Machine learning‑based prediction of survival prognosis in esophageal squamous cell carcinoma

Machine learning analysis of cardiovascular risk factors and their associations with hearing loss

Hearing loss poses immense burden worldwide and early detection is crucial. The accurate models identify high-risk groups, enabling timely intervention to improve quality of life. The subtle changes in hearing often go unnoticed, presenting a challenge for early hearing loss detection. While machine learning shows promise, prior studies have not leveraged cardiovascular risk factors known to impact hearing. As hearing outcomes remain challenging to characterize associations, we evaluated a new approach to predict current hearing outcomes through machine learning models using cardiovascular risk factors. The National Health and Nutrition Examination Survey (NHANES) 2012-2018 data comprising audiometric tests and cardiovascular risk factors was utilized. Machine learning algorithms were trained to classify hearing impairment thresholds and predict pure tone average values. Key results showed light gradient boosted machine performing best in classifying mild or greater impairment (> 25 dB HL) with 80.1% accuracy. It also classified > 16 dB HL and > 40 dB HL thresholds, with accuracies exceeding 77% and 86% respectively. The study also found that CatBoost and Gradient Boosting performed well in classifying hearing loss thresholds, with test set accuracies around 0.79 and F1-scores around 0.79-0.80. A multi-layer neural network emerged as the top predictor of pure tone averages, achieving a mean absolute error of just 3.05 dB. Feature analysis identified age, gender, blood pressure and waist circumference as key associated factors. Findings offer a promising direction for a clinically applicable tool, personalized prevention strategies, and calls for prospective validation.

Prognostic factors in locally advanced oesophageal squamous cell carcinoma: a clinical and radiomic analysis of neoadjuvant immunochemotherapy before surgery

Background

The treatment of locally advanced oesophageal squamous cell carcinoma (LAESCC) without distant metastasis remains a subject of debate. Neoadjuvant immunochemotherapy (NIC) combined with surgery is the preferred initial approach for managing LAESCC. However, information on the clinical efficacy and survival of patients with LAESCC treated with NIC followed by surgery is limited.

Methods

This retrospective analysis aimed to identify predictors NIC treatment effectiveness and on patient survival. We developed a Cox proportional hazards model and Kaplan-Meier curve to estimate progression-free survival (PFS) and overall survival (OS) following NIC treatment and surgery.

Results

Overall, 225 patients with LAESCC were divided into training (157) and test set (68) (7:3). After a median follow-up of 2.86 years, death was observed as a positive event in 41 patients (26.1%). It is statistically significant to construct a prediction model combining radiomics features pre- and post-NIC with clinical features to predict the PFS and OS of LAESCC. The combined model showed the highest performance in predicting both disease-free survival and OS compared with the clinical or radiomics models. multivariate Cox regression analysis identified smoking (HR = 1.417, 95% confidence interval [CI]: 0.875-2.293, p = 0.156), Ki67(HR = 2.426, 95% confidence interval [CI]: 1.506-3.908, p = 0.000) and postRad-S1 (HR = 1.867, 95% CI: 1.053-3.311, p = 0.033) as significant independent covariates associated with high PFS. While Ki67 and postRad-S2 were prognostic factors significantly associated with OS (HR = 1.521, 95% CI: 0.821-2.818, p = 0.183; HR = 1.912, 95% CI: 1.001-3.654, p = 0.050, respectively).

Conclusion

For patients with LAESCC treated with NIC followed by surgery, the combined model effectively evaluated the efficacy of NIC and predicted PFS and OS. Additionally, different independent predictors were associated with PFS and OS, providing clues for future studies.

