Previous studies mostly use single-type features to establish a prediction model. We aim to develop a comprehensive prediction model that effectively identify patients with poor prognosis for single hepatocellular carcinoma (HCC) based on artificial intelligence (AI). Patients and methods: 236 single HCC patients were studied to establish a comprehensive prediction model. We collected the basic information of patients and used AI to extract the features of magnetic resonance (MR) images.

The clinical model based on linear regression (LR) algorithm (AUC: 0.658, 95%CI: 0.5021-0.8137), the radiomics model and deep transfer learning (DTL) model based on light gradient-boosting machine (Light GBM) algorithm (AUC: 0.761, 95%CI: 0.6326-0.8886 and AUC: 0.784, 95%CI: 0.6587-0.9087, respectively) were the optimal prediction models. A comparison revealed that the integrated nomogram had the largest area under the receiver operating characteristic curve (AUC) (all P < 0.05). In the training cohort, the integrated nomogram was predictive of recurrence-free survival (RFS) as well as overall survival (OS) (C-index: 0.735 and 0.712, P < 0.001). In the test cohort, the integrated nomogram also can predict RFS and OS (C-index: 0.718 and 0.740, P < 0.001) in patients.

The integrated nomogram composed of signatures in the prediction models can not only predict the postoperative recurrence of single HCC patients but also stratify the risk of OS after the operation.

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.

The prognostic impact of preoperative gamma-glutamyl transpeptidase to platelet ratio (GPR) levels in patients with solitary hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC) following radical resection has not been established.

To examine the clinical utility of GPR for prognosis prediction in solitary HBV-related HCC patients.

A total of 1167 solitary HBV-related HCC patients were retrospectively analyzed. GPR levels were compared with 908 non-HCC individuals. Overall survival (OS) and recurrence-free survival (RFS) were evaluated, and cox proportional hazard model analyses were performed to identify independent risk factors. Differences in characteristics were adjusted by propensity score matching (PSM). Subgroup and stratified survival analyses for HCC risks were performed, and a linear trend of the hazard ratio (HR) according to GPR levels was constructed.

GPR levels of patients with solitary HBV-related HCC were higher than those with hepatic hemangiomas, chronic hepatitis B and healthy control (adjusted P < 0.05). Variable bias was diminished after the PSM balance test. The low GPR group had improved OS (P < 0.001) and RFS (P < 0.001) in the PSM analysis and when combined with other variables. Multivariate cox analyses suggested that low GPR levels were associated with a better OS (HR = 0.5, 95%CI: 0.36-0.7, P < 0.001) and RFS (HR = 0.57, 95%CI: 0.44-0.73, P < 0.001). This same trend was confirmed in subgroup analyses. Prognostic nomograms were constructed and the calibration curves showed that GPR had good survival prediction. Moreover, stratified survival analyses found that GPR > 0.6 was associated with a worse OS and higher recurrence rate (P for trend < 0.001).

Preoperative GPR can serve as a noninvasive indicator to predict the prognosis of patients with solitary HBV-related HCC.

The liver function reserve has a significant impact on the therapeutic effects of anti-programmed cell death-1 (PD-1) for hepatocellular carcinoma (HCC). This study aimed to comprehensively evaluate the ability of liver-function-based indicators to predict prognosis and construct a novel prognostic score for HCC patients with anti-PD-1 immunotherapy.

Between July 2018 and January 2020, patients diagnosed with HCC who received anti-PD-1 treatment were screened for inclusion in the study. The valuable prognostic liver-function-based indicators were selected using Cox proportional hazards regression analysis to build a novel liver-function-indicators-based signature (LFIS). Concordance index (C-index), the area under the receiver operating characteristic (ROC) curve (AUC), and Kaplan-Meier survival curves were utilized to access the predictive performance of LFIS.

A total of 434 HCC patients who received anti-PD-1 treatment were included in the study. The LFIS, based on alkaline phosphatase-to-albumin ratio index, Child-Pugh score, platelet-albumin score, aspartate aminotransferase-to-lymphocyte ratio index, and gamma-glutamyl transpeptidase-to-lymphocyte ratio index, was constructed and identified as an independent risk factor for patient survival. The C-index of LFIS for overall survival (OS) was 0.692, which was higher than the other single liver-function-based indicator. The AUC of LFIS at 6-, 12-, 18-, and 24-month were 0.74, 0.714, 0.747, and 0.865 for OS, respectively. Patients in the higher-risk LFIS group were associated with both worse OS and PFS. An online and easy-to-use calculator was further constructed for better application of the LFIS signature.

The LFIS score had an excellent prognosis prediction ability superior to every single liver-function-based indicator for anti-PD-1 treatment in HCC patients. It is a reliable, easy-to-use tool to stratify risk for OS and PFS in HCC patients who received anti-PD-1 treatment.

Hepatocellular carcinoma (HCC) presents a significant threat to individuals and healthcare systems due to its high recurrence rate. Accurate prognostic models are essential for improving patient outcomes. Gamma-glutamyl transpeptidase (GGT) and prealbumin (PA) are biomarkers closely related to HCC. This study aimed to investigate the predictive value of the GGT to PA ratio (GPR) and to construct prognostic nomograms for HCC patients without microvascular invasion.

We retrospectively analyzed data from 355 HCC patients who underwent radical hepatectomy at Shengjing Hospital of China Medical University between December 2012 and January 2021. Patients were randomly assigned to a training cohort (n = 267) and a validation cohort (n = 88). The linearity of GPR was assessed using restricted cubic spline (RCS) analysis, and the optimal cut-off value was determined by X-tile. Kaplan-Meier survival curves and log-rank tests were used to investigate the associations between GPR and both progression-free survival (PFS) and overall survival (OS). Cox multivariate regression analysis identified independent risk factors, enabling the construction of nomograms. Time-dependent receiver operating characteristic (ROC) and calibration curves were used to evaluate the accuracy of the nomograms. Decision curve analysis (DCA) assessed the predictive value of the models.

Patients were categorized into GPR-low and GPR-high groups based on a GPR value of 333.33. Significant differences in PFS and OS were observed between the two groups (both P < 0.001). Cox multivariate analysis identified GPR as an independent risk factor for both PFS (OR = 1.80, 95% CI: 1.24-2.60, P = 0.002) and OS (OR = 1.87, 95% CI: 1.07-3.26, P = 0.029). The nomograms demonstrated good predictive performance, with C-index values of 0.69 for PFS and 0.76 for OS. Time-dependent ROC curves and calibration curves revealed the accuracy of the models in both the training and validation cohorts, with DCA results indicating notable clinical value.

GPR emerged as an independent risk factor for both OS and PFS in HCC patients without microvascular invasion. The nomograms based on GPR demonstrated relatively robust predictive efficiency for prognosis.

Although patients with alpha-fetoprotein-negative hepatocellular carcinoma (AFPNHCC) have a favorable prognosis, a high risk of postoperative recurrence remains. We developed and validated a novel liver fibrosis assessment index, the direct bilirubin-gamma-glutamyl transpeptidase-to-platelet ratio (DGPRI). DGPRI was calculated for each of the 378 patients with AFPNHCC who underwent hepatic resection. The patients were divided into high- and low-score groups using the optimal cutoff value. The Lasso-Cox method was used to identify the characteristics of postoperative recurrence, followed by multivariate Cox regression analysis to determine the independent risk factors associated with recurrence. A nomogram model incorporating the DGPRI was developed and validated. High DGPRI was identified as an independent risk factor (hazard ratio = 2.086) for postoperative recurrence in patients with AFPNHCC. DGPRI exhibited better predictive ability for recurrence 1-5 years after surgery than direct bilirubin and the gamma-glutamyl transpeptidase-to-platelet ratio. The DGPRI-nomogram model demonstrated good predictive ability, with a C-index of 0.674 (95% CI 0.621-0.727). The calibration curves and clinical decision analysis demonstrated its clinical utility. The DGPRI nomogram model performed better than the TNM and BCLC staging systems for predicting recurrence-free survival. DGPRI is a novel and effective predictor of postoperative recurrence in patients with AFPNHCC and provides a superior assessment of preoperative liver fibrosis.

