Value of dual-energy computed tomography quantitative parameters in differentiating neoplastic from bland portal vein thrombosis

Introduction

Portal vein thrombosis (PVT) is the formation of thrombi in the main portal vein or its intrahepatic branches, caused by various etiologies with or without concurrent involvement of the splenic vein or superior mesenteric vein (1-3). Notably, PVT can arise from either bland thrombosis or neoplastic tumor invasion. Hepatocellular carcinoma (HCC), the sixth most common cancer globally and the third leading cause of cancer-related mortality (4), frequently invades hepatic vasculature, particularly the portal vein system, due to its anatomical and biological characteristics. Portal vein tumor thrombus (PVTT) is observed in approximately 44.0–62.2% of HCC patients (5). According to both the Barcelona Clinic Liver Cancer and China Liver Cancer Staging, the presence of PVTT generally indicates advanced disease, for which surgical resection is often not considered the primary treatment option (6,7). Conversely, bland PVT can often be effectively treated with anticoagulant therapy (1). The characteristic of the portal venous thrombus has significant implications for treatment strategy and prognosis, underscoring the importance of an accurate qualitative diagnosis.

Although biopsy remains the gold standard for diagnosing PVT, its invasiveness, risk of bleeding, sample errors, and potential for seeding metastases highlight the critical role of imaging techniques. Modalities such as ultrasound, conventional computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography (PET) are commonly used to identify PVT (8-12), but they lack quantitative indicators and are heavily operator-dependent, which can complicate the differentiation between bland and neoplastic PVT in routine practice. Thus, there is an urgent need for a highly accurate and user-friendly method to identify bland from neoplastic PVT.

In recent years, dual-energy computed tomography (DECT) has gained widespread acceptance as an advanced diagnostic technique. It utilizes two sets of X-ray tubes and detectors to distinguish tissues based on their attenuation characteristics at different X-ray energies, offering superior diagnostic capabilities compared to traditional CT (13). DECT can improve image quality, reduce beam-hardening, and provide quantitative parameters such as iodine concentration (IC), normalized iodine concentration (NIC), effective atomic number (Z), dual-energy index (DEI), and spectral slope (K), which are valuable for disease diagnosis, differentiation, therapeutic evaluation, and prognostic assessment (14,15). Researchers have successfully utilized DECT quantitative parameters to detect and assess tumor lesions, demonstrating promising results (16-18). Thus, based on the different mechanisms and constituent components of PVTT and bland PVT, estimation of the quantitative parameter differences using DECT may be able to distinguish them, which holds crucial clinical value for choosing treatment, forecasting survival, and assessing PVTT activity post-treatment.

However, research on the use of DECT quantitative parameters to differentiate between bland and neoplastic PVT is restricted, with only preliminary studies on IC having been conducted by Qian et al. (19) and Ascenti et al. (20). Hence, this study was conducted to evaluate the viability and diagnostic utility of DECT quantitative parameters in differentiating neoplastic from bland thrombus. We present this article in accordance with the STARD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2516/rc).

Methods

Participants and study design

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This retrospective study was approved by the Ethics Committee of The First Affiliated Hospital of Hunan Normal University Hunan Provincial People’s Hospital (No. [2024]-196), and the requirement to obtain informed consent from patients was waived by the Ethics Committee. In this investigation, we retrieved the data of 173 patients diagnosed with PVT who had undergone portal venous phase multi-phase enhanced DECT imaging. The data was sourced retrospectively from the Picture Archiving and Communication System (PACS) of the Imaging Department at Hunan Provincial People’s Hospital between August 2022 and November 2023 (Figure 1). Meanwhile, patient baseline data, including demographics (age, gender) and lab tests, were retrieved from our hospital’s Hospital Information System (HIS) e-medical record system and outpatient big data platform. All the participants in our study were selected consecutively. The inclusion criteria were as follows: (I) PVT present in the main branch and/or segmental branch; (II) complete clinical information; and (III) an enhanced DECT scan was performed. The exclusion criteria were as follows: (I) patients receiving liver transplantation; (II) PVT is too small to measure; (III) after treatment of liver neoplastic tumor; and (IV) poor image quality.

Figure 1 The flowchart of participants throughout the study. DECT, dual-energy computed tomography; PVT, portal vein thrombosis.

Reference standard for diagnosis and characterization of PVT

Two abdominal imaging specialists, one with 3 years of experience and the other with 10 years, independently classified the thrombus as bland or neoplastic based on predefined criteria. The reviewers were blinded to each patient’s clinical data and medical history and were not involved in the subsequent quantitative analysis. In cases where the results were inconsistent, a decision was made through consultation between the two parties.

