- Email: [email protected]
Ex-vivo MR imaging of liver intracellular contrast agents
Magnetic
PI1 SO730-725X( 96) 02007-G
ELSEVIER
l
Resonance Imaging, Vol. 15, No. 4, pp. 469-474, 1997 Copyright 0 1997 Elsevier Science Inc. P&ted in the USA. All rights reserved 0730-725X/97 $17.00 + .OO
Original Contribution EX-VIVO
MR IMAGING OF LIVER INTRACELLULAR CONTRAST AGENTS
MARIO MASCALCHI, * XIAN-NU JIN,$ CRISTIANA AGEN,$ PASQUALE PETRUZZI, * DENISE NARDINI,$ CARLO TESSA, * DAVIDE CARAMELLA, * AND CARLO BARTOLOZZI
*
*Cattedra di Radiologia, Universita di Pisa, TBeijing Tiantan Hospital, Capital Medical Institute, China, and SIstituto di Farmacologia, Universita di Pisa The objective is to evaluate whether an ex-vivo model can be used to test intracellular contrast agents for MR imaging of the liver. Tl weighted inversion recovery, proton density spin echo and T2* weighted gradient echo images of the liver were acquired at 0.5 T in 10 rats before and 30 min after intravenous injection of 0.075 mmol/kg Gadolinlum benzyloxypropionictetraacetate (Gd-BOPTA, n = 5) or 0.015 mmol/ kg dextran magnetite (DM, IZ = 5)) Four additional animals served as controls. After exsanguination and perfusion with saline and formalin, specimens of the liver and brain were embedded in an agar gel and examined with MR imaging one to three weeks later using the same protocol. In-vivo, the mean liver signal enhancement caused by Gd-BOPTA in Tl, proton density and T2* weighted images was +23%, +28% and -7O%, respectively. The mean liver signal enhancement caused by DM was -71%, -76% and -94%. In-vitro, no signal change was seen in the brain of animals injected with Gd-BOPTA and DM as compared to controls. Liver signal was increased by Gd-BOPTA and decreased by DM. Mean liver enhancement rate induced by Gd-BOPTA was +22%, +5% and +27% for Tl, proton density and T2* weighted images, respectively. Mean liver enhancement rate induced by DM was -279’ O, -19% and -31%. MR imaging signal changes induced by liver intracellular contrast agents are still appreciable in an ex-vivo model. The latter might be useful for for preliminary investigation of intracellular contrast agents for MR imaging of the liver. 0 1997 Elsevier Science, Inc. Keywords: Contrast dimeglumine.
media; Experimental
studies; Gd-BOPTA;
INTRODUCTION
Dextran
magnetite; Liver;
Gadobenate
zyloxypropionictetraacetate ( Gd-BOPTA) magnetite (DM), were assessed.
Contrast agents for magnetic resonance (MR) imaging of the liver have been developed and tested in a variety of animal models in viva.’ To investigate the distribution of newly designed MR contrast agents for lymph nodes in small-sized animals, Wagner et al.* devised an ex-vivo model in which, before MR imaging, excised organs are embedded in an agar gel matrix. The purpose of this study was to evaluate whether this model can be used to test intracellular contrast agents for MR imaging of the liver. Potential advantages of the exvivo model include elimination of motion artifacts and comparison of contrast enhancement in the same specimen using different pulse sequences. Two different intracellular contrast agents, namely Gadolinium-ben-
MATERIALS
and dextran
AND METHODS
Contrast Agents Gd-BOPTA (Gadobenate dimeglumine) was provided by Bracco SpA, (Milan, Italy) as a 0.25 mol/L aqueous solution and was administered at a dose of 0.075 mmol/kg. DM was obtained from Institute fur Diagnosik Forschung, (Berlin, Germany) as an aqueous solution containing an iron concentration of 0.5 mol/L. The particles had a core diameter of 5-10 nm (measured by electron microscopy) and a carboxydextran coat with a whole particle size of 40-80 nm (as Cattedra di Radiologia, Universita di Pisa, Via Roma 67, 56126 Pisa, Italia. Tel. 50-592509; Fax 50-596933.
31219.5;ACCEPTED 6126195. Address correspondence to Mario Mascalchi M.D., Ph.D.
