A Primer on Parenteral Radiopaque Contrast Agents

THE UNIVERSITY OF NEW MEXICO HEALTH SCIENCES CENTER COLLEGE OF PHARMACY ALBUQUERQUE, NEW MEXICO

Correspondence Continuing Education Courses For Nuclear Pharmacists and Nuclear Medicine Professionals

VOLUME X, NUMBER 2

Peter Dawson, BSc, PhD, CPhys, MBBS, FRCP, FRCR University College London Hospitals London, United Kingdom

The University of New Mexico Health Sciences Center College of Pharmacy is accredited by the American Council on Pharmaceutical Education as a provider of continuing pharmaceutical education. Program No. 039-000-02-005-H04. 3.0 Contact Hours or .30 CEUs

A Primer on Parenteral Radiopaque Contrast Agents By:

Peter Dawson, BSc, PhD, CPhys, MBBS, FRCP, FRCR University College London Hospitals London, United Kingdom

Coordinating Editor and Director of Pharmacy Continuing Education William B. Hladik III, MS, RPh, FASHP, FAPhA College of Pharmacy University of New Mexico Health Sciences Center

Managing Editor Julliana Newman, ELS Wellman Publishing, Inc. Albuquerque, New Mexico

Editorial Board George H. Hinkle, MS, RPh, BCNP, FASHP, FAPhA Jeffrey P. Norenberg, MS, PharmD, BCNP, FASHP Neil A. Petry, MS, RPh, BCNP, FAPhA Laura L. Boles Ponto, PhD, RPh Timothy M. Quinton, PharmD, MS, RPh, BCNP, FAPhA

Guest Reviewer Robert L. Siegle, MD Professor and Chairman Department of Radiologic Sciences Drexel University College of Medicine 245 North 15th Street, MailStop 206 Philadelphia, PA 19102-1192

While the advice and information in this publication are believed to be true and accurate at press time, the author(s), editors, or the publisher cannot accept any legal responsibility for any errors or omissions that may be made. The publisher makes no war- ranty, express or implied, with respect to the material contained herein.

STATEMENT OF OBJECTIVES

The purpose of this continuing education lesson is to increase the reader’s knowledge and understanding of the clinical applications of parenteral radiopaque contrast agents.

Upon completion of this lesson, the reader should be able to:

1. Discuss in broad terms the clinical applications of contrast agents.

2. Explain the significance of iodine, its use in radiopaque contrast agents and the variations

of carrier molecules used to deliver it.

3. Explain why relatively low osmolality and low chemotoxicity are desirable characteristics

of contrast agents and describe how the structure of the contrast molecule influences these

two characteristics.

4. Identify the risk factors for major reactions to contrast media and critically discuss

pre-treatment protocols that minimize the probability of occurrence of such reactions.

5. Discuss the clinical manifestations and incidence of side effects and adverse reactions

known to be associated with contrast media and identify their important causes and

mediating factors.

6. Describe the manifestations and treatment of anaphylactoid reactions.

7. Discuss, in descriptive terms, contrast agent pharmacokinetics. COURSE OUTLINE VII. ANAPHYLACTOID/ IDIOSYNCRATIC I. INTRODUCTION REACTIONS A. The Term "Contrast" A. Prophylaxis of B. Historical Development Anaphylactoid Reactions B. Treatment of II. DISADVANTAGES OF Anaphylactoid Reactions CONVENTIONAL IONIC CONTRAST AGENTS VIII. A NOTE ON THE CLINICAL USE OF CONTRAST AGENTS III. LOW-OSMOLAR AGENTS A. Non-Ionic Agents IX. A BRIEF NOTE ON B. Monoacid Dimeric Agents COAGULATION AND C. Non-Ionic Dimeric PLATELETS Contrast Agents X. THE METFORMIN ISSUE IV. CONTRAST AGENT TOXICITY – THE CONCEPT XI. CONCLUSIONS OF MOLECULAR TOXICITY XII. REFERENCES V. PHARMACOKINETICS XIII. QUESTIONS VI. OSMOLALITY, MOLECULAR TOXICITY, ORGAN SPECIFIC TOXICITIES AND HIGH DOSE TOXICITY OF CONTRAST AGENTS