The reference standards included pathology results and established multidetector computed tomography (MDCT) criteria for PVT, along with a comparative analysis of thrombus characteristics from MDCT scans performed within 1–3 months (19,21-23). The CT criteria for neoplastic thrombi are defined as follows: (I) direct invasion of tumor into the portal vein; (II) expansion of the affected vessel (vessel diameter >2.6 cm for the main portal vein; >2.3 cm for the right portal vein; >2.2 cm for the left portal vein, together with disproportionate vessel growth when compared to same-order portal vein branches in the same lobe without impacted); (III) an increase in CT attenuation values of at least 20 Hounsfield units (HU) in the dynamic arterial phase images compared to unenhanced images. If two or more of these criteria were met, the thrombus was classified as neoplastic; otherwise, it was considered bland. A thrombus was considered potentially neoplastic if its maximum transverse diameter increased by more than 30% within 3 months despite anticoagulant therapy, as compared to a prior MDCT examination. Conversely, small or stable thrombi were indicative of bland characteristics.

DECT imaging

All patients underwent dual-energy contrast-enhanced multiphase CT scanning (Somatom Force, Siemens Healthcare, Erlangen, Germany). The patient was positioned in a supine posture, with the scanning scope extending from the upper boundary of the xiphoid process to the lower pole of the kidney. The protocol included the acquisition of the non-contrast, arterial, and equilibrium phases using single-energy mode, whereas the portal venous phase images were obtained in dual-energy mode [scanning parameters: tube voltage 90 kilovolt peak (kVp)/150 kVp, automatic tube current modulation, pitch 0.6, rotation time 0.5 s/r, and slice thickness 5 mm].

Each patient was given a dose of 1–1.2 mL/kg of a non-ionic contrast medium (ioversol, 350 mg iodine/mL), administered via a high-pressure injector (Missouri-XD 2001; Ulrich Medical, Buchbrunnenweg, Germany) at a rate of 3.5 mL/s via an antecubital vein catheter. The arterial phase was initiated automatically 12 seconds after the supraceliac abdominal aorta reached a trigger threshold of 100 HU, using a bolus tracking technique. Images of the portal phase and equilibrium phase were acquired at 30 and 60 seconds after the completion of the arterial phase, respectively.

Data reconstruction and image analysis

Two radiologists, each with 3 years of abdominal imaging experience, assessed all PVTs while remaining unaware of patients’ clinical details. The portal venous phase dual-energy thin-slice images were imported into the Siemens syngo.via VB40 workstation, where the DECT applications were executed to assess the quantitative parameters of the DECT. The quantitative measurement of dual-energy parameters was performed using the default 50% iodine overlay images. The region of interest (ROI) for PVT was delineated on the largest cross-sectional area of the lesion, ensuring that areas of necrosis, calcification, blood vessels, or other elements that could introduce artifacts were excluded. The ROI was placed near the center of the lesion, encompassing at least two-thirds of its total area. For normalization, additional ROIs were replaced on the aorta and portal vein in each slice.

The average IC (mg/mL) was measured in the lesion, aorta, and normal portal vein using the “Liver-Virtual Non-Contrast (Liver-VNC)” mode. The NIC was calculated to minimize the effects of physiological variations among individuals with NIC-A defined as the ratio of IC in the lesion to that in the aorta (NIC-A = IC_lesion/IC_aorta) and NIC-V as the ratio of IC in the lesion to that in the portal vein (NIC-V = IC_lesion/IC_portal vein). The electron density (Rho), Z, and DEI values for lesion were derived using the “Rho/Z” mode. Another quantitative parameter utilized in the dual-energy analysis is the K, which was calculated using the formula [K = (CT40 KeV − CT100 keV)/60], where CT represents the attenuation values at monochromatic energy levels of 40 and 100 keV, respectively. All measurements were performed three times, and the average values were used for further analysis.

