Establishment and Validation of SSCLIP Scoring System to Estimate Survival in Hepatocellular Carcinoma Patients Who Received Curative Liver Resection

Construction of a random survival forest model based on a machine learning algorithm to predict early recurrence after hepatectomy for adult hepatocellular carcinoma

BACKGROUND AND AIMS

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.

METHODS

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.

RESULTS

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).

CONCLUSIONS

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.

AI-powered prediction of HCC recurrence after surgical resection: Personalised intervention opportunities using patient-specific risk factors

BACKGROUND

Hepatocellular carcinoma (HCC) recurrence following surgical resection remains a significant clinical challenge, necessitating reliable predictive models to guide personalised interventions. In this study, we sought to harness the power of artificial intelligence (AI) to develop a robust predictive model for HCC recurrence using comprehensive clinical datasets.

METHODS

Leveraging data from 958 patients across multiple centres in Australia and Hong Kong, we employed a multilayer perceptron (MLP) as the optimal classifier for model generation.

RESULTS

Through rigorous internal cross-validation, including a cohort from the Chinese University of Hong Kong (CUHK), our AI model successfully identified specific pre-surgical risk factors associated with HCC recurrence. These factors encompassed hepatic synthetic function, liver disease aetiology, ethnicity and modifiable metabolic risk factors, collectively contributing to the predictive synergy of our model. Notably, our model exhibited high accuracy during cross-validation (.857 ± .023) and testing on the CUHK cohort (.835), with a notable degree of confidence in predicting HCC recurrence within accurately classified patient cohorts. To facilitate clinical application, we developed an online AI digital tool capable of real-time prediction of HCC recurrence risk, demonstrating acceptable accuracy at the individual patient level.

CONCLUSION

Our findings underscore the potential of AI-driven predictive models in facilitating personalised risk stratification and targeted interventions to mitigate HCC recurrence by identifying modifiable risk factors unique to each patient. This model aims to aid clinicians in devising strategies to disrupt the underlying carcinogenic network driving recurrence.

Development and comprehensive validation of a predictive prognosis model for very early HCC recurrence within one year after curative resection: a multicenter cohort study

BACKGROUND

The high incidence of early recurrence after liver resection (LR) for hepatocellular carcinoma (HCC) is the main obstacle in achieving good long-term survival outcomes. The aim of the present study is to develop a prognostic model in predicting the risk of very early (1-year) recurrence.

MATERIAL AND METHODS

Consecutive patients who underwent LR for HCC with curative intent at multicenters in China were enrolled in this study. The VERM-pre (the Preoperative Very Early Recurrence Model of HCC) with good performance was derived and validated by internal and external cohorts retrospectively and by another two-center cohort prospectively.

RESULTS

Seven thousand four hundred one patients were enrolled and divided randomly into three cohorts. Eight variables (tumor diameter, tumor number, macrovascular invasion, satellite nodule, alpha-fetoprotein, level of HBV-DNA, γ-GT, and prothrombin time) were identified as independent risk factors for recurrence-free survival on univariate and multivariate analyses. The VERM-pre model was developed which showed a high capacity of discrimination (C-index: 0.722; AUROC at 1-year: 0.722)) and was validated comprehensively by the internal, external, and prospective cohorts, retrospectively. Calibration plots showed satisfactory fitting of probability of early HCC recurrence in the cohorts. Three risk strata were derived to have significantly different recurrence-free survival rates (low-risk: 80.4-85.4%; intermediate-risk: 59.7-64.8%; high-risk: 32.6-42.6%). In the prospective validation cohort, the swimming plot illustrated consistent outcomes with the beginning predictive score.

CONCLUSION

The VERM-pre model accurately predicted the 1-year recurrence rates of HCC after LR with curative intent. The model was retrospectively and prospectively validated and then developed as the online tool.

Preoperative Radiomics Nomogram Based on CT Image Predicts Recurrence-Free Survival After Surgical Resection of Hepatocellular Carcinoma

RATIONALE AND OBJECTIVES

To construct preoperative models based on CT radiomics, radiologic and clinical features to predict recurrence-free survival (RFS) after liver resection (LR) of BCLC 0 to B stage hepatocellular carcinoma (HCC) and to classify the prognosis.