1 A PRIMER ON PARENTERAL an image is not the norm and usually contrast RADIOPAQUE CONTRAST AGENTS agents may be relied upon to improve the images and their information content by By: enhancement of relevant image contrast. Peter Dawson, BSc, PhD, Sometimes reference is made to "nega- CPhys, MBBS, FRCP, FRCR tive" and "positive" contrast agents. This fre- University College London Hospitals quently causes confusion. As far as x-ray London, United Kingdom contrast agents are concerned "positive" agents are those that increase the absorption of x-rays while "negative" agents decrease x- INTRODUCTION ray absorption. The iodinated agents are pos- The Term "Contrast" itive contrast agents because iodine atoms, The word "contrast" should be explained which are incorporated into their chemical at the outset so that its use in the terms "con- structures, are efficient absorbers of x-rays in trast agent" and "contrast medium" is clearly the diagnostic energy range. (So, incidental- understood. The term contrast is a photo- ly, are the barium atoms in contrast agents graphic or, more generally, a medical imag- used for studies of the gut.) Gases, such as ing term. Contrast in an image of any kind is air or CO2, on the other hand, are negative what allows detail to be discerned between agents as they do not significantly absorb x- two or more adjacent regions of the image. rays and instead allow them to pass more Black chalk on a blackboard cannot be seen readily. Thus, these gases may be very useful because there is no contrast, however, white as well, for example, in providing the con- chalk on the same blackboard produces high trast necessary to differentiate the gastroin- contrast that is readily apparent to the human testinal tract from other adjacent anatomical eye. Fortunately, there is some natural con- structures. trast in medical images of all kinds. If there were no contrast, medical images would be Historical Development of little value. Thus in a computed tomogra- Without the availability of good and rel- phy (CT) image (which has far better resolu- atively safe intravascular contrast agents, tion and display of contrast than any ordi- diagnostic radiology would not have come to nary x-ray film) darker blood vessels may be occupy the central position it currently does seen crossing a lighter liver. A tumor in the in the medical diagnostic process. The great liver will frequently, but not reliably, be bulk of "interventional radiology" proce- darker than normal liver and so will be read- dures would certainly not have emerged ily apparent because of the natural contrast without the development of the contrast associated with this imaging technology. The agent industry. The fact that most radiolo- purpose of a contrast-enhancing agent ("con- gists often take the contribution of the contrast agent," "contrast medium") is, as the trast agent largely for granted may be seen as name suggests, to enhance such natural con- a tribute to the excellent qualities associated trast and render tissues of interest – usually with the current generation of these. those with pathology – more clearly visible. The need for intravascular contrast Now it must be said immediately that con- agents was realized very early in the history trast agents do not always succeed optimally of radiology, as witnessed by the perform- in this aim. Sometimes a tumor, to take the ance of angiograms on an amputated hand same example, may actually become less and on a cadaver kidney by Haschek and clearly visible after administration of a con- Lindenthal1 and Hicks and Addison,2 respec- trast agent. Clearly, much depends on the tively, both within a month of Roentgen’s blood supply and other structural character- report of the discovery of x-rays. The radio- istics of normal tissue and tumor. However, opaque suspensions used for these classic such worsening of the observed contrast in studies were not clinically practicable but the 2 achievements were remarkable. In subse- tional cell carcinoma,8 while local injections quent years, various agents were used to into breast and maxillary sinuses have been opacify the gastrointestinal tract, the bladder, followed by carcinoma developing at these ureters, renal pelvis and lymphatics by direct sites.9 Perivascular extravasation of contrast injection, but little progress was made with is associated with local granuloma forma- the development of agents which could be tion. safely injected into the circulation. While inorganic iodide solutions were Berberich and Hirsch3 used a strontium too toxic for general use, it was realized that bromide solution for the opacification of the elemental iodine is certainly an outstanding- peripheral veins and, in the same year, ly good choice for x-ray absorption. The K- Osborne, et al4 at the Mayo Clinic noted shell electron binding energy is approxi- opacification of the bladder in patients treat- mately 33 KeV, which is lower than, but ed with intravenous (IV) solutions of sodium close to, the mean energy used in diagnostic iodide for syphilis. They had unwittingly x-ray work, thereby maximizing the cross- performed the first IV urogram. However, section for photoelectric interactions. All the delineation of the upper urinary tract was other considerations such as chemistry aside, poor and sodium iodide solutions were rather iodine is an ideal choice of element for use in toxic for general use for urography. Enough an x-ray contrast agent at the energies typi- interest was generated, however, for Brooks5 cally used clinically. Barium, incidentally, to try sodium iodide solutions for peripheral used in studies of the gastrointestinal tract is angiography and in 1927 for Moniz to apply similarly an ideal choice in this same respect, them to carotid arteriography.6 Both proce- having a K-shell binding energy of ~ 35 dures had to be performed under general KeV. anesthesia and Moniz was so concerned If iodine was to be used, a suitable carri- about the toxicity of sodium iodide that he er molecule was clearly going to be neces- diverted his activity into one of the great sary. The first such emerged in 1925 when dead ends or "cul-de-sacs" of radiological Binz and Rath10 synthesized a large number history, namely the use of colloidal thorium of compounds, some of them pyridone deriv- dioxide ("Thorotrast") as a contrast agent.6 ative containing iodine, in the quest for a Thorium dioxide had the great advantage of being highly radio-opaque at a Figure 1. The contrast agents “Selectan neutral” time when radiographic film and equip- and “Uroselectan.” The first generation of ment produced relatively little image pyridone-based agents. contrast, and the agent had virtually no known acute side effects upon injection. Unfortunately, thorium is radioac- tive, being an emitter of α-particles and having a half-life of 1.4 x 1010 years. It is permanently retained within the reticuloendothelial system, largely in the liver and spleen, and is associated with treatment of syphilis. Two compounds in this induction of malignancies, the first of which series became known as "Selectan neutral" to be recorded in a human was a sarcoma of and "Uroselectan" as they were partially the liver some 20 years following its admin- selectively excreted in urine. These were istration.7 Other tumors described, and appar- used by a young American, Moses Swick, ently related to the agent, include hepatoma, working under the supervision of, first, cholangiocarcinoma and myeloid leukemia. Lichtwitz and, subsequently, von Retention of Thorotrast by urothelium fol- Lichtenberg in Berlin, for IV urography in lowing retrograde pyelography has also been the years 1927 to 1929.10 These mono-iodi- associated with the development of transi- nated pyridone compounds (Figure 1) were 3 superseded within 3 years or so by di-iodi- as the iothalamates and metrizoates, fol- nated pyridones of higher solubility, which lowed. As can be seen from Figure 3, the var- were also synthesized by Binz and Rath ious representatives of this type of contrast (Figure 2). These became the standard con- agent are all salts of the benzoic acids and trast agents for intravascular and other appli- differ only in minor ways in the substituent cations during the next two decades or more, side-chains at the 3 and 5 positions on the until they were replaced in the early 1950s benzene ring. The various competing com- by the benzene ring based compounds. In mercial formulations differ in being sodium fact, as early as 1933 Swick had suggested salts, meglumine (methylglucamine) salts or the use of iodinated benzene ring compounds mixed sodium and meglumine salts, and in the concentrations available. For exam- Figure 2. The contrast agents ‘Uroselectan B’ and ple, the iothalamate series is known “Diodone’. The second generation of pyridine- commercially as the Conray range. based agents. Conray 280 is a pure meglumine salt at a concentration of 280 mg iodine/mL solution, whereas Conray 420 is a pure sodium salt at a concentration of 420 mg iodine/mL. The iothalamates are isomers of the diatrizoates with the NHCOCH3 substituent group at position 5 merely changed to CONHCH3. The diatrizoates are known as "Hypaque" when marketed by and had developed monoiodo-hippurate for Amersham and as "Renografin" when mar- this purpose. Unfortunately, the very addi- keted by Bracco. Urografin 370, for exam- tion of the iodine to the hippurate substan- ple, is a mixed sodium and meglumine salt at tially increased the toxicity of the compound, a concentration of 370 mg iodine/mL solu- a point developed below. The compound tion. Hypaque 270 is a pure sodium salt at a was, therefore, unsuccessful in its original concentration of 270 mg iodine/mL solution. intention but became an important tool in The important points to consider in the renal physiology and opened the way for the choice of agent for a particular application eventual development of successful contrast are (i) the iodine concentration of the formu- agents of this type in the 1950s. lation, and (ii) the sodium and/or meglumine content. Sodium salts are Figure 3. The structures of benzene ring-based contrast agents. more toxic to vascular endothelium and to blood- R2 R3 Proper name Commercial name brain barrier and neural tis- H NHCOCH3 Acetrizoate Urokon, Diaginol sues and so should be CH3CONH NHCOCH3 Diatrizoate Urografin, Hypaque avoided in venography and CH3CONH CONHCH3 Iothalamate Conray cerebral angiography. CH3CONH NCOCH3 Metrizoate Isopaque, Triosil Meglumine salts should be CH3 used for these purposes. For cardiac work, a mix- In 1952 Wallingford discovered that the ture of sodium and meglumine should be introduction into the benzene ring of an chosen (for example, Urografin 370), since amide side-chain with acetylation produced both pure sodium and pure meglumine salts a low toxicity compound, acetrizoic acid are more cardiotoxic and are associated with (Figure 3).11 Hoppe subsequently developed high levels of ventricular fibrillation. Some a doubly substituted molecule diatrozic manufacturers have special formulations of acid12 and closely related compounds, such their agents, with calcium and magnesium 4 Table 1. Iodinated Contrast Agents Available Conventional (High Osmolality) Ionic Agents Sodium/Meglumine Diatrizoate (Renografin), Bracco Inc, Princeton NJ, USA Sodium/Meglumine Metrizoate (Hypaque), Amersham plc, UK Low Osmolality Non-Ionic Monomeric Agents Iopamidol (Isovue) Bracco Inc, Princeton, NJ, USA Iopromide (Ultravist) Berlex Inc, USA Iohexol (Omnipaque) Amersham plc, UK Ioversol (Optiray) Tyco Inc, St Louis, USA Low Osmolality Ionic Agents Sodium Meglumine Ioxaglate (Hexabrix), Guerbet, France Iso-osmolar Non-Ionic Dimeric Agents Iotrolan (Isovist) Berlex Inc, USA Iodixanol (Visipaque) Amersham plc, UK ions partially replacing sodium ions, in an repeatedly under fluoroscopy and must, attempt to reduce various aspects of toxicity. additionally, take pictures, very large total These ionic agents are sometimes referred to doses may be given–up to 1000 mL of 300 as "conventional" agents. Table 1 summa- mg iodine/mL strength. Of course, such rizes the proper names, commercial names doses are administered over a prolonged and ionic content of a number of these agents period of perhaps 2 hours or more. It would still available in the USA. not be appropriate to give recipes for all con- Agents of this kind were used almost trast-enhanced radiological procedures, not universally in diagnostic radiology for 30 only because of their number but because years or so, from their introduction in the different radiologists tend often to do things early 1950s until the introduction of the sec- differently (see Section VIII below). A tissue ond-generation “non-ionic” agents (see later) or blood concentration of at least 20 mg in the early 1980s. Even today they are occa- iodine/mL is essential to obtain sufficient sionally used but their only advantage over enhancement effect. the non-ionic agents is lower price. They were used for all intravascular applications, DISADVANTAGES OF CONVENTIONAL arteriographies and venographies, and for IONIC CONTRAST AGENTS body cavity studies such as cystography, An ideal contrast agent would provide a arthrography, sinography and fistulography. passive increase in x-ray absorption while They were not without their significant having absolutely no effect on any body sys- undesirable side effects (see next section). tem or function, biochemical or physiologi- For myelography they were too dangerous, cal. The early agents and even those used up causing severe irritation of neural tissues and to the 1980s were far from ideal as these seizures. Iodine concentrations in commer- agents were often associated with both sub- cial formulations range up to 370 mg jective and objective side effects of many iodine/mL and volumes used in various kinds. applications range from a few mL to 150 mL. At the typical iodine concentrations used The latter volume might be used to enhance in many applications, 300-400 mg a CT scan, for example. In complex angio- iodine/mL, the ionic contrast agents are sub- graphic procedures where the operator must stantially hyperosmolar with respect to plas- guide himself by injecting small volumes ma–up to seven or eight times in some 5 cases.13,14 This hyperosmolality is associated neuronal irritation is the seat of all idiosyn- with a number of adverse effects, both sub- cratic anaphylactoid adverse reactions (see jective and objective. Thus, when absolutely below) to contrast agents. or relatively high doses of the contrast medi- Contrast agents affect not only endothe- um are injected into the circulation, as in lial cells, but also erythrocytes, basophils many interventional studies or in cardiac and mast cells. Water is drawn out of ery- studies in infants and small children, the throcytes by hyperosmolar contrast agents, osmotic load, may be a threat to the patient. so that they become shrunken and deformed Acute pulmonary edema is a risk. In fact, the and thus lose the flexibility essential for plasma volume expands but this is partly them to negotiate the capillary circulation.19 compensated by the shrinkage of circulating The "sludging" of rigidified erythrocytes in cells. The effects of the hyperosmolar solu- the microcirculation of the lungs in pul- tions on smooth muscle produce a general- monary angiography has been related to the ized peripheral vasodilation with a some- (usually) transient pulmonary hypertension times profound fall in blood pressures fol- associated with pulmonary angiography. lowed by a reflex tachycardia, events which Additionally, in selective renal angiography are poorly tolerated by some patients.15 The this sludging effect has also been cited as a subjective counterparts of this are the feeling factor in contrast agent-associated nephro- of flushing and faintness experienced by toxicity. many patients. Other unpleasant and subjec- Basophils and mast cells are induced to tive side effects, such as sensation of sponta- release histamine, which may contribute to neous bladder emptying and metallic taste, the subjective side effects experienced by often cited by patients as most unpleasant, patients and perhaps, although not certainly, must be added.13 Pain on arterial injection to anaphylactoid reactions.20 may be severe. Clinical bronchospasm is In addition to the effects on the peripher- occasionally seen and may be severe (see al vasculature and on the expansion of the anaphylactoid reactions below). Nausea is blood volume, the hyperosmolality of con- not uncommon; frank vomiting is less com- trast agents is an important component of mon but very important both for the patient their cardiotoxicity (see below).21 and the technologists because it disrupts the Then there are a number of organ-specif- examination. ic toxic effects, which will be discussed At the microscopic level there may be below in Section VI. It may be wise to enter injury to endothelial cells which is largely, a caveat even at this stage. Osmolality is but not entirely, mediated by hyperosmolali- very important and in the next section we ty16 and which on venous injection may lead will discuss how it was reduced in newer to thrombophlebitis or frank venous throm- contrast agent formulations. However, there bosis.17 In arterial injections hyperosmolality are other factors contributing to the observed may be a factor in the post-angiographic pro- toxicity associated with these agents that will gression of stenosis to occlusion and to the be discussed later. Meanwhile the reader failure, by early closure, of angioplasties. should not take away an impression that high This injurious effect on endothelium is of osmolality is the only issue or the only cause particular significance in relation to the of the objective and subjective effects noted blood-brain barrier. There is considerable above. experimental work demonstrating the increased permeability of the blood-brain LOW OSMOLAR AGENTS barrier and the transfer of contrast agent into The various important subjective and the brain and into direct contact with neu- objective side effects of high-osmolar agents rons, which are very sensitive to contrast provided a powerful stimulus to formulate agents.18 Indeed, one hypothesis holds that low-osmolar contrast agents. The obvious blood-brain barrier injury and subsequent way to reduce the osmolality of solutions is 6 to dilute them, but this would also dilute the the fact that Nyegaard was already a produc- iodine concentration. Another approach was er of metrizoates (Isopaque) contrast agents. needed. The loss of solubility, inevitable with the elimination of the carboxyl group, was over- Non-Ionic Agents come by substituting a D-glucose group, Almen first suggested an approach to which provided many hydrophilic hydroxyl reducing the osmolality of contrast agents groups (Figure 4). Metrizamide had the seri- without diluting their iodine concentration.22 ous disadvantage of being unstable at high He pointed out that the osmolality of any temperatures, making its sterilization by solution is proportional to the number of par- autoclaving impossible, requiring the agent ticles in that solution and that, by definition, to be produced as a lyophilized powder. Its ionic agents dissociate into two particles, instability in solution, even at room temper- anion and cation, and thereby, double their ature, made it impossible to formulate ready- osmolality in solution. A non-ionic, non-dis- to-use solutions and the agent had to be sociating agent theoretically should have made up at the time of use from a powder half the osmolality at any iodine concentra- solute and a sterile solvent. It was therefore tion of an ionic agent. The first practical both expensive and inconvenient, and was result of this concept was metrizamide not much used in intravascular applications. (Amipaque), produced by Nyegaard (subse- However, its low neurotoxicity, compared to quently Nycomed, then taken over by the ionic water-soluble agents, led to a revo- Amersham) in Oslo. This was a 1-substituted lution in myelography. Although low-osmo- amide of metrizoic acid, a choice dictated by lar, non-ionic agents were otherwise little used in clinical practice, enough experimen- Figure 4. Metrizamide (Amipaque). tal and clinical work was performed to estab- The first non-ionic contrast agent. lish their considerable advantages in terms of

