Comparing the computed tomography radiologic features of cytokeratin 19-positive hepatocellular carcinoma to those of conventional hepatocellular carcinoma and intrahepatic cholangiocarcinoma

Introduction

Hepatocellular carcinoma (HCC) is the most common primary malignancy of the hepatobiliary system and is the fourth leading cause of cancer-related death (1). The incidence rates of HCC are very high in Asia, and the Chinese mainland has the highest number of cases due to the large number of people living with chronic hepatitis B virus (HBV) infection (2). The prognosis of early stage HCC patients who undergo therapy is acceptable; however, the prognosis of most other HCC patients is poor due to its highly aggressive nature (3). Non-surgical therapy, such as combined treatment, including targeted therapy and immunotherapy, has provided new hope in the treatment of advanced HCC (4,5). However, due to the high heterogeneity of malignant tumors, substantial clinical challenges remain in the treatment of HCC, particularly in terms of individualized therapy.

Cytokeratin 19–positive hepatocellular carcinoma (CK19⁺ HCC) is a novel subtype of HCC that is highly aggressive and has a poor clinical prognosis (6). CK19⁺ HCC is defined as the pathological manifestation of the morphological features of classic HCC in tumor cells, along with the immunohistochemical expression of bile duct cell markers (6,7). Notably, CK19⁺ HCC is differentiated from combined HCC and cholangiocarcinoma, the latter of which is characterized by the presence of both typical HCC and intrahepatic cholangiocarcinoma (ICC) components within a single tumor nodule. These two components of combined HCC and cholangiocarcinoma may either be mixed or located at different zones inside the tumor (8). Previous studies have shown that CK19⁺ HCC is highly invasive, prone to postoperative recurrence, and has a low overall survival rate (9-11). The poor prognosis of CK19⁺ HCC is related to the expression of specific genes and the activation of the epidermal growth factor-epidermal growth factor receptor signaling pathway (12-14). Thus, the preoperative diagnosis of CK19⁺ HCC could assist in the formulation of individualized treatment strategies to improve patient prognosis.

To date, only a few imaging studies on CK19⁺ HCC have been published (15,16). Chung et al. previously focused on the poor prognosis of hypervascular HCC on contrast-enhanced computed tomography (CT), of which about one-third of cases were CK19⁺ HCC (17). Later magnetic resonance imaging (MRI) studies showed that irregular margins, rim enhancement, a low apparent diffusion coefficient value, hepatobiliary phase hypointensity, and the absence of non-peripheral “washout” were predictive indicators of CK19⁺ HCC (18,19). Moreover, it has been suggested that the reduced mean diffusion value derived from diffusion kurtosis imaging could serve as a potential imaging marker for CK19⁺ HCC (20). Recently, several studies have used the radiomics approach to predict CK19⁺ HCC and reported good performance (21-25). However, the number of CK19⁺ HCC patients in these studies was relatively small (26). Additionally, most studies have employed MRI data-driven artificial intelligence algorithms in predictive models for diagnosing CK19⁺ HCC. To date, only limited research has been conducted on the morphological characteristics of CK19⁺ HCC, especially its CT features, which remain uncertain.

This study analyzed the CT features of CK19⁺ HCC compared to those of conventional hepatocellular carcinoma (CHCC) and ICC, and established a diagnostic model based on the imaging characteristics of this distinct HCC subtype. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-24-914/rc).

Methods

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of the First Affiliated Hospital, Zhejiang University School of Medicine (No. K2023-IRB-092). The Ethics Committee waived the requirement of informed consent due to the retrospective nature of the study.

Clinical patients

The data of the CK19⁺ HCC patients were retrospectively collected from the First Affiliated Hospital, Zhejiang University School of Medicine from June 2011 to June 2016. CK19⁺ HCC patients were included in the study if they met the following inclusion criteria: (I) had histological and immunohistochemical confirmation of CK19⁺ HCC based on resected tumor specimens. (The pathological criterion for CK19⁺ HCC was a histopathological diagnosis of HCC; the immunohistochemical criterion was strongly positive cholangiocyte marker CK19 results in more than 15% of the tumor cells, which were simultaneously positive for the hepatocyte markers; patients expressing both cholangiocyte markers and hepatocyte markers in different regions were excluded from the study) (6); (II) had undergone unenhanced and contrast-enhanced CT scans of the liver prior to surgery; and (III) had well-preserved CT data, allowing for further evaluation. CK19⁺ HCC patients were excluded from the study if they met the following exclusion criteria: (I) had been treated with preoperative ablation, radiotherapy, or another anti-tumor therapy before surgery; (II) had poor-quality imaging data or pathology specimens that did not support further evaluation; and/or (III) had recurrent or metastatic HCC. Ultimately, 104 CK19⁺ HCC patients were enrolled in this study.

