Do Risk Factors for HCC Impact the Association of CT/MRI LIRADS Major Features With HCC? An Individual Participant Data Meta-Analysis

Abstract

French

Background: Guidelines suggest the Liver Imaging Reporting and Data System (LI-RADS) may not be applicable for some populations at risk for hepatocellular carcinoma (HCC). However, data assessing the association of HCC risk factors with LI-RADS major features are lacking. Purpose: To evaluate whether the association between HCC risk factors and each CT/MRI LI-RADS major feature differs among individuals at-risk for HCC. Methods: Databases (MEDLINE, Embase, Cochrane Central Register of Controlled Trials, and Scopus) were searched from 2014 to 2022. Individual participant data (IPD) were extracted from studies evaluating HCC diagnosis using CT/MRI LI-RADS and reporting HCC risk factors. IPD from studies were pooled and modelled with one-stage meta-regressions. Interactions were assessed between major features and HCC risk factors, including age, sex, cirrhosis, chronic hepatitis B virus (HBV), and study location. A mixed effects model that included the major features, as well as separate models that included interactions between each risk factor and each major feature, were fit. Differences in interactions across levels of each risk factor were calculated using adjusted odds-ratios (ORs), 95% confidence-intervals (CI), and z-tests. Risk of bias was assessed using QUADAS-2. (Protocol: https://osf.io/tdv7j/). Results: Across 23 studies (2958 patients and 3553 observations), the associations between LI-RADS major features and HCC were consistent across several HCC risk factors (P-value range: .09-.99). A sensitivity analysis among the 4 studies with a low risk of bias did not differ from the primary analysis. Conclusion: The association between CT/MRI LI-RADS major features and HCC risk factors do not significantly differ in individuals at-risk for HCC. These findings suggest that CT/MR LI-RADS should be applied to all patients considered at risk by LI-RADS without modification or exclusions, regardless of the presence or absence of the risk factors evaluated in this study.

Keywords

LI-RADS CT/MRI HCC risk factors major features

Introduction

The Liver Imaging Reporting and Data System (LI-RADS, LR) is a framework to standardize the reporting and data collection of liver imaging results in patients at risk of hepatocellular carcinoma (HCC).¹ The CT/MRI LI-RADS diagnostic algorithms consist of 5 major features and 21 ancillary features. The American Association for the Study of Liver Diseases (AASLD) as well as the Organ Procurement and Transplantation Network (OPTN) have endorsed LI-RADS v2018 for diagnostic evaluation of individuals with liver lesions at-risk for HCC.^2,3

LI-RADS v2018 is currently applied to all “at-risk” individuals to non-invasively diagnose HCC. Similarly, other recent guidelines and societies such as those from the AASLD, European Association for the Study of Liver Disease (EASL), Japan Society of Hepatology (JSH), Asian Pacific Association for the Study of the Liver (APASL), and Korean Liver Cancer Association-National Cancer Center Korea (KLCSG-NCC), also support the non-invasive diagnoses of HCC based on imaging features.^4-6 However, each society uses slightly different risk factors to define the threshold for individual populations where they recommend imaging as a non-invasive HCC diagnosis (risk above threshold), which leads to different non-invasive diagnostic thresholds when compared to the approach of LI-RADS v2018.⁷

In particular, the AASLD recommends using LI-RADS imaging features to non-invasively diagnose HCC in individuals with cirrhosis from any etiology, and those with chronic hepatitis B Virus (HBV) at “high-risk” of HCC. The AASLD uses the PAGE-B score to stratify individuals into “high-risk” (PAGE >9) or “low-risk” (PAGE-B <9) categories. The PAGE-B scoring system considers platelet count, age, and gender for HBV individuals (PAGE-B) to stratify the 5-year risk of developing HCC.^2,8-12 However, AASLD recommends that “low-risk” individuals (PAGE-B ≤9) require an invasive biopsy to diagnose HCC.¹³ Contrary to these recent guidelines, LI-RADS permits a non-invasive diagnosis of HCC in “at-risk” populations which is defined as all individuals with chronic HBV, cirrhosis, or a personal history of HCC.^2,14 In summary, recent guidelines recommend that “low-risk” HBV-positive individuals should have a biopsy for HCC diagnosis, while LI-RADS allows for non-invasive diagnosis of HCC in these individuals.

There are many risk factors for HCC that potentially influence the association of LI-RADS imaging features with HCC that are not included in stratification frameworks. Factors such as excess alcohol consumption, metabolic dysfunction-associated steatotic liver disease (MASLD), family history of HCC, obesity, type 2 diabetes mellitus, and smoking among others, outlined in Tang et al, 2018, are not consistently used by major societies as risk factors when stratifying an individuals’ risk of developing HCC across North America, Asia, and Europe.⁷ Additionally, these major societies concur that the primary risk factor for HCC is cirrhosis caused by chronic HBV or chronic hepatitis C virus (HCV).^7,15,16 In the Eastern hemisphere, the primary risk factor for HCC emerges as cirrhosis secondary to chronic HBV, whereas in the Western hemisphere, the primary risk factors for HCC include cirrhosis secondary to chronic HCV as well as excess alcohol consumption and MASLD.¹⁷

Cirrhosis is the leading risk factor for HCC, with a 5-year cumulative risk of 5% to 30%.¹⁸ However, other non-liver disease risk factors, such as older age and male sex, are also strongly associated with HCC risk across various liver disease etiologies.^19,20 Additional risk factors have been incorporated into risk stratification models including obesity, type 2 diabetes mellitus, liver disease etiology, and alcohol consumption, among others.⁷ However, LI-RADS does not currently apply to other at-risk patient groups, such as those with vascular causes of cirrhosis like Fontan-associated heart disease or Budd-Chiari syndrome. Hence, analyzing a wider pool of HCC risk factors can enhance our understanding of the positive or negative associations of LI-RADS major features with HCC, which is crucial to refining tools, like LI-RADS, intended for diverse populations.²¹

The purpose of this study was to evaluate the association between HCC risk factors and each CT/MRI LI-RADS major feature among individuals “at-risk” for HCC. Specifically, it explored the association between LI-RADS major features both with a subset of HCC risk factors in the PAGE-B score, such as age and sex, as well as a broader scope of risk factors, such as cirrhosis and geographic location.

