Tool For Predicting Survival Research Articles

737. ETHNIC, GENDER AND TUMOR-SPECIFIC ASSESSMENT OF BODY COMPOSITION IS CRITICAL FOR CLINICAL RELEVANCE IN PATIENTS UNDERGOING THERAPY FOR ESOPHAGEAL

Abstract Background Growing evidence has supported a role for body composition in predicting survival in patients undergoing esophagectomy. A contemporary meta-analysis of published data (n=44 studies) conducted for this study confirmed an association between reduced muscle mass at diagnosis and worse overall survival (HR 1.72, 95%CI 1.49-1.97). Variation in methods of assessing and defining parameters of body composition remain however an important barrier to the wider adoption in routine clinical practice. This is particularly pertinent in esophageal cancer due to known sex, biological and geographical difference in tumor subtypes. This study seeks to define baseline variation in body composition parameters and to report their association with 90-day and overall survival in a large international cohort of patients undergoing esophagectomy for cancer. Methods Patients who underwent esophagectomy in fourteen high-volume centers between January 2007 to December 2019 were eligible for inclusion. Survival data were included up until June 2022. Skeletal muscle, visceral and subcutaneous adipose tissues within CT images, acquired at the time of diagnosis, were analyzed in accordance with a standardized protocol. Results 1716 patients were recruited to this study from three global regions: North America (n=386;22.5%), Europe (n=940;54.8%) and Asia (n=390;22.7%). Patients were predominantly male (79.5%) and adenocarcinoma was the most common histological subtype (66.6%). Wide variation of skeletal muscle (17.1-83.5cm2), visceral adiposity (0.2-170.1cm2) and subcutaneous adiposity (0.2-216.8cm2) was observed at presentation. Characteristics significantly associated with levels of muscle and adiposity were global region, sex, age, and histological subtype (P<0.001) (Figure 1). Compared to adenocarcinoma, squamous cell carcinoma was associated with significantly lower levels of muscle and adiposity, a finding that was independent of global region, sex, and age using a multivariable linear regression model (p<.0001). Median follow-up was 71.3 (95%CI 69.1-73.5) months with a median overall survival of 60.9 (95%CI 53.0-68.9) months. Reduced skeletal muscle and an excess of total adiposity at the time of diagnosis was associated with worse 90-day mortality and long-term survival. An interactive online survival prediction tool was developed to provide personalized survival estimates based on relevant predictors including skeletal muscle, total adiposity at diagnosis and other tumor and patient specific variables. Conclusions This is the first and largest study to adopt standardized methods to define international variation in parameters of body composition in esophageal cancer patients. Findings corroborate the potential role of body composition as a method of prognostication in esophageal cancer and presents a on-line tool for convenient integration into clinical practice.

CORR Insights®: Comparison of Two Modern Survival Prediction Tools, SORG-MLA and METSSS, in Patients With Symptomatic Long-bone Metastases Who Underwent Local Treatment With Surgery Followed by Radiotherapy and With Radiotherapy Alone.

CORR Insights®: Comparison of Two Modern Survival Prediction Tools, SORG-MLA and METSSS, in Patients With Symptomatic Long-bone Metastases Who Underwent Local Treatment With Surgery Followed by Radiotherapy and With Radiotherapy Alone.

Development and validation of novel interpretable survival prediction models based on drug exposures for severe heart failure during vulnerable period

