Dual-layer spectral detector computed tomography in distinguishing between bland and neoplastic portal vein thrombosis

Introduction

Liver malignancies, particularly hepatocellular carcinoma (HCC), can invade the hepatic vascular system, especially the portal vein, leading to the development of neoplastic portal vein thrombosis (PVT) (1-4). The presence of neoplastic PVT signifies an advanced stage of the disease and is associated with a poor prognosis (2,5,6).

Meanwhile, bland PVT, which is associated with slow portal blood flow, is observed in 20–40% of patients with HCC, particularly those with cirrhosis or those who have undergone splenectomy (1,7). Identifying neoplastic PVT and accurately differentiating it from bland PVT are crucial for staging, treatment decisions, and prognosis prediction in patients with suspected or confirmed liver malignancies (2,3,6,8). Although histopathological examination remains the gold standard for differentiating between bland and neoplastic PVT, imaging plays a vital role in noninvasive preoperative characterization.

The Liver Imaging Reporting and Data System (LI-RADS) tumor in vein (LR-TIV) defines unequivocal enhancing soft tissue in vein as TIV (9). However, despite having high specificity (96–99.8%), the LR-TIV has moderate sensitivity (64.4–85%) in detecting TIV or distinguishing bland thrombus from TIV based on conventional computed tomography (CT) or magnetic resonance imaging (MRI), which may lead to inappropriate treatment decisions in patients with missed TIV (10,11).

Dual-layer spectral detector CT (DLCT) is a novel spectral CT technology that utilizes dual-layer detectors to simultaneously collect low- and high-energy data, enabling “homogeneous, simultaneous, and co-directional” spectral data acquisition (12). Unlike traditional dual-energy CT (DECT), DLCT does not require preselected spectral acquisition modes. Instead, spectral images can be retrospectively generated from raw data. This feature renders the clinical application of spectral CT more convenient while ensuring that quantitative accuracy is unaffected by variations in acquisition parameters, potentially facilitating the detection of neoplastic PVT (13).

Previous studies have shown that quantitative DECT parameters for PVT (e.g., iodine density in late arterial and portal venous phases) help to distinguish neoplastic from bland PVT, with an excellent area under the receiver operating characteristic (ROC) curve (AUC) and sensitivity (7,14). However, the value of DLCT in the differential diagnosis of PVT has not been reported in the literature, particularly regarding the comparison of late arterial and portal venous phase parameters, and a comprehensive assessment of other DLCT quantitative parameters is lacking.

Therefore, the purpose of this study was to comprehensively assess the value of quantitative DLCT parameters in distinguishing between bland and neoplastic PVT, which could facilitate accurate staging and inform treatment decisions for these patients. We present this article in accordance with the STARD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-24-1890/rc).

Methods

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments and was approved by the Institutional Review Board of West China Hospital, Sichuan University (approval No. 2221). Informed consent was obtained from all participants.

Study design and patients

This single-center prospective study was conducted between April 2022 and August 2023. Consecutive patients were identified based on DLCT imaging reports that included findings of suspected PVT. Suspected PVT was considered to be the presence of filling defects or identifiable thrombosis in the portal vein on DLCT imaging reports. We then performed additional history and imaging reviews of these cases to determine the final study cohort. Patients were enrolled according to the following predefined inclusion criteria: (I) completion of contrast-enhanced liver DLCT and (II) identifiable PVT. Meanwhile, the exclusion criteria were as follows: (I) prior locoregional or systemic therapy for liver malignancies (e.g., liver resection, transarterial chemoembolization, and chemotherapy) and (II) imaging artifacts or insufficient quality for accurate quantitative analysis. Clinical and laboratory data, including patient demographics, liver disease etiologies, and tumor markers [e.g., α-fetoprotein (AFP) and aspartate aminotransferase (AST)], were prospectively collected within 1 month of imaging.

