Survival Prediction Model Research Articles

Key oncogenes and candidate drugs for hepatitis-B-driven hepatocellular carcinoma progression.

This study aimed to uncover the key hepatitis-B (HB)-related liver cancer (LC) promoting genes, and clarity their interrelationships, enrichments, impacts on LC immune infiltration, and potential drugs targeting these genes. The LC-survival associated genes were acquired from the LIHC samples of the TCGA-database; and HB related genes from the DisGeNET database. The intersection was used to screen the key genes. Using the 8 HB-LC genes, we constructed prognostic models for survival prediction of HBV positive patients with LIHC and performed enrichment analysis, interaction analysis, immune infiltration analysis, and potential drug digging from the GTRP and GDSC databases. In the core intersection of different sets. Based on these genes, prognostic cox regression models for OS and DFS were constructed. Overall, HB-LC genes were significantly negatively correlated with Th17, MAIT, monocytes, and CD4 Naive cells, while they were positively correlated with B cells, nTreg cells, and Tr1 cells. Among 8 genes, MKI67, EZH2, and CDCA5 were hub ones. Finally, 7 drugs target at least three HB-LC genes and can be used as novel drugs. Together, eight key HB-LC genes play important cancer-promoting roles in LC, which may be the molecular mechanism by which HBV drives the development of LC.

Predictive models for post-liver transplant survival using machine learning techniques in three critical time intervals

Predictive models for post-liver transplant survival using machine learning techniques in three critical time intervals

Development and validation of a prognostic prediction model for elderly gastric cancer patients based on oxidative stress biochemical markers

BackgroundThe potential of the application of artificial intelligence and biochemical markers of oxidative stress to predict the prognosis of older patients with gastric cancer (GC) remains unclear.MethodsThis retrospective multicenter study included consecutive patients with GC aged ≥ 65 years treated between January 2012 and April 2018. The patients were allocated into three cohorts (training, internal, and external validation). The GC-Integrated Oxidative Stress Score (GIOSS) was developed using Cox regression to correlate biochemical markers with patient prognosis. Predictive models for five-year overall survival (OS) were constructed using random forest (RF), decision tree (DT), and support vector machine (SVM) methods, and validated using area under the curve (AUC) and calibration plots. The SHapley Additive exPlanations (SHAP) method was used for model interpretation.ResultsThis study included a total of 1,859 older patients. The results demonstrated that a low GIOSS was a predictor of poor prognosis. RF was the most efficient method, with AUCs of 0.999, 0.869, and 0.796 in the training, internal validation, and external validation sets, respectively. The DT and SVM models showed low AUC values. Calibration and decision curve analyses demonstrated the considerable clinical usefulness of the RF model. The SHAP results identified pN, pT, perineural invasion, tumor size, and GIOSS as key predictive features. An online web calculator was constructed based on the best model.ConclusionsIncorporating the GIOSS, the RF model effectively predicts postoperative OS in older patients with GC and is a robust prognostic tool. Our findings emphasize the importance of oxidative stress in cancer prognosis and provide a pathway for improved management of GC.Trial registrationRetrospectively registered at ClinicalTrials.gov (trial registration number: NCT06208046, date of registration: 2024–05-01).

Effect of visceral fat area on prognosis of patients undergoing radical gastrectomy and construction of nomogram

