[Physico-chemical and toxicological profile of gadolinium chelates as contrast agents for magnetic resonance imaging]. / Physico-chimie et profil toxicologique d'agents de contraste pour l'imagerie par résonance magnétique, les chélates de gadolinium.

Gadolinium chelates (GC) are contrast agents widely used to facilitate or to enable diagnosis using magnetic resonance imaging (MRI). From a regulatory viewpoint, GC are drugs. GC have largely contributed to the success of MRI, which has become a major component of clinician's diagnostic armamentarium. GC are not metabolised and are excreted by the kidneys. They distribute into the extracellular compartment. Because of its high intrinsic toxicity, gadolinium must be administered as a chelate. GC can be classified according to two key molecular features: (a) nature of the chelating moiety: either macrocyclic molecules in which gadolinium is caged in the pre-organized cavity of the ligand, or linear, open-chain molecules, (b) ionicity: Gd chelates can be ionic (meglumine or sodium salts) or non-ionic. The thermodynamic and kinetic stabilities of the various GCs differ according to these structural characteristics. The kinetic stability of macrocyclic GCs is much higher than that of linear GCs and the thermodynamic stability of ionic GCs is generally higher than that of non-ionic GC, thus leading to a lower risk of gadolinium dissociation. This class of drugs has enjoyed an excellent reputation in terms of safety for a long time, until a causal link with a recently-described serious disease, nephrogenic systemic fibrosis (NSF), was evidenced. It is acknowledged that the vast majority of NSF cases are related to the administration of some linear CG in renally-impaired patients. Health authorities, worldwide, released recommendations which drastically reduced the occurrence of new cases.

[In vivo imaging for biodistribution and metabolism evaluations of new chemical entities]. / Applications de l'imagerie pour l'étude de la biodistribution et du métabolisme de nouvelles molécules.

Due to numerous technical developments, in vivo imaging is suitable for pharmacokinetic and metabolism studies of new chemical entities as well as for evaluating their pharmacological or biological effects. MRI, nuclear medicine, X-Ray, ultrasound and optical imaging are available for both clinical and experimental imaging with even higher performance. For all these imaging modalities, diagnostic agents are useful to improve contrast and specificity. Specific targeting of biological events is addressed by molecular imaging. From a pharmacodynamic perspective, radiolabeling of a new chemical entity allows in vivo visualization quantitative measure of its biodistribution, its elimination and its specific molecular binding. Non-invasive imaging methods are useful for longitudinal investigations of biological changes. Based on nanotechnologies, specificity of drug delivery can be monitored by imaging. New developments in hybrid imaging technologies as well as multimodal contrast agents reinforce in vivo experimental and clinical proof of mechanism of new chemical entities.

Magnetic resonance imaging of atherosclerotic plaque with ultrasmall superparamagnetic particles of iron oxide in hyperlipidemic rabbits.

BACKGROUND: Based on the observation that ultrasmall superparamagnetic particles of iron oxides (USPIOs) are phagocytosed by cells of the mononuclear phagocytic system, the purpose of this study was to evaluate their use as a marker of atherosclerosis-associated inflammatory changes in the vessel wall before luminal narrowing is present. METHODS AND RESULTS: Experiments were conducted on 6 heritable hyperlipidemic and 3 New Zealand White rabbits. 3D MR angiography (MRA) of the thoracic aorta was performed on all rabbits by use of a conventional paramagnetic contrast agent that failed to reveal any abnormalities. One week later, all rabbits except 1 of the hyperlipidemic animals were injected with a USPIO contrast agent (Sinerem, Guerbet) at a dose of 1 mmol Fe/kg. 3D MRA data sets collected over the subsequent 5 days showed increasing signal in the aortic lumen. Whereas the aortic wall of the control rabbits remained smooth and bright, marked susceptibility effects became evident on day 4 within the aortic walls of hyperlipidemic rabbits. Ex vivo imaging of aortic specimens confirmed the in vivo results. Histopathology documented marked Fe uptake in macrophages embedded in atherosclerotic plaque of the hyperlipidemic rabbits. Electron microscopy showed multiple cytoplasmic Fe particles in macrophages. No such changes were seen in control rabbits or in the hyperlipidemic rabbit that had not received Sinerem. CONCLUSIONS: USPIOs are phagocytosed by macrophages in atherosclerotic plaques of the aortic wall of hyperlipidemic rabbits in a quantity sufficient to cause susceptibility effects detectable by MRI.