At the same time, we retrospectively collected a series of data from CHCC and ICC patients at our institution. The pathological criterion for CHCC was a histological diagnosis of HCC; the immunohistochemical criterion was negative cholangiocyte marker CK19 results in the tumor cells, which were simultaneously positive for the hepatocyte marker or markers (6). The CHCC and ICC patients were then matched with the CK19⁺ HCC patients with a maximum tumor diameter difference of less than 0.3 cm. In total, 104 CHCC patients and 104 ICC patients were selected for inclusion in the control groups. The screening flowchart for the study is shown in Figure 1.

Figure 1 Flow diagram of study population. CK19⁺ HCC, cytokeratin 19-positive hepatocellular carcinoma; CHCC, conventional hepatocellular carcinoma; CT, computed tomography; ICC, intrahepatic cholangiocarcinoma.

Image acquisition

Different CT scanners were used at different periods, including an Aquilion 16 multi-slice spiral CT scanner (Toshiba Medical System, Japan) and a 256-layer iCT scanner (Philips Medical System, Netherlands). The patients were placed in the supine position, and the scanning ranged from the liver’s upper edge in the right diaphragmatic dome to the liver’s lower edge, involving the whole liver. The scanning parameters were as follows: tube voltage: 120 kV; tube current: 320 mA; helical pitch: 0.95, matrix: 512 × 512; layer thickness: 2–5 mm; and reconstruction distance: 2 mm. The contrast agent, iohexol, was injected intravenously using a high-pressure syringe at a flow rate of 2.5–3.0 mL/s with a total dose of 1.5 mL/kg.

CT image analysis

In this study, two junior radiologists (HB Zhang and LT Chen) independently conducted the CT radiologic evaluations without knowledge of the final pathological findings. If the two radiologists could not reach an agreement in relation to certain features, a senior radiologist (W.L.) was asked to determine the final evaluation results. The Liver Imaging Reporting and Data System (LI-RADS, version 2018) was used for the radiologic evaluation of the enrolled HCC patients (27). Similarly, we adopted the same CT feature analysis protocol to evaluate cases of ICC.

The CT features analyzed in both groups included tumor location, single-lobe involvement (defined as tumors entirely confined to one hepatic lobe across all imaging sequences), tumor size (measured as the maximum diameter on the largest cross-sectional image), tumor shape [categorized as regular (round or ovoid) or irregular (lobulated or with irregular margins)], border appearance (well-defined or ill-defined), and the presence or absence of intratumoral fat, hemorrhage, necrosis, or calcification. The additional evaluated features comprised bile duct dilatation, hepatic capsular retraction, arterial phase enhancement patterns [homogeneous hyperenhancement, rim enhancement, or transient hepatic attenuation difference (THAD)], pseudocapsule formation, corona-like enhancement, washout characteristics (peripheral or non-peripheral), progressive enhancement, and persistent hyperenhancement (defined as non-rim arterial hyperenhancement with sustained enhancement relative to background liver parenchyma in portal venous and delayed phases). The assessment also recorded satellite lesions, lymphadenopathy, and the LI-RADS classification.

Statistical analysis

The statistical analysis was performed using SPSS 27.0 software. The categorical data are expressed as the number (percentage), and were compared using the chi-square test. The normally distributed measurement data are described as the mean ± standard deviation, and were compared using the t-test. The non-normally distributed measurement data are described as the P₅₀ (P₂₅, P₇₅), and were analyzed using the Mann-Whitney U test. A multivariate logistic regression analysis was conducted to establish a predictive model for CK19⁺ HCC identification. Receiver operating characteristic (ROC) curves were constructed to evaluate the diagnostic performance of various indicators for CK19⁺ HCC detection, and the area under the curve (AUC), sensitivity, and specificity were calculated. Comparisons of the AUC values between different indicators were performed using DeLong’s test. Statistical significance was defined as P<0.05. To test the consistency of the two radiologists’ evaluation of the tumor CT features, a Kappa analysis was completed before the data analysis. Table S1 summarizes the results of the Kappa analysis.

