Drs. Chauhan, Sahani, and Saini are from the Division of
Abdominal Imaging and Intervention, Department of Radiology,
Massachusetts General Hospital, Boston, MA.
Liver MR imaging is primarily performed in oncologic patients
who are being worked up for liver resection or lesion ablation, and
it is imperative to provide accurate clinical information to decide
if patient is eligible for the therapy. In these patients, the goal
of liver imaging is to detect the exact number and location of
focal liver lesions.
1,2
MR contrast agents improve the liver MR examination in both of
these respects. Liver lesion characterization is particulary
important for preoperative imaging, since there is a high
prevalence of benign liver tumors.
3
This characterization is usually performed with gadolinium-chelate
(Gd) to study the tumor perfusion profile. MR imaging with
liver-specific contrast agents is a very sensitive technique for
detecting focal liver lesions.
4-6
Three classes of MR contrast agents are now available for liver
imaging: extracellular agents, hepatocyte targeted agents, and
reticuloendothelial system (RES) targeted agents. The purpose of
this review is to provide a practical approach for the use of
contrast agents for liver MR imaging.
Extracellular Agents
Three paramagnetic chelates of gadolinium have been approved for
liver imaging (Magnevist [Berlex Laboratories, Wayne, NJ], Omniscan
[Nycomed Amersham, Princeton, NJ], and ProHance [Bracco
Diagnostics, Princeton, NJ]).
6
At the recommended dose of 0.1 mmol/kg, each of these agents has an
excellent safety profile and can be used in patients with renal
insufficiency.
7
They are used just as iodinated contrast agents are used in CT.
Therefore, since gadolinium increases tissue signal intensity on
T1-weighted images, multiphase dynamic imaging is performed
4,6,8
after bolus administration of the contrast agent. The protocol at
our institution is to perform breath-hold, short TE, T1-weighted
gradient echo (GRE) imaging during the arterial (20 to 30 seconds
postinjection), portal venous (60 to 80 seconds postinjection), and
equilibrium phase (3 minutes postinjection) phases.
Extensive clinical experience with gadolinium chelates supports
its use for lesion detection and characterization. For lesion
characterization, lesion enhancement pattern is analyzed as some of
the liver tumors have a characteristic perfusion pattern. Van beers
et al
8
have reported >90% accuracy in differentiating hepatic
hemangiomas, benign liver cell tumors, and malignant liver lesions
with dynamic Gd-enhanced MR imaging.
Hemangiomas demonstrate dense peripheral nodular blush in the
arterial phase, which progresses centripetally on subsequent phase
images (figure 1). On the other hand, hypovascular metastases show
peripheral band-like enhancement with peripheral wash-out on
equilibrium phase images (figure 2). Hepatocellular carcinomas
(HCC) and hypervascular metastases may show a heterogeneous blush
in the arterial phase.
8
In contrast, focal nodular hyperplasia (FNH) has a more uniform
enhancement in the arterial phase with enhancement of the central
scar on equilibrium phase images (figure 3). With simple cysts, no
enhancement is seen in all phases of dynamic MR imaging. The
application of extracellular agents for lesion detection is most
effective in patients with cirrhosis (figure 4). Tang et al
9
have reported sensitivity and specificity of gadolinium-enhanced MR
imaging in detecting liver lesions is over 90%. In patients with a
normal liver, lesion detection is best acheived with liver-specific
contrast agents described below.
Hepatocyte-specific contrast agents
Hepatobiliary contrast agents are a heterogeneous group of
soluble paramagnetic molecules that are partially taken up by
hepatocytes and then excreted into the bile. At present only one
product (Teslascan, Nycomed, Amersham, Princeton, NJ), a chelate of
paramagnetic manganese, is approved for liver imaging. The
recommended dose is 10 µmol/kg (0.01 mmol/kg), which may be
administered as a rapid infusion over 1 to 2 minutes. It is
reasonably well tolerated, reported minor reactions include facial
flushing, nausea, vomiting, and transient increase in blood
pressure.
10-12
It is best to be cautious with patients with cholestasis, however,
unless extrahepatic cholestasis is corrected by means of adequate
bile drainage.