Statistical analysis

Statistical analyses were performed using SPSS software version 27.0 (IBM Corp, Armonk, NY, USA) and GraphPad Prism version 9.5.0 for Windows (GraphPad Software, Boston, MA, USA, www.graphpad.com) and R (version 4.2.2; R Foundation for Statistical Computing, Vienna, Austria) software. All P values were calculated using two-sided tests, and a P value less than 0.05 was considered statistically significant. The Mann-Whitney U test was used to analyze non-normally distributed continuous variables, whereas the paired t-test was employed to compare normally distributed variables. Patients were randomly split into training and test datasets at a ratio of 7:3. Univariable logistic regression was used to find risk factors for neoplastic PVT prediction, the backward-stepwise multivariable logistic regression was then employed to identify independent predictors and construct the predictive nomogram in training dataset. The diagnostic performance was assessed using the receiver operating characteristic (ROC) curves. The optimal cutoff value was determined by selecting the point corresponding to the maximum Youden index (sensitivity + specificity − 1). The model’s calibration was assessed via calibration curves, whereas its clinical utility was evaluated using decision curve analysis (DCA). The interobserver consistency of dual-energy CT quantitative parameters was assessed using the intraclass correlation coefficient (ICC). An ICC <0.5 indicated poor interobserver agreement, between 0.5 and 0.75 indicated moderate agreement, between 0.75 and 0.9 indicated good agreement, and greater than 0.9 indicated excellent agreement.

Results

The study cohort included 99 cases of neoplastic and 74 cases of bland PVTs with a mean age of 55.3±11.7 years in the neoplastic group and 58.6±10.7 years in the bland group. Statistically significant differences were observed between neoplastic and bland PVTs in terms of sex, alpha-fetoprotein (AFP), total bilirubin, and aspartate aminotransferase (AST) levels (P<0.05; Table 1).

The analysis of PVT parameters demonstrated a high level of interobserver agreement, with the ICC exceeding 0.9 for the measurements of IC, NIC-A, NIC-V, Rho, Z, DEI, and K by two radiologists. The DECT parameters of neoplastic and bland PVTs are summarized in Table 2. The values of IC, NIC-A, and NIC-V were significantly higher in neoplastic PVTs compared to bland PVTs (P<0.001; Figure 2). Additionally, Rho, Z, DEI, and K were also significantly higher in neoplastic PVTs compared to bland PVTs (P<0.001; Figure 3).

Figure 2 Box plot showing the range of IC (A), NIC-A (B), and NIC-V (C) for neoplastic and bland PVT on the portal phases. The IC, NIC-A, and NIC-V were significantly higher for neoplastic than for bland PVT on the portal phases. IC, iodine concentration; NIC-A, normalized iodine concentration in the aorta; NIC-V, normalized iodine concentration in the portal vein; PVT, portal vein thrombosis.

Figure 3 Box plot showing the range of Rho (A), Z (B), DEI (C), and K (D) for neoplastic and bland PVT on the portal phases. The Rho, Z, DEI, and K were significantly higher for neoplastic than for bland PVT on the portal phases. DEI, dual-energy index; K, spectral slope; PVT, portal vein thrombosis; Rho, electron density; Z, effective atomic number.

In the ROC analysis (Table 3), the areas under the ROC curves (AUC) for differentiating neoplastic from bland PVT were 0.963, 0.970, and 0.969 for IC, NIC-A, and NIC-V, respectively; and 0.732, 0.952, 0.949, and 0.933 for Rho, Z, DEI, and K, respectively (Figure 4). When the cut-off value for IC was set at 0.950 mg/mL, the diagnostic performance achieved a sensitivity of 96.0% and a specificity of 97.0%. At cut-off values of 19.2% for NIC-A and 24.3% for NIC-V, the diagnostic performance demonstrated sensitivities of 93.2% and 97.3%, and specificities of 96.0% and 91.9%, respectively. When the cut-off values for Rho and Z were set at 37.9 e⁻/cm³ and 8.05 respectively, the diagnostic performance yielded sensitivities of 68.9% and 96.0% and specificities of 69.7% and 89.9%, respectively. At cut-off values of 0.011 for DEI and 1.10 for K, the diagnostic performance showed sensitivities of 96.0% and 90.5%, and specificities of 87.9% and 88.9%, respectively. Examples of both bland and neoplastic PVT are illustrated in Figures 5,6.

Figure 4 ROC curve analyses of DECT parameters for distinguishing neoplastic from bland PVT. DECT, dual-energy computed tomography; DEI, dual-energy index; IC, iodine concentration; K, spectral slope; NIC-A, normalized iodine concentration in the aorta; NIC-V, normalized iodine concentration in the portal vein; PVT, portal vein thrombosis; Rho, electron density; ROC, receiver operating characteristic; Z, effective atomic number.

