Giovanni Morana, Andrea Mazzaro, Alex Faccinetto, Luigi Grazioli
When MRI started to be used in clinical applications, the higher contrast resolution provided and the multiparameter assessment seemed to make the use of contrast media avoidable. However, it was soon clear that in many cases the differentiation between different pathologic tissues was unsatisfactory, that led to the development of MRI Contrast Agents (CA).
In CT imaging, signal increase has a linear relationship with density and atomic weight of contrast agents. Differently from that, in MRI there are multiple and more complex parameters that affect signal intensity, not only related to proton density but also to proton relaxation. So, while in CT imaging contrast agents directly stop the photon-X crossing the target-organ, in MRI they affect the relaxation times of the nearby protons; therefore a single molecule of contrast agent can modify the relaxation time of a large amount of protons, needing a much lower concentration of contrast media to obtain a differentiation between normal and pathologic tissues than the one required for CT imaging.
MRI contrast agents are made up of atoms or molecules with a different action, not only capable to increase signal (positive CA) but also to decrease it (negative CA). The capability of contrast agents to increase T1-T2 relaxation time is called relaxivity, expressed as R1 or R2, whether it’s referred to the capability to increase T1 (R1) or T2 (R2) time of protons (Dawson; Xiao).
Most contrast agents influence both T1 and T2 relaxation times, although with prevalence of one of them. The resulting signal (positive or negative) depends on the prevalent T1 or T2 relaxation time activity on the nearby protons and on the sequence used (T1- or T2-weighted).
In accordance with their prevalent relaxation activity, MRI contrast agents can be divided as follows.
•Paramagnetic CA: characterized by a signal increase (positive contrast agents) in their distribution area on T1-weighted sequences. This category is represented by atoms belonging to transitional metals (iron and manganese) or lanthanides (gadolinium) groups, the latter also named rare earths elements. In these atoms the four unpaired electrons in the outer orbitals generate a magnetic moment that, when exposed to an external magnetic field, tends to be aligned with it, strengthening it. The magnetic field produced by an electron is clearly higher than the one produced by the adjacent water protons (about 1,000 times), resulting in a high capability to accelerate their relaxation times, especially R1. The magnitude of the interaction between the contrast medium and the adjacent water protons is inversely proportional to the sixth power of the distance existing between them (1/r6), therefore it’s mandatory to get a short distance between them to strengthen the relaxing effect of the contrast medium. There is a non-linear relationship between the concentration of the contrast agent and the relaxing effect: there is no paramagnetic effect at low concentrations while when the concentration increases it becomes much clearer. At high concentrations of contrast media the T2 effect prevails, with marked lowering of the signal even with in T1-weighted sequences. However, even with low paramagnetic CA concentrations it is possible to have a T2 effect, detectable only with T2w sequences. This effect can be used to reduce the background signal in the biliopancreatic region when performing MRCP (Magnetic Resonance CholangioPancreatography) studies: in case of high background signal, the administration of a few ml of paramagnetic contrast medium leads to a decrease of the background signal maintaining that one of the ducts. This is a consequence of the superparamagnetic effect shown during the interstitial distribution phase of the contrast medium (3-5 min after the injection) () (Czeyda-Pommersheim; Dawson; Xiao).
•Superparamagnetic CA: characterized by a signal reduction (negative contrast agents) in their distribution area on T2-weighted sequences. This category is represented by small iron ferromagnetic particles, called SuperParamagnetic Iron Oxide (SPIO) particles.
Unlike paramagnetic substances, in normal conditions these particles don’t have any magnetic activity, but when subject to a magnetic field they get a much higher magnetization than the paramagnetic particles, quickly reaching a non-linear saturation effect with the intensity of the magnetic field. The local magnetic fields produced in this way overlap with the external magnetic field, creating inhomogeneities that accelerate the proton relaxation processes, especially the transverse ones (dephasing), responsible for the T2-weighting of the images. This effect is particularly marked when using sequences that are sensitive to the inhomogeneity of the magnetic field (so-called T2* sequences: gradient-echo sequences), while T2-weighted sequences that use multiple refocusing pulses at 180° for a fast filling of K space (turbo spin-echo sequences) are less sensitive to such contrast media () (Wang; Xiao).
SPIO particle sizes influence relaxing activity: relatively large particles (diameter > 50 nm: SPIO) have a predominantly relaxing T2 activity (), while smaller particles (< 50 nm diameter: USPIO, Ultrasmall SuperParamagnetic Iron Oxide particles) also have a relaxing T1 activity. In USPIO particles the T1 or T2 activity is also linked to their concentration: when the contrast medium is diluted in the solution (as in the circulatory stream), the T1 effect is evident while when the USPIO particles are clustered inside lysosomes after being phagocitated by the macrophages of the reticuloendothelial system, the T2 effect prevails. This can be detected not only with T2 sequences () but also with T1-weighted sequences, in which it can be detected a decrease of the signal () (Wang).
Figure 1 – T2 effect of gadolinium. Magnetic Resonance CholangioPancreatography (MRCP). A, B) 2D MRCP. Before Gd injection (A), the background of the pancreas shows small signal inhomogeneity (*); 3-5 min after Gd injection (B), the background is more dark and the Wirsung duct more clearly visible. C, D) 3D MRCP: the signal of small veins can be appreciated (arrows) before Gd injection (C), which disappear after Gd injection (D).
Figure 2 – Response of various sequences to superparamagnetic CA. Patient with HCC (HepatoCellular Carcinoma). A, B) T2w sequence before (A) and after (B) SPIO: the presence of multiple refocalization pulses (ETL 15: fifteen 180 degrees RF refocalization pulses between two consecutive TR) significantly reduces the role of T2* relaxation induced by administration of iron oxide particles, with slight decrease of the signal after CA administration, without significant increase of C/N ratio between liver and lesion. C, D) T2* sequence before (C) and after (D) SPIO. C/N ratio between lesion and liver is still low before CA administration, but the lack of refocalization pulses makes this sequence sensible to the effect of SPIO particles, with significant decrease of the signal of liver parenchyma after CA administration and consequent increase of C/N ratio between the liver and the lesion.
Figure 3 – T1-T2 effect of USPIO particles. Patient with hemangioma. GRE 3D T1w sequences during administration of Resovist®: before CA injection (A) the lesion is hypointense; during CA administration a peripheral globular enhancement can be appreciated (B), more visible during portal phase (C). On T2w images before CA administration the lesion is markedly hyperintense (D), showing decrease of signal intensity after CA administration (E).
Figure 4 – T1-T2 effect of SPIO particles. Patient with metastases from breast tumor. A) At GRE T1w sequence before CA administration a hypointense lesion can be noted in segment VII. Another uncertain lesion in segment IV (arrow). B) GRE T2w fat-sat: the lesions appear hyperintense; it can be noted another small lesion in segment VII (arrow). C) After SPIO (Endorem®) the liver is hypointense due to T2* effect, but the lesions did not uptake the CA, thus being more visible, even the smallest ones (arrows). After SPIO the liver on GRE T1w (D) is hypointense due to T2* effect, thus the lesions are more visible. * = ghost artifact of the aorta.
