HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2393042218v1 240/1/116    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kim, S. H. Articles by Choi, B. I. Search for Related Content PubMed PubMed Citation Articles by Kim, S. H. Articles by Choi, B. I.

Hepatic Macrosteatosis: Predicting Appropriateness of Liver Donation by Using MR Imaging—Correlation with Histopathologic Findings¹

¹ From the Department of Radiology (S.H.K., J.M.L., J.K.H., J.Y.L., K.H.L., C.J.H., K.S. Shin, B.I.C.), Institute of Radiation Medicine (J.M.L., J.K.H., B.I.C.), Department of Surgery (J.Y.J., N.J.Y., K.S. Suh), and Department of Pathology (S.Y.J.), Seoul National University Hospital, 28 Yongon-dong, Chongno-gu, Seoul 110-744, Korea; Department of Radiology, Seoul National University Bundang Hospital, Seoul, Korea (K.H.L.); and Chungnam National University Hospital, DaeJeon, Korea (K.S. Shin). Received December 31, 2004; revision requested March 9, 2005; revision received March 23; accepted April 1; final version accepted August 1. Address correspondence to J.M.L. (e-mail: leejm{at}radcom.snu.ac.kr).

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
Purpose: To retrospectively evaluate the diagnostic performance of magnetic resonance (MR) imaging in predicting the appropriateness of liver donation in potential living liver donors by using histopathologic results as the reference standard.

Materials and Methods: This study was approved by institutional review board; all patients gave informed consent for the use of MR data for future research. Fifty-seven potential liver donors (40 male, 17 female; age range, 17–57 years; mean age, 32 years) underwent dual-echo 1.5-T MR imaging. Two radiologists qualitatively graded each MR image, with consensus for disagreements. Livers were assigned one of three degrees of hepatic steatosis on the basis of changes in hepatic signal intensity (SI) between in-phase and opposed-phase images. For quantitative analysis, a third radiologist calculated mean hepatic and mean splenic SI by averaging 25 hepatic regions of interest and three splenic regions of interest. Relative SI decrease (RSID) in the liver on opposed-phase images compared with in-phase images was calculated. Linear regression analysis was used to correlate RSID with the degree of total steatosis, macrosteatosis, and microsteatosis. Diagnostic performance for predicting the appropriateness of liver donation was analyzed.

Results: Histologic findings of macrosteatosis resulted in 52 patients being categorized as appropriate donors, with the remaining five being categorized as inappropriate donors. RSID was correlated with total steatosis (r = 0.850). When the RSID criterion for inappropriateness of liver donation was set at 20%, the sensitivity, specificity, and accuracy were 100%, 92.3%, and 93%, respectively. When RSID was used, four livers that had been misclassified as inappropriate for transplantation were found to have microsteatosis of various degrees and a less than moderate degree of macrosteatosis at histologic analysis. Qualitative and quantitative analyses were comparably accurate.

Conclusion: When an RSID criterion of less than 20% was used, dual-echo MR imaging facilitated the correct prediction of appropriateness of liver donation in 53 of 57 patients.

© RSNA, 2006

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
Living related donor liver transplantation is an innovative surgical technique that has profoundly affected organ waiting list times and has decreased waiting list mortality (1–3). Preoperative and radiologic considerations of donor selection for living donor liver transplantation include the presence and degree of hepatic steatosis, the presence of focal or other diffuse abnormalities, volumetric estimation of the liver, and information about variations in the vascular (arterial, portal, hepatic venous, and biliary) anatomy.

Of these considerations, perhaps most important is determining the presence and degree of hepatic steatosis. Excessive hepatic steatosis places the recipient at risk for primary dysfunction of the graft and affects the recovery of the donor after partial hepatectomy (4–7). Although the effect of microvesicular steatosis on liver injury remains controversial, a general consensus currently exists in the hepatic transplantation community that microsteatosis has less clinical importance than macrosteatosis, while moderate to severe (30%) macrosteatosis increases the risk of postoperative complications and patient death after liver transplantation (4–7).

Although the ability of using various imaging techniques to detect and quantify the degree of steatosis is well established (8–14), relatively few studies have been published in the literature on the radiologic assessment of the degree of steatosis in liver donors (10,15). Currently, dynamic contrast material–enhanced computed tomography (CT) and magnetic resonance (MR) imaging have been used as the primary preoperative imaging modalities for the evaluation of liver donors. Because MR cholangiography does not require the administration of contrast material (unlike CT cholangiography), MR imaging has been used as the sole imaging modality for comprehensive noninvasive evaluation of living liver donors (16–18). To our knowledge, however, only one study on the use of MR imaging to predict donor appropriateness with respect to hepatic steatosis has been published in the literature (15). Furthermore, this article did not address whether MR imaging could qualitatively demonstrate hepatic (macrovesicular or microvesicular) steatosis.

The purpose of our study was to retrospectively evaluate the diagnostic performance of MR imaging in predicting the appropriateness of liver donation in potential living liver donors by using histopathologic results as the reference standard.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
Patient Population
Between October 2002 and April 2004, 62 consecutive potential living liver donors who underwent preoperative MR imaging for liver donation were identified from a surgical database for liver transplantation. Among these 62 donors, five were excluded because only MR cholangiopancreatographic images were obtained. After the above exclusion, 57 donors (age range, 17–57 years; mean age, 32 years), including 40 male donors (age range, 17–49 years; mean age, 29 years) and 17 female donors (age range, 19–57 years; mean age, 38 years) in whom dual-echo chemical shift MR images and pathologic specimens were available, were retrospectively included in our study. Fifty of these patients had undergone hepatic resection for liver donation (right hemihepatectomy, 34 patients; left hemihepatectomy, seven patients; left lateral segmenectomy, five patients; extended right hepatectomy, three patients; and extended left hepatectomy, one patient). The remaining seven patients who were suspected of having hepatic steatosis and who subsequently underwent liver biopsy did not undergo surgery because of moderate (30%–59%) macrosteatosis (three patients), severe (60%) macrosteatosis (two patients), or complex hepatic vascular or biliary anatomy (two patients). The mean interval between MR imaging and biopsy was 14.5 days ± 18.8 (standard deviation) (range, 0–124 days). This study was performed in accordance with institutional review board guidelines of the Seoul National University Hospital and was approved by the institutional review board. Before undergoing MR imaging, all patients gave informed consent to allow their MR imaging data to be used for future research purposes.