Angiography contrast agents interact with serine proteinases. The molecular structure of the model system elastase/iohexol.

Pancreatic elastase was co-crystallised with iohexol, a tri-iodo benzenic contrast agent used in angiography analyses. The X-ray analysis of the complex reveals the presence of three molecules of iohexol associated with the proteinase with low occupancy factors. Two iohexol molecules are located in and near the active site of the enzyme and provide a model for explaining the inhibition of the hemostatic system, one of the major and inconvenient side effect associated with these chemicals.

Evaluation of histamine release following intravenous injection of ionic and nonionic contrast media.

To evaluate histamine release (HR) following injection of radiocontrast media and to test the predictive value of peak-expiratory-flow rate (PEFR) in detecting high-risk patients, the authors performed in two series 90 intravenous pyelograms (IVPs). In the first group, HR was measured by a fluorometric method after injection of Hexabrix (25 patients) and Iopamiron (25 patients). In the second group, HR measured by a radioimmunoassay, and PEFR measured by a peak flowmeter were investigated after injection of Hexabrix (10 patients), Telebrix (10 patients), Omnipaque (10 patients) and Iopamiron (10 patients). Histamine release in groups 1 and 2, and PEFR in group 2, were not significantly modified by the injection of each radiocontrast media. For the four patients (two per group) who experienced minor allergic side effects, the levels of HR and PEFR were always within the normal ranges.

Pharmacokinetics of three gadolinium chelates with different molecular sizes shortly after intravenous injection in rabbits: relevance to MR angiography.

RATIONALE AND OBJECTIVES: To mimic an MR angiographic protocol, an experimental model was developed in rabbits to determine gadolinium concentrations in blood early after injection. Contrast agents with different molecular sizes were compared. METHODS: The influence of injection parameters (injection rate, duration of injection, volume of injection, dose) was assessed, and two blood pool agents (P760, P717) were compared with the low-molecular-weight Gd-DOTA under similar conditions. RESULTS: Two main phases were identified: bolus and post-bolus. Injection parameters strongly influenced the pattern of the bolus phase, but the postbolus phase was sensitive only to the injected dose. The blood pool agents presented a bolus phase profile similar to that of Gd-DOTA, but 45 seconds after injection, 84% of P717 molecules were still present in the blood compartment, compared with 65% of P760 molecules and only 51% of Gd-DOTA molecules. CONCLUSIONS: Blood pool agents are useful for MR angiographic protocols involving the postbolus phase. In protocols in which only the bolus phase is imaged, blood pool agents provide similar data to Gd-DOTA, provided that the injection rate, iso-T1 efficiency dose, and volume are similar.

Effect of iodinated contrast media on blood clotting.

Recently, blood clot formation in catheters used for the injection of nonionic contrast media (CM) during angiography has been reported as being due to activation of hemostasis in the catheter. However, CM exhibit inhibitory properties regarding coagulation and platelet functions. The effect on blood clotting of iohexol, iopamidol, ioxaglate, diatrizoate, and ioxitalamate at a ratio of 10% v/v with nonanticoagulated human whole blood was evaluated using the kinetics of fibrinopeptide A (FpA) generation. Blood aliquots were taken every 2 minutes until blood clot occurred. Two groups of contrast media were identified: (1) iohexol and iopamidol, which increased the clotting time, and (2) ioxaglate, diatrizoate, and ioxitalamate, for which all clotting times were over 30 minutes and no FpA generation occurred.