Current Role of Artificial Intelligence in the Management of Esophageal Cancer

Background/Objectives: Esophageal cancer (EC) represents a major global contributor to cancer-related mortality. The advent of artificial intelligence (AI), including machine learning, deep learning, and radiomics, holds promise for enhancing treatment decisions and predicting outcomes. The aim of this review is to present an overview of the current landscape and future perspectives of AI in the management of EC. Methods: A literature search was performed on MEDLINE using the following keywords: "Artificial Intelligence", "Esophageal cancer", "Barrett's esophagus", "Esophageal Adenocarcinoma", and "Esophageal Squamous cell carcinoma". All titles and abstracts were screened; the results included 41 studies. Results: Over the past five years, the number of studies focusing on the application of AI to the treatment and prognosis of EC has surged, leveraging increasingly larger datasets with external validation. The simultaneous incorporation in AI models of clinical factors and features from several imaging modalities displays improved predictive performance, which may enhance patient outcomes, based on direct personalized therapeutic options. However, clinicians and researchers must address existing limitations, conduct randomized controlled trials, and consider the ethical and legal aspects that arise to establish AI as a standard decision-support tool. Conclusions: AI applications may result in substantial advances in EC management, heralding a new era. Considering the complexity of EC as a clinical entity, the evolving potential of AI is anticipated to ameliorate patients' quality of life and survival rates.

Development of Machine Learning Models for Predicting Radiation Dermatitis in Breast Cancer Patients Using Clinical Risk Factors, Patient-Reported Outcomes, and Serum Cytokine Biomarkers

BACKGROUND

Radiation dermatitis (RD) is a significant side effect of radiotherapy experienced by breast cancer patients. Severe symptoms include desquamation or ulceration of irradiated skin, which impacts quality of life and increases healthcare costs. Early identification of patients at risk for severe RD can facilitate preventive management and reduce severe symptoms. This study evaluated the utility of subjective and objective factors, such as patient-reported outcomes (PROs) and serum cytokines, for predicting RD in breast cancer patients. The performance of machine learning (ML) and logistic regression-based models were compared.

PATIENTS AND METHODS

Data from 147 breast cancer patients who underwent radiotherapy was analyzed to develop prognostic models. ML algorithms, including neural networks, random forest, XGBoost, and logistic regression, were employed to predict clinically significant Grade 2+ RD. Clinical factors, PROs, and cytokine biomarkers were incorporated into the risk models. Model performance was evaluated using nested cross-validation with separate loops for hyperparameter tuning and calculating performance metrics.

RESULTS

Feature selection identified 18 predictors of Grade 2+ RD including smoking, radiotherapy boost, reduced motivation, and the cytokines interleukin-4, interleukin-17, interleukin-1RA, interferon-gamma, and stromal cell-derived factor-1a. Incorporating these predictors, the XGBoost model achieved the highest performance with an area under the curve (AUC) of 0.780 (95% CI: 0.701-0.854). This was not significantly improved over the logistic regression model, which demonstrated an AUC of 0.714 (95% CI: 0.629-0.798).

CONCLUSION

Clinical risk factors, PROs, and serum cytokine levels provide complementary prognostic information for predicting severe RD in breast cancer patients undergoing radiotherapy. ML and logistic regression models demonstrated comparable performance for predicting clinically significant RD with AUC>0.70.

Downregulation of SMAD2 and SMAD4 is associated with poor prognosis and shorter survival in esophageal squamous cell carcinoma

BACKGROUND

Esophageal Squamous Cell Carcinoma (ESCC) presents a serious global health challenge, ranking among the most prevalent cancers worldwide. Small mothers against decapentaplegic 2 (SMAD2) and SMAD4 play a significant role in various types of cancer.

METHODS

This study performed relative mRNA expression level profiling of SMAD2 and SMAD4 using Real time quantitative polymerase chain reaction (qPCR) in tissue and blood of ESCC patients and analyzed their associations with numerous clinical and lifestyle parameters for evaluating prognostic significance along with survival and hazard outcomes.

RESULTS

SMAD2 and SMAD4 relative expression level showed downregulation in both tissue (85% and 87% respectively) and blood samples (80% and 79% respectively), and a significant positive correlation (p < 0.05) between their relative expression level was observed in both tissue and blood levels. Various clinicopathological parameters and food habits revealed significant association (p < 0.05) with SMAD2 and SMAD4 relative expression level. While analyzing survival and hazard in ESCC patients, various parameters revealed significant association (p < 0.05) in univariate model and histopathology grade, node stage, stage of metastasis, betel nut consumption, smoked food consumption and altered SMAD2 and SMAD4 relative expression level in tissue samples revealed significant association (p < 0.05) in the multivariate model, indicating their direct association with ESCC patients' survival and this makes them reliable predictors for ESCC prognosis.

