This article is mentioned in:

Introduction

The liver-specific magnetic resonance imaging (MRI) contrast agent gadolinium-ethoxybenzyl diethylenetriaminepentaacetic acid (Gd-EOB-DTPA, Primovist®; Bayer Schering Pharma, Berlin, Germany) is used to detect focal liver lesions (1–10) and evaluate the biliary tree (11,12). Intravenously injected Gd-EOB-DTPA is gradually absorbed by hepatocytes and finally excreted via the biliary tract. As a result of hepatocyte uptake, normal liver parenchyma exhibit T1 shortening, unlike focal liver lesions, including those caused by hepatic metastasis. Certain experimental animal studies have demonstrated that this contrast material has the potential to evaluate liver function or detect diffuse liver disease (13–18). Experimentally induced hepatic dysfunction decreases the degree of liver enhancement produced by Gd-EOB-DTPA and prolongs the washout of contrast material (17,18). In rats, liver enhancement with Gd-EOB-DTPA during MRI is delayed and prolonged following liver transplantation as compared with controls (14). The correlation between liver function and Gd-EOB-DTPA kinetics has been evaluated in animal models, but never in humans. The livers of patients with liver tumors are occasionally damaged, which might degrade the contrast between liver and lesion produced by the decreased hepatocyte uptake of Gd-EOB-DTPA. The present study investigated the correlation between hepatic function and liver parenchymal enhancement in Gd-EOB-DTPA-enhanced MRI.

Materials and methods

Patient population

The present study was performed in accordance with the recommendations of the Declaration of Helsinki. All patients gave their written, informed consent upon enrollment prior to undergoing MRI. In total, 49 consecutive patients with chronic hepatitis were referred to the Interventional Radiology unit at our institution between March 2, 2008 and June 30, 2009 for curative treatment of hepatocellular carcinoma (HCC). All patients except those with MRI contraindications (n=2), including claustrophobia and presence of a pacemaker, underwent Gd-EOB-DTPA-enhanced MRI prior to receiving treatment for HCC. Six (12%) of the 49 patients were considered ineligible for the study due to multifocal HCC (defined as the presence of >10 tumor nodules, n=5), or inadequate MRI results (unacceptable image quality due to motion artifacts, n=1). The remaining 41 patients who were able to undergo MRI and who met the study criteria (mean age, 71.9 years; range, 38–78 years) comprised the study cohort, which included 32 males (mean age, 73.5 years; range, 59–86) and 9 females (mean age, 65.9 years; range, 38–78). Three patients had hepatitis B, 36 had hepatitis C, and two had alcohol-related hepatitis.

MRI

MR images were obtained using a 1.5-T system (Signa HDx; GE Medical Systems, Milwaukee, WI, USA) with an eight-channel anteroposterior phased-array surface coil placed around the patient and covering the entire liver. Imaging protocols included unenhanced sequences [transaxial T2-weighted fast spin-echo (FSE) and in- and opposed-phase GRE sequences] and Gd-EOB-DTPA-enhanced dynamic three-dimensional (3D)-GRE sequences. Images were captured in the transverse plane during an end-expiratory breath-hold with a combined eight-coil element, anteroposterior phased-array surface coil. Using parallel imaging with sensitivity encoding by a factor of two, the total acquisition time was decreased to approximately 15 sec. A three-quarter field-of-view was used in the phase-encoding direction. Presaturation pulses were applied above and below the imaging volume to diminish flow artifacts. The patients were administered a 30 μmol/kg (0.12 ml/kg body weight) dose of Gd-EOB-DTPA (Primovist) via the antecubital vein at a rate of 2 ml/s through a 22-gauge intravenous catheter using a power injector (Spectris Solaris® EP; Medrad, Indianola, PA, USA), followed by a 40 ml saline flush at the same injection rate. Dynamic and delayed images were obtained using a fat-suppressed 3D T1-weighted GRE sequence with parallel imaging [LAVA™ EFGRE ASSET™ breath-hold; repetition time/echo time (TR/TE), 3.0/1.6 msec; flip angle, 15°; field of view, 42×42 cm; matrix, 384×256 interpolated to 512×512; thickness, 5 mm; overlap 2 mm, ASSET acceleration, 2.0]. T1-weighted dynamic GRE breath-hold images were captured at 30 and 180 sec following contrast material administration during the hepatic arterial dominant and equilibrium phases, respectively, and during the delayed hepatobiliary phase at 10, 20 and 30 min following injection.

