Gadoxetic Acid enhanced MRI of Hepatocellular Carcinoma: Value of Washout in Transitional and Hepatobiliary Phases

Transcription

1 ORIGINAL RESEARCH GASTROINTESTINAL IMAGING Gadoxetic Acid enhanced MRI of Hepatocellular Carcinoma: Value of Washout in Transitional and Hepatobiliary Phases Dong Hwan Kim, MD* Sang Hyun Choi, MD, PhD* So Yeon Kim, MD, PhD Min-Ju Kim, MSc Seung Soo Lee, MD, PhD Jae Ho Byun, MD, PhD From the Department of Radiology and Research Institute of Radiology (D.H.K., S.H.C., S.Y.K., S.S.L., J.H.B.) and Department of Clinical Epidemiology and Biostatistics (M.J.K.), University of Ulsan College of Medicine, Asan Medical Center, 88 Olympic-Ro 43-Gil, Songpa-Gu, Seoul 05505, Republic of Korea. Received November 10, 2018; revision requested January 9, 2019; final revision received February 8; accepted February 26. Address correspondence to S.Y.K. ( * D.H.K. and S.H.C. contributed equally to this work. Conflicts of interest are listed at the end of this article. See also the editorial by Fowler in this issue. Radiology 2019; 00:1 7 Content codes: Background: Current Liver Imaging Reporting and Data System guidelines define the washout appearance of gadoxetic acid enhanced MRI only during the portal venous phase. Defining washout only during the portal venous phase may lead to lower sensitivity for diagnosis of hepatocellular carcinoma (HCC). Purpose: To compare the diagnostic performances of three gadoxetic acid enhanced MRI criteria fono. that part was correcr HCC according to the phases during which washout appearance was determined. Materials and Methods: In this retrospective study, patients with a hepatic nodule detected at US surveillance for HCC from January to December 2012 underwent gadoxetic acid enhanced MRI. Three diagnostic MRI criteria for HCC were defined according to the phases during which washout appearance was observed, with the presence of arterial phase hyperenhancement and hypointensity noted (a) only during the portal venous phase, with washout confined to the portal venous phase; (b) during the portal venous phase or transitional phase, with washout extended to the transitional phase; or (c) during the portal venous, transitional, or hepatobiliary phase, with washout extended to the hepatobiliary phase. If a nodule showed marked T2 hyperintensity or a targetoid appearance, it was precluded from the diagnosis of HCC. The sensitivity and specificity were compared by using a generalized estimating equation. Results: A total of 178 patients were included (mean age 6 standard deviation, 55.3 years 6 9.1) with 203 surgically confirmed hepatic nodules (186 HCCs and 17 non-hccs) measuring 3.0 cm or smaller. The sensitivity with washout extended to the hepatobiliary phase (95.2% [177 of 186]) was better than that with washout extended to the transitional phase (90.9% [169 of 186]; P =.01) and washout confined to the portal venous phase (75.3% [140 of 186]; P,.01). The specificity with extensions of washout to the transitional phase and hepatobiliary phase (82% [14 of 17] for both) was similar to that obtained with washout confined to the portal venous phase (94.1% [16 of 17]) (P =.47). Conclusion: After exclusion of typical hemangiomas and nodules with a targetoid appearance, extending washout appearance to the transitional or hepatobiliary phase (instead of restricting it to the portal venous phase) allowed higher sensitivity without a reduction in specificity. RSNA, 2019 adoxetic acid enhanced MRI substantially improves Gthe sensitivity of diagnosis of hepatocellular carcinoma (HCC) by allowing the detection of signal intensity changes on hepatobiliary phase images (1 5). The latest versions of the major clinical guidelines proposed by the American Association for the Study of Liver Disease, the Liver Imaging Reporting and Data System (LI-RADS), and the European Association for the Study of the Liver incorporate the use of gadoxetic acid into their diagnostic algorithms (6 8). In these guidelines, the washout appearance of gadoxetic acid enhanced MRI can be depicted only during the portal