Toxicity Mechanisms of Gadolinium and Gadolinium-Based Contrast Agents—A Review

Gadolinium-based contrast agents (GBCAs) have been used for more than 30 years to improve magnetic resonance imaging, a crucial tool for medical diagnosis and treatment monitoring across multiple clinical settings. Studies have shown that exposure to GBCAs is associated with gadolinium release and tissue deposition that may cause short- and long-term toxicity in several organs, including the kidney, the main excretion organ of most GBCAs. Considering the increasing prevalence of chronic kidney disease worldwide and that most of the complications following GBCA exposure are associated with renal dysfunction, the mechanisms underlying GBCA toxicity, especially renal toxicity, are particularly important. A better understanding of the gadolinium mechanisms of toxicity may contribute to clarify the safety and/or potential risks associated with the use of GBCAs. In this work, a review of the recent literature concerning gadolinium and GBCA mechanisms of toxicity was performed.


Introduction
The paramagnetic properties of gadolinium (Gd (III)) has made it a crucial imaging aid tool for medical diagnosis and for treatment monitoring, across multiple clinical settings.
In the 1960s, the toxicity of Gd (III) in the salt form was reported in animal studies, limiting its use [1].The development of the first formulations of Gd (III) stabilized by chelating agents, in the late 1980s, renewed the applicability of Gd (III) as a contrast agent [1].
Contrast agents composed of chelated Gd (III), commonly referred to gadoliniumbased contrast agents (GBCAs), have been widely used in magnetic resonance imaging (MRI) for over three decades, supporting the diagnosis of tumors, central nervous system diseases, vascular diseases, bone marrow disorders, sclerosis, and cerebrovascular events [2], among other clinical conditions.Furthermore, recently, its potential applicability in the theranostic agents field has also been investigated [3,4].The design of GBCAs must consider that the release of Gd (III) from chelates should be low enough to be safe; thus, Gd (III) must attach firmly to a high affinity ligand to form a safe GBCA.By chelating Gd (III), its toxicity is reduced to a safe level, while maintaining the paramagnetic properties, which will increase the sensitivity and specificity of MRI diagnostic.
consider that the release of Gd (III) from chelates should be low enough to be safe; thus, Gd (III) must attach firmly to a high affinity ligand to form a safe GBCA.By chelating Gd (III), its toxicity is reduced to a safe level, while maintaining the paramagnetic properties, which will increase the sensitivity and specificity of MRI diagnostic.
Considering the chemical structure of the chelating molecule, GBCAs can be classified as linear or macrocyclic, depending on whether or not they have an open or an enclosing structure, respectively (Figure 1).Depending on their charge, they can be ionic, like the acidic GBCA, or non-ionic, like the chelating agents with amide or alcohol groups.Linear complexes are flexible open chains that do not bind robustly to Gd (III), while macrocyclic GBCAs, with pre-arranged rigid rings, present almost the ideal size to trap the ion, offering a stronger linkage to Gd (III).The development of macrocyclic chelates was prompted by the low stability of linear GBCAs.Indeed, Gd (III) dissociates more quickly and easily from linear chelates, leading to higher circulating levels and increased tissue uptake of free Gd (III), which may entail long-term disturbances in multiple organs [5].Studies with fibroblasts and macrophages showed that, following endosomal internalization into living cells, acyclic GBCAs are degraded much more rapidly than macrocyclic chelates [6].Accumulation of Gd (III) ion has been reported in kidney [7,8], brain [9], liver [10], skin [11], and bone tissue [12].Animal studies have shown that the amounts of Gd (III) retained in the organs are higher for linear GBCAs than for those with a macrocyclic structure [13][14][15].In postmortem studies of patients who died from nephrogenic systemic fibrosis (NSF), a clinical complication that can be observed in subjects with compromised renal function, after exposure(s) to GBCAs [7], Gd (III) was found in all analyzed tissues, showing very high levels in the kidney, heart, and blood vessels [16].The long-term retention of Gd (III) raises concerns about the safety of GBCAs, once the mobilization of such deposits may result in adverse events, with variable onset.A study conducted in an aquatic environment suggested that the chelating structure of the contrast agent may affect cell growth, also raising some concerns about the safety of the ligand [17].Most of the NSF reported cases were associated with the administration of non-ionic linear agents, such as gadodiamide and gadoversetamide, although some NSF cases have also been associated with gadopentetic acid, a linear ionic agent [18]; with macrocyclic GBCAs, there are less reported cases of NSF, and most of them are in gadolinium-exposed patients with renal insufficiency [19][20][21][22].
Reports of NSF occurrence in patients with advanced kidney disease exposed to GBCAs strengthened the concern on their nephrotoxicity [7,[19][20][21][22].A slower elimination of Gd (III), due to kidney dysfunction, increases the potential for Gd (III) accumulation in the kidney and other tissues [23].Accumulation of Gd (III) in the kidney, as well as in other organs, has also been reported in individuals without renal dysfunction, particularly in those submitted to repeated administrations of GBCAs [24].
Understanding the pathways involved in the toxicity of Gd (III) might help to clarify the clinical significance of its renal retention, allowing a more accurate assessment of the risks associated with GBCAs use.This review aims to identify, gather, and summarize the current scientific data available on Gd (III) and/or GBCAs mechanisms of toxicity.

