Multimodality 3D Tumor Segmentation in HCC Patients Treated with TACE

Deep Learning Combined with Radiologist's Intervention Achieves Accurate Segmentation of Hepatocellular Carcinoma in Dual-Phase Magnetic Resonance Images

Purpose

Segmentation of hepatocellular carcinoma (HCC) is crucial; however, manual segmentation is subjective and time-consuming. Accurate and automatic lesion contouring for HCC is desirable in clinical practice. In response to this need, our study introduced a segmentation approach for HCC combining deep convolutional neural networks (DCNNs) and radiologist intervention in magnetic resonance imaging (MRI). We sought to design a segmentation method with a deep learning method that automatically segments using manual location information for moderately experienced radiologists. In addition, we verified the viability of this method to assist radiologists in accurate and fast lesion segmentation.

Method

In our study, we developed a semiautomatic approach for segmenting HCC using DCNN in conjunction with radiologist intervention in dual-phase gadolinium-ethoxybenzyl-diethylenetriamine penta-acetic acid- (Gd-EOB-DTPA-) enhanced MRI. We developed a DCNN and deep fusion network (DFN) trained on full-size images, namely, DCNN-F and DFN-F. Furthermore, DFN was applied to the image blocks containing tumor lesions that were roughly contoured by a radiologist with 10 years of experience in abdominal MRI, and this method was named DFN-R. Another radiologist with five years of experience (moderate experience) performed tumor lesion contouring for comparison with our proposed methods. The ground truth image was contoured by an experienced radiologist and reviewed by an independent experienced radiologist.

Results

The mean DSC of DCNN-F, DFN-F, and DFN-R was 0.69 ± 0.20 (median, 0.72), 0.74 ± 0.21 (median, 0.77), and 0.83 ± 0.13 (median, 0.88), respectively. The mean DSC of the segmentation by the radiologist with moderate experience was 0.79 ± 0.11 (median, 0.83), which was lower than the performance of DFN-R.

Conclusions

Deep learning using dual-phase MRI shows great potential for HCC lesion segmentation. The radiologist-aided semiautomated method (DFN-R) achieved improved performance compared to manual contouring by the radiologist with moderate experience, although the difference was not statistically significant.

A Nomogram Based on Preoperative Lipiodol Deposition after Sequential Retreatment with Transarterial Chemoembolization to Predict Prognoses for Intermediate-Stage Hepatocellular Carcinoma

(1) Background: Conventional transarterial chemoembolization (cTACE) is the mainstay treatment for patients with Barcelona Clinic Liver Cancer (BCLC) B-stage hepatocellular carcinoma (HCC). However, BCLC B-stage patients treated with cTACE represent a prognostically heterogeneous population. We aim to develop and validate a lipiodol-deposition-based nomogram for predicting the long-term survival of BCLC B-stage HCC patients after sequential cTACE. (2) Methods: In this retrospective study, 229 intermediate-stage HCC patients from two hospitals were separately allocated to a training cohort (n = 142) and a validation cohort (n = 87); these patients underwent repeated TACE (≥4 TACE sessions) between May 2010 and May 2017. Lipiodol deposition was assessed by semiautomatic volumetric measurement with multidetector computed tomography (MDCT) before cTACE and was characterized by two ordinal levels: ≤50% (low) and >50% (high). A clinical lipiodol deposition nomogram was constructed based on independent risk factors identified by univariate and multivariate Cox regression analyses, and the optimal cutoff points were obtained. Prediction models were assessed by time-dependent receiver-operating characteristic curves, calibration curves, and decision curve analysis. (3) Results: The median number of TACE sessions was five (range, 4−7) in both cohorts. Before the TACE-3 sessions, the newly constructed nomogram based on lipiodol deposition achieved desirable diagnostic performance in the training and validation cohorts with AUCs of 0.72 (95% CI, 0.69−0.74) and 0.71 (95% CI, 0.68−0.73), respectively, and demonstrated higher predictive ability compared with previously published prognostic models (all p < 0.05). The prognostic nomogram obtained good clinical usefulness in predicting the patient outcomes after TACE. (4) Conclusions: Based on each pre-TACE lipiodol deposition, two sessions are recommended before abandoning cTACE or combining treatment for patients with intermediate-stage HCC. Furthermore, the nomogram based on pre-TACE-3 lipiodol deposition can be used to predict the prognoses of patients with BCLC B-stage HCC.