The classification of MRI contrast agents is based on their distribution and mechanism of action.
These compounds have a distribution kinetics that is similar to the one of iodine contrast agents used in CT and in conventional contrast radiography. Paramagnetic extracellular contrast agent currently used are based on gadolinium ion Gd+++. This ion has to be bound to chelates as it is toxic in free state (accumulation in reticuloendothelial system, especially of liver, spleen and bone marrow, inducing cellular necrosis) (Czeyda-Pommersheim; Guglielmo; Xiao).
After intravenous infusion CA quickly reaches a distribution equilibrium between intravascular and interstitial compartment, except for brain and testicles, as blood brain barrier and blood testis barrier are impermeable to these compounds (). These contrast agent particles have a low molecular weight (500 Dalton), allowing them to cross the pores of the capillary wall. Their elimination occurs almost exclusively through kidneys filtration, with 90% of infused dose that can be found in urine approximately 3 hours after the injection (Guglielmo).
Although originally developed for “static” imaging, to allow a better detection of brain lesions, where dynamic imaging was not required, the development of more performant gradients and fast sequences allowed the use of MRI contrast agents in the same way as iodine CT contrast agents. With GRE sequences, possibly 3D, it’s possible to analyze the enhancement of contrast agents in each single phase (). There are two techniques of synchronization for contrast agents: in the first the bolus arrival time is calculated previously (test bolus), in the second the bolus arrival is assessed in real time to make the sequence start, manually or automatically. A correct synchronization between the bolus arrival in arterial phase and the sequence acquisition is mandatory, depending on the imaging protocol: in MR-angio, the bolus arrival must coincide with the acquisition of the center of K space (pure arterial phase), while with other protocols, different synchronization is required (e.g., late arterial phase in hepatic imaging, pancreatic phase, etc.). It is important to be aware of the timing of acquisition of the center of K space after the start of the sequence: depending on the sequence, the center of K space can be acquired at the beginning, one third or half of acquisition time; this timing can be visualized in the console machine and it is used to calculate the delay between the contrast agent injection and the start of the sequence. While in the beginning of dynamic imaging the perfect synchronization of bolus arrival and center of K space acquisition was essential in order to achieve a high quality imaging, nowadays, the use of “center of K space first” imaging sequence (elliptical-centric ordering) is the preferred K space sampling method used for most CE-MRA (Contrast-Enhanced MR Angiography) examinations, which allows a simple detection of bolus arrival with one frame real time sequences, starting the acquisition of the sequence as soon as the bolus arrivals (Czeyda-Pommersheim; Guglielmo; Xiao).
Paramagnetic CA are the most used compounds for Dynamic Contrast-Enhanced-MR Imaging (DCE-MRI), allowing to obtain as much information as a contrast-enhanced CT scan. Due to their high tolerability, they can be used in patients who have a known allergy to iodinated contrast media. Their field of application consists most in oncological applications ( and ), where there is a need for lesion detection, characterization and extension, as well as many others pathological conditions, such as trauma, inflammations, and so on (Czeyda-Pommersheim; Guglielmo; Xiao).
Moreover, with 3D high resolution sequences is possible both to achieve an evaluation of the enhancement features of organs and to post-process images (MIP, VR) obtaining a vascular map of the area ().
Figure 5 – Extracellular paramagnetic CA. Meningioma. A large lesion in the right clinoid is appreciable on T2w axial image (A), which appears isointense on T1w images (B). After paramagnetic Gd-based CA the lesion shows marked enhancement (C, D), due to disruption of Blood Brain Barrier (BBB), while the normal parenchyma does not show enhancement, as intact BBB does not permit the passage of CA.
Figure 6 – Extracellular paramagnetic CA. Metastasis from carcinoid. On T2w images (A), a small hyperintense lesion is appreciable in the VIII segment, which appears hypointense on GRE T1w (B). During bolus injection of paramagnetic Gd-based CA the lesion appears hypervascular in the arterial phase (C), with wash-out in the portal phase (D), suggestive for the malignity of the lesion.
Figure 7 – Extracellular paramagnetic CA. Vascular reconstruction after dynamic imaging of the abdomen. Multifocal HCC. During arterial phase (A) multiple hypervascular lesions can be appreciated in a patient with cirrhosis. The MIP of the arterial phase (B) shows a significant stenosis at the origin of the celiac axis (arrow), which was confirmed during transarterial chemoembolization (C).
Figure 8 – Extracellular paramagnetic CA. Whole body MRA during first pass acquisition from cerebral circle to pedideal arteries.
Blood Pool CA (BPCA) are paramagnetic compounds that due to their size are not able to cross the capillary pores, so they remain confined mainly in the intravascular compartment for a variable time period, depending on the product used. These contrast media can be considered markers of the intravascular compartment, and only in the case of increased capillary permeability (for example, in the case of neoangiogenetic processes) can a passage of such contrast media be observed in the interstitial compartment, with a progressive increase in the signal of these tissues over time (Czeyda-Pommersheim; Xiao).
Retention of contrast media is obtained developing products consisting of compounds with a high, middle or low molecular size, the latter however has to reversible bind to serum albumin to increase their molecular size. Glomerular base membrane is still permeable to high molecular size CA, therefore they can be eliminated by kidneys. Molecules with a molecular weight > 70 kDa are not eliminated by kidneys, so they have to be previously metabolized.
Two different solutions were developed in the research of BPCA: gadolinium-based or iron-based.
•Gadolinium-based CA: their dimensions are approximately those of serum proteins such as albumin, and precisely the Gadolinium-DiethyleneTriaminePentacetic Acid (Gd-DTPA) chelated with albumin was the prototype of the CA blood pool. The only BPCA approved for clinical use was gadofosveset trisodium (Vasovist® or Ablavar®) that was a stable gadolinium chelate (Gd-DTPA) that reversibly binded to plasma albumin; the main applies of these compounds were the contrast-enhanced MR angiography. The manufacturer discontinued production in 2017 due to poor sales (Goyen; Xiao).
•Iron-based CA: ultrasmall superparamagnetic iron oxides (USPIO, < 50 nm diameter) were developed for contrast-enhanced MR imaging. The molecular weight, coating compound and its surface charge influence the pharmacokinetics and pharmacodynamics. The greater the molecular weight, the slower their renal elimination, thus the longer the plasmatic half-life. Moreover, the higher molecular dimensions of USPIO provide a more efficient relaxivity activity, allowing a 3-4 fold increase of enhancement compared to their analogue Gd-based CA. The most advance molecule was developed by Bayer with SHU 555C (Supravist®), made by USPIO (25 nm) coated by carboxydextran, with an iron concentration of 0.5 mmol Fe/ml, derived from Resovist®, a liver-specific SPIO CA (see below). However also these molecules were withdrawn from the market by the manufactures, mostly due to economical reasons (low volumes of sell) (Czeyda-Pommersheim; Xiao).