Reference standard

The classification of PVT as bland or neoplastic was based on predefined criteria, including imaging findings from DLCT, clinical context, and follow-up data. PVT was classified as bland or neoplastic primarily based on LR-TIV criteria (i.e., the presence of unequivocal enhancing soft tissue in the portal vein indicated neoplastic PVT) in high-risk patients [i.e., those with cirrhosis, chronic hepatitis B virus (HBV) infection, or current or prior HCC] (9), combined with the patient’s clinical history, prior and follow-up imaging studies, and histopathology (7,11,14,15). Follow-up imaging of ≥3 months was available for all cases included in the final analysis to ensure correct classification. Additionally, imaging findings were corroborated by clinical data, including (I) the patient’s treatment history (e.g., no history of antitumor therapy status) and (II) laboratory markers such as non-elevated serum AFP. Cases showing ambiguous enhancement or AFP-negative thrombus progression were excluded to minimize misclassification bias. The criteria applied are described below.

Neoplastic PVT

The primary imaging criteria for neoplastic PVT included the presence of unequivocal enhancing soft tissue in the portal vein, as defined by the LR-TIV criteria in high-risk patients (9).

The supplementary imaging evidence included substantial growth or progressive enhancement of the thrombus within 3 months in the context of concurrent liver malignancies or elevated serum AFP (7,11,14,15). For the follow-up imaging, we considered an increase in the longest transverse diameter of thrombus ≥30% or PVT enhancement ≥20 Hounsfield units (HU) in the arterial phase to be neoplastic PVT to ensure accuracy and specificity (7,14).

Histopathological confirmation included available for patients who underwent surgery.

Bland PVT

The imaging criteria for bland PVT included stability, shrinkage, or complete resolution of the thrombus on follow-up imaging without evidence of systemic or regional (e.g., radiotherapy) antitumor therapy (7,11,14,15).

The supplementary clinical evidence included the absence of systemic progression of the underlying malignancy, supported by treatment records and laboratory data. If available, follow-up imaging (minimum 3 months) was reviewed to confirm the stability or progression of PVT in order to minimize the risk of misclassification.

All cases were independently reviewed by two radiologists (7 and 10 years of experience in liver imaging) who reached a consensus regarding the classification of PVT as bland or neoplastic.

Imaging technique

Patients underwent DLCT with a predefined standardized scanning protocol. Contrast-enhanced multiphase CT, consisting of precontrast, arterial, and portal venous phases, was performed on a DLCT scanner (Spectral CT 7500, Philips Healthcare, Best, the Netherlands) in all patients. Nonionic contrast agent (1.0–1.2 mL/kg of Ultravist 370; Bayer AG, Leverkusen, Germany) was administered intravenously at a rate of 3.0 mL/s with an automated injector (MEDRAD Stellant, Bayer AG), followed by a 20-mL saline flush applied at the same rate. The acquisition of the arterial and portal venous phases was initiated by bolus tracking with automated triggering, which automatically began at 8 and 35 s, respectively, after a trigger threshold of 150 HU was reached in the supraceliac abdominal aorta. Image acquisition parameters included a tube voltage of 120 kV, automated tube current modulation, a collimation of 128 mm × 0.625 mm, and a reconstruction matrix of 512×512. Conventional images were reconstructed with the iDose 4 algorithm (Philips Healthcare), while spectral-based images (SBIs) were reconstructed via the spectral level 4 algorithm, with a slice thickness of 1 mm and an increment of 1 mm.

Imaging analysis

Quantitative measurements of thrombi were performed prospectively by two radiologists following a structured protocol. Regions of interest (ROIs) were placed on thrombi for quantitative analysis, with small or indistinct thrombi being avoided. PVTs that were smaller than 5 mm in diameter were considered too small to characterize due to limitations in imaging resolution and difficulty in ROI placement.

All image data were transferred to a workstation (IntelliSpace Portal version 12.0, Philips Healthcare) for postprocessing and analysis. Virtual monoenergetic images (40–90 keV), iodine density maps, effective atomic number (Zeff) maps of the arterial and portal venous phases, and arterial enhancement fraction (AEF) maps were automatically generated from SBI. Virtual monoenergetic images from 40 to 90 keV were analyzed based on evidence from a spectral CT study suggesting optimal contrast resolution and diagnostic utility within this range (16).