Figure 5 A 57-year-old man with a bland PVT. (A) CT image at the portal phase shows a bland PVT in the main branch (arrow). (B) IC, NIC-A, and NIC-V of the lesion (arrow) was 0.40 mg/mL, 7.60%, 7.50% on the “Liver-Virtual Non-Contrast (Liver-VNC)” mode image. (C) Rho, Z, and DEI of the lesion (arrow) was 29.4 e⁻/cm³, 7.57, 0.001 on the “Rho/Z” mode image. (D) The calculated portal-phase K spectral curve value was 0.239. CT, computed tomography; DEI, dual-energy index; HU, Hounsfield units; IC, iodine concentration; K, spectral slope; NIC-A, normalized iodine concentration in the aorta; NIC-V, normalized iodine concentration in the portal vein; PVT, portal vein thrombosis; Rho, electron density; Z, effective atomic number.

Figure 6 A 42-year-old woman with a neoplastic PVT. (A) CT image at the portal phase shows a neoplastic PVT in the right portal vein (arrow) and a well-defined round mass in the right lobe of liver (circle). (B) IC, NIC-A, and NIC-V of the lesion (arrow) was 2.20 mg/mL, 32.7%, 27.4% on the “Liver-Virtual Non-Contrast (Liver-VNC)” mode image. (C) Rho, Z, and DEI of the lesion (arrow) was 37.1 e⁻/cm³, 8.77, 0.022 on the “Rho/Z” mode image. (D) The calculated portal-phase K spectral curve value was 2.567. CT, computed tomography; DEI, dual-energy index; HU, Hounsfield units; IC, iodine concentration; K, spectral slope; NIC-A, normalized iodine concentration in the aorta; NIC-V, normalized iodine concentration in the portal vein; PVT, portal vein thrombosis; Rho, electron density; Z, effective atomic number.

Univariate logistic regression analysis of DECT quantitative parameters revealed that IC, NIC-A, NIC-V, Rho, Z, DEI, and K were all risk factors for PVTT. Multivariate logistic regression analysis identified NIC-V, Rho, and DEI as independent risk factors for PVTT (Table 4). A diagnostic model based on these DECT quantitative parameters was constructed using the training cohort and subsequently validated in the validation cohort. The ROC curve, calibration curve, and DCA of the model are illustrated in Figure 7. In training and test datasets, the model’s AUC, sensitivity, specificity, and accuracy reached 0.994, 98.59%, 97.96%, and 98.33% and 0.940, 100%, 92.00%, and 96.23%, respectively.

Figure 7 Nomogram for predicting PVTT patients. A nomogram (A) was constructed using NIC-V, Rho, and DEI. In both the training (B) and test (C) sets, the nomogram demonstrated good calibration. DCA was performed for the training (D) and test (E) sets. The x-axis shows the threshold probability and the y-axis represents the net benefit. ROC curves were used to evaluate and contrast the diagnostic performance of NIC-V, Rho, DEI, and the nomogram in the two sets (F, training sets; G, test sets). The AUCs of the nomogram were 0.994 and 0.940 in the training and test sets respectively. AUC, area under the curve; DCA, decision curve analysis; DEI, dual-energy index; NIC-V, normalized iodine concentration in the portal vein; PVTT, portal vein tumor thrombosis; Rho, electron density; ROC, receiver operating characteristic.

Discussion

Bland PVT most commonly occurs in individuals with cirrhosis, but it may also be seen in patients who have undergone splenectomy, hepatectomy, or liver cancers. In the early stages, bland PVT may be asymptomatic; however, it can exacerbate or lead to the development of portal hypertension (24-26). Patients with HCC are more prone to neoplastic PVT, which is directly induced by tumor invasion. Consequently, there were statistically significant differences in AFP levels between neoplastic and bland PVTs (P<0.001). The incidence of HCC is significantly higher in male patients compared to female patients. Only 28 of the 173 patients in this study were female, which may partially explain the statistical difference observed between individuals with bland and neoplastic PVTs (P=0.003).

The DECT iodine quantification technique employs spectrum-based material decomposition to accurately calculate the IC within a lesion. This method is less affected by technical factors, enabling the detection of subtle differences in iodine uptake. Overall, it is more precise than conventional contrast-enhanced measurements (20,27-29). The primary pathophysiological mechanisms contributing to bland PVT include alterations in portal blood flow, hypercoagulability, endothelial damage, and the activation of inflammatory responses. Bland PVT predominantly consists of quiescent fibrin or blood clots, leading to reduced blood flow (30). The etiology of neoplastic PVT is primarily related to the presence of primary tumor cells in the liver which increase pressure within the blood sinuses surrounding the tumor, obstruct blood return through the portal vein, alter the portal vein blood flow, and promote the dissemination of cancer cells into the portal vein system (31,32). Neoplastic PVT, in similarity with tumor tissue, is primarily supplied by the hepatic artery, resulting in a rich blood supply. Consequently, the iodine density of PVTs is thought to be related to the thrombus’ vascular distribution. This relationship can potentially help to distinguish different thrombus characteristics. The observed difference in iodine density measurements between neoplastic and bland thrombus in this study may be attributed to this factor (P<0.0001).

