The principal aim of the present study was to develop and validate a nomogram predicting overall survival (OS) in patients with α-fetoprotein (AFP)-negative hepatocellular carcinoma (AFP-NHCC) who experience dynamic changes in AFP level after hepatectomy. A cohort of 870 patients were enrolled and randomly assigned into a training cohort (n=600) and a validation cohort (n=270) at a 7:3 ratio. The key variables contributing to the nomogram were determined through random survival forest analysis and multivariate Cox regression. The discriminative ability of the nomogram was evaluated using time-dependent receiver operating characteristic curves and the area under the curves. Furthermore, the nomogram was comprehensively assessed using the concordance index (C-index), calibration curves and clinical decision curve analysis (DCA). Kaplan-Meier (KM) curves analysis was employed to discern survival rates across diverse risk strata of patients. Ultimately, the nomogram incorporated critical factors including sex, tumor size, globulin levels, gamma-glutamyl transferase and fibrinogen levels. In the training and validation cohorts, the C-indexes were 0.72 [95% confidence interval (CI): 0.685-0.755) and 0.664 (95% CI: 0.611-0.717], respectively, attesting to its predictive validity. The nomogram demonstrated excellent calibration and DCA further confirmed its clinical usefulness. Additionally, KM curve analysis unveiled statistically significant differences in OS among three distinct risk groups. In conclusion, the present study successfully formulated a nomogram predicting 3-, 5- and 8-year OS in patients with AFP-NHCC with dynamic changes in AFP level post-local resection. This model serves as a valuable tool for clinicians to promptly identify high-risk patients, thereby facilitating timely interventions and potentially enhancing patient survival outcomes.

To develop a novel hybrid model for preoperative prediction of tertiary lymphoid structures (TLSs) of hepatocellular carcinoma (HCC), and to identify patients who may benefit from postoperative targeted immunotherapy.

Retrospective data were gathered from 332 patients with HCC who underwent surgical resection and gadoxetate disodium (Gd-EOB-DTPA) enhanced MRI at two tertiary hospitals (training cohort, n = 205; internal validation cohort, n = 90; and external validation cohort, n = 37) between March 2020 and January 2023. Radiomic features were extracted from Gd-EOB-DTPA-enhanced MRI sequences. These signatures were integrated with clinical-radiologic (CR) factors into a hybrid model and nomogram for clinical application. The performance of the model was assessed using the area under the curve (AUC) and 95% confidence intervals (CI).

The hybrid model outperformed the radiomics and CR models in the training cohort (AUC = 0.860 [95% CI: 0.805, 0.904], 0.784 [95% CI: 0.721, 0.838], and 0.809 [95% CI: 0.748, 0.860]). The hybrid model showed optimal performance, with AUCs of 0.823 (95% CI: 0.728, 0.895) and 0.875 (95% CI: 0.725, 0.960) in the internal and external validation cohorts, respectively. The calibration curve demonstrated that the nomogram had good diagnostic ability, and decision curve analysis indicated good clinical utility across all cohorts. Importantly, patients with a predicted high risk of TLSs from the hybrid model gained a survival benefit from targeted immunotherapy.

The hybrid model showed satisfactory performance in predicting intra-tumoral TLS positivity and targeted immunotherapy benefit in patients with HCC, potentially assisting clinicians in selecting precise individualized therapies.

Question How can accurate preoperative risk stratification of tertiary lymphoid structures positivity HCC be achieved to support targeted immunotherapy decision-making? Findings A hybrid model combining radiomics model and clinical-radiological model may be a reliable marker for predicting tertiary lymphoid structures positivity HCC. Clinical relevance Using this hybrid model may be useful in predicting tertiary lymphoid structures and screening candidate patients for targeted immunotherapy based on multiparametric MRI, which has potential clinical value in guiding clinical decision-making and improving patient outcomes.

This study aims to develop and validate a predictive nomogram for early recurrence in hepatocellular carcinoma (HCC), utilizing gadoxetic acid-enhanced MRI and intravoxel incoherent motion (IVIM) imaging to improve preoperative assessment and decision-making.

From March 2018 and June 2022, a total of 245 patients with pathologically confirmed HCC, who underwent preoperative gadoxetic acid-enhanced MRI and IVIM, were retrospectively enrolled from two hospitals. These patients were divided into a training cohort (n = 160) and a validation cohort (n = 85). All patients were followed until death or the last follow-up date, with a minimum follow-up period of two years. Clinical indicators and pathologic information were compared between train cohort and validation cohort. Radiological features and diffusion parameters were compared between recurrence and non-recurrence groups using the chi-square test, Mann-Whitney U test and independent sample t test in training cohort. Univariate and multivariate analyses were performed to identify significant clinical-radiological variables associated with early recurrence in the training cohort. Based on these findings, a predictive nomogram integrating risk factors and diffusion parameters was developed. The predictive performance of the nomogram was evaluated in both the training and validation cohorts.

No statistically significant difference in clinical and pathologic characteristics were observed between the training and validation cohorts. In training cohort, significant differences were identified between the recurrence and non-recurrence groups in tumor size, nodule-in-nodule architecture, mosaic architecture, non-smooth tumor margin, intratumor necrosis, satellite nodule, and peritumoral hypo-intensity in the hepatobiliary phase (HBP). The results of multivariate analysis identified tumor size (HR, 1.435; 95 % CI, 0.702-2.026; p < 0.05), mosaic architecture (HR, 0.790; 95 % CI, 0.421-1.480; p < 0.05), non-smooth tumor margin (HR, 1.775; 95 % CI, 0.941-3.273; p < 0.05), intratumor necrosis (HR, 1.414; 95 % CI, 0.807-2.476; p < 0.05), satellite nodule (HR, 0.648; 95 % CI, 0.352-1.191; p < 0.01), peritumoral hypo-intensity on HBP (HR, 2.786; 95 % CI, 1.141-6.802; p < 0.001) and D (HR, 0.658; 95 % CI,0.487-0.889; p < 0.01) were the independent risk factor for recurrence. The nomogram exhibited excellent predictive performance with C-index of 0.913 and 0.875 in the training cohort and validation cohort, respectively. Also, based on the nomogram score, the patients were classified according to risk factor and the Kaplan-Meier curve analysis also showed that the nomogram had a good predictive efficacy.

The nomogram, integrating radiological risk factors and diffusion parameters, offers a reliable tool for preoperative prediction of early recurrence in HCC patients.

To develop an explainable fusion model that combines clinical, radiomic, and habitat features to predict postoperative early recurrence in hepatocellular carcinoma (HCC).

The bicentric retrospective study included 370 patients with surgically confirmed early-stage HCC who underwent gadoxetic acid-enhanced MRI. The patients were stratified into a training cohort (n=296) and an external validation cohort (n=74). From the hepatobiliary phase images, habitat and radiomics features were extracted across the entire tumor and used to construct radiomics and habitat models. Additionally, a clinical model was established utilizing relevant clinical features. Subsequently, all previously mentioned features were merged to construct the fusion model (HabRad_FB). Diagnostic performance of these models was assessed and compared using the area under the receiver operating characteristic curve (AUC), net reclassification index (NRI), and integrated discrimination improvement (IDI). The fusion model was then interpreted using SHapley Additive exPlanations (SHAP) analysis.

Tumor recurrence was observed in 73 out of 370 patients (19.7%; 55.2±11.3 years; male=333). Among all study cohorts, the HabRad_FB model showed the highest AUC (0.820-0.959), outperforming the clinical (0.517-0.729), radiomics (0.707-0.815), and habitat (0.729-0.861) models. The HabRad_FB model also demonstrated significant improvement in IDI in the training cohort and NRI in the validation cohort. SHAP force plots provided valuable insights into the interpretation of HabRad_FB model's predictions for early recurrence.