venous phase because of concern over the loss of specificity (9 12). Hypointensity in the transitional or hepatobiliary phase is merely considered an ancillary finding favoring malignancy (7,8). The definition of washout appearance at gadoxetic acid enhanced MRI is a hotly debated topic. A study published in 2015 (12) first raised concerns over mitigated specificity due to the inclusion of hypointensity at the transitional or hepatobiliary phase within the definition of washout appearance; the researchers reported a strikingly low specificity of 48.4% (12). However, indepth exploration of the data from this study revealed that most of the false-positive diagnoses were hemangiomas or cholangiocarcinomas. In actual clinical practice, these diseases could easily be excluded with the help of ancillary imaging findings, such as marked T2 hyperintensity or a targetoid appearance (8). In addition, another study reported a higher specificity, of 74.3%, with the same diagnostic criteria (5). Moreover, the incidence of non-hcc malignancy in patients to whom the diagnostic criteria are applied is as low as 2.4% 3.7% (13 15). In this regard, the fear of lower specificity by including washout in a later phase could be overblown. This copy is for personal use only. To order printed copies, contact reprints@rsna.org

2 Gadoxetic Acid enhanced MRI of Hepatocellular Carcinoma Abbreviations HCC = hepatocellular carcinoma, LI-RADS = Liver Imaging Reporting and Data System Summary When hepatocellular carcinoma is being diagnosed with use of gadoxetic acid enhanced MRI, extending washout appearance to the transitional phase or hepatobiliary phase (instead of restricting it to the portal venous phase) allows higher sensitivity without a reduction in specificity. Key Points nn nn Extending washout appearance to the transitional phase or hepatobiliary phase (sensitivity, 90.9% and 95.2%, respectively) instead of restricting it to the portal venous phase (sensitivity, 75.3%) allowed higher sensitivity for detection of hepatocellular carcinoma (P.01). After exclusion of typical hemangiomas and nodules with a targetoid appearance, extension of washout to the transitional and hepatobiliary phases did not significantly decrease specificity (82% for both phases) compared with washout confined to the portal venous phase (specificity, 94.1%) (P =.47). In addition, the strict adherence to defining washout only during the portal venous phase severely degrades the sensitivity of HCC diagnosis, lowering it to 66.3% 77.9% (12,16,17). The confinement of washout to the portal venous phase is contrary to the main reason for the recent popular use of gadoxetic acid: that is, the enhanced sensitivity achieved by adding the hepatobiliary phase. We hypothesized that extending washout appearance to the transitional or hepatobiliary phase would result in higher sensitivity without substantially reducing specificity for HCC at gadoxetic acid enhanced MRI. This study therefore aimed to compare the diagnostic performances of three different gadoxetic acid enhanced MRI criteria for HCC according to the phases during which washout appearance was determined. Materials and Methods Our institutional review board approved the study and waived the requirement for written informed consent because the study retrospectively used data from a clinical cohort of an HCC surveillance program. Study Population According to the protocol, patients at risk for HCC underwent surveillance US. When a hepatic nodule measuring at least 1 cm in diameter was detected at US, the patient was referred for further evaluation, which included dynamic CT or MRI. From January 2012 to December 2012, 888 consecutive patients who underwent gadoxetic acid enhanced MRI for the evaluation of suspicious nodules detected during US surveillance were identified as eligible participants. The current study population has been reported in two prior studies that validated the LI-RADS category 4 and 5 criteria and investigated outcomes of nodules that met the criteria for LI-RADS categories 4 and 5 according to treatment