CONCLUSIONS

This study's results revealed that downregulation of SMAD2 and SMAD4 is associated with poor prognosis and ESCC progression emphasizing their potential as potent prognostic factors for survival prediction as well as reliable biomarkers for screening.

Development and validation of a novel artificial intelligence algorithm for precise prediction the postoperative prognosis of esophageal squamous cell carcinoma

BACKGROUND

Esophageal squamous cell carcinoma (ESCC) is a highly aggressive malignancy, and current postoperative prognostic assessment methods remain unsatisfactory, underlining the urgent to develop a reliable approach for precision medicine. Given the similarities with gametogenesis, cancer/testis genes (CTGs) are acknowledged for regulation unrestrained multiplication and immune microenvironment during oncogenic processes. These processes are associated with advanced disease and poorer prognosis, indicating that CTGs could serve as ideal prognostic biomarkers in ESCC. The purpose of this study is to develop a novel clinically prognostic prediction system to facilitate the individualized postoperative care.

METHODS

We conducted LASSO regression analysis of protein-coding CTGs and clinical characteristics from 119 pathologically confirmed ESCC patients to recognize powerful predictive variables. We employed nine supervised machine learning classifiers and integrated best predictive machine learning classifiers by weighted voting method to construct an ensemble model called PPMESCC. Additionally, functional assay was conducted to examine the potential effect of top-ranking CTG HENMT1 in ESCC.

RESULTS

LASSO regression identified five CTGs and TNM stage as optimized prognostic features. Six machine learning classifiers were integrated to construct an ensemble model, PPMESCC, which exhibited outstanding performance in ESCC prediction. The AUC for PPMESCC was 0.9828 (95% confidence interval: 0.9608 to 0.9926), with an accuracy of 98.32% (95% CI: 96.64-99.16%) in the discovery cohort and 0.9057 (95% CI: 0.8897 to 0.9583) of AUC with an accuracy of 90% (95% CI: 89.08-93.28%) in validation cohort. In addition, the top-ranking CTG HENMT1 encodes 2'-O-methyltransferase of piRNAs that was confirmed positively correlated with the proliferation capacity of ESCC cells. Then we systematically screen piRNAs associated with esophageal carcinoma based on GWAS, eQTL-piRNA, and i2OM databases, and successfully discovered 8 piRNAs potentially regulated by HENMT1.

CONCLUSION

The study highlights the clinical utility of PPMESCC algorithm in prognostic prediction that may facilitate to establish the personalized screening and management strategies for postoperative ESCC patients.

Machine learning-driven prediction of medical expenses in triple-vessel PCI patients using feature selection

Revascularization therapies, such as percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG), alleviate symptoms and treat myocardial ischemia. Patients with multivessel disease, particularly those undergoing 3-vessel PCI, are more susceptible to procedural complications, which can increase healthcare costs. Developing efficient strategies for resource allocation has become a paramount concern due to tightening healthcare budgets and the escalating costs of treating heart conditions. Therefore, it is essential to develop an evaluation model to estimate the costs of PCI surgeries and identify the key factors influencing these costs to enhance healthcare quality. This study utilized the National Health Insurance Research Database (NHIRD), encompassing data from multiple hospitals across Taiwan and covering up to 99% of the population. The study examined data from triple-vessel PCI patients treated between January 2015 and December 2017. Additionally, six machine-learning algorithms and five cross-validation techniques were employed to identify key features and construct the evaluation model. The machine learning algorithms used included linear regression (LR), random forest (RF), support vector regression (SVR), generalized linear model boost (GLMBoost), Bayesian generalized linear model (BayesGLM), and extreme gradient boosting (eXGB). Among these, the eXGB model exhibited outstanding performance, with the following metrics: MSE (0.02419), RMSE (0.15552), and MAPE (0.00755). We found that the patient's medication use in the previous year is also crucial in determining subsequent surgical costs. Additionally, 25 significant features influencing surgical expenses were identified. The top variables included 1-year medical expenditure before PCI surgery (hospitalization and outpatient costs), average blood transfusion volume, ventilator use duration, Charlson Comorbidity Index scores, emergency department visits, and patient age. This research is crucial for estimating potential expenses linked to complications from the procedure, directing the allocation of resources in the future, and acting as an important resource for crafting medical management policies.

