Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print January 14, 2008, 10.1148/radiol.2463070077
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2463070077v1
246/3/926    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hagiwara, M.
Right arrow Articles by Taouli, B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hagiwara, M.
Right arrow Articles by Taouli, B.
(Radiology 2008;246:926-934.)
© RSNA, 2008

Technical Developments

Advanced Liver Fibrosis: Diagnosis with 3D Whole-Liver Perfusion MR Imaging—Initial Experience1

Mari Hagiwara, MD, Henry Rusinek, PhD, Vivian S. Lee, MD, PhD, Mariela Losada, MD, Michael A. Bannan, MD, Glenn A. Krinsky, MD 2, and Bachir Taouli, MD

1 From the Departments of Radiology (M.H., H.R., V.S.L., G.A.K., B.T.) and Pathology (M.L., M.A.B.), New York University Medical Center, 560 First Ave, New York, NY 10016. Received January 12, 2007; revision requested March 15; revision received May 15; accepted June 13; final version accepted September 7. Address correspondence to B.T. (e-mail: bachir.taouli{at}med.nyu.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE...
 References
 
Institutional review board approval and informed consent were obtained for this HIPAA–compliant study. The purpose of this study was to prospectively evaluate sensitivity and specificity of various estimated perfusion parameters at three-dimensional (3D) perfusion magnetic resonance (MR) imaging of the liver in the diagnosis of advanced liver fibrosis (stage ≥ 3), with histologic analysis, liver function tests, or MR imaging as the reference standard. Whole-liver 3D perfusion MR imaging was performed in 27 patients (17 men, 10 women; mean age, 55 years) after dynamic injection of 8–10 mL of gadopentetate dimeglumine. The following estimated perfusion parameters were measured with a dual-input single-compartment model: absolute arterial blood flow (Fa), absolute portal venous blood flow (Fp), absolute total liver blood flow (Ft) (Ft = Fa + Fp), arterial fraction (ART), portal venous fraction (PV), distribution volume (DV), and mean transit time (MTT) of gadopentetate dimeglumine. Patients were assigned to two groups (those with fibrosis stage ≤ 2 and those with fibrosis stage ≥ 3), and the nonparametric Mann-Whitney test was used to compare Fa, Fp, Ft, ART, PV, DV, and MTT between groups. Receiver operating characteristic curve analysis was used to assess the utility of perfusion estimates as predictors of advanced liver fibrosis. There were significant differences for all perfusion MR imaging–estimated parameters except Fp and Ft. There was an increase in Fa, ART, DV, and MTT and a decrease in PV in patients with advanced fibrosis compared with those without advanced fibrosis. DV had the best performance, with an area under the receiver operating characteristic curve of 0.824, a sensitivity of 76.9% (95% confidence interval: 46.2%, 94.7%), and a specificity of 78.5% (95% confidence interval: 49.2%, 95.1%) in the prediction of advanced fibrosis.

© RSNA, 2008


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE...
 References
 
Liver fibrosis is an important cause of morbidity, mortality, and increased health care costs in patients with chronic viral hepatitis (1). In these patients, the early detection of liver fibrosis and cirrhosis has important clinical implications in the determination of antiviral treatment options and patient prognosis (2). The diagnosis of liver fibrosis is usually based on histologic findings after liver biopsy (2–4). This procedure has inherent risks (5–8), and it is prone to interobserver variability and sampling errors (9–11). Morphologic criteria used to diagnose early and advanced cirrhosis with conventional magnetic resonance (MR) imaging have been described in several reports, most of which were written by the same investigative group (12–15). These criteria have a sensitivity of 68%–93% and a specificity of 77.4%–98%. Each value depends on the criterion used. However, none of these studies included patients with chronic hepatitis but without cirrhosis, and there are limited data on the use of MR imaging in the detection of advanced liver fibrosis and cirrhosis. For example, the results of a recent study (16) showed excellent (>90%) sensitivity and specificity in the diagnosis of advanced fibrosis with use of double contrast material–enhanced imaging; however, use of two contrast agents (in this case, gadodiamide and superparamagnetic iron oxide) increases the cost of the examination and is not always practical.

It has been shown that liver perfusion is altered in patients with chronic liver disease, and perfusion MR imaging has the potential to enable physicians to noninvasively detect and assess vascular alterations associated with collagen deposition in patients with chronic liver disease (17).

Three-dimensional (3D) perfusion MR imaging with coverage of the entire liver has the advantage of depicting perfusion changes in the liver on both a global basis and a regional basis. This could be useful in the diagnosis of fibrosis and the detection of interval changes during antiviral treatment monitoring. It could also be used to screen for hepatocellular carcinoma. Researchers from the same group have used single-section perfusion CT or MR imaging with limited liver coverage (18,19) in patients with liver disease. To our knowledge, however, the role of perfusion MR imaging in the diagnosis of liver fibrosis had not been examined.

