Skip to content

Gadolinium Contrast Agents Comparison Essay

Methodologic Considerations

Early studies to compare contrast agents for diagnostic efficacy were designed exclusively as interindividual parallel group studies in which each patient (with varying disease entities) was randomly assigned to receive just 1 of the 2 contrast agents.16⇓⇓⇓⇓⇓⇓⇓⇓–25 Unfortunately, studies of this type are subject to wide interpatient and interlesion variability, resulting in disparate variations within each arm of the study. Such studies do not permit reliable demonstrations of relevant differences between study groups and agents. Indeed, such parallel group studies can only demonstrate equivalence between study groups for equivalent doses of different GBCAs and not superiority or inferiority, even among GBCAs with widely varying relaxivity values.24,25

Direct comparison among different agents is accomplished more objectively with an intraindividual crossover study in which each patient receives both GBCAs in random order in 2 identical MR imaging examinations separated by just a few days. Such studies are better designed to isolate the GBCA as the only variable being assessed. Thus, variations due to patient-, disease-, or examination-related factors are eliminated or, at least, minimized. In such studies, any differences in findings can be attributed solely to the GBCA because all other variables are identical for the 2 examinations.

Summarized below are the findings of several such intraindividual crossover studies performed in human subjects for CNS evaluation of various GBCAs. In organizing this discussion, we realized that there is no standardized definition for use of the term, “high relaxivity.” It is the opinion of these authors that “high relaxivity” should be defined not numerically alone, but rather by an objectively proved ability of an agent to deliver increased clinical utility as measured by clinically relevant increases in signal enhancement (rather than merely small but statistically significant signal increases) or, preferably, an objectively measured increase in lesion number or lesion extent compared with other “standard” GBCAs.

Steady-State Imaging.

To analyze the possible differential roles of relaxivity and concentration in imaging performance, findings are presented for studies that compared the following: 1) Dotarem, Magnevist, Omniscan, OptiMARK, and ProHance (which all possess similar [standard] relaxivity values and are formulated at standard [0.5 mol/L] concentrations), 2) Gadavist (the GBCA with slightly higher relaxivity and the highest concentration [1 mol/L]), and 3) MultiHance (the GBCA with the highest relaxivity and formulated at standard concentration).

1) Intraindividual Crossover Studies Comparing GBCAs with Standard Relaxivity/Standard Concentration (Dotarem, Magnevist, Omniscan, OptiMARK, ProHance).

To the authors' knowledge, Greco et al26 performed the only intraindividual crossover study that compared GBCAs with standard relaxivity/standard concentration. In that study, 2 blinded readers intraindividually compared Magnevist (r1-relaxivity at 1.5T: 3.9–4.1 L × mmol−1 × s−1)11,14 with ProHance (r1-relaxivity at 1.5T: 4.1 L × mmol−1 × s−1 measured in bovine plasma at 37°C)11 in 80 subjects for the presence of disease, degree of enhancement, location and number of lesions, and additional information gained (definition of lesion borders, improved visualization, distinction of edema, disease classification, determination of recurrent tumor, and so forth). Neither reader noted any significant differences in terms of GBCA preference (readers 1 and 2 preferred ProHance over Magnevist in 2 and 4 cases, respectively, and Magnevist over ProHance in 1 and 2 cases, respectively), and no differences were noted between agents in terms of the additional information provided on postcontrast images (On-line Table 2).

2) Intraindividual Crossover Studies Comparing a GBCA with Slightly Higher Relaxivity and High Concentration versus Standard Relaxivity/Standard Concentration Agents (Gadavist versus Either Dotarem, Magnevist, Omniscan, OptiMARK, or ProHance).

To date, 5 published reports have described intraindividual crossover comparisons of Gadavist with standard relaxivity GBCAs.27⇓⇓⇓–31 Three of these looked specifically at the potential benefit of Gadavist versus Magnevist27,28 or ProHance29 for the detection and visualization of cerebral metastases and concluded, in each case, that Gadavist is advantageous for lesion detection primarily because of improved lesion conspicuity (On-line Table 2). However, the conclusions in a study by Anzalone et al27 were based solely on subjective assessment of images from 27 patients by 2 neuroradiologists in consensus, but unfortunately, no quantitative assessment of lesion enhancement was reported. Kim et al28 reported improved quantitative enhancement (lesion/brain contrast-to-noise ratio [CNR]) with Gadavist, but this study was retrospective and only compared double (0.2 mmol/kg body weight) doses of Gadavist and Magnevist. Furthermore, GBCA administration was not random (all patients received Magnevist for the first examination and Gadavist for the second). In the third study in patients with brain metastases, Katakami et al29 evaluated a larger number of patients by using a prospective design and concluded that a single 0.1-mmol/kg dose of Gadavist is noninferior to a double 0.2-mmol/kg dose of ProHance for lesion detection. However, despite administering a single 0.1-mmol/kg dose of ProHance as part of the study design, no assessment of single-dose ProHance images was performed. Thus, it is not possible to say whether a single dose of ProHance would have proved noninferior to a single dose of Gadavist by using their study design, sample size, and assessment methodology.

In a more recent single-center study in 51 patients with either primary or secondary brain tumors, 2 blinded readers each preferred Gadavist to ProHance in more patients in terms of subjective “preference in contrast enhancement,” “overall preference,” and “preference in diagnostic quality.”30 However, differences in quantitative enhancement were inconsistent and sequence-dependent, with a higher SNR for Gadavist noted only on a second T1-FLASH sequence at approximately 10 minutes postinjection, with no differences between agents seen on T1-spin echo or MPRAGE sequences. Indeed, Bayer HealthCare (manufacturer of gadobutrol, Gadavist) reported to the FDA that the performances of 0.1-mmol/kg doses of Gadavist and ProHance for brain tumor imaging are similar.2,31 In a prospective multicenter phase III study performed in 419 patients for the FDA approval of Gadavist for CNS imaging, 3 blinded readers each reported similar contrast enhancement, lesion border delineation, and lesion internal morphology and a similar overall accuracy of diagnosis when these 2 agents were administered at an equivalent dose of 0.1 mmol/kg body weight. The conclusion of the study was that Gadavist is noninferior to ProHance.2,31

Another recent report presented findings from a study that prospectively compared single-dose Gadavist (r1-relaxivity: 4.7–5.2 L × mmol−1 × s−1 in plasma at 37°C11,14) with single-dose Dotarem (r1-relaxivity: 3.6 L × mmol−1 × s−1) in 136 patients with cerebral neoplastic enhancing lesions.32 In this study, significant preference for Gadavist compared with Dotarem was noted by 2 of 3 blinded readers for overall reader preference. However, none of the 3 readers considered Gadavist superior to Dotarem for lesion delineation, while only 1 blinded reader noted a minimally significant preference for Gadavist for the definition of lesion internal structure. Quantitatively, the percentage lesion enhancement following Gadavist was approximately 9% higher than that following Dotarem, as expected from the differences in their respective relaxivities, but this yielded no significant difference between the 2 agents for measured CNR. Most important, no differences in the number of lesions detected with either agent were observed.