The HabRad_FB, an explainable fusion model, aids clinicians in accurately and non-invasively predicting the early recurrence of HCC preoperatively. This model might provide great potential in prognostic prediction and clinical management.

This study aims to develop a machine learning model that accurately diagnoses microvascular invasion (MVI) in hepatocellular carcinoma by using radiomic features from MVI-positive regions of interest (ROIs). Unlike previous studies, which do not account for the location and distribution of MVI, this research focuses on correlating preoperative imaging with postoperative pathological MVI. This study involves obtaining ex vivo 3D ultrasound images of 36 hepatic specimens from nine rabbits. These images are fused with whole-slide images to localize MVI regions precisely. The identified MVI regions are segmented into MVI-positive ROIs, with a 1:3 ratio of positive to negative ROIs. Radiomic features are extracted from each ROI, and 30 features highly associated with MVI are selected for model development. The performance of several machine learning models is evaluated using metrics such as sensitivity, specificity, accuracy, the area under the curve (AUC), and F1 score. The GBDT model achieves the best results, with an AUC of 0.91, an F1 score of 0.85, a sensitivity of 0.76, a specificity of 0.92, and an accuracy of 0.86. The high diagnostic accuracy of these models highlights the potential for future clinical application in the precise diagnosis of MVI using radiomic features from MVI-positive ROIs.

This study aimed to develop two preoperative magnetic resonance imaging (MRI) based models for detecting and classifying microvascular invasion (MVI) in early-stage small hepatocellular carcinoma (sHCC) patients.

MVI is graded as M0 (no invasion), M1 (invasion of five or fewer vessels located within 1 cm of the tumor's peritumoral region), and M2 (invasion of more than five vessels or those located more than 1 cm from the tumor's surface). This study enrolled 395 early-stage sHCC (≤ 3 cm) patients from three centers who underwent preoperative gadopentetate-enhanced MRI. From the first two centers, 310 patients were randomly divided into training (n = 217) and validation (n = 93) cohorts in a 7:3 ratio to develop the first model for predicting MVI presence. Among these, 153 patients with identified MVI were further divided into training (n = 112) and validation (n = 41) cohorts, using the same ratio, to construct the second model for MVI classification. An independent test cohort of 85 patients from the third center to validate both models. Univariate and multivariate logistic regression analyses identified independent predictors of MVI and its classification in the training cohorts. Based on these predictors, two nomograms were developed and assessed for their discriminative ability, calibration, and clinical usefulness. Besides, considering the two models are supposed applied in a serial fashion in the real clinical setting, we evaluate the performance of the two models together on the test cohorts by applying them simultaneously. Kaplan-Meier survival curve analysis was employed to assess the correlation between predicted MVI status and early recurrence, similar to the association observed with actual MVI status and early recurrence.

The MVI detection nomogram, with platelet count (PLT), activated partial thromboplastin time (APTT), rim arterial phase hyperenhancement (Rim APHE) and arterial peritumoral enhancement, achieved area under the curve (AUC) of 0.827, 0.761 and 0.798 in the training, validation, and test cohorts, respectively. The MVI classification nomogram, integrating Total protein (TP), Shape, Arterial peritumoral enhancement and enhancement pattern, achieved AUC of 0.824, 0.772, and 0.807 across the three cohorts. When the two models were applied on the test cohorts in a serial fashion, they both demonstrated good performance, which means the two models had good clinical applicability. Calibration and decision curve analysis (DCA) results affirmed the model's reliability and clinical utility. Notably, early recurrence was more prevalent in the MVI grade 2 (M2) group compared to the MVI-absent and M1 groups, regardless of the actual or predicted MVI status.

The nomograms exhibited excellent predictive performance for detecting and classifying MVI in patients with early-stage sHCC, particularly identifying high-risk M2 patients preoperatively.

Hepatocellular carcinoma (HCC) is one of the leading causes of cancer-related deaths globally, owing to its high recurrence rate of 50 to 70% within five years. Despite known associations of certain DNA damage and repair (DDR) genes with tumor recurrence and drug resistance, a comprehensive understanding of DDR pathways' role in predicting HCC recurrence and therapeutic responses remains elusive. Addressing this gap could offer significant advancements in prognostic and therapeutic strategies for HCC. This study used 769 RNA sequencing samples from public datasets and 53 samples from Xiangya Hospital for DDR model training and validation. It came out that DDR pathways were significantly enriched in samples with P53 mutations. Next, among the 173 combinations of algorithms and parameters, CoxBoost + RSF, Lasso [fold = 10] + RSF, and Lasso [fold = 50] + RSF demonstrated the best performance. The average AUC values of 1 to 5 years and the average concordance index (C-index) value were around 0.7. The risk scores were increased in tumors with recurrence, P53 mutation, and higher TNM stages. High-risk groups, characterized by enriched DDR pathways, exhibited lower CD8 + T cell infiltration and poorer responses to immunotherapy using atezolizumab and bevacizumab, emphasizing the potential of DDR signatures as valuable prognostic and therapeutic biomarkers. In conclusion, the DDR signatures associated with P53 mutations can predict recurrence and therapeutic response in HCC, highlighting their potential as prognostic and therapeutic biomarkers.

Extrahepatic recurrence (EHR) is a significant negative prognostic factor in hepatocellular carcinoma (HCC). Although EHR is commonly observed in high-risk patients following HCC hepatectomy, its occurrence without concurrent intrahepatic HCC remains poorly understood. Therefore, this study aims to examine the clinical characteristics and risk factors associated with EHR in patients without intrahepatic HCC at diagnosis.

This study included 1066 treatment-naïve patients who underwent curative hepatectomy for HCC at four tertiary academic centers between January 2004 and December 2019. After excluding those with intrahepatic recurrence (IHR), concurrent EHR, or incomplete clinical records, 569 patients were included in the final analysis. Risk factors for EHR were assessed using multivariate Cox regression over a median follow-up period of 3.91 years.

Among the cohort, 38 patients developed EHR post-surgery without residual intrahepatic HCC, with a median follow-up of 1.04 years. These patients experienced earlier initial HCC recurrence than those without EHR (1.73 vs. 4.43 years). Multivariate analysis revealed significant associations between EHR and microvascular invasion (hazard ratio [HR]: 2.418, p = 0.020), tumor necrosis (HR: 2.592, p = 0.009), and initial tumor staging beyond the Milan criteria (HR: 3.008, p = 0.001). Moreover, Cox regression analysis revealed that EHR strongly correlated with decreased post-hepatectomy survival (HR: 14.044, p < 0.001). Cumulative EHR and survival rates were closely linked to the number of risk factors present.

EHR without detectable IHR is significant and warrants close monitoring in high-risk patients.

Post-operative recurrence is a major clinical challenge with hepatocellular carcinoma (HCC). While currently unapproved, anti-programmed cell death 1 (PD-1) and anti-vascular endothelial growth factor combination adjuvant therapy showed promise. We initiated a phase I trial using sequential treatment with ropeginterferon alfa-2b (ropeg), a novel interferon-based antiviral and antitumor agent, followed by anti-PD-1 therapeutic antibody nivolumab as an adjuvant therapy for hepatitis B virus (HBV)-related HCC.

Patients who underwent surgical resection of HBV-related HCC with curative intent received sequential therapy with six doses of ropeg every two weeks at 450 μg, followed by three doses of nivolumab escalating from 0.3 to 0.75 mg/kg every two weeks. Safety, HBV surface antigen (HBsAg) loss or decrease, anti-HBV surface (HBs) antibodies, cancer recurrence, and survival were evaluated.

