WO1989011874A1

WO1989011874A1 - Magnetic resonance imaging

Info

Publication number: WO1989011874A1
Application number: PCT/GB1989/000617
Authority: WO
Inventors: Laurence David Hall; Julian Hugh Braybrook
Original assignee: Laurence David Hall; Julian Hugh Braybrook
Priority date: 1988-06-07
Filing date: 1989-06-02
Publication date: 1989-12-14
Also published as: GB8813425D0; AU3767089A

Abstract

A preparation suitable for clinical administration for use as a contrast agent in magnetic resonance imaging comprises a porous particulate substance with paramagnetic material bound with respect to the surface thereof, the preparation being such that no significant release of the paramagnetic material occurs in undesired locations in clinical use. Such materials can be safely used for clinical diagnostic purposes with humans and animals.

Description

Title: Magnetic Resonance Imaging

Field of the Invention

This invention relates to magnetic resonance imaging (MRI), and concerns magnetic resonance contrast agents suitable for clinical administration.

Background to the Invention

MRI is a well known technique, based on the variation of the magnetic properties of protons and other species in different local environmental conditions, which is used for various analytical purposes including clinical diagnosis. It is known to use contrast agents, such as certain metal ions, which act to enhance spin-lattice relaxation rates (R. values), and spin-spin relaxation rates (R₂ values), increasing sensitivity and enabling improved differentiation of structures. Paramagnetic materials such as ions of copper, manganese, iron and gadolinium are good contrast agents, but their potential toxicity presents difficulties for use in clinical applications.

The present invention aims to provide novel magnetic resonance contrast agents, suitable for clinical administra ion.

Summary of the Invention According to one aspect of the present invention there is provided a preparation suitable for clinical administration for use as a contrast agent in magnetic resonance imaging, comprising a porous particulate substance with paramagnetic material bound with respect to the surface thereof, the preparation being such that no significant release of the paramagnetic material occurs in undesired locations in clinical use.

Because the preparation is such that no significant release of the paramagnetic material occurs in undesired locations in clinical use, the preparation can be safely used for clinical diagnostic purposes with humans and animals^'.

Hence in a further aspect the present invention provides a method of diagnosis using MRI, wherein a preparation in accordance with the invention is administered to a human or animal body.

The method of administration will depend on the area to be diagnosed. For example the preparation may be administered orally or as an enema for imaging of regions of the gastro-intestinal (GI) tract. Alternatively, the preparation may be introduced to the circulatory system, eg by injection, for imaging of regions such as the reticulo-endithelical system (RES), lung, liver, spleen, pancreas, kidney, bone-marrow, lymphatic system or vascular system.

Various properties and characteristics of the preparation can be readily and precisely controlled and so can be varied appropriately to suit requirements depending on the intended use of the preparation.

The porous particulate substance may be selected from a wide range of polymers and other materials. Many suitable materials are commercially available, or can be made by known techniques.

Suitable non-polymeric porous particulate substances include eg microporous glasses, zeolites such as zeolite A and zeolite Z, pillared clays, crown ethers such as 18 crown 6 ether, and porphyrins, and derivatives thereof.

Suitable polymeric porous particulate substances include eg, serum albumin (eg human or bovine), polystyrene, polyacrylamides, polycyanoacrylates, polymethacrylates, starch, cellulose, dextran, alginates and other polysaccharides, polyacrolein, polypeptides, polyorthoesters and typical hydrogels, Sepharose (agarose) (Sepharose is a Trade Mark) and derivatives thereof.

It is generally desirable to use hydrophilic materials, such as polymers with hydrophilic backbones, as it is found that the enhancement of the spin-lattice relaxation rate increases with hydrophilicity. This accords with the intuitive expectation that the effect depends on the exchange rates of water through the sphere of influence of the bound paramagnetic material.