Methods

The LI-RADS individual participant data (IPD) database was leveraged from previous publications^8,22-28 in accordance with the Text Recycling Project.²⁹ There is no overlap with prior or ongoing studies’ objectives. This study is distinct from previous analyses conducted with our IPD database as it includes additional data from recently updated searches, employs different inclusion criteria, and follows a unique primary objective.

This study was approved by the institutional review board (IRB) at (The Ottawa Hospital). A protocol was posted on the Open Science Framework before initiating the study (https://osf.io/tdv7j/). Personal health information was not shared with the project team. Optimal methodological practices for diagnostic test accuracy of systematic reviews were employed.^30,31 Reporting was conducted via the Preferred Reporting Items for a Systematic Review and Meta-analysis of Individual Participant Data (PRISMA-IPD) and PRISMA for Diagnostic Test Accuracy (DTA) statements.^32-34 Data were collected and analyzed on a per-observation basis. Protocol deviations are listed in Supplemental Appendix E1.

Eligibility Criteria

The included CT/MRI studies reported major features as well as the percentage of HCC, overall malignancy for LI-RADS categories 1 to 5, LR-TIV (tumour in vein), and LR-M (probably or definitely malignant, not HCC specific) in individuals at-risk for HCC. CT and MRI techniques were evaluated for each study to ensure compliance with LI-RADS technical imaging guidelines.³⁵ Only liver observations categorized according to CT/MRI LI-RADS v2014/v2017/v2018 were included.^14,24,36,37 Participants were only those at-risk of HCC, as defined by LI-RADS, which encompasses individuals with cirrhosis, chronic hepatitis B viral infection, or a personal history of current or prior HCC.⁴ A reference standard was established to assess bias risk (Supplemental Appendix E2).²² Studies eligible for inclusion were evaluated by 2 authors (EL and JPS) using DistillerSR.³⁸

Study Selection and Data Collection

The search strategy, study selection, and data collection process are provided at https://osf.io/rpg8x and van der Pol et al, 2022.²² A previous search from January 2014 to December 2019 was combined with an updated search from January 2020 to January 2022 (Figure 1). Study data were stored and managed into a master individual participant database using Research Electronic Data Capture (REDCap) tools hosted at (The Ottawa Hospital).^39,40 IPD from corresponding study investigators and published reports were compared for each study. Data validation was performed by a research assistant (EL), graduate student (EL and JPS), and board-certified abdominal radiologist with experience in applying LI-RADS (AC). If data were unclear or inconsistent, investigators were contacted to resolve the differences. If data from multiple readers were available, data from one reader were chosen at random. The LI-RADS IPD steering committee was consulted regarding any issues related to LI-RADS applications and discussions regarding preliminary results.

Figure 1.

Flow diagram shows search results, study review, and exclusion criteria.

Risk of Bias and Applicability

The Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) tool was adapted for application to LI-RADS, which categorized potential sources of bias into 4 domains: patient selection, index test, reference standard, and flow and timing.^22,34 The risk of bias and applicability assessment was conducted in duplicate and independently by 2 authors (JPS, CvdP), each with experience conducting risk of bias assessments for diagnostic test accuracy studies. Differences between the independent authors were resolved by discussion with a third author (MDFM). A pilot of one study with subsequent discussion was performed by the 3 authors to improve the inter-observer agreement. A preferred composite reference standard was established to assess the risk of bias (Supplemental Appendix E2).²²

Outcome Measures

Multivariable outcomes were used to evaluate whether risk factors for HCC impacted the association of LI-RADS major features with HCC. The diagnosis of HCC was primarily based on histopathological findings from either a core needle biopsy, hepatectomy, or explant. If these were not available, a composite reference standard was utilized as detailed in Supplemental Appendix E2. The primary outcomes assessed in the patient population were multivariable-adjusted odds ratios (ORs) of major features with and without that HCC risk factor. These were used to determine the independent association of HCC with CT/MRI LI-RADS major features (size, arterial phase hyperenhancement [APHE], enhancing capsule, and washout) and their interaction with the following risk factors for HCC: age (1-year increase compared to the mean age of 59), sex, cirrhosis history, chronic HBV history, and study-location (North America, Asia, or Europe). Personal history of HCC and threshold growth were not included in the analysis due to limited data available in our database. Although our previous IPD meta-analysis did not find an independent association between threshold growth and HCC,²² differing results have been shown in a recent single-centre study by Choi et al, 2023.⁴¹ Despite this finding, threshold growth was not included in this study since our dataset had a limited sample size, which prevented us from obtaining meaningful results as the models did not converge. Notably, since platelet count was not reported in our data, we could not evaluate all risk factors included in the PAGE-B score, which has been proposed to identify “high-risk” individuals for LI-RADS diagnosis of HCC.

Statistical Analysis

To determine whether the multivariable-adjusted ORs reflecting the association of major features with HCC were impacted by risk factors for HCC, IPD were pooled across studies and modelled simultaneously with one-stage meta-regressions by interacting each of the following HCC risk factors: age, sex, cirrhosis history, chronic HBV history, and study location, individually, with CT/MRI LI-RADS major features. More specifically, a multivariable generalized linear model with mixed-effects (GLMM), clustering for patient, observation, and study level correlation effects, was used to create a “simple” model including the HCC risk factor and LI-RADS major feature variables. Then, separate GLMMs were created to add interaction effects between all major features and each risk factor (one GLMM per risk factor). To compare the goodness-of-fit between the multivariable GLMMs with and without interactions, and to see if it is beneficial to include interaction terms in our model, the likelihood ratio test (LRT) was used.

Adjusted ORs for all the variables for HCC were calculated with 95% confidence intervals (CIs). The strength of the independent association of the variables and interactions with the outcome of interest was determined based on the statistical significance and the magnitude of the adjusted ORs derived in the multivariable models. Differences between adjusted ORs of major feature association to HCC with and without HCC risk factors were determined by z-test.

Publication bias was not assessed as per contemporary guidance for DTA systematic reviews.³³ Tau², the estimate of between-study variance, was calculated using a maximum-likelihood estimator. All analyses were performed using the “lme4” package in R (R Core Team, version 4.2.2; R Foundation for Statistical Computing).^42-44 The level of significance was set at P < .05.

Sensitivity Analyses

A sensitivity analysis was performed by limiting the same analyses to studies at an overall low risk of bias. Results were reported if the number of positive observations allowed for the GLMM to converge and obtain stable estimates.