Quantitative image analysis

Two radiologists obtained signal intensity values for the liver using a monitor-defined region of interest (>50 pixels) while avoiding major intrahepatic vessels. The location of the region of interest was maintained as constant as possible for individual patients at different time points (Fig. 1). Relative enhancement (RE) of the liver was calculated using the equation: RE = SIpostcontrast/SIprecontrast where SIprecontrast is the signal intensity of the liver on the precontrast image, and SIpostcontrast is the signal intensity of the liver on the postcontrast image.

Figure 1 T1-weighted-GRE images at (A) the pre-contrast and (B) hepatocyte uptake phase (20 min) from a 62-year-old man with hepatocellular carcinoma. The location of the region of interest was maintained as constant as possible for each patient at different time points. GRE, gradient echo.

Liver function parameters

Blood serum parameters (total bilirubin, serum albumin and international normalized ratio of prothrombin time), the 15 min retention rate of indocyanine green test (ICG-R15 test; Daiichi-Sankyou, Tokyo, Japan) and the presence or absence of ascites were recorded to evaluate the degree of liver damage 1–5 days prior to undergoing Gd-EOB-DTPA-enhanced MRI. The degree of liver damage was assessed according to a system similar to the Child-Pugh classification except for inclusion of the ICG test; this evaluation procedure followed the algorithm for HCC treatment guidelines in Japan (19). Aspartate aminotransferase (AST) and alanine aminotransferase (ALT) were also recorded, as these parameters correlated with the degree of liver enhancement in animal studies using Gd-EOB-DTPA-enhanced MRI. Since all 41 patients included were referred from the Interventional Radiology unit, and MRI was part of the workup prior to administering treatment for HCC, these laboratory investigations were undertaken for clinical reasons.

Blood serum parameters, as well as the degree of liver damage, were compared with the quantitative parameter of liver enhancement observed on the Gd-EOB-DTPA images.

Statistical analysis

Data were statistically analyzed using the SPSS software package version 11.0 for Windows (SPSS Inc., Chicago, IL, USA). Continuous variables are presented as the mean and standard deviation (SD) as appropriate. The Student’s t-test was used to compare RE values among respective liver damage scores. Correlations between the continuous variables and the RE values were determined using univariate regression analyses. Independent determinants of RE values were determined by forward multiple stepwise regression analyses. In the two-tailed test, P<0.05 was considered to indicate a statistically significant difference.

Results

Time profile of RE values

Fig. 2 shows the time profile of RE values for the respective liver damage scores. Following injection of Gd-EOB-DTPA, T1-weighted 3D-GRE images revealed an early contrast effect in the liver parenchyma, with a steep increase in signal intensity at 30 sec. A further, although slower, increase for up to approximately 20 min following injection was followed by a plateau of enhancement in the two groups with liver damage. The maximal RE value was selected for each patient (8 and 33 patients at 20 and 30 min, respectively). Maximal RE values did not significantly differ between liver damage levels A and B. None of the study population had ascites or level C liver damage.

Figure 2 Time profile of the mean relative enhancement following administration of Gd-EOB-DTPA. The error bars are the standard deviation. The solid and dotted lines show liver damage levels A and B, respectively. Gd-EOB-DTPA, gadolinium-ethoxybenzyl diethylenetriaminepentaacetic acid.

Correlation between RE values and liver function parameters

Table I provides patient demographic data, and Table II shows the results of univariate and multivariate analyses for determinants of maximal RE values. The parameters that significantly related to maximal RE values were serum albumin (r=0.496, p=0.001), AST (r=−0.366, p=0.023), total bilirubin (r=−0.487, p=0.002) and ICG-R15 (r=−0.462, p=0.003). Multiple stepwise regression analysis revealed that serum albumin (p=0.002) and total bilirubin (p=0.001, inversely) among these parameters remained significantly (Table II) and independently related to maximal RE values (R2=0.287).

