An evaluation of gadolinium polyoxometalates as possible MRI contrast agent

Magnetic Resonance Imaging 20 (2002) 407– 412

An evaluation of gadolinium polyoxometalates as possible MRI contrast agent Jianghua Fenga, Xiaojing Lia, Fengkui Peia,*, Guoying Suna, Xu Zhangb, Maili Liub a

b

Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun, 130022, P. R. China Laboratory of Magnetic Resonance and Atomic and Molecular Physics, Wuhan Institute of Physics and Mathematics, Chinese Academy of Sciences, Wuhan, 430071, P. R. China Received 12 February 2002; accepted 25 May 2002

Abstract Two gadolinium polyoxometalates, K9GdW10O36 and K11[Gd(PW11O39)2], have been evaluated both in vivo and in vitro as candidates for tissue-specific MRI contrast agents. T1-relaxivities of 6.89 mM⫺1 䡠 s⫺1 for K9GdW10O36 and 5.27 mM⫺1 䡠 s⫺1 for K11[Gd(PW11O39)2] are slightly higher than that of the commercial MRI contrast agent (Gd-DTPA). Both compounds bind with bovine serum albumin and human serum transferrin and favorable liver-specific contrast enhancement in in vivo MRI with Sprague-Dawley rats after i.v. administration has been demonstrated. Imaging studies demonstrate that the two agents have a long residence time, showing MR signal enhancement in the liver for more than 40 min, longer than commercially available contrast agents. In vivo and in vitro assays showed that GdW10 and Gd(PW11)2 are promising liver-specific MRI contrast agents and GdW10 may be used in the diagnosis of the pathological state. However, with the higher acute toxicity, the two gadolinium polyoxometalates need to be modified and studied further before clinical use. © 2002 Elsevier Science Inc. All rights reserved. Keywords: Gadolinium polyoxometalates; Relaxivity; MRI; Contrast agent

1. Introduction Magnetic Resonance Imaging (MRI) is at present well established as a safe and efficient imaging technique for the human body in clinical diagnosis. However, the technique can be non-specific, that is, it may be impossible to distinguish between many pathological conditions, for example between cancer and the edema surrounding the cancer; active and inactive multiple sclerosis plaques; and bowel from other adjacent organs. Contrast agents have expanded the utility of MRI by helping to distinguish between normal and pathological conditions such as tumors and tissue damage [1] and functional information may also be derived from experiments using target-specific contrast agents. These agents enhance various portions of the MR image by changing, usually decreasing, the relaxation rate of protons in the immediate vicinity of the agent and, thus, allowing the area of interest to be much more conspicuous than surrounding tissue. Developing effective, non-toxic and organ-specific * Corresponding author. Tel.: ⫹86-431-5262219; fax: ⫹86-4315685653. E-mail address: [email protected] (F. K. Pei).

contrast agents for in vivo image enhancement has been an active area of research for two decades [2–11]. Most attention has been devoted to organic ligands for complexing gadolinium due to the kinetic stability of the complexes, the high relaxivities and possible target specificity. Four gadolinium chelates used as MR contrast agents (gadopentetate dimeglumine, gadoteridol, gadodiamide and gadoterate meglumine) for intravenous administration currently enjoy widespread approval and use [12]. However, contrast agents in clinical use suffer from several defects, for example, they are not tissue-specific, they are rapidly excreted, their synthesis requires a complex, tedious, and expensive process, and/or they may provoke allergic reactions in the recipient. Therefore, novel target-specific MRI contrast agents with favorable properties need to be developed. Here we have evaluated two lanthanide polyoxometalates (POMs), K9GdW10O36 (GdW10) and K11[Gd(PW11O39)2](Gd(PW11)2), in vivo as well as in vitro in regard to their potential application in medical imaging. Both showed favorable tissue-specificity to liver and kidney. Our decision to investigate lanthanide POM was primarily based on the fact that transition metal-substituted heteropoly complexes are regarded as purely inorganic analogs

0730-725X/02/$ – see front matter © 2002 Elsevier Science Inc. All rights reserved. PII: S 0 7 3 0 - 7 2 5 X ( 0 2 ) 0 0 5 2 1 - 0

