Formation of toxic iodinated disinfection by-products from compounds used in medical imaging.

Iodinated X-ray contrast media (ICM) were investigated as a source of iodine in the formation of iodo-trihalomethane (iodo-THM) and iodo-acid disinfection byproducts (DBPs), both of which are highly genotoxic and/or cytotoxic in mammalian cells. ICM are widely used at medical centers to enable imaging of soft tissues (e.g., organs, veins, blood vessels) and are designed to be inert substances, with 95% eliminated in urine and feces unmetabolized within 24 h. ICM are not well removed in wastewater treatment plants, such that they have been found at elevated concentrations in rivers and streams (up to 100 µg/L). Naturally occurring iodide in source waters is believed to be a primary source of iodine in the formation of iodo-DBPs, but a previous 23-city iodo-DBP occurrence study also revealed appreciable levels of iodo-DBPs in some drinking waters that had very low or no detectable iodide in their source waters. When 10 of the original 23 cities' source waters were resampled, four ICM were found--iopamidol, iopromide, iohexol, and diatrizoate--with iopamidol most frequently detected, in 6 of the 10 plants sampled, with concentrations up to 2700 ng/L. Subsequent controlled laboratory reactions of iopamidol with aqueous chlorine and monochloramine in the absence of natural organic matter (NOM) produced only trace levels of iodo-DBPs; however, when reacted in real source waters (containing NOM), chlorine and monochloramine produced significant levels of iodo-THMs and iodo-acids, up to 212 nM for dichloroiodomethane and 3.0 nM for iodoacetic acid, respectively, for chlorination. The pH behavior was different for chlorine and monochloramine, such that iodo-DBP concentrations maximized at higher pH (8.5) for chlorine, but at lower pH (6.5) for monochloramine. Extracts from chloraminated source waters with and without iopamidol, as well as from chlorinated source waters with iopamidol, were the most cytotoxic samples in mammalian cells. Source waters with iopamidol but no disinfectant added were the least cytotoxic. While extracts from chlorinated and chloraminated source waters were genotoxic, the addition of iopamidol enhanced their genotoxicity. Therefore, while ICM are not toxic in themselves, their presence in source waters may be a source of concern because of the formation of highly toxic iodo-DBPs in chlorinated and chloraminated drinking water.

Nitric oxide and prostanoids protect the renal outer medulla from radiocontrast toxicity in the rat.

Human radiocontrast nephrotoxicity is predicted by the presence of multiple risk factors, often associated with compromised renal circulation. To produce a simple model of radiocontrast nephropathy, rats were pretreated with indomethacin and N omega-nitro-L-arginine methyl ester (L-NAME, to inhibit nitric oxide synthesis) before the administration of iothalamate. Acute renal failure consistently developed, with a decline in creatinine clearance from 1.05 +/- 0.10 to 0.27 +/- 0.05 ml/min (P < 0.001) associated with selective necrosis of 49 +/- 9% of medullary thick ascending limbs. Hemodynamic studies using laser-Doppler probes revealed that when injected alone, iothalamate increased outer medullary blood flow to 196 +/- 25% of baseline (P < 0.001). Pretreatment by L-NAME or indomethacin both reduced basal medullary blood flow and transformed the medullary vasodilator response to radiocontrast into vasoconstriction, with a prolonged reduction of medullary blood flow to less then half of baseline. Combined administration of indomethacin, L-NAME, and iothalamate lowered medullary blood flow to 12 +/- 4% of baseline. We conclude that prostanoids and nitric oxide have an important protective role in the renal response to radiocontrast material. Reduced synthesis of these vasoactive substances in renal/vascular diseases may predispose patients to radiocontrast nephropathy.

Variable effects of radiological contrast media on thrombus growth in a rabbit jugular vein thrombosis model.

We studied the effect of an ionic high osmolar contrast medium (Ioxitalamate), an ionic low osmolar contrast medium (Ioxaglate) and various nonionic low osmolar contrast media (Iopamidol, Iopromide and Iohexol) on thrombus growth in a rabbit jugular vein thrombosis model. Thrombus growth was determined by the accretion of 125I-labeled fibrinogen onto autologous preformed thrombi in rabbit jugular veins at various time-intervals from 15 min up to 10 h after infusion of the study solution. The ionic low osmolar contrast medium markedly inhibited thrombus growth whereas all nonionic low osmolar contrast media promoted thrombus growth. The ionic high osmolar, contrast medium inhibited thrombus growth, but less than the ionic low osmolar contrast medium. Within the group of nonionic contrast media, the Iopamidol associated promotion of thrombus growth was significantly higher than the Iopromide or Iohexol associated effects. The simultaneous administration of the apparently most potent thrombus growth promoting contrast medium (i.e. Iopamidol) and heparin resulted in complete abolishment of the increase in thrombus growth. These results support the claims of prothrombotic properties of nonionic as compared to ionic contrast media and could explain the clinically encountered thromboembolic complications after the use of nonionic low osmolar contrast media.

