Predict Survival Outcomes Research Articles

Association of Increased CT-Attenuation of Visceral Adipose Tissue After Surgery with Poor Survival Outcomes in Patients with Stage II–III Gastric Cancer: A Retrospective Cohort Study

Background/Objectives: This study aimed to investigate whether post-operative changes in the computed tomography (CT)-attenuation of subcutaneous (SAT) and visceral (VAT) adipose tissues were significantly associated with recurrence-free survival (RFS), peritoneal RFS, and overall survival (OS) in patients with stage II–III gastric cancer. Methods: This retrospective study analyzed 243 patients with stage II–III gastric cancer who underwent curative surgery. CT-attenuation values of SAT (SAT HU) and VAT (VAT HU) were measured from non-contrast-enhanced abdominopelvic CT images taken pre-operatively and 6 months post-operatively. Changes in SAT HU (ΔSAT HU) and VAT HU (ΔVAT HU) between the two CT scans were calculated. The prognostic value of these variables for predicting survival outcomes was assessed. Results: Correlation analyses showed that both ΔSAT HU and ΔVAT HU were significantly positively correlated with T stage, TNM stage, and tumor size (p < 0.05). In the multivariate survival analysis, ΔVAT HU emerged as an independent significant predictor for RFS (p = 0.002, hazard ratio, 2.437), peritoneal RFS (p = 0.023, hazard ratio, 2.457), and OS (p = 0.043, hazard ratio, 2.204) after adjusting for age, sex, histopathological classification, T stage, and N stage. Patients with high ΔVAT HU had worse RFS, peritoneal RFS, and OS compared to those with low ΔVAT HU. Conclusions: Change in CT-attenuation of VAT following surgery was significantly correlated with tumor characteristics and was a significant predictor of RFS, peritoneal RFS, and OS in patients with stage II–III gastric cancer.

Peripheral blood PD-1+T lymphocytes as biomarkers in liquid biopsies for solid tumors: Clinical significance and prognostic applications.

A shift toward a T cell exhaustion phenotype is associated with the upregulation of expression of programmed cell death protein 1 (PD-1) on T lymphocytes in patients with malignant solid tumors. The interaction between PD-1 and programmed death-ligand 1 (PD-L1) inhibits PD-1+ T lymphocyte function, impacting their anti-tumor immune activity. Immune checkpoint inhibitors targeting PD-1/PD-L1 have revolutionized the treatment of various solid malignancies, improving therapeutic efficacy and survival outcomes. Peripheral blood analysis of liquid biopsies is being increasingly used to identify populations most likely to benefit from various treatment modalities. PD-1+ T lymphocytes represent the primary cell population responsive to immunotherapeutic interventions for patients with solid malignancies, as evidenced by the altered PD-1 expression levels and proportion of cells comprising the overall population of immunocytes. PD-1+ T cells in peripheral blood exert an associative and reciprocal predictive effect on homologous intratumoral cells. Distinct subpopulations of PD-1+T cells exhibit differential ability to proliferate in the periphery and can be characterized by tumor antigen-specific and exhaustion phenotypes. These characteristics have prognostic implications, aiding in the prediction of the efficacy of antitumor therapy and predicting survival outcomes. We highlight distinct subpopulations of PD-1+T cells, their exhaustion and antigen-specific phenotypes, and their dynamic changes over treatment, providing insights into their utility for tailoring personalized therapies. For the first time, this review discusses the role of peripheral PD-1+T lymphocytes as prognostic biomarkers in liquid biopsies, focusing on their clinical significance, predictive value during therapy, and future research directions.

Prognostic significance of three lymph node staging systems in pancreatic cancer with ≤ 12 and > 12 retrieved lymph nodes.