Fifteen eligible patients were enrolled. Most adverse events (AEs) were mild or moderate and no severe or serious AEs were observed. Alanine transaminase flares, including one grade 3 event as dose-limiting toxicity, were noted in five cases and the final recommended dose for anti-PD1 was determined at 0.75 mg/kg. Interestingly, all five cases had HBsAg clearance or reduction. All patients in the study were alive without cancer recurrence during a median follow-up of 1024 days with six patients surviving > 4 years and three for > 5 years.

This phase I trial supports the safety and clinical efficacy of sequential treatment with ropeg and nivolumab in post-resection HBV-related HCC. This regimen holds promise for further adjuvant therapy trials in HCC, both HBV-related and other types.

To develop an MRI-based score that enables individualized predictions of the survival benefit of wide over narrow resection margins.

This single-center retrospective study (December 2011 to May 2022) included consecutive patients who underwent curative-intent resection for single Barcelona Clinic Liver Cancer (BCLC) 0/A HCC and preoperative contrast-enhanced MRI. In patients with narrow resection margins, preoperative demographic, laboratory, and MRI variables independently associated with early recurrence-free survival (RFS) were identified using Cox regression analyses, which were employed to develop a predictive score (named "MARGIN"). Survival outcomes were compared between wide and narrow resection margins in a propensity-score matched cohort for the score-stratified low- and high-risk groups, respectively.

Four hundred nineteen patients (median age, 54 years; 361 men) were included, 282 (67.3%) undergoing narrow resection margins. In patients with narrow resection margins, age, alpha-fetoprotein (AFP) > 400 ng/mL, protein induced by vitamin K absence or antagonist-II (PIVKA-II) > 200 mAU/mL, radiological involvement of liver capsule, and infiltrative appearance were associated with early RFS (p values, 0.002-0.04) and formed the MARGIN score with a testing dataset C-index of 0.75 (95% CI: 0.65-0.84). In the matched cohort, wide resection margin was associated with improved early RFS rate for the high-risk group (MARGIN score ≥ - 1.3; 71.1% vs 41.0%; p = 0.02), but not for the low-risk group (MARGIN score < - 1.3; 79.7% vs 76.1%; p = 0.36).

In patients with single BCLC 0/A HCC, the MARGIN score may serve as promising decision-making to indicate the need for wide resection margins.

The MARGIN score has the potential to identify patients who would benefit more from wide resection margins than narrow resection margins, improving the postoperative survival of patients with single BCLC 0/A hepatocellular carcinoma (HCC).

Age, AFP, PIVKA-II, radiological involvement of liver capsule, and infiltrative appearance were associated with early RFS and formed the MARGIN score. The MARGIN score achieved a testing dataset C-index of 0.75. Wide resection margins were associated with improved early RFS for the high-risk group, but not for the low-risk group.

Tumor recurrence post-operation of hepatocellular carcinoma (HCC) impacts patient prognosis. Identifying and predicting 5-year HCC recurrence following surgery remains a substantial challenge.

We included 338 patients diagnosed with HCC who underwent surgery from January 2013 to December 2018. Traditional logistic regression, random forest (RF), and LASSO regression methods were used to develop a predictive model for 5-year recurrence. The findings were presented visually using nomogram. The accuracy and sensitivity of the predictive model were evaluated by receiver operating curves (ROC) and decision curve analysis (DCA).

Of the 338 patients, 172 (50.9%) experienced 5 years recurrence, with a gender distribution of 79.7% males. Univariate and multivariate logistic regression analysis identified that three independent predictors of 5-year HCC recurrence (all P < 0.001). The area under the curve (AUC) value of the model (Model-1) constructed was 0.678. Then we combined LASSO regression and RF construct a predictive model including six factors: age, transarterial chemoembolization (TACE), microvascular invasion (MVI), alcohol, size, and number. The AUC of the model (Model-2) constructed was 0.733. DeLong's test results showed that Model-2 had significantly better prediction ability compared with Model-1 (P = 0.004). DCA also demonstrated that Model-2 had better predictive accuracy (P < 0.05). Then we constructed a nomogram, and Kaplan-Meier analysis showed that patients in the low-risk group had significantly better prognosis than the high (P < 0.001).

The predictive accuracy of our model, incorporating factors, such as age, alcohol, size, number, MVI, and TACE, significantly enhances clinical practice management by accurately forecasting 5 years HCC recurrence.

Accurate prediction of microvascular invasion (MVI) in hepatocellular carcinoma (HCC) is crucial for guiding treatment. This study evaluates and compares the performance of clinicoradiologic, traditional radiomics, deep-learning radiomics, feature fusion, and decision fusion models based on multi-region MR habitat imaging using six machine-learning classifiers.

We retrospectively included 300 HCC patients. The intratumoral and peritumoral regions were segmented into distinct habitats, from which radiomics and deep-learning features were extracted using arterial phase MR images. To reduce feature dimensionality, we applied intra-class correlation coefficient (ICC) analysis, Pearson correlation coefficient (PCC) filtering, and recursive feature elimination (RFE). Based on the selected optimal features, prediction models were constructed using decision tree (DT), K-nearest neighbors (KNN), logistic regression (LR), random forest (RF), support vector machine (SVM), and XGBoost (XGB) classifiers. Additionally, fusion models were developed utilizing both feature fusion and decision fusion strategies. The performance of these models was validated using the area under the receiver operating characteristic curve (ROC AUC), calibration curves, and decision curve analysis.

The decision fusion model (VOI-Peri10-1) using LR and integrating clinicoradiologic, radiomics, and deep-learning features achieved the highest AUC of 0.808 (95% confidence interval [CI]: 0.807-0.912) in the test cohort, with good calibration (Hosmer-Lemeshow test, P > 0.050) and clinical net benefit.

The LR-based decision fusion model integrating clinicoradiologic, radiomics, and deep-learning features shows promise for preoperative prediction of MVI in HCC, aiding in patient outcome predictions and personalized treatment planning.

Hepatocellular carcinoma (HCC) exhibits high de novo recurrence rates post-resection. Current post-surgery recurrence prediction methods are limited, emphasizing the need for reliable biomarkers to assess recurrence risk. We aimed to develop methylation-based markers for classifying HCC patients and predicting their risk of de novo recurrence post-surgery.

In this retrospective cohort study, we analyzed data from HCC patients who underwent surgical resection in Korea, excluding those with recurrence within one year post-surgery. Using the Infinium Methylation EPIC array on 140 samples in the discovery cohort, we classified patients into low- and high-risk groups based on methylation profiles. Distinctive markers were identified through random forest analysis. These markers were validated in the cancer genome atlas (n=217), Validation cohort 1 (n=63) and experimental Validation using a methylation-sensitive high-resolution melting (MS-HRM) assay in Validation cohort 1 and Validation cohort 2 (n=63).

The low-risk recurrence group (methylation group 1; MG1) showed a methylation average of 0.73 (95% confidence interval [CI] 0.69-0.77) with a 23.5% recurrence rate, while the high-risk group (MG2) had an average of 0.17 (95% CI 0.14-0.20) with a 44.1% recurrence rate (P<0.03). Validation confirmed the applicability of methylation markers across diverse populations, showing high accuracy in predicting the probability of HCC recurrence risk (area under the curve 96.8%). The MS-HRM assay confirmed its effectiveness in predicting de novo recurrence with 95.5% sensitivity, 89.7% specificity, and 92.2% accuracy.

Methylation markers effectively classified HCC patients by de novo recurrence risk, enhancing prediction accuracy and potentially offering personalized management strategies.

To develop a diagnostic model for predicting the early recurrence of microvascular invasion (MVI)-negative hepatocellular carcinoma (HCC) after surgical resection, using the Liver Imaging Reporting and Data System (LI-RADS) version 2018.