modalities with gadoxetic acid enhanced MRI (13,18). From this population of 888 patients, a further selection was made according to the following inclusion criteria: (a) focal hepatic solid nodules observed at MRI, (b) nodule size of 1 3 cm, and (c) a maximum of three nodules observed at MRI. Definite cyst, vascular anomaly, perfusion alteration, hepatic fat deposition or sparing, hypertrophic pseudomass, confluent fibrosis, and focal scar were not considered solid nodules (8). This selection yielded 424 patients with 660 nodules. Among them, 43 patients with 56 nodules were excluded because they were considered to have hepatic hemangiomas (the nodules showed marked T2 hyperintensity comparable to cerebrospinal fluid and the typical enhancement pattern) (8). Then, 294 patients with 401 nodules who did not undergo surgery were further excluded. Finally, 203 nodules in 178 patients were included in our study (Fig 1). This process was made by a study coordinator (D.H.K., with 7 years of experience in hepatic imaging) who was not involved in the image analysis for this study. MRI Examination All patients underwent MRI with a 1.5-T unit (Magnetom Avanto; Siemens Medical Solutions, Erlangen, Germany) with a six-element phased-array torso coil as the receiver. Fat-suppressed T1-weighted fast low-angle shot MRI was performed with a three-dimensional volumetric interpolated breath-hold examination technique (repetition time msec/echo time msec, 4/1.5; flip angle, 10 ; field of view, mm; matrix, ; number of signals acquired, one; parallel imaging factor, 2) before and after administration of gadoxetic acid (Primovist; Bayer Schering Pharma, Berlin, Germany). Contrast material was administered as a bolus at a dose of mmol per kilogram body weight at a rate of 1 ml/sec via a 22-gauge intravenous cubital line and was followed by a 10-mL saline flush at the same flow rate, delivered by using a power injector. Images in the arterial phase were obtained 5 seconds after the peak time determined by means of a test bolus technique, and portal venous phase, transitional phase, and hepatobiliary phase images were obtained 50 seconds, 3 minutes, and 20 minutes, respectively, after contrast material injection. Images were acquired in the transverse plane with a section thickness of 4 mm and no gap. Other MRI sequences included an axial dual-echo T1- weighted breath-hold gradient-echo sequence for acquisition of in-phase and out-of-phase images, an axial respiratory-triggered turbo spin-echo T2-weighted sequence with fat saturation, an axial half-fourier acquisition single-shot turbo spin-echo T2- weighted sequence with fat saturation, and diffusion-weighted imaging with a respiratory-triggered single-shot echo-planar imaging sequence with b values of 0, 50, 500, and 900 sec/mm 2. Image Analysis The images obtained with gadoxetic acid enhanced dynamic MRI were reviewed by two board-certified abdominal radiologists (S.H.C. and J.H.B., with 10 and 22 years of experience in hepatic imaging, respectively) in consensus. The radiologists were blinded to any information on clinical history, US findings, or final diagnosis. If any interpretations demonstrated discrepancies between the reviewers, they re-evaluated the examinations together and reached a consensus. The number of hepatic nodules visible and the longest diameter at gadoxetic 2 radiology.rsna.org n Radiology: Volume 00: Number

3 Figure 1: Flow diagram of patient selection. HCC = hepatocellular carcinoma. acid enhanced dynamic MRI were recorded. For each hepatic nodule, the two reviewers analyzed the presence or absence of arterial phase hyperenhancement and hypointensity on images obtained during the portal venous, transitional, and hepatobiliary phases. Arterial phase hyperenhancement was defined as non-rimlike enhancement unequivocally greater in whole or in part than the surrounding liver tissue during the arterial phase (8). When a nodule showed high signal intensity on T1-weighted images before administration of gadoxetic acid, the arterial phase hyperenhancement was evaluated by using