CT pulmonary angiography with a macromolecular contrast medium: a comparative study versus iobitridol in rabbits.

RATIONALE AND OBJECTIVES: To compare the pharmacokinetics of a new macromolecular iodinated contrast medium, prototype P743, with a standard contrast agent (iobitridol) for spiral computed tomography pulmonary angiography in rabbits. METHODS: Manual injection was first used to test the performance of P743 even in cases of nonoptimal bolus timing. Then a protocol was designed to compare vessel enhancement in both first-pass and delayed scans for the two contrast agents with the help of a power injector. RESULTS: With manual fast injection, the first pass of iobitridol was observed only on proximal scans. Conversely, opacification of vessels was maintained during three spiral scans with P743 under the same injection conditions. When optimal bolus timing was performed, higher vessel enhancement was observed during bolus first pass with iobitridol (iodine dosage 250 mg I/kg) compared with P743 (150 mg I/kg). However, during the postbolus phase, the decrease in attenuation values was markedly faster with iobitridol than with P743. CONCLUSIONS: This study confirmed that P743 remains more intravascular than iobitridol, which may have clinical implications for the diagnosis of pulmonary embolism, for example.

Experimental evaluation of the vascular effects and transport of an iodinated macromolecular contrast medium.

RATIONALE AND OBJECTIVES: For assessment of the tissue blood pool and overall vascularity, macromolecular contrast media have significant advantages over low molecular weight contrast agents. The authors evaluated the vascular effect and transport of a new macromolecular contrast media (MMCM), an iodinated dextran polymer of 32 kDa. METHODS: The new MMCM was obtained from dextran activated by carboxy methylation, followed by linkage with triiodinated aminophtalamid conjugates. To detect whether the tracer induces vascular leakage, MMCM (350 mg I/kg) was administered intravenously in 10 mice, or applied on the cremaster muscle of 26 mice previously injected with carbon particles; after 30 or 45 minutes, the cremaster was fixed and examined by optical microscopy. For investigation of the vascular transport 3, 5, and 15 minutes after MMCM administration, various tissue fragments were processed and examined by electron microscopy. RESULTS: In all vascular examined, MMCM does not induce plasma extravasation and the probe was detected mostly within the vascular lumen. At the ultrastructural level, a small fraction of MMCM was found in endothelial plasmalemmal vesicles (endosome-like structures) and, in time, transcytosed to the subendothelial space. No intercellular junctions were permeated by MMCM. CONCLUSIONS: The MMCM induces no vascular leakage and it is retained mainly in the plasma. Transport of MMCM is restricted to endothelial vesicles, which may explain, in part, its prolonged vascular space retention.

Pilot MR evaluation of pharmacokinetics and relaxivity of specific blood pool agents for MR angiography.

RATIONALE AND OBJECTIVES: To evaluate the use of two new blood pool contrast agents (P760, P775) compared with a low-molecular-weight gadolinium chelate in MR angiography. METHODS: The r1 efficiency of P760 was evaluated in vitro at 1.5 T; 3D abdominal contrast-enhanced MR angiography with qualitative analysis was compared in four rabbits after injection of incremental doses of P760 and in one rabbit after Gd-DOTA. A dynamic MR study was performed using a 2D T1-weighted turbo-flash MR sequence after injection of P760, P775, and Gd-DOTA. Each compound was tested at equivalent doses in three rabbits to assess r1 efficiency. Quantitative analysis of signal intensity in the aorta, the inferior vena cava, the renal cortex, and the medulla was performed. RESULTS: In vitro, the r1 efficiency of P760 was 23.3 mmol(-1) x L x sec(-1) at 1.5 T. Injection of a dose of P760 10 times less than Gd-DOTA allowed similar vessel visualization. The signal intensity peak and first-pass contrast kinetics in the aorta and the inferior vena cava were similar with the three products. Compared with P760 and Gd-DOTA, P775 allowed a greater renal cortex signal intensity at the first pass and a faster decrease on delayed images. CONCLUSIONS: The superior r1 efficiency of P760 and P775 was confirmed in vitro and in vivo at 1.5 T compared with Gd-DOTA, and P775 proved to be a rapid-clearance blood pool agent.