Figure 6. Iopamidol (Niopam). Second generation non-ionic agent.

Figure 5. Metrizamide (Omnipaque). Iopromide (Ultravist). Second Second generation non-ionic agent. Figure 7. generation non-ionic agent.

7 reduced toxicity in a variety of applications, role in any toxic effects. Its role is to chelate including coronary and carotid arteriography any traces of heavy metal contamination, and peripheral venography. This stimulated which would promote degradation of the the production of a second generation of benzene rings with liberation of iodide. non-ionic contrast media in ready-to-use formulations, of which there are now a number Monoacid Dimeric Agents of representatives, including iohexol If a linking bridge joins two tri-iodinated (Omnipaque, Amersham) (Figure 5), iopami- benzoic acid groups, the result is a dimeric dol (Niopam, Bracco) (Figure 6) and iopro- acid. Thus, iocarmic acid, a meglumine salt mide (Ultravist, Schering) (Figure 7). These of a dimer produced by linking two iothala- second-generation compounds have the con- mic acids, was introduced as "Dimer X" in venience of being in ready-to-use solutions 1970 and used, in spite of the fact that it was and are considerably less expensive than ionic and relatively neurotoxic, as a myelo- metrizamide, although still significantly graphic agent. The ratio of number of iodine more expensive than the conventional ionic atoms to particles in solution from each mol- agents. They are all associated with a ecule in this compound was 6:3, an improve- markedly reduced incidence and severity of ment on the 3:2 ratio associated with the all the objective and subjective effects dis- conventional ionic monomeric contrast cussed above. Cost is the only factor that has agents. Subsequently, one of the carboxyl militated against replacing the conventional groups of such a dicarboxylic acid was sub- ionic agents with these second-generation stituted with a non-ionizing group to produce non-ionic compounds completely for all a monoacid dimer. This is ioxaglic acid applications.23, 24 (Figure 8), which is produced in a commer- A useful way to express the merits of any cial formulation as a mixture of sodium and contrast agent, in terms of contrast-providing meglumine salts of ioxaglic acid, known as potential, relative to the disadvantages in Hexabrix. This has a high iodine to particle terms of osmolality, is the ratio of iodine ratio of 6:2 (=3:1) and, while remaining an atoms provided per molecule (three in all ionic agent, this compound succeeds, as do cases so far considered) to the number of the non-ionic monomeric agents, in reducing particles derived from a molecular unit (two osmolality to approximately half that of con- in the case of the ionic agents and one, the ventional ionic agents at any iodine concen- whole molecule, in the case of the non-ionic tration. The osmolalities for several contrast agents). Thus, the ionic agents are some- agents, plotted as a function of iodine con- times referred to as 3:2 or ratio 1.5 agents centration, are shown in Figure 9. Sodium and the non-ionic variety as 3:1 or ratio 3 iothalamate represents the conventional agents. The higher the ratio the better, and ionic agents in this figure. As can be seen, 3:1 is clearly better than 3:2. the reduction in osmolality, at any given The non-ionic agents are formulated with iodine concentration, as compared with the a trace of calcium EDTA, which plays no conventional agent, is somewhat more than

Figure 8. Sodium meglumine ioxaglate (Hexabrix). A monoacid dimeric low-osmolality agent.

8 Figure 9. The osmolalities of some contrast agents as a function Non-Ionic Dimeric of iodine concentration at 37˚C. Contrast Agents One obvious next step in contrast agent development, which has been taken both by Schering and Nycomed, is to combine the dimeric approach of Hexabrix with the non- ionic concept.25 Simply taking a dimeric compound of the type discussed above and replacing both carboxyl groups with non-ionizing groups can do this. To maintain solubility, it is now essential to substitute around the molecule a large number of hydrophilic groups. The two commercial compounds of this type now the predicted 50% for all of the agents. This available are illustrated in Figure 10. Iotrolan is partly the result of the large size of all (Isovist, Schering) is available for myelo- these molecules and partly due to a degree of graphic and intravascular use and iodixanol molecular aggregation in solution. The dif- (Visipaque, Nycomed) is available for ferences between the osmolalities of the var- intravascular use. ious new agents are real but of little clinical The osmolalities of such non-ionic significance. The fact that Hexabrix has a dimers are lower than that of plasma (hypo- slightly lower osmolality at any iodine con- osmolar) at all iodine concentrations up to centration than any of the other agents might about 400 mg iodine/mL. This is the result of suggest clinical advantages, even if small. both a large molecular size and molecular Although Hexabrix is an acceptable arterio- aggregation producing a very high effective graphic agent associated with reduced pain molecular weight and hence a lower than and discomfort, on IV injection it displays predicted osmolality. This problem of hypo- significantly greater toxicity, particularly in osmolality is a new concern for the contrast causing nausea and vomiting, than do the agent industry but is easily overcome. The non-ionic agents. Therefore, it should not be addition of a little saline allows the osmolal- used for enhancement of computed tomogra- ity to be adjusted upwards to become identi- phy (CT). This fact is, at first, surprising and cal to that of plasma at any iodine concentra- the important reasons for this recommenda- tion and this is done in the commercial for- tion are discussed below. It must not be used mulations. The osmolality factor is thereby for myelography. completely eliminated. It should be noted that, while the non- Such agents are more viscous than other ionic monomeric agents and ioxaglate are contrast agents, largely because of their high described as low osmolality agets, this is a molecular weight but when heated to body relative term. They are all hyper-osmolar temperature (perhaps a good maneuver any- with respect to plasma. way) have acceptably low viscosities for clinical use.