3a) Intraindividual Crossover Studies Comparing the GBCA with the Highest Relaxivity/Standard Concentration versus GBCAs with Standard Relaxivity/Standard Concentration (MultiHance versus Dotarem, Magnevist, Omniscan, OptiMARK, or ProHance).

Numerous multicenter studies have compared MultiHance with standard GBCAs by using an intraindividual crossover study design with blinded image evaluation by fully independent experienced neuroradiologists.33⇓⇓⇓⇓⇓⇓⇓–41 All of these studies were designed to demonstrate superiority rather than noninferiority, and the findings of all concluded that MultiHance is significantly superior in terms of both qualitative enhancement (global diagnostic preference, lesion border delineation, definition of disease extent, visualization of lesion internal morphology, lesion contrast enhancement) and quantitative enhancement (CNR, lesion-to-background ratio) (On-line Table 2). In each of these studies, the authors concluded that the superiority of MultiHance was due to its higher r1 value.

3b) Intraindividual Crossover Study to Compare the Highest Relaxivity GBCA versus the GBCA Formulated at the Highest Concentration (MultiHance versus Gadavist).

A recent study by Seidl et al41 directly addressed the relative merits of high relaxivity versus high gadolinium concentration. In their randomized, double-blind, intraindividual crossover study, 123 patients each underwent 1 examination with 0.1-mmol/kg MultiHance and 1 examination with 0.1-mmol/kg Gadavist. Three blinded readers consistently demonstrated a highly significant (P < .0001) preference for MultiHance for all qualitative end points with good interreader agreement for all evaluations (On-line Table 2). In addition, significant superiority was noted for all quantitative assessments with a mean difference of approximately 22% in percentage lesion enhancement between MultiHance and Gadavist.

This study demonstrated that gadolinium concentration has little-to-no practical clinical impact on steady-state morphologic imaging and that at identical approved (0.1 mmol/kg) doses, the relaxivity of the GBCA is the dominant characteristic determining the degree of enhancement.

Perfusion Imaging.

Cerebral perfusion assessment by dynamic susceptibility contrast MR imaging is frequently used for evaluation of brain tumors, stroke, and degenerative diseases such as dementia. The technique is based on rapid intravenous injection of a GBCA and subsequent bolus tracking by using a fast susceptibility-weighted imaging sequence that uses the T2* relaxing properties of the GBCA. Following tracer kinetic modeling, parametric maps of mean transit time, regional cerebral blood volume (rCBV), and regional cerebral blood flow (rCBF) can be calculated by unfolding tissue concentration curves and the concentration curve of the feeding artery.

Compared with conventional morphologic (static) imaging, (dynamic) perfusion imaging is more dependent on the shape of the injected contrast bolus and thus on the rate at which GBCAs are injected. Additionally, higher administered GBCA concentration and higher relaxivity might each be beneficial in augmenting the signal loss associated with the first-pass contrast bolus through the tissues of interest.

Although interindividual parallel group studies have compared Gadavist with Magnevist at 1.5T,42,43 comparatively few intraindividual crossover studies have been performed to compare GBCAs for perfusion imaging. Those that have been performed have compared Gadavist with either Magnevist at 3T44 or MultiHance at 1.5T45 or 3T.46,47

1) Inter- and Intraindividual Crossover Studies Comparing GBCAs with Standard Relaxivity/High Concentration versus Standard Relaxivity/Standard Concentration (Gadavist versus Magnevist).

An interindividual parallel group comparison of Gadavist with Magnevist was first performed by Griffiths et al,42 who compared 10- and 20-mL injections of Gadavist with 20-mL injections of Magnevist (all at 5 mL/s, resulting in overall injection times of 2 and 4 seconds, respectively) in 6 groups of 6 patients (36 patients overall) at 1.5T to determine whether the higher Gd concentration of Gadavist was beneficial when using thinner sections (4 mm as opposed to 7 mm) for single-shot, gradient-recalled echo-planar imaging. They compared time-intensity curves calculated at regions of interest in the hemispheric white matter and thalamus in terms of maximum signal reduction (ie, the difference between mean baseline and minimum value on the time-intensity curve), full width at half minimum, and signal-to-noise measurements. No significant differences were found between 20 mL of Magnevist and 10 mL of Gadavist in terms of the maximum signal changes observed in either anatomic area and at either section thickness. On the other hand, the signal changes nearly doubled when 20-mL Gadavist was compared with 20-mL Magnevist (ie, when a 2-fold higher dose of Gadavist was used), indicating that the total amount (ie, dose) of Gd was the dominant factor in determining signal response rather than the administered concentration per se.

A second interindividual parallel group comparison of Gadavist and Magnevist at 1.5T was subsequently performed by the same group when investigating whether 2 gadolinium perfusion studies of the whole brain could be performed during the same table occupancy without degradation of the derived data in the second study.43 In this study, 12 patients each received 2 injections at a fixed rate of 5 mL/s of either 20-mL Magnevist (6 patients) or 10-mL Gadavist (6 patients), with each administration separated by 8 minutes. Although the study was not designed specifically to compare the 2 agents directly, the authors nevertheless showed no significant differences in either the maximum signal change or full width at half maximum with 10-mL Gadavist compared with 20-mL Magnevist.