This study aimed to evaluate and compare the predictive performance of negative lymph nodes (NLN), lymph node ratio (LNR), and N stage in pancreatic ductal adenocarcinoma (PDAC) among patients with ≤ 12 retrieved lymph nodes and those with > 12 retrieved lymph nodes. Moreover, the association between the three nodal staging systems and survival was also explored. Clinical data on patients diagnosed with PDAC between 2004 and 2020 were downloaded from the Surveillance, Epidemiology, and End Results (SEER) database. Cox regression was performed to identify independent predictors of cancer specific survival (CSS) and overall survival (OS). Survival probability was calculated and compared by the Kaplan-Meier method and log rank test. Akaike information criterion (AIC) and Harrell's C-index were used to evaluate the prognostic ability of each nodal staging system. All three lymph node staging systems were independent predictors of CSS and OS. A higher NLN, a lower N stage, and a lower LNR were associated with improved survival. Compared with N stage, LNR staging performed better with a lower AIC and higher C-index for predicting the prognosis regardless of the sufficiency of retrieved lymph nodes, while NLN staging performed poorly in both the training and validation set. Subgroup analyses showed that the NLN successfully predicted survival outcomes in both lymph node-positive and node-negative patients. LNR demonstrated better predictive performance in PDAC patients regardless of the sufficiency of retrieved lymph nodes. Notably, for stage N0 disease, NLN was a more important prognostic predictor. The combination of LNR and NLN may offer more precise information on lymph node staging than the current staging system.

From Prediction to Precision: Explainable AI-Driven Insights for Targeted Treatment in Equine Colic

Colic is a leading cause of mortality in horses, demanding precise and timely interventions. This study integrates machine learning and explainable artificial intelligence (XAI) to predict survival outcomes in horses with colic, using clinical, procedural, and diagnostic data. Random forest and XGBoost emerged as top-performing models, achieving F1 scores of 85.9% and 86.1%, respectively. SHAP (Shapley additive explanations) was employed to provide interpretable insights, offering both global and local explanations for model predictions. The analysis revealed that key features, such as pulse rate, lesion type, and total protein levels, significantly influenced survival likelihood. Local interpretations highlighted the unique contribution of clinical factors to individual cases, enabling personalized insights that guide targeted treatment strategies. These tailored predictions empower veterinarians to prioritize interventions based on the specific conditions of each horse, moving beyond generalized care protocols. By combining predictive accuracy with interpretability, this study advances precision veterinary medicine, enhancing outcomes for equine colic cases and setting a benchmark for future applications of AI in animal health.

Observation on the Therapeutic Efficacy of Camrelizumab Combined with Chemotherapy in Non-small Cell Lung Cancer and the Cutaneous Immune-related Adverse Events: A Retrospective Study.

Immunotherapy targeting PD-1/PD-L1 shows significant benefits in lung cancer. Cutaneous immune-related adverse events (irAEs) are frequent, early-developing side effects of ICIs, and their potential role as prognostic markers in non-small cell lung cancer (NSCLC) therapy requires further exploration. Data of patients with NSCLC treated with camrelizumab Combined with chemotherapy were collected at Xuzhou Medical University from 2019 to 2023. Cutaneous irAEs were monitored using CTCAE v5.0, and therapeutic efficacy was assessed using RECIST 1.1 criteria for ORR and PFS. Multivariable Cox regression analysis identified independent predictors of PFS, and a nomogram was constructed to predict survival outcomes. Data from 151 patients were analyzed. Significant differences in the objective response rate (ORR, P = 0.016) and progression-free survival (PFS, P < 0.0001) were detected between NSCLC patients, either with cirAEs or not. Besides, PFS was significantly different in NSCLC patients who were subgrouped by the time of first cutaneous irAEs occurrence (P = 0.011), duration of cutaneous irAEs (P = 0.002), grade of cutaneous irAEs (P = 0.002), the number of cutaneous irAEs(P = 0.021). The multivariable analysis also revealed that cirAEs were positively associated with survival outcomes (HR: 0.316, 95% CI, 0.193- 0.519, P＜0.001) for PFS. The nomogram was formulated based on the results of multivariate analysis and validated using an internal bootstrap resampling approach, which showed that the nomogram exhibited a sufficient level of discrimination according to the C-index 0.80 (95% CI, 0.748-0.850). The presence of cirAEs in NSCLC patients treated with camrelizumab combined with chemotherapy is indicative of better treatment efficacy and prognosis. This study supports the utility of cirAEs as biomarkers for predicting the validity of immunotherapy in NSCLC. It proposes a novel, multi-parameter prognostic model to assess patient outcomes more accurately.