9 Figure 10. Two non-ionic dimeric contrast agents (a) iotrolan and (b) ty). Furthermore, iodixanol. These prepared in formulations are iso-osmolar with plasma metrizamide is a non-ionic agent. It is clear that simply to be non-ionic is not enough, to be low osmolality is not enough, and to be low osmolality and non-ionic is not enough, necessarily, to achieve the low- est possible toxicity. The explanation is that contrast agents possess, by virtue of their chemical structure, an intrin- sic toxicity that is sometimes labeled chemotoxicity.26 This toxicity is agent-specific, It has been argued that the higher viscos- unlike hyperosmolality-related toxicity, ity, combined with the lower diffusibility of which is an entirely non-specific manifesta- such large molecules, is advantageous in tion of the high-concentration solutions and myelography. can be mimicked by high-concentration In summary, at this point, there are two solutions of other materials such as glucose. kinds of non-ionic agents–monomeric and A clue is that when Swick, as discussed dimeric. above, substituted an iodine atom in the hip- puric acid structure, a marked increase in CONTRAST AGENT TOXICITY – THE toxicity was the result. Iodine is a highly CONCEPT OF MOLECULAR TOXICITY hydrophobic atom and it has long been Table 2 shows how LD50 has increased known that the more hydrophobic agents (toxicity decreased) since the early 1930s. have a greater toxicity than the more This is a very crude, but important, measure hydrophilic ones. The relative hydrophobici- of contrast agent toxicity. It reveals an inter- ty/hydrophilicity is usually measured as a esting fact, namely that toxicity cannot be entirely a Table 2. The LD50 Values for a Number on Contrast function of osmolality. Agents From 1930 to the Present Day Metrizamide and sodium meglumine ioxaglate are Agent LD50 (g iodine/kg mouse) low-osmolality agents and, Sodium iomethamate (Uroselectan B) 2.0 indeed, have somewhat Iodopyracet (Diodone) 3.2 lower osmolalities at any Sodium acetrizoate (Urokon) 5.5 iodine concentration than Sodium diatrizoate 8.4 do the second-generation Sodium meglumine Ioxaglate (Hexabrix) 12.0 non-ionic agents, and yet Metrizamide (Amipaque) 14.0 they have significantly Iopamidol (Nipam) 21.0 lower LD50 (higher toxici- Iohexol (Omnipaque) 24.0 10 partition coefficient, which correlates very ecule renders it more toxic by enhancing the well with various measures of toxicity. The hydrophobicity of the molecular core. The addition of iodine to a benzene ring increas- non-ionic agents have lower chemotoxicity es the hydrophobicity and therefore the par- partly because they carry no charge to medi- tition coefficient. ate Coulomb interactions and partly because Another parameter that correlates with they shield the hydrophobic core with their the toxicity is the protein-binding capacity. hydrophilic substituents. The protein-bind- None of these contrast agents possess much ing capacity of the molecule correlates with potential to bind to proteins but significant its toxicity because it is this binding, which differences are to be found between different mediates the toxicity. The partition coeffi- agents, particularly between ionic and non- cient correlates because the more the ionic agents, and there is a close correlation hydrophobic core of the molecule is shield- between protein binding and toxicity. The ed, limiting its interaction with biological higher binding agents are more toxic than the systems, the more hydrophilic the contrast lower binding ones. The relationship agent is overall and the lower, therefore, is between this observation and the partition its partition coefficient. Conversely, the less coefficient has been explained in a synthesis efficiently the hydrophilic substituents shield of ideas as follows. the hydrophobic core, the more the The chemotoxicity of contrast agents hydrophobic interactions can take place, the appears to be mediated by non-specific weak more hydrophobic the molecule the greater interactions between the agents and various the partition coefficient. Furthermore, we biological molecules. This weak binding can now see why metrizamide had such a rel- would not be of any significance if contrast atively high toxicity for a non-ionic agent. agents were not used at such high concentra- All the hydrophilic groups are carried on a tions and total doses as compared with any glucose substituent at position 1, rather than other type of drug. The interaction is mediat- being distributed around the molecule, as is ed essentially by two elements. the case with the second-generation non- 1. There is a Coulomb (charge mediated) ionic agents. There is very little shielding of interaction in the case of the ionic agents the hydrophobic core of the molecule, which that is obviously eliminated in the non- therefore has a relatively high partition coef- ionic agents. ficient, high protein binding and high 2. There are hydrophobic interactions toxicity. between hydrophobic portions of the Figures 11 and 12 display the results of molecule, mainly its benzene ring/iodine two in vitro studies of chemotoxic manifes- core, and hydrophobic portions of the tations of contrast agents, namely the inter- biological molecules (the hydrophilic actions with an enzyme and with structural portions of both types of molecule being proteins of platelet surfaces. The first results largely solvated by water molecules). in an inhibition of the enzyme and the sec- The non-ionic agents have lower chemo- ond in an inhibition of the aggregation of toxicity, not only because they possess no platelets on stimulus. As can be seen, in both charge, but also because of the many cases there is a concentration-dependent hydrophilic groups they carry. The addi- effect (here expressed as iodine concentration of hydrophilic groups restores the tion) and there are marked differences solubility lost when the carboxyl group between different agents. The same hierar- of the precursor ionic agent is eliminated chy of magnitudes is seen in both assays: the and shields the hydrophobic core of the conventional ionic agents have the greatest molecule, thus limiting its availability for effect, the non-ionic agents the least effect, hydrophobic interactions. and the monoacid dimeric low-osmolality Several concepts can now be understood ionic agent, Hexabrix, has an intermediate from this. The addition of iodine to the mol- effect. 11 The ideal contrast agent, it Figure 11. The inhibition of enzyme acetylcholinesterase by should be remembered, should various contrast agents as a function of concentration. have no effects of any kind on biological systems. Although not quite meeting this ideal, the non-ionic agents, monomeric and dimeric, repre- sent the closest approach so far to the ideal.

PHARMACOKINETICS The pharmacokinetics, fitting a two-compartment model, of the iodinated x-ray contrast agents are the same as those of the gadolinium chelates and technetium 99mTc-pertechne- tate (DTPA).27 The two com- partments are plasma and interstitium (intra- and extra- vascular extra-cellular spaces). When contrast agent is deliv- Figure 12. The inhibition of platelet aggregation by various ered into the circulation it is contrast agents as a function of concentration (expressed diluted and mixed in the circu- as iodine). lating plasma volume. There is a very weak tendency to bind to circulating proteins. The agents do not enter cells in the circulation to any significant degree. The agents bind mini- mally to plasma proteins. Being small, the molecules (or ions) also leak rapidly across normal blood vessels into the extra-vascular extra-cellular space. Cells such as hepatocytes may take up 1%-2%, although these agents, to a very good approximation are extra-cellular agents. They may be seen as extra-cellular space markers. The agents do not cross the blood-brain barrier unless it is damaged. (Such damage is the basis for contrast Although illustrations will not be given enhancement of brain pathologies.) There is here, it may be noted that the low chemotox- no metabolism or significant breakdown of icity of the non-ionic monomeric agents is the molecules. Small amounts of free iodine shared, and perhaps shown to a greater in the iodide form [<0.005%] may be found extent, by the non-ionic dimeric agents. in commercial preparations; levels rise if the 12 containers are left in bright sunlight. OSMOLALITY, MOLECULAR TOXICITY, However, there is no further significant ORGAN-SPECIFIC TOXICITIES AND iodine release in vivo. Excretion is by pas- HIGH DOSE TOXICITY sive glomerular filtration with no active We have seen above that both hyperos- tubular excretion or reabsorption. The kid- molality and molecular toxicity contribute to ney may be viewed as a "target organ" in this the toxicity of iodinated contrast agents and sense. This route of excretion is the basis of that the non-ionic agents owe their reduced the technique of intravenous urography. In toxicity to a reduction in both. Some of the the presence of renal impairment there may toxic effects have been described above, be a significant liver uptake and excretion in including injurious effects on endothelium. bile, but otherwise there is little liver uptake The endothelium is now viewed as an or excretion. The gall bladder is occasional- "organ," the largest in the body in fact, and ly visualized on a plain abdominal film or this brings us to the important matter of other CT scan the day after contrast administration organ specific effects. Intravascular contrast in such patients. The gallbladder occasional- agents have a variety of organ-specific toxi- ly may be seen if normal function is cities. impaired by the contrast agent (contrast Endothelium: (the largest organ in the agent associated nephrotoxicity). body) Damage predisposes to thrombosis. Thus when a contrast agent is injected There is one hypothesis that endothelial into the circulation by bolus or rapid infusion injury and its effect on various activation it mixes in plasma and leaks into interstitium systems is the basis of major reactions.26 everywhere in the body. As plasma levels fall Cardiac: There are adverse effects on – as a result of the mixing and dilution itself, pump function and electrophysiology.21 as a result of the leak and, to a lesser extent, Renal: There may be temporary or occa- as a result of renal excretion – they eventual- sionally permanent impairment of renal ly fall below the levels in the interstitium and function. The effect is thought to be more then the agent leaks back into the circulation likely in older patients, in diabetics, in for ultimate excretion by the kidneys. patients with already impaired renal function If contrast agent is injected into any body and in dehydrated individuals.28-35 cavity, such as a joint, the spinal theca or an Neural Tissues: The blood brain barrier abscess cavity it will be reabsorbed into the protects the normal brain but contrast agents, circulation and excreted by the kidneys. particularly high osmolar ones, may damage There is no absorption from the normal, as the barrier and then pass. Injection into the opposed to perforated or inflamed, gut. The spinal theca of course bypasses the barrier peritoneum allows rapid intravasation into and brings contrast agent into direct contact the circulation. Thus if there is a gut perfora- with neural tissues. tion and oral contrast is administered, it leaks Liver: Toxicity is not readily apparent into the peritoneum, rapidly enters the circu- but it should be noted that there is a general lation and may be seen if normal. weak cytotoxic effect, greater with the ionic Only small amounts of contrast agent are agents, on cells of all types. found in breast milk. In any case, this seems In organ specific studies such as coro- to be of marginal importance since if this nary angiography, cerebral angiography, milk is given to a normal infant with a nor- renal angiography, low osmolality non-ionic mal gut there will be no significant systemic agents tend to be used because of their lower absorption by the infant. Contrast agents are toxicities. therefore not contraindicated in breast-feed- Organ-specific toxicities are most obvi- ing mothers. ously manifest in selective individual organ studies such as cardioangiography, cerebral