RECEIVED
469
Magnetic Resonance Imaging 0 Volume 15, Number 4, 1997
470
measured by laser light scattering). The administered dose of iron was 0.015 mmol/kg. Animals Female Wistar rats of about 300 g body weight without any pre-treatment were used. Ten animals were examined in-vivo before and after administration of Gd-BOPTA (n = 5) or DM (n = 5). Four animals served as controls. Liver and brain specimens of all these animals and of three additional animals not imaged in-vivo were then examined with the ex-vivo model detailed below. In-vivo Experiments Animals were anesthetised with urethane (lg/kg) intraperitoneal injection and fixed on a three-floor wooden support. This allows up to 6 animals to be imaged at the same time. The animals were imaged before and 30 min after injection of the contrast agents in the tail vein. MR Imaging Protocol MR imaging examinations were performed on a 0.5 T system (GE MR Max) using the head volume quadrature coil. Field of view was 17-22cm, matrix 128 x 192 and slice thickness 5 mm. Number of excitations ranged between 4 and 6. Tl weighted inversion recovery (IR) images were obtained with TR 1500 ms, TE 25 ms and inversion time 400 ms, proton density weighted spin echo (PD SE) images were obtained with TR 2000 ms and TE 25 ms, and T2* weighted gradient recalled echo (GRE) images were acquired with TR 600 ms, TE 30 ms and 30 degree flip angle. Total acquisition time for the three sequences was about 45 min. Ex-vivo Experiments After imaging the animals were removed from the magnet and after 0.1 ml heparin (5000 U/ml) administration to prevent blood clotting, the chest was opened and the aorta incannulated. Following recision of inferior vena cava, animals were perfused with saline, in
order to remove intravascular blood, until the liver discoloured. After perfusion with formaldehyde solution (4%) for fixation, one lobe of liver and entire brain and spleen were removed and put in formaldehyde solution (4%) for l-3 weeks. For preparation of the agar gel matrix, dry agar powder (20g) (Fluka Chemie AG, Buchs, Schweiz) was mixed in cold deionized water ( 11) doped with 1 ml of Gadolinium-DTPA (Magnevist, Schering AG, Berlin, Germany) and stirred with a magnetic mixer. The suspension was heated twice to the boil with a microwave oven. The mixture was cooled to about 80°C and poured into a plastic box ( 10 X 15 X 10 cm) to form a bubble-free layer of about 1 cm thickness. After solidification of this layer, tissue specimens were placed upon the layer and fixed with some drops of the liquid agar solution (80°C). After solidification of these drops another layer of the mixture was poured. The box was placed over a phantom containing NiCl solution (0.1%) and evaluated within two weeks of preparation with the same MR imaging protocol used for the in-vivo experiments. Data Analysis The liver signal intensity to noise ratios (S/N) before and after contrast agents administration in the live rats were computed on the operator console using regions of interest, and the signal enhancement was calculated according to the following formula: (S/N after contrast - S/N before contrast)/S/N before contrast X 100%. The liver and brain signal intensity to agar ratio (S/A) of the animals injected with contrast agents was compared to the S/A of the control animals. Mean signal enhancement was calculated according to the following formula: (S/A of injected animals - S/A of control animals)/S/A of control animals x 100%. RESULTS Mean and standard deviation of the liver signal intensity to noise ratio before and after Gd-BOPTA and DM in live animals and controls are reported in Table
Table 1. Liver S/N before and after intravenous administration of Gd-BOPIA Before Gd-BOPTA
and DM in vivo After
Control
Gd-BOPTA
Control
Pulse sequences
(n = 5)
(nD_M5)
(n = 4)
(n = 5)
(n = 4)
IR PD SE
39.1 + 26.3
22.1 + 5.8
24.0 k 2.9 179.4 ? 59.4
39.2 ? 16.2 244.3 2 57.5
35.6 + 2.9 54.0 ? 1.1 228.5 2 10.6
48.0 + 31.1 30.7 t 15.1 52.8 +- 29.8
GRE
The values are the mean k standard deviation.
6.38 ? 0.8 9.40 ? 4.2 14.5 Ifr 11.5
37.8 2 5.4 47.9 ? 7.9 214.0 2 19.7
Ex-vivo
MR
imaging
of liver
intracellular
contrast
agents 0 M. MASCALCHI
ET
AL.
471
Fig. 1. Axial images obtained with Tl weighted IR (TR 1500 ms, TE 25 ms and inversion time 400 ms) sequence in two animals before (top row) and after (bottom row) intravenous administration of 0.075 mmol/Kg Gd-BOPTA showing moderate homogeneous signal increase of the liver.
Fig. 2. Axial images obtained with T2 * weighted GRE (TR 600 ms, TE 30 ms and 30 degree flip angle) sequence in two animals before (top row ) and after (bottom row ) intravenous administration of 0.015 mmol/kg DM showing marked homogeneous signal decrease of the liver.
1. Gd-BOPTA increased the signal of the liver on Tl (Fig. 1) and PD weighted images, whereas it decreased the signal on T2* weighted images. The mean signal enhancement caused by Gd-BOPTA in Tl, PD and T2* weighted images was +23%, +28% and -70%. respectively. DM caused a marked decrease of the liver signal in all the sequences which was more pronounced in T2 * weighted images (Fig. 2). The mean signal enhancement induced by DM was -71%, -76% and -94% for Tl, PD and T2* weighted images. Liver and brain signal intensity to agar ratios for animals injected with Gd-BOPTA, DM and controls are reported in Table 2. No signal change of the brain was seen in specimens of animals injected with GdBOPTA and DM as compared to controls (Fig. 3). Liver signal was increased by Gd-BOPTA and reduced by DM on all the sequences. Conspicuity of the signal change induced by Gd-BOPTA and DM was higher for Tl (Fig. 3a) and T2* (Fig. 3c) weighted images, respectively. Mean liver enhancement rate induced by
Gd-BOPTA in the ex-vivo model was +22%, +5% and +27% for Tl, PD and T2 * weighted images. Mean liver enhancement rate induced by DM was -27%, - 19% and -31% for Tl, PD and T2 * weighted images. Signal changes similar to those observed in liver were presente in the spleen specimens (Fig. 3 ) .