These compounds were used in the assessment of vascular system with CE-MRA technique, both with first-pass () and with steady-state technique. The latter is characterized by a 3D high resolution acquisition during the phase of equilibrium concentration between arterial and venous compounds, that needs a longer time of acquisition but allows a high spatial resolution, obtaining submillimetric voxels (). Also with these CA it was possible to obtain a complete visualization of the most important arterial vessels of the body in a unique session (whole body MRA, ).
Figure 9 – Blood pool trisodium gadofosveset CA (Vasovist®). MR angiography. A) MIP sequence of arterial phase (first-pass): a slight parietal irregularity can be appreciated in the internal carotid artery, with non-significant reduction of caliber. B) SubMIP sequence in steady-state phase, with high resolution acquisition: the irregularity of the ulcerated plaque is better shown (arrow).
Figure 10 – Blood pool trisodium gadofosveset CA (Vasovist®). Whole body MRA during first pass acquisition from cerebral circle to pedideal arteries.
To improve the sensitivity and specificity of MRI in the study of focal hepatic lesions, CA with specific liver accumulation have been developed. Depending on their storage location, these CA can be divided as follows.
•CA with exclusive accumulation in the hepatocellular compartment: Manganese-DiPyridoxilDiPhosphate (Mn-DPDP) (Teslascan®). The manganese ion is chelated to 4 molecules of meglumine; it is infused slowly (2 ml – 3 ml/min in 10-20 min) at a concentration of 10 mmol/ml, and at a dose of 0.5 ml/kg. In the plasma the Mn++ ion contained in the molecule is released slowly from the DPDP chelator, replaced by zinc, whose affinity for the chelator is markedly higher than for manganese. The Mn++ ion therefore becomes available for uptake by parenchymal cells, particularly in liver, pancreas, kidneys, adrenals and myocardium. The Mn++ ion performs paramagnetic activity and at hepatic level the maximum enhancement is observed about 20 min from the end of the administration of the contrast medium, with a duration of about 4 hours, with the liver showing hyperintensity while the non-hepatocytic lesions do not show significant enhancement, thus increasing the contrast noise ratio. In clinical studies the administration of Mn-DPDP increased the identification of liver lesions compared to the pre-contrastographic phase (). On the other hand, numerous lesions of hepatocellular origin demonstrate significant enhancement after administration of Mn-DPDP, reducing C/N but improving the characterization capacity of lesions, between hepatocytes and non-hepatocytes (), although this contrast medium is limited in the differentiation between benign and malignant hepatocytic lesions (). The marketing authorization for Teslascan® has been withdrawn at the request of the marketing-authorization holder for commercial reason in 2012 (Bartolozzi).
Figure 11 – Liver assessment with hepatospecific CA Mn-DPDP (Teslascan®). Colorectal metastasis. A) On SE T1w sequence the lesion, a slightly hypointense one, is appreciable on segment 8, in paracaval, much more appreciable one hour after Mn-DPDP administration (B), where the liver is remarkably hyperintense while the lesion doesn’t show any significant enhancement (*).
Figure 12 – Liver assessment with hepatospecific Mn-DPDP (Teslascan®). Focal Nodular Hyperplasia (FNH). On GRE T1w coronal sequence (A) in the right hepatic lobe a slightly hypointense area can be appreciated. One hour after CA administration (B), the lesion shows a significant enhancement, higher than the surrounding parenchyma.
Figure 13 – Liver assessment with hepatospecific Mn-DPDP (Teslascan®). Well differentiated HCC with extrahepatic metastases. On GRE T1w sequence a large hypointense lesion in the right hepatic lobe can be appreciated (A); another lesion in the left flank (B). One hour after Mn-DPDP administration, both lesions (C, D) show significant contrast enhancement, similar to hepatic parenchyma.
•CA with mixed, extracellular and intrahepatocytic distribution: these compounds are injected into the bolus and provide for a first distribution in the vasculointerstitial compartment, in a similar way to the conventional extracellular distribution medium. Subsequently a part of the injected dose is picked up by the hepatocytes, where they exert a paramagnetic effect, with an increase in the signal of the hepatic parenchyma. Two molecules belong to this class:
–Gadobenate dimeglumine (Gadolinium-BenzylOxyPropionicTetra-Acetate – Gd-BOPTA -, MultiHance®) is a gadolinium chelate characterized by a weak and transient interaction with serum albumin (which gives Gd-BOPTA a greater relaxivity than other conventional paramagnetic contrasts) and an elimination profile that predicts about 96% of the injected dose eliminated via the kidney by glomerular filtration, while the remaining 2-4% is taken-up by functioning hepatocytes and eliminated through the biliary tract, which leads to an increase in enhancement of the hepatic parenchyma, starting approximately 40 min after administration and lasting about 3 hours. Gd-BOPTA behaves similarly to conventional paramagnetic contrast media during the dynamic phase, while in the hepatobiliary phase it improves the sensitivity of the MRI method in the identification of focal hepatic lesions and increases their specificity, contributing to a better characterization of the lesions, especially those that present an atypical dynamic enhancement pattern (). With Gd-BOPTA in the hepatobiliary phase only benign hepatocytic lesions appear hyperintense, while malignant hepatocytic lesions and non-hepatocyte lesions appear hypointense () (Agostini; Grazioli, 2001, 2005; Morana).
Figure 14 – Liver assessment with hepatospecific Gd-BOPTA (Multihance®). FNH. A) TSE T2w sequence: slightly hyperintense area at segment VII. B) GRE T1w sequence before CA administration: the lesion appears slightly hypointense. C, D) GRE T1w sequence during CA administration: arterial (C) and venous (D) phases. The lesion appears remarkably hypervascular, without typical characteristics of FNH (hypovascular central scar, enhancing during equilibrium phase). E) GRE T1w sequence 2 hours after administration of Gd-BOPTA: the lesion appears hyperintense. Hyperintensity in biliary phase is typical of benign liver lesions. The whole liver appears hyperintense due to uptake of CA. Ghost artifact of the aorta can be noticed (*).
Figure 15 – Liver assessment with hepatospecific Gd-BOPTA (Multihance®). HCC. A) GRE T1w sequence before CA administration: the lesion appears hypointense. B, C) GRE T1w sequence during CA administration: arterial (C) and venous (D) phases; the lesion appears homogeneously hypervascular (C), with corona enhancement during the portal phase (D). E) GRE T1w sequence 2 hours after administration of Gd-BOPTA: the lesion appears hypointense.