408

J.H. Feng et al. / Magnetic Resonance Imaging 20 (2002) 407– 412

of porphyrins [13] whose metal complexes are very promising tissue- and tumor-specific contrast agents [14 –16]. Anti-HIV and anti-viral activity is a common feature of 80% of POMs although the degree of effectiveness varies [17], which may be beneficial to the body. POMs are multidentate hard base ligands which could lead to kinetic stability due to strong bond formation to the hard acid Gd(III), moreover POM complexes have high water solubilities which may minimize uptake in lipophilic organs and target the blood pool for application in magnetic resonances angiography [18].

where RI(t) corresponds to the normalized signal intensity measured at time point t and RI(0) is the normalized signal intensity precontrast.

3. Results For paramagnetic aqueous solution, the solvent longitudinal relaxation rate, 1/T1, is linearly dependent on the concentration of the paramagnetic species ([M]) as described in equation (1): (1/T1)obsd ⫽ (1/T1)d ⫹ R1 ⫻ [M]

2. Materials and methods GdW10 and Gd(PW11)2 were synthesized and characterized according to the procedure given in ref. 19. GdDTPA complex was prepared according to a routine method [20]. Proton relaxivities of the two lanthanide polyoxometalates in D2O, bovine serum albumin (BSA) (0.725 mmol 䡠 L⫺1, in D2O) and human serum transferrin (hTF) (35 ␮mol 䡠 L⫺1, in D2O) solutions at 9.39 T were determined on Varian Unity400 NMR spectrometer at 25°C by using a standard inversion-recovery pulse sequence. T1 relaxation times were measured for five concentrations of each complex in the concentration range from 0 to 5.00 mM (0, 0.50, 1.00, 2.00, 3.50, and 5.00 mM) in unbuffered aqueous solutions. The relationship of relaxation times vs concentration for the investigated complexes was linear with the correlation coefficient greater than 0.994. The toxicity of the complexes was assessed by their median lethal dose (LD50). The LD50 of GdW10 and Gd(PW11)2 were estimated by single intravenous (tail vein) injection of different volumes of a 10 mmol 䡠 L⫺1 solution at a rate of 0.2 mL/min in Kunming male and female mice weighing 19 –22 g. The animals were observed for 14 days after the injection. MR imaging in rats was carried out on a Bruker BIOSPEC-47/30 MRI imager. A series of T1-weighted images of abdomen were obtained after intravenous injection (tail vein) of a total of 0.5 mL of a 0.02 mol 䡠 L⫺1 Gd(PW11)2 solution or a total of 1.0 mL of a 0.01 mol 䡠 L⫺1 GdW10 solution in five male Sprague-Dawley (SD) rats (90 –120 g), respectively. MR imaging after intravenous injection of GdDTPA (a total of 1.0 mL of a 0.01 mol 䡠 L⫺1) was also studied. MRI signal intensity enhancement was monitored up to 90 min with an image every 5 min. Axial imaging of 2-mm thick slices were acquired by multislice and multiecho (MSME) techniques using TR ⫽ 500 ms, TE ⫽ 15 ms, four averages. A field of view (FOV) 5 ⫻ 5 cm2 and a matrix of 128 ⫻ 256 were employed. A water tube was placed in the field of view as a phantom reference. Thus intensity enhancement (IE) of region of interest (RI) at time point t is expressed by IE ⫽ 100 (RI共t兲 ⫺ RI(0))/RI(0)

(1)