Consistent with the findings conducted by Qian et al. (19) and Ascenti et al. (20), both IC and NIC exhibited significant diagnostic value, sensitivity, and specificity in distinguishing between bland and neoplastic thrombi. The optimal threshold for IC in identifying thrombosis, as suggested in this study (0.950 mg/mL) differs from that reported by Qian et al. (19) (1.14 mg/mL), but is comparable to the level recommended by Ascenti et al. (20) (0.9 mg/mL). However, the proposed threshold for the optimal NIC-A in this research (0.192) is consistent with the threshold suggested by Qian et al. (19) (0.17). This could be attributed to a limitation of the current DECT technology, which is its inability to eliminate inter-vendor differences in material decomposition techniques; studies on patients have demonstrated substantial variation in iodine content measurements (29). Standardizing vascular IC is a widely acknowledged method that may improve the consistency of organ-based iodine measurements (33). Our research found that the diagnostic value of standardized IC (AUC =0.969) is slightly superior to that of IC (AUC =0.963). Therefore, we advocate using standardized IC instead of IC to differentiate between bland and neoplastic PVTs.

DECT Rho/Z analysis demonstrates a high precision in identifying the components of substances. Its primary applications currently involve the identification of thoracic and abdominal tumors, optimization of radiation therapy, and gout (34,35). However, there have been no reports on its application in the identification of PVT. Rho, which represents the electron density per unit volume, is a direct correlated with the CT values. Z, the atomic number of an element, demonstrates increased sensitivity to materials with relatively higher atomic numbers (36,37). The DEI is determined by comparing the relative differences in attenuation values of materials at various photon energies (38). It serves as a reliable and efficient marker for distinguishing between different material kinds. The difference in Rho, Z, and DEI observed in this research may be partially attributed to the distinct compositional differences between bland and neoplastic PVT. Among them, the Z-value and DEI exhibit good diagnostic value, sensitivity, and specificity; however, the Rho-value demonstrates relatively limited diagnostic effectiveness (AUC =0.732), indicating that further research is needed to establish its clinical applicability.

The CT energy spectrum curve shows the variation in tissue CT values at the single energy level and the attenuation of X-rays by different materials (39). The energy spectrum of a material is determined by its chemical composition, and X-rays of the same energy have different absorption coefficients when penetrating various tissues (40). This property allows the characteristics of the energy spectrum curve to be expressed mathematically, with the slope of the curve being used for the quantitative assessment of various compounds. Our investigation revealed a statistically significant difference in the K-value between the bland and neoplastic PVTs (P<0.0001). The K-value was found to be an effective marker for distinguishing bland and neoplastic PVTs, with an AUC of 0.933, sensitivities of 90.5%, and specificities of 88.9%.

This study developed a model based on DECT quantitative parameters to predict the probability of PVTT. The model demonstrated strong predictive potential, with AUC values of 0.994 and 0.940 in the training and testing cohorts, respectively. The most significant predictive features for tumor thrombosis included NIC-V, Rho, and DEI during the portal venous phase. Compared to individual quantitative parameters, the model exhibited superior diagnostic performance. For instance, although Rho alone showed limited diagnostic efficacy in differentiating portal vein thrombus, its integration into the model enhanced its auxiliary diagnostic value, providing a more comprehensive and reliable strategy for accurate characterization of portal vein thrombus. When making clinical diagnoses, instead of simply relying on individual parameters, we strongly recommend using this model. By incorporating multiple key DECT quantitative parameters, the model significantly improved the accuracy of tumor thrombosis prediction, offering valuable decision-making support for clinical practice. It can offer more reliable evidence for doctors to distinguish between neoplastic and bland PVT, enabling them to formulate more appropriate treatment plans. A visual nomogram further improved clinical applicability, offering a robust tool for accurate thrombus characterization and decision-making support.