subtraction of arterial phase images. Hypointensity during the portal venous, transitional, and hepatobiliary phases was defined as a visually assessed temporal reduction in signal intensity relative to the surrounding liver tissue (8). If the nodule showed a targetoid appearance according to LI-RADS version 2018 (8), it was considered a sign of non-hcc malignancy. Three different diagnostic MRI criteria for HCC were defined, as follows: (a) arterial phase hyperenhancement and hypointensity only during the portal venous phase, with the definition of washout being confined to the portal venous phase; (b) arterial phase hyperenhancement and hypointensity during the Kim et al portal venous or transitional phase, with the definition of washout being extended to the transitional phase; and (c) arterial phase hyperenhancement and hypointensity during the portal venous, transitional, or hepatobiliary phase, with the definition of washout being extended to the hepatobiliary phase. Histopathologic Diagnosis Histopathologic diagnoses after surgical resection were used as the standard of reference. Histopathologic diagnoses in our institution were made mainly by five pathologists who have more than 5 years of experience in liver pathology. Statistical Analysis Per-lesion estimates of diagnostic performance (sensitivity and specificity) were computed. Estimates and 95% confidence intervals were adjusted by using generalized estimating equations to account for data clustering effects. The sensitivity and specificity of each criterion were compared by using generalized estimating equations. For multiple comparison, Bonferroni-corrected P values are presented. P <.05 was considered to indicate a statistically significant difference. All statistical analyses were performed by using software (SAS, version 9.4; SAS Institute, Cary, NC). Results Patient Characteristics and Pathologic Findings There were 150 men (84%) and 28 women (16%) in the cohort. The mean patient age (6standard deviation) was 55.3 years (range, years). Hepatitis B was the predominant cause of chronic liver disease (155 of 178 patients [87.1%]), followed by alcoholic liver disease (10 of 178 [5.6%]). One hundred sixty patients had one nodule and 18 patients had two or three nodules. The clinical characteristics of the 178 patients are summarized in Table 1. A total of 203 nodules were found in 178 patients. The 203 nodules (mean size, 19.8 mm; range, mm) consisted of 186 HCCs, six dysplastic nodules, and 11 non-hcc malignancies (eight intrahepatic cholangiocarcinomas, one combined hepatocellular-cholangiocarcinoma, one metastasis from colon cancer, and one metastasis from breast cancer). Diagnostic Performance of the Three MRI Criteria for HCC During the arterial phase, 179 of the 186 HCC nodules (96.2%) showed typical arterial phase hyperenhancement. Of the 11 nodules with non-hcc malignancy, eight (73%) showed a rimlike enhancement during the arterial phase that allowed them to be easily distinguished from HCC, whereas only three (27%) showed arterial phase hyperenhancement mimicking the enhancement features of HCC. Radiology: Volume 00: Number n radiology.rsna.org 3

4 Gadoxetic Acid enhanced MRI of Hepatocellular Carcinoma Table 1: Clinical Characteristics Variable Value (n = 178) Age (y)* (33 77) Sex Men 150 (84) Women 28 (16) No. of nodules (89.9) 2 11 (6.2) 3 7 (3.9) Risk factors Hepatitis B 155 (87.1) Hepatitis C 5 (2.8) Hepatitis B and C 2 (1.1) Alcoholic liver disease 10 (5.6) Other 6 (3.4) Child-Pugh classification A 173 (97.2) B 5 (2.8) C 0 Tumor marker: AFP level (ng/ml) 11.3 ( ) Note. Except where indicated, data are numbers of patients, with percentages in parentheses. AFP = a-fetoprotein. * Data are means 6 standard deviations, with range in parentheses. "Other" includes liver cirrhosis caused by nonalcoholic fatty liver disease (n = 2) and cryptogenic liver cirrhosis (n = 4). Data are medians, with ranges in parentheses. The diagnostic performances of the three different MRI washout criteria (according to the phases during which the washout appearance was