DISCUSSION Gd-BOPTA is an octadentate chelate of Gadolinium which has a higher rate of biliary excretion than GdDTPA.3m5 In a previous in-vivo animal study using Tl weighted IR images, liver enhancement after intravenous injection of Gd-BOPTA was consistently higher than after injection of Gd-DTPA for doses between 0.25 and 1 mmol/kg with peak enhancement within 30 min after injection.4 DM consists of superparamagnetic magnetite (Fe,O,) cores coated with hydrophilic dextran.6-9 After intravenous administration, DM is captured by the reticuloendothelial cells of the liver and
Table 2. Liver and brain S/A of rats injected with Gd-BOPTA and DM in the ex-vivo model Brain
Liver Pulse sequences
Gd-BOPTA (n = 6)
(nD_M6)
Control (n = 5)
Gd-BOPTA (n = 6)
(nD:6)
Control (n = 5)
IR PD SE GRE
1.00 + 0.18 0.78 + 0.03 0.76 2 0.15
0.59 2 0.17 0.61 +- 0.13 0.41 ? 0.19
0.81 2 0.06 0.75 IL 0.11 0.59 ? 0.11
1.12 ? 0.14 0.84 -t 0.01 0.99 + 0.25
0.96 + 0.19 0.87 + 0.11 0.85 + 0.19
0.97 + 0.16 0.85 + 0.05 0.72 -e 0.08
The values
are the mean
t standard
deviation.
412
Magnetic ResonanceImaging 0 Volume 15, Number 4, 1997
Fig. 3. (a) Images of the ex-vivo model obtained with Tl v(eighted IR (‘TR 1500 ms, TE 25 ms and inversion time 400 ms; (b) PD SE (TR 2000 ms and TE 25 ms) and (c) T2* we:ighted GRE (TR 600 ms, TE 30 ms and 30 degree flip angle) sequence. High liver and spleen signal of animals injected P&h Gd-BOP TA and low signal of animals injected with DM are observed in all sequences. The former is more evident in ( a), the latta : in (c). The brain signal intensity is unaffected by contrast injection. The black dots (arrows) in (c) are due to magnetic SUIsceptibility effects of air bubbles. (Figure continues)
spleen and determines a strong shortening of the T2 relaxation of the nearby protons. This is better demonstrated by T2* weighted GRE than SE sequences because of the sensitivity of the former to magnetic susceptibility effects.g In our study Gd-BOPTA created positive liver enhancement on IR and PD SE images, whereas signal was decreased on GRE sequences in-vivo. This is consistent with theoretical analysis” taking into account the long TR, long ‘IX and flip angle of the GRE sequence we employed. The liver signal decrease observed with all sequences in our live animals after injection of DM is consistent with prior observations.6-g The effects of the contrast agents we observed in the liver of live animals were still appreciable in the specimens of the same animals examined with the exvivo model. A general decrease of the signal changes induced by the contrast agents was however observed.
In addition, in the ex-vivo model, Gd-BOPTA caused a slight increase of the liver signal on T2* weighted GRE images. Computation of Tl and T2 relaxation times could be useful in order to quantify these effects but were not performed in our study. The differences among in vivo and ex-vivo signal changes can be explained taking into account, on the one hand, the distribution of Gd-BOPTA and DM, and, on the other hand, the modification induced in the liver parenchyma by preparation of the ex-vivo model. Although Gd-BOPTA and DM are regarded as intracellular contrast agents, some amount of the injected contrast is contained in the intravascular and interstitial spaces until it is captured by target cells.” Thus, the overall liver signal change caused by these agents can be thought of as the summation of the effects of an intracellular and an extracellular component. The concentration of the contrast agent in the latter is decreased
Ex-vivo MR imaging of liver intracellular contrast agents 0 M. MASCALCHI
ETAL.
473
Second, the freezing of the contrast agent distribution in the ex-vivo model may be useful for determination of the contrast agent at different main magnetic fields not only in the case of superparamagnetic iron oxides l3 but also in the case of paramagnetic hepatobiliary contrast agents. In fact, as recently pointed out by Elster, I4 the main magnetic field of the MR system influences paramagnetic contrast effects due to its relationship with tissue relaxation rates. The ex-vivo preparation with its long maintanance (personal unpublished observation) could afford a reliable comparison of the effect of a given contrast at different field strengths, by simply serially examining the box with different MR systems. Finally, since the ex-vivo method is not suitable to investigate extracellular contrast agents as Gd-DTPA (unpublished observations), it might indirectly provide an estimation of the true extent of the intracellular compartment of newly designed liver contrast agents. Obviously, implications for clinical applications in humans of the observations made on the ex-vivo model will require validation with in-vivo animal experiments. Furthermore, the contribution, if any, of formalin fixation to the effects of the contrast agents observed in the ex-vivo model has to be addressed. In conclusion, our study indicates that an ex-vivo model can be useful for preliminary investigations of intracellular liver contrast agents.