Results

Study Selection and Characteristics

Of the 873 unique studies that were initially identified, 23 were selected for the final population because they reported risk factors for HCC, such as age, sex, cirrhosis history, chronic HBV history, and study location, and evaluated CT/MRI LI-RADS major features.²² These 23 studies comprised 22 full manuscripts and one published conference abstract. Characteristics of the included studies are in Table 1 and Supplemental Appendix E3. There were 2958 total patients with 3553 observations: 90 patients (133 observations) were imaged with CT, 2148 patients (2584 observations) were imaged with MRI, and 736 patients (836 observations) were imaged with both CT and MRI as seen in Figure 1. The prevalence of overall HCC, malignancy, and benign lesions for each modality and the study population are in Table 2.

Table 1.

Characteristics of Included Studies in Terms of Study Details, Imaging Technique, and Observation Data.

Study details				Imaging technique				Observation data
Ref no.	Country	Design	Prevailing risk factor	Modality	Contrast agent	LI-RADS version	No. of readers	No. of liver observations/patients	No. of HCCs
Alhassan A 2019⁴⁵	Canada	RC	Cirr >> HBV	CT	ECA	v2017	2	98/98	69
Allen BC 2018⁴⁶	USA	RCCon	Cirr >> HBV	MRI	HPB	v2014	3	117/64	18
Cerny M 2018⁴⁷	Canada	RC	Cirr >> HBV	MRI	ECA	v2014	2	193/98	108
Chen J 2019⁴⁸	China	RC	HBV > Cirr	MRI	ECA, HPB	v2018	2	149/149	149
Choi SH 2019⁴⁹	South Korea	RC	HBV > Cirr	MRI	HPB	v2018	NR	254/253	198
Kim DH 2021⁵⁰	South Korea	RC	Cirr > HBV	MRI	Primovist	v2018	3	114/114	0
Choi SH 2021 Hepint⁹	South Korea	RC	HBV >> Cirr	CT+MRI	Primovist	v2018	2	279/253	247
Choi SH 2022⁵¹	South Korea	RC	HBV >> Cirr	MRI	Primovist	v2018	2	291/222	208
Clarke CGD 2021⁵²	UK	RC	Cirr	MRI	HPB	v2018	2	105/47	70
Fraum TJ 2018⁵³	USA	RC	Cirr >> HBV	MRI	HPB	v2014	2	167/165	122
Fraum TJ 2020⁵⁴	USA	RCCon	Cirr >> HBV	CT+MRI	ECA	v2018	2	331/331	81
Lim H 2022⁵⁵	South Korea	RC	Cirr >> HBV	CT+MRI	CT: Ultravist, MRI: Primovist	v2018	2	226/162	130
Jiang H 2019⁵⁶	China	PC	HBV >> Cirr	MRI	HPB	v2018	2	220/220	201
Joo I 2017⁵⁷	South Korea	RC	HBV >> Cirr	MRI	CT: ECA, MRI: HPB	v2014	2	140/140	70
Kang HJ 2020⁵⁸	China	RC	At risk for HCC	MRI	ECA	v2014	2	19/19	15
Kierans AS 2019⁵⁹	USA	RC	Cirr > HBV	MRI	ECA, HPB	v2017	3	138/108	77
Kim YY 2019⁶⁰	South Korea	RC	Cirr > HBV	MRI	HPB	v2018	2	220/220	165
Ronot M 2018⁶¹	France	PC	Cirr >> HBV	CT+MRI	HPB	v2014	1	583/479	334
Rosiak G 2018⁶²	Poland	RC	Cirr >> HBV	MRI	HPB	v2017	2	69/32	50
Song JS 2019⁶³	South Korea	RC	HBV ~ Cirr	MRI	ECA, HPB	v2014	2	146/107	146
Stocker D 2020⁶⁴	Switzerland	RC	Cirr > HBV	MRI	ECA	v2018	4	65/64	27
van der Pol CB 2022²²	Canada	RC	HBV > Cirr	MRI	ECA, HPB	v2018	2	214/78	67
Zhang L 2019⁶⁵	China	RC	HBV ~ Cirr	MRI	ECA	v2018	2	82/82	82

Note. LI-RADS = Liver Imaging Reporting and Data System; HBV = hepatitis B virus; HCC = hepatocellular carcinoma; CT/MRI = Computed Tomography/Magnetic Resonance Imaging; CI = confidence interval; USA = Unites States of America; UK = United Kingdom; ECA = extracellular contrast agent; HPB = hepatobiliary contrast agent; Ref = reference; PC = prospective cohort; RC = retrospective cohort; RCCon = retrospective case-control; Cirr = cirrhosis; P = pathology; C = clinical; CCRS = composite reference standard.

Table 2.

Overall Characteristics of All Patients for CT/MRI LI-RADS Categories.

Variable	Total (all modalities)	CT	MRI	CT + MRI
Patients	2958^a	90	2148	736
Observations	3553	133	2584	836
Mean age ± SD	59.08 ± 10.93	62.92 ± 10.08	58.08 ± 11.12	61.58 ± 9.91
Female	22.60% (803/3553)	24.06% (32/133)	22.64% (585/2584)	22.37% (187/836)
Male	77.40% (2750/3553)	75.94% (101/133)	77.36% (1999/2584)	77.63% (649/836)
Prevalence of HCC	67.27% (2390/3553)	66.17% (88/133)	66.52% (1719/2584)	69.74% (583/836)
Prevalence of overall malignancy	9.96% (354/3553)	13.53% (18/133)	12.54% (324/2584)	1.44% (12/836)
Prevalence of benign lesions	15.28% (543/3553)	9.02% (12/133)	16.52% (427/2584)	12.44% (104/836)

Note. HBV = hepatitis B virus; CT/MRI = Computed Tomography/Magnetic Resonance Imaging; LI-RADS = Liver Imaging Reporting and Data System; HCC = hepatocellular carcinoma.

Total patients in the entire cohort does not have the same sum as each imaging modality summed separately since there were patients with multiple observations taken on different imaging modalities. There were 2958 total patients with 3553 observations, where 90 patients (133 observations) were imaged on CT, 2148 patients (2584 observations) were imaged on MRI, and 736 patients (836 observations) were imaged on both CT and MRI.