BackgroundWe aim to investigate the impact of visceral fat area (VFA) on the prognosis of patients following radical gastric resection and develop a nomogram prediction model to forecast the prognosis of gastric cancer patients.MethodsWe retrospectively analyzed 156 patients who underwent laparoscopic radical gastrectomy for distal gastric cancer in the 900th hospital of the Joint Logistics Support Force from April 2018 to April 2020. We collected the CT image data and clinicopathological data one week prior to the operation and then used software to calculate the VFA, dividing it into two groups: a low VFA group (n = 71) and a high VFA group (n = 85). We compared the clinicopathological characteristics and early postoperative complications of the two groups. The Pearson χ2 test was used to analyze the correlation between body mass index (BMI) and VFA. We used the Kaplan-Meier method to draw the survival curve, analyzed the independent risk factors affecting the prognosis of gastric cancer patients using univariate and multivariate Cox regression models, and established a nomogram model for patient prognosis prediction.ResultsThe results of CT showed that VFA value was (95.89 ± 41.40) cm², and body mass index (BMI) was positively correlated with VFA value (r = 0.291, P < 0.001). The ROC curve shows that VFA can predict the prognosis of patients with gastric cancer significantly better than BMI (AUC = 0.826 vs. AUC = 0.707, P = 0.016). The incidence of incision fat liquefaction, pancreatic fistula, and abdominal infection in the high VFA group was higher than that in the low VFA group (P < 0.05). We followed up with all patients for 0.5–48.5 months, with a median follow-up time of 30 months. We used the Kaplan-Meier method to draw the survival curve. The results showed that the overall survival rate of patients in the high VFA group was significantly higher than that in the low VFA group (χ2 = 38.208, P < 0.001), and the high BMI group was significantly higher than that in the low BMI group (χ2 = 29.767, P < 0.001). Age, the degree of differentiation, complications after surgery, VFA, ASA grading, and TNM staging were all found to have independent effects on the prognosis of gastric cancer patients (Multivariate Cox regression analysis). Multivariate Cox regression analysis led to the construction of a nomogram prediction model for the total survival of gastric cancer patients. Its internal verification C-index was 0.881 (95% CI: 0.852–0.910), and the calibration chart showed good consistency.ConclusionsAge, differentiation degree, postoperative complications, VFA, ASA grading, and TNM staging are independent influencing factors for the prognosis of patients with gastric cancer. The constructed nomogram has excellent prediction accuracy and is helpful to evaluate the prognosis of patients.

Development of a modified nutritional index model based on nutritional status and sarcopenia to predict long-term survival and chemotherapy benefits in elderly patients with advanced gastric cancer

BackgroundElderly patients with advanced gastric cancer have poor prognoses. This study aims to develop a prediction model for long-term survival after radical surgery and to identify patients who may benefit from chemotherapy. MethodsData from 555 elderly patients with advanced gastric cancer admitted to two medical centers from 2009 to 2018 were retrospectively analyzed. Sarcopenia was combined with the Controlling Nutritional Status (CONUT) score to create a modified nutritional index (mCONUT). Cox regression analyses were used to develop a novel nomogram prediction model (mCNS) that combined mCONUT, pN, and tumor size, and its performance was further verified both internally and externally. ResultsMultivariate Cox analysis revealed that tumor size, pN, and mCONUT were independent prognostic risk factors for overall survival (OS). The mCNS model showed good fit and high predictive value (AUC: training set 0.711; validation set 0.707), outperforming the pTNM model (p<0.05). To further investigate the association between the model and adjuvant chemotherapy, we categorized the model into two risk groups : a high-risk group and a low-risk group. Further analysis revealed that, in the low-risk group, the OS and recurrence-free survival(RFS) for patients receiving adjuvant chemotherapy was significantly better than that of those who did not receive chemotherapy (p=0.047,p=0.019). In the high-risk group, this result was not observed (p=0.120, p=0.053). ConclusionThe mCNS model has high predictive value in predicting long-term survival of elderly patients with advanced gastric cancer. Patients with mCNS-L were able to benefit from chemotherapy after laparoscopic radical gastrectomy.

External Validation of Twelve Existing Survival Prediction Models for Patients with Spinal Metastases.