Further, for some applications, eg use in the circulatory system, it is appropriate to use biodegradable materials such as the hydrogels, polylactic acid (polylactide ) , polyglycolic acid (polyglycolide) , polydactic acid-co- glycolic acid), poly(lac ide-co-glycolide ) or poly ortho- esters, poly acetals, synthetic polypeptides, cross-linked proteins, polycyanoacrylates, etc.

The porous particulate substance is conveniently of spherical configuration, but the form is not critical.

The size of the porous particulate substance may be selected to suit the intended use and typically may range from less than micron size (eg prepared using known emulsion polymerisation techniques) to millimetre size and over (eg prepared using known drop teqhniques). A common particle diameter size range is 5 to 500 urn. For oral administration particles are conveniently in the range of about 50 um to a few millimetres in diameter. For use in the circulatory system particles are conveniently in the range 0.1 to 100 um, generally 1 to 5 um, with particles for imaging the RES typically being about 0.1 um, and those for the lung 7 to 12 um. The size of the particles may affect the ultimate location of the material in use.

The surface area of the porous particulate substance can also be selected as appropriate depending on the intended use. For diagnostic purposes, the availability for interaction of the preparation with its surrounding media is important, with a large surface area giving a better availability for interaction and a stronger effect. However, for certain applications a large surface area is not a pre-requisite.

The porosity of the particulate substance and the nature of channels therein can also be appropriately selected, depending on the indended use. For example, the extent of porosity and tortuosity as well as the rate of diffusion through channels is important- The pores are preferably relatively large, being sufficiently large for access of molecules without being so large that leakage of attached material occurs, and are typically in the range of 30A to a few hundred A in diameter.

Porous particulate substance of desired chemical composition and with desired physical properties, such as particle size, surface area, pore properties etc, can either be obtained commercially or made using known techniques. The desired material may be made directly, or may be produced by surface modification of an existing material.

The paramagnetic material can be selected from a range of known materials having paramagnetic properties. These include ions of elements having an atomic number from 21 to 29, 42, 44, and from 58 to 70 such as chromium (III), manganese (II), iron (III), iron (II), cobalt (II), copper (II), nickel (II), praseodymium (III), neodymium (III), samarium (III) and ytterbium (III). Of particular interest are ions having a strong magnetic moment, such as gadolinium (III), terbium (III), dysprosium (JII), holmium (III) and erbium (III). A further class of paramagnetic materials is nitroxide stable free radicals, which can be divided into two groups: pyrrolidone-N-oxyl and piperidine-N-oxyl. These radicals have good chemical versatility and may be made tissue-specific, enabling targeting of materials.

The paramagnetic material may be bound directly to the particulate substance in appropriate cases, where the pore size of the particulate substance is such that the paramagnetic material can be held directly. For example, porphyrins have a suitable pore size for holdinq certain metal ions . Generally, however, the paramagnetic material is bound (covalently or otherwise) to a ligand or linking group, attached to the porous particulate substance. Suitable ligands include acids containing methylenephosphonic acid groups (-CH₂SO,Y) , methylenecarbohydroxamic acid groups (- CH-CONHOY) , carboxyethylidene groups (-CH₂CH₂COOY) , or carboxymethylene groups (-CH_COOY) of which at least 1,2 or 3 are bound to a nitrogen atom supporting the complex formation, eg EDTA, DTPA, etc. Such chelating acids form complex salts with paramagnetic metal ions.

Thus, chelating ligands/sites may be -S0₃Y, -COOY, -PO-FY, -CONHOY, where Y is a H atom, a metal equivalent and/or cation of an inorganic or organic base or amino acid as long as two of the substituents Y- are paramagnetic metal equivalents as aforementioned, or a nitroxide stable free radical.

If not all the acid H atoms of the complex-forming acid are substituted by paramagnetic material it is better for solubility if substitution of remaining H atoms by cations or inorganic and/or organic bases or amino acids, eg Li, Na, K primary, secondary and tertiary amines, lysine, arginine, ornithine, is completed.

