Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights.
Original Research n Gastrointestinal
Combined Use of MR Fat Quantification and MR Elastography in Living Liver Donors: Can It Reduce the Need for Preoperative Liver Biopsy?1 Purpose:
To evaluate the diagnostic performance of magnetic resonance (MR) fat quantification and MR elastography for the assessment of hepatic steatosis and fibrosis in living liver donor candidates.
Materials and Methods:
This retrospective study was approved by the institutional review board, and the requirement of informed consent was waived. Donors who underwent MR fat quantification and MR elastography at 1.5 T, followed by liver biopsy, were chronologically grouped into test and validation groups. In the test group (n = 362), MR fat fraction and liver stiffness were compared among donors with normal parenchyma (n = 244), simple steatosis (n = 71), steatosis with inflammatory activity (n = 21), nonalcoholic steatohepatitis (n = 17), and fibrosis (n = 9). Diagnostic performance of the two techniques was assessed by using receiver operating characteristic curve analysis for the detection of substantial steatosis (macrovesicular fat 10%) or fibrosis (F1) and was tested in a validation group (n = 34).
Results:
In the test group, donors with steatosis showed significantly higher fat fraction than donors without steatosis (P , .0001), and donors with fibrosis and nonalcoholic steatohepatitis showed significantly higher liver stiffness values than donors without fibrosis (P , .0001). Areas under the curve were 0.93 (cutoff value . 5.8%) for MR fat quantification and 0.85 (cutoff value . 1.94 kPa) for MR elastography. By using those values, the combination of the two techniques could be used to detect substantial steatosis or fibrosis with 100% sensitivity (12 of 12 patients, 95% confidence interval: 73.4%, 100%) and 100% negative predictive value (15 of 15 patients, 95% confidence interval: 78.0%, 100%) in the validation group.
Conclusion:
A combination of MR fat quantification and MR elastography can provide sufficient sensitivity to detect substantial steatosis or fibrosis (F1) in liver donor candidates.
1
From the Departments of Radiology (J.H.Y., J.M.L., J.K.H., B.I.C.), Surgery (K.S.S., K.W.L., N.J.Y.), and Pathology (K.B.L.), Seoul National University Hospital, Seoul, South Korea; and Institute of Radiation Medicine, Seoul National University College of Medicine, 101 Daehak-ro, Jongno-gu, Seoul 110-744, South Korea (J.M.L., J.K.H., B.I.C.). Received April 15, 2014; revision requested June 5; revision received November 16; accepted December 16; final version accepted December 29. This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2013R1A1A2A10066037). Address correspondence to J.M.L. (e-mail: [email protected]).
Imaging
Jeong Hee Yoon, MD Jeong Min Lee, MD Kyung-Suk Suh, MD Kwan-Woong Lee, MD Nam-Joon Yi, MD Kyung Bun Lee, MD Joon Koo Han, MD Byung Ihn Choi, MD
RSNA, 2015
q
Online supplemental material is available for this article.
RSNA, 2015
q
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
1
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
L
iver transplantation is widely performed around the world as a curative treatment strategy for liver failure and for the early stages of hepatocellular carcinoma (1,2). However, owing to widespread organ shortages reported for pediatric recipients, the living donor liver transplantation technique was developed, which has since rapidly expanded to include adult donors (3,4). A prerequisite to living donor liver transplantation is the thorough investigation of living liver donors with regard to their hepatic vascular anatomy, biliary anatomy, and presence of steatosis, to avoid morbidity and mortality of living donors and to prevent graft failure (5,6). Indeed, studies have shown that in addition to laboratory and psychosocial issues, vascular or biliary anatomic variations, inadequate remnant liver volumes (,30%), and abnormal histologic findings, such
Advances in Knowledge nn Living liver donors with hepatic fibrosis, nonalcoholic steatohepatitis (NASH), and steatosis with inflammatory activity (liver stiffness [LS] values of 2.54 kPa 6 0.34, 2.17 kPa 6 0.38, and 2.12 kPa 6 0.38, respectively) showed higher LS values than donors with simple steatosis and normal livers (1.84 kPa 6 0.30 and 1.68 kPa 6 0.26, respectively) (P , .0001). nn Areas under the curve were 0.93 (cutoff value . 5.8%) for MR fat quantification for the detection of substantial hepatic steatosis (10%) and 0.85 (cutoff value . 1.94 kPa) for MR elastography for the detection of hepatic fibrosis (F1); by using those values, MR elastography and MR fat quantification could be used to detect the presence of clinically significant hepatic steatosis or fibrosis with 100% sensitivity (95% confidence interval [CI]: 73.4%, 100%) and 100% negative predictive value (95% CI: 78.0%, 100%) in an independent validation set. 2
as hepatic steatosis or hepatic fibrosis, were the main reasons for exclusion from donation in potential adult living liver transplantation donors (7–9). Recently, nonalcoholic fatty liver disease (NAFLD), including nonalcoholic steatohepatitis (NASH), has increasingly been recognized as the most common chronic liver disease worldwide, with reports of an increasing prevalence of NASH in the United States (10). This increasing prevalence of NAFLD in the general population poses a risk for organ donation, since allograft steatosis has been reported to be associated with nonfunctioning grafts (11,12), and the first case of living liver transplantation donor death was reported in Japan in a donor with NASH (5). Computed tomography (CT) and magnetic resonance (MR) imaging have shown accuracy comparable to that of diagnostic angiography or conventional endoscopic cholangiography for the depiction of liver anatomy (13,14). However, the ability to diagnose steatosis, steatohepatitis, inflammation, or fibrosis by using noninvasive imaging tools remains a point of controversy (12,15). According to a previous study performed in a single transplantation center, the prevalence of NAFLD diagnosed with biopsy was 51.4% among 589 potential living liver donors (16). However, ultrasonography (US) and CT
Implications for Patient Care nn Preoperative assessment of the presence of hepatic fibrosis (F1), including NASH or substantial (10%) macrovesicular hepatic steatosis in living liver donors, could be achieved noninvasively by using MR fat quantification and MR elastography. nn Combined use of MR elastography and MR fat quantification could potentially reduce the clinical demand for routine liver biopsies in living liver donor candidates by providing information regarding the risk of substantial (10%) macrovesicular hepatic steatosis or hepatic fibrosis (F1).
Yoon et al
have shown limitations in the detection of clinically significant hepatic steatosis or NASH in liver donors (16–18); as a result, preoperative liver biopsy continues to be regarded as an essential part of living liver donor work-ups in many centers (17,19,20). On the other hand, there are also many centers where only selected potential donors with potential clinical risk factors, such as high body mass index (BMI), abnormal liver enzyme levels, or abnormal imaging findings at US or CT, undergo preoperative liver biopsy (12,21) because it is an invasive procedure with the problem of high interobserver variation (12,22). However, until now, there has been no clear consensus regarding the role of noninvasive tools in identifying living liver donor candidates at risk for poor recipient outcome while maximizing the donor pool (21). With recent advances in MR technology, studies have demonstrated promising results of MR spectroscopy Published online before print 10.1148/radiol.15140908 Content codes: Radiology 2015; 000:1–12 Abbreviations: ALP = alkaline phosphatase ALT = alanine transaminase APRI = AST-to-platelet ratio index AST = aspartate transaminase AUC = area under the receiver operating characteristic curve BMI = body mass index CI = confidence interval FF = fat fraction gGT = g-glutamyl transpeptidase LS = liver stiffness NAFLD = nonalcoholic fatty liver disease NASH = nonalcoholic steatohepatitis ROI = region of interest Author contributions: Guarantor of integrity of entire study, J.M.L.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; agrees to ensure any questions related to the work are appropriately resolved, all authors; literature research, J.H.Y., J.M.L., K.S.S., K.B.L., J.K.H.; clinical studies, J.H.Y., J.M.L., K.S.S., K.W.L., N.J.Y., K.B.L., J.K.H.; statistical analysis, J.H.Y.; and manuscript editing, J.H.Y., J.M.L., K.S.S., K.W.L., N.J.Y., K.B.L., B.I.C. Conflicts of interest are listed at the end of this article.
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
and the proton density fat fraction (FF) technique in quantifying hepatic steatosis (23,24). It has also been shown that MR elastography can be promising for the staging of hepatic fibrosis and the detection of NASH in patients with NAFLD (25,26). Considering that MR imaging is frequently used as a noninvasive imaging test for the evaluation of vascular and biliary anatomy for living liver donor candidates prior to liver transplantation without radiation-related biologic hazards, we postulated that the combined use of MR fat quantification and MR elastography may be able to demonstrate both hepatic steatosis and fibrosis in living liver donor candidates, thereby potentially reducing the necessity of liver biopsy when selecting ideal donor candidates. The purpose of this study, therefore, was to evaluate the diagnostic performance of MR fat quantification and MR elastography for the assessment of macrovesicular hepatic steatosis and fibrosis in living liver donor candidates.
Materials and Methods Study Population This retrospective study was approved by the institutional review board of our hospital, and the requirement for informed consent was waived. From January 2010 to September 2014, 404 living liver donor candidates who agreed to donate liver tissue and passed psychological evaluation tests underwent liver resections for donation (n = 348) or liver parenchymal biopsies (n = 56) after undergoing both MR fat quantification and MR elastography. Among them, eight donors were excluded because no tissue specimens were obtained owing to emergency liver transplantation (n = 5) or registration for exercise and diet modification programs was required to reduce steatosis after MR examinations (n = 3). Finally, 396 living liver donors were included in this study. These donors were divided into two groups, depending on the date of the MR examination for two independent investigations. The test group included 362 living liver
Yoon et al
Figure 1
Figure 1: Flow diagram of the study population. LB = liver biopsy, LT = liver transplantation, M:F = ratio of male to female patients.
donors (234 men and 128 women; mean age 6 standard deviation, 31.7 years 6 10.4 in men and 36.5 years 6 11.0 in women) who underwent MR examinations between January 2010 and November 2013. The validation group included 34 donors (26 men and eight women; mean age, 35.3 years 6 10.4 in men and 36.8 years 6 9.4 in women) who underwent MR examination between December 2013 and September 2014 (Fig 1). All donors denied having moderate or heavy alcohol consumption for 6 months prior to liver transplantation and did not show positive viral markers of hepatitis B or C in preoperative tests. The median interval between MR imaging and surgery or biopsy was 10 days (range, 0–66 days; 25th percentile, 4 days; 75th percentile, 18 days). BMI was also recorded in all donors (mean BMI, 23.4 kg/m2 6 3.1 in the test group and 23.5 kg/ m2 6 3.2 in the validation group). The following laboratory findings were also recorded: albumin level, total bilirubin
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
level, alkaline phosphatase (ALP), aspartate transaminase (AST), alanine transaminase (ALT), g-glutamyl transpeptidase (gGT), and platelet count. In addition, the AST-ALT ratio and ASTto-platelet ratio index (APRI) were calculated by using 40 IU/L (0.67 mkat/L) of AST as the upper normal limit.
MR Examination MR fat quantification technique.—Living liver donor candidates were asked to fast for at least 8 hours prior to MR imaging. Fat quantification was performed by using either the multiecho chemical-shift imaging sequence (n = 313, IDEAL-IQ; GE Healthcare, Waukesha, Wis) on a 1.5-T MR unit (Signa HDxt; GE Healthcare) or T2-corrected multiecho single-voxel hydrogen 1 (1H) MR spectroscopy (n = 83, Siemens Healthcare, Erlangen, Germany) on a 3.0-T MR unit (Magnetom Verio; Siemens Healthcare). MR elastography.—All MR elastography examinations were performed 3
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
on commercially available MR elastography hardware (MR Touch; GE Healthcare) with a 1.5-T imager (Signa HDxt; GE Healthcare) by using a twodimensional MR elastography protocol, which has been described previously, and the direct inversion algorithm (27). The two-dimensional MR elastography protocol used in our study was similar to the one described previously in the literature (28). In brief, imaging was performed in the supine position, with an acoustic pressureactivated passive driver placed against the body wall adjacent to the liver. The passive driver was connected to an active acoustic driver system located outside of the imaging room via a polyvinylchloride tube. Low-frequency longitudinal mechanical waves of 60 Hz were transmitted into the liver by a passive driver, as synchronized with the imaging sequence. Details of MR elastography acquisition parameters are provided in Appendix E1 (online).
Image Analysis Measurements of hepatic FF.—One attending radiologist (J.H.Y., with 8 years of clinical experience in MR image interpretation) who was blinded to the histologic results measured the hepatic FF on FF maps, which were generated automatically. Circular regions of interest (ROIs) approximately 300 mm2 6 38.4 in size (range, 250–400 mm2) were placed in the right liver lobe three times, and the mean value of the three measurements was used. If large hepatic vessels or focal liver lesions were seen on FF maps, vessels and lesions were carefully avoided. With MR spectroscopy, the automatically calculated hepatic FF (percentage) was used and displayed as a percentage in Digital Imaging and Communications in Medicine format (29). A detailed MR fat quantification protocol is given in Appendix E1 (online). We did not divide FF results from the two fat quantification techniques according to the literature (24,30,31). Measurement of the liver stiffness value.—As for MR elastography, liver stiffness (LS) values of the hepatic parenchyma were measured by placing 4
four ROIs on the elastogram. ROIs (mean area, 3837.8 mm2 6 1989.5; range, 1018.5–6352.3 mm2) were placed by one attending radiologist (J.H.Y.) who was blinded to the histologic findings. All ROIs were drawn in the area indicated to have high confidence and good signal-to-noise ratio, with stiffness outliers excluded on confidence maps (32) first and then copied to the corresponding position on stiffness maps that provided stiffness values in kilopascals. After reconfirming whether the ROIs were adequately placed in the right liver lobe, LS values (in kilopascals) were calculated as the median value of multiple ROIs.