Thus, the purpose of our study was to prospectively evaluate the sensitivity and specificity of various estimated perfusion parameters at 3D perfusion MR imaging of the liver in the diagnosis of advanced liver fibrosis (stage ≥ 3), with histologic analysis, liver function tests, or MR imaging as the reference standard.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE...
 References
 
Patients and Control Subjects
Twenty-eight consecutive patients (17 men, 11 women; mean age, 54 years) were prospectively enrolled in this study during a 12-month period from March 2004 to March 2005. Of these 28 patients, 20 (13 men, seven women; mean age, 55 years) had a known diagnosis of chronic liver disease. In these 20 patients, the underlying causes of disease were chronic viral hepatitis C (n = 16), viral hepatitis B (n = 1), nonalcoholic steatohepatitis (n = 1), and chronic autoimmune hepatitis (n = 2). These patients were recruited from the hepatology clinic during the initial examination of patients with newly diagnosed chronic hepatitis. This study was compliant with the Health Insurance Portability and Accountability Act. Our institutional review board approved this study, and written informed consent was obtained from all patients.

MR imaging was performed in eight patients (three men, five women; mean age, 52 years) with no history of liver disease and with normal liver function test results to evaluate benign liver lesions or nonhepatic disease. These eight patients were the control subjects. One control subject who had a large focal nodular hyperplasia (8 cm diameter) and a large hemangioma (9 cm diameter) was excluded. Two control subjects who had small (1 cm diameter) cysts or hemangiomas were not excluded; thus, seven control subjects (three men, four women; mean age, 55 years) were included in this study. The final cohort included 27 patients (16 men, 11 women; mean age, 55 years).

Two experienced hepatopathologists (M.L., M.A.B.; 8 and 12 years of postgraduate experience, respectively) working in consensus diagnosed liver fibrosis or cirrhosis by analyzing blind liver biopsy specimens (n = 12, specimens were obtained with a 20-gauge Menghini needle) or liver explant specimens (n = 4) and retrospectively staged the disease with the Batts-Ludwig classification system (20). This scoring system involves the use of a five-point scale for staging and grading. Staging refers to the degree of fibrosis: Stage 0 indicates no fibrosis; stage 1, portal fibrosis; stage 2, periportal fibrosis; stage 3, septal fibrosis; and stage 4, cirrhosis. Grading of activity refers to the degree of hepatocellular necroinflammatory activity: Grade 0 indicates no activity; grade 1, minimal activity; grade 2, mild activity; grade 3, moderate activity; and grade 4, severe activity.

The mean interval between liver biopsy or transplantation and MR imaging was 105 days (range, 1–480 days). Three patients who underwent liver transplantation accounted for the longest intervals. (The mean interval was 46 days if these three patients are excluded.) In these three patients, unequivocal diagnosis of liver cirrhosis was already established at MR imaging by using validated criteria (liver nodularity and signs of portal hypertension) (21). Four of the 20 patients, who did not undergo liver biopsy, had an unequivocal diagnosis of cirrhosis at MR imaging with use of the previously mentioned criteria (21). None of the patients had hepatocellular carcinoma.

3D Perfusion MR Imaging
MR imaging was performed by using a 1.5-T system with torso phased-array coils (Magnetom Symphony or Avanto; Siemens Medical Solutions, Erlangen, Germany). Whole-liver perfusion MR imaging was performed by using a 3D interpolated spoiled gradient-recalled-echo sequence in the coronal plane. Since portal venous flow is known to increase after a meal (22), the patients and control subjects were asked to fast for 6 hours before the study. The patients' arms were elevated to minimize aliasing artifacts. One acquisition was performed before contrast material injection, and the first contrast-enhanced acquisition started after injection of 8 or 10 mL (8 mL was used with the Symphony system, whereas 10 mL was used with the Avanto system) of gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ) followed by a 20-mL saline flush injected at a rate of 5 mL/sec with an MR-compatible power injector (Spectris; Medrad, Indianola, Pa). The following imaging parameters were used: 1.7–3.2/0.8 (repetition time msec/echo time msec), 9° flip angle, 128 x 256 matrix, 3.1 x 1.8-mm in-plane pixel size, 18 x 40-cm field of view, 18-cm slab thickness resulting in an interpolated 4-mm section thickness, and 720 Hz/pixel bandwidth. A parallel imaging technique (R factor of two with the Magnetom Symphony system and of three with the Avanto system) was performed with generalized autocalibrating partially parallel acquisition (GRAPPA; Siemens Medical Solutions). Between 36 and 40 coronal images (Fig 1) were acquired every 3.3–5.0 seconds for approximately 2 minutes. Patients were instructed to suspend respiration at end expiration to minimize misregistration artifacts during all 3D acquisitions. The first breath hold lasted approximately 25–30 seconds, and four subsequent breath holds lasted 20 seconds each, with 4 seconds of shallow respiration between each breath hold. Oxygen administration by means of a nasal cannula was offered routinely to patients before the examination to facilitate breath holding.