This retrospective study included 73 patients with MVI-negative HCC who underwent Gadoxetic acid-enhanced MRI (EOB-MRI) scanning before surgical resection. The clinical factors and LI-RADS v2018 MRI features associated with early recurrence were determined using univariable and multivariable analyses. A diagnostic model predicting early recurrence after surgical resection was developed, and its predictive ability was evaluated via a receiver operating characteristic curve. Then, the recurrence-free survival (RFS) rates were analyzed by Kaplan-Meier method.

In total, 26 (35.6%) patients were diagnosed with early recurrence according to the follow-up results. Infiltrative appearance and targetoid hepatobiliary phase (HBP) appearance were independent predictors associated with early recurrence (p < 0.05). For the established diagnostic model that incorporated these two significant predictors, the AUC value was 0.76 (95% CI: 0.64-0.85) for predicting early recurrence after resection, which was higher than the infiltrative appearance (AUC: 0.67, 95% CI: 0.55-0.78, p = 0.019) and targetoid HBP appearance (AUC: 0.68, 95% CI:0.57-0.79, p = 0.028). In the RFS analysis, patients with infiltrative appearance and targetoid HBP appearance showed significantly lower RFS rates than those without infiltrative appearance (2-year RFS rate, 48.0% vs. 72.0%; p = 0.009) and targetoid HBP appearance (2-year RFS rate, 60.0% vs. 35.0%; p = 0.003).

An EOB-MRI model based on infiltrative appearance and targetoid HBP appearance showed good performance in predicting early recurrence of HCC after surgery, which may provide personalized guidance for clinical treatment decisions in patients with MVI-negative HCC.

Liver cancer, particularly hepatocellular carcinoma (HCC), is a major health concern globally and in China, possibly shows recurrence after ablation treatment in high-risk patients. This study investigates the prognosis of early-stage male HCC patients with chronic hepatitis virus B (HBV) infection who also have long-term smoking and drinking habits, following local ablation treatment.

Data from 257 patients treated at Capital Medical University, Beijing Youan Hospital from 2014 to 2022 were retrospectively analyzed. We first screened the variables by Lasso regression and random survival forest (RSF), followed by multivariate Cox regression analysis. Based on the screened variables after these steps, we performed and validated a nomogram to predict the survival status of these patients.

Our results indicated that monocytes and globulin are risk factors while pre-albumin (PALB) is protective after selected by Lasso, RSF and multivariate Cox regression, providing a robust tool for predicting overall survival and guiding treatment for high-risk HCC patients. With promising discrimination, accuracy and clinical applicability, our model was translated into a nomogram for practical use.

Our prognostic model effectively identifies key risk factors such as monocytes, globulin and PALB, providing accurate predictions for HBV-induced male patients with smoking and drinking habits.

Adjuvant lenvatinib in combination with transarterial chemoembolization (TACE) has demonstrated prolonged disease-free survival in hepatocellular carcinoma patients at high risk of recurrence post-resection. Here, we present the case of a 68-year-old woman who developed serious side effects including pemphigus erythematosus (PE) linked to lenvatinib usage. Initially treated for breast cancer with radical surgery in April 2018 followed by adjuvant therapy, she was later diagnosed with liver cancer, initially mistaken for metastatic breast cancer to the liver. Although systemic treatment for advanced breast cancer was received, the tumor continued to progress and required partial removal of the liver after final evaluation. Subsequent pathology revealed hepatocellular carcinoma combined with risk factors for recurrence, prompting adjuvant therapy with TACE and oral lenvatinib. After three weeks of lenvatinib administration, the patient developed a skin rash diagnosed as PE through skin pathology. Treatment involved oral methylprednisolone, intravenous human immune globulin, and supportive care, resulting in a cure within a month. This unique case highlights the importance of further research not only on lenvatinib but also on monitoring and managing adverse reactions associated with targeted drugs to optimize patient safety and treatment outcomes.

The Oncology Grand Rounds series is designed to place original reports published in the Journal into clinical context. A case presentation is followed by a description of diagnostic and management challenges, a review of the relevant literature, and a summary of the authors' suggested management approaches. The goal of this series is to help readers better understand how to apply the results of key studies, including those published in Journal of Clinical Oncology, to patients seen in their own clinical practice.

Microvascular invasion (MVI) is linked to poor prognosis, early recurrence and post-surgical intrahepatic metastasis of hepatocellular carcinoma (HCC) but roles of tumor-associated endothelial cells (TECs) remain unclear. The aim of the current study was to investigate the role of TECs in microvascular invasion in HCC.

Single-cell RNA sequencing (scRNA-seq) data from three patients with MVI and two patients with non-MVI HCC were used to identify TECs subpopulations via Seurat R package. Using bioinformatics analysis identified co-expression modules associated with MVI in TECs. Differential gene expression analysis, KME values and Gene Expression Profiling Interactive Analysis (GEPIA) survival were utilized to identify genes with significant involvement. TECs subgroup developmental trajectory was analyzed using monocle2. Five additional spatial transcriptomics (ST) datasets and four HCC postoperative pathological specimens were used to validate the differential expression of subgroups of TECs and hub genes between MVI and non-MVI groups.

Distinct TECs subgroups had significant heterogeneity between datasets from MVI and non-MVI patients. MVI samples had TECs subgroups with increased levels of the epithelial-mesenchymal transition (EMT), endothelial cell migration and angiogenesis. Opposing EMT development was found in MVI TECs relative to non-MVI TECs. TM4SF1 was highly expressed in TECs undergoing the EMT and is thought to be linked to MVI.

TECs with elevated TM4SF1 expression facilitate MVI during HCC via an effect on the EMT, suggesting the potential of TM4SF1 as a therapeutic target.

Reportedly, there is an association between body metabolites and the risk of Hepatocellular Carcinoma (HCC) & Cholangiocarcinoma (CCA), possibly due to disrupted metabolic pathways leading to oxidative stress and an imbalance in cell proliferation and apoptosis, thereby increasing the risk of cancer. However, whether metabolites play a role in the onset of HCC or CCA remains inconclusive.

The aim of our study is to explore the potential causal relationship between metabolites and the risk of HCC&CCA.

Our study investigated the causal relationship between 1400 metabolites and HCC&CCA using publicly available genome-wide association study data. Single nucleotide polymorphisms (SNPs) associated with both metabolites and HCC&CCA were chosen as instrumental variables (IVs). The main approaches employed include inverse variance weighted (IVW), MR-Egger regression, and weighted median estimator (WME), with odds ratios (OR) used as the assessment criterion. Heterogeneity testing and sensitivity analyses were conducted to validate the results. We also conducted a reverse MR analysis to further validate the relationship between exposure and disease outcomes.

This Mendelian Randomization (MR) study indicates a significant causal relationship between 19 metabolites and the risk of HCC&CCA. Among them, the risk factors include "Bilirubin (E, Z or Z, E) levels," "Bilirubin (Z, Z) to taurocholate ratio," "Dimethylarginine (sdma + adma) levels," "N-methyltaurine levels," "4-vinylguaiacol sulfate levels," "Cholate to adenosine 3',5'-cyclic monophosphate (cAMP) ratio," "Glycohyocholate levels," "Cholesterol levels," and "4-methylguaiacol sulfate levels." The incidence risk of HCC and CCA increases with the elevation of these metabolites. Protective factors include "Ursodeoxycholate levels," "3-hydroxybutyroylglycine levels," "Linoleoylcholine levels," "Nonanoylcarnitine (C9) levels," "Pristanate levels," "Heptenedioate (C7:1-DC) levels," "Mannonate levels," "N-acetyl-L-glutamine levels," "Sphinganine levels," and "N-lactoyl isoleucine levels." The incidence risk of HCC and CCA potentially decreases as the levels of these metabolites increase. Heterogeneity tests show that most instrumental variables do not exhibit inter-gene heterogeneity, and the possibility of pleiotropy in the analysis is very low according to the sensitivity analysis. The reverse MR analysis did not yield positive results.