Liver Biopsy and Evaluation
Liver specimens were obtained in all 57 patients. Surgical hepatic specimens that were 1 cm³ in volume were obtained in 50 patients who underwent hepatic resection. Hepatic specimens were obtained from an area adjacent to the resection margin (right lobe, 37 patients; left lobe, 13 patients). Eleven patients who were suspected of having fatty changes at preoperative imaging, which included ultrasonography (US), CT, and MR imaging, underwent intercostal or subcostal liver biopsy of the right lobe with US guidance by using an 18-gauge automatic biopsy gun (ACECUT; TSK Laboratory, Tochigi, Japan). Four of these 11 patients underwent both percutaneous and surgical biopsy, while the remaining seven patients underwent percutaneous biopsy only. On average, 2.3 liver biopsies (range, 1–3) were performed. Specimens measuring 2 cm in length or longer were fixed in formalin and were stained with hematoxylin-eosin stain.

Sixty-one biopsy slides in 57 patients were examined retrospectively by a pathologist (S.Y.J.) who had 5 years of experience in evaluating the particular type (macrovesicular and microvesicular) and extent of steatosis. Hepatocytes that contained one large vacuole of fat that displaced the nuclei to the periphery of the cell were considered to be macrovesicular fat deposits. When the cytoplasm contained many small fatty inclusions and the nuclei remained in the center of the cell, steatosis was classified as microvesicular (19). Assessment of the quality of hepatic fat was determined by whether the deposited fat was macrovesicular or microvesicular. Macrosteatosis and microsteatosis were evaluated qualitatively in 10 consecutive fields (magnification x25). Steatosis, in addition to being evaluated with qualitative analysis, was classified into four quantitative groups on the basis of the percentage of hepatocytes that had fat vacuoles within the cytoplasm (normal, less than 5%; mild, between 5% and 29%; moderate, between 30% and 59%; and severe, 60% or more) (20).

Patients whose biopsy specimens demonstrated macrovesicular steatosis were divided into two groups (appropriate or inappropriate) on the basis of donor suitability according to the 30% cutoff established by several authors and adopted by our institution (4,6,21,22).

MR Imaging
MR imaging was performed by using one of two 1.5-T MR imaging systems (Magnetom Vision Plus [n = 44] or Sonata [n = 13]; Siemens, Erlangen, Germany). In the first year of this study, the Magnetom Vision Plus imager was used, and in the later part of the study the Sonata imager was used. In-phase and opposed-phase images were obtained by using dual-echo chemical shift gradient-echo MR imaging. Imaging parameters were applied as a default setting according to the optimized protocol recommended by the manufacturer. For the Magnetom Vision Plus unit, imaging parameters included a repetition time of 128 msec, dual-echo time of 2.7 and 5.3 msec, receiver bandwidth of 480 Hz/pixel, and acquisition time of 20 seconds. For the Sonata MR unit, imaging parameters included a repetition time of 110 msec, dual-echo time of 2.4 and 4.8 msec, receiver bandwidth of 380 Hz/pixel, and acquisition time of 18 seconds. Other imaging parameters were identical for both units, including a 70° flip angle, 154 x 256 matrix, one signal acquired, and 30–35-cm field of view with a body-array coil. Images were obtained in the transverse plane with a 6-mm section thickness and a 1.5-mm intersection gap. A dual-echo acquisition, during which in-phase and opposed-phase images were obtained during the same breath hold, was performed in all 57 patients.

MR Image Interpretation
All MR images were interpreted on a picture archiving and communications system workstation (PACS; Marotech, Seoul, Korea) by using qualitative and quantitative methods in order to determine the presence and degree of hepatic steatosis. All three of the radiologists who participated in MR image interpretation were blinded to the histopathologic and surgical findings.

For qualitative analysis, two abdominal radiologists (J.M.L. and J.Y.L., with 12 and 10 years of experience, respectively) independently evaluated MR images for the presence and degree of hepatic steatosis by noting the signal intensity (SI) of the liver relative to that of the paravertebral muscle and spleen on opposed-phase images. Discrepancies in qualitative MR image interpretations were resolved by consensus. In addition to the spleen, which is generally accepted as a reference to quantify the degree of hepatic steatosis, we also used the SI of the paravertebral muscle as one of the references to subdivide the degree of hepatic steatosis on opposed-phase MR images.

In previous studies, it has been well established that the longest T1 relaxation times at 1.5-T MR imaging, in descending order, are found in the spleen (1026–1057 msec), paravertebral muscle (856–891 msec), and liver (570–586 msec) (23,24). The expression of this phenomenon is that, on T1-weighted MR images, the highest SI is found in the liver, followed by the paravertebral muscle and the spleen.

On the basis of these results, if the SI of the liver was higher than that of both the paravertebral muscle and the spleen on opposed-phase images, the liver was considered to be normal. If the SI of the liver was higher than that of the paravertebral muscle and lower than that of the spleen, a diagnosis of mild steatosis was made. Therefore, when the SI of the liver was higher than or the same as that of the paravertebral muscle and spleen on opposed-phase images, the liver was considered to be appropriate for transplantation. However, if the SI of the liver was lower than that of both the paravertebral muscle and the spleen, the liver was considered to be more than moderately steatotic and inappropriate for transplantation.

Each MR image of the liver was classified according to the degree of steatosis by using these visual observations. In addition, hepatic SI on opposed-phase images was compared with hepatic SI on in-phase images, and the degree of steatosis was qualitatively graded as normal when SI increased on opposed-phase images, as mild when SI did not change, and as more than moderate when SI decreased on opposed-phase images. Accordingly, if hepatic SI on opposed-phase images was higher than or the same as that on in-phase images, the liver was considered to be appropriate for transplantation; if hepatic SI on opposed-phase images was lower than that on in-phase images, the liver was considered to be inappropriate for transplantation.

For quantitative analysis, one abdominal radiologist (S.H.K., with 7 years of experience) who did not participate in qualitative MR image analysis performed all MR imaging measurements. Twenty-five regions of interest (ROIs) were placed in both lobes of the liver on five transverse sections at different hepatic levels (five ROIs per section) (10). The ROI values were then averaged to obtain the mean hepatic SI. The spleen was used as an internal standard because it is present on most images of the liver and because it does not contain fat that is observable at MR imaging (25). The mean splenic SI was calculated by averaging three random ROI values of splenic SI on three transverse sections at different levels of the spleen (one ROI per section) (10).