REFERENCES Fig. 3. continued.
by exsanguination and perfusion and this could explain the lower “enhancement” observed in the ex-vivo as compared to in-vivo conditions. Another factor of possible relevance is the fixation in formalin of the specimens excised for the ex-vivo model. Tovi et al.‘* reported that formalin, a fixative that does not coagulate protein, leads to a slight decrease of the water content of tissues and can alter tissue relaxation rates. In our study fixation time in formalin ranged between 1 and 3 weeks. The effects of different fixation time have to be addressed in further studies. Several potential advantages exist for the ex-vivo model tested in this study for the preclinical evaluation of intracellular contrast agents for the liver. First, the ex-vivo model eliminates motion artifacts from bowel, respiratory motion and arterial pulsation encountered in-vivo. The reduced image blurring and improved spatial resolution might be especially valuable in evaluating models of multiple small (primary or secondary) liver neoplasms with potentials for better MR imaging-histopathological correlations.
1. Schmiedl, U.P.; Maravilla, K.R.; Nelson, J.A. Improved method for in-vivo magnetic resonance contrast media research. Invest. Radiol. 26:65-70; 1991. 2. Wagner, S.; Pfefferer, D.; Taupitz, M.; Kresse, M.; Lawaczeck, R.; Hamm, B.; Wolf, K.J. Intravenous MR-lymphography with ultrasmall iron oxide particles: In vivo and ex vivo MRI examinations in rats. In: Book of Abstracts: Eleventh annual scientific meeting of the Society of Magnetic Resonance in Medicine. Berlin, Germany: SMRM; 1992:570. 3. Vittadini, G.; Felder, E.; Tirone, P.; Lorusso, V. B 19036, a potential new hepatobiliary contrast agent for MR proton imaging. Invest. Radiol. 23( suppl 1):246-248; 1988. 4. Pavone, P.; Patrizio, G:; Buoni, C.; Tettamanti, E.; Passariello, R.; Musu, C.; Tirone, P.; Felder, E. Comparison of Gd-BOPTA with Gd-DTPA in MR imaging of rat liver. Radiology 176:61-64; 1990. 5. Vittadini, G.; Felder, E.; Musu, C; Tirone P. Preclinical profile of Gd-BOPTA: A liver specific MRI contrast agent. Invest. Radiol. 25:S59-S62; 1990. 6. Majumder, S.; Zoghbi, S.; Pope, C.F.; Gore, J.C. Quantitation of MR relaxation effects of iron oxide particles in liver and spleen. Radiology 169:653-655; 1988. 7. Magin, R.L.; Basic, G.; Niesman, M.R.; Alameda, J.C.; Wright, S.M.; Swartz, H.M. Dextran magnetite as a liver contrast agent. Magn. Res. Med. 2&l-16; 1991.
414
Magnetic Resonance Imaging 0 Volume 15, Number 4, 1997
8. Iannone,A.; Federico,M.; Tomasi,A.; Magin, R.; Casasco,A.; Calhgaro,A.; Vamrini,V. Detectionandquantitation in rat tissuesof the superparamagnetic magnetic resonance contrastagentdextranmagnetiteasdemontrated by electronspin resonancespectroscopy.Invest. Radiol. 27:450-455; 1992. 9. Kawamori,Y.; Matsui, 0.; Kadoya, M.; Yoshikawa,J.; Demachi,H.; Takashima, T. Differentiationof hepatocellular carcinomasfrom hyperplasticnodulesinducedin rat liver with ferrite-enhancedMR imaging. Radiology 183:65-72; 1992. 10. Davis,P.L.; Parker,D.L.; Nelson,J.A.; Gillen,J.S.;Runge, V.M. Interactionsof paramagnetic contrastagentsandthe spin echo pulse sequence.Invest. Radiol. 23:381-388; 1988.
11. Kreft, B.P.; Tanimoto,A.; Stark, D.D.; Baba, Y.; Zhao, L.H.; Chen,J.T.;Compton,C.C.;Finn, J.P.;Cavagna,F.M. Enhancementof tumor-liver contrast-to-noise ratio with gadobenate dimegluminein MR imagingof rats.J. Magn. Reson.Imaging.3:41-49; 1993. 12. Tovi, M.; Ericsson,A. Measurements of Tl andT2 over time in formalin-fixed human whole-brainspecimens. Acta. Radiol.33400-404; 1992. 13. Thickman,D.; Hendrick, R.E.; Jejian, K.A.; Schanker, C.S. Liver-lesiontissuecontraston MR images:Effect of iron oxide concentrationand magneticfield strength. Radiology 176557-562, 1990. 14. Elster,A.D. Field-strengthdependence of gadoliniumenhancement:Theory andimplications.Am. J. Neuroradiol. 15:1420-1423,1994.