BackgroundSevere heart failure (HF) has a higher mortality during vulnerable period while targeted predictive tools, especially based on drug exposures, to accurately assess its prognoses remain largely unexplored. Therefore, this study aimed to utilize drug information as the main predictor to develop and validate survival models for severe HF patients during this period.MethodsWe extracted severe HF patients from the MIMIC-IV database (as training and internal validation cohorts) as well as from the MIMIC-III database and local hospital (as external validation cohorts). Three algorithms, including Cox proportional hazards model (CoxPH), random survival forest (RSF), and deep learning survival prediction (DeepSurv), were applied to incorporate the parameters (partial hospitalization information and exposure durations of drugs) for constructing survival prediction models. The model performance was assessed mainly using area under the receiver operator characteristic curve (AUC), brier score (BS), and decision curve analysis (DCA). The model interpretability was determined by the permutation importance and Shapley additive explanations values.ResultsA total of 11,590 patients were included in this study. Among the 3 models, the CoxPH model ultimately included 10 variables, while RSF and DeepSurv models incorporated 24 variables, respectively. All of the 3 models achieved respectable performance metrics while the DeepSurv model exhibited the highest AUC values and relatively lower BS among these models. The DCA also verified that the DeepSurv model had the best clinical practicality.ConclusionsThe survival prediction tools established in this study can be applied to severe HF patients during vulnerable period by mainly inputting drug treatment duration, thus contributing to optimal clinical decisions prospectively.

Comparison of Two Modern Survival Prediction Tools, SORG-MLA and METSSS, in Patients With Symptomatic Long-bone Metastases Who Underwent Local Treatment With Surgery Followed by Radiotherapy and With Radiotherapy Alone.

Comparison of Two Modern Survival Prediction Tools, SORG-MLA and METSSS, in Patients With Symptomatic Long-bone Metastases Who Underwent Local Treatment With Surgery Followed by Radiotherapy and With Radiotherapy Alone.

Correlation between Palliative Performance Scale results and observed survival among Mexican patients with advanced cancer.

e24040 Background: Cancer is a common cause of death among Hispanics. Estimating patient survival is critical in cancer care since it allows better shared decision-making and avoids use of futile therapy. Palliative Performance Scale (PPS) has been widely used for survival prediction among patients with cancer. However, its performance has not been validated across populations in Latin America, which could have different outcomes. The aim of this study was to analyze the correlation between PPS score and observed survival among patients with advanced cancer to facilitate physician and patient decision-making near the end of life. Methods: We conducted a retrospective observational analysis of patients with advanced cancer included in “ Te Acompañamos” cohort, a multidisciplinary patient navigator-led supportive care program in a single center in Mexico City. We used descriptive statistics to analyze clinico-demographic variables, PPS score, observed survival in days (from recruitment to death). Pearson correlation coefficients were used for correlating observed survival in days and predicted survival using PPS. U de Mann Whitney was used for comparing survival days between PPS groups. Results: We analyzed 246 patients with advanced cancer included in “Te Acompañamos” who died between August 2017-January 2024. Among these patients, 143 (58%) were women. Mean age was 64 years (SD 13). 153 patients (62%) had gastrointestinal (GI), 43 (17%) gynecologic, 19 (7%) genitourinary, 5 (2%) lung, 5 (2%) thyroid, and 21 (8%) other solid malignancies. Pancreatic cancer was the most common cancer (n = 60, 24%). Patients with GI cancer more frequently reported pain (59%) and had an average PPS of 67.7%. Mean survival was 233 days (SD 247) and mean predicted survival in days according to PPS was 68 days (SD 20). Correlation between observed survival days and PPS-predicted survival was low (R = 0.29). Patients with PPS < 50% had mean survival of 74.6 days while patients with PPS > 50% of 243.2 days (p < 0.05). Patients aged < 60 years with GI tumors had a higher PPS score, while older patients had a lower score (74.6% vs 65%; p < 0.05). No other significant differences were found in other subgroups. Conclusions: Our results show weak correlation between observed and predicted survival in days using PPS. However, lower PPS was associated with decreased survival. Among patients with GI cancers we observed that older patients had the lowest PPS scores. This study emphasizes the need to develop and validate more precise tools for survival prediction across patient populations, including minorities such as the Hispanics. Our study suggests that strategies to diagnose and alleviate pain promptly in patients with advanced cancer near the end of life are urgently required.

