Predictor Of Survival In Patients Research Articles

The immune cell infiltration-associated molecular subtypes and gene signature predict prognosis for osteosarcoma patients

Host immune dysregulation involves in the initiation and development of osteosarcoma (OS). However, the exact role of immune cells in OS remains unknown. We aimed to distinguish the molecular subtypes and establish a prognostic model in OS patients based on immunocyte infiltration. The gene expression profile and corresponding clinical feature of OS patients were obtained from TARGET and GSE21257 datasets. MCP-counter and univariate Cox regression analyses were applied to identify immune cell infiltration-related molecular subgroups. Functional enrichment analysis and immunocyte infiltration analysis were performed between two subgroups. Furthermore, Cox regression and LASSO analyses were performed to establish the prognostic model for the prediction of prognosis and metastasis in OS patients. The subgroup with low infiltration of monocytic lineage (ML) was related to bad prognosis in OS patients. 435 DEGs were screened between the two subgroups. Functional enrichment analysis revealed these DEGs were involved in immune- and inflammation-related pathways. Three important genes (including TERT, CCDC26, and IL2RA) were identified to establish the prognostic model. The risk model had good prognostic performance for the prediction of metastasis and overall survival in OS patients. A novel stratification system was established based on ML-related signature. The risk model could predict the metastasis and prognosis in OS patients. Our findings offered a novel sight for the prognosis and development of OS.

Proteomic analysis of ascitic extracellular vesicles describes tumour microenvironment and predicts patient survival in ovarian cancer.

High-grade serous carcinoma of the ovary, fallopian tube and peritoneum (HGSC), the most common type of ovarian cancer, ranks among the deadliest malignancies. Many HGSC patients have excess fluid in the peritoneum called ascites. Ascites is a tumour microenvironment (TME) containing various cells, proteins and extracellular vesicles (EVs). We isolated EVs from patients' ascites by orthogonal methods and analyzed them by mass spectrometry. We identified not only a set of 'core ascitic EV-associated proteins' but also defined their subset unique to HGSC ascites. Using single-cell RNA sequencing data, we mapped the origin of HGSC-specific EVs to different types of cells present in ascites. Surprisingly, EVs did not come predominantly from tumour cells but from non-malignant cell types such as macrophages and fibroblasts. Flow cytometry of ascitic cells in combination with analysis of EV protein composition in matched samples showed that analysis of cell type-specific EV markers in HGSC has more substantial prognostic potential than analysis of ascitic cells. To conclude, we provide evidence that proteomic analysis of EVs can define the cellular composition of HGSC TME. This finding opens numerous avenues both for a better understanding of EV's role in tumour promotion/prevention and for improved HGSC diagnostics.

Exploration of YBX1 role in the prognostic value and immune characteristics by single-cell and bulk sequencing analysis for liver hepatocellular carcinoma.

Y-box binding protein 1 (YBX1) plays a variety of roles in progression of multiple tumors. However, the role of YBX1 in prognostic value and immune regulation for liver hepatocellular carcinoma (LIHC) remains unclear. The present study aimed to examine the effect of YBX1 on the regulation of tumor immunity and survival prediction in LIHC patients. YBX1-related expression profiles and single-cell and bulk sequencing analysis were performed using online databases. YBX1 expression was validated by a quantitative real-time PCR (qRT-PCR), western blotting and immunohistochemistry. Univariate/multivariate Cox regression analysis was performed to determine independent predictors of overall survival (OS). The ESTIMATE (i.e., Estimation of STromal and Immune cells in MAlignant Tumor tissues using Expression data) algorithm and Tumor Immune Dysfunction and Exclusion (TIDE) analysis were used to assess the relationships between YBX1 and LIHC immunity. YBX1 was over-expressed in LIHC tissues and cell lines. High YBX1 expression was significantly associated with poor OS. Univariate/multivariate Cox regression analysis revealed that YBX1 was an independent prognostic factor for LIHC. Gene set enrichment analysis revealed that YBX1 was associated with multiple signaling pathways correlated to LIHC. Additionally, YBX1 was expressed in multiple immune cells and was significantly correlated with immune cells, immune checkpoint markers and tumor immune microenvironment. The TIDE analysis demonstrated that LIHC patients with high YBX1 expression showed a higher T-cell dysfunction score and a higher exclusion score, as well as poorer immunotherapy response. YBX1 plays crucial oncogenic roles in LIHC and is closely associated with the immune defense system. YBX1 inhibition may serve as a potential treatment for LIHC.