We applied a different number of ROIs to the liver (25 ROIs) and to the spleen (three ROIs) because of the different volumes. ROI measurements were obtained in an area that was as large as possible and that contained only hepatic or splenic parenchyma, with no large vessels or biliary trees. The size of the ROIs ranged from 100 to 400 mm². The percentage decrease in normalized hepatic SI on opposed-phase images relative to that on in-phase images was calculated by using the following formula and was defined as the relative SI decrease (RSID): RSID = 100 · (L_in/S_in – L_op/S_op)/(L_in/S_in), where L_in and S_in represent SI on in-phase images of the liver and spleen, respectively, and L_op and S_op represent SI on opposed-phase images of the liver and spleen, respectively, (26).

RSID was used as a quantitative parameter for predicting the degree of hepatic steatosis. The threshold value for RSID that was considered diagnostic of an unsuitable liver donor was chosen retrospectively after the results were reviewed to maximize specificity while maintaining 100% sensitivity. When the RSID was less than 20%, the liver was considered to be appropriate for transplantation; when the RSID was 20% or more, the liver was considered to be inappropriate for transplantation.

Statistical Analysis
To determine whether MR imaging could qualitatively reflect hepatic steatosis (macrovesicular or microvesicular), RSID was correlated with the pathologic degree of steatosis (total, macrovesicular, and microvesicular) by using a linear regression analysis. The degree of steatosis, as determined by the pathologist, was used as a continuous variable in the regression models because the value of the RSID that was calculated by using the MR images was a continuous variable.

Pathologic grades were correlated with qualitative grades by the radiologists by using weighted and statistics and the Spearman rank correlation. We considered values of more than 0.81 to represent almost perfect agreement and values of 0.61–0.80 and 0.41–0.60 to represent substantial and moderate agreement, respectively. Values of less than 0.40 were considered to represent fair agreement (27). Because qualitative MR assessments of fatty liver (as determined by the radiologists) were classified by using three degrees of categoric variables, macrovesicular steatosis (as determined by the pathologist) was also classified by using three degrees of variables (normal, <5%; mild, 5%–29%; and more than moderate, 30%).

In addition, the distributions of RSID that were derived from quantitative MR image analysis among the four degrees of macrosteatosis (normal, mild, moderate, and severe) and between the two categories of transplant suitability (appropriate and inappropriate), which were classified according to the 30% cutoff for macrosteatosis, were analyzed by using the Kruskal-Wallis test and the Mann-Whitney U test, respectively. The sensitivity, specificity, positive predictive value, negative predictive value, and overall accuracy of the qualitative and quantitative analyses were also calculated. Finally, the degree of agreement between qualitative classification by the radiologists and quantitative classification on the basis of RSID value to determine appropriateness for liver transplantation was obtained by using statistics.

SPSS (version 11.0; SPSS, Chicago, Ill) and GraphPad InStat (version 3.0; GraphPad Software, San Diego, Calif) were used for statistical analysis. A P value of less than .05 was considered to indicate a statistically significant difference.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
Percutaneous and Surgical Biopsy Comparisons
Of the 57 potential donors, four underwent both surgical and percutaneous biopsy. The mean interval between biopsy procedures was 16.3 days (2, 13, 16, and 34 days). Pathologic grades with respect to macrosteatosis were identical for both biopsies in three of four patients (mild degree of macrosteatosis at both percutaneous and surgical biopsy). The pathologic grade in one patient was changed from mild at percutaneous biopsy to normal at surgical biopsy. However, no change in appropriateness for transplantation according to biopsy type was observed in any of the four patients.

Histopathologic Findings
For macrosteatosis, 46 of 57 patients were classified as normal, six as having mild macrosteatosis, three as having moderate macrosteatosis, and two as having severe macrosteatosis. For microsteatosis, 46 patients were classified as normal, nine as having mild microsteatosis, and two as having moderate microsteatosis. For total steatosis (macrovesicular and microvesicular), 25 patients were classified as normal, 23 as having mild steatosis, six as having moderate steatosis, and three as having severe steatosis. Thus, 52 patients were categorized as appropriate donors and five were categorized as inappropriate donors with respect to macrosteatosis according to histopathologic results.

Demographic Findings
No significant differences in age or sex were evident between the two groups. The mean age of patients whose livers were deemed appropriate for transplantation was 31.6 years ± 10.6 compared with 32.8 years ± 12.3 in patients whose livers were deemed inappropriate for transplantation (P = .75, Mann-Whitney U test). Of the 52 patients whose livers were deemed appropriate for transplantation, 37 (71%) were male and 15 (29%) were female; of the five patients whose livers were deemed inappropriate for transplantation, three (60%) were male and two (40%) were female (P = .63, Fisher exact test).

MR Assessments
Qualitative analysis.—For qualitative MR evaluation of SI in the liver relative to that in the paravertebral muscle and spleen on opposed-phase images, the two radiologists initially agreed in all but two cases, which were interpreted as normal and mild by one radiologist and as mild and moderate by the other radiologist. These two cases were confirmed as having a mild and a more than moderate degree of steatosis by consensus. For qualitative MR evaluation by using the differences in hepatic SI between in-phase and opposed-phase images, the two radiologists initially agreed in all but one case, which was interpreted as mild by one radiologist and moderate by the other radiologist. This case was confirmed as having a more than moderate degree of steatosis by consensus. Qualitative grades of macrosteatosis interpreted by the radiologists were compared with grades of macrosteatosis interpreted by the pathologist and are presented in Table 1.

View larger version (142K): [in this window] [in a new window] [Download PPT slide]   Figure 1a: Normal liver in 49-year-old woman. (a) In-phase (128/5.3 [repetition time msec/echo time msec]) and (b) opposed-phase (128/2.7) breath-hold transverse MR images show SI of the liver that is slightly higher in b than in a, indicating the absence of fat. Theoretically, hepatic SI should be the same on opposed-phase and in-phase images. However, the longer echo time for in-phase images slightly lessens SI in all organs because of the T2* effect for which the gradient-echo sequence is sensitive. In b, the liver has a higher SI than the paravertebral muscle and spleen. RSID is –3.95%, which represents appropriate suitability as a liver donor. Histologic specimen obtained during transplantation (not shown) revealed no fatty infiltration.

