Value of Diffusion-Weighted MRI for Assessing Liver Fibrosis and Cirrhosis
Abstract
OBJECTIVE. The objective of our study was to determine the usefulness of the apparent diffusion coefficient (ADC) of liver parenchyma for determining the severity of liver fibrosis.
MATERIALS AND METHODS. This study investigated 78 patients who underwent diffusion-weighted imaging (DWI) with 1.5-T MRI and pathologic staging of liver fibrosis based on biopsy. DWI was performed with b values of 50 and 400 s/mm2. ADCs of liver were measured using 2.0- to 3.0-cm2 regions of interest in the right and left lobes of the liver; the mean ADC value was used for analysis. Pathologic METAVIR scores for liver fibrosis stage were used as a reference standard.
RESULTS. The mean ADC values for fibrosis pathologically staged using the METAVIR classification system as F0 (n = 11), F1 (n = 16), F2 (n = 10), F3 (n = 14), and F4 (n = 27) were 125.9, 105.0, 104.5, 103.2, and 99.1 × 10-5 s/mm2, respectively. The correlation between the ADC values and the degree of liver fibrosis was moderate (Spearman's test, ρ = –0.36). There was a significant difference in ADC values between patients with nonfibrotic liver (F0) and those with cirrhotic liver (F4) (p = 0.008). The best cutoff ADC value to distinguish between these groups was 118 × 10-5 s/mm2. However, ADC values were not useful for differentiating viral hepatitis patients with F2 fibrosis or higher from those with a lower degree of fibrosis (area under the receiver operating characteristic curve [AUC] = 0.66) or for differentiating low-stage fibrosis in all patients from high-stage fibrosis in all patients (AUC = 0.54).
CONCLUSION. The ADCs in cirrhotic livers are significantly lower than those in nonfibrotic livers. However, ADC values measured using the current generation of scanners are not reliable enough to replace liver biopsy for staging hepatic fibrosis.
Introduction
Liver fibrosis is a consequence of sustained prolonged injury from a variety of causes, including alcohol- and drug-induced, viral, autoimmune, cholestatic, and metabolic diseases. Fibrosis indicates liver damage and is an important cause of portal hypertension. Progression of early fibrosis can be reversed by treatment with specific antifibrotic therapy or by removal of the cause, such as viral hepatitis or alcohol-induced disease [1–5]. Therefore, identification of the early stages of liver fibrosis is crucial.
Currently, liver biopsy is the reference standard for detecting and staging liver fibrosis. However, this method is limited by a small (1–5%) but definite rate of complications. In addition, a very small portion of the liver is sampled at biopsy, so biopsy results are subject to sampling variability related to the heterogeneity of the distribution of liver fibrosis [6]. Because of the limitations of liver biopsy, noninvasive methods including serologic tests [7], such as the FibroTest (Bio-Predictive), and advanced imaging methods are being investigated for the diagnosis of liver fibrosis [8–11]. FibroTest gives a score of liver fibrosis based on serum levels of α2-macroglobulin, α2-globulin (haptoglobin), γ-globulin, apolipoprotein A-I, γ-glutamyl transferase, and total bilirubin. Conventional CT and MRI are not sensitive to the early stages of fibrosis.
Liver fibrosis results in extracellular accumulation of collagen, glycosaminoglycans, and proteoglycans that may restrict the molecular diffusion of water, thus suggesting that diffusion-weighted imaging (DWI) may be useful for assessing fibrosis. However, DWI of the liver is beset with several problems. These problems include susceptibility to motion artifact and eddy currents and poor signal-to-noise ratio, particularly when strong diffusion-sensitizing gradients (i.e., high b values) are used in the scan sequence. Most studies of DWI have found that the apparent diffusion coefficient (ADC) of cirrhotic livers is significantly lower than that of normal livers [12–18], although no difference in ADCs was noted in one study [19]. A reduction in ADC values with increasing fibrosis has also been noted in animal models of chemically induced fibrosis [20, 21]. To our knowledge, only three human studies have assessed the correlation of ADCs with the various stages of hepatic fibrosis [15, 17, 19]. These studies were performed predominantly on patients with hepatitis C and healthy volunteers. The purpose of the current study was to determine the usefulness of ADC values for assessing the severity of liver fibrosis with different causes for chronic liver disease based on a large population of patients.