13
Liver uptake of manganese is rapid, with peak enhancement occurring
within 10 minutes. Teslascan increases liver signal intensity of
normal liver on T1-weighted images, an effect which is best seen on
fat-saturation techniques. Since T2-weighted images are not
affected, our approach is to inject the contrast agent immediately
after pre-contrast T1-weighted images.
Teslascan increases lesion-to-liver contrast because of
selective enhancement of normal liver signal intensity.
10,12
In addition to improved lesion detection, lesion localization is
facilitated by the high vessel-to-liver contrast. However, a
potentially more important application may be for the detection of
HCCs in patients with cirrhosis. In these patients, gadolinium
chelate-enhanced dynamic scanning is used for lesion detection. The
application of Teslascan in cirrhotic liver may be limited, since
the uptake of contrast in cirrhotic livers can be heterogeneous due
to areas of hepatic fibrosis, limiting the accuracy of lesion
detection. Hepatocellular lesions such as HCC, hepatocellular
adenomas (HCA), and FNH, however, often enhance with Teslascan and
may be obscured on postcontrast images.
11,12
In these patients, our limited experience suggests that lesion
detection and characterization is best performed with extracellular
agents.
RES-specific contrast agents
An alternative approach to liver imaging is to use particulate
superparamagnetic iron oxides (Feridex IV, Advanced Magnetics, NJ)
that are rapidly taken up by the RES, which includes hepatic
Kupffer cells. Due to its magnetic susceptibility effect, the agent
produces T2 shortening, thereby decreasing the liver signal
intensity on T2-weighted images.
5,14
This T2 effect of the iron oxide particles in the
reticuloendothelial cells lasts several hours following contrast
injection, therefore, we routinely perform precontrast imaging
first. Subsequently, patient is taken out of the magnet (MRI
gantry), iron oxide is slowly infused, and postcontrast imaging is
performed 30 to 45 minutes postinjection. The recommended dose is
10 mmol/kg administered as a slow infusion over 30 minutes. During
administration, patients should be monitored carefully as back pain
reportedly occurs in 3% to 4% of patients.
14-16
Since most focal liver lesions and tumors are devoid of Kupffer
cells, they maintain their native bright signal intensity after
iron oxide injection. Background liver parenchyma turns dark on
T2-weighted image, however, increasing conspicuity of liver tumors.
14-16
As T2-weighted images are prone to motion artifact, this degrades
liver signal-to-noise ratio. Therefore, postcontrast T2-weighted
imaging requires excellent motion artifact suppressed techniques.
T1-weighted images may also show signal loss, therefore
postcontrast T1 weighted imaging is not as important.
Clinical experience with Feridex has shown that it provides high
sensitivity for preoperative lesion detection.
16,17
Multicenter studies have shown post ferumoxides images to be more
accurate for detecting liver metastases than CT arterial
portography.
18
Ward et al
19
have also shown that MR sensitivity was significantly higher than
that of dual phase CT. Super paramagnetic iron oxides have limited
utility for liver lesion characterization, except for focal nodular
hyperplasia, where the central scar excludes iron particles and is
thus more demonstrable on post ferumoxides images.
20
The application of iron oxide in patients with cirrhosis may be
limited since the signal decrease on T2-weighted images in
cirrhotic livers can be heterogeneous due to areas of hepatic
fibrosis, limiting the accuracy of lesion detection.
9
In addition, a cirrhotic liver takes up less of the contrast agent
with relatively greater uptake by the spleen.
Clinical applications
For patients referred for liver MR, the first determination is
if the primary goal is lesion detection or characterization. When
lesion characterization is required, such a lesion that was
identified on CT or US but inadequately characterized, the
gadolinium protocol should be used. If the objective is lesion
detection for preoperative evaluation or if iodinated contrast
media at CT was contraindicated, then a Teslascan study is
performed.
Although clinical experience with Teslascan is not as large as
with iron oxide, multicenter studies have shown improved lesion
detection compared with unenhanced images.
21
Compared with iron oxides, however, T1-based hepatobiliary agents
are more effective because they show an increase in liver
signal-to-noise ratio. Thus, images are of better quality and
contrast-to-noise benefits are also provided. In addition to
improved lesion detection, hepatocellular tumor characterization is
possible, since these lesions enhance postcontrast injection.