From Pixels to Prognosis: A Survey on AI-Driven Cancer Patient Survival Prediction Using Digital Histology Images.

Survival analysis is an integral part of medical statistics that is extensively utilized to establish prognostic indices for mortality or disease recurrence, assess treatment efficacy, and tailor effective treatment plans. The identification of prognostic biomarkers capable of predicting patient survival is a primary objective in the field of cancer research. With the recent integration of digital histology images into routine clinical practice, a plethora of Artificial Intelligence (AI)-based methods for digital pathology has emerged in scholarly literature, facilitating patient survival prediction. These methods have demonstrated remarkable proficiency in analyzing and interpreting whole slide images, yielding results comparable to those of expert pathologists. The complexity of AI-driven techniques is magnified by the distinctive characteristics of digital histology images, including their gigapixel size and diverse tissue appearances. Consequently, advanced patch-based methods are employed to effectively extract features that correlate with patient survival. These computational methods significantly enhance survival prediction accuracy and augment prognostic capabilities in cancer patients. The review discusses the methodologies employed in the literature, their performance metrics, ongoing challenges, and potential solutions for future advancements. This paper explains survival analysis and feature extraction methods for analyzing cancer patients. It also compiles essential acronyms related to cancer precision medicine. Furthermore, it is noteworthy that this is the inaugural review paper in the field. The target audience for this interdisciplinary review comprises AI practitioners, medical statisticians, and progressive oncologists who are enthusiastic about translating AI-driven solutions into clinical practice. We expect this comprehensive review article to guide future research directions in the field of cancer research.

A novel signature integrated endoplasmic reticulum stress and apoptosis related genes to predict prognosis for breast cancer

BackgroundBreast cancer (BC) is the primary cause of cancer mortality. Herein, we aimed to establish and verify a prognostic model consisting of endoplasmic reticulum stress and apoptosis related genes (ERAGs) to predict patient survival. MethodsThe Cancer Genome Atlas (TCGA) database was used to download gene expression and clinical data to identify the differentially expressed genes (DEGs). Using univariate Cox regression analysis and the Least Absolute Shrinkage and Selection Operator (LASSO)-penalized Cox proportional hazards regression analysis, the prognostic ERAGs were screened. The predictive performance was evaluated using Kaplan-Meier (KM) survival and receiver operating characteristic (ROC) curve analysis. Furthermore, a nomogram model incorporating clinical parameters and risk scores was constructed and subsequently evaluated using ROC and KM analysis. The correlation analysis, mutation analysis, functional enrichment analysis, and immune infiltration analysis were employed to investigate the specific mechanism of ERAGs. We also used Quantitative Real-Time PCR (RT-qPCR) to verify the differential expression of DE-ERAGs between the breast cancer cell line and mammary epithelial cell line. ResultsWe constructed a prognostic signature comprising 16 ERAGs. ROC, KM analysis and the nomogram model demonstrated high effectiveness in accurately predicting the overall survival (OS) of BRCA patients. The results of these analysis could provide reference for further mechanism exploration. ConclusionWe developed and assessed a novel molecular predictive model for breast cancer that focuses on endoplasmic reticulum stress and apoptosis in this study. It is a valuable complement to the existing prognostic prediction models for breast cancer.

A Correlation Evaluation Between the Peripheral Blood Index and the Prognosis of Advanced Esophageal Squamous Cell Carcinoma Patients Treated with Camrelizumab.

This study aimed to investigate the relationship between peripheral blood indices and the efficacy and prognosis of advanced esophageal squamous cell carcinoma (ESCC) patients treated with camrelizumab. We retrospectively analyzed 64 patients who received camrelizumab for advanced ESCC at the Second People's Hospital of Lianyungang City between July 2020 and June 2022. The study included examination of the neutrophil-to-lymphocyte ratio (NLR), the platelet-to-lymphocyte ratio (PLR), the systemic inflammation index (SII), the lymph-to-monocytes ratio (LMR), the absolute lymphocyte count (ALC), and lactate dehydrogenase (LDH). We used multivariate logistic regression analysis to explore the link existing between peripheral blood and the efficacy of treatment. Determination of potential prognostic factors for Progression-free survival (PFS) and Overall survival (OS) using Cox regression analysis. The nomogram model was developed based on the results of the Cox multivariate analysis. Patients were divided into three groups according to the reduction in LDH and LDL levels before treatment, and the Kaplan-Meier survival curves for the three groups were compared and ROC curves for LDH combined with PLR were plotted. Lower LDH (OR=6.237, 95% CI: 1.625-23.944) were independently associated with disease control rates(DCR). LDH was independently correlated with PFS (HR: 0.227 95% CI: 0.099-0.517). LDH and PLR were independently linked to OS. The C index of the nomogram model is 0.819, indicating good predictive performance. Kaplan-Meier Survival Curve suggested better OS in patients with reduced pretreatment LDH and PLR. The area under the ROC curve showed that the LDH index combined with the PLR index predicts patient survival better than the index alone. LDH combined with PLR predicted prognosis in patients with ESCC treated with camrelizumab.

