical MRI. Various findings important for the understanding of signal intensities obtained in the hepatobiliary phase of human liver imaging are listed in the fol- lowing paragraphs.

First, we distinguish‘true concentrations’from‘probe concen- trations’in all compartments included in the liver. Concentra- tions calculated from signal intensities with clinical MR systems correspond to our‘probe concentrations’and are not true tissue concentrations. These systems, similarly to our probe, detect all molecules included in the ROI delineated by the system (probe or MRI) without any possibility to localize concentrations in each Table 2. Gd-DTPA and Gd-BOPTA transfer rates

Time* Gd-DTPA (B200) Gd-DTPA (B200TR) Gd-BOPTA (B200) Gd-BOPTA (B200TR)

SINELIMRLIVER, nmol/min 10 9 ± 5 9 ± 6

50 1267 ± 134 1117 ± 95

55 503 ± 83 409 ± 28

75 422 ± 42 424 ± 53

SINCLLIVER, ml/min 10 0.05 ± 0.02 0.04 ± 0.04

50 6.3 ± 0.7 5.6 ± 0.5

55 2.5 ± 0.4 2.1 ± 0.1

75 2.1 ± 0.3 2.1 ± 0.2

SINERLIVER, % 10 0.15 ± 0.08 0.15 ± 0.13

50 21 ± 2 19 ± 2

55 8 ± 1 7 ± 1

75 7 ± 1 7 ± 1

INTUPTRHC,μM/min 46-48 44 ± 3 37 ± 2

HCELIMRBC, nmol/min 50 27 ± 12 0 ± 0

55 223 ± 45 0 ± 0

75 354 ± 49 6 ± 1

95 77 ± 16 8 ± 2

105 42 ± 12 8 ± 2

HCCLBC, g/min 50 0.14 ± 0.06 0.00 ± 0.00

55 0.83 ± 0.25 0.00 ± 0.00

75 0.85 ± 0.25 0.00 ± 0.00

95 0.57 ± 0.11 0.01 ± 0.00

105 0.45 ± 0.10 0.01 ± 0.00

HCELIMRINT, nmol/min 85 40 ± 6 33 ± 2

95 16 ± 4 26 ± 1

105 8 ± 2 26 ± 5

HCCLINT, g/min 85 0.19 ± 0.03 0.03 ± 0.01

95 0.11 ± 0.02 0.02 ± 0.01

105 0.08 ± 0.02 0.02 ± 0.01

*Units in min. Gd-DTPA (200μM) is perfused from 0 to 10 min. Gd-BOPTA (200μM) is perfused from 45 to 75 min (perfusion period) and replaced by a rinse solution from 75 to 105 min (rinse period).

(7)

compartment. During Gd-DTPA perfusion, we calculated two different true concentrations in sinusoids and interstitium, apply- ing volume coefficients in both compartments. However, these results do not fit with the common knowledge of a unique concentration in the extracellular compartment, according to large fenestrations existing in normal sinusoids. True concentra- tions in interstitium remain to be defined, because these concen- trations regulate uptake transport into hepatocytes, similarly to drug concentrations surrounding isolated human hepatocytes (16,20).

Another finding is that clearance from sinusoids into liver is 50 times higher during Gd-BOPTA than Gd-DTPA perfusion. At steady state, Gd-BOPTASINCLLIVERis 2.5 mL/min when portal flow is 30 mL/min. By definition, it is independent of sinusoidal con- centrations, but other factors interfere with the values because

SINCLLIVER is higher 5 min after the start of Gd-BOPTA than at the end of perfusion. Moreover, Gd-BOPTASINCLLIVERis inhomo- geneous along sinusoids because few OATP transporters are expressed in hepatocytes around portal triads in humans and ro- dent livers, while the expression progressively increases until Gd- BOPTA reaches perivenous hepatocytes (21–23). Consequently, Gd-BOPTA clearance increases along sinusoids (Fig. 6). This find- ing suggests that a decreased enhancement in the hepatobiliary phase would be evidenced mainly when injury occurs in perivenous hepatocytes. This zonal distribution of Gd-BOPTA clearance is impossible to measure with the gamma probe, which targets a large volume of tissue. This zonal distribution can be modelled by several compartments inside livers (24,25).

Another observation is the low Gd-BOPTA extraction ratio (7%) in a single pass, which explains the high hepatic vein concentra- tions. Radioactivity is sensitive enough to detect Gd-BOPTA entry into hepatocytes 33 s after the start of perfusion. In clinical MRI, accumulation in hepatocytes should be high enough to be detected and hepatobiliary sequences are acquired 20 min (Gd- Figure 3. Gd-BOPTA concentrations over time in structures identified at

clinical MRI. Values are either probe concentrations (A, liver) or true con- centrations (B and C, bile and great vessels). Gd-BOPTA (200μM) is per- fused from 45 to 75 min (perfusion period) and replaced by a rinse solution from 75 to 105 min (rinse period). In C, the horizontal black bar illustrates Gd-BOPTA steady perfusion (200μM) in the portal vein from 45 to 75 min. Two groups of rats are studied: normal rats (B200,n= 5, black symbols) and rats lacking Mrp2 (B200TR,n= 3, grey symbols). PV, portal vein; HV, hepatic vein.

