DOI: 10.1148/rg.266065014
RadioGraphics 2006;26:1621-1636
© RSNA, 2006
MR Contrast Agents for Liver Imaging: What, When, How1
Sunil N. Gandhi, MD,
Michèle A. Brown, MD,
James G. Wong, MD,
Diego A. Aguirre, MD2 and
Claude B. Sirlin, MD
1 From the Department of Radiology, UCSD Medical Center, 200 W Arbor Dr, San Diego, CA 92103. Recipient of a Certificate of Merit award for an education exhibit at the 2004 RSNA Annual Meeting. Received February 6, 2006; revision requested March 16 and received May 5; accepted May 8. All authors have no financial relationships to disclose.
Address correspondence to C.B.S. (e-mail: csirlin{at}ucsd.edu).
 |
Abstract
|
|---|
The major classes of contrast agents currently used for magnetic resonance (MR) imaging of the liver include extracellular agents (eg, low-molecular-weight gadolinium chelates), reticuloendothelial agents (eg, ferumoxides), hepatobiliary agents (eg, mangafodipir), blood pool agents, and combined agents. Mechanisms of action, dosage, elimination, toxic effects, indications for use, and MR imaging technical considerations vary according to class. Gadolinium chelates are the most widely used. Ferumoxides are a useful adjunct for detection of hepatocellular carcinoma, particularly when used in combination with gadolinium to achieve improved lesion-to-liver contrast over that achievable with gadolinium alone. Mangafodipir is a prototype hepatobiliary agent that is taken up by lesions with functioning hepatocytes. It may be used for MR cholangiography as well as liver imaging. Although mangafodipir is no longer commercially available in the United States, it is currently marketed and used in Europe. Blood pool agents have not yet been approved for human use in the United States. However, a new combined MR contrast agent, gadobenate dimeglumine, recently was approved, and other agents are in various stages of development.
© RSNA, 2006
|
LEARNING OBJECTIVES FOR TEST 2
|
|---|
After reading this article and taking the test, the reader will be able to:- Describe the mechanisms of action of the major classes of MR contrast agents used in liver imaging.
- Recognize clinical indications for the use of a particular in MR contrast agent lesion detection and characterization.
- Discuss key concepts for the selection and optimization of pulse sequences to maximize contrast and avoid common pitfalls.
|
Introduction
|
|---|
Numerous magnetic resonance (MR) contrast agents are available to improve lesion detection and characterization at MR imaging of the liver. These agents can be divided broadly into the following five classes: extracellular agents, reticuloendothelial agents, hepatobiliary agents, blood pool agents, and combined agents (1). While blood pool and combined agents are mentioned briefly, the following discussion focuses on extra-cellular, reticuloendothelial, and hepatobiliary agents. The article reviews the mechanisms of action, potential adverse effects, and clinical indications for use of specific contrast agents in liver imaging, as well as MR imaging techniques that enhance the diagnostic effect of these agents.
|
Extracellular Agents
|
|---|
Extracellular agents are distributed within the extracellular interstitial space. Gadolinium chelates, which are formed by the chelation of gadolinium to organic ligands such as diethylenetriaminepentaacetic acid, constitute a class of extra-cellular agents (2). Although several formulations are available with different ligands (Table 1), their pharmacologic characteristics and imaging considerations are essentially identical (3,4). Cost varies, depending on the manufacturer, but gadolinium-based agents are the least expensive of the five classes of MR contrast agents discussed in this article (3,4).
Mechanism of Action
Gadolinium has seven unpaired electrons and is highly paramagnetic. Gadolinium shortens the T1 (spin-lattice) and T2 (spin-spin) relaxation times of adjacent water protons. These relaxation effects tend to cause signal enhancement at T1-weighted imaging (Fig 1) and signal loss at T2-weighted imaging (4–6). One important exception is short inversion time inversion recovery (STIR) sequences; with these sequences, T1 shortening due to gadolinium accumulation leads to signal loss (7). T1 shortening predominates at low concentrations of gadolinium, and T2 shortening predominates at high gadolinium concentrations (Fig 2); however, after in vivo administration of clinical doses of gadolinium chelates, the T1 shortening effect is observed in essentially all tissues. Thus, T1 is the imaging property that is routinely evaluated after the administration of extracellular agents (4).
View larger version (88K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1a. T1 shortening due to gadolinium in a 42-year-old woman with focal nodular hyperplasia. (a) Axial T1-weighted spoiled GRE image, obtained before the administration of gadolinium, shows no visible lesion. (b) Axial T1-weighted spoiled GRE image, obtained during the late hepatic arterial phase after gadolinium administration, shows enhancement of a liver lesion (arrow), a finding suggestive of focal nodular hyperplasia.
|
|
View larger version (87K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1b. T1 shortening due to gadolinium in a 42-year-old woman with focal nodular hyperplasia. (a) Axial T1-weighted spoiled GRE image, obtained before the administration of gadolinium, shows no visible lesion. (b) Axial T1-weighted spoiled GRE image, obtained during the late hepatic arterial phase after gadolinium administration, shows enhancement of a liver lesion (arrow), a finding suggestive of focal nodular hyperplasia.
|
|
View larger version (120K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a. Predominant T2 shortening due to gadolinium in a 58-year-old man. (a) Axial T2-weighted echo-train spin-echo image, obtained before the administration of gadolinium, depicts high signal intensity throughout the bladder. (b) Axial gadolinium-enhanced T2-weighted echo-train spin-echo image shows a dependent area with low signal intensity (arrow), a feature indicative of T2 shortening in the bladder because of a high concentration of excreted gadolinium chelate.
|
|
View larger version (108K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b. Predominant T2 shortening due to gadolinium in a 58-year-old man. (a) Axial T2-weighted echo-train spin-echo image, obtained before the administration of gadolinium, depicts high signal intensity throughout the bladder. (b) Axial gadolinium-enhanced T2-weighted echo-train spin-echo image shows a dependent area with low signal intensity (arrow), a feature indicative of T2 shortening in the bladder because of a high concentration of excreted gadolinium chelate.