A comparison between centralized and asynchronous federated learning approaches for survival outcome prediction using clinical and PET data from non-small cell lung cancer patients

Background and Objective: Survival analysis plays an essential role in the medical field for optimal treatment decision-making. Recently, survival analysis based on the deep learning (DL) approach has been proposed and is demonstrating promising results. However, developing an ideal prediction model requires integrating large datasets across multiple institutions, which poses challenges concerning medical data privacy.Methods: In this paper, we propose FedSurv, an asynchronous federated learning (FL) framework designed to predict survival time using clinical information and positron emission tomography (PET)-based features. This study used two datasets: a public radiogenic dataset of non-small cell lung cancer (NSCLC) from the Cancer Imaging Archive (RNSCLC), and an in-house dataset from the Chonnam National University Hwasun Hospital (CNUHH) in South Korea, consisting of clinical risk factors and F-18 fluorodeoxyglucose (FDG) PET images in NSCLC patients. Initially, each dataset was divided into multiple clients according to histological attributes, and each client was trained using the proposed DL model to predict individual survival time. The FL framework collected weights and parameters from the clients, which were then incorporated into the global model. Finally, the global model aggregated all weights and parameters and redistributed the updated model weights to each client. We evaluated different frameworks including single-client-based approach, centralized learning and FL.Results: We evaluated our method on two independent datasets. First, on the RNSCLC dataset, the mean absolute error (MAE) was 490.80±22.95 d and the C-Index was 0.69±0.01. Second, on the CNUHH dataset, the MAE was 494.25±40.16 d and the C-Index was 0.71±0.01. The FL approach achieved centralized method performance in PET-based survival time prediction and outperformed single-client-based approaches.Conclusions: Our results demonstrated the feasibility and effectiveness of employing FL for individual survival prediction in NSCLC patients, using clinical information and PET-based features.

Plasma THBS1 as a predictive biomarker for poor prognosis and brain metastasis in patients with HER2-enriched breast cancer.

Thrombospondin-1 (THBS1) is a secretory adhesive glycoprotein involved in the progression of multiple malignancies, including breast cancer. However, the clinical significance and prognostic role of plasma THBS1 in breast cancer have yet to be clarified. Plasma THBS1 levels in 627 breast cancer patients were analyzed by enzyme-linked immunosorbent assay. Bone marrow blood was drawn from the anterior/posterior superior iliac spine to detect the presence of disseminated tumor cells (DTCs). The effects of plasma THBS1 on the clinicopathological characteristics and survival prediction of breast cancer patients were explored. Plasma THBS1 did not correlate with overall survival, breast cancer-specific survival (BCSS), and distant disease-free survival (DDFS) in the entire breast cancer cohort. Notably, HER2-enriched patients with high-plasma THBS1 levels had significantly shorter BCSS (P = 0.027) and DDFS (P = 0.011) than those with low levels. Multivariate analyses revealed that plasma THBS1 was an independent prognostic marker of BCSS (P = 0.026) and DDFS (P = 0.007) in HER2-enriched patients. THBS1 levels were 24% higher in positive DTC patients than in negative DTC patients (P = 0.031), and high levels were significantly associated with poor BCSS in positive DTC patients (HR 2.08, 95% CI 1.17-3.71; P = 0.019). Moreover, high-plasma THBS1 levels were specifically associated with an increased occurrence of brain metastasis in HER2-enriched patients (P = 0.041). These findings suggest that plasma THBS1 may be serving as an unfavorable prognosis predictor for HER2-enriched breast cancer and justifies the need for further research.

Red Cell Distribution Width as a Predictor of Survival in Patients with Hepatocellular Carcinoma.