Table I Demographic data of the study subjects (n=41).

Table I

Demographic data of the study subjects (n=41).

Factors	Mean ± SD
Age (years)	71.9±9.31
AST (IU/l)	45.1±24.8
ALT (IU/l)	37.8±21.3
Total bilirubin	0.85±0.32
Serum albumin	3.8±0.43
ICG R15 (%)	26.5±13.2
Prothrombin time (%)	77.9±12.8

[i] SD, standard deviation; AST, aspartate aminotransferase; ALT, alanine aminotransferase; ICG R15 (%), 15-min retention rate of indocyanine green test.

Table II Determinants of maximal relative enhancement values by stepwise regression (forward selection).

Table II

Determinants of maximal relative enhancement values by stepwise regression (forward selection).

	Univariatea		Multivariateb
Factors	β	p	β	F	p
Age (years)	−0.002	NS	-	-	-
AST (IU/l)	−0.366	0.023	-	-	-
ALT (IU/l)	−0.002	NS	-	-	-
Total bilirubin	−0.487	0.002	−0.285	4.169	0.001
Serum albumin	0.496	0.001	0.224	4.61	0.002
ICG R15 (%)	−0.462	0.003	-	-	-
Prothrombin time (%)	0.004	NS	-	-	-
R2	-	-	0.287	-	-

a Univariate coefficients,

b Stepwise multivariate regression analysis was performed.

{ label (or @symbol) needed for fn[@id='tfn4-ol-03-05-0990'] } β indicates regression coefficient; F, F value of the overall regression model; NS, not significant; ICG R15 (%), 15-min retention rate of indocyanine green test; AST, aspartate aminotransferase; ALT, alanine aminotransferase.

Discussion

Our findings demonstrate that the degree of liver enhancement by Gd-EOB-DTPA during the hepatic uptake phase in humans depends on liver function. Certain experimental animal studies have described the potential of Gd-EOB-DTPA to evaluate liver function and diffuse hepatic disease (15–18). Hepatocyte uptake and biliary excretion of Gd-EOB-DTPA, bilirubin and ICG appear to be mediated by glutathione-s-transferase (20–22). Kim et al (17) reported that Gd-EOB-DTPA enhancement in animal livers with chemically induced hepatitis correlates with plasma bilirubin level and ICG clearance. Our findings in patients with hepatitis confirmed the results of Kim et al and the physiological mechanism of Gd-EOB-DTPA. The prognosis of patients with cirrhosis and HCC depends on residual liver function and tumor extension (23,24). Various prognostic staging systems for HCC combine the Child-Pugh liver function classification with tumor extension (23,24). The preoperative assessment of functional reserve is significant for estimating the extent of hepatectomy. Hepatic functional reserve is widely assessed using the ICG test, but technetium-99m-galactosyl human serum albumin liver scintigraphy appears to be equally effective for selecting candidates for hepatectomy (23,24). Findings of the current study have shown that Gd-EOB-DTPA-enhanced MRI findings correlated with total bilirubin values and ICG-R15, and that this modality has potential for use in the same manner as scintigraphy as a prognostic staging system and as a parameter for the preoperative assessment of hepatectomy. Shimizu et al (25) identified regional ischemic damage in the rat right hepatic lobe during the hepatic uptake phase of Gd-EOB-DTPA-enhanced MRI. The ability of Gd-EOB-DTPA-enhanced MRI to evaluate regional liver function might contribute to the preoperative assessment of hepatectomy. Tsuda et al (26) reported that the time to reach maximal enhancement and the half-life of such enhancement following Gd-EOB-DTPA injection are prolonged in the livers of rats with non-alcoholic steatohepatitis compared with the livers of rats with common fatty liver. Non-alcoholic steatohepatitis is now considered one of the most common types of chronic liver disease, which induces the development of liver cirrhosis and tumors (27–29). Although non-alcoholic steatohepatitis should be diagnosed early so that therapy begins early, the differences between non-alcoholic steatohepatitis and common fatty liver are not distinguishable by any radiological modality, including ultrasonography, computed tomography (CT) and MRI without contrast material (30,31). Gd-EOB-DTPA-enhanced MRI has the potential to differentially diagnose human non-alcoholic steatohepatitis and common fatty liver. Our study results demonstrate that Gd-EOB-DTPA-enhanced MRI was capable of estimating liver function in humans similar to that observed in animal studies.

