Prediction of high-stage liver fibrosis using ADC value on diffusion-weighted imaging and quantitative enhancement ratio at the hepatobiliary phase of Gd-EOB-DTPA

Abstract

Background

Recently, diffusion-weighted imaging (DWI) and quantitative enhancement ratio measured at the hepatobiliary phase (HBP) of Gd-EOB-DTPA–enhanced magnetic resonance imaging (MRI) has been established as an effective method for evaluating liver fibrosis.

Purpose

To evaluate which is a more favorable surrogate marker in predicting high-stage liver fibrosis, apparently diffusion coefficient (ADC) value or quantitative enhancement ratio measured on HBP.

Material and Methods

Eighty-three patients with 99 surgically resected hepatic lesions were enrolled in this study. DWI was performed with b-values of 100 and 800 s/mm². Regions of interest were set on ADC map, and the HBP of Gd-EOB-DTPA–enhanced MRI, to calculate ADC value, liver-to-muscle ratio (LMR), liver-to-spleen ratio (LSR), and contrast enhancement index (CEI) of liver. We compared these parameters between low-stage fibrosis (F0, F1, and F2) and high-stage fibrosis (F3 and F4). Receiver operating characteristic analysis was performed to compare the diagnostic performance when distinguishing low-stage fibrosis from high-stage fibrosis.

Results

LMR and CEI were significantly lower at high-stage fibrosis than at the low stage (P < 0.01 and P = 0.04, respectively), whereas LSR did not show a significant difference (P = 0.053). No significant difference was observed in diagnostic performance between LMR and CEI (P = 0.185). The best sensitivity and specificity, when an LMR of 2.80 or higher was considered to be low-stage fibrosis, were 82.4% and 75.6%, respectively. ADC value showed no significant differences among fibrosis grades (P = 0.320).

Conclusion

LMR and CEI were both adequate surrogate parameters to distinguish high-stage fibrosis from low-stage fibrosis.

Keywords

Liver fibrosis diffusion-weighted imaging (DWI)magnetic resonance imaging (MRI)Gd-EOB-DTPA enhancement ratio hepatosteatosis

Introduction

Liver fibrosis is a reaction of stellate cells for chronic inflammation caused by hepatitis B or C virus and many other causes. The fibrosis leads to cirrhosis, portal hypertension after many years, and eventually can be fatal. Moreover, the severe fibrosis restricts surgical indications. Liver biopsy is a widely accepted procedure for diagnosing and grading liver fibrosis. However, it is associated with major complications in 0.3% of patients and with mortality in 0.018% (1). Furthermore, because of the heterogeneity of liver fibrosis, sampling errors can also arise (2,3). Therefore, alternative non-invasive diagnostic methods that can precisely evaluate liver fibrosis are desirable. Some methods, such as elastography, diffusion-weighted imaging (DWI) and contrast enhancement index (CEI), have been introduced to evaluate liver fibrosis (4,5). DWI and CEI do not require special tools and can be performed within routine clinical work. DWI can evaluate the restricted diffusion by collagen fibers accumulated interstitially (4,6,7) in cirrhotic liver. The CEI for fibrosis evaluation is currently using gadolinium ethoxy-benzyl diethylenetriamine penta-acetic acid (Gd-EOB-DTPA). This method is affected by the number of functional hepatocytes. In case of cirrhosis, the number of hepatocytes decreases, and consequently, the enhancement at the hepatobiliary phase decreases (8 –10). Some indexes of the enhancement ratio have been proposed, such as the liver-to-muscle ratio (LMR), liver-to-spleen ratio (LSR), and CEI (8 –10). In these articles, the biopsy specimen was not always consistent with the data acquisition area on imaging. The direct comparison of the surgical resected specimen with that of the corresponding area on imaging is desirable. This method can set the region of interest (ROI) precisely and a sufficient amount of specimen for pathological evaluation can be obtained. Furthermore, previous papers included other bias, for example, all the selected patients had B or C virus infection or a habit of alcohol intake (8). Therefore, we intended to assess which index, which could be obtained during routine clinical work, was the best surrogate marker for evaluating liver fibrosis.