Results

The CK19⁺ HCC patients were aged from 22 to 71 years (mean age: 52.2 years), and 79 were male and 25 were female. Of these patients, 27 presented with symptoms, 88 with chronic HBV infection, 73 with cirrhosis, 13 with cholelithiasis, 16 with hypertension, 8 with type 2 diabetes, and 22 with other diseases. Unhealthy habits included smoking (42 patients) and alcohol abuse (33 patients). The preoperative laboratory tests of the patients showed that the serum alpha-fetoprotein (AFP) levels ranged from 0.8–87,500.0 ng/mL (normal reference: 0–20.0 ng/mL), the serum carbohydrate antigen 19-9 (CA19-9) levels ranged from 2.0–177.7 U/mL (normal reference: 0–37.0 U/mL), and the serum carcinoembryonic antigen (CEA) levels ranged from 0.5–38.7 ng/mL (normal reference: 0–5.0 ng/mL). The statistical results of the clinicopathological characteristics of the two groups, CK19⁺ HCC vs. CHCC and CK19⁺ HCC vs. ICC, are summarized in Table 1.

In the CK19⁺ HCC group, the tumors were located in the right lobe in 66 patients, and 90 patients showed single-lobe involvement. The tumor shape was regular in 58 patients, and the border was defined in 52 patients. On the unenhanced scans, there were 103 patients with hypodensity, 41 with homogeneous density, 2 with intratumoral fat, 4 with intratumoral hemorrhage, 64 with intratumoral necrosis, 3 with intratumoral calcification, 13 with bile duct dilatation, and 13 with hepatic capsular retraction. On the contrast-enhanced CT scans, there were 40 patients with homogeneous hyperenhancement in the arterial phase, 8 with rim enhancement in the arterial phase, 100 with heterogeneous enhancement, 5 with THAD, 13 with pseudocapsule formation, 9 with corona-like enhancement, 6 with peripheral washout, 92 with non-peripheral washout, 6 with progressive enhancement, 13 with satellite lesions, and 10 with lymphadenopathy. In the CK19⁺ HCC group, 94 patients were classified using the LI-RADS, of whom, 1 was classified as LR-4, 70 as LR-5, 17 as LR-TIV, and 6 as LR-M.

In the CHCC group, the tumors were located in the right lobe in 70 patients, and 82 patients showed single-lobe involvement. The tumor shape was regular in 61 patients, and the border was defined in 75 patients. On the unenhanced scans, there were 94 patients with hypodensity, 51 with homogeneous density, 6 with intratumoral fat, 16 with intratumoral hemorrhage, 21 with intratumoral necrosis, 7 with bile duct dilatation, and 3 with hepatic capsular retraction. On the contrast-enhanced CT scans, there were 57 patients with homogeneous hyperenhancement in the arterial phase, 6 with rim enhancement in the arterial phase, 90 with heterogeneous enhancement, 7 with THAD, 43 with pseudocapsule formation, 15 with corona-like enhancement, 3 with peripheral washout, 91 with non-peripheral washout, 6 with progressive enhancement, 4 with persistent hyperenhancement, 7 with satellite lesions, and 5 with lymphadenopathy. In the CHCC group, 92 patients were classified using the LI-RADS, of whom, 4 were classified as LR-4, 80 as LR-5, 3 as LR-TIV, and 5 as LR-M.

In the ICC group, the tumors were located in the right lobe in 51 patients, and 90 patients showed single-lobe involvement. The tumor shape was regular in 13 patients, and the border was defined in 22 patients. On the unenhanced scans, there were 97 patients with hypodensity, 40 with homogeneous density, 2 with intratumoral hemorrhage, 31 with intratumoral necrosis, 10 with intratumoral calcification, 60 with bile duct dilatation, and 26 with hepatic capsular retraction. The contrast-enhanced CT scans, there were 7 patients with homogeneous hyperenhancement in the arterial phase, 41 with rim enhancement in the arterial phase, 90 with heterogeneous enhancement, 41 with THAD, 6 with pseudocapsule formation, 7 with corona-like enhancement, 4 with non-peripheral washout, 99 with progressive enhancement, 1 with persistent hyperenhancement, 10 with satellite lesions, and 46 with lymphadenopathy. In the ICC group, 23 patients were classified using the LI-RADS, of whom 1 was classified as LR-3, 3 as LR-4, 4 as LR-5, 2 as LR-TIV, and 13 as LR-M.