Figure 1
View larger version (113K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: Perfusion MR images of the liver in a 66-year-old female patient with chronic hepatitis C and stage 2 fibrosis obtained by using a coronal 3D interpolated spoiled gradient-recalled-echo sequence (1.7–3.2/0.8, 9° flip angle, generalized autocalibrating partially parallel acquisition acceleration factor of three) to cover the entire liver before and after injection of 10 mL of gadopentetate dimeglumine. Selected time points from 20 measurements are shown in chronologic order. The total acquisition time was approximately 120 seconds. (Each volume was acquired in 5.6 seconds.) Progressive opacification of the aorta and hepatic artery (short arrows), portal vein (long arrow), and liver parenchyma and hepatic vein (dashed arrow) was noted.

 
The rationale behind the use of 3D acquisition was to ensure that the entire liver was imaged and to account for regional perfusion differences in the liver. The rationale behind the use of coronal imaging was to minimize flow-related enhancement of the aorta (23) (as substantiated by imaging two subjects and a flow phantom in the coronal and transverse planes). Three-dimensional perfusion MR imaging was incorporated into the routine MR imaging protocol and added approximately 5 minutes to the total imaging time, which did not exceed 45 minutes. The patient was asked to remain on the table for 5 minutes after administration of the initial contrast material bolus (used for perfusion MR imaging) to allow the gadopentetate dimeglumine to wash out from the hepatic parenchyma before routine contrast material injection.

Image analysis.—The images were transferred to a standard personal computer running a locally developed software package for image segmentation and coregistration. Regions of interest (ROIs) were drawn manually on the main portal vein (at the level of the porta hepatis), the proximal abdominal aorta (at the level of the celiac axis), and four areas of the liver (right anterior, right posterior, left medial, and left lateral lobes) by an expert observer (M.H., with 1 year of experience in liver MR imaging) who was unaware of the histologic findings. Mean ROI size was 8 mm in the main portal vein, 10 mm in the abdominal aorta, and 20 mm in the liver parenchyma. Given the small size of the hepatic artery, ROIs were drawn in the proximal abdominal aorta (at the level of the celiac axis), which was used as a surrogate for the hepatic arterial blood supply. ROIs for the portal vein were drawn for each time frame to correct for spatial misregistration artifacts due to respiratory motion. The aorta and liver ROIs were drawn for one time frame and then transferred to the remaining time frames and manually corrected, if necessary. Large vascular structures or liver lesions were avoided.

Conversion of signal intensity to gadopentetate dimeglumine concentration and calculation of perfusion MR imaging parameters.—ROI measurements were normalized by subtracting the ROI value of the initially obtained unenhanced image from the ROI values of the subsequently obtained contrast-enhanced images and dividing this difference by the initial ROI value. Signal intensity curves were then obtained from the ROI data. A linear relationship between signal intensity and gadopentetate dimeglumine concentration was assumed to exist for the range of expected contrast agent concentrations in the liver (0.0–0.5 mmol/L) and blood (0–5 mmol/L). These concentrations are based on the results of our prior work, in which we measured gadopentetate dimeglumine concentration in vivo and in vitro (24). Thus, our conversion was based on the following approximation: c = [k(S – S0)]/S0 (25), where c is the concentration, S0 is the precontrast signal intensity, S is the postcontrast signal intensity, and k is the scaling constant (0.395 for liver and 0.201 for blood). These constants are based on the results of our prior phantom and human calibration study (24). Resulting time-activity curves (Fig 2) were fitted by using a dual-input single-compartment model that was validated previously with radiolabeled microspheres in rabbits (17). This model reflects the dual blood supply from the portal vein and hepatic artery received by the liver. The general equation for the dual-input kinetic model is as follows:

Formula
where Ca(t), Cp(t), and Cl(t) represent the concentration of contrast material in the aorta, portal vein, and liver, respectively; {delta} represents the transit time from the aorta ROI to the liver ROI; k1a represents the aortic inflow rate constant; k1p represents the portal venous inflow rate constant; k2 represents the outflow rate constant; t represents time; t' is the integration variable; and dt' is the differential. A fitting procedure was performed to determine values for k1a, k1p, k2, and {delta} by using locally developed software written in C++ language. Our program combines the conventional simplex method (26) with a four-dimensional grid of starting parameter values to enable us to determine the best least-squares fit of the model curve Cl(t) to the measured data. The distribution volume (DV) of gadopentetate dimeglumine through the liver compartment was calculated as follows: DV = [100(k1a + k1p)]/k2. Mean transit time (MTT) was calculated as follows: MTT = 1/k2. MTT is the average time it takes a gadopentetate dimeglumine molecule to traverse the liver from arterial or portal venous entry to venous exit. The integration of these constants over the volume of the liver shows absolute arterial liver blood flow (Fa) (Fa = 6.103 x k1a) (measured in milliliters per 100 g per minute); absolute portal liver blood flow (Fp) (Fp = 6.103 x k1p) (measured in milliliters per 100 g per minute), and absolute total liver blood flow (Ft) (Ft = Fa + Fp) (measured in milliliters per 100 g per minute), as well as the arterial fraction (ART) (ART = 100 x Fa/(Fa + Fp) (measured as a percentage) and the portal venous fraction (PV) (PV = 100 – ART) (measured as a percentage).


Figure 2
View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2: Graph shows the concentration of gadopentetate dimeglumine ([Gd]) in the blood (aorta and portal vein) and liver parenchyma versus time curves for the aorta, portal vein, liver parenchyma (liver enhancement), and interpolated best fit dual-input model in a control patient. The gadopentetate dimeglumine concentration scale for blood is different from that for liver parenchyma.

 
Statistical Analysis
Statistical software (SAS, version 9.0; SAS Institute, Cary, NC) was used for all statistical computations. Statistical analysis was based on the average signal intensity derived from four ROI locations used to calculate perfusion parameters, as well as on a single ROI measurement for the estimation of segmental variability of liver perfusion. The estimated perfusion parameters (Fa, Fp, Ft, ART, PV, DV, and MTT) were compared between patient groups defined by using fibrosis stage (stage 0 vs stages 1–3 vs stage 4, and stage ≤ 2 vs stage ≥ 3) and necroinflammatory grade (grade 0 vs grade ≥ 1) by using exact significance levels from a Mann-Whitney test. Receiver operating characteristic curve analyses were used to assess the utility of Fa, Fp, Ft, ART, PV, DV, and MTT as predictors of advanced fibrosis and cirrhosis (stage ≥ 3). For each candidate predictor variable, receiver operating characteristic curve analysis was used to estimate (a) the area under the receiver operating characteristic curve, (b) the criterion (ie, the cutoff value for the predictor variable) associated with the highest sensitivity and specificity, and (c) the sensitivity and specificity associated with the selected criterion. All reported P values were calculated with two-sided tests, they were not subjected to multiple comparison correction (no Bonferroni correction was applied), and they were declared significant when they were less than .05. Segmental variation of estimated liver perfusion parameters was expressed in terms of the coefficient of variation (calculated by multiplying the standard deviation by 100% and dividing by the mean) derived from ROI measurements in four liver locations (right posterior, right anterior, left medial, and left lateral lobes) within a subject. Coefficient of variation values reported for each perfusion estimate were compared between subjects (stage ≤ 2 group vs stage ≥ 3 group) with the Mann-Whitney test.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE...
 References
 
Histologic Findings
The distribution of fibrosis stage was as follows: stage 0, 10 patients; stage 1, two patients; stage 2, two patients; stage 3, four patients; and stage 4, nine patients. The distribution of necroinflammatory changes was as follows: grade 0, eight patients; grade 1, two patients; grade 2, seven patients; grade 3, six patients; and grade 4, zero patients. The seven control patients were presumed to have no fibrosis and no hepatocellular necroinflammatory activity (stage 0, grade 0). Four patients who did not undergo biopsy and had an MR diagnosis of cirrhosis were classified as having stage 4 fibrosis; however, the grade of necroinflammatory activity was unknown. Three patients with chronic hepatitis did not have liver fibrosis or inflammation at biopsy.

Perfusion MR Imaging Findings
There was a trend toward an increase in arterial perfusion (ART and Fa), DV, and MTT and a decrease in portal venous perfusion (PV and Fp) with an increase in the degree of liver fibrosis, with significant differences (a) between patients who did not have fibrosis (stage 0) and those who had stage 1–3 fibrosis for DV (P = .02) and MTT (P = .02) and (b) between patients who did not have fibrosis (stage 0) and those who had stage 4 fibrosis for Fa, ART, PV, MTT, and DV (P = .004–.04) (Table 1). When patients were dichotomized with respect to fibrosis stage (those with fibrosis stage ≤ 2 vs those with fibrosis stage ≥ 3), there were significant differences for all estimated perfusion parameters except Fp and Ft, with a significant increase in arterial perfusion (ART and Fa), DV, and MTT and a significant decrease in portal venous perfusion (PV and Fp) in patients with advanced fibrosis and cirrhosis (stage ≥ 3) (Table 2).