|
|
Gadolinium chelates and iodinated contrast media share similar pharmacokinetics in the liver and throughout the body. Gadolinium enters the liver via the hepatic artery and portal vein and is freely redistributed from the vascular to the interstitial space. Whereas the iodine molecule is imaged at computed tomography (CT), in MR imaging it is the effect of gadolinium that is assessed rather than the molecule itself. Gadolinium exhibits an amplification effect, in which many adjacent water protons are relaxed by a single gadolinium atom. As a result, MR imaging is orders of magnitude more sensitive to the effect of gadolinium than is CT to the effect of iodine (8). This is advantageous for three reasons. First, subtle areas of contrast material accumulation may be depicted with MR imaging that would not be depicted with CT. Second, compared to the dose of iodine for CT, a far lower dose of gadolinium is required for MR imaging. Third, the blood pool remains visibly more enhanced than the parenchyma on equilibrium-phase images. Thus, blood vessels and hemangiomas tend to appear much brighter at delayed contrast-enhanced MR imaging than at CT (9,10).
Dosage and Elimination
The recommended dose of gadolinium chelate for liver imaging is 0.1 mmol/kg of body weight, or 0.2 mL/kg. A dose of 20 mL is effective in nearly all adults for liver imaging (3). The recommended injection rate is 2–3 mL/sec. Gadolinium is eliminated almost entirely through the kidneys (11). Unlike iodine, extracellular gadolinium chelates are not excreted by the hepatobiliary system and therefore are not observed in the gallbladder in the setting of renal failure (11).
Toxic Effects
A common misconception is that gadolinium chelates are nonnephrotoxic. At the low doses used clinically in MR imaging, gadolinium chelates are safe for patients with renal failure (12,13). However, at high doses similar to those of iodinated contrast material, gadolinium chelates probably are more nephrotoxic than are iodinated contrast agents. 
A 40% incidence of nephrotoxic effects has been reported with high-dose (up to 220 mmol) intraarterial gadolinium administration, because of the high osmolality of gadolinium chelates. As a result, the use of gadolinium chelates is not recommended at the high doses that are necessary for applications such as digital subtraction angiography (14).
At low doses, gadolinium chelates are safe. Potential adverse effects include transient headache, nausea, and emesis (14). Other reactions have been observed, but at frequencies of 1% or less. Anaphylaxis is exceedingly rare; to our knowledge, only one death due to gadolinium-related anaphylaxis has been reported (15). All gadolinium chelates are class C drugs (16); a teratogenic effect has been documented in animal studies, but no controlled human studies have been performed (6,14,16,17). For this reason, gadolinium chelates should be avoided in the first trimester and used later in pregnancy only when needed to establish a critical diagnosis that will significantly impact patient care (18). The effect of gadolinium excreted in breast milk is unknown; however, most manufacturers suggest that patients avoid breast-feeding for 48 hours after gadolinium chelate administration (19).
Indications for Use
There are multiple indications for the use of extracellular contrast agents in MR imaging of the liver (4). These include lesion detection, lesion characterization, and liver vasculature assessment. Gadolinium also may be helpful for the MR imaging evaluation of patients with known or suspected cholangitis (20).
MR Technical Considerations
T1-weighted imaging sequences are almost universally used for MR imaging after gadolinium administration. Improved lesion-to-liver contrast with the use of T2-weighted fast spin-echo imaging for the identification of solid liver lesions was reported previously (16,17), but the results have not been widely reproduced. Chemically selective fat saturation is critical for the evaluation of extra-hepatic processes at gadolinium-enhanced T1-weighted imaging. Although fat saturation is less critical for the evaluation of intrahepatic lesions, it is commonly used for all dynamic MR sequences and for delayed gadolinium-enhanced imaging of the liver because it helps improve the dynamic range (21). STIR sequences generally are not useful for gadolinium-enhanced imaging. Because gadolinium shortens T1 to a level approximating that in fat, an inversion time that nulls the signal from fat also nulls that from the gadolinium-enhanced tissue of interest (Fig 3) (7).
View larger version (112K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3a. Effect of gadolinium when used with a STIR sequence in a 37-year-old woman. (a) Axial T2-weighted STIR image, obtained before gadolinium administration, shows fluid-sensitive signal hyperintensity in the kidneys and to a lesser extent in the liver. The fat signal was intrinsically nulled by the short inversion time (180 msec). (b) Axial T2-weighted STIR image, obtained after gadolinium administration, depicts nulling of signal (arrows) both in fat and in gadolinium-enhanced tissue in the kidneys.
|
|
View larger version (100K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3b. Effect of gadolinium when used with a STIR sequence in a 37-year-old woman. (a) Axial T2-weighted STIR image, obtained before gadolinium administration, shows fluid-sensitive signal hyperintensity in the kidneys and to a lesser extent in the liver. The fat signal was intrinsically nulled by the short inversion time (180 msec). (b) Axial T2-weighted STIR image, obtained after gadolinium administration, depicts nulling of signal (arrows) both in fat and in gadolinium-enhanced tissue in the kidneys.
|
|
MR imaging of the liver should be performed both before and after the administration of extra-cellular contrast material. Dynamic imaging after gadolinium administration is essential for lesion detection and characterization (22). A suggested protocol for MR imaging of the liver is shown in Table 2. Appropriate evaluation of the liver requires imaging during the arterial phase of extra-cellular contrast enhancement, specifically the late hepatic arterial phase, when the portal vein is only slightly enhanced (23).
