Inter-reader agreement and imaging-pathology correlation of the LI-RADS M on gadoxetic acid-enhanced magnetic resonance imaging: efforts to improve diagnostic performance

Risk of Bias in Liver Imaging Reporting and Data System Studies Using QUADAS-2

Purpose: Use a tailored version of the Quality Assessment of Diagnostic Accuracy Studies tool to evaluate risk of bias and applicability across LIRADS related publications. Method: A tailored QUADAS-2 tool was created through consensus approach to assess risk of bias and applicability across 37 LI-RADS related publications. Studies were selected from 2017 to 2022 using the assistance of experienced hospital librarians to search for studies evaluating the diagnostic accuracy of CT, MRI, or contrast-enhanced ultrasound for HCC using LI-RADS through multiple different databases. QUADAS-2 assessments were performed in duplicate and independently by 2 authors with experience using the QUADAS-2 tool. Disagreements were resolved with a third expert reviewer. Consensus QUADAS-2 assessments were tabulated for each domain. Results: Using the tailored QUADAS-2 tool, 31 of the 37 included LI-RADS studies were assessed as high risk of bias, and 9 out of 37 studies demonstrated concerns for applicability. Patient selection (21 out of 37 studies) and flow/timing (24 out of 37 studies) domains demonstrated the highest risk of bias. 6 out of 37 studies in the index domain demonstrated high risk of bias. 2 out of 37 studies showed high risk of bias in the reference standard domain. Conclusion: A significant proportion of LI-RADS research is at risk of bias with concerns for applicability. Identifying risk of bias in such research is essential to recognize limitations of a study that may affect the validity of the results. Areas for improvement in LI-RADS research include reducing selection bias, avoiding inappropriate exclusions, and decreasing verification bias.

Inter-reader agreement for CT/MRI LI-RADS category M imaging features: a systematic review and meta-analysis

BACKGROUNDS/AIMS

To systematically evaluate inter-reader agreement in the assessment of individual liver imaging reporting and data system (LI-RADS) category M (LR-M) imaging features in computed tomography/magnetic resonance imaging (CT/MRI) LIRADS v2018, and to explore the causes of poor agreement in LR-M assignment.

METHODS

Original studies reporting inter-reader agreement for LR-M features on multiphasic CT or MRI were identified using the MEDLINE, EMBASE, and Cochrane databases. The pooled kappa coefficient (κ) was calculated using the DerSimonian-Laird random-effects model. Heterogeneity was assessed using Cochran's Q test and I2 statistics. Subgroup meta-regression analyses were conducted to explore the study heterogeneity.

RESULTS

In total, 24 eligible studies with 5,163 hepatic observations were included. The pooled κ values were 0.72 (95% confidence interval [CI], 0.65-0.78) for rim arterial phase hyperenhancement, 0.52 (95% CI, 0.39-0.65) for peripheral washout, 0.60 (95% CI, 0.50-0.70) for delayed central enhancement, 0.68 (95% CI, 0.57-0.78) for targetoid restriction, 0.74 (95% CI, 0.65-0.83) for targetoid transitional phase/hepatobiliary phase appearance, 0.64 (95% CI, 0.49-0.78) for infiltrative appearance, 0.49 (95% CI, 0.30-0.68) for marked diffusion restriction, and 0.61 (95% CI, 0.48-0.73) for necrosis or severe ischemia. Substantial study heterogeneity was observed for all LR-M features (Cochran's Q test, P<0.01; I2≥89.2%). Studies with a mean observation size of <3 cm, those performed using 1.5-T MRI, and those with multiple image readers, were significantly associated with poor agreement of LR-M features.

CONCLUSIONS

The agreement for peripheral washout and marked diffusion restriction was limited. The LI-RADS should focus on improving the agreement of LR-M features.

Imaged periductal infiltration: Diagnostic and prognostic role in intrahepatic mass-forming cholangiocarcinoma

Purpose

This study examines periductal infiltration in intrahepatic mass-forming cholangiocarcinoma (IMCC), focusing on its importance for differentiating hepatic tumors and its influence on post-surgical survival in IMCC patients.