Histologic Analysis Surgical liver biopsy was performed in the right lobe of the liver in all liver donors during the procedure, and approximately 1 mm3 of liver parenchyma was excised from the right lobe of the liver. US-guided percutaneous liver biopsy was performed prior to the procedure in 56 donors who showed abnormal biochemical liver function tests, such as increase of serum aminotransferase levels and evidence of hepatic steatosis seen on US, CT, or MR images in the right lobe of the liver by using an 18-gauge automated gun (TSK Laboratory, Tochigi, Japan) twice. If the candidate underwent percutaneous liver biopsy prior to surgery, areas of macrovesicular steatosis assessed from the percutaneous liver biopsy specimen were used for histologic analysis. The degree of macrovesicular steatosis was defined as the percentage of hepatocytes that contained intracellular macrovesicular fat droplets at hematoxylin-eosin staining (12). Although potential donors with mild steatosis are commonly denied from live donation, depending on graft volume and donor age (12), in our center, at least 10% macrovesicular steatosis was regarded as “substantial macrovesicular steatosis,” representing a threshold for donor candidates who required further evaluations or dietary interventions to achieve weight reduction prior to donation. NASH and steatosis with inflammatory activity were diagnosed on the
Yoon et al
basis of NAFLD activity scores (33). Donors who had NAFLD activity scores of at least 5 received diagnoses of NASH, and donors with NAFLD activity scores of 3 or 4 were classified as having steatosis with inflammatory activity. Donors with NAFLD activity score of 1 or 2 received diagnoses of simple steatosis. Hepatic fibrosis was assessed according to standardized guidelines proposed by the Korean Study Group for the Pathology of Digestive Diseases, which is similar to the METAVIR scoring system, as follows: F0 = no fibrosis, F1 = portal fibrosis, F2 = periportal fibrosis, F3 = septal fibrosis, and F4 = cirrhosis (34,35). Details of histologic analysis are given in Appendix E1 (online).
Statistical Analysis Correlation analyses were performed between FF on MR images and macrovesicular fat in the specimen, between BMI and macrovesicular hepatic steatosis, and between hepatic fibrosis and LS values (36). Donors were grouped into five groups on the basis of histologic findings of the liver (normal liver, simple steatosis, steatosis with inflammatory activity, NASH, and hepatic fibrosis). FF, LS values, degrees of macrovesicular steatosis, clinical findings, and laboratory findings, including the AST-ALT ratio and APRI score, were compared among the five groups by using either analysis of variance or Kruskal-Wallis tests, as well as post hoc analysis after a normality test. In this study, clinically significant hepatic steatosis was defined as macrovesicular fat of at least 10% on the basis of previous studies (37,38), and presence of hepatic fibrosis was defined as fibrosis score of at least F1. Factors associated with steatosis or fibrosis were determined by using multiple regression analysis. In addition, diagnostic performance of MR fat quantification and MR elastography was assessed by using receiver operating characteristic analysis to identify clinically significant macrovesicular hepatic steatosis or the presence of hepatic fibrosis. Finally, sensitivity, specificity, positive predictive value, and negative predictive value,
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
Study Results of the Test Group Demographics of living liver donors with different histologic findings.—Living liver donors received histologic diagnoses as follows: normal parenchyma (n = 244), simple steatosis (n = 71), steatosis with inflammatory activity (n = 21), NASH (n = 17; F1, n = 13; F2, n = 4), and hepatic fibrosis (n = 9; F1, n = 7; F2, n = 2) (33). Detailed demographics are summarized in Table 1. There were no significant differences in age among the groups, except between donors with normal liver parenchyma and donors with simple steatosis (P = .033). BMI of donors with normal parenchyma (22.8 kg/m2 6 2.9) was significantly lower than that in donors with NAFLD (steatosis with inflammatory activity and NASH, P , .0001 and P , .013, respectively), except in donors with simple steatosis (P = .051). None of the laboratory findings were found to be significantly different among donors with different histologic findings (Table 1). At pathologic examination, macrovesicular FF was shown to be higher in donors with NAFLD than in donors with normal parenchyma (P , .0001) or hepatic fibrosis (simple steatosis vs fibrosis, P , .0001; steatosis with inflammation vs fibrosis, P , .0001; and NASH vs fibrosis, P = .007). However, macrovesicular FF did not show significant differences between donors with NASH, steatosis with inflammatory activity, and simple steatosis (P = .066–.761). including 95% confidence interval (CI), were obtained in a validation set to predict either the presence of clinically significant hepatic steatosis or the presence of hepatic fibrosis by using cutoff values of MR fat quantification and MR elastography obtained in the test group. Details of the statistical analysis are provided in Appendix E1 (online). All statistical analyses were performed by using commercially available software (IBM SPSS, version 21, IBM, Armonk, NY; and MedCalc, version 12, MedCalc Software, Mariakerke, Belgium).
Results
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
156 88 32.2 6 10.6 (18–61) 22.8 6 2.9 (16.0–32.3) 232.6 6 49.9 (138.8–347.2) 4.33 6 0.36 (3.7–5.1) 0.82 6 0.48 (0.3–2.25) 64.85 6 20.29 (28–151) 22.68 6 14.36 (12–122) 23.03 6 18.70 (6–96) 26.84 6 21.62 (7–134) 1.18 6 0.63 (0.39–4.0) 0.25 6 0.21 (0.21–2.36) 3.34 v 1.83 (0.6–9.0) 1.69 6 1.51 (0.0–4.0) 1.68 6 0.26 (1.0–2.31)
Patient sex No. of men No. of women Age (y) BMI (kg/m2) Platelet count (3109/L) Albumin level (g/dL) Total bilirubin level (mg/dL) ALP (IU/L) AST (IU/L) ALT (IU/L) gGT (IU/L) AST-ALT ratio APRI score FF on MR images (%) Macrovesicular fat (%) LS value (kPa) 48 23 36.5 6 11.6 (18–61) 23.9 6 3.0 (16.7–32.9) 239.2 6 49.0 (172.4–379.8) 4.33 6 0.36 (3.7–5.1) 0.96 6 1.07 (0.3–2.7) 62.04 6 18.00 (30–182) 26.45 6 27.37 (10–144) 24.64 6 29.21 (10–71.5) 30.33 6 22.96 (9–89) 1.39 6 1.01 (0.42–7.0) 0.29 6 0.36 (0.13–0.93) 6.65 6 4.70 (2.0–23.9) 7.18 6 6.31 (5.0–40.0) 1.84 6 0.30 (1.32–2.58)
Simple Steatosis (n = 71) 16 5 34.2 6 7.7 (22–53) 26.1 6 3.6 (18.9–34.6) 249.2 6 64.4 (132.0–373.0) 4.41 6 0.49 (3.7–4.9) 1.06 6 0.80 (0.3–1.5) 71.43 6 21.09 (30–96) 25.57 6 17.47 (15–84) 27.57 6 22.36 (12–89.0) 36.76 6 30.86 (10–66) 1.11 6 0.47 (0.44–2.1) 0.24 6 0.23 (0.16–1.25) 10.90 6 4.50 (3.0–18.5) 13.12 6 9.06 (8.9–40.0) 2.12 6 0.38 (1.52–3.07)
Steatosis with Inflammatory Activity (n = 21)*
5 4 33.9 6 12.5 (20–52) 24.8 6 3.9 (19.1–29.0) 243.4 6 25.6 (158.0–339.0) 4.52 6 0.32 (3.7–4.9) 1.02 6 0.51 (0.3–2.5) 77.89 6 32.21 (30–77) 17.67 6 3.94 (16–94) 13.89 6 6.85 (7–101) 19.11 6 10.07 (10–122) 1.51 6 0.78 (0.83–3.0) 0.18 6 0.03 (0.24–1.27) 4.02 6 2.50 (1.0–7.7) 3.00 6 3.20 (0.0–8.0) 2.54 6 0.34 (1.96–3.00)
38.9 6 9.6 (20–52) 25.2 6 2.6 (20.7–29.9) 241.9 6 53.9 (180.0–310.0) 4.39 6 0.27 (3.4–4.9) 0.81 6 0.38 (0.4–1.6) 70.00 6 19.56 (46–119) 29.94 6 29.97 (13–46) 26.76 6 28.23 (11–46) 23.63 6 14.37 (10–134) 1.28 6 0.44 (0.61–1.95) 0.35 6 0.43 (0.20–1.50) 10.70 6 5.67 (1.1–23.0) 15.00 6 10.10 (7.0–40.0) 2.17 6 0.38 (1.47–2.80)
Fibrosis (n = 9)
9 8
NASH (n = 17)
.009‡ ,.001 .559 .494 .303 .086 .260 .527 .207 .162 .372 ,.001 ,.001 ,.001
…
P Value†
Differences among the five groups.
Only normal parenchyma and simple steatosis groups showed significant differences (P = .033).
†
‡
* Nonalcoholic fatty liver disease score 4.
Note.—Unless indicated otherwise, values are means 6 standard deviations, with ranges in parentheses. The upper normal limit of AST is 40 IU/L (0.67 mkat/L) at our institution. To convert milligrams per deciliter to micromoles per liter for total bilirubin level, multiply by 17.104. To convert international units per liter to microkatals per liter for ALP, AST, ALT, and gGT, multiply by 0.0167.
Normal Parenchyma (n = 244)
Parameter
FF on MR Images, Macrovesicular Fat, and LS Values at MR Elastography in Specimens from Living Liver Donors with Different Histologic Findings in the Test Group
Table 1
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors Yoon et al
5
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
Yoon et al
Figure 2
Figure 2: Box and whisker plots of (a) FF on MR images and (b) LS values at MR elastography for the five different histologic findings of the liver. Normal liver parenchyma and hepatic fibrosis had the lowest fat measurement, followed by simple steatosis (a). Steatosis with inflammatory activity and NASH showed the highest FF (a). Normal liver showed the lowest LS value, followed by simple steatosis. Steatosis with inflammatory activity, NASH, and hepatic fibrosis showed significantly higher LS values (b).
Hepatic FF values of living liver donor groups measured with MR techniques.—FF measured with MR techniques in donors with NAFLD was significantly higher than that in donors with normal liver parenchyma or donors with fibrosis (P , .001). Among donors with NAFLD, FF on MR images was significantly higher in donors with NASH (10.70% 6 5.67) and steatosis with inflammatory activity (10.90% 6 4.50) (P = .031) than in donors with simple steatosis (6.65% 6 4.70) (P = .006), similar to pathologic assessment (Figs 2a, 3–6). LS values of living liver donor groups measured with MR elastography.—Living liver donors with hepatic fibrosis showed significantly higher LS values (2.54 kPa 6 0.34) than those with normal parenchyma (1.68 kPa 6 0.26, P , .0001) and those with simple steatosis (1.84 kPa 6 0.30, P , .001). In addition, donors with NASH and steatosis with inflammatory activity showed significantly higher LS values (2.17 kPa 6 0.38 and 2.12 kPa 6 0.38, respectively; P = .004) than those with simple steatosis (P = .004). There were no significant 6
Figure 3
Figure 3: (a) FF map and (b) MR elastogram in a 51-year-old woman who donated the right lobe of her liver. At excisional liver biopsy, the liver demonstrated 2% macrovesicular fat and no evidence of hepatic fibrosis. The FF map (a) showed that the FF was 2.1%, and the LS value was 1.3 kPa on the elastogram (b), suggesting normal liver parenchyma without substantial fat infiltration.
differences in LS values among donors with NASH, steatosis with inflammatory activity (P = .812), and hepatic fibrosis (P = .204) (Figs 2b, 3–6).
Correlation between MR Values and Histologic Findings Correlation between FF on MR images, BMI, and fat in the specimen.— FF on MR images showed moderate
correlation with macrovesicular fat in the specimen (r = 0.77, 95% CI: 0.73, 0.81; P , .001). Only a weak correlation was found between BMI and macrovesicular fat (r = 0.27, 95% CI: 0.17, 0.36; P , .0001). Correlation between LS at MR elastography, fibrosis, and inflammatory activity.—LS values showed a significant correlation with hepatic fibrosis (r =
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
0.40, 95% CI: 0.31, 0.48; P , .0001). LS values also showed a significant correlation with inflammatory activity (r = 0.394, 95% CI: 0.30, 0.48; P , .0001).