In the normal liver, the portal vein delivers about 70% of the blood supply and the hepatic artery about 30%, with a small fraction delivered by peribiliary vessels and other collateral networks (24). Most hepatic tumors are fed by the hepatic artery; the disproportionate hepatic arterial flow to tumors compared with that to the normal parenchyma is the foundation for arterial and portal venous phase imaging. In the cirrhotic liver, the situation is more complex because portal venous flow tends to diminish and may be abrogated entirely, and there is a compensatory increase in hepatic arterial flow. Moreover, several cirrhosis-associated tumors, including some dysplastic nodules and some well-differentiated hepatocellular carcinoma (HCC) tumors, may receive most of their blood supply from the portal vein (25).
Three methods may be used to determine the acquisition delay necessary to obtain images during the late hepatic arterial phase: hazarding a "best" guess, fluoroscopic triggering, and timing with a test bolus (19,26,27). A reasonable best-guess estimate of the timing of the late hepatic arterial phase is 15–20 seconds, with adjustments for the injection rate, patient size, and cardiovascular status. The authors of the present article, and others (28), do not recommend the best-guess method because it does not consistently result in optimal timing. As a general rule, it is better to perform the image acquisition too late than too early; if in doubt about the optimal timing, one should err on the side of a longer acquisition delay. The use of fluoroscopic triggering is appropriate only for MR sequences with which the high-contrast central portion of k-space is filled first, at the beginning of the acquisition (26). Fluoroscopic triggering is infrequently used at the authors’ institution because patients may have difficulty hearing instructions for breathing during the fluoroscopic acquisition.
A test bolus, or timing run, provides the most accurate determination of the acquisition delay (19,27,29). The steps involved in performing a timing run are shown in Table 3. A small bolus (1–2 mL) of the contrast agent is administered, followed by a 20-mL saline flush, to determine the time to peak aortic enhancement (TTP). For pulse sequences with sequential 3D k-space encoding schemes in which the center of k-space corresponds to the middle of the acquisition time, the acquisition delay (D) is calculated by using two known variables—the injection time (IT) and acquisition time (AT)—in the following formula: D = IT/2 + TTP – AT/2 (Fig 4). This equation yields the acquisition delay for early hepatic arterial phase imaging. To achieve late hepatic arterial phase imaging (which generally is superior to early hepatic arterial phase imaging for liver tumor depiction), the authors routinely add 4 seconds to the delay estimated with the equation. The extra 4 seconds permit the transit of contrast material from the aorta to the target tissue (eg, liver tumor). For pulse sequences without sequential 3D k-space encoding, the formula can be generalized to D = IT/2 + TTP – TTC, where TTC is the time to center, or the time interval from the beginning of the acquisition to central k-space encoding.
View larger version (23K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4. Diagram shows the procedure used to calculate the delay from gadolinium (Gd) injection to peak aortic enhancement for pulse sequences with 3D sequential k-space encoding. The authors routinely add 4 seconds to the calculated delay to account for the transit of contrast material from the aorta to the target tissue (eg, liver tumor). AT = acquisition time, IT = injection time, TTP = time to peak aortic enhancement.
|
|
|
Reticuloendothelial Agents
|
|---|
Reticuloendothelial agents target the reticuloendothelial system, particularly the liver and spleen. The uptake of such agents, like that of technetium 99m sulfur colloid in nuclear medicine, reflects the number of functioning macrophages (30,31). Reticuloendothelial agents currently in clinical use include superparamagnetic iron oxide (SPIO) particles. These dextran-coated iron-based particles are 30–150 nm in diameter (32). Ferumoxides (Feridex; Berlex, Wayne, NJ) are the only reticuloendothelial agent approved for use in the United States (8). Ferucarbotran (Resovist; Schering Diagnostics, Berlin, Germany) was approved for use in Europe and Japan and is undergoing phase III clinical trials in the United States. In general, SPIO particles are the most expensive of the contrast agent classes discussed in this article.
Mechanism of Action
SPIO particles are phagocytosed by macrophages throughout the body but are preferentially entrapped by Kupffer cells, which line the hepatic sinusoids. The reason for this affinity is unclear, but it may be related to the particle size (1,33).
SPIO particles act as a negative contrast agent. Their superparamagnetic properties cause local magnetic field inhomogeneity and result in considerable T2 and T2* shortening (30). Tissues that accumulate SPIO particles thus show reduced signal intensity, particularly on T2- and T2*-weighted images (4,22), and to a lesser extent on T1-weighted images (34,35). While applying experimental sequences with an echo time (TE) of less than 0.1 msec, the authors anecdotally observed that SPIO particles caused mild signal intensity enhancement due to weak T1 shortening effects.
Dosage and Elimination
The recommended dose of ferumoxides is 0.56 mg/kg body weight, or 0.05 mL/kg (36). The manufacturer recommends dilution of the dose in 100 mL of a 5% dextrose solution and intravenous infusion over 30–60 minutes. Imaging is typically performed 1–4 hours after infusion. Elimination occurs through intrinsic iron metabolism. The liver appears darkest on T2*- and T2-weighted images in the first 24 hours after the infusion of ferumoxides. Complete metabolism requires 14–28 days, but the signal intensity of the liver typically returns to the normal (unenhanced) level within 7–14 days (30,37).
Toxic Effects
The most common complication of the administration of SPIO particles is acute severe low back pain (4%), which leads to discontinuation of the infusion in 2.5% of patients (38,39). Occurrences of this complication can be minimized with a slow infusion of the agent (13). At the authors’ institution, SPIO particles have been administered to more than 1500 subjects since January 2000. There have been no serious adverse effects. The utility of SPIO particles in imaging of patients with hemochromatosis has not been studied but nevertheless is questionable, since the liver in a patient with iron deposition disease already shows a signal loss at MR imaging.
Labeled a class C drug, ferumoxides should not be used in pregnant patients. The effect on nursing infants of ferumoxides excreted in breast milk is unknown.