View this table:
[in this window]
[in a new window]

  		Table 1. Distribution of Estimated Liver Perfusion Parameters in 27 Patients Stratified with Respect to the Stage of Liver Fibrosis

 

View this table:
[in this window]
[in a new window]

  		Table 2. Distribution of Estimated Perfusion Parameters in 27 Patients Dichotomized with Respect to Fibrosis Stage

 
Receiver operating characteristic analysis showed that the three best estimated parameters used to predict advanced liver fibrosis stage were DV, MTT, and Fa (area under the receiver operating characteristic curve, 0.791–0.824; sensitivity, 76.9%–84.6%; and specificity, 71.4%–78.5%) (Table 3; Figs 3, 4). There was no significant difference between patients stratified by grade of necroinflammatory changes in any of the estimated perfusion parameters (grade 0 vs grade ≥ 1, P = .20–.80).


View this table:
[in this window]
[in a new window]

  		Table 3. Receiver Operating Characteristic Curve Analysis to Assess Utility of Estimated Perfusion Parameters as Predictors of Advanced Liver Fibrosis

 

Figure 3
View larger version (16K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3: Flowchart shows the patient and control populations, index test results (based on DV of gadopentetate dimeglumine), and reference standard (biopsy, MR imaging, or liver function tests). FNH = focal nodular hyperplasia, pMRI = perfusion MR imaging.

 

Figure 4
View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4: Receiver operating characteristic curves of the three best estimated perfusion parameters used for the prediction of severe liver fibrosis (stage ≥ 3). The parameters were DV, Fa, and MTT (areas under the receiver operating characteristic curve: 0.824, 0.791, and 0.775, respectively).

 
Segmental Variations in Estimated Liver Perfusion Parameters
There was more heterogeneous segmental liver perfusion in patients with fibrosis stage greater than or equal to 3 than in patients with fibrosis stage less than or equal to 2 for Fa, ART, PV, and DV, with significant results for only PV and DV (P = .03, Table 4).


View this table:
[in this window]
[in a new window]

  		Table 4. Segmental Variation of Estimated Liver Perfusion Parameters Expressed as Coefficients of Variation Derived from ROI Measurements at Four Liver Locations in One Subject

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE...
 References
 
In our initial experience, we demonstrated alterations in Fp, Fa, Ft, MTT, and DV of gadopentetate dimeglumine in patients with advanced liver fibrosis and cirrhosis, with good sensitivity and specificity for the diagnosis of advanced fibrosis obtained with at least three parameters (DV, Fa, and MTT).

Liver fibrosis and cirrhosis are associated with marked alterations in the vascular architecture, including collagen deposition in the space of Disse, loss of fenestration in the sinusoidal endothelium, sinusoidal basement membrane formation, and formation of basal laminas, which is associated with increased transit time of small and large molecules (19,27,28). In patients with cirrhosis, the observed decrease in portal flow is thought to result in a compensatory response of hepatic arterial vasodilation, with subsequent elevation of the hepatic arterial fraction (hepatic arterial buffer response), resulting in the maintenance of total liver blood flow (29). While this has been shown in an animal model of cirrhosis (30), it has not been correlated with the histologic stage of fibrosis in humans.

There are limited data on liver perfusion MR imaging in patients with focal or diffuse liver disease; however, the results of several studies are promising (17,18,31,32). Scharf et al (31) used a single-section T1-weighted gradient-recalled-echo technique to perform gadopentetate dimeglumine–enhanced perfusion MR imaging with 2-second temporal resolution in nine pigs prior to and after partial portal vein occlusion. Their results correlated well with those obtained with surgically placed ultrasonographic flow probes. Jackson et al (32) proposed the use of a 3D MR imaging technique to study perfusion in primary and metastatic lesions in cirrhotic and noncirrhotic patients (n = 14). They showed that lesion-specific perfusion parameters, including endothelial permeability and relative tumor blood volume, were reproducible. However, they assumed the hepatic artery was the only source of blood flow to hepatic lesions. This assumption could contribute to possible errors in calculated perfusion parameters for liver parenchyma. Materne et al (17) performed single-section liver perfusion MR imaging in rabbits with use of dynamic T1-weighted gradient-recalled-echo contrast-enhanced MR imaging. They showed that arterial, portal, and total hepatic blood flow parameters correlated well with parameters obtained by using radiolabeled microspheres (17). In a subsequent study, the same group (18) used a similar technique to examine 46 patients with cirrhosis. Their findings showed global, arterial, and portal perfusion, as well as MTT, correlated with the severity of disease, as assessed with the Child-Pugh class and the degree of portal hypertension. However, histologic evaluation of the degree of fibrosis was not correlated with perfusion parameters.