Material and Methods

Patients

Our institutional review board approved this retrospective study and informed consent was waived. We reviewed the imaging and clinical records. The inclusion criteria were as follows: participants who were examined with both abdominal magnetic resonance imaging (MRI) and hepatic resection because of suspicion of tumor between 2009 and 2014. The interval between abdominal MRI and surgery was three months. Two hundred patients were selected consecutively. The exclusion criteria were as follows: insufficient amount of specimen to evaluate pathologically; no tumor in specimen; needle biopsy; more than ten tumors in the liver; did not receive Gd-EOB-DTPA (Primovist®, Bayer Healthcare, Berlin, Germany); implementation of preoperative portal embolization; arterial chemo-embolization before the operation. Afterwards, we also excluded tumors which were located on the lateral segment of the liver because of the effect on the apparent diffusion coefficient (ADC) from the heartbeat (11). Finally, 83 patients with 99 lesions were enrolled in the study. Sixty-eight patients had one tumor, 14 patients had two tumors, and one patient had three tumors. The participants consisted of 59 men and 24 women. The mean age was 67.0 years (range = 28–87 years). The pathological results of tumors were as follows: hepatocellular carcinoma (n = 34); metastatic liver tumor (n = 55); and miscellaneous (gallbladder cancer, intrahepatic cholangiocellular carcinoma, combined hepatocellular and cholangiocellular carcinoma, angiomyolipoma, and others) (n = 10). The underlying liver diseases were alcoholism (n = 8), hepatitis C (n = 10), and hepatitis B (n = 5). Thirty-three patients (39.4%) underwent preoperative chemotherapy; most patients were given the FOLFOX regimen (5-fluorouracil, folinic acid, and oxaliplatin) (12). The mean time interval between the operation and the MRI study was 28 days (range = 2–79 days).

MRI technique

All MRI examinations were performed using a 1.5 T superconductive MRI scanner (Avanto; Siemens, Erlangen, Germany) with a 32-channel body array coil. The maximum gradient of the system was 45 mT/m and the maximum slew rate was 200 T/m/s. T1-weighted (T1W) images were acquired under breath-holding, including in-phase and opposed-phase images, and the parameters were as follows: repetition/echo time (TR/TE) = 120/4.76, 2.38 ms; flip angle = 75°; 1 averaging; matrix = 320 × 224; parallel acquisition technique with reduction factor 2 using the generalized autocalibration partially parallel acquisition (GRAPPA) algorithm; slice thickness = 6 mm; slice gap = 1.2 mm; and acquisition time = 13 s. T2-weighted (T2W) images were acquired with Sampling Perfection with Application optimized Contrasts using different flip angle Evolution (SPACE) under breath synchronization using two-dimensional prospective acquisition correction (2D-PACE); the parameters were as follows: TR/TE = 1600/149 ms; flip angle = 120°; flip angle mode = constant; matrix = 320 × 227; 1.4 averaging; field of view (FOV) = 400 × 275; slice per slab = 64; slice oversampling = 12.5%; PAT factor 2 with the GRAPPA algorithm; bandwidth = 710 Hz/pixel; slice thickness = 3 mm. DWI was also acquired under breath synchronization using 2D-PACE; the parameters were as follows: TR/TE = 3000/71 ms; flip angle = 120°; matrix = 128 × 128; slice thickness = 6 mm; intersection gap = 1 mm; 6 signals acquired; FOV = 40 cm; 30 sections in 5–8 min; chemical shift selective method; PAT factor 2 with the GRAPPA algorithm; b-factor = 100 and 800 s/mm². The DWI motion-probing gradient pulses were placed in three orthogonal axes, and DWI was reconstructed by combining these three images. Quantitative ADC maps were created on a voxel-by-voxel basis using the software on the scanner. Dynamic Gd-EOB-DTPA–enhanced MRI was performed with a T1W 3D gradient echo sequence with fat saturation and volumetric interpolated breath-hold examination (VIBE). The sequence parameters were as follows: TR/TE = 3.96/1.79 ms; slice thickness = 2 mm; FOV = 400 mm; effective matrix size = 320 × 224; signal average = 1; acquisition time = 19 s; k-space trajectory = linear filling. Gd-EOB-DTPA (0.025 mmol/kg) was injected at a rate of 1 or 2 mL/s via the antecubital vein, followed by 40 mL of physiological saline at the same injection rate as the contrast media. Dynamic study included the hepatic arterial-dominant phase, portal-dominant phase, and the period 4 min after injection of the contrast material and hepatobiliary phase. The hepatobiliary phase was obtained at 20 min after injection of Gd-EOB-DTPA.