observed) for HCC are summarized in Table 2. Extension of the washout period to the hepatobiliary phase provided higher sensitivity (95.2% [177 of 186]) than did extension of washout to just the transitional phase (90.9% [169 of 186]; P =.01) (Fig 2), whereas extending washout to the transitional phase showed higher sensitivity than washout confined to the portal venous phase (75.3% [140 of 186]; P,.01). Extended washout to the transitional and hepatobiliary phases showed lower specificity (82% [14 of 17]) than did washout confined to the portal venous phase (94% [16 of 17]), but the difference was not significant (P =.47) (Fig 3). There was one false-positive diagnosis with use of washout confined to the portal venous phase, an intrahepatic cholangiocarcinoma. When the washout appearance was extended to the transitional phase, two additional false-positive diagnoses occurred (one intrahepatic cholangiocarcinoma and one dysplastic nodule) (Fig 3); however, 29 additional HCCs were diagnosed. When the washout appearance was extended to the hepatobiliary phase, no further false-positive diagnoses were found compared with extended washout to the transitional phase, whereas eight additional HCCs were diagnosed. In comparison with washout confined to the portal venous phase, the extended washout to the hepatobiliary phase depicted 37 additional HCCs at a cost of only two false-positive diagnoses. There were still nine falsenegative diagnoses when the washout was extended to the hepatobiliary phase; all of them were hypovascular HCCs that did not show arterial phase hyperenhancement. Discussion In our study, we retrospectively compared the diagnostic performance of three different gadoxetic acid enhanced MRI criteria for the diagnosis of HCC according to the phases during which the washout appearance was determined. The diagnostic criterion with extended washout to the hepatobiliary phase resulted in higher sensitivity (95.2% [177 of 186]) than did washout confined to the portal venous phase (75.3% [140 of 186]; P,.01) and washout extended to the transitional phase (90.9% [169 of 186]; P =.01). The specificities of the washout extended to the transitional and hepatobiliary phase (82% [14 of 17] for both) were lower than that of the washout confined to the portal venous phase (94% [16 of 17]), but the differences were not significant (P =.47). Twenty-nine additional HCCs were depicted when the washout was extended to the transitional phase, and 37 additional HCCs were depicted when washout was extended to the hepatobiliary phase, with a cost of only two false-positive diagnoses in comparison with washout confined to the portal venous phase. Therefore, at gadoxetic acid enhanced MRI, extension of the determination of a washout appearance to the transitional or hepatobiliary phase, as opposed to just the portal venous phase, may not degrade the specificity of HCC diagnosis while improving sensitivity. Several recently published meta-analyses found overall sensitivities for gadoxetic acid enhanced MRI diagnosis of HCCs ranging from 87% to 96% (19 22); those findings are consistent with the highest sensitivity of 95.2% (177 of 186) achieved with extending washout to the hepatobiliary phase in our study (19 22). However, when the conventional definition of washout confined only to the portal venous phase was applied, our sensitivity results showed a significant reduction to 75.3% (140 of 186). This result is similar to findings of previous studies that used the same diagnostic criterion (12,16,17). Even in the case of extracellular contrast-enhanced MRI, the washout appearance is better seen during the delayed phase than during the portal venous phase, with 19% 64% of HCCs exhibiting washout only on delayed-phase images (23,24). It therefore does not seem appropriate to confine the timing of washout to only the portal venous phase when gadoxetic acid is used. Joo et al (12) reported a specificity of 48.4% for extended washout to the hepatobiliary phase, a finding that differs substantially from our result. This difference can largely be