Risk of Bias and Applicability

Of the 23 included studies, 17% (4/23) were at low risk of bias, and 96% (22/23) had low applicability concerns (Figure 2 and Supplemental Appendix E4). The overall assessment determined that 83% (19/23) of studies were at high risk of bias with 63% (12/19) of those studies being at high risk of bias in more than one domain. The domain most often at risk for bias was study flow and timing (65% of included studies, 15/23), which were primarily due to inappropriate or unclear intervals between the index test and reference standard (33% of these studies, 5/15). Specifically, 2 studies had an index imaging to pathological reference standard time interval ranging from zero to greater than 2 years for HCC (25, 32). Patient and observation selection were unclear or at high risk of bias for 65% (15/23) of included studies due to case-control study design. The reference standard domains were at low risk of bias for 78% (18/23) of studies.

Figure 2.

Risk of bias assessment for 23 included studies using the Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) tool which consists of 4 domains.

Association Between Major Features and Risk Factors on CT/MRI With HCC

The adjusted ORs for the associations between CT/MRI major features with HCC risk factors compared to the reference risk factor level are presented in Table 3. When comparing the GLMMs that included interactions between major features to patient age, sex, and HBV history to the GLMM without interactions, no significant differences were found in the associations of major features and risk factors with HCC to the associations of major features alone (P-value = .13, P-value = .22, and P-value = .35, respectively). GLMMs that included interactions between major features and cirrhosis and study location showed significantly greater accuracy to the IPD compared to GLMMs without these interactions (P-value = .04 and P < .01, respectively; Supplemental Appendices E5 and E6). The “P-value range” represents the range of P-values for a specific risk factor across the major features as shown in Table 3. For the following risk factors, no differences in association of major features with HCC were identified: increase in age (1-year increase compared to the mean patient age of 59, P-value range = .88-.99), sex (male vs female, P-value range = .63-.97), cirrhosis status (patients with vs without cirrhosis, P-value range = .30-.77), chronic HBV status (patients with vs without chronic HBV, P-value range = .41-.98), and study location (studies conducted in Asia or Europe vs those in North America, P-value range = .09-.98 and P-value range = .23-.98, respectively).

Table 3.

Multivariable Logistic Regression With Mixed-Effects and Interaction Analysis for CT/MRI LI-RADS Major Features, Except Threshold Growth, and Risk Factors Influencing the Association to HCC. Models Included 2958 Total Patients With 3553 Observations.

Interaction model^a
APHE (no APHE—reference)		P-value*
For a 1-y increase in age	At mean age = 59.03
5.94 [4.63-7.61]	5.84 [4.65-7.32]	.92
Male	Female
5.58 [2.03-15.36]	7.27 [4.52-11.69]	.64
With cirrhosis	Without cirrhosis
6.86 [2.30-20.49]	3.60 [2.15-6.04]	.30
With chronic HBV	Without chronic HBV
6.03 [2.83-12.85]	5.97 [4.41-8.08]	.98
Location-Asia	Location-North America
7.24 [2.86-18.30]	3.78 [2.50-5.73]	.21
Location-Europe	Location-North America
8.08 [2.54-25.67]	3.78 [2.50-5.73]	.23
Lesion size 10-19 mm (<10 mm—reference)		P-value*
For a 1-y increase in age	At mean age = 59.03
2.34 [1.64-3.36]	2.43 [1.75-3.38]	.88
Male	Female
2.09 [0.51-8.58]	3.07 [1.58-5.97]	.63
With cirrhosis	Without cirrhosis
2.10 [0.27-16.53]	4.46 [1.63-12.21]	.52
With chronic HBV	Without chronic HBV
3.11 [1.09-8.89]	1.95 [1.32-2.87]	.41
Location-Asia	Location-North America
1.98 [0.62-6.30]	2.15 [1.35-3.45]	.90
Location-Europe	Location-North America
3.80 [0.69-20.91]	2.15 [1.35-3.45]	.53
Lesion size ≥20 mm (<10 mm—reference)		P-value*
For a 1-y increase in age	At mean age = 59.03
3.44 [2.41-4.92]	3.53 [2.55-4.90]	.92
Male	Female
3.48 [0.87-13.95]	2.80 [1.46-5.36]	.78
With cirrhosis	Without cirrhosis
3.07 [0.44-21.19]	6.00 [2.35-15.36]	.54
With chronic HBV	Without chronic HBV
3.97 [1.39-11.29]	3.05 [2.05-4.53]	.64
Location-Asia	Location-North America
1.86 [0.58-5.96]	5.58 [3.45-9.03]	.09
Location-Europe	Location-North America
8.72 [1.50-50.85]	5.58 [3.45-9.03]	.63
Enhancing capsule (no enhancing capsule—reference)		P-value*
For a 1-y increase in age	At mean age = 59.03
3.83 [2.82-5.18]	3.84 [2.91-5.07]	.99
Male	Female
3.91 [1.09-13.99]	3.82 [2.10-6.97]	.97
With cirrhosis	Without cirrhosis
4.11 [1.07-15.80]	3.09 [1.63-5.84]	.71
With chronic HBV	Without chronic HBV
3.26 [1.29-8.23]	4.20 [2.94-5.98]	.62
Location-Asia	Location-North America
4.02 [1.20-13.48]	3.12 [1.83-5.33]	.71
Location-Europe	Location-North America
3.17 [0.84-11.96]	3.12 [1.83-5.33]	.98
Washout (no washout—reference)		P-value*
For a 1-y increase in age	At mean age = 59.03
4.26 [3.35-5.41]	4.27 [3.43-5.32]	.99
Male	Female
4.16 [1.57-11.04]	4.43 [2.81-7.00]	.91
With cirrhosis	Without cirrhosis
4.47 [1.51-13.29]	3.75 [2.24-6.29]	.77
With chronic HBV	Without chronic HBV
3.63 [1.75-7.64]	4.56 [3.43-6.08]	.57
Location-Asia	Location-North America
2.93 [1.11-7.73]	5.81 [3.74-9.02]	.21
Location-Europe	Location-North America
7.64 [2.46-23.69]	5.81 [3.74-9.02]	.66

Note. Reference categories: Enhancing capsule (reference category = no enhancing capsule), APHE (reference category = no APHE), lesion size (<10 mm—reference, 10-19 mm, ≥20 mm), washout (reference category = no nonperipheral washout), age (measured continuously and centred around the mean), sex (reference category = women), cirrhosis history (reference category = no cirrhosis history), chronic HBV history (reference category = no history of HBV), and study location (North America—reference, Asia, Europe).