PI1 SO730-725X( 96) 02007-G
ELSEVIER
l
Resonance Imaging, Vol. 15, No. 4, pp. 469-474, 1997 Copyright 0 1997 Elsevier Science Inc. P&ted in the USA. All rights reserved 0730-725X/97 $17.00 + .OO
Original Contribution EX-VIVO
MR IMAGING OF LIVER INTRACELLULAR CONTRAST AGENTS
MARIO MASCALCHI, * XIAN-NU JIN,$ CRISTIANA AGEN,$ PASQUALE PETRUZZI, * DENISE NARDINI,$ CARLO TESSA, * DAVIDE CARAMELLA, * AND CARLO BARTOLOZZI
*
*Cattedra di Radiologia, Universita di Pisa, TBeijing Tiantan Hospital, Capital Medical Institute, China, and SIstituto di Farmacologia, Universita di Pisa The objective is to evaluate whether an ex-vivo model can be used to test intracellular contrast agents for MR imaging of the liver. Tl weighted inversion recovery, proton density spin echo and T2* weighted gradient echo images of the liver were acquired at 0.5 T in 10 rats before and 30 min after intravenous injection of 0.075 mmol/kg Gadolinlum benzyloxypropionictetraacetate (Gd-BOPTA, n = 5) or 0.015 mmol/ kg dextran magnetite (DM, IZ = 5)) Four additional animals served as controls. After exsanguination and perfusion with saline and formalin, specimens of the liver and brain were embedded in an agar gel and examined with MR imaging one to three weeks later using the same protocol. In-vivo, the mean liver signal enhancement caused by Gd-BOPTA in Tl, proton density and T2* weighted images was +23%, +28% and -7O%, respectively. The mean liver signal enhancement caused by DM was -71%, -76% and -94%. In-vitro, no signal change was seen in the brain of animals injected with Gd-BOPTA and DM as compared to controls. Liver signal was increased by Gd-BOPTA and decreased by DM. Mean liver enhancement rate induced by Gd-BOPTA was +22%, +5% and +27% for Tl, proton density and T2* weighted images, respectively. Mean liver enhancement rate induced by DM was -279’ O, -19% and -31%. MR imaging signal changes induced by liver intracellular contrast agents are still appreciable in an ex-vivo model. The latter might be useful for for preliminary investigation of intracellular contrast agents for MR imaging of the liver. 0 1997 Elsevier Science, Inc. Keywords: Contrast dimeglumine.
media; Experimental
studies; Gd-BOPTA;
INTRODUCTION
Dextran
magnetite; Liver;
Gadobenate
zyloxypropionictetraacetate ( Gd-BOPTA) magnetite (DM), were assessed.
Contrast agents for magnetic resonance (MR) imaging of the liver have been developed and tested in a variety of animal models in viva.’ To investigate the distribution of newly designed MR contrast agents for lymph nodes in small-sized animals, Wagner et al.* devised an ex-vivo model in which, before MR imaging, excised organs are embedded in an agar gel matrix. The purpose of this study was to evaluate whether this model can be used to test intracellular contrast agents for MR imaging of the liver. Potential advantages of the exvivo model include elimination of motion artifacts and comparison of contrast enhancement in the same specimen using different pulse sequences. Two different intracellular contrast agents, namely Gadolinium-ben-
MATERIALS
and dextran
AND METHODS
Contrast Agents Gd-BOPTA (Gadobenate dimeglumine) was provided by Bracco SpA, (Milan, Italy) as a 0.25 mol/L aqueous solution and was administered at a dose of 0.075 mmol/kg. DM was obtained from Institute fur Diagnosik Forschung, (Berlin, Germany) as an aqueous solution containing an iron concentration of 0.5 mol/L. The particles had a core diameter of 5-10 nm (measured by electron microscopy) and a carboxydextran coat with a whole particle size of 40-80 nm (as Cattedra di Radiologia, Universita di Pisa, Via Roma 67, 56126 Pisa, Italia. Tel. 50-592509; Fax 50-596933.
31219.5;ACCEPTED 6126195. Address correspondence to Mario Mascalchi M.D., Ph.D.
