Contrast Agents in MRI"

Contrast Agents

in Magnetic Resonance Imaging

Prepared by Andrei Volkov

May 23, 1997

Overview

From the Beginning

Principles of Magnetic Resonance Imaging (MRI)

Comparison of MRI to X-Ray Computer Tomography (CT)

Methods of Image Enhancement

Contrast Agents for MRI

Paramagnetic Contrast Agents

General considerations

Gadolinium(III) complexes

Monocrystalline Iron Oxide Nanocompounds

Metalloporphyrines of Iron(III) and Manganese(III)

Native Proteins Acting as Contrast Agents

Other Contrast Agents

Conclusion

References

Overview

From the Beginning

Noninvasive study of insides of a human body has always

been a challenge to medical doctors, scientists and, later,

designers of commercial devices. Discovery of X-Rays at the

end of nineteenth century has brought about a powerful and,

as it seemed to be at the time, ultimate instrument of such

study. Soon, however, it became clear that X-ray radiography

may hardly be called noninvasive due to destructive effect

on the tissues caused by ionizing radiation. Today's X-Ray

techniques, although much more safe and sophisticated than

before, still employ the same kind of radiation and

constitute the same kind of health risks as years ago.

Discovery of nuclear magnetic resonance in mid-forties

did not seem to be of much importance immediately. Not after

a long time, however, it was realized that it has tremendous

potential for a great variety of applications. Ideas about

its possible use to study of human body were conceived in

sixties, but it was in seventies when the power of computers

allowed these prospects to be fulfilled.

Nuclear magnetic resonance employs low-intensity

radiofrequency electromagnetic waves to study substances

placed in a strong magnetic field. Neither radiation

employed, nor the strong magnetic field have been proven to

be harmful to living organisms in any way so far. Remarkable

results have been achieved using this most noninvasive

technique during the last decade, yet more horizons of its

use still remain unexplored.

Principles of Magnetic Resonance Imaging (MRI)

Much like X-Ray Computer Tomography (CT), MRI is a

method of obtaining images of the body in thin slices 1. It

measures the characteristics of hydrogen nuclei of water and

nuclei with similar chemical shifts, modified by chemical

environment across the slice. This is the major difference

of MRI from proton NMR spectroscopy. The latter measures the

characteristics of any hydrogen nuclei depending on their

position in the molecule. Instead of obtaining information

about chemical shifts and coupling constants, MRI gives

spatial distribution of the intensity of water proton signal

in the volume of the body. This signal intensity depends on

the amount of water in the given place and on the magnetic

relaxation times T1 and T2 which in turn are influenced by a

range of factors 3,4. Deducing these factors in each case is

the ultimate goal of the radiologist trying to diagnose the

problem by looking at the MR image.

Comparison of MRI to X-Ray Computer Tomography (CT)

The main advantage of the MRI is that it is harmless to

the patient. Of course, it would be of little use if MRI did

not offer sufficient diagnostic power. In many cases MRI is

the only way to do unambiguous diagnosis, especially in

detection of cerebral abnormalities, multiple sclerosis and

lesions in sites often obscured by bone artifact on CT. MRI

has inherently superior contrast scale compared to that of

computed tomography 2. Because of that, despite of relatively

high cost of equipment and maintenance, it finds its way to

more hospitals, clinics and research facilities in developed

countries.

Plain MRI also has a few disadvantages when compared to

contrast-enhanced computed tomography. In head examination,

tumors cannot be reliably distinguished from the

surroundings, while in abdominal images it may be difficult

to identify the loops of small bowel making the diagnosis

of lesions uncertain. These shortcomings have led to much

effort to develop ways to improve the MRI image quality.

Methods of Image Enhancement

Both CT and MRI are methods based on image generation

by computers. Therefore, choice of computational method

become important for obtaining quality images. These various

methods constitute the whole separate area and will not be

considered here.

The signal intensity in all X-Ray-based methods

reflects the electron density. Since the early years of X-

Ray radiography a number of substances containing heavy

elements with large number of electrons were employed to

achieve greater contrast of images. In thirties a Nobel

Prize was awarded for the development of contrast-enhanced X-

Ray angiography. The method involved injection of strontium

salts into the neck arteria and immediately taking X-Ray

pictures of the head. In such a way the blood vessels of the

brain become clearly visible on the image.

Since the signal intensity in MRI depends not only on

the amount of water in a given place, but also on the

magnetic relaxation times T1 and T2 , there is more

opportunity to change the picture. The overall quality of

image is strongly dependent on hardware design, especially

on transmitting and receiving coil design, and the pulse

sequence employed to take particular picture 3,4. For

example, blood can appear black, gray and white - depending

on the pulse sequence, velocity of flow and orientation of

flow to the imaging plane 5-7.