–Gadolinium-EthOxyBenzyl DiethyleneTriaminePentacetic Acid (Gd-EOB-DTPA) (Primovist®): similarly to Gd-BOPTA, this contrast agent after bolus injection is initially distributed in the vasculointerstitial compartment, with a behavior that is completely superimposable to conventional paramagnetic contrast agents; subsequently, however, 50% of the injected dose is taken up by the hepatocytes through the bilirubin carrier and bilary eliminated. The increase of the hepatic parenchymal enhancement is earlier than Gd-BOPTA and at 20 min after its administration an optimal increase of the Signal Intensity – SI – of the liver is obtained which lasts about 2 hours. Similarly to Gd-BOPTA, Gd-EOB-DTPA improves the sensitivity and specificity of MR in the identification and characterization of focal liver lesions ( and ) (Agostini).
Figure 16 – Liver assessment with hepatospecific Gd-EOB-DTPA (Primovist®). FNH. A) TSE T2w sequence: slightly hyperintense area at segment V. B) GRE T1w sequence before CA administration: the lesion appears slightly hypointense. C, D) GRE T1-weighted sequence during CA administration: arterial (C) and venous (D) phases. The lesion appears remarkably hypervascular in arterial phase and isointense in venous phase. E) GRE T1w sequence 20 min after CA administration: the lesion appears isointense for CA uptake.
Figure 17 – Liver assessment with hepatospecific Gd-EOB-DTPA (Primovist®). HCC. A) TSE T2w: a large hyperintense lesion in the dome of the liver. B) GRE T1w sequence before CA administration: the lesion appears hypointense. C, D) GRE T1w sequence during CA administration: arterial (C) and venous (D) phases. The lesion appears remarkably hypervascular in arterial phase, with wash-out and pseudocapsule in venous phase. E) GRE T1w sequence 20 min after CA administration: enhancement of the liver due to uptake of CA. The lesion appears hypointense.
CA with distribution in the ReticulousEndothelial System (RES)
They consist of particles of iron oxide (SPIO), with superparamagnetic activity. After intravenous injection, the SPIO particles are phagocytosed by the macrophage cells present in the RES at the hepatic, splenic, medullary and lymph node levels. The susceptibility effect of the iron particles results in signal reduction in normal liver on T2- or T2*-weighted MR imaging. Malignant lesions lack normal Küpfer cells and still show relatively high signal intensity on post-contrast imaging, while the surrounding normal liver parenchyma shows a marked decrease of signal intensity on T2-T2* images, thus increasing the C/N. Two products were available in the market, one with slow administration (Endorem®), with a slow progressive uptake of the iron particles (SPIO, 50-180 nm) and a correspondent decrease of T2-T2* signal intensity starting from one hour to 6 hours after administration (); the other one (Resovist®) was developed to allow a fast injection, by using smaller iron oxide particles (45-60 nm), with a double enhancement effect: a T1 during dynamic imaging, thanks to the high concentration of iron oxide particles in the extracellular compartment, and a late T2 effect, due to the accumulation of SPIO in the reticuloendothelial system; thus with this CA was possible to have both T1 dynamic a T2 liver-specific informations (). SPIO contrast agents have largely been withdrawn from the market and they are no longer currently available for routine diagnostic use (Czeyda-Pommersheim; Lencioni; Xiao).
CA with lymph node distribution
A further development of such contrast media is constituted by ultrasmall iron oxide particles (USPIO < 50 nm). The size of USPIO allows a different uptake: while 50 nm are still uptaken by RES of liver and spleen, smaller particles (20 nm) are picked up by macrophages present at the lymph node level. Thus, after intravenous administration they accumulate in the lymph nodes, inducing a reduction in T2-T2* signal intensity. By contrast, metastatic lymph nodes, in which macrophages are substituted by tumor cells, do not show significant changes in signal intensity in these sequences (). Sinerem® was the first RM contrast agent developed for the study of lymph nodes, although it never reached the market and it is still unavailable (Sigal).
Figure 18 – Lymphnodal assessment with Sinerem®. Armpit metastatic lymph node in patient with breast tumor. Images acquired 24 hours after CA administration. A) With SE T1w sequence, a wedge-shaped lymph node can be noted in the left armpit. B) With GRE T2* sequence, the lymph node is remarkably hyperintense in its metastatic component, with decrease of signal intensity in its normal portion. C) Macroscopic biopsy specimen, showing the difference between metastatic and normal portion.
Their use is mainly required in the study of the biliopancreatic district and of the intestinal loops.
•The correct visualization of the biliopancreatic ducts with heavily T2-weighted sequences is often hindered by the liquid content of stomach and duodenum. To avoid that, oral CAs that selectively suppress the signal from the intestinal contents are used. This effect can be obtained with a large number of products, both superparamagnetic and paramagnetic also, taking advantage of their superparamagnetic effect in T2-weighted sequences (). These substances are therefore represented either by solutions of gadolinium chelates or by fruit juices particularly rich in manganese (black bilberry, pineapple) or by commercial products containing of iron oxide (Lumirem®) (Czeyda-Pommersheim; Wang; Xiao).
Figure 19 – MRCP for assessment of pancreatic ducts. A gadolinium-based solution is used to mask intestinal fluids. Cystic pancreatic lesion. A) Axial GRE T1-weighted fat-sat sequence, arterial phase: hyperintensity of the lumen of the duodenal loop (arrows), due to T1 effect of the solution. Cistic lesion (*). Hepatic artery (arrowheads). B) At MRCP, the bright signal of the fluids of the stomach and duodenum are cancelled by the T2 effect of the gadolinium. The Santorini duct can be appreciated (arrows).
•The MR study of intestinal loops requires an optimal distension of the loops themselves, with good differentiation between the intestinal lumen and its wall. At present there are no standardized protocols for carrying out the dedicated study of small intestine with MRI. The contrast medium can be alternatively administered through a nose-jejunal catheter or usually by oral route. Oral administration has the advantage of not requiring exposure to radiant dose for intubation, adequately opacifies the stomach and all the duodenum, frequent locations of localization of chronic inflammatory diseases. However, it imposes the association with an antiperistaltic to reduce movement artifacts: the most suitable for this purpose is glucagon intramuscularly or iv for its instant and prolonged effect, but considering the non-negligible cost, often the choice falls on the hyoscine-butylbromide (Buscopan®) by intramuscular or iv. Buscopan® is contraindicated in acute angle glaucoma, in conditions stenosing the gastrointestinal canal, in paralytic ileus, in ulcerative colitis, in megacolon, in myasthenia gravis and in cardiac tachyarrhythmias. Different types of oral CA have been suggested and are classified as positive, negative and biphasic (Laghi).