RECEIVED
469
Magnetic Resonance Imaging 0 Volume 15, Number 4, 1997
470
measured by laser light scattering). The administered dose of iron was 0.015 mmol/kg. Animals Female Wistar rats of about 300 g body weight without any pre-treatment were used. Ten animals were examined in-vivo before and after administration of Gd-BOPTA (n = 5) or DM (n = 5). Four animals served as controls. Liver and brain specimens of all these animals and of three additional animals not imaged in-vivo were then examined with the ex-vivo model detailed below. In-vivo Experiments Animals were anesthetised with urethane (lg/kg) intraperitoneal injection and fixed on a three-floor wooden support. This allows up to 6 animals to be imaged at the same time. The animals were imaged before and 30 min after injection of the contrast agents in the tail vein. MR Imaging Protocol MR imaging examinations were performed on a 0.5 T system (GE MR Max) using the head volume quadrature coil. Field of view was 17-22cm, matrix 128 x 192 and slice thickness 5 mm. Number of excitations ranged between 4 and 6. Tl weighted inversion recovery (IR) images were obtained with TR 1500 ms, TE 25 ms and inversion time 400 ms, proton density weighted spin echo (PD SE) images were obtained with TR 2000 ms and TE 25 ms, and T2* weighted gradient recalled echo (GRE) images were acquired with TR 600 ms, TE 30 ms and 30 degree flip angle. Total acquisition time for the three sequences was about 45 min. Ex-vivo Experiments After imaging the animals were removed from the magnet and after 0.1 ml heparin (5000 U/ml) administration to prevent blood clotting, the chest was opened and the aorta incannulated. Following recision of inferior vena cava, animals were perfused with saline, in
order to remove intravascular blood, until the liver discoloured. After perfusion with formaldehyde solution (4%) for fixation, one lobe of liver and entire brain and spleen were removed and put in formaldehyde solution (4%) for l-3 weeks. For preparation of the agar gel matrix, dry agar powder (20g) (Fluka Chemie AG, Buchs, Schweiz) was mixed in cold deionized water ( 11) doped with 1 ml of Gadolinium-DTPA (Magnevist, Schering AG, Berlin, Germany) and stirred with a magnetic mixer. The suspension was heated twice to the boil with a microwave oven. The mixture was cooled to about 80°C and poured into a plastic box ( 10 X 15 X 10 cm) to form a bubble-free layer of about 1 cm thickness. After solidification of this layer, tissue specimens were placed upon the layer and fixed with some drops of the liquid agar solution (80°C). After solidification of these drops another layer of the mixture was poured. The box was placed over a phantom containing NiCl solution (0.1%) and evaluated within two weeks of preparation with the same MR imaging protocol used for the in-vivo experiments. Data Analysis The liver signal intensity to noise ratios (S/N) before and after contrast agents administration in the live rats were computed on the operator console using regions of interest, and the signal enhancement was calculated according to the following formula: (S/N after contrast - S/N before contrast)/S/N before contrast X 100%. The liver and brain signal intensity to agar ratio (S/A) of the animals injected with contrast agents was compared to the S/A of the control animals. Mean signal enhancement was calculated according to the following formula: (S/A of injected animals - S/A of control animals)/S/A of control animals x 100%. RESULTS Mean and standard deviation of the liver signal intensity to noise ratio before and after Gd-BOPTA and DM in live animals and controls are reported in Table
Table 1. Liver S/N before and after intravenous administration of Gd-BOPIA Before Gd-BOPTA
and DM in vivo After
Control
Gd-BOPTA
Control
Pulse sequences
(n = 5)
(nD_M5)
(n = 4)
(n = 5)
(n = 4)
IR PD SE
39.1 + 26.3
22.1 + 5.8
24.0 k 2.9 179.4 ? 59.4
39.2 ? 16.2 244.3 2 57.5
35.6 + 2.9 54.0 ? 1.1 228.5 2 10.6
48.0 + 31.1 30.7 t 15.1 52.8 +- 29.8
GRE
The values are the mean k standard deviation.
6.38 ? 0.8 9.40 ? 4.2 14.5 Ifr 11.5
37.8 2 5.4 47.9 ? 7.9 214.0 2 19.7
Ex-vivo
MR
imaging
of liver
intracellular
contrast
agents 0 M. MASCALCHI
ET
AL.
471
Fig. 1. Axial images obtained with Tl weighted IR (TR 1500 ms, TE 25 ms and inversion time 400 ms) sequence in two animals before (top row) and after (bottom row) intravenous administration of 0.075 mmol/Kg Gd-BOPTA showing moderate homogeneous signal increase of the liver.
Fig. 2. Axial images obtained with T2 * weighted GRE (TR 600 ms, TE 30 ms and 30 degree flip angle) sequence in two animals before (top row ) and after (bottom row ) intravenous administration of 0.015 mmol/kg DM showing marked homogeneous signal decrease of the liver.
1. Gd-BOPTA increased the signal of the liver on Tl (Fig. 1) and PD weighted images, whereas it decreased the signal on T2* weighted images. The mean signal enhancement caused by Gd-BOPTA in Tl, PD and T2* weighted images was +23%, +28% and -70%. respectively. DM caused a marked decrease of the liver signal in all the sequences which was more pronounced in T2 * weighted images (Fig. 2). The mean signal enhancement induced by DM was -71%, -76% and -94% for Tl, PD and T2* weighted images. Liver and brain signal intensity to agar ratios for animals injected with Gd-BOPTA, DM and controls are reported in Table 2. No signal change of the brain was seen in specimens of animals injected with GdBOPTA and DM as compared to controls (Fig. 3). Liver signal was increased by Gd-BOPTA and reduced by DM on all the sequences. Conspicuity of the signal change induced by Gd-BOPTA and DM was higher for Tl (Fig. 3a) and T2* (Fig. 3c) weighted images, respectively. Mean liver enhancement rate induced by
Gd-BOPTA in the ex-vivo model was +22%, +5% and +27% for Tl, PD and T2 * weighted images. Mean liver enhancement rate induced by DM was -27%, - 19% and -31% for Tl, PD and T2 * weighted images. Signal changes similar to those observed in liver were presente in the spleen specimens (Fig. 3 ) .