This study aimed to assess the prognostic and predictive value of a circulating hematological signature (CHS) and to develop a CHS-based nomogram for predicting prognosis and guiding individualized chemotherapy in non-metastatic nasopharyngeal carcinoma (NPC) patients.

NPC patients were recruited between January 2014 and December 2017 at the Jiangxi Cancer Hospital. The CHS was constructed based on a series of hematological indicators. The nomogram was developed by CHS and clinical factors.

A total of 779 patients were included. Three biomarkers were selected by least absolute shrinkage and selection operator regression, including prognostic nutritional index, albumin-to-fibrinogen ratio, and prealbumin-to-fibrinogen ratio, were used to construct the CHS. The patients in the low-CHS group had better 5-year DMFS and OS than those in the high-CHS group in the training (DMFS: 85.0% vs 56.6%, p<0.001; OS: 90.3% vs 65.4%, p<0.001) and validation cohorts (DMFS: 92.3% vs 43.6%, p<0.001; OS: 92.1% vs 65.5%, p<0.001). The nomogram_CHS showed better performance than clinical stage in predicting distant metastasis (concordance index: 0.728 vs 0.646). In the low-TRS (total risk scores) group, the patients received RT alone, CCRT and IC plus CCRT had similar 5-year DMFS and OS (p>0.05). In the middle-TRS group, the patients received RT alone had worse 5-year DMFS (58.7% vs 80.8% vs 90.8%, p=0.002) and OS (75.0% vs 94.1% vs 95.0%, p=0.001) than those received CCRT or IC plus CCRT. In the high-TRS group, the patients received RT alone and CCRT had worse 5-year DMFS (18.6% vs 31.3% vs 81.5%, p<0.001) and OS (26.9% vs 53.2% vs 88.8%, p<0.001) than those received IC plus CCRT.

The developed nomogram_CHS had satisfactory prognostic accuracy in NPC patients and may individualize risk estimation to facilitate the identification of suitable IC candidates.

We initiated this multicenter study to integrate important risk factors to create a nomogram for hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC) for clinician decision-making.

Between April 2011 and March 2022, 2281 HCC patients with an HBV-related diagnosis were included. All patients were randomly divided into two groups in a ratio of 7:3 (training cohort, n = 1597; validation cohort, n = 684). The nomogram was built in the training cohort via Cox regression model and validated in the validation cohort.

Multivariate Cox analyses revealed that the portal vein tumor thrombus, Child-Pugh class, tumor diameter, alanine aminotransferase level, tumor number, extrahepatic metastases, and therapy were independent predictive variables impacting overall survival. We constructed a new nomogram to predict 1-, 2-, and 3-year survival rates based on these factors. The nomogram-related receiver operating characteristics (ROC) curves indicated that the area under the curve (AUC) values were 0.809, 0.806, and 0.764 in predicting 1-, 2-, and 3-year survival rates, respectively. Furthermore, the calibration curves revealed good agreement between real measurements and nomogram predictions. The decision curve analyses (DCA) curves demonstrated excellent therapeutic application potential. In addition, stratified by risk scores, low-risk groups had longer median OS than medium-high-risk groups (p < 0.001).

The nomogram we constructed showed good performance in predicting the 1-year survival rate for HBV- related HCC.