Elevated neutrophil-percentage-to-albumin ratio predicts increased all-cause and cardiovascular mortality in hypertensive patients: Evidence from NHANES 1999-2018.

This study investigates the association between the neutrophil-percentage-to-albumin ratio (NPAR) and all-cause mortality in patients with hypertension, and its relationship with cardiovascular mortality. This study examined data from 18,469 adults with hypertension in the National Health and Nutrition Examination Survey (1999-2018) and the mortality data from the National Death Index. The link between NPAR and mortality risk was visualized using restricted cubic splines. The optimal NPAR cut-off value for the prediction of survival outcomes was identified via maximally selected rank statistics. We employed weighted multivariate Cox regression and subgroup analyses to evaluate the relationship between NPAR and risk of all-cause and cardiovascular mortality. We assessed NPAR's predictive accuracy for survival outcomes using time-related receiver operating characteristic analysis. During a median follow-up of 105months, 31.8% of 18,469 participants died, with 8.9% from cardiovascular causes. Restricted cubic splines analysis showed a positive link between NPAR and both all-cause and cardiovascular mortality. Cox models indicated that higher NPAR (>1602.08) significantly raised risks of all-cause (HR 1.80, 95% CI 1.54-2.12, p<0.0001) and cardiovascular mortality (HR 1.54, 95% CI 1.24-1.91, p<0.0001). The stability of results was confirmed through stratified and interaction analyses. The area under the curve for 3-, 5-, and 10-year survival were 0.67, 0.65, and 0.63 for all-cause mortality and 0.61, 0.62, and 0.63 for cardiovascular mortality. Elevated NPAR independently raises the risk of all-cause and cardiovascular mortality in hypertensive patients, indicating its potential in the clinic as a practical tool for predicting long-term mortality risks and influencing treatment strategies.

Commentator Discussion: Preoperatively predicting survival outcome for clinical stage IA pure-solid non–small cell lung cancer by radiomics-based machine learning

Commentator Discussion: Preoperatively predicting survival outcome for clinical stage IA pure-solid non–small cell lung cancer by radiomics-based machine learning

Large Unstained Cells (LUC): A Novel Predictor of CDK4/6 Inhibitor Outcomes in HR+ HER2-Negative Metastatic Breast Cancer.

Background: Although CDK4/6 inhibitors combined with endocrine therapies have improved outcomes in HR+ HER2-negative metastatic breast cancer, predictive biomarkers for treatment response and adverse effects remain limited. This study assessed the prognostic and predictive value of large unstained cells (LUC), a subset of white blood cells that may reflect immune status or treatment response. Methods: A retrospective analysis of 210 patients with HR+ HER2-negative metastatic breast cancer treated with CDK 4/6 inhibitors between 2021 and 2024 was conducted. Clinical data, including demographics, tumor characteristics, and treatment regimens, were analyzed. Based on LUC levels, progression-free survival (PFS), overall survival (OS), and adverse events were evaluated. Results: The cohort had a median age of 57, of which 78% were postmenopausal. Common metastatic sites included bone (67%) and liver (24%). At a median follow-up of 18.5 months, the PFS and OS rates were 65% and 83%. Patients with low LUC levels had significantly shorter PFS (OR: 1.91; p = 0.014) and OS (OR: 2.39; p = 0.012), while high LUC levels correlated with a lower incidence of grade 3 neutropenia (OR: 0.49; p = 0.017). Liver metastasis and prior treatments were also linked to shorter survival. Conclusions: LUC levels emerge as a promising biomarker for predicting survival outcomes and the risk of neutropenia in HR+ HER2-negative metastatic breast cancer patients treated with CDK 4/6 inhibitors and endocrine therapy, showing their potential to guide personalized treatment approaches.

Mechanistic Learning for Predicting Survival Outcomes in Head and Neck Squamous Cell Carcinoma.