View larger version (150K): [in this window] [in a new window] [Download PPT slide]   Figure 1b: Normal liver in 49-year-old woman. (a) In-phase (128/5.3 [repetition time msec/echo time msec]) and (b) opposed-phase (128/2.7) breath-hold transverse MR images show SI of the liver that is slightly higher in b than in a, indicating the absence of fat. Theoretically, hepatic SI should be the same on opposed-phase and in-phase images. However, the longer echo time for in-phase images slightly lessens SI in all organs because of the T2* effect for which the gradient-echo sequence is sensitive. In b, the liver has a higher SI than the paravertebral muscle and spleen. RSID is –3.95%, which represents appropriate suitability as a liver donor. Histologic specimen obtained during transplantation (not shown) revealed no fatty infiltration.

View larger version (142K): [in this window] [in a new window] [Download PPT slide]   Figure 2a: Mild steatosis in 27-year-old man. (a) In-phase (110/4.8) and (b) opposed-phase (110/2.4) breath-hold transverse MR images show SI of the liver that is similar in both b and a, indicating mild (5%–29%) accumulation of fat. In b, the liver has intermediate SI compared with the paravertebral muscle and spleen. RSID is 14.15%, which represents appropriate suitability as a liver donor with respect to steatosis. (c) Histologic specimen obtained during transplantation reveals 10% macrosteatosis and 5% microsteatosis. (Hematoxylin-eosin stain; original magnification, x100.)

View larger version (151K): [in this window] [in a new window] [Download PPT slide]   Figure 2b: Mild steatosis in 27-year-old man. (a) In-phase (110/4.8) and (b) opposed-phase (110/2.4) breath-hold transverse MR images show SI of the liver that is similar in both b and a, indicating mild (5%–29%) accumulation of fat. In b, the liver has intermediate SI compared with the paravertebral muscle and spleen. RSID is 14.15%, which represents appropriate suitability as a liver donor with respect to steatosis. (c) Histologic specimen obtained during transplantation reveals 10% macrosteatosis and 5% microsteatosis. (Hematoxylin-eosin stain; original magnification, x100.)

View larger version (145K): [in this window] [in a new window] [Download PPT slide]   Figure 2c: Mild steatosis in 27-year-old man. (a) In-phase (110/4.8) and (b) opposed-phase (110/2.4) breath-hold transverse MR images show SI of the liver that is similar in both b and a, indicating mild (5%–29%) accumulation of fat. In b, the liver has intermediate SI compared with the paravertebral muscle and spleen. RSID is 14.15%, which represents appropriate suitability as a liver donor with respect to steatosis. (c) Histologic specimen obtained during transplantation reveals 10% macrosteatosis and 5% microsteatosis. (Hematoxylin-eosin stain; original magnification, x100.)

View larger version (171K): [in this window] [in a new window] [Download PPT slide]   Figure 3a: Severe macrosteatosis and moderate microsteatosis in 49-year-old man. (a) In-phase (128/5.3) and (b) opposed-phase (128/2.7) breath-hold transverse MR images show SI of the liver that is markedly hypointense in b compared with a, indicating a more than moderate degree of fat accumulation (30%). In b, the SI of the liver is lower than that of the paravertebral muscle and spleen. RSID is 66.39%, which represents an inappropriate suitability as a liver donor with respect to steatosis. (c) Histologic specimen obtained at subsequent liver biopsy reveals 60% macrosteatosis and 30% microsteatosis. (Hematoxylin-eosin stain; original magnification, x100.)

View larger version (181K): [in this window] [in a new window] [Download PPT slide]   Figure 3b: Severe macrosteatosis and moderate microsteatosis in 49-year-old man. (a) In-phase (128/5.3) and (b) opposed-phase (128/2.7) breath-hold transverse MR images show SI of the liver that is markedly hypointense in b compared with a, indicating a more than moderate degree of fat accumulation (30%). In b, the SI of the liver is lower than that of the paravertebral muscle and spleen. RSID is 66.39%, which represents an inappropriate suitability as a liver donor with respect to steatosis. (c) Histologic specimen obtained at subsequent liver biopsy reveals 60% macrosteatosis and 30% microsteatosis. (Hematoxylin-eosin stain; original magnification, x100.)

View larger version (131K): [in this window] [in a new window] [Download PPT slide]   Figure 3c: Severe macrosteatosis and moderate microsteatosis in 49-year-old man. (a) In-phase (128/5.3) and (b) opposed-phase (128/2.7) breath-hold transverse MR images show SI of the liver that is markedly hypointense in b compared with a, indicating a more than moderate degree of fat accumulation (30%). In b, the SI of the liver is lower than that of the paravertebral muscle and spleen. RSID is 66.39%, which represents an inappropriate suitability as a liver donor with respect to steatosis. (c) Histologic specimen obtained at subsequent liver biopsy reveals 60% macrosteatosis and 30% microsteatosis. (Hematoxylin-eosin stain; original magnification, x100.)

View larger version (172K): [in this window] [in a new window] [Download PPT slide]   Figure 4a: Moderate macrosteatosis in 37-year-old man who was misclassified as being an appropriate donor due to uneven fatty infiltration at qualitative analysis. (a) In-phase (128/5.3) and (b) opposed-phase (128/2.7) breath-hold transverse MR images show unevenly decreased SI of the liver in b (arrows). In particular, the posterior segment of the liver is spared of fat; therefore, the liver was graded as normal because the radiologists judged the SI of the liver to be higher than that of the paravertebral muscle and spleen. However, SI of the liver in b is relatively hypointense compared with that in a. Even if such uneven hepatic SI could be shown to be caused by coil inhomogeneity and uneven fatty infiltration, the main cause of uneven hepatic SI in this case should be considered to be uneven fatty infiltration because an absolute decrease in SI was observed throughout the entire liver in b. RSID is 20.98%, which represents inappropriate suitability as a liver donor. Histologic specimen (not shown) obtained at subsequent liver biopsy revealed 30% macrosteatosis and 20% microsteatosis.

View larger version (142K): [in this window] [in a new window] [Download PPT slide]   Figure 4b: Moderate macrosteatosis in 37-year-old man who was misclassified as being an appropriate donor due to uneven fatty infiltration at qualitative analysis. (a) In-phase (128/5.3) and (b) opposed-phase (128/2.7) breath-hold transverse MR images show unevenly decreased SI of the liver in b (arrows). In particular, the posterior segment of the liver is spared of fat; therefore, the liver was graded as normal because the radiologists judged the SI of the liver to be higher than that of the paravertebral muscle and spleen. However, SI of the liver in b is relatively hypointense compared with that in a. Even if such uneven hepatic SI could be shown to be caused by coil inhomogeneity and uneven fatty infiltration, the main cause of uneven hepatic SI in this case should be considered to be uneven fatty infiltration because an absolute decrease in SI was observed throughout the entire liver in b. RSID is 20.98%, which represents inappropriate suitability as a liver donor. Histologic specimen (not shown) obtained at subsequent liver biopsy revealed 30% macrosteatosis and 20% microsteatosis.