DISCUSSION Gd-BOPTA is an octadentate chelate of Gadolinium which has a higher rate of biliary excretion than GdDTPA.3m5 In a previous in-vivo animal study using Tl weighted IR images, liver enhancement after intravenous injection of Gd-BOPTA was consistently higher than after injection of Gd-DTPA for doses between 0.25 and 1 mmol/kg with peak enhancement within 30 min after injection.4 DM consists of superparamagnetic magnetite (Fe,O,) cores coated with hydrophilic dextran.6-9 After intravenous administration, DM is captured by the reticuloendothelial cells of the liver and
Table 2. Liver and brain S/A of rats injected with Gd-BOPTA and DM in the ex-vivo model Brain
Liver Pulse sequences
Gd-BOPTA (n = 6)
(nD_M6)
Control (n = 5)
Gd-BOPTA (n = 6)
(nD:6)
Control (n = 5)
IR PD SE GRE
1.00 + 0.18 0.78 + 0.03 0.76 2 0.15
0.59 2 0.17 0.61 +- 0.13 0.41 ? 0.19
0.81 2 0.06 0.75 IL 0.11 0.59 ? 0.11
1.12 ? 0.14 0.84 -t 0.01 0.99 + 0.25
0.96 + 0.19 0.87 + 0.11 0.85 + 0.19
0.97 + 0.16 0.85 + 0.05 0.72 -e 0.08
The values
are the mean
t standard
deviation.
412
Magnetic ResonanceImaging 0 Volume 15, Number 4, 1997
Fig. 3. (a) Images of the ex-vivo model obtained with Tl v(eighted IR (‘TR 1500 ms, TE 25 ms and inversion time 400 ms; (b) PD SE (TR 2000 ms and TE 25 ms) and (c) T2* we:ighted GRE (TR 600 ms, TE 30 ms and 30 degree flip angle) sequence. High liver and spleen signal of animals injected P&h Gd-BOP TA and low signal of animals injected with DM are observed in all sequences. The former is more evident in ( a), the latta : in (c). The brain signal intensity is unaffected by contrast injection. The black dots (arrows) in (c) are due to magnetic SUIsceptibility effects of air bubbles. (Figure continues)
spleen and determines a strong shortening of the T2 relaxation of the nearby protons. This is better demonstrated by T2* weighted GRE than SE sequences because of the sensitivity of the former to magnetic susceptibility effects.g In our study Gd-BOPTA created positive liver enhancement on IR and PD SE images, whereas signal was decreased on GRE sequences in-vivo. This is consistent with theoretical analysis” taking into account the long TR, long ‘IX and flip angle of the GRE sequence we employed. The liver signal decrease observed with all sequences in our live animals after injection of DM is consistent with prior observations.6-g The effects of the contrast agents we observed in the liver of live animals were still appreciable in the specimens of the same animals examined with the exvivo model. A general decrease of the signal changes induced by the contrast agents was however observed.
In addition, in the ex-vivo model, Gd-BOPTA caused a slight increase of the liver signal on T2* weighted GRE images. Computation of Tl and T2 relaxation times could be useful in order to quantify these effects but were not performed in our study. The differences among in vivo and ex-vivo signal changes can be explained taking into account, on the one hand, the distribution of Gd-BOPTA and DM, and, on the other hand, the modification induced in the liver parenchyma by preparation of the ex-vivo model. Although Gd-BOPTA and DM are regarded as intracellular contrast agents, some amount of the injected contrast is contained in the intravascular and interstitial spaces until it is captured by target cells.” Thus, the overall liver signal change caused by these agents can be thought of as the summation of the effects of an intracellular and an extracellular component. The concentration of the contrast agent in the latter is decreased
Ex-vivo MR imaging of liver intracellular contrast agents 0 M. MASCALCHI
ETAL.
473
Second, the freezing of the contrast agent distribution in the ex-vivo model may be useful for determination of the contrast agent at different main magnetic fields not only in the case of superparamagnetic iron oxides l3 but also in the case of paramagnetic hepatobiliary contrast agents. In fact, as recently pointed out by Elster, I4 the main magnetic field of the MR system influences paramagnetic contrast effects due to its relationship with tissue relaxation rates. The ex-vivo preparation with its long maintanance (personal unpublished observation) could afford a reliable comparison of the effect of a given contrast at different field strengths, by simply serially examining the box with different MR systems. Finally, since the ex-vivo method is not suitable to investigate extracellular contrast agents as Gd-DTPA (unpublished observations), it might indirectly provide an estimation of the true extent of the intracellular compartment of newly designed liver contrast agents. Obviously, implications for clinical applications in humans of the observations made on the ex-vivo model will require validation with in-vivo animal experiments. Furthermore, the contribution, if any, of formalin fixation to the effects of the contrast agents observed in the ex-vivo model has to be addressed. In conclusion, our study indicates that an ex-vivo model can be useful for preliminary investigations of intracellular liver contrast agents.