–Positive contrast agents increase the signal intensity of the lumen. Most are paramagnetic substances based on gadolinium chelate, iron or manganese; these compounds increase the signal in the T1-weighted sequences (); this allows to distinguish the hyperintense lumen from the wall of the loops which instead shows signal hypointensity in the T1-weighted sequences. The main limit of this category consists in the fact that the iv administration of paramagnetic contrast medium, carried out in order to obtain adequate information on the parietal enhancement (index of disease activity in inflammatory bowel diseases), determines an increase in the signal of the intestinal wall that appears difficult to dissociate from the signal increase of the intestinal lumen induced by these contrast agents. In an attempt to meet the criteria of low cost, good tolerability and good opacity of the loops of the small intestine, a large variety of food products with a high content of manganese, such as cranberry juice, milk, ice cream, vegetable oil and green tea, were tested. These foods show an acceptable paramagnetic effect, but they are not suitable for use because their signal varies along the gastrointestinal tract. In particular, cranberry juice is useful in the assessment of the first digestive tract. One of the first formulations on the market was a mixture of gadopentate dimeglumine and mannitol (Magnevist Enteral®), which produces an osmotic effect and therefore ensures a good relaxation and opacification of all the handles. However, this product is no more available in the market, as it has been withdrawn by the manufacturer (Laghi).
–Negative contrast agents: they decrease the signal intensity in T1 and in T2 through three possible mechanisms, on the basis of which they can be classified as superparamagnetic agents (based on iron oxide particles), diamagnetic (barium), or substitutes for the hydrogen of the intestine (gas, perflubron). Most negative contrast media are substances with a superparamagnetic effect. The use of negative oral contrast agents facilitates the detection of pathologies of wall and of serosa, but, due to the iron content of some contrast agents, the sequences can present ferromagnetic artifacts, thus obscuring the slight alterations of the mucosa. In chronic inflammatory bowel disease a double contrast effect can be obtained by combining the use of negative oral contrast medium with fat-suppressing sequences (which cancels the signal from the visceral fat); in this way it is possible to maximize the visualization of the inflamed, hyperintense wall in the T2-weighted sequences. Furthermore, in the T1-weighted sequences the use of gadolinium-based mdc iv increases the contrast between the normal and pathological wall, while the lumen maintains a low signal (Czeyda-Pommersheim; Xiao).
▪Lumirem® consists of a suspension of superparamagnetic particles containing iron oxide crystals with a 10 nm diameter forming 300 nm aggregates capable of decreasing the T2 relaxation time, but reducing T1 relaxation time also (). A homogeneous distribution along the intestinal skein takes an hour to complete. The adverse effects are mostly digestive disorders, such as nausea, vomiting (in 1-5% of the patients), flatulence, and sense of abdominal filling. It was withdrawn from the market in 2007, due to limited request and high costs of productions (Wang).
▪Perflubron® was an inert liquid that does not contain hydrogen atoms, radiopaque due to the presence of the bromine atom in the active ingredient (C8F17Br). This agent demonstrates an optimal negative opacification both with T1- and T2-weighted sequences. Due to its high cost, it was withdrawn from the market in 1996 (Czeyda-Pommersheim; Xiao).
A solution of 1,350 ml at 2% of barium sulfate, fractionated into 3 intakes starting 90 min before the investigation, can be used to distend the intestinal loops, especially the distal ones and the colon. If associated with the use of gadolinium iv, it allows to easily distinguish abnormalities of the intestinal wall. The advantages of using this contrast medium can be summarized in the ease of use, in the low cost, and in the relative tolerability by the patient.
–Biphasic CA behaves as a positive or negative depending on the sequence applied: hypointense in T1 images, hyperintense in T2 images; in this way the intestinal lumen appears hypointense in T1 (with good explorability of the intestinal wall after administration of paramagnetic contrast medium) and hyperintense in T2 (with good contrast compared to the wall that appears hypointense).
▪PEG, PolyEthylene Glycol (Selg-Esse®, macrogol), is an aqueous solution with 0.5% methylcellulose containing non-digestible, non-absorbable and non-fermentable hydrophilic macromolecules; it distends the intestine optimally reaching also the terminal ileum (). Immediately before the exam, 750-1,000 ml are administered orally in adults, 10 ml per kg of body weight in pediatric patients. The dose can be increased by a second oral administration in case of insufficient opacification of the intestinal loops. Its use is contraindicated in occlusive or stenotic forms of the intestine, in gastric stasis, in dynamic ileum, in gastrointestinal perforations, in acute colitis and in toxic megacolon (Laghi).
▪Another oral solution is Ferric Ammonium Citrate (FAC, FerriSeltz®). It is a solution of ferric ammonium citrate used to enhance the delineation of the bowel. Due to the presence of iron, it shows a double effect: with T1-weighted sequences the predominantly positive enhancement helps to distinguish organs and tissues that are adjacent to the upper regions of the gastrointestinal tract; with T2-weighted sequences it reduces the bright signal of fluid-filled gastrointestinal tracts, and it is useful in the MRCP studies. It demonstrates an optimal distribution in the first intestinal tract, but the simultaneous visualization of both proximal and distal parts occurs only in 25% of cases; this behavior makes it effective in the study of the first routes, but not of the terminal ileum, a frequent site of chronic inflammatory pathologies.
Figure 20 – Intestinal loops assessment with MRI after administration of Lumirem®. Patient with Crohn disease. A) Steady-state sequence (true-FISP) T1- T2-weighted. A marked thickening of a jejunal loop can be appreciated (arrows), whose lumen is dark due to the T2* effect of iron oxide crystals. The lumen is dark (*) also on GRE T1w sequence due to iron oxide crystals effect (B), thus allowing a clear visualization of the enhancement of the thickened walls after Gd-based CA administration (C).
Figure 21 – Intestinal loops assessment with MRI after administration of aqueous solution of PEG. Patient with Crohn disease. A) Coronal steady-state sequence (true-FISP) T1- T2-weighted: the walls of the last ileal loop appear to be thickened (arrows), with a hyperintense lumen. B, C) Coronal VIBE T1w sequences before (B) and after (C) paramagnetic CA administration: it can be noticed a remarkable enhancement of the walls of distal ileum (arrows), involving the ileocecal valve. Intestinal lumen appears hypointense.
MR contrast media are an important tool in the diagnostic management of focal lesions. Unfortunately, several CA have been withdrawn from the market, thus the possibility of different information on focal lesions have been reduced. However, the available CA are the most important and allow to best manage the diagnostic approach to disease with MRI.
Alfonso Ragozzino, Marta Puglia, Anna Giacoma Tucci
Since their clinical introduction in 1987, low-dose Gadolinium-Based Contrast Agents (GBCAs) have been considered extremely safe. However, adverse reactions may rarely occur. According to the new European Society of Urogenital Radiology (ESUR) guidelines on contrast agent, adverse reactions are divided as follows.
•General adverse reactions
–acute reactions: within one hour of administration;
–late reactions: between one hour and a week;
–very late reactions: over a week (Nefrogenic Systemic Fibrosis, NSF).
•Renal adverse reactions
–Post-Contrast Acute Kidney Injury (PC-AKI).
The overall incidence of adverse reactions to GBCAs is 0.07-0.3%, with approximately 9.2-0.52 adverse events per 10,000 administrations.
Immediate acute adverse reactions occur within one hour of gadolinium contrast agents administration. Their etiopathogenesis is multifactorial, but they are mostly due to hypersensitivity.
The frequency of acute immediate adverse reactions from hypersensitivity after injection of gadolinium compounds is rather low compared to that observed with iodine-based contrast media, with an incidence reported between 0.004 and 0.7% (0.01-0.3%). Most reactions are cutaneous (75-90% of acute reactions) mild and transient; even more rare are severe anaphylactoid reactions (0.001-0.01%), with only one case of laryngospasm reported in one article (0.002-0.005%) (Kalaiselvan).
The risk factors associated with immediate adverse reactions are female sex (65-81%), previous reactions to GBCAs (7-8%), history of asthma (3-11%), allergic rhinitis (3%) and chronic urticaria (2%), hypersensitivity to drugs (2%).
The clinical manifestations of these reactions () are divided into mild, moderate, and severe. The most frequent skin manifestation is urticaria (63-90%), followed by rash (20%), itching (22%), and transient facial edema (6%).
Rare is the incidence of moderate reactions (severe vomiting, marked urticaria, bronchospasm, facial-laryngeal edema, vasovagal crisis) or severe reactions (hypotensive shock, respiratory and cardiac arrest, convulsions).
After a moderate or severe acute adverse reaction an allergy test should be done to support the evidence of allergy. It can be performed evaluating levels of hystamine and tryptase from blood test within 2 hours after contrast agent injection, or by using prick and intradermal skin tests from 1 to 6 months after the injection.
The risk of adverse reactions is not influenced by the osmolality of the contrast agent, as the low volumes used do not affect the osmolar load. In addition, there is no difference in the incidence of adverse reactions between gadolinium-based extracellular compounds. According to a recent meta-analysis (Behzadi), the rate of immediate allergic adverse events can be associated with protein binding, as linear ionic GBCAs without protein binding showed a lower rate of allergic reactions.
The incidence of deaths from GBCAs is extremely rare with reported value per million dose of 0.15% for gadodiamide, 0.19% for gadoversetimide, 0.97% for gadopentetate dimeglumine, 0.002% for gadobenate dimeglumine, 0.7% for gadoteridol. In a recent meta-analysis (Behzadi) data were collected from 9 articles, with only 2 deaths reported, which corresponds to a mortality rate of approximately 2.8 per 1 million administrations.
acute adverse reactions
•transient vasovagal reactions
•diffuse and marked urticaria
•facial laryngeal edema
•vasovagal reaction that requires treatment
A late adverse reaction to contrast agent is defined as a reaction which occurs 1 h to 1 week after injection. Late skin reactions have not been described after GBCAs. A variety of late symptoms (nausea, vomiting, headache, musculoskeletal pains, fever) have been described following the injection, but many are not related to the contrast medium.
Adverse reactions that occur more than a week after administration of the contrast are defined as very late.
For gadolinium compounds there is only one type of very late adverse reaction, called nephrogenic systemic fibrosis. It can start a few days after exposure for up to 2-3 months; rarely, it can occur even after years. It consists of a series of early alterations (pain, itching, edema, erythema of the skin, which usually originates from the lower limbs) and late manifestations (fibrotic thickening of the skin and subcutaneous tissues, fibrosis of the internal organs). Cases of death have been reported due to severe involvement of internal organs.
What has one learned over the years about the possible link between GBCAs and NSF?
•There is a correlation between GBCAs and NSF: in fact, no new cases have been registered since August 2008 or after the pubblication of the guidelines by the American College of Radiology and the European Society of Urology regarding the use of GBCAs in patients with renal insufficiency.
•The onset of NSF is recorded only in patients with severe renal insufficiency, especially with estimated Glomerular Filtration Rate (eGFR) < 15 ml/min/1.73 m2 and in patients on hemodialysis.
•Cases of NSF have never been recorded in patients with normal renal function.
•GBCAs with low-kinetic stability present a higher risk of developing NSF. Gadodiamide is held responsible for most cases of NSF. NSF also occurred after use of gadopentetate dimeglumine and gadoversetamide. The estimated incidence in patients with severe renal failure was of 3-18% after gadodiamide and of 0.1-1% after gadopentetate dimeglumine.
•Higher cumulative doses of GBCAs with low-kinetic stability are linked to higher risk of NSF in patients with severe renal insufficiency, but NSF is also possible with a single dose of contrast of gadolinium compounds.
•Consider a possible alternative test to contrast-based-MR not only in terms of sensitivity and specificity, but also in terms of safety and risk of morbidity and mortality, as MR examinations without contrast can offer, in certain circumstances, comparable diagnostic information to the investigation with contrast agents.
Post-contrast acute kidney injury is defined as an increase in serum creatinine > 0.3 mg/dl (or > 26.5 μmol/l) or > 1.5 times the baseline, within 48-72 hours of intravascular administration of a contrast medium.
The risk of PC-AKI is very low when using gadolinium-based contrast agents in approved doses.
To reduce the risk of acute reactions to gadolinium-based contrast agents, the following precautions should be taken.
•For patients with a high risk of adverse reactions, the use of a unhenanced investigation should be considered. In case of previous allergic reactions to a given GBCAs, a different compound should be used, preferably after consulting an allergist specialist.
•In general, there is no scientific evidence in favour of premedication, which is therefore not recommended. The fundamental precaution to be implemented is that of having drugs and aids for immediate intervention in the event of an adverse reaction.
•The patient must be kept under observation for about 30 min after the contrast administration.
•Fasting before the administration of gadolinium compounds is not recommended.
•If an adverse reaction occurs, the name of the compound, the dose used, the clinical details of the reaction and the precautions and medications administered must be recorded. If the reaction was particularly severe or unusual, all the details should be sent to the register of the pharmacovigilance authority.
•The extravasation of gadolinium is better tolerated than the extravasation of iodine contrast media, due to the lower osmolarity and very low volumes used. The main risk factors are related to technical problems (for example, use of automatic injectors, suboptimal injection sites) or to the patient (venous fragility, impaired lymphatic and/or venous drainage). In most cases edema and erythema are the most common manifestations; severe manifestations are very rare, such as skin ulceration and soft tissue necrosis. In such cases, early identification is important.
With regard to the evaluation of renal function, in recent years many efforts have been made to clarify problems regarding the correct use of contrast agents. However, some issues still remain open.
•Which is the most reliable factor for measure renal function? eGFR, calculated from the serum Creatinine (sCr), is the recommended method to estimate renal function before contrast agent administration.