In terms of the CT features, hypodensity (99.0% vs. 90.4%, P=0.005), intratumoral necrosis (61.5% vs. 20.2%, P<0.001), hepatic capsular retraction (12.5% vs. 2.9%, P=0.009), and heterogeneous enhancement (96.2% vs. 86.5%, P=0.014) were more common in CK19⁺ HCC than CHCC, while a defined border (50.0% vs. 72.1%, P=0.001), intratumoral hemorrhage (3.8% vs. 15.4%, P=0.005), pseudocapsule formation (12.5% vs. 41.3%, P<0.001), and homogeneous hyperenhancement in the arterial phase (38.5% vs. 54.8%, P=0.018) were less common. In addition, the proportion of patients with LR-TIV was higher in the CK19⁺ HCC group than the CHCC group (16.3% vs. 2.9%, P<0.001).

There were significant differences between the CK19⁺ HCC and ICC groups in terms of lesion location (right lobe: 63.5% vs. 49.0%, P=0.036), shape (regular: 55.8% vs. 12.5%, P<0.001), margins (defined: 50% vs. 21.2%, P<0.001), intratumoral necrosis (61.5% vs. 29.8%, P<0.001), intratumoral calcification (2.9% vs. 9.6%, P=0.045), bile duct dilatation (12.5% vs. 57.7%, P<0.001), hepatic capsular retraction (12.5% vs. 25.0%, P=0.021), homogeneous hyperenhancement in the arterial phase (38.5% vs. 6.7%, P<0.001), rim enhancement in the arterial phase (7.7% vs. 39.4%, P<0.001), heterogeneous enhancement (96.2% vs. 86.5%, P=0.014), THAD (4.8% vs. 39.4%, P<0.001), peripheral washout (5.8% vs. 0.0%, P=0.038), non-peripheral washout (88.5% vs. 3.8%, P<0.001), progressive enhancement (5.8% vs. 95.2%, P<0.001), and lymphadenopathy (9.6% vs. 44.2%, P<0.001). The proportion of LR-TIV was higher (16.3% vs. 1.9%, P<0.001) and the proportion of LR-5 was lower (67.3% vs. 76.9%, P<0.001) in the CK19⁺ HCC than the ICC group. The statistical results comparing the radiologic features of the CK19⁺ HCC group to those of the CHCC and ICC groups are summarized in Table 2.

The statistical analysis revealed significant differences between the CK19⁺ HCC group and the non-CK19⁺ hepatic tumor group (comprising the CHCC and ICC patients). No significant differences were observed between the two groups in terms of gender, symptoms, cholelithiasis, type 2 diabetes, other diseases, smoking, or alcohol abuse (P>0.05). However, statistically significant differences were observed between the groups in terms of age, chronic HBV infection, cirrhosis, hypertension, and CEA, CA19-9, and AFP levels (P<0.05) (Table 3).

In terms of the imaging features, compared to the non-CK19⁺ hepatic tumor group, the CK19⁺ HCC group exhibited higher proportions of regular morphology, hypodensity, intratumoral necrosis, heterogeneous enhancement, peripheral washout, non-peripheral washout, and LR-5 and LR-TIV in the LI-RADS classification, but exhibited lower proportions of adjacent bile duct dilation, rim arterial enhancement, THAD, pseudocapsule formation, progressive enhancement, and lymphadenopathy (P<0.05) (Table 4).