Integrating single-cell RNA-Seq and machine learning to dissect tryptophan metabolism in ulcerative colitis

BACKGROUND

Ulcerative colitis (UC) is a persistent inflammatory bowels disease (IBD) characterized by immune response dysregulation and metabolic disruptions. Tryptophan metabolism has been believed as a significant factor in UC pathogenesis, with specific metabolites influencing immune modulation and gut microbiota interactions. However, the precise regulatory mechanisms and key genes involved remain unclear.

METHODS

AUCell, Ucell, and other functional enrichment algorithms were utilized to determine the activation patterns of tryptophan metabolism at the UC cell level. Differential analysis identified key genes associated with tryptophan metabolism. Five machine learning algorithms, including Random Forest, Boruta algorithm, LASSO, SVM-RFE, and GBM were integrated to identify and categorize disease-specific characteristic genes.

RESULTS

We observed significant heterogeneity in tryptophan metabolism activity across cell types in UC, with the highest activity levels in macrophages and fibroblasts. Among the key tryptophan metabolism-related genes, CTSS, S100A11, and TUBB were predominantly expressed in macrophages and significantly upregulated in UC, highlighting their involvement in immune dysregulation and inflammation. Cross-analysis with bulk RNA data confirmed the consistent upregulation of these genes in UC samples, highly indicating their relevance in UC pathology and potential as targets for therapeutic intervention.

CONCLUSIONS

This study is the first to reveal the heterogeneity of tryptophan metabolism at the single-cell level in UC, with macrophages emerging as key contributors to inflammatory processes. The identification of CTSS, S100A11, and TUBB as key regulators of tryptophan metabolism in UC underscores their potential as biomarkers and therapeutic targets.

Development of a machine learning-based risk model for postoperative complications of lung cancer surgery

PURPOSE

To develop a comorbidity risk score specifically for lung resection surgeries.

METHODS

We reviewed the medical records of patients who underwent lung resections for lung cancer, and developed a risk model using data from 2014 to 2017 (training dataset), validated using data from 2018 to 2019 (validation dataset). Forty variables were analyzed, including 35 factors related to the patient's overall condition and five factors related to surgical techniques and tumor-related factors. The risk model for postoperative complications was developed using an elastic net regularized generalized linear model. The performance of the risk model was evaluated using receiver operating characteristic curves and compared with the Charlson Comorbidity Index (CCI).

RESULTS

The rate of postoperative complications was 34.7% in the training dataset and 21.9% in the validation dataset. The final model consisted of 20 variables, including age, surgical-related factors, respiratory function tests, and comorbidities, such as chronic obstructive pulmonary disease, a history of ischemic heart disease, and 12 blood test results. The area under the curve (AUC) for the developed risk model was 0.734, whereas the AUC for the CCI was 0.521 in the validation dataset.

CONCLUSIONS

The new machine learning model could predict postoperative complications with acceptable accuracy.

CLINICAL REGISTRATION NUMBER

2020-0375.

Establishment and validation of the prognostic risk model based on the anoikis-related genes in esophageal squamous cell carcinoma

BACKGROUND

Esophageal squamous cell carcinoma (ESCC) is a malignant condition in humans. Anoikis-related genes (ARGs) are crucial to cancer progression. Therefore, more studies on the relationship between ARGs and ESCC are warranted.

METHODS

The study acquired ESCC-related transcriptome data from TCGA. Differentially expressed ARGs (DE-ARGs) were obtained by differential analysis and candidates were filtered out by survival analysis. Prognostic genes were determined by Cox and LASSO regression. A risk model was constructed based on prognostic gene expressions. An immune infiltration study was done to explain how these genes contribute to ESCC development. The IC50 test was adopted to assess the clinical response of chemotherapy drugs. Single cell analysis was performed on the GSE145370 dataset. Moreover, the prognostic gene expressions were detected by qRT-PCR.