Gd (III) Mechanisms of Toxicity
To understand the cellular and molecular mechanisms of action underlying the toxic effects of Gd (III) and/or GBCAs, we performed a bibliographic search, considering in vitro and in vivo mechanistic studies, in the databases PubMed, Scopus, and Web of Science.Keywords were specifically used for each database in order to retrieve all studies containing information on Gd (III) or GBCA exposure.From this search, only non-human experimental mechanistic studies were included in this review.
Table 1 summarizes, chronologically and alphabetically (first author's surname), the studies for Gd (III) mechanisms of toxicity deemed relevant for the purpose of this review.

Reference Study Design Main Findings
Badger et al., 1997 [109] Liver microsomes and hepatocytes isolated from control male and female Sprague Dawley rats and rats administered with a single dose of GdCl 3 GdCl 3 treatment reduced the activity of total hepatic microsomal cytochrome P450 and aniline hydroxylase; it also reduced the susceptibility of hepatocytes to the cytotoxicity induced by CCl 4 , but not by CdCl 2 Spencer et al., 1997 [110] Male and female Sprague Dawley rats treated with a single administration of GdCl 3 Deposition in capillary beds of the lung and kidney, and in the liver and spleen, with signs of necrosis in both organs; phagocytosis by the mononuclear phagocytic system was also observed Rai et al., 1996 [111] Rats treated with a single dose of GdCl 3 Distribution of Kupfer cells in the liver and changes in their phenotype towards a more pro-inflammatory one (↑TNFα, ↓IL-10) A total of 93 studies were herein reviewed (Table 1), including studies concerning exposure to non-chelated Gd (III) (n = 54) and/or to GBCAs (n = 51).Sixty-four of these involved in vitro studies, using either established cell lines, primary cultures, or isolated tissues exposed to Gd (III) or GBCAs; two studies assessed the cellular mechanisms in hepatic material isolated from rats administered with Gd (III); 22 studies were conducted in vivo, using different species of animals; and five studies included both in vitro and in vivo models.Of note, in 15 of the in vivo studies, the effect of repeated administrations was evaluated, and animal models of renal failure were used in 6 studies.
Gd (III) was shown to interfere with calcium homeostasis as well: competition of Gd (III) with calcium, needed for cellular processes, was highlighted as a potential mechanism of cytotoxicity [49,58,89,105,114,115].Promotion of calcium influx was also reported [79], along with inhibition of mitochondrial calcium-activated F 1 F O -ATPase and desensitization of the permeability transition pore to calcium by binding to F 1 [32], which is also in line with the reported mitochondrial dysfunction.Gd (III) may block calcium transport in tissues with a lower excretion rate, increasing toxicity; it may inhibit some enzymes that are activated by calcium, interfering with the reticuloendothelial system, as well as with other calcium-dependent biological processes [33,36].It can, also, disturb physiological processes, like contraction of smooth, skeletal, and cardiac muscles; transmission of nervous influx; and blood coagulation [116].Furthermore, cell culture studies have shown that Gd (III) may lead to abnormal calcification of several types of cultured cells, inducing calcium deposition [83].NSF may be, at least in part, a consequence of this alteration in the calcification process, which promotes hardening of the skin and fibrotic changes in other tissues and organs.