Adjusted odds ratios were calculated from separate GLMMs for each HCC-relevant risk factor interacting with major features.

P-values were calculated using the z-scores from the difference of adjusted odds ratios.

Sensitivity Analyses

When restricting analyses to 4 studies considered low-risk of bias (Supplemental Appendices E9 and E13), the low-risk population (695 patients, 862 observations) showed no difference between adjusted ORs of major features with and without HCC risk factors: increase in age (1-year increase compared to the mean patient age of 59, P-value range = .88-1.00), sex (male vs female patients, P-value range = .31-.73), chronic HBV status (patients with vs without chronic HBV, P-value range = .30-.99), and study location (studies conducted in Europe vs those in North America, P-value range = .37-1.00). Results for the study location-Asia variable could not be obtained as there were no low-risk of bias studies originating from this study region. Additionally, P-values for the difference between adjusted ORs with and without cirrhosis could not be obtained due to the small sample size and wide confidence intervals ranging from 0 to infinity. Considerable variability persists in this sensitivity analysis, as indicated by the range of τ² for interactions being 1.92 to 2.31 (Supplemental Appendix E11).

Discussion

This IPD meta-analysis of more than 3000 observations found that the assessed risk factors (age, sex, cirrhosis history, chronic HBV history, and study location) did not influence the strength of association of CT/MRI LI-RADS major features with HCC. Importantly, this finding persisted in a sensitivity analysis of the risk of bias studies, suggesting that the CT/MRI LI-RADS major features are associated with HCC independent of the assessed risk factors. Guidelines that recommend altering LI-RADS application based on patient risk factors should consider this evidence.

Prior studies have assessed risk factors for HCC. Based on the risk factors with sufficient data in our IPD analysis, Tang et al, 2018, identified increased age, male sex, cirrhosis, and chronic HBV as risk factors that increased the incidence of HCC. Additionally, they found that the age-standardized incidence rate of HCC was highest in Asia (20 cases per 100 000), lower in North America (10-15 cases per 100 000), and lowest in Southern and Western Europe (10 cases per 100 000).⁷ Although these factors are associated with HCC risk, our study importantly demonstrated that these HCC risk factors do not influence the strength of the association between CT/MRI LI-RADS major features and the diagnosis of HCC. Therefore, irrespective of an individual’s HCC risk factors, LI-RADS 5 (Definite HCC) criteria can be applied to all individuals considered “at-risk” for developing HCC.

Using the QUADAS-2 tool, most studies (19 of 23) had a high risk of bias. However, a sensitivity analysis with a low risk of bias yielded the same outcomes, suggesting that the association of CT/MRI LI-RADS major features with HCC is not influenced by HCC risk factors. Despite this, we acknowledge that the high proportion of high risk of bias studies dampens the strength of conclusions. The authors note that at present, there is no other body of evidence at a low risk of bias to refer to as an alternative. This should serve as a call to liver imaging researchers to strive for more high-quality research based on the areas identified as commonly at risk of bias.^27,66 The main sources of bias were patient and observation selection as well as flow and timing, as a considerable proportion of the observations were at high risk or unclear. Patient selection bias can occur when patients are included based on imaging and pathology availability, which leads to potentially excluding those meeting LI-RADS criteria but managed conservatively without pathology. Observation selection bias may arise when inclusion or exclusion criteria favour liver observations with a high suspicion of malignancy (LI-RADS 5), potentially underrepresenting less suspicious lesions (LI-RADS 3). Moreover, many studies had an inappropriate interval between multiphase CT, MRI, and reference standard. In future research, it is crucial to address these sources of bias to ensure more robust results.

Our study had important limitations in addition to the risk of bias concerns described above. First, many datasets were unavailable for analysis (less than half of the eligible data sets were made available by their original authors), which precluded more detailed analyses. In addition, a lack of clarity regarding patient risk factors further precluded the use of several data sets. As per our protocol, we intend to conduct a follow-up report for CEUS once we collect more data that facilitates meaningful analysis. Additionally, threshold growth, while not frequently evaluated in the literature, was not included as a major feature in this multivariable analysis due to the low sample size. However, in our prior IPD analysis, threshold growth was not independently associated with HCC.²² Furthermore, there is variation in surveillance recommendations in non-HBV patients with advanced fibrosis in the absence of cirrhosis. We found the presence of cirrhosis did not impact the association of LI-RADS major features but this analysis included patients with HBV infection. We did not have sufficient numbers to evaluate non-cirrhotic metabolic dysfunction-associated steatotic liver disease (MASLD) or HCV infection, and this serves as an area for future research on LI-RADS.⁶⁷ Moreover, although the inclusion of certain at-risk patients, such as those with vascular causes of cirrhosis, in the LI-RADS framework is beyond the scope of this study, our findings suggest that these populations might warrant consideration in future research to expand the LI-RADS framework. Finally, at present, the LI-RADS IPD does not have data on platelet count, which prevents the calculation of PAGE-B scores and comparison of diagnostic performance in the lower-risk HBV population (PAGE-B ≤9) and other at-risk populations. Therefore, a direct assessment of the AASLD-recommended exclusion of the PAGE-B ≤9 cohort from the LI-RADS application could not be performed.^2,3 Notably, 2 of the 3 variables that contribute to the PAGE-B score (age and sex) were not associated with differences in the utility of major features.

In conclusion, risk factors for HCC do not impact the strength of the association between CT/MRI LI-RADS major features and the diagnosis of HCC. However, a high proportion of the studies included were susceptible to bias, and we could not conduct an analysis stratified by PAGE-B score. These represent opportunities for further research. Nevertheless, this study suggests that the LI-RADS v2018 CT/MRI approach to evaluating major features applies to all populations currently deemed at-risk by LI-RADS. Notably, patients at low risk for HCC, for whom LI-RADS is not presently applicable, were not included and generalizations to this population cannot be made based on the results of this study.