•In adults ≥ 18 years the Chronic Kidney Disease-EPIdemiology collaboration (CKD-EPI) formula is recommended to calculate eGFR (ESUR guidelines 10.0 version)
eGFR (ml/min/1.73 m2) =
female sCr ≤ 62 μmol/l: 144 × (sCr/62) − 0.329 × 0.993 age
female sCr > 62 μmol/l: 144 × (sCr/62) − 1.209 × 0.993 age
male sCr ≤ 80 μmol/l: 141 × (sCr/80) − 0.411 × 0.993 age
male sCr > 80 μmol/l: 141 × (sCr/80) − 1.209 × 0.993 age
(sCr in μmol/l; age in years; all equations x 1.159 if African American race)
Alternatively, the Modification of Diet in Renal Disease (MDRD) formula can be used
eGFR (ml/min/1.73 m2) = 175 × serum creatinine (m2/dl) − 1.154 × age (years) − 0.203 × 0.742 (if female patient) and × 1.21 (if African American patient)
•In children, the revised Schwartz formula is recommended to calculate eGFR
eGFR (ml/min/1.73 m2) = 36.5 × length/sCr
(sCr in μmol/l; length in cm)
Serum creatinine values can be normal even in patients with low eGFR, therefore it is not a safe value.
•Which patients should be evaluated for renal function? All patients should be screened, although it is not mandatory in patients who are given low-risk GBCAs.
•Is the questionnaire sufficient? Yes, if the questionnaire contains the evaluation of renal function (eGFR) and information about previous history of moderate or severe reaction to GBCAs, history of atopy requiring treatment, unstable asthma, end-stage renal failure (eGFR < 30 ml/min/1.73 m2) or if the patient is on dialysis.
It is essential also to provide to the patient the informed consent with precise information about methods, purposes, diagnostic alternatives to the MRI investigation and potential risk of NSF following the use of gadolinium-based contrast agents, with adequate illustration of the symptoms of onset. Furthermore, written documentation of consent with signature of acceptance of the investigation or non-acceptance is essential.
Management of the dose
Even if GBCAs are considered very safe, the administrations need some precautions. The dose of the contrast agent must be calculated on the basis of the patient’s weight, avoiding doses > 0.3 mmol/kg. Generally, it is preferable to use low-risk GBCAs.
Dialysis. According to the new ESUR guidelines (2018), a dialysis treatment should be performed as soon as possible after administration of GBCAs in patients already on dialysis. All gadolinium compounds can be removed by hemodialysis or peritoneal dialysis. However, there is no evidence that hemodialysis protects from kidney failure from the AKI-PC and the NSF.
In clinical practice, it may be necessary to use iodine and gadolinium contrast agents on the same day. Since iodine and gadolinium-based contrast agents are excreted within 4 hours of administration, to reduce any potential for nephrotoxicity, the following precautions are recommended:
•in patients with normal or moderately reduced kidney function (eGFR > 30 ml/min/1.73 m2), there should be 4-hour interval between two subsequent injections of iodine and gadolinium contrast agents;
•in patients with severely reduced kidney function (eGFR < 30 ml/min/1.73 m2 or on dialysis), 7 days should pass between injections of iodine and gadolinium contrast agents.
The same rule must be applied for two consecutive gadolinium administrations.
In clinical practice, it may be necessary to face the need to administer contrast to a woman in pregnancy or in breastfeeding.
In general, the cardinal principle that guides the choice of administering a contrast agent is to never deny an examination useful for the diagnosis. Therefore, according to the new guidelines, contrast agent may be given when there is a very strong indication for performing an MR, using the minimum possible dose of a macrocyclic gadolinium contrast agent. For their low-risk profile, no neonatal tests is required after the administration to the mother during pregnancy.
Breastfeeding is considered relatively safe after administration of macrocyclic gadolinium chelates, and it is not necessary to stop breastfeeding or remove milk in the next hours after the administration of the contrast agent.
In women with renal insufficiency, both during pregnancy and breastfeeding, the contrast agent should not be administered.
The authors compared a group of patients, who had undergone at least 6 MR examinations with gadobenate dimeglumine or gadodiamide, with another group that didn’t receive contrast agent. The first had high signal intensity in T1w sequences, finding not visible in the second group of patients. These results were also found and confirmed in subsequent studies, where the accumulation of gadolinium in the brain was confirmed by mass spectrometry.
The main conclusion was that the stability of the various gadolinium compounds is related to the structure of the ligand and influences the risk of deposition in different tissues, mainly in brain, skin and bones. In vitro and non-clinical studies have also shown that linear GBCAs have a higher dissociation constant, as they can release gadolinium from ligand molecules to a greater extent than macrocyclic contrast agents. However, there are currently no data attributing neurological effects, such as cognitive or movement disorders related to brain gadolinium storage. As a consequence of these reports, the Pharmacovigilance Risk Assessment Committee (EMA, 10 March 2017) initially suspended all linear compounds. In July 2017, the initial conclusions were partially revised with the European Medicines Agency (EMA, 7 July 2017), and the final revision of the applications of GBCAs was published, reaffirming the need for evaluate the cost/benefit ratio in advance, as well as the need to use in all patients the lowest dosage of contrast agent necessary for an effective diagnostic result (EMA, 21 July 2017). Linear compound are allowed only if they have hepatospecific use (gadobenate dimeglumine and gadoxetic acid).
Figure 22 – T1w high signal intensity in cerebellar dentate nuclei, in 3 different patients affected by multiple sclerosis, after numerous gadolinium administrations.
On the basis of scientific evidence, the Committee for Medicinal Products for Human Use of the European Medicines Agency has classified gadolinium-based contrast agents on the basis of their safety profile, in particular, on the basis of the risk of occurrence of nephrogenic systemic fibrosis which can be high, intermediate and low. As previously mentioned, GBCA applications have also changed in relation to the greater ability of linear compounds to release gadolinium into the tissues.
•Gadodiamide (Omniscan®); ligand: non-ionic linear chelate (DTPA-BMA*).
•Gadopentetate dimeglumine (Magnevist®); ligand: ionic linear chelate (DTPA).
•Gadoversetamide (Optimark®); ligand: non-ionic linear chelate (DTPA-BMEA**).
For these agents their marketing authorizations for intravenous administration have been suspended in Europe as a precautionary measure and should no longer be used. Optimark® has been withdrawn from the European market. European medicines agency has allowed the use of Magnevist® for arthrography.
*BMA = Bis-MethylAmide.
**BMEA = Bis-MethoxyEthylAmine.
•Gadobenate dimeglumine (Multihance®); ligand: ionic linear chelate (BOPTA).
•Gadoxetate disodium (Primovist®); ligand: ionic linear chelate (DTPA).
The use of these contrast agents is allowed for the study of the liver and where assessment of the hepatospecific phase is necessary. Departments that continue to use these contrast media with indications other than liver imaging should make arrangements to switch to other authorized construction agents. Where Multihance® has been used in relation to its greater relaxation (for example, in the central nervous system, in the breast and in vascular imaging), radiologists will have to determine which dosage of other contrast media should be used to obtain an equivalent improvement of their study protocols, always keeping in mind that the lowest and most effective dose possible should be used.
•Gadoterate meglumine (Dotarem®); ligand: cyclic ionic chelate (DOTA).
•Gadoteridol (Prohance®); ligand: non-ionic cyclic chelate (HP-DO3A***).