View larger version (159K): [in this window] [in a new window] [Download PPT slide]   Figure 5a: Moderate macrosteatosis in 50-year-old woman who was misclassified as an appropriate donor due to suspicious fatty infiltration of paravertebral muscle on qualitative analysis. (a) In-phase (128/5.3) and (b) opposed-phase (128/2.7) breath-hold transverse MR images show that hepatic SI was hypointense in b compared with a, indicating a more than moderate degree of steatosis. SI of the paravertebral muscle in b is more dramatically decreased compared with that in a. Therefore, the radiologists misjudged the liver to be appropriate for donation at qualitative analysis by comparing the SI of the three organs in b alone. RSID is 21.93%, which represents inappropriate suitability as a liver donor. Histologic specimen (not shown) obtained at subsequent liver biopsy revealed 30% macrosteatosis and 20% microsteatosis.

View larger version (171K): [in this window] [in a new window] [Download PPT slide]   Figure 5b: Moderate macrosteatosis in 50-year-old woman who was misclassified as an appropriate donor due to suspicious fatty infiltration of paravertebral muscle on qualitative analysis. (a) In-phase (128/5.3) and (b) opposed-phase (128/2.7) breath-hold transverse MR images show that hepatic SI was hypointense in b compared with a, indicating a more than moderate degree of steatosis. SI of the paravertebral muscle in b is more dramatically decreased compared with that in a. Therefore, the radiologists misjudged the liver to be appropriate for donation at qualitative analysis by comparing the SI of the three organs in b alone. RSID is 21.93%, which represents inappropriate suitability as a liver donor. Histologic specimen (not shown) obtained at subsequent liver biopsy revealed 30% macrosteatosis and 20% microsteatosis.

In addition, 47 patients showed either higher hepatic SI on opposed-phase images than on in-phase images (31 patients) or similar hepatic SI on in-phase and opposed-phase images (16 patients); these patients were considered to have normal or mild steatosis (Table 2). In terms of macrosteatosis, all of these 47 patients proved to be appropriate donors at pathologic analysis. Of the 10 patients who showed a decrease in hepatic SI on opposed-phase images compared with in-phase images, five (50%) proved to be inappropriate donors. In four of the remaining five patients who were considered to have a more than moderate degree of steatosis, histologic analysis revealed a moderate degree of total steatosis (30%, 30%, 40%, and 45%), even though those patients should have been classified as appropriate donors with respect to macrosteatosis (15%, 20%, 20%, and 15%, respectively) (Fig 6). Accordingly, for the classification of patients as inappropriate liver donors on the basis of a decrease in hepatic SI on opposed-phase images compared with in-phase images, the sensitivity, specificity, positive predictive value, negative predictive value, and overall accuracy were 100% (five of five), 90.4% (47 of 52), 50% (five of 10), 100% (47 of 47), and 91.2% (52 of 57), respectively.

View larger version (135K): [in this window] [in a new window] [Download PPT slide]   Figure 6a: Mild macrosteatosis (20%) and mild microsteatosis (20%) in 19-year-old man who was misclassified as an inappropriate donor at qualitative and quantitative analysis. (a) In-phase (110/4.8) and (b) opposed-phase (110/2.4) breath-hold transverse MR images show SI of the liver that is hypointense in b compared with a. In b, the liver has lower SI than the paravertebral muscle and spleen. RSID is 32.34%, which represents inappropriate suitability as a liver donor with respect to steatosis. Histologic specimen (not shown) revealed 20% macrosteatosis and 20% microsteatosis, which indicates that the organ is appropriate for donation with respect to macrosteatosis.

View larger version (143K): [in this window] [in a new window] [Download PPT slide]   Figure 6b: Mild macrosteatosis (20%) and mild microsteatosis (20%) in 19-year-old man who was misclassified as an inappropriate donor at qualitative and quantitative analysis. (a) In-phase (110/4.8) and (b) opposed-phase (110/2.4) breath-hold transverse MR images show SI of the liver that is hypointense in b compared with a. In b, the liver has lower SI than the paravertebral muscle and spleen. RSID is 32.34%, which represents inappropriate suitability as a liver donor with respect to steatosis. Histologic specimen (not shown) revealed 20% macrosteatosis and 20% microsteatosis, which indicates that the organ is appropriate for donation with respect to macrosteatosis.

View larger version (23K): [in this window] [in a new window] [Download PPT slide]   Figure 7: Scatter plot and regression lines show a positive correlation between pathologic grade of total steatosis and RSID of the liver on opposed-phase images compared with in-phase images (y = 8.731 + 1.044x, r2 = 0.742, P < .001) The solid line is the regression line, the dotted line is the 95% confidence interval, and the dashed line is the 95% prediction interval. The horizontal line is the zero line of total steatosis.

Agreement between Qualitative and Quantitative Classification
There was almost perfect agreement ( = 0.937, P < .001) between qualitative classification (appropriate vs inappropriate) for determining the appropriateness of liver donation with respect to hepatic steatosis on the basis of changes in hepatic SI on opposed-phase images compared with in-phase images and quantitative classification on the basis of RSID value.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
An interesting result of our study is the fact that a decrease in SI on opposed-phase images reflects the total amount of hepatic steatosis, regardless of subtype (ie, macrovesicular or microvesicular fat). Indeed, by using a linear regression analysis, we found that RSID, which is used as a quantitative MR imaging parameter to predict the degree of hepatic steatosis, proved to be clearly associated with the total amount of fat accumulation (ß = 0.705, P < .05). Even though the effect of moderate to severe microvesicular steatosis in liver transplantation has not been fully investigated, it is generally accepted that microvesicular steatosis has a minimal effect on posttransplantation morbidity or mortality and that the use of donor livers with microvesicular steatosis could expand the donor pool (21,28,29). Therefore, in radiologic studies that assess hepatic steatosis in potential liver donors, the different types of hepatic steatosis should be considered and treated separately, just as they are in many surgical and pathology articles (21,28,29). To our knowledge, only one radiology article has been published on whether radiologic modalities can be used to differentiate hepatic steatosis as being either macrovesicular or microvesicular (30).