Survival prediction models for patients with spinal metastases may inform patients and clinicians in shared decision-making. To externally validate all existing survival prediction models for patients with spinal metastases DESIGN: Prospective cohort study using retrospective data PATIENT SAMPLE: 953 patients OUTCOME MEASURES: Survival in months, area under the curve (AUC), and calibration intercept and slope METHOD: This study included patients with spinal metastases referred to a single tertiary referral center between 2016-2021. Twelve models for predicting 3, 6, and 12-month survival were externally validated Bollen, Mizumoto, Modified Bauer, New England Spinal Metastasis Score, Original Bauer, Oswestry Spinal Risk Index (OSRI), PathFx, Revised Katagiri, Revised Tokuhashi, Skeletal Oncology Research Group Machine Learning Algorithm (SORG-MLA), Tomita, and Van der Linden. Discrimination was assessed using (AUC) and calibration using the intercept and slope. Calibration was considered appropriate if calibration measures were close to their ideal values with narrow confidence intervals. In total, 953 patients were included. Survival was 76.4% at 3 months (728/953), 62.2% at 6 months (593/953), and 50.3% at 12 months (479/953). Revised Katagiri yielded AUCs of 0.79 (95%CI 0.76;0.82) to 0.81 (95%CI 0.79;0.84), Bollen yielded AUCs of 0.76 (95%CI 0.73;0.80) to 0.77 (95%CI 0.75;0.80), and OSRI yielded AUCs of 0.75 (95%CI 0.72;0.78) to 0.77 (95%CI 0.74;0.79). The other nine prediction models yielded AUCs ranging from 0.59 (95%CI 0.55;0.63) to 0.76 (95%CI 0.74;0.79). None of the twelve models yielded appropriate calibration. Twelve survival prediction models for patients with spinal metastases yielded poor to fair discrimination and poor calibration. Survival prediction models may inform decision-making in patients with spinal metastases, provided that recalibration using recent patient data is performed. This study was funded by the AOSpine under the Discovery & Innovation award (AOS-DIA-22-012-TUM). A total amount of CHF 40,000 ($45,000) was received.

Machine Learning-Based Real-Time Survival Prediction for Gastric Neuroendocrine Carcinoma.

This study aimed to develop a dynamic survival prediction model utilizing conditional survival (CS) analysis and machine learning techniques for gastric neuroendocrine carcinomas (GNECs). Data from the Surveillance, Epidemiology, and End Results (SEER) database (2004-2015) were analyzed and split into training and validation groups (7:3 ratio). CS profiles for patients with GNEC were examined in the full cohort. We utilized random survival forests (RSFs) and least absolute shrinkage and selection operator (LASSO) regression, alongside stepwise Cox regression, for variable selection. A CS-based nomogram was developed on the basis of key prognostic factors, followed by risk stratification and model validation. We included 654 patients with GNEC in our study, with 457 assigned to the training set and 197 to the validation set. The CS analysis demonstrated that the probability of achieving 5-year CS improved from 48% immediately after diagnosis to 68%, 81%, 88%, and 94% after surviving an additional year (i.e., at 1, 2, 3, and 4 years, respectively). Through the use of RSFs and LASSO regression, combined with multivariable regression analysis, we identified the optimal combination of prognostic factors, which included age, tumor grade, tumor stage, surgery, and chemotherapy. Utilizing these prognostic indicators, we successfully developed a nomogram model that incorporated CS and effectively stratified these patients by risk. Subsequent performance analyses further validated the superior efficacy of the nomogram. Our study highlights the value of CS in GNEC prognosis. The nomogram offers dynamic, individualized survival predictions, supporting personalized treatment strategies.

Development and external validation of a model for lung transplantation-free survival prediction of primary Sjögren's syndrome-associated interstitial lung disease: a prospective cohort study

