It is known that gadolinium-based contrast agents have found their widest application in MRI studies [1]. Although gadolinium is present in them as a chelate, one should bear in mind that the toxicity of this rare earth element in its free form can be compared to that of mercury and lead [2] and that the stability of gadolinium-based magnetic resonance contrast agents (MRCAs) varies and is determined by two major factors: 1) a chemical structure of a chelator; 2) a presence of some organic and non-organic ligands in the medium that can compete for binding to Gd³⁺ ions or a chelating compound thus facilitating Gd³⁺ release.

Using an unstable contrast agent can be life threatening for patients with impaired renal function since free gadolinium retains in tissues and can cause nephrogenic systemic fibrosis [3, 4, 5].

Recent studies demonstrated an increased intensity signal in such brain structures as globus pallidus and dentate nucleus on unenhanced T₁-weighted MR images in patients [6] or laboratory animals [7] who had received low stability linear MRCAs before, which is possibly related to Gd³⁺ depositing. After administration of high stability macrocyclic MRCAs, no such “residual” increased signal was observed. It is also known that gadolinium release from MRCAs depends on the presence of various ions in the surrounding medium [8]. Therefore, a complex study on how the above mentioned factors interact can shed some light on the dynamics of Gd³⁺ release from a chelate complex in various media, as well as estimate the risk of administering certain contrast agents to patients with renal insufficiency or conditions accompanied by increased zinc or calcium ions concentration in blood. Improving the stability of these contrast agents, as by means of adding a substance with strong chelating properties, is also important. Polyvinylpyrrolidone (PVP) with its chelating and detoxifying properties can be regarded as such a substance [9].

The aim of this study is to conduct the comparative analysis of the stability of Gd³⁺-based MRCAs in the presence of zinc ions, calcium ions and PVP in water, phosphate buffer solution and human serum.

METHODS

The following linear Gd3⁺-based MRCAs were studied: gadopentetate dimeglumine (Magnevist 0.5 М, Bayer, Germany); gadobenate dimeglumine (MultiHance 0.5 M, Bracco, Italy); sodium gadopentetate + PVP (Dipentast 0.125 М, Epidbiomed Group of Companies, OOO, Russia); gadopentetate-β-cyclodextrin (Cyclogadopentetate 0.125 М, Epidbiomed Group of Companies, OOO, Russia), and gadobutrol, a macrocyclic MRCA (Gadovist 1 М, Bayer, Germany)

Contrast agents stability was assessed by proton NMR relaxometry (Minispec mq 20, Bruker, Germany). Gadolinuim release from a chelate affects proton relaxation times in the medium [10]. T₁ relaxation time was measured since MR signal intensity depends on this parameter. Stability assays of the substances listed above were performed in distilled water (pH 6.0), phosphate buffer and blood serum (рН 7.4). In the experiments with zinc, stability of five MRCAs was assessed, i. e. gadopentetate dimeglumine, sodium gadopenetate with PVP, gadopentetate-β-cyclodextrin, gadobutrol and gadobenic acid, whereas in the experiments with calcium only gadopentate dimeglumine was involved.

To obtain 0.2 M phosphate buffer (pH 7.4), aqueous solutions of NaH₂PO₄ and Nа₂HPO₄ were prepared [11]. Blood serum was obtained from the patients of A. N. Ryzhikh State Scientific Centre for Coloproctology. All donors signed the informed consent to their biological material being used in the scientific research under the conditions of respecting their privacy and confidentiality. Blood was collected in sterile tubes with a clot activator and a barrier gel. Serum was obtained by centrifuging blood at 1200 g for 10 minutes and stored frozen at –20 °C for no more than 10 days. Prior to freezing, serum samples were tested for albumin concentration on Spotchem EZ SР-4430 clinical chemistry analyzer (Arkray Inc., Japan). Then the samples were diluted in phosphate buffer until albumin concentration of 10^–4 M (close to physiological) was obtained.

A 200 mM ZnCl aqueous solution (Komponent-reaktiv, Russia) was prepared by dissolving the weighted amount of 2.7 g in 100 ml distilled water. The final concentration of ZnCl₂ in the sample was 2 mM. While adjusting ZnCl₂ final concentration, we drew on the study by M. Taupitz et al. [12] that demonstrated the most illustrative results at this particular ZnCl₂ concentration. The concentration of the initial CaCl₂ aqueous solution (Komponent-reaktiv, Russia) was also 200 mM (2.2 g CaCl₂ in 100 ml distilled water), the final concentration in the sample was 2 mM. The initial aqueous solution of PVP (Kollidon 17 PF, BASF) was prepared by dissolving 500 mg PVP powder in 1 ml distilled water.

To assess the stability of the studied MRCAs, two samples were prepared simultaneously. The first sample was a 0.2 mM MRCA solution. Т₁ relaxation time of the 0.2 mM MRCA solution was measured at 40 °C (temperature value in the sample chamber of the MR relaxometer). Then a zinc chloride or calcium chloride solution was added to the sample until the final concentration of 2 mM was reached; then relaxation time was measured again. After that the sample was incubated in the thermostat at 40 °C; Т₁ measurements were repeated in 1, 2 and 24 hours. The second sample was similar to the first one, the difference being a PVP solution with a final concentration of 10mg/ml added to it after adding zinc chloride or calcium chloride. In the second sample relaxation time was measured at the same time points.

Within the framework of this study all experiments were repeated sixfold to improve the reliability of the results. Using Statistica 10 software, mean values and standard deviations were computed. Because of the normal distribution of the obtained data (in all cases of sample checks using the Kolmogorov–Smirnov test, the p-value was substantially higher than 0.05), a statistical significance of differences between the means was determined by Student’s t-test, the difference being significant with p < 0,05.

RESULTS

Effect of zinc ions on MRCAs stability

In distilled water T₁ longitudinal relaxation time of all linear MRCAs shortened by an average of 23–28 % (fig. 1) in the absence of PVP 24 hours after the addition of zinc chloride. In the gadopentetate dimeglumine sample T₁ value lowered by 25.7 ± 0.6 %, in the sodium gadopentetate sample — by 28.1 ± 0.7 %, in the Cyclogadopentetate sample (CGP)— by 22.0 ± 0.5 %, in the gadobenate dimeglumine sample— by 24.8 ± 0.4 %, respectively. For macrocyclic gadobutrol T₁ did not change significantly.

In phosphate buffer without PVP, T₁ of all linear MRCAs lowered by an average of 13–19 % 24 hours after the addition of zinc chloride. We observed a reduction in T₁ by 18.1 ± 0.7 % in the gadopentetate dimeglumine sample, a reduction by 19.3 ± 0.8 % in the sodium gadopentetate and PVP sample, a reduction by 12.8 ± 0.6 % in the CGP sample, a reduction by 15.9 ± 0.5 % in the gadobenic acid sample. In the gadobutrol sample T₁ did not undergo any significant alterations (fig. 2). The lowest T₁ value was observed 1 and 24 hours after the addition of zinc chloride as opposed to its immediate reduction in the previous series of experiments with distilled water being a medium.