In vitro interactions of gadolinium DOTA meglumine and gadolinium DTPA meglumine on hemostasis.

The interactions of two gadolinium complexes (Gd-DOTA meglumine and Gd-DTPA meglumine) with hemostatic function have been analyzed using: (1) coagulation reactions (extrinsic and intrinsic pathways and fibrinoformation) and (2) platelet function investigations (aggregation, release of Ca++ and ATP after stimulation with collagen 2.5 micrograms/mL). The data obtained with Gd-DTPA meglumine (Mgl) exhibited a significant increase of the intrinsic coagulation pathway and a delay in fibrinoformation, although there is no alteration of the effect of thrombin on fibrinogen (fibrinopeptid A determinations). Platelet aggregation and release are moderately modified. In contrast, Gd-DOTA Mgl exerts no effect on the coagulation system and only minor effects on platelet functions. It is suggested that at least one mechanism involves the complexation of ionized calcium because Gd-DTPA Mgl and Gd-DOTA Mgl complex, respectively, 45% and 23% of ionized calcium in plasma. However, other mechanisms such as an alteration of fibrin polymerization are not unlikely.

In vitro evaluation of vascular permeability to contrast media using cultured endothelial cell monolayers.

OBJECTIVE: The efficiency of contrast agents in medical imaging depends on their distribution into vascular and interstitial compartments. The aim of this study was to compare in vitro endothelial permeability to different classes of contrast agents with various vascular persistence properties: a triiodinated nonionic monomer (ioversol), an iodinated dextran polymer (P604), and an iron oxide nanoparticle (sinerem). METHODS: Permeability studies, through collagen-coated filters with or without porcine aortic endothelial cell monolayer, were carried out by placing each filter-ring (luminal chamber) into a beaker containing a culture medium (abluminal chamber). Contrast media, diluted in the culture medium, were added to the luminal chamber. Aliquots were sampled from the abluminal chamber for contrast agent determinations. The volume cleared of the compound was calculated from the luminal side to the abluminal side. Parallel permeability tests to [3H]-H2O and Evans blue albumin were performed as references. Finally, the modulatory effect of bradykinin on endothelial permeability to albumin or to contrast agents was studied. RESULTS: The volume cleared of ioversol, P604, and sinerem through membrane filters was decreased by 19.6%, 32.1%, and 52.0%, respectively, in the presence of a cell monolayer. Bradykinin (10(-6) M) significantly increased permeability to albumin, ioversol, and sinerem. Ioversol and sinerem induced a significant decrease in permeability to albumin. CONCLUSIONS: A relation between the molecular size of the contrast agents tested and their endothelial permeability can be established with this in vitro model.

Physicochemical and biological evaluation of P792, a rapid-clearance blood-pool agent for magnetic resonance imaging.

RATIONALE AND OBJECTIVES: To summarize the physicochemical characterization, pharmacokinetic behavior, and biological evaluation of P792, a new monogadolinated MRI blood-pool agent. METHODS: The molecular modeling of P792 was described. The r1 relaxivity properties of P792 were measured in water and 4% human serum albumin at different magnetic fields (20, 40, 60 MHz). The stability of the gadolinium complex was assessed. The pharmacokinetic and biodistribution profiles were studied in rabbits. Renal tolerance in dehydrated rats undergoing selective intrarenal injection was evaluated. Hemodynamic safety in rats and in vitro histamine and leukotriene B4 release were also tested. RESULTS: The mean diameter of P792 is 50.5 A and the r1 relaxivity of this monogadolinium contrast agent is 29 L x mmol(-1) x s(-1) at 60 MHz. The stability of the gadolinium complex in transmetallation is excellent. The pharmacokinetic and biodistribution profiles are consistent with that of a rapid-clearance blood-pool agent: P792 is mainly excreted by glomerular filtration, and its diffusion across normal endothelium is limited. Renal and hemodynamic safety is comparable to that of the nonspecific agent gadolinium-tetraazacyclododecane tetraacetic acid. No histamine or leukotriene B4 release was found in RBL-2H3 isolated mastocytes. CONCLUSIONS: The relaxivity of P792 at clinical field is very high for a monogadolinium complex without protein binding. The pharmacokinetic and biodistribution profiles are consistent with those of a rapid-clearance blood-pool agent. Its initial safety profile is satisfactory. Experimental and clinical studies are underway to confirm the potential of P792 in MRI.