Evaluation of MRI and liver fibrosis

Two radiologists (with 4 and 25 years of experience, respectively) and a pathologist (with 17 years of experience) discussed where to set the ROIs on liver parenchyma to evaluate fibrosis, comparing MR images with the picture of resected specimen and pathology slides. After reaching consensus, the first ROI was set on ADC map at a distance of 5–10 mm from the tumor where fibrosis grade was assessed pathologically. We set a circular ROI not larger than 1 cm² on MRI (Fig. 1). The mean signal intensity (SI) was used as data. The second ROI was set almost on the same location of the first ROI, but at the hepatobiliary phase (SI liver). The third and fourth ROI were set at the erector spinae muscle and spleen at the hepatobiliary phase, respectively (SI muscle, SI spleen). The fifth ROI was set on the same location as the second ROI on the pre-enhanced MRI on T1W 3D-VIBE (SI pre-liver). The ROIs corresponding to each other were placed using the copy and paste function, and minor adjustment was performed carefully by hand to remove errors caused by breath movement, without changing the form and size of the ROIs. All ROIs were set at the same slice if possible, and were set avoiding the lateral segment of the liver and major vessels. The ratios of the third, fourth, and fifth ROI to the second ROI were calculated as the LMR (10,13), LSR (8), and CEI (14) as represented by the following formulas:

\begin{matrix} LMR = SI (liver) / SI (muscle) \\ LSR = SI (liver) / SI (spleen) \\ CEI = SI (liver) / SI (pre - liver) \end{matrix}

Fig. 1.

ROI setting. Five ROIs set on ADC map and hepatobiliary phase image. Three ROIs were set at near the tumor on ADC map (a), hepatobiliary phase (b), and pre-enhanced MRI (d). The two remaining ROIs were set at erector spine and spleen on hepatobiliary phase (c).

Liver fibrosis was classified pathologically into five stages according to the Inuyama criteria (15), as follows: no fibrosis (F0); portal fibrosis widening (F1); portal fibrosis widening with bridging fibrosis (F2); bridging fibrosis plus lobular distortion (F3); and cirrhosis (F4). The pathologist mentioned above performed the pathological evaluation. Then, 14, 53, 15, six, and 11 specimens were classified into stages F0, F1, F2, F3, and F4 fibrosis, respectively. The numbers of patients with liver specimens at fibrosis stage F0, F1, F2, F3, and F4 were 11, 29, 13, six, and nine, respectively. Another 14 patients had two resected specimens and one patient had three resected specimens. Those patients had the following combinations of fibrosis stages: two patients had one at F0 and one at F1, nine patients had both specimens at F1, two patients had one at F1 and one at F2, one patient had both at F4, and one patient had two at F1 and one at F0. Thirty-one patients had liver fibrosis F ≥ 2. Of these, 23 patients had underlying liver disease as mentioned above, and the remaining eight patients had no underlying liver disease. ADC value is known to be affected by fat deposition; therefore, the degree of fatty change (16) was also pathologically evaluated using the classification proposed by Brunt et al. (17). The numbers of specimens that were classified into grades 0–3 were 17, 70, 10, and two, respectively (Table 1). The effect of fat deposition on every parameter was evaluated only in the F1 fibrosis group because the populations of the other fibrosis groups were too small to evaluate the effect of fat deposition statistically.

Table 1.

Distribution of specimens according to degree of fatty change and fibrosis stage.