attributed to their inclusion of hemangiomas and intrahepatic cholangiocarcinoma. Their study did not consider the ancillary imaging features of these diseases, such as marked T2 hyperintensity or a targetoid appearance (8). We believe that our results better reflect actual clinical practice because we comprehensively considered ancillary imaging findings. A recent study (25) using an approach similar to that used in our study demonstrated results consistent with ours. When the performances of different diagnostic criteria are compared, percentages such as sensitivity and specificity are usually used as measurements. However, from a practical 4 radiology.rsna.org n Radiology: Volume 00: Number

5 Kim et al Table 2: Diagnostic Performances of the Three MRI Criteria Diagnostic Criteria Washout confined to PVP Washout extended to TP Washout extended to HBP Sensitivity (%) Specificity (%) No. of False-Positive Value P Value* Value P Value* No. of HCCs Diagnoses 75.3 (140/186), (16/17) One (IHCC) [68.2, 81.2] [67.6, 99.2] 90.9 (169/186) (14/17) Three (two IHCCs [85.8, 94.2] [56.5, 94.4] and one dysplastic nodule) 95.2 (177/186) [91.2, 97.4], (14/17) [56.5, 94.4] Three (two IHCCs and one dysplastic nodule) Note. Numbers in parentheses are raw data, and numbers in brackets are 95% confidence intervals. HBP = hepatobiliary phase, HCC = hepatocellular carcinoma, IHCC = intrahepatic cholangiocarcinoma, PVP = portal venous phase, TP = transitional phase. * P values were obtained after adjustment for multiple comparison (Bonferroni correction). Pathologic diagnoses are given in parentheses. Comparison between washout confined to the portal venous phase and washout extended to the transitional phase. Comparison between washout extended to the transitional phase and washout extended to the hepatobiliary phase. Comparison between washout extended to the hepatobiliary phase and washout confined to the portal venous phase. Figure 2: Axial images obtained with gadoxetic acid enhanced MRI in 52-year-old man with chronic hepatitis B and surgically confirmed hepatocellular carcinoma (HCC). (a) Image obtained in arterial phase shows hyperenhancement of 2-cm nodule (arrow) in hepatic segment VII. (b d) Nodule (arrow) does not have washout appearance on images obtained in portal venous (b) and transitional (c) phases, whereas it shows hypointensity in hepatobiliary phase (d). (e) T2-weighted image (repetition time msec/echo time msec, 1900/88) shows nodule (arrow) with mild hyperintensity. If the criterion of washout confined to portal venous phase is applied, the nodule would not be diagnosed as HCC. Extending washout to hepatobiliary phase enabled diagnosis of HCC in this patient. point of view, the real numbers of patients who will benefit or be harmed by changes to the criteria are more meaningful than proportions. Even though the specificity (in percentages) seemed to show a substantial reduction when washout was extended to the hepatobiliary phase, only two non-hcc nodules were actually misdiagnosed as HCC. This can be explained by the fact that non-hcc malignancies are far less frequently encountered in patients at risk for HCC. If the extended washout criteria were applied to real clinical situations, the number of patients who would benefit from an additional diagnosis of HCC exceeds the number that would be harmed by a falsepositive diagnosis. Radiology: Volume 00: Number n radiology.rsna.org 5

6 Gadoxetic Acid enhanced MRI of Hepatocellular Carcinoma Figure 3: Axial images obtained with gadoxetic acid enhanced MRI in 44-year-old man with chronic hepatitis B and surgically confirmed cholangiocarcinoma. (a) Image obtained in arterial phase shows 2-cm hyperintense nodule (arrow) in hepatic segment VIII. (b) Image obtained in portal venous phase does not show washout (arrow), whereas images in (c) transitional and (d) hepatobiliary phases show washout appearance of the nodule (arrow). (e) T2-weighted image (repetition time msec/ echo time msec, 1900/88) shows nodule (arrow) with intermediate T2 hyperintensity. Extending washout to the