Our study has unveiled the intricate causal relationships between metabolites and the risk of HCC&CCA. Through our analysis, we identified nine metabolites, including "Bilirubin (E, Z or Z, E) levels," "Dimethylarginine (sdma + adma) levels," "Cholesterol levels,"ect, as risk factors for HCC&CCA. The incidence risk of HCC and CCA increases with their elevation. On the other hand, ten metabolites, such as "Ursodeoxycholate levels," "Linoleoylcholine levels," "Pristanate levels," ect, were identified as protective factors for HCC&CCA. The risk of developing HCC and CCA decreases with an increase in these metabolites. In conclusion, these findings further explore the physiological metabolic pathways underlying the pathogenesis of HCC and CCA, emphasizing future research directions. They pave the way for researchers to delve into the biological mechanisms of these diseases, facilitating early intervention and treatment strategies for these conditions.

To investigate the predictive value of radiomics models based on intra-tumoral ecological diversity (iTED) and temporal characteristics for assessing microvascular invasion (MVI) in patients with hepatocellular carcinoma (HCC).

We retrospectively analyzed the data of 398 HCC patients who underwent dynamic contrast-enhanced MRI with Gd-EOB-DTPA (training set: 318; testing set: 80). The tumors were segmented into five distinct habitats using case-level clustering and a Gaussian mixture model was used to determine the optimal clusters based on the Bayesian information criterion to produce an iTED feature vector for each patient, which was used to assess intra-tumoral heterogeneity. Radiomics models were developed using iTED features from the arterial phase (AP), portal venous phase (PVP), and hepatobiliary phase (HBP), referred to as MiTED-AP, MiTED-PVP, and MiTED-HBP, respectively. Additionally, temporal features were derived by subtracting the PVP features from the AP features, creating a delta-radiomics model (MDelta). Conventional radiomics features were also extracted from the AP, PVP, and HBP images, resulting in three models: MCVT-AP, MCVT-PVP, and MCVT-HBP. A clinical-radiological model (CR model) was constructed, and two fusion models were generated by combining the radiomics or/and CR models using a stacking algorithm (fusion_R and fusion_CR). Model performance was evaluated using AUC, accuracy, sensitivity, and specificity.

The MDelta model demonstrated higher sensitivity compared to the MCVT-AP and MCVT-PVP models. No significant differences in performance were observed across different imaging phases for either conventional radiomics (p = 0.096-0.420) or iTED features (p = 0.106-0.744). Similarly, for images from the same phase, we found no significant differences between the performance of conventional radiomics and iTED features (AP: p = 0.158; PVP: p = 0.844; HBP: p = 0.157). The fusion_R and fusion_CR models enhanced MVI discrimination, achieving AUCs of 0.823 (95% CI: 0.816-0.831) and 0.830 (95% CI: 0.824-0.835), respectively.

Delta radiomics features are temporal and predictive of MVI, providing additional predictive information for MVI beyond conventional AP and PVP features. The iTED features provide an alternative perspective in interpreting tumor characteristics and hold the potential to replace conventional radiomics features to some extent for MVI prediction.

Novel biomarkers reflecting liver fibrosis and the immune microenvironment may correlate with the risk of hepatocellular carcinoma (HCC) recurrence. This study aimed to evaluate the prognostic value of serum biomarkers in predicting HCC recurrence.

Serum biomarkers, including M2BPGi, IL-6, IL-10, CCL5, VEGF-A, soluble PD-1, PD-L1, TIM-3, and LAG-3, were measured in 247 patients with HCC undergoing surgical resection. Factors associated with recurrence-free survival (RFS) and overall survival (OS) were evaluated. The ERASL-post model and IMbrave050 criteria were used to define HCC recurrence risk groups.

Serum M2BPGi levels significantly correlated with FIB-4 score, aspartate transaminase-to-platelet ratio index, ALBI score, alpha-fetoprotein (AFP), alanine transaminase, aspartate transaminase, IL-10, CCL5, VEGF-A, soluble PD-1, PD-L1, TIM-3, and LAG-3 levels. M2BPGi, VEGF-A, soluble PD-1, and TIM-3 levels significantly correlated with RFS. In multivariate analysis, M2BPGi >1.5 COI (hazard ratio [HR] = 2.100, p < 0.001), tumor size >5 cm (HR = 1.859, p = 0.002), multiple tumors (HR = 2.562, p < 0.001), AFP >20 ng/mL (HR = 2.141, p < 0.001), and microvascular invasion (HR = 1.954, p = 0.004) were independent predictors of RFS. M2BPGi levels significantly stratified the recurrence risk in ERASL-post and IMbrave050 risk groups. An M2BPGi-based model could significantly discriminate RFS in the overall cohort as well as in the IMbrave050 low- and high-risk groups. M2BPGi >1.5 COI was also an independent predictor of OS after resection (HR = 2.707, p < 0.001).

Serum M2BPGi levels significantly correlated with surrogate markers of liver fibrosis, liver function, and immunology. M2BPGi is a significant predictor of HCC recurrence and survival after resection and could be incorporated into recurrence-prediction models.

Hepatocellular carcinoma (HCC) remains a significant global health burden, with unmet clinical needs despite the availability of multiple therapeutic options.

We present the case of an 85-year-old male diagnosed with HCC and bilateral lung metastases following hepatectomy. The patient responded favorably to treatment with icaritin and thalidomide, which resulted in a reduction in alpha-fetoprotein (AFP) levels and tumor size. This treatment achieved partial remission, with a progression-free survival (PFS) of 24 months and an overall survival (OS) of 33 months. Unfortunately, the patient ultimately passed away due to a cerebral infarction unrelated to cancer progression.

This case underscores the potential of icaritin as a therapeutic option for HCC patients with compromised health status. The combination of icaritin and thalidomide demonstrated promising efficacy in this real-world scenario. Multidisciplinary combination treatment strategies incorporating icaritin merit further exploration, given its immunomodulatory effects and favorable safety profile.

Large hepatocellular carcinoma (HCC) exhibits heterogeneous morphologies and varied responses to treatment. We evaluated outcomes of patients with different large HCC classifications receiving surgical resection (SR) or transarterial chemoembolization plus ablation (TA).

Patients with HCC ≥ 5 cm receiving SR or TA between May 2016 and December 2020 at one center were analyzed retrospectively and with propensity score matching (PSM). Overall survival (OS) and progression-free survival (PFS) of the 2 treatment groups were compared. Tumors were classified according to imaging morphology and gross pathology: Type I, simple nodular; Type II, simple nodular with extranodular growth or confluent multinodular; Type III, infiltrative.

Of 644 patients, 374 met the inclusion criteria (300 received SR and 74 received TA). Before PSM, median follow-up was 51.2 (IQR 29.6-65.3) months, and the SR group had longer OS (HR 2.13, 95% CI 1.44-3.15, p<0.001) and PFS (HR 2.31, 95% CI 1.66-3.20, p<0.001) than the TA group; after PSM these differences were not significant (all p>0.05). Infiltrative HCC (Type III) was an independent negative prognostic factor for OS and PFS. Within both treatment groups, patients with infiltrative HCC had shorter OS and PFS than patients with non-infiltrative HCC (Types I and II) (all p<0.001).

For patients with HCC ≥ 5 cm, tumor classification is an important prognostic factor. In patients with non-infiltrative HCC, TA and SR had comparable OS after PSM. For patients with infiltrative HCC, TA and SR had limited efficacy.