MATERIALS AND METHODS

This study retrospectively analyzed 161 HCC patients who underwent radical LR. Two methods, the least absolute shrinkage and selection operator and random survival forest analysis, were performed for radiomics signature (RS) construction. Univariate and multivariate stepwise Cox regression analyses were performed to establish a combined nomogram (RCN) of RS and clinical parameters and a clinical nomogram (CN). The performance of the models was assessed comprehensively using Harrell's concordance index (C-index), the calibration curve, and decision curve analysis. The discrimination accuracy of the models was compared using integrated discrimination improvement index (IDI). The risk stratification effect was assessed with Kaplan-Meier survival analysis and subgroup analysis.

RESULTS

The RCN achieved a C-index of 0.792/0.758 in the training/validation set, which was higher than the CN, RS, and BCLC stage system. The discriminatory accuracy of the RCN was improved when compared to the CN, RS, and BCLC staging systems (IDI > 0). Decision curve analysis reflected the clinical net benefit of the RCN. The RCN allows risk stratification of patients in different clinical subgroups.

CONCLUSION

The integrated model combining RS and clinical factors can more effectively predict RFS after LR of BCLC 0 to B stage HCC patients and can effectively stratify the prognostic risk.

Development of ensemble learning models for prognosis of hepatocellular carcinoma patients underwent postoperative adjuvant transarterial chemoembolization

Background

Postoperative adjuvant transarterial chemoembolization (PA-TACE) has been increasing widely used to improve the prognosis of hepatocellular carcinoma (HCC) patients. However, clinical outcomes vary from patient to patient, which calls for individualized prognostic prediction and early management.

Methods

A total of 274 HCC patients who underwent PA-TACE were enrolled in this study. The prediction performance of five machine learning models was compared and the prognostic variables of postoperative outcomes were identified.

Results

Compared with other machine learning models, the risk prediction model based on ensemble learning strategies, including Boosting, Bagging, and Stacking algorithms, presented better prediction performance for overall mortality and HCC recurrence. Moreover, the results showed that the Stacking algorithm had relatively low time consumption, good discriminative ability, and the best prediction performance. In addition, according to time-dependent ROC analysis, the ensemble learning strategies were found to perform well in predicting both OS and RFS for the patients. Our study also found that BCLC Stage, hsCRP/ALB and frequency of PA-TACE were relatively important variables in both overall mortality and recurrence, while MVI contributed more to the recurrence of the patients.

Conclusion

Among the five machine learning models, the ensemble learning strategies, especially the Stacking algorithm, could better predict the prognosis of HCC patients following PA-TACE. Machine learning models could also help clinicians identify the important prognostic factors that are clinically useful in individualized patient monitoring and management.

Understanding the causes of recurrent HCC after liver resection and radiofrequency ablation

INTRODUCTION

Surgical resection and radiofrequency ablation are preferred options for early-stage disease, with 5-year recurrence rates as high as 70% when patients are treated according to guidelines. With increasing availability of therapeutic options, including but not limited to, immune-checkpoint inhibitors (ICI), tyrosine kinase inhibitors, antiangiogenics, and adoptive cell therapies, understanding the causes of recurrence and identifying its predictors should be priorities in the hepatocellular carcinoma (HCC) research agenda.

AREAS COVERED

Current knowledge of HCC predictors of recurrence is reviewed, and recent insights about its underlying mechanisms are presented. In addition, results from recent clinical trials investigating treatment combinations are critically appraised.

EXPERT OPINION

HCC recurrence is either due to progressive growth of microscopic residual disease, or to de novo cancer development in the context of a diseased liver, each occurring in an early (<2years) vs. late (≥2 years) fashion. Collectively, morphological, proteomic, and transcriptomic data suggest vascular invasion and angiogenesis as key drivers of HCC recurrence. Agents aimed at blocking either of these two hallmarks should be prioritized at the moment of early-stage HCC clinical trial design. Emerging results from clinical trials testing ICI in early-stage HCC underscore the importance of defining the best treatment sequence and the most appropriate combination strategies. Lastly, as different responses to systemic therapies are increasingly defined according to the HCC etiology, patient enrolment into clinical trials should take into account the biological characteristics of their inherent disease.