Methods

Eighty-three consecutive patients with IMCC (n = 43) and liver cancer whose preoperative images showed intrahepatic bile duct dilatation adjacent to the tumor for differential diagnosis from hepatocellular carcinoma (HCC) [n = 21], metastatic liver cancer (MLC) [n = 16] and combined hepatocellular-cholangiocarcinoma (cHCC-CC) [n = 3] were enrolled. CT and MRI findings of simple bile duct compression, imaged periductal infiltration, and imaged intrabiliary growth adjacent to the main tumor were reviewed. Clinicopathological and imaging features were compared in each group. The sensitivity, specificity, and odds ratio were calculated for each imaging finding of IMCC versus the other tumor groups. Overall survival was compared between cases of IMCC with and without imaged periductal infiltration.

Results

Simple bile duct compression and imaged intrabiliary growth were more frequently observed in HCC than in the others (p < 0.0001 and 0.040, respectively). Imaged periductal infiltration was observed more often in histopathologically confirmed large-duct type IMCC than in the small-duct type IMCC (p = 0.034). Multivariable analysis demonstrated that only imaged periductal infiltration (odds ratio, 50.67) was independently correlated with IMCC. Patients with IMCC who had imaged periductal infiltration experienced a poorer prognosis than those without imaged periductal infiltration (p = 0.0034).

Conclusion

Imaged periductal infiltration may serve as a significant marker for differentiating IMCC from other liver cancers. It may also have the potential to predict post-surgical outcomes in patients with IMCC.

A Multicenter Assessment of Interreader Reliability of LI-RADS Version 2018 for MRI and CT

Background Various limitations have impacted research evaluating reader agreement for Liver Imaging Reporting and Data System (LI-RADS). Purpose To assess reader agreement of LI-RADS in an international multicenter multireader setting using scrollable images. Materials and Methods This retrospective study used deidentified clinical multiphase CT and MRI and reports with at least one untreated observation from six institutions and three countries; only qualifying examinations were submitted. Examination dates were October 2017 to August 2018 at the coordinating center. One untreated observation per examination was randomly selected using observation identifiers, and its clinically assigned features were extracted from the report. The corresponding LI-RADS version 2018 category was computed as a rescored clinical read. Each examination was randomly assigned to two of 43 research readers who independently scored the observation. Agreement for an ordinal modified four-category LI-RADS scale (LR-1, definitely benign; LR-2, probably benign; LR-3, intermediate probability of malignancy; LR-4, probably hepatocellular carcinoma [HCC]; LR-5, definitely HCC; LR-M, probably malignant but not HCC specific; and LR-TIV, tumor in vein) was computed using intraclass correlation coefficients (ICCs). Agreement was also computed for dichotomized malignancy (LR-4, LR-5, LR-M, and LR-TIV), LR-5, and LR-M. Agreement was compared between research-versus-research reads and research-versus-clinical reads. Results The study population consisted of 484 patients (mean age, 62 years ± 10 [SD]; 156 women; 93 CT examinations, 391 MRI examinations). ICCs for ordinal LI-RADS, dichotomized malignancy, LR-5, and LR-M were 0.68 (95% CI: 0.61, 0.73), 0.63 (95% CI: 0.55, 0.70), 0.58 (95% CI: 0.50, 0.66), and 0.46 (95% CI: 0.31, 0.61) respectively. Research-versus-research reader agreement was higher than research-versus-clinical agreement for modified four-category LI-RADS (ICC, 0.68 vs 0.62, respectively; P = .03) and for dichotomized malignancy (ICC, 0.63 vs 0.53, respectively; P = .005), but not for LR-5 (P = .14) or LR-M (P = .94). Conclusion There was moderate agreement for LI-RADS version 2018 overall. For some comparisons, research-versus-research reader agreement was higher than research-versus-clinical reader agreement, indicating differences between the clinical and research environments that warrant further study. © RSNA, 2023 Supplemental material is available for this article. See also the editorials by Johnson and Galgano and Smith in this issue.

Optimizing CT and MRI criteria for differentiating intrahepatic mass-forming cholangiocarcinoma and hepatocellular carcinoma

BACKGROUND

Accurate diagnosis of intrahepatic mass-forming cholangiocarcinoma (IMCC) is crucial with regard to the choice of patient management and treatment options.

