Microvascular Invasion Is a Better Predictor of Tumor Recurrence and Overall Survival Following Surgical Resection for Hepatocellular Carcinoma Compared to the Milan Criteria

Gene expression signature as a surrogate marker of microvascular invasion on routine hepatocellular carcinoma biopsies

BACKGROUND & AIMS

Microvascular invasion (MVI), a major risk factor for tumor recurrence after surgery in hepatocellular carcinoma (HCC), is only detectable by microscopic examination of the surgical specimen. We aimed to define a transcriptomic signature associated with MVI in HCC than can be applied to formalin-fixed paraffin-embedded (FFPE) biopsies for use in clinical practice.

METHODS

To identify a gene expression signature related to MVI by using NanoString technology, we selected a set of 200 genes according to the literature and RNA-sequencing data obtained from a cohort of 150 frozen HCC samples previously published. We used 178 FFPE-archived HCC samples, including 109 surgical samples for the training set and 69 paired pre-operative biopsies for the validation set. In 14 cases of the training set, a paired biopsy was available and was also analyzed.

RESULTS

We identified a 6-gene signature (ROS1, UGT2B7, FAS, ANGPTL7, GMNN, MKI67) strongly associated with MVI in the training set of FFPE surgical HCC samples, with 82% accuracy (sensitivity 82%, specificity 81%, AUC 0.82). The NanoString gene expression was highly correlated in 14 paired surgical/biopsy HCC samples (mean R: 0.97). In the validation set of 69 FFPE HCC biopsies, the 6-gene NanoString signature predicted MVI with 74% accuracy (sensitivity 73%, specificity 76%, AUC 0.74). Moreover, on multivariate analysis, the MVI signature was associated with overall survival in both sets (hazard ratio 2.29; 95% CI 1.03-5.07; p = 0.041).

CONCLUSION

We defined a 6-gene signature that can accurately predict MVI in FFPE HCC biopsy samples, which is also associated with overall survival, although its survival impact must be confirmed by extensive study with further clinical data.

LAY SUMMARY

Microvascular invasion, a major risk factor for tumor recurrence after surgery in hepatocellular carcinoma, is only detectable by microscopic examination of a surgical specimen. In this study, we defined a relevant surrogate signature of microvascular invasion in hepatocellular carcinoma that may be applied in clinical practice with routine tumor biopsy and integrated into the therapeutic strategy.

Preoperative Microvascular Invasion Prediction to Assist in Surgical Plan for Single Hepatocellular Carcinoma: Better Together with Radiomics

BACKGROUND

Prediction models with or without radiomic analysis for microvascular invasion (MVI) in hepatocellular carcinoma (HCC) have been reported, but the potential for model-predicted MVI in surgical planning is unclear. Therefore, we aimed to explore the effect of predicted MVI on early recurrence after anatomic resection (AR) and non-anatomic resection (NAR) to assist surgical strategies.

METHODS

Patients with a single HCC of 2-5 cm receiving curative resection were enrolled from 2 centers. Their data were used to develop (n = 230) and test (n = 219) two prediction models for MVI using clinical factors and preoperative computed tomography images. The two prediction models, clinico-radiologic model and clinico-radiologic-radiomic (CRR) model (clinico-radiologic variables + radiomic signature), were compared using the Delong test. Early recurrence based on model-predicted high-risk MVI was evaluated between AR (n = 118) and NAR (n = 85) via propensity score matching using patient data from another 2 centers for external validation.

RESULTS

The CRR model showed higher area under the curve values (0.835-0.864 across development, test, and external validation) but no statistically significant improvement over the clinico-radiologic model (0.796-0.828). After propensity score matching, difference in 2-year recurrence between AR and NAR was found in the CRR model predicted high-risk MVI group (P = 0.005) but not in the clinico-radiologic model predicted high-risk MVI group (P = 0.31).

CONCLUSIONS

The prediction model incorporating radiomics provided an accurate preoperative estimation of MVI, showing the potential for choosing the more appropriate surgical procedure between AR and NAR.

Prediction of Microvascular Invasion and Its M2 Classification in Hepatocellular Carcinoma Based on Nomogram Analyses

Background and Aims

As a key pathological factor, microvascular invasion (MVI), especially its M2 grade, greatly affects the prognosis of liver cancer patients. Accurate preoperative prediction of MVI and its M2 classification can help clinicians to make the best treatment decision. Therefore, we aimed to establish effective nomograms to predict MVI and its M2 grade.

Methods

A total of 111 patients who underwent radical resection of hepatocellular carcinoma (HCC) from January 2017 to December 2019 were retrospectively collected. We utilized logistic regression and least absolute shrinkage and selection operator (LASSO) regression to identify the independent predictive factors of MVI and its M2 classification. Integrated discrimination improvement (IDI) and net reclassification improvement (NRI) were calculated to select the potential predictive factors from the results of LASSO and logistic regression. Nomograms for predicting MVI and its M2 grade were then developed by incorporating these factors. Area under the curve (AUC), calibration curve, and decision curve analysis (DCA) were respectively used to evaluate the efficacy, accuracy, and clinical utility of the nomograms.

Results

Combined with the results of LASSO regression, logistic regression, and IDI and NRI analyses, we founded that clinical tumor-node-metastasis (TNM) stage, tumor size, Edmondson–Steiner classification, α-fetoprotein (AFP), tumor capsule, tumor margin, and tumor number were independent risk factors for MVI. Among the MVI-positive patients, only clinical TNM stage, tumor capsule, tumor margin, and tumor number were highly correlated with M2 grade. The nomograms established by incorporating the above variables had a good performance in predicting MVI (AUC_MVI = 0.926) and its M2 classification (AUC_M2 = 0.803). The calibration curve confirmed that predictions and actual observations were in good agreement. Significant clinical utility of our nomograms was demonstrated by DCA.

Conclusions

The nomograms of this study make it possible to do individualized predictions of MVI and its M2 classification, which may help us select an appropriate treatment plan.

A Non-Invasive Nomogram for Preoperative Prediction of Microvascular Invasion Risk in Hepatocellular Carcinoma

Background

Microvascular invasion (MVI) is a significant predictive factor for early recurrence, metastasis, and poor prognosis of hepatocellular carcinoma. The aim of the present study is to identify preoperative factors for predicting MVI, in addition to develop and validate non-invasive nomogram for predicting MVI.

Methods

A total of 381 patients with resected HCC were enrolled and divided into a training cohort (n = 267) and a validation cohort (n = 114). Serum VEGF-A level was examined by enzyme-linked immunosorbent assay (ELISA). Risk factors for MVI were assessed based on univariate and multivariate analyses in the training cohort. A nomogram incorporating independent risk predictors was established and validated.

Result

The serum VEGF-A levels in the MVI positive group (n = 198) and MVI negative group (n = 183) were 215.25 ± 105.68 pg/ml and 86.52 ± 62.45 pg/ml, respectively (P <0.05). Serum VEGF-A concentration ≥138.30 pg/ml was an independent risk factor of MVI (OR: 33.088; 95%CI: 12.871–85.057; P <0.001). Higher serum concentrations of AFP and VEGF-A, lower lymphocyte count, peritumoral enhancement, irregular tumor shape, and intratumoral artery were identified as significant predictors for MVI. The nomogram indicated excellent predictive performance with an AUROC of 0.948 (95% CI: 0.923–0.973) and 0.881 (95% CI: 0.820–0.942) in the training and validation cohorts, respectively. The nomogram showed a good model fit and calibration.

Conclusions

Higher serum concentrations of AFP and VEGF-A, lower lymphocyte count, peritumoral enhancement, irregular tumor shape, and intratumoral artery are promising markers for MVI prediction in HCC. A reliable non-invasive nomogram which incorporated blood biomarkers and imaging risk factors was established and validated. The nomogram achieved desirable effectiveness in preoperatively predicting MVI in HCC patients.

Prognostic Value of Microvascular Invasion in Eight Existing Staging Systems for Hepatocellular Carcinoma: A Bi-Centeric Retrospective Cohort Study

Background

Microvascular invasion (MVI) is a significant risk factor affecting survival outcomes of patients after R0 liver resection (LR) for hepatocellular carcinoma (HCC). However, whether the existing staging systems of hepatocellular carcinoma can distinguish the prognosis of patients with MVI and the prognostic value of MVI in different subtypes of hepatocellular carcinoma remains to be clarified.

Methods

A dual-center retrospective data set of 1,198 HCC patients who underwent R0 LR was included in the study between 2014 and 2016. Baseline characteristics and staging information were collected. Homogeneity and modified Akaike information criterion (AICc) were compared between each system. And the prognostic significance of MVI for overall survival (OS) was studied in each subgroup.

Results

In the entire cohort, there were no significant survival differences between Cancer of the Liver Italian Program (CLIP) score 2 and 3 (p = 0.441), and between Taipei Integrated Scoring System (TIS) score 3 and 4 (p = 0.135). In the MVI cohort, there were no significant survival differences between Barcelona Clinic Liver Cancer stages B and C (p=0.161), CLIP scores 2 and 3 (p = 0.083), TIS scores 0 and 1 (p = 0.227), TIS scores 2 and 3 (p =0.794), Tokyo scores 3 and 4 (p=0.353), and American Joint Committee on Cancer Tumor-Node-Metastasis 7^th stage I and II (p=0.151). Among the eight commonly used HCC staging systems, the Hong Kong Liver Cancer (HKLC) staging system showed the highest homogeneity and the lowest AICc value in both the entire cohort and MVI cohort. In each subgroup of the staging systems, MVI generally exhibited poor survival outcomes.

Conclusions

The HKLC staging system was the most accurate model for discriminating the prognosis of MVI patients, among the eight staging systems. Meanwhile, our findings suggest that MVI may be needed to be incorporated into the current HCC staging systems as one of the grading criteria.

Microvascular Invasion in Hepatocellular Carcinoma: A Review of Its Definition, Clinical Significance, and Comprehensive Management

Hepatocellular carcinoma (HCC) is one of the most common types of malignancies in the world, and most HCC patients undergoing liver resection relapse within five years. Microvascular invasion (MVI) is an independent factor for both the disease-free survival and overall survival of HCC patients. At present, the definition of MVI is still controversial, and a global consensus on how to evaluate MVI precisely is needed. Moreover, this review summarizes the current knowledge and clinical significance of MVI for HCC patients. In terms of management, antiviral therapy, wide surgical margins, and postoperative transcatheter arterial chemoembolization (TACE) could effectively reduce the incidence of MVI or improve the disease-free survival and overall survival of HCC patients with MVI. However, other perioperative management practices, such as anatomical resection, radiotherapy, targeted therapy and immune therapy, should be clarified in future investigations.