Complex-forming acids required for the preparations may^' be manufactured in a manner known per se. Also known are techniques for conjugation to biomolecules, ie, nucleophilic groups of biomolecules, eg, amino, hydroxy, etc, with an activated derivative of complex-forming acid, eg, acid chlorides, acid anhydrides, activated esters, etc. Also known per se is the manufacture of some of the complex salts (excess acid groups should be converted to neutral complex salts using cation-forming inorgainc and/or organic bases or amino acids).

The ligand may be attached directly to the porous particulate substance in known manner, or may be linked thereto via a spacer arm. Suitable spacers include, eg, hydrophilic groups such as the bis oxirane, 1 ,4-bis-(2, 3- epoxypropoxy)-butane (an ether linkage), 6-amino-hexanoic acid and hexa-methylenedia ine, and hydrophobic groups such as a pentyl grouping (consisting of 5 methylene groups) .

The properties of the ligand, if present, and spacer arm, if present, affect the properties of the preparation such as hydrophilicity/hydrophobicity and the strength with which the paramagnetic material is bound, and so may be selected to produce a preparation with desired properties .

The porous particulate substance with bound paramagnetic material may be provided with an outer coating of controlled and variable prorsity, primarily to prevent unwanted release of the paramagnetic material in use, but possibly also for controlling timing of release of the paramagnetic material and/or for targeting the preparation to particular sites such as certain tissues. If the particulate substance is non-biodegradable a coating is generally essential: with biodegradable substances coatings are optional. Suitable coating materials include, eg, known enteric coatings, which may be applied in known manner, eg by a spray coating technique. The nature of the coating will depend on the intended use. For instance preparations for use in the GI tract must generally be coated to enable them to withstand the highly acid (about pH 1.3) environment of the stomach. Suitable coatings for this purpose include cellulose acetate coatings eg phthallate, butyrate (resistant to acid pH; dissolves thereafter), ethyl cellulose (only swells), coatings derived from copolymers of methacrylic acid, amino-ethyl methacrylates and- neutral esters of acrylic and methacrylic acid, the Eudragit resins (these materials are both resistant to acid and swell). For gastric protection, such coatings typically have a thickness in the range 5 to 15 um. Preparations for use in the circulatory system require biodegradable particulate material, so coating is optional: a typical coating suitable for this use is polylysine. However coatings may be desirable in some cases not only for tissue recognition (possibly by addition of suitable groups to the coating), but possibly also for protection from the reticuloendothelial system.

It is somewhat surprising that coated material will function effectively as a contrast agent, as it might be expected that the coating would inhibit the interaction of the paramagnetic material with protons or other species. Although it is found that the presence of a coating reduces relaxivity as compared with uncoated material, the preparations can nevertheless still function as effective contrast agents.

In a preferred aspect, the present invention provides a preparation suitable for clinical administration for use as a contrast agent in magnetic resonance imaging, comprising a porous particulate substance with paramagnetic material bound with respect to the surface thereof, and an outer coating which acts to prevent significant release of the paramagnetic material in undesired locations in clinical use.

In a further aspect, the invention also provides a preparation suitable for clinical administration for use as a contrast agent in magnetic resonance imaging, comprising a biodegradable porous particulate substance with paramagnetic material bound with respect to the surface thereof, the preparation being such that no significant release of the paramagnetic material occurs in undesired locations in clinical use.

By regulating the various factors discussed above it is clear that the characteristics and behaviour of the preparation can be precisely controlled and tailored to the intended clinical use, so that the invention provides useful preparations for clinical diagnostic purposes. For example, for use of the preparation enterically for the GI tract, the preparation can be formulated to control distribution and rate of transport of the particulate matter as well as control, if appropriate, the release of material at a specific locus. For use in the circulatory system, the preparation can be formulated to prevent any osmotic shock upon administration of such matter and to aid the hydrophilic/hydrophobic environment.

