Article Text

Original research
Diagnostic values of contrast-enhanced MRI and contrast-enhanced CT for evaluating the response of hepatocellular carcinoma after transarterial chemoembolisation: a meta-analysis
  1. Chao Zhang1,
  2. Xin Chen1,
  3. Jukun Wang1,
  4. Tao Luo2
  1. 1Xuanwu Hospital Capital Medical University, Beijing, China
  2. 2Department of General Surgery, Xuanwu Hospital Capital Medical University, Beijing, China
  1. Correspondence to Dr Tao Luo; 3402171856{at}qq.com

Abstract

Objectives To assess and compare the diagnostic value of contrast-enhanced MRI (CEMRI) and contrast-enhanced CT (CECT) for evaluating the response of hepatocellular carcinoma (HCC) after transarterial chemoembolisation (TACE).

Design Systematic review and meta-analysis.

Data sources PubMed, Embase, the Cochrane Library, CNKI and Wanfang databases were systematically searched from inception to 1 August 2023.

Eligibility criteria Studies with any outcome that demonstrates the diagnostic performance of CEMRI and CECT for HCC after TACE were included.

Data extraction and synthesis Two authors independently extracted the data and assessed the quality of included studies. Study quality was assessed using Quality Assessment of Diagnostic Accuracy Studies-2. The diagnostic performance of CEMRI and CECT for the response of HCC was investigated by collecting true and false positives, true and false negatives, or transformed-derived data from each study to calculate specificity and sensitivity. Other outcomes are the positive likelihood ratio/negative likelihood ratio (NLR), the area under the receiver operating characteristic curve (AUC) for diagnostic tests and the diagnostic OR (DOR). Findings were summarised and synthesised qualitatively according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines.

Results This study included 5843 HCC patients diagnosed with CEMRI or CECT and treated with TACE from 36 studies. The mean proportion of men in the total sample was 76.3%. The pool sensitivity, specificity and AUC of CEMRI in diagnosing HCC after TACE were 0.92 (95% CI: 0.86 to 0.96), 0.94 (95% CI: 0.86 to 0.98) and 0.98 (95% CI: 0.96 to 0.99). The pool sensitivity, specificity and AUC of CECT in diagnosing HCC after TACE were 0.74 (95% CI: 0.68 to 0.80), 0.98 (95% CI: 0.93 to 1.00) and 0.90 (95% CI: 0.88 to 0.93).

Conclusions In conclusion, this study found that both CEMRI and CECT had relatively high predictive power for assessing the response of HCC after TACE. Furthermore, the diagnostic value of CEMRI may be superior to CECT in terms of sensitivity, AUC, DOR and NLR.

  • Hepatobiliary disease
  • Hepatology
  • Hepatobiliary tumours

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Strengths and limitations of this study

  • Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement was used in reporting this systematic review and meta-analysis, and a comprehensive literature search was conducted for this study.

  • The quality of the systematic review is limited by the quality of the included studies, such as the potential for selection bias, information bias and confounding bias in the original studies.

  • There is some heterogeneity among studies due to the patient characteristics and study designs.

  • Publication bias was inevitable as we did not search the grey literature.

Introduction

Hepatocellular carcinoma (HCC) is the sixth most prevalent cancer and the second leading cause of cancer death worldwide, accounting for approximately 740 000 deaths annually.1 Liver resection, radiofrequency ablation and liver transplantation are potential curative strategies for HCC based on the Barcelona Clinic Liver Cancer staging classification.2 However, most patients with HCC are at intermediate to advanced stages at diagnosis, with multiple lesions. Less than 40% of HCC cases are suitable candidates for treatment using surgical resection or liver transplantation.3

Transarterial chemoembolisation (TACE) is widely used as a first-line treatment and palliative therapy for patients with unresectable HCC without vascular invasion, blocking the tumour-feeding artery by increasing the concentrations of chemotherapeutic agents at the tumour site.4 TACE can be a bridge to liver transplantation.5 Unfortunately, the progression of HCC after TACE remains high.6–8 Therefore, early and accurate evaluation of HCC after TACE is essential for assessing the therapeutic response and guiding further treatment strategies.

Contrast-enhanced computed tomography (CECT) is commonly used to assess tumour lesions, including the response to TACE.9–19 Lipiodol is used as an angiographic contrast medium to guide TACE, but since CECT also uses an iodine-based contrast medium, lipiodol deposition in the liver can obscure the enhancement of viable HCC tissue when CECT is performed to evaluate the response to TACE. Recently, contrast-enhanced magnetic resonance imaging (CEMRI) has been proposed as an alternative to assess the tumour response after TACE.10–12 14 16 19–24 It provides high-quality liver imaging with high intrinsic soft tissue contrast and spatial resolution, and it is not affected by lipiodol deposition.25

Still, whether the diagnostic value of CEMRI for evaluating the response of HCC after TACE was superior to CECT remains unclear. Therefore, the present systematic review and meta-analysis were undertaken to update the diagnostic values of CEMRI and CECT for evaluating the response of HCC after TACE, and the diagnostic value was compared using indirect analysis methods.