•Gadobutrol (Gadovist®); ligand: non-ionic cyclic chelate (BT-DO3A****).
***HP-DO3A = HydroxyPropyl-tetrazacycloDOdecane-triAcetate.
****BT-DO3A = Butrol-TetrazacycloDOdecanetriacetic Acid.
These GBCAs have demonstrated a high safety profile and still remain on the market. In particular, macrocyclic contrast agents (gadobutrol, gadoteric acid and gadoteridol) have maintained their indications as they are more stable. Their use is recommended only when it is not possible to obtain diagnostic information with imaging only without contrast and the lowest and most effective dose for diagnosis should always be used. They can be used even in the pediatric population and in pregnant women to provide essential diagnostic information.
This distinction recognizes as a key element the different stability of gadolinium compounds. Stability affects the different ability of the chelate to keep the gadolinium ion bound and, therefore, is able to influence the number of dissociation reactions and exchange with endogenous ions (transmetallation phenomenon).
Agostini A, Kircher MF, Do R et al: Magnetic resonance imaging of the liver (including biliary contrast agents). Part 1: technical considerations and contrast materials. Semin Roentgenol 2016, 51(4): 308-316. doi: 10.1053/j.ro.2016.05.015.
Bartolozzi C, Donati F, Cioni D et al: Detection of colorectal liver metastases: a prospective multicenter trial comparing unenhanced MRI, MnDPDP-enhanced MRI, and spiral CT. Eur Radiol 2004, 14: 14-20. .
Czeyda-Pommersheim F, Martin DR, Costello JR, Kalb B: Contrast agents for MR imaging. Magn Reson Imaging Clin N Am 2017, 25(4): 705-711. doi: 10.1016/j.mric.2017.06.011.
Dawson P, Cosgrove DO, Grainger RG (eds): Textbook of contrast media. Dunitz, Londra, 2002.
Grazioli L, Morana G, Federle MP et al: Focal nodular hyperplasia: morphologic and functional information from MR imaging with gadobenate dimeglumine. Radiology 221(3): 731-739, 2001.
Grazioli L, Morana G, Kirchin MA, Schneider G: Accurate differentiation of focal nodular hyperplasia from hepatic adenoma at gadobenate dimeglumine-enhanced MR imaging: prospective study. Radiology 2005, 236(1): 166-177.
Guglielmo FF, Mitchell DG, Gupta S: Gadolinium contrast agent selection and optimal use for body MR imaging. Radiol Clin North Am 2014, 52(4): 637-656. doi: 10.1016/j.rcl.2014.02.004.
Laghi A, Paolantonio P, Iafrate F et al: Oral contrast agents for magnetic resonance imaging of the bowel. Top Magn Reson Imaging 2002, 13(6): 389-396. doi: 10.1097/00002142-200212000-00003.
Lencioni R, Della Pina C, Bruix J et al: Clinical management of hepatic malignancies: ferucarbotran-enhanced magnetic resonance imaging versus contrast-enhanced spiral computed tomography. Dig Dis Sci 2005, 50(3): 533-537.
Morana G, Grazioli L, Testoni M et al: Contrast agents for hepatic magnetic resonance imaging. Top Magn Reson Imaging 2002, 13(3): 117-150.
Sigal R, Vogl T, Casselman J et al: Lymph node metastases from head and neck squamous cell carcinoma: MR imaging with ultrasmall superparamagnetic iron oxide particles (Sinerem MR) – results of a phase-III multicenter clinical trial. Eur Radiol 2002, 12: 1104-1113. .
Xiao YD, Paudel R, Liu J et al: MRI contrast agents: classification and application (review). Int J Mol Med 2016, 38(5): 1319-1326. doi: 10.3892/ijmm.2016.2744.
Safety of MRI contrast agents
Behzadi AH, Zhao Y, Farooq Z, Prince MR: Immediate allergic reactions to gadolinium-based contrast agents: a systematic review and meta-analysis. Radiology 2018, 286: 471-482.
Costello JR, Kalb B, Martin DR: Incidence and risk factors for gadolinium-based contrast agent immediate reaction. Top Magn Reson Imaging 2016, 25: 257-263.
European Medicines Agency: PRAC concludes assessment of gadolinium agents used in body scans and recommends regulatory actions, including suspension for some marketing authorisations. Pharmacovigilance Risk Assessment Committee. EMA/157486/2017, 10 March 2017.
European Medicines Agency: PRAC confirms restrictions on the use of linear gadolinium agents. Benefit-risk balance of certain linear gadolinium agents no longer favourable. EMA/424715/2017, 7 July 2017.
European Medicines Agency: EMA’s final opinion confirms restrictions on use of linear gadolinium agents in body scans. Recommendations conclude EMA’s scientific review of gadolinium deposition in brain and other tissues. EMA/457616/2017, 21 July 2017.
Forghani R: Adverse effects of gadolinium based-contrast agent: changes in practice patterns. Top Magn Reson Imaging 2016, 25: 163-169.
Gulani V, Calamante F, Shellock FG et al: Gadolinium deposition in the brain: summary of evidence and recommendations. Lancet Neurol 2017, 16: 564-570.
Iordache AM, Docea AO, Buga AM et al: The incidence of skin lesions in contrast media-induced chemical hypersensitivity. Experimental And Therapeutic Medicine 2019, 17: 1113-1124.
Jost G, Frenzel T, Boyken J et al: Gadolinium presence in the brain after administration of the liver-specific gadolinium-based contrast agent gadoxetate: a systematic comparison to multipurpose agents in rats. Invest Radiol 2019, 54: 468-474.
Kalaiselvan V, Sharma S, Singh GN: Adverse reactions to contrast media: an analysis of spontaneous reports in the database of the pharmacovigilance programme of India. Drug Saf 37: 703-710, 2014.
Kanda T, Ishii K, Kawaguchi H et al: High signal intensity in the dentate nucleus and globus pallidus on unenhanced T1-weighted MR images; relationship with increasing cumulative dose of gadolinium-based contrast material. Radiology 2014, 270: 834-841.
Khawaja AZ, Cassidy DB, Al Shakarchi J et al: Revisiting the risks of MRI with gadolinium based contrast agents: review of literature and guidelines. Insights Imaging 2015, 6: 553-558.
McDonald JS, Hunt CH, Kolbe AB et al: Acute adverse events following gadolinium-based contrast agent administration: a single-center retrospective study of 281 945 injections. Radiology 2019, 292(3): 620-627.
Schieda N, Krishna S, Davenport MS: Update on gadolinium-based contrast agent-enhanced imaging in the genitourinary system. American Journal of Roentgenology 2019, 212: 1223-1233.
Tedeschi E, Caranci F, Giordano F et al: Gadolinium retention in the body: what we know and what we can do. Radiol Med 2017, 122(8): 589-600.
Thomsen HS: Nephrogenic systemic fibrosis: a serious late adverse reactions to gadodiamide. Eur Radiol 2006, 16: 2619-2621.