Degree of fatty change	Fibrosis stage
Degree of fatty change	F0	F1	F2	F3	F4
None	4	9	2	1	1
Mild	10	37	10	5	8
Moderate	0	5	3	0	2
Severe	0	2	0	0	0
Total	14	53	15	6	11

Statistical analysis

Data are represented as the mean ± standard deviation. Statistical analyses were mainly performed with SPSS version 22.0.0.0 (IBM Corp., Armonk, NY, USA). One-way analysis of variance (ANOVA) with Tukey post hoc comparison was used to correlate the parameters with histologic grade of fibrosis, and the unpaired t-test to compare the parameters of low-stage fibrosis (F0, F1, and F2) with those of high-stage fibrosis (F3 and F4). Pearson’s correlation was used to evaluate the effect of fat deposition on all parameters. A value of P < 0.05 was considered significant. Receiver operating characteristic curve (ROC) analysis was also performed to determine the sensitivity and specificity for diagnosing high-stage fibrosis. Statistical analyses of ROC were performed with EZR software (Saitama Medical Center, Jichi Medical University, http://www.jichi.ac.jp/saitama-sct/Saita maHP.files/statmedEN.html), which is a graphical user interface for R (The R Foundation for Statistical Computing, version 2.13.0).

Results

ADC value and fibrosis stage

ADC values were 0.85 ± 0.20, 0.82 ± 0.19, 0.81 ± 0.18, 0.70 ± 0.29, and 0.91 ± 0.16 × 10⁻³ mm²/s in fibrosis stages F0, F1, F2, F3, and F4, respectively. No significant differences were observed in ADC values among fibrosis stages (P = 0.320) (Fig. 2a), nor between high and low fibrosis grade (P = 0.634) (Fig. 2b).

Fig. 2.

ADC value and fibrosis stage. No significant differences were observed in ADC values among fibrosis stages (P = 0.320) (a), nor between high-grade and low-grade fibrosis (P = 0.634) (b).

CEI and fibrosis stage

CEI values were 2.17 ± 0.20, 2.30 ± 0.35, 2.15 ± 0.38, 2.11 ± 0.53, and 1.90 ± 0.35 in fibrosis stages F0, F1, F2, F3, and F4, respectively. CEI values showed significant differences among fibrosis stages (P = 0.013). A significant difference was observed between F1 and F4 (P = 0.014) (Fig. 3a). CEI values of low-stage fibrosis and high-stage fibrosis were 2.25 ± 0.34 and 2.00 ± 0.42, respectively, and a significant difference was observed (P = 0.04) (Fig. 3b).

Fig. 3.

CEI and fibrosis stage. CEI values showed significant differences among fibrosis stages (P = 0.013). A significant difference was observed between F1 and F4 (P = 0.014) (a). CEI values of low-stage fibrosis and high-stage fibrosis showed a significant difference (P = 0.04) (b).

LMR and fibrosis stage

LMR values were 3.16 ± 0.60, 3.31 ± 0.60, 2.93 ± 0.56, 2.72 ± 0.60, and 2.48 ± 0.49 in fibrosis stages F0, F1, F2, F3, and F4, respectively. LMR values showed significant differences among fibrosis stages (P < 0.001). Significant differences were observed between F0 and F4 (P = 0.047), and between F1 and F4 (P < 0.001) (Fig. 4a). LMR values of low-stage fibrosis and high-stage fibrosis were 3.21 ± 0.60 and 2.57 ± 0.53, respectively, and a significant difference was observed (P < 0.01) (Fig. 4b).

Fig. 4.

LMR and fibrosis stage. LMR values showed significant differences among fibrosis stages (P < 0.001). Significant differences were observed between F0 and F4 (P = 0.047), and between F1 and F4 (P < 0.001) (a). LMR values of low-stage fibrosis and high-stage fibrosis showed significant difference (P < 0.01) (b).

LSR and fibrosis stage

LSR values were 2.76 ± 0.77, 3.00 ± 0.75, 2.49 ± 0.56, 2.94 ± 0.54, and 2.24 ± 0.49 in fibrosis stages F0, F1, F2, F3, and F4, respectively. LSR values showed significant differences among fibrosis stages (P = 0.008). A significant difference was observed between F1 and F4 (P = 0.014) (Fig. 5a). LSR values of low-stage fibrosis and high-stage fibrosis were 2.86 ± 0.74 and 2.49 ± 0.60, respectively, and no significant difference was observed (P = 0.053) (Fig. 5b).