Background and Objectives. Hepatocellular carcinoma (HCC) and the intrahepatic biliary tract cancers are estimated to rank sixth for incidence among solid cancers worldwide, and third for mortality rates. A critical issue remains the need for accurate biomarkers for risk stratification and overall prognosis. The aim of this study was to investigate the ability of a biomarker of heterogeneity of the size of red blood cells, the red cell distribution width (RDW), to predict survival in patients with HCC. Materials and Methods. A consecutive series of patients with a histologic diagnosis of HCC were included into this study irrespective of their age, stage of the disease, and treatment administered, and followed-up for a period of three years. Demographic, anthropometric [age, sex, body mass index (BMI)], and clinical data (Charlson Comorbidity Index, Child-Pugh score, etc.), along with laboratory tests were retrieved from clinical records. Results. One-hundred and four patients were included in this study. Among them, 54 (69%) were deceased at the end of the follow-up. Higher RDW values, but not other hematological and biochemical parameters, were significantly associated with mortality in both univariate and multivariate analysis. The optimal RDW cut-off value identified with the Youden test for survival was 14.7%, with 65% sensitivity and 74% specificity (AUC = 0.718, 95% CI 0.622-0.802, p < 0.001). Kaplan-Meier survival curves showed significantly lower survival with higher RDW values (HR = 3.5204; 95% CI 1.9680-6.2975, p < 0.0001) with a mean survival of 30.9 ± 9.67 months for patients with RDW ≤ 14.7% and 22.3 ± 11.4 months for patients with RDW > 14.7%. Conclusions. The results of our study showed that RDW can perform better than other blood-based biomarkers in independently predicting prognosis in patients with HCC.

Screening of immune related gene and survival prediction of lung adenocarcinoma patients based on LightGBM model

Lung cancer is one of the malignant tumors with the greatest threat to human health, and studies have shown that some genes play an important regulatory role in the occurrence and development of lung cancer. In this paper, a LightGBM ensemble learning method is proposed to construct a prognostic model based on immune relate gene (IRG) profile data and clinical data to predict the prognostic survival rate of lung adenocarcinoma patients. First, this method used the Limma package for differential gene expression, used CoxPH regression analysis to screen the IRG to prognosis, and then used XGBoost algorithm to score the importance of the IRG features. Finally, the LASSO regression analysis was used to select IRG that could be used to construct a prognostic model, and a total of 17 IRG features were obtained that could be used to construct model. LightGBM was trained according to the IRG screened. The K-means algorithm was used to divide the patients into three groups, and the area under curve (AUC) of receiver operating characteristic (ROC) of the model output showed that the accuracy of the model in predicting the survival rates of the three groups of patients was 96%, 98% and 96%, respectively. The experimental results show that the model proposed in this paper can divide patients with lung adenocarcinoma into three groups [5-year survival rate higher than 65% (group 1), lower than 65% but higher than 30% (group 2) and lower than 30% (group 3)] and can accurately predict the 5-year survival rate of lung adenocarcinoma patients.

A novel ferroptosis-related gene signature for overall survival prediction in patients with gastric cancer.

The global diagnosis rate and mortality of gastric cancer (GC) are among the highest. Ferroptosis and iron-metabolism have a profound impact on tumor development and are closely linked to cancer treatment and patient's prognosis. In this study, we identified six PRDEGs (prognostic ferroptosis- and iron metabolism-related differentially expressed genes) using LASSO-penalized Cox regression analysis. The TCGA cohort was used to establish a prognostic risk model, which allowed us to categorize GC patients into the high- and the low-risk groups based on the median value of the risk scores. Our study demonstrated that patients in the low-risk group had a higher probability of survival compared to those in the high-risk group. Furthermore, the low-risk group exhibited a higher tumor mutation burden (TMB) and a longer 5-year survival period when compared to the high-risk group. In summary, the prognostic risk model, based on the six genes associated with ferroptosis and iron-metabolism, performs well in predicting the prognosis of GC patients.

Expression of key unfolded protein response genes predicts patient survival and an immunosuppressive microenvironment in glioblastoma

Abstract Background Dysregulation of cellular processes related to protein folding and trafficking leads to the accumulation of misfolded proteins in the endoplasmic reticulum (ER), triggering ER stress. Cells cope with ER stress by activating the unfolded protein response (UPR), a signaling pathway that has been implicated in a variety of diseases, including cancer. However, the role of the UPR in cancer initiation and progression is still unclear. Methods Here we used bulk and single cell RNA sequencing data to investigate ER stress-related gene expression in glioblastoma, as well as the impact key UPR genes have on patient survival. Results ER stress-related genes are highly expressed in both cancer cells and tumor-associated macrophages, with evidence of high intra- and inter-tumor heterogeneity. High expression of the UPR-related genes HSPA5, P4HB, and PDIA4 was identified as risk factors while high MAPK8 (JNK1) expression was identified as a protective factor in glioblastoma patients, indicating UPR genes have prognostic potential in this cancer type. Finally, expression of XBP1 and MAPK8, two key downstream targets of the ER sentinel IRE1α, correlates with the presence of immune cell types associated with immunosuppression and a worse patient outcome. This suggests that the expression of these genes is associated with an immunosuppressive tumor microenvironment and uncover a potential link between stress response pathways, tumor microenvironment and glioblastoma patient survival. Conclusions We performed a comprehensive transcriptional characterization of the unfolded protein response in glioblastoma patients and identified UPR-related genes associated with glioblastoma patient survival, providing potential prognostic and predictive biomarkers as well as promising targets for developing new therapeutic interventions in glioblastoma treatment.

Association of preoperative muscle-adipose index measured by computed tomography with survival in patients with esophageal squamous cell carcinoma

BackgroundConventional nutritional metrics are closely associated with the prognosis of patients with radically resected esophageal squamous cell carcinoma (ESCC). Nevertheless, the prognostic implications of muscle and adipose tissue composite indexes in ESCC remain unknown.MethodsWe retrospectively analyzed clinicopathological data of 304 patients who underwent resected ESCC. To obtain measurements of the muscle and adipose indexes, preoperative computed tomography (CT) images were used to quantify skeletal-muscle adipose tissue. The diagnostic threshold for muscle-adipose imbalance was determined using X-tile software and used to analyze the association between the muscle-adipose index (MAI) and survival. Instantaneous risk of recurrence was assessed using a hazard function. We constructed a nomogram based on the MAI and other clinical characteristics and established a novel predictive model with independent prognostic factors. The prognostic capabilities of these nomograms were evaluated using calibration curves, receiver operating characteristic (ROC) curves, and decision-curve analysis (DCA).ResultsThe overall survival (OS) and disease-free survival (DFS) rates in the muscle-adipose-balanced group were significantly better than those in the muscle-adipose-imbalanced group. Multivariate analyses revealed that the MAI, prognostic nutritional index (PNI), tumor stage, and tumor differentiation were independent prognostic factors for OS and DFS in patients with resected ESCC (P < 0.05). The nuclear density curve indicated a lower risk of recurrence for patients in the muscle-adipose-balanced group than that for their imbalanced counterparts. Conversely, the nuclear density curve for PNI was confounded. Postoperative radiotherapy- (RT) benefit analysis demonstrated that patients with ESCC in the muscle-adipose-balanced group could benefit from adjuvant RT.ConclusionThis study confirmed that preoperative MAI could serve as a useful independent prognostic factor in patients with resected ESCC. A nomogram based on the MAI and other clinical characteristics could provide individualized survival prediction for patients receiving radical resection. Timely and appropriate nutritional supplements may improve treatment efficacy.

A novel peroxisome-related gene signature predicts clinical prognosis and is associated with immune microenvironment in low-grade glioma.