A novel classification in predicting prognosis and guiding postoperative management after R0 liver resection for patients with hepatocellular carcinoma and microvascular invasion

BACKGROUND

Microvascular invasion (MVI) is a significant risk factor affecting survival outcomes of patients after R0 liver resection (LR) for hepatocellular carcinoma (HCC). The current classification of MVI is not refined enough to prognosticate long-term survival of these patients, and a new MVI classification is needed.

METHODS

Patients with HCC who underwent R0 LR at the Eastern Hepatobiliary Surgery Hospital from January 2013 to December 2013 and with resected specimens showing MVI were included in this study with an aim to establish a novel MVI classification. The classification which was developed using multivariate cox regression analysis was externally validated.

RESULTS

There were 180 patients in the derivation cohort and 131 patients in the external validation cohort. The following factors were used for scoring: α-fetoprotein level (AFP), liver cirrhosis, tumor number, tumor diameter, MVI number, and distance between MVI and HCC. Three classes of patients could be distinguished by using the total score: class A, ≤3 points; class B, 3.5-5 points and class C, >5 points with distinct long-term survival outcomes (median recurrence free survival (mRFS), 22.6, 10.2, and 1.9 months, P < 0.001). The predictive accuracy of this classification was more accurate than the other commonly used classifications for HCC patients with MVI. In addition, the mRFS of class C patients was significantly prolonged (1.9 months vs. 6.2 months, P < 0.001) after adjuvant transcatheter arterial chemoembolization (TACE).

CONCLUSIONS

A novel MVI classification was established in predicting prognosis of HCC patients with MVI after R0 LR. Adjuvant TACE was useful for class C patients.

A Novel Four-Gene Signature as a Potential Prognostic Biomarker for Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) is a common malignant tumor with high incidence and mortality rates. However, a reliable prognostic signature has not yet been confirmed. Thus, the purpose of the present study was to develop a biomarker with high specificity and sensitivity for the diagnosis and prognosis of patients with HCC. The mRNA expression profiles of HCC were obtained from the GSE19665, GSE41804, and TCGA databases. Subsequently, 193 differentially expressed genes (DEGs) were identified from the intersection of the data from the three datasets. Bioinformatics analysis showed that the identified DEGs are related to the cell cycle, oocyte meiosis, and p53 signaling pathway, among other factors, in cancers. A protein-protein interaction (PPI) and a functional analysis were performed to investigate the biological function of the DEGs and obtain the candidate genes using the MCODE of Cytoscape. The candidate genes were introduced into the TCGA database for survival analysis, and the four candidate genes that were hub genes and meaningful for survival were retained for further verification. We validated the gene and protein expression and determined the prognosis of our patient cohort. In addition, we evaluated the biological functions regulating tumor cell proliferation and metastasis in vitro. According to the ROC curve analysis of gene expression in clinical samples, it was found that the four genes can be used to predict the diagnosis. A survival analysis based on data from the TCGA database and clinical samples showed that the four genes may be used as biomarkers for providing prognoses for patients. The cell functional experiments revealed that these four genes were related to tumor proliferation, migration, and invasion. In conclusion, the genes identified in the present study could be used as markers to diagnose and predict the prognosis of patients with HCC and guide targeted therapy.

Efficacy and Safety of Central Memory T Cells Combined With Adjuvant Therapy to Prevent Recurrence of Hepatocellular Carcinoma With Microvascular Invasion: A Pilot Study

Background

Postoperative adjuvant transcatheter arterial chemoembolization (TACE) following curative hepatectomy has been reported to improve the clinical outcomes of hepatocellular carcinoma (HCC) patients with microvascular invasion (MVI), but more endeavors are required to achieve greater clinical benefit. Central memory T-cell (Tcm) self-transfusion has shown superior antitumor activity in several preclinical studies; however, clinical studies are rare. The aim of this study was to evaluate the clinical benefit and safety of combination treatment with Tcm self-transfusion and TACE as adjuvant treatment in HCC patients with MVI after curative hepatectomy.

Methods

From October 2016 to September 2018, primary HCC patients with histologically confirmed MVI who underwent curative hepatectomy at the Cancer Hospital of the Chinese Academy of Medical Sciences were recruited for this study. The patients were divided into a Tcm group (combined Tcm self-transfusion with TACE treatment) or a control group (TACE treatment alone) according to their willingness. The recurrence-free survival (RFS), quality-of-life (QOL) score, and adverse events of each patient were recorded within 2 years.

Results

A total of 52 patients were enrolled, and 48 were eligible for the final data analysis. The median follow-up time was 20.5 months (95% CI: 17.05–22.55 months). The median RFS time was 9.5 months in the control group; the cutoff date was not reached in the Tcm group (when the follow-up duration was 12 months, p = 0.049, HR = 0.40; 95% CI: 0.16–0.99). Compared with the control group, 1- and 2-year RFS rates were higher in the Tcm group (72.0% vs. 46.4% and 58.18% vs. 39.14%, respectively). Multivariate analysis did not indicate that Tcm treatment was an independent prognostic factor associated with HCC recurrence (p = 0.107, HR = 2.312; 95% CI: 0.835–6.400), which might be due to the small sample size of this study. Nevertheless, Tcm treatment effectively improved a reduced QOL due to HCC and liver function injury. Finally, the safety profile of Tcm treatment in this study was good, without any serious adverse events.

Conclusions

This pilot study showed that Tcm self-transfusion combined with TACE treatment might be a beneficial adjuvant therapy with good safety for primary HCC patients with MVI after curative hepatectomy.

Trial registration number

NCT03575806

A predictive and prognostic model for hepatocellular carcinoma with microvascular invasion based TCGA database genomics

Background

Microvascular invasion (MVI) adversely affects postoperative long-term survival outcomes in patients with hepatocellular carcinoma (HCC). There is no study addressing genetic changes in HCC patients with MVI. We first screened differentially expressed genes (DEGs) in patients with and without MVI based on TCGA data, established a prediction model and explored the prognostic value of DEGs for HCC patients with MVI.

Methods

In this paper, gene expression and clinical data of liver cancer patients were downloaded from the TCGA database. The DEG analysis was conducted using DESeq2. Using the least absolute shrinkage and selection operator, MVI-status-related genes were identified. A Kaplan-Meier survival analysis was performed using these genes. Finally, we validated two genes, HOXD9 and HOXD10, using two sets of HCC tissue microarrays from 260 patients.

Results

Twenty-three MVI-status-related key genes were identified. Based on the key genes, we built a classification model using random forest and time-dependent receiver operating characteristic (ROC), which reached 0.814. Then, we performed a survival analysis and found ten genes had a significant difference in survival time. Simultaneously, using two sets of 260 patients’ HCC tissue microarrays, we validated two key genes, HOXD9 and HOXD10. Our study indicated that HOXD9 and HOXD10 were overexpressed in HCC patients with MVI compared with patients without MVI, and patients with MVI with HOXD9 and 10 overexpression had a poorer prognosis than patients with MVI with low expression of HOXD9 and 10.

Conclusion

We established an accurate TCGA database-based genomics prediction model for preoperative MVI risk and studied the prognostic value of DEGs for HCC patients with MVI. These DEGs that are related to MVI warrant further study regarding the occurrence and development of MVI.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12885-021-09047-1.

Serum Sorbitol Dehydrogenase as a Novel Prognostic Factor for Hepatocellular Carcinoma after Surgical Resection

The majority of patients with hepatocellular carcinoma (HCC) undergoing curative resection experience tumor recurrence. To examine the association between preoperative serum sorbitol dehydrogenase (SORD), a liver-derived enzyme that reflects liver damage, and recurrence of HCC after curative resection, 92 patients were randomly selected who underwent curative resection for HCC between 2011 and 2012 from a prospective registry. Recurrence-free survival (RFS) was compared based on serum SORD levels. Cox proportional hazard models were used to investigate prognostic factors for RFS. During a median follow-up duration of 57.1 months, 43 patients experienced HCC recurrence. Patients with serum SORD ≥15 ng/mL (HR, 3.46; 95% CI, 1.76-6.81; p < 0.001) had worse RFS compared with patients with serum SORD <15 ng/mL. Serum AFP and SORD levels were two independent prognostic factors for RFS. When patients were stratified by baseline serum SORD and AFP levels, patients with serum AFP levels ≥400 ng/mL and serum SORD levels ≥15 ng/mL had a distinctly poor prognosis with the lowest RFS rates (HR, 22.08; 95% CI, 6.91-70.50; p < 0.001). Baseline serum SORD is an effective prognostic factor for HCC after resection. It may help guide patient selection for surgery, especially when combined with serum AFP levels.

Development of pre and post-operative nomograms to predict individual survival for ideal liver resection candidates with hepatocellular carcinoma

BACKGROUND

Liver resection is currently the only recommended treatment option for solitary hepatocellular carcinoma (HCC) at an early stage, with well-preserved liver function and no clinically significant portal hypertension. However, this population is heterogeneous, rendering it crucial to develop a risk stratification tool. Therefore, this study aimed to develop preoperative and post-operative nomograms to predict individual survival and stratify patient risk in the ideal candidates for liver resection.

METHODS

A total of 1405 ideal liver resection candidates were recruited. Independent risk factors were identified by Cox regression model and used to establish two ideal liver resection for overall survival (IROS) nomograms in training cohort. Model performance was assessed by discrimination, calibration, clinical usefulness. The two model also compared with six other prognostic nomograms and six other staging systems.

RESULTS

Multivariate COX analysis revealed that ALP, ln(AFP), NrLR, PNI, ln(tumor size), microvascular invasion, Edmondson-Steiner grade and tumour capsular were the independent risk factors associated with mortality. 5 preoperative variables were incorporated to construct IROS-pre model; All eight available variables were used to draw IROS-post model. The C-index, K-index, time-dependent AUC and DCA of the two models showed significantly better predictive performances than other models. The models could stratify patients into three different risk groups. The web-based tools are convenient for clinical practice.