It is hard to influence the proton density in the

tissue. Because of that, the most effect on changing

appearance of different tissues on the MR image is achieved

by changing the magnetic relaxation times T1 and T2 of the

protons in tissue-contained water 2.

Therefore, application of contrast substances in MRI

should change tissue characteristics, as compared to

application of radiation-absorbing substances in CT .8

Contrast Agents for MRI

Paramagnetic Contrast Agents

General considerations

Spin-lattice relaxation time T1 and spin-spin

relaxation time T2 may be shortened considerably in presence

of paramagnetic species. The resulting effects can be seen

best in NMR spectroscopy. While shortening of T1 leads to

increase in signal intensity, shortening of T2 produces

broader lines with decreased intensity 9. The net result is a

nonlinear relationship between signal intensity and the

concentration of the contrast agent 10. At low

concentrations, an increase in contrast agent provides an

increase in signal intensity due to effect on T1 until the

optimal concentration is reached. Further increase in

concentration reduces the signal because of effect on T2.

Therefore, in clinical practice it is possible to achieve

less than optimal contrast effect and even produce a

negative contrast effect. This dictates the use of agents

that have a relatively greater effect on T1 than on T2, as

well as the use of pulse sequences that emphasize the

changes in T1.

Paramagnetic species are species which have unpaired

electrons. It may be simple substance (i.e. molecular

oxygen), stable radical (i.e. nitroxide radical) or metal

ion (i.e. many transition metal ions).

Radicals generally cause damage to the living tissues.

Therefore, they are not suitable candidates for medical MRI

purposes. The paramagnetic effect of oxygen, although

demonstrable, seems too weak for practical applications 11.

Paramagnetic metal ions do show suitable effect which

depends on number of unpaired electrons in the ion. The

following table 12 shows some of the paramagnetic metal ions.

There are more transition metals and lanthanide metals

with unpaired spins, but for the metal to be effective as a

relaxation agent the electron spin-relaxation time must

match the Larmor frequency of the protons. This condition is

met better for the Fe3+, Mn2+ and Gd3+ ions (about 10-8 - 10-

10 s, others have 10-11 - 10-12 s) 8.

The main problem with paramagnetic heavy metal ions in

their native form is their toxicity. Investigation has

focused on the development of stable paramagnetic ion

complexes. Both the metal ion and the ligand usually exhibit

substantial toxicity in the unbound state. Together,

however, they may create a thermodynamically and kinetically

stable compound which is much less toxic. Complexation of

the metal ion with organic ligand, while considerably

decreasing toxicity, may alter paramagnetic properties of

the metal.

Chromium-EDTA complex was the first such agent tried,

but problems with synthesis and long-term stability

prevented its clinical application 12.

Gadolinium-DTPA complex, a renally excreted chelate

with a very high formation constant (Table 2) 18, had

sufficiently favorable properties to be approved by Food and

Drug Administration of USA for use in cranial disease

diagnostics in mid-1988.

Another, relatively new type of paramagnetic contrast

agents is so-called superparamagnetic iron oxide (SPIO)

based colloids . They consist of nonstoichiometric

microcrystalline magnetite cores which are coated with

dextranes or siloxanes. Use of these colloids as tissue-

specific contrast agents is now a well-established area of

pharmaceutical development 13-15.

Gadolinium(III) complexes

The prominent feature of Gadolinium(III) is the high

number of unpaired electrons - seven. The Gd3+ ion retains a

number of unpaired spins when bound to the organic ligand.

The free Gd3+ ion is extremely toxic. Number of its

complexes are very stable and thus exhibit much less

toxicity. As mentioned before, Gd-DTPA complex has been

approved to clinical use and is now marketed in USA under

the name "Magnevist".

The relationship between the thermodynamic stability of

the complex and the acute toxicity in vivo seems to be more

complex, as demonstrated by study 16 on that matter. Authors

investigated stability of the series of complexes of several

ligands with Gd3+ versus acute toxicity on mice. They also

attempted to measure the rate of transmetallation of Gd3+ by

Cu2+ and selectivities of different ligands towards Gd3+ as

compared to Zn3+, Cu3+, Ca3+. Iron(III) was not considered

because it is tightly bound in vivo by the storage proteins

ferritin and hemosiderin and is essentially unavailable for

interaction with Gd3+ complexes. The main competitor to Gd3+

was found to be Zn2+, and the most important thermodynamic

criterion of toxicity is the selectivity of the ligand for

Gd3+ over other endogenous metal ions. Zinc

transmetallation was found to be the most likely mode of

both acute and subchronic toxicity in experiments on rats.

Complexes whose structure make in vivo transmetallation

reactions much slower than renal excretion rates have

significantly improved toxicities than would be predicted by

thermodynamics. Slower clearance from the body is likely to

significantly increase the toxicity of any Gd3+ complex.

The following Figure 2 16 shows other ligands often used in

gadolinium studies.

