- Email: [email protected]
Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis
Clinical Radiology xxx (xxxx) xxx
Contents lists available at ScienceDirect
Clinical Radiology journal homepage: www.clinicalradiologyonline.net
Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis X. Li a, H. Liu b, R. Wang a, J. Yang a, c, Y. Zhang a, **, 1, C. Li a, c, *, 1 a
Department of Diagnostic Radiology, The First Affiliated Hospital of Xi’an Jiaotong University, Xi’an, Shaanxi Province, China b Department of Radiology, The Third Hospital of Hebei Medical University, Shijiazhuang, Hebei Province, China c Department of Biomedical Engineering, The Key Laboratory of Biomedical Information Engineering of the Ministry of Education, School of Life Science and Technology, Xi’an Jiaotong University, Xi’an, China
art icl e i nformat ion Article history: Received 28 March 2019 Accepted 12 November 2019
AIM: To identify and evaluate the efficiency of the most commonly used parameters applied to gadoxetate disodium (Gd-EOB-DTPA)-enhanced magnetic resonance imaging (MRI) for liver fibrosis (LF) staging. MATERIALS AND METHODS: Literature searches of PubMed, Web of Science, Embase, and MEDLINE databases from January 2004 to August 2018 were conducted. The applied parameters during imaging were noted and summarised. Studies using the most commonly used parameter were included. Extractive data were combined as pooled weighted mean difference (WMD) to determine the benefits in LF staging. The pooled sensitivity, specificity, and summary receiver operating characteristics (SROC) curve were calculated. RESULTS: Among 57 relevant studies, the contrast enhancement index (CEI) was a relatively commonly used parameter. It was calculated as SIpost/SIpre, where SIpost and SIpre are the liverto-muscle signal intensity ratios in the hepatocyte phase and pre-enhanced images, respectively. Six studies were included. F0 was regarded as normal liver, F1 as mild LF, F2 as moderate LF, and F3 and F4 as advanced LF. Comparisons of WMD revealed significant differences between F1e2 and F3e4. For stage F1, the pooled sensitivity, specificity, and areas under SROC curve were 0.58, 0.84, and 0.85, respectively; the corresponding values for F2 were 0.57, 0.68, and 0.76, while those for F3 were 0.61, 0.75, and 0.72. CONCLUSION: The methodology and parameters used for Gd-EOB-DTPA-enhanced MRI for LF staging are diverse, but the CEI was a relatively common parameter. Overall, there is evidence to support use of CEI, but more evidence from larger studies is needed. Ó 2019 The Royal College of Radiologists. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
* Guarantor and correspondent: C. Li, Department of Diagnostic Radiology, The First Affiliated Hospital of Xi’an Jiaotong University, Yanta West Road 277, Xi’an, Shaanxi, China. Tel.: þ86-15389042141. ** Guarantor and correspondent: Y. Zhang, Department of Diagnostic Radiology, The First Affiliated Hospital of Xi’an Jiaotong University, Yanta West Road 277, Xi’an, Shaanxi, China. Tel.: þ86-18991232590. E-mail addresses: [email protected] (Y. Zhang), [email protected] (C. Li). 1 These authors contributed equally to this work. https://doi.org/10.1016/j.crad.2019.11.001 0009-9260/Ó 2019 The Royal College of Radiologists. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
2
X. Li et al. / Clinical Radiology xxx (xxxx) xxx
Introduction Liver fibrosis (LF) involves the accumulation of an extracellular matrix resulting in fibrogenesis, which is created in response to acute or chronic liver injury. Fibrosis progresses at variable rates and eventually develops into cirrhosis, which in turn leads to hepatic dysfunction.1,2 It is reported that early LF is manageable, its progression can be controlled, and it is reversible in a large percentage of patients3; however, the probability of reversing advanced LF is low,4,5 and so early detection of LF has important clinical implications. A liver biopsy is the reference-standard technique for diagnosing LF; however, this procedure is limited by its invasive nature, sampling errors associated with small tissue samples, and the risk of complications.6 Several noninvasive diagnostic methods, including direct/indirect serum markers, serum panels, and imaging techniques, are now being utilised to avoid the shortcomings of biopsies.7,8 Conventional magnetic resonance imaging (MRI) is one of the non-invasive methods, but this provides data only on the changes in liver morphology. Some magnetic resonance techniques can provide information on liver function in addition to morphology, such as contrast-enhanced MRI.9,10 Gadoxetate disodium (Gd-EOB-DTPA; marketed as Primovist in Europe and Eovist in the United States, Bayer HealthCare) is a liver-specific contrast agent, and the uptake of Gd-EOB-DTPA by hepatocytes mainly occurs via the organic anion transporter polypeptides (OATP)1B1 and B3 located at the sinusoidal membrane. Biliary excretion occurs via the multidrug resistance-associated proteins, MRP2, at the canalicular membrane. Gd-EOB-DTPA can therefore be utilised for evaluating liver function and the staging of LF.11 The application of T1-weighted MRI (T1WI) and relaxation time12e14 or histogram15,16 analysis revealed a correlation between the stage of LF and the signal intensity (SI) of the liver during Gd-EOB-DTPA-enhanced MRI17e19; however, current research uses various parameters for assessing liver function and thus lacks uniformity, which limits the further application of this technique.20,21 This present systematic review and meta-analysis was conducted with the aim of detecting the parameters used in Gd-EOB-DTPA-enhanced MRI that are most suitable for evaluating LF. The diagnostic performance of the selected parameters was also determined.
Materials and methods The present meta-analysis was performed in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement. No particular ethical considerations were required as this meta-analysis did not involve any identifiable patient information.
Literature search criteria A literature search was performed of the PubMed, Web of Science, Embase, and MEDLINE databases for articles
published from January 2004 (start of clinical application) to August 2018. Articles pertaining to the diagnostic accuracy of Gd-EOB-DTPA-enhanced MRI for staging LF were considered relevant for inclusion in the analysis. Crossreferencing from the reference lists in the included original articles and associated review articles was also performed to identify additional eligible articles. Search query terms were developed based on specific keywords (“gadoxetic acid,” “gadoxetate,” “Gd-EOB-DTPA,” and “hepatic fibrosis”) and medical subject headings (“liver fibrosis” and “magnetic resonance imaging”). The following specific search query was designed for retrieving relevant data from the four databases: (“liver fibrosis” OR “hepatic fibrosis”) AND (“gadoxetic acid” OR “Gd-EOB-DTPA” OR “gadoxetate”) AND (“magnetic resonance imaging” OR “MRI”).
Inclusion and exclusion criteria The relevance of each article was assessed based on its title and abstract, followed by a full-text evaluation. Articles were included based on the pre-specified inclusion/exclusion criteria described below. The database search was performed by three investigators independently, and any ambiguities were resolved through discussion. The following inclusion criteria were applied: (1) objective: to determine the diagnostic performance of GdEOB-DTPA-enhanced MRI in LF staging; (2) patients: patients with pathologically staged LF or healthy volunteers with no history of chronic hepatic disease. All of them underwent Gd-EOB-DTPA-enhanced MRI; (3) parameters: summary of the parameters used for Gd-EOB-DTPAenhanced MRI, and the study that applied the most frequently used parameter was included; (4) study type: randomised controlled trials or original research; (5) language: article written in English. The exclusion criteria were as follows: (1) incompleteness of data preventing extraction and analysis; (2) repeated articles reporting on the same population.
Data extraction The studies identified in the databases were screened manually to avoid repetition, and checked manually for relevancy by three independent investigators who also performed the database searches, and then independently extracted the data related to the most frequently used parameter. In addition to the baseline characteristics, the investigators also extracted the results, which included the mean and standard deviation (SD) values for the each stage of LF, and also the cut-off value, true-positive (TP) value, false-positive (FP) value, false-negative (FN) value, and truenegative (TN) value for each compared group. Histopathology results were considered the reference standard in all of the included studies. Consistent with published research, F0 was regarded as a normal liver, and F1 as mild LF, F2 as moderate LF, and F3 and F4 as advanced LF.22,23 The data for stages F0, F1, F2, and F3e4 were therefore extracted and analysed.
Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
X. Li et al. / Clinical Radiology xxx (xxxx) xxx
Quality assessment The overall methodological quality and bias of the included articles were assessed using the revised tool for the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2), which evaluates the risk of bias for four domains and the clinical applicability for three domains of the study characteristics.24
Statistical analysis All of the data were arranged into two sets: (1) demography and (2) diagnostic performance. Data from the included studies were combined and expressed as pooled weighted mean difference (WMD) values with 95% CIs (confidence intervals) using RevMan (Review Manager, version 5.3, Cochrane Collaboration) to determine the benefit of using the most frequently used parameter for staging LF. Studies were weighted by the inverse variance, and a fixed-effects model was initially used. Owning to different stages of LF being analysed in different studies, the pooled mean and SD of the most frequently used parameter in each compared group were calculated. The pooled sensitivity and specificity with 95% CIs for the most frequently used parameter were calculated based on different stages of LF using MetaDiSc (version 1.4). The summary receiver operating characteristics (SROC) curve was constructed to perform a comprehensive evaluation of the diagnostic performance of the most frequently used parameter for LF staging. Heterogeneity was evaluated using Cochrane’s Q-test, where p<0.1 indicates the presence of heterogeneity. The I2 value was used to quantitatively evaluate heterogeneity, with a fixed-effects model used for meta-analysis when I2<25% and a random-effects model applied when I2¼25e50%. When I2>50%, the sources of heterogeneity were analysed first, and then a random-effects model was used once it was determined that there was no obvious clinical heterogeneity and the sources of heterogeneity could not be found.