Two radiologists with 7 and 10 years of experience in liver imaging who were blinded to clinical and pathological information independently performed measurements after a 2-month washout period following patient enrollment. Both radiologists underwent a consensus training session to standardize ROI placement protocols. ROIs were manually drawn within the PVT on the arterial and portal venous phase images while AEF maps were drawn at two consecutive slices representing the maximum cross-section of the thrombus as follows: (I) for the arterial and portal venous phase images, the ROIs were initially placed on iodine density maps (the area with the highest density of portal vein emboli) for ease of localizing the PVT and then automatically coregistered to conventional and virtual monoenergetic (40–90 keV) images and Zeff maps (at the same position and slice) via vendor software, minimizing registration-related misalignment. Interphase consistency was visually verified. (II) The size and position of the ROIs remained almost consistent between the two slices and between the arterial and portal venous phase images and AEF maps. (III) ROIs covered as many areas of PVT as possible, with a minimum size of 10 mm². (IV) The vessel borders, necrosis, calcification, and artifacts, if any, were carefully avoided. The averaged measurement from the two slices and two radiologists were used for further analyses. For each patient, the CT attenuation values of conventional and virtual monoenergetic (40–90 keV) images, iodine density, Zeff, and AEF values were documented. Additionally, the maximum PVT diameter on cross-sectional iodine density maps was recorded. Radiological cirrhosis was diagnosed based on typical imaging findings (e.g., liver surface nodularity) with or without signs of portal hypertension (e.g., splenomegaly and gastroesophageal varices) (17).

Statistical analysis

Continuous variables are presented as the mean ± standard deviation or as median values with ranges, as appropriate, while categorical variables are presented as numbers with percentages. Continuous variables were compared via the Student t-test or the Mann-Whitney test, whereas categorical variables were compared via the Chi-squared test or the Fisher exact test, as appropriate.

Interobserver agreement was assessed via the intraclass correlation coefficient (ICC) for continuous variables. An ICC of 0.8–1.0, 0.6–0.79, 0.4–0.59, 0.2–0.39, and 0–0.19 was considered excellent, good, moderate, fair, and poor agreement, respectively.

Correlations between quantitative DLCT parameters and neoplastic PVT were assessed via Pearson or Spearman correlation analysis, with patient age and sex being controlled for. To improve the diagnostic performance of single quantitative parameters, independent variables from the correlation analysis were entered into multivariable linear regression analysis with backward stepwise selection. Subsequently, a combined score was developed based on significant variables weighted by their β regression coefficients. Optimal thresholds for continuous variables were determined by ROC analysis with the Youden index. Diagnostic performances were evaluated according to the AUC, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV).

All statistical analyses were performed via SPSS software version 26.0 (IBM Corp., Armonk, NY, USA), MedCalc version 20.100-64-bit (MedCalc Software, Ostend, Belgium), and R statistical software (The R Foundation for Statistical Computing, Vienna, Austria). A two-sided P value <0.05 was considered statistically significant.

Results

Patients

A total of 81 patients (mean age 54±11 years; 64 males) were included in this study, including 31 patients with bland PVT and 50 patients with neoplastic PVT (Table 1 and Figure 1). Among them, 85% (69/81) had HBV infection and 84% (68/81) had radiological or pathological cirrhosis. Liver tumor was detected in 77% (62/81) of the patients.

Figure 1 Flowchart of the study cohort. CT, computed tomography; LI-RADS, Liver Imaging Reporting and Data System; ROI, region of interest; TIV, tumor in vein.

There was a higher proportion of female patients in the bland PVT group than in the neoplastic PVT group (42% vs. 8%, P<0.001). Patients with bland PVT had a greater proportion of non-HBV etiologies (19% vs. 0%, P=0.002), lower AFP (median 3.3 vs. 932.0 ng/mL, P<0.001) and AST (median 40.0 vs. 68.5 U/L, P=0.004) levels, smaller PVT (mean diameter 13.6 vs. 16.5 mm, P=0.02), and a lower incidence of liver tumor (39% vs. 100%, P<0.001). There was no difference in other clinical characteristics between the bland and neoplastic PVT groups.

Comparison of quantitative parameters between the bland PVT and neoplastic PVT groups

The CT attenuation values of conventional and virtual monoenergetic (40–90 keV) images, iodine density, Zeff, and AEF values were significantly higher in the neoplastic PVT group than in the bland PVT group (all P values ≤0.001) (Table 2).

Interobserver agreement of quantitative parameters

Interobserver agreement was good for arterial phase Zeff (ICC 0.796), while excellent for all other arterial and portal venous phase quantitative parameters derived from DLCT (ICC 0.924–0.973) (Table S1).