Materials and Methods
Patients
For this retrospective HIPAA-compliant study, we reviewed the imaging and clinical records of 183 patients who underwent DWI between October 2006 and March 2007. Institutional review board permission was obtained for retrospective assessment of imaging and clinical data, along with a waiver of informed consent. Patients were excluded if pathologic confirmation of liver fibrosis stage within 6 months of MRI was not available (n = 87), if there was thrombosis of the main portal vein or its right or left branches (n = 3), or if the DW images contained artifacts that precluded ADC measurements (n = 15). The remaining 78 patients were included in this study.
The study group was composed of 55 men and 23 women. The mean age was 53 years (range, 28–74 years). The cause of liver disease was alcohol, n = 13; nonalcoholic steatohepatitis, n = 11; hepatitis C, n = 35 (includes four patients with combined hepatitis C and alcohol cause); hepatitis B, n = 8; autoimmune hepatitis, n = 6; and miscellaneous (including cryptogenic, primary sclerosing cholangitis, and drug-induced), n = 5.
The mean time interval between histopathology and MRI study was 2.4 months (range, 0–6 months). There were 11, 16, 10, 14, and 27 patients with stage F0, F1, F2, F3, and F4 fibrosis, respectively. Twenty-eight of the patients had undergone liver transplantation at least 6 months before MRI. For all patients, histology results from either a core biopsy of liver performed using ultrasound guidance or an explant were available, as discussed later in this article.
MRI Examination
All MRI examinations were performed using a 1.5-T 18-channel MRI scanner (Magnetom Avanto, Siemens Healthcare) with high performance gradients (maximum gradient, 45 mT/m; maximum slew rate, 200 T/m/s). Transverse breath-hold single-shot echo-planar imaging was performed using a six-element phased-array coil. Frequency-selective fat saturation was used to reduce chemical shift artifacts. The parameters used were TR range/TE range, 14–1,600/52–62; field of view, 36–40 cm; matrix, 192 × 115; number of excitations, 2; slice width, 6 mm; gap, 2 mm; and receiver bandwidth, 1,736 Hz per pixel. Parallel imaging was performed using the generalized auto calibrating partially parallel acquisition (GRAPPA) with an acceleration factor (r) of 2 as well as 75% partial-phase Fourier and 80% phase resolution to reduce TE and minimize artifacts. Typically, three breath-hold acquisitions, each lasting 21 seconds, were performed to cover the entire liver. Each acquisition was obtained in the same liver location at b values of 50 and 400 s/mm2 with three orthogonal diffusion gradients. These b values were used to obtain ADC values that were more sensitive to diffusion than would be possible if b values of 0 and 400 s/mm2 had been used [22]. Quantitative ADC maps were created on a voxel-by-voxel basis using the software on the scanner (Syngo vB13A, Siemens Healthcare).
A single reader who was blinded to the histology results placed a region of interest (ROI) in each lobe of the liver. The ROIs were between 2–5 cm2 in size. The central vascular region, potential masses, and peripheral 2 cm of the liver were avoided. All patients underwent routine MRI including gadolinium-enhanced sequences. These sequences were used to clarify the liver outline and detect liver masses. The mean of the right and left lobe ADCs was used for statistical analysis.
Histopathology
The stage of fibrosis was determined on the basis of the pathology of an explant (n = 5) or an 18-gauge ultrasound-guided core biopsy liver sample (n = 73). If more than one biopsy was performed, the result of the biopsy temporally closest to the MRI examination was used. The pathology specimens were scored by an experienced histopathologist using the METAVIR classification system [23], which classifies fibrosis in five stages: F0, no fibrosis; F1, portal fibrosis without septa; F2, few septa; F3, bridging fibrosis without cirrhosis; and F4, cirrhosis with architectural distortion. The equivalent to the METAVIR score was recorded by the hepatopathologist in cases with a cause other than viral hepatitis.