Results Study selection and characteristics The literature search of the four databases yielded 267 records. Screening for duplicates resulted in the removal of 73 articles, and a further 137 studies were excluded for the reasons listed in Fig 1. Among the remaining 57 potentially relevant studies, the contrast enhancement index (CEI) was the most frequently used parameter, being applied in 10 studies (Table 1). The CEI was calculated as SIpost/SIpre, where SIpost and SIpre are the liver-to- paraspinal muscle SI ratios in T1WI of hepatocyte-phase images and preenhanced images, respectively. Six studies17,18,25e28 were finally included in the analysis after excluding a further four studies (Fig 1). Hepatocyte-phase images were obtained at 20 min after injection of Gd-EOB-DTPA in those studies.
3
The characteristics of the included studies are presented in Table 2. Nojiri et al. randomly divided all subjects into an estimation model and a validation model, and only data for the former model26 were used in the present study. The present meta-analysis included 695 patients with mean ages of 42.3e71 years in the various studies, and men:women gender ratio of 18:7. Of the six included studies, five studies used 1.5 T MRI and one study used 3 T MRI. Three of the six included studies applied a liver acquisition with volume acceleration (LAVA) sequence. Two studies did not use the LAVA sequence, but the repetition time (TR) and echo time (TE) values of the sequences used were similar. Only one study used higher TR and TE. Histopathology was considered as the reference standard in all of the studies, which used different equipment. The METAVIR scoring system was applied in five studies, with the sixth study employing the Batts and Ludwig system, but the detailed scoring standard was similar in all six studies.
Quality assessment The results of the quality assessment performed using the QUADAS-2 tool are summarised in Fig 2, with the risk of bias presented on the left and concerns regarding applicability presented on the right. A substantial risk of bias regarding the index test was observed, probably due to the cut-off value not being pre-specified. A considerable risk of bias was also observed for patient selection, since for most studies it was not reported whether a consecutive or random sample of patients was enrolled; however, there were few applicability concerns regarding patient selection, the index test, or the reference standard for all of the studies.
Diagnostic performance of Gd-EOB-DTPA-enhanced MRI The mean CEI was used for staging LF. The diagnostic performances of Gd-EOB-DTPA-enhanced MRI in patients with LF for stage F1 versus F0 and stage F3e4 versus F1e2 are presented in Fig 3a and 3b, respectively. For LF of stage F1 compared with stage F0, the pooled WMD was e0.09 (95% CI¼e0.20 to 0.02, p¼0.11). A high degree of heterogeneity was found among the relevant studies: p¼0.03, I2¼67%. The mean CEI of Gd-EOB-DTPA-enhanced MRI in all of the individual studies differed between LF of stage F3-4 and stage F1-2: WMD¼e0.11, 95% CI¼e0.16 to e0.06, and p<0.0001. A moderate heterogeneity was found among the relevant studies: p¼0.11, I2¼50%.
Diagnostic accuracy of Gd-EOB-DTPA-enhanced MRI Detailed data on the cut-off value, TP, FP, FN, and TN for each compared group are presented in Table 3. Fig 4 presents the pooled sensitivity and specificity values of GdEOB-DTPA-enhanced MRI in LF staging. In patients with LF stage 1 (F0 versus F1e4), the pooled sensitivity and specificity were 0.58 (95% CI¼0.54 to 0.63, I2 ¼ 94%) and 0.84 (95% CI¼0.73 to 0.92, I2 ¼ 0.0%), respectively (Fig 4a). The pooled sensitivity of Gd-EOB-DTPA-enhanced MRI was 0.57 (95% CI¼0.53 to 0.62, I2 ¼ 90.0%), which was lower in
Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
4
X. Li et al. / Clinical Radiology xxx (xxxx) xxx
Figure 1 Flow diagram of study selection. Some relative studies applied several parameters, so the number of applied parameter was not consistent with the number of relative studies. SIpost signal intensity in hepatobiliary phase. SIpre pre-enhanced signal intensity. Table 1 Studies using the contrast enhancement index (CEI) for liver fibrosis staging. First author
Title
Chen
Clinical and histological implications of delayed hepatobiliary enhancement on magnetic resonance imaging with gadolinium ethoxybenzyl diethylenetriaminepentaacetic acid Comparison of magnetic resonance elastography and gadoxetate disodium-enhanced magnetic resonance imaging for the evaluation of hepatic fibrosis Gd-EOB-DTPA-enhanced MR imaging: prediction of hepatic fibrosis stages using liver contrast enhancement index and liver-to-spleen volumetric ratio Noninvasive assessment of hepatic fibrosis using gadoxetate-disodium-enhanced 3T MRI Hepatic enhancement of Gd-EOB-DTPA enhanced 3 tesla MR Imaging: assessing severity of liver cirrhosis Staging liver fibrosis by using liver-enhancement ratio of gadoxetic acid-enhanced MR imaging: comparison with aspartate aminotransferase-to-platelet ratio index Noninvasive evaluation of hepatic fibrosis in hepatitis C virus-infected patients using ethoxybenzylmagnetic resonance imaging Three-tesla magnetic resonance elastography for hepatic fibrosis: comparison with diffusion-weighted imaging and gadoxetic acid-enhanced magnetic resonance imaging Staging hepatic fibrosis: comparison of gadoxetate disodium-enhanced and diffusion-weighted MR imagingdpreliminary observations Comparison of the efficacy of Gd-EOB-DTPA enhanced magnetic resonance imaging and magnetic resonance elastography in the detection and staging of hepatic fibrosis
Choi
Goshima
Jang Lee Motosugi
Nojiri
Park
Watanabe
Wu
patients with LF 2 (F0e1 versus F2e4) than in patients with LF 1. The pooled specificity in patients with LF 2 was 0.68 (95% CI¼0.61 to 0.75, I2 ¼91.1%; Fig 4b). The pooled sensitivity increased to 0.61 (95% CI¼0.57 to 0.66, I2 ¼ 66.8%) in patients with LF 3 (F0e2 versus F3e4), with a pooled specificity of 0.75 (95% CI¼0.69 to 0.79, I2 ¼ 81.3%; Fig 4c). An overall trend of the pooled sensitivity increasing with the LF stage was observed. The area under the SROC curve (AUC) for the CEI when evaluating the LF stage was 0.85 for F0 versus F1e4, 0.76 for F0e1 versus F2e4, and 0.72 for F1e2 versus F3e4 (Fig 5). The Q* values for the corresponding AUCs were 0.78, 0.70, and 0.67, respectively.
Heterogeneity analysis and publication bias The threshold effect is one of the most important causes of heterogeneity in diagnostic tests.29 Due to the small number of studies included in each group in the present study, threshold, meta-regression, and sensitivity analyses could not be performed. The inclusion of <10 studies further meant that the statistical power for calculating publication bias was very low, and so this was not assessed or plotted.
Discussion To the authors’ knowledge, this was the first metaanalysis on selecting the most commonly used parameters in Gd-EOB-DTPA-enhanced MRI for staging LF. The present systematic review and meta-analysis found that the CEI was
Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
X. Li et al. / Clinical Radiology xxx (xxxx) xxx
5
Table 2 Basic characteristics of included literatures. Study
Year Study design N
Age, mean (SD/Scope)
Sex (M/F)
Equipment
Chen
2012 Prospective
42.3 (10.6)
39/17
1.5 T Siemens 209
Choi
2013 Retrospective 173 57.2 (10.4)
129/44 1.5 T GE
Jang
2013 Retrospective 113 56.0 (12.5)
73/40
56
TR (ms) TE (ms) Reference Scoring standard system
ROI ROI location number
4.8
Pathology METAVIR 3
4.2
1.9
Pathology METAVIR 3
3.0 T GE
4.0
2.0
Motosugia 2011 Retrospective 100 64.6 (28e83) ND Nojirib 2013 Prospective 149 71 (8) 99/50
1.5 T GE 1.5 T Philips
ND 4.2
ND 2.0
Pathology Batts and 4 Ludwig system Pathology METAVIR 3 Pathology METAVIR 3
Wu
1.5 T Siemens 3.7
1.4
Pathology METAVIR 1
2017 Retrospective 104 60.6 (10.6)
87/17
Left, right anterior, and right posterior lobe Left, right anterior, and right posterior lobe Right anterior, right posterior, left lateral, and left medial segments Right lobe Left, right anterior, and right posterior lobe Right lobe
TR, repetition time TE, echo time; ND not documented; ROI, region of interest. a This study used liver acquisition with volume acceleration sequence, consistent with studies by Jang YJ et al. and Choi YR et al. b This study randomly divided into model estimation and model validation, and only model estimation data were used in present.
Figure 2 Quality assessments with QUADAS-2. The substantial risk of bias derived from index test (left part) and the applicability concerns were low in all the studies (right part).