Indications for Use
Most liver tumors—whether benign or malignant, primary or metastatic—are deficient in Kupffer cells and do not exhibit SPIO particle uptake (40,41). Thus, after an infusion of SPIO particles, liver tumors appear relatively hyperintense because the background liver darkens preferentially. The primary exception to this rule occurs in focal nodular hyperplasia, in which SPIO particles may accumulate, with a resultant isointense or even hypointense appearance of lesions in comparison with the normal liver parenchyma (42).
SPIO particles are indicated to varying degrees for the detection of HCC and metastases and the characterization of focal nodular hyperplasia (4). At the authors’ institution, they are used most routinely to aid in the detection of HCC in high-risk patients. The detection of HCC in cirrhotic patients may be difficult with gadolinium alone, because of the parenchymal changes (fibrosis and regenerating nodules) and altered liver perfusion (collateral vessels, increased hepatic arterial flow relative to portal venous flow) caused by cirrhosis (43). The use of SPIO particles may help improve HCC detection in such patients (Fig 5).
View larger version (179K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a. Increased conspicuity of HCC in a 54-year-old man after the administration of SPIO particles and gadolinium, relative to conspicuity with SPIO alone. (a) Axial T1-weighted spoiled GRE image shows a lesion in the right hepatic lobe (arrow), a finding that is difficult to characterize without the use of SPIO particles or gadolinium. (b) Axial T1-weighted spoiled GRE image, obtained after the injection of SPIO particles and before that of gadolinium, shows negative enhancement of the background liver tissue and clearly depicts the lesion (arrow). (c) Axial T1-weighted spoiled GRE image, obtained during the late hepatic arterial phase of enhancement after the administration of SPIO particles and gadolinium, shows increased conspicuity of the lesion (arrow).
|
|
View larger version (179K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b. Increased conspicuity of HCC in a 54-year-old man after the administration of SPIO particles and gadolinium, relative to conspicuity with SPIO alone. (a) Axial T1-weighted spoiled GRE image shows a lesion in the right hepatic lobe (arrow), a finding that is difficult to characterize without the use of SPIO particles or gadolinium. (b) Axial T1-weighted spoiled GRE image, obtained after the injection of SPIO particles and before that of gadolinium, shows negative enhancement of the background liver tissue and clearly depicts the lesion (arrow). (c) Axial T1-weighted spoiled GRE image, obtained during the late hepatic arterial phase of enhancement after the administration of SPIO particles and gadolinium, shows increased conspicuity of the lesion (arrow).
|
|
View larger version (181K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5c. Increased conspicuity of HCC in a 54-year-old man after the administration of SPIO particles and gadolinium, relative to conspicuity with SPIO alone. (a) Axial T1-weighted spoiled GRE image shows a lesion in the right hepatic lobe (arrow), a finding that is difficult to characterize without the use of SPIO particles or gadolinium. (b) Axial T1-weighted spoiled GRE image, obtained after the injection of SPIO particles and before that of gadolinium, shows negative enhancement of the background liver tissue and clearly depicts the lesion (arrow). (c) Axial T1-weighted spoiled GRE image, obtained during the late hepatic arterial phase of enhancement after the administration of SPIO particles and gadolinium, shows increased conspicuity of the lesion (arrow).
|
|
SPIO particles have shown utility for the detection of metastases in some patient series (44); however, they are not routinely used for this purpose at our institution. The characterization of focal nodular hyperplasia also is possible with SPIO particles, but their utility is unpredictable (42,45,46). Some focal nodular hyperplasias exhibit SPIO uptake, a feature that helps corroborate the diagnosis (Fig 6). However, not all focal nodular hyperplasias accumulate SPIO particles, and some well-differentiated HCC tumors, dysplastic nodules, and hepatic adenomas may exhibit uptake.
View larger version (110K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6a. Characterization of a liver mass with the use of SPIO particles and gadolinium in a 52-year-old woman with renal cell carcinoma. (a) Axial 2D T1-weighted GRE image, obtained before the administration of SPIO particles and gadolinium, demonstrates a mass (arrow) with signal isointense to that of liver. (b) Axial 2D T1-weighted GRE image, obtained after the injection of SPIO particles, shows a signal intensity loss in the mass (arrow). (c, d) Axial 3D T1-weighted GRE images, obtained after the administration of SPIO particles and gadolinium, demonstrate homogeneous avid enhancement of the mass during the late hepatic arterial phase (arrow in c) and prompt washout of gadolinium during the portal venous phase (arrow in d). The accumulation of SPIO particles in the mass is suggestive of focal nodular hyperplasia, a diagnosis confirmed by the gadolinium enhancement characteristics, and helps exclude metastasis. Although uptake of SPIO particles is suggestive of focal nodular hyperplasia, the differential diagnosis includes well-differentiated HCC, dysplastic nodule, and hepatic adenoma.
|
|
View larger version (121K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6b. Characterization of a liver mass with the use of SPIO particles and gadolinium in a 52-year-old woman with renal cell carcinoma. (a) Axial 2D T1-weighted GRE image, obtained before the administration of SPIO particles and gadolinium, demonstrates a mass (arrow) with signal isointense to that of liver. (b) Axial 2D T1-weighted GRE image, obtained after the injection of SPIO particles, shows a signal intensity loss in the mass (arrow). (c, d) Axial 3D T1-weighted GRE images, obtained after the administration of SPIO particles and gadolinium, demonstrate homogeneous avid enhancement of the mass during the late hepatic arterial phase (arrow in c) and prompt washout of gadolinium during the portal venous phase (arrow in d). The accumulation of SPIO particles in the mass is suggestive of focal nodular hyperplasia, a diagnosis confirmed by the gadolinium enhancement characteristics, and helps exclude metastasis. Although uptake of SPIO particles is suggestive of focal nodular hyperplasia, the differential diagnosis includes well-differentiated HCC, dysplastic nodule, and hepatic adenoma.