Preparations in accordance with the invention can be used in generally conventional manner for diagnostic purposes, particularly in the GI tract and circulatory system, with improved results as compared with existing materials used for these purposes. Typical uses of these materials include study of a wide variety of pathology in tioth the GI tract and the circulatory system and associated organs, such as identification/localisation of ulcers, tumours, obstructions, organ disease and abnormality. The properties of preparations in accordance with the invention additionally mean they can be used in various novel ways, eg for studying processes occuring in the circulatory system such as phagocytosis, adhesion and filtration, and for measurement of transit times through, and spatial distribution of pH within, specific regions of the GI tract to provide information useful for estimating nutritional uptake and pharmaceutical delivery.

The invention will be further described, by way of example, with reference to the accompanying drawings, in which:

Figure 1 is a graph showing the variation of the proton relaxation. rates of water induced by: (A) lO M solutions of the free metal ions, (B) lOmM solutions of the free metal ions in 2% carboxy ethylcellulose, (C) 24% w/v suspensions, in 2% carboxymethlcellulσse, of resin that had been fully saturated with metal ions;

Figure 2 is a graph showing the variation of the proton spin-lattice relaxation rates of water induced by a variation of the concentration of a fully saturated, manganese-bound resin, suspended in 2% carboxymethylcellulose;

Figure 3 is a graph showing the variation of the proton relaxation rates of water induced by 24 % w/v suspensions, in 2% carboxymethylcellulose, of: (A) Resonium A resinr (B) AGMP-50 resin, (C) Chelex 100 resin, each of which had been fully saturated with metal ions; Figure 4 is a graph showing variation of the proton relaxation rates of water induced by: (A) lOmM solutions of the free metal ions, (B) 24% w/v suspentions, in 2% carboxy ethylellulose, of Chelating Sepharose 6B resin that had been fully saturated with metal ions; and

Figure 5 is a graph showing the variation of the proton spin-lattice relaxation rates of water induced by metal- bound sulphonated polystyrene resins pre-and post-acid treatment.

Examples

Samples of sulphonated polystyrene resin ( inthrop Laboratories; mesh size 85-170um; pore size 30 A; C 36.6%: H 4.92%; S 10.86%) were immersed in solutions of the appropriate metal salts for 2.5 days, washed with distilled water, and then vacuum dried. The amount of sequestered metal ions was determined by atomic absorption and ultra-violet spectroscopy.

Nuclear magnetic resononce (NMR) relaxation measurements were made at 26°C with a Varian VXR-300 spectrometer using the inversion recovery sequence (see references 1 and 2) with phase-cycling and a composite 180° pulse (see references 3 to 5), and the R.. values were calculated from an exponential fit of the data using the Varian software. Standard 5mm NMR tubes were fitted with polytetraflouroethylene plugs to keep the sample within the region of the reciever coil (see reference 2).

Two series of comparisons serve to illustrate the effects of these metallated resin beads on the R-, values of the protons of water. Figure 1 summarises the relative efficacy of different metal ions, both sequestered and free. In accord with many previous studies (see references 2 and 6), the diamagnetic ions have little effect and the relaxation efficiency of the different paramagnetic species increases with the number of unpaired electrons. Comparison of curves (A) and (B) shows that the relaxation efficiency of lOmM metal ions in aqueous solution is not significantly-influenced by the presence of 2% carboxymethylcellulose gel. The data in curve (C) shows that when the metal ions are bound to the resin they have a lower relaxivity than when free (having allowed for the excess of metal of metal ions present on the resin) - Nevertheless, they still cause substantial increases in the spin-lattice relaxation rate of water from its normal value (0.34s^-1); Fe¹¹: 10.0% (440mM) Fe²⁺, x75; Cu¹¹: 10.4% (400mM) Cu²⁺, χ85; Mn¹¹: 10.0% (450mM) Mn²⁺, xlOO; Fe¹¹¹: 7.5% (330mM) Fe³⁺, xlOO; Gd¹¹¹: 7.5% (120mM) Gd³⁺, xllO.