Other plausible mechanisms underlying Gd (III) toxicity include blockage of adenosine diphosphate and adenosine triphosphate (ATP) hydrolysis through stimulation of angiotensin II AT1 receptors [78]; inhibition of ATP-permeable channels [106]; interference with the epithelial Na + -channel's activity [63]; downregulation of RhoA, mTORC1, and mTORC2 proteins [31]; and inhibition on both inward and outward ionic current through Gd (III) accumulation at the binding site of the Na + -Ca 2+ exchanger protein that carries the current [104].It may also interfere with the mobilization of iron [29,67], as it has been associated with total iron-binding capacity (TIBC) decrease [75], and to increases in serum iron, ferritin [41,75,80], and transferrin saturation [75].It also has an effect on the differentiation of mononuclear cells into ferroportin-expressing fibrocytic cells [67] and the differentiation of mononuclear cells into collagen-secreting cells, with increased expression of iron metabolism proteins and of angiogenic and osteoblast-lineage markers [73].Iron involvement in Gd (III) toxicity is in line with the transmetallation theory, by which endogenous metals, like iron, zinc, copper, magnesium, or calcium, attract the ligand, replacing gadolinium with the release of free Gd (III), which may deposit in different tissues.Finally, metabolic dysfunction, affecting lipid metabolism [38,52], and glycolytic and redox metabolic pathways were also highlighted [81].Tubulin was pointed to as a potential Gdbinging protein, at least in the NIH-3T3 cells (mouse embryonic fibroblasts); this binding might inhibit the assembling of tubulin or depolymerize microtubules in cells [44].
The pathophysiology of NSF remains poorly clarified and appears to be independent of sex, race, or age [9].The dissociation of Gd (III) from Gd-chelates, which has been highlighted as the primary etiology, is more likely to occur in patients with renal dysfunction, who have a reduced excretion rate, allowing a longer retention that facilitates in vivo ion dissociation, when compared to those with normal renal function.
Few studies using animal models of renal failure have addressed the impact of kidney disease in the toxicity of GBCAs [11,35,57,67,75,76].Nonetheless, they suggested that, in the case of renal disease, GBCAs decreased renal function [75], triggered skin fibrosis [11,57], increased the number of fibrocytes (related to the oxidative stress environment) [76], enhanced the differentiation of mononuclear cells into ferroportin-expressing fibrocystic cells [67], produced renal tube vacuolization [11], and caused disturbances in iron metabolism and TBARS values [75], as well as increased neutrophil count and neutrophil elastase activity [35].
Accumulation of Gd (III) in the kidney, as well as in other organs, has also been described in individuals without renal dysfunction [24], particularly in individuals submitted to repeated administrations of GBCAs [24].According to Roberts et al., in subjects with normal renal function, exposure to large cumulative doses of GBCAs can lead to the deposition of Gd (III) in the skin and brain [117].The observation of renal damage and tissue accumulation of Gd (III) after GBCA exposure, in subjects without previous renal disease, suggests the involvement of other nephrotoxic mechanisms, beyond the decrease in Gd (III) elimination, due to impaired renal function.
Cell culture studies using Gd (III)-exposed HK-2 cells (human proximal tubular cell line) reported increased oxidative stress, mitochondrial dysfunction, cell death by apoptosis, switching to necrosis at higher Gd (III) levels, and autophagic activation.Disturbance of the lipid metabolism was also observed, with intracellular accumulation of lipid droplets and upregulation of genes related to both lipogenesis and lipolysis; moreover, increased expression of the modulators of various signaling pathways involved in the development and progression of renal disease, including inflammation, hypoxia, and fibrosis, were also detected, even at subtoxic concentrations [38].