PURPOSE

To evaluate the feasibility and diagnostic performance of the LI-RADS M (LR-M) targetoid criteria on computed tomography (CT) and gadoxetic acid-enhanced magnetic resonance imaging (EOB-MRI) in differentiating IMCC from hepatocellular carcinoma (HCC).

MATERIAL AND METHODS

A total of 118 patients with IMCC and HCC were included who underwent CT and EOB-MRI examinations. Multivariate analysis was used to determine the strongest predictors differentiating IMCC from HCC. Using these predictors, a predictive model for differentiating IMCC from HCC was constructed and the performance of the model was confirmed using the receiver operating characteristic curve.

RESULTS

Multivariate analyses revealed rim-like arterial phase hyperenhancement (rim APHE) on CT and rim APHE, delayed central enhancement (DCE), and targetoid hepatobiliary phase (HBP) on MRI as independent variables significantly differentiating IMCC from HCC. The multivariate logistic regression model incorporating the three variables on EOB-MRI was constructed with an area under the curve (AUC) of 0.946, sensitivity of 87.80%, specificity of 92.21%, and accuracy of 94.60%. Per the DeLong test, the multivariate logistic regression model showed significantly higher AUC than rim APHE on CT (0.946 vs. 0.871; P = 0.008) and MRI (0.946 vs. 0.876; P = 0.003), whereas rim APHE on CT and MRI did not differ significantly (P = 0.809).

CONCLUSION

The multivariate logistic regression model based on rim APHE, DCE, and targetoid HBP on EOB-MRI can effectively distinguish IMCC from HCC and is superior to any other targetoid appearance criterion of LI-RADS on CT and EOB-MRI.

Percentages of Hepatocellular Carcinoma in LI-RADS Categories with CT and MRI: A Systematic Review and Meta-Analysis

Background The Liver Imaging Reporting and Data System (LI-RADS) CT and MRI algorithm applies equally to CT, MRI with extracellular contrast agents (ECA-MRI), and MRI with gadoxetate (Gx-MRI). Purpose To estimate pooled percentages of hepatocellular carcinoma (HCC) and overall malignancy for each LI-RADS category with CT and MRI. Materials and Methods MEDLINE and EMBASE databases were searched for research articles (January 2014-April 2021) reporting the percentages of observations in each LI-RADS category with use of versions 2014, 2017, or 2018. Study design, population characteristics, imaging modality, reference standard, and numbers of HCC and non-HCC malignancies in each category were recorded. A random-effects model evaluated the pooled percentage of HCC and overall malignancy for each category. Results There were 49 studies with 9620 patients and a total 11 562 observations, comprising 7921 HCCs, 1132 non-HCC malignancies, and 2509 benign entities. No HCC or non-HCC malignancies were reported with any modality in the LR-1 category. The pooled percentages of HCC for CT, ECA-MRI, and Gx-MRI, respectively, were 10%, 6%, and 1% for LR-2 (P = .16); 48%, 31%, and 38% for LR-3 (P = .42); 76%, 64%, and 77% for LR-4 (P = .62); 96%, 95%, and 96% for LR-5 (P = .76); 88%, 76%, and 78% for LR-5V or LR-TIV (tumor in vein) (P = .42); and 20%, 30%, and 35% for LR-M (P = .32). Most LR-M (93%-100%) and LR-5V or LR-TIV (99%-100%) observations were malignant, regardless of modality. Conclusion There was no difference in percentages of hepatocellular carcinoma and overall malignancy between CT, MRI with extracellular contrast agents, and MRI with gadoxetate for any Liver Imaging Reporting and Data System categories. © RSNA, 2023 Supplemental material is available for this article See also the editorial by Ronot in this issue.

MRI-Based Radiomics Models to Discriminate Hepatocellular Carcinoma and Non-Hepatocellular Carcinoma in LR-M According to LI-RADS Version 2018