CONCLUSIONS

These two nomograms were developed to estimate survival probability and stratify three strata with significantly different outcomes, outperforming other models in training and validation cohorts, as well as different subgroups. Both IROS models will help guide individualized follow-up.

Impact of prognostic nutritional index on the recurrence of hepatocellular carcinoma after a curative resection

Backgrounds/Aims

The purpose of this retrospective study was to determine the association between prognostic nutritional index (PNI) and recurrence of hepatocellular carcinoma after a curative resection.

Methods

Between 2007 to 2019, 130 patients who underwent curative hepatectomy for hepatocellular carcinoma were enrolled. PNI was calculated. Its cutoff value was identified through receiver operating characteristic curve analysis. According to PNI, patients were divided into two groups. Univariate and multivariate analyses were performed to identify independent risk factors for recurrence.

Results

The cutoff value of PNI was 52. In univariate analysis, alcoholic liver cirrhosis (p = 0.041), protein induced by vitamin K antagonist- II &ge; 200 (p = 0.012), indocyanine green retention test (ICG R15) >10% (p = 0.001), estimated blood loss &ge; 800 mL (p = 0.037), tumor size (p = 0.001), microvascular invasion (p = 0.023), T-stage (p = 0.001), and PNI < 52 (p = 0.001) were significant factors affecting the recurrence. In multivariate analysis, alcoholic liver cirrhosis (p = 0.046), ICG R15 >10% (p = 0.025), T-stage (p = 0.003), and PNI < 52 (p = 0.046) were independent prognostic factors for disease-free survival.

Conclusions

PNI, a nutritional and immunologic factor, is an independent prognostic factor that can predict the recurrence of hepatocellular carcinoma in patients undergoing a curative resection.

A study of generalization and compatibility performance of 3D U-Net segmentation on multiple heterogeneous liver CT datasets

BACKGROUND

Most existing algorithms have been focused on the segmentation from several public Liver CT datasets scanned regularly (no pneumoperitoneum and horizontal supine position). This study primarily segmented datasets with unconventional liver shapes and intensities deduced by contrast phases, irregular scanning conditions, different scanning objects of pigs and patients with large pathological tumors, which formed the multiple heterogeneity of datasets used in this study.

METHODS

The multiple heterogeneous datasets used in this paper includes: (1) One public contrast-enhanced CT dataset and one public non-contrast CT dataset; (2) A contrast-enhanced dataset that has abnormal liver shape with very long left liver lobes and large-sized liver tumors with abnormal presets deduced by microvascular invasion; (3) One artificial pneumoperitoneum dataset under the pneumoperitoneum and three scanning profiles (horizontal/left/right recumbent position); (4) Two porcine datasets of Bama type and domestic type that contains pneumoperitoneum cases but with large anatomy discrepancy with humans. The study aimed to investigate the segmentation performances of 3D U-Net in: (1) generalization ability between multiple heterogeneous datasets by cross-testing experiments; (2) the compatibility when hybrid training all datasets in different sampling and encoder layer sharing schema. We further investigated the compatibility of encoder level by setting separate level for each dataset (i.e., dataset-wise convolutions) while sharing the decoder.

RESULTS

Model trained on different datasets has different segmentation performance. The prediction accuracy between LiTS dataset and Zhujiang dataset was about 0.955 and 0.958 which shows their good generalization ability due to that they were all contrast-enhanced clinical patient datasets scanned regularly. For the datasets scanned under pneumoperitoneum, their corresponding datasets scanned without pneumoperitoneum showed good generalization ability. Dataset-wise convolution module in high-level can improve the dataset unbalance problem. The experimental results will facilitate researchers making solutions when segmenting those special datasets.

CONCLUSIONS

(1) Regularly scanned datasets is well generalized to irregularly ones. (2) The hybrid training is beneficial but the dataset imbalance problem always exits due to the multi-domain homogeneity. The higher levels encoded more domain specific information than lower levels and thus were less compatible in terms of our datasets.

Radiomics in hepatocellular carcinoma: A state-of-the-art review

Hepatocellular carcinoma (HCC) is the most common cancer and the second major contributor to cancer-related mortality. Radiomics, a burgeoning technology that can provide invisible high-dimensional quantitative and mineable data derived from routine-acquired images, has enormous potential for HCC management from diagnosis to prognosis as well as providing contributions to the rapidly developing deep learning methodology. This article aims to review the radiomics approach and its current state-of-the-art clinical application scenario in HCC. The limitations, challenges, and thoughts on future directions are also summarized.

Prediction of Microvascular Invasion of Hepatocellular Carcinoma Based on Preoperative Diffusion-Weighted MR Using Deep Learning

RATIONALE AND OBJECTIVES

To investigate the value of diffusion-weighted magnetic resonance imaging for the prediction of microvascular invasion (MVI) of Hepatocellular Carcinoma (HCC) using Convolutional Neural Networks (CNN).

MATERIAL AND METHODS

This study was approved by the local institutional review board and the patients' informed consent was waived. Consecutive 97 subjects with 100 HCCs from July 2012 to October 2018 with surgical resection were retrieved. All subjects with diffusion-weighted imaging (DWI) examinations were performed with single-shot echo-planar imaging in a breath-hold routine. DWI parameters were three b values of 0,100,600 sec/mm². First, apparent diffusion coefficients (ADC) images were computed by mono-exponentially fitting the three b-value points. Then, multiple 2D axial patches (28 × 28) of HCCs from b0, b100, b600, and ADC images were extracted to increase the dataset for training the CNN model. Finally, the fusion of deep features derived from three b value images and ADC was conducted based on the CNN model for MVI prediction. The data set was split into the training set (60 HCCs) and the independent test set (40 HCCs). The output probability of the deep learning model in the MVI prediction of HCCs was assessed by the independent student's t-test for data following a normal distribution and Mann-Whitney U test for data violating the normal distribution. Receiver operating characteristic curve and area under the curve (AUC) were also used to assess the performance for MVI prediction of HCCs in the fixed test set.

RESULTS

Deep features in b600 images yielded better performance (AUC = 0.74, p = 0.004) for MVI prediction than b0 (AUC = 0.69, p = 0.023) and b100 (AUC = 0.734, p = 0.011). Comparatively, deep features in the ADC map obtained lower performance (AUC = 0.71, p = 0.012) than that of the higher b value images (b600) for MVI prediction. Furthermore, the fusion of deep features from the b0, b100, b600, and ADC images yielded the best results (AUC = 0.79, p = 0.002) for MVI prediction.

CONCLUSION

Fusion of deep features derived from DWI images concerning the three b-value images and the ADC image yields better performance for MVI prediction.

Baseline HBV-DNA load plus AST/ALT ratio predicts prognosis of HBV-related hepatocellular carcinoma after hepatectomy: A multicentre study

Hepatitis B viral (HBV) load and hepatic enzymes play a critical role in hepatocellular carcinoma (HCC) development. However, the clinical significance of these in HBV-related HCC patients after hepatectomy remains unclear. In this study, we analysed 1,940 HBV-related HCC patients who underwent hepatectomy from four hospitals in west China. Risk classification was constructed based on baseline HBV-DNA load and AST/ALT ratio. Based on the HBV-DNA load and AST/ALT ratio classification, four types with distinguishable prognoses were established. Type 1 patients had the best prognosis with 5-year overall survival (OS) of 69.8%, followed by type 2 and type 3 patients, whereas type 4 patients had the worst prognosis with 5-year OS of 42.7%. Similarly, the four types had statistically different recurrence-free survival. This classification was significantly associated with HCC recurrence (hazard ratio [HR]:1.492, p < .001) and long-term survival (HR: 1.574, p = .001). Pathologically, type 4 correlated with more advanced tumours considering tumour size and microvascular invasion than those in type 1, 2, or 3. Moreover, type 4 patients had more severe hepatic inflammation in underlying liver. Conversely, type 1 patients had an active tumour immune microenvironment as indicated by more CD8+ T cell infiltration and less PD-L1 expression. In conclusion, the classfication based on baseline HBV-DNA load and AST/ALT ratio could effectively stratify HBV-related HCC patients with distinguishable prognoses after hepatectomy.

Prognostic value of baseline imaging and clinical features in patients with advanced hepatocellular carcinoma

Aims

To investigate the prognostic value of baseline imaging features for overall survival (OS) and liver decompensation (LD) in patients with hepatocellular carcinoma (HCC).

Design

Patients with advanced HCC from the SORAMIC trial were evaluated in this post hoc analysis. Several radiological imaging features were collected from baseline computed tomography (CT) and magnetic resonance imaging (MRI) imaging, besides clinical values. The prognostic value of these features for OS and LD (grade 2 bilirubin increase) was quantified with univariate Cox proportional hazard models and multivariate Least Absolute Shrinkage and Selection Operator (LASSO) regression.

Results

Three hundred and seventy-six patients were included in this study. The treatment arm was not correlated with OS. LASSO showed satellite lesions, atypical HCC, peritumoral arterial enhancement, larger tumour size, higher albumin–bilirubin (ALBI) score, liver–spleen ratio <1.5, ascites, pleural effusion and higher bilirubin values were predictors of worse OS, and higher relative liver enhancement, smooth margin and capsule were associated with better OS. LASSO analysis for LD showed satellite lesions, peritumoral hypointensity in hepatobiliary phase, high ALBI score, higher bilirubin values and ascites were predictors of LD, while randomisation to sorafenib arm was associated with lower LD.

Conclusions

Imaging features showing aggressive tumour biology and poor liver function, in addition to clinical parameters, can serve as imaging biomarkers for OS and LD in patients receiving sorafenib and selective internal radiation therapy for HCC.

β-Catenin Expression and Its Association with Prognostic Factors in Hepatocellular Carcinoma: A Study on Alpha-fetoprotein, Histologic Grade, and Microvascular Invasion

Background. Hepatocellular carcinoma (HCC), the most common primary liver cancer. In addition to its high incidence, the disease burden is high due to its poor prognosis and high recurrence rate. Some of the currently known clinicopathologic prognostic factors include alpha-fetoprotein (AFP) level, histologic grade, and microvascular invasion. At the molecular level, β-catenin is one of the most common driver mutation found in HCC. The Wnt/β-catenin pathway regulates cellular processes related to initiation, growth, survival, migration, differentiation, and apoptosis. Although the underlying pathogenesis of hepatocarcinogenesis is known, clinical application warrants a greater understanding of the molecular characteristics and tumor phenotype, especially for determining the prognosis. This study aims to analyze the expression of β-catenin and its association with AFP, histologic grade, and microvascular invasion. Materials and methods. Thirty-five samples of surgically resected HCCs at Cipto Mangunkusumo National Referral Hospital were examined. Diagnoses were made based on histopathological and immunohistochemical findings followed by β-catenin staining. β-catenin expression was analyzed to determine difference between variables. Results and conclusions. Here we show that β-catenin expression is significantly associated with low serum alpha-fetoprotein and well to moderate differentiation Implications. Strong nuclear β-catenin expression implies better prognosis in HCC.