Figure 3 Pooled WMD in CEI of Gd-EOB-DTPA enhanced MRI. (a) Comparison of stage F0 with F1. (b) Comparison of stage F1e2 and F3e4.
the most commonly used parameter in Gd-EOB-DTPAenhanced MRI for evaluating liver function. Further, the WMD between the CEI at different stages showed that it
could be used for the staging of LF. The AUCs of the different compared groups revealed obvious differences between the normal and fibrotic liver, mild and moderate LF, and
Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
6
X. Li et al. / Clinical Radiology xxx (xxxx) xxx
Table 3 Detail data of included studies for pooled sensitivity and specificity. Stages
First author
Cut-off
TP
FP
FN
TN
Se (%)
Sp (%)
F0 versus F1e4
Chen Choi Jang Motosugi Wu Chen Choi Jang Motosugi Wu Chen Chen Choi Jang Motosugi Wu Chen
ND 1.613 1.322 1.91 1.75 ND 1.609 1.3169 1.76 ND 1.57 ND 1.637 1.3173 1.76 ND 1.55
20 69 44 73 75 26 63 46 50 69 44 4 66 38 43 60 38
1 3 1 4 1 25 6 1 14 6 3 5 23 4 21 15 6
27 81 56 11 26 0 73 46 17 41 45 10 38 29 14 31 34
8 20 12 12 2 5 31 20 19 33 11 37 47 42 22 43 26
42.6 46.0 44.0 86.9 74.3 100 46.3 50.0 74.6 62.7 49.4 28.6 62.8 56.7 75.4 65.9 52.8
88.9 85.0 92.3 75.0 66.7 16.7 82.4 93.5 56.0 84.6 80.0 88.1 68.2 91.3 50.0 74.1 81.3
F0e1 versus F2e4
F0e2 versus F3e4
TP, true-positive value; FP, false-positive value; FN, false-negative value; TN, true-negative value; Se, sensitivity; Sp, specificity; ND, not documented.
moderate and advanced LF, thus indicating it can be used to evaluate the LF stage. The published literature includes multiple MRI parameters for staging LF (Fig 1). There are numerous parameters related to Gd-EOB-DTPA-enhanced MRI, such as the CEI, liver-to-spleen SI ratio, post-contrast to pre-contrast ratio of the liver SI, liver-to-intervertebral disc SI ratio,26 liver-tospine SI ratio,25 and texture analysis parameters.30 In addition, the following parameters are also used to assess liver function, such as the arterial blood flow and slope derived from semi-quantitive perfusion MRI include,31 and the T1 relaxation time from T1 mapping13,14,19; however, these parameters have been used only a few times, and they have also been used inconsistently across studies. Among all of the parameters mentioned above, the CEI is the only one that has been evaluated comprehensively in multiple studies. The CEI was therefore chosen as the evaluable parameter for determining the diagnostic performance of Gd-EOB-DTPA-enhanced MRI for staging LF. Gd-EOB-DTPA is a liver-specific MRI contrast agent that is used to produce liver-enhanced images that reflect both liver perfusion and the condition of normal liver cells.11 After being injected intravenously, Gd-EOB-DTPA is gradually taken up by hepatocytes, and eventually is excreted via the biliary pathway and kidneys. Decreases in hepatic enhancement in hepatocyte-phase images suggest the impairment of Gd-EOB-DTPA uptake by the liver, whereas the prolongation of liver enhancement suggests that GdEOB-DTPA excretion into bile is also impaired.32 This impaired Gd-EOB-DTPA uptake by the liver would be caused by a decrease in the number of normal hepatocytes or a decrease in Gd-EOB-DTPA uptake by hepatocytes, and impaired Gd-EOB-DTPA excretion may be caused by hepatocyte dysfunction.32 All of these conditions are associated with a progressive increase in inflammation and the accumulation of extracellular matrix during LF that would influence the function, perfusion, vascular permeability, and extracellular diffusion of the liver.33 The change in liver SI
after injecting Gd-EOB-DTPA would be mainly related to liver function, but it also appears to be related to several other factors such as individual indocyanine green clearance and plasma bilirubin levels.34 Due to the influence of the liver function in individual patients and the possibility of other liver diseases being present, the liver SI ratio of before and after enhancement was lacked uniform comparison criteria when selected as an index for evaluating liver function. The above-described situation meant that previous studies have used SI values from various other tissues as reference values for achieving relative objectivity when assessing liver function. Among these various SI parameters, the SI of paraspinal muscle is more stable than that for other organs such as the spleen or liver; this is because the paraspinal muscle SI is minimally affected by age and liver function.35 Moreover, Noren et al. demonstrated that the CEI had a better diagnostic accuracy than the SI ratio in one phase.20 Therefore, since it corresponds to the ratio of SI before and after enhancement, the CEI could be a better index for evaluating the stage of LF. Previous research has also demonstrated that the CEI is useful for predicting both the entire and the regional hepatic functional reserves,36,37 and it might also be useful in evaluating the regional LF. Future prospective multicentre studies should further explore the diagnostic performance of Gd-EOB-DTPA-enhanced MRI. Xiao et al.38 conducted a systemic review and metaanalysis to compare the diagnostic accuracy of laboratory tests, shear-wave elastography (SWE), and magnetic resonance elastography (MRE) for assessments of LF. Their results suggested that MRE and SWE have the highest diagnostic accuracy for staging fibrosis; however, MRE is not used widely, and the sequence parameters applied have not been consistent.39 In addition, successful SWE imaging is dependent on the acquisition of clear two-dimensional images, and so the imaging success rate is low for patients with severe cirrhosis. Moreover, the measured values are affected by various factors such as the skill of the operator, external pressure, respiration, and food intake.40,41 T1 mapping was another method used to stage LF; however, the scan sequence are diverse, such as look-locker sequences, modified look-locker inversion recovery (MOLLI) sequence, volumetric T1 mapping with fast multisection B1 inhomogeneity correction,42,43 and many factors limit its clinical application, such as relatively long acquisition time, the influence of magnetic field inhomogeneity, and the inability to perform a full liver scan, etc. The use of Gd-EOB-DTPA-enhanced MRI to evaluate LF provides not only morphological information but also information on the liver function, and also allows the evaluation of focal hepatic lesions. We conducted this metaanalysis in order to detect the evaluable parameters used in Gd-EOB-DTPA-enhanced MRI for staging LF and to determine the diagnostic performance of the selected parameters. Future prospective studies should further explore the diagnostic performance of different technologies in staging LF. The present study was subject to some limitations. Firstly, six studies were included in the evaluation.
Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
X. Li et al. / Clinical Radiology xxx (xxxx) xxx
7
Figure 4 Pooled sensitivity and specificity of CEI in Gd-EOB-DTPA enhanced MRI. (a) Comparison of stage F0 with F1e4. (b) Comparison of stage F0e1 with F2e4. (c) Comparison of stage F0e2 with F3e4. Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
8
X. Li et al. / Clinical Radiology xxx (xxxx) xxx
(Indegene, Bangalore, India) for providing medical writing support and technical assistance in the development of this manuscript.
Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.crad.2019.11.001.
References
Figure 5 Summary receiver operating characteristics (SROC) curve of CEI.
Considering the limited number of studies included, regression analysis was not undertaken to evaluate the impact of relevant factors on the results. CEI may be impacted by different field strengths and imaging parameters, so the study using 3 T MRI and the study using higher TR and TE values were removed from the analysis. The results show that removal of these studies had little effect on the meta-analysis results (Electronic Supplementary Material). Secondly, retrospective studies have also been included, which will inevitably reduce the level of evidence. This is because there are too few prospective studies that meet the inclusion criteria, so further prospective studies are needed for liver function assessment. Thirdly, our analysis was limited to comparisons of clinically relevant studies (F0 versus F1e4, F0e1 versus F2e4, and F1e2 versus F3e4) due to a lack of studies comparing individual periods. We compare liver function in groups rather than in pairs. There are two reasons for this: one is that only one of the six included studies compared liver function across all stages. The second is that clinicians tend to obtain differences in liver function between adjacent stages rather than differences between all stages. In conclusion, the methodology and parameters applied in published studies for LF staging using Gd-EOB-DTPAenhanced MRI are diverse, but the CEI was a relatively common parameter. Overall, there is evidence to support use of CEI in LF staging, but more evidence from larger studies is needed.
Conflict of interest The authors declare no conflict of interest.
