Tyler J Fraum1, Roberto Cannella2,3, Daniel R Ludwig4, Richard Tsai4, Muhammad Naeem4, Maverick LeBlanc4, Amber Salter5, Allan Tsung6, Anup S Shetty4, Amir A Borhani2, Alessandro Furlan2, Kathryn J Fowler7. 1. Mallinckrodt Institute of Radiology, Washington University School of Medicine, 510 S. Kingshighway Blvd, Campus Box 8131, Saint Louis, MO, 63110, USA. fraumt@wustl.edu. 2. Department of Radiology, University of Pittsburgh Medical Center, 3708 Fifth Ave, Pittsburgh, PA, 15213, USA. 3. Department of Radiology, University of Palermo, University Hospital "Paolo Giaccone", Via del Vespro 129, 90127, Palermo, Italy. 4. Mallinckrodt Institute of Radiology, Washington University School of Medicine, 510 S. Kingshighway Blvd, Campus Box 8131, Saint Louis, MO, 63110, USA. 5. Division of Biostatistics, Washington University School of Medicine, 660 S. Euclid Ave., CB 8067, St. Louis, MO, 63110, USA. 6. Department of Surgery, Ohio State University Medical Center, N924 Doan Hall, 410 W 10th Ave, Columbus, OH, 43210, USA. 7. Department of Radiology, University of California San Diego, 200 W Arbor Dr., San Diego, CA, 92103, USA.
Abstract
OBJECTIVES: To determine whether the LI-RADS imaging features of primary liver carcinomas (PLCs) other than hepatocellular carcinoma (non-HCC PLCs) differ between patients considered high risk (RF+) versus not high risk (RF-) for HCC and to compare rates of miscategorization as probable or definite HCC between the RF+ and RF- populations. METHODS: This retrospective study included all pathology-proven non-HCC PLCs imaged with liver-protocol CT or MRI from 2007 to 2017 at two liver transplant centers. Patients were defined per LI-RADS v2018 criteria as RF+ or RF-. Two independent, blinded readers (R1, R2) categorized 265 lesions using LI-RADS v2018. Logistic regression was utilized to assess for differences in imaging feature frequencies between RF+ and RF- patients. Fisher's exact test was used to assess for differences in miscategorization rates. RESULTS: Non-HCC PLCs were significantly more likely to exhibit nonrim arterial phase hyperenhancement (R1: OR = 2.94; R2: OR = 7.09) and nonperipheral "washout" (R1: OR = 3.65; R2: OR = 7.69) but significantly less likely to exhibit peripheral "washout" (R1: OR = 0.30; R2: OR = 0.10) and delayed central enhancement (R1: OR = 0.18; R2: OR = 0.25) in RF+ patients relative to RF- patients. Consequently, non-HCC PLCs were more often miscategorized as probable or definite HCC in RF+ versus RF- patients (R1: 23.3% vs. 3.6%, p < 0.001; R2: 11.0% vs. 2.6%, p = 0.009). CONCLUSIONS: Non-HCC PLCs are more likely to mimic HCCs on CT and MRI in the LI-RADS target population than in patients without LI-RADS-defined HCC risk factors. KEY POINTS: • The presence of LI-RADS-defined risk factors for HCC tends to alter the imaging appearances of non-HCC PLCs, resulting in higher frequencies of major features and lower frequencies of LR-M features. • Non-HCC PLCs are more likely to be miscategorized as probable or definite HCC in the LI-RADS target population than in patients without LI-RADS-defined HCC risk factors.
OBJECTIVES: To determine whether the LI-RADS imaging features of primary liver carcinomas (PLCs) other than hepatocellular carcinoma (non-HCCPLCs) differ between patients considered high risk (RF+) versus not high risk (RF-) for HCC and to compare rates of miscategorization as probable or definite HCC between the RF+ and RF- populations. METHODS: This retrospective study included all pathology-proven non-HCCPLCs imaged with liver-protocol CT or MRI from 2007 to 2017 at two liver transplant centers. Patients were defined per LI-RADS v2018 criteria as RF+ or RF-. Two independent, blinded readers (R1, R2) categorized 265 lesions using LI-RADS v2018. Logistic regression was utilized to assess for differences in imaging feature frequencies between RF+ and RF- patients. Fisher's exact test was used to assess for differences in miscategorization rates. RESULTS: Non-HCCPLCs were significantly more likely to exhibit nonrim arterial phase hyperenhancement (R1: OR = 2.94; R2: OR = 7.09) and nonperipheral "washout" (R1: OR = 3.65; R2: OR = 7.69) but significantly less likely to exhibit peripheral "washout" (R1: OR = 0.30; R2: OR = 0.10) and delayed central enhancement (R1: OR = 0.18; R2: OR = 0.25) in RF+ patients relative to RF- patients. Consequently, non-HCCPLCs were more often miscategorized as probable or definite HCC in RF+ versus RF- patients (R1: 23.3% vs. 3.6%, p < 0.001; R2: 11.0% vs. 2.6%, p = 0.009). CONCLUSIONS: Non-HCCPLCs are more likely to mimic HCCs on CT and MRI in the LI-RADS target population than in patients without LI-RADS-defined HCC risk factors. KEY POINTS: • The presence of LI-RADS-defined risk factors for HCC tends to alter the imaging appearances of non-HCCPLCs, resulting in higher frequencies of major features and lower frequencies of LR-M features. • Non-HCCPLCs are more likely to be miscategorized as probable or definite HCC in the LI-RADS target population than in patients without LI-RADS-defined HCC risk factors.
Authors: Sang Hyun Choi; Seung Soo Lee; So Hyun Park; Kang Mo Kim; Eunsil Yu; Yangsoon Park; Yong Moon Shin; Moon-Gyu Lee Journal: Radiology Date: 2018-11-13 Impact factor: 11.105
Authors: Kathryn J Fowler; Arman Sheybani; Rex A Parker; Sean Doherty; Elizabeth M Brunt; William C Chapman; Christine O Menias Journal: AJR Am J Roentgenol Date: 2013-08 Impact factor: 3.959
Authors: Anne R J Péporté; Wieland H Sommer; Konstantin Nikolaou; Maximilian F Reiser; Christoph J Zech Journal: Eur J Radiol Date: 2012-11-16 Impact factor: 3.528
Authors: H T Sorensen; S Friis; J H Olsen; A M Thulstrup; L Mellemkjaer; M Linet; D Trichopoulos; H Vilstrup; J Olsen Journal: Hepatology Date: 1998-10 Impact factor: 17.425