BackgroundThe primary Sjögren's Syndrome-associated Interstitial Lung Disease (pSS-ILD) patients have a worse prognosis than pSS patients without pulmonary involvement. This study aims to establish a prediction nomogram for early prediction of lung transplantation (LTx)-free survival in pSS-ILD patients.MethodsThe training cohort comprised 260 patients from China-Japan Friendship Hospital between January 1, 2016, and June 30, 2022, while the external validation cohort consisted of 135 patients from Beijing Chaoyang Hospital between January 1, 2007, and December 31, 2012. Univariable Cox regression analysis and least absolute shrinkage and selection operator (LASSO) were employed for variable selection, and a nomogram model was developed to predict the 1 -, 3 -, 5-year LTx-free survival. Both discrimination and calibration of the nomogram were assessed using concordance index (C-index), area under the curve (AUC), calibration curve and decision curve analysis (DCA).ResultsMultivariable Cox regression demonstrated that elevated age, oxygenation index, CA125 and fibrosis score were independent risk factors for LTx-free survival in pSS-ILD patients. The C-index displayed of the training and validation cohorts were 0.812 and 0.809. The 1-, 3-, 5-year AUC values in training and validation cohorts were 0.781, 0.874, 0.909 and 0.793, 0.826, 0.863 respectively. Bias-corrected curve is close to the ideal curve and revealed a strong consistency between actual observation and prediction.

Increased nerve density adversely affects outcome in colorectal cancer and denervation suppresses tumor growth

BackgroundThe colon and rectum are highly innervated, with neural components within the tumor microenvironment playing a significant role in colorectal cancer (CRC) progression. While perineural invasion (PNI) is associated with poor prognosis in CRC, the impact of nerve density and diameter on tumor behavior remains unclear. This study aims to evaluate the prognostic value of nerve characteristics in CRC and to verify the impact of nerves on tumor growth.MethodsTissue samples from 129 CRC patients were stained with immunofluorescent markers NF-L and S100B to detect nerves. Nerve diameter and density were measured and normalized. Kaplan-Meier survival analysis and Cox regression models were used to identify prognostic factors. Prognostic models were established using receiver operating characteristic (ROC) curve analysis to assess the predictive value of neural factors. A murine chemical denervation model was employed to disrupt sympathetic nerves using 6-hydroxydopamine, inhibit muscarinic receptor 3 with darifenacin, and ablate sensory neurons with capsaicin.ResultsThe total nerve density was 0.72 ± 0.59/mm², with intratumoral (0.42 ± 0.40/mm²) being significantly lower than extratumoral regions (1.00 ± 0.75/mm²). The average nerve diameter was 28.14 ± 6.04 μm, with no significant difference between intratumoral (28.2 ± 7.65 μm) and extratumoral regions (27.86 ± 6.72 μm). PNI was observed in 65 patients (50.4%). PNI and high normalized nerve density (NND) were associated with shorter overall survival and disease-free survival in CRC patients, with PNI identified as an independent prognostic factor. Patients with PNI exhibit higher NND. Incorporating PNI and NND into ROC curve analysis improved the sensitivity and specificity of survival predictions. In the murine model, chemical denervation of sympathetic, parasympathetic, and sensory nerves significantly reduced rectal tumor volume.ConclusionsPNI and NND are critical factors influencing CRC patient survival and enhance the accuracy of survival prediction models. Moreover, chemical denervation effectively inhibits rectal tumor growth in vivo, highlighting the potential of neural targeting as a therapeutic strategy in CRC.

Expert Judgment Supporting a Bayesian Network to Model the Survival of Pancreatic Cancer Patients.

Purpose: Pancreatic cancer is known for its poor prognosis. The most effective treatment combines surgery with peri-operative chemotherapy. Current prognostic tools are designed to predict patient outcomes and inform treatment decisions based on collected data. Bayesian networks (BNs) can integrate objective data with subjective clinical insights, such as expert opinions, or they can be independently based on either element. This pilot study is one of the first efforts to incorporate expert opinions into a prognostic model using a Bayesian framework. Methods: A clinical hybrid BN was selected to model the long-term overall survival of pancreatic cancer patients. The SHELF expert judgment method was employed to enhance the BN's effectiveness. This approach involved a two-phase protocol: an initial single-center pilot phase followed by a definitive international phase. Results: Experts generally agreed on the distribution shape among the 12 clinically relevant predictive variables identified for the BN. However, discrepancies were noted in the tumor size, age, and ASA score nodes. With regard to expert concordance for each node, tumor size, and ASA score exhibited absolute concordance, indicating a strong consensus among experts. Ca19.9 values and resectability status showed high concordance, reflecting a solid agreement among the experts. The remaining nodes showed acceptable concordance. Conclusions: This project introduces a novel clinical hybrid Bayesian network (BN) that incorporates expert elicitation and clinical variables present at diagnosis to model the survival of pancreatic cancer patients. This model aims to provide research-based evidence for more reliable prognosis predictions and improved decision-making, addressing the limitations of existing survival prediction models. A validation process will be essential to evaluate the model's performance and clinical applicability.