Low-grade glioma (LGG), a common primary tumor, mainly originates from astrocytes and oligodendrocytes. Increasing evidence has shown that peroxisomes function in the regulation of tumorigenesis and development of cancer. However, the prognostic value of peroxisome-related genes (PRGs) in LGG has not been reported. Therefore, it is necessary to construct a prognostic risk model for LGG patients based on the expression profiles of peroxisome-related genes. Our study mainly concentrated on developing a peroxisome-related gene signature for overall survival (OS) prediction in LGG patients. First, according to these peroxisome-related genes, all LGG patients from The Cancer Genome Atlas (TCGA) database could be divided into two subtypes. Univariate Cox regression analysis was used to find prognostic peroxisome-related genes in TCGA_LGG dataset, and least absolute shrinkage and selection operator Cox regression analysis was employed to establish a 14-gene signature. The risk score based on the signature was positively associated with unfavorable prognosis. Then, multivariate Cox regression incorporating additional clinical characteristics showed that the 14-gene signature was an independent predictor of LGG. Time-dependent ROC curves revealed good performance of this prognostic signature in LGG patients. The performance about predicting OS of LGG was validated using the GSE107850 dataset derived from the Gene Expression Omnibus (GEO) database. Furethermore, we constructed a nomogram model based on the gene signature and age, which showed a better prognostic power. Gene ontology (GO) and Kyoto Encylopedia of Genes and Genomes (KEGG) analyses showed that neuroactive ligand-receptor interaction and phagosome were enriched and that the immune status was decreased in the high-risk group. Finally, cell counting kit-8 (CCK8) were used to detect cell proliferation of U251 and A172 cells. Inhibition of ATAD1 (ATPase family AAA domain-containing 1) and ACBD5 (Acyl-CoA binding-domain-containing-5) expression led to significant inhibition of U251 and A172 cell proliferation. Flow cytometry detection showed that ATAD1 and ACBD5 could induce apoptosis of U251 and A172 cells. Therefore, through bioinformatics methods and cell experiments, our study developed a new peroxisome-related gene signature that migh t help improve personalized OS prediction in LGG patients.

Radiomics Features on Enhanced Computed Tomography Predict FOXP3 Expression and Clinical Prognosis in Patients with Head and Neck Squamous Cell Carcinoma.

Forkhead box P3 (FOXP3) has been identified as a novel molecular marker in various types of cancer. The present study assessed the expression of FOXP3 in patients with head and neck squamous cell carcinoma (HNSCC) and its potential as a clinical prognostic indicator, and developed a radiomics model based on enhanced computed tomography (CT) imaging. Data from 483 patients with HNSCC were downloaded from the Cancer Genome Atlas for FOXP3 prognostic analysis and enhanced CT images from 139 patients included in the Cancer Imaging Archives, which were subjected to the maximum relevance and minimum redundancy and recursive feature elimination algorithms for radiomics feature extraction and processing. Logistic regression was used to build a model for predicting FOXP3 expression. A prognostic scoring system for radiomics score (RS), FOXP3, and patient clinicopathological factors was established to predict patient survival. The area under the receiver operating characteristic (ROC) curve (AUC) and calibration curve and decision curve analysis (DCA) were used to evaluate model performance. Furthermore, the relationship between FOXP3 and the immune microenvironment, as well as the association between RS and immune checkpoint-related genes, was analyzed. Results of analysis revealed that patients with HNSCC and high FOXP3 mRNA expression exhibited better overall survival. Immune infiltration analysis revealed that FOXP3 had a positive correlation with CD4 + and CD8 + T cells and other immune cells. The 8 best radiomics features were selected to construct the radiomics model. In the FOXP3 expression prediction model, the AUC values were 0.707 and 0.702 for the training and validation sets, respectively. Additionally, the calibration curve and DCA demonstrated the positive diagnostic utility of the model. RS was correlated with immune checkpoint-related genes such as ICOS, CTLA4, and PDCD1. A predictive nomogram was established, the AUCs were 0.87, 0.787, and 0.801 at 12, 24, and 36months, respectively, and DCA demonstrated the high clinical applicability of the nomogram. The enhanced CT radiomics model can predict expression of FOXP3 and prognosis in patients with HNSCC. As such, FOXP3 may be used as a novel prognostic marker to improve individualized clinical diagnosis and treatment decisions.

Single-cell and whole-transcriptome sequencing of lymph node metastasis-related gene signature: A large-scale pan-cancer cohort, machine learning and experimental study: [RETRACTED