Differentiating hepatocellular carcinoma (HCC) from other primary liver malignancies in the Liver Imaging Reporting and Data System (LI-RADS) M (LR-M) tumours noninvasively is critical for patient treatment options, but visual evaluation based on medical images is a very challenging task. This study aimed to evaluate whether magnetic resonance imaging (MRI) models based on radiomics features could further improve the ability to classify LR-M tumour subtypes. A total of 102 liver tumours were defined as LR-M by two radiologists based on LI-RADS and were confirmed to be HCC (n = 31) and non-HCC (n = 71) by surgery. A radiomics signature was constructed based on reproducible features using the max-relevance and min-redundancy (mRMR) and least absolute shrinkage and selection operator (LASSO) logistic regression algorithms with tenfold cross-validation. Logistic regression modelling was applied to establish different models based on T2-weighted imaging (T2WI), arterial phase (AP), portal vein phase (PVP), and combined models. These models were verified independently in the validation cohort. The area under the curve (AUC) of the models based on T2WI, AP, PVP, T2WI + AP, T2WI + PVP, AP + PVP, and T2WI + AP + PVP were 0.768, 0.838, 0.778, 0.880, 0.818, 0.832, and 0.884, respectively. The combined model based on T2WI + AP + PVP showed the best performance in the training cohort and validation cohort. The discrimination efficiency of each radiomics model was significantly better than that of junior radiologists’ visual assessment (p < 0.05; Delong). Therefore, the MRI-based radiomics models had a good ability to discriminate between HCC and non-HCC in LR-M tumours, providing more options to improve the accuracy of LI-RADS classification.

Risk stratification of LI-RADS M and LI-RADS 4/5 combined hepatocellular cholangiocarcinoma: prognostic values of MR imaging features and clinicopathological factors

OBJECTIVES

To investigate the role of clinicopathological factors and MR imaging factors in risk stratification of combined hepatocellular cholangiocarcinoma (cHCC-CCA) patients who were classified as LR-M and LR-4/5.

METHODS

We retrospectively identified consecutive patients who were confirmed as cHCC-CCA after surgical surgery in our institution from June 2015 to November 2020. Two radiologists evaluated the preoperative MR imaging features independently, and each lesion was assigned with a LI-RADS category. Preoperative clinical data were also collected. Multivariate Cox proportional hazards model was applied to separately identify the independent factors correlated with the recurrence of cHCC-CCAs in LR-M and LR-4/5. Risk stratifications were conducted separately in LR-M and LR-4/5. Recurrence-free survival (RFS) rates and overall survival (OS) rates were analyzed by using the Kaplan-Meier survival curves and log-rank test.

RESULTS

A total of 131 patients with single primary lesion which met the 2019 WHO classification criteria were finally included. Corona enhancement, delayed central enhancement, and microvascular invasion (MVI) were identified as predictors of RFS in LR-M. Mosaic architecture, CA19-9, and MVI were independently associated with RFS in LR-4/5. Based on the number of these independent predictors, patients were stratified into favorable-outcome groups (LR-ML subgroup and LR-4/5L subgroup) and dismal-outcome groups (LR-MH subgroup and LR-4/5H subgroup). The corresponding median RFS for LR-ML, LR-MH, LR-5L, and LR-5H were 25.6 months, 8.2 months, 51.7 months, and 18.1 months.

CONCLUSION

Our study explored the prognostic values of imaging and clinicopathological factors for LR-M and LR-4/5 cHCC-CCA patients, and different survival outcomes were observed among four subgroups after conducting risk stratifications.

KEY POINTS

• Corona enhancement, delayed central enhancement, and MVI were identified as predictors of RFS in cHCC-CCAs which were classified into LR-M. Mosaic architecture, CA19-9, and MVI were independently associated with RFS in cHCC-CCAs which were classified into LR-4/5. • Based on the identified risk factors, LR-M and LR-4/5 cHCC-CCA patients could be stratified into two subgroups respectively, with significantly different RFS and OS. • cHCC-CCA patients from LR-M did not always have worse RFS and OS than those from LR-4/5 in some cases.

Added-value of ancillary imaging features for differentiating hepatocellular carcinoma from intrahepatic mass-forming cholangiocarcinoma on Gd-BOPTA-enhanced MRI in LI-RADS M

OBJECTIVE

To identify the reliable imaging features and added-value of ancillary imaging features for differentiating hepatocellular carcinoma (HCC) and intrahepatic mass-forming cholangiocarcinoma (IMCC) assigned to LI-RADS M on Gd-BOPTA-enhanced MRI.

METHODS

This retrospective study included 116 liver observations assigned to LI-RADS M, including 82 HCC and 34 IMCC histologically confirmed. Before and after adding ancillary imaging features, all variables with a p-value of < 0.05 in univariable analysis were entered into a multivariable logistic regression analysis to build diagnostic model 1 and model 2 to find reliable predictors of HCC diagnosis. Receiver operating characteristic (ROC) analysis and the DeLong test were used to compare the two models.