#1756 Development and internal validation of machine learning algorithms for mortality prediction model of people with DM and CKD

Abstract Background and Aims Diabetes Mellitus (DM) and Chronic Kidney Disease (CKD) are leading causes of morbidity and mortality, presenting challenges in patient management. The aim of this study is to develop a machine learning-based predictive model for mortality in DM and CKD patients, improving early intervention and treatment personalization. Method 3,637 participants with DM and CKD from the National Health and Nutrition Examination Survey (NHANES) from 1999 to 2018 were included. All-cause mortality was ascertained by linkage to National Death Index records through 31 December 2019 in NHANES. The dataset included clinical profiles, demographic details, and laboratory results of patients with DM and CKD. We performed 100 random groupings, dividing the data into a 75% training set and a 25% validation set each time. We then visualized the results of 100 AUC measurements using a boxplot. The objective was to forecast patient survival over 1, 3, 5, and 10-year horizons using a variety of machine learning algorithms. The algorithms tested included Logistic Regression (LR), Decision Tree (DT), Random Forest (RF), K-Nearest Neighbors (KNN), Gaussian Naive Bayes (GNB), Multinomial Naive Bayes (MNB), Bayesian Network Classifier (BNC), AdaBoost (Ada), Gradient Boosting (Gradient), and Extreme Gradient Boosting (XG). The performance of each algorithm was evaluated based on the Area Under the Receiver Operating Characteristic (ROC) Curve (AUC), with the results visualized in boxplot format. Results Upon result shown as the boxplots, it becomes evident that the Random Forest (RF) algorithm consistently exhibits one of the highest AUCs on average across all prediction intervals, suggesting a strong and stable predictive capacity. In the 1-year survival prediction, the RF algorithm demonstrates a highest AUC, implying consistent performance. The 3 and 5-year predictions reveal a slight dip in AUC for most models, including RF, but it still remains one of the top-performing algorithms with fewer outliers than XG and Gradient Boosting. The mean AUC of RF is particularly noteworthy at the 10-year mark, where most algorithms struggle with prediction accuracy, as indicated by lower median AUCs and greater variability. Notably, the RF algorithm identified six major factors for 1-year survival prediction: systolic blood pressure, uric acid, blood urea nitrogen, lactate dehydrogenase, lymphocyte count and total cholesterol, as well as three major factors for 10-year survival prediction: age, lymphocyte and gamma-glutamyl transferase. Conclusion Overall, the study indicates that while machine learning can be a powerful tool for survival prediction in patients with DM and CKD, the choice of algorithm is crucial, with RF standing out for its consistent and reliable performance.

Machine Learning-based Online Survival Prediction Tool for Adrenocortical Carcinoma

Machine Learning-based Online Survival Prediction Tool for Adrenocortical Carcinoma

A Prognosis and Immunological Study of Cellular Aging Related lncRNAs in Lung Adenocarcinoma