13 angiography, etc and, since they are dose- here, giving scope for higher doses to be dependent, play a role in the high-dose sys- used if the examination is vital and demands temic toxicity which is encountered in some it. There is clearly a question of clinical circumstances as discussed below. judgment to be exercised in the context of the individual patient, the importance of the High-Dose Toxicity of Contrast Agents procedure and the consequences of not pro- While much energy has been expended ceeding with it. The time over which the on the problem of idiosyncratic/anaphylac- total dose is to be administered is also clear- toid reactions to contrast agents (see below), ly an important consideration, remembering particularly with regard to the elucidation of that the half-life of contrast agents in the cir- their underlying mechanisms, their predic- culation is approximately 90 minutes.27 One tion and their prophylaxis and management, other important factor now is the availability much less attention has been paid to the of the better low-osmolar non-ionic contrast problem of the dose-dependent side effects agents. These have a higher LD50, present a of the agents. In terms of frequency, dose- lower osmotic load and contain little or no dependent effects are of greater importance sodium. Whatever arbitrary upper limit in than idiosyncratic reactions in a number of mg iodine/kg is set for the conventional patient groups who are well represented in agents, it seems likely that twice or three radiology, such as those with poor cardiac times this dose of a new agent can be used. reserve, impaired renal or hepatic function, Thus if 1000 mg iodine/kg may be given in and multi-system organ failure.36 Serious the form of a conventional agent, possibly idiosyncratic reactions are uncommon (see 2500 mg iodine kg1 of a new agent may be below) but higher and higher doses of the given. If this is in the form of, say, agents are being used in more and more Omnipaque 300 (Iohexol 300 mg patients in the field of interventional radiolo- iodine/mL), this would mean a volume dose gy. Even without such obvious underlying for a 70-kg man of (2500 x 70) /300 ≈ 600 risk factors, many patients may be at risk mL. from dose-dependent contrast medium toxic- Even higher doses may presumably be ity, including sodium and/or osmotic over- used if the contrast medium is administered load in complex or prolonged procedures over an extended period, for example in a 2 involving large total doses of contrast medi- or 3 hour long complicated procedure. Such um.36 high doses should never be given without serious thought but, if the procedure is vital Theoretical Considerations for diagnosis or therapy considerations, they In a typical IV urogram, 300 mg may then be given with caution, detailed iodine/kg may be used. For a 70-kg man this decisions being tailored to the individual represents 21g iodine, which could be pro- patient. A reasonable precaution, with the vided, for example, by 50 mL of sodium risk of renal function in mind, is to make iothalamate containing 420 mg iodine/mL. sure that the patient is well hydrated at the Several rules of thumb are in use to define an start of the procedure (see below). Note that upper limit for the total contrast medium even a modest dose of a low osmolar agent dose. A maximum dose for a healthy adult of may put patients with poor cardiac function some three times the above level, which is into pulmonary edema. 1000 mg iodine/kg, is sometimes suggested. This may be compared with the measure- ANAPHYLACTOID/IDIOSYNCRATIC ment of the LD50 of ~ 8000 mg iodine/kg for REACTIONS conventional contrast agents, a comparison Idiosyncratic reactions may occur fol- of uncertain relevance to the clinical situa- lowing the administration of any contrast tion but the only guide available. There agent.27 Clinical manifestations include would appear to be some safety margins severe bronchospasm, angioedema and car- 14 diovascular collapse. They have close paral- is definably at increased risk of such a reac- lels with anaphylactic responses and may be tion. clinically indistinguishable from them but they are not anaphylactic. For this reason Prophylaxis of Anaphylactoid Reactions they are sometimes described as anaphylac- If anxiety does have a role to play, sim- toid reactions. Although unpredictable, such ple reassurance may achieve worthwhile reactions are more likely to occur in certain results. The role of corticosteroids as pro- patient groups: (i) those that have reacted on phylaxis against such reactions in at-risk a previous occasion to a contrast agent (the patients has been the subject of considerable recurrence rate at a second contrast agent controversy.40 It is widely believed that corti- administration is thought to be only ~ 20%), costeroids are effective in this regard but the and (ii) those with established allergies to evidence for it is not substantial. The paper other drugs or agents and (iii) asthmatic and by Lasser and colleagues41 describing a mul- atopic individuals. Whether or not cardiac ticenter study in the USA is usually cited, but disease is a risk factor is controversial with this paper has not been without its critics. It different studies reaching different conclu- has been argued that even if its evidence is sions. These reactions are essentially dose- accepted at face value, the reduction in risk independent and may even occur following is little more than 50%, whereas a 6-10 fold subcutaneous or intradermal test doses. decrease in risk is achieved by using non- Occasionally, reactions follow non-vascular ionic contrast agents.40 Some, as yet incon- procedures such as percutaneous transhepat- clusive, evidence has been claimed for the ic cholangiography, arthrography and hys- efficacy of antihistamines as prophylaxis. terosalpingography, presumably as a result Since anxiety is said to play a role, reassur- of intravasation in these cases. It is important ance to the patient may be as valuable as any to realize that such reactions may be delayed other maneuver. up to 30 min and the patient should never be left entirely unsupervised during or immedi- Treatment of Anaphylactoid Reactions ately following a procedure. If a patient collapses following contrast The mechanisms of these reactions agent administration, first principles must be remain obscure in spite of a considerable applied – clear and maintain the airway and amount of study and the elaboration of a perform CPR. Rapid plasma volume expan- number of hypotheses. Apart from the citing sion is known to be effective. A first line of various mediators of the immune system a drug in cardiovascular collapse and severe role has been suggested (supported by a bronchospasm in these circumstances is great deal of anecdotal evidence) for anxiety adrenaline (1:1000 deep intra-muscular or, and the possibility that a substantial propor- with care, 1:10,000 (suitably diluted) intra- tion of those events involving cardiovascular venously and titrated against effect). collapse are of primary cardiac rather than Corticosteroids are not front line drugs but allergic origin. However, it has now been may have a useful and rapid effect on bron- clarified that the non-ionic monomeric con- chospasm and cardiovascular collapse. A trast agents are associated with a significant- variety of other drugs may also have a role. ly reduced risk of such reactions, particular- The reader is referred for details of recom- ly in definably at-risk patients. This certain- mended management to Guidelines issued, ly is the outcome of large-scale clinical trials variously, by The American College of of ionic versus non-ionic contrast agents in Radiology and The Royal College of Japan38 and Australia.39 It can be firmly stat- Radiologists in the UK. The following, orig- ed, as a consequence, that non-ionic agents inally developed by the author and repro- should be used in all cases when the patient duced with permission, is taken from the latter guidelines.

15 Management of Adverse Reactions to safe and effective with a quoted adverse Intravascular Contrast Agents reaction rate of 5.8%, the vast majority being Conventional ionic iodinated intravascu- of a minor nature. The adverse reaction rate lar contrast media have been available for associated with the newer non-ionic agents is over 35 years and have proved to be very generally considered to be of the order of one

Table 3. Management of Adverse Reactions to Intravascular Contrast Agents Symptoms Treatment Nausea/vomiting Reassurance Retain intravenous (IV) access and observe Anti-emetics rarely necessary Mild scattered hives/urticaria Routine treatment not necessary Retain IV access. Observe. If troublesome, antihistamine – chloropheniramine maleate 10-20 mg or promethazine hydrochloride 25-50 mg (max 100 mg) by slow IV injection Severe generalized urticaria IV antihistamines as above. With addition of IV Hydrocortisone 100 mg IV Mild wheeze 100% oxygen by MC mask (unless evidence of chronic obstructive airways disease with hypercapnia) Beta-2 agonist by inhaler e.g. salbutamol by nebulizer (5 mg in 2 ml saline). Repeat as necessary. Hypotension with bradycardia Raise the patient’s feet (vasovagal reaction/faint) 100% oxygen by MC mask (unless evidence of chronic obstructive airways disease with hypercapnia) IV fluids (preferably colloid e.g. Gelofusine – 10 mL/Kg rapidly. THE HELP OF AN ANESTHESIOLOGIST SHOULD BE REQUESTED IF RESPONSE IS NOT RAPID. ECG monitoring and oximetry should be instigated. Atropine 0.6mg IV. Repeat at 5 minute intervals up to 3 mg total. Hypotension alone – 100% oxygen by MC mask not vasovagal but no other IV fluids rapidly rapidly as above of Ephedrine 30 mg signs anaphylactoid reaction diluted and given incrementally to response. It is usually given in 3 to 6 mg increments providing that the patient is not tachycardic. (In tachycardia methoxamine 5-10mg IV at rate of 1 mg/min may be more appropriate.) IF THERE IS NO RESPONSE THE CRASH TEAM SHOULD BE SUMMONED TO DIRECT THE SUPERVISION OF PRESSOR AGENTS: Vasopressor, e.g. dobutamine 2.5-10 micrograms/kg/min or dopamine 2-5 micrograms/kg/min titrated to response. ECG Oximetry and blood pressure monitoring should be established. NOTE ANTIHISTAMINES AND HYDROCORTISONE SHOULD NOT BE MIXED IN THE SAME SYRINGE. IF THIS IS DONE, IT CAUSES PRECIPITATION (Continued on next page)

16 fifth that with the older conventional quickly and appropriately. ionic agents. Uncommonly, a patient will The exact mechanisms of contrast reac- develop a more serious reaction to tions remain undefined, but several factors either type of agent that may prove life- may contribute, with the individual patient’s threatening and which must be treated ‘reactivity’ being the most important.