Diagnostic Performance of MR Fat Quantification and MR Elastography for the Detection of Substantial Hepatic Steatosis or Presence of Hepatic Fibrosis (F1) in Comparison with Clinical Data For the detection of substantial hepatic steatosis (macrovesicular fat 10%), the MR fat quantification technique showed a higher area under the receiver operating characteristic curve (AUC) than did BMI (0.93 vs 0.69, P , .0001) and higher sensitivity (90.2% [46 of 51 patients] vs 68.6% [35 of 51 patients], P = .0039; Table 2) by using a cutoff value of 5.8%. The AUC of LS at MR elastography for detecting the presence of hepatic fibrosis (F1) was 0.85 (95% CI: 0.81, 0.89), showing 84.6% sensitivity (22 of 26 patients) by using a cutoff value of 1.94 kPa (Table 2). According to multiple regression analyses, FF, LS value, and gGT were associated with macrovesicular fat (R2 = 0.599, P , .0001), and LS value, FF, and age were associated with fibrosis stage (R2 = 0.350, P , .0001). Only FF and LS values were significantly associated with the presence of hepatic fibrosis (F1) or substantial hepatic steatosis (10%) at multiple logistic regression analysis (AUC, 0.935; 95% CI: 0.905, 0.958), providing 96.9% sensitivity (62 of 64 patients, 95% CI: 89.1, 99.5) and 99.0% negative predictive value (204 of 206 patients, 95% CI: 96.5, 99.9; Table 2). By using previously obtained cutoff values (fat on MR images . 5.8%; LS value . 1.94 kPa), the combination of MR fat quantification and MR elastography showed comparable results with the model: 96.9% sensitivity (62 of 64 patients, 95% CI: 89.1, 99.5) and 99.0% negative predictive value (208 of 210 patients, 95% CI: 96.6, 99.9; Table 2). Study Results of the Validation Group In the validation set (n = 34), FF showed significant correlation with the histologic fat amount (r = 0.91, P , .0001, 95%
Yoon et al
Figure 4
Figure 4: (a) FF map and (b) MR elastogram in a 42-year-old woman who underwent percutaneous liver biopsy for preoperative work-up. On the FF map, the liver showed clinically significant fatty infiltration (14.6%) (a), and the LS value was 1.7 kPa at MR elastography (b). Macrovesicular FF was 15% in the specimen, but neither lobular activity nor other evidence of NASH was found.
Figure 5
Figure 5: (a) FF map and (b) MR elastogram in a 31-year-old man who underwent preoperative percutaneous liver biopsy. On the FF map, FF was estimated to be 12.8% (a), and LS was 2.8 kPa at MR elastography (b). After liver biopsy, the patient received a diagnosis of NASH stage 1c.
Figure 6
Figure 6: (a) FF map and (b) MR elastogram in a 21-year-old man who had no history of liver disease. On the FF map, estimated fat fraction was 3% (a), but LS at MR elastography was 2.6 kPa (b). At liver biopsy, the patient received a diagnosis of early-stage hepatic fibrosis (F1, mild activity) without substantial fat infiltration (macrovesicular fat, 1%; microvesicular fat, 1%).
CI: 0.83, 0.96), and LS values showed significant correlation with fibrosis (r = 0.73, P , .0001, 95% CI: 0.51, 0.85).
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
FF, LS values, and other parameters are summarized in Table 3. By using the cutoff values derived from test data set (fat 7
8
†
Note.—Numbers in parentheses are the data used to calculate percentages. Numbers in brackets are 95% CIs.
Findings were considered positive when a donor showed either more than 5.98% fat with the MR fat quantification tool or stiffness of more than 1.94 kPa at MR elastography.
Discussion
* Classification cutoff value was set as 0.051 (when the probability that the patient had either fibrosis or substantial steatosis was more than 5%).
99.0 (204/206) [96.5, 99.9] 99.0 (208/210) [96.6, 99.9] 39.7 (62/156) [32.0, 47.9] 40.8 (62/152) [32.9, 49.1] 96.9 (62/64) [89.1, 99.5] 96.9 (62/64) [89.1, 99.5] … … 0.935 [0.91, 0.96] …
Yoon et al
on MR images . 5.8%; LS value . 1.94 kPa), a combination of the MR fat quantification technique and MR elastography showed 100% sensitivity (12 of 12 patients, 95% CI: 73.4%, 100%), 68.2% specificity (15 of 22 patients, 95% CI: 45.1%, 86.1%), 63.2% positive predictive value (12 of 19 patients, 95% CI: 19.8%, 53.5%), and 100% negative predictive value (15 of 15 patients, 95% CI: 78.0%, 100%).
68.5 (204/298) [62.9, 73.7] 69.8 (208/298) [64.2, 74.8]
98.5 (258/262) [96.1, 99.6] 22.0 (22/100) [14.3, 31.4] .1.94 0.85 [0.81, 0.89]
84.6 (22/26) [65.1, 95.5]
76.8 (258/336) [71.9, 81.2]
98.1 (261/266) [95.7, 99.4] 93.0 (214/230) [89.0, 96.0] 47.9 (46/96) [37.6, 58.4] 26.5 (35/132) [19.2, 34.9] .5.8 .24.09
Detection of substantial hepatic steatosis (10%) FF on MR images BMI Detection of the presence of hepatic fibrosis (F1) LS value at MR elastography Detection of a clinically significant pathologic condition (either substantial steatosis or presence of fibrosis) Multiple logistic regression model* Use of previously obtained two cutoff values†
0.93 [0.90, 0.95] 0.69 [0.64, 0.74]
90.2 (46/51) [78.6, 96.7] 68.6 (35/51) [54.0, 80.5]
83.9 (261/311) [79.4, 87.8] 68.8 (214/311) [63.3, 73.9]
Negative Predictive Value (%) Positive Predictive Value (%) Specificity (%) Sensitivity (%) Cutoff Value AUC Parameter
Diagnostic Performance of FF with the MR Fat Quantification Technique, BMI, and LS Value at MR Elastography for Detection of Either Substantial Hepatic Steatosis or Presence of Hepatic Fibrosis in the Test Group
Table 2
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
Our study demonstrated that a combination of MR fat quantification and MR elastography was able to provide high sensitivity (96.9%, 62 of 64 patients) for the prediction of substantial hepatic steatosis or hepatic fibrosis (F1) in living liver donor candidates in the test group, and the validation data set showed 100% sensitivity (12 of 12 patients, 95% CI: 73.4%, 100%) and 100% negative predictive value (15 of 15 patients, 95% CI: 78.0%, 100%). In addition, both examinations showed high accuracy in the detection of clinically significant hepatic steatosis (AUC, 0.93), as well as the presence of hepatic fibrosis (AUC, 0.85). Our study results show good agreement with previous studies, which also demonstrated excellent accuracy in the estimation of hepatic FF by using multiecho chemicalshift imaging with multipeak fat spectrum modeling or T2-corrected multiecho single-voxel 1H MR spectroscopy (39,40), as well as significant correlation between LS values at MR elastography and the degree of hepatic fibrosis at pathologic examination (25,41). Considering that both methods are achieved within a breath hold and cover a larger area than percutaneous liver biopsy, as well as the high accuracy in the prediction of substantial hepatic steatosis, MR fat quantification and MR elastography may serve as a screening tool for the selection of donors without a risk of substantial hepatic steatosis and/or hepatic fibrosis. In addition, our study also showed that there were significant differences in LS values measured with MR elastography between donors with NASH and simple steatosis (P = .023). radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
This result is consistent with the results of prior studies, in which simple steatosis was also distinguished from NASH by using MR elastography (26,42). Given that it is important to differentiate donor candidates with simple steatosis from those with NASH, since there have been reports that sole hepatic steatosis is not a contradiction for living donor liver transplantation (43), as well as the report of mortality in a donor with NASH (44), our study results suggest that LS measurements obtained with MR elastography could provide substantial clinical value in the selection of living liver donors by showing significant differences in LS values between donors with NASH and simple steatosis (P = .023). Therefore, we believe that MR elastography may be able to provide much value in the evaluation of liver donor candidates, in addition to MR fat quantification techniques. In living donor liver transplantation, maximizing donor safety is of paramount importance (12,44). To guarantee donor safety, routine liver biopsies have been recommended in previous studies (17,20). However, this policy provides a paradox: To maximize donor safety, an invasive preoperative examination, with the possibility of morbidity and mortality, should be performed (20). Indeed, concerns of possible morbidity and mortality keep personnel in many liver transplantation programs— up to 26%—from performing routine liver biopsies. Instead, they resort to an inconsistent selection criterion that may cause postoperative complications (21). Many transplantation centers have reported unexpected nonsteatotic hepatopathy demonstrated at liver biopsy in donor candidates (17,20) to be one of the reasons for supporting routine liver biopsies in donor candidates. However, our study results showed that the combined use of MR fat quantification techniques and MR elastography provided high sensitivity for the detection of steatotic hepatopathy, as well as hepatic fibrosis without steatosis. Furthermore, we found that MR elastography and MR fat quantification techniques provided better performance than clinical and laboratory findings,
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
15 7 34.7 6 10.3 (20–58) 22.4 6 2.3 (19.6–27.3) 238 6 46.3 (172–362) 4.6 6 0.2 (4.2–5.0) 1.0 6 0.6 (0.4–2.7) 55.5 6 14 (36.2–91.2) 22.5 6 12.3 (14.0–69.4) 25.2 6 26.4 (11.1–134.8) 30.6 6 31.7 (11.0–137.2) 1.03 6 0.25 (0.52–1.46) 0.28 6 0.33 (0.11–1.65) 5.19 6 1.40 (3.72–8.46) 2.59 6 1.65 (0.0–4.0) 1.76 6 0.19 (1.66–1.91)
Patient sex No. of men No. of women Age (y) BMI (kg/m2) Platelet count (3109/L) Albumin level (g/dL) Total bilirubin level (mg/dL) ALP (IU/L) AST (IU/L) ALT (IU/L) gGT (IU/L) AST-ALT ratio APRI score FF on MR images (%) Macrovesicular fat (%) LS value (kPa) (30–41) (21.6–22.2) (249–400) (4.6–4.9) (0.7–1.0) (43.0–50.0) (25–32) (40–50) (27–94) (0.63–0.64) (0.31–0.40) (8.45–17.64) (10.0–35.0) (1.66–1.91)
2 0
Simple Steatosis (n = 2) 1 0 38 27.1 260 4.6 0.5 78 35 83 65 0.42 0.8 8.5 10 2.17
Steatosis with Inflammatory Activity* (n = 1)
31.8 6 5.5 (23–37) 25.3 6 4.4 (19.6–30.8) 257.3 6 53.7 (213–352) 4.7 6 0.1 (4.5–4.8) 1.1 6 0.9 (0.4–2.7) 47.8 6 8.6 (32.0–55.0) 20 6 4.4 (13.0–26.0) 29.3 6 16.5 (14.0–61.0) 48.4 6 36.2 (24.0–108.0) 0.77 6 0.18 (0.43–0.93) 0.29 6 0.18 (0.16–0.63) 12.7 6 5.84 (8.90–24.48) 17.83 6 14.77 (7.0–45.0) 2.18 6 0.40 (1.92–2.95)
5 1
NASH (n = 6)
50 6 8.5 (42–59) 27.5 6 2.9 (24.3–29.9) 235 6 15.1 (221–251) 4.5 6 0.3 (4.3–4.9) 1.3 6 0.6 (0.9–2.0) 62.3 6 11.4 (53.0–75.0) 23.7 6 2.3 (21.0–25.0) 24.3 6 7.7 (18.0–33.0) 45 6 13.2 (35.0–60.0) 1.02 6 0.23 (0.76–1.17) 0.26 6 0.09 (0.18–0.35) 5.45 6 2.34 (2.89–7.47) 4.0 6 3.60 (1.0–8.0) 2.29 6 0.35 (2.00–2.68)
3 0
Fibrosis (n = 3)
.115 .017 .220 .846 .746 .165 .680 .177 .499 .011 .540 ,.0001 ,.0001 .0001
P Value†
†
Difference among the five groups.
* NAFLD score 4.
Note.—Values are means 6 standard deviations, with ranges in parentheses. A P value less than .05 indicates a statistically significant difference. Pairwise comparison could not be performed, because one group (steatosis with inflammatory activity) included only one donor. The upper normal limit of AST is 40 IU/L (0.67 mkat/L) at our institution. To convert milligrams per deciliter to micromoles per liter for total bilirubin level, multiply by 17.104. To convert international units per liter to microkatals per liter for ALP, AST, ALT, and gGT, multiply by 0.0167.