Furthermore, although the reference standard in this study is based on conventional MDCT, which necessitates two CT scans within 1–3 months, this approach is time-consuming, labor-intensive, and increases radiation exposure risks. In contrast, DECT enables comprehensive evaluation with a single examination, significantly improving diagnostic efficiency while minimizing radiation exposure. This advancement optimizes the diagnostic workflow, enhancing both patient safety and comfort. Consequently, DECT demonstrates substantial clinical value in differentiating PVT from tumor thrombosis, offering critical insights for related clinical practices.

Additionally, MRI special sequences, such as diffusion-weighted imaging (DWI) and susceptibility-weighted imaging (SWI), exhibit distinct capabilities in discerning between neoplastic and bland PVT. DWI reflects water molecule diffusion: Catalano et al. (22) found that the apparent diffusion coefficient (ADC) values of PVTT were similar to those of HCC, whereas the ADC values of bland PVT were higher. SWI is sensitive to magnetic-field-distorting substances: Huang et al. (41) found that SWI can effectively distinguish bland PVT from neoplastic PVT. Bland PVT shows low signals, whereas neoplastic PVT shows high signals, with diagnostic performance equal to or better than DWI. However, MRI sequences are affected by patient movement and metal implants, and their interpretation is complex. DECT, used in this study, has certain equipment-related issues and radiation exposure problems. In clinical practice, considering both modalities based on patient conditions can enhance the diagnostic accuracy for differentiating these thromboses.

There are some limitations in this study. First, since our study was conducted at a single center and had a retrospective design, external validation data are not available. Second, our research included individuals with PVT of various etiologies. Although we conducted a broad comparison between bland and neoplastic PVTs, we did not perform any further etiology subgroup analysis based on specific etiologies. Finally, unlike previous studies that focused on the advanced stage of hepatic artery involvement, our study specifically analyzed the stage of the portal vein. Additional research is required to assess if the diagnostic value of examining the portal vein phase surpasses or falls short of that of the late hepatic artery phase.

Conclusions

The findings of this study suggest that employing dual-energy MDCT with iodine quantification, electron density, effective atomic number, spectral slope, and dual energy index is a precise and reliable method for differentiating between bland and neoplastic PVT without the need for invasive procedures.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STARD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2516/rc

Funding: This study was supported by Science and Technology Program of Hunan Province (No. 2021SK50927).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2516/coif). All authors report that this study was supported by the Science and Technology Program of Hunan Province (No. 2021SK50927). The authors have no other 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. This retrospective study was approved by the Ethics Committee of The First Affiliated Hospital of Hunan Normal University Hunan Provincial People’s Hospital (No. [2024]-196), and the requirement for the provision of informed consent by patients was waived by the Ethics Committee.