Statistical Analysis
Statistical analysis was performed using statistics software (MedCalc, version 9.2, MedCalc Software). The data were checked for distribution and skewness. Because the ADC values were not normally distributed, nonparametric tests were used. Mann-Whitney and Kruskal-Wallis tests were used to determine the significance of the differences in ADC values between different stages of fibrosis. For patients with viral hepatitis, comparisons were made between F0 and F1 (which typically are not treated with antiviral therapy) and F2–F4 (which may be treated with antiviral agents). In addition, for all patients, comparisons were made between low-stage fibrosis (F1 and F2) and high-stage fibrosis (F3 and F4). Spearman's rank-order test was used to determine the correlation between ADC values and the degree of fibrosis. Receiver operating characteristic (ROC) curves were used to evaluate the usefulness of ADC measurements for predicting the presence of stage F3 or F4 fibrosis and the presence of stage F2 or higher fibrosis (in viral hepatitis). The value of ADC that separates these groups with the best possible specificity and sensitivity was calculated. A p value of < 0.05 was considered significant.
Results
ADC Values and Liver Fibrosis Stages
Figure 1 is the box-and-whisker plot of the ADC values in patients grouped by fibrosis stage. The mean (SD) ADC values were 125.9 (17.2), 105.0 (16.1), 104.5 (18.5), 103.2 (15.9), and 99.1 (12.0) × 10-5 s/mm2 in fibrosis stages F0, F1, F2, F3, and F4, respectively. The ADC values in each fibrosis group overlapped substantially except between stage F0 and stage F3 or F4 (Figs. 2A, 2B, 2C and 3A, 3B, 3C). There was also a significant difference between ADC values of nonfibrotic (F0) and cirrhotic (F4) livers (p = 0.008, Mann-Whitney test). However, no differences were seen in the ADC values of F0 compared with F1 (p = 0.06), F1 compared with F2 (p = 0.85), F2 compared with F3 (p = 0.84), and F3 compared with F4 (p = 0.45). The Spearman's rank-order correlation (ρ) between fibrosis stage and ADC values was –0.36 (95% CI of ρ = –0.537 to –0.146). Taking all fibrosis groups, there was no significant difference in ADC values between the groups (p = 0.32, Kruskal-Wallis test).
ROC curves were used to analyze the usefulness of ADC values for predicting fibrosis stage. Figure 4 shows the ROC curve with ADC for the prediction of the absence of fibrosis (F0) or cirrhosis (F4) in all patients. Figure 5 shows the ROC curve in the prediction of F2 fibrosis and higher stages in patients with hepatitis B or hepatitis C. Due to considerations of cost, toxicity, and efficacy, antiviral therapy is given to patients with stage F2 fibrosis and above [24]. Figure 6 shows the ROC curve in discriminating high-stage fibrosis (F3, F4) from low-stage fibrosis (F1, F2) in all patients. Table 1 gives the area under the ROC curve (AUC) and the ADC value that resulted in the maximized sensitivity and specificity for differentiating fibrosis stages.
| Stage of Fibrosis | AUC (95% CI) | ADC Value (10-5 s/mm2) | Sensitivity (%) (95% CI) | Specificity (%) (95% CI) |
|---|---|---|---|---|
| Nonfibrotic (F0) vs cirrhosis (F4) | 0.919 (0.784-0.982) | ≤ 115 | 96.3 (81.0-99.4) | 81.8 (48.2-97.2) |
| F0 and F1 vs F2 or highera | 0.686 (0.571-0.787) | ≤ 103 | 72.6 (58.3-84.1) | 59.3 (38.8-77.6) |
| Low-stageb vs high-stagec fibrosis | 0.656 (0.496-0.794) | ≤ 98 | 51.7 (32.5-70.5) | 71.4 (41.9-91.4) |
a
Patients with viral hepatitis (see text for explanation).
b
F1 and F2.
c
F3 and F4.