Using the HCC classification (0 = non-CK19⁺ hepatic tumor, 1 = CK19⁺ HCC) as the dependent variable, the statistically significant variables from Tables 4,5 were included as independent variables in a stepwise multivariate logistic regression analysis. The results showed that intratumoral necrosis, THAD, pseudocapsule formation, progressive enhancement, and LR-TIV in the LI-RADS classification were independent predictors for CK19⁺ HCC identification (P<0.05). The logistic regression equation for CK19⁺ HCC discrimination is shown in Table 5, and is expressed as follows:

P=1/[1+e−(−0.413+1.732X1−1.625X2−1.3X3−2.893X4+1.849X5)][1]

The ROC curve analysis and DeLong’s test revealed that the combined prediction model achieved the highest AUC [0.867, 95% confidence interval (CI): 0.827–0.907] for identifying CK19⁺ HCC, with a sensitivity of 88.46% and a specificity of 69.71%. The AUC of the combined model significantly surpassed those of individual predictors (intratumoral necrosis, THAD, pseudocapsule formation, progressive enhancement, and LR-TIV) (Z=7.624, 10.901, 12.278, 7.369, and 12.868; P<0.05) (Table 6 and Figure 2).

Figure 2 Predictive performance of imaging features for CK19+ HCC, showing AUC values of 0.683 for intratumoral necrosis (A), 0.591 for transient hepatic attenuation difference (B), 0.555 for pseudocapsule formation (C), 0.724 for progressive enhancement (D), 0.570 for LR-TIV (E), and 0.867 for the joint prediction model (F). AUC, area under the curve; CI, confidence interval; CK19⁺ HCC, cytokeratin 19-positive hepatocellular carcinoma; LI-RADS, Liver Imaging Reporting and Data System; LR-TIV, Liver Imaging Reporting and Data System – Category tumor in vein; ROC, receiver operating characteristic.

Discussion

This study showed that CK19⁺ HCC has specific CT features and clinical characteristics. While CK19⁺ HCC shares bile duct markers with ICC, it predominantly presents with intratumoral necrosis and LR-TIV, demonstrating lower frequencies of THAD, pseudocapsule formation, and progressive enhancement. The integrated prediction model incorporating these imaging features was able to effectively discriminate between CK19⁺ HCC and non-CK19 hepatic tumors (AUC: 0.867). This diagnostic model enables the preoperative detection of CK19⁺ HCC, which could aid in the individualized clinical management of this uncommon HCC subtype.

In the current study, intratumoral necrosis served as a valuable diagnostic feature, revealing the highly aggressive biological behavior of CK19⁺ HCC on CT. Previous studies have demonstrated that CK19+ HCC exhibits significantly more irregular morphology and infiltrative borders compared to CK19-negative HCC, with statistically significant differences (18,22). Similarly, another study reported lobular morphology in 23.5% (12/51) of CK19⁺ HCC cases (19). Conversely, Wang et al. found no significant difference in morphological irregularity between CK19⁺ HCC and CK19-negative HCC (21). We found that tumors in CHCC exhibited more regular morphology than those in CK19+ HCC. more prevalent in CHCC than CK19⁺ HCC. Two previous studies reported significant discrepancies in intratumoral necrosis in CK19⁺ HCC (5/38 vs. 13/25) (18,23), and that the necrosis rate was significantly lower in CK19⁺ HCC than CK19-negative HCC (18). Conversely, our study found that the prevalence of necrosis was markedly higher in CK19⁺ HCC (63.5%) than CHCC, which might be attributable to tumor ischemia from rapid CK19⁺ HCC cell proliferation. These interstudy discrepancies in imaging characteristics may reflect variations in sample sizes and imaging protocols across studies.

Our findings showed that pseudocapsule absence occurred more frequently in CK19⁺ HCC than CHCC on contrast-enhanced CT. Previous studies have reported inconsistent pseudocapsule incidence rates in CK19⁺ HCC based on MRI findings, while CT-based data remain scarce. Three independent CK19⁺ HCC studies reported pseudocapsule frequencies of 60.0% (18/30), 75.4% (43/57), and 60.0% (15/25), respectively, and found no significant differences between the CK19⁺ HCC and CK19-negative HCC groups (21-23). Another series of 51 CK19⁺ HCC cases demonstrated pseudocapsule visualization on MRI in 96.1% of cases (19). However, Choi et al. reported substantially lower pseudocapsule prevalence (26.3%) in CK19⁺ HCC versus CK19-negative HCC (10/38 vs. 186/204) (18). We found that the frequency of pseudocapsule formation was significantly lower in the CK19⁺ HCC group than the CHCC group. We attributed this lower pseudocapsule detection rate on CT to histopathological tumor infiltration patterns and microvascular invasion (15). Additionally, a greater proportion of CK19⁺ HCC patients exhibited LR-TIV, which is a diagnostic marker for differentiating between CK19⁺ HCC and non-CK19⁺ hepatic tumors.