Correlation of quantitative parameters with neoplastic PVT

All arterial and portal venous phase quantitative parameters derived from DLCT were positively correlated with neoplastic PVT (r=0.411–0.700; all P values <0.001) (Table 3). AEF demonstrated the highest correlation coefficient (r=0.700), followed by the CT attenuation value of the arterial-phase 40-keV virtual monoenergetic images (r=0.670).

Diagnostic performances of single quantitative parameters for neoplastic PVT

Among the single quantitative parameters derived from DLCT, the CT attenuation values of the arterial-phase 40-keV virtual monoenergetic images demonstrated the highest AUC of 0.938 [95% confidence interval (CI): 0.862–0.980; sensitivity 84.0%; specificity 96.8%], followed by AEF, with an AUC of 0.930 (95% CI: 0.851–0.975; sensitivity 90.0%; specificity 93.6%) (Table 4 and Figure 2). The arterial phase quantitative parameters had higher diagnostic AUCs than did the portal venous phase parameters (0.892–0.938 vs. 0.866–0.890). The specificity of arterial phase quantitative parameters for diagnosing neoplastic PVT was higher than that of portal venous phase parameters (93.6–96.8% vs. 80.7–83.9%), while the sensitivity was lower (74.0–86.0% vs. 88.0–96.0%).

Figure 2 ROC curves of DLCT quantitative parameters from the arterial phase, AEF, and the combined score for diagnosing neoplastic PVT. The combined score derived from CT attenuation values of conventional and virtual monoenergetic (40–90 keV) images, iodine density, Zeff of the arterial phase, and AEF showed excellent diagnostic efficiency for neoplastic PVT (AUC 0.979). AEF, arterial enhancement fraction; AUC, area under the receiver operating characteristic curve; CT, computed tomography; DLCT, dual-layer spectral detector computed tomography; PVT, portal vein thrombosis; ROC, receiver operating characteristic; Zeff, effective atomic number.

Development and evaluation of the combined score for diagnosing neoplastic PVT

To improve the diagnostic performance of CT attenuation values of the arterial phase images while minimizing the workload of delineating ROI and calculation, all arterial phase quantitative parameters {i.e., the CT attenuation values of conventional and virtual monoenergetic [40–90 keV] images, iodine density, and Zeff of the arterial phase} and AEF were input into multivariable regression analyses, all of which were significantly associated with neoplastic PVT. Based on these variables, the combined score was formulated in the form of a linear equation as follows: combined score = 1,963.819 − 0.431 × (CT value_A_conventional) + 3.938 × (CT value_A_40 keV) + 2.126 × (CT value_A_50 keV) + 1.007 × (CT value_A_60 keV) − 25.997 × (CT value_A_70 keV) + 8.865 × (CT value_A_80 keV) + 10.53 × (CT value_A_90 keV) + 46.854 × (Iodine density_A) − 271.27 × (Zeff_A) + 0.119 × (AEF). In the above formula, “(CT value_A_conventional/40 keV/50 keV/60 keV/70 keV/80 keV/90 keV)” denote the CT attenuation values (HU) measured in the arterial phase under conventional imaging and at the specified monoenergetic levels (keV), respectively. “(Iodine density_A)” and “(Zeff_A)” denote the iodine concentration (mg/mL) measured on the arterial-phase iodine density map reconstructed from DLCT, and Zeff derived from the same voxel set in the arterial phase. “(AEF)” is denoted the absolute enhancement ratio of arterial-phase to portal-venous-phase CT attenuation increments (%).

To facilitate the calculation of the combined score, we developed a program (https://thecodexbygit.github.io/DLCT-PVT/DLCT-PVT-ScoreCalc.html) called the “Combined Score Calculator”. Figures S1-S3 provide detailed examples of how the Combined Score Calculator is used.

With 0.24 serving as the optimal threshold, the combined score yielded an AUC of 0.979 (95% CI: 0.919–0.998) for distinguishing neoplastic PVT, outperforming all individual DLCT parameters (AUC 0.844–0.938; all P values <0.05) except for AEF (AUC 0.930; P=0.13) (Table 4 and Figure 2). The sensitivity, specificity, PPV, and NVP of the combined score were 94.0%, 100.0%, 100.0%, and 96.4%, respectively.

Figures 3,4 show the DLCT images of patients with bland and neoplastic PVT, respectively.