Effect of the Cause of Liver Disease on ADC Values
We assessed the effect of the cause of liver disease on the ADC values of patients with high-stage fibrosis (F3 and F4). The cause of liver disease in the 41 patients with stage F3 and F4 fibrosis were hepatitis C, n = 21; hepatitis B, n = 3; alcohol, n = 5; nonalcoholic steatohepatitis, n = 10; and primary biliary cirrhosis, n = 2. There were no significant differences in ADC values between patients with viral hepatitis and those without (p = 0.79, Mann-Whitney test). There was no difference in ADC values of patients with hepatitis C and those with alcohol-induced highstage liver fibrosis (p = 0.58) or between patients with alcohol-induced high-stage liver fibrosis and those with nonalcoholic steatohepatitis–induced high-stage liver fibrosis (p = 0.51).
Discussion
The results of several studies have shown that the ADC values of cirrhotic patients are lower than those of noncirrhotic patients or of healthy volunteers [12–14, 18]. However, these studies did not assess the usefulness of the ADC in evaluating the intermediate stages of fibrosis. To our knowledge, only three studies have assessed the changes in ADC values with the different stages of fibrosis. Taouli et al. [15] assessed seven control subjects and 23 patients with hepatitis-related liver disease using multiple b values of 0–1,000 s/mm2. Although there was a significant difference in the ADC of the F0 and F1 group compared with the ADC of the F2 to F4 group, there was much overlap in the ADC values of individual patients in each group. The Spearman's rank-order correlation of ADC with fibrosis score (Batts-Ludwig classification) was moderate (ρ = –0.45 to –0.65). Lewin et al. [17] compared DWI in 54 hepatitis C patients and 20 healthy volunteers with ultrasound elastography (FibroScan, EchoSens) and FibroTest. They found that DWI was equivalent to these tests in detecting high stages (F3 and F4, METAVIR classification system) of fibrosis. Again, there was significant overlap of ADC values between patients with F0, F1, and F2 fibrosis. Boulanger et al. [19] could not find a difference between ADC values and fibrosis scores in 18 patients with hepatitis C and 10 healthy volunteers. For their study, they used smaller b values (0 and 250 s/mm2) and a different pathologic scoring system (Ishak scale).
Our findings, in common with those of most studies, showed a significant difference in the ADC values of nonfibrotic (F0) and cirrhotic (F4) patients. With an ADC value of 115 × 10-5 s/mm2, it was possible with a high degree of sensitivity (96.3%) and specificity (81.8%) to differentiate F0 fibrosis from F4 fibrosis. However, there was substantial overlap in the ADC values of F1 to F4. No cutoff ADC value could reasonably separate low-stage fibrosis from high-stage fibrosis or predict which patients with viral hepatitis would benefit from antiviral therapy (F2 to F4) [24]. In this respect, we found that DWI did not give information that was superior to that obtained using conventional imaging methods or that could replace core liver biopsy, which is the current reference standard for liver fibrosis staging.
There is a potential for ADC values to be affected not only by the stage of fibrosis but also by the cause of disease. For instance, increased fatty infiltration in patients with nonalcoholic steatohepatitis may show ADC values that are different from the ADC values of those with untreated viral hepatitis, in whom necrosis and inflammation may be a more prominent pathologic process. In our analysis, we did not find a significant difference in ADC values of high-stage fibrosis patients without viral hepatitis and those with viral hepatitis.
Among the many difficulties with hepatic DWI is the lack of a standardized imaging protocol. Studies performed on different MRI scanners using different software for ADC maps result in different ADC values. For instance, the ADC values of normal liver ranged from 95 to 345 × 10-5 s/mm2 [14, 25]. In most studies, including our study, the ADC of normal liver varied between 120 and 170 × 10-5 s/mm2 [12, 13, 15, 17, 18].