We employed a mechanistic learning approach, integrating on-treatment tumor kinetics (TK) modeling with various machine learning (ML) models to address the challenge of predicting post-progression survival (PPS)-the duration from the time of documented disease progression to death-and overall survival (OS) in Head and Neck Squamous Cell Carcinoma (HNSCC). We compared the predictive power of model-derived TK parameters versus RECIST and assessed the efficacy of nine TK-OS ML models against conventional survival models. Data from 526 advanced HNSCC patients treated with chemotherapy and cetuximab in the TPExtreme trial were analyzed using a double-exponential model. TK parameters from the first line and maintenance (TKL1) or after four cycles (TK4) were used to predict PPS and post-cycle 4 OS (OS4), combined with 12 baseline parameters. While ML algorithms underperformed compared to the Cox model for PPS, a random survival forest was superior for OS prediction using TK4 and surpassed RECIST-based metrics. This model demonstrated unbiased OS4 prediction, suggesting its potential for improving HNSCC treatment evaluation. Trial Registration: ClinicalTrials.gov identifier: NCT02268695.

Prognostic value of early response in predicting survival in hepatocellular carcinoma patients treated with selective internal radiation therapy.

This study evaluates the prognostic value of tumor response on CT at 3 months, assessed by Response Evaluation Criteria in Solid Tumors (RECIST), modified RECIST (mRECIST), and Liver Imaging Reporting and Data System Treatment Response Algorithm (LR-TRA) in patients with hepatocellular carcinoma (HCC) treated with selective internal radiation therapy (SIRT). A retrospective analysis was conducted on 102 HCC patients treated with SIRT between 2018 and 2020. RECIST, mRECIST, and LR-TRA were assessed at 3 months post-SIRT. Overall survival (OS) and progression-free survival (PFS) were assessed using Kaplan-Meier analysis and Cox proportional-hazards models. The median age was 71 years, and most patients (90%) had advanced-stage tumors (Barcelona Clinic Liver Cancer-C). After a median follow-up of 32.0 months (95% CI: 16.8-60.9), 60/102 patients died (59%), and 90/102 patients showed tumor progression (88%). Median OS was 20.4 months (95% CI: 15.4-33.0), and median PFS was 14.5 months (95% CI: 6.5-24.5); 1-year OS and PFS rates were 65.6% and 50.7%. Multivariable analysis revealed that early response according to RECIST 1.1 (HR 1.66, p = 0.30), mRECIST (HR 1.40, p = 0.215), and LR-TRA (HR 0.67, p = 0.30) were not predictors of OS. Disease progression evaluated by RECIST (HR 2.55, p < 0.001) and mRECIST (HR 2.53, p < 0.001), bilirubin levels (HR 1.03, p < 0.001), and prothrombin time (HR 0.98, p = 0.005) were predictors of OS. For PFS, neither RECIST nor mRECIST response, disease progression, nor LR-TRA viability were predictors. In this advanced-stage HCC population, early response assessed by RECIST, mRECIST, and LR-TRA criteria did not predict OS or PFS after SIRT. However, early disease progression and liver function indicators were prognostic factors for OS. QuestionHow well does early tumor response, assessed at 3 months post-selective internal radiation therapy (SIRT), predict survival in advanced hepatocellular carcinoma (HCC) patients? Findings Early response, assessed by RECIST, mRECIST, and LR-TRA, did not predict overall or progression-free survival; disease progression and liver function indicators were significant predictors. Clinical relevance This study highlights the limitations of early imaging criteria in predicting survival outcomes in advanced HCC post-SIRT, suggesting the need for alternative or complementary prognostic indicators to guide treatment decisions.

Machine Learning Models in Sepsis Outcome Prediction for ICU Patients: Integrating Routine Laboratory Tests-A Systematic Review.