Material and methods

Data sources, search strategy and selection criteria

This systematic review and meta-analysis was reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Statement.26 PubMed, Embase, the Cochrane Library, the CNKI and Wanfang databases were systematically searched on 1 August 2023 without restrictions on publication language or date. The search strategy combined subject headings and keywords to represent the following concepts: (“magnetic resonance” OR “computed tomography”) AND “liver cancer” AND “transarterial chemoembolization”. In addition, the reference lists of relevant reviews were manually searched to identify any additional studies that met the inclusion criteria. One relevant review article, titled ‘Meta-analysis of the lesion survival or recurrence value of enhanced CT and enhanced MRI for hepatocellular carcinoma’ by Li Dong, Luo Tianyang and Ran Yawei, was included in the manual search.27

The inclusion criteria were as follows: (1) participants: patients with HCC treated with TACE; (2) diagnostic tools: CEMRI or CECT and (3) outcomes: any outcome that demonstrates the diagnostic performance of CEMRI and CECT for HCC after TACE, including true and false positive, true and false negative, or data obtained through transformation.

The exclusion criteria were as follows: (1) participants: mixed populations of HCC or other diagnose unless data can be extracted separately; (2) outcomes: studies lacking relevant outcome data on diagnostic performance of CEMRI and CECT for HCC after TACE and (3) studies reporting only non-original data (eg, reviews, commentaries, editorials, meeting abstracts, etc).

The data comes from published articles and does not require ethical approval.

Data collection and quality assessment

Two authors independently extracted the data and assessed the quality of included studies. Any conflict was resolved by discussion until a consensus was reached. We collected the following data items from each study: first author’s name, publication year, country, study design, sample size, mean age, percentage male, number of nodules, mean tumour diameter, TACE emulsion, the time interval between diagnosis and TACE, diagnostic tool, reference standard, true and false positive, and true and false negative. We used the Quality Assessment of Diagnostic Accuracy Studies-2 to assess the quality of the studies. It was based on seven items, each being answered as low risk, unclear risk, or high risk.28

Statistical analysis

The diagnostic parameters for CEMRI and CECT were calculated based on the number of true and false positives and true and false negatives. Summary sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR) and the area under the receiver operating characteristic curve (AUC) of CEMRI and CECT were calculated using the bivariate generalised linear mixed model. The random-effects model was applied to calculate the pooled diagnostic OR (DOR).19 29 30 I2 and Q statistics were used for evaluating the heterogeneity, and I2>50.0% and p<0.10 indicated significant statistical heterogeneity.31 32 Subgroup analyses were also conducted to assess the diagnostic performance of CEMRI and CECT according to publication year, study design, mean age and percentage of males. The risk of publication bias was assessed using a funnel plot and Deeks’ asymmetry test.33 The p values for all pooled results were two-sided, and the inspection level was 0.05. Statistical analysis was conducted using Stata (SE 15.0).

Patient and public involvement

None.

Results

Literature search

After an initial search in five databases, a total of 7876 studies were identified, of which 2284 were from PubMed, 5317 were from Embase, 214 were from the Cochrane Library, 3 were from CNKI, 58 were from WANGFANG and 4 from manually screened. According to predetermined inclusion and exclusion criteria, 7880 article abstracts were screened. Among them, 1912 were excluded due to duplication, while 1837 were excluded due to inappropriate research types such as conferences, reviews, protocols, editorials, letters, note or meta-analyses. Due to population, intervention, comparison, outcomes and study design (PICOS) not meeting the inclusion criteria, 4095 studies were further excluded (figure 1). Finally, 36 studies involving 5843 patients were included in the meta-analysis.11–14 16 18 20–22 24 34–39 The completed search strategy is in online supplemental file 1.

Figure 1

Literature search and study selection.

Study characteristics

The characteristics of included studies and patients are summarised in table 1 (detailed in online supplemental table 2). Of the 36 included studies, 15 had a prospective design, and the remaining 21 had a retrospective design. 27 studies were conducted in Asia, and the remaining nine in Europe, the USA or Egypt. These studies recruited 2109 nodules, with 12–297 nodules included in each study. 16 studies reported the diagnostic value of CEMRI, and 26 studies reported the diagnostic value of CECT. The quality of included studies is shown in online supplemental table 3. 19 studies had a high risk of bias in patient selection, as details of patient recruitment were not reported. 31 studies had a high risk of bias for indicator testing due to the uncertainty of detecting HCC after TACE blinding. The risk of bias for reference standards as well as process and timing was low in all included studies. The risk of bias was generally low for patient selection and reference standards in terms of applicability, while there was an unclear risk for index testing.