Supplemental Material

sj-docx-1-caj-10.1177_08465371241306297 – Supplemental material for Do Risk Factors for HCC Impact the Association of CT/MRI LIRADS Major Features With HCC? An Individual Participant Data Meta-Analysis

Supplemental material, sj-docx-1-caj-10.1177_08465371241306297 for Do Risk Factors for HCC Impact the Association of CT/MRI LIRADS Major Features With HCC? An Individual Participant Data Meta-Analysis by Robert G. Adamo, Eric Lam, Jean-Paul Salameh, Christian B. van der Pol, Stacy M. Goins, Haben Dawit, Andreu F. Costa, Brooke Levis, Amit G. Singal, Victoria Chernyak, Claude B. Sirlin, Mustafa R. Bashir, An Tang, Ayman Alhasan, Brian C. Allen, Caecilia S. Reiner, Christopher Clarke, Daniel R. Ludwig, Milena Cerny, Jin Wang, Sang Hyun Choi, Tyler J. Fraum, Bin Song, Ijin Joo, So Yeon Kim, Heejin Kwon, Hanyu Jiang, Hyo-Jin Kang, Andrea S. Kierans, Yeun-Yoon Kim, Maxime Ronot, Joanna Podgórska, Grzegorz Rosiak, Ji Soo Song and Matthew D. F. McInnes in Canadian Association of Radiologists Journal

Footnotes

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Amit Singal has served as a consultant or on advisory boards for Genentech, AztraZeneca, Eisai, Exelixis, Bayer, Elevar, Boston Scientific, Sirtex, Histosonics, FujiFilm Medical Sciences, Exact Sciences, Roche, Abbott, Glycotest, Freenome, and GRAIL.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: CIHR Operating Research Grant, Radiological Society of North America (RSNA) R&E Foundation Research Scholar Grant, Mach-Gaensslen Foundation of Canada Grant, Fonds de recherche du Québec en Santé (FRQS #298509 to Dr. An Tang), National Institute of Health (hU01 CA283935 to Dr. Amit Singal).

Protocol Registration

Open Science Framework: ().

ORCID iDs

Robert G. Adamo

Christian B. van der Pol

Victoria Chernyak

An Tang

Supplemental Material

Supplemental material for this article is available online.

References

American College of Radiology. Liver Reporting & Data System. 2018. Accessed October 30, 2020. https://www.acr.org/Clinical-Resources/Reporting-and-Data-Systems/LI-RADS

Singal

Llovet

Yarchoan

, et al. AASLD practice guidance on prevention, diagnosis, and treatment of hepatocellular carcinoma. Hepatology. 2023;78(6):1922-1965. doi:10.1097/HEP.0000000000000466

Kierans

Chernyak

Mendiratta-Lala

Sirlin

Hecht

Fowler

. The Organ Procurement and Transplantation Network hepatocellular carcinoma classification: alignment with Liver Imaging Reporting and Data System, current gaps, and future direction. Liver Transpl. 2023;29(2):206-216. doi:10.1002/lt.26570

EASL Clinical Practice Guidelines on non-invasive tests for evaluation of liver disease severity and prognosis - 2021 update. J Hepatol. 2021;75(3):659-689. doi:10.1016/j.jhep.2021.05.025

Hasegawa

Takemura

Yamashita

, et al. Clinical practice guidelines for hepatocellular carcinoma: the Japan Society of Hepatology 2021 version (5th JSH-HCC guidelines). Hepatol Res. May 2023;53(5):383-390. doi:10.1111/hepr.13892

Omata

Cheng

Kokudo

, et al. Asia-Pacific clinical practice guidelines on the management of hepatocellular carcinoma: a 2017 update. Hepatol Int. 2017;11(4):317-370. doi:10.1007/s12072-017-9799-9

Tang

Hallouch

Chernyak

Kamaya

Sirlin

. Epidemiology of hepatocellular carcinoma: target population for surveillance and diagnosis. Abdom Radiol (NY). 2018;43(1):13-25. doi:10.1007/s00261-017-1209-1

van der Pol

Lim

Sirlin

, et al. Accuracy of the liver imaging reporting and data system in computed tomography and magnetic resonance image analysis of hepatocellular carcinoma or overall malignancy-a systematic review. Gastroenterology. 2019;156(4):976-986. doi:10.1053/j.gastro.2018.11.020

Lee

Kim

Shin

, et al. CT and MRI liver imaging reporting and data system version 2018 for hepatocellular carcinoma: a systematic review with meta-analysis. J Am Coll Radiol. 2020;17(10):1199-1206. doi:10.1016/j.jacr.2020.06.005

10.

Shin

Lee

Bae

, et al. Contrast-enhanced ultrasound liver imaging reporting and data system for diagnosing hepatocellular carcinoma: a meta-analysis. Liver Int. 2020;40(10):2345-2352. doi:10.1111/liv.14617

11.

Tang

Bashir

Corwin

, et al. Evidence supporting LI-RADS major features for CT- and MR imaging-based diagnosis of hepatocellular carcinoma: a systematic review. Radiology. 2018;286(1):29-48. doi:10.1148/radiol.2017170554

12.

Shin

Lee

Yoon

Chung

Choi

Park

. LI-RADS major features on MRI for diagnosing hepatocellular carcinoma: a systematic review and meta-analysis. J Magn Reson Imaging. 2021;54(2):518-525. doi:10.1002/jmri.27570

13.

Moctezuma-Velázquez

Lewis

Lee

, et al. Non-invasive imaging criteria for the diagnosis of hepatocellular carcinoma in non-cirrhotic patients with chronic hepatitis B. JHEP Rep. 2021;3(6):100364. doi:10.1016/j.jhepr.2021.100364

14.

American College of Radiology. Liver Imaging Reporting and Data Systems (LI-RADS) v2018. Accessed September 20, 2022. https://www.acr.org/Clinical-Resources/Reporting-and-Data-Systems/LI-RADS

15.

Toh

Wong

EYT

Wong

, et al. Global epidemiology and genetics of hepatocellular carcinoma. Gastroenterology. 2023;164(5):766-782. doi:10.1053/j.gastro.2023.01.033

16.

Fattovich

Stroffolini

Zagni

Donato

. Hepatocellular carcinoma in cirrhosis: incidence and risk factors. Gastroenterology. 2004;127(5 Suppl 1):S35-S50. doi:10.1053/j.gastro.2004.09.014

17.

Choo

Tan

Goh

BKP

Tai

Zhu

. Comparison of hepatocellular carcinoma in Eastern versus Western populations. Cancer. 2016;122(22):3430-3446. doi:10.1002/cncr.30237

18.

