Document Type

Article

Publication Date

12-14-2020

Comments

This article is the author’s final published version in World Journal of Gastroenterology, Volume 26, Issue 46, December 2020, Pages 7325-7337.

The published version is available at https://doi.org/10.3748/wjg.v26.i46.7325. Copyright © Yang et al.

Abstract

BACKGROUND: Combined hepatocellular-cholangiocarcinoma (CHC) is a rare type of primary liver cancer. Due to its complex histopathological characteristics, the imaging features of CHC can overlap with those of hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC).

AIM: To investigate the possibility and efficacy of differentiating CHC from HCC and ICC by using contrast-enhanced ultrasound (CEUS) Liver Imaging Reporting and Data System (LI-RADS) and tumor biomarkers.

METHODS: Between January 2016 and December 2019, patients with histologically confirmed CHC, ICC and HCC with chronic liver disease were enrolled. The diagnostic formula for CHC was as follows: (1) LR-5 or LR-M with elevated alpha-fetoprotein (AFP) and carbohydrate antigen 19-9 (CA19-9); (2) LR-M with elevated AFP and normal CA19-9; or (3) LR-5 with elevated CA19-9 and normal AFP. The sensitivity, specificity, accuracy and area under the receiver operating characteristic curve were calculated to determine the diagnostic value of the criteria.

RESULTS: After propensity score matching, 134 patients (mean age of 51.4 ± 9.4 years, 108 men) were enrolled, including 35 CHC, 29 ICC and 70 HCC patients. Based on CEUS LI-RADS classification, 74.3% (26/35) and 25.7% (9/35) of CHC lesions were assessed as LR-M and LR-5, respectively. The rates of elevated AFP and CA19-9 in CHC patients were 51.4% and 11.4%, respectively, and simultaneous elevations of AFP and CA19-9 were found in 8.6% (3/35) of CHC patients. The sensitivity, specificity, positive predictive value, negative predictive value, accuracy and area under the receiver operating characteristic curve of the aforementioned diagnostic criteria for discriminating CHC from HCC and ICC were 40.0%, 89.9%, 58.3%, 80.9%, 76.9% and 0.649, respectively. When considering the reported prevalence of CHC (0.4%-14.2%), the positive predictive value and NPV were revised to 1.6%-39.6% and 90.1%-99.7%, respectively.

CONCLUSION: CHCs are more likely to be classified as LR-M than LR-5 by CEUS LI-RADS. The combination of the CEUS LI-RADS classification with serum tumor markers shows high specificity but low sensitivity for the diagnosis of CHC. Moreover, CHC could be confidently excluded with high NPV.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License

PubMed ID

33362387

Language

English

Included in

Radiology Commons

Share

COinS