Fig. 5.

LSR and fibrosis stage. LSR values showed significant differences among fibrosis stages (P = 0.008). A significant difference was observed between F1 and F4 (P = 0.014) (a). LSR values of low-stage fibrosis and high-stage fibrosis showed no significant difference (P = 0.053) (b).

Comparison of ROC between LMR and CEI

The sensitivity and specificity, when a CEI of 2.05 or higher was considered to be low-stage fibrosis, were 76.5% and 75.6%, respectively. On the other hand, sensitivity and specificity, when an LMR of 2.80 or higher was considered to be low-stage fibrosis, were 82.4% and 75.6%, respectively. No significant difference was observed between CEI and LMR in diagnostic performance for distinguishing low-stage fibrosis and high-stage fibrosis (P = 0.185) (Fig. 6).

Fig. 6.

ROC curve between LMR and CEI. The sensitivity and specificity, when a CEI of 2.05 or higher was considered to be low-stage fibrosis, were 76.5% and 75.6%, respectively. On the other hand, sensitivity and specificity, when an LMR of 2.80 or higher was considered to be low-stage fibrosis, were 82.4% and 75.6%, respectively. No significant difference was observed between CEI and LMR in diagnostic performance for distinguishing low-stage fibrosis and high-stage fibrosis (P = 0.185).

Effect of fatty change on every parameter at stage 1 fibrosis

Table 1 shows the distribution of specimens according to degree of fat steatosis and fibrosis grade. The r-values of ADC, CEI, LMR, and LSR with fat deposition were r = –0.218, 0.116, −0.044, and −0.077, respectively, and only ADC had a significant correlation with fat deposition (P = 0.03) (Fig. 7).

Fig. 7.

Effect of fatty change at stage 1 fibrosis. Only ADC had a significant correlation with fat deposition (P = 0.03).

Discussion

This study revealed no significant correlation between ADC value and fibrosis stage. Several papers reported that significant correlation was observed (4,5,7), although very few papers reported no correlation (18). We supposed that the reasons why no significant correlation was observed between liver fibrosis and ADC value were as follows. First, the direct comparison method we adopted in our study was different from previous reports. Second, steatosis might have an effect on the ADC value. Poyraz et al. suggested that steatosis decreases ADC value because increased fat content of hepatocytes and extracellular fat accumulation reduce the interstitial space and restrict water diffusion (16,19). The results of this study are consistent with those of previous reports. Third, iron deposition might affect ADC value. ADC value was extremely low in a few cases, although with low fibrosis grade and steatosis grade. We supposed it to be the T2* shortening effect (20), because the signal intensity of liver parenchyma on in-phase imaging was lower than that of opposed-phase imaging. Actually, chronic liver disease can lead to iron overload (21), and this is critical for quantitative evaluation using ADC value.

LMR and CEI could differentiate high-grade fibrosis from low-grade fibrosis better than ADC value. Signal intensity at the hepatobiliary phase is affected by the number of functional hepatocytes. Therefore, the signal intensity of liver decreases in liver fibrosis (22). LMR and CEI have some advantages compared with ADC values, that is, steatosis, iron deposition, perfusion effect from neighboring major vessels, and heartbeat do not affect those parameters. The only disadvantage is that contrast media is necessary. LSR showed no significant difference between high-grade fibrosis and low-grade fibrosis. Nojiri et al. also reported that LSR is the worst parameter among some enhancement ratios (13). Some studies of elastography showed that spleen stiffness is comparable to liver stiffness (23,24). It indicates that spleen also changes pathologically along with the change in liver parenchyma. We supposed that spleen showed variable enhancement that was dependent on liver fibrosis and portal hypertension, and consequently LSR showed inaccurate data. On the other hand, muscle showed minimum enhancement and pathological change, and it is almost independent of liver disease. Therefore, LMR appeared to show consistent data.