Background: Recently, cellular senescence has been recognized as a novel hallmark of cancer. However, the mechanisms that control cellular senescence remain poorly understood. This study seeks to discover long non-coding RNAs (lncRNAs) that are linked to senescence and can predict outcomes in lung adenocarcinoma (LUAD) patients. Methods: Using RNA sequencing data from the Cancer Genome Atlas Lung Adenocarcinoma (TCGA-LUAD) and senescence genes from the CellAge database, a group of lncRNAs related to senescence was initially identified. Subsequently, through univariate and multivariate Cox regression analyses, a senescence-related lncRNA signature (SenLncSig) linked to the prognosis of LUAD was developed. LUAD patients were categorized into high-risk and low-risk groups based on the median risk score of SenLncSig. Kaplan-Meier analysis was employed to assess differences in overall survival (OS) between these subgroups. Additionally, disparities in Gene Set Enrichment Analysis (GSEA), immune infiltration, scores from the tumor immune dysfunction and exclusion (TIDE) module, and choices of chemotherapy and targeted therapy were also evaluated between the high-risk and low-risk groups. Results: Based on univariate and multivariate Cox regression analyses, we established a prognostic model composed of seven senescence-related lncRNAs, named SenLncSig, which includes AL031775.2, AC026355.1, AL365181.2, AC090559.1, AL606489.1, AL513550.1, and C20orf197. This model divides patients into high-risk and low-risk groups, with the high-risk group showing significantly shorter overall survival compared to the low-risk group, with a hazard ratio (HR) of 1.326. The predictive accuracy of SenLncSig was validated through ROC curves and principal component analysis (PCA). Additionally, we developed an intuitive survival prediction tool that combines SenLncSig with clinical pathological features to assess patients' overall survival (OS). Enrichment analysis (GSEA) revealed that SenLncSig is involved in multiple pathways related to tumor immune modulation. Risk model analysis indicated significant differences between the high and low-risk groups in terms of immune status and responses to immunotherapy, chemotherapy, and targeted therapy. Conclusion: In this study, a lncRNA signature named as SenLncSig was developed that not only identifies the senescence phenotype and predicts prognosis but also has the capability to forecast the response of LUAD to immunotherapy.

CORR Insights®: Are Current Survival Prediction Tools Useful When Treating Subsequent Skeletal-related Events From Bone Metastases?

CORR Insights®: Are Current Survival Prediction Tools Useful When Treating Subsequent Skeletal-related Events From Bone Metastases?

Integration of pathologic characteristics, genetic risk and lifestyle exposure for colorectal cancer survival assessment

The development of an effective survival prediction tool is key for reducing colorectal cancer mortality. Here, we apply a three-stage study to devise a polygenic prognostic score (PPS) for stratifying colorectal cancer overall survival. Leveraging two cohorts of 3703 patients, we first perform a genome-wide survival association analysis to develop eight candidate PPSs. Further using an independent cohort with 470 patients, we identify the 287 variants-derived PPS (i.e., PPS287) achieving an optimal prediction performance [hazard ratio (HR) per SD = 1.99, P = 1.76 × 10−8], accompanied by additional tests in two external cohorts, with HRs per SD of 1.90 (P = 3.21 × 10−14; 543 patients) and 1.80 (P = 1.11 × 10−9; 713 patients). Notably, the detrimental impact of pathologic characteristics and genetic risk could be attenuated by a healthy lifestyle, yielding a 7.62% improvement in the 5-year overall survival rate. Therefore, our findings demonstrate the integrated contribution of pathologic characteristics, germline variants, and lifestyle exposure to the prognosis of colorectal cancer patients.

Identifying optimal surgical approach among T1N2-3M0 non-small cell lung cancer patients: a population-based analysis.

Whether stage T1N2-3M0 non-small cell lung cancer (NSCLC) patients could benefit from surgery and the optimal surgical procedure have remained controversial and unclear. This study aimed to investigate whether stage T1N2-3M0 NSCLC can benefit from different surgery types and develop a tool for survival prediction. The Surveillance, Epidemiology, and End Results (SEER) database was used to identify patients diagnosed with stage T1N2-3M0 NSCLC between 2000 and 2015. A 1:1 propensity score-matched (PSM) analysis was used to balance the distribution of clinical characteristics. Survival analyses were performed by using the Kaplan-Meier (KM) curves and Cox proportional hazards regression. All patients were randomly split at a ratio of 7:3 into training and validation cohorts. The nomogram was constructed by integrating all independent predictors for overall survival (OS) and cancer-specific survival (CSS). The model's performance was evaluated by discrimination, calibration ability, and risk stratification ability. A total of 4,671 patients were enrolled. After 1:1 PSM, the distribution proportions of clinical characteristics in 1,146 patients were balanced (all P>0.05). The non-surgical approach was associated with worse survival compared with sublobectomy and lobectomy in the unmatched and matched cohorts. The multivariate Cox analysis showed that sublobectomy and lobectomy were both related to better OS and CSS rates compared with no surgery (P<0.001). Moreover, the results of subgroup analyses based on age, N stage, and radiotherapy or chemotherapy strategy were consistent. A total of 801 patients were included in the training cohort and 345 cases constituted the validation cohort. The nomogram constructed for the 1-, 3-, and 5-year OS and CSS prediction showed good discrimination, performance, and calibration both in the training and validation sets. Significant distinctions in survival curves between different risk groups stratified by prognostic scores were also observed (all P<0.001). Stage T1N2-3M0 NSCLC patients could benefit from sublobectomy or lobectomy, and lobectomy provides better survival benefits. We developed and validated nomograms, which could offer clinicians instructions for strategy making.