Identification and validation of mitochondria-related LncRNA signatures as a novel prognostic model for glioma.

A predictive model for long-term survival is needed, and mitochondrial dysfunction is a key feature of cancer metabolism, though its link to glioma is not well understood. The aim of this study was to identify the molecular characteristics associated with glioma prognosis and explore its potential function. We analyzed RNA-seq data from The Cancer Genome Atlas and identified differentially expressed mitochondrial long noncoding RNAs (lncRNAs) using R's 'limma' package. A prognostic model was developed using 10 selected lncRNAs and validated with Cox regression and least absolute shrinkage and selection operator algorithm. The model's efficacy was assessed using Kaplan-Meier and receiver operating characteristic curve analyses, and its correlation with immune cell profiles and drug sensitivity was explored. A 10-mitochondria-related LncRNA signature was generated. The median risk score values are used to classify glioma samples into low-risk and high-risk groups. In breast patients, the signature-based risk score demonstrated a more potent ability to predict survival than conventional clinicopathological features. Furthermore, we noted a substantial disparity in the number of immune cells, including B cells, CD8+T cells, and macrophages, between the two groups. In addition, the high-risk group exhibited lower half-maximal inhibitory concentration values for specific chemotherapy medications, including bortezomib, luminespib, rapamycin, and 5-fluorouracil. Our study elucidates the diagnostic and prognostic value of mitochondria-related-lncRNAs in the promotion, suppression, and treatment of glioma.

A prognostic survival model based on endocrine-related gene expression in acute myelogenous leukemia.

Accurate prediction of survival in patients with acute myelogenous leukemia (AML) is challenging. Therefore, we developed a predictive survival model using endocrine-related gene expression to identify an endocrine signature for accurate stratification of AML prognosis. RNA matrices and clinical data for AML were downloaded from a training dataset (GEO) and two validation datasets (TCGA and TARGET). In relation to the survival outcome, a risk model was constructed by incorporating seven endocrine-related genes. The model exhibited favorable predictive efficacy in estimating 5-year survival rates, as demonstrated by both the training and validation cohorts. Multivariable analysis revealed that the endocrine signature demonstrated autonomous prognostic significance in the aforementioned cohorts. Prediction accuracy for 5-year overall survival increased using a nomogram combining endocrine risk score and classical prognostic factors compared with using classical prognostic factors alone. The model predictions were confirmed using AML cell lines. The endocrine-related prognostic model established in this study improves AML survival prediction accuracy.

Machine Learning Algorithms in Controlled Donation After Circulatory Death Under Normothermic Regional Perfusion: A Graft Survival Prediction Model.

Several scores have been developed to stratify the risk of graft loss in controlled donation after circulatory death (cDCD). However, their performance is unsatisfactory in the Spanish population, where most cDCD livers are recovered using normothermic regional perfusion (NRP). Consequently, we explored the role of different machine learning-based classifiers as predictive models for graft survival. A risk stratification score integrated with the model of end-stage liver disease score in a donor-recipient (D-R) matching system was developed. This retrospective multicenter cohort study used 539 D-R pairs of cDCD livers recovered with NRP, including 20 donor, recipient, and NRP variables. The following machine learning-based classifiers were evaluated: logistic regression, ridge classifier, support vector classifier, multilayer perceptron, and random forest. The endpoints were the 3- and 12-mo graft survival rates. A 3- and 12-mo risk score was developed using the best model obtained. Logistic regression yielded the best performance at 3 mo (area under the receiver operating characteristic curve = 0.82) and 12 mo (area under the receiver operating characteristic curve = 0.83). A D-R matching system was proposed on the basis of the current model of end-stage liver disease score and cDCD-NRP risk score. The satisfactory performance of the proposed score within the study population suggests a significant potential to support liver allocation in cDCD-NRP grafts. External validation is challenging, but this methodology may be explored in other regions.