LMR and CEI could differentiate high-grade fibrosis from low-grade fibrosis better than ADC, although the condition of the participants varied. Chemotherapy-associated liver injury is a different entity from viral or alcohol-induced liver injury. Several pathological changes, such as steatosis and sinusoidal dilatation, in chemotherapy-associated liver injury have been reported in several papers (25,26). Furthermore, fibrosis is associated with neoadjuvant 5-fluorouracil-based chemotherapy in combination with oxaliplatin (27). These pathological changes may affect the ADC and liver parenchymal enhancement. Actually, the present study clarified that ADC was affected by steatosis. Miscellaneous conditions as in the present study might lead to significant complexity. However, the usefulness of LMR and CEI was demonstrated.

There were some limitations in this study. First, some patients (40% of all participants) received chemotherapy and some bias might be included. Almost all of these patients received FOLFOX; however, the sinusoidal dilatation was not observed pathologically in all patients. Therefore, we believe that the effect of the chemotherapy was limited. Second, this study was retrospective and all the participants received surgery. This led to the small number of high-grade fibrosis specimens. This might include some bias. Furthermore, the sub-groups became small in number. However, we supposed that having sufficient specimens of liver parenchyma to evaluate the fibrosis and direct comparison were advantages of this study.

In conclusion, LMR and CEI were both adequate surrogate parameters to distinguish high-stage fibrosis from low-stage fibrosis, and these values are available in routine clinical examinations. On the other hand, the ADC value was ineffective in predicting the fibrosis grade. Our results included the potential bias from hepatic steatosis, which affected the ADC value of some patients in the present study.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Wong

Bennett

Koff

et al.

Pretreatment evaluation of chronic hepatitis C: risks, benefits, and costs. JAMA 1998; 280: 2088–2093.

Regev

Berho

Jeffers

et al.

Sampling error and intraobserver variation in liver biopsy in patients with chronic HCV infection. Am J Gastroenterol 2002; 97: 2614–2618.

Howlett

Drinkwater

Lawrence

et al.

Findings of the UK national audit evaluating image-guided or image-assisted liver biopsy. Part II. Minor and major complications and procedure-related mortality. Radiology 2013; 266: 226–235.

Luciani

Vignaud

Cavet

et al.

Liver cirrhosis: intravoxel incoherent motion MR imaging–pilot study. Radiology 2008; 249: 891–899.

Bakan

Inci

Bakan

et al.

Utility of diffusion-weighted imaging in the evaluation of liver fibrosis. Eur Radiol 2012; 22: 682–687.

Sandrasegaran

Akisik

Lin

et al.

Value of diffusion-weighted MRI for assessing liver fibrosis and cirrhosis. Am J Roentgenol 2009; 193: 1556–1560.

Lewin

Poujol-Robert

Boelle

et al.

Diffusion-weighted magnetic resonance imaging for the assessment of fibrosis in chronic hepatitis C. Hepatology 2007; 46: 658–665.

Goshima

Kanematsu

Watanabe

et al.

Gd-EOB-DTPA-enhanced MR imaging: prediction of hepatic fibrosis stages using liver contrast enhancement index and liver-to-spleen volumetric ratio. J Magn Reson Imaging 2012; 36: 1148–1153.

Nishie

Asayama

Ishigami

et al.

MR prediction of liver fibrosis using a liver-specific contrast agent: Superparamagnetic iron oxide versus Gd-EOB-DTPA. J Magn Reson Imaging 2012; 36: 664–671.

10.

Watanabe

Kanematsu

Goshima

et al.

Staging hepatic fibrosis: comparison of gadoxetate disodium-enhanced and diffusion-weighted MR imaging–preliminary observations. Radiology 2011; 259: 142–150.

11.

Kwee

Takahara

Niwa

et al.

Influence of cardiac motion on diffusion-weighted magnetic resonance imaging of the liver. MAGMA 2009; 22: 319–325.

12.

Chen

Liu

et al.

Preoperative evaluation of the histological grade of hepatocellular carcinoma with diffusion-weighted imaging: a meta-analysis. PLoS One 2015; 10: e0117661–e0117661.