A small-scale intraindividual crossover comparison of these 2 GBCAs was recently performed at 3T by Giesel et al44 in 11 patients (6 with intra-axial tumors, 5 with extra-axial tumors), who each underwent examinations with 5-mL Gadavist and 10-mL Magnevist by using a T2*-weighted, gradient recalled-echo, echo-planar technique. As in the studies by Griffiths et al,42,43 the injection rate for both agents was 5 mL/s. However, unlike Griffiths et al, the authors reported significantly higher maximal signal changes for Gadavist in both gray and white matter and noted that Gadavist depicted a larger number of “hot spots” (areas with higher blood perfusion in the tumor) on color-coded maps than Magnevist in most of the 6 intra-axial tumors. The authors concluded that the higher concentration of Gadavist offers advantages over standard-concentration Magnevist for delineation of gray and white matter and for the demarcation of highly vascularized tumor tissue and that these advantages are due to an improved bolus effect with increased intravascular concentration during the first pass.

2) Intraindividual Crossover Studies Comparing High Relaxivity/Standard Concentration versus Standard Relaxivity/High Concentration (MultiHance versus Gadavist).

Early intraindividual crossover studies to compare Gadavist and MultiHance were performed independently by Essig et al45 and Thilmann et al46 in healthy volunteers at 1.5T and 3T, respectively. In the study by Thilmann et al,46 16 healthy volunteers each underwent 3 DSC-MR imaging examinations separated by at least 3 days, receiving a single (0.1-mmol/kg; 7-mL) dose of Gadavist, a double (14-mL) dose of Gadavist, and a single (14-mL) dose of MultiHance, each at an injection rate of 5 mL/s (ie, resulting in injection times of 1.4, 2.8, and 2.8 seconds, respectively). Quantitative determinations based on signal intensity/time curves were made at regions of interest on gray and white matter and specific arteries selected for perfusion analysis. Additionally, gray-scale and color-coded maps of regional cerebral blood volume and regional cerebral blood flow were calculated and compared.

Quantitative analysis revealed nearly identical signal intensity/time curves for the 2 single-dose examinations. No differences were noted in terms of maximal relative signal drop, full width at half maximum, or signal-to-noise ratio of the concentration curve at maximum concentration. Likewise, qualitative evaluation of rCBV and rCBF maps by 2 experienced blinded radiologists revealed no differences between the 2 single-dose examinations with no advantage noted for either of the 2 GBCAs. More pronounced signal drops (52% versus 32%) and better quality perfusion maps (rCBV and rCBF) were obtained with double-dose Gadavist compared with either single-dose examination, though both single-dose examinations were considered suitable for diagnostic purposes.

More recently, Wirestam et al47 performed further evaluations of data acquired by Thilmann et al46 and confirmed that double-dose Gadavist results in higher absolute CBV, CBF, and mean transit time than single-dose Gadavist and that no significant differences are apparent between single-dose Gadavist and single-dose MultiHance.

Similar findings and conclusions to those of Thilmann et al46 were made by Essig et al45 in a study comprising 12 healthy male volunteers who each underwent 4 highly standardized perfusion MR imaging examinations with 0.1- and 0.2-mmol/kg doses of Gadavist and MultiHance, each administered at 5 mL/s. As in the study by Thilmann et al,46 a single dose of both agents was shown to be sufficient to achieve high-quality, diagnostically valid perfusion maps. Again, no differences were noted between single doses of the 2 agents for any quantitative parameter (susceptibility effect [percentage signal drops of approximately 30%], rCBV, and rCBF values) apart from full width at half maximum, which was significantly greater for MultiHance. Likewise, 2 off-site blinded readers found no significant differences between Gadavist and MultiHance in terms of image quality, adequacy of white-to-gray matter differentiation, or subjective preference for 1 agent or the other in terms of CBV and CBF image sets. Better overall image quality was noted with double (0.2 mmol/kg) doses of the 2 agents, for which a slightly higher susceptibility effect was seen with Gadavist. Nevertheless, the authors considered that double doses of the 2 agents provided no clinical benefit over single-dose examinations. The authors also concluded that single doses of both agents were effective at inducing sufficient signal drop on T2* EPI sequences to permit robust and reproducible quantification of perfusion parameters. Moreover, they concluded that the greater volume of injection of MultiHance had no disadvantage and gave comparable perfusion values to those obtained with the more highly concentrated Gadavist.

Contrast-Enhanced MR Angiography.

Similar to perfusion imaging, dynamic bolus contrast-enhanced MR angiography is a rapid imaging technique in which images are acquired during the first pass of a GBCA through the vessels of interest. However, unlike DSC perfusion imaging, the level of enhancement is dependent on the r1 of the agent rather than the r2* value. Accordingly, image quality and diagnostic performance are dependent not only on the image acquisition parameters but also on the contrast-injection protocol. Thus, while advances in sequence design can lead to marked improvements in the spatial and temporal resolution of vessel images, it remains fundamental that bolus timing and the peak concentration of intraluminal contrast coincide with the acquisition of the lower order phase-encoding steps of k-space image acquisition. To achieve MR angiograms with adequate homogeneous arterial contrast and without image artifacts, contrast bolus timing must achieve high concentration and a relatively constant plateau during acquisition of the central part of k-space, which contributes most of the image contrast. In addition, evidence appears to support the need to maintain a high level of Gd in the vessels during much of the higher order phase-encoding acquisition to minimize vessel edge blurring that can reduce vessel detail and visualization of smaller vessels.48

For the purposes of the present article, the focus will mainly be on studies comparing GBCAs for CE-MRA of the intracranial and supra-aortic vessels. However, the underlying principles of GBCA administration and image acquisition are common to all CE-MRA examinations across all vascular territories.

Intraindividual Crossover Studies Comparing GBCAs for CE-MRA of the Supra-Aortic Vessels

Of the few intraindividual crossover studies performed in the supra-aortic vessels, most have compared MultiHance with Magnevist.49⇓–51 In a very early study of 12 patients referred for CE-MRA of the carotid arteries, Pediconi et al49 compared a single 0.1-mmol/kg dose of MultiHance with a double (0.2-mmol/kg) dose of Magnevist and found superior quantitative and qualitative enhancement with MultiHance for carotid time-resolved CE-MRA. Both doses of GBCAs were administered at a fixed rate of 2 mL/s, and the better imaging performance was ascribed to the higher r1 of MultiHance. In that study, a single 0.1-mmol/kg dose of MultiHance would have been administered during 7.5 seconds for a 75-kg patient. Conversely, the double 0.2-mmol/kg dose of Magnevist would have been administered during 15 seconds, potentially resulting in exclusion of a portion of the increased Magnevist dose from the central part of the k-space during the MRA acquisition. On the other hand, the extended injection time for double-dose Magnevist would have provided double the window of opportunity to correctly “catch” the highest intraluminal GBCA concentration and may have contributed to better vessel wall sharpness, though this was not evaluated in the study. Empiric adjustments to optimize signal by using a specific pulse sequence, gadolinium agent, acquisition timing, and injection parameters are, therefore, critical in achieving best image quality.