Abstract 864: Pan-cancer analysis of lactic acid metabolism landscape across the Cancer Genome Atlas

Abstract Introduction: Metabolic reprogramming plays an important role in tumor occurrence and progression. Our studies confirmed the survival state of patients with different metabolic systems was heterogeneous and glycolytic metabolism may cause tumor progression. Based on these findings, lactic acid metabolism in glycolysis aroused our interest. This study evaluated the potential association between lactate metabolism gene and tumor clinical process in hepatocellular carcinoma, and proved that it can be used as an effective tool for survival prediction and treatment guidance of 32 kinds of tumor patients by establishing lactate score prognosis model. Methods: We used the expression profile, proteomics and epigenetics of HCC lactate gene set in the Cancer Genome Atlas Database to systematically describe the changes of lactate related genes. The differentially expressed genes were then verified in 32 clinical tumor data sets. A prognostic model of tumor lactic acid infiltration was constructed to evaluate its correlation with the prognosis of patients. Finally, we analyzed the effects of lame score, immunotherapy and drug response on the clinical process of tumor. Results: Firstly, we analyzed the differential expression of lactate metabolism genes in HCC and found that the prognosis of patients with high lactate infiltration was significantly lower than that of mild infiltration. AF angiogenesis factor and EMT factor in high infiltration may be the main reason for their differential expression. We constructed a tumor lactic acid prognosis model through six significantly different lactic acid genes, and found that lame score can be used as a predictor of immunotherapy. In addition, we proved the role of lame score in Pan cancer. PGA was positively correlated with lame score in most tumors. The prognosis of 16 tumors with high lame score group was significantly lower than that of low lame score group. There were significant differences in lame score between metastatic and non metastatic cancer groups of Cesc, kirc and kirp. The high and low risk groups of some tumors showed different dryness indices and dampness. At the same time, methylation gene modification has a significant impact on the Lame score of pan cancer, and the Lame score is used to link the drug tolerance and specific pathway enrichment of pan cancer tumor cell lines, confirming the targeted effect of Lame score on tumor treatment. Finally, we used the clinical characteristics and immunohistochemical analysis of a cohort study of real patients with breast cancer endocrine therapy (Tamoxifen) and hepatocellular carcinoma targeted therapy (Lenvartinib) to verify the clinical treatment prediction ability of lame score. Conclusions: Our study provides a comprehensive analysis of lactate metabolic across cancers and highlights the potential treatment of lactate targeted therapies for cancers. Citation Format: Xiaolin Wei, Lei Cai, Qian He. Pan-cancer analysis of lactic acid metabolism landscape across the Cancer Genome Atlas [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 864.

Are Current Survival Prediction Tools Useful When Treating Subsequent Skeletal-related Events From Bone Metastases?

Are Current Survival Prediction Tools Useful When Treating Subsequent Skeletal-related Events From Bone Metastases?