Acknowledgements This work was supported by the Shaanxi Provincial Key Research and Development Program (2019SF-007). The authors thank Mr Karan Sharma and Dr Amit Bhat
1. Duval F, Moreno-Cuevas JE, Gonzalez-Garza MT, et al. Liver fibrosis and mechanisms of the protective action of medicinal plants targeting inflammation and the immune response. Int J Inflamm 2015;2015:943497. 2. Ghany MG, Strader DB, Thomas DL, et al. American Association for the Study of Liver D. Diagnosis, management, and treatment of hepatitis C: an update. Hepatology 2009;49(4):1335e74. 3. Czaja AJ. Hepatic inflammation and progressive liver fibrosis in chronic liver disease. World J Gastroenterol 2014;20(10):2515e32. 4. Afdhal NH. Diagnosing fibrosis in hepatitis C: is the pendulum swinging from biopsy to blood tests? Hepatology 2003;37(5):972e4. 5. Kim AI, Saab S. Treatment of hepatitis C. Am J Med 2005;118(8):808e15. 6. Saleh HA, Abu-Rashed AH. Liver biopsy remains the gold standard for evaluation of chronic hepatitis and fibrosis. J Gastrointestin Liver Dis 2007;16(4):425e6. 7. Lurie Y, Webb M, Cytter-Kuint R, et al. Non-invasive diagnosis of liver fibrosis and cirrhosis. World J Gastroenterol 2015;21(41):11567e83. 8. Papastergiou V, Tsochatzis E, Burroughs AK. Non-invasive assessment of liver fibrosis. Ann Gastroenterol 2012;25(3):218e31. 9. Chung SR, Lee SS, Kim N, et al. Intravoxel incoherent motion MRI for liver fibrosis assessment: a pilot study. Acta Radiol 2015;56(12):1428e36. 10. Li Z, Sun J, Chen L, et al. Assessment of liver fibrosis using pharmacokinetic parameters of dynamic contrast-enhanced magnetic resonance imaging. J Magn Reson Imaging 2016;44(1):98e104. 11. Van Beers BE, Pastor CM, Hussain HK. Primovist, eovist: what to expect? J Hepatol 2012;57(2):421e9. 12. Okada M, Murakami T, Yada N, et al. Comparison between T1 relaxation time of Gd-EOB-DTPA-enhanced MRI and liver stiffness measurement of ultrasound elastography in the evaluation of cirrhotic liver. J Magn Reson Imaging 2015;41(2):329e38. 13. Ding Y, Rao SX, Zhu T, et al. Liver fibrosis staging using T1 mapping on gadoxetic acid-enhanced MRI compared with DW imaging. Clin Radiol 2015;70(10):1096e103. 14. Yang L, Ding Y, Rao S, et al. Staging liver fibrosis in chronic hepatitis B with T1 relaxation time index on gadoxetic acid-enhanced MRI: comparison with aspartate aminotransferase-to-platelet ratio index and FIB4. J Magn Reson Imaging 2017;45(4):1186e94. 15. Kim H, Park SH, Kim EK, et al. Histogram analysis of gadoxetic acidenhanced MRI for quantitative hepatic fibrosis measurement. PLoS One 2014;9(12):e114224. 16. Asayama Y, Nishie A, Ishigami K, et al. Histogram analysis of noncancerous liver parenchyma on gadoxetic acid-enhanced MRI: predictive value for liver function and pathology. Abdom Radiol 2016;41(9):1751e7. 17. Choi YR, Lee JM, Yoon JH, et al. Comparison of magnetic resonance elastography and gadoxetate disodium-enhanced magnetic resonance imaging for the evaluation of hepatic fibrosis. Invest Radiol 2013;48(8):607e13. 18. Jang Y-J, Cho SH, Bae JH, et al. Noninvasive assessment of hepatic fibrosis using gadoxetate-disodium-enhanced 3T MRI. Ann Hepatol 2013;12(6):926e34. 19. Ding Y, Rao S, Yang L, et al. Comparison of the effect of region-of-interest methods using gadoxetic acid-enhanced MR imaging with diffusionweighted imaging on staging hepatic fibrosis. Radiol Med 2016;121(11):821e7. 20. Noren B, Forsgren MF, Dahlqvist Leinhard O, et al. Separation of advanced from mild hepatic fibrosis by quantification of the hepatobiliary uptake of Gd-EOB-DTPA. Eur Radiol 2013;23(1):174e81.
Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
X. Li et al. / Clinical Radiology xxx (xxxx) xxx 21. Feier D, Balassy C, Bastati N, et al. Liver fibrosis: histopathologic and biochemical influences on diagnostic efficacy of hepatobiliary contrastenhanced MR imaging in staging. Radiology 2013;269(2):459e67. 22. Kumazawa K, Edamoto Y, Yanase M, et al. Liver analysis using gadoliniumethoxybenzyl-diethylenetriamine pentaacetic acid-enhanced magnetic resonance imaging: correlation with histological grading and quantitative liver evaluation prior to hepatectomy. Hepatol Res 2012;42(11):1081e8. 23. Zhang B, Liang L, Dong Y, et al. Intravoxel incoherent motion MR imaging for staging of hepatic fibrosis. PLoS One 2016;11(1):e0147789. 24. Whiting PF, Rutjes AW, Westwood ME, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011;155(8):529e36. 25. Motosugi U, Ichikawa T, Oguri M, et al. Staging liver fibrosis by using liver-enhancement ratio of gadoxetic acid-enhanced MR imaging: comparison with aspartate aminotransferase-to-platelet ratio index. Magn Reson Imaging 2011;29(8):1047e52. 26. Nojiri S, Kusakabe A, Fujiwara K, et al. Noninvasive evaluation of hepatic fibrosis in hepatitis C virus-infected patients using ethoxybenzylmagnetic resonance imaging. J Gastroenterol Hepatol 2013;28(6):1032e9. 27. Chen BB, Hsu CY, Yu CW, et al. Clinical and histologic implications of delayed hepatobiliary enhancement on magnetic resonance imaging with gadolinium ethoxybenzyl diethylenetriaminepentaacetic acid. Invest Radiol 2012;47(11):649e55. 28. Wu WP, Hoi CI, Chen RC, et al. Comparison of the efficacy of Gd-EOBDTPA-enhanced magnetic resonance imaging and magnetic resonance elastography in the detection and staging of hepatic fibrosis. Medicine 2017;96(42):e8339. 29. Zamora J, Abraira V, Muriel A, et al. Meta-DiSc: a software for metaanalysis of test accuracy data. BMC Med Res Methodol 2006;6:31. 30. Wu Z, Matsui O, Kitao A, et al. Hepatitis C related chronic liver cirrhosis: feasibility of texture analysis of MR images for classification of fibrosis stage and necroinflammatory activity grade. PLoS One 2015;10(3). 31. Chen BB, Hsu CY, Yu CW, et al. Dynamic contrast-enhanced magnetic resonance imaging with Gd-EOB-DTPA for the evaluation of liver fibrosis in chronic hepatitis patients. Eur Radiol 2012;22(1):171e80. 32. Kim T, Murakami T, Hasuike Y, et al. Experimental hepatic dysfunction: evaluation by MRI with Gd-EOB-DTPA. J Magn Reson Imaging 1997;7(4):683e8.
9
33. Chen C, Wang B, Shi D, et al. Initial study of biexponential model of intravoxel incoherent motion magnetic resonance imaging in evaluation of the liver fibrosis. Chin Med J 2014;127(17):3082e7. 34. Watanabe H, Kanematsu M, Goshima S, et al. Staging hepatic fibrosis: comparison of gadoxetate disodium-enhanced and diffusion-weighted MR imaging-preliminary observations. Radiology 2011;259(1):142e50. 35. Harada TL, Saito K, Araki Y, et al. Prediction of high-stage liver fibrosis using ADC value on diffusion-weighted imaging and quantitative enhancement ratio at the hepatobiliary phase of Gd-EOB-DTPAenhanced MRI at 1.5 T. Acta Radiol 2018;59(5):509e16. 36. Tajima T, Takao H, Akai H, et al. Relationship between liver function and liver signal intensity in hepatobiliary phase of gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid-enhanced magnetic resonance imaging. J Comput Assist Tomogr 2010;34(3):362e6. 37. Bae KE, Kim SY, Lee SS, et al. Assessment of hepatic function with GdEOB-DTPA-enhanced hepatic MRI. Dig Dis 2012;30(6):617e22. 38. Xiao G, Zhu S, Xiao X, et al. Comparison of laboratory tests, ultrasound, or magnetic resonance elastography to detect fibrosis in patients with nonalcoholic fatty liver disease: a meta-analysis. Hepatology 2017;66(5):1486e501. 39. Kim YS, Song JS, Kannengiesser S, et al. Comparison of spin-echo echoplanar imaging and gradient recalled echo-based MR elastography at 3 Tesla with and without gadoxetic acid administration. Eur Radiol 2017;27(10):4120e8. 40. Gersak MM, Badea R, Lenghel LM, et al. Influence of food intake on 2-D shear wave elastography assessment of liver stiffness in healthy subjects. Ultrasound Med Biol 2016;42(6):1295e302. 41. Chang JM, Moon WK, Cho N, et al. Clinical application of shear wave elastography (SWE) in the diagnosis of benign and malignant breast diseases. Breast Cancer Res Treat 2011;129(1):89e97. 42. Yoon JH, Lee JM, Paek M, et al. Quantitative assessment of hepatic function: modified look-locker inversion recovery (MOLLI) sequence for T1 mapping on Gd-EOB-DTPA-enhanced liver MR imaging. Eur Radiol 2016;26(6):1775e82. 43. Yoon JH, Lee JM, Kim E, et al. Quantitative liver function analysis: volumetric T1 mapping with fast multisection B1 inhomogeneity correction in hepatocyte-specific contrast-enhanced liver MR imaging. Radiology 2017;282(2):408e17.
Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
Contents lists available at ScienceDirect
Clinical Radiology journal homepage: www.clinicalradiologyonline.net
Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis X. Li a, H. Liu b, R. Wang a, J. Yang a, c, Y. Zhang a, **, 1, C. Li a, c, *, 1 a
Department of Diagnostic Radiology, The First Affiliated Hospital of Xi’an Jiaotong University, Xi’an, Shaanxi Province, China b Department of Radiology, The Third Hospital of Hebei Medical University, Shijiazhuang, Hebei Province, China c Department of Biomedical Engineering, The Key Laboratory of Biomedical Information Engineering of the Ministry of Education, School of Life Science and Technology, Xi’an Jiaotong University, Xi’an, China
art icl e i nformat ion Article history: Received 28 March 2019 Accepted 12 November 2019
AIM: To identify and evaluate the efficiency of the most commonly used parameters applied to gadoxetate disodium (Gd-EOB-DTPA)-enhanced magnetic resonance imaging (MRI) for liver fibrosis (LF) staging. MATERIALS AND METHODS: Literature searches of PubMed, Web of Science, Embase, and MEDLINE databases from January 2004 to August 2018 were conducted. The applied parameters during imaging were noted and summarised. Studies using the most commonly used parameter were included. Extractive data were combined as pooled weighted mean difference (WMD) to determine the benefits in LF staging. The pooled sensitivity, specificity, and summary receiver operating characteristics (SROC) curve were calculated. RESULTS: Among 57 relevant studies, the contrast enhancement index (CEI) was a relatively commonly used parameter. It was calculated as SIpost/SIpre, where SIpost and SIpre are the liverto-muscle signal intensity ratios in the hepatocyte phase and pre-enhanced images, respectively. Six studies were included. F0 was regarded as normal liver, F1 as mild LF, F2 as moderate LF, and F3 and F4 as advanced LF. Comparisons of WMD revealed significant differences between F1e2 and F3e4. For stage F1, the pooled sensitivity, specificity, and areas under SROC curve were 0.58, 0.84, and 0.85, respectively; the corresponding values for F2 were 0.57, 0.68, and 0.76, while those for F3 were 0.61, 0.75, and 0.72. CONCLUSION: The methodology and parameters used for Gd-EOB-DTPA-enhanced MRI for LF staging are diverse, but the CEI was a relatively common parameter. Overall, there is evidence to support use of CEI, but more evidence from larger studies is needed. Ó 2019 The Royal College of Radiologists. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
* Guarantor and correspondent: C. Li, Department of Diagnostic Radiology, The First Affiliated Hospital of Xi’an Jiaotong University, Yanta West Road 277, Xi’an, Shaanxi, China. Tel.: þ86-15389042141. ** Guarantor and correspondent: Y. Zhang, Department of Diagnostic Radiology, The First Affiliated Hospital of Xi’an Jiaotong University, Yanta West Road 277, Xi’an, Shaanxi, China. Tel.: þ86-18991232590. E-mail addresses: [email protected] (Y. Zhang), [email protected] (C. Li). 1 These authors contributed equally to this work. https://doi.org/10.1016/j.crad.2019.11.001 0009-9260/Ó 2019 The Royal College of Radiologists. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
2
X. Li et al. / Clinical Radiology xxx (xxxx) xxx
Introduction Liver fibrosis (LF) involves the accumulation of an extracellular matrix resulting in fibrogenesis, which is created in response to acute or chronic liver injury. Fibrosis progresses at variable rates and eventually develops into cirrhosis, which in turn leads to hepatic dysfunction.1,2 It is reported that early LF is manageable, its progression can be controlled, and it is reversible in a large percentage of patients3; however, the probability of reversing advanced LF is low,4,5 and so early detection of LF has important clinical implications. A liver biopsy is the reference-standard technique for diagnosing LF; however, this procedure is limited by its invasive nature, sampling errors associated with small tissue samples, and the risk of complications.6 Several noninvasive diagnostic methods, including direct/indirect serum markers, serum panels, and imaging techniques, are now being utilised to avoid the shortcomings of biopsies.7,8 Conventional magnetic resonance imaging (MRI) is one of the non-invasive methods, but this provides data only on the changes in liver morphology. Some magnetic resonance techniques can provide information on liver function in addition to morphology, such as contrast-enhanced MRI.9,10 Gadoxetate disodium (Gd-EOB-DTPA; marketed as Primovist in Europe and Eovist in the United States, Bayer HealthCare) is a liver-specific contrast agent, and the uptake of Gd-EOB-DTPA by hepatocytes mainly occurs via the organic anion transporter polypeptides (OATP)1B1 and B3 located at the sinusoidal membrane. Biliary excretion occurs via the multidrug resistance-associated proteins, MRP2, at the canalicular membrane. Gd-EOB-DTPA can therefore be utilised for evaluating liver function and the staging of LF.11 The application of T1-weighted MRI (T1WI) and relaxation time12e14 or histogram15,16 analysis revealed a correlation between the stage of LF and the signal intensity (SI) of the liver during Gd-EOB-DTPA-enhanced MRI17e19; however, current research uses various parameters for assessing liver function and thus lacks uniformity, which limits the further application of this technique.20,21 This present systematic review and meta-analysis was conducted with the aim of detecting the parameters used in Gd-EOB-DTPA-enhanced MRI that are most suitable for evaluating LF. The diagnostic performance of the selected parameters was also determined.
Materials and methods The present meta-analysis was performed in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement. No particular ethical considerations were required as this meta-analysis did not involve any identifiable patient information.
Literature search criteria A literature search was performed of the PubMed, Web of Science, Embase, and MEDLINE databases for articles
published from January 2004 (start of clinical application) to August 2018. Articles pertaining to the diagnostic accuracy of Gd-EOB-DTPA-enhanced MRI for staging LF were considered relevant for inclusion in the analysis. Crossreferencing from the reference lists in the included original articles and associated review articles was also performed to identify additional eligible articles. Search query terms were developed based on specific keywords (“gadoxetic acid,” “gadoxetate,” “Gd-EOB-DTPA,” and “hepatic fibrosis”) and medical subject headings (“liver fibrosis” and “magnetic resonance imaging”). The following specific search query was designed for retrieving relevant data from the four databases: (“liver fibrosis” OR “hepatic fibrosis”) AND (“gadoxetic acid” OR “Gd-EOB-DTPA” OR “gadoxetate”) AND (“magnetic resonance imaging” OR “MRI”).
Inclusion and exclusion criteria The relevance of each article was assessed based on its title and abstract, followed by a full-text evaluation. Articles were included based on the pre-specified inclusion/exclusion criteria described below. The database search was performed by three investigators independently, and any ambiguities were resolved through discussion. The following inclusion criteria were applied: (1) objective: to determine the diagnostic performance of GdEOB-DTPA-enhanced MRI in LF staging; (2) patients: patients with pathologically staged LF or healthy volunteers with no history of chronic hepatic disease. All of them underwent Gd-EOB-DTPA-enhanced MRI; (3) parameters: summary of the parameters used for Gd-EOB-DTPAenhanced MRI, and the study that applied the most frequently used parameter was included; (4) study type: randomised controlled trials or original research; (5) language: article written in English. The exclusion criteria were as follows: (1) incompleteness of data preventing extraction and analysis; (2) repeated articles reporting on the same population.
Data extraction The studies identified in the databases were screened manually to avoid repetition, and checked manually for relevancy by three independent investigators who also performed the database searches, and then independently extracted the data related to the most frequently used parameter. In addition to the baseline characteristics, the investigators also extracted the results, which included the mean and standard deviation (SD) values for the each stage of LF, and also the cut-off value, true-positive (TP) value, false-positive (FP) value, false-negative (FN) value, and truenegative (TN) value for each compared group. Histopathology results were considered the reference standard in all of the included studies. Consistent with published research, F0 was regarded as a normal liver, and F1 as mild LF, F2 as moderate LF, and F3 and F4 as advanced LF.22,23 The data for stages F0, F1, F2, and F3e4 were therefore extracted and analysed.
Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
X. Li et al. / Clinical Radiology xxx (xxxx) xxx
Quality assessment The overall methodological quality and bias of the included articles were assessed using the revised tool for the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2), which evaluates the risk of bias for four domains and the clinical applicability for three domains of the study characteristics.24
Statistical analysis All of the data were arranged into two sets: (1) demography and (2) diagnostic performance. Data from the included studies were combined and expressed as pooled weighted mean difference (WMD) values with 95% CIs (confidence intervals) using RevMan (Review Manager, version 5.3, Cochrane Collaboration) to determine the benefit of using the most frequently used parameter for staging LF. Studies were weighted by the inverse variance, and a fixed-effects model was initially used. Owning to different stages of LF being analysed in different studies, the pooled mean and SD of the most frequently used parameter in each compared group were calculated. The pooled sensitivity and specificity with 95% CIs for the most frequently used parameter were calculated based on different stages of LF using MetaDiSc (version 1.4). The summary receiver operating characteristics (SROC) curve was constructed to perform a comprehensive evaluation of the diagnostic performance of the most frequently used parameter for LF staging. Heterogeneity was evaluated using Cochrane’s Q-test, where p<0.1 indicates the presence of heterogeneity. The I2 value was used to quantitatively evaluate heterogeneity, with a fixed-effects model used for meta-analysis when I2<25% and a random-effects model applied when I2¼25e50%. When I2>50%, the sources of heterogeneity were analysed first, and then a random-effects model was used once it was determined that there was no obvious clinical heterogeneity and the sources of heterogeneity could not be found.