CK19⁺ HCC may display atypical imaging characteristics, but the diagnostic utility of these features requires further validation. While rim enhancement has traditionally been regarded as a hallmark radiologic feature of ICC and metastatic lesions, emerging evidence supports its predictive value for CK19⁺ HCC on contrast-enhanced MRI (18,21-23). These studies reported rim enhancement in approximately one-third of the CK19⁺ HCC cases, which significantly exceeded the frequency observed in the CK19-negative HCC cohorts (18,21-23). Previous research has established an association between arterial phase rim enhancement in HCC and both poor histological differentiation and an unfavorable prognosis (28,29). Notably, our study observed rim enhancement in only 7.7% of the CK19⁺ HCC patients on contrast-enhanced CT. This discrepancy in the prevalence of rim enhancement across studies may reflect variations in imaging modalities. One study similarly identified progressive enhancement as diagnostically significant for CK19⁺ HCC, but neither our study nor other studies have replicated this observation (21-23).

The plasma AFP level is a valuable marker for the preoperative evaluation of CK19⁺ HCC. Previous studies have reported that up to 84.3% of CK19⁺ HCC patients exhibit elevated AFP, establishing it as a valuable independent predictor (19). In the same period (2019–2021), a number of different studies have reported that CK19⁺ HCC patients frequently present with high AFP levels, and the AFP level has been shown to be the only valuable clinical feature to improve the predictive efficacy of radiomics models (21-23). Similarly, the present study also found that the preoperative AFP levels of the patients were significantly higher in the CK19⁺ HCC group than the non-CK19⁺ hepatic tumor group. High AFP levels in HCC cases are associated with poor differentiation and microvascular invasion (30). Thus, patients with AFP-elevated HCCs, including CK19⁺ HCC, require individualized treatment to improve their prognosis.

Our retrospective analysis had several limitations. First, while the incidence of CK19⁺ HCC is low, our study included over 100 patients, representing the largest radiological series reported. However, the 1:1:1 comparison design for CK19⁺ HCC, CHCC, and ICC may limit the clinical applicability of our diagnostic models. Second, technological advancements in CT hardware and software during the extended study period could influence imaging interpretation. Third, quantitative radiomics features were not incorporated, despite their demonstrated predictive value for CK19+ HCC. Thus, in our future investigations, we intend to expand sample sizes through multicenter collaborations to develop quantitative CT-based diagnostic models for CK19⁺ HCC.

Conclusions

CK19⁺ HCC frequently presents with intratumoral necrosis, LR-TIV, minimal THAD, and either pseudocapsule formation or progressive enhancement on contrast-enhanced CT. Collectively, our findings should enable the construction of diagnostically valuable models.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-24-914/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-24-914/dss