Pathogenesis of postmyelographic arachnoiditis.

The causal relationship between a myelogram with an aqueous contrast medium and postmyelographic arachnoiditis has not been clarified. Primates after myelography with metrizamide or meglumine iocarmate were studied with fluorescent microscopy, energy dispersion analysis and with scanning and transmission electron and light microscopy. All controls and, up to the eighth day after myelography, all treated animals had normal-appearing arachnoid membranes by microscopy. After eight days the treated animals had progressively more severe arachnoid fibrosis. Energy dispersive analysis revealed iodine in the arachnoid only up to 24 hours after myelography. Fluorescent microscopy revealed no evidence of immune complexes in the arachnoid. The chronic effects of water-soluble media on the arachnoid are not apparently mediated by contrast medium persisting in the arachnoid or by immune complexes.

Histologic effects of Amipaque (metrizamide) and various contrast media on mouse peritoneum.

The histologic effects of Amipaque and various commonly used contrast media on mouse peritoneum were studied by intraperitoneal injections of Amipaque 370 mg I/ml and 170 mg I/ml, Conray 300 mg I/ml, Conray Meglumine 282 mg I/ml, Urografin 60% and 76%, Gastrografin and barium sulfate. Amipaque and the other water-soluble media cause no deleterious effects, whereas barium sulfate evokes granulomatous inflammatory reaction. Amipaque is suitable for gastrointestinal studies when leakage into the peritoneal cavity is expected.

Experimental tissue damage after subcutaneous injection of water soluble contrast media.

Various water soluble contrast media (WSCM) were injected subcutaneously into 970 hind feet of 485 rats. Gross morphologic changes were seen after the injection and analyzed as a function of various physicochemical characteristics of WSCM. The WSCM of larger volume, higher osmolality, higher iodine content, and meglumine salts rather than sodium salts caused more severe tissue damage; younger rats showed more severe tissue damage by WSCM of high osmolality.

In vitro testing for the risk of arachnoiditis from myelographic contrast media.

The risk of arachnoiditis from aqueous myelographic contrast media has been assayed reliably only in experimental animals. The effect of contrast media on protein and collagen production by fibroblasts in vitro was studied. Iocarmate, metrizamide, and iopamidol added to the culture medium caused cells to produce more protein and collagen. The degree to which the contrast medium stimulated collagen production correlated with the risk of arachnoiditis from the intrathecal use of the contrast medium. In vitro testing appears to be an effective assay for arachnoiditis.

Ultrastructural effects of Conray 60 and Pantopaque on ependyma and choroid plexus following intraventricular injections.

Electron microscope evaluation of choroid plexus and ependyma following single cerebral intraventricular injections of Conray 60 and Pantopaque was carried out on 35 rats and 4 dogs. The animals were sacrificed at periods ranging from 1 hour to 4 months. Conray was not detected with the light or electron microscope; however, Pantopaque was presumptively localized as electron-dense masses associated with lipid-like bodies at the ventricular interface of both choroidal and ependymal epithelium. Conray 60 injections consistently induced convulsions in rats. Histological studies demonstrated moderate cellular damage and multilayered proliferation of ependymal epithelium. Morphological damage following Pantopaque was more severe and widespread in both choroid plexus and ependyma suggesting that, of the two agents, Conray may have the greater clinical potential provided that the associated convulsions are controlled by appropriate measures.

Scanning electron microscopic (SEM) studies on the effect of x-ray contrast media on aortic endothelium in the rat.

A study comparing the effects of diatrizoate, ioxitalamate, metrizamide, ioxaglate, and a 22% sorbitol on aortic endothelia of the rat was performed. In each case 1 ml of the respective solution was injected in a single dose into the aorta. Endothelial damage was seen after injection of metrizamide, diatrizoate, and ioxitalamate. After injection of ioxaglate or sorbitol, only minor endothelial changes were observed.