The optimal technique of DWI for assessing hepatic fibrosis has not yet been established. We used a breath-hold single-shot echo-planar technique with b values of 50 and 400 s/mm2. The signal-to-noise ratio will be improved by the use of respiratory-triggered sequences, which require a 4- to 8-minute acquisition time [26]. The choice of b values affects the calculated ADC [27]. ADC measurements using images obtained with higher b values are more sensitive to diffusion and are generally smaller than those obtained using b values of less than 500 s/mm2. However, in the assessment of hepatic fibrosis, there may be an advantage to calculating ADC values using an intermediate b value of 400 s/mm2 compared with a higher b value of 800 s/mm2 [27]. Hepatic perfusion is known to be substantially reduced in cirrhotic patients compared with healthy subjects [28]. A study on rats showed that the ADC values were inversely correlated with hepatic fibrosis in live rats but not after death [21]. In addition, a study on humans suggested that a decline in perfusion may be a significant contributor to the reduced ADC seen in cirrhotic patients [29]. These studies suggest that the assessment of microvascular perfusion as well as of true diffusion is important in trying to differentiate a liver without fibrosis from those with mild or severe fibrosis. Lower b values are more sensitive to the effects of perfusion and may allow better discrimination of hepatic fibrosis stage.
We acknowledge some limitations of our study. The study was retrospective and hence the time interval between MRI and histologic confirmation of liver fibrosis varied, with a mean interval of 2.4 months and a maximum interval of 6 months. Despite this relatively small time interval, medical therapy for fibrosis or progression of disease may have led to discordance between the degree of fibrosis at the time of biopsy and that at the time of MRI examination. This putative change in fibrosis stage must be considered in the context of the numerous pitfalls in using liver biopsy as a reference standard, including the sampling variability depending on the site of liver biopsied and the interobserver variability in reporting fibrosis stage [6].
In conclusion, we found that ADC values can be used to distinguish nonfibrotic liver from cirrhotic liver. We believe that ADC measurements taken at a single time frame using 1.5-T MRI cannot be used to reliably distinguish among the intermediate stages of fibrosis. Future research should concentrate on the usefulness of serial ADC measurements to determine whether changes in ADC values correlate with changes in the degree of fibrosis.
Footnote
Address correspondence to K. Sandrasegaran ([email protected]).
References
1.
Malekzadeh R, Mohamadnejad M, Nasseri-Moghaddam S, et al. Reversibility of cirrhosis in autoimmune hepatitis. Am J Med 2004; 117:125 –129
2.
Wanless IR, Nakashima E, Sherman M. Regression of human cirrhosis: morphologic features and the genesis of incomplete septal cirrhosis. Arch Pathol Lab Med 2000; 124:1599 –1607
3.
Dufour JF, DeLellis R, Kaplan MM. Regression of hepatic fibrosis in hepatitis C with long-term interferon treatment. Dig Dis Sci 1998; 43:2573 –2576
4.
Rambaldi A, Gluud C. Colchicine for alcoholic and non-alcoholic liver fibrosis and cirrhosis. Cochrane Database Syst Rev 2001;CD002148
5.
Safadi R, Zigmond E, Pappo O, Shalev Z, Ilan Y. Amelioration of hepatic fibrosis via beta-glucosylceramide-mediated immune modulation is associated with altered CD8 and NKT lymphocyte distribution. Int Immunol 2007; 19:1021 –1029
6.
Afdhal NH, Nunes D. Evaluation of liver fibrosis: a concise review. Am J Gastroenterol 2004; 99:1160 –1174
7.
Vallet-Pichard A, Mallet V, Nalpas B, et al. FIB-4: an inexpensive and accurate marker of fibrosis in HCV infection—comparison with liver biopsy and FibroTest. Hepatology 2007; 46:32–36
8.
Friedrich-Rust M, Ong MF, Herrmann E, et al. Real-time elastography for noninvasive assessment of liver fibrosis in chronic viral hepatitis. AJR 2007; 188:758 –764
9.
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
10.
Grier S, Lim AK, Patel N, et al. Role of microbubble ultrasound contrast agents in the non-invasive assessment of chronic hepatitis C–related liver disease. World J Gastroenterol 2006; 12:3461 –3465
11.
Rouviere O, Yin M, Dresner MA, et al. MR elastography of the liver: preliminary results. Radiology 2006; 240:440–448
12.