Hung

Liang

Hsu

, et al. Association between complicated liver cirrhosis and the risk of hepatocellular carcinoma in Taiwan. PLoS One. 2017;12(7):e0181858. doi:10.1371/journal.pone.0181858

19.

Kohli

. The relationship between hepatocellular carcinoma and hepatitis B and C virus. Gastroenterol Hepatol (N Y). 2016;12(2):116-118.

20.

Michielsen

. Viral hepatitis B and hepatocellular carcinoma. Acta Gastroenterol Belg. 2011;74(1):4-8.

21.

Ramalho

Matos

AlObaidy

Velloni

Altun

Semelka

. Magnetic resonance imaging of the cirrhotic liver: diagnosis of hepatocellular carcinoma and evaluation of response to treatment - part 2. Radiol Bras. 2017;50(2):115-125. doi:10.1590/0100-3984.2015.0140

22.

van der Pol

McInnes

MDF

Salameh

, et al. CT/MRI and CEUS LI-RADS major features association with hepatocellular carcinoma: individual patient data meta-analysis. Radiology. 2022;302(2):326-335. doi:10.1148/radiol.2021211244

23.

van der Pol

McInnes

MDF

Salameh

J-P

, et al. Impact of reference standard on CT, MRI, and contrast-enhanced US LI-RADS diagnosis of hepatocellular carcinoma: a meta-analysis. Radiology. 2022;303(3):544-545. doi:10.1148/radiol.212340

24.

Goins

Adamo

Lam

, et al. Conversion strategy for LI-RADS category 5 observations across versions 2014, 2017, and 2018. Radiology. 2023;307(4):e222971. doi:10.1148/radiol.222971

25.

Goins

Jiang

van der Pol

, et al. Individual participant data meta-analysis of LR-5 in LI-RADS version 2018 versus revised LI-RADS for hepatocellular carcinoma diagnosis. Radiology. 2023;309(3):e231656. doi:10.1148/radiol.231656

26.

Goins

Jiang

van der Pol

, et al. Comparative performance of 2018 LI-RADS versus modified LIRADS (mLI-RADS): an individual participant data meta-analysis. J Magn Reson Imaging. 2024;60(3):1082-1091. doi:10.1002/jmri.29167

27.

van der Pol

Costa

Lam

Dawit

Bashir

McInnes

MDF

. Best practice for MRI diagnostic accuracy research with lessons and examples from the LI-RADS individual participant data group. J Magn Reson Imaging. 2024;60(1):21-28. doi:10.1002/jmri.29049

28.

Dawit

Lam

McInnes

MDF

, et al. LI-RADS CT and MRI ancillary feature association with hepatocellular carcinoma and malignancy: an individual participant data meta-analysis. Radiology. 2024;310(2):e231501. doi:10.1148/radiol.231501

29.

Hall

Moskovitz

Pemberton

. Understanding text recycling: a guide for researchers. Text Recycling Research Project. 2021:1.

30.

de Vet

HCW

Eisinga

Riphagen

Aertgeerts

Pewsner

. Chapter 7: Searching for studies. In: Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy Version 0.4. The Cochrane Collaboration, 2008.

31.

McInnes

Bossuyt

. Pitfalls of systematic reviews and meta-analyses in imaging research. Radiology. 2015;277(1):13-21. doi:10.1148/radiol.2015142779

32.

McGrath

Alabousi

Skidmore

, et al. Recommendations for reporting of systematic reviews and meta-analyses of diagnostic test accuracy: a systematic review. Syst Rev. 2017;6(1):194. doi:10.1186/s13643-017-0590-8

33.

McInnes

Moher

Thombs

McGrath

Bossuyt

. Preferred Reporting Items for a Systematic-Review and Meta-analysis of Diagnostic Test Accuracy studies: the PRISMA-DTA statement. JAMA. 2018;319(4):388-396.

34.

Cohen

Deeks

Hooft

, et al. Preferred reporting items for journal and conference abstracts of systematic reviews and meta-analyses of diagnostic test accuracy studies (PRISMA-DTA for Abstracts): checklist, explanation, and elaboration. BMJ. 2021;372:n265. doi:10.1136/bmj.n265

35.

Kambadakone

Fung

Gupta

, et al. LI-RADS technical requirements for CT, MRI, and contrast-enhanced ultrasound. Abdom Radiol (NY). 2018;43(1):56-74. doi:10.1007/s00261-017-1325-y

36.

American College of Radiology. Liver Imaging Reporting and Data Systems (LI-RADS) v2014. Accessed September 15, 2022. https://nrdr.acr.org/lirads/

37.

American College of Radiology. Liver Imaging Reporting and Data Systems (LI-RADS) v2017. Accessed September 15, 2022. https://www.acr.org/Clinical-Resources/Reporting-and-Data-Systems/LI-RADS

38.

DistillerSR. Version 2.35. DistillerSR Inc.; 2022. Accessed April-June 2022. https://www.distillersr.com/

39.

Harris

Taylor

Thielke

Payne

Gonzalez

Conde

. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377-381. doi:10.1016/j.jbi.2008.08.010

40.

Harris

Taylor

Minor

, et al. The REDCap consortium: building an international community of software platform partners. J Biomed Inform. 2019;95:103208. doi:10.1016/j.jbi.2019.103208

41.

Choi

Kim

, et al. Value of threshold growth as a major diagnostic feature of hepatocellular carcinoma in LI-RADS. J Hepatol. 2023;78(3):596-603. doi:10.1016/j.jhep.2022.11.006

42.

Lüdecke

. ggeffects: tidy data frames of marginal effects from regression models. J Open Source Softw. 2018;3:772. doi:10.21105/joss.00772

43.

Viechtbauer

. Conducting meta-analyses in R with the metafor package. J Stat Softw. 2010;36:1-48.

44.

Bates

Mächler

Bolker

Walker

. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67(1):1-48. doi:10.18637/jss.v067.i01

45.

Alhasan

Cerny

Olivié

, et al. LI-RADS for CT diagnosis of hepatocellular carcinoma: performance of major and ancillary features. Abdom Radiol (NY). 2019;44(2):517-528. doi:10.1007/s00261-018-1762-2

46.

Allen

Jaffe

Miller

Mazurowski

Bashir

. Comparison of visualization rates of LI-RADS version 2014 major features with IV gadobenate dimeglumine or gadoxetate disodium in patients at risk for hepatocellular carcinoma. AJR Am J Roentgenol. 2018;210(6):1266-1272. doi:10.2214/ajr.17.18981

47.