Preclinical profile of the monodisperse iodinated macromolecular blood pool agent P743.

RATIONALE AND OBJECTIVES: To summarize the chemical synthesis, physicochemical characterization, pharmacokinetic behavior, and biological evaluation of P743, a new macromolecular iodinated contrast medium. METHODS: The synthesis and molecular modeling of the iodinated macromolecule P743 are described. The pharmacokinetic profile was established in rabbits and rats. Acute toxicity in mice, renal tolerance in normal rabbits, and renal tolerance in uninephrectomized, dehydrated rats undergoing selective intrarenal injection was evaluated. In vitro permeability effects on isolated mastocytes and on the coagulation pathways were carried out. Computed tomography vascular imaging was performed after intravenous injection of P743 (300 mg I/kg) in rabbits and compared with the nonspecific nonionic agent iobitridol. RESULTS: P743 is a monodisperse, macromolecular iodinated contrast medium. In both rabbits and rats, P743 showed a pharmacokinetic profile consistent with that of a rapid-clearance blood-pool agent. Its diffusion through the endothelium was found to be low in vitro, thus confirming early confinement of this macromolecule, unlike nonspecific contrast media. In both species, P743 was excreted by glomerular filtration. Acute toxicity disclosed no mortality at the highest volume that could be injected into mice, leading to a median lethal dose greater than 8.9 g I/kg. Renal tolerance was found to be good in both euvolemic rabbits and uninephrectomized, dehydrated rats. No histamine or leukotriene B4 release was found on RBL-2H3 isolated mastocytes. P743 did not interfere with the coagulation pathways. Imaging experiments confirmed that P743 remains in the vascular compartment for a longer time than does iobitridol, thus allowing vascular enhancement that is twice as high as that of iobitridol in the recirculation phase. CONCLUSIONS: The pharmacokinetic and imaging profiles of P743, a new, monodisperse, macromolecular blood-pool iodinated contrast medium, were consistent with those of a rapid-clearance blood-pool agent. Its initial safety profile is satisfactory. Further experimental imaging studies are required to define the clinical interest in such molecules.

Involvement of protein solvation in the interaction between a contrast medium (iopamidol) and fibrinogen or lysozyme.

The interaction between proteins and a radiological commonly-used contrast medium (iopamidol) have been studied by calorimetry. When aqueous solutions of fibrinogen or of lysozyme (20 g/l) are mixed with an aqueous solution of iopamidol (1,3-5 benzendicardoxamid,N,N'-bis[2-hydroxy-1-(hydroxymethyl)ethyl]-5- [(2-hydroxy-1-oxopropyl)amino]-2,4,6-triiodo) in the clinical blood concentration range (26-485 mM), isothermal calorimetry reveals a weak endothermal interaction at a high concentration of iopamidol for both proteins. This endothermal effect does not appear to be due to direct protein-iopamidol association. Differential scanning calorimetry confirms the influence of iopamidol by the change in protein unfolding in the presence of contrast medium, and suggests alterations in the protein solvation as a mechanism. Dilution studies indicate that iopamidol can influence protein solvation even when water molecules are present in a molecular excess of 1000. The influence of iopamidol on the availability of water molecules and the absence of direct interaction with the protein molecules is shown by Raman spectroscopy of two amino acids in the presence of iopamidol. The spectrum of alanine is unchanged at any iopamidol concentration studied, whereas the spectrum lines due to the thiol group of cysteine are shifted in a manner consistent with altered solvation.