Funding: This work was funded by the Huadong Medicine Joint Fund of the Zhejiang Provincial Natural Science Foundation of China (grant No. LHDMZ24H160001), and the Key Research Project of Jinhua Municipal Bureau of Science and Technology (grant Nos. 2023-3-046 and 2023-3-052).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-24-914/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of the First Affiliated Hospital, Zhejiang University School of Medicine (No. K2023-IRB-092). The Ethics Committee waived the requirement of informed consent due to the retrospective nature of the study.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018;68:394-424. [Crossref] [PubMed]
2. Chen W, Zheng R, Baade PD, Zhang S, Zeng H, Bray F, Jemal A, Yu XQ, He J. Cancer statistics in China, 2015. CA Cancer J Clin 2016;66:115-32. [Crossref] [PubMed]
3. McGlynn KA, Petrick JL, El-Serag HB. Epidemiology of Hepatocellular Carcinoma. Hepatology 2021;73:4-13. [Crossref] [PubMed]
4. Patauner S, Scotton G, Notte F, Frena A. Advanced hepatocellular carcinoma treatment strategies: Are transarterial approaches leading the way? World J Gastrointest Oncol 2025;17:99834. [Crossref] [PubMed]
5. Lin ZP, Hu XL, Chen D, Huang DB, Zou XG, Zhong H, Xu SX, Chen Y, Li XQ, Zhang J. Efficacy and safety of targeted therapy plus immunotherapy combined with hepatic artery infusion chemotherapy (FOLFOX) for unresectable hepatocarcinoma. World J Gastroenterol 2024;30:2321-31. [Crossref] [PubMed]
6. Lu XY, Xi T, Lau WY, Dong H, Zhu Z, Shen F, Wu MC, Cong WM. Hepatocellular carcinoma expressing cholangiocyte phenotype is a novel subtype with highly aggressive behavior. Ann Surg Oncol 2011;18:2210-7. [Crossref] [PubMed]
7. Zhou J, Sun H, Wang Z, Cong W, Wang J, Zeng M, et al. Guidelines for the Diagnosis and Treatment of Hepatocellular Carcinoma (2019 Edition). Liver Cancer 2020;9:682-720. [Crossref] [PubMed]
8. Brunt E, Aishima S, Clavien PA, Fowler K, Goodman Z, Gores G, et al. cHCC-CCA: Consensus terminology for primary liver carcinomas with both hepatocytic and cholangiocytic differentation. Hepatology 2018;68:113-26. [Crossref] [PubMed]
9. Uenishi T, Kubo S, Yamamoto T, Shuto T, Ogawa M, Tanaka H, Tanaka S, Kaneda K, Hirohashi K. Cytokeratin 19 expression in hepatocellular carcinoma predicts early postoperative recurrence. Cancer Sci 2003;94:851-7. [Crossref] [PubMed]
10. Lee JI, Lee JW, Kim JM, Kim JK, Chung HJ, Kim YS. Prognosis of hepatocellular carcinoma expressing cytokeratin 19: comparison with other liver cancers. World J Gastroenterol 2012;18:4751-7. [Crossref] [PubMed]
11. Sun DW, Zhang YY, Sun XD, Chen YG, Qiu W, Ji M, Lv GY. Prognostic value of cytokeratin 19 in hepatocellular carcinoma: A meta-analysis. Clin Chim Acta 2015;448:161-9. [Crossref] [PubMed]
12. Yoneda N, Sato Y, Kitao A, Ikeda H, Sawada-Kitamura S, Miyakoshi M, Harada K, Sasaki M, Matsui O, Nakanuma Y. Epidermal growth factor induces cytokeratin 19 expression accompanied by increased growth abilities in human hepatocellular carcinoma. Lab Invest 2011;91:262-72. [Crossref] [PubMed]
13. Zhang J, Qi YP, Ma N, Lu F, Gong WF, Chen B, Ma L, Zhong JH, Xiang BD, Li LQ. Overexpression of Epcam and CD133 Correlates with Poor Prognosis in Dual-phenotype Hepatocellular Carcinoma. J Cancer 2020;11:3400-6. [Crossref] [PubMed]
14. Minagawa T, Yamazaki K, Masugi Y, Tsujikawa H, Ojima H, Hibi T, Abe Y, Yagi H, Kitago M, Shinoda M, Itano O, Kitagawa Y, Sakamoto M. Activation of extracellular signal-regulated kinase is associated with hepatocellular carcinoma with aggressive phenotypes. Hepatol Res 2020;50:353-64. [Crossref] [PubMed]
15. Wang Y, Wang X, Huang X, Zhang J, Hu J, Qi Y, Xiang B, Wang Q. Integrated Genomic and Transcriptomic Analysis reveals key genes for predicting dual-phenotype Hepatocellular Carcinoma Prognosis. J Cancer 2021;12:2993-3010. [Crossref] [PubMed]