Normal Parenchyma (n = 22)
Parameter
FF on MR Images, Macrovesicular Fat, and LS Values at MR Elastography in Specimens from Living Liver Donors with Different Histologic Findings in the Validation Group
Table 3
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors Yoon et al
9
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
such as BMI or increased aminotransferase levels, in the detection of substantial hepatic steatosis, as well as the presence of hepatic fibrosis. These results are in good agreement with those of a previous study, which showed that while there is a significant correlation between BMI and degree of hepatic steatosis, 74% of overweight donors (BMI . 25) had 0%–10% steatosis (17). Considering that MR imaging is noninvasive, has no radiation hazard, and can be performed in a short period of time, we believe it would be valuable to implement MR elastography and fat quantification techniques in preoperative MR imaging protocols for living liver donor work-ups, as they can be used as a screening tool for the preoperative selection of candidates for liver biopsy among liver transplant donors. We anticipate that our study results can also serve as a reference for the selection of living liver donors in liver transplantation centers that do not perform routine liver biopsy. Implementation of MR elastography and MR quantification tools in preoperative liver donor evaluation strategies has other advantages over liver biopsy, which is currently the standard of reference for evaluation of NAFLD. To improve postoperative outcomes and solve the problem of organ shortages, intensive dietary and lifestyle management is currently performed in donors with hepatic steatosis (45–47). Although this approach can expand the pool of marginal living liver donors (45), follow-up should be performed by monitoring laboratory findings, weight changes, or repeated liver biopsies. Therefore, there is a great clinical need for a noninvasive assessment tool to be able to grade the improvement of NAFLD after exercise and diet modification. By using liver fat quantification and MR elastography, we can further reduce the number of repeated liver biopsies. Another advantage of MR imaging compared with liver biopsy is its all-in-one evaluation capability. Liver biopsies usually require short-term hospitalization, which increases medical costs, recipients’ economic burden, and donor inconvenience. MR imaging, on the other hand, 10
can provide a comprehensive, one-stop evaluation, including MR arteriography, MR cholangiography, MR fat quantification, and MR elastography, which can show anatomic and histologic information to surgeons, thereby shortening the preoperative work-up process and decreasing the inconvenience to potential donors who have to undergo variable imaging tests. Our study results of MR elastography are slightly different from the results of previous studies on the evaluation of NAFLD. First, in our study, MR elastography findings did not allow us to differentiate donors that had NASH from those that had steatosis with inflammatory activity, although NASH showed higher LS values than did simple steatosis. Our results are different from those of a previous study, which demonstrated the early detection of NASH in patients with nonalcoholic fatty liver disease by using MR elastography (26). This discrepancy can be explained by the differing study populations (living liver donor candidates vs patients) and study designs (prospective vs retrospective) between our studies. Although our study population included only living liver donor candidates who had no clinical suspicion of liver disease, the study population in the previous study was very heterogeneous and included patients clinically suspected of having liver diseases. Furthermore, in our study, the severity of NASH was mild and demonstrated F1 and F2 scores only. Considering that inflammation is a well-known confounding factor of hepatic fibrosis that increases LS values, MR elastography may have a limitation in the differentiation of NASH with early stage of fibrosis from steatosis with inflammatory activity (25,48). Second, the cutoff value of LS for detecting hepatic fibrosis (F1) from F0 or a normal liver was 1.94 kPa, which is lower than the 2.87 kPa (25) and 2.4 kPa (41) values reported in previous studies. The reason for the lower LS values is not clear, but we can assume that different study populations might have led to different cutoff values. In this study, young and healthy donors were included, whereas
Yoon et al
patients with chronic liver disease were included in those previous studies. Considering that inflammatory activity could increase LS values without fibrosis (49), cutoff values for detection of hepatic fibrosis from F0 in previous studies could be higher than cutoff values for detection of hepatic fibrosis in living liver donors with normal parenchyma in our study. This study has several limitations. First, since it was a retrospective, single-center study, there may have been selection bias. Second, the study population included a large portion of donors with normal livers but did not include many candidates who had more than 30% macrovesicular steatosis or clinically significant fibrosis (.F2). Therefore, this deviated study population may have affected the results of cutoff values in both examinations. However, considering that only potential living liver donors without significantly abnormal biochemical liver test findings or any potential illnesses are considered to undergo imaging workup in many transplantation centers, we believe that our study results are closer to the real clinical practice of donor selection in liver transplantation centers. Third, we performed MR elastography at 1.5 T by using the twodimensional technique, and it has been reported that different MR elastography techniques provided significantly different LS values. As a result, we cannot simply extrapolate the cutoff value to the results obtained with different methods. In conclusion, a combination of MR fat quantification and MR elastography can provide sufficient sensitivity to detect substantial (10%) macrovesicular hepatic steatosis or fibrosis (F1) in living liver donor candidates, potentially reducing the necessity of liver biopsy. Acknowledgment: We thank Chris Woo, BA, for his editorial assistance. Disclosures of Conflicts of Interest: J.H.Y. disclosed no relevant relationships. J.M.L. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: author received nonfinancial technical support from Siemens Healthcare; author received grants from Donseo Medical, CMS, Acuzen, Starmed, RF Medical, and Bayer
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
Healthcare; author received personal fees from Bayer Healthcare for lectures. Other relationships: disclosed no relevant relationships. K.S.S. disclosed no relevant relationships. K.W.L. disclosed no relevant relationships. N.J.Y. disclosed no relevant relationships. K.B.L. disclosed no relevant relationships. J.K.H. disclosed no relevant relationships. B.I.C. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: author received a grant from Samsung Electronics. Other relationships: disclosed no relevant relationships.
References 1. Mazzaferro V, Regalia E, Doci R, et al. Liver transplantation for the treatment of small hepatocellular carcinomas in patients with cirrhosis. N Engl J Med 1996;334(11):693–699. 2. Adam R, McMaster P, O’Grady JG, et al. Evolution of liver transplantation in Europe: report of the European Liver Transplant Registry. Liver Transpl 2003;9(12):1231– 1243. 3. Trotter JF. Selection of donors and recipients for living donor liver transplantation. Liver Transpl 2000;6(6 Suppl 2):S52–S58. 4. Kim KW, Lee J, Lee H, et al. Right lobe estimated blood-free weight for living donor liver transplantation: accuracy of automated blood-free CT volumetry—preliminary results. Radiology 2010;256(2):433–440. 5. Akabayashi A, Slingsby BT, Fujita M. The first donor death after living-related liver transplantation in Japan. Transplantation 2004;77(4):634. 6. Kiuchi T, Tanaka K, Ito T, et al. Small-forsize graft in living donor liver transplantation: how far should we go? Liver Transpl 2003;9(9):S29–S35. 7. Sharma A, Ashworth A, Behnke M, Cotterell A, Posner M, Fisher RA. Donor selection for adult-to-adult living donor liver transplantation: well begun is half done. Transplantation 2013;95(3):501–506. 8. Tsang LL, Chen CL, Huang TL, et al. Preoperative imaging evaluation of potential living liver donors: reasons for exclusion from donation in adult living donor liver transplantation. Transplant Proc 2008;40(8):2460– 2462. 9. Fan ST, Lo CM, Liu CL, Yong BH, Chan JK, Ng IO. Safety of donors in live donor liver transplantation using right lobe grafts. Arch Surg 2000;135(3):336–340. 10. Kemmer N, Neff GW, Franco E, et al. Nonalcoholic fatty liver disease epidemic and its implications for liver transplantation. Transplantation 2013;96(10):860–862.
Yoon et al
11. Said A. Non-alcoholic fatty liver disease and liver transplantation: outcomes and advances. World J Gastroenterol 2013;19(48): 9146–9155.
23. Tang A, Tan J, Sun M, et al. Nonalcoholic fatty liver disease: MR imaging of liver proton density fat fraction to assess hepatic steatosis. Radiology 2013;267(2):422–431.
12. McCormack L, Dutkowski P, El-Badry AM, Clavien PA. Liver transplantation using fatty livers: always feasible? J Hepatol 2011;54(5):1055–1062.
24. Reeder SB, Cruite I, Hamilton G, Sir lin CB. Quantitative assessment of liver fat with magnetic resonance imaging and spectroscopy. J Magn Reson Imaging 2011; 34(4):729–749.
13. Song GW, Lee SG, Hwang S, et al. Pre operative evaluation of biliary anatomy of donor in living donor liver transplantation by conventional nonenhanced magnetic resonance cholangiography. Transpl Int 2007;20(2):167–173. 14. Lee SS, Kim TK, Byun JH, et al. Hepatic arteries in potential donors for living related liver transplantation: evaluation with multidetector row CT angiography. Radiology 2003;227(2):391–399. 15. Schwenzer NF, Springer F, Schraml C, Stefan N, Machann J, Schick F. Non-invasive assessment and quantification of liver steatosis by ultrasound, computed tomography and magnetic resonance. J Hepatol 2009;51(3):433–445. 16. Lee JY, Kim KM, Lee SG, et al. Preva lence and risk factors of non-alcoholic fatty liver disease in potential living liver donors in Korea: a review of 589 consecutive liver biopsies in a single center. J Hepatol 2007;47(2):239–244. 17. Ryan CK, Johnson LA, Germin BI, Mar cos A. One hundred consecutive hepatic biopsies in the workup of living donors for right lobe liver transplantation. Liver Transpl 2002;8(12):1114–1122. 18. Park SH, Kim PN, Kim KW, et al. Mac rovesicular hepatic steatosis in living liver donors: use of CT for quantitative and qualitative assessment. Radiology 2006;239(1): 105–112. 19. Brandhagen D, Fidler J, Rosen C. Evaluation of the donor liver for living donor liver transplantation. Liver Transpl 2003;9(10 Suppl 2):S16–S28. 20. Nadalin S, Malagó M, Valentin-Gamazo C, et al. Preoperative donor liver biopsy for adult living donor liver transplantation: risks and benefits. Liver Transpl 2005;11(8):980–986. 21. Brown RS Jr, Russo MW, Lai M, et al. A survey of liver transplantation from living adult donors in the United States. N Engl J Med 2003;348(9):818–825. 22. Brancatelli G. Science to practice: Should biopsy be performed in potential liver donors when unenhanced CT shows an unacceptable degree of steatosis for transplantation? Radiology 2006;239(1):1–2.
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
25. Kim BH, Lee JM, Lee YJ, et al. MR elastography for noninvasive assessment of hepatic fibrosis: experience from a tertiary center in Asia. J Magn Reson Imaging 2011; 34(5):1110–1116. 26. Chen J, Talwalkar JA, Yin M, Glaser KJ, Sanderson SO, Ehman RL. Early detection of nonalcoholic steatohepatitis in patients with nonalcoholic fatty liver disease by using MR elastography. Radiology 2011;259(3):749– 756. 27. Manduca A, Oliphant TE, Dresner MA, et al. Magnetic resonance elastography: noninvasive mapping of tissue elasticity. Med Image Anal 2001;5(4):237–254. 28. Yin M, Talwalkar JA, Glaser KJ, et al. Assessment of hepatic fibrosis with magnetic resonance elastography. Clin Gastroenterol Hepatol 2007;5(10):1207–1213, e2. 29. Pineda N, Sharma P, Xu Q, Hu X, Vos M, Martin DR. Measurement of hepatic lipid: high-speed T2-corrected multiecho acquisition at 1H MR spectroscopy—a rapid and accurate technique. Radiology 2009; 252(2):568–576. 30. Hu HH, Kim HW, Nayak KS, Goran MI. Comparison of fat-water MRI and singlevoxel MRS in the assessment of hepatic and pancreatic fat fractions in humans. Obesity (Silver Spring) 2010;18(4):841–847. 31. Elhamy R, Heba GH, Sirlin CB, et al. Agreement of 2-, 3-, 4-, 5- and 6-echo MRIPDFF with MRS-PDFF in 580 adults with known or suspected non-alcoholic fatty liver disease (NAFLD) [abstr]. In: Proceedings of the Twenty-Second Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2014; 0138. 32. Hines CD, Bley TA, Lindstrom MJ, Reeder SB. Repeatability of magnetic resonance elastography for quantification of hepatic stiffness. J Magn Reson Imaging 2010; 31(3):725–731. 33. Kleiner DE, Brunt EM, Van Natta M, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology 2005;41(6):1313–1321.
11
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
34. Bedossa P, Poynard T. An algorithm for the grading of activity in chronic hepatitis C. The METAVIR Cooperative Study Group. Hepatology 1996;24(2):289–293. 35. Park YN, Kim HG, Chon CY, et al. Histological grading and staging of chronic hepatitis standardized guideline proposed by the Korean Study Group for the Pathology of Digestive Diseases. Korean J Pathol 1999; 33(5):337–346. 36. Zou KH, Tuncali K, Silverman SG. Correlation and simple linear regression. Radiology 2003;227(3):617–622. 37. Pascher A, Sauer IM, Walter M, et al. Donor evaluation, donor risks, donor outcome, and donor quality of life in adult-to-adult living donor liver transplantation. Liver Transpl 2002;8(9):829–837. 38. Valentín-Gamazo C, Malagó M, Karliova M, et al. Experience after the evaluation of 700 potential donors for living donor liver transplantation in a single center. Liver Transpl 2004;10(9):1087–1096. 39. Kang BK, Yu ES, Lee SS, et al. Hepatic fat quantification: a prospective comparison of
12
magnetic resonance spectroscopy and analy sis methods for chemical-shift gradient echo magnetic resonance imaging with histologic assessment as the reference standard. Invest Radiol 2012;47(6):368–375. 40. Kühn JP, Hernando D, Muñoz del Rio A, et al. Effect of multipeak spectral modeling of fat for liver iron and fat quantification: correlation of biopsy with MR imaging results. Radiology 2012;265(1):133–142. 41. Huwart L, Sempoux C, Salameh N, et al. Liver fibrosis: noninvasive assessment with MR elastography versus aspartate aminotransferase-to-platelet ratio index. Radiology 2007;245(2):458–466. 42. Yin M, Chen J, Glaser KJ, Talwalkar JA, Ehman RL. Abdominal magnetic resonance elastography. Top Magn Reson Imaging 2009; 20(2):79–87. 43. Angele MK, Rentsch M, Hartl WH, et al. Effect of graft steatosis on liver function and organ survival after liver transplantation. Am J Surg 2008;195(2):214–220. 44. Yamamoto K, Takada Y, Fujimoto Y, et al. Nonalcoholic steatohepatitis in donors for
Yoon et al
living donor liver transplantation. Transplantation 2007;83(3):257–262. 45. Hwang S, Lee SG, Jang SJ, et al. The effect of donor weight reduction on hepatic steatosis for living donor liver transplantation. Liver Transpl 2004;10(6):721–725. 46. Nakamuta M, Morizono S, Soejima Y, et al. Short-term intensive treatment for donors with hepatic steatosis in living-donor liver transplantation. Transplantation 2005; 80(5):608–612. 47. Jin YJ, Kim KM, Hwang S, et al. Exercise and diet modification in non-obese non-alcoholic fatty liver disease: analysis of biopsies of living liver donors. J Gastroenterol Hepatol 2012;27(8):1341–1347. 48. Vizzutti F, Arena U, Marra F, Pinzani M. Elastography for the non-invasive assessment of liver disease: limitations and future developments. Gut 2009;58(2):157–160. 49. Wang Y, Ganger DR, Levitsky J, et al. Assessment of chronic hepatitis and fibrosis: comparison of MR elastography and diffusion-weighted imaging. AJR Am J Roentgenol 2011;196(3):553–561.