In a separate study the relative amount of resin fully saturated with the respective metal, and suspended in 2% carboxymethylcellulose gel, was varied. In the case of Mn(II) suspensions, the R,-enhancements were linear up to 16% w/v, but thereafter formed a plateau (Figure 2), in contrast to the linear results obtained with soluble paramagnetic ions (see references 7 and 8).

It would be expected that as the concentration of etallated resin increases there would be an increase in the relaxivity values (due to the increased amount of paramagnetic metal ions present). Nevertheless, a point arises when the resin material is present in such quantities that some falls out of suspension. Assuming that the observed relaxivity values are an average of the slower motion of the bulk water and the faster motion of the water near to the paramagnetic centres, the difficulty in maintaining a suspension would mean that the slower motion would begin to dominate, thus causing the plateau effect .

This work demonstrates that concentrated suspensions of this resin, saturated with any of the paramagnetic metals used here, substantially enhance the relaxation properties of aqueous media. The metallated resins appear to be stable indefinitely in water and, at neutral pH, they show no sign of leaching of metal ions. In the presence of 2% carboxymethylcullulose they have the consistency of a thick gel.

Further experiments were carried out to compare, at two different frequencies, the effect on the NMR spin-lattice relaxation rates of water, in 2% carboxymethylcellulose suspensions, of different solid paramagnetic resins which vary in the type of polymer backbone, the type of ligand present and the presence, or absence, or a spacer arm.

The commercial samples used were: (A) Resonium A, a sulphonated polystyrene resin (Winthrop Laboratories; mesh size 85-170um pore size 30 A), (B) AGMP-50, a macroporous sulphonated polystyrene resin (Bio-Rad Laboratories Ltd.; mesh size 200-400um; effective pore size, large), (C) Chelex 100 resin based on a polystyrene lattice with imino-diacetic acid ligands (Bio-Rad Laboratories Ltd.; mesh size 200-400um; effective pore size, large), and (D) Chelating Sepharose (Sepharose is a Trade Mark) resin composed of imino-diacetic acid functional groups coupled by hydrophilic spacer arms to epoxy-activated Sepharose 6! (Pharmacia Ltd.; mesh size 45-160um; exclusion limit 10,000-5,500,000). Each of the samples was immersed in solutions of the appropriate metal salts for 2 days, thoroughly rinsed with an excess of distilled water, vacuum dried and stored in a dessicator. The various extents of metallation were determined by atomic absorption spectroscopy.

All NMR relaxation measurements were made under normal operating conditions at both 26°C with a Varian VXR-300 spectrometer (299.984 MHz; 7 Tesla), and 21°C with an Oxford Instruments horizontal bore magnet linked to an Oxford Research Systems Biospec console (84.851 MHz; 2 Tesla). The inversion recovery sequence with phase cycling.and a composite 180° pulse was used and the R.. values calculated from an exponential fit of the data. In all measurements, a 24% suspension of the fully metallated resin in 2% carboxymethlcellulose gel was used.

All three of the fully saturated metal-bound polystyrene resins are less efficient than when the metal ions are free, but a substantial effect in the relaxation rate of water from its normal value (R₁=0.34s~ ) can still be detected (Figure 3). The polymers saturated with Fe(III) gave the following relaxation enhancements. At 299.984 MHz, Fe³⁺-RESONIUM A: 7.53% Fe³⁺, R_χ xlOO; Fe³⁺-AGMP-50: 5.74% Fe³⁺, R_{L X}28; Fe³⁺-CHELEX 100: 5.08% Fe³⁺, R_χ χ7. Thus, for this series, the overall relaxivity is related directly, albeit not linearly, to the amount of Fe(III) complexed.