RESULTS

53 DE-ARGs were screened and four candidate genes including PBK, LAMC2, TNFSF10 and KL were obtained. Cox and LASSO regression identified the two prognostic genes, TNFSF10 and PBK. Immuno-infiltration analysis revealed positive associations of PBK with Macrophages M0 cells, and TNFSF10 with Macrophages M1 cells. The IC50 values of predicted drugs, in the case of Tozasertib 1096 and WIKI4 1940, were significantly variant between risk groups. Single cell analysis revealed that TNFSF10 and PBK levels were higher in epithelial cells than in other cells. The prognostic genes expression results by qRT-PCR were compatible with the dataset analysis.

CONCLUSION

The study established an ARG prognosis model of ESCC. It provided a reference for the research of ARGs in ESCC.

MiR-101-3p Promotes Tumor Cell Proliferation and Migration via the Wnt Signal Pathway in MNNG-Induced Esophageal Squamous Cell Carcinoma

N-methyl-n'-nitroso-n'-nitroso guanidine (MNNG) can induce esophageal squamous cell carcinoma (ESCC), and microRNAs are associated with the development of ESCC and may serve as potential tumor prognostic markers. Thus, the aim of this study was to evaluate the potential function of miR-101-3p in MNNG-induced ESCC. An investigation of risk factors in patients with ESCC was carried out and the concentration of nine nitrosamines in urine samples was detected by the SPE-GC-MS technique. Then, we performed cancer tissue gene sequencing analysis, and RT-qPCR verified the expression level of miR-101-3p. Subsequently, the relationship between miR-101-3p potential target genes and the ESCC patients' prognosis was predicted. Finally, we investigated the function of miR-101-3p in MNNG-induced ESCC pathogenesis and the regulatory mechanism of the signaling pathway by in vivo and in vitro experiments. The results revealed that high dietary nitrosamine levels are high-risk factors for ESCC. MiR-101-3p is down-regulated in ESCC tissues and cells, and its potential target genes are enriched in cell migration and cancer-related pathways. MiR-101-3p target genes include AXIN1, CK1, and GSK3, which are involved in the regulation of the Wnt signaling pathway. MiR-101-3p overexpression promotes apoptosis and inhibits the proliferation and migration of Eca109 cells. The Wnt pathway is activated after subchronic exposure to MNNG, and the Wnt pathway is inhibited by the overexpression of miR-101-3p in Eca109 cells. Down-regulated miR-101-3p may exert tumor suppressive effects by regulating the Wnt pathway and may be a useful biomarker for predicting ESCC progression.

Integrated machine learning developed a prognosis-related gene signature to predict prognosis in oesophageal squamous cell carcinoma

The mortality rate of oesophageal squamous cell carcinoma (ESCC) remains high, and conventional TNM systems cannot accurately predict its prognosis, thus necessitating a predictive model. In this study, a 17-gene prognosis-related gene signature (PRS) predictive model was constructed using the random survival forest algorithm as the optimal algorithm among 99 machine-learning algorithm combinations based on data from 260 patients obtained from TCGA and GEO. The PRS model consistently outperformed other clinicopathological features and previously published signatures with superior prognostic accuracy, as evidenced by the receiver operating characteristic curve, C-index and decision curve analysis in both training and validation cohorts. In the Cox regression analysis, PRS score was an independent adverse prognostic factor. The 17 genes of PRS were predominantly expressed in malignant cells by single-cell RNA-seq analysis via the TISCH2 database. They were involved in immunological and metabolic pathways according to GSEA and GSVA. The high-risk group exhibited increased immune cell infiltration based on seven immunological algorithms, accompanied by a complex immune function status and elevated immune factor expression. Overall, the PRS model can serve as an excellent tool for overall survival prediction in ESCC and may facilitate individualized treatment strategies and predction of immunotherapy for patients with ESCC.

Artificial intelligence enhances the management of esophageal squamous cell carcinoma in the precision oncology era

Esophageal squamous cell carcinoma (ESCC) is the most common histological type of esophageal cancer with a poor prognosis. Early diagnosis and prognosis assessment are crucial for improving the survival rate of ESCC patients. With the advancement of artificial intelligence (AI) technology and the proliferation of medical digital information, AI has demonstrated promising sensitivity and accuracy in assisting precise detection, treatment decision-making, and prognosis assessment of ESCC. It has become a unique opportunity to enhance comprehensive clinical management of ESCC in the era of precision oncology. This review examines how AI is applied to the diagnosis, treatment, and prognosis assessment of ESCC in the era of precision oncology, and analyzes the challenges and potential opportunities that AI faces in clinical translation. Through insights into future prospects, it is hoped that this review will contribute to the real-world application of AI in future clinical settings, ultimately alleviating the disease burden caused by ESCC.