Integrated clinical and prognostic analyses of mTOR/Hippo pathway core genes in hepatocellular carcinoma

Hepatocellular carcinoma (HCC) is one of the most aggressive and dismal cancers globally. Emerging evidence has established that mTOR and Hippo pathways are oncogenic drivers of HCC. However, the prognostic value of these pathways in HCC remains unclear. In this study, we aimed to develop a gene signature utilizing the mTOR/Hippo genes for HCC prognostication. A multiple stage strategy was employed to screen, and a 12-gene signature based on mTOR/Hippo pathways was constructed to predict the prognosis of HCC patients. The risk scores calculated by the signature were inversely correlated with patient prognosis. Validation of the signature in independent cohort confirmed its predictive power. Further analysis revealed molecular differences between high and low-risk groups at genomic, transcriptomic, and protein-interactive levels. Moreover, immune infiltration analysis revealed an immunosuppressive state in the high-risk group. Finally, the gene signature could predict the sensitivity to current chemotherapeutic drugs. This study demonstrated that combinatorial mTOR/Hippo gene signature was a robust and independent prognostic tool for survival prediction of HCC. Our findings not only provide novel insights for the molecular understandings of mTOR/Hippo pathways in HCC, but also have important clinical implications for guiding therapeutic strategies.

Cost-effective prognostic evaluation of breast cancer: using a STAR nomogram model based on routine blood tests.

Breast cancer (BC) is the most common and prominent deadly disease among women. Predicting BC survival mainly relies on TNM staging, molecular profiling and imaging, hampered by subjectivity and expenses. This study aimed to establish an economical and reliable model using the most common preoperative routine blood tests (RT) data for survival and surveillance strategy management. We examined 2863 BC patients, dividing them into training and validation cohorts (7:3). We collected demographic features, pathomics characteristics and preoperative 24-item RT data. BC risk factors were identified through Cox regression, and a predictive nomogram was established. Its performance was assessed using C-index, area under curves (AUC), calibration curve and decision curve analysis. Kaplan-Meier curves stratified patients into different risk groups. We further compared the STAR model (utilizing HE and RT methodologies) with alternative nomograms grounded in molecular profiling (employing second-generation short-read sequencing methodologies) and imaging (utilizing PET-CT methodologies). The STAR nomogram, incorporating subtype, TNM stage, age and preoperative RT data (LYM, LYM%, EOSO%, RDW-SD, P-LCR), achieved a C-index of 0.828 in the training cohort and impressive AUCs (0.847, 0.823 and 0.780) for 3-, 5- and 7-year OS rates, outperforming other nomograms. The validation cohort showed similar impressive results. The nomogram calculates a patient's total score by assigning values to each risk factor, higher scores indicating a poor prognosis. STAR promises potential cost savings by enabling less intensive surveillance in around 90% of BC patients. Compared to nomograms based on molecular profiling and imaging, STAR presents a more cost-effective, with potential savings of approximately $700-800 per breast cancer patient. Combining appropriate RT parameters, STAR nomogram could help in the detection of patient anemia, coagulation function, inflammation and immune status. Practical implementation of the STAR nomogram in a clinical setting is feasible, and its potential clinical impact lies in its ability to provide an early, economical and reliable tool for survival prediction and surveillance strategy management. However, our model still has limitations and requires external data validation. In subsequent studies, we plan to mitigate the potential impact on model robustness by further updating and adjusting the data and model.

A survival nomogram model for patients with resectable non-small cell lung cancer and lymph node metastasis (N1 or N2) based on the Surveillance, Epidemiology, and End Results Database and single-center data.