|
|
View larger version (87K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6c. Characterization of a liver mass with the use of SPIO particles and gadolinium in a 52-year-old woman with renal cell carcinoma. (a) Axial 2D T1-weighted GRE image, obtained before the administration of SPIO particles and gadolinium, demonstrates a mass (arrow) with signal isointense to that of liver. (b) Axial 2D T1-weighted GRE image, obtained after the injection of SPIO particles, shows a signal intensity loss in the mass (arrow). (c, d) Axial 3D T1-weighted GRE images, obtained after the administration of SPIO particles and gadolinium, demonstrate homogeneous avid enhancement of the mass during the late hepatic arterial phase (arrow in c) and prompt washout of gadolinium during the portal venous phase (arrow in d). The accumulation of SPIO particles in the mass is suggestive of focal nodular hyperplasia, a diagnosis confirmed by the gadolinium enhancement characteristics, and helps exclude metastasis. Although uptake of SPIO particles is suggestive of focal nodular hyperplasia, the differential diagnosis includes well-differentiated HCC, dysplastic nodule, and hepatic adenoma.
|
|
View larger version (101K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6d. Characterization of a liver mass with the use of SPIO particles and gadolinium in a 52-year-old woman with renal cell carcinoma. (a) Axial 2D T1-weighted GRE image, obtained before the administration of SPIO particles and gadolinium, demonstrates a mass (arrow) with signal isointense to that of liver. (b) Axial 2D T1-weighted GRE image, obtained after the injection of SPIO particles, shows a signal intensity loss in the mass (arrow). (c, d) Axial 3D T1-weighted GRE images, obtained after the administration of SPIO particles and gadolinium, demonstrate homogeneous avid enhancement of the mass during the late hepatic arterial phase (arrow in c) and prompt washout of gadolinium during the portal venous phase (arrow in d). The accumulation of SPIO particles in the mass is suggestive of focal nodular hyperplasia, a diagnosis confirmed by the gadolinium enhancement characteristics, and helps exclude metastasis. Although uptake of SPIO particles is suggestive of focal nodular hyperplasia, the differential diagnosis includes well-differentiated HCC, dysplastic nodule, and hepatic adenoma.
|
|
MR Technical Considerations
SPIO particles may be used alone or in combination with gadolinium (43,47). Contrast-enhanced MR imaging with the use of SPIO particles should be performed with T2- or T2*-weighted sequences. However, this technique has important intrinsic disadvantages. First, lesion characterization is limited because most lesions do not exhibit SPIO uptake and thus appear hyperintense on T2-weighted images. Second, well-differentiated HCC may accumulate SPIO particles and thus appear invisible against the normal liver background (48). Conflicting results have been reported for the accuracy of HCC detection with the use of gadolinium versus that with the use of ferumoxides at single-agent contrast-enhanced MR imaging (49,50). However, it was reported that, in at least one series, gadolinium-enhanced imaging was superior to ferumoxides-enhanced imaging for the detection of small HCC tumors (51), albeit with reduced specificity (52). Given these conflicting results and the increased cost and time of imaging with ferumoxides, we believe that gadolinium is preferable to SPIO particles as a single contrast agent used for the detection of HCC at MR imaging.
In our opinion, SPIO particles are most useful when combined with gadolinium to create a double-contrast effect. With this technique, the SPIO particles are infused first and are followed later by an infusion of gadolinium. The two agents synergistically improve lesion-to-liver contrast on dynamic T1-weighted images because the background liver is darkened by the SPIO particles while the lesion of interest is lightened by gadolinium (43,53).
The depiction of HCC, in particular, may be improved with the use of both contrast agents (19). The SPIO particles improve the visibility of hypovascular HCC tumors that would be missed at imaging with gadolinium alone (Fig 7), and the gadolinium allows detection of hypervascular, well-differentiated HCC tumors that would be missed at imaging with the use of SPIO particles alone. Disadvantages of the combined-contrast technique include higher costs due to the use of a second contrast agent, a slight increase in the overall imaging time, and impaired visibility of gadolinium washout. Some investigators question the incremental benefit of using SPIO particles for imaging of a cirrhotic liver, because the chronically diseased liver is frequently iron laden and thus has short intrinsic T2 and T2* relaxation times (4,28). However, the authors have consistently observed significant benefits with the use of SPIO particles at MR imaging, even in cirrhotic livers (Fig 5). A suggested protocol for double-contrast MR imaging is shown in Table 4.
View larger version (164K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7a. TE-dependent negative contrast enhancement effect of SPIO particles in a 56-year-old man with cirrhosis and HCC. (a) Axial 2D T1-weighted GRE image, obtained with a TE of 2.2 msec after the administration of SPIO particles, depicts a mild signal intensity loss in liver tissue and consequent mild enhancement of contrast between the liver and HCC (arrow), which does not accumulate SPIO particles. The effect was increased on images obtained with a TE of 6.6 msec (b) and 8.8 msec (c).
|
|
View larger version (164K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7b. TE-dependent negative contrast enhancement effect of SPIO particles in a 56-year-old man with cirrhosis and HCC. (a) Axial 2D T1-weighted GRE image, obtained with a TE of 2.2 msec after the administration of SPIO particles, depicts a mild signal intensity loss in liver tissue and consequent mild enhancement of contrast between the liver and HCC (arrow), which does not accumulate SPIO particles. The effect was increased on images obtained with a TE of 6.6 msec (b) and 8.8 msec (c).
|
|
View larger version (165K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7c. TE-dependent negative contrast enhancement effect of SPIO particles in a 56-year-old man with cirrhosis and HCC. (a) Axial 2D T1-weighted GRE image, obtained with a TE of 2.2 msec after the administration of SPIO particles, depicts a mild signal intensity loss in liver tissue and consequent mild enhancement of contrast between the liver and HCC (arrow), which does not accumulate SPIO particles. The effect was increased on images obtained with a TE of 6.6 msec (b) and 8.8 msec (c).
|
|
Imaging before SPIO administration is not routinely performed at the authors’ institution in high-risk patients undergoing evaluation for HCC. Instead, because the liver-darkening effect of SPIO particles is more pronounced as the TE increases, gradient-recalled-echo (GRE) imaging is performed with multiple variations of the TE to assess the SPIO particle uptake indirectly (Fig 7). If it is determined that imaging without SPIO particle–based enhancement would yield additional information, unenhanced imaging may be performed 1–2 weeks later to allow clearance of the previously administered SPIO particles.