A novel prediction model for survival in individual patients with cardiac resynchronization therapy with a defibrillator: Analysis of the new Japan cardiac device treatment registry database.

Accurate prediction for survival in individualized patients with cardiac resynchronization therapy with a defibrillator (CRT-D) is difficult. We analyzed the New Japan cardiac device treatment registry (JCDTR) database to develop a survival prediction model for CRT-D recipients. Four hundred and eighty-two CRT-D recipients, at the implantation year 2018-2021, with a QRS width ≥120 ms and left ventricular ejection fraction (LVEF) ≤35% at baseline, were analyzed. During an average follow-up of 21 ± 10 months, death occurred in 66 of 482 CRT-D patients (14%). A prediction model estimating annual survival probability was developed using Cox regression with internal validation. With seven explanation predictors (age >75 years, serum creatinine >1.4 mg/dL, blood hemoglobin <12 g/dL, heart rate ≥90/min, LVEF, prior NSVT, and QRS width <150 ms), the model distinguished patients with and without all-cause death, with an optimism-corrected C-statistics of 0.766, 0.764, and 0.768, and calibration slope of 1.01, 1.00, and 1.00 at 1 year, 2 years, and 3 years. Additionally, we have devised the calculator of survival probability for individual CRT-D recipients. Using routine available variables, we have developed a survival prediction model for individual CRT-D recipients.

Development of metastasis and survival prediction model of luminal and non-luminal breast cancer with weakly supervised learning based on pathomics.

Breast cancer stands as the most prevalent form of cancer among women globally. This heterogeneous disease exhibits varying clinical behaviors. The stratification of breast cancer patients into risk groups, determined by their metastasis and survival outcomes, is pivotal for tailoring personalized treatments and therapeutic interventions. The pathological sections of radical specimens encompass a diverse range of histological information pertinent to the metastasis and survival of patients. In this study, our objective is to develop a deep learning model utilizing pathological images to predict the metastasis and survival outcomes for breast cancer patients. This study utilized pathological sections from 204 radical mastectomy specimens obtained between January 2013 and December 2014 at the Second Affiliated Hospital of the Medical College of Zhejiang University. The 204 pathological slices were scanned and transformed into whole slide imaging (WSI), with manual labeling of all tumor areas. The WSI was then partitioned into smaller tiles measuring 512 × 512 pixels. Three networks, namely Densely Connected Convolutional Network 121 (DenseNet121), Residual Network (ResNet50), and Inception_v3, were assessed. Subsequently, we combined patch-level predictions, probability histograms, and Term Frequency-Inverse Document Frequency (TF-IDF) features to create comprehensive participants representations. These features served as the foundational input for developing a machine learning algorithm for metastasis analysis and a Cox regression model for survival analysis. Our results show that the Inception_v3 model shows a particularly robust patch recognition ability for estrogen receptor (ER) recognition. Our pathological model shows high accuracy in predicting tumor regions. The train area under the curve (AUC) of the Inception_v3 model based on supervised learning is 0.975, which is higher than the model established by weakly supervised learning. But the AUC of the metastasis prediction in training and testing sets is higher than value based on supervised learning. Furthermore, the C-index of the survival prediction model is 0.710 in the testing sets, which is also better than the value by supervised learning. Our study demonstrates the significant potential of deep learning models in predicting breast cancer metastasis and prognosis, with the pathomic model showing high accuracy in identifying tumor areas and ER status. The integration of clinical features and pathomics signature into a nomogram further provides a valuable tool for clinicians to make individualized treatment decisions.

Construction of an Oxidative Stress Risk Model to Analyze the Correlation Between Liver Cancer and Tumor Immunity.