where (1/T1)obsd and (1/T1)d are the measured solvent relaxation rate in the presence and absence of the paramagnetic species, respectively. The relaxivity, R1, is commonly defined as the slope of (1/T1)obsd vs [M] in units of mM⫺1 䡠 s⫺1, which reflects the relaxation enhancement ability of a paramagnetic compound. Compared to GdDTPA (T1 ⫽ 4.1 mM⫺1 䡠 s⫺1 in water) [21], the two lanthanide polyoxometalates have a higher relaxivity. The T1-relaxivities of GdW10 were 6.89 mM⫺1 䡠 s⫺1 in D2O, 7.47 mM⫺1 䡠 s⫺1 in 0.725 mmol 䡠 L⫺1 BSA, and 7.30 mM⫺1 䡠 s⫺1 in 35 ␮mol 䡠 L⫺1 hTF. The corresponding T1-relaxivities of Gd(PW11)2 were 5.27, 5.60 and 6.13 mM⫺1 䡠 s⫺1, respectively. From the results, we can deduce that protein binding contents of GdW10 were 5.6% for BSA and 5.3% for hTF by using of the method described in ref. 20. These were 4.2% and 10.9% for Gd(PW11)2, respectively. An intravenous LD50 of about 3.0 mmol/kg body wt for Gd(PW11)2 and 0.42 mmol/kg body wt for GdW10 was estimated. Death occurred within the first 12 h after injection, and no long-term toxicity and effects was seen for the survivors during the 14-day observation time. GdW10, Gd(PW11)2 and GdDTPA have been evaluated as magnetic resonance imaging contrast agents in SD rat liver, kidney and stomach. As shown in Figs. 1– 4, signal enhancement in the T1-weighted image was obtained in the abdomen after i.v. administration of contrast agents. The doses of GdW10, Gd(PW11)2, and GdDTPA were 0.087 ⫾ 0.005, 0.092 ⫾ 0.002 and 0.088 ⫾ 0.013 mmol Gd/kg, respectively, these were a little lower than the typical clinical dose (0.1 mmol Gd/kg). The signal intensity of the rat liver parenchyma and kidney increased shortly after injection of GdW10. Fig. 1 shows the T1-weighted image of the rat liver and stomach for precontrast and after application of GdW10. The maximum enhancement for liver was found in 25– 60 min after injection, and this stronger enhancement persisted throughout the imaging period (Fig. 3). The experimental dose of GdW10 induced 45.7% ⫾ 8.6% enhancement in signal intensity of the liver, and 44.7% ⫾ 6.3% enhancement for the kidney. There is a large variation of percentage enhancement values between the rats as indicated by the large

J.H. Feng et al. / Magnetic Resonance Imaging 20 (2002) 407– 412

409

Fig. 1. MR images of a transverse section of a Sprague-Dawley rat showing the liver and stomach using GdW10 as the contrast enhancement agent. Sections (1) through (6) are shown as follows: (1) control, prior to injection; (2) post-injection, 15 min; (3) post-injection, 30 min; (4) post-injection, 45 min; (5) post-injection, 65 min; (6) post-injection, 80 min.

standard deviations, but the enhancement patterns were essentially the same in the different rats. The T1-weighted images of the rat kidney before and after intravenous injection of Gd(PW11)2 are shown in Fig. 2. Liver and kidney all demonstrate early and long-time enhancement during the experimental time (Fig. 3 and Fig. 4). The enhancement patterns of liver in the different rats were very similar with peak percentage enhancement (37.2% ⫾ 3.9%) reached within 20 –55 min, it decreased slowly thereafter. A mean percentage enhancement for kidney of close to 40% was reached within the first 20 min. Enhancement persisted throughout the imaging period. The standard deviations of the percentage enhancement values of the abdomenal tissue were large, but the enhancement patterns in the different rats were essentially the same. Rat stomach images were also obtained before and after injection of GdW10 (see the circular region in the rightlower corner in Fig. 1(1– 6)). Of the five SD rats tested, two gave dramatic enhancement for stomach after intravenous injection (128.3% in 40 –70 min, and 372.3% in 50 – 80 min respectively!); the others had no obvious enhancement. GdDTPA induced a strong signal enhancement in kidney and also in the whole hepatic parenchyma (Fig. 3 and 4). But the enhancement rapidly decreased following contrast injection.