Pitfalls
While MR imaging with SPIO particles is most useful in the setting of cirrhosis according to some authors, heterogeneous SPIO uptake may complicate interpretation (54,55). Severe liver dysfunction with or without cirrhosis may result in poor overall SPIO uptake, which would limit the utility of the agent (54). In such cases, the spleen demonstrates preferential SPIO uptake. As a result, the spleen darkens to a greater extent than does the liver. The effect is analogous to the "colloid shift" in nuclear imaging with technetium 99m sulfur colloid (Fig 8).
View larger version (130K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8a. Effect of SPIO particles in a 72-year-old man with cirrhosis and severe liver dysfunction. (a) Axial 2D T1-weighted GRE image, obtained with a TE of 4.4 msec after the administration of SPIO particles, shows an MR imaging analog of "colloid shift": The signal loss in the spleen (arrow) is greater than that in the liver. SPIO particles accumulate in the spleen because of Kupffer cell dysfunction in the setting of severe liver disease. (b) Axial 2D T1-weighted GRE image, obtained with a TE of 8.8 msec after the administration of SPIO particles, shows increased contrast between the liver and spleen (arrow).
|
|
View larger version (121K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8b. Effect of SPIO particles in a 72-year-old man with cirrhosis and severe liver dysfunction. (a) Axial 2D T1-weighted GRE image, obtained with a TE of 4.4 msec after the administration of SPIO particles, shows an MR imaging analog of "colloid shift": The signal loss in the spleen (arrow) is greater than that in the liver. SPIO particles accumulate in the spleen because of Kupffer cell dysfunction in the setting of severe liver disease. (b) Axial 2D T1-weighted GRE image, obtained with a TE of 8.8 msec after the administration of SPIO particles, shows increased contrast between the liver and spleen (arrow).
|
|
SPIO particles cause signal intensity loss; thus, SPIO-enhanced images tend to be signal poor. Depending on the indication, the voxel size or the acquisition time may have to be increased to compensate. Signal intensity loss in the liver also may lead to obscuration of the intrahepatic bile ducts by blooming artifacts. Accordingly, MR cholan-giopancreatographic sequences should be performed before SPIO particles accumulate in the liver. Hepatic signal intensity loss also may result in the partial obscuration of focal liver lesions and cause underestimation of lesion size.
|
Hepatobiliary Agents
|
|---|
Hepatobiliary agents are paramagnetic compounds that are taken up by functioning hepatocytes and excreted in bile. Agents of this class increase the signal intensity of the liver, bile ducts, and some hepatocyte-containing lesions at T1-weighted imaging (56–58).
Manganese is chelated to dipyridoxyl diphosphate to produce the prototype hepatobiliary agent known as mangafodipir trisodium (Tesla-scan; Nycomed Amersham, Oslo, Norway). Regrettably, this agent was recently removed from the U.S. market, possibly because of insufficient demand; however, it is still marketed and used in Europe. Moreover, future hepatobiliary agents are under development that likely will have properties similar to those of mangafodipir trisodium (56,59).
Mechanism of Action
As a result of its five unpaired electrons, manganese is moderately paramagnetic. It shortens the T1 and T2 relaxation times of water protons. T1 shortening predominates at low manganese concentrations, resulting in high signal intensity on T1-weighted images; and T2 shortening predominates at high concentrations, resulting in low signal intensity on T2-weighted images (Fig 9) (56).
View larger version (130K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9a. T2 shortening predominates with a high concentration of manganese in a 34-year-old woman evaluated because of a history of biliary obstruction. (a) Coronal T2-weighted single-shot echo-train spin-echo image demonstrates high signal intensity in the common bile duct (arrow) and mild biliary dilatation. (b) Coronal T2-weighted single-shot echo-train spin-echo image, obtained after the administration of mangafodipir, shows a signal intensity loss in the common bile duct (arrow), a finding secondary to a high concentration of manganese.
|
|
View larger version (115K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9b. T2 shortening predominates with a high concentration of manganese in a 34-year-old woman evaluated because of a history of biliary obstruction. (a) Coronal T2-weighted single-shot echo-train spin-echo image demonstrates high signal intensity in the common bile duct (arrow) and mild biliary dilatation. (b) Coronal T2-weighted single-shot echo-train spin-echo image, obtained after the administration of mangafodipir, shows a signal intensity loss in the common bile duct (arrow), a finding secondary to a high concentration of manganese.
|
|
Dipyridoxyl diphosphate has a chemical structure similar to that of vitamin B6 (60). After intravenous injection, mangafodipir trisodium is taken up by functioning hepatocytes through vitamin B6 receptors. Extrahepatic uptake is observed when some of the manganese dissociates from its ligand within the blood circulation. Uptake of free Mn+2 ions occurs in the pancreas, heart, and liver through nonspecific transport mechanisms (56, 57,60).
Dosage and Elimination
The recommended adult dose of mangafodipir is 5 µmol/kg body weight, which corresponds to 0.5 mL/kg. The dose is administered with a relatively slow injection over 10–20 minutes. Hepatic enhancement begins at approximately 1 minute after administration, peaks at approximately 15 minutes, and persists for several hours (56).