Hepatocellular carcinoma (HCC) remains one of the most lethal cancers globally. Despite advancements in immunotherapy, the prognosis for patients with HCC continues to be poor. As oxidative stress plays a significant role in the onset and progression of various diseases, including metabolism-related HCC, comprehending its mechanism in HCC is critical for effective diagnosis and treatment. This study utilized the TCGA dataset and a collection of oxidative stress genes to determine the expression of oxidative stress-related genes in HCC and their association with overall survival using diverse bioinformatics methods. A novel prognostic risk model was developed, and the TCGA cohort was divided into high-risk and low-risk groups based on each tumor sample's risk score. Levels of immune cell infiltration and the expression of immune checkpoint-related genes in different risk subgroups were analyzed to investigate the potential link between tumor immunity and oxidative stress-related features. The expression of model genes in actual samples was validated through immunohistochemistry, and their mRNA and protein expression levels were measured in cell cultures. Four oxidative stress-related genes (EZH2, ANKZF1, G6PD, and HMOX1) were identified and utilized to create a predictive risk model for HCC patient overall survival, which was subsequently validated in an independent cohort. A significant correlation was found between the expression of these prognostic genes and the infiltration of tumor immune cells. Elevated expression of EZH2, ANKZF1, G6PD, and HMOX1 was observed in both HCC tissues and cell lines. The combined assessment of EZH2, ANKZF1, G6PD, and HMOX1 gene expression can serve as a model to evaluate the risk of oxidative stress in HCC. Furthermore, there is a notable correlation between the expression of these risk model genes and tumor immunity.

Development and verification of prognostic nomogram for extraskeletal Ewing's sarcoma based on the SEER database

BackgroundExtraskeletal Ewing's sarcoma (EES) is a rare tumor, and there is currently no predictive model for overall survival of EES patients. This study sought to use data from the Surveillance, Epidemiology, and End Results (SEER) database to develop a clinical predictive model that could be used to assess the prognosis of EES patients. MethodsWe selected and downloaded prognostic data on 356 patients diagnosed with extraskeletal Ewing’s sarcoma based on screening criteria,These patients were distributed between 2004 and 2015. 356 patients were randomly divided into a training cohort (70%) and an internal validation cohort (30%). After univariate or multifactor Cox regression analysis, the relevant variables were screened and a nomogram was constructed. The consistency index (C-index), calibration chart and receiver operating characteristic (ROC) curve were used to evaluate the established nomogram. The clinical utility of the model was verified by clinical decision curve.Study conducted and outcomes reported according to STROBE statement. ResultsAfter multifactor regression analysis, we identified five factors that were significantly associated with EES prognosis, and subsequently established a nomogram. Verification data showed that the C-index of this nomogram was 0.829 (95% CI 0.774 -- 0.884). the AUCs of the nomogram for predicting the 3- and 5-year OS were 0.91 and 0.863. the calibration curves and Decision curve analysis showed that nomogram could predict the prognosis of EES patients. ConclusionStage, age, tumour size, chemotherapy, and surgery may be independent prognostic factors for EES. our study produced a survival nomogram that can be used to predict the prognosis of patients with EES and validated its performance, which may help clinicians evaluate the condition of patients with EES and choose personalized treatment.

Comparison of in vitro cell survival predictions using Monte Carlo methods for proton irradiation.