Results Study selection and characteristics The literature search of the four databases yielded 267 records. Screening for duplicates resulted in the removal of 73 articles, and a further 137 studies were excluded for the reasons listed in Fig 1. Among the remaining 57 potentially relevant studies, the contrast enhancement index (CEI) was the most frequently used parameter, being applied in 10 studies (Table 1). The CEI was calculated as SIpost/SIpre, where SIpost and SIpre are the liver-to- paraspinal muscle SI ratios in T1WI of hepatocyte-phase images and preenhanced images, respectively. Six studies17,18,25e28 were finally included in the analysis after excluding a further four studies (Fig 1). Hepatocyte-phase images were obtained at 20 min after injection of Gd-EOB-DTPA in those studies.
3
The characteristics of the included studies are presented in Table 2. Nojiri et al. randomly divided all subjects into an estimation model and a validation model, and only data for the former model26 were used in the present study. The present meta-analysis included 695 patients with mean ages of 42.3e71 years in the various studies, and men:women gender ratio of 18:7. Of the six included studies, five studies used 1.5 T MRI and one study used 3 T MRI. Three of the six included studies applied a liver acquisition with volume acceleration (LAVA) sequence. Two studies did not use the LAVA sequence, but the repetition time (TR) and echo time (TE) values of the sequences used were similar. Only one study used higher TR and TE. Histopathology was considered as the reference standard in all of the studies, which used different equipment. The METAVIR scoring system was applied in five studies, with the sixth study employing the Batts and Ludwig system, but the detailed scoring standard was similar in all six studies.
Quality assessment The results of the quality assessment performed using the QUADAS-2 tool are summarised in Fig 2, with the risk of bias presented on the left and concerns regarding applicability presented on the right. A substantial risk of bias regarding the index test was observed, probably due to the cut-off value not being pre-specified. A considerable risk of bias was also observed for patient selection, since for most studies it was not reported whether a consecutive or random sample of patients was enrolled; however, there were few applicability concerns regarding patient selection, the index test, or the reference standard for all of the studies.
Diagnostic performance of Gd-EOB-DTPA-enhanced MRI The mean CEI was used for staging LF. The diagnostic performances of Gd-EOB-DTPA-enhanced MRI in patients with LF for stage F1 versus F0 and stage F3e4 versus F1e2 are presented in Fig 3a and 3b, respectively. For LF of stage F1 compared with stage F0, the pooled WMD was e0.09 (95% CI¼e0.20 to 0.02, p¼0.11). A high degree of heterogeneity was found among the relevant studies: p¼0.03, I2¼67%. The mean CEI of Gd-EOB-DTPA-enhanced MRI in all of the individual studies differed between LF of stage F3-4 and stage F1-2: WMD¼e0.11, 95% CI¼e0.16 to e0.06, and p<0.0001. A moderate heterogeneity was found among the relevant studies: p¼0.11, I2¼50%.
Diagnostic accuracy of Gd-EOB-DTPA-enhanced MRI Detailed data on the cut-off value, TP, FP, FN, and TN for each compared group are presented in Table 3. Fig 4 presents the pooled sensitivity and specificity values of GdEOB-DTPA-enhanced MRI in LF staging. In patients with LF stage 1 (F0 versus F1e4), the pooled sensitivity and specificity were 0.58 (95% CI¼0.54 to 0.63, I2 ¼ 94%) and 0.84 (95% CI¼0.73 to 0.92, I2 ¼ 0.0%), respectively (Fig 4a). The pooled sensitivity of Gd-EOB-DTPA-enhanced MRI was 0.57 (95% CI¼0.53 to 0.62, I2 ¼ 90.0%), which was lower in
Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
4
X. Li et al. / Clinical Radiology xxx (xxxx) xxx
Figure 1 Flow diagram of study selection. Some relative studies applied several parameters, so the number of applied parameter was not consistent with the number of relative studies. SIpost signal intensity in hepatobiliary phase. SIpre pre-enhanced signal intensity. Table 1 Studies using the contrast enhancement index (CEI) for liver fibrosis staging. First author
Title
Chen
Clinical and histological implications of delayed hepatobiliary enhancement on magnetic resonance imaging with gadolinium ethoxybenzyl diethylenetriaminepentaacetic acid Comparison of magnetic resonance elastography and gadoxetate disodium-enhanced magnetic resonance imaging for the evaluation of hepatic fibrosis Gd-EOB-DTPA-enhanced MR imaging: prediction of hepatic fibrosis stages using liver contrast enhancement index and liver-to-spleen volumetric ratio Noninvasive assessment of hepatic fibrosis using gadoxetate-disodium-enhanced 3T MRI Hepatic enhancement of Gd-EOB-DTPA enhanced 3 tesla MR Imaging: assessing severity of liver cirrhosis Staging liver fibrosis by using liver-enhancement ratio of gadoxetic acid-enhanced MR imaging: comparison with aspartate aminotransferase-to-platelet ratio index Noninvasive evaluation of hepatic fibrosis in hepatitis C virus-infected patients using ethoxybenzylmagnetic resonance imaging Three-tesla magnetic resonance elastography for hepatic fibrosis: comparison with diffusion-weighted imaging and gadoxetic acid-enhanced magnetic resonance imaging Staging hepatic fibrosis: comparison of gadoxetate disodium-enhanced and diffusion-weighted MR imagingdpreliminary observations Comparison of the efficacy of Gd-EOB-DTPA enhanced magnetic resonance imaging and magnetic resonance elastography in the detection and staging of hepatic fibrosis
Choi
Goshima
Jang Lee Motosugi
Nojiri
Park
Watanabe
Wu
patients with LF 2 (F0e1 versus F2e4) than in patients with LF 1. The pooled specificity in patients with LF 2 was 0.68 (95% CI¼0.61 to 0.75, I2 ¼91.1%; Fig 4b). The pooled sensitivity increased to 0.61 (95% CI¼0.57 to 0.66, I2 ¼ 66.8%) in patients with LF 3 (F0e2 versus F3e4), with a pooled specificity of 0.75 (95% CI¼0.69 to 0.79, I2 ¼ 81.3%; Fig 4c). An overall trend of the pooled sensitivity increasing with the LF stage was observed. The area under the SROC curve (AUC) for the CEI when evaluating the LF stage was 0.85 for F0 versus F1e4, 0.76 for F0e1 versus F2e4, and 0.72 for F1e2 versus F3e4 (Fig 5). The Q* values for the corresponding AUCs were 0.78, 0.70, and 0.67, respectively.
Heterogeneity analysis and publication bias The threshold effect is one of the most important causes of heterogeneity in diagnostic tests.29 Due to the small number of studies included in each group in the present study, threshold, meta-regression, and sensitivity analyses could not be performed. The inclusion of <10 studies further meant that the statistical power for calculating publication bias was very low, and so this was not assessed or plotted.
Discussion To the authors’ knowledge, this was the first metaanalysis on selecting the most commonly used parameters in Gd-EOB-DTPA-enhanced MRI for staging LF. The present systematic review and meta-analysis found that the CEI was
Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
X. Li et al. / Clinical Radiology xxx (xxxx) xxx
5
Table 2 Basic characteristics of included literatures. Study
Year Study design N
Age, mean (SD/Scope)
Sex (M/F)
Equipment
Chen
2012 Prospective
42.3 (10.6)
39/17
1.5 T Siemens 209
Choi
2013 Retrospective 173 57.2 (10.4)
129/44 1.5 T GE
Jang
2013 Retrospective 113 56.0 (12.5)
73/40
56
TR (ms) TE (ms) Reference Scoring standard system
ROI ROI location number
4.8
Pathology METAVIR 3
4.2
1.9
Pathology METAVIR 3
3.0 T GE
4.0
2.0
Motosugia 2011 Retrospective 100 64.6 (28e83) ND Nojirib 2013 Prospective 149 71 (8) 99/50
1.5 T GE 1.5 T Philips
ND 4.2
ND 2.0
Pathology Batts and 4 Ludwig system Pathology METAVIR 3 Pathology METAVIR 3
Wu
1.5 T Siemens 3.7
1.4
Pathology METAVIR 1
2017 Retrospective 104 60.6 (10.6)
87/17
Left, right anterior, and right posterior lobe Left, right anterior, and right posterior lobe Right anterior, right posterior, left lateral, and left medial segments Right lobe Left, right anterior, and right posterior lobe Right lobe
TR, repetition time TE, echo time; ND not documented; ROI, region of interest. a This study used liver acquisition with volume acceleration sequence, consistent with studies by Jang YJ et al. and Choi YR et al. b This study randomly divided into model estimation and model validation, and only model estimation data were used in present.
Figure 2 Quality assessments with QUADAS-2. The substantial risk of bias derived from index test (left part) and the applicability concerns were low in all the studies (right part).