Table 3. Management of Adverse Reactions to Intravascular Contrast Agents (continued) Symptoms Treatment Angioedema/ IV fluids Urticaria/ Beta-2 agonist by nebulizer as above Bronchospasm/ IV antihistamines Hypotension proceeding to IV hydrocortisone 500 mg Anaphylactoid reaction ECG monitoring, oximetry, and blood pressure monitoring should be established. CALL THE CRASH TEAM* The crash team will then consider the use of adrenaline 1 mL of 1 in 10,000 [100 microgram by slow IV injection (10 microgram/min); in severe cases further aliquots up to a total of 1 mg may be required.]** *The use of adrenaline when appropriate should not be delayed if there are medical staff in attendance. **The use of IV (1/10,000) vs deep IM (1/1,000) adrenaline is controversial. Some authorities advocate the IM route. Others argue that if medically qualified staff are in attendance the IV route should be used with due caution. Unconscious/ Standard cardiopulmonary resuscitation Unresponsive/ measures must be employed Pulseless/ 1. CRASH TEAM summoned Collapsed patient The crash team will then institute standard resuscitation procedures: 2. Establish airway 3. Initiate ventilation 4. Precordial thump followed by external cardiac massage 5. IV fluids established 6. Cardiac monitoring by ECG 7. DC cardioversion if in ventricular fibrillation (200 to 360 joules) 8. IV adrenaline/isoprenaline/lignocaine/calcium chloride as necessary 9. Consider administration of hydrocortisone/antihistamines during course of a successful resuscitation. Seizure May be the consequence of hypotension and primary treatment should be on above lines. 100% oxygen by MC mask. If convulsions continue, anticonvulsant may be given – e.g. diazepam IV 5-10 mg, initially although much higher doses may be needed. Second-line drugs such as thiopentone may be required but by this time the patient should be intubated and ventilated.

17 Anecdotally, stress and fear increase a Aminophylline is an agent that has fallen patient’s susceptibility to a contrast reaction. into disrepute in recent years since it may be Histamine release, complement activation, associated with hypotension and collapse coagulation and fibrinolytic system activa- and may only serve to exacerbate the situa- tion and prekallikrein to kallikrein transfor- tion. mation with bradykinin generation may all be elements involved. Central nervous sys- General Points Worth Reinforcing tem direct toxic effects have also been cited Depending on the severity of the reaction as a primary trigger. The very diversity of and the length of treatment required, an postulated reaction mechanisms appears to intravenous cannula should be inserted as have produced confusion about manage- soon as is practicable and drugs administered ment. Identification and symptomatic char- through this rather than the needle, which acterization of a reaction are the key first carries the risk of cutting out. steps and should be followed by ad hoc man- Contrast agents should not be injected in agement based on general principles. an isolated clinical setting. Facilities for Symptom complexes are listed together resuscitation should be available if required. with their suggested management. They are Non-ionic contrast medium should be used if placed in approximately ascending order of a full cardiac arrest team is not available. severity, no attempt being made to integrate The patient should never be left alone symptomatology into a coherent etiological following injection particularly during the scheme. first 10 minutes. Remember serious reac- DRUG DOSAGES WHERE GIVEN tions may sometimes by considerably ARE APPLICABLE TO ADULTS. SUIT- delayed. ABLE ADVICE SHOULD BE TAKEN IN Equipment and drugs routinely used in PEDIATRIC PATIENTS. (Anti-emetics the management of medical emergencies may be dangerous in children as there is a should be immediately available. One trol- high incidence of extra-pyramidal side- ley serving a unit with scattered rooms on effects to such drugs as perphenazine.) several floors is not acceptable. Equipment and drugs should be regular- General Comments ly checked by a designated person(s) and Corticosteroids are, quite correctly, recorded as checked. thought of as slow-acting drugs with a lag All persons involved in the daily running time after injection of at least 6 hours. of a department should be aware of the site However, there is no doubt they may work of resuscitation equipment. very rapidly indeed when administered intra- All personnel should attend a basic venously in severe asthma and in bron- resuscitation course on a regular basis. chospasm related to drug reactions and Each department should have a specific should therefore not be withheld. protocol for dealing with various reactions, IV adrenaline is sometimes considered a which should be updated periodically. dangerous drug associated with cardiac The person who administers the contrast arrhythmias. This is true and it should not be should have a basic medical history about used in minor reactions but only in severe the patient, particularly relating to previous life threatening bronchospasm and/or cardio- allergies and be adequately trained in resus- vascular collapse when the potential benefits citation procedures. greatly outweigh the risks. Doses may have Suggested drugs for the emergency box to be high and may need to be repeated. to be placed in each room where intravenous Under normal circumstances Adrenaline contrast agents are used: should only be given by a medical or other ➣ Chlorpheniramine maleate suitably trained practitioner. ➣ Promethazine ➣ Hydrocortisone 18 ➣ Atropine abilities and sizes of vascular and interstitial ➣ Salbutamol inhaler spaces) it also usually increases the contrast ➣ Salbutamol nebulizer between normal and abnormal. In other ➣ Ephedrine words, the agents enhance what is really ➣ Normal saline important, namely the conspicuity of abnor- ➣ Gelofusine malities in the image. The following will usually be reserved for use by the crash team: A BRIEF NOTE ON COAGULATION ➣ Methoxamine AND PLATELETS ➣ Dobutamine All intravascular x-ray contrast agents ➣ Dopamine inhibit coagulation and platelet aggrega- ➣ Adrenaline (1 in 10,000) tion.42 Historically, this has been thought ➣ Adrenaline 1 in 1,000, at least 10 ampules advantageous in angiography since it helps The advice of the local consultant anes- prevent formation of clots in catheters and thesiologist designated to take responsibility syringes, which might be accidentally inject- for anesthesiology matters within the radiol- ed into the patient sometimes with cata- ogy department, should be sought when for- strophic results. The non-ionic agents have a mulating local policies. less anticoagulant properties than do the ionic agents. This caused a considerable con- A NOTE ON THE CLINICAL USE OF troversy, more in the United States than else- CONTRAST AGENTS where because of the medico-legal climate, It is difficult to convey in a short survey with some expressing anxieties that non-ion- the range of application and enormous ics might be unsafe in angiography. It has importance of contrast agents in clinical even been erroneously suggested that non- practice. Other than in such simple imaging ionic agents have some positive pro-coagu- as chest or bone x-rays, scarcely an x-ray lant or pro-thrombotic effect. The simple image is ever obtained without the concomi- facts are that anticoagulant properties are a tant administration of a contrast agent, intra- function of chemotoxicity. Therefore, the vascular, oral or both. Direct injection into greater the anticoagulant properties are, the blood vessels via a variety of catheter tech- more toxic the contrast agent. The key to niques makes blood vessels clearly visible. safe angiography lies not in a return to more This whole field is called angiography. toxic agents but in the use of less toxic non- Injection into natural body cavities, such as ionic agents together with good technique.42 joints (arthrography) or the spinal theca (myelography) helps delineate normal and THE METFORMIN ISSUE abnormal anatomy. Iodinated intra-vascular The biguanide agent for the treatment of agents are handled and excreted almost NIDDM, Phenformin, was associated with a exclusively by the kidneys and during their significant incidence of Type B lactic acido- excretion following a simple intravenous sis. The mortality of this condition is some injection render the whole renal tract – kid- 50%.43 The drug was withdrawn in 1977. The neys, ureters, and bladder – visible on x-ray related drug, Metformin, is also a recognized images (Intravenous Urography – IVU). cause of this serious condition but much less Virtually all body CT scans are "enhanced" frequently than Phenformin. The incidence is by the administration of both an intravascu- reported to be ~0.03 per 1000 patients lar contrast agent and an oral agent. This not years.44 In the reported cases there do almost only helps clarify which soft tissues are invariably exist contraindications to the very blood vessels or gut but, because normal and use of Metformin.44-47 Nevertheless, when the abnormal tissues (such as, say, tumors) han- drug was introduced into the USA, the man- dle contrast agents somewhat differently ufacturer, Bristol Myers Squibb, put a warn- (because of their differing capillary perme- ing into the package insert regarding the co- 19 administration of contrast agents, a warning lactate clearance, liver dysfunction may based on the idea that contrast agents are increase the risk of lactic acid accumula- thought associated with renal function tion in patients taking Metformin. This is impairment, the sine qua non of lactic acido- particularly likely if the clearance of the sis. The medico-legal context, in which drug is reduced by renal disease or if car- American medicine is conducted, combined diac impairment reduces renal perfusion. interestingly with the Internet, generated and ➣ The renal function of patients undergo- spread anxiety rapidly. These anxieties then ing contrast enhancement may become spread to the United Kingdom, notwith- impaired, particularly if there is pre- standing the fact that Metformin had been in existing renal disease.48 Patients with dia- use there for many years. In response to betes are more likely to have some many enquiries the Royal College of degree of impairment of renal function Radiologists felt obliged to put out some than the normal population and are said guidelines in 1996. These guidelines caused to be at greater risk of contrast nephropa- a great deal of controversy since they con- thy. siderably complicated the lives of radiologists dealing with such patients who are by Is there a risk in practice? no means rare. Was all this necessary and ➣ As long ago as 1977, sporadic cases were was there a real problem? reported of lactic acidosis in patients receiving Metformin.49,50 The cases were Theoretical Mechanisms apparently precipitated by the adminis- Theoretically the combination of tration of iodinated x-ray contrast agents. Metformin and x-ray contrast agents may The cause then suggested was as indicat- have adverse effects because contrast agents ed above: that contrast agents impair, or supposedly impair or further impair renal further impair, renal function and hence function and hence Metformin excretion. Metformin excretion. ➣ Metformin is a well-recognized cause of ➣ Further cases have been reported occa- Type B lactic acidosis in which mortality sionally throughout the 1980s and is greater than 50%.43 Type B lactic aci- 1990s.51, 52 dosis is much less likely to occur with McCartney et al carefully reviewed the Metformin than its predecessor, literature in this field.53 They found that in Phenformin, which was withdrawn in more than a quarter of a century there were 1977. only 18 reports of Metformin-induced lactic ➣ The causal mechanisms may be related acidosis (MALA) after contrast usage. In to increased peripheral anaerobic glycol- only one case was renal function normal at ysis, and inhibition of hepatic gluconeo- the start of the examination. They concluded genesis from lactate, associated with a that at least 16, and probably 17, had con- reduction in hepatic intracellular pH. The traindications to Metformin use in the first latter might further compromise hepatic place. Their conclusion was essentially that lactate metabolism. the problem almost invariably lies in the pre- ➣ Metformin is excreted unchanged by the scribing of Metformin rather than in the use kidney and reduced renal function is a of the contrast agent. sine qua non for the development of lactic acidosis. Mild degrees of liver dys- What should be done? function do not appear to substantially Modified RCR guidelines suggest dis- predispose to lactic acidosis in general. continuation of Metformin at the time of the Arterial pH is often raised in hepatic fail- investigation and withholding it for 48 hours. ure because of respiratory alkalosis. For those with abnormal renal function it Nonetheless, because chronic liver dis- should only be reinstated when the renal ease is associated with reduced rates of function is found to be normal. This at least 20 has the merit for the radiologist of passing CONCLUSIONS the responsibility to the physician. The water-soluble intra-vascular iodinat- McCartney and associates53 suggest ed x-ray contrast-enhancing agents are what, in effect, is another way of passing essential tools in medical imaging. Their responsibility and inconvenience to the chemistry and pharmacokinetics are very referring clinician. They suggest that since simple and, superficially speaking, they may no patient with impaired renal function be said to have no pharmacology. Their man- should be on Metformin the clinician be ufacturers prefer us to think of them as some- asked to state that the patient is on thing other than "drugs" – hence the term Metformin and has normal renal function. In "agent". However, as has been described, this case, they argue, there is no evidence of they have a considerable variety of interest- a need to stop Metformin. If the renal func- ing and clinically important effects that tion is not normal the physician should be should be understood by anyone administer- assessing whether Metformin is appropriate ing them. Indeed they are associated, albeit therapy anyway. Emergency patients should rarely, with significant morbidity and mortal- undergo the study and the Metformin should ity. However, their enormous usefulness in be stopped as a precaution. Further manage- adding value to x-ray examinations out- ment again will revert to the clinical team. weighs the very small risks associated with This all seems eminently sensible from the their use. radiologist’s point of view.