Cerny

Bergeron

Billiard

, et al. LI-RADS for MR imaging diagnosis of hepatocellular carcinoma: performance of major and ancillary features. Radiology. 2018;288(1):118-128. doi:10.1148/radiol.2018171678

48.

Chen

Zhou

Kuang

, et al. Liver imaging reporting and data system category 5: MRI predictors of microvascular invasion and recurrence after hepatectomy for hepatocellular carcinoma. AJR Am J Roentgenol. 2019;213(4):821-830. doi:10.2214/ajr.19.21168

49.

Choi

Kim

Byun

, et al. Subtraction arterial images of hepatocyte-specific contrast-enhanced MRI: added value for the diagnosis of hepatocellular carcinoma in the LI-RADS v2018. Acta Radiol. 2019;54(2):127-136.

50.

Kim

Choi

Kim

, et al. Combined hepatocellular-cholangiocarcinoma: magnetic resonance imaging features and prognosis according to risk factors for hepatocellular carcinoma. J Magn Reson Imaging. 2021;53(6):1803-1812. doi:10.1002/jmri.27528

51.

Choi

Kang

Kim

. Comparison of gadoxetate disodium-enhanced MRI sequences for measuring hepatic observation size and its implication of LI-RADS classification. Abdom Radiol (NY). 2022;47(3):1024-1031. doi:10.1007/s00261-021-03403-2

52.

Clarke

CGD

Albazaz

Smith

, et al. Comparison of LI-RADS with other non-invasive liver MRI criteria and radiological opinion for diagnosing hepatocellular carcinoma in cirrhotic livers using gadoxetic acid with histopathological explant correlation. Clin Radiol. 2021;76(5):333-341. doi:10.1016/j.crad.2020.12.007

53.

Fraum

Tsai

Rohe

, et al. Differentiation of hepatocellular carcinoma from other hepatic malignancies in patients at risk: diagnostic performance of the liver imaging reporting and data system version 2014. Radiology. 2018;286(1):158-172. doi:10.1148/radiol.2017170114

54.

Fraum

Cannella

Ludwig

, et al. Assessment of primary liver carcinomas other than hepatocellular carcinoma (HCC) with LI-RADS v2018: comparison of the LI-RADS target population to patients without LI-RADS-defined HCC risk factors. Eur Radiol. 2020;30(2):996-1007. doi:10.1007/s00330-019-06448-6

55.

Lim

Kwon

Cho

Kim

Kang

Kim

. Added value of enhanced CT on LR-3 and LR-4 observation of Gd-EOB-DTPA MRI for the diagnosis of HCC: are CT and MR washout features interchangeable? Br J Radiol. 2022;95(1132):20210738. doi:10.1259/bjr.20210738

56.

Jiang

Liu

Chen

, et al. Man or machine? Prospective comparison of the version 2018 EASL, LI-RADS criteria and a radiomics model to diagnose hepatocellular carcinoma. Cancer Imaging. 2019;19(1):84. doi:10.1186/s40644-019-0266-9

57.

Joo

Lee

Ahn

Lee

Han

. Liver imaging reporting and data system v2014 categorization of hepatocellular carcinoma on gadoxetic acid-enhanced MRI: comparison with multiphasic multidetector computed tomography. J Magn Reson Imaging. 2017;45(3):731-740. doi:10.1002/jmri.25406

58.

Kang

Kim

Joo

Han

. Additional value of contrast-enhanced ultrasound (CEUS) on arterial phase non-hyperenhancement observations (≥ 2 cm) of CT/MRI for high-risk patients: focusing on the CT/MRI LI-RADS categories LR-3 and LR-4. Abdom Radiol (NY). 2020;45(1):55-63. doi:10.1007/s00261-019-02132-x

59.

Kierans

Makkar

Guniganti

, et al. Validation of Liver Imaging Reporting and Data System 2017 (LI-RADS) criteria for imaging diagnosis of hepatocellular carcinoma. J Magn Reson Imaging. 2019;49(7):e205-e215. doi:10.1002/jmri.26329

60.

Kim

Roh

. Hepatocellular carcinoma versus other hepatic malignancy in cirrhosis: performance of LI-RADS version 2018. Radiology. 2019;291(1):72-80. doi:10.1148/radiol.2019181995

61.

Ronot

Fouque

Esvan

Lebigot

Aubé

Vilgrain

. Comparison of the accuracy of AASLD and LI-RADS criteria for the non-invasive diagnosis of HCC smaller than 3 cm. J Hepatol. 2018;68(4):715-723. doi:10.1016/j.jhep.2017.12.014

62.

Rosiak

Podgorska

Rosiak

Cieszanowski

. Comparison of LI-RADS v.2017 and ESGAR guidelines imaging criteria in HCC diagnosis using MRI with hepatobiliary contrast agents. Biomed Res Int. 2018;2018:7465126. doi:10.1155/2018/7465126

63.

Song

Choi

Hwang

Choi

. LI-RADS v2014 categorization of hepatocellular carcinoma: intraindividual comparison between gadopentetate dimeglumine-enhanced MRI and gadoxetic acid-enhanced MRI. Eur Radiol. 2019;29(1):401-410. doi:10.1007/s00330-018-5559-z

64.

Stocker

Becker

Barth

, et al. Does quantitative assessment of arterial phase hyperenhancement and washout improve LI-RADS v2018-based classification of liver lesions? Eur Radiol. 2020;30(5):2922-2933. doi:10.1007/s00330-019-06596-9

65.

Zhang

Kuang

Chen

, et al. The role of preoperative dynamic contrast-enhanced 3.0-T MR imaging in predicting early recurrence in patients with early-stage hepatocellular carcinomas after curative resection. Front Oncol. 2019;9:1336. doi:10.3389/fonc.2019.01336

66.

Costa

McInnes

MDF

van der Pol

, et al. Individual participant data meta-analyses for diagnostic accuracy research: challenges and lessons learned from the LI-RADS IPD group. Radiol Imaging Cancer. 2024;6(3):e240015. doi:10.1148/rycan.240015

67.

Hosseini Shabanan

Martins

Wolfson

, et al. MASLD: what we have learned and where we need to go—a call to action. Radiographics. 2024;44(11):e240048. doi:10.1148/rg.240048

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.12 MB