To our knowledge, we are the first to report that perfusion MR imaging can be used to diagnose liver fibrosis. We chose a 3D acquisition to ensure coverage of the entire liver and to enable us to detect regional changes in liver perfusion for PV and DV, which were more heterogeneous in patients with advanced fibrosis and cirrhosis. Collagen deposition in the space of Disse is associated with restricted diffusion of small particles in the extravascular space in fibrotic and cirrhotic livers; thus, there is an expected increase in MTT of small gadopentetate dimeglumine chelates throughout the liver (17,18,33,34). Our findings were similar. With endothelial defenestration and collagenization of the space of Disse, particles should theoretically be restricted to the intravascular space and should therefore be restricted in their distribution throughout the liver. Thus, there is an expected decrease in DV in association with liver fibrosis and cirrhosis, as seen in rabbits with use of a high-molecular-weight contrast agent; however, there is no decrease in DV with use of low-molecular-weight contrast agents, such as gadopentetate dimeglumine or gadoterate meglumine (33). In another study (18), the same group demonstrated a slight increase in mean DV in patients with cirrhosis versus those without cirrhosis (14.72% ± 6.39 [standard deviation] vs 11.43% ± 4.48); however, this increase was not significant. In our study, DV increased significantly in patients with advanced fibrosis and proved to be a useful parameter. The increase in DV could be explained by a decreased hepatic outflow in patients with advanced fibrosis (possibly from small venous compression in fibrosis and cirrhosis) and an increased interstitial volume.

The potential clinical applications of perfusion MR imaging are numerous, and they need to be further investigated and validated. Three-dimensional perfusion MR imaging with coverage of the entire liver could be useful in the detection of interval changes during antiviral treatment monitoring in patients with chronic viral hepatitis, and it could play a role in the detection and characterization of angiogenic activity in patients with hepatocellular carcinoma or metastatic lesions (34–36).

Our study had several limitations. First, this is a report of our initial experience, and our results are limited by the sample size—in particular, the small number of patients with intermediate levels of liver fibrosis—and by the fact that our results are likely to include one or more type I errors since no formal multiple hypothesis correction was used. Thus, a large-scale confirmation study needs to be performed to validate the significant correlation between estimated perfusion parameters and increasing degrees of fibrosis, as well as to determine the role of perfusion MR imaging in clinical practice. Second, patients with normal liver function and some patients with cirrhosis did not undergo liver biopsy for either ethical or practical reasons. However, the control patients without liver disease had normal liver function test results and no history of chronic hepatitis. The four patients with cirrhosis had definite findings of cirrhosis at imaging. Third, the estimated perfusion parameters did not show differences between patients stratified by single fibrosis stage (except between control subjects and patients with cirrhosis), and differences were present only when patients were combined (stage ≤ 2 vs stage ≥ 3 or stage 0 vs stage 1–3). Fourth, interobserver variability in estimated perfusion parameters was not assessed in this study. Fifth, the extensive postprocessing required to obtain perfusion parameters is a substantial barrier to the potential widespread clinical use of perfusion MR imaging. Postprocessing entails the transfer of data multiple times, with each transfer requiring its own data conversion.

Future work will need to focus on improved image postprocessing, with automated registration, vessel segmentation, and parametric mapping of perfusion parameters performed on a voxel-by-voxel basis. This could improve assessment of regional differences in patients with fibrosis, help to assess differentiation of mild and moderate fibrosis, and be useful in the detection and characterization of liver lesions. In addition, we are currently investigating a technique to improve gadopentetate dimeglumine quantification with T1-based methods (37).

In conclusion, we found that whole-liver 3D perfusion MR imaging is a feasible and noninvasive imaging modality in which multiple perfusion parameters can be measured and potentially used as biomarkers of liver fibrosis, with DV, Fa, and MTT having the best sensitivity and specificity in the diagnosis of advanced fibrosis. With the development of misregistration correction software and volumetric color-based parametric maps, perfusion MR imaging may become a clinical imaging tool that can be used to evaluate chronic liver disease.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE...
 References
 


   IMPLICATION FOR PATIENT CARE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE...
References
 


   FOOTNOTES
 

Abbreviations: ART = arterial fraction • DV = distribution volume • Fa = absolute arterial liver blood flow • Fp = absolute portal liver blood flow • Ft = absolute total liver blood flow • MTT = mean transit time • PV = portal venous fraction • ROI = region of interest • 3D = three-dimensional

2 Current address: Department of Radiology, Valley Hospital, Ridgewood, NJ. Back

Guarantors of integrity of entire study, H.R., V.S.L., B.T.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, M.H., H.R., V.S.L., G.A.K., B.T.; clinical studies, M.H., H.R., M.L., M.A.B., G.A.K., B.T.; experimental studies, G.A.K.; statistical analysis, B.T.; and manuscript editing, M.H., H.R., V.S.L., M.L., B.T.