Thrombotic risk associated with the use of iodinated contrast media in interventional cardiology: pathophysiology and clinical aspects.

A review of the current knowledge of the anti-thrombotic properties of iodinated contrast media (CM) has been conducted. CM are classified according to their chemical structure, either ionic or non-ionic (monomeric or dimeric). Numerous in vitro and in vivo data show that, although all CM have anti-coagulant properties, ionic molecules are more potent than non-ionic and, furthermore, do not activate resting platelets, unlike non-ionic agents. These properties may lead to a decrease in thrombus formation during interventional procedures. Several clinical trials have shown that CM may play a role in the occurrence of acute thrombotic complications but also in delayed ischaemic events during interventional procedures. A recent meta-analysis showed that, compared to non-ionic monomers, ionic low-osmolar CM reduce the rate of coronary artery abrupt closure, but no significant difference was found with respect to ischaemic complications. Ionic CM lead to a lower deposit of thrombotic materials on catheters and guide-wires. To date, clinical data comparing ionic CM and non-ionic dimers are scarce, significantly heterogeneous and, unlike experimental data, they do not show differences between both classes of CM. Further studies are required to better understand the precise mechanisms of such interactions and to analyse the effect of CM when new antiplatelet agents or new procedures (stenting) are used, to comply with new clinical strategies.

Effects of non-ionic monomeric and dimeric iodinated contrast media on renal and systemic haemodynamics in rats.

Non-ionic dimeric contrast media (CM) are a new class of CM which are iso-osmolar with plasma. The aim of this study was to investigate their effects on systemic and renal haemodynamics. The non-ionic dimeric CM iodixanol and the non-ionic monomeric agent iobitridol (both at a dose of 1,600 mgI/kg) were compared in terms of their effects on systemic blood pressure (BP) and renal blood flow (RBF) in two strains of rats (Wistar and Sprague Dawley). Iodixanol significantly lowered BP in Wistar rats (-33 +/- 9% of baseline, 10 min post-injection, P < 0.001 vs. saline and iobitridol). Iobitridol had virtually no effect on BP. Iobitridol and iodixanol significantly decreased RBF. This effect was more marked following injection of the dimer rather than the monomer (iodixanol: -32 +/- 13% iobitridol: -20 +/- 4 of baseline at 16 min, P < 0.05). For both agents, RBF was still decreased 50 min following injection (iodixanol: -30 +/- 11%, and iobitridol: -20 +/- 5% of baseline). Iodixanol also decreased RBF in Sprague Dawley rats, while BP remained unchanged. This suggests that changes in BP/RBF autoregulation do not account for the renal haemodynamic effects of this agent. When measured 2 h following injection, the iodixanol-induced renal hypoperfusion was still detectable (-29% vs. saline-treated rats), although not significant (P = 0.06). This effect was no longer observed 4 h following injection. Increasing the saline infusion rate (18 mL/h vs. 2 mL/h) during the experiment did not significantly decrease the effects of iodixanol on BP and RBF in Wistar rats. In spite of its iso-osmolality, iodixanol, a non-ionic dimeric CM, depressed RBF and BP significantly more than iobitridol, a monomeric non-ionic agent, in Wistar rats. This effect was long-lasting and was not alleviated by increasing the hydration rate.

In vitro comparison of the effects of contrast media on coagulation and platelet activation.