4. Discussion Serum albumin is the richest protein of human blood plasma, and human serum transferring (hTF) is the most important binding protein with lanthanide complexes [22]. Both would play a crucial role on the uptake, transportation, biodistribution and excretion of contrast agent in the human body [20]. Therefore, to gain insight into the effects of contrast agents in the body, the study of the interaction between contrast agents and BSA and hTF is very important. Significant enhancement of solvent proton relaxation rates was produced for the two gadolinium POMs in BSA and hTF solution, indicating the formation of paramagnetic macromolecular adducts [9,23]. An obvious effect in the liver after injection of GdW10 and Gd(PW11)2 was probably due to the protein binding with serum albumin and serum transferrin and thus leading to effective transport by the blood to the liver. Similar to Gd-EOB-DTPA, GdW10 and Gd(PW11)2 presumably enters the hepatocytes by a specific carrier-mediated mechanism [24]. This carrier is known to eliminate anionic exobiotic molecules from the body. According to the relaxivities data and the results of the in vivo MRI studies, hTF probably has a larger effect than serum albumin in transporting the compounds in the body. In contrast to the rapid rate of metabolization of GdDTPA, the

410

J.H. Feng et al. / Magnetic Resonance Imaging 20 (2002) 407– 412

Fig. 2. MR images of a transverse section of a Sprague-Dawley rat showing the kidney using Gd(PW11)2 as the contrast enhancement agent. Sections (1) through (6) are shown as follows: (1) control, prior to injection; (2) post-injection, 15 min; (3) post-injection, 30 min; (4) post-injection, 45 min; (5) post-injection, 60 min; (6) post-injection, 75 min.

animal group receiving the two agents displayed an increase in the MR signal intensity from the liver shortly after injection with the peak enhancement in liver persisting for a longer time. Therefore, the timing of contrast injection and

long data acquisition times become less significant while the optimal imaging window becomes feasible in tens of minutes rather than seconds. As well as the liver, the kidney was also enhanced, indicating that GdW10 and Gd(PW11)2 were being removed

Fig. 3. Time dependence of the GdDTPA (filled white), Gd(PW11)2 (filled light gray), and GdW10 (filled gray) induced hepatic intensity enhancement obtained from the T1-weighted images of the type of shown in Fig. 1.

Fig. 4. Time dependence of the GdDTPA (filled white), Gd(PW11)2 (filled light gray), and GdW10 (filled gray) induced renal intensity enhancement obtained from the T1-weighted images of the type shown in Fig. 2.

J.H. Feng et al. / Magnetic Resonance Imaging 20 (2002) 407– 412

from the blood and eliminated, at least in part, through the kidney. However, in contrast to GdDTPA (enhancement in the kidney, 82.2% ⫾ 10.9% in 10 – 45 min after injection), the enhancement of the POMs in the kidney was lower. This probably supports the hypothesis that the two gadolinium POMs were also excreted through the hepatocytes into the bile and feces. Elimination of the two gadolinium POMs is probably through both the liver’s biliary system and the kidney, similar to the effect on porphyrin complexes [14]. From Fig. 4, the rapid excretion rate of GdDTPA from the body can be seen. Conversely, the MRI intensity enhancement of GdW10 and Gd(PW11)2 has not significantly decreased during the imaging period, indicating that the renal metabolization rate of the contrast agent was very slow, which can be useful when studying variations in organ blood volume and capillary permeability [25]. The high variability of the percentage enhancement values in the different rats is most likely related to the effect of differing body weight, on the dose, in the different rats. Furthermore the kidney is a very small structure, a partial volume effect [25] is also partially responsible for the high variability of the percentage enhancement value from this region. Variability of the values will be more significant in clinical diagnosis. Interestingly, the signal intensity in images of the stomach of the two rats obviously increased after injection of GdW10. It is very likely that the strong signal intensity of the stomach after injection of GdW10 implied a pathological state of the stomach, such as edema, cyst or tumor [26]. It is also possible that the gastric walls of the two rats were much thicker than that of the other rats or folded to a certain extent during the experiment, this would cause more accumulation of contrast agent into the stomach and bloodstream and affect the results of the imaging. Unfortunately, the experimented rats were not studied further. The good stability, considerably water solubility (20 mM corresponding to GdW10 and 90 mM corresponding to Gd(PW11)2), and the effectiveness over a wide pH range of the two gadolinium POMs was observed in our previous experiments [19]. These are very useful during the practical use of contrast agents [27]. From the LD50 of GdW10 and Gd(PW11)2, it is more toxic than the gadolinium (III) compounds currently used as clinical MRI contrast agents [1]. However, it is less toxic than manganese polyhydroxylamide porphyrin, which has an LD50 value of 0.1 mmol/kg [14]. The toxicity of gadolinium polyoxometalates may account for the challenges to GdPOMs by metal ions [18], causing disruption of the biochemical equilibrium, but it is very likely that the high anionic charge on GdPOMs also contributes to its toxicity. From the different roles of net charge, the size, the metal type and the rotational correlation time on the relaxivity and stability of these contrast agents, the design of non-ionic and multi-centric lanthanide POMs is in progress to develop more favorable tissue-specific MRI contrast agents.