Figure 3 Pooled WMD in CEI of Gd-EOB-DTPA enhanced MRI. (a) Comparison of stage F0 with F1. (b) Comparison of stage F1e2 and F3e4.
the most commonly used parameter in Gd-EOB-DTPAenhanced MRI for evaluating liver function. Further, the WMD between the CEI at different stages showed that it
could be used for the staging of LF. The AUCs of the different compared groups revealed obvious differences between the normal and fibrotic liver, mild and moderate LF, and
Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
6
X. Li et al. / Clinical Radiology xxx (xxxx) xxx
Table 3 Detail data of included studies for pooled sensitivity and specificity. Stages
First author
Cut-off
TP
FP
FN
TN
Se (%)
Sp (%)
F0 versus F1e4
Chen Choi Jang Motosugi Wu Chen Choi Jang Motosugi Wu Chen Chen Choi Jang Motosugi Wu Chen
ND 1.613 1.322 1.91 1.75 ND 1.609 1.3169 1.76 ND 1.57 ND 1.637 1.3173 1.76 ND 1.55
20 69 44 73 75 26 63 46 50 69 44 4 66 38 43 60 38
1 3 1 4 1 25 6 1 14 6 3 5 23 4 21 15 6
27 81 56 11 26 0 73 46 17 41 45 10 38 29 14 31 34
8 20 12 12 2 5 31 20 19 33 11 37 47 42 22 43 26
42.6 46.0 44.0 86.9 74.3 100 46.3 50.0 74.6 62.7 49.4 28.6 62.8 56.7 75.4 65.9 52.8
88.9 85.0 92.3 75.0 66.7 16.7 82.4 93.5 56.0 84.6 80.0 88.1 68.2 91.3 50.0 74.1 81.3
F0e1 versus F2e4
F0e2 versus F3e4
TP, true-positive value; FP, false-positive value; FN, false-negative value; TN, true-negative value; Se, sensitivity; Sp, specificity; ND, not documented.
moderate and advanced LF, thus indicating it can be used to evaluate the LF stage. The published literature includes multiple MRI parameters for staging LF (Fig 1). There are numerous parameters related to Gd-EOB-DTPA-enhanced MRI, such as the CEI, liver-to-spleen SI ratio, post-contrast to pre-contrast ratio of the liver SI, liver-to-intervertebral disc SI ratio,26 liver-tospine SI ratio,25 and texture analysis parameters.30 In addition, the following parameters are also used to assess liver function, such as the arterial blood flow and slope derived from semi-quantitive perfusion MRI include,31 and the T1 relaxation time from T1 mapping13,14,19; however, these parameters have been used only a few times, and they have also been used inconsistently across studies. Among all of the parameters mentioned above, the CEI is the only one that has been evaluated comprehensively in multiple studies. The CEI was therefore chosen as the evaluable parameter for determining the diagnostic performance of Gd-EOB-DTPA-enhanced MRI for staging LF. Gd-EOB-DTPA is a liver-specific MRI contrast agent that is used to produce liver-enhanced images that reflect both liver perfusion and the condition of normal liver cells.11 After being injected intravenously, Gd-EOB-DTPA is gradually taken up by hepatocytes, and eventually is excreted via the biliary pathway and kidneys. Decreases in hepatic enhancement in hepatocyte-phase images suggest the impairment of Gd-EOB-DTPA uptake by the liver, whereas the prolongation of liver enhancement suggests that GdEOB-DTPA excretion into bile is also impaired.32 This impaired Gd-EOB-DTPA uptake by the liver would be caused by a decrease in the number of normal hepatocytes or a decrease in Gd-EOB-DTPA uptake by hepatocytes, and impaired Gd-EOB-DTPA excretion may be caused by hepatocyte dysfunction.32 All of these conditions are associated with a progressive increase in inflammation and the accumulation of extracellular matrix during LF that would influence the function, perfusion, vascular permeability, and extracellular diffusion of the liver.33 The change in liver SI
after injecting Gd-EOB-DTPA would be mainly related to liver function, but it also appears to be related to several other factors such as individual indocyanine green clearance and plasma bilirubin levels.34 Due to the influence of the liver function in individual patients and the possibility of other liver diseases being present, the liver SI ratio of before and after enhancement was lacked uniform comparison criteria when selected as an index for evaluating liver function. The above-described situation meant that previous studies have used SI values from various other tissues as reference values for achieving relative objectivity when assessing liver function. Among these various SI parameters, the SI of paraspinal muscle is more stable than that for other organs such as the spleen or liver; this is because the paraspinal muscle SI is minimally affected by age and liver function.35 Moreover, Noren et al. demonstrated that the CEI had a better diagnostic accuracy than the SI ratio in one phase.20 Therefore, since it corresponds to the ratio of SI before and after enhancement, the CEI could be a better index for evaluating the stage of LF. Previous research has also demonstrated that the CEI is useful for predicting both the entire and the regional hepatic functional reserves,36,37 and it might also be useful in evaluating the regional LF. Future prospective multicentre studies should further explore the diagnostic performance of Gd-EOB-DTPA-enhanced MRI. Xiao et al.38 conducted a systemic review and metaanalysis to compare the diagnostic accuracy of laboratory tests, shear-wave elastography (SWE), and magnetic resonance elastography (MRE) for assessments of LF. Their results suggested that MRE and SWE have the highest diagnostic accuracy for staging fibrosis; however, MRE is not used widely, and the sequence parameters applied have not been consistent.39 In addition, successful SWE imaging is dependent on the acquisition of clear two-dimensional images, and so the imaging success rate is low for patients with severe cirrhosis. Moreover, the measured values are affected by various factors such as the skill of the operator, external pressure, respiration, and food intake.40,41 T1 mapping was another method used to stage LF; however, the scan sequence are diverse, such as look-locker sequences, modified look-locker inversion recovery (MOLLI) sequence, volumetric T1 mapping with fast multisection B1 inhomogeneity correction,42,43 and many factors limit its clinical application, such as relatively long acquisition time, the influence of magnetic field inhomogeneity, and the inability to perform a full liver scan, etc. The use of Gd-EOB-DTPA-enhanced MRI to evaluate LF provides not only morphological information but also information on the liver function, and also allows the evaluation of focal hepatic lesions. We conducted this metaanalysis in order to detect the evaluable parameters used in Gd-EOB-DTPA-enhanced MRI for staging LF and to determine the diagnostic performance of the selected parameters. Future prospective studies should further explore the diagnostic performance of different technologies in staging LF. The present study was subject to some limitations. Firstly, six studies were included in the evaluation.
Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
X. Li et al. / Clinical Radiology xxx (xxxx) xxx
7
Figure 4 Pooled sensitivity and specificity of CEI in Gd-EOB-DTPA enhanced MRI. (a) Comparison of stage F0 with F1e4. (b) Comparison of stage F0e1 with F2e4. (c) Comparison of stage F0e2 with F3e4. Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
8
X. Li et al. / Clinical Radiology xxx (xxxx) xxx
(Indegene, Bangalore, India) for providing medical writing support and technical assistance in the development of this manuscript.
Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.crad.2019.11.001.
References
Figure 5 Summary receiver operating characteristics (SROC) curve of CEI.
Considering the limited number of studies included, regression analysis was not undertaken to evaluate the impact of relevant factors on the results. CEI may be impacted by different field strengths and imaging parameters, so the study using 3 T MRI and the study using higher TR and TE values were removed from the analysis. The results show that removal of these studies had little effect on the meta-analysis results (Electronic Supplementary Material). Secondly, retrospective studies have also been included, which will inevitably reduce the level of evidence. This is because there are too few prospective studies that meet the inclusion criteria, so further prospective studies are needed for liver function assessment. Thirdly, our analysis was limited to comparisons of clinically relevant studies (F0 versus F1e4, F0e1 versus F2e4, and F1e2 versus F3e4) due to a lack of studies comparing individual periods. We compare liver function in groups rather than in pairs. There are two reasons for this: one is that only one of the six included studies compared liver function across all stages. The second is that clinicians tend to obtain differences in liver function between adjacent stages rather than differences between all stages. In conclusion, the methodology and parameters applied in published studies for LF staging using Gd-EOB-DTPAenhanced MRI are diverse, but the CEI was a relatively common parameter. Overall, there is evidence to support use of CEI in LF staging, but more evidence from larger studies is needed.
Conflict of interest The authors declare no conflict of interest.