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
Original Research n Gastrointestinal
Combined Use of MR Fat Quantification and MR Elastography in Living Liver Donors: Can It Reduce the Need for Preoperative Liver Biopsy?1 Purpose:
To evaluate the diagnostic performance of magnetic resonance (MR) fat quantification and MR elastography for the assessment of hepatic steatosis and fibrosis in living liver donor candidates.
Materials and Methods:
This retrospective study was approved by the institutional review board, and the requirement of informed consent was waived. Donors who underwent MR fat quantification and MR elastography at 1.5 T, followed by liver biopsy, were chronologically grouped into test and validation groups. In the test group (n = 362), MR fat fraction and liver stiffness were compared among donors with normal parenchyma (n = 244), simple steatosis (n = 71), steatosis with inflammatory activity (n = 21), nonalcoholic steatohepatitis (n = 17), and fibrosis (n = 9). Diagnostic performance of the two techniques was assessed by using receiver operating characteristic curve analysis for the detection of substantial steatosis (macrovesicular fat 10%) or fibrosis (F1) and was tested in a validation group (n = 34).
Results:
In the test group, donors with steatosis showed significantly higher fat fraction than donors without steatosis (P , .0001), and donors with fibrosis and nonalcoholic steatohepatitis showed significantly higher liver stiffness values than donors without fibrosis (P , .0001). Areas under the curve were 0.93 (cutoff value . 5.8%) for MR fat quantification and 0.85 (cutoff value . 1.94 kPa) for MR elastography. By using those values, the combination of the two techniques could be used to detect substantial steatosis or fibrosis with 100% sensitivity (12 of 12 patients, 95% confidence interval: 73.4%, 100%) and 100% negative predictive value (15 of 15 patients, 95% confidence interval: 78.0%, 100%) in the validation group.
Conclusion:
A combination of MR fat quantification and MR elastography can provide sufficient sensitivity to detect substantial steatosis or fibrosis (F1) in liver donor candidates.
1
From the Departments of Radiology (J.H.Y., J.M.L., J.K.H., B.I.C.), Surgery (K.S.S., K.W.L., N.J.Y.), and Pathology (K.B.L.), Seoul National University Hospital, Seoul, South Korea; and Institute of Radiation Medicine, Seoul National University College of Medicine, 101 Daehak-ro, Jongno-gu, Seoul 110-744, South Korea (J.M.L., J.K.H., B.I.C.). Received April 15, 2014; revision requested June 5; revision received November 16; accepted December 16; final version accepted December 29. This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2013R1A1A2A10066037). Address correspondence to J.M.L. (e-mail: [email protected]).
Imaging
Jeong Hee Yoon, MD Jeong Min Lee, MD Kyung-Suk Suh, MD Kwan-Woong Lee, MD Nam-Joon Yi, MD Kyung Bun Lee, MD Joon Koo Han, MD Byung Ihn Choi, MD
RSNA, 2015
q
Online supplemental material is available for this article.
RSNA, 2015
q
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
1
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
L
iver transplantation is widely performed around the world as a curative treatment strategy for liver failure and for the early stages of hepatocellular carcinoma (1,2). However, owing to widespread organ shortages reported for pediatric recipients, the living donor liver transplantation technique was developed, which has since rapidly expanded to include adult donors (3,4). A prerequisite to living donor liver transplantation is the thorough investigation of living liver donors with regard to their hepatic vascular anatomy, biliary anatomy, and presence of steatosis, to avoid morbidity and mortality of living donors and to prevent graft failure (5,6). Indeed, studies have shown that in addition to laboratory and psychosocial issues, vascular or biliary anatomic variations, inadequate remnant liver volumes (,30%), and abnormal histologic findings, such
Advances in Knowledge nn Living liver donors with hepatic fibrosis, nonalcoholic steatohepatitis (NASH), and steatosis with inflammatory activity (liver stiffness [LS] values of 2.54 kPa 6 0.34, 2.17 kPa 6 0.38, and 2.12 kPa 6 0.38, respectively) showed higher LS values than donors with simple steatosis and normal livers (1.84 kPa 6 0.30 and 1.68 kPa 6 0.26, respectively) (P , .0001). nn Areas under the curve were 0.93 (cutoff value . 5.8%) for MR fat quantification for the detection of substantial hepatic steatosis (10%) and 0.85 (cutoff value . 1.94 kPa) for MR elastography for the detection of hepatic fibrosis (F1); by using those values, MR elastography and MR fat quantification could be used to detect the presence of clinically significant hepatic steatosis or fibrosis with 100% sensitivity (95% confidence interval [CI]: 73.4%, 100%) and 100% negative predictive value (95% CI: 78.0%, 100%) in an independent validation set. 2
as hepatic steatosis or hepatic fibrosis, were the main reasons for exclusion from donation in potential adult living liver transplantation donors (7–9). Recently, nonalcoholic fatty liver disease (NAFLD), including nonalcoholic steatohepatitis (NASH), has increasingly been recognized as the most common chronic liver disease worldwide, with reports of an increasing prevalence of NASH in the United States (10). This increasing prevalence of NAFLD in the general population poses a risk for organ donation, since allograft steatosis has been reported to be associated with nonfunctioning grafts (11,12), and the first case of living liver transplantation donor death was reported in Japan in a donor with NASH (5). Computed tomography (CT) and magnetic resonance (MR) imaging have shown accuracy comparable to that of diagnostic angiography or conventional endoscopic cholangiography for the depiction of liver anatomy (13,14). However, the ability to diagnose steatosis, steatohepatitis, inflammation, or fibrosis by using noninvasive imaging tools remains a point of controversy (12,15). According to a previous study performed in a single transplantation center, the prevalence of NAFLD diagnosed with biopsy was 51.4% among 589 potential living liver donors (16). However, ultrasonography (US) and CT
Implications for Patient Care nn Preoperative assessment of the presence of hepatic fibrosis (F1), including NASH or substantial (10%) macrovesicular hepatic steatosis in living liver donors, could be achieved noninvasively by using MR fat quantification and MR elastography. nn Combined use of MR elastography and MR fat quantification could potentially reduce the clinical demand for routine liver biopsies in living liver donor candidates by providing information regarding the risk of substantial (10%) macrovesicular hepatic steatosis or hepatic fibrosis (F1).
Yoon et al
have shown limitations in the detection of clinically significant hepatic steatosis or NASH in liver donors (16–18); as a result, preoperative liver biopsy continues to be regarded as an essential part of living liver donor work-ups in many centers (17,19,20). On the other hand, there are also many centers where only selected potential donors with potential clinical risk factors, such as high body mass index (BMI), abnormal liver enzyme levels, or abnormal imaging findings at US or CT, undergo preoperative liver biopsy (12,21) because it is an invasive procedure with the problem of high interobserver variation (12,22). However, until now, there has been no clear consensus regarding the role of noninvasive tools in identifying living liver donor candidates at risk for poor recipient outcome while maximizing the donor pool (21). With recent advances in MR technology, studies have demonstrated promising results of MR spectroscopy Published online before print 10.1148/radiol.15140908 Content codes: Radiology 2015; 000:1–12 Abbreviations: ALP = alkaline phosphatase ALT = alanine transaminase APRI = AST-to-platelet ratio index AST = aspartate transaminase AUC = area under the receiver operating characteristic curve BMI = body mass index CI = confidence interval FF = fat fraction gGT = g-glutamyl transpeptidase LS = liver stiffness NAFLD = nonalcoholic fatty liver disease NASH = nonalcoholic steatohepatitis ROI = region of interest Author contributions: Guarantor of integrity of entire study, J.M.L.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; agrees to ensure any questions related to the work are appropriately resolved, all authors; literature research, J.H.Y., J.M.L., K.S.S., K.B.L., J.K.H.; clinical studies, J.H.Y., J.M.L., K.S.S., K.W.L., N.J.Y., K.B.L., J.K.H.; statistical analysis, J.H.Y.; and manuscript editing, J.H.Y., J.M.L., K.S.S., K.W.L., N.J.Y., K.B.L., B.I.C. Conflicts of interest are listed at the end of this article.
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
and the proton density fat fraction (FF) technique in quantifying hepatic steatosis (23,24). It has also been shown that MR elastography can be promising for the staging of hepatic fibrosis and the detection of NASH in patients with NAFLD (25,26). Considering that MR imaging is frequently used as a noninvasive imaging test for the evaluation of vascular and biliary anatomy for living liver donor candidates prior to liver transplantation without radiation-related biologic hazards, we postulated that the combined use of MR fat quantification and MR elastography may be able to demonstrate both hepatic steatosis and fibrosis in living liver donor candidates, thereby potentially reducing the necessity of liver biopsy when selecting ideal donor candidates. The purpose of this study, therefore, was to evaluate the diagnostic performance of MR fat quantification and MR elastography for the assessment of macrovesicular hepatic steatosis and fibrosis in living liver donor candidates.
Materials and Methods Study Population This retrospective study was approved by the institutional review board of our hospital, and the requirement for informed consent was waived. From January 2010 to September 2014, 404 living liver donor candidates who agreed to donate liver tissue and passed psychological evaluation tests underwent liver resections for donation (n = 348) or liver parenchymal biopsies (n = 56) after undergoing both MR fat quantification and MR elastography. Among them, eight donors were excluded because no tissue specimens were obtained owing to emergency liver transplantation (n = 5) or registration for exercise and diet modification programs was required to reduce steatosis after MR examinations (n = 3). Finally, 396 living liver donors were included in this study. These donors were divided into two groups, depending on the date of the MR examination for two independent investigations. The test group included 362 living liver
Yoon et al
Figure 1
Figure 1: Flow diagram of the study population. LB = liver biopsy, LT = liver transplantation, M:F = ratio of male to female patients.
donors (234 men and 128 women; mean age 6 standard deviation, 31.7 years 6 10.4 in men and 36.5 years 6 11.0 in women) who underwent MR examinations between January 2010 and November 2013. The validation group included 34 donors (26 men and eight women; mean age, 35.3 years 6 10.4 in men and 36.8 years 6 9.4 in women) who underwent MR examination between December 2013 and September 2014 (Fig 1). All donors denied having moderate or heavy alcohol consumption for 6 months prior to liver transplantation and did not show positive viral markers of hepatitis B or C in preoperative tests. The median interval between MR imaging and surgery or biopsy was 10 days (range, 0–66 days; 25th percentile, 4 days; 75th percentile, 18 days). BMI was also recorded in all donors (mean BMI, 23.4 kg/m2 6 3.1 in the test group and 23.5 kg/ m2 6 3.2 in the validation group). The following laboratory findings were also recorded: albumin level, total bilirubin
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
level, alkaline phosphatase (ALP), aspartate transaminase (AST), alanine transaminase (ALT), g-glutamyl transpeptidase (gGT), and platelet count. In addition, the AST-ALT ratio and ASTto-platelet ratio index (APRI) were calculated by using 40 IU/L (0.67 mkat/L) of AST as the upper normal limit.
MR Examination MR fat quantification technique.—Living liver donor candidates were asked to fast for at least 8 hours prior to MR imaging. Fat quantification was performed by using either the multiecho chemical-shift imaging sequence (n = 313, IDEAL-IQ; GE Healthcare, Waukesha, Wis) on a 1.5-T MR unit (Signa HDxt; GE Healthcare) or T2-corrected multiecho single-voxel hydrogen 1 (1H) MR spectroscopy (n = 83, Siemens Healthcare, Erlangen, Germany) on a 3.0-T MR unit (Magnetom Verio; Siemens Healthcare). MR elastography.—All MR elastography examinations were performed 3
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
on commercially available MR elastography hardware (MR Touch; GE Healthcare) with a 1.5-T imager (Signa HDxt; GE Healthcare) by using a twodimensional MR elastography protocol, which has been described previously, and the direct inversion algorithm (27). The two-dimensional MR elastography protocol used in our study was similar to the one described previously in the literature (28). In brief, imaging was performed in the supine position, with an acoustic pressureactivated passive driver placed against the body wall adjacent to the liver. The passive driver was connected to an active acoustic driver system located outside of the imaging room via a polyvinylchloride tube. Low-frequency longitudinal mechanical waves of 60 Hz were transmitted into the liver by a passive driver, as synchronized with the imaging sequence. Details of MR elastography acquisition parameters are provided in Appendix E1 (online).
Image Analysis Measurements of hepatic FF.—One attending radiologist (J.H.Y., with 8 years of clinical experience in MR image interpretation) who was blinded to the histologic results measured the hepatic FF on FF maps, which were generated automatically. Circular regions of interest (ROIs) approximately 300 mm2 6 38.4 in size (range, 250–400 mm2) were placed in the right liver lobe three times, and the mean value of the three measurements was used. If large hepatic vessels or focal liver lesions were seen on FF maps, vessels and lesions were carefully avoided. With MR spectroscopy, the automatically calculated hepatic FF (percentage) was used and displayed as a percentage in Digital Imaging and Communications in Medicine format (29). A detailed MR fat quantification protocol is given in Appendix E1 (online). We did not divide FF results from the two fat quantification techniques according to the literature (24,30,31). Measurement of the liver stiffness value.—As for MR elastography, liver stiffness (LS) values of the hepatic parenchyma were measured by placing 4
four ROIs on the elastogram. ROIs (mean area, 3837.8 mm2 6 1989.5; range, 1018.5–6352.3 mm2) were placed by one attending radiologist (J.H.Y.) who was blinded to the histologic findings. All ROIs were drawn in the area indicated to have high confidence and good signal-to-noise ratio, with stiffness outliers excluded on confidence maps (32) first and then copied to the corresponding position on stiffness maps that provided stiffness values in kilopascals. After reconfirming whether the ROIs were adequately placed in the right liver lobe, LS values (in kilopascals) were calculated as the median value of multiple ROIs.