Unraveling the ecological landscape of mast cells in esophageal cancer through single-cell RNA sequencing

Background

Esophageal cancer (EC) is a major health issue, ranking seventh in incidence and sixth in mortality worldwide. Despite advancements in multidisciplinary treatment approaches, the 5-year survival rate for EC remains low at 21%. Challenges in EC treatment arise from late-stage diagnosis, high malignancy, and poor prognosis. Understanding the tumor microenvironment is critical, as it includes various cellular and extracellular components that influence tumor behavior and treatment response. Mast cells (MCs), as tissue-resident immune cells, play dual roles in tumor dynamics. High-throughput single-cell RNA sequencing offers a powerful tool for analyzing tumor heterogeneity and immune interactions, although its application in EC is limited.

Methods

In this study, we investigated the immune microenvironment of EC using single-cell RNA sequencing and established a comprehensive immune profile. We also performed analysis of upstream transcription factors and downstream pathway enrichment to further comprehensively decipher MCs in EC. Besides, we performed knockdown experiments to explore the role of epidermal growth factor receptor (EGFR) signaling pathway in MCs-tumor cell interactions, highlighting its potential as a prognostic marker. Finally, we constructed a prognostic model for EC, which provided valuable suggestions for the diagnosis and prognosis of EC.

Results

Our analysis identified 11 major cell types, of which MCs were particularly present in pericarcinoma tissues. Further grouping of the 5,001 MCs identified 8 distinct subtypes, including SRSF7-highly expressed MCs, which showed strong tumor preference and potential tumor-promoting properties. Moreover, we identified the key signaling receptor EGFR and validated it by in vitro knockdown experiments, demonstrating its cancer-promoting effects. In addition, we established an independent prognostic indicator, SRSF7+ MCs risk score (SMRS), which showed a correlation between high SMRS group and poor prognosis.

Conclusion

These findings illuminate the complex interactions within the tumor microenvironment of EC and suggest that targeting specific MCs subtypes, particularly via the EGFR signaling pathway, may present novel therapeutic strategies. This study establishes a comprehensive immune map of EC, offering insights for improved treatment approaches.

Prognosis and risk factor assessment of patients with advanced lung cancer with low socioeconomic status: model development and validation

BACKGROUND

Lung cancer, a major global health concern, disproportionately impacts low socioeconomic status (SES) patients, who face suboptimal care and reduced survival. This study aimed to evaluate the prognostic performance of traditional Cox proportional hazards (CoxPH) regression and machine learning models, specifically Decision Tree (DT), Random Forest (RF), Support Vector Machine (SVM), and Extreme Gradient Boosting (XGBoost), in patients with advanced lung cancer with low SES.

DESIGN

A retrospective study.

METHOD

The 949 patients with advanced lung cancer with low SES who entered the hospice ward of a tertiary hospital in Wuhan, China, from January 2012 to December 2021 were randomized into training and testing groups in a 3:1 ratio. CoxPH regression methods and four machine learning algorithms (DT, RF, SVM, and XGBoost) were used to construct prognostic risk prediction models.

RESULTS

The CoxPH regression-based nomogram demonstrated reliable predictive accuracy for survival at 60, 90, and 120 days. Among the machine learning models, XGBoost showed the best performance, whereas RF had the lowest accuracy at 60 days, DT at 90 days, and SVM at 120 days. Key predictors across all models included Karnofsky Performance Status (KPS) score, quality of life (QOL) score, and cough symptoms.

CONCLUSIONS

CoxPH, DT, RF, SVM, and XGBoost models are effective in predicting mortality risk over 60-120 days in patients with advanced lung cancer with low SES. Monitoring KPS, QOL, and cough symptoms is crucial for identifying high-risk patients who may require intensified care. Clinicians should select models tailored to individual patient needs and preferences due to varying prediction accuracies.

REPORTING METHOD

This study was reported in strict compliance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline.