Gd3+ has been shown to inhibit Ca2+ binding to

mammalian cardiac sarcoplasmic reticulum. The mechanism of

toxicity could involve hemodynamic disruption.

An example 19 of two another perspective ligands is

shown on Figure 3.

Potential improvement also lies in the idea of

covalently coupling the ligand to protein to generate tissue-

specific contrast agents 20.

Understandably, search for new potential ligands for

Gd3+ complexation is still hot area of investigation. The

complexes are evaluated against Gd-DTPA, the only approved

compound for human use so far. Criteria include

thermodynamic stability, rates of excretion, toxicity,

lipophilicity, biodistribution, percent change in MR signal

intensity. Some of the complexes are slightly better than Gd-

DTPA in particular tests, others are a little worse.

However, neither of them got as much attention as Gd-DTPA

itself 21. This complex is far not be the best choice as of

today, but it is relatively well-known choice, widely used

in many MRI facilities on a daily basis. Accumulated

experience allows in many cases offset the shortcomings of

Gd-DTPA, so the agent continues to play indispensable role

in modern MR imaging.

Monocrystalline Iron Oxide Nanocompounds

Compounds called MION constitute relatively new but

rapidly evolving area in MRI contrast agents. As compared to

the single approved gadolinium-containing complex, there are

variety of MION (also called SPIO - SuperParamagnetic Iron

Oxide) reagents available on the market. Flashy names

Feridex I.V�., Endorem�, Gastromark�, Lumirem�, Sinerem�

and more patents pending tell us that the last word in the

area is yet to be said.

These compounds consist of nonstoichiometric

microcrystalline magnetite cores which are coated with

dextranes (in ferumoxides) or siloxanes (in ferumoxils) 22.

SPIO agents are much more effective in MR relaxation than

their paramagnetic counterparts. Since they are

nonstoichiometric, there was and still is much interest in

studying these compounds with all the vast array of modern

physical-chemical methods: single crystal X-ray diffraction,

powder X-ray diffraction, Mossbauer spectroscopy,

transmission electron microscopy, dynamic light scattering,

atomic adsorption spectroscopy, spectrophotometry, electron

microscopy, superconducting quantum interference devices

etc 22,23-25.

The compositions and physiochemical properties of

nonstoichiometric magnetites are continuously variable

between those of Fe₃O₄ and Fe₂O₃. Conceptually, these

cation-deficient, inverse-spinel phases are formed by

partial oxidation of Fe(II) in stoichiometric

magnetite24.The Fe2+ content is typically 8-15 mol%. The

lattice parameters of these colloids also fall between those

of Fe₃O₄ and Fe₂O₃.

The particles are usually of varying sizes from several

to several hundred nanometers. They are irregular in shape

and highly light-absorbing. They have no magnetic hysteresis

at ambient temperatures, which is characteristic of

superparamagnetic materials. Mossbauer spectra are

characteristic of small (<10nm) magnetic domains which

undergo superparamagnetic relaxation or collective magnetic

excitation on the Mossbauer time scale 23. There is no

evidence of covalent bonding between the iron surface and

the surrounding dextran 25.

SPIO compounds are promising contrast agents since

their properties may be fine-tuned for the specific

application. They are non-toxic and rapidly cleared from the

organism. Experiments have been successful in receptor-

specific SPIO delivery 26.

Metalloporphyrines of Iron(III) and Manganese(III)

Porphyrins have been known for decades as indicators of

various metabolic disorders and disease states 27,28. They

are used in photodynamic therapy of tumors 29. Low toxicity

of metalloporphyrins and their selective retention in tumors

have led recently to their study as a MRI contrast media 30-35.

Article 34 contains theoretical treatment of relaxivity

of some of the metalloporphyrins. This is a fairly new

area of development, but metalloporphyrins of Mn(III) and

Fe(III) show favorable properties as MRI contrast agents for

tumor detection. No doubt, it will lead to new discoveries

soon.

Native Proteins Acting as Contrast Agents

Heme-containing proteins may act as "natural" contrast

agents, just like previously discussed iron(III) porphyrins.

Hematomas are easily identified by MRI. Paramagnetic

deoxyhemoglobin within intact blood cells causes a local

region of high magnetic field 36 This results in rapid

dephasing of water protons diffusing in the region of acute

hematoma with shortening of T2, which results in low signal.

A similar phenomenon is observed when magnetically

susceptible ferritin is deposited in macrophages in

hemochromatosis.

Although these not a specifically designed contrast

agents, their properties may be successfully used for

diagnostics in specific cases.

Gastrointestinal Contrast Agents

The gastrointestinal tract cannot be reliably studied

by MRI without the use of contrast agents. Oral contrast

agents may dramatically improve utility of MRI for gastro-

diagnostics. The only clinically approved agents for that

purpose are soluble iron compounds. (ferrous gluconate,

ferric ammonium citrate) and Gd-DTPA. There is a problem

with dosage of iron salts, which may not exceed the levels

above those when iron supplementation is used. There are no

particulate agents approved for oral use yet.