411

5. Conclusion The two lanthanide polyoxometalates, GdW10 and Gd(PW11)2, were evaluated in vivo and in vitro as candidates for tissue-specific MRI contrast agents. The measured relaxivities for GdW10 and Gd(PW11)2 in D2O, bovine serum albumin and human serum transferrin showed favorable relaxation ability and binding affinity with BSA and hTF. Our preliminary in vitro and in vivo studies with GdW10 and Gd(PW11)2 have proven that the two agents could be promising liver-specific MRI contrast agents and GdW10 may be used in the diagnosis of pathological states. These agents 1. are completely soluble in water and suitable for injection in an in vivo experiment; 2. display a higher relaxivity than that of the widely used contrast agent (GdDTPA); 3. show a long residence time in the body; and 4. induce strong signal enhancement in liver, further GdW10 may be helpful in the diagnosis of the pathological state of the stomach. However, acute toxicity studies showed that the lanthanide polyoxometalates were more toxic than the gadolinium (III) compounds currently being used as MRI contrast agents. By appropriate modification, the lanthanide polyoxometalates may be applied to clinical use.

Acknowledgments This work was supported by State Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics and Science and Technology Foundation of Changchun City.

References [1] Lauffer RB. Paramagnetic metal complexes as water proton relaxation agents for NMR imaging: theory and design. Chem Rev 1987; 87:901–27. [2] Caravan PJ, Ellison J, McMurry J, Lauffer RB. Gadolinium(III) chelates as MRI contrast agents: structure, dynamics, and applications. Chem Rev 1999;99:2293–352. [3] Zhang SR, Sun DY, Li XY, Pei FK, Liu SY. Synthesis and solvent enhanced relaxation property of water-soluble endohedral metallofullerenols. Fullerene Sci Techn 1997;5(7):1635– 43. [4] Reynolds CH, Annan N, Beshah K, Huber JH, Shaber SH, Lenkinski RE, Wortman JA. Gadolinium-loaded nanoparticles: new contrast agents for magnetic resonance imaging. J Am Chem Soc 2000;122: 8940 –5. [5] Wolfgang S, Sanjay S, Jan P, Ralph W, Mukesh H, William MS, Peter FH. MR imaging of the liver with Gd-BOPTA: quantitative analysis of T1-weighted images at two different doses. J Magn Reson Imag 1999;10:80 –3. [6] Lucia JMK, Joost D, Rob JVDG, Albert DR. Blood pool contrast agent CMD-A2-Gd-DOTA-enhanced MR imaging of infarcted myocardium in pigs. J Magn Reson Imag 1999;10:170 –7.