The ability to predict survival in patients with lymph node metastasis has long been elusive. After surgery, the basis for decision-making on the combination treatment of patients is not clear. The purpose of this study was thus to build a survival nomogram model to effectively predict the overall survival (OS) of patients with non-small cell lung cancer (NSCLC) and lymph node metastasis. The number of dissected lymph nodes (NDLN), number of positive lymph nodes (NPLN), lymph node ratio (LNR), and log odds of positive lymph nodes (LODDS) were included in this study to determine the risk factors in patients with advanced NSCLC. The data of 5,132 patients with NSCLC and lymph node metastasis (N1 or N2) were extracted from the Surveillance, Epidemiology, and End Results (SEER) database according to inclusion and exclusion criteria and used as the training cohort. We enrolled 117 patients from the First Affiliated Hospital, Zhejiang University School of Medicine as the external validation cohort. Receiver operating characteristic (ROC) analyses were performed to determine the best cutoff values for predicting the prognosis of patients with NSCLC. Based on the risk factors affecting prognosis, a nomogram was constructed using univariate and multivariate Cox proportional hazard regression models. The discrimination ability of the nomogram was evaluated with the concordance index (C-index) and calibration curves. For the independent risk factors, survival curves were drawn using Kaplan-Meier analysis. ROC curve analysis showed that the optimal NPLN cut-off value was 4, LNR was 0.26, and LODDS was -0.25, respectively. However, LNR was nonsignificant in multivariate analysis, with a P value of 0.274. The novel survival nomogram model included seven independent risk factors, among which were NPLN, LODDS, and chemotherapy. Model 4, which included N stage, NPLN, and LODDS, had a higher likelihood ratio (LR) and C-index than did the other models. The C-index was 0.648 [95% confidence interval (CI): 0.636-0.659] in the training cohort and 0.807 (95% CI: 0.751-0.863) in the external validation cohort, showing good prognostic accuracy and discrimination ability. According to the median risk score, the patients in the training cohort and external validation cohort were divided into high-risk and low-risk groups, between which significant differences in OS were found. In the training cohort, age, sex, T stage, N stage, NPLN, LODDS, and chemotherapy were significantly associated with OS (P<0.001). In the external validation cohort, T stage, NPLN, LODDS, and chemotherapy were found to be correlated with OS. The NPLN and LODDS nomogram is an accurate survival prediction tool for patients with N1 or N2 NSCLC. Patients with lymph node metastasis can benefit from chemotherapy, but no evidence shows that radiotherapy is necessary for patients with resectable NSCLC.

Conditional survival analysis and dynamic survival prediction for intracranial solitary-fibrous tumor/hemangiopericytoma

BackgroundAs the form of World Health Organization Central Nervous System (WHO CNS) tumor classifications is updated, there is a lack of research on outcomes for intracranial combined solitary-fibrous tumor and hemangiopericytoma (SFT/HPC). This study aimed to explore conditional survival (CS) pattern and develop a survival prediction tool for intracranial SFT/HPC patients.MethodsData of intracranial SFT/HPC patients was gathered from the Surveillance, Epidemiology, and End Results (SEER) program of the National Cancer Institute. The patients were split into training and validation groups at a 7:3 ratio for our analysis. CS is defined as the likelihood of surviving for a specified period of time (y years), given that the patient has survived x years after initial diagnosis. Then, we used this definition of CS to analyze the intracranial SFT/HPC patients. The least absolute shrinkage and selection operator (LASSO) regression and best subset regression (BSR) were employed to identify predictive factors. The Multivariate Cox regression analysis was applied to establish a novel CS-based nomogram, and a risk stratification system was developed using this model.ResultsFrom the SEER database, 401 patients who were diagnosed with intracranial SFT/HPC between 2000 and 2019 were identified. Among them, 280 were included in the training group and 121 were included in the internal validation group for analysis. Our study revealed that in intracranial SFT/HPC, 5-year survival rates saw significant improvement ranging from 78% at initial diagnosis to rates of 83%, 87%, 90%, and 95% with each successive year after surviving for 1–4 years. The LASSO regression and BSR identified patient age, tumor behavior, surgery and radiotherapy as predictors of CS-based nomogram development. A risk stratification system was also successfully constructed to facilitate the identification of high-risk patients.ConclusionThe CS pattern of intracranial SFT/HPC patients was outlined, revealing a notable improvement in 5-year survival rates after an added period of survival. Our newly-established CS-based nomogram and risk stratification system can provide a real-time dynamic survival estimation and facilitate the identification of high-risk patients, allowing clinicians to better guide treatment decision for these patients.