RESULTS

Forty-nine of 82(59.8%) HCCs had a considerably higher frequency of enhancing "capsule" compared with IMCCs (p < 0.001). Based on LI-RADS major and LR-M features and clinical-pathologic factors, an elevated AFP level (OR = 10.676, 95%CI = 2.125-4.470, p = 0.004) and enhancing "capsule" (OR = 20.558, 95%CI = 4.470-94.550, p < 0.001) were extracted as independent risk factors in Model 1. After adding ancillary imaging features, Male (OR = 23.452, 95%CI = 1.465-375.404, p = 0.026), enhancing "capsule" (OR = 13.161, 95%CI = 1.725-100.400, p = 0.013), septum (OR = 17.983, 95%CI = 1.049-308.181, p = 0.046), small-scale central HBP hyperintensity (OR = 44.386, 95%CI = 1.610-1223.484, p = 0.025) were confirmed as independent significant variables associated with HCC. Model 2 demonstrated significantly superior AUC (0.918 vs 0.845, p = 0.021) compared with Model 1. When any two or more predictors in model 2 were satisfied, sensitivity was 91.46%, and accuracy was at the top (87.93%).

CONCLUSION

Enhancing "capsule" was a reliable imaging feature to help identify HCC. Adding ancillary imaging features improved sensitivity and accuracy for HCC diagnosis with differentiation from IMCC in LR-M.

CT/MRI and CEUS LI-RADS Major Features Association with Hepatocellular Carcinoma: Individual Patient Data Meta-Analysis

Background The Liver Imaging Reporting and Data System (LI-RADS) assigns a risk category for hepatocellular carcinoma (HCC) to imaging observations. Establishing the contributions of major features can inform the diagnostic algorithm. Purpose To perform a systematic review and individual patient data meta-analysis to establish the probability of HCC for each LI-RADS major feature using CT/MRI and contrast-enhanced US (CEUS) LI-RADS in patients at high risk for HCC. Materials and Methods Multiple databases (MEDLINE, Embase, Cochrane Central Register of Controlled Trials, and Scopus) were searched for studies from January 2014 to September 2019 that evaluated the accuracy of CT, MRI, and CEUS for HCC detection using LI-RADS (CT/MRI LI-RADS, versions 2014, 2017, and 2018; CEUS LI-RADS, versions 2016 and 2017). Data were centralized. Clustering was addressed at the study and patient levels using mixed models. Adjusted odds ratios (ORs) with 95% CIs were determined for each major feature using multivariable stepwise logistic regression. Risk of bias was assessed using Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) (PROSPERO protocol: CRD42020164486). Results A total of 32 studies were included, with 1170 CT observations, 3341 MRI observations, and 853 CEUS observations. At multivariable analysis of CT/MRI LI-RADS, all major features were associated with HCC, except threshold growth (OR, 1.6; 95% CI: 0.7, 3.6; P = .07). Nonperipheral washout (OR, 13.2; 95% CI: 9.0, 19.2; P = .01) and nonrim arterial phase hyperenhancement (APHE) (OR, 10.3; 95% CI: 6.7, 15.6; P = .01) had stronger associations with HCC than enhancing capsule (OR, 2.4; 95% CI: 1.7, 3.5; P = .03). On CEUS images, APHE (OR, 7.3; 95% CI: 4.6, 11.5; P = .01), late and mild washout (OR, 4.1; 95% CI: 2.6, 6.6; P = .01), and size of at least 20 mm (OR, 1.6; 95% CI: 1.04, 2.5; P = .04) were associated with HCC. Twenty-five studies (78%) had high risk of bias due to reporting ambiguity or study design flaws. Conclusion Most Liver Imaging Reporting and Data System major features had different independent associations with hepatocellular carcinoma; for CT/MRI, arterial phase hyperenhancement and washout had the strongest associations, whereas threshold growth had no association. © RSNA, 2021 Online supplemental material is available for this article.