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. de Franchis R, Bosch J, Garcia-Tsao G, Reiberger T, Ripoll C, Baveno V. II Faculty. Baveno VII - Renewing consensus in portal hypertension. J Hepatol 2022;76:959-74. [Crossref] [PubMed]
2. Odriozola A, Puente Á, Cuadrado A, Rivas C, Anton Á, González FJ, Pellón R, Fábrega E, Crespo J, Fortea JI. Portal Vein Thrombosis in the Setting of Cirrhosis: A Comprehensive Review. 2022;11:6435.
3. Prakash S, Bies J, Hassan M, Mares A, Didia SC. Portal vein thrombosis in cirrhosis: A literature review. Front Med (Lausanne) 2023;10:1134801. [Crossref] [PubMed]
4. Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2022. CA Cancer J Clin 2022;72:7-33. [Crossref] [PubMed]
5. Sun J, Yang L, Shi J, Liu C, Zhang X, Chai Z, Lau WY, Meng Y, Cheng SQ. Postoperative adjuvant IMRT for patients with HCC and portal vein tumor thrombus: An open-label randomized controlled trial. Radiother Oncol 2019;140:20-5. [Crossref] [PubMed]
6. Department of Medical Administration National Health Commission of the People's Republic of China. Guidelines for the diagnosis and treatment of primary liver cancer (2024 edition). China Journal of General Surgery 2024;33:475-530.
7. Llovet JM, Kelley RK, Villanueva A, Singal AG, Pikarsky E, Roayaie S, Lencioni R, Koike K, Zucman-Rossi J, Finn RS. Hepatocellular carcinoma. Nat Rev Dis Primers 2021;7:6. [Crossref] [PubMed]
8. Chen J, Zhu J, Zhang C, Song Y, Huang P. Contrast-enhanced ultrasound for the characterization of portal vein thrombosis vs tumor-in-vein in HCC patients: a systematic review and meta-analysis. Eur Radiol 2020;30:2871-80. [Crossref] [PubMed]
9. Wu B, Zhang Y, Tan H, Shi H. Value of (18)F-FDG PET/CT in the diagnosis of portal vein tumor thrombus in patients with hepatocellular carcinoma. Abdom Radiol (NY) 2019;44:2430-5. [Crossref] [PubMed]
10. 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]
11. Kim JH, Lee JM, Yoon JH, Lee DH, Lee KB, Han JK, Choi BI. Portal Vein Thrombosis in Patients with Hepatocellular Carcinoma: Diagnostic Accuracy of Gadoxetic Acid-enhanced MR Imaging. Radiology 2016;279:773-83. [Crossref] [PubMed]
12. Bae JS, Lee JM, Yoon JH, Jang S, Chung JW, Lee KB, Yi NJ, Lee JH. How to Best Detect Portal Vein Tumor Thrombosis in Patients with Hepatocellular Carcinoma Meeting the Milan Criteria: Gadoxetic Acid-Enhanced MRI versus Contrast-Enhanced CT. Liver Cancer 2020;9:293-307. [Crossref] [PubMed]
13. Forghani R, De Man B, Gupta R. Dual-Energy Computed Tomography: Physical Principles, Approaches to Scanning, Usage, and Implementation: Part 1. Neuroimaging Clin N Am 2017;27:371-84. [Crossref] [PubMed]
14. Martin SS, Kolaneci J, Czwikla R, Booz C, Gruenewald LD, Albrecht MH, Thompson ZM, Lenga L, Yel I, Vogl TJ, Wichmann JL, Koch V. Dual-Energy CT for the Detection of Portal Vein Thrombosis: Improved Diagnostic Performance Using Virtual Monoenergetic Reconstructions. Diagnostics (Basel) 2022;12:1682. [Crossref] [PubMed]
15. Hamid S, Nasir MU, So A, Andrews G, Nicolaou S, Qamar SR. Clinical Applications of Dual-Energy CT. Korean J Radiol 2021;22:970-82. [Crossref] [PubMed]
16. Tsurumaru D, Nishimuta Y, Kai S, Oki E, Minoda Y, Ishigami K. Clinical significance of dual-energy dual-layer CT parameters in differentiating small-sized gastrointestinal stromal tumors from leiomyomas. Jpn J Radiol 2023;41:1389-96. [Crossref] [PubMed]
17. Nagayama Y, Inoue T, Oda S, Tanoue S, Nakaura T, Ikeda O, Yamashita Y. Adrenal Adenomas versus Metastases: Diagnostic Performance of Dual-Energy Spectral CT Virtual Noncontrast Imaging and Iodine Maps. Radiology 2020;296:324-32. [Crossref] [PubMed]
18. Li Z, Chen Y, Zhang Y, Shi J, Wan Y. Quantitative energy spectrum CT in differential diagnosis of aldosterone-producing adenoma and cortisol-producing adenoma. Quant Imaging Med Surg 2023;13:5012-21. [Crossref] [PubMed]
19. 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]
20. 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]
21. Shah ZK, McKernan MG, Hahn PF, Sahani DV. Enhancing and expansile portal vein thrombosis: value in the diagnosis of hepatocellular carcinoma in patients with multiple hepatic lesions. AJR Am J Roentgenol 2007;188:1320-3. [Crossref] [PubMed]
22. Catalano OA, Choy G, Zhu A, Hahn PF, Sahani DV. Differentiation of malignant thrombus from bland thrombus of the portal vein in patients with hepatocellular carcinoma: application of diffusion-weighted MR imaging. Radiology 2010;254:154-62. [Crossref] [PubMed]