Development and Validation of a Nomogram for Predicting Survival Based on Ferritin and Transferrin Ratio in Breast Cancer Patients.

Our objective is to develop a predictive model utilizing the ferritin and transferrin ratio (FTR) and clinical factors to forecast overall survival (OS) in breast cancer (BC) patients. We conducted a retrospective analysis of clinical data from 2858 BC patients diagnosed between 2013 and 2021. Subsequently, the cohort of 2858 BC patients underwent random assignment into distinct subsets: a training cohort comprising 2002 patients and a validation cohort comprising 856 patients, maintaining a proportional ratio of 7:3. Employing multivariable Cox regression analysis within the training cohort, we derived a prognostic nomogram. The predictive performance was assessed using calibration curves, C-index, and decision curve analysis. The final prognostic model included the TNM stage, subtype, hemoglobin levels, and the ferritin-transferrin ratio. The nomogram achieved a C-index of .794 (95% CI: .777-.810). The nomogram demonstrated superior predictive accuracy for OS at 3, 5, and 7 years for BC, with area under the time-dependent curves of .812, .782, and .773, respectively. These values notably outperformed those of the conventional TNM stage. Decision curve analysis reaffirmed the greater net benefit of our nomogram compared to the TNM stage. These findings were subsequently validated in the independent validation cohort. The FTR-based prognostic model may predict a patient's OS better than the TNM stage in a clinical setting. The nomogram can provide an early, affordable, and reliable tool for survival prediction, as well as aid clinicians in treatment option-making and prognosis evaluation. However, further multi-center prospective trials are required to confirm the reliability of the existing nomogram.

Clinical and prognostic significance analysis of glycolysis-related genes in HNSCC.

Head and neck squamous cell carcinoma (HNSCC) represents one of the most malignant cancers worldwide, with poor survival. Experimental evidence implies that glycolysis/hypoxia is associated with HNSCC. In this study, we aimed to construct a novel glycolysis-/hypoxia-related gene (GHRG) signature for survival prediction of HNSCC. A multistage screening strategy was used to establish the GHRG prognostic model by univariate/least absolute shrinkage and selection operator (LASSO)/step multivariate Cox regressions from The Cancer Genome Atlas cohort. A nomogram was constructed to quantify the survival probability. Correlations between risk score and immune infiltration and chemotherapy sensitivity were explored. We established a 12-GHRG mRNA signature to predict the prognosis in HNSCC patients. Patients in the high-risk score group had a much worse prognosis. The predictive power of the model was validated by external HNSCC cohorts, and the model was identified as an independent factor for survival prediction. Immune infiltration analysis showed that the high-risk score group had an immunosuppressive microenvironment. Finally, the model was effective in predicting chemotherapeutic sensitivity. Our study demonstrated that the GHRG model is a robust prognostic tool for survival prediction of HNSCC. Findings of this work provide novel insights for immune infiltration and chemotherapy of HNSCC, and may be applied clinically to guide therapeutic strategies.

Whole-Liver Based Deep Learning for Preoperatively Predicting Overall Survival in Patients with Hepatocellular Carcinoma.

Survival prediction is crucial for treatment decision making in hepatocellular carcinoma (HCC). We aimed to build a fully automated artificial intelligence system (FAIS) that mines whole-liver information to predict overall survival of HCC. We included 215 patients with preoperative contrast-enhance CT imaging and received curative resection from a hospital in China. The cohort was randomly split into developing and testing subcohorts. The FAIS was constructed with convolutional layers and full-connected layers. Cox regression loss was used for training. Models based on clinical and/or tumor-based radiomics features were built for comparison. The FAIS achieved C-indices of 0.81 and 0.72 for the developing and testing sets, outperforming all the other three models. In conclusion, our study suggest that more important information could be mined from whole liver instead of only the tumor. Our whole-liver based FAIS provides a non-invasive and efficient overall survival prediction tool for HCC before the surgery.