Elimination occurs through the biliary system (59% is eliminated within 5 days) (Fig 10) and to a lesser extent through the kidneys (15% in 24 hours) (43). Renal excretion may occur preferentially in patients with hepatic insufficiency (Fig 11). Biliary excretion is first visible at 5 minutes after injection. Complete delineation of the biliary system may require more than 15 minutes (61, 62).
View larger version (97K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10a. Biliary excretion of manganese in a 42-year-old man. (a) Axial 2D T1-weighted GRE image, obtained before mangafodipir administration, shows low signal intensity in the common hepatic duct. (b–d) Axial contrast-enhanced 2D T1-weighted GRE images, obtained at 5 minutes (b), 10 minutes (c), and 15 minutes (d) after mangafodipir administration, show the common hepatic duct with faint enhancement at 10 minutes (arrow in c) and strong enhancement at 15 minutes (arrow in d). Parenchymal liver enhancement is apparent at 5 minutes (b) and steadily increases with time over the 15-minute observation period.
|
|
View larger version (110K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10b. Biliary excretion of manganese in a 42-year-old man. (a) Axial 2D T1-weighted GRE image, obtained before mangafodipir administration, shows low signal intensity in the common hepatic duct. (b–d) Axial contrast-enhanced 2D T1-weighted GRE images, obtained at 5 minutes (b), 10 minutes (c), and 15 minutes (d) after mangafodipir administration, show the common hepatic duct with faint enhancement at 10 minutes (arrow in c) and strong enhancement at 15 minutes (arrow in d). Parenchymal liver enhancement is apparent at 5 minutes (b) and steadily increases with time over the 15-minute observation period.
|
|
View larger version (112K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10c. Biliary excretion of manganese in a 42-year-old man. (a) Axial 2D T1-weighted GRE image, obtained before mangafodipir administration, shows low signal intensity in the common hepatic duct. (b–d) Axial contrast-enhanced 2D T1-weighted GRE images, obtained at 5 minutes (b), 10 minutes (c), and 15 minutes (d) after mangafodipir administration, show the common hepatic duct with faint enhancement at 10 minutes (arrow in c) and strong enhancement at 15 minutes (arrow in d). Parenchymal liver enhancement is apparent at 5 minutes (b) and steadily increases with time over the 15-minute observation period.
|
|
View larger version (118K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10d. Biliary excretion of manganese in a 42-year-old man. (a) Axial 2D T1-weighted GRE image, obtained before mangafodipir administration, shows low signal intensity in the common hepatic duct. (b–d) Axial contrast-enhanced 2D T1-weighted GRE images, obtained at 5 minutes (b), 10 minutes (c), and 15 minutes (d) after mangafodipir administration, show the common hepatic duct with faint enhancement at 10 minutes (arrow in c) and strong enhancement at 15 minutes (arrow in d). Parenchymal liver enhancement is apparent at 5 minutes (b) and steadily increases with time over the 15-minute observation period.
|
|
View larger version (128K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11a. Renal excretion of manganese after mangafodipir administration in a 51-year-old man with liver dysfunction. Axial (a) and sagittal oblique (b) 3D T1-weighted spoiled GRE images demonstrate the renal excretion of mangafodipir (arrow) in the setting of liver failure. The liver appears less enhanced than normal, a result of hepatocyte dysfunction. Variable enhancement of the renal cortex, adrenal gland, and gastric mucosa is normal and is due to the nonspecific uptake of dissociated manganese ions.
|
|
View larger version (101K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11b. Renal excretion of manganese after mangafodipir administration in a 51-year-old man with liver dysfunction. Axial (a) and sagittal oblique (b) 3D T1-weighted spoiled GRE images demonstrate the renal excretion of mangafodipir (arrow) in the setting of liver failure. The liver appears less enhanced than normal, a result of hepatocyte dysfunction. Variable enhancement of the renal cortex, adrenal gland, and gastric mucosa is normal and is due to the nonspecific uptake of dissociated manganese ions.
|
|
Toxic Effects
Potential adverse effects of mangafodipir are mild and transient and include flushing, nausea, increased blood pressure and heart rate, and dizziness (63,64). Transaminase and bilirubin levels may increase transiently (65).
Mangafodipir is labeled a class C drug and should not be administered to pregnant women. The effect on nursing infants of excreted mangafodipir in breast milk is unknown (56).
Indications for Use
Because mangafodipir is taken up only by hepatocytes, hepatocyte-containing masses (eg, HCC, focal nodular hyperplasia, and hepatic adenoma) appear enhanced in 75% of cases, whereas metastases generally are not enhanced (56). Thus, one major indication for the use of mangafodipir is to characterize lesions as hepatocellular or nonhepatocellular (Figs 12, 13). Another major indication is surveillance of the liver for metastases, particularly in patients with colorectal adenocarcinoma (4).
View larger version (125K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12a. Positive contrast enhancement due to manganese in a 46-year-old woman with focal nodular hyperplasia. (a) Axial 2D T1-weighted GRE image, obtained before mangafodipir administration, shows an ill-defined mass (arrows) in the liver. (b) Axial 2D T1-weighted GRE image, obtained after mangafodipir administration, demonstrates positive contrast enhancement due to manganese uptake in the liver lesion (arrows). This finding helps confirm the presence of hepatocytes within the lesion and, thereby, exclude the presence of a metastasis.
|
|
View larger version (136K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12b. Positive contrast enhancement due to manganese in a 46-year-old woman with focal nodular hyperplasia. (a) Axial 2D T1-weighted GRE image, obtained before mangafodipir administration, shows an ill-defined mass (arrows) in the liver. (b) Axial 2D T1-weighted GRE image, obtained after mangafodipir administration, demonstrates positive contrast enhancement due to manganese uptake in the liver lesion (arrows). This finding helps confirm the presence of hepatocytes within the lesion and, thereby, exclude the presence of a metastasis.
