Intravenous contrast agents are important in the performance of liver magnetic resonance (MR) imaging. These agents differ in their physical properties. These differences can be exploited to optimize imaging protocols based on indications for examination. Institutional protocols should be designed to take advantage of the variety of available contrast agent types. Important contrast agent properties including relaxivity and biodistribution are discussed in this article. Practitioners administering contrast agents for MR imaging should be familiar with potential safety issues and establish rational safety guidelines based on available data. Precautions in at-risk populations are discussed, and sample institutional guidelines are provided.
Key points
- •
Intravenous contrast agents are important in the performance of high-quality liver MR imaging.
- •
Contrast agents differ in their physical properties, biodistribution, and safety profiles and these differences can be exploited to optimize MR imaging examinations.
- •
Practitioners administering contrast agents for MR imaging should be familiar with potential safety issues and establish rational guidelines for administration based on available safety and efficacy data.
Introduction
Magnetic resonance (MR) imaging is an important tool for diagnosing and staging disease and for monitoring response to therapy. In the liver, MR imaging is frequently used for the characterization of masses, detection of primary and metastatic lesions, assessment of vascular structures and lesion vascularity, and evaluation of the biliary tree. MR imaging has an important role in the work-up of common clinical conditions, including abdominal pain, jaundice, and abnormal liver function tests.
Compared with other imaging modalities, MR imaging has advantages for liver imaging in terms of the variety of available image contrast mechanisms. Conventional T 1 -weighted and T 2 -weighted imaging have been supplemented over time by additional contrast methods including chemical shift imaging, T 2 */susceptibility-weighted imaging, and diffusion-weighted imaging, providing a wealth of synergistic data for characterizing disease. In addition, dynamic contrast-enhanced (DCE) sequences are considered vital for the characterization of most liver lesions, providing insights into vascular features and improving detection of subtle abnormalities.
Most (although not all) of the contrast agents used clinically in liver MR imaging are composed of the rare earth metal gadolinium bound within an organic chelate. Other types of contrast agents, based on manganese and iron, have been available, but gadolinium-based agents remain the most commonly used. The various contrast agents share many physical features but have important differences in relaxivity, protein binding, and biodistribution. These properties provide critical information for diagnosis in clinical liver imaging. This article reviews the characteristics of contrast agents used in liver MR imaging and their utility in clinical imaging. It also reviews the toxicities associated with these contrast agents and considerations for their use in at-risk populations.
Introduction
Magnetic resonance (MR) imaging is an important tool for diagnosing and staging disease and for monitoring response to therapy. In the liver, MR imaging is frequently used for the characterization of masses, detection of primary and metastatic lesions, assessment of vascular structures and lesion vascularity, and evaluation of the biliary tree. MR imaging has an important role in the work-up of common clinical conditions, including abdominal pain, jaundice, and abnormal liver function tests.
Compared with other imaging modalities, MR imaging has advantages for liver imaging in terms of the variety of available image contrast mechanisms. Conventional T 1 -weighted and T 2 -weighted imaging have been supplemented over time by additional contrast methods including chemical shift imaging, T 2 */susceptibility-weighted imaging, and diffusion-weighted imaging, providing a wealth of synergistic data for characterizing disease. In addition, dynamic contrast-enhanced (DCE) sequences are considered vital for the characterization of most liver lesions, providing insights into vascular features and improving detection of subtle abnormalities.
Most (although not all) of the contrast agents used clinically in liver MR imaging are composed of the rare earth metal gadolinium bound within an organic chelate. Other types of contrast agents, based on manganese and iron, have been available, but gadolinium-based agents remain the most commonly used. The various contrast agents share many physical features but have important differences in relaxivity, protein binding, and biodistribution. These properties provide critical information for diagnosis in clinical liver imaging. This article reviews the characteristics of contrast agents used in liver MR imaging and their utility in clinical imaging. It also reviews the toxicities associated with these contrast agents and considerations for their use in at-risk populations.
Contrast agent properties
Table 1 lists a variety of intravenous gadolinium-based contrast agents (GBCAs) used in clinical imaging. Most of these agents are not approved by the US Food and Drug Administration (FDA) specifically for liver imaging. Most of these agents have received FDA approval for central nervous system applications (gadoterate meglumine, gadobutrol, gadopentetate dimeglumine, gadobenate dimeglumine, gadodiamide, gadoversetamide, and gadoteridol). Some are approved for general body indications (gadopentetate dimeglumine and gadodiamide), whereas others are approved for specific liver indications (gadoxetate disodium and gadoversetamide) or aortoiliac and peripheral vascular disease (gadofosveset trisodium).