Moteki T, Horikoshi H. Evaluation of hepatic lesions and hepatic parenchyma using diffusion-weighted echo-planar MR with three values of gradient b-factor. J Magn Reson Imaging 2006; 24:637 –645
13.
Namimoto T, Yamashita Y, Sumi S, Tang Y, Takahashi M. Focal liver masses: characterization with diffusion-weighted echo-planar MR imaging. Radiology 1997; 204:739–744
14.
Koinuma M, Ohashi I, Hanafusa K, Shibuya H. Apparent diffusion coefficient measurements with diffusion-weighted magnetic resonance imaging for evaluation of hepatic fibrosis. J Magn Reson Imaging 2005; 22:80 –85
15.
Taouli B, Tolia AJ, Losada M, et al. Diffusion-weighted MRI for quantification of liver fibrosis: preliminary experience. AJR 2007; 189:799 –806
16.
Ichikawa T, Haradome H, Hachiya J, Nitatori T, Araki T. Diffusion-weighted MR imaging with a single-shot echoplanar sequence: detection and characterization of focal hepatic lesions. AJR 1998; 170:397 –402
17.
Lewin M, Poujol-Robert A, Boelle PY, et al. Diffusion-weighted magnetic resonance imaging for the assessment of fibrosis in chronic hepatitis C. Hepatology 2007; 46:658–665
18.
Girometti R, Furlan A, Bazzocchi M, et al. Diffusion-weighted MRI in evaluating liver fibrosis: a feasibility study in cirrhotic patients. Radiol Med 2007; 112:394–408
19.
Boulanger Y, Amara M, Lepanto L, et al. Diffusion-weighted MR imaging of the liver of hepatitis C patients. NMR Biomed 2003; 16:132 –136
20.
Guan S, Zhou KR, Zhao WD, Peng WJ, Tang F, Mao J. Magnetic resonance diffusion-weighted imaging in the diagnosis of diffuse liver diseases in rats. Chin Med J (Engl) 2005; 118:639–644
21.
Annet L, Peeters F, Abarca-Quinones J, Leclercq I, Moulin P, Van Beers BE. Assessment of diffusion-weighted MR imaging in liver fibrosis. J Magn Reson Imaging 2007; 25:122–128
22.
Yanagisawa O, Shimao D, Maruyama K, Nielsen M, Irie T, Niitsu M. Diffusion-weighted magnetic resonance imaging of human skeletal muscles: gender-, age- and muscle-related differences in apparent diffusion coefficient. Magn Reson Imaging 2009; 27:69–78
23.
[No authors listed]. Intraobserver and interobserver variations in liver biopsy interpretation in patients with chronic hepatitis C. The French METAVIR Cooperative Study Group. Hepatology 1994; 20(1 Pt 1):15 –20
24.
Kim AI, Saab S. Treatment of hepatitis C. Am J Med 2005; 118:808 –815
25.
Murtz P, Flacke S, Traber F, van den Brink JS, Gieseke J, Schild HH. Abdomen: diffusion-weighted MR imaging with pulse-triggered single-shot sequences. Radiology 2002; 224:258–264
26.
Kandpal H, Sharma R, Madhusudhan KS, Kapoor KS. Respiratory-triggered versus breath-hold diffusion-weighted MRI of liver lesions: comparison of image quality and apparent diffusion coefficient values. AJR 2009; 192:915–922
27.
Girometti R, Furlan A, Esposito G, et al. Relevance of b-values in evaluating liver fibrosis: a study in healthy and cirrhotic subjects using two single-shot spin-echo echo-planar diffusion-weighted sequences. J Magn Reson Imaging 2008; 28:411 –419
28.
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
29.
Luciani A, Vignaud A, Cavet M, et al. Liver cirrhosis: intravoxel incoherent motion MR imaging—pilot study. Radiology 2008; 249:891–899
Information & Authors
Information
Published In
Copyright
© American Roentgen Ray Society.
History
Submitted: January 21, 2009
Accepted: June 4, 2009
First published: November 23, 2012
Keywords
Authors
Metrics & Citations
Metrics
Citations
Export Citations
To download the citation to this article, select your reference manager software.