HALP score consisting of hemoglobin content, albumin content, lymphocyte count, and platelet count can comprehensively evaluate the inflammatory response and nutritional status. Many researchers have indicated that the HALP score is an effective predictor of the overall prognosis of various tumors. However, there is no relevant research to suggest whether the HALP score can predict the prognosis of patients with hepatocellular carcinoma (HCC).

We retrospectively analyzed 273 HCC patients who underwent surgical resection. Hemoglobin content, albumin content, lymphocyte count, and platelet count in peripheral blood were measured for each patient. The relationship between the HALP score and overall survival (OS) was investigated.

With a mean of 56.69 ± 1.25 months follow-up, the 1-, 3-, and 5-year OS was 98.9%, 76.9%, and 55.3% for all patients, respectively. HALP scores (HR=1.708, 95% CI=1.192-2.448, P=0.004) were significant independent risk factors of OS. The 1-, 3-, and 5-year OS were 99.3%, 84.3%, and 63.4% for patients with high HALP scores; and 98.6%, 69.8%, and 47.5% for patients with low HALP scores, respectively (P=0.018). In TNM I-II stage patients, compared with high HALP scores, low HALP scores have worse OS (P=0.039). In AFP positive patients, compared with high HALP scores, low HALP scores have worse OS (P=0.042).

Our research showed the preoperative HALP score is an independent predictive factor of overall prognosis, and a low HALP score indicates a worse prognosis in HCC patients who underwent surgical resection.

Hepatocellular carcinoma (HCC) is the predominant form of primary liver cancer. Although many surgical and nonsurgical therapeutic options have been established for treating HCC, the overall prognosis for HCC patients receiving different treatment modalities remains inadequate, which causes HCC to remain among the most life-threatening human cancers worldwide. Therefore, it is vitally important and urgently needed to develop valuable and independent prognostic biomarkers for the early prediction of poor prognosis in HCC patients, allowing more time for more timely and appropriate treatment to improve the survival of patients. As the most abundant protein in plasma, human serum albumin (ALB) is predominantly expressed by the liver and exhibits a wide variety of essential biological functions. It has been well recognized that serum ALB level is a significant independent biomarker for a broad spectrum of human diseases including cancer. Moreover, ALB has been commonly used as a potent biomaterial and therapeutic agent in clinical settings for the treatment of various human diseases. This review provides a comprehensive summary of the evidence from the up-to-date published literature to underscore the prognostic significance of serum ALB level and various ALB-based mono- and combination biomarkers in the prediction of the prognosis of HCC patients after treatment with different surgical, locoregional, and systemic therapies.

Most microvascular invasion (MVI)-predicting models have not considered MVI classification, and thus do not reflect true MVI effects on prognosis of patients with hepatocellular carcinoma (HCC). We aimed to develop a novel MVI-predicting model focused on MVI classification, hoping to provide useful information for clinical treatment strategy decision-making.

A retrospective study was conducted with data from two Chinese medical centers for 800 consecutive patients with HCC (derivation cohort) and 250 matched patients (external validation cohort). MVI-associated variables were identified by ordinal logistic regression. Predictive models were constructed based on multivariate analysis results and validated internally and externally. The models' discriminative ability and calibration ability were examined.

Four factors associated independently with MVI: tumor diameter, tumor number, serum lactate dehydrogenase (LDH) ≥ 176.58 U/L, and γ-glutamyl transpeptidase (γ-GGT). Area under the curve (AUC)s for our M2, M1, and M0 nomograms were 0.864, 0.648, and 0.782. Internal validation of all three models was confirmed with AUC analyses in D-sets (development datasets) and V-sets (validation datasets) and C-indices for each cohort. GiViTI calibration belt plots and Hosmer-Lemeshow (HL) chi-squared calibration values demonstrated good consistency between observed frequencies and predicted probabilities for the M2 and M0 nomograms. Although the M1 nomogram was well calibrated, its discrimination was poor.

We developed and validated MVI prediction models in patients with HCC that differentiate MVI classification and may provide useful guidance for treatment planning.