Predictors of early and late hepatocellular carcinoma recurrence

Hepatocellular carcinoma (HCC) is the most frequent liver neoplasm, and its incidence rates are constantly increasing. Despite the availability of potentially curative treatments (liver transplantation, surgical resection, thermal ablation), long-term outcomes are affected by a high recurrence rate (up to 70% of cases 5 years after treatment). HCC recurrence within 2 years of treatment is defined as "early" and is generally caused by the occult intrahepatic spread of the primary neoplasm and related to the tumor burden. A recurrence that occurs after 2 years of treatment is defined as "late" and is related to de novo HCC, independent of the primary neoplasm. Early HCC recurrence has a significantly poorer prognosis and outcome than late recurrence. Different pathogenesis corresponds to different predictors of the risk of early or late recurrence. An adequate knowledge of predictive factors and recurrence risk stratification guides the therapeutic strategy and post-treatment surveillance. Patients at high risk of HCC recurrence should be referred to treatments with the lowest recurrence rate and when standardized to combined or adjuvant therapy regimens. This review aimed to expose the recurrence predictors and examine the differences between predictors of early and late recurrence.

Novel model combining contrast-enhanced ultrasound with serology predicts hepatocellular carcinoma recurrence after hepatectomy

BACKGROUND

Surgery is the primary curative option in patients with hepatocellular carcinoma (HCC). However, recurrence within 2 years is observed in 30%-50% of patients, being a major cause of mortality.

AIM

To construct and verify a non-invasive prediction model combining contrast-enhanced ultrasound (CEUS) with serology biomarkers to predict the early recurrence of HCC.

METHODS

Records of 744 consecutive patients undergoing first-line curative surgery for HCC in one institution from 2016-2018 were reviewed, and 292 local patients were selected for analysis. General characteristics including gender and age, CEUS liver imaging reporting and data system (LIRADS) parameters including wash-in time, wash-in type, wash-out time, and wash-out type, and serology biomarkers including alanine aminotransferase, aspartate aminotransferase, platelets, and alpha-fetoprotein (AFP) were collected. Univariate analysis and multivariate Cox proportional hazards regression model were used to evaluate the independent prognostic factors for tumor recurrence. Then a nomogram called CEUS model was constructed. The CEUS model was then used to predict recurrence at 6 mo, 12 mo, and 24 mo, the cut-off value was calculate by X-tile, and each C-index was calculated. Then Kaplan-Meier curve was compared by log-rank test. The calibration curves of each time were depicted.

RESULTS

A nomogram predicting early recurrence (ER), named CEUS model, was formulated based on the results of the multivariate Cox regression analysis. This nomogram incorporated tumor diameter, preoperative AFP level, and LIRADS, and the hazard ratio was 1.123 (95% confidence interval [CI]: 1.041-1.211), 1.547 (95%CI: 1.245-1.922), and 1.428 (95%CI: 1.059-1.925), respectively. The cut-off value at 6 mo, 12 mo, and 24 mo was 100, 80, and 50, and the C-index was 0.748 (95%CI: 0.683-0.813), 0.762 (95%CI: 0.704-0.820), and 0.762 (95%CI: 0.706-0.819), respectively. The model showed satisfactory results, and the calibration at 6 mo was desirable; however, the calibration at 12 and 24 mo should be improved.

CONCLUSION

The CEUS model enables the well-calibrated individualized prediction of ER before surgery and may represent a novel tool for biomarker research and individual counseling.

Growing Human Hepatocellular Tumors Undergo a Global Metabolic Reprogramming

Hepatocellular carcinoma (HCC) is a common malignancy with poor prognosis, high morbidity and mortality concerning with lack of effective diagnosis and high postoperative recurrence. Similar with other cancers, HCC cancer cells have to alter their metabolism to adapt to the changing requirements imposed by the environment of the growing tumor. In less vascularized regions of tumor, cancer cells experience hypoxia and nutrient starvation. Here, we show that HCC undergoes a global metabolic reprogramming during tumor growth. A combined proteomics and metabolomics analysis of paired peritumoral and tumor tissues from 200 HCC patients revealed liver-specific metabolic reprogramming and metabolic alterations with increasing tumor sizes. Several proteins and metabolites associated with glycolysis, the tricarboxylic acid cycle and pyrimidine synthesis were found to be differentially regulated in serum, tumor and peritumoral tissue with increased tumor sizes. Several prognostic metabolite biomarkers involved in HCC metabolic reprogramming were identified and integrated with clinical and pathological data. We built and validated this combined model to discriminate against patients with different recurrence risks. An integrated and comprehensive metabolomic analysis of HCC is provided by our present work. Metabolomic alterations associated with the advanced stage of the disease and poor clinical outcomes, were revealed. Targeting cancer metabolism may deliver effective therapies for HCC.

Development and Validation of a Machine Learning Prognostic Model for Hepatocellular Carcinoma Recurrence After Surgical Resection

Surgical resection remains primary curative treatment for patients with hepatocellular carcinoma (HCC) while over 50% of patients experience recurrence, which calls for individualized recurrence prediction and early surveillance. This study aimed to develop a machine learning prognostic model to identify high-risk patients after surgical resection and to review importance of variables in different time intervals. The patients in this study were from two centers including Eastern Hepatobiliary Surgery Hospital (EHSH) and Mengchao Hepatobiliary Hospital (MHH). The best-performed model was determined, validated, and applied to each time interval (0-1 year, 1-2 years, 2-3 years, and 3-5 years). Importance scores were used to illustrate feature importance in different time intervals. In addition, a risk heat map was constructed which visually depicted the risk of recurrence in different years. A total of 7,919 patients from two centers were included, of which 3,359 and 230 patients experienced recurrence, metastasis or died during the follow-up time in the EHSH and MHH datasets, respectively. The XGBoost model achieved the best discrimination with a c-index of 0.713 in internal validation cohort. Kaplan-Meier curves succeed to stratify external validation cohort into different risk groups (p < 0.05 in all comparisons). Tumor characteristics contribute more to HCC relapse in 0 to 1 year while HBV infection and smoking affect patients' outcome largely in 3 to 5 years. Based on machine learning prediction model, the peak of recurrence can be predicted for individual HCC patients. Therefore, clinicians can apply it to personalize the management of postoperative survival.

Prognostic Nomograms for Patients with Hepatocellular Carcinoma After Curative Hepatectomy, with a Focus on Recurrence Timing and Post-Recurrence Management

Background

Prognoses of patients with hepatocellular carcinoma (HCC) after curative hepatectomy remain unsatisfactory because of the high incidence of postoperative recurrence. Published predictive systems focus on pre-resection oncological characteristics, ignoring post-recurrence factors.

Purpose

This study aimed to develop prognostic nomograms for 3- and 5-year overall survival (OS) of patients with HCC after curative hepatectomy, focusing on potentially influential post-recurrence factors.

Patients and Methods

Clinicopathological and postoperative follow-up data were extracted from 494 patients with HCC who underwent curative hepatectomy between January 2012 and June 2019. Early recurrence (ER) and late recurrence (LR) were defined as recurrence at ≤2 and >2 years, respectively, after curative hepatectomy. Nomograms for the prediction of 3- and 5-year OS were established based on multivariate analysis. The areas under time-dependent receiver operating characteristic curves (AUCs) for the nomograms were calculated independently to verify predictive accuracy. The nomograms were internally validated based on 2000 bootstrap resampling of 75% of the original data.

Results

In total, 494 patients with HCC who underwent curative hepatectomy met the eligibility criteria. Cox proportional hazard regression analysis identified factors potentially influencing 3- and 5-year OS. Multivariate analysis indicated that patient age, Hong Kong Liver Cancer stage, γ-glutamyl transferase (γ-GGT) level, METAVIR inflammation activity grade, ER and post-recurrence treatment modality were influencing factors for 3-year OS (AUC, 0.891; 95% CI, 0.8364-0.9447). γ-GGT > 60 U/L, hepatectomy extent, LR and post-recurrence treatment modality were influencing factors for 5-year OS (AUC, 0.864; 95% CI, 0.8041-0.9237). Calibration plots showed satisfactory concordance between the predicted and actual observation cohorts.

Conclusion

We propose new prognostic nomograms for OS prediction with a focus on the differentiation of recurrence timing and post-recurrence management. These nomograms overcome the shortcomings of previous predictive nomograms and significantly improve predictive accuracy.

Novel inflammation-based prognostic nomograms for individualized prediction in hepatocellular carcinoma after radical resection

Background

The prognosis for patients with hepatocellular carcinoma (HCC) after liver resection ranges widely and is unsatisfactory. This study aimed to develop two novel nomograms that combined tumor characteristics and inflammation-related indexes to predict overall survival (OS) and recurrence-free survival (RFS).

Methods

In total, 3,071 patients who underwent radical resection were recruited. Independent risk factors were identified by Cox regression analysis and used to conduct prognostic nomograms. The C-index, time-dependent areas under the receiver operating characteristic curve (time-dependent AUC), decision curve analysis (DCA), and calibration curves were used to assess the performance of the nomograms.

Results

Multivariate analysis revealed that alpha-fetoprotein (AFP), resection margin, neutrophil times γ-glutamyl transpeptidase-to-lymphocyte ratio (NrLR), platelet-to-lymphocyte ratio (PLR), γ-glutamyl transpeptidase-to-platelet ratio (GPR), tumor size, tumor number, microvascular invasion, and Edmondson-Steiner grade were the independent risk factors associated with OS. The independent risk factors associated with RFS were hepatitis, AFP, albumin-bilirubin (ALBI), NrLR, PLR, PNI, GPR, tumor size, tumor number, microvascular invasion, and Edmondson-Steiner grade. The C-index of the nomograms in the training and validation cohort were 0.71 [95% confidence interval (CI): 0.70–0.73] and 0.71 (95% CI: 0.69–0.74) for the OS, and 0.71 (95% CI: 0.70–0.73) and 0.74 (95% CI: 0.72–0.76) for RFS, respectively. The C-index, time-dependent AUC, and DCA of the nomograms showed significantly better predictive performances than those of commonly used staging systems. The models could stratify patients into three different risk groups. The web-based tools are convenient for clinical practice.

Conclusions

Two novel nomograms in which integrated inflammation-related indexes and accessible clinical parameters were developed to predict OS and RFS in HCC patients who underwent radical resection. Such models will help guide postoperative individualized follow-up and adjuvant therapy.

NUF2 is a valuable prognostic biomarker to predict early recurrence of hepatocellular carcinoma after surgical resection

Early tumor recurrence after curative surgical resection poses a great challenge to the clinical management of hepatocellular carcinoma (HCC). We conducted whole genome expression microarrays on 64 primary HCC tumors with clinically defined recurrence status and cross-referenced with RNA-seq data from 18 HCC tumors in the Cancer Genome Atlas project. We identified a 77-gene signature, which is significantly associated with early recurrent (ER) HCC tumors. This ER-associated signature shows significant enrichment in genes involved in cell cycle pathway. We performed receiver operating characteristic (ROC) analysis to evaluate the prognostic biomarker potential of these 77 genes and Pearson correlation analysis to identify 11 close clusters. The one gene with the best area under the ROC curve in each of the 11 clusters was selected for validation using reverse-transcription quantitative PCR in an independent cohort of 24 HCC tumors. NUF2 was identified to be the minimal biomarker sufficient to discriminate ER tumors from LR tumors. NUF2 in combination with liver cirrhosis could significantly improve the detection of ER tumors with an AUROC of 0.82 and 0.85 in the test and validation cohort, respectively. In conclusion, NUF2 in combination with liver cirrhosis is a promising prognostic biomarker for early HCC recurrence.