Figure 3 DLCT images of a 53-year-old woman with newly formed bland thrombus in the left branch of the portal vein about 1 month after resection of HCC. Circular ROIs (red circles) were manually drawn within the portal vein thrombus on the precontrast (J), arterial (A-D), and portal venous phase (F-I) images and AEF map (E). AEF, arterial enhancement fraction; AP, arterial phase; DLCT, dual-layer spectral detector computed tomography; HCC, hepatocellular carcinoma; HU, Hounsfield units; PV, portal venous phase; ROI, region of interest; Z effective, effective atomic number.

Figure 4 DLCT images of a 30-year-old man with HCC and neoplastic thrombus in the right secondary branch of the portal vein. Circular ROIs (red circles) were manually drawn within the portal vein thrombus on the precontrast (J), arterial (A-D), and portal venous phase (F-I) images and AEF map (E). AEF, arterial enhancement fraction; AP, arterial phase; DLCT, dual-layer spectral detector computed tomography; HCC, hepatocellular carcinoma; HU, Hounsfield units; PV, portal venous phase; ROI, region of interest; Z effective, effective atomic number.

Discussion

Accurate differentiation between bland and neoplastic PVT is crucial for treatment decisions in patients with suspected or confirmed liver malignancies. However, the value of DLCT in the differential diagnosis of PVT remains unknown. By prospectively analyzing 81 patients with macroscopic PVT on DLCT, we found that quantitative DLCT parameters were different between the bland and neoplastic groups (all P values ≤0.001). The interobserver agreement of DLCT quantitative parameters ranged from good to excellent (ICC 0.796–0.973). Arterial phase quantitative parameters, as compared to portal venous phase parameters, exhibited higher diagnostic AUCs (0.892–0.938 vs. 0.866–0.890) and specificities (specificity 93.6–96.8% vs. 80.7–83.9%) but relatively lower sensitivities for neoplastic PVT (sensitivity 74.0–86.0% vs. 88.0–96.0%). The combined score derived from DLCT quantitative parameters achieved the highest AUC of 0.979 (95% CI: 0.919–0.998) for diagnosing neoplastic PVT, potentially improving the diagnostic performance of single DLCT quantitative parameters.

In DECT, material decompensation based on high- and low-energy data enables the precise quantification of iodine concentration in tissues (16). Iodine density has been reported to be a predictor of Ki-67 expression (18), microvascular invasion (MVI) (19), and treatment outcomes after antitumor therapy (20,21) and as mentioned above, as a marker for distinguishing between bland and neoplastic PVT (7,14). Virtual monoenergetic images approximate the appearance of images acquired with a monochromatic X-ray beam and improve soft tissue contrast while maintaining low image noise at low energy levels (40–70 keV) (16). Virtual monoenergetic images provide additional information in the evaluation of the macrotrabecular-massive subtype of HCC (22). AEF, defined as the ratio of enhancement in the arterial phase to that of the portal venous phase, indirectly reflects the ratio of hepatic arterial perfusion to total hepatic perfusion (23). AEF has been reported to correlate with response to transarterial chemoembolization in HCC (24) and recurrence after radiofrequency ablation in liver metastasis (23). Zeff characterizes the photon attenuation properties of a compound as a whole and facilitates tissue characterization and discrimination (12,25). Zhu et al. found that arterial phase Zeff was highly predictive of MVI in HCC (26). Our study was the first to examine and confirm the value of virtual monoenergetic images, Zeff, and AEF in distinguishing between bland and neoplastic PVT.

DLCT is an advanced CT technique that simultaneously obtains two photon spectra via a dual-layer detector without the need for additional equipment or scanning protocols, as required in traditional DECT. In our study, quantitative DLCT parameters were significantly different between the bland and neoplastic PVT groups and effectively distinguished between them. This might be attributable to the high sensitivity of DECT to changes in iodine concentration and the angiogenesis associated with neoplastic PVT.