Figure 4. Gd-BOPTA true concentrations in hepatocytes over time (A) according to concomitant Gd-BOPTA elimination rates from sinusoids to liver (SINELIMRLIVER) and Gd-BOPTA elimination rates from hepatocytes to bile canaliculi (HCELIMRBC) (B). Two groups of rats are studied: normal rats (B200,n= 5, black symbols) and rats lacking Mrp2 (B200TR,n= 3, grey symbols). Livers are perfused with 200μM Gd-BOPTA (45 to 75 min) and rinse solution (75 to 105 min).

C. M. PASTOR

(8)

EOB-DTPA) or 120 min (Gd-BOPTA) after injection. However, early hepatocyte uptake has been recently evidenced (26).

The best interval to measure signal intensity enhancement in human livers remains puzzling. We detected Gd-BOPTA entry into hepatocytes between 1 and 3 min after the start of Gd- BOPTA perfusion. During this interval, Gd-BOPTA had time to distribute into the extracellular volume and few Gd-BOPTA molecules reached bile canaliculi. A similar interval is chosen in SPECT imaging following the injection of 99mTc-mebrofenin, which is transported by the same transporters as Gd-BOPTA and Gd-EOB-DTPA (27). We also show that Gd-BOPTA uptake rates from interstitium to hepatocytes (INTUPTRHC) or liver (INTUPTRLIVER) were similar, suggesting that signal intensity enhancement in clinical MRI correctly estimates Gd-BOPTA uptake rates from interstitium into hepatocytes.

Finally, Gd-BOPTA true concentrations in hepatocytes depend onSINCLLIVER,HCCLBCandHCCLINT. Gd-BOPTA accumulation in he- patocytes over time is linear when bile clearance is impeded (9,13). In contrast, the inflection of hepatocyte concentration curve over time we describe during Gd-BOPTA perfusion corresponds to the concomitant Gd-BOPTA bile excretion, which minimizes concentration increase.

Gd-BOPTA efflux back from hepatocytes into sinusoids is less important because HCCLINT is 15% when HCCLINT is 85%.

Thus, signal intensities in clinical imaging estimate hepato- cyte concentrations generated by simultaneous Gd-BOPTA hepatocyte uptake and exit rates. This conclusion is proven in hepatocarcinomas that retain OATP1B1/B3 expression on tumour cells. These lesions are either hypointense if Mrp2

expression is increased (increased Gd-EOB-DTPA bile excre- tion) or hyperintense when Mrp2 expression is decreased (28).

4.1. Limitations of the study

Gd-BOPTA perfusion over 30 min in rat perfused liver does not mimic clinical protocols. Moreover, the hepatic artery is not per- fused and the dual imput of Gd-BOPTA entry into livers is not respected. Portal flow rate is constant over the protocol, while food intake may increase portal blood flow in humans. The absence of albumin in the perfusion solution provides a maximal Gd-BOPTA entry into hepatocytes, while Gd-BOPTA uptake into human hepatocytes depends on plasma albumin concentrations.

REFERENCES

1. Watanabe A, Ramalho M, Alobaidy M, Kim HJ, Velloni FG, Semelka RC. Magnetic resonance imaging of the cirrhotic liver: an update.

World J Hepatol 2015; 7: 468487.

2. Choi JY, Lee JM, Sirlin CB. CT and MR imaging diagnosis and staging of hepatocellular carcinoma: part II. Extracellular agents, hepatobiliary agents, and ancillary imaging features. Radiology 2014; 273: 3050.

3. Verloh N, Haimerl M, Rennert J, Müller-Wille R, Nießen C, Kirchner G, Scherer MN, Schreyer AG, Stroszczynski C, Fellner C, Wiggermann P.

Impact of liver cirrhosis on liver enhancement at Gd-EOB-DTPA en- hanced MRI at 3 Tesla. Eur J Radiol 2013; 82: 17101715.

4. Filippone A, Blakeborough A, Breuer J, Grazioli L, Gschwend S, Hammerstingl R, Heinz-Peer G, Kittner T, Laghi A, Leen E, Lencioni R, Lucidarme O, Remplik P, Robinson PJ, Ruehm SG, Schaefer F, Stoupis C, Tombach B, Valette PJ, Zech CJ, Huppertz A. Enhancement of liver parenchyma after injection of hepatocyte-specific MRI con- trast media: a comparison of gadoxetic acid and gadobenate dimeglumine. J Magn Reson Imaging 2010; 31: 356364.