Development of a novel prognostic nomogram for the early recurrence of liver cancer after curative hepatectomy

BACKGROUND

Hepatocellular carcinoma (HCC) is one of the most common malignant cancers worldwide. Curative resection is an effective treatment but HCC recurrence rates remain high. This study aimed to establish a novel prognostic nomogram to assess the risk of recurrence in patients following curative resection.

METHODS

A total of 410 patients undergoing HCC curative resection were recruited from the Guangdong Provincial People's Hospital (GDPH). The cohort was divided into a training group (n=291) and a validation group (n=97). The risk factors for HCC early recurrence within 1 year of curative hepatectomy were identified. Finally, a multivariate prognostic nomogram was developed and validated.

RESULTS

Age, tumor number, tumor capsule, portal vein tumor thrombi, pathological grade, vascular tumor emboli, activated partial thromboplastin time (APTT), and tumor size were identified as independent prognostic risk factors for HCC early recurrence within 1 year of curative hepatectomy. The area under the receiver operating characteristic (ROC) curve (AUC) was 0.806 [95% confidence interval (CI): 0.755 to 0.857; P<0.001], and no AUC/ROC statistical difference was detected between the training and validation sets.

CONCLUSIONS

The nomogram effectively predicted postoperative HCC recurrence within 1 year after curative hepatectomy, which may be a useful tool for the postoperative treatment or follow up for HCC patients.

Impact of everolimus on survival after liver transplantation for hepatocellular carcinoma

BACKGROUND/AIMS

This study aimed to investigate whether everolimus (EVR) affects long-term survival after liver transplantation (LT) in patients with hepatocellular carcinoma (HCC).

METHODS

The data from 303 consecutive patients with HCC who had undergone LT from January 2012 to July 2018 were retrospectively reviewed. The patients were divided into two groups: 1) patients treated with EVR in combination with calcineurin inhibitors (CNIs) (EVR group; n=114) and 2) patients treated with CNI-based therapy without EVR (non-EVR group; n=189). Time to recurrence (TTR) and overall survival (OS) after propensity score (PS) matching were compared between the groups, and prognostic factors for TTR and OS were evaluated.

RESULTS

The EVR group exhibited more aggressive tumor biology than the non-EVR group, such as a higher number of tumors (P=0.003), a higher prevalence of microscopic vascular invasion (P=0.017) and exceeding Milan criteria (P=0.029). Compared with the PS-matched non-EVR group, the PS-matched EVR group had significantly better TTR (P<0.001) and OS (P<0.001). In multivariable analysis, EVR was identified as an independent prognostic factor for TTR (hazard ratio [HR], 0.248; P=0.001) and OS (HR, 0.145; P<0.001).

CONCLUSION

Combined with CNIs, EVR has the potential to prolong long-term survival in patients undergoing LT for HCC. These findings warrant further investigation in a well-designed prospective study.

Systematic review: radiomics for the diagnosis and prognosis of hepatocellular carcinoma

BACKGROUND

Advances in imaging technology have the potential to transform the early diagnosis and treatment of hepatocellular carcinoma (HCC) through quantitative image analysis. Computational "radiomic" techniques extract biomarker information from images which can be used to improve diagnosis and predict tumour biology.

AIMS

To perform a systematic review on radiomic features in HCC diagnosis and prognosis, with a focus on reporting metrics and methodologic standardisation.

METHODS

We performed a systematic review of all full-text articles published from inception through December 1, 2019. Standardised data extraction and quality assessment metrics were applied to all studies.

RESULTS

A total of 54 studies were included for analysis. Radiomic features demonstrated good discriminatory performance to differentiate HCC from other solid lesions (c-statistics 0.66-0.95), and to predict microvascular invasion (c-statistic 0.76-0.92), early recurrence after hepatectomy (c-statistics 0.71-0.86), and prognosis after locoregional or systemic therapies (c-statistics 0.74-0.81). Common stratifying features for diagnostic and prognostic radiomic tools included analyses of imaging skewness, analysis of the peritumoural region, and feature extraction from the arterial imaging phase. The overall quality of the included studies was low, with common deficiencies in both internal and external validation, standardised imaging segmentation, and lack of comparison to a gold standard.

CONCLUSIONS

Quantitative image analysis demonstrates promise as a non-invasive biomarker to improve HCC diagnosis and management. However, standardisation of protocols and outcome measurement, sharing of algorithms and analytic methods, and external validation are necessary prior to widespread application of radiomics to HCC diagnosis and prognosis in clinical practice.

Transarterial Chemoembolization Combined With Radiofrequency Ablation Versus Repeat Hepatectomy for Recurrent Hepatocellular Carcinoma After Curative Resection: A 10-Year Single-Center Comparative Study

Background

The purpose of this study was to compare the efficacy and safety of transarterial chemoembolization (TACE) in combination with radiofrequency ablation (RFA) (TACE-RFA) and repeat hepatectomy in the treatment of recurrent hepatocellular carcinoma (HCC) after curative resection.

Methods

This retrospective study evaluated consecutive medical records of patients who received either TACE-RFA or repeat hepatectomy between January 2010 and May 2021. Overall survival (OS), progression-free survival (PFS), and complications were compared.

Results

Of the 2672 patients who received either TACE-RFA or repeat hepatectomy, 111 eligible patients were included in our study, 63 in the TACE-RFA group and 48 in the repeat hepatectomy group. The median OS was 38 months in the TACE-RFA group and 42 months in the repeat hepatectomy group, with no statistically difference between the two groups (P=0.45). Meanwhile, there was also no statistically significant difference in PFS between the two groups (P=0.634). Although both groups achieved similar outcomes, the rate of major complications was significantly higher in the repeat hepatectomy group (P=0.003).

Conclusions

Patients with recurrent HCC in the TACE-RFA group and the repeat hepatectomy group had similar OS and PFS regardless of the patient’s tumor diameter, but the TACE-RFA group was safer and more minimally invasive.

Preoperative Prediction of Microvascular Invasion in Patients With Hepatocellular Carcinoma Based on Radiomics Nomogram Using Contrast-Enhanced Ultrasound

PURPOSE

This study aimed to develop a radiomics nomogram based on contrast-enhanced ultrasound (CEUS) for preoperatively assessing microvascular invasion (MVI) in hepatocellular carcinoma (HCC) patients.

METHODS

A retrospective dataset of 313 HCC patients who underwent CEUS between September 20, 2016 and March 20, 2020 was enrolled in our study. The study population was randomly grouped as a primary dataset of 192 patients and a validation dataset of 121 patients. Radiomics features were extracted from the B-mode (BM), artery phase (AP), portal venous phase (PVP), and delay phase (DP) images of preoperatively acquired CEUS of each patient. After feature selection, the BM, AP, PVP, and DP radiomics scores (Rad-score) were constructed from the primary dataset. The four radiomics scores and clinical factors were used for multivariate logistic regression analysis, and a radiomics nomogram was then developed. We also built a preoperative clinical prediction model for comparison. The performance of the radiomics nomogram was evaluated via calibration, discrimination, and clinical usefulness.

RESULTS

Multivariate analysis indicated that the PVP and DP Rad-score, tumor size, and AFP (alpha-fetoprotein) level were independent risk predictors associated with MVI. The radiomics nomogram incorporating these four predictors revealed a superior discrimination to the clinical model (based on tumor size and AFP level) in the primary dataset (AUC: 0.849 vs. 0.690; p < 0.001) and validation dataset (AUC: 0.788 vs. 0.661; p = 0.008), with a good calibration. Decision curve analysis also confirmed that the radiomics nomogram was clinically useful. Furthermore, the significant improvement of net reclassification index (NRI) and integrated discriminatory improvement (IDI) implied that the PVP and DP radiomics signatures may be very useful biomarkers for MVI prediction in HCC.

CONCLUSION

The CEUS-based radiomics nomogram showed a favorable predictive value for the preoperative identification of MVI in HCC patients and could guide a more appropriate surgical planning.

A new model based on gamma-glutamyl transpeptidase to platelet ratio (GPR) predicts prognostic outcome after curative resection of solitary hepatocellular carcinoma

BACKGROUND

This study intends to explore the potential clinical value of gamma-glutamyl transpeptidase to platelet ratio (GPR) and the new multi-factor scoring model for recurrence and prognosis prediction in solitary HCC patients who received radical resection.

METHODS

This study retrospectively analyzed 295 HCC patients after curative resection. According to the Receiver Operating Characteristic (ROC) curve, the optimal cut-off value of GPR for predicting prognosis of HCC after resection was determined. The Kaplan Meier method and Cox regression analysis were performed to assess the important potential factors in the prognosis of HCC and determine the independent risk factors. Assign a value to each independent risk factor and establish a new scoring model. Then, using GPR and the new scoring model to evaluate overall survival (OS) and postoperative recurrence probability.

RESULTS

When GPR's cut-off value was selected as 0.30, its predictive efficiency for postoperative prognosis was more favorable than those of other cut-off values (0.76, 0.84 and 0.94). GPR, tumor size, microvascular invasion and neutrophil to lymphocyte ratio (NLR) were identified as independent prognostic predictors. Using these variables, a novel prognostic scoring model was devised and established to identify different levels of risk: high, intermediate and low risk groups. We found that patients with high GPR level and of high risk group would have a poorer OS and a higher recurrence rate after radical resection.

CONCLUSIONS

GPR may serve as a promising predictor for postoperative prognosis and recurrence probability of HCC, and the new prognostic scoring model may be available for postoperative management among HCC patients.

Prognostic and predictive significance of the tumor microenvironment in hepatocellular carcinoma

BACKGROUND

Identification of molecular markers that reflect the characteristics of the tumor microenvironment (TME) may be beneficial to predict the prognosis of post-operative hepatocellular carcinoma (HCC) patients.