Authors stated no financial relationship to disclose.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE...
 References
 

  1. Afdhal NH, Nunes D. Evaluation of liver fibrosis: a concise review. Am J Gastroenterol 2004;99:1160–1174. [CrossRef][Medline]
  2. Kim AI, Saab S. Treatment of hepatitis C. Am J Med 2005;118:808–815. [CrossRef][Medline]
  3. Strader DB, Wright T, Thomas DL, Seeff LB. Diagnosis, management, and treatment of hepatitis C. Hepatology 2004;39:1147–1171. [CrossRef][Medline]
  4. Gish RG. Current treatment and future directions in the management of chronic hepatitis B viral infection. Clin Liver Dis 2005;9:541–565, v. [CrossRef][Medline]
  5. Piccinino F, Sagnelli E, Pasquale G, Giusti G. Complications following percutaneous liver biopsy: a multicentre retrospective study on 68,276 biopsies. J Hepatol 1986;2:165–173. [CrossRef][Medline]
  6. Janes CH, Lindor KD. Outcome of patients hospitalized for complications after outpatient liver biopsy. Ann Intern Med 1993;118:96–98. [Abstract/Free Full Text]
  7. Froehlich F, Lamy O, Fried M, Gonvers JJ. Practice and complications of liver biopsy: results of a nationwide survey in Switzerland. Dig Dis Sci 1993;38:1480–1484. [CrossRef][Medline]
  8. Cadranel JF, Rufat P, Degos F; for the Group of Epidemiology of the French Association for the Study of the Liver (AFEF). Practices of liver biopsy in France: results of a prospective nationwide survey. Hepatology 2000;32:477–481. [CrossRef][Medline]
  9. Maharaj B, Maharaj RJ, Leary WP, et al. Sampling variability and its influence on the diagnostic yield of percutaneous needle biopsy of the liver. Lancet 1986;1:523–525. [Medline]
  10. Regev A, Berho M, Jeffers LJ, et al. Sampling error and intraobserver variation in liver biopsy in patients with chronic HCV infection. Am J Gastroenterol 2002;97:2614–2618. [CrossRef][Medline]
  11. Bedossa P, Dargere D, Paradis V. Sampling variability of liver fibrosis in chronic hepatitis C. Hepatology 2003;38:1449–1457. [Medline]
  12. Ito K, Mitchell DG, Gabata T, Hussain SM. Expanded gallbladder fossa: simple MR imaging sign of cirrhosis. Radiology 1999;211:723–726. [Abstract/Free Full Text]
  13. Ito K, Mitchell DG, Gabata T. Enlargement of hilar periportal space: a sign of early cirrhosis at MR imaging. J Magn Reson Imaging 2000;11:136–140. [CrossRef][Medline]
  14. Ito K, Mitchell DG. Hepatic morphologic changes in cirrhosis: MR imaging findings. Abdom Imaging 2000;25:456–461. [CrossRef][Medline]
  15. Awaya H, Mitchell DG, Kamishima T, Holland G, Ito K, Matsumoto T. Cirrhosis: modified caudate-right lobe ratio. Radiology 2002;224:769–774. [Abstract/Free Full Text]
  16. Aguirre DA, Behling CA, Alpert E, Hassanein TI, Sirlin CB. Liver fibrosis: noninvasive diagnosis with double contrast material–enhanced MR imaging. Radiology 2006;239:425–437. [Abstract/Free Full Text]
  17. Materne R, Smith AM, Peeters F, et al. Assessment of hepatic perfusion parameters with dynamic MRI. Magn Reson Med 2002;47:135–142. [CrossRef][Medline]
  18. Annet L, Materne R, Danse E, Jamart J, Horsmans Y, Van Beers BE. Hepatic flow parameters measured with MR imaging and Doppler US: correlations with degree of cirrhosis and portal hypertension. Radiology 2003;229:409–414. [Abstract/Free Full Text]
  19. Van Beers BE, Leconte I, Materne R, Smith AM, Jamart J, Horsmans Y. Hepatic perfusion parameters in chronic liver disease: dynamic CT measurements correlated with disease severity. AJR Am J Roentgenol 2001;176:667–673. [Abstract/Free Full Text]
  20. Batts KP, Ludwig J. Chronic hepatitis: an update on terminology and reporting. Am J Surg Pathol 1995;19:1409–1417. [Medline]
  21. Ito K, Mitchell DG, Siegelman ES. Cirrhosis: MR imaging features. Magn Reson Imaging Clin N Am 2002;10:75–92, vi. [CrossRef][Medline]