Accurate identification of microvascular invasion (MVI) in hepatocellular carcinoma (HCC) is crucial for treatment and prognosis. Single-modality and feature fusion models using manual segmentation fail to provide insights into MVI. This study aims to develop a DeepLab V3+ model for automated segmentation of HCC magnetic resonance (MR) images and a decision fusion model to predict MVI and early recurrence (ER).

This retrospective study included 209 HCC patients (146 in the training and 63 in the test cohorts). The performance of DeepLab V3+ for HCC MR image segmentation was evaluated using Dice Loss and F1 score. Intraclass correlation coefficients (ICCs) assessed feature extraction reliability. Spearman's correlation analyzed the relationship between tumor volumes from automated and manual segmentation, with agreement evaluated using Bland-Altman plots. Model performance was assessed using the area under the receiver operating characteristic curve (ROC AUC), calibration curves, and decision curve analysis. A nomogram predicted ER of HCC after surgery, with Kaplan-Meier analysis for 2-year recurrence-free survival (RFS).

The DeepLab V3+ model demonstrated high segmentation accuracy, with strong agreement in feature extraction (ICC: 0.802-0.999). The decision fusion model achieved AUCs of 0.968 and 0.878 for MVI prediction, and the nomogram for predicting ER yielded AUCs of 0.782 and 0.690 in the training and test cohorts, respectively, with significant RFS differences between the risk groups.

The DeepLab V3+ model accurately segmented HCC. The decision fusion model significantly improved MVI prediction, and the nomogram offered valuable insights into recurrence risk for clinical decision-making.

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

This study aimed to develop a novel nomogram to predict recurrence-free survival (RFS) for microvascular invasion (MVI)-negative hepatocellular carcinoma (HCC) patients after curative resection.

A total of 143 pathologically confirmed MVI-negative HCC patients were analyzed retrospectively. Baseline MRI features and inflammatory markers were collected. We used univariable and multivariable Cox regression analysis to identify the independent risk factors for RFS. And we established a nomogram based on significant MRI features and inflammatory marker. The receiver operating characteristic (ROC) curve, concordance index (C-index) and calibration curve were used to evaluate the predictive accuracy and discriminative ability of the nomogram. The decision curve analysis (DCA) was performed to validate the clinical utility of the nomogram.

In multivariate Cox regression analysis, neutrophil-to-lymphocyte ratio (NLR) (P = 0.018), tumor size (P = 0.002), and tumor capsule (P = 0.000) were independent significant variables associated with RFS. Nomogram with independent factors was developed and achieved a good C-index of 0.730 (95% confidence interval [CI]: 0.656-0.804) for predicting RFS. In ROC analysis, the areas under curve of the nomogram for 1-, 3- and 5-year RFS prediction were 0.725, 0.784 and 0.798, respectively. The risk score calculated by nomogram could divide MVI-negative HCC patients into high-risk group or low-risk group (P < 0.0001). DCA analysis revealed that the nomogram could increase net benefit and exhibited a wider range of threshold probabilities by the risk stratification than the independent risk factors in the prediction of MVI-negative HCC recurrence.

The nomogram prognostic model based on MRI features and NLR for predicting RFS showed high accuracy in MVI-negative HCC patients after curative resection. It can help clinicians make treatment decisions for MVI-negative HCC patients and identify high-risk patients for timely intervention.

To assess tumor progression in patients with hepatocellular carcinoma (HCC) without macrovascular invasion who underwent treatment with conventional transarterial chemoembolization (cTACE) based on microvascular invasion (MVI) risk within 2 years.

This retrospective investigation comprised adult patients with HCC who had either liver resection or cTACE as their first treatment from January 2016 to December 2021. A predictive model for MVI was developed and validated using preoperative clinical and MRI data from patients with HCC treated with liver resection. The MVI predictive model was applied to patients with HCC receiving cTACE, and differences in tumor progression between the MVI high- and low-risk groups were examined throughout 2 years.

The MVI prediction model incorporated nonsmooth margin, intratumoral artery, incomplete or absent tumor capsule, and tumor DWI/T2WI mismatch. The area under the receiver operating characteristic curve (AUC) for the prediction model, in the training cohort, was determined to be 0.904 (95% CI, 0.862-0.946), while in the validation cohort, it was 0.888 (0.782-0.994). Among patients with HCC undergoing cTACE, those classified as high risk for MVI possessed a lower rate of achieving a complete response after the first tumor therapy and a higher risk of tumor progression within 2 years.

The MVI prediction model developed in this study demonstrates a considerable degree of accuracy. Patients at high risk for MVI who underwent cTACE treatment exhibited a higher risk of tumor progression within 2 years.

Many patients with hepatocellular carcinoma (HCC) experience recurrence after curative-intent resection or ablation, with a poor prognosis. Real-world patterns of recurrence and the prognostic significance of early recurrence in U.S. clinical practice have not been well characterized.

This retrospective observational study was designed to evaluate the impact of recurrence on overall survival (OS) among patients with HCC following initial curative-intent resection or ablation. We used the Surveillance, Epidemiology, and End Results cancer registry linked with Medicare claims (January 1, 2010-December 31, 2019). Eligible patients (≥66 years) diagnosed with HCC (2010-2017) had liver resection or ablation within 180 days of diagnosis. Patients were stratified by recurrence status using diagnosis- and treatment-based definitions of recurrence. Early or late recurrence was defined as within 1 year or after 1 year, respectively. Adjusted OS analyses used multivariable Cox regression models.

A total of 1,146 patients were included. During a median overall follow-up of 35.2 months, 736 (64%) patients had a recurrence, of whom 380 (52%) had early recurrence (within 1 year). In the adjusted analysis, patients with recurrence had a 2.24-fold higher risk of death (95% confidence interval 1.85, 2.71; P < 0.001). Patients with early recurrence had a 1.39-fold higher risk of death (95% confidence interval 1.14, 1.68; P < 0.001) than those with late recurrence.

Recurrence and the timing of recurrence are significant predictors of increased mortality risk for patients with HCC following initial curative-intent resection or ablation, highlighting the need for effective adjuvant therapies that may delay or avoid recurrences.

To develop and validate a nomogram for predicting recurrence in individuals suffering single hepatocellular carcinoma (HCC) after curative hepatectomy.

A retrospective analysis was conducted on 189 patients with single HCC undergoing curative resection in our center were randomized into training and validation cohorts. P53 status was determined using immunohistochemistry. Clinical data, such as age, and gender were collected. MRI findings, such as tumor size, intratumoral arteries, the presence of peritumoral enhancement and intratumoral necrosis were also recorded. Nomograms were established based on the predictors selected in the training cohort, and receiver operating characteristic (ROC) curve analyses were used to compare the predictive ability among single predictors and nomogram model. The Kaplan-Meier method was used to assess the impact of each predictor and nomogram model on HCC recurrence. The results were validated in the validation cohort.

Multivariate Cox regression analysis showed that P53 (P < 0.001), tumor size (P = 0.009), and intratumoral artery (P = 0.026) were the independent risk factors for HCC recurrence. The nomogram model demonstrated favorable C-index of 0.740 (95 %CI:0.653-0.826) and 0.767 (95 %CI: 0.633-0.900) in the training and validation cohorts, and the areas under the curve was 0.740 and 0.752, which was better than the performance of P53 and MR factors alone. Calibration curves indicated a good agreement between observed actual outcomes and predicted values. Kaplan-Meier curves indicated that nomogram model was powerful in discrimination and clinical usefulness.

The integrated nomogram combining P53 status and MRI findings can be a valuable prognostic tool for predicting postoperative recurrence of single HCC.