A recent study by Li et al50 in 46 patients compared single-dose MultiHance and double-dose Magnevist. In this study, the 2-fold greater volume of Magnevist required to achieve a double dose was injected at a 2-fold faster rate to achieve comparable bolus geometry for the 2 examinations in each patient. Three blinded readers in the study found no differences between single-dose MultiHance and double-dose Magnevist for any qualitative parameter (vessel anatomic delineation, detection/exclusion of pathology, and global preference) or for quantitative measures of contrast enhancement (SNR, CNR). Indirect support for the findings of Li et al50 comes from a study by Bültmann et al,51 who compared single 0.1-mmol/kg doses of MultiHance and Magnevist across 19 arteries/arterial segments (comprising the internal carotid arteries; anterior, middle, and posterior cerebral arteries; vertebral arteries; and basilar artery) in 12 healthy volunteers at 3T. Maximum-intensity-projection images acquired with MultiHance were found to be markedly superior in terms of mean technical quality and vessel delineation to those acquired with Magnevist. Likewise the relative CNR was significantly greater with MultiHance, with overall increases of 23.3%, 26.7%, and 28.5% noted for the internal carotid, middle cerebral, and basilar arteries, respectively.

More recently, Kramer et al52 compared Gadavist with both MultiHance and Dotarem in 20 healthy volunteers at 3T. Although the total dose of each GBCA administered was 0.1 mmol/kg body weight, at variance with previous studies, the authors acquired both static CE-MRA and dynamic (time-resolved) CE-MRA images with 0.07 mmol/kg injected initially for the acquisition of static images followed by a further 0.03 mmol/kg for the acquisition of dynamic MRA. A fixed injection rate of 2 mL/s was used for both injections with all 3 GBCAs, and determinations were made of both quantitative and qualitative end points. Qualitative assessment of static images by 3 blinded readers found Gadavist to be not significantly different from MultiHance but superior to Dotarem, while few differences were noted between MultiHance and Dotarem. In terms of quantitative assessment of static images, a higher SNR with Gadavist was noted in the proximal ICA but not in the distal ICA, while the CNR with Gadavist was not significantly different from that with MultiHance but significantly superior to that with Dotarem. Similar findings were obtained for dynamic MRA. Finally, no differences were noted between the different GBCAs in terms of vessel sharpness.

The manner in which the contrast agents were administered for this study is not one routinely used in clinical practice. Nevertheless, it supports the advantages potentially gained with increased GBCA concentration if data acquisition can be appropriately timed to the shortened first pass of contrast bolus.

CE Accreditation Information

Date of Release:    1/15/14
Date of Expiration: 1/31/16
Estimated time to complete: 1 Hour

Target Audience

Radiologic Technologists

Learning Objectives

    Upon completion, participants should:
        • Understand the physicochemical similarities and differences among the available GBCAs
        • Understand how the chemical design of each agent affects its relaxivity
        • Understand the important practical issues of stability and safety

To claim credits, visit


Dr. Tweedle is a Stefanie Spielman Professor of Radiology at The Ohio State University, Columbus, OH; Dr. Kanal is Professor of Radiology and Neuroradiology at the University of Pittsburgh Medical Center, Pittsburgh, PA; and Dr. Muller is in the Department of General, Organic & Biochemical Chemistry, University of Mons, Mons, Belgium.


The use of gadolinium-based contrast agents (GBCAs) to enhance the sensitivity and specificity of magnetic resonance imaging (MRI) has been part of standard clinical practice for over 2 decades. Currently, there are 9 GBCAs approved by the U.S. Food and Drug Administration (FDA) that vary in a number of properties, some of which may significantly impact their clinical utility, particularly for specific applications. So what criteria should radiologists use to select a GBCA? Of the various physicochemical properties, the criteria most important to radiologists for optimization of diagnostic efficacy and patient safety are relaxivity and stability.

A GBCA with a higher relaxivity provides increased signal intensity, greater contrast enhancement, and improved diagnostic efficacy.1-3 In addition, the higher signal seen with higher-relaxivity agents affords the potential to use lower doses in patients at risk of developing nephrogenic systemic fibrosis (NSF).4-9 Stability is an important consideration because free gadolinium (Gd3+) is toxic and, therefore, the ability of the ligand to bind tightly to the Gd ion is an important safety consideration.10 Moreover, since 2006, when an association was made between the development of NSF and the administration of Gd,11 stability has become an even greater concern in the selection of a GBCA.

The 9 available GBCAs can be classified on the basis of relaxivity into 3 groups: macrocyclic standard relaxivity agents (including gadoterate meglumine [Dotarem],12 gadobutrol [Gadavist],13 and gadoteridol [ProHance]),14 linear standard relaxivity agents (gadopentetate dimeglumine [Magnevist]),15 gadodiamide [Omniscan],16 and gadoversetamide [OptiMARK]),17 and high-relaxivity agents, which interact with or overtly bind to proteins (the blood pool agent gadofosveset trisodium (Ablavar),18 the liver imaging agent gadoxetic acid (Eovist),19 and the multipurpose agent gadobenate dimeglumine (MultiHance)20 (Table 1). While all 3 of these high-relaxivity agents are linear in structure, there are some interesting observations regarding unconfounded NSF association – or lack thereof – with these agents that will be covered below.

Here we describe the differences among these 3 groups of GBCAs, including the advantages and disadvantages of their physicochemical profiles as they pertain to the selection of a GBCA. In addition, we summarize the published evidence surrounding the efficacy and safety of these 3 groups of agents. Finally, we discuss considerations when adopting a new GBCA for a radiology practice in order to maximize diagnostic yield while minimizing patient risk.