Background. Sepsis presents significant diagnostic and prognostic challenges, and traditional scoring systems, such as SOFA and APACHE, show limitations in predictive accuracy. Machine learning (ML)-based predictive survival models can support risk assessment and treatment decision-making in the intensive care unit (ICU) by accounting for the numerous and complex factors that influence the outcome in the septic patient. Methods. A systematic literature review of studies published from 2014 to 2024 was conducted using the PubMed database. Eligible studies investigated the development of ML models incorporating commonly available laboratory and clinical data for predicting survival outcomes in adult ICU patients with sepsis. Study selection followed the PRISMA guidelines and relied on predefined inclusion criteria. All records were independently assessed by two reviewers, with conflicts resolved by a third senior reviewer. Data related to study design, methodology, results, and interpretation of the results were extracted in a predefined grid. Results. Overall, 19 studies were identified, encompassing primarily logistic regression, random forests, and neural networks. Most used datasets were US-based (MIMIC-III, MIMIC-IV, and eICU-CRD). The most common variables used in model development were age, albumin levels, lactate levels, and ventilator. ML models demonstrated superior performance metrics compared to conventional methods and traditional scoring systems. The best-performing model was a gradient boosting decision tree, with an area under curve of 0.992, an accuracy of 0.954, and a sensitivity of 0.917. However, several critical limitations should be carefully considered when interpreting the results, such as population selection bias (i.e., single center studies), small sample sizes, limited external validation, and model interpretability. Conclusions. Through real-time integration of routine laboratory and clinical data, ML-based tools can assist clinical decision-making and enhance the consistency and quality of sepsis management across various healthcare contexts, including ICUs with limited resources.

Association of NLR with all-cause and cardiovascular mortality in adults with coronary heart disease: 1999-2018 NHANES data analysis.

The neutrophil-to-lymphocyte ratio (NLR) is an important inflammatory marker. However, the relationship between NLR and the prognosis of patients with coronary heart disease (CHD) remains unclear. The purpose of this study is to explore the relationship between NLR and all-cause mortality and cardiovascular mortality in CHD patients. This study analyzed data from 1625 CHD patients who participated in the National Health and Nutrition Examination Survey from 1999 to 2018. Multivariate Cox regression analysis was used to explore the relationship between mortality risk and NLR. The optimal NLR cutoff value related to survival outcomes was determined using the maximum selected rank method. Restricted cubic spline analysis was performed to investigate the correlation between NLR and mortality risk in CHD patients. Moreover, subgroup analyses were conducted to assess the relationship between NLR and all-cause and cardiovascular mortality in different populations. Additionally, time-dependent receiver operating characteristic curves were used to evaluate the accuracy of NLR in predicting survival outcomes. During a median follow-up of 88 months, a total of 475 patients experienced all-cause mortality, and 278 patients experienced cardiovascular mortality. After adjusting for confounding factors, compared with CHD patients with higher NLR, those with lower NLR had a 43% reduced risk of all-cause mortality (hazard ratio: 0.57, 95% CI: 0.41-0.8) and a 51% reduced risk of cardiovascular mortality (hazard ratio: 0.49, 95% CI: 0.3-0.78). Kaplan-Meier analysis showed that the survival rate in the high NLR group was significantly lower in terms of all-cause and cardiovascular mortality rates than in the low NLR group (P < .0001). The results of the restricted cubic spline analysis indicated a nonlinear relationship between NLR and all-cause mortality as well as cardiovascular mortality in CHD patients. In addition, receiver operating characteristic analysis showed that the area under the curve for all-cause mortality at 3 years, 5 years, and 10 years were 0.596, 0.591, and 0.604, while the area under the curve for cardiovascular mortality were 0.623, 0.617, and 0.623, in CHD patients. Elevated NLR is associated with increased risk of cardiovascular and all-cause mortality in CHD patients, and NLR can independently predict the prognosis of CHD patients.