13.

Nojiri

Kusakabe

Fujiwara

et al.

Noninvasive evaluation of hepatic fibrosis in hepatitis C virus-infected patients using ethoxybenzyl-magnetic resonance imaging. J Gastroenterol Hepatol 2013; 28: 1032–1039.

14.

Tamada

Ito

Higaki

et al.

Gd-EOB-DTPA-enhanced MR imaging: evaluation of hepatic enhancement effects in normal and cirrhotic livers. Eur J Radiol 2011; 80: e311–316.

15.

Ichida

FTT

Omata

Ichida

et al.

New Inuyama classification; new criteria for histological assessment of chronic hepatitis. Int Hepatol Commun 1996; 6: 112–119.

16.

Poyraz

Onur

Kocakoc

et al.

Diffusion-weighted MRI of fatty liver. J Magn Reson Imaging 2012; 35: 1108–1111.

17.

Moore

Khatri

Madhuranthakam

et al.

Added value of diffusion-weighted acquisitions in MRI of the abdomen and pelvis. Am J Roentgenol 2014; 202: 995–1006.

18.

Boulanger

Amara

Lepanto

et al.

Diffusion-weighted MR imaging of the liver of hepatitis C patients. NMR Biomed 2003; 16: 132–136.

19.

Hansmann

Hernando

Reeder

. Fat confounds the observed apparent diffusion coefficient in patients with hepatic steatosis. Magn Reson Med 2013; 69: 545–552.

20.

Chandarana

Mussi

et al.

The effect of liver iron deposition on hepatic apparent diffusion coefficient values in cirrhosis. Am J Roentgenol 2012; 199: 803–808.

21.

Verloh

Haimerl

Zeman

et al.

Multivariable analysis of clinical influence factors on liver enhancement of Gd-EOB-DTPA-enhanced 3T MRI. Rofo 2015; 187: 29–35.

22.

Tschirch

Struwe

Petrowsky

et al.

Contrast-enhanced MR cholangiography with Gd-EOB-DTPA in patients with liver cirrhosis: visualization of the biliary ducts in comparison with patients with normal liver parenchyma. Eur Radiol 2008; 18: 1577–1586.

23.

Chen

Kuo

et al.

Early response of hepatocellular carcinoma to transcatheter arterial chemoembolization: choline levels and MR diffusion constants–initial experience. Radiology 2006; 239: 448–456.

24.

Nakanishi

Chuma

Hige

et al.

Relationship between diffusion-weighted magnetic resonance imaging and histological tumor grading of hepatocellular carcinoma. Ann Surg Oncol 2012; 19: 1302–1309.

25.

Aloysius

Zaitoun

Beckingham

et al.

The pathological response to neoadjuvant chemotherapy with FOLFOX-4 for colorectal liver metastases: a comparative study. Virchows Arch 2007; 451: 943–948.

26.

Aloia

Sebagh

Plasse

et al.

Liver histology and surgical outcomes after preoperative chemotherapy with fluorouracil plus oxaliplatin in colorectal cancer liver metastases. J Clin Oncol 2006; 24: 4983–4990.

27.

Kandutsch

Klinger

Hacker

et al.

Patterns of hepatotoxicity after chemotherapy for colorectal cancer liver metastases. Eur J Surg Oncol 2008; 34: 1231–1236.

Prediction of high-stage liver fibrosis using ADC value on diffusion-weighted imaging and quantitative enhancement ratio at the hepatobiliary phase of Gd-EOB-DTPA–enhanced MRI at 1.5 T

Abstract

Background

Purpose

Material and Methods

Results

Conclusion

Keywords

Introduction

Material and Methods

Patients

MRI technique

Evaluation of MRI and liver fibrosis

Statistical analysis

Results

ADC value and fibrosis stage

CEI and fibrosis stage

LMR and fibrosis stage

LSR and fibrosis stage

Comparison of ROC between LMR and CEI

Effect of fatty change on every parameter at stage 1 fibrosis

Discussion

Footnotes

Declaration of Conflicting Interests

Funding

References