Concerns about the Use of GBCAs
The ability of Gd (III) to be retained in body tissues following its detachment from linear GBCAs led the European Medicines Agency (EMA) to recommend a restriction in their use [118].Some linear structure contrast agents, namely gadodiamide and gadoversetamide, were suspended.According to the EMA, the use of gadoxetic and gadobenic acid should be restricted to liver MRIs, as they undergo biliary excretion, meeting an important diagnostic need; gadopentetic acid should be restricted to intra-articular administration for MRI of the joints, since the dose necessary for this exam is very low.Moreover, the EMA recommended the use of agents with a macrocyclic molecular structure (such as gadoteric acid, gadobutrol, and gadoteridol), at the lowest dose necessary for diagnosis, and only if this is not possible without resorting to contrast agents.
Although no restrictions were made for the use of macrocyclic GBCAs, a few human and animal studies have already demonstrated that their use leads to Gd (III) retention in body tissues [119], which was also reported in patients with normal renal function [120].Following the administration of macrocyclic GBCAs in rats, organ tissue (e.g., brain and renal, hepatic, and splenic tissues) deposition was observed [121].In adults and children, after multiple administrations of macrocyclic GBCAs, an increase in signal intensity on unenhanced T1-weighted magnetic resonance in the brain was detected [122][123][124][125], suggesting Gd (III) retention at this organ.Ex vivo analysis of brain and bone tissues from patients administered with the macrocyclic GBCAs, gadobutrol or gadoteridol, showed Gd (III) deposition [120].The development of NSF, following the use of macrocyclic GBCAs, has also been reported, although data is not always consensual [126].
Nevertheless, it was reported that both linear and macrocyclic GBCAs stimulated the expression of multiple type I interferon-regulated genes and of numerous chemokines, cytokines, and growth factors in normal human blood monocytes [98]; in addition, supernatants recovered from monocyte cell cultures exposed to both types of GBCAs stimulated the expression, in normal dermal fibroblasts, of types I and III collagen, fibronectin, and α-smooth muscle actin [72].Although gadodiamide, a linear GBCA, led to greater skin fibrosis and dermal cellularity than the macrocyclic gadoteridol, both led to renal proximal tubule vacuolization and increased fibronectin accumulation [11].In addition, kidneys showed a significantly higher Gd (III) content after administration of gadodiamide and of the macrocyclic gadobutrol, as compared to gadobenate dimeglumine administration [127].
Repeated exposure to the macrocyclic GBCAs gadoteric acid or gadobutrol caused elevation of oxidative stress and inflammation in the brain [39]; although neurotoxicity was more prominent for linear GBCAs, both linear and macrocyclic GBCAs triggered neuronal cell death, through activation of apoptosis [34].
In rat testis, gadoteric acid and gadodiamide induced apoptosis in the Leydig cells, increased serum calcium levels, and reduced testosterone levels [58].These GBCAs were also able to trigger hepatocellular necrosis and apoptosis, causing liver damage [10].
It should be taken in consideration that different profiles have been reported for macrocyclic GBCAs; for instance, gadobutrol appears to easily release Gd (III), while gadoterate meglumine showed the best performance concerning the complex stability [49]; gadoteric acid neurotoxic potential was found to be higher than that of gadobutrol [39].It is important to highlight that each GBCA has its own properties and its own behavior regarding in vivo retention or deposition.The results reported for one GBCA cannot be extrapolated for all GBCAs [128].For instance, gadoteric acid undergoes a much faster residual excretion from the body than linear GBCAs [129].Gadoteridol was found to be eliminated more rapidly from rat cerebellum, cerebrum, and skin, compared to gadoteric acid and to gadobutrol, in the first 5 weeks after repeated administration of these macrocyclic GBCAs, resulting in lower levels of retained Gd (III) in these tissues [130].The faster clearance of gadoteridol has been attributed to its lower viscosity, molecular weight, and osmolality [131].Even among macrocyclic GBCAs, there are differences in their clearance and in the amount of Gd (III) retention [130].
The concerns regarding GBCA safety are driving research to find other solutions, with better safety and pharmacokinetic profiles, improving their performance and/or reducing the administered dose; these novel agents may include macro-and supramolecular multimeric Gd (III) complexes (dendrimers, polymers, carbon nanostructures, micelles, and liposomes) [132].The encapsulation of Gd (III) into nanoparticulates is another approach considered to overcome the poor selective tissue labeling and localization associated with GBCAs [133].Smart radiotherapy biomaterials loaded with Gd-based nanoparticles were also investigated for use in MRI scans, revealing a great potential [134].The interest of liposomal formulations application in MRI has been increasing [135].For instance, the use of nanoliposomal Gd (III) did not present adverse effects on human-derived hepatocytelike HepaRG cells and macrophages, although in vitro studies are needed to evaluate its safety [136].
Another approach to reduce GBCAs toxicity is the association with antioxidant or metal chelation agents.A study, in rats with renal failure (5/6 nephrectomy), exposed to a single dose of gadoteric acid, showed that the effects in renal function improved by treatment with the antioxidant N-acetylcysteine [75].Renal failure (5/6 nephrectomy) mice exposed repeatedly to gadodiamide developed NSF, infiltration of ferroportin-expressing fibrocyte-like cells, and iron accumulation in the skin; these effects were less pronounced in the group treated with gadodiamide plus deferiprone [67], a metal chelating agent used in clinical practice to treat iron overload, able to avoid Gd (III) tissue deposition.In accordance, the addition to gadodiamide of the chelating agent DTPA (diethylenetriaminepentaacetic acid) reduced transmetallation of this GBCA [137], suggesting that the use of metal chelates may help to reduce, possibly even eliminate, Gd (III) retention by tissues.