Machine learning approach identifies inflammatory gene signature for predicting survival outcomes in hepatocellular carcinoma

BackgroundHepatocellular carcinoma (HCC) is a leading cause of cancer-related deaths worldwide, often linked to chronic inflammation. Our study aimed to probe inflammation pathways at the genetic level and pinpoint biomarkers linked to HCC patient survival.MethodsWe analyzed gene transcriptome data from 246 resectable stage I and II HCC patients from The Cancer Genome Atlas (TCGA). After selecting 917 inflammation-related genes (IRGs), we identified 104 differentially expressed genes (DEGs) through differential expression analysis. Two significant prognostic DEGs, S100A9 and PBK, were identified using LASSO and Cox regression, forming the basis of a risk score model. We conducted functional enrichment and immune landscape analyses, validated our findings on 170 patients from the GSE14520 dataset, and performed mutational analysis using TCGA somatic mutation data.ResultsWe analyzed 296 samples (246 HCC, 50 normal liver), showing significant survival differences between high and low-risk groups based on our risk score model. Functional enrichment analysis unveiled inflammation-associated pathways. Validation using the GSE14520 dataset confirmed our risk score’s predictive ability, and we explored clinical correlations.ConclusionOur study delineates inflammation-related genomic changes in HCC, unveiling prognostic biomarkers with potential therapeutic implications. These findings deepen our understanding of HCC molecular mechanisms and may guide personalized therapeutic approaches, ultimately improving patient outcomes.

An integrated machine learning framework for developing and validating a prognostic risk model of gastric cancer based on endoplasmic reticulum stress-associated genes

BackgroundGastric cancer (GC), a prevalent and deadly malignancy, demonstrates poor survival outcomes. Evidence has emerged indicating that disruptions in endoplasmic reticulum homeostasis are significantly implicated in the onset and progression of various oncological conditions. This study was designed to construct a prognostic model based on genes related to endoplasmic reticulum stress(ERS) to predict survival outcomes in patients with GC. MethodsExpression profiling data for GC samples were extracted and analyzed from TCGA-STAD, revealing 214 genes related to endoplasmic reticulum stress that show differential expression when compared with normal gastric tissue. Building on these insights, a prognostic model was formulated using data from TCGA-STAD and validated through subsequent analyses of GEO datasets. The tumor immune dysfunction and exclusion(TIDE) algorithm was applied to determine the susceptibility of individuals in high- and low-risk categories to immunotherapy. The presence of immune and stromal cells within the tumor microenvironment was assessed with the aid of the ESTIMATE algorithm. Sensitivity variations to prevalent anticancer drugs between the risk groups were evaluated using the Genomics of Drug Sensitivity in Cancer(GDSC) database, and prospective therapeutic agents were confirmed through molecular docking techniques. ResultsThirty-one endoplasmic reticulum stress (ERS)-related differentially expressed genes (DEGs) crucial for prognosis in GC were pinpointed. These DEGs were then used to construct a prognostic model and were considered as independent prognostic factors for GC patients. This risk model proved to have a good predictive performance for estimating the overall survival of these patients. The patients placed into the high-risk group showed worse results and lower sensitivity to immunotherapy. Moreover, five specific targeted therapy drugs, namely BMS-754807, Dasatinib, JQ1, AZD8055 and SB505124, produced better results in the treatment of the high-risk group of patients. ConclusionsA new molecular prognostic model associated with ERS was established and validated for GC and showed relatively good discriminative and predictive ability. This model greatly expands the collection of weapons in the armoury of prognostic analysis in GC.

Nomograms for predicting the prognosis in multiple primary esophageal squamous cell carcinoma

Background Due to its rarity, it is challenging to predict the survival of patients with synchronous multiple primary esophageal squamous carcinomas (SMPESCs). We aimed to construct nomograms to predict survival outcomes and help to make therapeutic strategy for patients with SMPESCs. Materials and Methods The clinical and survival data of 135 patients with SMPESCs were analyzed retrospectively. Univariate and multivariate Cox analyses were used to identify independent prognostic factors. Nomograms were constructed to predict 1-year, 3-year and 5-year disease-free survival (DFS) and overall survival (OS). In addition, we further evaluated the effect of postoperative adjuvant therapy on SMPESCs patients with lymph node metastasis. Results In univariate and multivariate analyses of DFS and OS, age, site of the main lesion, lymph node metastasis, total number of lymph nodes dissected, lactate dehydrogenase level and lymphocyte-to-monocyte ratio were identified as independent prognostic factors. These characteristics were further included to establish nomograms. For the internal validation of the nomogram predictions of survival outcomes, the concordance indices were 0.752 and 0.756, respectively. Decision curve analysis also proved the efficacy of the nomograms. Furthermore, adjuvant therapy had a statistically significant benefit for OS but not DFS in patients with lymph node metastasis. Conclusions These nomograms could effectively predict the 1-year, 3-year and 5-year survival outcomes of patients with SMPESCs. Furthermore, adjuvant therapy has the potential to improve OS in patients with lymph node metastasis.