According to the LR-TIV and other proposed imaging diagnostic criteria for neoplastic PVT, unequivocal enhancement is the key radiological finding to differentiate bland from neoplastic PVT (9,10,27). However, few studies have compared the value of arterial and that of portal venous phases in detecting PVT enhancement. Gawande et al. reported that the AUCs for differentiating benign and malignant PVT were similarly high across phases of contrast in the measurement of signal intensity on pre- and contrast-enhanced MRI (15). Regarding DECT, Ascenti et al. (7) and Qian et al. (14) separately reported the value of iodine density in the arterial and portal venous phases for characterizing PVT. In our study, we found that arterial phase quantitative parameters had lower sensitivities, but higher specificities, as compared to the portal venous phase parameters. Notably, the CT attenuation values of arterial-phase, 40-keV images emerged as the most effective single DLCT quantitative parameter for diagnosing neoplastic PVT. To further improve its diagnostic performance while minimizing the workload of delineating ROI and calculation, we developed a combined score by incorporating arterial phase DLCT quantitative parameters and AEF. The combined score achieved a higher AUC of 0.979 and a more balanced sensitivity and specificity (94.0% and 100.0%, respectively) for diagnosing neoplastic PVT, suggesting the complementarity of arterial and portal venous phase images as well as different DLCT quantitative parameters. We would like to clarify that the development of the combined score was intended not as a final predictive tool but rather as a proof-of-concept demonstration to highlight the added diagnostic value of integrating multiple spectral parameters.

Previous studies have reported conflicting interobserver agreement and unsatisfactory sensitivity of the imaging feature of enhancing soft tissue in vein for the detection and differentiation of neoplastic PVT. Ichikawa et al. (11) found that in gadoxetic acid-enhanced MRI, the enhancing soft tissue in vein had moderate interobserver agreement (ICC 0.63; 95% CI: 0.54–0.71), with sensitivity ranging from 62% to 93% across readers. In contrast, a study by Bae et al. (10) reported substantial interobserver agreement for both CT (k=0.80; 95% CI: 0.77–0.83) and gadoxetic acid-enhanced MRI (k=0.78; 95% CI: 0.75–0.81), with a sensitivity of 64.4% (95% CI: 54.2–73.6) for CT and 62.4% (95% CI: 52.2–71.8) for MRI. In our study, the interobserver agreement was excellent for all DLCT quantitative parameters (ICC 0.924–0.973) except for arterial phase Zeff (ICC 0.796). The sensitivity and specificity of the combined score derived from quantitative DLCT parameters reached 94.0% and 100.0%, respectively. These findings suggest that DLCT may more reliably differentiate bland from neoplastic PVT than traditional imaging features. However, further validation is needed to confirm our results.

Certain limitations to this study should be noted. First, as we employed a single-center, prospective design and a small sample size, there was a lack of both internal and external validation cohort to test and refine our findings. Moreover, all patients with neoplastic PVT had concurrent liver tumors, which might have introduced selection bias. Additionally, our study was conducted with a single scanner and scan parameter settings, limiting the generalizability of our findings. Therefore, future large-scale multicenter studies are warranted to enroll a more diverse sample, including nonmalignant causes and extrahepatic malignancies. Second, there were no pathological results used as the reference standard due to the risks associated with portal vein biopsy and the unresectability of the patients in the neoplastic PVT group. Instead, we applied established imaging-based diagnostic criteria supported by longitudinal follow-up, consistent with previous literature (7,11,14,15). However, in rare cases, neoplastic PVT with slow growth and less angiogenetic, as well as nonneoplastic PVT with growth and enhancement, may also be present. Although we endeavored to improve the accuracy of the reference standard with LR-TIV criteria by combining it with patient history and previous follow-up imaging and histopathological examinations, the aforementioned cases might have still been misdiagnosed. Third, we sampled the PVT using manually drawn circular ROIs, which might not have sufficiently represented the entire PVT and could have resulted in misregistration between the arterial and portal venous phase images and AEF maps. The diagnostic utility of DLCT-derived quantitative parameters in cases of small PVT may be limited, as reduced ROI size could increase the likelihood of measurement errors.

Conclusions

The quantitative DLCT parameters demonstrated excellent performance in distinguishing between bland and neoplastic PVT and may thus aid in the treatment allocation for patients with suspected or confirmed liver malignancies. The combined score derived from the arterial phase quantitative DLCT parameters and AEF achieved the highest diagnostic performance. Further large-scale multicenter studies are needed to confirm our results.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STARD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-24-1890/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-24-1890/dss

Funding: This work was supported by the National Natural Science Foundation of China (No. U22A20343), the China Post-doctoral Science Foundation (No. 2023M732435), the Science and Technology Department of Sichuan Province (No. 2022YFS0071), the Science and Technology Department of Guangxi Zhuang Autonomous Region (No. 2025GXNSFAA069531), the Science and Technology Department of Hainan Province (No. ZDYF2024SHFZ052), the Development Project of Hainan Provincial Clinical Medical Center, and the Post-Doctoral Station Development Project of Sanya (No. 23CZ009).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-24-1890/coif). X.Z. and H.L. are current employees of Philips Healthcare. The other 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 institutional review board of West China Hospital, Sichuan University (No. 2221) and informed consent was obtained from all individual participants.