Acknowledgements This work was supported by the Shaanxi Provincial Key Research and Development Program (2019SF-007). The authors thank Mr Karan Sharma and Dr Amit Bhat
1. Duval F, Moreno-Cuevas JE, Gonzalez-Garza MT, et al. Liver fibrosis and mechanisms of the protective action of medicinal plants targeting inflammation and the immune response. Int J Inflamm 2015;2015:943497. 2. Ghany MG, Strader DB, Thomas DL, et al. American Association for the Study of Liver D. Diagnosis, management, and treatment of hepatitis C: an update. Hepatology 2009;49(4):1335e74. 3. Czaja AJ. Hepatic inflammation and progressive liver fibrosis in chronic liver disease. World J Gastroenterol 2014;20(10):2515e32. 4. Afdhal NH. Diagnosing fibrosis in hepatitis C: is the pendulum swinging from biopsy to blood tests? Hepatology 2003;37(5):972e4. 5. Kim AI, Saab S. Treatment of hepatitis C. Am J Med 2005;118(8):808e15. 6. Saleh HA, Abu-Rashed AH. Liver biopsy remains the gold standard for evaluation of chronic hepatitis and fibrosis. J Gastrointestin Liver Dis 2007;16(4):425e6. 7. Lurie Y, Webb M, Cytter-Kuint R, et al. Non-invasive diagnosis of liver fibrosis and cirrhosis. World J Gastroenterol 2015;21(41):11567e83. 8. Papastergiou V, Tsochatzis E, Burroughs AK. Non-invasive assessment of liver fibrosis. Ann Gastroenterol 2012;25(3):218e31. 9. Chung SR, Lee SS, Kim N, et al. Intravoxel incoherent motion MRI for liver fibrosis assessment: a pilot study. Acta Radiol 2015;56(12):1428e36. 10. Li Z, Sun J, Chen L, et al. Assessment of liver fibrosis using pharmacokinetic parameters of dynamic contrast-enhanced magnetic resonance imaging. J Magn Reson Imaging 2016;44(1):98e104. 11. Van Beers BE, Pastor CM, Hussain HK. Primovist, eovist: what to expect? J Hepatol 2012;57(2):421e9. 12. Okada M, Murakami T, Yada N, et al. Comparison between T1 relaxation time of Gd-EOB-DTPA-enhanced MRI and liver stiffness measurement of ultrasound elastography in the evaluation of cirrhotic liver. J Magn Reson Imaging 2015;41(2):329e38. 13. Ding Y, Rao SX, Zhu T, et al. Liver fibrosis staging using T1 mapping on gadoxetic acid-enhanced MRI compared with DW imaging. Clin Radiol 2015;70(10):1096e103. 14. Yang L, Ding Y, Rao S, et al. Staging liver fibrosis in chronic hepatitis B with T1 relaxation time index on gadoxetic acid-enhanced MRI: comparison with aspartate aminotransferase-to-platelet ratio index and FIB4. J Magn Reson Imaging 2017;45(4):1186e94. 15. Kim H, Park SH, Kim EK, et al. Histogram analysis of gadoxetic acidenhanced MRI for quantitative hepatic fibrosis measurement. PLoS One 2014;9(12):e114224. 16. Asayama Y, Nishie A, Ishigami K, et al. Histogram analysis of noncancerous liver parenchyma on gadoxetic acid-enhanced MRI: predictive value for liver function and pathology. Abdom Radiol 2016;41(9):1751e7. 17. Choi YR, Lee JM, Yoon JH, et al. Comparison of magnetic resonance elastography and gadoxetate disodium-enhanced magnetic resonance imaging for the evaluation of hepatic fibrosis. Invest Radiol 2013;48(8):607e13. 18. Jang Y-J, Cho SH, Bae JH, et al. Noninvasive assessment of hepatic fibrosis using gadoxetate-disodium-enhanced 3T MRI. Ann Hepatol 2013;12(6):926e34. 19. Ding Y, Rao S, Yang L, et al. Comparison of the effect of region-of-interest methods using gadoxetic acid-enhanced MR imaging with diffusionweighted imaging on staging hepatic fibrosis. Radiol Med 2016;121(11):821e7. 20. Noren B, Forsgren MF, Dahlqvist Leinhard O, et al. Separation of advanced from mild hepatic fibrosis by quantification of the hepatobiliary uptake of Gd-EOB-DTPA. Eur Radiol 2013;23(1):174e81.
Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001
X. Li et al. / Clinical Radiology xxx (xxxx) xxx 21. Feier D, Balassy C, Bastati N, et al. Liver fibrosis: histopathologic and biochemical influences on diagnostic efficacy of hepatobiliary contrastenhanced MR imaging in staging. Radiology 2013;269(2):459e67. 22. Kumazawa K, Edamoto Y, Yanase M, et al. Liver analysis using gadoliniumethoxybenzyl-diethylenetriamine pentaacetic acid-enhanced magnetic resonance imaging: correlation with histological grading and quantitative liver evaluation prior to hepatectomy. Hepatol Res 2012;42(11):1081e8. 23. Zhang B, Liang L, Dong Y, et al. Intravoxel incoherent motion MR imaging for staging of hepatic fibrosis. PLoS One 2016;11(1):e0147789. 24. Whiting PF, Rutjes AW, Westwood ME, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011;155(8):529e36. 25. Motosugi U, Ichikawa T, Oguri M, et al. Staging liver fibrosis by using liver-enhancement ratio of gadoxetic acid-enhanced MR imaging: comparison with aspartate aminotransferase-to-platelet ratio index. Magn Reson Imaging 2011;29(8):1047e52. 26. Nojiri S, Kusakabe A, Fujiwara K, et al. Noninvasive evaluation of hepatic fibrosis in hepatitis C virus-infected patients using ethoxybenzylmagnetic resonance imaging. J Gastroenterol Hepatol 2013;28(6):1032e9. 27. Chen BB, Hsu CY, Yu CW, et al. Clinical and histologic implications of delayed hepatobiliary enhancement on magnetic resonance imaging with gadolinium ethoxybenzyl diethylenetriaminepentaacetic acid. Invest Radiol 2012;47(11):649e55. 28. Wu WP, Hoi CI, Chen RC, et al. Comparison of the efficacy of Gd-EOBDTPA-enhanced magnetic resonance imaging and magnetic resonance elastography in the detection and staging of hepatic fibrosis. Medicine 2017;96(42):e8339. 29. Zamora J, Abraira V, Muriel A, et al. Meta-DiSc: a software for metaanalysis of test accuracy data. BMC Med Res Methodol 2006;6:31. 30. Wu Z, Matsui O, Kitao A, et al. Hepatitis C related chronic liver cirrhosis: feasibility of texture analysis of MR images for classification of fibrosis stage and necroinflammatory activity grade. PLoS One 2015;10(3). 31. Chen BB, Hsu CY, Yu CW, et al. Dynamic contrast-enhanced magnetic resonance imaging with Gd-EOB-DTPA for the evaluation of liver fibrosis in chronic hepatitis patients. Eur Radiol 2012;22(1):171e80. 32. Kim T, Murakami T, Hasuike Y, et al. Experimental hepatic dysfunction: evaluation by MRI with Gd-EOB-DTPA. J Magn Reson Imaging 1997;7(4):683e8.
9
33. Chen C, Wang B, Shi D, et al. Initial study of biexponential model of intravoxel incoherent motion magnetic resonance imaging in evaluation of the liver fibrosis. Chin Med J 2014;127(17):3082e7. 34. Watanabe H, Kanematsu M, Goshima S, et al. Staging hepatic fibrosis: comparison of gadoxetate disodium-enhanced and diffusion-weighted MR imaging-preliminary observations. Radiology 2011;259(1):142e50. 35. Harada TL, Saito K, Araki Y, et al. Prediction of high-stage liver fibrosis using ADC value on diffusion-weighted imaging and quantitative enhancement ratio at the hepatobiliary phase of Gd-EOB-DTPAenhanced MRI at 1.5 T. Acta Radiol 2018;59(5):509e16. 36. Tajima T, Takao H, Akai H, et al. Relationship between liver function and liver signal intensity in hepatobiliary phase of gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid-enhanced magnetic resonance imaging. J Comput Assist Tomogr 2010;34(3):362e6. 37. Bae KE, Kim SY, Lee SS, et al. Assessment of hepatic function with GdEOB-DTPA-enhanced hepatic MRI. Dig Dis 2012;30(6):617e22. 38. Xiao G, Zhu S, Xiao X, et al. Comparison of laboratory tests, ultrasound, or magnetic resonance elastography to detect fibrosis in patients with nonalcoholic fatty liver disease: a meta-analysis. Hepatology 2017;66(5):1486e501. 39. Kim YS, Song JS, Kannengiesser S, et al. Comparison of spin-echo echoplanar imaging and gradient recalled echo-based MR elastography at 3 Tesla with and without gadoxetic acid administration. Eur Radiol 2017;27(10):4120e8. 40. Gersak MM, Badea R, Lenghel LM, et al. Influence of food intake on 2-D shear wave elastography assessment of liver stiffness in healthy subjects. Ultrasound Med Biol 2016;42(6):1295e302. 41. Chang JM, Moon WK, Cho N, et al. Clinical application of shear wave elastography (SWE) in the diagnosis of benign and malignant breast diseases. Breast Cancer Res Treat 2011;129(1):89e97. 42. Yoon JH, Lee JM, Paek M, et al. Quantitative assessment of hepatic function: modified look-locker inversion recovery (MOLLI) sequence for T1 mapping on Gd-EOB-DTPA-enhanced liver MR imaging. Eur Radiol 2016;26(6):1775e82. 43. Yoon JH, Lee JM, Kim E, et al. Quantitative liver function analysis: volumetric T1 mapping with fast multisection B1 inhomogeneity correction in hepatocyte-specific contrast-enhanced liver MR imaging. Radiology 2017;282(2):408e17.
Please cite this article as: Li X et al., Gadoxetate-disodium-enhanced magnetic resonance imaging for liver fibrosis staging: a systematic review and meta-analysis, Clinical Radiology, https://doi.org/10.1016/j.crad.2019.11.001