Final Considerations
Considering the usefulness of contrast agents, the lack of safer alternatives to GBCAs and the higher prevalence of renal complication in GBCA-exposed patients, especially in patients with renal insufficiency when the incidence of chronic kidney disease is increasing worldwide, the studies on the molecular and cellular mechanisms underlying Gd (III) cytotoxicity for each GBCA, as well as their pharmacological effects, are warranted.
This review provides an overview of the available evidence regarding the toxicity mechanisms of Gd (III) and GBCAs determined using in vitro and in vivo models, providing scientific grounds for the development of counteracting therapeutic measures.
It is clear that, compared to GBCAs with macrocyclic structures, the linear GBCAs are more unstable and, thereby, have shown higher Gd (III) retention and cytotoxicity in the organs.
Cell cultures with macrophages and renal and endothelial cells demonstrate that GBCA toxicity seems to involve pro-inflammatory and pro-fibrotic mechanisms.Despite several studies involving cell cultures, fewer have tackled in vivo evaluation using animal models, particularly addressing renal function.Current available data indicate that single exposure to macrocyclic GBCAs seems safe in animals with normal renal function.However, the toxicity at long-term Gd (III) retention deserves more investigation, both in cases with normal and decreased renal function.
In patients with moderate/severe renal disease, GBCA exposure may further compromise renal function, but the effect in preexisting mild kidney disease is not so clear.Although there are studies reporting nephrotoxicity and impaired renal function associated with repeated administrations of GBCAs, the frequency of exposure used in most research studies poorly mimics the use of these agents in clinical practice, and some research studies were carried out in models of advanced stage of renal disease.Also, the use of different lengths of exposure to GBCAs makes the interpretation and comparison between studies difficult.The effect of repeated administrations in mild kidney disease using standardized exposures to contrast agents deserves further study.Finally, considering the increasing prevalence of chronic kidney disease worldwide and that most of the complications following GBCA exposure are associated with renal dysfunction, the mechanisms underlying GBCA toxicity, especially renal toxicity, need further research studies.

Figure 1 .
Figure 1.Chemical structures of linear and macrocyclic gadolinium-based contrast agents, their brand names, and registering pharmaceutical companies.Figure 1.Chemical structures of linear and macrocyclic gadolinium-based contrast agents, their brand names, and registering pharmaceutical companies.

Figure 1 .
Figure 1.Chemical structures of linear and macrocyclic gadolinium-based contrast agents, their brand names, and registering pharmaceutical companies.Figure 1.Chemical structures of linear and macrocyclic gadolinium-based contrast agents, their brand names, and registering pharmaceutical companies.

Table 1 .
In vitro and in vivo studies concerning gadolinium mechanisms of toxicity.
A reversible decrease in liver O 2 consumption, accompanied by a decline in MT cytochromes c1 and c