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. Lu J, Zhang XP, Zhong BY, Lau WY, Madoff DC, Davidson JC, Qi X, Cheng SQ, Teng GJ. Management of patients with hepatocellular carcinoma and portal vein tumour thrombosis: comparing east and west. Lancet Gastroenterol Hepatol 2019;4:721-30. [Crossref] [PubMed]
2. EASL-ILCA Clinical Practice Guidelines on the management of intrahepatic cholangiocarcinoma. J Hepatol 2023;79:181-208. [Crossref] [PubMed]
3. Beaufrère A, Calderaro J, Paradis V. Combined hepatocellular-cholangiocarcinoma: An update. J Hepatol 2021;74:1212-24. [Crossref] [PubMed]
4. Lee KF, Chu W, Lai PB. Portal vein tumor thrombus in colorectal liver metastasis. Am J Surg 2005;190:364-5. [Crossref] [PubMed]
5. Llovet JM, Bustamante J, Castells A, Vilana R, Ayuso Mdel C, Sala M, Brú C, Rodés J, Bruix J. Natural history of untreated nonsurgical hepatocellular carcinoma: rationale for the design and evaluation of therapeutic trials. Hepatology 1999;29:62-7. [Crossref] [PubMed]
6. Singal AG, Llovet JM, Yarchoan M, Mehta N, Heimbach JK, Dawson LA, Jou JH, Kulik LM, Agopian VG, Marrero JA, Mendiratta-Lala M, Brown DB, Rilling WS, Goyal L, Wei AC, Taddei TH. AASLD Practice Guidance on prevention, diagnosis, and treatment of hepatocellular carcinoma. Hepatology 2023;78:1922-65. [Crossref] [PubMed]
7. Ascenti G, Sofia C, Mazziotti S, Silipigni S, D'Angelo T, Pergolizzi S, Scribano E. Dual-energy CT with iodine quantification in distinguishing between bland and neoplastic portal vein thrombosis in patients with hepatocellular carcinoma. Clin Radiol 2016;71:938.e1-9. [Crossref] [PubMed]
8. EASL Clinical Practice Guidelines. Management of hepatocellular carcinoma. J Hepatol 2018;69:182-236. [Crossref] [PubMed]
9. 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]
10. Bae JS, Lee JM, Jeon SK, Yoo J, Park SJ, Yoon JH, Joo I, Lee KB, Kim H. LI-RADS Tumor in Vein at CT and Hepatobiliary MRI. Radiology 2022;302:107-15. [Crossref] [PubMed]
11. Ichikawa S, Sirlin CB, Cunha GM, Wolfson T, Fowler KJ, Morisaka H, Onohara K, Shimizu T, Omiya Y, Sugiyama H, Okumura A, Onishi H. Inter-observer agreement and accuracy of LI-RADS v2018 for differentiating tumor in vein from bland thrombus using gadoxetic acid-enhanced magnetic resonance imaging. Abdom Radiol (NY) 2023;48:2557-69. [Crossref] [PubMed]
12. Rassouli N, Etesami M, Dhanantwari A, Rajiah P. Detector-based spectral CT with a novel dual-layer technology: principles and applications. Insights Imaging 2017;8:589-98. [Crossref] [PubMed]
13. Liu LP, Shapira N, Halliburton SS, Meyer S, Perkins A, Litt HI, Kauczor HU, Leiner T, Stiller W, Noël PB. Spectral performance evaluation of a second-generation spectral detector CT. J Appl Clin Med Phys 2024;25:e14300. [Crossref] [PubMed]
14. Qian LJ, Zhu J, Zhuang ZG, Xia Q, Cheng YF, Li JY, Xu JR. Differentiation of neoplastic from bland macroscopic portal vein thrombi using dual-energy spectral CT imaging: a pilot study. Eur Radiol 2012;22:2178-85. [Crossref] [PubMed]