|
|
View larger version (118K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13a. Negative enhancement characteristics of focal nodular hyperplasia and hemangioma after mangafodipir administration. (a, b) Axial 2D T1-weighted spoiled GRE images obtained in a 45-year old woman. Image obtained before mangafodipir administration (a) demonstrates an ill-defined hepatic mass (arrow). Image obtained after mangafodipir administration (b) shows enhancement of contrast between normal liver tissue and the mass (arrow), which was known to represent focal nodular hyperplasia. Note the absence of uptake in the central region of scar tissue, which lacks hepatocytes. (c, d) Axial 2D T1-weighted spoiled GRE images obtained in a 38-year-old woman. Image obtained before mangafodipir administration (c) demonstrates a low-signal-intensity hepatic mass (arrow). Image obtained after mangafodipir administration (d) depicts enhancement of the liver parenchyma and an absence of manganese uptake in the mass (arrow), a nonspecific feature that could be indicative of either a metastasis or a hemangioma. Dynamic gadolinium-enhanced images (not shown) helped confirm the diagnosis of a hemangioma.
|
|
View larger version (117K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13b. Negative enhancement characteristics of focal nodular hyperplasia and hemangioma after mangafodipir administration. (a, b) Axial 2D T1-weighted spoiled GRE images obtained in a 45-year old woman. Image obtained before mangafodipir administration (a) demonstrates an ill-defined hepatic mass (arrow). Image obtained after mangafodipir administration (b) shows enhancement of contrast between normal liver tissue and the mass (arrow), which was known to represent focal nodular hyperplasia. Note the absence of uptake in the central region of scar tissue, which lacks hepatocytes. (c, d) Axial 2D T1-weighted spoiled GRE images obtained in a 38-year-old woman. Image obtained before mangafodipir administration (c) demonstrates a low-signal-intensity hepatic mass (arrow). Image obtained after mangafodipir administration (d) depicts enhancement of the liver parenchyma and an absence of manganese uptake in the mass (arrow), a nonspecific feature that could be indicative of either a metastasis or a hemangioma. Dynamic gadolinium-enhanced images (not shown) helped confirm the diagnosis of a hemangioma.
|
|
View larger version (123K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13c. Negative enhancement characteristics of focal nodular hyperplasia and hemangioma after mangafodipir administration. (a, b) Axial 2D T1-weighted spoiled GRE images obtained in a 45-year old woman. Image obtained before mangafodipir administration (a) demonstrates an ill-defined hepatic mass (arrow). Image obtained after mangafodipir administration (b) shows enhancement of contrast between normal liver tissue and the mass (arrow), which was known to represent focal nodular hyperplasia. Note the absence of uptake in the central region of scar tissue, which lacks hepatocytes. (c, d) Axial 2D T1-weighted spoiled GRE images obtained in a 38-year-old woman. Image obtained before mangafodipir administration (c) demonstrates a low-signal-intensity hepatic mass (arrow). Image obtained after mangafodipir administration (d) depicts enhancement of the liver parenchyma and an absence of manganese uptake in the mass (arrow), a nonspecific feature that could be indicative of either a metastasis or a hemangioma. Dynamic gadolinium-enhanced images (not shown) helped confirm the diagnosis of a hemangioma.
|
|
View larger version (122K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13d. Negative enhancement characteristics of focal nodular hyperplasia and hemangioma after mangafodipir administration. (a, b) Axial 2D T1-weighted spoiled GRE images obtained in a 45-year old woman. Image obtained before mangafodipir administration (a) demonstrates an ill-defined hepatic mass (arrow). Image obtained after mangafodipir administration (b) shows enhancement of contrast between normal liver tissue and the mass (arrow), which was known to represent focal nodular hyperplasia. Note the absence of uptake in the central region of scar tissue, which lacks hepatocytes. (c, d) Axial 2D T1-weighted spoiled GRE images obtained in a 38-year-old woman. Image obtained before mangafodipir administration (c) demonstrates a low-signal-intensity hepatic mass (arrow). Image obtained after mangafodipir administration (d) depicts enhancement of the liver parenchyma and an absence of manganese uptake in the mass (arrow), a nonspecific feature that could be indicative of either a metastasis or a hemangioma. Dynamic gadolinium-enhanced images (not shown) helped confirm the diagnosis of a hemangioma.
|
|
A third indication relies on the excretion of mangafodipir through the biliary system. Contrast-enhanced cholangiography may be performed with mangafodipir trisodium when unenhanced MR cholangiopancreatography does not lead to a diagnosis. This generally requires delayed imaging at 20 minutes or more after administration of the contrast agent. Specific indications include evaluation for suspected biliary leak (65) and diagnosis of biliary atresia as the cause of neonatal cholestasis (66).
MR Technical Considerations
Mangafodipir should be administered before imaging with gadolinium. The authors recommend 2D axial T1-weighted or 3D multiplanar spoiled GRE imaging at 3, 5, and 10 minutes (for lesion detection and characterization) or at 5, 10, 15, and 20 minutes (for biliary evaluation) after the administration of mangafodipir. One then may proceed with the protocol described previously for gadolinium-enhanced liver imaging. Because mangafodipir enhances the liver signal, mangafodipir-enhanced images are intrinsically signal rich, which permits high-resolution acquisition.
Pitfalls
Because of the T2 shortening effect of manganese, mangafodipir-enhanced MR cholangiography should be performed after unenhanced MR cholangiography; images obtained with routine T2-weighted MR cholangiopancreatographic sequences after mangafodipir administration may be of insufficient quality for diagnosis (Fig 9) (56,61). In patients in whom the presence of a biliary leak is suspected, the intrinsically high T1 signal intensity of a hematoma and of any iodinated contrast material administered during conventional cholangiography may mimic a bile leak at mangafodipir-enhanced cholangiography. In addition, a significant delay in image acquisition may be required at mangafodipir-enhanced cholangiography in the setting of hepatic insufficiency (64).
Top
Abstract
LEARNING OBJECTIVES FOR TEST...