Contrast media induced pulmonary edema. Comparison of ionic and nonionic agents in an animal model.

High intravenous doses of diatrizoate are known to induce pulmonary edema in the rat. The newer generation of contrast media--nonionics and monovalent dimers--are considered less toxic than diatrizoate. In this study we evaluated the degree of pulmonary edema induced by a high dose (6 g I/kg) of these new agents and found that Ioxaglate produced higher lung weights than Renografin 60 and Iopamidol. Iohexol and Amipaque did not induce a significant degree of edema. The model used in this study demonstrates distinct differences in pulmonary toxicity among these new agents, when given in doses exceedingly higher than given in clinical practice.

Toxicity of X-ray contrast media in cell cultures.

Monolayer cell culture techniques for neuronal and glial cells and fibroblasts were used for the investigation of the direct toxicity of x-ray contrast media. The effect of various concentrations of sodium diatrizoate, methylglucamine salts of diatrizoate, iothalamate, iodamide, metrizamide, iocarmate, ioserate, and ioglycamate were studied. The intravenous biliary contrast medium--ioglycamate--was most toxic to all cell types. All the contrast media studied were somewhat toxic to neurons in the concentration of 50 mmol/l and showed variable glial toxicity. In equiosmolar concentrations the contrast media studied showed toxicity on neurons in the following order: iothalamate, diatrizoate, iodamide, ioserate, iocarmate, ioglycamate. Fresh metrizamide solution in equimolar concentrations showed the least toxic effects on neurons but interfered considerably with the growth and survival of glial cells. Metrizamide solution proved to be highly toxic to neurons when refrigerated for two weeks. Methylgucamine iocarmate produced a peculiar swelling of fibroblasts.

Release of serotonin from human platelets in vitro by radiographic contrast media.

Both ionic and nonionic, monomeric and dimeric contrast media were found to release serotonin from intact human platelets in vitro. The monomeric contrast media were compared at the concentration range of 25 mg I/ml. Iothalamate was the strongest and the statistically equal metrizamide iopamidol, and P-297 were the weakest releasers. Monomeric and dimeric contrast media were compared at concentration ranges of 50 and 100 mg I/ml. They ranked, in descending order of serotonin releasing potency: iodipamide, iothalamate, P-127, iopamidol, and a statistically indistinguishable group of the monoacid dimer P-286, the nonionic dimer ZK 74 435, and metrizamide. The capability of contrast media to release serotonin seems to be a composite result of their specific physical and molecular structural properties.

Effects of radiographic contrast media on monolayer cell cultures.

The relationship between iodine concentration, osmolality, and toxicity for nine different contrast media was studied. High osmolal conventional ionic contrast media (Na-metrizoate, Na-iothalamate, meglumine/Na-diatrizoate, meglumine-calcium-metrizoate) and the new low osmolal nonionic (Metrizamide, iopamidol, iohexol) and ionic dimer (Meglumine/Na-ioxaglate) contrast media were tested. Monolayer cell cultures of human cervical carcinoma in situ cells were used as a test system. The toxicity of contrast media on cell cultures was strongly dependent on the osmolality, and different contrast media with the same osmolality had about similar effects on the cell cultures. However, contrast media seem to have some additional and more specific effects since equiosmolal saline and mannitol were better tolerated. When the toxicity was related solely to iodine concentration it emerged that the new low osmolal contrast media were much better tolerated than the high osmolal conventional contrast media.

Comparative evaluation of the effects of an ionic vs. a nonionic contrast medium on the venous endothelium. Preliminary scanning electron microscopic observations.

Isolated segments of the right jugular veins of six mongrel dogs were exposed to solutions of 60% diatrizoate (3 dogs) and 60% iopamidol (3 dogs) in vivo. Normal blood flow was re-established after 3 minutes of exposure to the contrast material. The left jugular veins served as controls. Veins were harvested at 1, 24, and 48 hours and studied by light and scanning electron microscopy. Changes consisting of cellular swelling, denudation, platelet aggregation and fibrin deposition were uniform and prominent with diatrizoate. Response to iopamidol was minimal consisting only of cellular swelling. The study suggests that post-phlebographic thrombophlebitis may be reduced by the use of nonionic contrast materials.

Experimental nephrotoxicity of the radiocontrast agents iohexol, ioxaglate, and iothalamate. An in vitro and in vivo study.