16. Yoneda N, Matsui O, Kobayashi S, Kitao A, Kozaka K, Inoue D, Yoshida K, Minami T, Koda W, Gabata T. Current status of imaging biomarkers predicting the biological nature of hepatocellular carcinoma. Jpn J Radiol 2019;37:191-208. [Crossref] [PubMed]
17. Chung GE, Lee JH, Yoon JH, Myung SJ, Lee K, Jang JJ, Lee JM, Kim SH, Suh KS, Kim YJ, Lee HS. Prognostic implications of tumor vascularity and its relationship to cytokeratin 19 expression in patients with hepatocellular carcinoma. Abdom Imaging 2012;37:439-46. [Crossref] [PubMed]
18. Choi SY, Kim SH, Park CK, Min JH, Lee JE, Choi YH, Lee BR. Imaging Features of Gadoxetic Acid-enhanced and Diffusion-weighted MR Imaging for Identifying Cytokeratin 19-positive Hepatocellular Carcinoma: A Retrospective Observational Study. Radiology 2018;286:897-908. [Crossref] [PubMed]
19. Hu XX, Wang WT, Yang L, Yang ZX, Liang HY, Ding Y, Ji Y, Zeng MS, Rao SX. MR features based on LI-RADS identify cytokeratin 19 status of hepatocellular carcinomas. Eur J Radiol 2019;113:7-14. [Crossref] [PubMed]
20. Chen J, Liu D, Guo Y, Zhang Y, Guo Y, Jiang M, Dai Y, Yao X. Preoperative identification of cytokeratin 19 status of hepatocellular carcinoma based on diffusion kurtosis imaging. Abdom Radiol (NY) 2023;48:579-89. [Crossref] [PubMed]
21. Wang HQ, Yang C, Zeng MS, Rao SX, Ji Y, Weng X, Wang JY, Sheng RF. Magnetic resonance texture analysis for the identification of cytokeratin 19-positive hepatocellular carcinoma. Eur J Radiol 2019;117:164-70. [Crossref] [PubMed]
22. Wang W, Gu D, Wei J, Ding Y, Yang L, Zhu K, Luo R, Rao SX, Tian J, Zeng M. A radiomics-based biomarker for cytokeratin 19 status of hepatocellular carcinoma with gadoxetic acid-enhanced MRI. Eur Radiol 2020;30:3004-14. [Crossref] [PubMed]
23. Chen Y, Chen J, Zhang Y, Lin Z, Wang M, Huang L, Huang M, Tang M, Zhou X, Peng Z, Huang B, Feng ST. Preoperative Prediction of Cytokeratin 19 Expression for Hepatocellular Carcinoma with Deep Learning Radiomics Based on Gadoxetic Acid-Enhanced Magnetic Resonance Imaging. J Hepatocell Carcinoma 2021;8:795-808. [Crossref] [PubMed]
24. Yang F, Wan Y, Xu L, Wu Y, Shen X, Wang J, Lu D, Shao C, Zheng S, Niu T, Xu X. MRI-Radiomics Prediction for Cytokeratin 19-Positive Hepatocellular Carcinoma: A Multicenter Study. Front Oncol 2021;11:672126. [Crossref] [PubMed]
25. Wu Q, Yu YX, Zhang T, Zhu WJ, Fan YF, Wang XM, Hu CH. Preoperative Diagnosis of Dual-Phenotype Hepatocellular Carcinoma Using Enhanced MRI Radiomics Models. J Magn Reson Imaging 2023;57:1185-96. [Crossref] [PubMed]
26. Huang X, Long L, Wei J, Li Y, Xia Y, Zuo P, Chai X. Radiomics for diagnosis of dual-phenotype hepatocellular carcinoma using Gd-EOB-DTPA-enhanced MRI and patient prognosis. J Cancer Res Clin Oncol 2019;145:2995-3003. [Crossref] [PubMed]
27. Chernyak V, Fowler KJ, Kamaya A, Kielar AZ, Elsayes KM, Bashir MR, Kono Y, Do RK, Mitchell DG, Singal AG, Tang A, Sirlin CB. Liver Imaging Reporting and Data System (LI-RADS) Version 2018: Imaging of Hepatocellular Carcinoma in At-Risk Patients. Radiology 2018;289:816-30. [Crossref] [PubMed]
28. Hwang SH, Rhee H. Radiologic features of hepatocellular carcinoma related to prognosis. J Liver Cancer 2023;23:143-56. [Crossref] [PubMed]
29. Wei H, Yang T, Chen J, Duan T, Jiang H, Song B. Prognostic implications of CT/MRI LI-RADS in hepatocellular carcinoma: State of the art and future directions. Liver Int 2022;42:2131-44. [Crossref] [PubMed]
30. Ma WJ, Wang HY, Teng LS. Correlation analysis of preoperative serum alpha-fetoprotein (AFP) level and prognosis of hepatocellular carcinoma (HCC) after hepatectomy. World J Surg Oncol 2013;11:212. [Crossref] [PubMed]