The aim of this study was to compare the in vitro effects of different classes of contrast media on both the blood coagulation system and on platelet function. Global tests (APTT, TT) and FpA and F1 + 2 generation measurements showed that ioxaglate (ionic dimer) presents the highest anticoagulant potential. The anticoagulant effects of nonionic agents were less marked, iodixanol (nonionic dimer) being significantly less anticoagulant than iohexol (nonionic monomer). Major platelet activation was observed with release of PF4, serotonin and PDGF-AB when iohexol was incubated for 1 min in whole blood. Iodixanol showed no effect over the same period, while moderate platelet activation was observed after 30 min. Under the same experimental conditions, ioxaglate had no effect on platelets even after incubation for 30 min, whereas activation was observed with 9 g/l saline control at this time. Prevention of thrombin formation and platelet activation is only achieved with ioxaglate, the ionic dimer. These findings may be clinically important in the thrombotic environment of radiological procedures and may explain the increased thrombotic risks observed with nonionic agents in interventional procedures.

Effect of ioxaglate--an ionic low osmolar contrast medium--on fibrin polymerization in vitro.

Ioxaglate, an iodinated contrast agent, decreases the rates of fibrin clot formation induced by thrombin or reptilase. This effect is not related to an increase in the ionic strength of the medium since a specific control of equivalent composition does not induce such variation. The concentration of ioxaglate which led to a 50% decrease of the control clot turbidity induced by thrombin was 17.5 +/- 2 mM. Macroscopically, clots formed with ioxaglate were larger and less turbid than the isotonic control. An increase in fibrin fibre diameters and a decrease in their densities were observed. During the fibrin polymerization process, all the fibrinogen was converted into fibrin, as for both the control and ioxaglate quantitative analysis of clots and supernatants showed (1) an identical quantity of FpA in clot supernatants, (2) the same quantities of protein incorporated into clots, and (3) no trace of fibrin monomers in the clot supernatants. Furthermore, dissolution in urea of clots formed in the presence of ioxaglate occurred more rapidly than in the control. Total incorporation of fibrinogen into clots, associated with a decrease in clot turbidity, indicated the existence of a qualitative abnormality in the construction of the three-dimensional fibrin structure. Using differential scanning calorimetry, it was observed that the two domains (D and E) of fibrinogen were modified by ioxaglate, showing the absence of specificity in the interaction between ioxaglate and a particular domain.

Interference of magnetic resonance imaging contrast agents with the serum calcium measurement technique using colorimetric reagents.

A possible interaction between either linear (Gd-DTPA-BMA and Gd-DTPA) or macrocyclic (Gd-DOTA) gadolinium complexes used as magnetic resonance imaging (MRI) contrast agents and colorimetric technique reagents for the measurement of serum calcium was evaluated on human serum pools, and its mechanism was investigated by means of UV spectrometry and electro-spray ionization mass spectrometry (ESI-MS). The highest concentration tested was 2.5 mM (corresponding to a putative strictly intravascular distribution of the compound) and the lowest dose was 0.2 mM (i.e. about two elimination half lives). Serum calcium was dosed in duplicate by conventional colorimetric techniques involving o-cresol-phthalein complexone (OCP) or methylthymol blue (MTB) as reagents. No interference was detected when mixing Gd DOTA with serum, whatever the concentration. Gd DTPA (2.5 mM) did not interfere with the colorimetric technique either. Conversely, the Gd DTPA-BMA solution induced a concentration-related variation in apparent calcium levels. In the UV experiments, solutions of 2.5 mM MRI contrast media were mixed with OCP or MTB and UV absorption spectra were recorded between 400 and 800 nm. For Gd-DOTA/OCP and Gd-DOTA/MTB, no significant variations in the absorbance were detected. However, in the presence of Gd DTPA BMA, the absorbance of OCP and MTB showed substantial and immediate variations over time. The ESI-MS studies showed a complete displacement of Gd3+ ion in the case of Gd-DTPA BMA. In the presence of OCP, we observed the disappearance of Gd-DTPA BMA and the formation of the free ligand DTPA BMA and a new complex Gd OCP with an original stoichiometry of 2/2. Such a phenomenon did not occur in the case of Gd DOTA and Gd DTPA. The decomplexation of Gd-DTPA BMA in the presence of OCP can probably be explained by the weaker thermodynamic stability of Gd-DTPA BMA compared to that of Gd-DOTA and Gd DTPA.