The authors compared the renal toxicity of the low osmolality radiocontrast agents, iohexol and ioxaglate, and the ionic agent, iothalamate, at equivalent iodine dose, using experimental models in vitro and in vivo. In isolated perfused rat kidneys, all agents induced comparable biphasic hemodynamic changes, associated with similar declines in glomerular filtration rate (GFR) and tubular necrosis. In two different in vivo models (using multiple insults combined with the administration of radiocontrast), iothalamate appeared to induce more severe morphologic injury. Despite similar nephrotoxic potential in vitro, the newer radiocontrast agents, iohexol and ioxaglate, cause in vivo less renal injury than iothalamate in the experimental models.

Effects of radiographic contrast agents on spinal cord physiology.

Segmental reflexes in the spinal cords of cats anesthetized with chloralose were used to evaluate the neurophysiologic effects of radiographic contrast agents. The exposed lumbar spinal cord was bathed with concentrations of ionic and nonionic agents including saline, sodium meglumine diatrizoate, meglumine iothalamate, meglumine iocarmate, and metrizamide. The following responses were evaluated: flexor and extensor monosynaptic reflex; polysynaptic flexion reflexes; spontaneous ventral root activity. Hypertonic solutions generally produced a transient decrease in all reflex activity for up to 1 hour. Isotonic solutions produced no significant changes in the monosynaptic responses, but an increase in amplitude of polysynaptic responses, and increased spontaneous activity. The usual facilitory effects of flexion reflex on the flexor monosynaptic reflex were unchanged, but the expected inhibitory effect of flexion reflex on the extensor monosynaptic reflex was changed to excitatory. The relative ability to produced these effects was sodium meglumine diatrizoate greater than meglumine iothalamate greater than meglumine iocarmate greater than metrizamide.

Toxicity studies of coronary arteriographic media in dogs.

Sequential injections of different contrast material in the canine coronary artery were performed while ventricular pressure and EKG were recorded. Larger volumes of contrast material caused more fibrillation and more drop in pressure. Preliminary myocardial infarction increased pressure effects. Diatrizoate methylglucamine (Renografin-76) caused more fibrillation and lowering of ventricular pressure than iothalamate or metrizoate Coronar. Calcium lessened the ventricular pressure depression from iothalamate.

Contrast, coagulation, and fibrinolysis.

Some adverse clinical effects of intravascular radiologic contrast agents have been attributed to their interference with the normal hemostatic processes. This study compares the effects of the low osmolality agents with those of the conventional agents by in vitro studies of platelet function, fibrin formation, and fibrinolytic activation. In various degrees, all the contrast agents studied inhibit platelet aggregation and fibrin formation but show virtually no direct activation of fibrinolysis. The new low osmolality agents generally show lesser inhibitory effects on the hemostatic mechanisms. Some clinical implications are discussed.

Tolerability of hypertonic and isotonic contrast media injected intravenously. A comparative study in the dog.

We examined the use of isotonic and hypertonic contrast media injected intravenously in the dog from the standpoint of cardiovascular tolerance after right atrial injections performed at 2.56 and 5.12 g I/second. The parameters measured were lead II of the electrocardiograph, heart rate, pulmonary and abdominal arterial pressure, and aortic flow. Three contrast media, ioxitalamate, ioxaglate, and iopamidol (two ionic and one nonionic), were compared, either concentrated (32% iodine) or dilute and isotonic with plasma (ioxaglate 160 mg I/mL and iopamidol 128 mg I/mL). At an injection rate of 5.12 g I/second, iopamidol-128 showed lower electrophysiologic tolerability and caused a higher increase in aortic flow than ioxitalamate 160 or ioxaglate 160. These effects may explain the lower radiographic efficacy observed with iopamidol-128 in previous digital subtraction angiography studies.

Chemotoxicity of contrast media and clinical adverse effects: a review.

The clinical effects of contrast agents not only result from high osmolality, but also from their own specific pharmacology, which mediates chemotoxic effects. In this review, the chemotoxic effects of the new nonionic agent, iohexol, are compared with those of standard ionic and other low osmolality contrast agents, ionic and nonionic. Iohexol has the lowest chemotoxicity of any agent yet synthesized. Its low systemic toxicity is the combined result of low chemotoxicity and low osmolality. Mechanisms of severe adverse reactions are reviewed, including the views of Lasser and Lalli, and the view that emphasizes the importance of cardiotoxic and hemodynamic effects. It is concluded that whichever view is taken of the mechanisms of severe adverse reactions, the new nonionic agents are likely to be safer than the ionic agents now in use.