Macrocyclic GBCAs

Three macrocyclic agents are currently approved by the FDA: gadoterate meglumine (Dotarem), gadobutrol (Gadavist), and gadoteridol (ProHance). In terms of relaxivity, all are nonprotein binding and therefore standard relaxivity, with r1 relaxivities in the range of 4 to 5 L·mmol-1s-1.21 In terms of stability, macrocyclic agents utilize a closed, cage-like ligand that surrounds and binds tightly to the Gd ion, resulting in highly stable complexes (Table 2).10 Many published studies exist to support the high stability of these macrocyclic agents relative to the linear agents and, in terms of clinical value, these data can be considered to exist within a hierarchy, from in vitro, test-tube data at the bottom all the way up to human in vivo data at the top (Figure 1). When assessing the various data, it is important to recognize that higher-level data (eg, clinical experience) always trump data lower on the hierarchy (eg, test tube numbers).

Test-tube data derived from GBCAs in solution demonstrate that as a class, the macrocyclic agents have high thermodynamic binding constants (log Ktherm and log Kcond [essentially log Ktherm measured at physiologic pH]) (Table 2).10 In vivo, the dissociation of GBCAs into gadolinium ion and ligand can be facilitated by a number of competing endogenous metals, such as zinc, copper, calcium, and iron, all of which may work simultaneously to destabilize the complex and lead to its dissociation. This displacement of the gadolinium ion from its ligand by other metals through competitive ionic binding is termed transmetallation (or dissociation or dechelation), and transmetallation has been studied both in vivo and in vitro. In vitro, copper and zinc ion stress test data demonstrate that in the presence of these competitors, the macrocyclic agents gadoterate meglumine (Dotarem) and gadoteridol (ProHance) remain essentially intact (<1% reaction), while the less stable linear agents gadodiamide (Omniscan) and gadopentetate dimeglumine (Magnevist) are more highly reactive.22 In vivo data from mice and rats also demonstrate that up to 14 days after injection of radiolabeled GBCAs, the lowest levels of residual Gd are observed with the macrocyclic agents23 (Figure 2).

Human ex vivo data also demonstrate that transmetallation can occur with the low stability, linear agents gadodiamide (Omniscan) and gadoversetamide (OptiMARK). In this case, the data come from the observation that these 2 agents (and not the more stable agents) interfere in the colorimetric serum assay for calcium.24 The mechanism for this interference is believed to be removal of the Gd from the ligand-Gd complex by the specific dye used in the assay, preventing the dye from binding calcium.24 The result is a spurious hypocalcemia result for the patient, with potentially harmful clinical consequences. More recently, it was demonstrated that up to 20% of the nonionic linear GBCAs dissociate in human plasma up to 2 weeks after administration, while for the ionic linear agents and the macrocyclics, that number was closer to 2% and 0%, respectively (Figure 3).25 Human in vivo data, the most credible data, also exist to support the high stability of the macrocyclic agents: White and colleagues demonstrated 4 times more Gd3+ deposited in the bone of hip replacement patients after administration of gadodiamide (Omniscan) vs gadoteridol (ProHance).26

Taken together, all of these data suggest that transmetallation occurs both ex vivo and in vivo, and that with low stability linear GBCAs, there is more marked displacement of the gadolinium ion from its ligand by other metals. It is notable that the only GBCA with a triple dose indication is one of the macrocyclic agents, gadoteridol (ProHance)14 (the approval for triple dose indication for the linear agent gadodiamide [Omniscan] was withdrawn by the FDA in December, 2010).

It should also be noted that Pietsch and colleagues recently used a sensitive in vivo animal model to derive elimination time-courses for Gd in the skin of rats and found significantly higher nmolGd/g of skin at both Day 35 and Day 364 for the less stable agents compared with the macrocyclic agents (Table 3).27 However, a very small amount of Gd was still present in the skin of rats administered a macrocyclic agent on day 364, indicating that although the risk of dissociation with macrocyclic GBCAs is very low, this “low risk” is not equivalent to “no risk” and therefore, in patients at highest risk of NSF, it is still best to carefully assess the risk vs the benefit of injecting any GBCA.

Standard- and High-Relaxivity Linear Agents

The linear agents can be divided into nonprotein binding, standard-relaxivity GBCAs and protein interacting/binding, high-relaxivity agents. Tables 2 and 4, respectively, show the stability and r1 relaxivity measurements for the various linear agents. The thermodynamic stability constant (Ktherm) is a measure of stability: a lower thermodynamic stability constant indicates that the Gd3+ ion will be more readily released.28 However, the thermodynamic stability constant does not take pH into account. The conditional stability constant (Kcond) is a measure of the stability of a complex at physiologic pH and, therefore, Kcond is considered a more relevant stability parameter. Note that the nonionic linear GBCAs gadodiamide (Omniscan) and gadoversetamide (OptiMARK) have the lowest conditional stability constants, and this low stability is believed to be related to the higher prevalence of NSF cases associated with these GBCAs.29 Based on in vitro Zn2+ transmetallation data (Figure 4), gadodiamide (Omniscan) has lower stability and gadobenate dimeglumine (MultiHance) has higher stability compared with gadopentetate dimeglumine (Magnevist).30 However, based on higher level, human ex vivo data, there is essentially no difference in the amounts of Gd3+ released among the ionic linear agents gadobenate dimeglumine (MultiHance), gadopentetate dimeglumine (Magnevist), and gadofosveset trisodium (Ablavar) when measured in native human serum at 37°C (Figure 3).25

How does protein interaction influence relaxivity? Interaction with serum proteins, most notably human serum albumin (HSA), greatly increases the effective size of the Gd-ligand complex, reducing its molecular motion, leading to greater T1 shortening and substantially increased signal-intensity enhancement.31 Gadofosveset trisodium (Ablavar), gadoxetate meglumine (Eovist), and gadobenate dimeglumine (MultiHance) all interact with HSA, the first strongly and the last 2 weakly.30

Figure 5 shows the relaxivity profiles of solutions containing 1mM of either gadobutrol (Gadavist) or gadobenate dimeglumine (MultiHance) in the absence and presence 4% HSA.30 Note that for gadobutrol (Gadavist), as well as for all other nonprotein binding GBCAs, the presence of HSA has no effect on the relaxivity profiles. On the other hand, for gadobenate dimeglumine (MultiHance), the presence of HSA results in a notable spike in measured relaxivity precisely in the range (≈10-150 MHz) of magnetic field strengths used in clinical practice (0.47T = 20 MHz; 3T = 127.5 MHz). It is this protein interaction, and resulting increased relaxivity, that likely explains results of intraindividual crossover studies demonstrating greater signal intensity and diagnostic performance for the high-relaxivity multipurpose GBCA gadobenate dimeglumine (MultiHance) vs other GBCAs in CNS,1-3 breast,32 and liver,33 as well as for MRA of various vascular territories.34-38

Clinical Considerations in GBCA Selection

Returning now to our 3 groups of agents, macrocyclic agents with standard relaxivity, linear agents with standard relaxivity, and linear agents with higher relaxivity, how does one select a GBCA for routine use applications?