Patients undergoing curative-intent surgery for hepato-pancreato-biliary (HPB) malignancies may achieve statistical cure i.e., a mortality risk which aligns with the general population.

To summarize the results of different cure models in HPB malignancies.

We conducted a systematic literature search and selected studies on curative-intent surgery (hepatic resection, HR, or liver transplantation, LT) for HPB malignancies including a cure model in their analysis. The review protocol was registered in PROSPERO (CRD42024528694).

Eleven studies reporting a cure model after HPB surgery for malignancy were included: 6 on hepatocellular carcinoma (HCC) two on biliary tract cancers (BTC), one on pancreatic neuroendocrine tumors (pNET), one on pancreatic ductal adenocarcinoma (PDAC), and one on colorectal liver metastases (CRLM). In terms of OS, the cure fraction of HCC is 63.4 %-75.8 % with LT and 31.8 %-40.5 % with HR, achieved within 7.2-10 years and 7-14.4 years respectively. The cure fraction of intrahepatic cholangiocarcinoma is 9.7 % in terms of DFS, but largely depends on tumor stage. PDAC and pNET display a cure fraction of 20.4 % and 57.1 % respectively in terms of DFS, confirming the impact of histotype on DFS.

Statistical cure for hepato-pancreato-biliary cancers can be achieved with surgery. The probability of cure depends on the interplay between tumor stage and aggressiveness, effectiveness of the surgical treatment and persistence of chronic conditions after surgery.

To evaluate the efficacy of re-resection in recurrent hepatocellular carcinoma (rHCC), identify prognostic factors, and provide clinical guidance.

A retrospective analysis was conducted on 130 rHCC patients undergoing re-resection and 60 primary HCC patients undergoing initial hepatectomy at Peking University People's Hospital (2014-2022). Disease-free survival (DFS) and overall survival (OS) were compared. Prognostic factors were identified using univariate and multivariate COX regression analyses.

Baseline characteristics were comparable between groups (P > 0.05). DFS was similar between groups (30.8 vs. 32.2 months, P = 0.612). The 1-year, 2-year, and 3-year DFS rates for the re-resection group were 88.5 %, 64.9 %, and 56.7 %, respectively, versus 88.3 %, 65.0 %, and 53.3 % for the primary resection group. OS was lower in the re-resection group (36.1 vs. 47.2 months, P = 0.041) with 1-year, 2-year, and 3-year OS rates of 90.8 %, 73.1 %, and 60.0 %, compared to 95.0 %, 80.0 %, and 68.3 % for the primary resection group. Significant factors affecting DFS were Child-Pugh classification (P = 0.044), time to recurrence (P = 0.002), tumor differentiation (P = 0.044), and satellite nodules (P = 0.019). Factors influencing OS included Child-Pugh classification (P = 0.040), time to recurrence (P = 0.002), and tumor differentiation (P = 0.032).

Re-resection is an effective treatment option for rHCC, with favorable outcomes as measured by DFS and OS, though OS is lower compared to initial hepatectomy. Key prognostic factors include Child-Pugh classification, time to recurrence, tumor differentiation, and satellite nodules.

Hepatocellular carcinoma (HCC) exhibits a propensity for early recurrence following liver resection, resulting in a bleak prognosis. At present, majority of the predictive models for the early postoperative recurrence of HCC rely on the linear assumption of the Cox Proportional Hazard (CPH) model. However, the predictive efficacy of this model is constrained by the intricate nature of clinical data. The present study aims to investigate the efficacy of the random survival forest (RSF) model, which is a machine learning algorithm, in predicting the early postoperative recurrence of HCC, and compare its performance with that of the traditional CPH model. This analysis seeks to elucidate the potential advantages of the RSF model over the CPH model in addressing this clinical challenge.

The present retrospective cohort study was conducted at a single center. After excluding 41 patients, a total of 541 patients were included in the final model construction and subsequent analysis. The patients were randomly divided into two groups at a 7:3 ratio: training group (n = 378) and validation group (n = 163). The least absolute shrinkage and selection operator (LASSO) regression was used to identify the risk factors in the training group. Then, the identified factors were used to develop the RSF and CPH regression models. The predictive ability of the model was assessed using the concordance index (C-index). The accuracy of the model predictions was evaluated using the receiver operating characteristic curve (ROC) and area under the receiver operating characteristic curve (AUC). The clinical practicality of the model was measured by decision curve analysis (DCA), and the overall performance of the model was evaluated using the Brier score. The RSF model was visually represented using the Shapley additive explanations (SHAP) framework. Then, the RSF, CPH regression, and albumin-bilirubin (ALBI) grade models were compared.

The following variables were examined by LASSO regression: alpha fetoprotein (AFP), gamma-glutamyl transpeptidase to platelet ratio (GPR), blood transfusion (BT), microvascular invasion (MVI), large vessel invasion (LVI), Edmondson-Steiner (ES) grade, liver capsule invasion (LCI), satellite nodule (SN), and Barcelona clinic liver cancer (BCLC) grade. Then, a RSF model was developed using 500 trees, and the variable importance (VIMP) ranking was MVI, LCI, SN, BT, BCLC, ESG, AFP, GPR and LVI. After these aforementioned factors were applied, the RSF and CPH regression models were developed and compared using the ALBI grade model. The C-index for the RSF model (0.896 and 0.798, respectively) outperformed that of the CPH regression model (0.803 and 0.772, respectively) and ALBI grade model (0.517 and 0.515, respectively), in both the training and validation groups. Three time points were selected to assess the predictive capabilities of these models: 6, 12 and 18 months. For the training group, the AUC value for the RSF model at 6, 12 and 18 months was 0.971 (95% CI: 0.955-0.988), 0.919 (95% CI: 0.887-0.951) and 0.899 (95% CI: 0.867-0.932), respectively. For the validation cohort, the AUC value for the RSF model at 6, 12 and 18 months was 0.830 (95% CI: 0.728-0.932), 0.856 (95% CI: 0.787-0.924) and 0.832 (95% CI: 0.764-0.901), respectively. The AUC values were higher in the RSF model, when compared to the CPH regression model and ALBI grade model, in both groups. The DCA results revealed that the net clinical benefits associated to the RSF model were superior to those associated to the CPH regression model and ALBI grade model in both groups, suggesting a higher level of clinical utility in the RSF model. The Brier score for the RSF model at 6, 12 and 18 months was 0.062, 0.125 and 0.178, respectively, in the training group, and 0.111, 0.128 and 0.149, respectively, in the validation group. In summary, the RSF model demonstrated superior performance, when compared to the CPH regression model and ALBI grade model. Furthermore, the RSF model demonstrated superior predictive ability, accuracy, clinical practicality, and overall performance, when compared to the CPH regression model and ALBI grade model. In addition, the RSF model was able to successfully stratify patients into three distinct risk groups (low-risk, medium-risk and high-risk) in both groups (p < 0.001).

The RSF model demonstrates efficacy in predicting early recurrence following HCC surgery, exhibiting superior performance, when compared to the CPH regression model and ALBI grade model. For patients undergoing HCC surgery, the RSF model can serve as a valuable tool for clinicians to postoperatively stratify patients into distinct risk categories, offering guidance for subsequent follow-up care.

HCC recurrence frequently occurs after curative surgery. Histological microvascular invasion (MVI) predicts recurrence but cannot provide preoperative prognostication, whereas clinical prediction scores have variable performances.