MRI-diagnosis of category LR-M observations in the Liver Imaging Reporting and Data System v2018: a systematic review and meta-analysis

OBJECTIVES

We performed a meta-analysis to determine the probability of hepatocellular carcinoma (HCC) and non-HCC malignancies in Liver Imaging Reporting and Data System (LI-RADS) category M (LR-M) observations and the frequency of defined LR-M imaging features on MRI using LI-RADS v2018.

METHODS

We searched the MEDLINE and EMBASE databases to identify studies published from 1 January 2018 to 16 March 2021 reporting the probability of category LR-M in HCC and non-HCC malignancies on MRI. The pooled percentages of HCC and non-HCC malignancies in the LR-M observations were evaluated. Meta-regression analysis was performed to identify factors for study heterogeneity. The frequencies of defined LR-M imaging features were also calculated. Risk of bias and concerns regarding applicability were evaluated using the Quality Assessment of Diagnostic Accuracy Studies-2 tool.

RESULTS

We identified 18 studies reporting the diagnostic performance of the LR-M category (3,812 observations in 3,615 patients), with nine studies reporting the frequencies of LR-M imaging features. The pooled percentages of HCC and non-HCC malignancies in the LR-M observations were 29% (95% confidence interval [CI], 21-38%) and 67% (95%CI, 57-77%), respectively. The study type and inclusion of benign lesions were significant factors for study heterogeneity. Of the 10 LR-M imaging features, rim arterial phase hyperenhancement (APHE) showed the highest frequency in non-HCC malignancies (68%; 95%CI, 61-75%).

CONCLUSIONS

The LR-M category was commonly used to characterize non-HCC malignancies, but also included 29% of HCC. The frequencies of the different LR-M imaging features were variable, with rim APHE showing the highest frequency in non-HCC malignancies.

KEY POINTS

• In the LR-M category using LI-RADS v2018 for MRI, the pooled percentage of malignancies in general was 96%, with 29% HCC and 67% non-HCC malignancies, while the remaining 4% was benign entity. • The study type and inclusion of benign lesions were significant factors contributing to substantial heterogeneity among included studies. • The frequencies of the different LR-M imaging features were variable, with rim APHE showing the highest frequency in non-HCC malignancies.

Multi-arterial phase MRI depicts inconsistent arterial phase hyperenhancement (APHE) subtypes in liver observations of patients at risk for hepatocellular carcinoma

OBJECTIVES

According to LI-RADS, a major discriminating feature between hepatocellular carcinoma (HCC) and non-HCC malignancies is the subtype of arterial phase hyperenhancement (APHE). The aim of this study was to investigate whether APHE subtypes are consistent across multi-arterial phase (mHAP) MRI acquisitions while evaluating reader agreement. Secondarily, we investigated factors that may affect reader agreement for APHE subtype.

METHODS

In this retrospective study, consecutive patients with liver cirrhosis and focal observations who underwent mHAP were included. Five radiologists reviewed MR images in 2 reading sessions. In reading session 1, individual AP series were reviewed and scored for presence of APHE and subtype. In reading session 2, readers scored observations' major and ancillary features and LI-RADS category in the complete MRI examination. Reader agreement was calculated using Fleiss' kappa for binary outcomes and Kendall's coefficient of concordance for LI-RADS categories. Univariate mixed effects logistic regressions were performed to investigate factors affecting agreement.

RESULTS

In total, 61 patients with 77 focal observations were analyzed. Of observations unanimously scored as having APHE, 27.7% showed both rim and nonrim subtypes on mHAP. Inter-reader agreement for APHE subtype ranged from 0.49 (95% CI: 0.33, 0.64) to 0.57 (95% CI: 0.40, 0.74) between reading sessions. Observation size had a trend level effect on rim APHE agreement (p = 0.052).

CONCLUSION

Approximately 1/3 of observations demonstrated inconsistent APHE subtype during mHAP acquisition. Small lesions were particularly challenging. Further guidance on APHE subtype classification, especially when applied to mHAP, could be a focus of LI-RADS refinement.

KEY POINTS

• In a cohort of patients at risk for HCC, 28% of the observations showed inconsistent arterial phase hyperenhancement (APHE) subtypes (rim and nonrim) on multi-arterial phase imaging according to the majority score of 5 independent readers. • Inconsistent APHE subtypes may challenge reliable imaging diagnosis, i.e., LI-RADS categorization, of focal liver observations in patients at risk for HCC.