transitional or hepatobiliary phase would have resulted in a false-positive diagnosis of hepatocellular carcinoma. Our study has some limitations. First, the inclusion of only hepatic nodules histopathologically confirmed through surgery might have introduced a selection bias. Particularly, most of our study population (173 of 178 [97.2%]) had Child-Pugh class A liver disease. Thus, our results might not be valid for patients with poorer liver function. However, this was unavoidable if ideal reference standards were to be established because biopsy-based tissue samples of combined hepatocellular-cholangiocarcinoma can lead to an incorrect diagnosis as a result of intratumoral heterogeneity (26). Second, our study is limited by its retrospective nature, although we tried our best to simulate the clinical situation and patients to whom the diagnostic criteria for HCC are mainly applied. Third, only a limited number of non-hcc nodules were included in our study, reflecting the fact that most nodules in patients at risk for HCC are HCC. However, the nonsignificant differences of specificities in our study may be attributed to the small number of non-hcc nodules. In conclusion, when hepatocellular carcinoma is diagnosed by using gadoxetic acid enhanced MRI, extending washout appearance to the transitional or hepatobiliary phase (instead of restricting it to the portal venous phase) allows higher sensitivity without a significant reduction in specificity. Given the growing use of gadoxetic acid enhanced MRI for improving the sensitivity of hepatocellular carcinoma detection, we believe it is time to reconsider the timing of washout appearance. A multicenter study with a prospective design is warranted to test the clinical effects of the modified diagnostic criteria as described by our study. Author contributions: Guarantor of integrity of entire study, S.Y.K.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; agrees to ensure any questions related to the work are appropriately resolved, all authors; literature research, D.H.K., S.H.C., S.Y.K.; clinical studies, D.H.K., S.H.C., S.Y.K., S.S.L., J.H.B.; statistical analysis, D.H.K., S.H.C., S.Y.K., M.J.K.; and manuscript editing, D.H.K., S.Y.K. Disclosures of Conflicts of Interest: D.H.K. disclosed no relevant relationships. S.H.C. disclosed no relevant relationships. S.Y.K. disclosed no relevant relationships. M.J.K. disclosed no relevant relationships. S.S.L. disclosed no relevant relationships. J.H.B. disclosed no relevant relationships. References 1. Hidaka M, Takatsuki M, Okudaira S, et al. The expression of transporter OATP2/ OATP8 decreases in undetectable hepatocellular carcinoma by Gd-EOB-MRI in the explanted cirrhotic liver. Hepatol Int 2013;7(2): Chou R, Cuevas C, Fu R, et al. Imaging techniques for the diagnosis of hepatocellular carcinoma: a systematic review and meta-analysis. Ann Intern Med 2015;162(10): Kierans AS, Kang SK, Rosenkrantz AB. The diagnostic performance of dynamic contrast-enhanced MR imaging for detection of small hepatocellular carcinoma measuring up to 2 cm: a meta-analysis. Radiology 2016;278(1): Hanna RF, Miloushev VZ, Tang A, et al. Comparative 13-year meta-analysis of the sensitivity and positive predictive value of ultrasound, CT, and MRI for detecting hepatocellular carcinoma. Abdom Radiol (NY) 2016;41(1): Choi SH, Byun JH, Lim YS, et al. Diagnostic criteria for hepatocellular carcinoma 3 cm with hepatocyte-specific contrast-enhanced magnetic resonance imaging. J Hepatol 2016;64(5): Marrero JA, Kulik LM, Sirlin CB, et al. Diagnosis, staging, and management of hepatocellular carcinoma: 2018 practice guidance by the American Association for the Study of Liver Diseases. Hepatology 2018;68(2): radiology.rsna.org n Radiology: Volume 00: Number