Using Routinely Collected Health Records to Identify the Fine-Resolution Spatial Patterns of Soil-Transmitted Helminth Infections in Rwanda

Background. Soil-transmitted helminths (STH) are parasitic diseases with significant public health impact. Analysis is generally based on cross-sectional prevalence surveys; outcomes are mostly aggregated to larger spatial units. However, recent research demonstrates that infection levels and spatial patterns differ between STH species and tend to be localized. Methods. Incidence data of STHs including roundworm (Ascaris lumbricoides), whipworm (Trichuris trichiura) and hookworms per primary health facility for 2008 were linked to spatially delineated primary health center service areas. Prevalence data per district for individual and combined STH infections from the 2008 nationwide survey in Rwanda were also obtained. Results. A comparison of reported prevalence and incidence data indicated significant positive correlations for roundworm (R2 = 0.63) and hookworm (R2 = 0.27). Weak positive correlations were observed for whipworm (R2 = 0.02) and the three STHs combined (R2 = 0.10). Incidence of roundworm and whipworm were found to be focalized with significant spatial autocorrelation (Moran’s I > 0: 0.05−0.38 and p ≤ 0.03), with (very) high incidence rates in some focal areas. In contrast, hookworm incidence is ubiquitous and randomly distributed (Moran’s I > 0: 0.006 and p = 0.74) with very low incidence rates. Furthermore, an exploratory regression analysis identified relationships between helminth infection cases and potential environmental and socio-economic risk factors. Conclusions. Findings show that the spatial distribution of STH incidence is significantly associated with soil properties (sand proportion and pH), rainfall, wetlands and their uses, population density and proportion of rural residents. Identified spatial patterns are important for guiding STH prevention and control programs.

Role of three-dimensional printing and artificial intelligence in the management of hepatocellular carcinoma: Challenges and opportunities

Hepatocellular carcinoma (HCC) constitutes the fifth most frequent malignancy worldwide and the third most frequent cause of cancer-related deaths. Currently, treatment selection is based on the stage of the disease. Emerging fields such as three-dimensional (3D) printing, 3D bioprinting, artificial intelligence (AI), and machine learning (ML) could lead to evidence-based, individualized management of HCC. In this review, we comprehensively report the current applications of 3D printing, 3D bioprinting, and AI/ML-based models in HCC management; we outline the significant challenges to the broad use of these novel technologies in the clinical setting with the goal of identifying means to overcome them, and finally, we discuss the opportunities that arise from these applications. Notably, regarding 3D printing and bioprinting-related challenges, we elaborate on cost and cost-effectiveness, cell sourcing, cell viability, safety, accessibility, regulation, and legal and ethical concerns. Similarly, regarding AI/ML-related challenges, we elaborate on intellectual property, liability, intrinsic biases, data protection, cybersecurity, ethical challenges, and transparency. Our findings show that AI and 3D printing applications in HCC management and healthcare, in general, are steadily expanding; thus, these technologies will be integrated into the clinical setting sooner or later. Therefore, we believe that physicians need to become familiar with these technologies and prepare to engage with them constructively.

Early prediction of 1-year tumor response of hepatocellular carcinoma with lipiodol deposition pattern through post-embolization cone-beam computed tomography during conventional transarterial chemoembolization

OBJECTIVE

To evaluate whether parenchyma-to-lipiodol ratio (PLR) and lesion-to-lipiodol ratio (LLR) on C-arm cone-beam computed tomography (CBCT) can predict 1-year tumor response in patients with hepatocellular carcinoma (HCC) treated with conventional transcatheter arterial chemoembolization (cTACE).

METHODS

This retrospective analysis included 221 HCC target lesions within up-to-seven criteria in 80 patients who underwent cTACE with arterial-phase CBCT and unenhanced CBCT after cTACE from 2015 to 2018. PLR and LLR of every tumor slice were obtained through mean density division of liver parenchyma and tumor enhancement with intratumoral lipiodol deposition. The cutoff values (COVs) of maximal PLR and LLR of every tumor were analyzed using Youden's index. The reliability of COV, correlations between the related parameters, and 1-year progression were assessed through interobserver agreement and multivariate analysis. COV validity was verified using the chi-square test and Cramer's V coefficient (V) in the validation cohort.