412

J.H. Feng et al. / Magnetic Resonance Imaging 20 (2002) 407– 412

[7] Susanne O, James F, Timothy PLR, Michael FW, Kenneth DA, Jeffry SM, Mark AI, Robert CB. Comparison of gadopentetate dimeglumine and albumin-(Gd-DTPA)30 for microvessel characterization in an intracranial glioma model. J Magn Reson Imag 1998;8(4):799 – 806. [8] Anderson SA, Rader RK, Westlin WF, Null C, Jackson D, Lanza GM, Wickline SA, Kotyk JJ. Magnetic resonance contrast enhancement of neovasculature with v3-targeted nonoparticles. Magn Reson Med 2000;44:433–9. [9] Bligh SWA, Chowdhury AHMS, Kennedy D, Luchinat C, Parigi C. Non-ionic bulky Gd(III) DTPA-bisamide complexes as potential contrast agents for magnetic resonance imaging. Magn Reson Med 1999; 41:767–73. [10] Yu X, Song SK, Chen JJ, Scott MJ, Fuhrhop RJ, Hall CS, Gaffney PJ, Wickline SA, Lanza GM. High-resolution MRI characterization of human thrombus using a novel fibrin-targeted paramagnetic contrast agent. Magn Reson Med 2000;44:867–72. [11] Emmanuel F, Hana L, Anne FJ, Roger D, Soraya B, Claire C, Regine F, Christophe R, Michel D, Chantal R. A new gadolinium-based contrast agent for magnetic resonance imaging of brain tumors: kinetic study on a C6 rat glioma model. J Magn Reson Imag 2001;14: 97–105. [12] Runge VM. Safety of approved MR contrast media for intravenous injection. J Magn Reson Imag 2000;12:205–13. [13] Gilbert J, Dassab D, Pawlow JH, Santora BP, Fiel RJ, Joshi VN, Kozik M. First characterization of the influence of paramagnetic heteropoly complexes of tungsten on relaxation times of water. Separation of various contributions to relaxivities. Inorg Chem 1995;34: 924 –7. [14] Bradshaw JE, Gillogly KA, Wilson LJ, Kumar K, Wan XM, Tweedle MF, Hernandez C, Bryant RG. New non-ionic water-soluble porphyrins: evaluation of manganese(III) polyhydroxylamide porphyrins as MRI contrast agents. Inorg Chim Acta 1998;275–276:106 –16. [15] Luo Y, Mei EW, Zhuo RX. Studies on water-soluble metalloporphyrins as tumor targeting magnetic resonance imaging contrast agents. Chem J Chin Univ 1995;16(10):1629 –32. [16] Gahrouei DS, Williams M, Rizvi S, Allen BJ. In vivo studies of Gd-DTPA-monoclonal antibody and Gd-porphyrins: potential mag-

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26] [27]

netic resonance imaging contrast agents for melanoma. J Magn Reson Imag 2001;14:169 –74. Hill CL, Weeks MS, Schinazi RF, Paymond F. Anti-HIV-1 activity, toxicity, and stability studies of representatives structural families of polyoxometalates. J Med Chem 1990;33(10):2767–72. Crook WJ. Studies of lanthanide interactions with polyoxometalates: evaluation as magnetic resonance imaging agents. Doctoral dissertation, The Florida State University, 1995. pp. 1–2. Feng JH, Li XJ, Jing FY, Pei FK. Synthesis, characterization, stability and relaxation behaviors of rare earth heteropolymetalates. Acta chimica sinica 2000;58:423– 6. Li XY, Li XJ, Zhang SR, Pei FK. NMR relaxation studies of GdDTPA in human serum albumin solution. Polyhedron 1999;18: 695–7. Koenig SH, Baglin C, Brown RD, Brewer CF. Magnetic field dependence of solvent proton relaxation induced by Gd3⫹ and Mn2⫹ complexes. Magn Reson Med 1984;1(4):496 –501. Meng L. Distribution of rare earth species in blood and some biological effect. Post-doctoral thesis, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, 1997. Bligh SWA, Harding CT, Sadler PJ, Bulman RA, Bydder GM, Pennock JM, Kelly JD, Latham IA, Marriott JA. Use of paramagnetic chelated metal derivatives of polysaccharides and spin-labeled polysaccharides as contrast agents in magnetic resonance imaging. Magn Reson Med 1991;17:516 –32. Weinmann HJ, Giampieri GS, Willich HS, Vogler H, Frenzel T, Geies H. A new lipophilic gadlinium chelate as a tissue-specific contrast medium for MRI. Magn Reson Med 1991;22:233–7. Li KCP, Quisling RG, Armitage FE, Richardson D, Mladinich C. In vivo MR evaluation of Gd-DTPA conjugated to dextran in normal rabbits. Magn Reson Imag 1992;10:439 – 44. Yang GF, Jin BS. Magnetic resonance imaging diagnostics. Shanxi Science and Technology Press: Xi’an, 1991. Saab-Ismail NH, Simor T, Gaszner B, Lorand T, Szollosy M, Elgavish GA. Synthesis and in vivo evaluation of new contrast agents for cardiac MRI. J Med Chem 1999;42:2852– 61.