It has been well established that greater relaxivity leads to higher signal intensity enhancement. Of the multipurpose linear agents, gadobenate dimeglumine (MultiHance) has the highest r1 relaxivity at 1.5T, and this high relaxivity persists at 3T.39 Tables 5 and 6 summarize results from the most recent, robust intraindividual crossover studies comparing the efficacy of gadobenate dimeglumine (MultiHance) with standard relaxivity agents for MR imaging of the CNS (Table 5) and vasculature (Table 6). Examples of image pairs from published comparative studies are shown in Figure 6. From this body of evidence, one can conclude that at typical so-called T1-weighted imaging approaches, a single dose of gadobenate dimeglumine (MultiHance) is very roughly diagnostically equivalent to a double dose of a standard relaxivity agent, and that it provides better qualitative and quantitative diagnostic MR images when compared to an equivalent dose of a standard-relaxivity agent, for a variety of applications.


Regarding safety profiles, all of the GBCAs are considered to have low and comparable acute adverse event (AE) rates, regardless of whether the data are derived from product package inserts,12-20 from clinical trial safety data,40 or from professional society guidelines.41 The vast majority of AEs are mild, with the most common being transient injection site discomfort, nausea with or without vomiting, headache, paresthesia, dizziness, and itching.41

Upon adoption of a new drug, reported AE rates may go up initially but over time, tend to return to baseline. This was first observed by Weber in 1984 and has thus been termed the Weber effect.42 Davenport and colleagues observed the Weber effect at their institution when switching from gadopentetate dimeglumine (Magnevist) to gadobenate dimeglumine (MultiHance).43 Specifically, they noted that the reaction rates for gadobenate dimeglumine (MultiHance) peaked in the second year after it replaced gadopentetate dimeglumine (Magnevist), and then declined in the third year, ultimately returning to a baseline rate that did not differ significantly from the original baseline rate of gadopentetate dimeglumine (Magnevist) (Figure 7). The reasons for the Weber effect are complex and not completely understood, but it is generally believed to be related to unfamiliarity with a new drug, or new contrast agent in this case, and/or concern from staff regarding the change.

Recently, the incidence of AEs was prospectively studied at 10 sites associated with the University of Pittsburgh Medical Center (UPMC).44 A summary of the findings is shown in Table 7 and Figure 8. Most noteworthy is the gradual decline in AE rates over time, indicative of the Weber effect. Specifically, for data collected from 2005 through 2006, the overall AE rate was 178/23,553 (0.8%), but when the data collection was extended through 2011, the overall AE rate declined to 474/125,308 (< 0.4%).45 The incidence of serious AEs from this prospective study with gadobenate dimeglumine (MultiHance) was much lower (8/23,553 [0.03%]) and similar to the incidence of serious AEs with gadopentetate dimeglumine (Magnevist; 1/4,892 [0.02%]), obtained retrospectively, consistent with a lack of Weber effect for serious AEs. In addition, during this same time period, the incidence of anaphylactoid events at UPMC was 0.004% or approximately 4/100,000, a rate that is much lower than that observed with either high-osmolar iodinated contrast media (≈100/100,000) or low-osmolar iodinated contrast media (≈20/100,000).29

The most significant potential nonacute event associated with Gd exposure remains NSF. The factors most closely associated with the development of NSF include severe renal failure in the patient, exposure to multiple/high doses of a GBCA, and administration of a less stable GBCA.41,46 However, the majority of patients with severe renal failure administered multiple and/or high doses of a relatively unstable GBCA do not develop NSF, so it is likely that there remain unidentified factors that contribute to the development of NSF.

Per the American College of Radiology, GBCAs can be grouped into 3 groups according to their NSF risk (Figure 9).41 These groupings are based on the numbers of unconfounded, single-agent cases of NSF recorded for each agent. However, it is also interesting to examine the number of unconfounded NSF cases and the number of doses distributed, but to do so using the grouping system introduced in the introduction (Figure 10). When one does so, there are 2 apparent observations: first, and not surprisingly, the incidence of NSF with the macrocyclic agents is extremely low considering the number of administered doses and, second, that the number of NSF cases is surprisingly low in the group of intermediate-stability, protein interacting GBCAs, including gadofosveset trisodium (Ablavar), gadoxetate meglumine (Eovist), and gadobenate dimeglumine (MultiHance), considering their number of administered doses, suggesting perhaps that there is something protective about protein binding that mitigates or prevents the development of NSF. Any evidence for such protection, or any possible mechanism, remains to be established.


Chemical stability and in vivo relaxivity are the GBCA properties most relevant for selection of a contrast agent for MRI. Agents with a macrocyclic structure are the most highly stable and, as they do not bind to serum proteins, they all possess similar, relatively low relaxivity. Relatively lower stability linear agents may be of standard or high relaxivity. No cases of NSF have been reported after the prior unconfounded administration of any of the high-relaxivity protein interacting agents, the vascular imaging agent gadofosveset trisodium (Ablavar), the hepatic imaging agent gadoxetate meglumine (Eovist), or the CNS/multipurpose agent gadobenate dimeglumine (MultiHance). In terms of NSF, “low risk” does not appear to be equivalent to “no risk”; some level of caution is still warranted for all agents in the highest-risk patients.