Histologic Analysis Surgical liver biopsy was performed in the right lobe of the liver in all liver donors during the procedure, and approximately 1 mm3 of liver parenchyma was excised from the right lobe of the liver. US-guided percutaneous liver biopsy was performed prior to the procedure in 56 donors who showed abnormal biochemical liver function tests, such as increase of serum aminotransferase levels and evidence of hepatic steatosis seen on US, CT, or MR images in the right lobe of the liver by using an 18-gauge automated gun (TSK Laboratory, Tochigi, Japan) twice. If the candidate underwent percutaneous liver biopsy prior to surgery, areas of macrovesicular steatosis assessed from the percutaneous liver biopsy specimen were used for histologic analysis. The degree of macrovesicular steatosis was defined as the percentage of hepatocytes that contained intracellular macrovesicular fat droplets at hematoxylin-eosin staining (12). Although potential donors with mild steatosis are commonly denied from live donation, depending on graft volume and donor age (12), in our center, at least 10% macrovesicular steatosis was regarded as “substantial macrovesicular steatosis,” representing a threshold for donor candidates who required further evaluations or dietary interventions to achieve weight reduction prior to donation. NASH and steatosis with inflammatory activity were diagnosed on the
Yoon et al
basis of NAFLD activity scores (33). Donors who had NAFLD activity scores of at least 5 received diagnoses of NASH, and donors with NAFLD activity scores of 3 or 4 were classified as having steatosis with inflammatory activity. Donors with NAFLD activity score of 1 or 2 received diagnoses of simple steatosis. Hepatic fibrosis was assessed according to standardized guidelines proposed by the Korean Study Group for the Pathology of Digestive Diseases, which is similar to the METAVIR scoring system, as follows: F0 = no fibrosis, F1 = portal fibrosis, F2 = periportal fibrosis, F3 = septal fibrosis, and F4 = cirrhosis (34,35). Details of histologic analysis are given in Appendix E1 (online).
Statistical Analysis Correlation analyses were performed between FF on MR images and macrovesicular fat in the specimen, between BMI and macrovesicular hepatic steatosis, and between hepatic fibrosis and LS values (36). Donors were grouped into five groups on the basis of histologic findings of the liver (normal liver, simple steatosis, steatosis with inflammatory activity, NASH, and hepatic fibrosis). FF, LS values, degrees of macrovesicular steatosis, clinical findings, and laboratory findings, including the AST-ALT ratio and APRI score, were compared among the five groups by using either analysis of variance or Kruskal-Wallis tests, as well as post hoc analysis after a normality test. In this study, clinically significant hepatic steatosis was defined as macrovesicular fat of at least 10% on the basis of previous studies (37,38), and presence of hepatic fibrosis was defined as fibrosis score of at least F1. Factors associated with steatosis or fibrosis were determined by using multiple regression analysis. In addition, diagnostic performance of MR fat quantification and MR elastography was assessed by using receiver operating characteristic analysis to identify clinically significant macrovesicular hepatic steatosis or the presence of hepatic fibrosis. Finally, sensitivity, specificity, positive predictive value, and negative predictive value,
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
Study Results of the Test Group Demographics of living liver donors with different histologic findings.—Living liver donors received histologic diagnoses as follows: normal parenchyma (n = 244), simple steatosis (n = 71), steatosis with inflammatory activity (n = 21), NASH (n = 17; F1, n = 13; F2, n = 4), and hepatic fibrosis (n = 9; F1, n = 7; F2, n = 2) (33). Detailed demographics are summarized in Table 1. There were no significant differences in age among the groups, except between donors with normal liver parenchyma and donors with simple steatosis (P = .033). BMI of donors with normal parenchyma (22.8 kg/m2 6 2.9) was significantly lower than that in donors with NAFLD (steatosis with inflammatory activity and NASH, P , .0001 and P , .013, respectively), except in donors with simple steatosis (P = .051). None of the laboratory findings were found to be significantly different among donors with different histologic findings (Table 1). At pathologic examination, macrovesicular FF was shown to be higher in donors with NAFLD than in donors with normal parenchyma (P , .0001) or hepatic fibrosis (simple steatosis vs fibrosis, P , .0001; steatosis with inflammation vs fibrosis, P , .0001; and NASH vs fibrosis, P = .007). However, macrovesicular FF did not show significant differences between donors with NASH, steatosis with inflammatory activity, and simple steatosis (P = .066–.761). including 95% confidence interval (CI), were obtained in a validation set to predict either the presence of clinically significant hepatic steatosis or the presence of hepatic fibrosis by using cutoff values of MR fat quantification and MR elastography obtained in the test group. Details of the statistical analysis are provided in Appendix E1 (online). All statistical analyses were performed by using commercially available software (IBM SPSS, version 21, IBM, Armonk, NY; and MedCalc, version 12, MedCalc Software, Mariakerke, Belgium).
Results
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
156 88 32.2 6 10.6 (18–61) 22.8 6 2.9 (16.0–32.3) 232.6 6 49.9 (138.8–347.2) 4.33 6 0.36 (3.7–5.1) 0.82 6 0.48 (0.3–2.25) 64.85 6 20.29 (28–151) 22.68 6 14.36 (12–122) 23.03 6 18.70 (6–96) 26.84 6 21.62 (7–134) 1.18 6 0.63 (0.39–4.0) 0.25 6 0.21 (0.21–2.36) 3.34 v 1.83 (0.6–9.0) 1.69 6 1.51 (0.0–4.0) 1.68 6 0.26 (1.0–2.31)
Patient sex No. of men No. of women Age (y) BMI (kg/m2) Platelet count (3109/L) Albumin level (g/dL) Total bilirubin level (mg/dL) ALP (IU/L) AST (IU/L) ALT (IU/L) gGT (IU/L) AST-ALT ratio APRI score FF on MR images (%) Macrovesicular fat (%) LS value (kPa) 48 23 36.5 6 11.6 (18–61) 23.9 6 3.0 (16.7–32.9) 239.2 6 49.0 (172.4–379.8) 4.33 6 0.36 (3.7–5.1) 0.96 6 1.07 (0.3–2.7) 62.04 6 18.00 (30–182) 26.45 6 27.37 (10–144) 24.64 6 29.21 (10–71.5) 30.33 6 22.96 (9–89) 1.39 6 1.01 (0.42–7.0) 0.29 6 0.36 (0.13–0.93) 6.65 6 4.70 (2.0–23.9) 7.18 6 6.31 (5.0–40.0) 1.84 6 0.30 (1.32–2.58)
Simple Steatosis (n = 71) 16 5 34.2 6 7.7 (22–53) 26.1 6 3.6 (18.9–34.6) 249.2 6 64.4 (132.0–373.0) 4.41 6 0.49 (3.7–4.9) 1.06 6 0.80 (0.3–1.5) 71.43 6 21.09 (30–96) 25.57 6 17.47 (15–84) 27.57 6 22.36 (12–89.0) 36.76 6 30.86 (10–66) 1.11 6 0.47 (0.44–2.1) 0.24 6 0.23 (0.16–1.25) 10.90 6 4.50 (3.0–18.5) 13.12 6 9.06 (8.9–40.0) 2.12 6 0.38 (1.52–3.07)
Steatosis with Inflammatory Activity (n = 21)*
5 4 33.9 6 12.5 (20–52) 24.8 6 3.9 (19.1–29.0) 243.4 6 25.6 (158.0–339.0) 4.52 6 0.32 (3.7–4.9) 1.02 6 0.51 (0.3–2.5) 77.89 6 32.21 (30–77) 17.67 6 3.94 (16–94) 13.89 6 6.85 (7–101) 19.11 6 10.07 (10–122) 1.51 6 0.78 (0.83–3.0) 0.18 6 0.03 (0.24–1.27) 4.02 6 2.50 (1.0–7.7) 3.00 6 3.20 (0.0–8.0) 2.54 6 0.34 (1.96–3.00)
38.9 6 9.6 (20–52) 25.2 6 2.6 (20.7–29.9) 241.9 6 53.9 (180.0–310.0) 4.39 6 0.27 (3.4–4.9) 0.81 6 0.38 (0.4–1.6) 70.00 6 19.56 (46–119) 29.94 6 29.97 (13–46) 26.76 6 28.23 (11–46) 23.63 6 14.37 (10–134) 1.28 6 0.44 (0.61–1.95) 0.35 6 0.43 (0.20–1.50) 10.70 6 5.67 (1.1–23.0) 15.00 6 10.10 (7.0–40.0) 2.17 6 0.38 (1.47–2.80)
Fibrosis (n = 9)
9 8
NASH (n = 17)
.009‡ ,.001 .559 .494 .303 .086 .260 .527 .207 .162 .372 ,.001 ,.001 ,.001
…
P Value†
Differences among the five groups.
Only normal parenchyma and simple steatosis groups showed significant differences (P = .033).
†
‡
* Nonalcoholic fatty liver disease score 4.
Note.—Unless indicated otherwise, values are means 6 standard deviations, with ranges in parentheses. The upper normal limit of AST is 40 IU/L (0.67 mkat/L) at our institution. To convert milligrams per deciliter to micromoles per liter for total bilirubin level, multiply by 17.104. To convert international units per liter to microkatals per liter for ALP, AST, ALT, and gGT, multiply by 0.0167.
Normal Parenchyma (n = 244)
Parameter
FF on MR Images, Macrovesicular Fat, and LS Values at MR Elastography in Specimens from Living Liver Donors with Different Histologic Findings in the Test Group
Table 1
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors Yoon et al
5
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
Yoon et al
Figure 2
Figure 2: Box and whisker plots of (a) FF on MR images and (b) LS values at MR elastography for the five different histologic findings of the liver. Normal liver parenchyma and hepatic fibrosis had the lowest fat measurement, followed by simple steatosis (a). Steatosis with inflammatory activity and NASH showed the highest FF (a). Normal liver showed the lowest LS value, followed by simple steatosis. Steatosis with inflammatory activity, NASH, and hepatic fibrosis showed significantly higher LS values (b).
Hepatic FF values of living liver donor groups measured with MR techniques.—FF measured with MR techniques in donors with NAFLD was significantly higher than that in donors with normal liver parenchyma or donors with fibrosis (P , .001). Among donors with NAFLD, FF on MR images was significantly higher in donors with NASH (10.70% 6 5.67) and steatosis with inflammatory activity (10.90% 6 4.50) (P = .031) than in donors with simple steatosis (6.65% 6 4.70) (P = .006), similar to pathologic assessment (Figs 2a, 3–6). LS values of living liver donor groups measured with MR elastography.—Living liver donors with hepatic fibrosis showed significantly higher LS values (2.54 kPa 6 0.34) than those with normal parenchyma (1.68 kPa 6 0.26, P , .0001) and those with simple steatosis (1.84 kPa 6 0.30, P , .001). In addition, donors with NASH and steatosis with inflammatory activity showed significantly higher LS values (2.17 kPa 6 0.38 and 2.12 kPa 6 0.38, respectively; P = .004) than those with simple steatosis (P = .004). There were no significant 6
Figure 3
Figure 3: (a) FF map and (b) MR elastogram in a 51-year-old woman who donated the right lobe of her liver. At excisional liver biopsy, the liver demonstrated 2% macrovesicular fat and no evidence of hepatic fibrosis. The FF map (a) showed that the FF was 2.1%, and the LS value was 1.3 kPa on the elastogram (b), suggesting normal liver parenchyma without substantial fat infiltration.
differences in LS values among donors with NASH, steatosis with inflammatory activity (P = .812), and hepatic fibrosis (P = .204) (Figs 2b, 3–6).
Correlation between MR Values and Histologic Findings Correlation between FF on MR images, BMI, and fat in the specimen.— FF on MR images showed moderate
correlation with macrovesicular fat in the specimen (r = 0.77, 95% CI: 0.73, 0.81; P , .001). Only a weak correlation was found between BMI and macrovesicular fat (r = 0.27, 95% CI: 0.17, 0.36; P , .0001). Correlation between LS at MR elastography, fibrosis, and inflammatory activity.—LS values showed a significant correlation with hepatic fibrosis (r =
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
0.40, 95% CI: 0.31, 0.48; P , .0001). LS values also showed a significant correlation with inflammatory activity (r = 0.394, 95% CI: 0.30, 0.48; P , .0001).