21 REFERENCES 11.Wallingford VH. Iodinated acy- laminobenzoic acids as x-ray contrast 1. Haschek E, Lindenthal TO. Ein Betrag media. J Am Pharmacol Assoc. zur praktischen Verwerthung der 1952;42:721-728. Photographie nach Röntgen. Wien Klin. Wschr. 1896;9:63-64. 12. Hoppe JO. Some pharmacological aspects of radio opaque compounds. 2. Burrows EH. Pioneers and Early Years. Ann NY Acad Sci. 1959;78:727-739. Colophon Ltd 1986. Alderney. 13. Grainger RG. Osmolality of intravascu- 3. Berberich J, Hirsch S. Die lar radiological contrast media. Br J Röntgenographische Darstellung der Radiol. 1980;53:739-46. Arterien und am Lebenden. Klin Wschr. 1923;9:226-228. 14. Dawson P, Grainger RG, Pitfield J. The new low osmolar contrast media. A sim- 4. Osborne ED, Sutherland CG, Scholl AJ, ple guide. Clin Radiol. 1983a;34:221- Rowntree LG. Roentgenography of uri- 224. nary tract during excretion of sodium iodide. JAMA. 1923;80:368-373. 15. Snowdon SL and Whitehouse GH. Blood pressure changes resulting from 5. Brooks B. Intra-arterial injection of aortography. JR Soc Med. 1981;74:419- sodium iodide. JAMA. 1924;83:1016- 421. 1019. 16. Raininko R. Role of hyperosmolality: 6. Moniz E. Die Cerebrale Arteriographie the endothelial injury caused by angio- und Phlebographie. Springer Verlag; graphic contrast media. Acta Radiol. Berlin: 1940. 1979;20: 410-416.

7. MacMahon HE, Murphy AS, Bates MI. 17. Laerum F, Holm HA. Endothelial cell sarcoma of the liver Postphlebographic thrombosis. following thorotrast injection. Am J Radiology. 1981;140:651-654. Pathol. 1947;23:585-611. 18. Sage MR. Neuroangiography. In: 8. Wenz W. Tumors of the kidney follow- Skucas J, ed. Radiographic Contrast ing retrograde pyelography with col- Agents, 2nd edition. Aspen; Rockville: loidal thorium dioxide Ann NY Acad 1989;170-188. Sci. 1967;145:8806-8810. 19. Dawson P, Harrison JG, Weisblatt E. 9. Swarm RL. Colloidal thorium dioxide. Effect of contrast media on red cell fil- In: International Encyclopaedia of terability and morphology. Br J Radiol. Pharmocolgy and Therapeutics. Vol. 2 1983b;56:707-710. Radiocontrast Agents. Pergamon Press 1971, Oxford. 20. Assem ESK, Bray K, and Dawson P. The release of histamine from human 10. Grainger RG. Intravascular contrast basophils by radiological contrast media – the past, the present and the agents. Br J Radiol. 1983;39:987-991. future. Br J Radiol. 1982;55:1-18. 21. Dawson P. Cardiovascular effects of contrast agents. Am J Cardiol. 1989;64:2E-9E. 22 22. Almen T. Contrast Agent design: some 31. Oldroyd S, Haylor J, Morcos SK. Effect aspects of the synthesis of water- solu- of Bosentan, an orally active endothelin ble contrast agents of low osmolality. J antagonist: effect on the renal response Theor Biol. 1969;24:216-226. to contrast media. Radiology. 1995; 196:661-665. 23. Swanson DP, Jakubowski DL, Shoup LK, Thrall JH. Bid purchasing of radio- 32. Morcos SK, Oldroyd S, Haylor J. Effect pharmaceuticals and radiopaque con- of radiographic contrast media on trast media. Am J Hosp Pharm. endothelium derived nitric oxide- 1985:42(7);1532-1536. dependent renal vasodilatation. Br J Radiol. 1997;70:154-159 24 Swanson DP, Thrall JH, Shetty PC. Evaluation of intravascular low-osmo- 33. Morcos SK, Thomsen HS, Webb JA. lality contrast agents. Clin Pharmacol. Contrast-media-induced nephrotoxici- 1986:Nov 5(11);877-891. ty: a consensus report. Eur Radiol. 1999;9:1602-1613. 25. Dawson P, Howell MJ. The pharmacology of the non-ionic dimersa new class 34. Oldroyd S, Fang L, Haylor J, Yates MS, of contrast agents. Br J Radiol. El Nahas AM, Morcos SK. Effects of 1986;59:987-991. adenosine receptor antagonists on the responses to contrast media in the iso- 26. Dawson P. Chemotoxicity of contrast lated rat kidney. Clin Sci (Lond). media and clinical adverse effects: A 2000;98:303-311. review. Invest Radiol. 1985; 20: 52-9. 35. Morcos SK. Contrast media-induced 27. Dawson P. ATextbook of Contrast nephrotoxicity: an overview. Kidney Media. ISIS Medical Media Ltd, Forum. 2000;2:49-52. Oxford, 1999. 36. Dawson P and Hemingway AP, Contrast 28. Oldroyd S, Haylor J, Morcos SK, el doses in interventional radiology. J Nahas AM. Reduced depression of Intervent Radiol. 1987;2:145-146. renal function by iotrolan in the isolated rat kidney. Eur J Radiol. 1994;18:64- 37. Ansell G, Tweedie MCK, West CR, et 69. al. The current status of reactions to contrast media. Invest Radiol. 29. Oldroyd S, Slee SJ, Haylor J, Morcos 1980;15:S32-S39. SK, Wilson C. Role for endothelin in the renal responses to radiocontrast 38. Katayama H, Yamaguchi K, Kozuka T, media in the rat. Clin Sci (Lond). et al. Adverse reactions to ionic and 1994;87:427-434. nonionic contrast media. Radiology. 1990;175:621-628. 30. Morcos SK, Brown PWG, Oldroyd S, El Nahas AM, Haylor J. Relationship 39. Palmer FJ. The RACR survey of intra- between the diuretic effect of radiocon- venous contrast media reactions. Final trast media and their ability to increase report. Australas Radiol. 1988;32:426- renal vascular resistance. Br J Radiol. 428. 1995;68:850-853.

23 40. Dawson P, Sidhu P. Is there a role for 47. Data Sheet, Metformin Hydrochloride. corticosteroid prophylaxis in patients at Bristol Myers Squibb, London: 1997. increased risk of adverse reactions to intravascular contrast agents? Clin 48. Dawson P, Trewhalla M. Intravascular Radiol. 1993;48:225-226. contrast agents and renal failure. Clin Radiol. 1990;41:373-375. 41. Lasser EC, Berry CC, Talner LB, et al. Pretreatment with corticosteroids to 49. Prinseau J, Jardin F, Sanchez A, et al. alleviate reactions to intravenous con- Acute renal failure after intravenous trast material. N Engl J Med. urography in diabetics. Incidence of 1987;317:845-849. lactic acidosis in metformin treatment. Ann Med Interne (Paris). 1977; 42. Dawson P, Strickland NH. Thrombo- 128:173-177. embolic phenomena in clinical angiography: role of materials and technique. J 50. Assan R, Heuclin C, Ganeval D, et al. Vasc Interv Radiol. 1991; 2:125-135. Metformin-induced lactic acidosis in the presence of acute renal failiure. 43. Monson P. Selected side effects: II Diabetologia. 1977;13(3):211-217. Metformin and lactic acidosis. Prescribers Journal. 1993;33:170-173. 51. Jamet P, Lebas de Lacour JC, Christoforov B, Stern M. Acidose lac- 44. Bailey CJ, Turner RC. Metformin. N tique mortelle après urographie intra- Engl J Med. 1996;334(9):574-579. veineuse chez une diabétique recevant de la Metformin. Sem Hop. 45. Hermann LF, Melander A. Biguanides. 1980;56:473-474 (French). Basic actions and clinical uses. In: Alberti KGMM, Zimmer P, DeFronzo 52. Sirtori CR, Pasik C. Re-evaluation of RA, et al. International Textbook of biguanide, metformin: mechanism of Diabetes Mellitus, 2nd ed. Chichester: action and tolerability. Pharmacol Res. John Wiley and Sons; 1997:841-864. 1994;30:187-228.

46. Association of British Pharmaceutical 53. McCartney MM, Gilbert FJ, Murchison Industry Compendium Data Sheets. LE, et al. Metformin and contrast 1998:664-665. media—a dangerous combination? Clin Radiol. 1999;54:29-33.

24 QUESTIONS 6. Which of the following statements regarding intravascular x-ray contrast 1. Iodine is an ideal element for use in x- agents is not true? ray contrast agents because: a. inhibit coagulation a. its atomic number is high b. inhibit platelet aggregation b. it is a halogen c. non-ionics are better anticoagu- c. it is non-reactive chemically lants than ionics d. it has an appropriate K-shell d. ioxaglate is a better anticoagulant binding energy than the non-ionics e. it is a β emitter. e. anticoagulant effect is a marker for toxicity 2. Which of the following is not a common side effect of intravascular x-ray 7. On intrasvascular contrast agent phar- contrast agents? macokinetics. The agents: a. metallic taste a. can readily negotiate the blood b. flushing brain barrier c. bronchospasm b. may be found in high concentra- d. vomiting tions in bile e. pain on arterial injection c. are described by a 2-compartment model 3. Which of the following is not a clear- d. are reabsorbed by renal tubular cut risk factor for anaphylactoid reac- cells tions to intravascular x-ray contrast e. take part in an enterohepatic agents? circulation a. a previous reaction b. asthma 8. Intravascular x-ray contrast agents c. cardiac disease a. are strongly bound to circulating d. allergies to other drugs or agents plasma proteins e. atopy b. stimulate histamine release c. are blood pool agents 4. Intravascular iodinated x-ray contrast d. are not significantly excreted by agents: the liver in renal failure a. distribute themselves between e. contain no free iodine cells and interstitium b. distribute themselves between 9. Organ-specific toxicities of intravas- plasma and interstitium cular x-ray contrast agents include: c. are taken up avidly by hepatocytes a. cardiotoxicity d. are excreted by renal tubules b. hepatotoxicity e. break down in the body c. nephrotoxicity d. neurotoxicity 5. Which of the following is not thought e. endothelium toxicity to be associated with intravascular contrast agent nephrotoxicity? a. all of the above a. dose-dependent b. a, b, and c only b. more likely in the elderly c. a, b, and d only c. more likely in patients with d. a, c, d, and e only pre-existing renal disease e. c, d, and e only d. more likely if the patient is dehydrated e. more likely in patients taking aspirin

25 10. On intravascular contrast agent chemotoxicity. This is: a. mediated by hydrophilic 14. Intravascular iodinated x-ray contrast substituents agents: b. mediated by hydrophobic parts of a. are absorbed from the normal gut the molecule b. are absorbed across the peri- c. higher in non-ionic agents than the toneum ionic agents c. freely cross the normal blood brain d. higher in non-ionic dimers than barrier monomers d. freely enter normal cells e. mediated by additives in the e. freely enter abnormal cells formulation 15. Regarding ‘low osmolality’ intrasvas- 11. High osmolar intravascular x-ray con- cular x-ray contrast agents, they are: trast agents: a. always non-ionic a. expand the plasma volume b. always dimeric b. cause red cells to expand c. always iso-osmolar with plasma c. inhibit histamine release d. monomers are always hyperosmo- d. do not predispose to venous lar with respect to plasma and thrombosis body fluids e. do not cause arterial pain e. formulation additive free 12. The structures of modern intravascular 16. Concerning ionic intravascular con- contrast agents are: trast agents: a. based on the pyridone ring a. these should be used for b. based on the benzene ring myelography c. relatively hydrophilic b. pure sodium salts should be used d. all dimeric for neuroangiography c. pure meglumine salts are ideal for e. relatively hydrophobic coronary angiography a. a. a and b only d. meglumine salts are more painful b. b and c only in arteriography than sodium c. c and d only salts d. d and e only e. meglumine salts are less irritant to e. none of the above veins than sodium salts 13. Which of the following statements 17. Concerning ‘low osmolar’ intravascu- regarding intravascular contrast agent lar x-ray contrast agents: toxicity is not true? a. non-ionic monomers are iso- a. mediated by high osmolality osmolar with body fluids b. a function of charge b. ioxaglate should be used for c. a function of hydrophobicity myelography d. decreased by addition of iodine to c. ioxaglate is best used the molecule intravenously e. dose-and concentration-dependent d. not all ionize in solution e. are never hyperosmolar with respect to body fluids

26 18. Intravascular x-ray contrast agents are: 22. Which is not true regarding treatment a. readily absorbed across inflamed of major anaphylactoid reactions? gut a. corticosteroids are the first line b. not absorbed from joint spaces treatment c. excreted by the kidneys following b. adrenaline is the first line interthecal injection treatment d. found in high concentrations in c. fluid replacement is important breast milk d. maintenance of airway and CPR e. taken up avidly by hepatocytes are of secondary importance e. adrenaline may be given intra- a. a and b only venously or intramuscularly b. a and c only c. b and c only a. a and d only d. b and d only b. b and c only e. d and e only c. c and a only d. d and e only 19. Anaphylactoid reactions to intrasvas- e. all of the above cular x-ray contrast agents: a. occur more commonly in 23. Which of the following is not true asthmatics regarding prophylaxis against major b. are more likely with non-ionic reactions? agents a. corticosteroids have an undisputed c. have a 90% recurrence rate after a role to play previous major reaction b. antihistamines may be useful d. corticosteroid prophylaxis reduces c. reassuring the nervous patient may the risk by a factor of 10 be helpful e. non-ionics are never associated d. the best ‘prophylaxis’ is the use of with them a non-ionic agent e. if corticosteroids are used they 20. Toxicity of intravascular x-ray con- may be given just before the trast agents is not mediated by: contrast agent a. hyperosmolality b. molecular weight a. a and b only c. chemotoxicity b. b and d only d. charge c. a and e only e. all of the above d. c and d only e. b and e only 21. Idiosyncratic reactions to intravascular x-ray contrast agents are: 24. On additives in intravascular x-ray a. anaphylactic in nature contrast agents: b. all due to cardiotoxic effects a. non-ionic agents contain no c. dose dependent formulation additives d. more common in atopic b. EDTA is added to all agents individuals c. Ca EDTA is added to non-ionic e. usually predictable agents d. non-ionic dimers are additive free e. all of the above.

27 25. Regarding high dose and relatively high doses: a. 1000 mgI/Kgms is a typical dose in IV urography b. high doses may cause acute pulmonary edema c. renal function is not a concern in high dose d. hepatic function is a concern in high dose e. ionic agents are safer if high doses are to be used

28 NOTES NOTES