7 Kim et al 7. European Association for the Study of the Liver. EASL clinical practice guidelines: management of hepatocellular carcinoma. J Hepatol 2018;69(1): American College of Radiology. CT/MRI LI-RADS v2018 core. org/-/media/acr/files/rads/li-rads/li-rads-2018-core.pdf?la=en. Accessed November 6, Golfieri R, Grazioli L, Orlando E, et al. Which is the best MRI marker of malignancy for atypical cirrhotic nodules: hypointensity in hepatobiliary phase alone or combined with other features? Classification after Gd-EOB-DTPA administration. J Magn Reson Imaging 2012;36(3): Granito A, Galassi M, Piscaglia F, et al. Impact of gadoxetic acid (Gd-EOB- DTPA) enhanced magnetic resonance on the non-invasive diagnosis of small hepatocellular carcinoma: a prospective study. Aliment Pharmacol Ther 2013; 37(3): Kwon HJ, Byun JH, Kim JY, et al. Differentiation of small ( 2 cm) hepatocellular carcinomas from small benign nodules in cirrhotic liver on gadoxetic acid enhanced and diffusion-weighted magnetic resonance images. Abdom Imaging 2015; 40(1): Joo I, Lee JM, Lee DH, Jeon JH, Han JK, Choi BI. Noninvasive diagnosis of hepatocellular carcinoma on gadoxetic acid enhanced MRI: can hypointensity on the hepatobiliary phase be used as an alternative to washout? Eur Radiol 2015;25(10): Choi SH, Byun JH, Kim SY, et al. Liver Imaging Reporting and Data System v2014 with gadoxetate disodium enhanced magnetic resonance imaging: validation of LI-RADS category 4 and 5 criteria. Invest Radiol 2016;51(8): Darnell A, Forner A, Rimola J, et al. Liver Imaging Reporting and Data System with MR imaging: evaluation in nodules 20 mm or smaller detected in cirrhosis at screening US. Radiology 2015;275(3): Cerny M, Bergeron C, Billiard JS, et al. LI-RADS for MR imaging diagnosis of hepatocellular carcinoma: performance of major and ancillary features. Radiology 2018;288(1): Joo I, Lee JM, Lee DH, Jeon JH, Han JK. Retrospective validation of a new diagnostic criterion for hepatocellular carcinoma on gadoxetic acid enhanced MRI: can hypointensity on the hepatobiliary phase be used as an alternative to washout with the aid of ancillary features? Eur Radiol 2019;29(4): Min JH, Kim JM, Kim YK, et al. Prospective intraindividual comparison of magnetic resonance imaging with gadoxetic acid and extracellular contrast for diagnosis of hepatocellular carcinomas using the Liver Imaging Reporting and Data System. Hepatology 2018;68(6): Choi SH, Byun JH, Lim YS, et al. Liver Imaging Reporting and Data System: patient outcomes for category 4 and 5 nodules. Radiology 2018;287(2): Lee YJ, Lee JM, Lee JS, et al. Hepatocellular carcinoma: diagnostic performance of multidetector CT and MR imaging a systematic review and meta-analysis. Radiology 2015;275(1): Junqiang L, Yinzhong W, Li Z, et al. Gadoxetic acid disodium (Gd-EOBDTPA) enhanced magnetic resonance imaging for the detection of hepatocellular carcinoma: a meta-analysis. J Magn Reson Imaging 2014;39(5): Duncan JK, Ma N, Vreugdenburg TD, Cameron AL, Maddern G. Gadoxetic acidenhanced MRI for the characterization of hepatocellular carcinoma: a systematic review and meta-analysis. J Magn Reson Imaging 2017;45(1): Liu X, Jiang H, Chen J, Zhou Y, Huang Z, Song B. Gadoxetic acid disodium enhanced magnetic resonance imaging outperformed multidetector computed tomography in diagnosing small hepatocellular carcinoma: a meta-analysis. Liver Transpl 2017;23(12): Jang HJ, Kim TK, Khalili K, et al. Characterization of 1-to 2-cm liver nodules detected on HCC surveillance ultrasound according to the criteria of the American Association for the Study of Liver Disease: is quadriphasic CT necessary? AJR Am J Roentgenol 2013;201(2): Luca A, Caruso S, Milazzo M, et al. Multidetector-row computed tomography (MDCT) for the diagnosis of hepatocellular carcinoma in cirrhotic candidates for liver transplantation: prevalence of radiological vascular patterns and histological correlation with liver explants. Eur Radiol 2010;20(4): Chen N, Motosugi U, Morisaka H, et al. Added value of a gadoxetic acid enhanced hepatocyte-phase image to the LI-RADS system for diagnosing hepatocellular carcinoma. Magn Reson Med Sci 2016;15(1): Brunt E, Aishima S, Clavien PA, et al. chcc-cca: consensus terminology for primary liver carcinomas with both hepatocytic and cholangiocytic differentation. Hepatology 2018;68(1): Radiology: Volume 00: Number n radiology.rsna.org 7