RESULTS

Standard COVs of PLR and LLR were 0.149 and 1.4872, respectively. Interobserver agreement of COV for PLR and LLR was near perfect (kappa > 0.9). Multivariate analysis suggested that COV of PLR is an independent predictor (odds ratio = 1.23532×10¹⁴, p = 4.37×10^-7). COV of PLR showed strong consistency, correlation with 1-year progression in prediction model (V = 0.829-0.776; p < 0.0001), and presented as an effective predictor in the validation cohort (V = 0.766; p < 0.0001).

CONCLUSION

The COV of PLR (0.149) assessed through immediate post-embolization CBCT is an objective, effective, and approachable predictor of 1-year HCC progression after cTACE.

KEY POINTS

• The maximal PLR value indicates the least lipiodol-distributed region in an HCC tumor. The maximal LLR value indicates the least lipiodol-deposited region in the tumor due to incomplete lipiodol delivery. PLR and LLR are concepts like signal-to-noise ratio to characterize the lipiodol retention pattern objectively to predict 1-year tumor progression immediately without any quantification software for 3D image analysis immediately after cTACE treatment. • COV of PLR can facilitate the early prediction of tumor progression/recurrence and indicate the section of embolized HCC, providing the operator's good targets for sequential cTACE or combined ablation. • The validation cohort in our study verified standard COVs of PLR and LLR. The validation process was more convincing and delicate than that of previous retrospective studies.

Magnetically responsive, sorafenib loaded alginate microspheres for hepatocellular carcinoma treatment

This study aimed to develop sorafenib loaded magnetic microspheres for the treatment of hepatocellular carcinoma. To achieve this goal, superparamagnetic iron oxide nanoparticles (SPIONs) were synthesised and encapsulated in alginate microspheres together with an antineoplastic agent, sorafenib. In the study, firstly SPIONs were synthesised and characterised by dynamic light scattering, energy-dispersive X-ray spectroscopy, and scanning electron microscopy. Then, alginate-SPIONs microspheres were developed, and further characterised by electron spin resonance spectrometer and vibrating sample magnetometer. Besides the magnetic properties of SPIONs, alginate microspheres with SPIONs were also found to have magnetic properties. The potential use of microspheres in hyperthermia treatment was then investigated and an increase of about 4°C in the environment was found out. Drug release studies and cytotoxicity tests were performed after sorafenib was encapsulated into the magnetic microspheres. According to release studies, sorafenib has been released from microspheres for 8 h. Cytotoxicity tests showed that alginate-SPION-sorafenib microspheres were highly effective against cancerous cells and promising for cancer therapy.

Integrating baseline MR imaging biomarkers into BCLC and CLIP improves overall survival prediction of patients with hepatocellular carcinoma (HCC)

OBJECTIVES

We aimed to evaluate the independent predictive role of baseline imaging biomarkers for overall survival (OS) and transplant-free survival (TFS) in patients with HCC and assess the incremental value of these biomarkers to current staging systems.

METHODS

In this retrospective IRB approved study, the clinical, laboratory, and imaging parameters of 304 HCC patients were collected. Cox regression model was utilized to identify the potential predictors of survival. Recursive partitioning test was utilized to identify the optimal ADC cutoff for stratifying patients' OS. Patients were stratified based on Barcelona Clinic Liver Cancer (BCLC) and Cancer of the Liver Italian Program (CLIP). Binary ADC value (above vs. below the cutoff) and tumor margin (well- vs. ill-defined) were integrated into BCLC and CLIP. OS and TFS was compared for patients based on standard criteria with and without imaging biomarkers.

RESULTS

At baseline, patients with low tumor ADC and well-defined tumor margin (favorable imaging biomarkers) had longer survival, as compared to those with high ADC and ill-defined tumor margin (unfavorable imaging biomarkers) (median OS of 43 months vs. 7 months, respectively) (p < 0.001). Tumor ADC and tumor margin remained strong independent predictors of survival after adjustment for demographics, BCLC and CLIP staging, and tumor burden. Incorporating ADC and tumor margin improved performance of OS prediction by 9% in BCLC group and 6% in CLIP group.

CONCLUSION

Incorporating ADC and tumor margin to current staging systems for HCC significantly improve prediction of OS and TFS of these criteria.