The prognostic role of KRAS and GNAS in the prediction of survival outcomes in PMP.

The prognostic role of KRAS and GNAS in the prediction of survival outcomes in PMP.

Tissue Cadherin 17 (CDH17): An Important Prognostic Determinant of Colorectal Cancer Using Digital Image Analysis

ABSTRACTBackgroundColorectal cancer (CRC) is a commonly diagnosed malignancy with significant mortality rates worldwide. The identification of robust prognostic biomarkers for prediction of survival outcomes and recurrence can aid disease management and improve patients' quality of living.AimThis study aimed to investigate the prognostic value of cadherin 17 (CDH17) tissue expression in CRC patients by utilizing a standardized automated immunohistochemistry (IHC) platform integrated with a digitalized scoring system.Methods and ResultsIHC was conducted to assess CDH17 expression on tumor tissues obtained from 150 retrospective CRC cases. A computer‐assisted imaging analysis was performed to quantify CDH17 tissue expression using an IHC scoring algorithm known as the Membrane (M) Score. The relationship between CDH17 M Score and clinicopathological factors such as TNM staging, distant metastasis, and recurrence status was analyzed. The prognostic value of CDH17 M Score was determined by Kaplan–Meier curves and Cox regression analysis for overall survival (OS) and recurrence‐free survival (RFS). Comparison of M Score and pathologist visual scoring was made to assess the validity of software‐derived IHC scoring. CDH17 expression, as measured by M Score, was found to be significantly increased in tumor tissues compared to adjacent normal tissues, and its expression is associated with advanced staging and distant metastasis (normal vs. tumor, p = 0.0011; Stages IV vs. I–III, p &lt; 0.0162; with vs. without metastasis, p = 0.0026; Mann–Whitney U test). Prognostic analysis revealed that high CDH17 M Score was associated with poor OS and RFS (OS, p = 0.0118; RFS, p = 0.0021; log‐rank test). Furthermore, multivariate Cox regression analysis identified that CDH17 M Score was an independent prognostic predictor for OS (HR = 2.296, 95% CI = 1.154–4.968, p = 0.0240) and RFS (HR = 2.489, 95% CI = 1.062–6.494, p = 0.0447).ConclusionIncreased CDH17 M Score was associated with advanced tumor staging and poor survival outcomes using an automated IHC system integrated with a digital image analysis software, highlighting CDH17 could serve as an independent prognostic marker for CRC patients.

Mucoepidermoid carcinoma: Enhancing diagnostic accuracy and treatment strategy through machine learning models and web-based prognostic tool.

Oral cancer, particularly mucoepidermoid carcinoma (MEC), presents diagnostic challenges due to its histological diversity and rarity. This study aimed to develop machine learning (ML) models to predict survival outcomes for MEC patients and pioneer a clinically accessible prognostic tool. Using the SEER database (2000-2020), we constructed predictive models with five ML algorithms: Random Forest Classifier (RFC), Gradient Boosting Classifier (GBC), Logistic Regression (LR), K-Nearest Neighbors (KNN), and Multilayer Perceptron (MLP). Predictive variables were identified via Cox regression, and Kaplan-Meier analysis assessed survival trends. Model performance was validated through the area under the curve (AUC) of receiver operating characteristic (ROC) curves. This study included 1314 patients diagnosed with MEC of the oral cavity. The RFC demonstrated the highest predictive accuracy (AUC = 0.55), followed by the GBC and RFC (AUC = 0.53). The most affected primary site was the hard palate, followed by the retromolar and cheek mucosa. Survival rates varied with the treatment modality, with the highest rates observed in patients undergoing surgery alone. ML models have identified age, sex, and metastasis as significant prognostic factors influencing survival outcomes, underscoring the complexity and heterogeneity of MEC. This study highlights ML's potential to enhance survival predictions and personalize treatment for MEC patients. We developed the first web-based prognostic tool, providing a novel, accessible solution for improving clinical decision-making in MEC.