OBJECTIVE AND METHODS

A total of 100 tissue samples from HCC patients were separately stained by immunohistochemistry to examine the expression levels of CD56, CD8α, CD68, FoxP3, CD31 and pan-Keratin. The prognostic values were analyzed by Cox regression and the Kaplan-Meier method.

RESULTS

Univariate and multivariate logistic analysis showed that FoxP3 was the independent factor associated with microvascular invasion (MVI), tumor size and envelop invasion; CD68 was associated with envelope invasion and AFP. Kaplan-Meier survival curves revealed that CD68 and FoxP3 expression were significantly associated with relapse free survival (RFS) of HCC patients (P< 0.05). The ROC curve indicated that the combination of tumor number, MVI present and CD68 expression yielded a ROC curve area of 82.3% (86.36% specificity, 68.75% sensitivity) to evaluate the prognosis of HCC patients, which was higher than the classifier established by the combination of tumor number and MVI (78.8% probability, 63.64% specificity and 85.42% sensitivity).

CONCLUSIONS

Our study indicated that CD68 and FoxP3 are associated with prognosis of HCC patients, and CD68 can be considered as a potential prognostic and predictive biomarker.

CKS1B promotes the progression of hepatocellular carcinoma by activating JAK/STAT3 signal pathway

Hepatocellular carcinoma (HCC) is a malignancy of considerable concern due to its continuous increase in morbidity and mortality. This study attempts to identify the molecules that play a key role in the progression of HCC, explore its potential mechanism, and provide more target choices for targeted therapy. Using overexpression plasmid and shRNA, CKS1B was respectively overexpressed and knocked down to explore its biological function roles in HCC progression and development. MTT and colony formation assays showed that knockdown of CKS1B inhibited the survival and proliferation of HCC cell lines (Hep3B and Huh7). The flow cytometry and western blot analysis showed that knockdown of CKS1B significantly induced the apoptosis of Hep3B and Huh7 cells. The wound healing and transwell invasion assays showed that knockdown of CKS1B had a significant inhibitory effect on the migration and invasion of Hep3B and Huh7 cells. These functional tests confirmed that CKS1B acts as an oncogene that regulates the malignant progression of HCC. Moreover, this study also demonstrated that knockdown of CKS1B inhibited the activation of JAK/STAT3 pathway, evidenced by the significantly downregulated p-STAT3 protein expression. Furthermore, knockdown of CKS1B also downregulated STAT3 target genes TIMP-1, Bcl-2 and VEGF, which were involved in controlling cell apoptosis and migration. On the contrary, overexpression of CKS1B caused the completely opposite results. Taken together, CKS1B acts as an oncogene to promote the proliferation and metastasis of HCC cells by activating JAK/STAT3 signaling pathway.

Association of type 2 diabetes mellitus with incidences of microvascular invasion and survival outcomes in hepatitis B virus-related hepatocellular carcinoma after liver resection: A multicenter study

BACKGROUND

Microvascular invasion (MVI) adversely affects long-term survival in patients with hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC). This study aimed to examine the association between preoperative type 2 diabetes mellitus (T2DM) with incidences of MVI and prognosis in HBV-related HCC after liver resection (LR).

MATERIAL AND METHODS

Data of HBV-related HCC patients who underwent LR as an initial therapy from four hospitals in China were retrospectively collected. Clinicopathological factors associated with the incidence of MVI were identified using univariate and multivariate logistic regression analysis. The recurrence-free survival (RFS) and overall survival (OS) curves between different cohorts of patients were generated using the Kaplan-Meier method and compared using the log-rank test.

RESULTS

Of 1473 patients who were included, 219 (14.9%) patients had T2DM. Preoperative T2DM, HBV DNA load, antiviral treatment, AFP level, varices, and tumor encapsulation were identified to be independent predictors of the incidence of MVI. Patients with HBV-related HCC and T2DM had a higher incidence of MVI (65.8%) than those without T2DM (55.4%) (P = 0.004). The RFS and OS were significantly worse in patients with T2DM than those without T2DM (median RFS: 11.1 vs 16.7 months; OS: 26.4 vs 42.6 months, both P < 0.001). Equivalent results were obtained in HCC patients with MVI who had or did not have T2DM (median RFS: 10.0 vs 15.9 months; OS: 24.5 vs 37.9 months, both P < 0.001).

CONCLUSIONS

Preoperative T2DM was an independent risk factor of incidence of MVI. Patients with HBV-related HCC and T2DM had worse prognosis than those without T2DM after LR.

Deep convolutional neural network for preoperative prediction of microvascular invasion and clinical outcomes in patients with HCCs

OBJECTIVES

We aimed to develop and validate a deep convolutional neural network (DCNN) model for preoperative prediction of microvascular invasion (MVI) in hepatocellular carcinoma (HCC) and its clinical outcomes using contrast-enhanced computed tomography (CECT) in a large population of candidates for surgery.

METHODS

This retrospective study included 1116 patients with HCC who had undergone preoperative CECT and curative hepatectomy. Radiological (R), DCNN, and combined nomograms were constructed in a training cohort (n = 892) respectively based on clinicoradiological factors, DCNN probabilities, and all factors; the performance of each model was confirmed in a validation cohort (n = 244). Accuracy and the AUC to predict MVI were calculated. Disease-free survival (DFS) and overall survival (OS) after surgery were recorded.

RESULTS

The proportion of MVI-positive patients was respectively 38.8% (346/892) and 35.7 % (87/244) in the training and validation cohorts. The AUCs of the R, DCNN, and combined nomograms were respectively 0.809, 0.929, and 0.940 in the training cohorts and 0.837, 0.865, and 0.897 in the validation cohort. The combined nomogram outperformed the R nomogram in the training (p < 0.001) and validation (p = 0.009) cohorts. There was a significant difference in DFS and OS between the R, DCNN, and combined nomogram-predicted groups with and without MVI (p < 0.001).

CONCLUSIONS

The combined nomogram based on preoperative CECT performs well for preoperative prediction of MVI and outcome.

KEY POINTS

• A combined nomogram based on clinical information, preoperative CECT, and DCNN can predict MVI and clinical outcomes of patients with HCC. • DCNN provides added diagnostic ability to predict MVI. • The AUCs of the combined nomogram are 0.940 and 0.897 in the training and validation cohorts, respectively.

The impact of liver fibrosis on microvascular invasion and prognosis of hepatocellular carcinoma with a solitary nodule: a Surveillance, Epidemiology, and End Results (SEER) database analysis

BACKGROUND

The pathogenesis of non-cirrhotic hepatocellular carcinoma (HCC) with a high recurrence remains controversial, while microvascular invasion (MVI) is highly suggestive of tumor recurrence. This study aimed to investigate the effects of liver fibrosis on MVI and prognosis in HCC.

METHODS

Based on the data of HCC in the Surveillance, Epidemiology, and End Results (SEER) database [2004-2015], multivariate logistic regression was used for correlation analysis. Survival was analyzed by Log-Rank test and Cox regression, and decision curve analysis and receiver operating characteristic curves were established to evaluate alternative diagnostic and prognostic strategies.

RESULTS

The study included 1,492 patients with MVI (17.8%) or without MVI (82.2%) for HCC with a solitary nodule. Liver fibrosis was significantly correlated with the occurrence of MVI, and the risk of MVI in patients with a fibrosis score F5-6 was lower than in those with a score of F0-4 (OR =0.651, 95% CI: 0.492-0.860). Combining liver fibrosis could improve the prediction performance of MVI risk models, but liver fibrosis was less associated with survival outcomes in comparison with other tumor characteristics.

CONCLUSIONS

Lower liver fibrosis correlated with a higher risk of MVI in HCC with a solitary nodule and was a good indicator for improving the performance of MVI risk models. However, it was not a prognostic sensitive indicator.

Artificial intelligence in assessment of hepatocellular carcinoma treatment response

Artificial Intelligence (AI) continues to shape the practice of radiology, with imaging of hepatocellular carcinoma (HCC) being of no exception. This article prepared by members of the LI-RADS Treatment Response (TR LI-RADS) work group and associates, presents recent trends in the utility of AI applications for the volumetric evaluation and assessment of HCC treatment response. Various topics including radiomics, prognostic imaging findings, and locoregional therapy (LRT) specific issues will be discussed in the framework of HCC treatment response classification systems with focus on the Liver Reporting and Data System treatment response algorithm (LI-RADS TRA).

Microvascular invasion may be the determining factor in selecting TACE as the initial treatment in patients with hepatocellular carcinoma

ABSTRACT

The aim of this study was to investigate factors affecting tumor necrosis with transcatheter arterial chemoembolization (TACE). Factors associated with early hepatocellular carcinoma recurrence after curative hepatectomy were also evaluated.Data of 51 patients who underwent surgery after a single session of TACE at a single university hospital were retrospectively analyzed. Factors that might affect tumor necrosis were determined by evaluating the TACE approach and by analyzing computed tomography and TACE findings, pathologic reports, and laboratory findings.In univariate analysis, microvascular invasion (MVI), radiological capsule appearance on the computed tomography, chronic hepatitis B, diabetes mellitus and serum albumin, MVI were significantly associated with tumor necrosis by TACE (P < .02). In multivariate analysis, MVI was the only statistically significant factor in TACE-induced tumor necrosis (P = .001). In univariate and multivariate analysis, MVI was the strongest factor for recurrence-free survival rate within 2 years (P = .008, P = .002).MVI could be a crucial factor in determining TACE as an initial treatment for hepatocellular carcinoma. MVI is also a strong indicator of recurrence within 2 years after curative hepatic resection.

A Nomogram Based on Combining Clinical Features and Contrast Enhanced Ultrasound LI-RADS Improves Prediction of Microvascular Invasion in Hepatocellular Carcinoma

PURPOSES

To establish a predictive model incorporating clinical features and contrast enhanced ultrasound liver imaging and reporting and data system (CEUS LI-RADS) for estimation of microvascular invasion (MVI) in hepatocellular carcinoma (HCC) patients.

METHODS

In the retrospective study, 127 HCC patients from two hospitals were allocated as training cohort (n=98) and test cohorts (n=29) based on cutoff time-point, June 2020. Multivariate regression analysis was performed to identify independent indicators for developing predictive nomogram models. The area under receiver operating characteristic (AUC) curve was also determined to establish the diagnostic performance of different predictive models. Corresponding sensitivities and specificities of different models at the cutoff nomogram value were compared.