KEY POINTS

• ADC and tumor margin are predictors of overall survival in HCC patients, independent of clinical, laboratory, and other imaging variables. • Adding ADC and tumor margin improved the prognostic value of BCLC and CLIP criteria by 9% and 6%, respectively. • High ADC and ill-defined tumor margin at baseline predicted poor survival, regardless of patient's liver function and general health status.

Consensus recommendations of three-dimensional visualization for diagnosis and management of liver diseases

Three-dimensional (3D) visualization involves feature extraction and 3D reconstruction of CT images using a computer processing technology. It is a tool for displaying, describing, and interpreting 3D anatomy and morphological features of organs, thus providing intuitive, stereoscopic, and accurate methods for clinical decision-making. It has played an increasingly significant role in the diagnosis and management of liver diseases. Over the last decade, it has been proven safe and effective to use 3D simulation software for pre-hepatectomy assessment, virtual hepatectomy, and measurement of liver volumes in blood flow areas of the portal vein; meanwhile, the use of 3D models in combination with hydrodynamic analysis has become a novel non-invasive method for diagnosis and detection of portal hypertension. We herein describe the progress of research on 3D visualization, its workflow, current situation, challenges, opportunities, and its capacity to improve clinical decision-making, emphasizing its utility for patients with liver diseases. Current advances in modern imaging technologies have promised a further increase in diagnostic efficacy of liver diseases. For example, complex internal anatomy of the liver and detailed morphological features of liver lesions can be reflected from CT-based 3D models. A meta-analysis reported that the application of 3D visualization technology in the diagnosis and management of primary hepatocellular carcinoma has significant or extremely significant differences over the control group in terms of intraoperative blood loss, postoperative complications, recovery of postoperative liver function, operation time, hospitalization time, and tumor recurrence on short-term follow-up. However, the acquisition of high-quality CT images and the use of these images for 3D visualization processing lack a unified standard, quality control system, and homogeneity, which might hinder the evaluation of application efficacy in different clinical centers, causing enormous inconvenience to clinical practice and scientific research. Therefore, rigorous operating guidelines and quality control systems need to be established for 3D visualization of liver to develop it to become a mature technology. Herein, we provide recommendations for the research on diagnosis and management of 3D visualization in liver diseases to meet this urgent need in this research field.

Paraplegia after transcatheter artery chemoembolization in a child with clear cell sarcoma of the kidney: A case report

BACKGROUND

Transcatheter arterial chemoembolization (TACE) is a common treatment for inoperable malignant renal tumors. However, a series of complications may follow the TACE treatment. Spinal cord injury caused by the embolization of intercostal or lumbar arteries is extremely rare.

CASE SUMMARY

We describe a case with quite uncommon spinal cord injury after TACE in a 3-year-old child with clear cell sarcoma of the kidney. Sensory impairment beneath the T10 dermatomes and paraplegia on the day after TACE were found in this patient. Unfortunately, sustained paraplegia still existed for more than 2 mo after TACE despite the large dose of steroids and supportive therapy.

CONCLUSION

We should draw attention to an uncommon complication of paraplegia after TACE treatment in malignant renal tumors. Although it is rare, the result is disastrous.

Which Criteria Applied in Multi-Phasic CT Can Predict Early Tumor Response in Patients with Hepatocellular Carcinoma Treated Using Conventional TACE: RECIST, mRECIST, EASL or qEASL?

PURPOSE

Our study aimed to evaluate quantitative tumor response assessment (quantitative EASL-[qEASL]) on computed tomography (CT) images in patients with hepatocellular carcinoma (HCC) treated using conventional transarterial chemoembolization (cTACE), compared to existing 1-dimensional and 2-dimensional methods (RECIST, mRECIST, EASL).

MATERIALS AND METHODS

In this IRB-approved, single-institution retrospective cohort study, 52 consecutive patients with intermediate-stage HCC were consecutively included. All patients underwent contrast-enhanced CT scan at baseline and 4 weeks after cTACE.