Recurr-NET, a multimodal multiphasic residual-network random survival forest deep-learning model incorporating preoperative CT and clinical parameters, was developed to predict HCC recurrence. Preoperative triphasic CT scans were retrieved from patients with resected histology-confirmed HCC from 4 centers in Hong Kong (internal cohort). The internal cohort was randomly divided in an 8:2 ratio into training and internal validation. External testing was performed in an independent cohort from Taiwan.Among 1231 patients (age 62.4y, 83.1% male, 86.8% viral hepatitis, and median follow-up 65.1mo), cumulative HCC recurrence rates at years 2 and 5 were 41.8% and 56.4%, respectively. Recurr-NET achieved excellent accuracy in predicting recurrence from years 1 to 5 (internal cohort AUROC 0.770-0.857; external AUROC 0.758-0.798), significantly outperforming MVI (internal AUROC 0.518-0.590; external AUROC 0.557-0.615) and multiple clinical risk scores (ERASL-PRE, ERASL-POST, DFT, and Shim scores) (internal AUROC 0.523-0.587, external AUROC: 0.524-0.620), respectively (all p < 0.001). Recurr-NET was superior to MVI in stratifying recurrence risks at year 2 (internal: 72.5% vs. 50.0% in MVI; external: 65.3% vs. 46.6% in MVI) and year 5 (internal: 86.4% vs. 62.5% in MVI; external: 81.4% vs. 63.8% in MVI) (all p < 0.001). Recurr-NET was also superior to MVI in stratifying liver-related and all-cause mortality (all p < 0.001). The performance of Recurr-NET remained robust in subgroup analyses.

Recurr-NET accurately predicted HCC recurrence, outperforming MVI and clinical prediction scores, highlighting its potential in preoperative prognostication.

Assessment of recurrence risk after liver resection (LR) is critical in hepatocellular carcinoma (HCC), particularly with the advent of effective adjuvant therapy. The aim of this study was to analyze the clinical and pathological factors associated with recurrence, aggressive recurrence, and survival after LR.

We performed a retrospective study in which all single HCC (BCLC-0/A) patients treated with LR between February 2000 and November 2020 were included. The main clinical variables were recorded. Histological features were blindly evaluated by two independent pathologists. Aggressive recurrence was defined as those that exceeded the Milan criteria at 1st recurrence.

A total of 218 patients were included (30% BCLC 0 and 70% BCLC A), median (IQR) tumor size of 28 (19-42 mm). The prevalence of microvascular invasion and/or satellitosis (mVI/S) was 39%, with a kappa-index between both pathologists of 0.8. After a median follow-up of 49 (23-85) months, 61/218 (28%) patients died, 32/218 (15%) underwent liver transplantation, 127 (58%) developed HCC recurrence. The prevalence of aggressive recurrence was 35% (44/127 Milan-out, with 20 cases at advanced stage), and the 5-year survival rate was 81%. The presence of mVI/S was the only independent predictor of recurrence (hazard ratio [HR] 1.83, 95% CI 1.28-2.61, p <0.001), aggressive recurrence (HR 3.31, 95% CI 1.74-6.29, p <0.001) and mortality (HR 2.23, 95% CI 1.27-3.91, p = 0.005). The macrotrabecular-massive subtype was significantly associated with a higher prevalence of mVI/S, Edmonson Steiner grade III-IV, AFP values and vessels that encapsulate tumor clusters, but not with recurrence, aggressive recurrence, or overall survival.

The presence of mVI/S was the only independent risk factor for aggressive recurrence and mortality. This has important implications for early-stage patient management, especially in the setting of adjuvant immunotherapy or ab initio LT.

Assessment of recurrence risk after liver resection is crucial in patients with hepatocellular carcinoma. Patients with a high risk of recurrence are candidates for liver transplantation as an ab initio indication or for the potential use of adjuvant therapy. Aggressive recurrences, defined as those exceeding the Milan criteria at first recurrence, have a significant impact on overall survival (OS). Fifty-eight percent of patients experienced hepatocellular carcinoma recurrence, with a prevalence of aggressive recurrence at the first occurrence standing at 35%. After a median follow-up of 49 (23-85) months, 61 (28%) patients died, and 32 (15%) underwent liver transplantation, resulting in a 5-year OS rate of 81%. Microvascular invasion and/or satellitosis was present in 39% of our cohort and was the only independent predictor of recurrence, aggressive recurrence, and OS on multivariate analysis. This is important as it could be used to guide therapeutic management.

Narrow resection margin hepatocellular carcinoma (NRM-HCC) has a high incidence of early recurrence. Our study was designed to identify prognostic factors in patients with NRM-HCC, establish and validate a nomogram model to predict early recurrence of NRM-HCC patients.

We retrospectively analyzed data from 2957 NRM-HCC patients who underwent radical hepatectomy at three medical centers between December 2009 and January 2015. Patients were randomly assigned to a training cohort (n = 2069) and a validation cohort (n = 888). Using univariate and multivariate COX regression to determine early relapse factors in NRM-HCC patients, and used these factors to construct a nomogram. The accuracy of the prediction was evaluated using the C-index, receiver operating characteristic (ROC) and calibration curve. Decision curve analysis (DCA) assessed the predictive value of the models. Finally, the recurrence-free survival of different risks was analyzed using Kaplan-Meier (K-M) method.

The nomogram of NRM model contains alpha-fetoprotein (AFP), alkaline phosphatase (ALP), tumor size, tumor number, microvascular invasion (MVI), tumor capsular, and satellite nodules. The model shows good discrimination with C-indexes of 0.71 (95% CI: 0.69-0.72) and 0.72 (95% CI: 0.70-0.75) in the train cohort and test cohort respectively. Decision curve analysis demonstrated that the model is clinically useful and the calibration of our model was favorable. Our model stratified patients into two different risk groups, which exhibited significantly different early recurrence. The web-based tools are convenient for clinical practice.

NRM model demonstrated favorable performance in predicting early recurrence in NRM-HCC patients. This novel model will be helpful to guide postoperative follow-up and adjuvant therapy.

To explore the role of deep learning (DL) and radiomics-based integrated approach based on contrast enhanced magnetic resonance imaging (CEMRI) for predicting early recurrence (ER) in hepatocellular carcinoma (HCC) patients after curative resection.

Total 165 HCC patients (ER, n = 96 vs. non-early recurrence (NER), n = 69) were retrospectively collected and divided into a training cohort (n = 132) and a validation cohort (n = 33). From pretreatment CEMR images, a total of 3111 radiomics features were extracted, and radiomics models were constructed using five machine learning classifiers (logistic regression, support vector machine, k-nearest neighbor, extreme gradient Boosting, and multilayer perceptron). DL models were established via three variations of ResNet architecture. The clinical-radiological (CR), radiomics combined with clinical-radiological (RCR), and deep learning combined with RCR (DLRCR) models were constructed. Model discrimination, calibration, and clinical utilities were evaluated by receiver operating characteristic curve, calibration curve, and decision curve analysis, respectively. The best-performing model was compared with the widely used staging systems and preoperative prognostic indexes.

The RCR model (area under the curve (AUC): 0.841 and 0.811) and the optimal radiomics model (AUC: 0.839 and 0.804) achieved better performance than the CR model (AUC: 0.662 and 0.752) in the training and validation cohorts, respectively. The optimal DL model (AUC: 0.870 and 0.826) outperformed the radiomics model in the both cohorts. The DL, radiomics, and CR predictors (aspartate aminotransferase (AST) and tumor diameter) were combined to construct the DLRCR model. The DLRCR model presented the best performance over any model, yielding an AUC, an accuracy, a sensitivity, a specificity of 0.917, 0.886, 0.889, and 0.882 in the training cohort and of 0.844, 0.818, 0.800, and 0.846 in the validation cohort, respectively. The DLRCR model achieved better clinical utility compared to the clinical staging systems and prognostic indexes.

Both radiomics and DL models derived from CEMRI can predict HCC recurrence, and DL and radiomics-based integrated approach can provide a more effective tool for the precise prediction of ER for HCC patients undergoing resection.