Lymph node metastasis (LNM) is an independent prognostic factor in numerous types of cancer. Therefore, a LNM-related gene-based nomogram may precisely predict survival and drug sensitivity, and reveal the mechanism underlying LNM. Gene sequencing profiles of pan-cancer data (33 cancer types) were acquired from The Cancer Genome Atlas UCSC Xena database. Patients were classified into primary (N = 10,071) and testing (N = 5,036) cohorts. The lymph node score (LNscore) was established via single-cell RNA sequencing, whole-transcriptome sequencing, machine learning, and Cox regression analyses. A novel prognosis model, formulated by incorporating the LNscore and clinical characteristics, was evaluated using the concordance index, calibration curve, and decision curve analysis. Moreover, patients were assigned into high- and low-risk groups according to the median LNscore. We investigated these two groups for survival prognosis, functional enrichment, immune infiltration, and drug sensitivity. In addition, we silenced and overexpressed insulin like growth factor 2 mRNA binding protein 2 (IGF2BP2). We also analyzed the behavior of breast cancer (BRCA) cells regarding lymphatic metastasis and lymphangiogenesis in vitro. IGF2BP2 stimulated the proliferation of BRCA cells via 5-Ethynyl-2'-deoxyuridine and Cell Counting Kit-8 experiments. A LNM-related set of 12 genes was identified and utilized to determine the LNscore. The concordance-index of both cohorts in the LNscore-based model was >0.7. The immune landscape revealed that the sensitivity to immunotherapy might be better in the high-risk group versus the low-risk group. In addition, we discovered that IGF2BP2 was overexpressed in BRCA tissues and significantly associated with poor survival. Functional analysis indicated that IGF2BP2 promoted BRCA cell migration and proliferation. Additionally, IGF2BP2 accelerated lymphatic metastasis and lymphangiogenesis in vivo. A novel LNscore-based model was established via comprehensive analysis of LNM-related genes. This model can accurately predict patient survival and drug sensitivity, and reveal the mechanism of LNM in the pan-cancer setting.

Development and internal validation of machine learning models for personalized survival predictions in spinal cord glioma patients

BACKGROUND CONTEXTNumerous factors have been associated with the survival outcomes in patients with spinal cord gliomas (SCG). Recognizing these specific determinants is crucial, yet it is also vital to establish a reliable and precise prognostic model for estimating individual survival outcomes. OBJECTIVEThe objectives of this study are twofold: first, to create an array of interpretable machine learning (ML) models developed for predicting survival outcomes among SCG patients; and second, to integrate these models into an easily navigable online calculator to showcase their prospective clinical applicability. STUDY DESIGNThis was a retrospective, population-based cohort study aiming to predict the outcomes of interest, which were binary categorical variables, in SCG patients with ML models. PATIENT SAMPLEThe National Cancer Database (NCDB) was utilized to identify adults aged 18 years or older who were diagnosed with histologically confirmed SCGs between 2010 and 2019. OUTCOME MEASURESThe outcomes of interest were survival outcomes at three specific time points postdiagnosis: 1, 3, and 5 years. These outcomes were formed by combining the “Vital Status” and “Last Contact or Death (Months from Diagnosis)” variables. Model performance was evaluated visually and numerically. The visual evaluation utilized receiver operating characteristic (ROC) curves, precision-recall curves (PRCs), and calibration curves. The numerical evaluation involved metrics such as sensitivity, specificity, accuracy, area under the PRC (AUPRC), area under the ROC curve (AUROC), and Brier Score. METHODSWe employed five ML algorithms—TabPFN, CatBoost, XGBoost, LightGBM, and Random Forest—along with the Optuna library for hyperparameter optimization. The models that yielded the highest AUROC values were chosen for integration into the online calculator. To enhance the explicability of our models, we utilized SHapley Additive exPlanations (SHAP) for assessing the relative significance of predictor variables and incorporated partial dependence plots (PDPs) to delineate the influence of singular variables on the predictions made by the top performing models. RESULTSFor the 1-year survival analysis, 4,913 patients [5.6% with 1-year mortality]; for the 3-year survival analysis, 4,027 patients (11.5% with 3-year mortality]; and for the 5-year survival analysis, 2,854 patients (20.4% with 5-year mortality) were included. The top models achieved AUROCs of 0.938 for 1-year mortality (TabPFN), 0.907 for 3-year mortality (LightGBM), and 0.902 for 5-year mortality (Random Forest). Global SHAP analyses across survival outcomes at different time points identified histology, tumor grade, age, surgery, radiotherapy, and tumor size as the most significant predictor variables for the top-performing models. CONCLUSIONSThis study demonstrates ML techniques can develop highly accurate prognostic models for SCG patients with excellent discriminatory ability. The interactive online calculator provides a tool for assessment by physicians (https://huggingface.co/spaces/MSHS-Neurosurgery-Research/NCDB-SCG). Local interpretability informs prediction influences for a given individual. External validation across diverse datasets could further substantiate potential utility and generalizability. This robust, interpretable methodology aligns with the goals of precision medicine, establishing a foundation for continued research leveraging ML's predictive power to enhance patient counseling.