RESULTS

Median follow-up period was 13.5 months (range 1.2-54.1). RECIST, mRECIST and EASL identified progression in 2 (4%), 1 (2%) and 1 (2%) patients, respectively, whereas qEASL identified 10 (19%) patients. qEASL was the only tumor response method able to predict survival among different tumor response groups (P < 0.05), whereas RECIST, mRECIST and EASL did not (P > 0.05). Both EASL and qEASL were able to identify responders and non-responders and were predictive of survival (P < 0.05). Multivariate analysis showed that progression was an independent predictor of overall survival with hazard ratio of 1.9 (P = 0.025). Patients who demonstrated progression with qEASL had significantly shorter survival than those with non-progression (7.6 vs. 20.4 months, P = 0.012). Similar multivariate analysis using RECIST, mRECIST and EASL could not be performed because too few patients were categorized as progressive disease.

CONCLUSION

qEASL could be applied on CT images to assess tumor response following cTACE and is a more sensitive biomarker to predict survival and identify tumor progression than RECIST, mRECIST and EASL at an early time point.

LEVEL OF EVIDENCE

Level 2a, retrospective cohort study.

Schistosomiasis mansoni incidence data in Rwanda can improve prevalence assessments, by providing high-resolution hotspot and risk factors identification

Background

Schistosomiasis mansoni constitutes a significant public health problem in Rwanda. The nationwide prevalence mapping conducted in 2007–2008 revealed that prevalence per district ranges from 0 to 69.5% among school children. In response, mass drug administration campaigns were initiated. However, a few years later some additional small-scale studies revealed the existence of areas of high transmission in districts formerly classified as low endemic suggesting the need for a more accurate methodology for identification of hotspots. This study investigated if confirmed cases of schistosomiasis recorded at health facility level can be used to, next to existing prevalence data, detect geographically more accurate hotspots of the disease and its associated risk factors.

Methods

A GIS-based spatial and statistical analysis was carried out. Confirmed cases, recorded at primary health facilities level, were combined with demographic data to calculate incidence rates for each of 367 health facility service area. Empirical Bayesian smoothing was used to deal with rate instability. Incidence rates were compared with prevalence data to identify their level of agreement. Spatial autocorrelation of the incidence rates was analyzed using Moran’s Index, to check if spatial clustering occurs. Finally, the spatial relationship between schistosomiasis distribution and potential risk factors was assessed using multiple regression.

Results

Incidence rates for 2007–2008 were highly correlated with prevalence values (R² = 0.79), indicating that in the case of Rwanda incidence data can be used as a proxy for prevalence data. We observed a focal distribution of schistosomiasis with a significant spatial autocorrelation (Moran’s I > 0: 0,05–0.20 and p ≤ 0,05), indicating the occurrence of hotspots. Regarding risk factors, it was identified that the spatial pattern of schistosomiasis is significantly associated with wetland conditions and rice cultivation.

Conclusion

In Rwanda the high density of health facilities and the standardized microscopic laboratory diagnostic allow the derived data to be used to complement prevalence studies to identify hotspots of schistosomiasis and its associated risk factors. This type of information, in turn, can support disease control interventions and monitoring.

The impact of antiangiogenic therapy combined with Transarterial Chemoembolization on enhancement based quantitative tumor response assessment in patients with hepatocellular carcinoma

PURPOSE

To investigate whether bevacizumab compromises early response assessment after Transarterial Chemoembolization (TACE) in patients with hepatocellular carcinoma by 3D quantitative European Association for the Study of the Liver (qEASL) criteria in comparison to other imaging-based criteria.

MATERIALS AND METHODS

Each of 14 patients receiving TACE and bevacizumab was matched with two patients receiving TACE alone. Baseline and Follow-up MRI was retrospectively analyzed regarding qEASL and other imaging-based criteria.

RESULTS

Percentage-based qEASL achieved significant separation in both therapy arms (p=0.046 and p=0.015). Response and Overall Survival showed similar association among treatment groups (p=0.749).

CONCLUSIONS

Anti-angiogenic therapy with bevacizumab does not impede early response assessment by qEASL.

3D Quantitative tumour burden analysis in patients with hepatocellular carcinoma before TACE: comparing single-lesion vs. multi-lesion imaging biomarkers as predictors of patient survival

OBJECTIVES

To compare the ability of single- vs. multi-lesion assessment on baseline MRI using 1D- and 3D-based measurements to predict overall survival (OS) in patients with hepatocellular carcinoma (HCC) before transarterial chemoembolization (TACE).