Diagnostic Performance of MR Fat Quantification and MR Elastography for the Detection of Substantial Hepatic Steatosis or Presence of Hepatic Fibrosis (F1) in Comparison with Clinical Data For the detection of substantial hepatic steatosis (macrovesicular fat 10%), the MR fat quantification technique showed a higher area under the receiver operating characteristic curve (AUC) than did BMI (0.93 vs 0.69, P , .0001) and higher sensitivity (90.2% [46 of 51 patients] vs 68.6% [35 of 51 patients], P = .0039; Table 2) by using a cutoff value of 5.8%. The AUC of LS at MR elastography for detecting the presence of hepatic fibrosis (F1) was 0.85 (95% CI: 0.81, 0.89), showing 84.6% sensitivity (22 of 26 patients) by using a cutoff value of 1.94 kPa (Table 2). According to multiple regression analyses, FF, LS value, and gGT were associated with macrovesicular fat (R2 = 0.599, P , .0001), and LS value, FF, and age were associated with fibrosis stage (R2 = 0.350, P , .0001). Only FF and LS values were significantly associated with the presence of hepatic fibrosis (F1) or substantial hepatic steatosis (10%) at multiple logistic regression analysis (AUC, 0.935; 95% CI: 0.905, 0.958), providing 96.9% sensitivity (62 of 64 patients, 95% CI: 89.1, 99.5) and 99.0% negative predictive value (204 of 206 patients, 95% CI: 96.5, 99.9; Table 2). By using previously obtained cutoff values (fat on MR images . 5.8%; LS value . 1.94 kPa), the combination of MR fat quantification and MR elastography showed comparable results with the model: 96.9% sensitivity (62 of 64 patients, 95% CI: 89.1, 99.5) and 99.0% negative predictive value (208 of 210 patients, 95% CI: 96.6, 99.9; Table 2). Study Results of the Validation Group In the validation set (n = 34), FF showed significant correlation with the histologic fat amount (r = 0.91, P , .0001, 95%
Yoon et al
Figure 4
Figure 4: (a) FF map and (b) MR elastogram in a 42-year-old woman who underwent percutaneous liver biopsy for preoperative work-up. On the FF map, the liver showed clinically significant fatty infiltration (14.6%) (a), and the LS value was 1.7 kPa at MR elastography (b). Macrovesicular FF was 15% in the specimen, but neither lobular activity nor other evidence of NASH was found.
Figure 5
Figure 5: (a) FF map and (b) MR elastogram in a 31-year-old man who underwent preoperative percutaneous liver biopsy. On the FF map, FF was estimated to be 12.8% (a), and LS was 2.8 kPa at MR elastography (b). After liver biopsy, the patient received a diagnosis of NASH stage 1c.
Figure 6
Figure 6: (a) FF map and (b) MR elastogram in a 21-year-old man who had no history of liver disease. On the FF map, estimated fat fraction was 3% (a), but LS at MR elastography was 2.6 kPa (b). At liver biopsy, the patient received a diagnosis of early-stage hepatic fibrosis (F1, mild activity) without substantial fat infiltration (macrovesicular fat, 1%; microvesicular fat, 1%).
CI: 0.83, 0.96), and LS values showed significant correlation with fibrosis (r = 0.73, P , .0001, 95% CI: 0.51, 0.85).
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
FF, LS values, and other parameters are summarized in Table 3. By using the cutoff values derived from test data set (fat 7
8
†
Note.—Numbers in parentheses are the data used to calculate percentages. Numbers in brackets are 95% CIs.
Findings were considered positive when a donor showed either more than 5.98% fat with the MR fat quantification tool or stiffness of more than 1.94 kPa at MR elastography.
Discussion
* Classification cutoff value was set as 0.051 (when the probability that the patient had either fibrosis or substantial steatosis was more than 5%).
99.0 (204/206) [96.5, 99.9] 99.0 (208/210) [96.6, 99.9] 39.7 (62/156) [32.0, 47.9] 40.8 (62/152) [32.9, 49.1] 96.9 (62/64) [89.1, 99.5] 96.9 (62/64) [89.1, 99.5] … … 0.935 [0.91, 0.96] …
Yoon et al
on MR images . 5.8%; LS value . 1.94 kPa), a combination of the MR fat quantification technique and MR elastography showed 100% sensitivity (12 of 12 patients, 95% CI: 73.4%, 100%), 68.2% specificity (15 of 22 patients, 95% CI: 45.1%, 86.1%), 63.2% positive predictive value (12 of 19 patients, 95% CI: 19.8%, 53.5%), and 100% negative predictive value (15 of 15 patients, 95% CI: 78.0%, 100%).
68.5 (204/298) [62.9, 73.7] 69.8 (208/298) [64.2, 74.8]
98.5 (258/262) [96.1, 99.6] 22.0 (22/100) [14.3, 31.4] .1.94 0.85 [0.81, 0.89]
84.6 (22/26) [65.1, 95.5]
76.8 (258/336) [71.9, 81.2]
98.1 (261/266) [95.7, 99.4] 93.0 (214/230) [89.0, 96.0] 47.9 (46/96) [37.6, 58.4] 26.5 (35/132) [19.2, 34.9] .5.8 .24.09
Detection of substantial hepatic steatosis (10%) FF on MR images BMI Detection of the presence of hepatic fibrosis (F1) LS value at MR elastography Detection of a clinically significant pathologic condition (either substantial steatosis or presence of fibrosis) Multiple logistic regression model* Use of previously obtained two cutoff values†
0.93 [0.90, 0.95] 0.69 [0.64, 0.74]
90.2 (46/51) [78.6, 96.7] 68.6 (35/51) [54.0, 80.5]
83.9 (261/311) [79.4, 87.8] 68.8 (214/311) [63.3, 73.9]
Negative Predictive Value (%) Positive Predictive Value (%) Specificity (%) Sensitivity (%) Cutoff Value AUC Parameter
Diagnostic Performance of FF with the MR Fat Quantification Technique, BMI, and LS Value at MR Elastography for Detection of Either Substantial Hepatic Steatosis or Presence of Hepatic Fibrosis in the Test Group
Table 2
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
Our study demonstrated that a combination of MR fat quantification and MR elastography was able to provide high sensitivity (96.9%, 62 of 64 patients) for the prediction of substantial hepatic steatosis or hepatic fibrosis (F1) in living liver donor candidates in the test group, and the validation data set showed 100% sensitivity (12 of 12 patients, 95% CI: 73.4%, 100%) and 100% negative predictive value (15 of 15 patients, 95% CI: 78.0%, 100%). In addition, both examinations showed high accuracy in the detection of clinically significant hepatic steatosis (AUC, 0.93), as well as the presence of hepatic fibrosis (AUC, 0.85). Our study results show good agreement with previous studies, which also demonstrated excellent accuracy in the estimation of hepatic FF by using multiecho chemicalshift imaging with multipeak fat spectrum modeling or T2-corrected multiecho single-voxel 1H MR spectroscopy (39,40), as well as significant correlation between LS values at MR elastography and the degree of hepatic fibrosis at pathologic examination (25,41). Considering that both methods are achieved within a breath hold and cover a larger area than percutaneous liver biopsy, as well as the high accuracy in the prediction of substantial hepatic steatosis, MR fat quantification and MR elastography may serve as a screening tool for the selection of donors without a risk of substantial hepatic steatosis and/or hepatic fibrosis. In addition, our study also showed that there were significant differences in LS values measured with MR elastography between donors with NASH and simple steatosis (P = .023). radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
This result is consistent with the results of prior studies, in which simple steatosis was also distinguished from NASH by using MR elastography (26,42). Given that it is important to differentiate donor candidates with simple steatosis from those with NASH, since there have been reports that sole hepatic steatosis is not a contradiction for living donor liver transplantation (43), as well as the report of mortality in a donor with NASH (44), our study results suggest that LS measurements obtained with MR elastography could provide substantial clinical value in the selection of living liver donors by showing significant differences in LS values between donors with NASH and simple steatosis (P = .023). Therefore, we believe that MR elastography may be able to provide much value in the evaluation of liver donor candidates, in addition to MR fat quantification techniques. In living donor liver transplantation, maximizing donor safety is of paramount importance (12,44). To guarantee donor safety, routine liver biopsies have been recommended in previous studies (17,20). However, this policy provides a paradox: To maximize donor safety, an invasive preoperative examination, with the possibility of morbidity and mortality, should be performed (20). Indeed, concerns of possible morbidity and mortality keep personnel in many liver transplantation programs— up to 26%—from performing routine liver biopsies. Instead, they resort to an inconsistent selection criterion that may cause postoperative complications (21). Many transplantation centers have reported unexpected nonsteatotic hepatopathy demonstrated at liver biopsy in donor candidates (17,20) to be one of the reasons for supporting routine liver biopsies in donor candidates. However, our study results showed that the combined use of MR fat quantification techniques and MR elastography provided high sensitivity for the detection of steatotic hepatopathy, as well as hepatic fibrosis without steatosis. Furthermore, we found that MR elastography and MR fat quantification techniques provided better performance than clinical and laboratory findings,
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
15 7 34.7 6 10.3 (20–58) 22.4 6 2.3 (19.6–27.3) 238 6 46.3 (172–362) 4.6 6 0.2 (4.2–5.0) 1.0 6 0.6 (0.4–2.7) 55.5 6 14 (36.2–91.2) 22.5 6 12.3 (14.0–69.4) 25.2 6 26.4 (11.1–134.8) 30.6 6 31.7 (11.0–137.2) 1.03 6 0.25 (0.52–1.46) 0.28 6 0.33 (0.11–1.65) 5.19 6 1.40 (3.72–8.46) 2.59 6 1.65 (0.0–4.0) 1.76 6 0.19 (1.66–1.91)
Patient sex No. of men No. of women Age (y) BMI (kg/m2) Platelet count (3109/L) Albumin level (g/dL) Total bilirubin level (mg/dL) ALP (IU/L) AST (IU/L) ALT (IU/L) gGT (IU/L) AST-ALT ratio APRI score FF on MR images (%) Macrovesicular fat (%) LS value (kPa) (30–41) (21.6–22.2) (249–400) (4.6–4.9) (0.7–1.0) (43.0–50.0) (25–32) (40–50) (27–94) (0.63–0.64) (0.31–0.40) (8.45–17.64) (10.0–35.0) (1.66–1.91)
2 0
Simple Steatosis (n = 2) 1 0 38 27.1 260 4.6 0.5 78 35 83 65 0.42 0.8 8.5 10 2.17
Steatosis with Inflammatory Activity* (n = 1)
31.8 6 5.5 (23–37) 25.3 6 4.4 (19.6–30.8) 257.3 6 53.7 (213–352) 4.7 6 0.1 (4.5–4.8) 1.1 6 0.9 (0.4–2.7) 47.8 6 8.6 (32.0–55.0) 20 6 4.4 (13.0–26.0) 29.3 6 16.5 (14.0–61.0) 48.4 6 36.2 (24.0–108.0) 0.77 6 0.18 (0.43–0.93) 0.29 6 0.18 (0.16–0.63) 12.7 6 5.84 (8.90–24.48) 17.83 6 14.77 (7.0–45.0) 2.18 6 0.40 (1.92–2.95)
5 1
NASH (n = 6)
50 6 8.5 (42–59) 27.5 6 2.9 (24.3–29.9) 235 6 15.1 (221–251) 4.5 6 0.3 (4.3–4.9) 1.3 6 0.6 (0.9–2.0) 62.3 6 11.4 (53.0–75.0) 23.7 6 2.3 (21.0–25.0) 24.3 6 7.7 (18.0–33.0) 45 6 13.2 (35.0–60.0) 1.02 6 0.23 (0.76–1.17) 0.26 6 0.09 (0.18–0.35) 5.45 6 2.34 (2.89–7.47) 4.0 6 3.60 (1.0–8.0) 2.29 6 0.35 (2.00–2.68)
3 0
Fibrosis (n = 3)
.115 .017 .220 .846 .746 .165 .680 .177 .499 .011 .540 ,.0001 ,.0001 .0001
P Value†
†
Difference among the five groups.
* NAFLD score 4.
Note.—Values are means 6 standard deviations, with ranges in parentheses. A P value less than .05 indicates a statistically significant difference. Pairwise comparison could not be performed, because one group (steatosis with inflammatory activity) included only one donor. The upper normal limit of AST is 40 IU/L (0.67 mkat/L) at our institution. To convert milligrams per deciliter to micromoles per liter for total bilirubin level, multiply by 17.104. To convert international units per liter to microkatals per liter for ALP, AST, ALT, and gGT, multiply by 0.0167.