In marked contrast, the metal-bound Chelating Sepharose resin in the same form of suspension shows a very much larger increase in the relaxation rate of water from its normal value (R. χl!60), even though the amount of F ( 111) complexed (2.53% Fe ) was less than half that of the least effective of the polystyrene resins. Further, the relaxivity is much larger than even that of free Fe(III) ions in solution (Figure 4). Very similar results were obtained at 84.851 MHz. Furthermore, similar findings were obtained for other metal ions such as Gd(III), although some of the formulations no longer had the consistency of the thick gel.

These results demonstrate that the efficiency with which a solid metal chelate can enhance the spin-lattice relaxation rate of water increases with increase in the hydrophilicity of the polymer surface. This accords with the intuitive expectation that the phenomenon depends on the exchange rates of water through the sphere of influence of the bound metal ions. It is therefore suggested that paramagnetic metals bound to hydrophilic resins are most suitable as contrast agents for magnetic resonance imaging of the gastro-intestinal tract of man.

Further work illustrates that encapsulation, by acid- stable materials, of solid paramagnetic metal complexes based on sulphonated polystyrene ion-exchange resin, prevents the significant demetallation which otherwise occurs when the complexes are subjected to an acidic environment equivalent to that of the human stomach (pH 1-2). In turn, this protects the ability of those complexes to enhance the nuclear relaxation rate of water.

It is common pharmaceutical practice to use encapsula ion, by materials which are resistant to the acid pH of the human stomach (pH 1-2), to protect labile, physiologic lly active substances during their passage through that part of the gastro-intestinal tract, so that they can be released unchanged when they encounter the more basic medium of the small intestine (pH 6-8). The following work demonstrates that encapsulation of metallated Resonium A with acid-stable material prevents the substantial reduction in their relaxation enhancement which otherwise occurs on treatment with aqueous acid.

The commercial sulphonated polystyrene resin, Resonium A (Winthrop Laboratories; mesh size 85-170; pore size 30A), was fully saturated with metal ions according to the technique previously described. Separate portions of these resins were coated with cellulose acetate butyrate (see reference 9) cellulose acetate phthallate (see references 10 and 11), ethyl cellulose (see references 10, 12 and 13), or a variety of Eudragit resin (see references 14 and 15). Each sample was then gently agitated in aqueous hydrochloric acid at pH 1.6 for 2 hours after which the resin was rinsed successively with distilled, deionised water and dried under vacuum overnight.

Suspensions of the metallated resins, 24 w/w %, in 2% carboxymethylcellulose, were taken and proton spin-lattice relaxation measurements of the water were made using the inversion recovery sequence with phase cycling. The R..- values were calculated from an exponential fit of the data. An Oxford Instruments 31cm, horizontal bore magnet operating at 2T and linked to an Oxford Research Systems Biospec console (84.851 MHz) was used with an inductively- coupled, split-ring resonator probe.

Treatment of the uncoated, metallated- resins with acid reduced the relaxation enhancement of the resin by nearly 50%. Results are shown in Figure 5. For example, the R, value of water was increased to 32s~ by the Fe(III)- Resonium A (7^ Fe(III)) prior to acid treatment and to 20s~ after acid treatment. However, this relaxation effect is still much larger than that of water (R. = 0.34s^"1) .

Each of the coated, metallated resins had significantly decreased relaxivity as compared with the uncoated material, the reduction depending on the type of coating. Furthermore, each coating prevented the acidic solution from leaching out any significant proportion of the metal ions. For example, the relaxation rates of water observed in the presence of samples of coated Fe(III)-Resonium A were as follows: (a) cellulose acetate butyrate coating; pre-acid, R_.= 5.5s~ ; post-acid, R,=4.9s~ / (b) cellulose acetate phthallate coating; pre-acid,

; post- acid, R, = 5,8s , (c) ethyl cellulose coating; pre-acid,

R,= 2.9s * post-acid, R,=2.6s~ (d) ethyl cellulose plus plasticiser coating; pre-acid, R = 2.7s —1; post-acid, R, =2.5s —1, and (e) Eudragit L100 resin coating; pre-acid,

R. = 7.4s -1; post-acid, R, = 11.1s-1, although, over a period of a few hours the R, values for this coating increased gradually to a value of about 41.7s in 3 hours, indicating increased passage of water through the coating due to increased porosity. The cellulose acetate and Eudragit resin coatings are soluble in the pH of the human intestine (pH 6-8) and treatment at that pH for 1 hour almost totally restored the relaxivity of the beads to their original value.