| Agent | Trade Name | Manufacturer | r 1 at 1.5 T (37 ° C) in Blood (L mmol −1 s −1 ) | Chemical Structure |
|---|---|---|---|---|
| Gadofosveset trisodium | Ablavar | Lantheus Medical Imaging | 19 | Linear |
| Gadoterate meglumine | Dotarem | Guerbet | 4.2 | Cyclic |
| Gadoxetate disodium | Eovist | Bayer Healthcare | 7.3 | Linear |
| Gadobutrol | Gadavist | Bayer Healthcare | 5.3 | Cyclic |
| Gadopentetate dimeglumine | Magnevist | Bayer Healthcare | 4.3 | Linear |
| Gadobenate dimeglumine | MultiHance | Bracco Diagnostics | 6.7 | Linear |
| Gadodiamide | Omniscan | GE Healthcare | 4.6 | Linear |
| Gadoversetamide | OptiMARK | Mallinckrodt Pharmaceuticals | 5.2 | Linear |
| Gadoteridol | Prohance | Bracco Diagnostics | 4.4 | Cyclic |
Ferumoxytol, an iron-based agent, is not currently approved for any imaging indications, although its use has been reported in patients who cannot receive a GBCA or for other unique applications, such as brain tumor, atherosclerotic plaque, and lymph node imaging. Although other iron-based agents (ferucarbotran, ferumoxide) and manganese-based agents (mangafodipir trisodium) have shown utility in liver and lymph node imaging, these compounds are no longer available in the United States.
Relaxivity
The GBCAs are paramagnetic substances that induce changes in the local magnetic field, which in turn shortens the T 1 , T 2 , and T 2 * relaxation times of nearby nuclei. Their T 1 -shortening property is the most clinically useful, and the enhancement effect seen on T 1 -weighted images is used to detect the presence of GBCA and thus to evaluate tissue perfusion characteristics. The effect on T 2 relaxation by GBCAs is very small and is typically ignored. For gradient echo pulse sequences with very short echo times, the effect on T 2 * can also generally be ignored. Pulse sequences can be modified to take advantage of the T 2 *-shortening effect to precisely measure T 2 *, and by extension the concentration of GBCA, but this application is most frequently used in dynamic susceptibility contrast (DSC) perfusion measures in the brain, and is rarely used in clinical liver imaging.
Compared with GBCAs, iron-based agents are ferromagnetic substances and have a much stronger T 2 *-shortening effects, and signal loss can be used to differentiate tissues that take up these agents to greater or lesser degrees. In addition, ferumoxytol causes adequate T 1 shortening to provide strong enhancement on T 1 -weighted images for vascular imaging.
The amount of subjective and quantitative enhancement seen on T1-weighted images after contrast administration depends on tissue concentration of the contrast agent, agent relaxivity, and magnetic field strength. The relaxivity constant r 1 describes the inherent enhancement power of an agent, independent of the amount of the contrast material present. Most GBCAs have r 1 values between 4.2 L mmol −1 s −1 and 5.3 L mmol −1 s −1 in blood at 1.5 T. However, because of transient binding interactions with serum albumin, 3 commonly used agents have higher r 1 relaxivities: gadobenate dimeglumine (6.7 L mmol −1 s −1 ), gadoxetate disodium (7.3 L mmol −1 s −1 ), and gadofosveset trisodium (19 L mmol −1 s −1 ). Because the enhancement effect also depends on the concentration of the agent, this advantage is reduced or negated for gadoxetate disodium and gadofosveset trisodium at FDA-approved doses, because their approved dosing (0.025 mmol/kg for gadoxetate disodium; 0.03 mmol/kg for gadofosveset trisodium) is much lower than that of the other GBCAs (0.1 mmol/kg).
In addition to being modified by protein-binding effects, r 1 relaxivity also depends on magnetic field strength, and decreases at higher field strength. Although this results in lower r 1 values at 3 T compared with 1.5 T for all of the GBCAs, tissue T 1 times are also substantially longer at high field strength, resulting in a net increase in enhancement at 3 T for GBCAs. Thus there is generally thought to be an advantage to performing contrast-enhanced MR imaging and MR angiography at 3 T, although other effects including increased susceptibility, energy deposition caused by radiofrequency pulses, and the dielectric effect must be considered.
Biodistribution
Extracellular contrast agents
Known as extracellular or nonspecific contrast agents, most of the GBCAs in clinical use share a similar biodistribution with regard to the liver and other organs. Soon after bolus intravenous injection, the extracellular agents first reach the liver via the hepatic arteries. The result is mild enhancement of the hepatic parenchyma with preferentially greater enhancement of those solid liver lesions with arterial-dominant vascular supply, including focal nodular hyperplasia (FNH) and most hepatocellular carcinomas (HCCs) and hepatocellular adenomas (HCAs). This phenomenon is known as arterial hyperenhancement and typically occurs at about 20 to 30 seconds following intravenous contrast media administration. The hepatic arterial phase is particularly important for lesion detection and characterization, as well as delineation of normal and variant hepatic arterial anatomy.