PATIENT OR PUBLIC CONTRIBUTION

No patient or public contribution.

Metabolism score and machine learning models for the prediction of esophageal squamous cell carcinoma progression

The incomplete prediction of prognosis in esophageal squamous cell carcinoma (ESCC) patients is attributed to various therapeutic interventions and complex prognostic factors. Consequently, there is a pressing demand for enhanced predictive biomarkers that can facilitate clinical management and treatment decisions. This study recruited 491 ESCC patients who underwent surgical treatment at Huashan Hospital, Fudan University. We incorporated 14 blood metabolic indicators and identified independent prognostic indicators for overall survival through univariate and multivariate analyses. Subsequently, a metabolism score formula was established based on the biochemical markers. We constructed a nomogram and machine learning models utilizing the metabolism score and clinically significant prognostic features, followed by an evaluation of their predictive accuracy and performance. We identified alkaline phosphatase, free fatty acids, homocysteine, lactate dehydrogenase, and triglycerides as independent prognostic indicators for ESCC. Subsequently, based on these five indicators, we established a metabolism score that serves as an independent prognostic factor in ESCC patients. By utilizing this metabolism score in conjunction with clinical features, a nomogram can precisely predict the prognosis of ESCC patients, achieving an area under the curve (AUC) of 0.89. The random forest (RF) model showed superior predictive ability (AUC = 0.90, accuracy = 86%, Matthews correlation coefficient = 0.55). Finally, we used an RF model with optimal performance to establish an online predictive tool. The metabolism score developed in this study serves as an independent prognostic indicator for ESCC patients.

Explainable machine learning predicts survival of retroperitoneal liposarcoma: A study based on the SEER database and external validation in China

OBJECTIVE

We have developed explainable machine learning models to predict the overall survival (OS) of retroperitoneal liposarcoma (RLPS) patients. This approach aims to enhance the explainability and transparency of our modeling results.

METHODS

We collected clinicopathological information of RLPS patients from The Surveillance, Epidemiology, and End Results (SEER) database and allocated them into training and validation sets with a 7:3 ratio. Simultaneously, we obtained an external validation cohort from The First Affiliated Hospital of Naval Medical University (Shanghai, China). We performed LASSO regression and multivariate Cox proportional hazards analysis to identify relevant risk factors, which were then combined to develop six machine learning (ML) models: Cox proportional hazards model (Coxph), random survival forest (RSF), ranger, gradient boosting with component-wise linear models (GBM), decision trees, and boosting trees. The predictive performance of these ML models was evaluated using the concordance index (C-index), the integrated cumulative/dynamic area under the curve (AUC), and the integrated Brier score, as well as the Cox-Snell residual plot. We also used time-dependent variable importance, analysis of partial dependence survival plots, and the generation of aggregated survival SHapley Additive exPlanations (SurvSHAP) plots to provide a global explanation of the optimal model. Additionally, SurvSHAP (t) and survival local interpretable model-agnostic explanations (SurvLIME) plots were used to provide a local explanation of the optimal model.

RESULTS

The final ML models are consisted of six factors: patient's age, gender, marital status, surgical history, as well as tumor's histopathological classification, histological grade, and SEER stage. Our prognostic model exhibits significant discriminative ability, particularly with the ranger model performing optimally. In the training set, validation set, and external validation set, the AUC for 1, 3, and 5 year OS are all above 0.83, and the integrated Brier scores are consistently below 0.15. The explainability analysis of the ranger model also indicates that histological grade, histopathological classification, and age are the most influential factors in predicting OS.

CONCLUSIONS

The ranger ML prognostic model exhibits optimal performance and can be utilized to predict the OS of RLPS patients, offering valuable and crucial references for clinical physicians to make informed decisions in advance.