Other Contrast Agents

Examples of contrast agents other than paramagnetic

involve rather mechanical action than changing the

characteristics of the surrounding tissue. They are thus

similar to agents used in X-Ray methods. One recent example

successfully utilized vegetable oil for rectal MRI

applications 37. Other possibilities include insufflation of

air to achieve better contrast of intestinal walls 38.

Conclusion

Careful and exacting clinical trials are necessary to

determine suitability of every potential contrast agents for

MR imaging. It is the matter of fact that contrast

enhancement is already playing a significant role in the

clinical use of MRI and this role seem to become more

important as new techniques and applications are being

developed.

References

1. Wong W.S; Tsukuda J.S; Kortman K.E; Bradley W.G.

Practical Magnetic Resonance Imaging. A Case Study Approach,

Aspen Publishers Inc., USA, 1987, pp 1.

2. Magnetic Resonance Imaging and Spectroscopy, Editors:

Mettler F.A.; Muroff L.R.; Kulkarm M.V., Churchill Livingstone, New York, 1986, pp 231.

3. Bradley W.G. Noninvasive Medical Imaging 1984, 1, pp

193-204.

4. Advanced Imaging Techniques, Editors: Wehrli F.W.;

MacFall J.; Newton T.H; Potts D.G., Clavadel Press, San Francisco, 1983, vol.2, pp 81-118.

5. Bradley W.G.; Waluch V. Radiology 1985, 154, 443-450.

6. Waluch V.; Bradley W.G. J. Comput. Assist. Tomogr.

1984, 8, 594-598.

7. Bradley W.G.; Waluch V. AJR 1984, 143, 1167-1174.

8. Magnetic Resonance Imaging. Theory and Practice.

Vlaardingerbroek M.T.; Boer, den J.A., Springer Verlag, Germany, 1996, pp 242-243.

9. Basic One- and Two-Dimensional NMR spectroscopy.

Friebolin H., VCH, Germany, 1993.

10. Runge V.M.; Clanton J.A.; Herzer W.A. et al. Radiology

1984, 153, 171.

11. Mendonca-Dias M.H.; Gaggelli E.; Lauterbur P.C.

Semin.Nucl.Med. 1983.

12. Runge V.M.; Clanton J.A.; Lukehart C.M. et al. AJR,

1983, 141, 1209.

13. Ferrucci J.T.; Stark D.D. AJR, 1990, 155, 943-950.

14. Runge V.M.; Rijcken T.H.P.; Davidoff A.; Wells J.V.;

Stark D.D. JMRI, 1994, 4, pp 281-289.

15. Hahn P.F. AJR, 1991, 156, 252-254.

16. Cacheris W.; Quay S.; Rocklage S. MRI, 1990, 8, 467-481

17. Courtesy of Shering AG, West Berlin. In Ref. 2., page 236.

18. Krasnow N. Biochem.Biophys.Acta 1972, 282, 187-194.

19. Wiegers C.B. et al. MRI, 1992, 10, 903-911

20. Spanoghe M. et al. MRI, 1992, 10, 913-917

21. Carollo B. et al. MRI, 1990, 8, 381-393

22. Shen T. et al. MRM, 1993, 29, 599-604

23. Jung C.W.; Jacobs P. MRI, 1995, 13, 661-674

24. Murad E.; Johnson J.H. In: Long G.L. (Ed.) Mossbauer

Spectroscopy Applied to Inorganic Chemistry, New York, Plenum Press, 1984,

Chapter 12

25. Jung C.W. MRI, 1995, 13, 675-691

26. Bader A. et al. MRI, 1995, 13, 991-1002

27. Policard A. Comp.Rend.Soc.Biol. 1924, 91, 1423-1424

28. Figge F. et al. Proc.Soc.Exp.Biol.Med. 1948, 68, 640

29. Berns M.W. Hematoporphyrin Derivative Photoradiation

Therapy of Cancer, New York, Alan R. Liss, 1984

30. Chen C. et al. FEBS Lett. 1984, 168, 70-74

31. Cohen J.C. et al. US Patent No 4,986,256

32. Patronas N.J. et al. Cancer Treat.Reports 1986, 70, 391-395

33. Place D.A. et al. MRI, 1992, 10, 919-928

34. Mercier G.A. Jr. MRI, 1995, 13, 807-817

35. Saini S.K. MRI, 1995, 13, 985-990

36. Gomori J.M. et al. Radiology 1985, 157, 87-93

37. Pokieser P. et al. MRI, 1995, 13, 979-984

38. Butch R. et al AJR, 1986, 146, 1155-1160