Table 1

Characteristics of included studies and patients

Contrast-enhanced MRI

The pooled sensitivity and specificity for detecting HCC after TACE were 0.92 (95% CI: 0.86 to 0.96; I2=92.61%; p<0.01) and 0.94 (95% CI: 0.86 to 0.98; I2=95.04%; p<0.01), respectively (figure 2). In addition, the pooled PLR and NLR for CEMRI were 15.90 (95% CI: 6.20 to 40.74; I2=93.73%; p<0.01) and 0.08 (95% CI: 0.04 to 0.16; I2=94.00%; p<0.01), respectively (online supplemental figure 1). The DOR for CEMRI was 192.49 (95% CI: 46.27 to 800.71; I2=100.00%, p<0.01) (online supplemental figure 2). Finally, the AUC for CEMRI was 0.98 (95% CI: 0.96 to 0.99) (figure 4A).

Figure 2

Summary sensitivity and specificity of CEMRI. CEMRI, contrast-enhanced MRI.

Contrast-enhanced CT

The pooled sensitivity and specificity for detecting HCC after TACE were 0.74 (95% CI: 0.68 to 0.80; I2=86.04%; p<0.01) and 0.98 (95% CI: 0.93 to 1.00; I2=73.67%; p<0.01), respectively (figure 3). In addition, the pooled NLR and PLR for CECT were 41.87 (95% CI: 10.52 to 166.68; I2=62.32%; p<0.01) and 0.26 (95% CI: 0.21 to 0.33; I2=86.99%; p<0.01), respectively (online supplemental figure 3). The DOR for CECT was 192.49 (95% CI: 46.27 to 800.71; I2=100.00%, p<0.01) (online supplemental figure 4). Finally, the AUC for CECT was 0.90 (95% CI: 0.88 to 0.93) (figure 4B).

Figure 3

Summary sensitivity and specificity of CECT. CECT, contrast-enhanced CT.

Figure 4

(A) AUC of CEMRI and (B) AUC of CECT. AUC, area under the receiver operating characteristic curve; CECT, contrast-enhanced CT; CEMRI, contrast-enhanced MRI; SENS, sensitivity; SPEC, specificity;SROC,summary receiver operating characteristic.

Subgroup analysis and meta-regression

Because of the apparent heterogeneity, we used regression analysis and subgroup analysis to explore possible sources of heterogeneity, and the results of CEMRI and CECT are shown in online supplemental figures 5 and 6, respectively. Meta-regression results showed mean age as a possible source of heterogeneity. The results of the subgroup analysis showed that CEMRI showed a higher sensitivity than CECT in all subgroups. In contrast, CEMRI was not superior to CECT in terms of specificity in the ‘year published in 2010 or after’ group, the ‘mean age<60’ group and the ‘percentage of males<80.0’ group. The results of the subgroup analysis are shown in online supplemental table 4.

The pre-test and post-test probabilities were evaluated by Fagan’s nomogram. We set the pre-test probability of CEMRI and CECT at 50% and the results show that the post-test probability is 94% for CEMRI and 98% for CECT (online supplemental figures 7 and 8).

Summary LRP and LRN for the Index Test are shown in online supplemental figures 9 and 10. The findings show the importance of each study for drawing diagnostic inferences.

Publication bias

Publication bias for CEMRI and CECT is shown in online supplemental figures 11 and 12. There was no significant publication bias for studies using CECT (p=0.15) to detect HCC. There was a significant publication bias for studies using CEMRI (p=0.04).

Discussion

This comprehensive systematic review and meta-analysis examined the diagnostic value of CEMRI and CECT in assessing the response to HCC after TACE. A total of 5843 patients from 36 studies were identified, and we observed different patient characteristics in the included studies. In this study, both CEMRI and CECT were found to have relatively high diagnostic value in detecting HCC after TACE. Specifically, there were subtle differences between both CECT and CEMRI in assessing the value of HCC after TACE, with MRI likely having higher sensitivity, DOR and AUC, and reduced NLR. However, CECT may be more valuable in detecting HCC after TACE in terms of specificity, PLR. Due to the limitations of the original study, we did not have the means to make a direct comparison of the diagnostic value of the two diagnostic methods.