Creation of a machine learning-based prognostic prediction model for various subtypes of laryngeal cancer

Depending on the source of the blastophore, there are various subtypes of laryngeal cancer, each with a unique metastatic risk and prognosis. The forecasting of their prognosis is a pressing issue that needs to be resolved. This study comprised 5953 patients with glottic carcinoma and 4465 individuals with non-glottic type (supraglottic and subglottic). Five clinicopathological characteristics of glottic and non-glottic carcinoma were screened using univariate and multivariate regression for CoxPH (Cox proportional hazards); for other models, 10 (glottic) and 11 (non-glottic) clinicopathological characteristics were selected using least absolute shrinkage and selection operator (LASSO) regression analysis, respectively; the corresponding survival models were established; and the best model was evaluated. We discovered that RSF (Random survival forest) was a superior model for both glottic and non-glottic carcinoma, with a projected concordance index (C-index) of 0.687 for glottic and 0.657 for non-glottic, respectively. The integrated Brier score (IBS) of their 1-year, 3-year, and 5-year time points is, respectively, 0.116, 0.182, 0.195 (glottic), and 0.130, 0.215, 0.220 (non-glottic), demonstrating the model's effective correction. We represented significant variables in a Shapley Additive Explanations (SHAP) plot. The two models are then combined to predict the prognosis for two distinct individuals, which has some effectiveness in predicting prognosis. For our investigation, we established separate models for glottic carcinoma and non-glottic carcinoma that were most effective at predicting survival. RSF is used to evaluate both glottic and non-glottic cancer, and it has a considerable impact on patient prognosis and risk factor prediction.

Genetic and molecular characterization of metabolic pathway-based clusters in esophageal squamous cell carcinoma

Esophageal squamous cell carcinoma (ESCC) is one of the most aggressive types of squamous cell carcinoma and represents a significant proportion of esophageal cancer. Metabolic reprogramming plays a key role in the occurrence and development of ESCC. Unsupervised clustering analysis was employed to stratify ESCC samples into three clusters: MPC1-lipid type, MPC2-amino acid type, and MPC3-energy type, based on the enrichment scores of metabolic pathways extracted from the Reactome database. The MPC3 cluster exhibited characteristics of energy metabolism, with heightened glycolysis, cofactors, and nucleotide metabolism, showing a trend toward increased aggressiveness and poorer survival rates. On the other hand, MPC1 and MPC2 primarily involved lipid and amino acid metabolism, respectively. In addition, liquid chromatography‒mass spectrometry-based metabolite profiles and potential therapeutic agents were explored and compared among ESCC cell lines with different MPCs. MPC3 amplified energy metabolism markers, especially carnitines. In contrast, MPC1 and MPC2 predominantly had elevated levels of lipids (primarily triacylglycerol) and amino acids, respectively. Furthermore, MPC3 demonstrated a suboptimal clinical response to PD-L1 immunotherapy but showed increased sensitivity to the doramapimod chemotherapy regimen, as evident from drug sensitivity evaluations. These insights pave the way for a more personalized therapeutic approach, potentially enhancing treatment precision for ESCC patients.

Prediction of five-year survival among esophageal cancer patients using machine learning

Background and aim

Considering the silent progression of esophageal cancer, the survival prediction of this disease is crucial in enhancing the quality of life of these patients globally. So far, no prediction solution has been introduced for the survival of EC in Iran based on the machine learning approach. So, this study aims to develop a prediction model for the five-year survival of EC based on the ML approach to promote clinical outcomes and various treatment and preventive plans.

Material and methods

In this retrospective study, we investigated the 1656 cases of survived and non-survived EC patients belonging to Imam Khomeini Hospital in Sari City from 2013 to 2020. The multivariable regression analysis was used to select the best predictors of five-year survival. We leveraged random forest, eXtreme Gradient Boosting, support vector machine, artificial neural networks, Bayesian networks, J-48 decision tree, and K-nearest neighborhood to develop the prediction models. To get the best model for predicting the five-year survival of EC, we compared them using the area under the receiver operator characteristics.

Results

The age at diagnosis, body mass index, smoking, obstruction, dysphagia, weight loss, lymphadenopathy, chemotherapy, radiotherapy, family history of EC, tumor stage, type of appearance, histological type, grade of differentiation, tumor location, tumor size, lymphatic invasion, vascular invasion, and platelet albumin ratio were considered as the best predictors associated with the five-year survival of EC based on the regression analysis. In this respect, the random forest with the area under the receiver operator characteristics of 0.95 was identified as a superior model.

Conclusion

The experimental results of the current study showed that the random forest could have a significant role in enhancing the quality of care in EC patients by increasing the effectiveness of follow-up and treatment measures introduced by care providers.