  22. Burkart DJ, Johnson CD, Reading CC, Ehman RL. MR measurements of mesenteric venous flow: prospective evaluation in healthy volunteers and patients with suspected chronic mesenteric ischemia. Radiology 1995;194:801–806. [Abstract/Free Full Text]
  23. Ivancevic MK, Zimine I, Foxall D, et al. Inflow effect in first-pass cardiac and renal MRI. J Magn Reson Imaging 2003;18:372–376. [CrossRef][Medline]
  24. Bokacheva L, Rusinek H, Chen Q, et al. Quantitative determination of Gd-DTPA concentration in T1-weighted MR renography studies. Magn Reson Med 2007;57:1012–1018. [CrossRef][Medline]
  25. Jones RA, Easley K, Little SB, Scherz H, Kirsch AJ, Grattan-Smith JD. Dynamic contrast-enhanced MR urography in the evaluation of pediatric hydronephrosis. I. Functional assessment. AJR Am J Roentgenol 2005;185:1598–1607.
  26. Nelder JA, Mead R. A simplex method for function minimization. Comp J 1965;7:308–313.
  27. Orrego H, Blendis LM, Crossley IR, et al. Correlation of intrahepatic pressure with collagen in the Disse space and hepatomegaly in humans and in the rat. Gastroenterology 1981;80:546–556. [Medline]
  28. Martinez-Hernandez A. The hepatic extracellular matrix. II. Electron immunohistochemical studies in rats with CCl4-induced cirrhosis. Lab Invest 1985;53:166–186.
  29. Gulberg V, Haag K, Rossle M, Gerbes AL. Hepatic arterial buffer response in patients with advanced cirrhosis. Hepatology 2002;35:630–634. [CrossRef][Medline]
  30. Richter S, Mucke I, Menger MD, Vollmar B. Impact of intrinsic blood flow regulation in cirrhosis: maintenance of hepatic arterial buffer response. Am J Physiol Gastrointest Liver Physiol 2000;279:G454–G462. [Abstract/Free Full Text]
  31. Scharf J, Zapletal C, Hess T, et al. Assessment of hepatic perfusion in pigs by pharmacokinetic analysis of dynamic MR images. J Magn Reson Imaging 1999;9:568–572. [CrossRef][Medline]
  32. Jackson A, Haroon H, Zhu XP, Li KL, Thacker NA, Jayson G. Breath-hold perfusion and permeability mapping of hepatic malignancies using magnetic resonance imaging and a first-pass leakage profile model. NMR Biomed 2002;15:164–173. [CrossRef][Medline]
  33. Van Beers BE, Materne R, Annet L, et al. Capillarization of the sinusoids in liver fibrosis: noninvasive assessment with contrast-enhanced MRI in the rabbit. Magn Reson Med 2003;49:692–699. [CrossRef][Medline]
  34. Pandharipande PV, Krinsky GA, Rusinek H, Lee VS. Perfusion imaging of the liver: current challenges and future goals. Radiology 2005;234:661–673. [Abstract/Free Full Text]
  35. Morgan B, Thomas AL, Drevs J, et al. Dynamic contrast-enhanced magnetic resonance imaging as a biomarker for the pharmacological response of PTK787/ZK 222584, an inhibitor of the vascular endothelial growth factor receptor tyrosine kinases, in patients with advanced colorectal cancer and liver metastases: results from two phase I studies. J Clin Oncol 2003;21:3955–3964. [Abstract/Free Full Text]
  36. Padhani AR, Dzik-Jurasz A. Perfusion MR imaging of extracranial tumor angiogenesis. Top Magn Reson Imaging 2004;15:41–57. [CrossRef][Medline]
  37. Bokacheva L, Huang AJ, Chen Q, et al. Single breath-hold T1 measurement using low flip angle TrueFISP. Magn Reson Med 2006;55:1186–1190. [CrossRef][Medline]



This article has been cited by other articles:


Home page
 Am. J. Roentgenol.Home page
B. Taouli, R. L. Ehman, and S. B. Reeder
Advanced MRI Methods for Assessment of Chronic Liver Disease
Am. J. Roentgenol., July 1, 2009; 193(1): 14 - 27.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2463070077v1
246/3/926    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hagiwara, M.
Right arrow Articles by Taouli, B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hagiwara, M.
Right arrow Articles by Taouli, B.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE