Risk of Thrombogenicity among Nonionic Radiocontrast Agents

*Alexander Georgakis, MD, *Rasih Atilla Ener, MD, Jianguo Jin, PhD, Satya Kunapuli, PhD, *David Fiss, MD, Stephen H. Leech, MD, *Nelson M. Wolf, MD, *William A. Van Decker, MD
*Alexander Georgakis, MD, *Rasih Atilla Ener, MD, Jianguo Jin, PhD, Satya Kunapuli, PhD, *David Fiss, MD, Stephen H. Leech, MD, *Nelson M. Wolf, MD, *William A. Van Decker, MD

Contrast media plays an important role in defining coronary anatomy in diagnostic and interventional cardiac catheterization procedures as well as in radiology. Iodine-containing compounds have been shown to have radio-opaque properties that attenuate X-rays during radiographic examinations. The first organic contrast compound, lipiodol, was discovered in 1901, but was not used until years later in radiological studies.1 With the introduction of water-soluble compounds in later years, Dr. Seldinger was the first person who used contrast media in cardiac catheterization.2 At that time, sodium diatrizoate was the prototype of the tri-iodinated benzene ring compounds used in imaging.
All contrast agents generally possess the same structural composition of a tri-iodinated benzene ring (monomer) or 2 bound benzene rings (dimer) mixed with buffers and stabilizers, all of which can be classified either as ionic or nonionic. They can also be classified separately as high-osmolar, low-osmolar or iso-osmolar agents. More recently, their viscosity has emerged as one of the most important properties which, in turn, can be subdivided into low- and high-viscosity agents.
Early contrast agents were ionic, monomeric and high in osmolality, causing multiple side effects in the human body. But in an effort to improve their safety profile, particularly in an effort to reduce allergic reactions, hemodynamic side effects and contrast-induced nephropathy, nonionic and low-osmolar compounds were developed. Metrizamide was the first nonionic, monomeric, low-osmolar agent, but since it was unstable in solutions, other low-osmolar agents were later developed.3 More recently, a dimeric, nonionic, iso-osmolar agent has been introduced to the imaging world at the expense of increased viscosity and cost. Although the safety profile of these nonionic agents has been shown to be more tolerable, there have been several reports suggesting that they may have increased thrombogenicity. This increased thrombogenicity is of particular concern given the fact that these agents are used to diagnose and treat intracoronary thrombosis following plaque rupture/erosion, the cornerstone mechanism of acute coronary syndromes.4–11
All of these nonionic agents have either low-osmolar or iso-osmolar properties with different levels of viscosity. Among these, iodixanol has been classified as iso-osmolar. It is not clear whether differences in these properties may potentiate thrombosis and ultimately affect short- and long-term cardiovascular outcomes, which is particularly important in cardiovascular imaging.
In this study, we sought to investigate the differences in low-osmolar versus iso-osmolar nonionic contrast agents with regard to their thrombogenicity/platelet reactivity.


We studied 6 different nonionic radiocontrast agents (iodixanol is an iso-osmolar agent, whereas the rest are low-osmolar agents) at 10% and 50% concentrations in mixture with platelets to see the differences in their aggregation curves in response to various agonist stimulations. The initial 10% dilution ratio approximates the in vivo dilution of 500 cc of contrast media used in 5 liters of circulating blood of a patient undergoing a cardiac catheterization procedure. The 50% concentration simulates the moment in time when the vessel is being injected.
Normal whole blood is used from subjects with no known hematological disorder, coagulopathy, systemic infection or exposure to anticoagulant or antiplatelet agents. The blood samples are also tested for the following coagulation paramaters in comparison to normal values: prothrombin time (PT), activated partial thromboplastin time (aPTT), bleeding time, complete blood count (CBC), serum metabolic panel and lipid profile.
Six nonionic contrast aents that are marketed for radiographic imaging are used in this comparative in vitro study to assess platelet activation/aggregation properties. Among these, iodixanol is an iso-osmolar contrast agent. The others, even though they are misnomenclatured as low-osmolar agents (lower than old ionic high-osmolar agents, but still higher than plasma osmolality), have an approximately two-fold higher osmolality than plasma.

The following contrast agents are listed in the order of the analysis performed:

1. iodixanol (Visipaque™, GE Healthcare)
2. iohexol (Omnipaque™, GE Healthcare)
3. ioxilan (Oxilan®, Guerbet, LLC)
4. iopamidol (Isovue®, Bracco Diagnostics, Inc.)
5. iopromide (Ultravist, Bayer Schering Pharma, AG)
6. ioversol (Optiray, Mallinckrodt Medical, Inc.)

Experimental Protocol

Step 1. Isolation of platelets. Whole blood was drawn from healthy, normal, consenting human volunteers into tubes containing one-sixth volume of ACD (2.5 g of sodium citrate, 1.5 g of citric acid, and 2 g of glucose in 100 ml of deionized water) using a 14 gauge needle. Blood was centrifuged (Eppendorf 5810R centrifuge, Hamburg, Germany) at 230 x g for 20 minutes at room temperature to obtain platelet-rich plasma. Platelet-rich plasma (PRP), along with 1 mM acetylsalicylic acid, were incubated for 45 minutes at 37°C. The platelet-rich plasma was then centrifuged for 10 minutes at 980 x g at room temperature to pellet the platelets. Platelets were then resuspended in HEPES-Tyrode’s buffer in the concentrations noted below (138 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, 3 mM NaH2PO4, 5 mM glucose, 10 mM Hepes, pH 7.4, 0.2% bovine serum albumin) containing 0.05 unit/ml of apyrase. A 50% radiocontrast concentration was prepared using 250 µl of cells in mixture with 250 µl of intravenous contrast agent. A 10% concentration mixture was achieved using 250 µl of cells, 200 µl of Tyrode’s buffer, and 50 µl of intravenous contrast agent. As a control, 250 µl of cells in mixture with 250 µl Tyrode’s buffer is used. The platelet pellets in mixture with Tyrode’s buffer, along with 10%/50% concentrations of intravenous contrast agents, were then studied under a light aggregometer.
Aggregometry. Aggregation of 0.5 ml of washed platelets was analyzed using a PICA lumiaggregometer (Chrono-Log model 440S, Chrono-Log Corp., Havertown, Pennsylvania). Aggregation was measured using light transmission under stirring conditions at 37°C. Agonists were added simultaneously for platelet stimulation (AYPGKF 500 µM, 2 Mes ADP 100 nM, collagen 20 µg). Aggregometer output was recorded using a Kipp & Zonen type BD 12E flatbed chart recorder (SCI-TEC, Saskatoon, Canada) set at 0.2 mm/sec. The aggregation curves are reported (Figures 1–5).
Step 2. Intracellular calcium release. For platelet measurements, PRP was incubated with 2 µm fura-2 (to make the intracellularly-released calcium fluorescence) and 1 mM acetylsalicylic acid (to prevent thromboxane generation) for 45 minutes at 37ºC. Platelets were then isolated and resuspended in Tyrode’s buffer in mixture with different intravenous contrast agents at 50% concentration exactly as described above (since only 50% contrast concentration inhibited platelet aggregation, this concentration was used in this test), and a control was also prepared in a similar fashion. With the addition of three different agonists, changes in fluorescence were measured in terms of intracellular calcium release using an Aminco Bowman Series 2 luminescense spectrometer with a waterjacketed cuvette holder, equipped with a thermostat, at 37ºC and set at constant stirring. Sample volumes of 0.5 ml were analyzed with an excitation wavelength of 340 nm and an emission wavelength of 510 nm. Fluorescense measurements were converted to calcium concentrations using the equation reported by Grynkiewicz et al,12 where Fmin and Fmax were determined with each platelet preparation. The results are shown as summed/mean graphic depiction.


At 10% radiocontrast media concentrations, adenosine diphosphate (ADP) stimulation essentially resulted in similar aggregation curves compared to the control. Although the differences among the agents were less impressive, ioversol, iopamidol and ioxilan had better anti-agregogenic properties. At 50% concentration with ADP stimulation, all the contrast agents inhibited aggregation in a similar fashion, although iodixanol and ioxilan appeared to have more significant effects than the others.
At 10% concentration, AYPGKF stimulation caused more or less the same aggregation curves among all the contrast media compared to the control. Although the differences were minor, it appeared that ADP aggregation curves at 10% concentration. At 50% concentration with AYPGKF stimulation, all the contrast agents showed similarly inhibited aggregation, although iodixanol had a greater effect on inhibition than the other agents.
Although we did not perform the last test with 10% contrast concentration using collagen, it appeared to have a similar effect to that of ADP stimulation with 10% concentration. At 50% radiocontrast media concentration using collagen stimulation, all the contrast agents had similar inhibited aggregation, although iodixanol and ioxilan appeared to have slightly more prominent effects than the others.
In summary (after adjusting for baseline shifts), at 50% dye concentrations, all of the radiocontrast agents inhibited platelet aggregation (this inhibition was more pronounced with iodixanol, followed by ioxilan). At 10% dye concentration, all the radiocontrast media showed a response similar to the normal control with agonist stimulation (although the differences among them were less remarkable, iopamidol, ioversol and ioxilan had the greatest platelet inhibition effect at this concentration). It is not known whether these minor differences will translate into favorable clinical outcomes.
The second step of this study was performed in order to identify the mechanism of action in regard to aggregation inhibition at the 50% intravenous dye concentration. Therefore, we sought to investigate intracellular calcium release in the activation process of platelets. Using three different agonist stimulations at 50% radiocontrast media concentration, we observed inhibition of intracellular calcium mobilization with all the contrast agents (Figures 6, 7 and 8).


Radiocontrast agents are widely used today in radiologic imaging. Early studies with radiographic contrast agents showed that they inhibit blood coagulation and activate the complement system.13,14 Since Dr. Seldinger's utilization of radiocontrast media in coronary angiography beginning in 1956, much has been learned regarding these agents’ side effects. Early contrast media, which was ionic and high in osmolality (even though it had a platelet inhibitory effect) had a major side effect profile and is thus no longer used in the modern imaging era.15
In our study, we investigated the differences among the 6 nonionic radiocontrast agents currently used throughout the world with regard to their thrombogenicity and platelet reactivity. We chose 10% and 50% concentrations in order to simulate the steady-state condition (we use intravenous contrast agents in the approximate amount of 10% total blood volume in a typical coronary interventional case) and the moment of injection into the coronary vessel, respectively. ADP, AYPGKF (a protease activated receptor 4-activating peptide of the thrombin receptor) and collagen were used as stimulants.
All of the nonionic contrast agents revealed an inhibitory effect on platelets at 50% contrast concentration, which appeared to return to near normal (similar response to the control) at the 10% steady-state condition. There were only minor differences among the 6 contrast agents for which the translational clinical significance is unknown at present. These data also show that none of the nonionic contrast agents tested added additional thrombogenic risk in the setting of acute coronary syndromes in terms of platelet activation. If anything, there was a transient/reversible inhibition of aggregation, at least initially, as the vessel was being injected (at higher contrast concentration), which is a potentially desirable effect, especially in therapeutic interventions.
At 50% contrast concentration, all of the nonionic contrast agents inhibited aggregation, whereas at the 10% contrast concentration level, all agents showed similar aggregation curves compared to the normal control.
In the event of an acute coronary syndrome, at least initially and theoretically, iodixanol may offer some advantage over the other agents when injected into the vessel (50% concentration) since it showed the highest levels of aggregation inhibition according to the above-noted curves. As the steady-state condition occurred, this favorable anti-aggregatory effect appeared to be lost over time. Therefore, whether the use of this agent is justified despite its higher cost depends on patient characteristics and physician preference. The second most effective agent at 50% concentration using ADP and collagen stimulants (but not necessarily with AYPGKF) appeared to be ioxilan due to its platelet aggregation inhibitory effect. In the steady-state condition (10% radiocontrast concentration), although the differences among the agents were less remarkable, iopamidol, ioversol and ioxilan had the best anti-aggregatory effect.
It is interesting to note that all of the nonionic contrast agents had an inhibitory effect on calcium mobilization, which contributes to platelet shape change once the platelets are stimulated. Although some of the agonists are more dependent on calcium mobilization than others for their downstream signaling pathways (including the resultant platelet shape change), it was very surprising to observe that platelets still underwent shape change despite the inhibition of calcium mobilization with some of the intravenous contrast agents — the mechanism of which requires further exploration.

The mechanisms by which radiographic contrast media inhibit platelet function remain to be further identified. Some of the mechanisms that have been postulated include, but are not limited, to:15

1. Chelation of calcium (essential for platelet aggregation and secretion during stimulation of PRP) may possibly account for the inhibition of platelet function by some contrast agents;
2. impaired thromboxane formation;
3. ionicity versus nonionicity of contrast agent;
4. binding of these iodinated compounds to membrane proteins;
5. contrast media are known to avidly bind to plasma proteins, causing inhibition of contrast media-induced platelet activation, therefore, their presence or absence would substantially alter the results; 6. other mechanisms.

Study limitations. The primary limitation of our study is the lack of multiple controls, which could have affected test results due to interindividual variability. The second limitation is the absence of flow cytometric evaluation as a second control. Retrospectively, we noticed that it did not show fluorescence properties in the process of running the intracellular calcium release test. Therefore, it can still be a good study subject for future. Finally, it would have been interesting to study the specific pathway involved. Acknowledgements. We also acknowledge Michael McCool, RN, and John Gaughn, PhD, for their help.






  1. Bonnemain B, Guerber M. (The history of lipiodol 1901–1994, or how a medication may evolve with the times). Rev His Pharm 1995;42:159–170.
  2. Doby T. A tribute to Sven-Ivar Seldinger. Am J Roentgenol 1984;142:1–4.
  3. McClennan BL, Preston M. Hickey memorial lecture. Ionic and non-ionic iodinated contrast media: Evolution and strategies for use. Am J Roentgenol 1990;155:225–233.
  4. Esplugas E, Cequier A, Jara F, et al. Risk of thrombosis during coronary angioplasty with low osmolality contrast media. Am J Cardiol 1991;68:1020–1024.
  5. Hwang MH, Piao ZE, Murdock DK, et al. The potential risk of thrombosis during coronary angiography using nonionic contrast media. Cathet Cardiovasc Diagn 1989;16:209–213.
  6. Voeltz MD, Nelson MA, McDaniel MC, Manoukian SV. The important properties of contrast media: Focus on viscocity. J Invasive Cardiol 2007;19(Suppl A):1A–9A.
  7. Albanese JR, Venditto JA, Patel GC, Ambrose JA. Effects of ionic and nonionic contrast media on in vitro and in vivo platelet activation. Am J Cardiol 1995;76:1059–1063.
  8. LeFeuvre C, Batisse A, Collet JP, et al. Cardiac events after low osmolar ionic or isosmolar nonionic contrast media utilization in the current era of coronary angioplasty. Catheter Cardiovasc Interv 2006;67:852–858.
  9. Stormorken H. Sakariassen KS. In vitro platelet degranulation by contrast media: Clinical relevance? [comment] Circulation 1994;90:1580–1581.
  10. Chronos NA, Godall AH, Wilson DJ, et al. Profound platelet degranulation is an important side effect of some types of contrast media used in interventional cardiology [comment]. Circulation 1993;88:2035–2044.
  11. Dalby MC, Davidson SJ, Burman JF, et al. Systemic platelet effects of contrast media: Implications for cardiologic research and clinical practice. Am Heart J 2002;143: E1.
  12. Grynkiewicz G, Poenie M, Tsien RY. A new generation of Ca2+ indicators with greatly improved fluorescence properties. J Biol Chem 1985;260:3440–3450.
  13. Aquejouf O, Doutremepuich F, Azougagh OF, Doutremepuich C. Thrombogenicity of ionic and non-ionic contrast media tested in a laser induced rat thrombosis model. Thromb Res 1995;77:259.
  14. Ardissini D, Merlini PA, Coppola R, et al. Tissue factor antigen and activity in human coronary atherosclerotic plaques. Lancet 1997;349:769.
  15. Rao AK, Rao VM, Willis J, et al. Inhibition of platelet function by contrast media: Iopamidol and ioxaglate versus iothalamate. Radiology 1985;156;311–313.