The gadolinium-diethylenetriaminepentaacetic (Gd-DTPA) derivative Gd-EOB-DTPA is comparable to Gd-DTPA in terms of being highly hydrophilic and water-soluble. In addition, a lipophilic ethoxybenzyl group enables selective intracellular uptake by hepatocytes (2,18). Therefore, images of the early dynamic perfusion and late hepatocyte uptake phases can be captured following a single injection of Gd-EOB-DTPA (1,2). The first phase within approximately 3 min following a bolus injection is equivalent to that of Gd-DTPA (1,2). Focal lesions are more effectively identified using Gd-EOB-DTPA than contrast-enhanced dynamic-CT, with high diagnostic reliability and superiority (4–10). Although valid direct comparisons of Gd-DTPA and Gd-EOB-DTPA are rare, at least one study has proven that there are similar perfusion-phase tumor enhancement characteristics following the injection of the two contrast materials (2). Focal lesions with hepatocellular function, including focal nodular hyperplasia, adenoma and well-differentiated HCC, all absorb Gd-EOB-DTPA during the hepatocyte uptake phase. However, conditions without such hepatocellular function, including moderately or poorly differentiated HCC and liver metastases, do not absorb contrast material during the uptake phase (4–6,9). These imaging features are useful in characterizing focal lesions (4–6,9). Consequently, the early dynamic perfusion and late hepatocyte uptake phases are useful for detecting and characterizing focal hepatic lesions. In addition, Gd-EOB-DTPA has the same favorable safety profile as Gd-DTPA (1,3,6,7,22). However, the delay in the hepatocyte uptake phase wastes more than 20 min (4,6,8–10). Certain investigators achieve the hepatocyte uptake phase at 10 and 20 min following injection (4,8,9). The uptake rates of Gd-EOB-DTPA are similar at 10 and 20 min in 88 and 90% of focal nodular hyperplasia lesions, respectively (9). Huppertz et al (8) interpreted the enhancement characteristics of focal liver lesions at 10 and 20 min without a distinction in image quality. Differences in tumor-liver contrast-to-noise ratios in 23 liver metastases were not significant between 10 and 45 min following injection (4). The results of the present study suggest that a normally functioning liver uptakes considerable amounts of contrast material during the early period (10 min) following injection, which generates favorable contrast between focal lesions and the surrounding liver. Further studies are required to determine the image acquisition time appropriate to individual liver functions to decrease the duration required during the procedure. Once the time that is currently wasted obtaining the hepatocyte uptake phase image is shortened, Gd-EOB-DTPA-enhanced MRI may become as significant as ultrasonography, CT and MRI in the diagnosis of focal hepatic lesions (32–35).

This study has two limitations. Firstly, as the patient population comprised candidates for HCC therapy, no study subjects had extremely poor liver function (level C liver damage). Secondly, we did not observe Gd-EOB-DTPA washout from the liver. Although Gd-EOB-DTPA is absorbed by hepatocytes and excreted from the rat liver within approximately 60 min (16,17,26), these processes occur over 6 h in humans (1). The washout of Gd-EOB-DTPA is also prolonged in animals with a damaged liver (16,17,26). Human patients would not be able to tolerate such protracted procedures.

In conclusion, the degree of liver enhancement with Gd-EOB-DTPA correlates with the level of liver function. Future clinical investigations are required to further evaluate the usefulness of Gd-EOB-DTPA-enhanced MRI as a test for diffuse liver disease and as a prognostic staging system for HCC. Decreasing the examination duration with rapid hepatocyte uptake phase images should also be investigated for patients with normal liver function.

References