Results similar to the above were found for materials saturated with other paramagnetic metal ions.

These results indicate that encapsula ion of a solid paramagnetic metal complex based on an ion-exchange resin protects it from demetallation at pH 1.3, albeit with some reduction of its relaxivity. The fact that full relaxivity can be restored following treatment at the pH of the human small intestine (pH 6-8) confirms that the coatings are removed at that pH. The effects on the relaxivity of the resin beads of adding and removing these coatings is consistent with a model in which the overall enhancement of the relaxation rate of the water depends on the extent with which it makes contact with the solvation shell of the paramagnetic ions. Materials such ds these should thus be suitable as contrast agents for magnetic resonance imaging of the gastro-intestinal tract of man.

Similar results have been obtained by examination of enhancement of spin-spin relaxation rates (R_ values), the R₂ values showing increased relaxation efficiency compared with the R-. values.

References

1. R. L. Void, J. S. Waugh, M. P. Klein, and D. E. Phelps, J. Chem. Phys . , 1968, 48, 3831.

2. M. L. Martin, J. J. Delpeuch, and G. J. Martin, "Practical NMR Spectroscopy", Heyded and Son Ltd. London. 1980.

3. R. Freeman, S. P. Kempsell, and M. H. Levitt, J. Mag. Res. , 1980, 38, 453.

4. M. H. Levitt, and R. Freeman, J. Mag. Res., 1981, 43, 65.

5. R. Tycko, H. M. Choo, E. Schneider, and A Pines, J. Mag. Res., 1985, 61, 90.

6. A. Abragam, "Principles of Nuclear Magnatism", Oxford Univ. Press. Oxford. 1983.

7. V. M. Runge, R. G. Stewart, J. A. Clanton et al., Radiology, 1983, 147,787.

8. V. M. Runge, M. A. Foster, J. A. Clanton et al . , Int. J. Nucl. Med. Biol., 1985, 12, 37.

9. D. L. Gardner, R. D. Falb, B. C. Kim, and D. C. Emmerling, Trans. A er. Soc. Artif. Int. Organs, 1971, 17, 239.

10. H. Dahlstrom, and S. Eriksson, Acta Phar . Suecica, 1971 , 8 ( 5 ) , 505 .

11. C. A. Clark, British patent 1,218,102 (1971).

12. G. P. D'Onofrio, R. C. Oppenheim, and N. E. Bateman, Int. J. Pharm., 1979, 2(2), 91.

13. Y. Raghunathan, L. Amsel, O. Hinsvark, and W. Bryant, J. Pharm. Sci., 1981, 70(4), 379.

14. K. Lehmann, Practical Course in Lacquer Coating. Darmstadt; Rohm Pharma GmbH. 1986.

15. K. Lehmann and D. Dreher, Int. J. Pharm. Tech. +

Proc. Mfr., 1981, 2(4), 31-43

Claims

1. A preparation suitable for clinical administration for use as a contrast agent in magnetic resonance imaging, comprising a porous particulate substance with paramagnetic material bound with respect to the surface thereof, the preparation being such that no significant release of the paramagnetic material occurs in undesired locations in clinical use.

2. A preparation according to claim 1, wherein the porous particulate substance is selected from the following: microporous glasses, zeolites such as zeolite A and zeolite Z, pillared clays, crown ethers such as 18 crown 6 ether, and porphyrins and derivatives thereof.