Figure 5. Clearances from hepatocyte into bile canaliculi (A,HCCLBC, g/min) and interstitium (B,HCCLINT, g/min). Gd-BOPTA (200μM) was perfused from 45 to 75 min (perfusion period) and replaced by KHB perfusion from 75 to 105 min (rinse period). Gd-BOPTAHCCLINTwas measured only during the rinse period, because during the perfusion period Gd-BOPTA efflux from hepatocytes is mixed with contrast agents perfused through the portal vein. Normal rats (n= 5, black symbols) and rats lacking Mrp2 (n= 3, white symbols).

Figure 6. Simplification of increased clearances into liver (SINCLLIVER) along sinusoids at steady state (30 min after the start of Gd-BOPTA perfusion).

Light grey, grey and black arrows illustrate low, medium and high expression of OATP transporters respectively. Zonal clearances are calculated accord- ing to decreased Gd-BOPTA concentrations along sinusoids. The small difference of Gd-BOPTA concentrations between the hepatic venule and portal venule reflects a low extraction ratio (7%). Values are collected from normal rats.

(9)

5. Van Beers BE, Pastor CM, Hussain HK. Primovist, Eovist: what to expect? J Hepatol 2012; 57: 421429.

6. Pastor CM, Müllhaupt B, Stieger B. The role of organic anion trans- porters in diagnosing liver diseases by magnetic resonance imaging.

Drug Metab Dispos 2014; 42: 675684.

7. Leonhardt M, Keiser M, Oswald S, Kühn J, Jia J, Grube M, Kroemer HK, Siegmund W, Weitschies W. Hepatic uptake of the magnetic reso- nance imaging contrast agent Gd-EOB-DTPA: role of human organic anion transporters. Drug Metab Dispos 2010; 38: 10241028.

8. Nassif A, Jia J, Keiser M, Oswald S, Modess C, Nagel S, Weitschies W, Hosten N, Siegmund W, Kühn JP. Visualization of hepatic uptake transporter function in healthy subjects by using gadoxetic acid- enhanced MR imaging. Radiology 2012; 264: 741750.

9. Jia J, Puls D, Oswald S, Jedlitschky G, Kühn JP, Weitschies W, Hosten N, Siegmund W, Keiser M. Characterization of the intestinal and hepatic uptake/efflux transport of the magnetic resonance imaging contrast agent gadolinium-ethoxylbenzyl-diethylenetriamine-pentaacetic acid.

Invest Radiol 2014; 49: 7886.

10. Forsgren MF, Dahlqvist Leinhard O, Dahlström N, Cedersund G, Lundberg P. Physiologically realistic and validated mathematical liver model reveals hepatobiliary transfer rates for Gd-EOB-DTPA using human DCE-MRI data. PLoS One 2014; 9 e95700.

11. Steingoetter A, Menne D, Braren RF. Assessing antiangiogenic ther- apy response by DCE-MRI: development of a physiology driven multi-compartment model using population pharmacometrics. PLoS One 2011; 6 e26366.

12. Ulloa JL, Stahl S, Yates J, Woodhouse N, Kenna JG, Jones HB, Waterton JC, Hockings PD. Assessment of gadoxetate DCE-MRI as a biomarker of hepatobiliary transporter inhibition. NMR Biomed 2013; 26: 12581270.

13. Saito S, Obata A, Kashiwagi Y, Abe K, Murase K. Dynamic contrast- enhanced MRI of the liver in Mrp2-deficient rats using the hepatobiliary contrast agent Gd-EOB-DTPA. Invest Radiol 2013; 48:

548553.

14. Ingrisch M, Sourbron S. Tracer-kinetic modeling of dynamic contrast- enhanced MRI and CT: a primer. J Pharmacokinet Pharmacodyn 2013; 40: 281300.

15. Dahlqvist Leinhard O, Dahlstrom N, Kihlberg J, Sandstrom P, Brismar TB, Smedby O, Lundberg P. Quantifying differences in hepatic uptake of the liver specific contrast agents Gd-EOB-DTPA and Gd- BOPTA: a pilot study. Eur Radiol 2012; 22: 642653.

16. Chu X, Korzekwa K, Elsby R, Fenner K, Galetin A, Lai Y, Matsson P, Moss A, Nagar S, Rosania GR, Bai JP, Polli JW, Sugiyama Y, Brouwer KL. Intracellular drug concentrations and transporters: measurement, modeling, and implications for the liver. Clin Pharmacol Ther 2013;

94: 126141.