15. Gawande R, Jalaeian H, Niendorf E, Olgun D, Krystosek L, Rubin N, Spilseth B. MRI in differentiating malignant versus benign portal vein thrombosis in patients with hepatocellular carcinoma: Value of post contrast imaging with subtraction. Eur J Radiol 2019;118:88-95. [Crossref] [PubMed]
16. Greffier J, Villani N, Defez D, Dabli D, Si-Mohamed S. Spectral CT imaging: Technical principles of dual-energy CT and multi-energy photon-counting CT. Diagn Interv Imaging 2023;104:167-77. [Crossref] [PubMed]
17. Zheng T, Qu Y, Chen J, Yang J, Yan H, Jiang H, Song B. Noninvasive diagnosis of liver cirrhosis: qualitative and quantitative imaging biomarkers. Abdom Radiol (NY) 2024;49:2098-115. [Crossref] [PubMed]
18. Xu L, Yu Sun F, Xian Li G, Zai Yang L, Xiao Zhuo Y, Xiao Hu WXH, Wen X. Correlation between Spectral CT Parameters and Ki67 Expression in Hepatocellular Carcinoma. Curr Med Imaging 2023; Epub ahead of print. [Crossref]
19. Yang CB, Zhang S, Jia YJ, Yu Y, Duan HF, Zhang XR, Ma GM, Ren C, Yu N. Dual energy spectral CT imaging for the evaluation of small hepatocellular carcinoma microvascular invasion. Eur J Radiol 2017;95:222-7. [Crossref] [PubMed]
20. Li S, Yuan L, Yue M, Xu Y, Liu S, Wang F, Liu X, Wang F, Deng J, Sun Q, Liu X, Xue C, Lu T, Zhang W, Zhou J. Early evaluation of liver metastasis using spectral CT to predict outcome in patients with colorectal cancer treated with FOLFOXIRI and bevacizumab. Cancer Imaging 2023;23:30. [Crossref] [PubMed]
21. Luo N, Li W, Xie J, Fu D, Liu L, Huang X, Su D, Jin G. Preoperative normalized iodine concentration derived from spectral CT is correlated with early recurrence of hepatocellular carcinoma after curative resection. Eur Radiol 2021;31:1872-82. [Crossref] [PubMed]
22. Li M, Fan Y, You H, Li C, Luo M, Zhou J, Li A, Zhang L, Yu X, Deng W, Zhou J, Zhang D, Zhang Z, Chen H, Xiao Y, Huang B, Wang J, Dual-Energy CT. Deep Learning Radiomics to Predict Macrotrabecular-Massive Hepatocellular Carcinoma. Radiology 2023;308:e230255. [Crossref] [PubMed]
23. Mahnken AH, Klotz E, Schreiber S, Bruners P, Isfort P, Günther RW, Wildberger JE. Volumetric arterial enhancement fraction predicts tumor recurrence after hepatic radiofrequency ablation of liver metastases: initial results. AJR Am J Roentgenol 2011;196:W573-9. [Crossref] [PubMed]
24. Chai B, Xiang D, Wang W, Ren Y, Wang F, Wang J, Zhou G, Zheng C. Arterial enhancement fraction in evaluating the therapeutic effect and survival for hepatocellular carcinoma patients treated with DEB-TACE. Cancer Imaging 2022;22:38. [Crossref] [PubMed]
25. Liu T, Hong G, Cai W. A comparative study of effective atomic number calculations for dual-energy CT. Med Phys 2021;48:5908-23. [Crossref] [PubMed]
26. Zhu Y, Feng B, Cai W, Wang B, Meng X, Wang S, Ma X, Zhao X. Prediction of Microvascular Invasion in Solitary AFP-Negative Hepatocellular Carcinoma ≤ 5 cm Using a Combination of Imaging Features and Quantitative Dual-Layer Spectral-Detector CT Parameters. Acad Radiol 2023;30:S104-16. [Crossref] [PubMed]
27. Sherman CB, Behr S, Dodge JL, Roberts JP, Yao FY, Mehta N. Distinguishing Tumor From Bland Portal Vein Thrombus in Liver Transplant Candidates With Hepatocellular Carcinoma: the A-VENA Criteria. Liver Transpl 2019;25:207-16. [Crossref] [PubMed]