REFERENCES Fig. 3. continued.
by exsanguination and perfusion and this could explain the lower “enhancement” observed in the ex-vivo as compared to in-vivo conditions. Another factor of possible relevance is the fixation in formalin of the specimens excised for the ex-vivo model. Tovi et al.‘* reported that formalin, a fixative that does not coagulate protein, leads to a slight decrease of the water content of tissues and can alter tissue relaxation rates. In our study fixation time in formalin ranged between 1 and 3 weeks. The effects of different fixation time have to be addressed in further studies. Several potential advantages exist for the ex-vivo model tested in this study for the preclinical evaluation of intracellular contrast agents for the liver. First, the ex-vivo model eliminates motion artifacts from bowel, respiratory motion and arterial pulsation encountered in-vivo. The reduced image blurring and improved spatial resolution might be especially valuable in evaluating models of multiple small (primary or secondary) liver neoplasms with potentials for better MR imaging-histopathological correlations.
1. Schmiedl, U.P.; Maravilla, K.R.; Nelson, J.A. Improved method for in-vivo magnetic resonance contrast media research. Invest. Radiol. 26:65-70; 1991. 2. Wagner, S.; Pfefferer, D.; Taupitz, M.; Kresse, M.; Lawaczeck, R.; Hamm, B.; Wolf, K.J. Intravenous MR-lymphography with ultrasmall iron oxide particles: In vivo and ex vivo MRI examinations in rats. In: Book of Abstracts: Eleventh annual scientific meeting of the Society of Magnetic Resonance in Medicine. Berlin, Germany: SMRM; 1992:570. 3. Vittadini, G.; Felder, E.; Tirone, P.; Lorusso, V. B 19036, a potential new hepatobiliary contrast agent for MR proton imaging. Invest. Radiol. 23( suppl 1):246-248; 1988. 4. Pavone, P.; Patrizio, G:; Buoni, C.; Tettamanti, E.; Passariello, R.; Musu, C.; Tirone, P.; Felder, E. Comparison of Gd-BOPTA with Gd-DTPA in MR imaging of rat liver. Radiology 176:61-64; 1990. 5. Vittadini, G.; Felder, E.; Musu, C; Tirone P. Preclinical profile of Gd-BOPTA: A liver specific MRI contrast agent. Invest. Radiol. 25:S59-S62; 1990. 6. Majumder, S.; Zoghbi, S.; Pope, C.F.; Gore, J.C. Quantitation of MR relaxation effects of iron oxide particles in liver and spleen. Radiology 169:653-655; 1988. 7. Magin, R.L.; Basic, G.; Niesman, M.R.; Alameda, J.C.; Wright, S.M.; Swartz, H.M. Dextran magnetite as a liver contrast agent. Magn. Res. Med. 2&l-16; 1991.
414
Magnetic Resonance Imaging 0 Volume 15, Number 4, 1997
8. Iannone,A.; Federico,M.; Tomasi,A.; Magin, R.; Casasco,A.; Calhgaro,A.; Vamrini,V. Detectionandquantitation in rat tissuesof the superparamagnetic magnetic resonance contrastagentdextranmagnetiteasdemontrated by electronspin resonancespectroscopy.Invest. Radiol. 27:450-455; 1992. 9. Kawamori,Y.; Matsui, 0.; Kadoya, M.; Yoshikawa,J.; Demachi,H.; Takashima, T. Differentiationof hepatocellular carcinomasfrom hyperplasticnodulesinducedin rat liver with ferrite-enhancedMR imaging. Radiology 183:65-72; 1992. 10. Davis,P.L.; Parker,D.L.; Nelson,J.A.; Gillen,J.S.;Runge, V.M. Interactionsof paramagnetic contrastagentsandthe spin echo pulse sequence.Invest. Radiol. 23:381-388; 1988.
11. Kreft, B.P.; Tanimoto,A.; Stark, D.D.; Baba, Y.; Zhao, L.H.; Chen,J.T.;Compton,C.C.;Finn, J.P.;Cavagna,F.M. Enhancementof tumor-liver contrast-to-noise ratio with gadobenate dimegluminein MR imagingof rats.J. Magn. Reson.Imaging.3:41-49; 1993. 12. Tovi, M.; Ericsson,A. Measurements of Tl andT2 over time in formalin-fixed human whole-brainspecimens. Acta. Radiol.33400-404; 1992. 13. Thickman,D.; Hendrick, R.E.; Jejian, K.A.; Schanker, C.S. Liver-lesiontissuecontraston MR images:Effect of iron oxide concentrationand magneticfield strength. Radiology 176557-562, 1990. 14. Elster,A.D. Field-strengthdependence of gadoliniumenhancement:Theory andimplications.Am. J. Neuroradiol. 15:1420-1423,1994.