RESULTS

In the training cohort, clinical information (larger tumor size, higher AFP level) and CEUS LR-M were significantly correlated with the presence of MVI (all p<0.05). By incorporating clinical information and CEUS LR-M, the predictive model (LR-M+Clin) achieved a desirable diagnostic performance (AUC=0.80 and 0.84) in both cohorts at nomogram cutoff score value of 89. The sensitivity of LR-M+Clin when predicting MVI in HCC patients was higher than that of the clinical model alone (86.7% vs. 46.7%, p=0.027), while specificities were 78.6% and 85.7% (p=0.06), respectively, in the test cohort. In addition, LR-M+Clin exhibited similar AUC and specificity, but a significantly higher sensitivity (86.7%) than those of LR-M alone and LR-5(No)+Clin (both sensitivities=73.3%, both p=0.048).

CONCLUSION

The predictive model incorporating CEUS LR-M and clinical features was able to predict the MVI status of HCC and is a potential reliable preoperative tool for informing treatment.

Implication of human endogenous retrovirus W family envelope in hepatocellular carcinoma promotes MEK/ERK-mediated metastatic invasiveness and doxorubicin resistance

Human endogenous retrovirus (HERVs), originating from exogenous retroviral infections of germ cells millions of years ago, have the potential for human diseases. Syncytin-1, an envelope protein encoded by the HERV W family, participates in the contexts of schizophrenia, multiple sclerosis, diabetes, and several types of cancers. Nevertheless, there is no report on the expression pattern and potential mechanism of Syncytin-1 in HCC. Here we found Syncytin-1 expression was up-regulated in HCC compared to adjacent non-tumorous tissues, especially in advanced HCC. Syncytin-1 was an independent risk factor to predict vascular invasion, metastasis, larger tumor size, and poor prognosis in HCC patients. Further analysis discovered that Syncytin-1 overexpression positively associated with HCC patients with serum HBsAg positive. Functional experiments in vitro and in vivo demonstrated that Syncytin-1 enhanced cell proliferation, metastasis, and tumorigenicity in HCC. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis suggested that the mitogen-activated protein kinase (MEK)/extracellular signal-regulated protein kinase (ERK) pathway was involved in HCC. Our clinical data indicated that the levels of phosphorylation MEK1/2 and ERK1/2 were increased in HCC comparing with adjacent non-tumorous tissues. It showed the linear correlation between Syncytin-1 expression and upregulated MEK1/2 and ERK1/2 phosphorylation levels in HCC. Furthermore, Syncytin-1 activated MEK/ERK pathway in HCC cells. In-depth research showed that the inflammation-activated MEK/ERK pathway was essential in Syncytin-1 promoted hepatocarcinogenesis. Syncytin-1 suppressed doxorubicin-induced apoptosis via MEK/ERK cascade. In conclusion, Syncytin-1 promoted HCC progression and doxorubicin resistance via the inflammation-activated MEK/ERK pathway. Our findings revealed that Syncytin-1 was a potential prognostic biomarker and therapeutic target for HCC.

Advances in Diagnostic and Interventional Radiology in Hepatocellular Carcinoma

BACKGROUND

Hepatocellular carcinoma (HCC) is one of the leading causes of cancer-related deaths, and radiological imaging and locoregional therapies are essential for the management of patients with HCC.

SUMMARY

In cirrhotic patients, a characteristic imaging pattern establishes the noninvasive diagnosis of HCC with acceptable sensitivity and high specificity. In addition to diagnosis, imaging is used in the staging of patients and treatment allocation. Multiparametric MRI with hepatospecific contrast agents improves lesion detection, characterization, and treatment allocation; recently described imaging criteria allow identification of precursor lesions. Radiofrequency ablation (RFA) and transarterial chemoembolization (TACE) have been established in the treatment of patients with HCC at the early and intermediate stages, respectively. Microwave ablation has been described as an alternative to RFA in selected cases. Imaging-guided brachytherapy, a catheter-based radiotherapy technique, offers advantages to overcome some limitations of the aforementioned therapies, including the tumor location and size. Currently, no adjuvant therapy is recommended after RFA or TACE, but several new drugs are under evaluation. Furthermore, although the exact role of selective internal radiation therapy (SIRT) in HCC still needs to be defined, it is an alternative to systemic agents in patients with intolerance, and additional benefit has been shown in selected subgroups. Additionally, SIRT offers an alternate to TACE with higher objective response rates in patients who needs bridging before transplantation.

KEY MESSAGES

New imaging criteria improved lesion detection in patients at a risk for HCC, and advances in interventional therapies expanded the range of patients eligible for locoregional treatments.

Liver transplantation for hepatocellular carcinoma using grafts from uncontrolled circulatory death donation

Controversy exists regarding whether the rate of hepatocellular carcinoma (HCC) recurrence after orthotopic liver transplantation (OLT) differs when using livers from donation after controlled circulatory death (DCD) versus livers from donation after brain death (DBD). The aim of this cohort study was to analyze rates of HCC recurrence, patient survival, and graft survival after OLT for HCC, comparing recipients of DBD livers (n = 103) with recipients of uncontrolled DCD livers (uDCD; n = 41). No significant differences in tumor size, tumor number, serum alpha-fetoprotein, proportion of patients within Milan criteria, or pre-OLT bridging therapies were identified between groups, although the waitlist period was significantly shorter in the uDCD group (p = 0.040). HCC recurrence was similar between groups. Patient survival was similar between groups, but graft survival was lower in the uDCD group. Multivariate analysis identified recipient age (p = 0.031), pre-OLT bridging therapy (p = 0.024), and HCC recurrence (p = 0.048) as independent risk factors for patient survival and pre-OLT transarterial chemoembolization (p = 0.045) as the single risk factor for HCC recurrence. In conclusion, similar patient survival and lower graft survival were observed in the uDCD group. However, the use of uDCD livers appears to be justified due to a shorter waitlist time, and lower waitlist dropout and HCC recurrence rates.

Preoperative combi-elastography for the prediction of early recurrence after curative resection of hepatocellular carcinoma

PURPOSE

To estimate the prognostic value of preoperative combi-elastography for early recurrence (ER) of hepatocellular carcinoma (HCC) after radical resection.

METHODS

A total of 94 HCC patients undergoing hepatectomy from January to August 2019 were included. The combined elastography (ARIETTA 850, Hitachi Healthcare) was used for real-time tissue elastography and shear wave measurement analysis. Six elastography related indicators were calculated. The patients were randomly divided into a training and a validation group in a 7:3 ratio and prediction model was assessed about discrimination capability by using area under the receiver operating curve. Univariate and multivariate analyses were performed to determine the prognostic value of clinicopathological factors, laboratory tests, and elastography for HCC ER.

RESULTS

The Vs, E, F, and A indexes were significantly higher in patients with ER than in those without ER (P = 0.002, P = 0.002, P < 0.001, and P < 0.001, respectively). Multivariate logistic regression analysis indicated that microvascular invasion (MVI, odds ratio [OR] = 3.964, 95% confidence interval [CI] = 1.326-11.845; P = 0.010) and the F index (OR = 9.533, 95%CI = 1.921-47.296; P = 0.006) were independent predictors of ER in HCC. A ER prediction model based on laboratory tests, MVI and F index were moderate [area under curves (AUCs) in training and validation cohort were 0.829(95%CI: 0.723-0.935; P < 0.001) and 0.846 (95%CI: 0.699-0.994; P = 0.002), respectively].

CONCLUSION

Preoperative combi-elastography analysis could be used as a potential prognostic tool for HCC ER and assist in clinical decision-making.

Adjuvant therapy using ex vivo-expanded allogenic natural killer cells in hepatectomy patients with hepatitis B virus related solitary hepatocellular carcinoma: MG4101 study

Backgrounds/Aims

Fewer reports have been published regarding hepatectomy patients with solitary hepatocellular carcinoma (HCC) who received immunotherapeutic agents as adjuvant therapy. We evaluated the safety and efficacy of ex vivo-expanded allogenic natural killer (NK) cells in those patients with modified International Union Against Cancer (UICC) stage T3.

Methods

From August 2014 to October 2015, five patients who underwent hepatic resection received ex vivo-expanded allogenic NK cells. Patients received five rounds of NK cells (2-3×10⁷ cells/kg) at postoperative 4, 6, 8, 12, and 16 weeks. This study is registered with ClinicalTrials.gov, number NCT02008929.

Results

The median age of the five patients (three men and two women) was 44.8 years (range, 36-54 years). All had hepatitis B virus-related HCC, and the median tumor size was 2.2 cm (range, 2.1-8.2 cm). None of the patients had any adverse events. HCC recurrence developed in two patients at one year after hepatic resection, but four patients were alive at 3 years. The two recurrence-free patients showed a higher ratio of CD8+ T lymphocyte populations before and after administration of ex vivo-expanded allogenic NK cells compared with the three patients who experienced recurrence.

Conclusions

Immunotherapy using ex vivo-expanded allogenic NK cells in hepatectomy patients can be used safely. Further studies should be investigated for efficacy.

Revisiting Surgical Strategies for Hepatocellular Carcinoma With Microvascular Invasion

Although liver resection (LR) and liver transplantation (LT) are widely considered as potentially curative therapies for selected patients with hepatocellular carcinoma (HCC); however, there is still high risk of tumor recurrence in majority of HCC patients. Previous studies demonstrated that the presence of microvascular invasion (MVI), which was defined as the presence of tumor emboli within the vessels adjacent to HCC, was one of the key factors of early HCC recurrence and poor surgical outcomes after LR or LT. In this review, we evaluated the impact of current MVI status on surgical outcomes after curative therapies and aimed to explore the surgical strategies for HCC based on different MVI status with evidence from pathological examination. Surgical outcomes of HCC patients with MVI have been described as a varied range after curative therapies due to a broad spectrum of current definitions for MVI. Therefore, an international consensus on the validated definition of MVI in HCC is urgently needed to provide a more consistent evaluation and reliable prediction of surgical outcomes for HCC patients after curative treatments. We concluded that MVI should be further sub-classified into MI (microvessel invasion) and MPVI (microscopic portal vein invasion); for HCC patients with MPVI, local R0 resection with a narrow or wide surgical margin will get the same surgical results. However, for HCC patients with MI, local surgical resection with a wide and negative surgical margin will get better surgical outcomes. Nowadays, MVI status can only be reliably confirmed by histopathologic evaluation of surgical specimens, limiting its clinical application. Taken together, preoperative assessment of MVI is of utmost significance for selecting a reasonable surgical modality and greatly improving the surgical outcomes of HCC patients, especially in those with liver cirrhosis.