Soon after the hepatic arterial phase, additional contrast medium arrives in the liver via the portal vein from the splenic and bowel vasculature, causing further hepatic enhancement and creating the portal venous phase appearance. In this phase, many arterially enhancing lesions become less enhanced than the liver because of strong enhancement of the liver and weak additional enhancement of these lesions from the portal venous supply, resulting in the washout appearance. The portal venous phase typically occurs at about 60 to 70 seconds following intravenous contrast media administration.
Following the portal venous phase, the administered extracellular GBCA recirculates and equilibrates between arterial and venous distributions, as well as the extravascular, extracellular interstitium, resulting in the interstitial or equilibrium phase appearance. In this phase, washout of hypervascular liver lesions may be better depicted than in the portal venous phase. The equilibrium phase may first be seen approximately 90 seconds following intravenous contrast injection in many patients. The preferred timing for the equilibrium phase can be variable from institution to institution. The choice of timing depends on balancing the greater sensitivity for washout against reduced liver enhancement/signal-to-noise ratio with later timing. Most institutions have reported 3 to 8 minutes as their preferred postcontrast timing for the equilibrium phase.
Taken together, the portal venous and equilibrium phases are used for both liver lesion characterization and assessment of the portal and hepatic veins. Liver lesions that maintain enhancement similar to the blood pool are often (but not always) benign hemangiomas or vascular shunts, whereas hypervascular lesions with washout often represent more ominous entities, such as HCCs and metastases.
Representative liver MR imaging performed with an extracellular contrast agent is shown in Fig. 1 .
Hepatobiliary agents
Two commercially available GBCAs, gadobenate dimeglumine and gadoxetate disodium, have the additional properties of hepatocyte uptake and biliary excretion, and thus have been termed hepatobiliary, hepatocyte-specific, or liver-specific agents. Both initially distribute in a manner similar to the extracellular contrast agents and show both the arterial phase enhancement and portal venous phase washout features of liver lesions.
After reaching the hepatic sinusoids/interstitium, these agents are taken up directly into functioning hepatocytes by adenosine triphosphate (ATP)–driven cellular receptors, and are further excreted into the biliary canaliculi. It has been shown that gadoxetate disodium is taken up into hepatocytes by the organic anion-transporting polypeptide family of transport proteins and excreted into the biliary system by multidrug resistance–associated protein-2. Hyperintensity or hypointensity of liver lesions in the hepatobiliary phase of enhancement reflects the expression of these receptors or, in the case of FNH, trapping of the agent within disordered biliary canaliculi. Approximately 5% of the gadobenate dimeglumine injected is taken up by the normal liver, whereas 50% of injected gadoxetate disodium is taken up by the normal liver. After a variable period of time, the so-called hepatocyte or hepatobiliary phase is reached, in which T1-weighted imaging the liver and bile ducts are enhanced because of GBCA accumulation, whereas the blood vessels are dark because of GBCA clearance by the liver and kidneys. In this phase, many solid lesions (hemangiomas and most HCCs and HCAs) are visualized as hypointense to the liver parenchyma because of poor GBCA accumulation, whereas other lesions (FNH and a minority of HCCs) are isointense or hyperintense to the hepatic parenchyma.
Because of the higher liver extraction fraction of gadoxetate disodium, hepatocyte phase imaging can be performed at around 20 minutes after injection (or sooner) in most patients, compared with 1 to 2 hours after injection for gadobenate dimeglumine. The strength and timing of peak hepatic enhancement depend on the individual patient’s expression of uptake receptors, presence and severity of chronic liver disease, and other factors. As a result, the precise time at which an adequate hepatocyte phase is achieved is highly patient dependent.
Because hepatocyte extraction of hepatobiliary contrast agents begins with the first pass of the agents through the liver, this phenomenon can be superimposed on the washout phenomenon. For gadobenate dimeglumine, hepatocyte uptake is a slow process, and thus washout can reliably be assessed independently of hepatocyte uptake for several minutes following injection. However, hepatocyte uptake with gadoxetate disodium occurs much more rapidly. As a result, a true equilibrium or interstitial phase may not be observed with gadoxetate disodium, but rather a transitional phase is seen in which the appearance of the liver and liver lesions is determined by a combination of vascular equilibrium, interstitial leakage, and hepatocyte uptake.
In addition, imaging of the biliary tree has been described in the hepatobiliary phase using these contrast agents. This method may have value for showing sites of bile duct injury and bile leakage. The T1-weighted MR cholangiogram may also be complimentary to T 2 -weighted MR cholangiopancreatography sequences for evaluating the biliary system, particularly if the T 2 -weighted sequences are compromised by excessive/irregular respiration or other artifacts.
Representative liver MR imaging performed with a hepatobiliary contrast agent is shown in Fig. 2 .