  1. Maravilla KR, Maldjian JA, Schmalfuss IM, et al. Contrast enhancement of central nervous system lesions: multicenter intraindividual crossover comparative study of two MR contrast agents. Radiology. 2006;240:389-400.
  2. Rowley HA, Scialfa G, Gao PY, et al. Contrast-enhanced MR imaging of brain lesions: a large scale intraindividual crossover comparison of gadobenate dimeglumine versus gadodiamide. AJNR Am J Neuroradiol. 2008;29:1684-1691.
  3. Seidl Z, Vymazal J, Mechl M, et al. Does higher gadolinium concentration play a role in the morphologic assessment of brain tumors? Results of a multicenter intraindividual crossover comparison of gadobutrol versus gadobenate dimeglumine (the MERIT Study). AJNR Am J Neuroradiol. 2012;33:1050-1058.
  4. Pediconi F, Fraioli F, Catalano C, et al. Gadobenate dimeglumine (Gd-DTPA) vs gadopentetate dimeglumine (Gd-BOPTA) for contrast-enhanced magnetic resonance angiography (MRA): Improvement in intravascular signal intensity and contrast to noise ratio. Radiol Med (Torino). 2003;106:87-93.
  5. Prokop M, Schneider G, Vanzulli A, et al. Contrast-enhanced MR angiography of the renal arteries: blinded multicenter crossover comparison of gadobenate dimeglumine and gadopentetate dimeglumine. Radiology. 2005;234:399-408.
  6. Achenbach M, Figiel JH, Burbelko M, Heverhagen JT. Prospective comparison of image quality and diagnostic accuracy of 0.5 molar gadobenate dimeglumine and 1.0 molar gadobutrol in contrast-enhanced run-off magnetic resonance angiography of the lower extremities. J Magn Reson Imaging. 2010;32:1166-1171.
  7. Woodard PK, Chenevert TL, Sostman HD, et al. Signal quality of single dose gadobenate dimeglumine pulmonary MRA examinations exceeds quality of MRA performed with double dose gadopentetate dimeglumine. Int J Cardiovasc Imaging. 2012;28:295-301.
  8. Li Y, Li X, Li D, et al. Multicenter, intraindividual comparison of single-dose gadobenate dimeglumine and double-dose gadopentetate dimeglumine for MR angiography of the supra-aortic arteries (the Supra-Aortic VALUE study). AJNR Am J Neuroradiol. 2013;34:847-854.
  9. Wang J, Yan F, Liu J, et al. Multicenter, intra-individual comparison of single dose gadobenate dimeglumine and double dose gadopentetate dimeglumine for MR angiography of the peripheral arteries (the Peripheral VALUE Study). J Magn Reson Imaging. 2013;38:926-937.
  10. Tweedle MF. The ProHance story: the making of a novel MRI contrast agent. Eur Radiol. 1997;7 Suppl 5:225-30.
  11. Grobner T. Gadolinium--a specific trigger for the development of nephrogenic fibrosing dermopathy and nephrogenic systemic fibrosis? Nephrol Dial Transplant. 2006;21:1104-1108.
  12. Dotarem® (gadoteric acid) [product information]. Wayne, NJ: Guerbet; March 2013.
  13. Gadavist™ (gadobutrol) injection [prescribing information]. Wayne, NJ; Bayer HealthCare Pharmaceuticals; March 2011.
  14. ProHance® (gadoteridol) injection, 279.3 mg/mL [prescribing information]. Princeton, NJ: Bracco Diagnostics Inc.; revised September 2011.
  15. Magnevist® (brand of gadopentetate dimeglumine) injection [prescribing information].Wayne, NJ: Bayer HealthCare Pharmaceuticals; March 2012.
  16. Omniscan™ (gadodiamide) injection [prescribing information]. Princeton, NJ: Amersham Health Inc.; December 2010.
  17. OptiMARK® 0.5 mmol/mL (gadoversetamide injection) [prescribing information]. St. Louis, MO: Mallinckrodt Inc.; November 2010.
  18. Ablavar® (gadofosveset trisodium) [prescribing information]. North Billerica, MA: Lantheus Medical Imaging, Inc.; February 2011.
  19. Eovist® (gadoxetate disodium) [prescribing information]. Wayne, NJ: Bayer HealthCare Pharmaceuticals; November 2011.
  20. MultiHance® (gadobenate dimeglumine) injection, 529 mg/mL [prescribing information]. Princeton, NJ; Bracco Diagnostics Inc.; July 2012.
  21. Rohrer M, et al. Comparison of magnetic properties of MRI contrast media solutions at different magnetic field strengths. Invest Radiol. 2005;40:715-724. 5.
  22. Tweedle MF, Hagan JJ, Kumar K, Mantha S, Chang CA. Reaction of gadolinium chelates with endogenously available ions. Magn Reson Imaging. 1991;9(3):409-15.
  23. Tweedle MF, Wedeking P, Kumar K. Biodistribution of radiolabeled, formulated gadopentetate, gadoteridol, gadoterate, and gadodiamide in mice and rats. Invest Radiol. 1995 Jun;30(6):372-80.
  24. Prince MR, Erel HE, Lent RW, Blumenfeld J, Kent KC, Bush HL, Wang Y. Gadodiamide administration causes spurious hypocalcemia. Radiology. 2003 Jun;227(3):639-46.
  25. Frenzel T, Lengsfeld P, Schirmer H, Hütter J, Weinmann HJ. Stability of gadolinium-based magnetic resonance imaging contrast agents in human serum at 37 degrees C. Invest Radiol. 2008;43:817-828.
  26. White GW, Gibby WA, Tweedle MF. Comparison of Gd(DTPA-BMA) (Omniscan) versus Gd(HP-DO3A) (ProHance) relative to gadolinium retention in human bone tissue by inductively coupled plasma mass spectroscopy. Invest Radiol. 2006 Mar;41(3):272-8.
  27. Pietsch H, Pering C, Lengsfeld P, Walter J, Steger-Hartmann T, Golfier S, Frenzel T, Hütter J, Weinmann HJ, Sieber MA. Evaluating the role of zinc in the occurrence of fibrosis of the skin: a preclinical study. J Magn Reson Imaging. 2009 Aug;30(2):374-83. doi: 10.1002/jmri.21845.
  28. Ersoy H1, Rybicki FJ. Biochemical safety profiles of gadolinium-based extracellular contrast agents and nephrogenic systemic fibrosis. J Magn Reson Imaging. 2007;26:1190-1197.
  29. Idée JM, Port M, Robic C, Medina C, Sabatou M, Corot C. Role of thermodynamic and kinetic parameters in gadolinium chelate stability. J Magn Reson Imaging. 2009;30:1249-1258.
  30. Laurent S, Elst LV, Muller RN. Comparative study of the physicochemical properties of six clinical low molecular weight gadolinium contrast agents. Contrast Med Mol Imaging. 2006;1:128-137.