The value of the signal intensity of peritumoral tissue on Gd-EOB-DTPA dynamic enhanced MRI in assessment of microvascular invasion and pathological grade of hepatocellular carcinoma

The aim of the study was to assess the potential role of preoperative gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid (Gd-EOB-DTPA) dynamic enhanced MR imaging for diagnosing microvascular invasion (MVI) and pathological grade of hepatocellular carcinoma (HCC).A total of 113 consecutive HCC patients confirmed by histopathology underwent preoperative Gd-EOB-DTPA dynamic enhanced MRI were included. Signal intensity (SI) of peritumoral, normal liver tissue and tumor parenchyma during arterial phase and hepatobiliary phase (HBP) were analyzed. The receiver operating characteristic (ROC) curves were performed to assess the potential diagnostic capability for MVI and pathological grade of HCC. Kaplan-Meier method was performed to estimate the recurrence-free survival rate and compared using the log rank test.SI ratio of peritumoral tissue to normal liver in arterial phase (SIAp/Al) was independently associated with MVI [odds ratio (OR) = 3.115, 95% confidence interval (CI): 1.867-5.198] and pathological grades (OR = 1.437, 95% CI: 1.042-1.981). The area under the curve (AUC) of SIAp/Al was equivalent to the SI of tumor parenchyma on arterial phase (SIAt) in distinguishing low and high pathological grades. However, the AUC of SIAp/Al (0.851) was larger than peritumoral hypointensity on HBP (0.668) for distinguishing MVI. The recurrence-free survival rate of HCC patients with SIAp/Al<1.1 was higher than HCC with SIAp/Al≥1.1(P = .025).The SIAp/Al in preoperative Gd-EOB-DTPA dynamic enhanced MR imaging is a potential diagnosis marker for MVI and pathological grade of HCC noninvasively. The higher SIAp/Al may predict the poor prognosis of HCC after surgery.

Arterial Spin Labeling MRI for Predicting Microvascular Invasion of T1 Staging Renal Clear Cell Carcinoma Preoperatively

BACKGROUND

Microvascular invasion (MVI) is a valuable factor for T1 staging renal clear cell carcinoma (ccRCC) operation strategy decision, which is confirmed histopathologically post-operation. This study aimed to prospectively evaluate the performance of arterial spin labeling (ASL) MRI for predicting MVI of T1 staging ccRCC preoperatively.

METHODS

16 volunteers and 39 consecutive patients were enrolled. MRI examinations consisted of ASL (three post label delays separately) of the kidney, followed by T1 and T2-weighted imaging. Two sessions of ASL were used to evaluate the reproducibility on volunteers. Renal blood flow of renal cortex, medulla, the entire and solid part of the tumor were measured on ASL images. Conventional imaging features were extracted. MVI and WHO/ISUP classification were evaluated histopathologically. A paired t-test was used to compare the renal cortex and medulla between ASL 1 and ASL 2. The reproducibility was assessed using the intraclass correlation. Differences in mean perfusion between the entire and the solid parts of tumors with or without MVI were assessed separately using Student's t test. The diagnostic performance was assessed. Logistic regression analysis was used to indicate the independent prediction index for MVI.

RESULTS

The two sessions of ASL showed no significant difference between the mean cortex values of RBF. The cortical RBF measurements demonstrated good agreement. 12 ccRCCs presented with MVI histopathologically. Mean perfusion of the solid part of tumors with MVI were 536.4 ± 154.8 ml/min/100 g (PLD1), 2912.5 ± 939.3 ml/min/100 g (PLD2), 3280.3 ± 901.2 ml/min/100 g (PLD3). Mean perfusion of the solid part of tumors without MVI were 453.5 ± 87.2 ml/min/100 g (PLD1), 1043.6 ± 695.8 ml/min/100 g (PLD2), 1577.6 ± 1085.8 ml/min/100 g (PLD3). These two groups have significant difference at all the PLDs (p < 0.05). The RBF of PLD1 of the solid part of tumor perfusion showed well diagnostic performance for predicting MVI: sensitivity 75%, specificity 100%, positive predictive value 66.7%, and negative predictive value 95.7%. The maximum diameter of the tumor, ill-defined margin, and the solid part of tumor perfusion were the independent prediction index for MVI.

CONCLUSION

ASL MR imaging has good reproducibility for renal cortex, and good diagnostic performance for predicting MVI for ccRCC.

Prediction of Microvascular Invasion in Hepatocellular Carcinoma via Deep Learning: A Multi-Center and Prospective Validation Study

Simple Summary

Microvascular invasion (MVI) is an independent risk factor for postoperative recurrence of hepatocellular carcinoma (HCC). Preoperative knowledge of MVI would assist with tailored surgical strategy making to prolong patient survival. Previous radiological studies proved the role of noninvasive medical imaging in MVI prediction. However, hitherto, deep learning methods remained unexplored for this clinical task. As an end-to-end self-learning strategy, deep learning may not only achieve improved prediction accuracy, but may also visualize high-risk areas of invasion by generating attention maps. In this multicenter study, we developed deep learning models to perform MVI preoperative assessments using two imaging modalities—computed tomography (CT) and gadoxetic acid-enhanced magnetic resonance imaging (EOB-MRI). A head-to-head prospective validation was conducted to verify the validity of deep learning models and achieve a comparison between CT and EOB-MRI for MVI assessment. The findings put forward a better understanding of MVI preoperative prediction in HCC management.

Abstract

Microvascular invasion (MVI) is a critical risk factor for postoperative recurrence of hepatocellular carcinoma (HCC). Preknowledge of MVI would assist tailored surgery planning in HCC management. In this multicenter study, we aimed to explore the validity of deep learning (DL) in MVI prediction using two imaging modalities—contrast-enhanced computed tomography (CE-CT) and gadoxetic acid-enhanced magnetic resonance imaging (EOB-MRI). A total of 750 HCCs were enrolled from five Chinese tertiary hospitals. Retrospective CE-CT (n = 306, collected between March, 2013 and July, 2019) and EOB-MRI (n = 329, collected between March, 2012 and March, 2019) data were used to train two DL models, respectively. Prospective external validation (n = 115, collected between July, 2015 and February, 2018) was performed to assess the developed models. Furthermore, DL-based attention maps were utilized to visualize high-risk MVI regions. Our findings revealed that the EOB-MRI-based DL model achieved superior prediction outcome to the CE-CT-based DL model (area under receiver operating characteristics curve (AUC): 0.812 vs. 0.736, p = 0.038; sensitivity: 70.4% vs. 57.4%, p = 0.015; specificity: 80.3% vs. 86.9%, p = 0.052). DL attention maps could visualize peritumoral high-risk areas with genuine histopathologic confirmation. Both DL models could stratify high and low-risk groups regarding progression free survival and overall survival (p < 0.05). Thus, DL can be an efficient tool for MVI prediction, and EOB-MRI was proven to be the modality with advantage for MVI assessment than CE-CT.

Texture Analysis Based on Gd-EOB-DTPA-Enhanced MRI for Identifying Vessels Encapsulating Tumor Clusters (VETC)-Positive Hepatocellular Carcinoma

PURPOSE

To determine the potential findings associated with vessels encapsulating tumor clusters (VETC)-positive hepatocellular carcinoma (HCC), with particular emphasis on texture analysis based on gadolinium-ethoxybenzyl-diethylenetriamine pentaacetic acid (Gd-EOB-DTPA)-enhanced MRI.

METHODS

Eighty-one patients with VETC-negative HCC and 52 patients with VETC-positive HCC who underwent Gd-EOB-DTPA-enhanced MRI before curative partial hepatectomy were retrospectively evaluated in our institution. MRI texture analysis was performed on arterial phase (AP) and hepatobiliary phase (HBP) images. The least absolute shrinkage and selection operator (LASSO) logistic regression was used to select texture features most useful for identifying VETC-positive HCC. Univariate and multivariate analyses were used to determine significant variables for identifying the VETC-positive HCC in clinical factors and the texture features of MRI. Receiver operating characteristic (ROC) analysis and DeLong test were performed to compare the identified performances of significant variables for identifying VETC-positive HCC.

RESULTS

LASSO logistic regression selected 3 features in AP and HBP images, respectively. In multivariate analysis, the Log-sigma-4.0-mm-3D first-order Kurtosis derived from AP images (odds ratio [OR] = 4.128, P = 0.001) and the Wavelet-LHL-GLDM Dependence Non Uniformity Normalized derived from HBP images (OR = 2.280, P = 0.004) were independent significant variables associated with VETC-positive HCC. The combination of the two texture features for identifying VETC-positive HCC achieved an AUC value of 0.844 (95% confidence interval CI, 0.777, 0.910) with a sensitivity of 80.8% (95% CI, 70.1%, 91.5%) and specificity of 74.1% (95% CI, 64.5%, 83.6%).

CONCLUSION

Texture analysis based on Gd-EOB-DTPA-enhanced MRI can help identify VETC-positive HCC.

Considerable effects of imaging sequences, feature extraction, feature selection, and classifiers on radiomics-based prediction of microvascular invasion in hepatocellular carcinoma using magnetic resonance imaging

BACKGROUND

Microvascular invasion (MVI) has a significant effect on the prognosis of hepatocellular carcinoma (HCC), but its preoperative identification is challenging. Radiomics features extracted from medical images, such as magnetic resonance (MR) images, can be used to predict MVI. In this study, we explored the effects of different imaging sequences, feature extraction and selection methods, and classifiers on the performance of HCC MVI predictive models.

METHODS

After screening against the inclusion criteria, 69 patients with HCC and preoperative gadoxetic acid-enhanced MR images were enrolled. In total, 167 features were extracted from the MR images of each sequence for each patient. Experiments were designed to investigate the effects of imaging sequence, number of gray levels (Ng), quantization algorithm, feature selection method, and classifiers on the performance of radiomics biomarkers in the prediction of HCC MVI. We trained and tested these models using leave-one-out cross-validation (LOOCV).