Novel models based on machine learning to predict the prognosis of metaplastic breast cancer.

Metaplastic breast cancer (MBC) is a rare and highly aggressive histological subtype of breast cancer. There remains a significant lack of precise predictive models available for use in clinical practice. This study utilized patient data from the SEER database (2010-2018) for data analysis. We utilized prognostic factors to develop a novel machine learning model (CatBoost) for predicting patient survival rates. Simultaneously, our hospital's cohort of MBC patients was utilized to validate our model. We compared the benefits of radiotherapy among the three groups of patients. The CatBoost model we developed exhibits high accuracy and correctness, making it the best-performing model for predicting survival outcomes in patients with MBC (1-year AUC=0.833, 3-year AUC=0.806; 5-year AUC=0.810). Furthermore, the CatBoost model maintains strong performance in an external independent dataset, with AUC values of 0.937 for 1-year survival, 0.907 for 3-year survival, and 0.890 for 5-year survival, respectively. Radiotherapy is more suitable for patients undergoing breast-conserving surgery with M0 stage [group1: (OS:HR=0.499, 95%CI 0.320-0.777 p<0.001; BCSS: HR=0.519, 95%CI 0.290-0.929 p=0.008)] and those with T3-4/N2-3M0 stage undergoing mastectomy [group2: (OS:HR=0.595, 95%CI 0.437-0.810 p<0.001; BCSS: HR=0.607, 95%CI 0.427-0.862 p=0.003)], compared to patients with stage T1-2/N0-1M0 undergoing mastectomy [group3: (OS:HR=1.090, 95%CI 0.673-1.750 p=0.730; BCSS: HR=1.909, 95%CI 1.036-3.515 p=0.038)]. We developed three machine learning prognostic models to predict survival rates in patients with MBC. Radiotherapy is considered more appropriate for patients who have undergone breast-conserving surgery with M0 stage as well as those in stage T3-4/N2-3M0 undergoing mastectomy.

Prognostic prediction for HER2-low breast cancer patients using a novel machine learning model

BackgroundsTo develop a machine learning (ML) model for predicting the prognosis of breast cancer (BC) patients with low human epidermal growth factor receptor 2 (HER2) expression, and to investigate the association between clinicopathological characteristics and outcomes in HER2-low BC (HLBC) patients.MethodsA retrospective analysis was conducted on data from 998 female HLBC patients treated at the Breast Center of the Fourth Hospital of Hebei Medical University (Hebei, China) between January 1, 2017, and December 31, 2020. To address class imbalance, the synthetic minority over-sampling technique was applied. Feature selection was performed using the least absolute shrinkage and selection operator, followed by construction of the prediction model using the random forest algorithm. Model performance, including specificity and accuracy, was assessed using the receiver operating characteristic (ROC) curve and confusion matrix, comparing it against other ML models. Additionally, the log-rank test was employed to examine the relationship between selected features and patient outcomes in HLBC.ResultsThe random survival forest model demonstrated superior accuracy and specificity in predicting survival outcomes for HLBC patients. Compared with other ML models, it achieved more precise predictions of Disease-Free Survival (DFS) at 1, 2, and 3 years, with the area under the ROC curve (AUC) in the test and training cohorts measured at 0.726 and 0.819, 0.712 and 0.776, and 0.685 and 0.774, respectively. The analysis further identified a strong correlation between poor prognosis in HLBC patients and factors such as axillary lymph node dissection, family history, elevated topoisomerase (TOPO)-2 expression, advanced clinical stage, negative progesterone receptor status, P53 mutation, and increased Ki67 expression, observed across both cohorts.ConclusionsA novel ML model was developed for accurate prognosis prediction in HLBC patients, offering valuable insights into prognostic risk factors. This model equips clinicians with enhanced data to guide treatment decisions, ultimately contributing to improved patient outcomes.