It is possible to combine theoretical models with Monte Carlo simulations to investigate the relationship between radiation-induced initial DNA damage and cell survival. Several combinations of models have been proposed in recent years, sparking interest in comparing their predictions in view of future clinical applications. Two in silico methods for calculating cell survival fractions were optimized for proton irradiation of the Chinese hamster V79 cell line, for LET values ranging from 3.40 and 100 keV/μm. These methods, based on different Monte Carlo codes and theoretical models, were benchmarked against published V79 cell survival data to identify the sources of discrepancies. The predictive capacities of the methods were evaluated for several proton LET values using an external dataset. After recalibrating model parameters, multiple methods were assessed. This approach helped identify sources of variation, the main one being the simulated number of DSBs, which differed by a factor up to 3 between the two Monte Carlo codes. In this process a new method was defined, that, in all but one case, allows for a reduction in prediction error of up to 56%. Additionally, a freely available GUI for computing cell survival was refined, to facilitate further comparison of diverse theoretical models. The systematic comparison of two predictive chains, characterized by distinct applicability ranges and features, was conducted. Optimization and analysis of various combinations were undertaken to elucidate differences. Addressing and minimizing such discrepancies will be crucial for further enhancing the reliability of predictive models of cell survival, aiming for biologically informed treatment planning.

Predictive models are indeed useful for causal inference.

The subject of investigating causation in ecology has been widely discussed in recent years, especially by advocates of a structural causal model (SCM) approach. Some of these advocates have criticized the use of predictive models and model selection for drawing inferences about causation. We argue that the comparison of model-based predictions with observations is a key step in hypothetico-deductive (H-D) science and remains a valid approach for assessing causation. We draw a distinction between two approaches to inference based on predictive modeling. The first approach is not guided by causal hypotheses and focuses on the relationship between a (typically) single response variable and a potentially large number of covariates. We agree that this approach does not yield useful inferences about causation and is primarily useful for hypothesis generation. The second approach follows a H-D framework and is guided by specific hypotheses about causal relationships. We believe that this has been, and continues to be, a useful approach to causal inference. Here, we first define different kinds of causation, arguing that a "probability-raisers-of-processes" definition is especially appropriate for many ecological systems. We outline different scientific "designs" for generating the observations used to investigate causation. We briefly outline some relevant components of the SCM and H-D approaches to investigating causation, emphasizing a H-D approach that focuses on modeling causal effects on vital rate (e.g., rates of survival, recruitment, local extinction, colonization) parameters underlying system dynamics. We consider criticisms of predictive modeling leveled by some SCM proponents and provide two example analyses of ecological systems that use predictive modeling and avoid these criticisms. We conclude that predictive models have been, and can continue to be, useful for providing inferences about causation.

Predicting Survival Outcomes for Patients with Ovarian Cancer Using National Cancer Registry Data from Taiwan: A Retrospective Cohort Study.

Ovarian cancer is one of the top seven causes of cancer deaths. Incidence of ovarian cancer varies by ethnicity, where Asian women demonstrate lower incidence rates than non-Hispanic Blacks and Whites. Survival prediction models for ovarian cancer have been developed for Caucasians and Black populations using national databases; however, whether these models work for Asians is unclear. Therefore, a retrospective cohort study was conducted to develop survival prediction models for patients with epithelial ovarian cancer from a Taiwan Cancer Registry (TCR) who underwent de-bulking and chemotherapy, with the aim to identify variables that can predict prognosis accurately. Patients diagnosed with OC from TCR were included. Two prognostic models (M1 and M2) were developed: M1 utilized clinical variables only, M2 additionally included cancer-specific variables with the aim to improve the accuracy. All methods were repeated independently for patients with only serous ovarian cancer. All findings for model M1 were validated among Black, White, and Asian populations from Surveillance, Epidemiology, and End Results (SEER) database and 10-fold internal cross-validations. Due to absence of cancer-specific site variables in SEER, model M2 was only internally validated. Cox-proportional hazards regression analysis was performed and a stepwise strategy with Akaike-information criterion was used to select appropriate variables as predictors to develop both M1 and M2. The c-index values of both models were >0.7 in both TCR and SEER populations for epithelial ovarian cancer. Calibration analysis demonstrated good prediction performance with the proportional difference between predicted and observed survival to be <5%. The performance was similar for the subset of patients with serous epithelial ovarian cancer. Notably, no significant racial differences were observed. The prognostic models proposed in this study can potentially be used for identifying patients, especially from Taiwan, at higher risk of ovarian cancer mortality early on, leading to improved prognosis, through shared decision-making between physicians and patients.