RESULTS

The radiomics model based on the images of the hepatobiliary phase (HBP) had better predictive performance than those based on the arterial phase (AP), portal venous phase (PVP), and pre-enhanced T1-weighted images [area under the receiver operating characteristic (ROC) curve (AUC) =0.792 vs. 0.641/0.634/0.620, P=0.041/0.021/0.010, respectively]. Compared with the equal-probability and Lloyd-Max algorithms, the radiomics features obtained using the Uniform quantization algorithm had a better performance (AUC =0.643/0.666 vs. 0.792, P=0.002/0.003, respectively). Among the values of 8, 16, 32, 64, and 128, the best predictive performance was achieved when the Ng was 64 (AUC =0.792 vs. 0.584/0.697/0.677/0.734, P<0.001/P=0.039/0.001/0.137, respectively). We used a two-stage feature selection method which combined the least absolute shrinkage and selection operator (LASSO) and recursive feature elimination (RFE) gradient boosting decision tree (GBDT), which achieved better stability than and outperformed LASSO, minimum redundancy maximum relevance (mRMR), and support vector machine (SVM)-RFE (stability =0.967 vs. 0.837/0.623/0.390, respectively; AUC =0.850 vs. 0.792/0.713/0.699, P=0.142/0.007/0.003, respectively). The model based on the radiomics features of HBP images using the GBDT classifier showed a better performance for the preoperative prediction of MVI compared with logistic regression (LR), SVM, and random forest (RF) classifiers (AUC =0.895 vs. 0.850/0.834/0.884, P=0.558/0.229/0.058, respectively). With the optimal combination of these factors, we established the best model, which had an AUC of 0.895, accuracy of 87.0%, specificity of 82.5%, and sensitivity of 93.1%.

CONCLUSIONS

Imaging sequences, feature extraction and selection methods, and classifiers can have a considerable effect on the predictive performance of radiomics models for HCC MVI.

Radiomics, Radiogenomics, and Next-Generation Molecular Imaging to Augment Diagnosis of Hepatocellular Carcinoma

Ultrasound, computed tomography, magnetic resonance imaging, and [F]F-fluorodeoxyglucose positron emission tomography are invaluable in the clinical evaluation of human cancers. Radiomics and radiogenomics tools may allow clinicians to standardize interpretation of these conventional imaging modalities, while better linking radiographic hallmarks to disease biology and prognosis. These advances, coupled with next-generation positron emission tomography imaging tracers capable of providing biologically relevant tumor information, may further expand the tools available in our armamentarium against human cancers. We present current imaging methods and explore emerging research that may improve diagnosis and monitoring of local, oligometastatic, and disseminated cancers exhibiting heterogeneous uptake of [F]F-fluorodeoxyglucose, using hepatocellular carcinoma as an example.

Nomogram development and validation to predict hepatocellular carcinoma tumor behavior by preoperative gadoxetic acid-enhanced MRI

OBJECTIVES

Pretreatment evaluation of tumor biology and microenvironment is important to predict prognosis and plan treatment. We aimed to develop nomograms based on gadoxetic acid-enhanced MRI to predict microvascular invasion (MVI), tumor differentiation, and immunoscore.

METHODS

This retrospective study included 273 patients with HCC who underwent preoperative gadoxetic acid-enhanced MRI. Patients were assigned to two groups: training (N = 191) and validation (N = 82). Univariable and multivariable logistic regression analyses were performed to investigate clinical variables and MRI features' associations with MVI, tumor differentiation, and immunoscore. Nomograms were developed based on features associated with these three histopathological features in the training cohort, then validated, and evaluated.

RESULTS

Predictors of MVI included tumor size, rim enhancement, capsule, percent decrease in T1 images (T1_D%), standard deviation of apparent diffusion coefficient, and alanine aminotransferase levels, while capsule, peritumoral enhancement, mean relaxation time on the hepatobiliary phase (T1_E), and alpha-fetoprotein levels predicted tumor differentiation. Predictors of immunoscore included the radiologic score constructed by tumor number, intratumoral vessel, margin, capsule, rim enhancement, T1_D%, relaxation time on plain scan (T1_P), and alpha-fetoprotein and alanine aminotransferase levels. Three nomograms achieved good concordance indexes in predicting MVI (0.754, 0.746), tumor differentiation (0.758, 0.699), and immunoscore (0.737, 0.726) in the training and validation cohorts, respectively.

CONCLUSION

MRI-based nomograms effectively predict tumor behaviors in HCC and may assist clinicians in prognosis prediction and pretreatment decisions.

KEY POINTS

• This study developed and validated three nomograms based on gadoxetic acid-enhanced MRI to predict MVI, tumor differentiation, and immunoscore in patients with HCC. • The pretreatment prediction of tumor microenvironment may be useful to guide accurate prognosis and planning of surgical and immunological therapies for individual patients with HCC.

Using deep learning to predict microvascular invasion in hepatocellular carcinoma based on dynamic contrast-enhanced MRI combined with clinical parameters

PURPOSE

Microvascular invasion (MVI) is a critical determinant of the early recurrence and poor prognosis of patients with hepatocellular carcinoma (HCC). Prediction of MVI status is clinically significant for the decision of treatment strategies and the assessment of patient's prognosis. A deep learning (DL) model was developed to predict the MVI status and grade in HCC patients based on preoperative dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and clinical parameters.

METHODS

HCC patients with pathologically confirmed MVI status from January to December 2016 were enrolled and preoperative DCE-MRI of these patients were collected in this study. Then they were randomly divided into the training and testing cohorts. A DL model with eight conventional neural network (CNN) branches for eight MRI sequences was built to predict the presence of MVI, and further combined with clinical parameters for better prediction.

RESULTS

Among 601 HCC patients, 376 patients were pathologically MVI absent, and 225 patients were MVI present. To predict the presence of MVI, the DL model based only on images achieved an area under curve (AUC) of 0.915 in the testing cohort as compared to the radiomics model with an AUC of 0.731. The DL combined with clinical parameters (DLC) model yielded the best predictive performance with an AUC of 0.931. For the MVI-grade stratification, the DLC models achieved an overall accuracy of 0.793. Survival analysis demonstrated that the patients with DLC-predicted MVI status were associated with the poor overall survival (OS) and recurrence-free survival (RFS). Further investigation showed that hepatectomy with the wide resection margin contributes to better OS and RFS in the DLC-predicted MVI present patients.

CONCLUSION

The proposed DLC model can provide a non-invasive approach to evaluate MVI before surgery, which can help surgeons make decisions of surgical strategies and assess patient's prognosis.

A new horizon in risk stratification of hepatocellular carcinoma by integrating vessels that encapsulate tumor clusters and microvascular invasion

BACKGROUND

Vessels that encapsulate tumor clusters (VETC) is a novel described vascular pattern different from microvascular invasion (MVI) for patients with hepatocellular carcinoma (HCC). The prognostic value of integrating VETC and MVI (VETC-MVI model) in HCC patients after resection remains unclear.

METHODS

From January 2013 to December 2016, 498 HCC patients who underwent curative resection were enrolled from five academic centers and stratified into different groups according to their VETC and MVI statuses. Overall survival (OS), disease-free survival (DFS), and early and late recurrence rates were evaluated.

RESULTS

The patients were divided into four subgroups: VETC^-/MVI^- (n = 277, 55.6%), VETC^-/MVI⁺ (n = 110, 22.1%), VETC⁺/MVI^- (n = 53, 10.6%), and VETC⁺/MVI⁺ (n = 58, 11.6%). The patients in the VETC⁺/MVI^- and VETC^-/MVI⁺ groups had similar long-term outcomes (OS: p = 0.402; DFS: p = 0.990), VETC^-/MVI^- patients showed the best prognosis, and VETC⁺/MVI⁺ patients had the worst prognosis. Further analysis revealed that the VETC-MVI model showed a similar stratification ability for early recurrence but not for late recurrence. The area under the curve values for early recurrence was 0.70, 0.63 and 0.64 for the VETC-MVI model, VETC, and MVI, respectively (VETC-MVI model vs VETC: p < 0.001; VETC-MVI model vs MVI: p = 0.004; VETC vs MVI: p = 0.539). Multivariate Cox regression analysis showed that the VETC-MVI model successfully predicted OS, DFS and early recurrence.

CONCLUSIONS

VETC status provides additional discriminative information for patients with either MVI^- or MVI⁺. A combination of VETC and MVI may help classify subtypes and predict the prognosis of HCC patients.

Comparison of Conventional Gadoxetate Disodium-Enhanced MRI Features and Radiomics Signatures With Machine Learning for Diagnosing Microvascular Invasion

OBJECTIVE. This study aimed to determine the best model for predicting microvascular invasion (MVI) of hepatocellular carcinoma (HCC) using conventional gadolinium-ethoxybenzyl-diethylenetriamine pentaacetic acid (gadoxetate disodium)-enhanced MRI features and radiomics signatures with machine learning. MATERIALS AND METHODS. This retrospective study included 269 patients with a postoperative pathologic diagnosis of HCC. Gadoxetate disodium-enhanced MRI features were assessed, including T1 relaxation time, tumor margin, tumor size, peritumoral enhancement, peritumoral hypointensity, and ADC. Radiomics models were constructed and validated by machine learning. The least absolute shrinkage and selection operator (LASSO) was used for feature selection, and radiomics-based LASSO models were constructed with six classifiers. Predictive capability was assessed using the ROC AUC. RESULTS. Histologic examination confirmed MVI in 111 (41.3%) of the 269 patients. ADC value, nonsmooth tumor margin, and 20-minute T1 relaxation time showed diagnostic accuracy with AUC values of 0.850, 0.847, and 0.846, respectively (p < .05 for all). A total of 1395 quantitative imaging features were extracted. In the hepatobiliary phase (HBP) model, the support vector machine (SVM), extreme gradient boosting (XGBoost), and logistic regression (LR) classifiers showed greater diagnostic efficiency for predicting MVI, with AUCs of 0.942, 0.938, and 0.936, respectively (p < .05 for all). CONCLUSION. ADC value, nonsmooth tumor margin, and 20-minute T1 relaxation time show high diagnostic accuracy for predicting MVI. Radiomics signatures with machine learning can further improve the ability to predict MVI and are best modeled during HBP. The SVM, XGBoost, and LR classifiers may serve as potential biomarkers to evaluate MVI.