Machine Learning System for the Effective Diagnosis and Survival Prediction of Breast Cancer Patients

Breast cancer is one of the most significant global health challenges. Effective diagnosis and prognosis prediction are crucial for improving patient outcomes in the case of this disease. As machine learning (ML) has significantly improved prediction models in many disciplines, the goal of this study is to develop a ML system for medical specialists that can accurately predict tumor diagnosis and patient survival for breast cancer patients. For the training of diagnosis and survival prediction, five algorithmic models—decision tree (DT), random forest (RF), naive bayes (NB), support vector machines (SVMs), and gradient boosting—were trained with 569 records from the Breast Cancer Wisconsin dataset and 1,980 records from the Breast Cancer Gene Expression Profiles dataset. The results showed that the NB model exhibited better performance for tumor diagnosis, achieving an accuracy of 95.0%, while RF presented the best results for patient survival, with an accuracy of 76.0%. A survey of medical experts’ experience with the resulting system showed high scores in reliability, performance, satisfaction, usability, and efficiency, confirming that ML systems have the potential to improve breast cancer patient outcomes.

Evaluation of the A2B Score for Prediction of Survival in Patients With Heart Failure in a Nationwide Cohort in Japan.

Although a tool for sharing patient prognosis among all medical staff is desirable in heart failure (HF) cases, only a few simple HF prognostic scores are available. We previously presented the A2B score, a simple user-friendly HF risk score, and validated it in a small single-center cohort. In the present study, we validated it in a larger nationwide cohort. We examined the 2-year mortality in relation to the A2B scores in 3483 patients from a Japanese nationwide cohort and attempted to stratify their prognoses according to the scores. The A2B score was determined by assigning points for age, anemia, and brain natriuretic peptide (BNP) level at discharge: age (<65 years, 0; 65-74 years, 1; ≥75 years, 2), anemia (hemoglobin ≥12 g/dL, 0; 10-11.9 g/dL, 1; <10 g/dL, 2), and BNP (<200 pg/mL, 0; 200-499 pg/mL, 1; ≥500 pg/mL, 2). Hemoglobin and BNP levels were applied to the data at discharge. The 2-year survival rates for A2B scores 1, 2, 3, 4, 5, and 6 were 94.1%, 83.2%, 74.1%, 63.5%, 51.6%, and 41.5%, respectively; the mortality rate increased by ≈10% for each point increase (c-index, 0.702). The A2B score was applicable in HF cases with reduced or preserved ejection fraction and remained useful when BNP was substituted with N-terminal proBNP (c-index, 0.749, 0.676, and 0.682, respectively). The A2B score showed a good prognostic value for HF in a large population even when BNP was replaced with N-terminal proBNP.

A novel nomogram for prediction of intrahepatic recurrence-free survival in patients with HCC followed by radiofrequency ablation.

Although radiofrequency ablation (RFA) has been considered as the favourable treatment option for hepatocellular carcinoma (HCC), there still exist some challenges for new recurrence after RFA. The present study aims to determine the factors affecting recurrence and develop an effective model to predict intrahepatic recurrence-free survival (RFS). Patients with HCC followed by RFA between 2000 and 2021 were included in this study. Multivariable Cox regression analysis was used to determine the independent prognostic factors and establish the nomogram predicting intrahepatic RFS after RFA. The predictive performance of the nomogram was assessed according to the C-index, calibration plots, and Kaplan-Meier curves stratified by the tertiles. A total of 801 sessions in 660 patients (including 1155 lesions) were enrolled into this study. Intrahepatic new recurrence was observed in all patients during the follow-up, and the mean intrahepatic RFS was 21.9 months in the present cohort. According to multivariate COX regression analysis, five independent prognostic factors affecting intrahepatic RFS were determined, including age, Child-Pugh class, tumour distribution, number of tumours, and a-fetoprotein (AFP). Based on all independent prognostic factors, the nomogram model was developed and evaluated, which achieved favourable discrimination and calibration. This study established five independent prognostic factors and constructed a nomogram model to predict intrahepatic RFS for HCC patients followed by RFA. It could better help clinicians select RFA candidates, as well as offering the important information about whether patients need receive comprehensive treatment to prevent new recurrence after RFA. (1) In this study, 5 preoperative clinic-pathological variables were determined as the independent prognostic factors affecting RFS after RFA in the current largest sample size. (2) Based on these independent prognostic factors, a prognostic nomogram predicting RFS after RFA was established, which may be used to select patients who benefit from RFA and could help both surgeons and patients provide useful information for choosing the personalized treatment.