Normal Parenchyma (n = 22)
Parameter
FF on MR Images, Macrovesicular Fat, and LS Values at MR Elastography in Specimens from Living Liver Donors with Different Histologic Findings in the Validation Group
Table 3
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors Yoon et al
9
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
such as BMI or increased aminotransferase levels, in the detection of substantial hepatic steatosis, as well as the presence of hepatic fibrosis. These results are in good agreement with those of a previous study, which showed that while there is a significant correlation between BMI and degree of hepatic steatosis, 74% of overweight donors (BMI . 25) had 0%–10% steatosis (17). Considering that MR imaging is noninvasive, has no radiation hazard, and can be performed in a short period of time, we believe it would be valuable to implement MR elastography and fat quantification techniques in preoperative MR imaging protocols for living liver donor work-ups, as they can be used as a screening tool for the preoperative selection of candidates for liver biopsy among liver transplant donors. We anticipate that our study results can also serve as a reference for the selection of living liver donors in liver transplantation centers that do not perform routine liver biopsy. Implementation of MR elastography and MR quantification tools in preoperative liver donor evaluation strategies has other advantages over liver biopsy, which is currently the standard of reference for evaluation of NAFLD. To improve postoperative outcomes and solve the problem of organ shortages, intensive dietary and lifestyle management is currently performed in donors with hepatic steatosis (45–47). Although this approach can expand the pool of marginal living liver donors (45), follow-up should be performed by monitoring laboratory findings, weight changes, or repeated liver biopsies. Therefore, there is a great clinical need for a noninvasive assessment tool to be able to grade the improvement of NAFLD after exercise and diet modification. By using liver fat quantification and MR elastography, we can further reduce the number of repeated liver biopsies. Another advantage of MR imaging compared with liver biopsy is its all-in-one evaluation capability. Liver biopsies usually require short-term hospitalization, which increases medical costs, recipients’ economic burden, and donor inconvenience. MR imaging, on the other hand, 10
can provide a comprehensive, one-stop evaluation, including MR arteriography, MR cholangiography, MR fat quantification, and MR elastography, which can show anatomic and histologic information to surgeons, thereby shortening the preoperative work-up process and decreasing the inconvenience to potential donors who have to undergo variable imaging tests. Our study results of MR elastography are slightly different from the results of previous studies on the evaluation of NAFLD. First, in our study, MR elastography findings did not allow us to differentiate donors that had NASH from those that had steatosis with inflammatory activity, although NASH showed higher LS values than did simple steatosis. Our results are different from those of a previous study, which demonstrated the early detection of NASH in patients with nonalcoholic fatty liver disease by using MR elastography (26). This discrepancy can be explained by the differing study populations (living liver donor candidates vs patients) and study designs (prospective vs retrospective) between our studies. Although our study population included only living liver donor candidates who had no clinical suspicion of liver disease, the study population in the previous study was very heterogeneous and included patients clinically suspected of having liver diseases. Furthermore, in our study, the severity of NASH was mild and demonstrated F1 and F2 scores only. Considering that inflammation is a well-known confounding factor of hepatic fibrosis that increases LS values, MR elastography may have a limitation in the differentiation of NASH with early stage of fibrosis from steatosis with inflammatory activity (25,48). Second, the cutoff value of LS for detecting hepatic fibrosis (F1) from F0 or a normal liver was 1.94 kPa, which is lower than the 2.87 kPa (25) and 2.4 kPa (41) values reported in previous studies. The reason for the lower LS values is not clear, but we can assume that different study populations might have led to different cutoff values. In this study, young and healthy donors were included, whereas
Yoon et al
patients with chronic liver disease were included in those previous studies. Considering that inflammatory activity could increase LS values without fibrosis (49), cutoff values for detection of hepatic fibrosis from F0 in previous studies could be higher than cutoff values for detection of hepatic fibrosis in living liver donors with normal parenchyma in our study. This study has several limitations. First, since it was a retrospective, single-center study, there may have been selection bias. Second, the study population included a large portion of donors with normal livers but did not include many candidates who had more than 30% macrovesicular steatosis or clinically significant fibrosis (.F2). Therefore, this deviated study population may have affected the results of cutoff values in both examinations. However, considering that only potential living liver donors without significantly abnormal biochemical liver test findings or any potential illnesses are considered to undergo imaging workup in many transplantation centers, we believe that our study results are closer to the real clinical practice of donor selection in liver transplantation centers. Third, we performed MR elastography at 1.5 T by using the twodimensional technique, and it has been reported that different MR elastography techniques provided significantly different LS values. As a result, we cannot simply extrapolate the cutoff value to the results obtained with different methods. In conclusion, a combination of MR fat quantification and MR elastography can provide sufficient sensitivity to detect substantial (10%) macrovesicular hepatic steatosis or fibrosis (F1) in living liver donor candidates, potentially reducing the necessity of liver biopsy. Acknowledgment: We thank Chris Woo, BA, for his editorial assistance. Disclosures of Conflicts of Interest: J.H.Y. disclosed no relevant relationships. J.M.L. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: author received nonfinancial technical support from Siemens Healthcare; author received grants from Donseo Medical, CMS, Acuzen, Starmed, RF Medical, and Bayer
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
Healthcare; author received personal fees from Bayer Healthcare for lectures. Other relationships: disclosed no relevant relationships. K.S.S. disclosed no relevant relationships. K.W.L. disclosed no relevant relationships. N.J.Y. disclosed no relevant relationships. K.B.L. disclosed no relevant relationships. J.K.H. disclosed no relevant relationships. B.I.C. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: author received a grant from Samsung Electronics. Other relationships: disclosed no relevant relationships.
References 1. Mazzaferro V, Regalia E, Doci R, et al. Liver transplantation for the treatment of small hepatocellular carcinomas in patients with cirrhosis. N Engl J Med 1996;334(11):693–699. 2. Adam R, McMaster P, O’Grady JG, et al. Evolution of liver transplantation in Europe: report of the European Liver Transplant Registry. Liver Transpl 2003;9(12):1231– 1243. 3. Trotter JF. Selection of donors and recipients for living donor liver transplantation. Liver Transpl 2000;6(6 Suppl 2):S52–S58. 4. Kim KW, Lee J, Lee H, et al. Right lobe estimated blood-free weight for living donor liver transplantation: accuracy of automated blood-free CT volumetry—preliminary results. Radiology 2010;256(2):433–440. 5. Akabayashi A, Slingsby BT, Fujita M. The first donor death after living-related liver transplantation in Japan. Transplantation 2004;77(4):634. 6. Kiuchi T, Tanaka K, Ito T, et al. Small-forsize graft in living donor liver transplantation: how far should we go? Liver Transpl 2003;9(9):S29–S35. 7. Sharma A, Ashworth A, Behnke M, Cotterell A, Posner M, Fisher RA. Donor selection for adult-to-adult living donor liver transplantation: well begun is half done. Transplantation 2013;95(3):501–506. 8. Tsang LL, Chen CL, Huang TL, et al. Preoperative imaging evaluation of potential living liver donors: reasons for exclusion from donation in adult living donor liver transplantation. Transplant Proc 2008;40(8):2460– 2462. 9. Fan ST, Lo CM, Liu CL, Yong BH, Chan JK, Ng IO. Safety of donors in live donor liver transplantation using right lobe grafts. Arch Surg 2000;135(3):336–340. 10. Kemmer N, Neff GW, Franco E, et al. Nonalcoholic fatty liver disease epidemic and its implications for liver transplantation. Transplantation 2013;96(10):860–862.
Yoon et al
11. Said A. Non-alcoholic fatty liver disease and liver transplantation: outcomes and advances. World J Gastroenterol 2013;19(48): 9146–9155.
23. Tang A, Tan J, Sun M, et al. Nonalcoholic fatty liver disease: MR imaging of liver proton density fat fraction to assess hepatic steatosis. Radiology 2013;267(2):422–431.
12. McCormack L, Dutkowski P, El-Badry AM, Clavien PA. Liver transplantation using fatty livers: always feasible? J Hepatol 2011;54(5):1055–1062.
24. Reeder SB, Cruite I, Hamilton G, Sir lin CB. Quantitative assessment of liver fat with magnetic resonance imaging and spectroscopy. J Magn Reson Imaging 2011; 34(4):729–749.
13. Song GW, Lee SG, Hwang S, et al. Pre operative evaluation of biliary anatomy of donor in living donor liver transplantation by conventional nonenhanced magnetic resonance cholangiography. Transpl Int 2007;20(2):167–173. 14. Lee SS, Kim TK, Byun JH, et al. Hepatic arteries in potential donors for living related liver transplantation: evaluation with multidetector row CT angiography. Radiology 2003;227(2):391–399. 15. Schwenzer NF, Springer F, Schraml C, Stefan N, Machann J, Schick F. Non-invasive assessment and quantification of liver steatosis by ultrasound, computed tomography and magnetic resonance. J Hepatol 2009;51(3):433–445. 16. Lee JY, Kim KM, Lee SG, et al. Preva lence and risk factors of non-alcoholic fatty liver disease in potential living liver donors in Korea: a review of 589 consecutive liver biopsies in a single center. J Hepatol 2007;47(2):239–244. 17. Ryan CK, Johnson LA, Germin BI, Mar cos A. One hundred consecutive hepatic biopsies in the workup of living donors for right lobe liver transplantation. Liver Transpl 2002;8(12):1114–1122. 18. Park SH, Kim PN, Kim KW, et al. Mac rovesicular hepatic steatosis in living liver donors: use of CT for quantitative and qualitative assessment. Radiology 2006;239(1): 105–112. 19. Brandhagen D, Fidler J, Rosen C. Evaluation of the donor liver for living donor liver transplantation. Liver Transpl 2003;9(10 Suppl 2):S16–S28. 20. Nadalin S, Malagó M, Valentin-Gamazo C, et al. Preoperative donor liver biopsy for adult living donor liver transplantation: risks and benefits. Liver Transpl 2005;11(8):980–986. 21. Brown RS Jr, Russo MW, Lai M, et al. A survey of liver transplantation from living adult donors in the United States. N Engl J Med 2003;348(9):818–825. 22. Brancatelli G. Science to practice: Should biopsy be performed in potential liver donors when unenhanced CT shows an unacceptable degree of steatosis for transplantation? Radiology 2006;239(1):1–2.
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
25. Kim BH, Lee JM, Lee YJ, et al. MR elastography for noninvasive assessment of hepatic fibrosis: experience from a tertiary center in Asia. J Magn Reson Imaging 2011; 34(5):1110–1116. 26. Chen J, Talwalkar JA, Yin M, Glaser KJ, Sanderson SO, Ehman RL. Early detection of nonalcoholic steatohepatitis in patients with nonalcoholic fatty liver disease by using MR elastography. Radiology 2011;259(3):749– 756. 27. Manduca A, Oliphant TE, Dresner MA, et al. Magnetic resonance elastography: noninvasive mapping of tissue elasticity. Med Image Anal 2001;5(4):237–254. 28. Yin M, Talwalkar JA, Glaser KJ, et al. Assessment of hepatic fibrosis with magnetic resonance elastography. Clin Gastroenterol Hepatol 2007;5(10):1207–1213, e2. 29. Pineda N, Sharma P, Xu Q, Hu X, Vos M, Martin DR. Measurement of hepatic lipid: high-speed T2-corrected multiecho acquisition at 1H MR spectroscopy—a rapid and accurate technique. Radiology 2009; 252(2):568–576. 30. Hu HH, Kim HW, Nayak KS, Goran MI. Comparison of fat-water MRI and singlevoxel MRS in the assessment of hepatic and pancreatic fat fractions in humans. Obesity (Silver Spring) 2010;18(4):841–847. 31. Elhamy R, Heba GH, Sirlin CB, et al. Agreement of 2-, 3-, 4-, 5- and 6-echo MRIPDFF with MRS-PDFF in 580 adults with known or suspected non-alcoholic fatty liver disease (NAFLD) [abstr]. In: Proceedings of the Twenty-Second Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2014; 0138. 32. Hines CD, Bley TA, Lindstrom MJ, Reeder SB. Repeatability of magnetic resonance elastography for quantification of hepatic stiffness. J Magn Reson Imaging 2010; 31(3):725–731. 33. Kleiner DE, Brunt EM, Van Natta M, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology 2005;41(6):1313–1321.
11
GASTROINTESTINAL IMAGING: MR Fat Quantification and MR Elastography in Living Liver Donors
34. Bedossa P, Poynard T. An algorithm for the grading of activity in chronic hepatitis C. The METAVIR Cooperative Study Group. Hepatology 1996;24(2):289–293. 35. Park YN, Kim HG, Chon CY, et al. Histological grading and staging of chronic hepatitis standardized guideline proposed by the Korean Study Group for the Pathology of Digestive Diseases. Korean J Pathol 1999; 33(5):337–346. 36. Zou KH, Tuncali K, Silverman SG. Correlation and simple linear regression. Radiology 2003;227(3):617–622. 37. Pascher A, Sauer IM, Walter M, et al. Donor evaluation, donor risks, donor outcome, and donor quality of life in adult-to-adult living donor liver transplantation. Liver Transpl 2002;8(9):829–837. 38. Valentín-Gamazo C, Malagó M, Karliova M, et al. Experience after the evaluation of 700 potential donors for living donor liver transplantation in a single center. Liver Transpl 2004;10(9):1087–1096. 39. Kang BK, Yu ES, Lee SS, et al. Hepatic fat quantification: a prospective comparison of
12
magnetic resonance spectroscopy and analy sis methods for chemical-shift gradient echo magnetic resonance imaging with histologic assessment as the reference standard. Invest Radiol 2012;47(6):368–375. 40. Kühn JP, Hernando D, Muñoz del Rio A, et al. Effect of multipeak spectral modeling of fat for liver iron and fat quantification: correlation of biopsy with MR imaging results. Radiology 2012;265(1):133–142. 41. Huwart L, Sempoux C, Salameh N, et al. Liver fibrosis: noninvasive assessment with MR elastography versus aspartate aminotransferase-to-platelet ratio index. Radiology 2007;245(2):458–466. 42. Yin M, Chen J, Glaser KJ, Talwalkar JA, Ehman RL. Abdominal magnetic resonance elastography. Top Magn Reson Imaging 2009; 20(2):79–87. 43. Angele MK, Rentsch M, Hartl WH, et al. Effect of graft steatosis on liver function and organ survival after liver transplantation. Am J Surg 2008;195(2):214–220. 44. Yamamoto K, Takada Y, Fujimoto Y, et al. Nonalcoholic steatohepatitis in donors for
Yoon et al
living donor liver transplantation. Transplantation 2007;83(3):257–262. 45. Hwang S, Lee SG, Jang SJ, et al. The effect of donor weight reduction on hepatic steatosis for living donor liver transplantation. Liver Transpl 2004;10(6):721–725. 46. Nakamuta M, Morizono S, Soejima Y, et al. Short-term intensive treatment for donors with hepatic steatosis in living-donor liver transplantation. Transplantation 2005; 80(5):608–612. 47. Jin YJ, Kim KM, Hwang S, et al. Exercise and diet modification in non-obese non-alcoholic fatty liver disease: analysis of biopsies of living liver donors. J Gastroenterol Hepatol 2012;27(8):1341–1347. 48. Vizzutti F, Arena U, Marra F, Pinzani M. Elastography for the non-invasive assessment of liver disease: limitations and future developments. Gut 2009;58(2):157–160. 49. Wang Y, Ganger DR, Levitsky J, et al. Assessment of chronic hepatitis and fibrosis: comparison of MR elastography and diffusion-weighted imaging. AJR Am J Roentgenol 2011;196(3):553–561.
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