17. Planchamp C, Hadengue A, Stieger B, Bourquin J, Vonlaufen A, Frossard J-L, Quadri R, Becker CD, Pastor CM. Function of both

sinusoidal and canalicular transporters controls the concentration of organic anions within hepatocytes. Mol Pharmacol 2007; 71:

10891097.

18. Pastor CM, Wissmeyer M, Millet P. Concentrations of Gd-BOPTA in cholestatic fatty rat livers: role of transport functions through mem- brane proteins. Contrast Media Mol Imaging 2013; 8: 147156.

19. Blouin A, Bolender RP, Weibel ER. Distribution of organelles and membranes between hepatocytes and nonhepatocytes in the rat liver parenchyma. A stereological study. J Cell Biol 1977; 72: 441455.

20. Liu L, Pang KS. An integrated approach to model hepatic drug clear- ance. Eur J Pharm Sci 2006; 29: 215230.

21. Donner MG, Keppler D. Up-regulation of basolateral multidrug resis- tance protein 3 (Mrp3) in cholestatic rat liver. Hepatology 2001; 34:

351359.

22. Vander Borght S, Libbrecht L, Blokzijl H, Faber KN, Moshage H, Aerts R, Van Steenbergen W, Jansen PL, Desmet VJ, Roskams TA. Diagnos- tic and pathogenetic implications of the expression of hepatic trans- porters in focal lesions occurring in normal liver. J Pathol 2005; 207:

471482.

23. Yoneda N, Matsui O, Kitao A, Kozaka K, Gabata T, Sasaki M, Nakanuma Y, Murata K, Tani T. Beta-catenin-activated hepatocellular adenoma showing hyperintensity on hepatobiliary-phase gadoxetic- enhanced magnetic resonance imaging and overexpression of OATP8. Jpn J Radiol 2012; 30: 777782.

24. Watanabe T, Kusuhara H, Sugiyama Y. Application of physiologically based pharmacokinetic modeling and clearance concept to drugs showing transporter-mediated distribution and clearance in humans. J Pharmacokinet Pharmacodyn 2010; 37: 575590.

25. Kusuhara H, Sugiyama Y. Pharmacokinetic modeling of the hepatobiliary transport mediated by cooperation of uptake and efflux transporters. Drug Metab Rev 2010; 42: 539550.

26. Schalkx HJ, Van Stralen M, Coenegrachts K, Van Den Bosch MA, Van Kessel CS, Van Hillegersberg R, Van Erpecum KJ, Verkooijen HM, Pluim JP, Veldhuis WB, Van Leeuwen MS. Liver perfusion in dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI): compar- ison of enhancement in Gd-BT-DO3A and Gd-EOB-DTPA in normal liver parenchyma. Eur Radiol 2014; 24: 21462156.

27. De Graaf W, Bennink Copyright © 2016 John Wiley & Sons, Ltd. RJ, Veteläinen R, Van Gulik TM. Nuclear imaging techniques for the as- sessment of hepatic function in liver surgery and transplantation. J Nucl Med 2010; 51: 742752.

28. Tsuboyama T, Onishi H, Kim T, Akita H, Hori M, Tatsumi M, Nakamoto A, Nagano H, Matsuura N, Wakasa K, Tomoda K. Hepatocellular carcinoma: hepatocyte-selective enhancement at gadoxetic acid- enhanced MR imagingcorrelation with expression of sinusoidal and canalicular transporters and bile accumulation. Radiology 2010; 255: 824833.

C. M. PASTOR

Références

Documents relatifs

The Cu 2+ ions are added as tracers for absorption titration because their com- plexation with DOTA and NODA molecules causes an intense sig- nal in the visible region that is speci fi

Pour sortir de la confidentialité, quelques groupes ont lancé des actions vers le grand public ou les IUFM :le GD 44,le GLEM et le GD 42 ont organisé des salons ouverts à tous,

Détermine (schéma à l’appui) en quel point arrive BONA s’il double

Le maximum de pluie, à Clarens et à Sion, tombe donc en été, tandis qu'au Od St-Bernard il a lieu en automne, rappro- chant cette station du régime des pays méditerranéens. Dans

In livers lacking Mrp2 and perfused with Bopta, liver concentrations only rely on the hepatocyte and extracel- lular compartments. The low Bopta concentrations in bile canaliculi

In comparison, contemporary series report recurrence following hepatectomy for CRLM shifted towards the remnant liver, with only 26% of patients presenting with isolated lung

We define new pharmacokinetic pa- rameters to quantify the multiple resistance-associated protein 2 (Mrp2) function, and present how transfer rates generate the Gd-BOPTA

Moreover, the drug concentrations generated by the hepatocyte uptake clear- ances are simultaneously modified by efflux clearances from hepatocytes into bile canaliculi and back