23. Sandrasegaran K, Tahir B, Nutakki K, Akisik FM, Bodanapally U, Tann M, Chalasani N. Usefulness of conventional MRI sequences and diffusion-weighted imaging in differentiating malignant from benign portal vein thrombus in cirrhotic patients. AJR Am J Roentgenol 2013;201:1211-9. [Crossref] [PubMed]
24. Kuboki S, Shimizu H, Ohtsuka M, Kato A, Yoshitomi H, Furukawa K, Takayashiki T, Takano S, Okamura D, Suzuki D, Sakai N, Kagawa S, Miyazaki M. Incidence, risk factors, and management options for portal vein thrombosis after hepatectomy: a 14-year, single-center experience. Am J Surg 2015;210:878-85.e2. [Crossref] [PubMed]
25. Kumar A, Sharma P, Arora A. Review article: portal vein obstruction--epidemiology, pathogenesis, natural history, prognosis and treatment. Aliment Pharmacol Ther 2015;41:276-92. [Crossref] [PubMed]
26. Swinson B, Waters PS, Webber L, Nathanson L, Cavallucci DJ, O'Rourke N, Bryant RD. Portal vein thrombosis following elective laparoscopic splenectomy: incidence and analysis of risk factors. Surg Endosc 2022;36:3332-9. [Crossref] [PubMed]
27. Chandarana H, Megibow AJ, Cohen BA, Srinivasan R, Kim D, Leidecker C, Macari M. Iodine quantification with dual-energy CT: phantom study and preliminary experience with renal masses. AJR Am J Roentgenol 2011;196:W693-700. [Crossref] [PubMed]
28. Ascenti G, Mileto A, Krauss B, Gaeta M, Blandino A, Scribano E, Settineri N, Mazziotti S. Distinguishing enhancing from nonenhancing renal masses with dual-source dual-energy CT: iodine quantification versus standard enhancement measurements. Eur Radiol 2013;23:2288-95. [Crossref] [PubMed]
29. Toia GV, Mileto A, Wang CL, Sahani DV. Quantitative dual-energy CT techniques in the abdomen. Abdom Radiol (NY) 2022;47:3003-18. [Crossref] [PubMed]
30. Turon F, Driever EG, Baiges A, Cerda E, García-Criado Á, Gilabert R, et al. Predicting portal thrombosis in cirrhosis: A prospective study of clinical, ultrasonographic and hemostatic factors. J Hepatol 2021;75:1367-76. [Crossref] [PubMed]
31. Wang JC, Xia AL, Xu Y, Lu XJ. Comprehensive treatments for hepatocellular carcinoma with portal vein tumor thrombosis. J Cell Physiol 2019;234:1062-70. [Crossref] [PubMed]
32. Zhang M, Ding Q, Bian C, Su J, Xin Y, Jiang X. Progress on the molecular mechanism of portal vein tumor thrombosis formation in hepatocellular carcinoma. Exp Cell Res 2023;426:113563. [Crossref] [PubMed]
33. Lennartz S, Parakh A, Cao J, Zopfs D, Große Hokamp N, Kambadakone A. Inter-scan and inter-scanner variation of quantitative dual-energy CT: evaluation with three different scanner types. Eur Radiol 2021;31:4438-51. [Crossref] [PubMed]
34. Goo HW, Goo JM, Dual-Energy CT. New Horizon in Medical Imaging. Korean J Radiol 2017;18:555-69. [Crossref] [PubMed]
35. Sakata D, Haga A, Kida S, Imae T, Takenaka S, Nakagawa K. Effective atomic number estimation using kV-MV dual-energy source in LINAC. Phys Med 2017;39:9-15. [Crossref] [PubMed]
36. Hua CH, Shapira N, Merchant TE, Klahr P, Yagil Y. Accuracy of electron density, effective atomic number, and iodine concentration determination with a dual-layer dual-energy computed tomography system. Med Phys 2018;45:2486-97. [Crossref] [PubMed]
37. Garcia LI, Azorin JF, Almansa JF. A new method to measure electron density and effective atomic number using dual-energy CT images. Phys Med Biol 2016;61:265-79. [Crossref] [PubMed]
38. Xu C, Kong L, Deng X. Dual-Energy Computed Tomography For Differentiation Between Osteoblastic Metastases and Bone Islands. Front Oncol 2022;12:815955. [Crossref] [PubMed]
39. Leng S, Yu L, Wang J, Fletcher JG, Mistretta CA, McCollough CH. Noise reduction in spectral CT: reducing dose and breaking the trade-off between image noise and energy bin selection. Med Phys 2011;38:4946-57. [Crossref] [PubMed]
40. Tarhan NC, Hatipoğlu T, Ercan E, Bener M, Keleş G, Başaran C, Bilezikçi B. Correlation of dynamic multidetector CT findings with pathological grades of hepatocellular carcinoma. Diagn Interv Radiol 2011;17:328-33. [PubMed]
41. Huang C, Xiao X, Guo M, Hu X, Liu C, Wang J, Zhang H, Li X, Cai P. Value of susceptibility-weighted imaging in differentiating benign from malignant portal vein thrombosis. Quant Imaging Med Surg 2023;13:2688-96. [Crossref] [PubMed]