3. A preparation according to claim 1, wherein the porous particulate substance is selected from the following: serum albumin (human or bovine), polystyrene, polyacrylamides, polycyanoacrylates, poly ethacrylates, starch, cellulose, dextran, alginates and other polysaccharides, polyacrolein, polypeptides, polyorthoesters and hydrogels, Sepharose (agarose) and derivatives thereof.

4. A preparation according to claim 1, wherein the porous particulate substance is a hydrophilic material.

5. A preparation according to claim I, wherein thπ porous particulate substance is a biodegradable material selected from the following: the hydrogels, polylactic acid (polylactide) , polyglycolic acid (polyglycolide) , poly(lactic acid-co-glycolic acid), poly(lactide-cσ- glycolide) or poly ortho-esters, poly acetals, synthetic polypeptides, cross-linked proteins, and polycyanoacrylates.

6. A preparation according to claim 1, wherein the porous particulate substance is of spherical configuration.

7. A preparation according to claim 1, comprising particles having a diameter in the range 0.1 to 500 u .

8. A preparation according to claim 1, wherein the particulate substance has pores with a diameter in the range 30A to a few hundred A.

9. A preparation according to claim 1, wherein the paramagnetic material is selected from the following: ions of elements having an atomic number from 21 to 29, 42, 44 and from 58 to 70 such as chromium (III), manganese (II), iron (III), iron (II), cobalt (II), copper (II), nickel (II), praseodymium (III), neodymium (III), samarium (III) and ytterbium (III), ions having a strong magnetic moment, such as galolinium (III), terbium (III), dysprosium (III), holmium (III) and erbium (III), and nitroxide stable free radicals: pyrrolidone-N-oxyl and piperidine-N-oxyl.

10. A preparation according to claim 1, wherein the paramagnetic material is bound to a ligand attached to the porous particulate substance.

11. A preparation according to claim 10, wherein the ligand is selected from the following: acids containing methylenephosphonic acid groups (-CH-SO^y) , methylenecarbohydroxamic acid groups (-CH-CONHOY) , carboxyethylidene groups (-CH-CH_^COOY) , or carboxy ethylene groups (-CH₂COOY) of which at least 1,2 or 3 are bound to a nitrogen atom supporting the complex formation, eg EDTA, DTPA.

12. A preparation according to claim 10, wherein the ligand is attached to the porous particulate substance via a spacer arm.

13. A preparation according to claim 12, wherein the spacer is selected from the following: hydrophilic groups such as the bis oxirane, 1,4-bis-(2,3-epoxypropoxy)- butane, 6-amino-hexanoic acid and hexa-methylenediamine, and hydrophobic groups such as a pentyl grouping (consisting of 5 methylene groups).

14. A preparation according to claim 1, wherein the porous particulate substance with bound paramagnetic material is provided with an outer coating.

15. A preparation according to claim 14, wherein the coating is selected from the following: cellulose acetate coatings eg phthallate, butyrate, ethyl cellulose, coatings derived from copolymers of methacrylic acid, a ino-ethyl ethacrylates and neutral esters of acrylic and methacrylic acid (the Eudragit resins), and polylysine.

16. A preparation according to claim 14, wherein the coating has a thickness in the range 5 to 15 um.

17. A preparation suitable for clinical administr tion for use as a contrast agent in magnetic resonance imaging, comprising a porous particulate substance with paramagnetic material bound with respect to the surface thereof, and an outer coating which acts to prevent significant release of the paramagnetic material in undesired locations in clinical use.

18. A preparation suitable for clinical administration for use as a contrast agent in magnetic resonance imaging, comprising a biodegradable porous particulate substance with paramagnetic material bound with respect to the surface thereof, the preparation being such that no significant release of the paramagnetic material occurs in undesired locations in clinical use.

19. A method of diagnosis using MRI, wherein a preparation in accordance with claim 1, 17 or 18 is administered to a human or animal body.