Although we prove in our study that MR imaging cannot be used to differentiate macrovesicular from microvesicular fat accumulation, RSID was revealed to be a good indicator in confirming that a donor liver was appropriate for transplantation with respect to macrosteatosis. Potential donor livers with an RSID of less than 20% on in-phase and opposed-phase images appear to be definitely appropriate for transplantation. Therefore, unnecessary liver biopsy can be obviated in such cases.

Furthermore, donor selection by using MR imaging might also be more accurate than donor selection by using biopsy because the latter is limited by sampling errors. By comparison, MR estimations for predictions of appropriateness as a liver donor by using RSID may inevitably overrepresent the degree of macrosteatosis, especially in patients who have various degrees of microsteatosis. Considering the shortage of liver donors, it would also be useful to discuss the overestimation of hepatic steatosis that can be caused by clinically insignificant microsteatosis and the subsequent exclusion of appropriate liver donors. Therefore, in potential liver donors with an RSID of 20% or more on MR images, it must be determined whether a moderate or severe degree of macrosteatosis increases the value of RSID or if various degrees of microsteatosis affect the value of RSID. Although percutaneous liver biopsy may adequately perform this function, other complementary imaging studies should be considered for potential liver donors who have a high RSID because living donor safety is of paramount importance, and the avoidance of unnecessary procedures is essential for lessening overall donor morbidity.

In a previous study, we proved that US grades for fatty liver, as determined by radiologists, are significantly correlated with the grade of macrovesicular fat accumulation in hepatocytes and not with microvesicular fat accumulation (30). Furthermore, artificial neural networks applied to US results have shown excellent performance in determining the appropriateness of liver donors with respect to macrosteatosis on the basis of multiple variables related to laboratory and US features. Therefore, US may be a good imaging option to determine donor inappropriateness for those patients who have been judged as unsuitable donors as a result of steatosis at MR imaging. This too would help to minimize the potential biopsy-related risks.

Differences in the ability to differentiate between types of accumulated fat within the hepatocytes by using US and MR imaging may be partly related to the different physics of the two modalities. In US, transmitted ultrasound beams are reflected from tissue interfaces and thus create echogenicity. Fatty infiltration has generally been thought to cause increased hepatic echogenicity by increasing the tissue interface between fat globules and other substances within the hepatocytes. More recently, however, such tissue interfaces and increased echogenicity at US were thought to depend on the size of the fat droplets within the hepatocyte. Pamilo et al (31) proved the latter hypothesis by observing that increased hepatic echogenicity is caused by large fat droplets only (ie, those exceeding 100 µm in diameter on US images). In other words, microvesicular fat accumulation has little effect on the increased echogenicity of the liver at US. On the other hand, assessment of fat within the liver or other organs at MR imaging is based on the characteristic differences in the resonant frequencies of water (hydroxyl group) and fat (methylene group). Therefore, the opposed-phase images accurately reflect the difference in the water and fat contents, regardless of the type of accumulated fat (32). Our result correlates well with that fact.

The dual-echo chemical shift gradient-echo MR imaging technique that we used has great advantages compared with older techniques because it provides in-phase and opposed-phase images in a single breath hold by using two different echo times; thus, imaging is not compromised by section misregistration, which can occur when two breath-hold sequences are performed and inevitably reduce the accuracy of the quantitative measurement of fat. The other advantage of the dual-echo technique is that apparent differences in contrast enhancement between corresponding in-phase and opposed-phase images are caused only by the chemical shift of the lesion and are not influenced by other factors.

Nonetheless, chemical shift MR imaging has some limitations as a quantitative tool. The first limitation is the theoretic effect of the ambiguity error—that is, because the opposed-phase images accurately reflect the differences in water and fat signals, it is not possible to identify whether fat or water is the dominant signal. If the fat fraction in the liver is greater than the water fraction in the liver, then the decrease in SI on opposed-phase images can cause an underestimation of the degree of steatosis. The ambiguity error can influence results in areas with very abundant steatosis. Furthermore, the inflection point at which SI paradoxically increases with larger amounts of fat is unknown in vivo. Even though, in our study, such paradoxical SI increases did not occur, even in the two patients with fat more than 81% total steatosis, we should keep in mind that such ambiguity could be a source of misinterpretation until the inflection point is clearly defined for the in vivo setting, such as with fatty liver.

Recently, three-point and two-point Dixon techniques, which do not have the ambiguity of in-phase and opposed-phase imaging that is related to large amounts of fat or to MR spectroscopy, have been implemented for accurate quantification of fat contents within a voxel (32,33). Therefore, a prospective and large study that uses the above techniques would be needed to quantify the exact amount of intrahepatic fat without any ambiguity-related potential risk. Second, voxels at the interface between fat and water may show a marked loss in SI on opposed-phase images, thereby causing the familiar "etching" artifact around the liver, regardless of fat content. Although this may be more problematic in smaller organs, such as the adrenal glands (34,35), it would have little effect on relatively larger organs, such as the liver.

Another interesting finding in the present study is that the visual assessment of hepatic steatosis from observing the difference in heptaic SI between in-phase and opposed-phase images was comparably accurate relative to quantitative assessment by using ROI measurement, although perception error additionally occurred in one patient at qualitative analysis. However, qualitative analysis from observing SI in the liver compared with the paravertebral muscle and spleen on opposed-phase images alone was less accurate than the other qualitative analysis method. This is related to the fact that the use of opposed-phase images alone for the assessment of hepatic steatosis can lead to misdiagnosis because the SI of the liver, paravertebral muscle, and spleen can be changed by various conditions, such as iron deposition, edema, focal lesions, dense fibrosis, and fat infiltration.

In one patient who had uneven fatty infiltration, hepatic macrosteatosis was underestimated; this patient was classified as an appropriate donor, even though hepatic histologic results showed moderate macrosteatosis (30%). In addition, another patient who had a moderate degree of macrosteatosis (30%) was misclassified as an appropriate donor. After retrospective analysis of the MR images in the latter case, a relatively lower SI was observed on the opposed-phase image in the paravertebral muscle compared with the liver. Therefore, we assume that mild fatty degeneration of the paravertebral muscle may have led to the misclassification of this patient, even though we have no histologic proof of that assumption. Actually, the paravertebral muscle is not a commonly used reference to quantify hepatic steatosis; rather, it is used to assess hepatic iron content (36,37). In the present study, however, we identified the potential that the paravertebral muscle may have a supplementary role in subdividing the degree of hepatic steatosis. Such a subdivision of hepatic steatosis could also be successfully achieved with another qualitative analysis by using the change in hepatic SI between in-phase and opposed-phase images.