Finally, if a lesion must characterized on a study using
Teslascan, immediate dynamic imaging with gadolinium is possible
(figure 6), although this is rarely required. AR
References
1. Malt R: Surgery for hepatic neoplasms. N Engl J Med
313:1591-1596, 1985.
2. Sugarbaker P: Surgical decision making for large bowel cancer
metastatic to the liver. Radiology 174:621-626, 1990.
3. Karhunen P: Benign hepatic tumors and tumor-like conditions
in men. J Clin Pathol 39:183-187, 1986.
4. Beers BV, Gallez B, Pringot J: Contrast-enhanced MR imaging
of the liver. Radiology 203: 297-306, 1997.
5. Hahn P, Saini S: Liver-specific MR imaging contrast agents.
Radiol Clin North Am 36:287-297, 1998.
6. Mitchell D: Hepatobiliary contrast material: A magic bullet
for sensitivity and specificity? Radiology 188: 21-22, 1993.
7. Nelson K, Gifford L, Lauber-Huber C, et al: Clinical safety
of Gadopentetate dimeglumine. Radiology 196:439-443, 1995.
8. Van Beers B, Demeure R, Pringot J, et al: Dynamic spin-echo
imaging with Gd-DTPA: Value in the differentiation of hepatic
tumors. AJR 154:515-519, 1990.
9. Tang Y, Yamashita Y, Arakawa A, et al: Detection of
hepatocellular carcinoma arising in cirrhotic livers: Comparison of
gadolinium- and ferumoxides-enhanced MR imaging. AJR 172:1547-1554,
1999.
10. Bernardino M, Young S, Lee J, Weinreb J: Hepatic MR imaging
with Mn-DPDP: Safety, image quality, and sensitivity. Radiology
183:53-58, 1992.
11. Murakami T, Baron RL, Peterson MS, et al: Hepatocellular
carcinoma: MR imaging with mangafodipir trisodium (Mn-DPDP).
Radiology 200:69-77, 1996.
12. Rofsky N, Weinreb J, Bernardino M, et al: Hepatocellular
tumors: Characterization with Mn-DPDP-enhanced MR imaging.
Radiology 188:53-59, 1993.
13. Ni Y, Petre C, Lukito G, et al: Effect of manganese
dipyridoxal diphosphate on liver magnetic resonance imaging and
serum bilirubin in rats with removable billiary obstruction. Acad
Radiol 2:300-305, 1995.
14. Ferrucci JT, Stark DD: Iron oxide enhanced imaging of the
liver and spleen: Review of first 5 years. AJR 155:943-950,
1990.
15. Vogl T, Hammerstingl R, Schwarz W, et al: Superparamagnetic
iron oxide-enhanced versus gadolinium-enhanced MR imaging for
differential diagnosis of focal liver lesions. Radiology
198:881-887, 1996.
16. Hagspiel K, Neidl K, Eichenberger A, et al: Detection of
liver metastases: Comparison of superparamagnetic iron
oxide-enhanced and unenhanced MR imaging at 1.5 T with dynamic CT,
intraoperative US, and percutaneous US. Radiology 196:471-478,
1995.
17. Ros P, Freeny P, Harms S, et al: Hepatic MR imaging with
ferumoxides: A multicenter clinical trial of the safety and
efficacy in the detection of focal hepatic lesions. Radiology
196:481-488, 1995.
18. Seneterre E, Taourel P, Bouvier Y, et al: Detection of
hepatic metastases: Ferumoxides-enhanced MR imaging versus
unenhanced MR imaging and CT during arterial portography. Radiology
200:785-792, 1996.
19. Ward J, Naik KS, Guthrie JA, et al: Hepatic lesion
detection: Comparison of MR imaging after the administration of
superparamagnetic iron oxide with dual-phase CT by using
alternative-free response receiver operating characteristic
analysis. Radiology 210:459-466, 1999.
20. Grandin C, Van Beers B, Pauwels S, et al: Ferumoxides and
Tc-99m sulfur colloid: Comparison of the tumor-to-liver uptake in
focal nodular hyperplasia. J Magn Reson Imaging 7(1):125-129,
1997.
21. Rummeny EJ, Ehrenheim C, Gehl HB, et al: Manganese-DPDP as
hepatobiliary contrast agent in MR imaging of liver tumors: Results
of clinical phase II trials in Germany including 141 patients.
Invest Radiol 27:879-886, 1992.