A review of prognostic scores after liver resection in hepatocellular carcinoma: the MSKCC, SLICER and SSCLIP scores

Predicting prognosis in hepatocellular carcinoma (HCC) aids clinical decision-making and stratifies patient follow-up plans. There are currently three prognostic scores specific to liver resection of HCC published in the literature: the MSKCC, SLICER and SSCLIP scores. In this review, we highlight the methodology employed in the construction of these scores and discuss the strengths and weaknesses of each. Current limitations to prognostic scores include the inability to differentiate between early and late recurrences of HCC, the failure to account for the impact of aetiology of HCC and the assumption that ethnicity has no impact on disease process. Further scientific discoveries in these areas will allow for improvement in prognostication. The SLICER score is currently the most comprehensive. External validation of each score in cohorts of patients with different baseline demographics and clinical characteristics will be required to examine their stability and reliability.

Low T3 levels as a predictor marker predict the prognosis of patients with acute ischemic stroke

BACKGROUND

Early and accurate prediction of outcome in acute stroke is important. The aim of this prospective study was to explore the correlation between serum triiodothyronine level and prognosis in acute ischemic stroke patients.

METHODS

A prospective observational study which included 359 consecutive patients with acute ischemic stroke from December 2014 to November 2015 was interrogated. Serum triiodothyronine (T₃) concentrations were measured on admission to understand their value in predicting functional outcome within 90 d using multivariable models adjusted for confounding factors. Receiver operating characteristic (ROC) curves were calculated to define the best cut-off value of triiodothyronine to predict outcome. The accuracy of the test was assessed measuring the area under the ROC curve (AUROC).

RESULTS

Triiodothyronine was significantly decreased in patients with an unfavorable functional outcome as compared to patients with a favorable functional outcome within 90 d (p = 0.01). Binary logistic regression analyses revealed that lower triiodothyronine concentrations on admission were associated with a risk for poor outcomes (OR 0.05, 95% CI 0.01-0.25; p < 0.01). In addition, in ROC curve analysis, triiodothyronine may improve the National Institutes of Health Stroke Scale (NIHSS) score in predicting functional outcome. The combined model AUROC was 0.84 for 30 d and 0.91 for 90 d, which were both significantly higher than the AUROCs of original NIHSS (0.83 and 0.87), triiodothyronine (0.64 and 0.69) and age (0.57 and 0.68) (all p < 0.05).

CONCLUSIONS

Low serum triiodothyronine levels can be a predictive marker of short-term outcome after ischemic stroke. A combined model (triiodothyronine, age and NIHSS score) can add significant additional predictive information to the clinical score of the NIHSS.

Stratified neutrophil-to-lymphocyte ratio accurately predict mortality risk in hepatocellular carcinoma patients following curative liver resection

OBJECTIVES

Neutrophil lymphocyte ratio (NLR) has been shown to predict prognosis of cancers in several studies. This study was designed to evaluate the impact of stratified NLR in patients who have received curative liver resection (CLR) for hepatocellular carcinoma (HCC).

METHODS

A total of 1659 patients who underwent CLR for suspected HCC between 2007 and 2014 were reviewed. The preoperative NLR was categorized into quartiles based on the quantity of the study population and the distribution of NLR. Hazard ratios (HRs) and 95% confidence intervals (CIs) were significantly associated with overall survival (OS) and derived by Cox proportional hazard regression analyses. Univariate and multivariate Cox proportional hazard regression analyses were evaluated for association of all independent parameters with disease prognosis.

RESULTS

Multivariable Cox proportional hazards models showed that the level of NLR (HR = 1.031, 95%CI: 1.002-1.060, P = 0.033), number of nodules (HR = 1.679, 95%CI: 1.285-2.194, P<0.001), portal vein thrombosis (HR = 4.329, 95%CI: 1.968-9.521, P<0.001), microvascular invasion (HR = 2.527, 95%CI: 1.726-3.700, P<0.001) and CTP score (HR = 1.675, 95%CI: 1.153-2.433, P = 0.007) were significant predictors of mortality. From the Kaplan-Meier analysis of overall survival (OS), each NLR quartile showed a progressively worse OS and apparent separation (log-rank P=0.008). The highest 5-year OS rate following CLR (60%) in HCC patients was observed in quartile 1. In contrast, the lowest 5-year OS rate (27%) was obtained in quartile 4.

CONCLUSIONS

Stratified NLR may predict significantly improved outcomes and strengthen the predictive power for patient responses to therapeutic intervention.