Although the differentiation between normal findings and mild hepatic steatosis for determining appropriateness as a liver donor was not critical in our study, this information would be more important if the threshold for inappropriateness as a liver donor was set at a lower level (eg, 5% macrosteatosis) than was used in our study. Further study is necessary to prove this assertion. In addition, the combined interpretation of the results from the two qualitative methods and from quantitative analysis may increase the diagnostic performance of MR imaging in predicting the appropriateness of transplantation with respect to macrosteatosis. For example, in two patients who were judged as being inappropriate donors on the basis of qualitative and quantitative findings, which were obtained by observing and calculating the difference in hepatic SI between in-phase and opposed-phase images, the SI of the liver was between that of the paravertebral muscle and spleen; these patients were correctly classified as appropriate donors with the other type of qualitative analysis by using paravertebral muscle.

One limitation of our study is that the study population, especially those who were deemed inappropriate donors, is relatively small. This undesired occurrence is related, in part, to how physicians determine the maximal threshold of hepatic macrosteatosis that is acceptable in living liver donors. At our institution, the maximal amount of hepatic macrosteatosis accepted in living liver donors is 30%. Although this cutoff has been widely accepted within the liver transplantation community, this cutoff is not entirely fixed. If we used a lower cutoff value (eg, 15%), which represents a fairly conservative approach for donor and recipient safety and subsequently restricts the liver donor pool, the number of patients classified as inappropriate donors would have increased. However, considering the shortage of potential liver donors and the general acceptance of 30% as the maximal threshold of hepatic macrosteatosis, this would not be an appropriate alternative; rather, further prospective studies with a larger number of cases should be performed.

Second, liver biopsies are subject to a well-recognized sampling error due to the inhomogeneous distribution of fat in the liver; moreover, the process of fat infiltration is continuous, and there may be temporal variations in fat content. In the present study, however, no change between being classified as an appropriate or inappropriate donor according to biopsy results was observed in any of the four patients who underwent both percutaneous and surgical biopsy. The results in these four patients indicated that temporal variability was not a serious problem. However, the small number of cases (four) and the short interval between the two biopsies (mean, 16.3 days) prevented our generalizing this result.

Third, although we identified the potential that the paravertebral muscle may have a supplementary role in subdividing the degree of hepatic steatosis, various conditions involving paravertebral muscle—including fat infiltration, muscle edema, or dense fibrosis—can lead to misinterpretation. Therefore, a combined interpretation of the two qualitative methods may be mandatory to predict a patient's appropriateness as a liver donor by using MR imaging.

Fourth, as a technical limitation, we should bear in mind the ambiguity (see above) of using in-phase and opposed-phase imaging to quantify fat. We also recognize the limitations of our study that were imposed by the consensus interpretation of MR imaging results.

In summary, by using an RSID of less than 20% for the appropriateness of donation, dual-echo MR imaging was used to correctly predict the suitability of donor livers in 53 of 57 patients. Although dual-echo chemical shift gradient-echo MR imaging demonstrates the total amount of hepatic steatosis regardless of type, potential liver donors with a normal RSID may be definitely appropriate for liver donation with respect to hepatic steatosis. However, other complementary imaging studies or liver biopsy will be necessary in a small portion of potential donors with high RSID in order to determine the exact proportion of macrovesicular fat accumulation within the hepatocyte.

	FOOTNOTES

 

Abbreviations: ROI = region of interest • RSID = relative SI decrease • SI = signal intensity

Authors stated no financial relationship to disclose.

Author contributions: Guarantors of integrity of entire study, S.H.K., J.M.L., K.S. Suh, B.I.C.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, S.H.K., C.J.H., J.Y.J., K.S. Shin; clinical studies, J.M.L., J.Y.L., J.Y.J., N.J.Y., K.S. Suh, S.Y.J.; statistical analysis, K.H.L.; and manuscript editing, J.M.L., J.K.H., J.Y.L., N.J.Y., K.S. Suh, B.I.C.