METHODS

This retrospective analysis included 122 patients. A quantitative 3D analysis was performed on baseline MRI to calculate enhancing tumour volume (ETV [cm(3)]) and enhancing tumour burden (ETB [%]) (ratio between ETV [cm(3)] and liver volume). Furthermore, enhancing and overall tumour diameters were measured. Patients were stratified into two groups using thresholds derived from the BCLC staging system. Statistical analysis included Kaplan-Meier plots, uni- and multivariate cox proportional hazard ratios (HR) and concordances.

RESULTS

All methods achieved good separation of the survival curves (p < 0.05). Multivariate analysis showed an HR of 5.2 (95 % CI 3.1-8.8, p < 0.001) for ETV [cm(3)] and HR 6.6 (95 % CI 3.7-11.5, p < 0.001) for ETB [%] vs. HR 2.6 (95 % CI 1.2-5.6, p = 0.012) for overall diameter and HR 3.0 (95 % CI 1.5-6.3, p = 0.003) for enhancing diameter. Concordances were highest for ETB [%], with no added predictive power for multi-lesion assessment (difference between concordances not significant).

CONCLUSION

3D quantitative assessment is a stronger predictor of survival as compared to diameter-based measurements. Assessing multiple lesions provides no substantial improvement in predicting OS than evaluating the dominant lesion alone.

KEY POINTS

• 3D quantitative tumour assessment on baseline MRI predicts survival in HCC patients. • 3D quantitative tumour assessment predicts survival better than any current radiological method. • Multiple lesion assessment provides no improvement than evaluating the dominant lesion alone. • Measuring enhancing tumour volume in proportion to liver volume reflects tumour burden.

Hepatic imaging following intra-arterial embolotherapy

PURPOSE

To discuss guidelines and salient imaging findings of solid tumors treated with common intra-arterial procedures used in interventional oncology.

METHODS

A meticulous literature search of PubMed-indexed articles was conducted. Key words included "imaging + embolization," "imaging + TACE," "imaging + radioembolization," "imaging + Y90," "mRECIST," and "EASL." Representative post-treatment cross-sectional images were obtained from past cases in this institution.

RESULTS

Intra-arterial therapy (IAT) in interventional oncology includes bland embolization, chemoembolization, and radioembolization. Solid tumors of the liver are the primary focus of these procedures. Cross-sectional CT and/or MR are the main modalities used to image tumors after treatment. Traditional size-based response criteria (WHO and RECIST) alone are of limited utility in determining response to IAT; tumoral necrosis and enhancement must be considered. Specifically for HCC, the EASL and mRECIST guidelines are becoming widely adopted response criteria to assess these factors. DWI, FDG-PET, and CEUS are modalities that play an adjunctive but controversial role.

CONCLUSIONS

Radiologists must be aware that the different forms of intra-arterial therapy yield characteristic findings on cross-sectional imaging. Knowledge of these findings is integral to accurate assessment of tumor response and progression.

Assessment of the feasibility of TACE combined with intratumoral injection of cisplatin in hepatocellular carcinoma

The feasibility of transcatheter arterial chemo-embolization (TACE) combined with intratumoral injection of cisplatin as treatment for hepatocellular carcinoma. 30 cases receiving TACE were denoted the TACE group, another 30 cases receiving TACE combined with an intratumoral multi-point injection of cisplatin were denoted the TACE/cisplatin group. Cases with partial remission/complete remission (PR/CR) were analyzed using 2 tests; alpha fetoprotein (AFP), aspartate amino transferase (AST), total bilirubin (TBIL), erythrocyte, and platelet levels were detected and the differences between two groups were analyzed using the Student's t-test; cases with complications, including intrahepatic metastasis (IM), upper gastrointestinal bleeding (UGB), and liver failure were also counted. The correlation of clinical parameters with PR/CR was analyzed using multifactorial correlation analysis. Cases with PR/CR in the TACE/cisplatin group were significantly more than in TACE group, accompanied by significant declination in FAP. There were no significant differences of AST, ALT, TBIL, blood urea nitrogen (BUN), white blood cells (WBC), red blood cells (RBC), and platelets (PLT) between two groups; 3 cases with IM, one case with UGB and one case with LF were found in the TACE group, but only 1 case with IM was found in the TACE/cisplatin group. In addition, tumor stage was correlated with PR/CR. We concluded that TACE combined with intratumoral injection of cisplatin was more effective than TACE, and with fewer complications and side effects.