  31. Giesel FL, von Tengg-Kobligk H, Wilkinson ID, et al. Influence of human serum albumin on longitudinal and transverse relaxation rates (r1 and r2) of magnetic resonance contrast agents. Invest Radiol. 2006;41:222-228.
  32. Martincich L, Faivre-Pierret M, Zechmann CM, et al. Multicenter, double-blind, randomized, intraindividual crossover comparison of gadobenate dimeglumine and gadopentetate dimeglumine for breast MR imaging (DETECT Trial). Radiology. 2011;258:396-408.
  33. Schneider G, Maas R, Schultze Kool L, et al. Low-dose gadobenate dimeglumine versus standard dose gadopentetate dimeglumine for contrast-enhanced magnetic resonance imaging of the liver: an intra-individual crossover comparison. Invest Radiol. 2003;38:85-94.
  34. Knopp MV, Schoenberg SO, Rehm C, et al. Assessment of gadobenate dimeglumine for magnetic resonance angiography: phase I studies. Invest Radiol. 2002;37:706-715.
  35. Knopp MV, Giesel FL, von Tengg-Kobligk H, et al. Contrast-enhanced MR angiography of the run-off vasculature: intraindividual comparison of gadobenate dimeglumine with gadopentetate dimeglumine. J Magn Reson Imaging. 2003;17:694-702.
  36. Prokop M, Schneider G, Vanzulli A, et al. Contrast-enhanced MR angiography of the renal arteries: blinded multicenter crossover comparison of gadobenate dimeglumine and gadopentetate dimeglumine. Radiology. 2005;234:399-408.
  37. Bueltmann E, Erb G, Kirchin MA, Klose U, Naegele T. Intra-individual crossover comparison of gadobenate dimeglumine and gadopentetate dimeglumine for contrast-enhanced magnetic resonance angiography of the supraaortic vessels at 3 Tesla. Invest Radiol. 2008;43:695-702.
  38. Gerretsen SC, le Maire TF, Miller S, et al. Multicenter, double-blind, randomized, intraindividual crossover comparison of gadobenate dimeglumine and gadopentetate dimeglumine for MR angiography of peripheral arteries. Radiology. 2010;255:988-1000.
  39. Pintaske J, Martirosian P, Graf H, et al. Relaxivity of gadopentetate dimeglumine (Magnevist), gadobutrol (Gadovist), and gadobenate dimeglumine (MultiHance) in human blood plasma at 0.2, 1.5, and 3 Tesla [erratum in Invest Radiol. 2006;41:859]. Invest Radiol. 2006;41:213-221.
  40. Shellock FG, Parker JR, Venetianer C, et al. Safety of gadobenate dimeglumine (MultiHance): summary of findings from clinical studies and postmarketing surveillance. Invest Radiol. 2006;41:500-509.
  41. American College of Radiology (ACR). ACR Manual on Contrast Media. 2013; version 9:81-98. Available at:
  42. Weber JCP. Epidemiology of Adverse Reactions to Nonsteroidal Antiinflammatory Drugs. New York, NY: Raven Press; 1984:1–7. Rainsford KD, Velo GP, eds. Advances in Inflammation Research; vol 6.
  43. Davenport MS, Dillman JR, Cohan RH, et al. Effect of abrupt substitution of gadobenate dimeglumine for gadopentetate dimeglumine on rate of allergic-like reactions. Radiology. 2013;266:773-782.
  44. Bleicher AG, Kanal E. Assessment of adverse reaction rates to a newly approved MRI contrast agent: review of 23,553 administrations of gadobenate dimeglumine. AJR Am J Roentgenol. 2008 Dec;191(6):W307-11.
  45. (E. Kanal, unpublished data).
  46. Thomsen HS. Nephrogenic systemic fibrosis: a serious late adverse reaction to gadodiamide. Eur Radiol. 2006;16:2619-2621.
  47. Rumboldt Z, et al. Multicenter, double-blind, randomized, intra-individual crossover comparison of gadobenate dimeglumine and gadopentetate dimeglumine in MRI of brain tumors at 3 Tesla. JMRI. 2009;29:760-767.
  48. Attenberger UI, Michaely HJ, Wintersperger BJ, et al. Three-dimensional contrast-enhanced magnetic-resonance angiography of the renal arteries: interindividual comparison of 0.2 mmol/kg gadobutrol at 1.5 T and 0.1 mmol/kg gadobenate dimeglumine at 3.0 T. Euro Radiol. 2008;18(6):1260-8.
  49. Bueltmann E, Erb G, Kirchin MA, Klose U, Naegele T. Intra-individual crossover comparison of gadobenate dimeglumine and gadopentetate dimeglumine for contrast-enhanced magnetic resonance angiography of the supraaortic vessels at 3 Tesla. Invest Radiol. 2008;43:695-702.
  50. Spampinato MV, Nguyen SA, Rumboldt Z. Comparison of gadobenate dimeglumine and gadodiamide in the evaluation of spinal vascular anatomy with MR angiography. AJR Am J Neuroradiol. 2010;31:1151-1156.
  51. Schneider G, Maas R, Schultze Kool L, et al. Low-dose gadobenate dimeglumine versus standard dose gadopentetate dimeglumine for contrast-enhanced magnetic resonance imaging of the liver: an intra-individual crossover comparison. Invest Radiol. 2003;38:85-94.
  52. Li Y, Li X, Li D, et al. Multicenter, intraindividual comparison of single-dose gadobenate dimeglumine and double-dose gadopentetate dimeglumine for MR angiography of the supra-aortic arteries (the Supra-Aortic VALUE study). AJNR Am J Neuroradiol. 2013;34:847-854.
  53. Wang J, Yan F, Liu J, et al. Multicenter, intra-individual comparison of single dose gadobenate dimeglumine and double dose gadopentetate dimeglumine for MR angiography of the peripheral arteries (the Peripheral VALUE Study). J Magn Reson Imaging. 2013;38:926-937.

This article was published in a supplement to the May 2014 issue of Applied Radiology and was sponsored by an educational grant from Bracco.

Back To Top

About the Author

Michael F. Tweedle, PhD, Emanuel Kanal, MD, FACR, FISMRM, and Robert Muller, PhD

Dr. Tweedle is a Stefanie Spielman Professor of Radiology at The Ohio State University, Columbus, OH; Dr. Kanal is Professor of Radiology and Neuroradiology at the University of Pittsburgh Medical Center, Pittsburgh, PA; and Dr. Muller is in the Department of General, Organic & Biochemical Chemistry, University of Mons, Mons, Belgium.