The diagnostic values of CEMRI and CECT for detecting HCC after TACE have already been illustrated in previous studies.40 41 A meta-analysis by Zhong et al included 11 studies comparing the diagnostic value of contrast-enhanced ultrasonography (CEUS) and CECT for detecting HCC after TACE. They pointed out that the sensitivity (CEUS, 0.97; CECT, 0.72) and negative predictive value (CEUS, 0.90; CECT, 0.51) of CEUS were better than those of CECT, while the specificity (CEUS, 0.86; CECT, 0.99) of CECT was higher than that of CEUS.40 Moreover, Liu et al conducted a meta-analysis of 13 studies and found that CEMRI had a relatively high diagnostic value (sensitivity: 91%; specificity: 93%; PLR: 12.22; NLR: 0.09; DOR: 126.99; AUC: 0.97) for diagnosing HCC after TACE and hence could be regarded as an alternative diagnostic method.41 However, one of the above studies reported on the diagnostic performance of CECT and CEUS, while the other reported that of CEMRI.

Accurate evaluation of postoperative tumour status is important to guide further treatment strategies and digital subtraction angiography is regarded as the gold standard for diagnosing residual lesions.42 However, healthcare workers should limit digital subtraction angiography because it is invasive, and patients are exposed to radiation. A non-invasive diagnostic tool to detect HCC after TACE is a better option. In our study, we noted the diagnostic values of CEMRI and CECT were relatively high for detecting HCC after TACE. The diagnosis of patients using CECT was fast and reflected the characteristics of HCC after TACE.16 Multi-slice spiral CT with multiphase-enhanced scanning and angiography is widely used in clinical practice to assess tumour response after TACE.38 However, iodized deposition can affect the normal display of active tumour tissues during enhanced scanning, leading to false-negative or false-positive results in the iodized deposition area and inflammatory granulation tissue.43

MRI involves the vertical scanning of the liver and is associated with high resolution of soft tissues, thus clearly displaying the anatomical features of lesions and surrounding tissues. However, non-enhanced MRI cannot detect small lesions, fibrous hyperplasia and inflammatory response in the subcapsular or fibrous compartmentation; therefore, CEMRI is more recommended when these lesions are suspected.44 In addition, we noted CEMRI had high sensitivity and AUC and low NLR for detecting tumours compared with CECT. The possible reason is that CEMRI uses differences in water molecule movement in different tissues to visualise the location of lesions and abnormal tissue structures. Moreover, the formation and integrity of the fibrous tumour envelope significantly correlated with the prognosis of HCC, and the intact tumour envelope could inhibit tumour growth and block the establishment of collateral circulation. Images from CEMRI can reflect the presence of necrosis, liquefaction, haemorrhage and lesions, thus revealing significant tumour envelope lesions after TACE.

We also performed subgroup analyses of the diagnostic performance of CEMRI and CECT. Differences in outcomes were inconsistent at a mean age ≥60.0 years and a proportion of men <80.0%. Several reasons could explain the above results: (1) the mean age of patients was associated with the severity of HCC, and patients with tumours would visit more frequently for follow-up and (2) the diagnostic value of CEMRI and CECT might differ according to the proportion of men, and the potential sex-related difference could explain the prognosis of HCC in various stages.

This analysis provided comprehensive results regarding the diagnostic values of CEMRI and CECT for detecting tumours after TACE. In clinical practice, since CEMRI has high sensitivity and AUC and low NLR for detecting tumours compared with CECT, the results could suggest that CEMRI could be used first after TACE to determine the response to TACE, and CECT could be used for confirmation if necessary. Of course, MRI is not readily available everywhere, and follow-up after TACE should be based on the availability of imaging modalities and patient convenience.

Notwithstanding, it has several limitations. First, the review was based on prospective and retrospective studies. Therefore inevitably suffers from recall or confounding factor bias. Second, although meta-regression and subgroup analyses were performed, the heterogeneity across studies could not be fully explained due to variations in characteristics such as mean tumour diameter, TACE emulsion or the time interval between diagnosis and TACE. Third, as the available data were insufficient to directly compare the diagnostic value of CEMRI and CECT. Finally, the overall results show a significant publication bias, which may be influenced by a number of factors such as missing grey literature.

In conclusion, this study found that both CEMRI and CECT had relatively high predictive power for assessing the response of HCC after TACE. Furthermore, the diagnostic value of CEMRI may be superior to CECT in terms of sensitivity, AUC, DOR and NLR. Further large-scale prospective studies are needed to directly compare the diagnostic performance of CEMRI with CECT for HCC after TACE.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

Ethics statements

Patient consent for publication

Ethics approval

The manuscript we submitted this time is a meta-analysis, which is known as one of the secondary research. It does not involve human participants, so it does not need the approval of the ethics body.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • TL is the guarantor. CZ contributed to conception and design, provision of study materials or patients, and manuscript writing. XC contributed to conception and design, collection and assembly of data, data analysis and interpretation, and manuscript writing. JW contributed to conception and design, data analysis and interpretation, and manuscript writing. TL is the guarantor and contributed to conception and design, administrative support, and manuscript writing. All authors read and approved the final manuscript.

  • Funding This work was supported by the National Natural Science Foundation of China (grant number 81470587).

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.