	References

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 

1. Smith B. Segmental liver transplantation from a living donor. J Pediatr Surg 1969;4:126–132.[CrossRef][Medline]
2. Ryckman FC, Flake AW, Fisher RA, Tchervenkov JI, Pedersen SH, Balistreri WF. Segmental orthotopic hepatic transplantation as a means to improve patient survival and diminish waiting-list mortality. J Pediatr Surg 1991;26:422–427.[CrossRef][Medline]
3. Emond JC, Heffron TG, Kortz EO, et al. Improved results of living-related liver transplantation with routine application in a pediatric program. Transplantation 1993;55:835–840.[Medline]
4. D'Alessandro AM, Kalayoglu M, Sollinger HW, et al. The predictive value of donor liver biopsies for the development of primary nonfunction after orthotopic liver transplantation. Transplantation 1991;51:157–163.[Medline]
5. Ploeg RJ, D'Alessandro AM, Knechtle SJ, et al. Risk factors for primary dysfunction after liver transplantation: a multivariate analysis. Transplantation 1993;55:807–813.[Medline]
6. Marsman WA, Wiesner RH, Rodriguez L, et al. Use of fatty donor liver is associated with diminished early patient and graft survival. Transplantation 1996;62:1246–1251.[CrossRef][Medline]
7. Garcia Urena MA, Colina Ruiz-Delgado F, Moreno GE, et al. Hepatic steatosis in liver transplant donors: common feature of donor population? World J Surg 1998;22:837–844.[CrossRef][Medline]
8. Scatarige JC, Scott WW, Donovan PJ, Siegelman SS, Sanders RC. Fatty infiltration of the liver: ultrasonographic and computed tomographic correlation. J Ultrasound Med 1984;3:9–14.[Abstract]
9. Taylor KJ, Riely CA, Hammers L, et al. Quantitative US attenuation in normal liver and in patients with diffuse liver disease: importance of fat. Radiology 1986;160:65–71.[Abstract/Free Full Text]
10. Limanond P, Raman SS, Lassman C, et al. Macrovesicular hepatic steatosis in living related liver donors: correlation between CT and histologic findings. Radiology 2004;230:276–280.[Abstract/Free Full Text]
11. Kawata R, Sakata K, Kunieda T, Saji S, Doi H, Nozawa Y. Quantitative evaluation of fatty liver by computed tomography in rabbits. AJR Am J Roentgenol 1984;142:741–746.[Abstract/Free Full Text]
12. Dixon WT. Simple proton spectroscopic imaging. Radiology 1984;153:189–194.[Abstract/Free Full Text]
13. Levenson H, Greensite F, Hoefs J, et al. Fatty infiltration of the liver: quantification with phase-contrast MR imaging at 1.5 T vs biopsy. AJR Am J Roentgenol 1991;156:307–312.[Abstract/Free Full Text]
14. Longo R, Pollesello P, Ricci C, et al. Proton MR spectroscopy in quantitative in vivo determination of fat content in human liver steatosis. J Magn Reson Imaging 1995;5:281–285.[Medline]
15. Rinella ME, McCarthy R, Thakrar K, et al. Dual-echo, chemical shift gradient-echo magnetic resonance imaging to quantify hepatic steatosis: implications for living liver donation. Liver Transpl 2003;9:851–856.[CrossRef][Medline]
16. Lee VS, Morgan GR, Teperman LW, et al. MR imaging as the sole preoperative imaging modality for right hepatectomy: a prospective study of living adult-to-adult liver donor candidates. AJR Am J Roentgenol 2001;176:1475–1482.[Abstract/Free Full Text]
17. Cheng YF, Chen CL, Huang TL, et al. Single imaging modality evaluation of living donors in liver transplantation: magnetic resonance imaging. Transplantation 2001;72:1527–1533.[CrossRef][Medline]
18. Fulcher AS. CT cholangiography: advantages and disadvantages compared with MR cholangiography in the evaluation of the biliary tract following liver transplantation. Liver Transpl 2004;10:1071–1072.[CrossRef][Medline]
19. Burt AD, Mutton A, Day CP. Diagnosis and interpretation of steatosis and steatohepatitis. Semin Diagn Pathol 1998;15:246–258.[Medline]
20. Adam R, Reynes M, Johann M, et al. The outcome of steatotic grafts in liver transplantation. Transplant Proc 1991;23:1538–1540.[Medline]
21. Fishbein TM, Fiel MI, Emre S, et al. Use of livers with microvesicular fat safely expands the donor pool. Transplantation 1997;64:248–251.[CrossRef][Medline]
22. Miki C, Iriyama K, Mirza DF, et al. Postperfusion energy metabolism of steatotic graft and its relation to early graft viability following liver transplantation. Dig Dis Sci 1998;43:74–79.[CrossRef][Medline]
23. Bluml S, Schad LR, Stepanow B, Lorenz WJ. Spin-lattice relaxation time measurement by means of a TurboFLASH technique. Magn Reson Med 1993;30:289–295.[Medline]
24. de Bazelaire CM, Duhamel GD, Rofsky NM, Alsop DC. MR imaging relaxation times of abdominal and pelvic tissues measured in vivo at 3.0 T: preliminary results. Radiology 2004;230:652–659.[Abstract/Free Full Text]
25. Stark DD, Wittenberg J, Middleton MS, Ferrucci JT Jr. Liver metastases: detection by phase-contrast MR imaging. Radiology 1986;158:327–332.[Abstract/Free Full Text]
26. Mitchell DG, Kim I, Chang TS, et al. Fatty liver: chemical shift phase-difference and suppression magnetic resonance imaging techniques in animals, phantoms, and humans. Invest Radiol 1991;26:1041–1052.[Medline]
27. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159–174.[CrossRef][Medline]
28. Koneru B, Dikdan G. Hepatic steatosis and liver transplantation current clinical and experimental perspectives. Transplantation 2002;73:325–330.[CrossRef][Medline]
29. Selzner M, Clavien PA. Fatty liver in liver transplantation and surgery. Semin Liver Dis 2001;21:105–113.[CrossRef][Medline]
30. Kim SH, Lee JM, Kim JH, et al. Appropriateness of a donor liver with respect to macrosteatosis: application of artificial neural networks to US images—initial experience. Radiology 2005;234:793–803.[Abstract/Free Full Text]
31. Pamilo M, Sotaniemi EA, Suramo I, Lahde S, Arranto AJ. Evaluation of liver steatotic and fibrous content by computerized tomography and ultrasound. Scand J Gastroenterol 1983;18:743–747.[Medline]
32. Zhang X, Tengowski M, Fasulo L, Botts S, Suddarth SA, Johnson A. Measurement of fat/water ratios in rat liver using 3D three-point dixon MRI. Magn Reson Med 2004;51:697–702.[CrossRef][Medline]
33. Schlemmer HP, Sawatzki T, Sammet S, et al. Hepatic phospholipids in alcoholic liver disease assessed by proton-decoupled 31P magnetic resonance spectroscopy. J Hepatol 2005;42:752–759.[CrossRef][Medline]
34. Namimoto T, Yamashita Y, Mitsuzaki K, et al. Adrenal masses: quantification of fat content with double-echo chemical shift in-phase and opposed-phase FLASH MR images for differentiation of adrenal adenomas. Radiology 2001;218:642–646.[Abstract/Free Full Text]
35. Haider MA, Ghai S, Jhaveri K, Lockwood G. Chemical shift MR imaging of hyperattenuating (>10 HU) adrenal masses: does it still have a role? Radiology 2004;231:711–716.[Abstract/Free Full Text]
36. Ernst O, Sergent G, Bonvarlet P, Canva-Delcambre V, Paris JC, L'Hermine C. Hepatic iron overload: diagnosis and quantification with MR imaging. AJR Am J Roentgenol 1997;168:1205–1208.[Abstract/Free Full Text]
37. Gandon Y, Olivie D, Guyader D, et al. Non-invasive assessment of hepatic iron stores by MRI. Lancet 2004;363:357–362.[CrossRef][Medline]

This article has been cited by other articles:

A. Nickkholgh, J. Weitz, J. Encke, P. Sauer, A. Mehrabi, M. W. Buchler, J. Schmidt, and P. Schemmer Utilization of extended donor criteria in liver transplantation: a comprehensive review of the literature Nephrol. Dial. Transplant., September 1, 2007; 22(suppl_8): viii29 - viii36. [Abstract] [Full Text]